Вы находитесь на странице: 1из 271



M any p r o f o u n d statements can be m a d e about w a t e r and its importance in our lives. Its beneficial
qualities are many. However, routinely handling large quantities of waters of vastly differing
compositions at m i n i m u m cost and with a m i n i m u m of operating problems often presents some
tremendous problems.


Water is often called the universal solvent because it has the p o w e r to dissolve virtually all inor-
ganic substances to some extent. In its pure state it has the following physical properties:

Physical Properties of Water
Property Value
Molecular Weight 18
Density @ 4C 1 g/mL
Freezing Point 0C [32F]
Boiling Point 100C [212F]

Water exhibits properties that might not be considered normal when compared with other com-
pounds, as can be seen in Table 2.2.

Comparison of Boiling Points v e r s u s Molecular Weight
Compound Formula Boiling Point, C Molecular Weight
Water H->0 100 18
Hydrogen Sulfide H2S -61.8 34
Hydrogen Selenide H 2 Se -41 81
Ammonia NH, -33 17
Methane CH 4 -160 16

Water would be expected to have a much lower boiling point.

The unique properties of w a t e r are due to the presence of "hydrogen bonding". While w e a k e r
an the atomic bonding between the h y d r o g e n and o x y g e n " a t o m s within each water molecule, the


hydrogen bonds are relatively s t r o n g m binding together the water molecules. Hence, the chemical
formula might even be written as: ( H 2 b ) n where "n" represents an unspecified number.

The structural formula could then be shown as:



- where the solid lines represent strong intermolecular b o n d s and the dotted lines s h o w the w e a k e r
intra-molecular hydrogen bonding. The hydrogen b o n d i n g results in tighter packing of the water mole-
cules, so that it displays the properties of a liquid instead of those of a gas. Heat is required to rupture
these b o n d s in order for water to vaporize.

Hydrogen bonding also gives water a high surface tension, because the water molecules have a
strong affinity for each other and do not w a n t to leave the surface of the liquid body. This high sur-
face tension causes water to rise up a capillary tube, or f o r m rounded droplets on clean solid surfaces.
It is the same property that hinders the wetting of particles in the water and hence increases the ability
of water to suspend solid particles.

H y d r o g e n bonding contributes significantly to water's high:

Freezing point
Boiling point
Heat capacity
Latent heat of vaporization

and the high amount of heat needed to melt ice. T h e s e all contribute to the value of water as a heating
and cooling medium.

Most water handling p r o b l e m s in the oilfield arise from the fact that water is such a superb sol-
vent. Both produced waters and surface waters contain considerable quantities of impurities. It has
had ample contact with soil and rock formations and has dissolved certain c o m p o u n d s . In addition, jt
usually contains some suspended solids and dissolved gases. The later often gives rise to corrosion of
metal in contact with the water. Microbiological growths often proceed very readily in water. As con
ditions of temperature and pressure change, many of the dissolved c o m p o u n d s m a y b e c o m e insoluble
to s o m e degree, precipitate f r o m the water and f o r m scalc. T h e n u m b e r and combination of difficulties
which m a y arise in water handling are e n o r m o u s .

In the oilfield we use water for many purposes. The most c o m m o n are:

Injection into subsurface formations to increase oil recovery and/or maintain reservoir pres-
Injection into subsurface formations to dispose of waste waters.
Disposal of waste waters into surface waters.
Cooling of natural gas engine jackets, compressor cylinders, natural gas and other process
Feedwater for boilers and steam generators.



Regardless of the application, we have two primary goals from an operational standpoint:
Avoid plugging and deposition of solids in lines, vessels and wells.
Prevent corrosion of surface and d o w n h o l e equipment.

O n e of the first items of interest in water handling is to sample it and determine its composition.
This is our primary means of detecting present and potential problems. However, the water sample
must be representative of the water of interest nr our analysis will lead to false conclusions. The im-
portance of good sampling practices cannot be overemphasized. An extremely accurate chemical anal-
ysis of a water sample, followed by a brilliant assessment of the problems indicated by the analysis, is
worthless if the sample does not represent the water in your system.

Sample Containers
Clean (preferably n e w ) plastic bottles with tightly fitting plastic caps are r e c o m m e n d e d for rou-
tine water samples. 500 mL or one liter [1 pint or 1 quart] bottles are c o m m o n l y used. Several differ-
ent types of plastic and different cap designs are available. O n c e you have selected a particular bottle
and cap combination, fill it with water, cap it, and squeeze it to m a k e sure that the cap will not pop off
or leak. Wide-mouth bottles are usually easier to fill, but often have less dependable caps.

Label the bottle (not just the cap) so that the sample can be identified. Cardboard mailing tubes
or cartons should be obtained if the samples are to be shipped to a laboratory or carried a considerable

If the sample is to be analyzed for oil content or f o r the concentration of any other organic con-
stituent, a glass bottle should be used. Oil or other organic materials will adhere to the walls of a plas-
tic container or be absorbed by it. Extra precautions are necessary in packing the sample for shipment.
Freezing can be a problem. Use a plastic cap with a plastic liner.

Never use a metal container or a metal cap. They will corrode and contaminate the w a t e r with
corrosion products.

Sample Volume
A m i n i m u m sample volume of 500 mL [one pint] is recommended for routine analysis.

If you are going to personally p e r f o n n on-site analyses, you can take as much as you think y o u
need. If only one or two analyses are to be performed, only a f e w hundred milliliters m a y be neces-
sary. Also, since you are on site, you can run back to the sampling point and obtain some more water
(provided the water system is not changing composition rapidly). But r e m e m b e r that if you send water
samples to a laboratory several miles away and y o u d o n ' t send enough to permit t h e m to p e r f o n n all of
the analyses you want, you are simply out of luck. A l w a y s send m o r e than enough if there is any
doubt in y o u r mind as to the m i n i m u m v o l u m e required.

Sampling Procedure
It a sampling valve is available, connect a piece of plasjic tubing to the end of the valve. A
small nipple or hose connector is handy for this purpose. O p e n the v a l v e and let the water run for at
least one minute. Watch to if the color of the_water is chanpinp If it is. wait until the O I M K


c o m e s constant. R e m e m b e r that there is probably, some debris in the valve or in the bottom of the line
ancfthis w i l l l i a v e to be flushed out before y o u can get a g o o d sample.

Once you have a representative water flow, the following sampling procedures are r e c o m m e n d e d :
Sample for Routine Analysis Rinse the bottle out three times, then place the end of the
hose in the bottom of the bottle and let the bottle overflow for an estimated 10 volumes.
Then slowly pull out the hose and quickly cap it to minimize (1) oxygen contamination
and (2) the escape of dissolved gases.
Sample for Oil-in-Water Analysis Fill a clean glass bottle to the neck directly f r o m the
sample point and cap quickly. Do not rinse or overflow the bottle with the water to be
sampled as in the case of the sampling procedure given for routine analyses. Oil will tend
to adhere to the bottle wall and separate f r o m the water. If the bottle is filled more than
once, oil f r o m each filling is likely to stay in the bottle, giving erroneously high results.
Alternately, free oil may be carried out of the bottle by overflowing, giving low results.
Sample for Bacterial Analysis T h i s procedure is covered in detail in Chapter 5.
Once the sample is capped, prepare a label immediately and attach it to the bottle or label the
bottle with an indelible, smear-proof marking pen. M a k e sure the label is securely glued to the bottle.
You can mark the cap for convenience if you wish, but r e m e m b e r that caps can be inadvertently

If the sample is to be sent to a laboratory for analysis it should be accompanied by a Water Sam-
ple Description Form detailing the sampling conditions. A typical example is given in Appendix 5
(page 390).

Some other hints:

1. Take wellhead samples at the wellhead, not at the heater-treater or at the tank.
2. If a tank is to be sampled, sample it at several levels, preferably top, middle and bottom.
This can require special equipment. You m a y be able to borrow a " t h i e f ' (sampling de-
vice) f r o m a gauger and clean it thoroughly b e f o r e use in order to obtain a sample f r o m the
center of a conventional tank.
3. Take samples f r o m the system when it is operating normally. It is often most convenient
to take samples when the system is shut down. Needless to say, this type of sample is of
questionable value. M a k e sure the flow rate is " n o r m a l " and that nothing unusual is going
on upstream of the sampling point.
4. If you are sampling a surface water as a possible source f o r a waterflood r e m e m b e r that the
water composition (especially the turbidity, o x y g e n concentration and microbial popula-
tion) may change considerably with the time of the year. It is advisable to sample at sev-
eral points over a period of several months.
5. Waters for disposal may change considerably with time. Plant waste waters m a y show a
cyclic change in composition due to ion exchange regeneration, cooling tower b l o w d o w n
cycles, or plant cleaning operations. O n c e again, a series of samples over a period of time
is recommended. Conversations with plant personnel may enrich y o u r intuition as to w h e n
to take samples which will reflect the greatest variations in disposal water composition.
This can influence your water treating procedure.




Water analyses are routinely carried out in laboratories by skilled chemists. They are capable of
m a k i n g extremely accurate measurements on the water sample provided. However, m a n y water prop-
erties can change very quickly after sampling. Typical are j)H, temperature, dissolved gas content, sus-
pended solids and bacterial population. This means that many of the properties which are of greatest
concern to us can be determined accurately only by measuring them on-site. Therefore, a thorough
analysis of any water usually involves both laboratory and field analyses.

It is extremely important that any person involved in a water injection or disposal project have an
understanding of:

The constituents and properties of water of greatest importance in water systems

The significance of each
T h e analytical methods typically used, along with their strengths and weaknesses
If these items are understood, then it is possible to specify which analyses are needed and to un-
derstand the significance of the results.

Primary Constituents and Properties

In water handling operations we are primarily concerned with those ions and physical properties
which are important f r o m the standpoint of plugging or corrosion. Table 2.3 is a list of the most im-

It may also be desirable to measure the amount of chlorine (a bactericide) or the concentration of
treating chemicals present in order to monitor their effectiveness.

In steam operations and cooling water systems the amount of silica (Si02) in the water is also
important, as it can f o r m deposits.

TABLE 2 . 3
Primary Constituents and Properties of Oilfield Waters
Cations Anions Other Properties
Sodium (Na + ) Chloride (CI") PH
Calcium ( C a ^ ) Sulfate (S0 4 = ) Bacterial Population
Magnesium (Mg ++ ) l Bicarbonate (HC0 3 ") Suspended Solids: Amount, size, shape, composition
Iron (Fe +++ & F e ^ ) 7Carbonate (C0 3 = ) Turbidity
Barium (Ba+4~) Water Quality (see page 39)
Strontium (Sr ++ ) Dissolved Oxygen
Dissolved Carbon Dioxide
Total Sulfides as H 2 S
Oil Content
Total Dissolved Solids (TDS)
Specific Gravity
Resistivity (Conductivity)
Silica xw.1 ],, j ^


Significance of Constituents and Properties (-.I)

a. Sodium is a m a j o r constituent in oilfield waters, but does not normally cause any prob-
lems. About the only exception is the precipitation of NaCl f r o m extremely salty brines.
b. T h e calcium ion is a m a j o r constituent of oilfield brines and m a y be as high as 30 000
mg/L although its concentration is normally considerably lower. T h e calcium ion is of ma-
j o r importance because it readily combines with bicarbonate, carbonate or sulfate ions and
precipitates to f o r m adherent scales or suspended solids.
c. Magnesium ions are usually present in m u c h lower concentrations than calcium. T h e y tend
to add to C a C O j scaling problems by co-precipitating with the calcium ion. It is very
c o m m o n to find m a g n e s i u m in calcium carbonate scafes.

Magnesium ions can sometimes decrease the amount of C a S 0 4 , B a S 0 4 and S r S 0 4 scales.

They do so by f o r m i n g "ion pairs" with the sulfate ion. Essentially, the m a g n e s i u m ion
has the ability to f o r m a c o m p o u n d which remains in solution. T h e sulfate ions which are
tied up with m a g n e s i u m are not available to f o r m sulfate scales.
d. T h e natural iron content of formation waters normally is quite low and its presence is usu-
ally indicative of corrosion. It is usually present in solution as ferrous (Fc ) or ferric
( F e + i ~ ) ions, the latter being due to the presence of oxygen (air). Iron can often be present
in suspension, as a precipitated c o m p o u n d . "Iron counts" are often used to detect and
monitor corrosion in a water system. Thej>rescnce of precipitated iron c o m p o u n d s is one
of the major causes of formation plugging.
e. Barium is of importance primarily because of its ability to combine with the sulfate ion to
form barium sulfate, which is extremely insoluble. Even small quantities can present se-
vere problems.
f. Strontium, like b a r i u m and calcium, can combine with the sulfate ion to f o r m insoluble
strontium sulfate. Although j r r o r e soluble than barium sulfate, it is often found in scales^
mixed with barium sulfate.

a. T h e chloride ion is nearly always the m a j o r anion in produced brines and is sometimes
present as a major constituent in fresh waters (although at relatively low concentrations).
T h e m a j o r source of the chloride ion is NaCl, so the chloride ion concentration is used as a
measure of water salinity.

Although salt deposition can be a problem, it is normally of little consequence. T h e pri-

mary problem associated with the chloride ion is that the conductivity of the water in-
creases as it gets saltier. Hence, high chloride concentrations allow faster corrosion
reactions. Also, the chloride ion is a stable constituent and its concentration is one of the
easier w a y s of identifying a water.
b. The sulfate_ion is a problem because of its ability to react with calcium, barium or stron-
tium to form insoluble sulfate scales. It also serves as a source of o x y g e n for sulfate re-
c. T h e bicarbonate ion can react with calcium, magnesium, iron, barium and strontium ions
to f o r m insoluble^ carbonate scales. It is present in virtually all waters. Bicarbonate ion
concentration is sometimes called methyl orange alkalinity.



d. Like the bicarbonate ion, the carbonate ion can also react with calcium, magnesium, iron,
barium and strontium ions to form insoluble scales. Carbonate ions are rarely present in
produced waters because the pH is usually too low (< 8.3). Carbonate ion concentration is
sometimes called phenophthalein alkalinity.

Other Properties
a. T h e pH value of a water is extremely important for several reasons. The solubility of
C a C p 3 and iron c o m p o u n d s is highly dependent on pH. The higher the p H . the greater the
tendency for precipitation.

As the pH decreases (becomes more acidic ) the scaling tendency of the water is decreased,
but its coirosivity is increased. Most oilfield waters have a pH b e t w e e n 4 and 8.

Both H2S and C O ; are "acid" gases as they tend to lower the pH of water (make it more
acid) when they dissolve in water. They partially ionize when they dissolve and the degree
of ionization is reflected by the p H . This is important in predicting their effect on corro-
sion and suspended solids.

Since pH values usually change rapidly after a sample is withdrawn from a pressurized
system (due to the escape of dissolved acid gases), pH values should be measured immedi-
ately after the sample is taken.

b. T h e presence of bacteria may result in corrosion and/or plugging. A more detailed treat-
ment of this subject is given in Chapter 5.
c. The quantity of suspended solids which can be filtered f r o m a given v o l u m e of water us-
ing a m e m b r a n e filter is one basis for estimating the plugging tendency of a water. A 0.45
pm pore-size filter is c o m m o n l y used.
d. It is possible to estimate the particle size distribution of the suspended solids in a water
sample by various techniques. A knowledge of the particle size distribution can be very
helpful in determining the need for filtration and in filter selection. It is also useful in
monitoring filter performance.
e. Determination of particle shape by visual or scanning electron microscopy is very helpful
in the determination of filtration needs. It is usually used in conjunction with particle size
f. Determination of the composition of the suspended solids makes it possible to ascertain
their origin (corrosion products, scale particles, formation sand, etc.) so that proper reme-
dial action can be taken. K n o w l e d g e of their chcmical composition is also important from
the standpoint of designing a cleanout procedure should plugging occur.
g. Turbidity simply means that the water is not "clear'" and that it contains undissolved matter
such as suspended solids, dispersed oil or gas bubbles. It is a measure of the degree of
"cloudiness" of the water. Turbidity indicates the possibility of formation plugging in in-
jection operations. Turbidity measurements are often used to monitor filter performance.
b. Water quality is a measure of the relative degree of plugging which occurs when a given
volume of water is passed through a m e m b r a n e filter of a given pore size. A pore size of
0.45 pm is most c o m m o n l y used. The utility of water quality testing lies largely in its use
as a comparative measurement.

Dissolved oxygen contributes significantly to the corrosivity of a water. Also, if dissolved

iron is present in a water, the entry of oxygen into the system can result in the precipitation

2 23

of dissolved iron a s _ i n s o l u b k iron oxides which may result in plugging. Oxygen also
facilitates the growth of aerobic bacteria.

j. Dissolved carbon dioxide influences the p H , corrosivity and CaCO} scaling tendency of a
k. The presence of sulfides in water will increase its corrosivity. Dissolved sulfides exist in
water as a mixture of HS~ ions and dissolved H^S gas at the pH values normally found in
oilfield waters, and the total concentration of both species is usually measured and re-
ferred to as "total sulfides". It m a y be present naturally in the water or it m a y be gener-
ated by sulfate~reducihg bacteria. If a normally sweet (free of H?S) water begins to show
traces of H2S, this indicates that sulfate-reducing bacteria are probably at work s o m e w h e r e
in the system busily corroding holes in your piping and vessels. In addition, iron sulfide
will be generated as a corrosion product, and it is a very efficient p l u g g i n g agent.

I. T h e presence of dispersed or emulsified oil in water often presents problems when inject-
ing produced waters.

Oil in water can cause decreased injectivity in several ways. It can cause "emulsion
b l o c k s " in the formation. It serves as an excellent glue for certain solids, such as iron sul-
fide, thereby increasing their plugging efficiency.

When water is being injected into an aquifer with no initial oil saturation, oil in the water
can be trapped in the pores of the formation rock around the wellbore. This creates an oil
saturation, which can reduce injectivity.

An analysis for oil content should be conducted on any water, regardless of origin. There
are many ways in which water can become contaminated with oil.

When produced water is disposed into surface waters, the concentration of oil in the water
is usually limited by government regulation.

m. T h e temperature of the water affects the scaling tendency, the pH and the solubility of
gases in water. The specific gravity of w a t e r is also a function of temperature.

II. The total dissolved solids is simply the total amount of matter dissolved in a given v o l u m e
of water. It can be calculated by taking the s u m of the concentrations of all cations and
anions shown on the water analysis report, or it can be measured by evaporating a sample
of water to dryness and weighing the residue. It can also be determined f r o m specific
gravity measurements.

0. Specific Gravity
_ . Densitv of Water Sample
Specific Gravity = (2.1)
Density of Pure Water

Density is simply weight per unit volume. Pure water weighs 1 g/mL or 1000 kg/m 3 .
Thus, a specific gravity greater than 1.0 means that the water sample of interest is more
dense, or weighs more than an equal volume of pure water. Since the waters we are inter-
ested in contain dissolved solids, they are always more dense than pure water and therefore
have a specific gravity greater than 1.0.

T h e magnitude of the specific gravity is a direct indicator of the total amount of solids dis-
solved in the water. Therefore, comparison of the specific gravity of several waters gives a
rapid estimate of the relative amounts of solids dissolved in the waters.



A graph of total dissolved solids as a function of specific gravity is given in Appendix 6

(page 391). This is an empirical correlation based on actual density measurements of oil-
field waters.

Specific gravity is also a u s e f u l check on the accuracy of laboratory water analyses. If the
quantity of dissolved solids calculated f r o m the laboratory analysis drastically disagrees
with the amount estimated f r o m the specific gravity correlation, there is a good chance that
the analysis is in error.
p. T h e resistance to electrical current flow is a function of the n u m b e r of ions dissolved in
the water. T h e lower the resistance the higher the concentration of ions dissolved in the
water. It is a method of estimating total dissolved solids which is primarily applied to
fresh, low T D S waters. Resistivity measurements are much m o r e sensitive to small
changes in T D S in fresh waters than specific gravity.

A chart showing the resistivities of sodium chloride solutions as a function of temperature

is given in Appendix 7 (page 392).
q. Conductivity values are measured by s o m e laboratories instead of resistivity. Conductivity
is the reciprocal of resistivity and can be derived from resistivity values using the follow-
ing formula:

Conductivity (|imho /cm) = 1^000 (2.2)

Resistivity (ohm - m)
r. Silica occurs in most well waters and can be a serious source of scale deposition in cooling
waters and in steam boilers. It normally does not present any problems in water injection

Water Analysis Techniques

Summary of Methods Used

R e c o m m e n d e d techniques for measuring the concentration of the ions of interest are given in API
RP 45, Analysis oj Oilfield Waters.In addition, meters may be used to determine dissolved oxygen
or pH; membrane filter analysis can be used to measure the suspended solids/-' 3 >

T h e types of techniques nonnally used to analyze an oilfield water sample are summarized in
Table 2.4.

Chemical Quantitative Analysis

Quantitative analysis is simply the determination of h o w much of each of the individual constitu-
ents are present in a given sample. Three basic types of methods are c o m m o n l y used to analyze water
samples in the field. More sophisticated instrumental techniques m a y be used in an analytical labora-

An "indicator" is added to the water sample and a standard solution is then added drop-by-drop
from a calibrated dispenser until the solution changes color. The point at which the solution changes
color is called the endpoint. This simply m e a n s that you are finished titrating. The volume of stan-
dard solution used to reach the endpoint is noted and the amount of the unknown species present is


S u m m a r y of Analysis Methods
Determination Analysis Method
Sodium Calculation, spectroscopic, gravimetric
Calcium Titration, gravimetric, spectroscopic
Magnesium Titration, gravimetric, spectroscopic
Iron Colorimetric, titration, spectroscopic
Barium Turbidimetric, spectroscopic
Strontium Spectroscopic
Chloride Titration
Sulfate Turbidimetric, gravimetric
Bicarbonate Titration
Carbonate Titration
pH pH Meter, colorimetric, pH paper
Bacterial Population Culturing, microscopic, others
Suspended Solids Concentration Gravimetric
Particle Size Coulter Counter, light scattering, microscopy
Particle Shape Microscopy
Suspended Solids Composition Chemical analyses
Turbidity Turbidimetric
Water Quality Membrane filter test
Dissolved Oxygen Oxygen meter, titration, colorimetric
Carbon Dioxide Titration
Total Sulfides Alkaseltzer test, colorimetric, titration
Oil-in-water Colorimetric, spectroscopic, gravimetric
Temperature Thermometer
Total Dissolved Solids Calculation, gravimetric, conductivity
Specific Gravity Hydrometer
Resistivity Resistivity cell, calculation
Silica Gravimetric, colorimetric, spectroscopic

A titration is simply a method of determining the concentration of a particular substance in solu-

tion by reacting it with a k n o w n material. Reactions are used which are k n o w n to be complete at a
given p H . A pH meter is normally used to detect the endpoint in the laboratory. However, color
indicators (substances that c h a n g e color at a specific pH value) are used for analyses carried out in the

T h e colors of some c o m m o n l y used indicators are shown in Table 2.5 as a function of p H . The
pH value at which the color change occurs is designated with a L 'T."

An example of a titration is the determination of bicarbonate ion ( H C 0 3 ~ ) concentration by titra-

tion with HC1:

HCO~ + H+ H20 + C02

T h e reaction is complete at a pH of 4.5, so methyl orange is used as the indicator.



Colors of Indicators
Increasingly Acid Increasingly Alkaline

H h 1 1 1 1 1 h 1-i H 1
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Litmus Red -T <- -Blue

Phenolphthalein Colorless T -Red

Methyl Orange Pink, Rod >T <- -Orange

Methyl Violet Y, G, BT -Violet

Alizarin Yellow Colorless >T < -Yellow

The concentration of the ion of interest is calculated based on the fact that one equivalent weight
of the k n o w n , or standard solution, will react with one equivalent weight of the ion in the water sam-

E standard = E sample
Where: ^standard =
number of equivalent weights of titrant (standard solution) added to water sample
to reach endpoint
Example number of equivalent weights of ion of interest in water sample

By definition:

Number of Gram Equivalent Weights of Ion (E)

Normalitv(N) = (2.3)
Liters of Solution ( L)


^ ^ standard and L
\E = NxL]J sample
Setting them equal to each other:

standard t^^ sample

' r 1000m/l 1000ml

N xLx i NxLx
L Jstandard sample

[N x mL] , =\N x mL\ ,

l J standard L J sample

rL N x mL\J J
x = N sample
1 " sample


T h e normality of the standard solution is known. The volume of the standard solution required
to reach the endpoint and the volume of the water sample is measured. The equation can then be
solved for the normality of the ion of interest in the water sample. Substituting:

= N sample
L sample

" 1 '
[YV x mL\standard
mL sample _L_ sample

This enables us to calculate the n u m b e r of equivalent weights of the ion of interest per liter of
water. In order to determine the weight of the ion present in a liter, it is necessary to multiply both
sides of the equation by the gram equivalent weight of the ion and convert grams to milligrams.

1000 mg E_ 1000 mg
[N x mL] standard x W ( g ) x xE1V(g)x
mL sample g L sample

All of this is the long way to arrive at the final formula which we use to calculate the results
from a titration:

E W 1 0 0 0
* '^standard * f * ^
mL of sample L

The API r e c o m m e n d s the use of mg/L as the unit of concentration for water analyses. Therefore,
it is suggested that all titration results be reported as mg/liter of the substance determined.

Sometimes p p m are used interchangeably with m g / L . This is correct only when the specific grav-
ity of the water is very near 1.0, since:

mg/L = ppm x S.G. (2.5)

In most titrations certain ions other than the ion of interest will react with the standard solution.
This can be a serious p r o b l e m if the interfering ions are present in significant quantities relative to the
amount of the ion of interest present. Table 2.6 lists some c o m m o n titrations and the interferences for

Common Titrations and I n t e r f e r e n c e s ' 2 ^
Ion Titrant Indicator Interferences
cor H 2 S 0 4 or HC1 Phenolphthalein Borate, silicate, sulfide and phosphate will be
Methyl Purple or Methyl included in the values for carbonate and bicar-
HCOj- H 2 S0 4 or HC1 bonate as are volatile fatty acid anions.
Cal-Red, Calcon or Barium and strontium are included with the cal-
Ca ++ EDTA
Murexide cium determination. Iron also interferes. Can
be masked by triethanolamine when more than
Mg + + EDTA Eriochrome Black T 20 mg/L is present.
Bromides, iodides, thiocyanates, phosphates,
Potassium Chromate carbonates and sulfides precipitate silver ions.
cr Silver Nitrate
(K 2 Cr0 4 ) Iron, barium , lead, and bismuth precipitate
chromate indicator.



Colorim etric A n afysis

Colorimetric methods are widely used because of their simplicity. This type of test is based on
addition of a material to the water sample which will react specifically with the species in question and
produce a colored solution. T h e intensity of the color is proportional to the amount of the species
present. The concentration of the substance in the w a t e r is f o u n d by comparing the color of the sam-
ple with the color standards of k n o w n concentration.

For example:
Fe++ + ortho - phenananthroline > Orange red

pH -I- various indictors > Spectrum of color

Battery-powered spectrophotometers or colorimeters are also available and can be used for many
test procedures. The color-developing reagents are available in pre-weighed packets w h i c h are ex-
tremely convenient for field use.

O n e of the most convenient colorimetric analysis procedures employs an extremely clever unit-
ized reagent and sampling system called a CHEMet Each C H E M e t is a 7 mm diameter glass am-
poule with a tapered, prescored tip. Color-forming reagents are sealed inside the ampoule under vac-
uum. W h e n the user snaps the tapered tip of the ampoule in the sample, v a c u u m pulls the sample in

Filling is instantaneous and complete except

for a small bubble of inert gas. Sample and reagent
Vacuum CHEMet
are mixed by tilting the ampoule and allowing the
bubble to travel f r o m end to end several times. A f -
ter waiting a specified period for color develop- Color-Forming
ment (usually two minutes or less) the analyst Reagent
quantifies the result using a color comparator sup-
plied with the kit.

Although C H E M e t s are available for a fairly

wide range of analyses, they have proven to be in-
valuable for the detection of low levels of dis-
solved oxygen. Test kits are available with ranges
a s low a s 0-20 p p b (0-0.02 ppm). T h e results are F i g u r e 2-1 Using a CHEMet
expressed a s m g / L 0 2 .

The results are not affected by temperature, salinity, other dissolved gases (including FhS), or the
presence of sulfite oxygen scavengers.

Oxidizing agents, such as chlorine, can cause high results unless a special formulation of the re-
agent is used.

Turbidimetric A n a lysis
In this type of analysis, a reagent is added to the water which will react with the ion of interest to
lorm a finely divided precipitated solid. T h e precipitate creates a cloudy solution and the degree of
cloudiness, or turbidity, is proportional to the amount of the ion present.

For example, turbidimetric analysis can be used to determine sulfate concentration.

SO; + Ba++ - BaSO,


Of the three analytical methods,

Titration is the most accurate.
Colorimetric techniques are the most sensitive (can detect the smallest quantities).
Turbidimetric analyses are the least accurate.

Expression of Results
The concentration of different species in a water can be expressed in several different ways.
1. Milligrams per Liter (mg/L)

This is the unit of concentration recommended by the API for oilfield waters.
2. Milliequivalents per Liter (meq/L)
Some laboratories also report results in milliequivalents per liter.

m !L
meq/L = f (2.6)
Equivalent weight
3. Parts per Million (ppm)

If the water is very fresh, the specific gravity is essentially 1.0, and p p m and mg/L are
equal. However, as the T D S of the water increases, the specific gravity increases and the
units b e c o m e increasingly different.
PPm= ( )
4. Equivalents per Million (epm)

This unit of concentration is calculated as follows:

epm = (2.8)
Equivalent weight
5. Grains/U.S. Gallon

Grains/gallon (gr/gal.) are units which are seldom used in routine oilfield water analysis
work. However, they are used in calculations involving water softening by ion exchange.

gr/U.S.gal.= '-^~ (2.9)

6. ppm as CaCO.i

Watch out for this unit system as it is frequently used in cooling w a t e r and boiler water
work. When the concentration of an ion is expressed in ppm as C a C O j this m e a n s that its
concentration has been calculated using the equivalent weight of C a C O j (50) instead of the
equivalent weight of the ion in question. This is done by multiplying the concentration of
the ion by the ratio of the equivalent weight of CaCO3 to the equivalent weight of the ion
of interest.

For example:

ppm Ca++ as CaCO3 = ppm Ca++ x^ (2.10)

= 2.5x ppm Ca++



In our work we are usually interested in just the reverse of the above procedure, i.e., con-
verting results reported to us in p p m as CaCC>3 to ppm of the ion. T h e r e f o r e , we do just
the reverse of the above.

ppm Ca++ = ppm Ca++ as CaCO, x (2-11)

= 0.4 x ppm Ca++ as CaCO,

A conversion table for converting ppm as C a C 0 3 to p p m of the ion is given in Appendix 8
(page 393).

Calcium carbonate equivalent concentration units are used in water softening calculations
and is the standard method of expressing hardness and alkalinity.
7. Hypothetical Salt Combinations
Sometimes the reported results are expressed in ppm of hypothetical salts. T h e chemist
m a y a s s u m e that all of the C a + + present will be present as calcium bicarbonate,
Ca(HCC>3)2, for instance. This is not true, since the calcium is present in solution as the
calcium ion, not as a salt. M a n y hypothetical combinations of ions are possible f r o m a sin-
gle ionic analysis. This unit system is not recommended.
8. Hardness

Total hardness is normally the sum of the calcium and m a g n e s i u m concentrations ex-
pressed as equivalent C a C 0 3 .

(flp/n Ca++ x 2 . 5 ) + (ppm Mg++ x 4 . l ) = Total Hardness, ppm as CaCO3 (2.12)

[Hardness m a y also be referred to as "temporary" or "carbonate" hardness and "permanent"

or "non-carbonate" hardness. As the n a m e s imply, the temporary hardness is associated
with calcium and m a g n e s i u m bicarbonate. The permanent hardness refers to other salts of
Ca and Mg, such as the chlorides and sulfates.]

9. Alkalinity

Alkalinity in waters is usually attributed to TABLE 2.7

the p r e s e n c e and inter-relationship of bicar- Ions Causing Alkalinity
bonate, carbonate, and hydroxyl ions. Alka-
linity is N O T a measure of pH. However, pH Ions Causing Alkalinity
each of these ions exists only in a given pH 9.6-14.0 OH and CO.,=
range. A general idea of the ions contribut- 8.3-9.6 HC0 3 - and C0 3 =
ing to alkalinity as a function of the pH of 4.5-8.3 HCO 3 -
the water is as s h o w n in Table 2.7.
T h e s e relationships are not exact, and alkalinity is defined as the capacity of a water to re-
act with hydrogen ions. The distribution of H C 0 3 ~ and C 0 3 = ions as a function of pH is
shown in Figure 3.1 (page 59).

Alkalinity is d e t e n n i n e d by titrating with a standard acid to two different endpoints or pFI


* P Alkalinity. A measure of the number of equivalents of acid required to lower the

pH of the water to approximately 8.3 (the phenolphthalein, or " P " endpoint).
M_Alkalinity. A measure of the number of equivalents of acid required to lower
the pH to approximately 4.5 (the methyl orange or " M " endpoint).


Most natural waters have a pH less than 8.3. and hence have a P alkalinity of zero and
contain no C 0 3 = . This means that they show only M alkalinity, which is attributed solely
to H C C V .
Alkalinity and acidity as a function of pH is shown in Figure 2.2.

|| fl |1
- I x
3 II
i*X 0. a LU

1 Acidity OH Alkalinity
5000 500 50 mg/l CaCOg mg/l CaC0 3 " 50 500 5000
_ (C0 2 escapes quickly Free CO z gas - Hydroxyl Alkalinity
into environment)
Bicarbonate Alkalinity
Free Mineral
- Carbonate Alkalinity
"Acidity (FMA)'
10 11 12 13 14

Figure 2.2 Alkalinity and Acidity R a n g e s

Although absent f r o m surface waters and most aquifer waters, carboxylic acids are c o m -
monly present in produced waters (water produced with oil and/or gas) and are usually re-
ferred to as volatile fatty acids (VFA's). E v e n though they are classified as acids they are
not actually present as acids in produced water because they have ionized to yield a hydro-
gen ion and a volatile fatty acid anion ("VFA a n i o n " ) / - ' ' " For example, in the case of
acetic acid:

CH3 COOH^> H+ + CH3 COO -

Acetic Acid Acetate Ion

The VFA anions c o m m o n l y found in produced water contribute to alkalinity and are listed
in Table 2.8.

Volatile Fatty Acid Anions
VFA Anions Formula
Acetate Ion CH3COO-
Propionate Ion CH3CH2COO-
Butyrate Ion CH3CH2CH2COO-

Alkalinity in produced waters is the sum of the concentrations of the bicarbonate ion and
the V F A anions.

Alkalinity = HCO~ + Acetate + Propionate + Butyrate (2.13)

The acetate ion usually comprises the m a j o r portion of the VFA anions in produced waters
with considerably smaller quantities of propionate and even smaller concentrations of
butyrate ions.



The concentration of the VFA anions in produced water can be substantial relative to the
bicarbonate concentration. W h e n this occurs, the bicarbonate ion concentration determined
by titration with acid will be significantly in error.

Accurate determination of the H C 0 3 ~ concentration in these types of waters requires the

measurement of the VFA anion concentration so that it can be subtracted from the total al-
kalinity. This measurement is not routinely performed and must be requested if desired.
This can be quite important when m a k i n g calcium carbonate scaling calculations.

Graphical Presentation of Results Water Patterns

Water analyses were often expressed graphically. T h e
diagram, or pattern, obtained by graphically plotting the re- TABLE 2.9
sults of a water analysis can often highlight important points Water Analysis
about the analysis that might be missed by simply reading the
toil mg/L mcq/L
report. Pattern comparison is also an easy way to quickly
spot differences in two or more waters. Na + 93 230 4053
Ca++ 5173 258
There are many different water analysis diagrams in Mg~ 620 51
use. However, the Stiff Method^- 5 ' was adopted by the API Fe (Total) 12 0.6
and was probably the most universally used method in the
ci- 153 175 4320
oilfield. It is the only one presented here. A logarithmic plot
HCOjf 195 3
of meq/liter of the various ions is most often used, although
sor 1910 40
linear plots are also used. T h e concentrations of N a + , Ca + + ,
co.r 0 0
Mg + + , Fe (total), CI", H C 0 3 ~ SOA~ and C 0 3 = given in Table
2.9 are plotted as s h o w n in Figure 2.3.

Logarithmic Pattern (meq/1)

Ca \ HCO,
Mg \ SO.

Fe CO,
o o o
s g

o Linear Pattern (meq/1) o"

10 10
100020 1000
100 1.0

Mg so
10 10

Fe co3
1.0 1.0

Figure 2.3 Graphical Presentation of Water Analysis Stiff Method


T h e concentrations of the cations are plotted to the left of the center line, while the anion concen-
trations are plotted to the right of the center line. Water patterns are not as commonly used today as
they were in past years. They are mentioned here and in R P 4 5 ' : : j for historical purposes. Today.jiva-
ter comparisons are more likely to be made by computer programs or by comparing specific constitu-
ents in the water samples.

Water Analysis Reports

Appendix 9 (pages 394-395) contains examples of two water analysis reports. T h e first is pre-
sented on a m o d i f i e 3 A P I W a t c r Analysis Report form. T h e second report presents the same data in a
slightly different f o n n a t used by a commercial water analysis laboratory.

Both reports contain the following:

Sample identification information.
Quantitative analysis of water sample. T h e degree of detail will vary.
A graphical representation of the analysis a water pattern.
The Stiff Method is used in both cases, although the logarithmic pattern is used in the first f o r m
while the linear pattern is used in the second.

T h e API report f o r m is r e c o m m e n d e d for your o w n field analysis work. You m a y find it desir-
able to eventually modify the f o r m to contain space for reporting additional information pertinent to
your o w n operation.

Oil-in-Water Analysis
T h e oil content of a water depends on h o w it is measured. There is no absolute value.

Produced water contains both dispersed oil and '"dissolved oil". T h e dispersed oil can be either
free oil which will separate relatively easily, or emulsified oil which will be more difficult to separate.
The dissolved oil consists of hydrocarbons, phenols, organic acids, and low molecular weight aromatic
c o m p o u n d s such as benzene and toluene.

Normal oil/water separation equipment cannot r e m o v e dissolved organic compounds. However,

values of "oil and grease" include both the dispersed oil and dissolved organic compounds. Hence,
equipment p e r f o r m a n c e should not be j u d g e d by "oil and g r e a s e " values.

Oil and Grease/Petroleum Hydrocarbons

In the determination of oil and grease, an absolute quantity of a specific substance is not mea-
sured. Rather, the amount of substances with similar physical characteristics are d e t e n n i n e d quantita-
tively on the basis of their c o m m o n solubility in a specific solvent. Oil and grease m a y includc hydro-
carbons, fatty acids, soaps, fats, waxes, oils, and any other material that is extracted by the solvent
f r o m an acidified sample unless it evaporates during the test procedure. Oil and grease is defined by
the method used for its determination.

In methods which include a heating stage there is the risk of losing amounts of any hydrocarbon
that evaporates close to the heating temperature. Hence, the results are usually considered to be more
accurate by spectro-photometric analyses rather than by gravimetric a p p r o a c h e s / - ^

113-Trichloro-trifiuoroethane (typically under the trade names of Freon 113, or Arklone-P) was
used worldwide as an extraction solvent for many years, but it is no longer acceptable for environmen-
tal reasons. Its use was banned by the Montreal Protocol (1987, with the 5th revision in 1999; signed



by more than 180 nations) which is aimed at reducing the manufacture and use of substances that de-
plete the o z o n e layer. Previous methodology based on chlorofluorocarbon-113 included:
EPA Method 413.1; a gravimetric (by weight) approach
EPA M e t h o d 413.2; infrared transmission comparison with standards
EPA Method 502E, for "Petroleum Hydrocarbons" by first removing water soluble polar
organics; either gravimetric or infrared.

Total Petroleum Hydrocarbons (TPH) in Water

T h e replacement analytical procedure n o w being used in the U.S. is the EPA Method
1664A'- in which the extraction solvent is n-hexane. The method is briefly as follows. One liter
of water sample is acidified then serially extracted (three times) with n-hexane. The extract is dried
over sodium sulfate and the solvent evaporated. The weighed residue gives the result as n - h e x a n e
extractable material ( H E M ) .

In order to discount the amount of water soluble polar organics, the H E M is redissolved in n-hex-
ane and treated with silica gel. The polar materials adsorb on the silica gel. The solution is then fil-
tered, evaporated and again the residue is weighed. T h e resultant silica gel treated H E M ( S G T - H E M )
reflects the amount of non-polar hydrocarbons in the original water sample.
S o m e limitations of the method'- are:
It is not applicable to organics that volatilize at temperatures higher than 85C. Having a
higher boiling point than Freon, the hexane takes longer to evaporate and with this there is
some loss of light hydrocarbons. Hence, a lower apparent "oil and grease" measurement is
recorded with this method when c o m p a r e d to Method 413.1.
n-hexane is flammable.
The procedure takes m o r e time, since there are three extractions and the water must be
drained f r o m the extraction flask at each stage in order to recover the hexane layer (unlike
the freon which is heavier than water). This adds to cost.
The quick infrared method previously used on offshore platforms cannot be used, since it
relies on measuring the stretching frequency of C - H bonds.
Some organic species that were soluble in Freon are not soluble in n-hexane. Hence, it is
not possible to draw direct comparisons between current results and those determined by
previous extraction methods.
Different procedures and discharge limits apply elsewhere in the world. For example, countries
involved in the North Sea oil and gas industry have signed up to the O S P A R (Oslo-Paris) Convention.

With the demise of "freon", OSPAR is introducing a new reference method (based on ISO
9377-2) oil-in-water analysis from January 1, 2007. It is a pentane extraction followed by gas chroma-
tography-flame ionization detection ( G C - F I D ) analysis. In the meantime, the U.K. North Sea sector is
using a tetrachloroethylene ( T T C E or TC'E) extraction followed by IR spectrometry. In the N o r w e g i a n
sector the authorities refused to allow the use of TTCE. Subsequently, at least one producer (Norsk
Hydro) implemented ISO-9377-2 as of January 2003.

OSPAR Recommendation 2001/1 referred to "total quantity of oil in produced water", which is
interpreted as "the sum of the aliphatic and aromatic portions , '. Hence, it is important to define analy-
sis methods for aromatic c o m p o u n d s and especially specific groups of c o m p o u n d s such as B T E X (ben-
zene. toluene, ethylbenzene and xylene), PAHs (polynuclear aromatic hydrocarbons) and N P D s
naphthalenes/phenanthrenes/dibenzothiophenes). These groups of aromatic c o m p o u n d s are known to


have carcinogenic properties towards h u m a n s and are suspected of other health related problems. They
are toxic and bio-accumulate.

The U.S. EPA lists 16 specific aromatic products^ for which periodic analyses must be submitted,
under the terms of Gulf of Mexico ( G O M ) discharge permits for offshore production platforms.
C h a n g e s in U . K . regulations, within the "Offshore Petroleum Activities (Offshore Pollution Prevention
and Control) Regulations ( O P P C ) 2004, in short O P P C 2004, introduces a mandatory requirement to
analyze these aromatic c o m p o u n d s bi-annually. As well as the 16 listed by the U.S. EPA, a further
eleven alkyl-substituted c o m p o u n d s c o m m o n in oil pollution are included in the list for analysis' 7 ^.

Some of the approved analytical m e t h o d s ' - - i n the U.K. include:

Dispersed oil (aliphatics)

- Current DTI method; a single peak infrared analysis after extraction with
tetrachloroethylene (TTCF. or T C E )
- F r o m January 2007; OSPAR method (ISO-9337-2 [Mod]), n-pentane extraction
then analysis by G C - F I D (Gas Chromatography - Flame Ionization Detection)
- ISO method 11423-1; based on GC separation and MS (mass spectrometry) analy-
PAH and N P D
- Draft ISO standard, ISO WD 14653-2 (1997); GC then MS detection
Organic acids and alkylphenols
- By G C / M S methods
H e a v y Metals
- By ICP/AES (Inductively Coupled Plasma/Atomic Emission Spectroscopy).
Similar approaches will be used in other countries.

Studies in N o r w a y ' - ' h a v e shown a correlation between the dispersed oil concentration and that
of the PAHs and alkylated phenols. The implication is that technologies which enhance dispersed oil
removal, will also remove a large fraction of the PAHs and heavier alkylated phenols.

Oil Content by Colorimetric Analysis

T h e water sample is not acidified. Oil is extracted f r o m a water sample using a dense solvent
such as trichloroethane. The color intensity of the extract is compared with standards prepared with
the same oiTtcTdetermine the concentration.

This is a very useful field technique and is c o m m o n l y used to assess equipment p e r f o r m a n c e ;

however, the sensitivity of the method is limited with light colored oils such as condensates.

Suspended Solids Analysis

Techniques for suspended solids determination and characterization are not covered in API RP
4 5 / - -'' Because of the extreme importance of suspended solids control in water handling operations, a
brief summary of techniques used to determine the amount and nature of suspended solids in a water is
presented in the following sections.



Suspended Solids Concentration

T h e concentration of suspended solids in a water is determined by passing a k n o w n quantity of
water through a m e m b r a n e filter and determining the weight of solids collected on the filter. The
weight of solids collected (mg) divided by the quantity of water passed through the filter (liters) gives
the suspended solids concentration in mg/L. Standard methods for this test are given in N A C E Stan-
dard T M 0 1 7 3 published by the National Association of Corrosion Engineers/- 7 ^

It is strongly recommended that water be flowed directly from the system through the membrane,
when possible, as shown in Figure 2.4. This approach eliminates problems of secondaiy precipitation
associated with sample aging and contact with atmospheric oxygen.

Suspended solids are sometimes collected by flowing a water sample f r o m the system into a clear
plastic cylinder which has been purged with nitrogen or some other inert gas. Water is then displaced
f r o m the cylinder through a 0.45 |_tm m e m b r a n e filter by applying approximately 20 psi pressure to the
cylinder with a nitrogen bottle. This method is often used at low-pressure sample points although it is
sometimes used at high-pressure sample points as well.

T h e use of a pressurized plastic cylinder m a y give acceptable results in s o m e waters but it is not
recommended for general use. It is particularly unreliable for produced waters. The basic problem

Low-Pressure Assembly High-Pressure Assembly

Pressure Gauge Pressure Gauge

(Full Scale = 50-100 psi) (Full Scale = 50-100 psi)

Note: Piping is typically brass

and needle valve is
stainless steel
Tight fitting piece ol glass,
plastic, or stainless tubing
through bored hole in
rubber stopper

Do not measure water volumes

in this (lask. Manual vacuum pump
if required

Pour water into a graduated Graduated Cylinder

cylinder at the end of the test
lor a more accurate

Figure 2.4 Membrane Filter Test A p p a r a t u s


arises f r o m the fact that a large volume of water is transferred f r o m the system into the plastic cylinder
(typically several liters) but only a small portion of the water (often a few hundred milliliters in poor
quality waters) passes through the membrane filter before it plugs and flow ceases. Two things typi-
cally occur:

1. Oil tends to rise to the surface of the water and to coat the walls of the plastic cylinder.
The net result is that the amount of oil passing through the m e m b r a n e filter is less than
occurs w h e n the sample is drawn directly f r o m the system. This effects the permeability of
the filter cake and h e n c e the v o l u m e of water which will flow through the cake at a given
pressure drop.
2. Suspended solids tend to segregate in the cylinder, with the larger particles settling to the
bottom. This gives a non-representative sample.
Although this technique is widely used, and is indeed a N A C E - r e c o m m e n d e d alternate method,
we strongly r e c o m m e n d that suspended solids be collected directly f r o m the system using an apparatus
such as one of those shown in Figure 2.4.

T h e most c o m m o n reason for using the pressurized cylinder method is that system pressure be-
comes irrelevant. Furthermore, the pressure drop across the membrane filter is always the same, which
is very handy w h e n water quality testing (flow rate vs. cumulative volume at a constant AP) is per-
formed. A pressure drop of 20 psi is typically chosen because it is the value recommended by N A C E
for water quality testing.

However, the pressure drop across the m e m b r a n e filter is not important when the sole purpose is
to determine the suspended solids concentration in the water. If the pressure at the sample point is
low. the pressure drop across the m e m b r a n e filter can be increased by using a hand-operated vacuum
p u m p as shown in Figure 2.4.

T h e test is normally conducted using a single pre-weighed m e m b r a n e filter when the suspended
solids content is 1 m g / L or greater. When the concentration is less than 1.0 m g / L , matched weight
m e m b r a n e s or pre-weighed pairs are preferred.

It is critical that the increase in weight of the m e m b r a n e filter due to solids filtered from the wa-
ter be sufficient to provide analytical accuracy. For example, a 47 mm diameter, 0.45 pm pore size,
m e m b r a n e filter weighs about 100 mg. At least 2 mg of solids should be filtered f r o m the water so
that the weight gain will be at least 2%. This m e a n s that the suspended solids concentration deter-
mines the amount of w a t e r which must be filtered to provide the m i n i m u m weight gain required.

Elution Loss

There is one complicating factor when large quantities of water are passed through the standard
cellulose acetate/cellulose nitrate m e m b r a n e filter: the water dissolves a small amount of the filter, re-
sulting in a small loss of weight. This weight loss is referred to as elation loss. This loss is not sig-
nificant when the amount of solids collected on the filter is large. However, elution loss can be quite
significant in waters containing low levels of suspended solids.

It has been c o m m o n practice to deal with this p r o b l e m by running matched-weight filters. Filtra-
tion through two matched-weight m e m b r a n e filters mounted in series subjects both filters to the same
elution loss but all solids are collected on the top filter. Therefore, the differential weight is the sus-
pended solids corrected for elution loss.

However, it is not necessary that the two filters be matched in weight. If both are pre-weighed,
the elution loss correction can be determined from the bottom filter and applied to the top filter.



Another option is to use a different m e m b r a n e material, such as polyvinylidene difluoride

( P V D F ) which is not subject to elution loss.

Chemical Composition of Suspended Solids

An analysis procedure for the determination of solids composition is given in TMO 1 7 3 - 9 2 / 2 /j?>
Determination of hydrocarbon solubles, acid solubles, organics and insoluble residue should be
supplemented by chemical analyses to identify the m a j o r components of each fraction. Of specific in-
terest is iron, sulfate and carbonate, supplemented by determinations of calcium, barium, and strontium
as deemed necessary. Analysis for silica and protein may also be of interest.

W h e n m e m b r a n e filters are used in brines, the m e m b r a n e must first be washed with deionized
water prior to w e i g h i n g and analysis. The purpose of the w a s h is to r e m o v e water-soluble salts which
precipitated in the filter w h e n the water evaporated from the wet membrane during drying. T h e y are
not a part of the suspended solids.

A typical analysis format for suspended solids is s h o w n in Table 2.10.

TABLE 2.10
Artalysis of S u s p e n d e d Solids
Component mg/L mg/L
Total Acid Solubles 1.50
Calcium as CaC0 3 0.13
Iron as FeS 0.40
Unidentified 0.97
Total Organics 3.55
Solvent Soluble (Oil) 1.70
Ignition Loss 1.85
Total Acid Insolubles 0
Total Suspended Solids 5.05

Water Quality
In water injection systems, water quality is a measure of the relative degree of plugging w h i c h
occurs when a given v o l u m e of water is passed through a m e m b r a n e filter of a given pore size. A pore
size of 0.45 pm is most c o m m o n l y used.

T h e National Association of Corrosion Engineers has approved a standard method for running the
test/ Several m o d i f i e d methods are used. However, the basic test consists of forcing a given vol-
ume of water through a filter under constant pressure. T h e cumulative v o l u m e through the filter is re-
corded as a function of time, and the flow rate for each time increment is calculated f r o m the data.
Flow rate is plotted versus cumulative volume throughput on a semi-log plot. The slope of the line in-
dicates the "quality." or degree of plugging which occurred with that particular water sample, as illus-
trated in Figure 2.5.

It is also a measure of the penneability of the filter cake formed on the membrane. T h e steeper
the slope, the less permeable the cake formed by the solids.

The important point to be made about water quality testing is that it is a way of comparing the
relative tendency of different waters to plug a membrane filter. It does not necessarily have any corre-
lation with the tendency of a water to plug the formation.


Cumulative Throughput, mL

F i g u r e 2.5 Water Quality Plot

TABLE 2.11
Interpretation of Water Quality C u r v e s in Figure 2.5
Curve Quality
1 Excellent. No plugging occuned since the flow rate remained constant throughout the test.
y Poorer than Curve 1. The flow rate decreased as the cumulative volume throughput increased in-
dicating plugging of the filter.
3 Poorest of the three curves. The flow rate dropped much more rapidly indicating faster plugging.

The utility of water quality testing lies largely in its use as a comparative test.
Water quality testing can be used to detect changes in a single water at a given point in a
system over a period of time.
Tests can be run at various points through a system to detect changes w h i c h m a y be occur-
ring between the water source and the injection wells.
Different waters can be compared.
Through experience, m i n i m u m water quality standards may be set for a specific area.
Careful correlation of water quality measurements with injectivity data can be very helpful
in determining filtration requirements.
A n y insoluble material in a water will reduce its quality. Corrosion products, water-formed
scales, clay, silt, oil, insoluble treating chemicals, bacterial growths or algae will all contribute to de-
creased quality.

It should be obvious that the quality required will be largely determined by the reservoir perme-
ability. Tight, low permeability zones will require better water than a high permeability reservoir.
What is considered acceptable quality in one area (or zone) might quickly plug a different formation.



Cerini Slope M e a s u r e m e n t s

In 1953, William F. Cerini proposed a novel means of measuring the slope of water quality
c u r v e s / - A l t h o u g h he used a sintered glass disc rather than m e m b r a n e filters for his test work, his
slope measurement technique is still widely used in water-quality work.

The w a t e r quality data is plotted in the normal w a y on two cycle semi-log graph paper. A best fit
straight line is drawn through the points, and the slope is then measured as though it were a linear Car-
tesian plot.

Both the vertical and horizontal components of the slope are measured using the same linear
scale. The slope value is then calculated by dividing the vertical component (y) by the horizontal dis-
tance (x). As long as the curve slopes d o w n w a r d to the right, the slope value is always negative, and
is c o m m o n l y called the Cerini Slope.

T h e Cerini Slope of the curve shown in Figure 2.6 is - 2.4/1.5 = 1.6.

Cerini Slope values are a function of scales used on both the ordinate and the abscissa. There-
fore, slope values can be c o m p a r e d only when the scales are fixed.

Relative Plugging Index

A m o c o developed a method of rating water quality called the Relative Plugging Index, or

Figure 2.6 M e a s u r e m e n t of Cerini Slope


RPI = TSS - MTSN (2.14)

Where: TSS = total suspended solids, ppm

MTSN = millipore test slope number

M i l l i p o r e Test S l o p e N u m b e r ( M T S N )

The M T S N is the Cerini Slope of a conventional water quality curve, with fixed ordinate and ab-
scissa scales:

Vertical " y " axis (Ordinate): 1 cycle/5 inches

Horizontal " x " axis (Abscissa): 500 mL/inch

T h e M T S N is determined as follows:
1. Plot the water quality data, log How rate (mL/sec) vs cumulative v o l u m e ( m L ) , on 2-cycle
semi-log graph paper.
2. D r a w a best fit straight line through the data points in the latter portion of the curve as
illustrated in f i g u r e 2.6.
3. Select t w o points on the curve. T h e coordinates of the selected points are designated as
(Q a , V a ) and (Q b , V b ).
4. Calculate M T S N directly f r o m the data or determine it graphically.
T h e simplest way to determine the M T S N is to calculate it using the following equation:

2500 l o g ( Q J Q , )
MTSN = (2]5)
' a

Use of this equation permits you to use any ordinate and abscissa scales you choose.

If you wish to determine the M T S N graphically, it is necessary to use graph paper with an ordi-
nate scale of 1 cycle/5 inches, which is usually quite convenient, since that is the normal grid size on
2-cycle semi-log graph paper (8-1/2 x 11 sheet).

There are t w o options regarding the abscissa scale:

1. Use an abscissa scale of 500 mL/inch.
2. Use an abscissa scale other than 500 mL/inch and calculate the equivalent distance on a
500 mL/inch scale using the following formula:
V -V
x v(inches) = (2.16)
500 mL/inch
M T S N is determined graphically as follows:
1. Measure the vertical distance between Q a and Qb, y.
2. Measure the horizontal distance between V a and Vb, x. If the scale on the horizontal axis
is other than 500 mL/in calculate the difference between the two volumes in milliliters,
and divide by 500 to obtain the correct value of x in inches.

3. MTSN T h e M T S N is always negative.




Example 2.1: Given the water quality data in Table 2.12:

T h e data is plotted in Figure 2.6. The M T S N of the water quality curve is calculated as follows:
2 5 0 0 X O g ( 2
3^- ' -3/0-7 6
^- 2 5 0 0 x 0
- 4 8
- 0 8
1500-3000 1500

The M T S N can also be determined graphically:

1. y = 2.4 inches 3. MTSN = = ^ = -0.8
x 3.0
3 0 0 0 / - 1 5 0 0 mL . r
2. .Y = = 3.0 inches
500 mL/inch

TABLE 2.12
Water Quality Data
Flow Rate Cumulative Volume
(niL/sec) (mL)
3.40 500
2.90 1000
2.30 1500
1.55 2000
1.10 2500
0.76 3000

Water Quality Rating Guide

A water quality rating guide proposed by A m o c o is given in Table 2 . 1 3 / - L v

TABLE 2.13
Water Quality Rating Guide
RPI General Quality Rating
<3 Excellent
3-10 Good to Fail-
10-15 Questionable
> 15 Poor

Well I m p a i r m e n t P r e d i c t i o n

Another approach to the presentation and inteipretation of water quality data was developed by
Barkman and D a v i d s o n / - They developed methods and theory which can be used to interpret water
quality data obtained with m e m b r a n e filters or cores to predict well impairment f r o m suspended solids.

Although this method can be used to m a k e relative predictions, the calculated half-life values
have proven to be unreliable in practice.

Particle Size Analysis

Solid particles f o u n d in injection waters range in diameter f r o m less than one micron up to sev-
eral hundred microns. Figure 2.7 shows the diameters of s o m e c o m m o n l y encountered particles for

Although many of the particles found in injection waters are approximately spherical, many are not.

Microns ^ Scanning Electron Microscope j Optical Microscope ^ Visible to Naked Eye
(|am) 0.01 0.1 1.0 10 100 1000
i i i mil I I 11 ml l l I 11 III l l l l l III! I I l l l III LLLl

Clay Silt Fine Sand Coarse Sand

Relative Smog Clouds & Fog | Mist Drizzle | Rain
of Tobacco Smoke Milled Flour
Materials Atmospheric Dust

Bacteria Human Hair

Figure 2.7 Typical Diameter of C o m m o n Particles

The description of the " s i z e " of a non-spherical particle is one of the fundamental p r o b l e m s of
particle technology. T h e most c o m m o n approach is to describe the particle by a sphere which will
have at least one property in c o m m o n with it. Thus, a single particle m a y have several "equivalent
sphere sizes" d e p e n d i n g on the method used for its m e a s u r e m e n t , as shown in Table 2.14.

TABLE 2.14
Diameters Used for Particle Characteristicsf 2 - ,7 < )


True dxhxw=1x1x2

Microscope Projected-Area
Diameter dc =1.58

Microscope Maximum
F e r e f 3 Dia. d, = 2.23

Sedimentation Stokes
O = 1.43


Coulter Volume dv = 1.55

Counter Diameter

Sieve Mesh-Size
Diameter n d = 1.00

Diameter O = 1.77

Maximum distance between parallel tangents.

S t o k e s diameter is included in the expression 3 p d, which is the
force (resistance to motion) exerted on a particle.



Note that for the particle s h o w n the " d i a m e t e r " varies by more than a factor of two among the
various measurement techniques.

There are a n u m b e r of techniques which are used to determine the size of particles suspended in
water. T h e y are compared in Table 2.15 and discussed in the following paragraphs.

Microscopic Techniques

The use of visual microscopy is probably the oldest technique used to examine particles in oil-
field waters. However, it is a rather tedious w a y to measure particle size distributions, as m a n y parti-
cles must be examined if the distribution determined is to be statistically valid. It is m o r e generally
used to determine the shape and nature of the particles and to get some idea of their general size range.

Scanning electron microscopy (SEM) is a popular method of particle examination. Much higher
magnifications are possible than with conventional microscopy, and the depth of field is far superior.

Figure 2.8 is an S E M photograph of suspended solids filtered f r o m A e g e a n seawater, while Fig-

ure 2.9 is an S E M photo of solids filtered f r o m North Sea Water.

Both waters contained virtually identical concentrations of suspended solids, but the shapes and
types of particles were quite different.

Figure 2.8 S u s p e n d e d Solids Filtered from Figure 2.9 S u s p e n d e d Solids Filtered from
Aegean Seawater north S e a Water
Magnification = 1000 x Magnification = 1000 x
(Courtesy Brown 5 Root, Inc.) (Courtesy Nalfloc Ltd.)

TABLE 2.15
Comparison of Commonly Used Particle Analysis T e c h n i q u e s

Size Maximum
Technique Range Sample Requirements Data Produced Instrument Portability Magnification

Visible Usually filtered solids on surface of Particle shape and size. Easily portable. 800 x
Light membrane filter although drops of Requires AC power for (Dry objective)
Microscope water can be examined. illuminating lamp. 1500 X
(Oil immersion)

Scanning Filtered solids on surface of membrane Particle shape and size. Permanent laboratory 100 000 x
Electron filter. Must be plated with Au or Pd installation.
Microscope prior to examination. Sample examined
under vacuum. CO
Coulter 0.5 pm Representative water sample. Counts Particle size distribution or Portable.
Counter to particles in liquid. Water must have total number greater than a Requires AC power.
400 mm sufficient TDS to be electrically manually set threshold. o>
conductive, z
Spectrex l pm Representative water sample. Counts Total number of particles Portable. z
Particle to particles in liquid. greater than a manually set Requires AC power. r<
Counter 100 pm threshold. CO

HIAC and 2 pm Flowing water sample piped through Particle size distribution Portable.
Royco to flow cell or batch sample. Most require AC power.
Counters 900 pm Some battery powered
units available.


Coulter Counter

Two clectrodes are immersed in a beaker of the w a t e r of interest, which m u s t contain sufficient
dissolved ions to easily conduct electrical current.

T h e negative electrode is located inside a glass tube which is sealed except for a tiny hole or ori-
fice on the side of the tube. The positive electrode is located in the water sample beaker. (Figure 2.10)

F i g u r e 2.10 Coulter Counter Schematic (Courtesy Coulter Eiectonics Ltd.)

A constant electrical current is passed f r o m the positive electrode to the negative electrode
through the orifice. W h e n a non-conductivc particle passes through the orifice it causes a change in
electrical resistance between the two electrodes which is proportional to the v o l u m e of the particle.

A fixed v o l u m e of water containing suspended particles is forced through the orifice. As each
particle passes through the orifice, the increased resistance results in a voltage pulse w h i c h is propor-
tional to particle volume. The series of pulses produced by a series of particles passing through the or-
ifice are electronically scaled and counted, yielding a particle size distribution.

It must be realized that the particle " d i a m e t e r " given by the counter is the diameter of a fictitious
sphere with the same volume as the real particle. T h u s , the more non-spherical the panicle, the greater
the error.

It should be noted that the Coulter Counter cannot distinguish between solid particles, oil drop-
lets and gas bubbles. Therefore, w a t e r must be degassed prior to analysis and accurate particle counts
cannot be obtained in waters containing oil droplets.

Light Scattering C o u n t e r s

Instruments such as those manufactured by H I A C ( H I A C Instruments Div., Pacific Scientific Co.,

Montclair, California) and Royco (Royco Instruments, Inc., M e n l o Park, California) use the principle
of light absorption/total scatter, or light blockage, to detect particles in a fluid. Water is flowed
through a sensor cell and as each particle passes through the intense beam of light in the sensor, it
scatters the light. The instrument measures the magnitude of each pulse of scattered light which is
proportional to the surface area of the particle. T h e particle diameter given by the instrument is the
diameter of a sphere with the same surface area as the particle.


The Spectrex Laser Particle Counter (Spectrex Corp.. Redwood City, California) operates on the
same principle, but uses a laser as a light source. Furthermore, in this instrument, the water sample re-
mains stationary while the laser scans the sample. A sample container of water is placed in the instru-
ment and the laser is activated. The beam revolves rapidly as s h o w n in Figure 2.11, thus scanning a
fixed volume of the sample. It gives a count of the total number of particles per milliliter above a cer-
tain manually set threshold between 1 pm and 100 p m .

As in the case of the Coulter Counter, light-scattering p a n i c l e counters cannot distinguish be-
tween solid particles, oil droplets and gas bubbles. Therefore, water must be degassed prior to analysis
and accurate particle counts cannot be obtained in waters containing oil droplets.

M e m b r a n e Filter A n a l y s i s

M e m b r a n e filters should not be used to estimate the size of suspended solid particles in water, as
the results have little meaning.

Two techniques are used.

1. Filter a water through two m e m b r a n e filters in series, the second filter having a smaller
pore size than the first, and measure the amount of suspended solids on each filter. The as-
sumption is that the results can be used to estimate the concentration of particles in differ-
ent size ranges.

For example, filter one liter of water through a 5 pm pore-size filter followed by a 2 pm
pore-size filter as s h o w n in Figure 2.12. Determine the weight of suspended solids on
each filter.
The results are erroneously interpreted as follows:

| Particle Diameter Concentration

5 > pm 10.5 mg/L
| 5 pm > D > 2 pm 3.7 mg/L J




10.5 mg
Suspended 5 (.im
Solids Pore-Size
' Membrane

3.7 mg
Suspended 2 urn
Solids Pore-Size

1 Liter

Figure 2.12 Filtration Through M e m b r a n e Filters in S e r i e s to Determine Particle S i z e s

This concept is analogous to a sieve analysis for sizing diy particles. However, the set of
sieves is shaken vigorously throughout the sieving operation which prevents bridging and
gives each particle in the sample the opportunity to pass through a given mesh-size.

Particle bridging occurs when particles are filtered f r o m a liquid slurry which often results
in the removal of particles smaller than the m e m b r a n e pore size. The filter cake f o n n e d on
the m e m b r a n e becomes the filter.
2. T h e second technique employed is to filter a given water through a m e m b r a n e of a given
pore size. T h e n a second sample is filtered through a different pore size. Consider the
example s h o w n in Figure 2.13.

Water Sample Water Sample

No. 1 No. 2
10.5 mg 14.2 mg
Suspended Suspended
Solids Solids
5 jam 2 nm
Pore-Size Pore-Size
Membrane Membrane
T Filter

1 Liter 1 Liter
of of
Water Water

Figure 2.13 Filtration Through M e m b r a n e filters in Parallel to Determine Particle S i z e s


The results are erroneously interpreted as follows:

Particle Diameter Concentration

D > 5 pm 10.5 mg/L
1 5 pm > D > 2 pm 14.2 - 10.5 = 3.7 mg/L j

Once again, the assumption is m a d e that all particles with diameters less than the pore-size of the
m e m b r a n e filter being used will pass through the filter. However, this is not true because of particle

Filtration of particles f r o m a liquid slurry is a process which is distinctly different f r o m a sieve

analysis of dry particles and should never be used to estimate particle sizes.

Turbidity is an optical property of a liquid which is related to the ability of undissolved particles
to scatter light.

In most cases we are interested in a correlation between turbidity and suspended solids content.
Unfortunately, no general correlation is possible because turbidity values are dependent upon the size,
color, shape and refractive index of the particles, as well as the refractive index of the carrier medium.
It is also dependent upon the wavelength of the incident light and the orientation of the scattered light
detector, which means that different types of instruments can give different readings. A correlation be-
tween turbidity and suspended solids concentration can be m a d e for a specific water containing a given
quantity and size distribution of suspended particles. However, since turbidity values are a function of
particle size, a change in particle size distribution will alter the measured turbidity.

Figure 2.14 illustrates the relationship between particle size and the amount of scattered light for
a constant weight concentration of suspended particles. The optical response m a y be divided into three

R e m e m b e r that these relationships apply when the concentration of suspended solids remains
constant and only particle size changes. A b o v e 1.0 p m , for example, the amount of light scattered is
greater for a larger particle because it is proportional to surface area. However, as particle size in-
creases, it takes fewer particles to give the same weight of suspended solids. The net result is a de-
crease in the total amount of light scattered.

When a ray of light hits a particle it is scattered in all directions. However, it is not scattered uni-
formly in all directions, with the result that the scattered light reading obtained for a given particle will
be a function of the position of the detector.

For particles less than 0.1 pm diameter, the forward and backscattered intensities are equal, with
half that intensity in the perpendicular direction (side scatter) as shown in Figure 2.15.

The pattern is not affected by particle shape as long as no dimension of the particle is greater
than 0.1 pm.

For particles larger than 0.1 p m , the amount of f o i w a r d scatter increases relative to the amount of
back and side-scatter (Figure 2.16).

The shape and orientation of non-spherical particles will m o d i f y the pattern, but in practical situ-
ations, the r a n d o m orientation of the particles will generate essentially the s a m e results as spherical
particles of the same average volume.


Figure 2.14 Scattered Light as a Function of Particle Size (Courtesy Sigrist-Photometer AG)

TABLE 2.16
Scattered Light vs. Particle Size
Particle Size Amount of Scattered Light
< 0 . 1 jim Increases with particle size
0.1 jam-1.0 pm Complex transition zone
> 1.0 |im Proportional to total surface area. Decreases
with particle size


Diameter / Wavelength
= 0.1, 1.0, 10.0

Not To Scale

Figure 2.16 Angular Distribution of Scattered Light for Different Particle Diameters
(Courtesy Monitor Technology, Inc.)

There are many turbidity instniments on the market. T h e majority used in w a t e r injection sys-
tems either measure side scatter or forward scatter. An instrument which measures side scattered light
at an angle of 90 to the incident light is called a nephelometer. Forward scatter instruments usually
measure the scattered light at an angle of 15-30 to the incident light beam.

Forward Scatter
Incident Light
90 Side


Figure 2.17 Scattered Light Direction

Selection of an instrument for a particular water should be based on a k n o w l e d g e of the nature

and size distribution of the suspended particles, and a careful investigation of the response characteris-
tics of each instrument being considered.

T h e units in which turbidity m e a s u r e m e n t s

are reported can be extremely confusing.

T h e father of m o d e m turbidimeters is the

Jackson Candle Turbidimeter. It consists of a verti-
cal glass tube, graduated in units of length, and a
candle as s h o w n in Figure 2.18.

A sample of liquid is poured into the tube un-

til the candlc can no longer be clearly seen. The
height of liquid at which this occurs is dependent
upon its turbidity. A standard table is used to con-
vert this height into Jackson Candle Units (JCU's),
also called Jackson Turbidity Units (JTU's). With
the advent of m o d e m turbidimeters, other units be-
gan to appear. These units are based on calibra-
tions made with standard suspensions of materials
Figure 2.18 Jackson Candle Turbidimeter
such as kieselguhr (diatomaceous earth or SiOi),



fullers earth, or f o r m a z i n (a polymer suspension). Of the many units in use, the following are the most
FTU Formazin turbidity unit. An instrument is calibrated in F T U ' s with a standard sus-
pension of formazin. Unfortunately, at least three different "standard"" suspensions are
used, resulting in three different formazin units.
N T U Nephelometer turbidity units are used solely for nephelometers. Nephelometers
can be calibrated with kieselguhr, fullers earth or formazin.
ppm Parts per million. Actually ppm of the calibration suspension, which usually is
diatomaceous earth (SiO;>) or fullers earth. Its use w a s originally intended to detect break-
through of DE particles downstream of DE filters. It is not the concentration of the sus-
pended solids in the water.
To make matters m o r e confusing, readings taken with a particular instrument in a given water
may not agree with readings taken with another type of instrument in the same water, even when the
same turbidity units are being used. This can result f r o m differences in calibration techniques, instru-
ment design or the angle at which the scattered light sensor is located. Conversion a m o n g different
unit systems must be made with extreme caution.

In summary, turbidimeters are extremely valuable instruments for monitoring suspended solids
levels in water systems. However, a separate correlation between suspended solids concentration and
turbidity values must be made for each water using a specific turbidity measuring instrument.

Field Water Analyses

As previously stated, certain water properties change very quickly after sampling. The determina-
tions listed in Table 2.17 must be made on-site for m a x i m u m accuracy:
T h e suspended solids should be collected in the field with a m e m b r a n e filter by allowing a
stream of water to flow through the filter. The filter paper and the collected solids are nor-
mally taken to a laboratory for analysis.
Particle size distribution carried out with either a Coulter Counter or a light scattering de-
vice must be performed on a fresh sample to minimize the effect of precipitation of solids
after sampling.
The culture media should be inoculated in the field immediately after sampling, if possible.
(See Chapter 5).
It is difficult to overemphasize the importance of field analyses. Certain properties may begin to
change within minutes; others will be stable for several hours; and some are stable indefinitely.

TABLE 2.17
On-Site M e a s u r e m e n t s
Paremeters to be Measured On-Site
PH Temperature
Carbonate Ion Suspended Solids Concentration^"
Bicarbonate Ion Particle Size Distribution^ : >
Dissolved Oxygen Turbidity
Dissolved Carbon Dioxide Water Quality
Total Sulfides as H 2 S Bacteria Counts' 231


For example, the pH of produced water will usually begin to rise immediately after the sample is
taken. This results f r o m the fact that dissolved acid gases begin to escape from solution as soon as the
sample is removed f r o m the system due to the reduction in pressure.

The rise in pH along with an increase in temperature may result in the formation of calcium car-
bonate scale. Bacteria in the sample may begin to multiply within a f e w hours or they may begin
to die depending on specific conditions. Little or no F b S or C O : m a y be found in the sample after
a short time period.

If the sample contains dissolved iron, exposure to atmospheric oxygen will result in the precipita-
tion of the iron due to reaction with oxygen.

4 Fe++ + 10 H20 + 0: 4 Fe(OH)}+8H+

The generation of hydrogen ions will cause the pH of the sample to drop and lead to an error in
the measured value. This is yet another reason w h y the pH should be measured on-site in a flowing

Thus, if the sample were transported to a laboratory and analyzed after several days, the resulting
analyses would give a very distorted picture of the water as it actually existed in the system, because
of the changes in water composition which took place after the sample was removed f r o m the system.
Had it been a sample of f r e s h surface water or a sample of seawafer, different changes would be antici-
pated. However, regardless of the water source, field analyses are absolutely necessary for an accurate
and m e a n i n g f u l assessment of oilfield water problems.

Field water analysis kits are available f r o m several manufacturers, as are oxygen analysis equip-
ment, pH meters, and membrane filter equipment. This equipment, and s o m e o n e skilled in its use, is
required during initial sampling, for monitoring of system p e r f o r m a n c e , and for trouble shooting. If
you or s o m e o n e in your organization do not h a v e field analysis skills, then reliable outside personnel
should be retained to do the work for you. T h e important point is that on-site measurement of certain
water properties is an integral part of the successful design and operation of a water injection system.
M a k e certain that this point is not overlooked, and that the measurements are carried out by
experienced personnel.

General Instructions for Making Analytical Determinations in the Field

T h e a x i o m that "cleanliness is next to Godliness" is the first rule of life to an analytical chemist.
Although we may occasionally get a bit sloppy in performing analyses in the field, it is important that
we follow this rule as closely as possible in our o w n analytical work.

The laboratory analyst uses glass beakers, flasks and burettes which he carefully washes and
cleans between each usage. This m o d e of operation is unsuitable in the field. Experience has shown
that the use of plastic beakers (disposable), disposable plastic or glass syringes, plastic bottles (dispos-
able) and glass bottles where necessary (disposable) is much easier. Our motto is " N e v e r use anything
twice." Throw it away rather than risk contaminating the sample.

T h e reasoning behind this preoccupation with cleanliness is extremely sound. If you go to a

great deal of t i m e and trouble to obtain an uncontaminated, representative water sample and
then use a dirty beaker or burette to run your analysis, you have just wasted a lot of time. Your
sample is contaminated by any material which may have remained in the analytical ware f r o m the pre-
vious analysis. It m a y or m a y not have mined your present analysis, but the chances are good that it



has. There are m a n y sources of uncertainty under the best of circumstances. Do not use dirty labora-
tory w a r e and avoid at least one of the k n o w n sources of error.

It is permissible to re-use some of the disposable beakers or syringes w h e n analyzing a given wa-
ter sample. Between determinations they should be carefully rinsed with deionized water or preferably
with some of the water sample of interest. They must be thrown away after completing the analysis of
the sample.

Some discretion is required in the question of cleanliness. If you want to take the time and trou-
ble to clean up y o u r plastic beakers and syringes between determinations, you can do so, just as is
done with glassware in a laboratory. However, it is simply often very inconvenient or impossible to do
so in the field. The use of inexpensive, disposable materials gives you the alternative of starting with
new, clean materials each time. This procedure o f f e r s the dual advantage of saving you time and trou-
ble, and of minimizing the possibility of contamination.

Use of Syringes
Syringes are convenient for measuring water sample volumes in the field instead of pipettes or
burettes. A sample v o l u m e of 10 mL or 25 mL is most c o m m o n .

Syringes can also be used for titrating instead of a burette. One-mL and 5-mL syringes are most
common for this purpose.

You should always use at least one-half of the v o l u m e of any syringe, whether y o u are measuring
a sample or titrating. With less than half of a syringe volume, you lose accuracy. U s e a smaller
syringe instead.

Normally, w h e n measuring a water sample volume, a needle is not used with the syringe. Titra-
tions are made with syringes equipped with needles so that the titrant may be dispensed drop-by-drop.
The use of syringes is detailed in Appendix 10 (page 396).

Final Point
Poor Sample Poor Data => Poor System Management

2.1 Ostroff, A. G.: Introduction to Oilfield Water Technology, Second Edition, National Association of Cor-
rosion Engineers, Houston (1979).
2.2 API RP 45-98: API Recommended Practice for Analysis of Oil-Field Waters, American Petroleum Insti-
tute, Dallas (1998).
2.3 Kemmer, F. N. (Editor): The Nalco Water Handbook, McGraw-Hill, New York (1979).
2.4 Stephenson, M. T.: "Components of Produced Water: A Compilation of Industry Studies," J. Pet. Tech.,
(May 1992), 548.
2.5 Stiff, H. A.: "The Interpretation of Chemical Water Analysis by Means of Patterns," Trans. AIME, Vol.
2.6 Bradley, B. W.: Two Oilfield Water Systems, Robert E. Krieger Publishing Co., Malabar, Fl. (1987) 221.
2.7 USEPA, "Method 1664: n-Hexane Extractable Material (HEM) and Silica Gel Treated n-Hexane
Extractable Material (SGT-HEM) by Extraction and Gravimetry (Oil and Grease and Total Petroleum Hy-
drocarbons)", at www.epa.gov/waterscience/methods/ 1664fs.html.
2.X USEPA, "Analytical Method Guidance for EPA Method 1664A Implementation and Use (40 CFR part
136)", at www.epa.gov/waterscience/methods/1664guide.pdf


Oily Water

Figure 3.18 Production Systems Potential Scale Inhibitor Injection Points

Effective inhibitor concentrations are typically between 5-50 mg/L

U s e correctly sized chemical dosing p u m p s
Use correctly sized injection quills (made of suitable metallurgy).
For the protection of downhole equipment, scale inhibitor application can be:
By injection into the lift-gas feeding to gas-lifted wells
Through capillary tubing (Macaroni String)
Via "squeeze treatment".

As well as being limited to gas-lifted production facilities, the first approach needs an inhibitor
solvent system which will not flash {evaporate) under the conditions prevailing in the gas-lift lines.
Also, there won't be any treatment below the bottom gas-lift mandrel. The use of thin-bore (capillary)
tubing to inject chemicals downhole might at first appear to be a sensible option. However, it is ex-
pensive to install, in the field there are practical p r o b l e m s during its installation and the thin tubing is
prone to damage. Squeeze treatment has been successfully used for m a n y years.

Scale inhibitor squeeze treatment consists of p u m p i n g a solution of the scale inhibitor (typically 5
to 2 5 % dissolved in water) into the production formation rock. T h e most commonly used water is sea
water, especially offshore. This is followed by an over-flush of fluid, often water, but diesel or stabi-
lized crude oil can be used. T h e basic principle is to displace the scale inhibitor solution into the res-
ervoir so that it is retained and then f l o w s back over a long period of time (months, or years). H o w -
ever, time is not the best measure of "Squeeze Life", since water rates change with time. Hence,
squeeze life is n o w usually expressed in terms of cumulative volumes of water treated/produced. T h e
well is usually "squeezed" again w h e n the scale inhibitor residual concentration in the produced water
falls to a pre-agreed level, below which scale formation might resume. It is important to note that
great care is essential in the planning and execution of scale inhibitor squeeze treatments.



A squeeze treatment w o u l d be considered successful when the following have b e e n achieved:

Inhibitor solution was p u m p e d into the reservoir without any unexpected or unacceptable
back-pressure (that might occur due to plugging of the well)
The well was returned to production at an acceptable oil (and water) production rate
T h e produced water contains sufficient scale inhibitor on a continuous basis for an ex-
tended period of time.
Benefits resulting f r o m successful squeeze treatments include the minimizing of scale deposition
in the perforations, subsurface safety valve (SSSV), tubing string, c h o k e and topside flow-lines.

Within water injection systems there is potential for scaling at points downstream f r o m the
deaeration process, depending on the source water chemistry. However, the m a i n worry tends to be
"downhole" w h e n the injection water mixes with formation water. Even then continuous inhibitor in-
jection is not considered to be economically viable in the long term. A c o m p r o m i s e approach is to in-
ject scale inhibitor at a typical rate of 25 mg/L into the first 0.5-1.0 million barrels of injected water.

Organic scale inhibitors are supplied as liquids with the exception of certain polymers which can
be obtained in p o w d e r form. The powder is dissolved in water prior to application.

Organic inhibitors are applied to surface facilities and injection wells by continuous injection us-
ing standard chemical pumps. It is essential that p u m p s be maintained and checked frequently to en-
sure that the proper amount of chemical is b e i n g injected. This is especially important for scale inhibi-
tors since they must be injected on a continuous basis to be effective. Therefore, the success of any
scale inhibition program will depend on the percentage of time which the pump is operating properly.
If the chemical p u m p is down half of the time, the most effective scale inhibitor cannot be expected to
give good results.

In deep w a t e r sub-sea completions, the low temperatures and high pressures often lead to the
need for hydrate control. Hydrates are ice-like solids c o m p o s e d of cages of water molecules stabilized
by the presence of small hydrocarbon molecules trapped within these "cages". Methanol, mono-ethyl-
ene Glycol ( M E G ) , or tri-ethylene glycol (TEG), are used to suppress hydrate formation. T h e presence
of these liquids adversely affects scale inhibition. There are two main effects. Firstly, the carbonate
and sulfate scale salts are even less soluble in the presence of these organic liquids, especially with
methanol. Secondly, standard scale inhibitors tend to be less soluble in the water/methanol (or glycol)
mixture, so their effectiveness is reduced. It is necessary to work with the speciality chemical suppli-
ers in order to select suitable scale (and corrosion) inhibitors for these systems.

The m i n i m u m effective scale inhibitor concentration is often difficult to determine with precision.
Some c o m m o n l y used monitoring methods are:

Pipe Spools or Nipples

Often used for corrosion monitoring, spools are simply short sections of pipe installed at
various locations which may be removed and inspected for scale buildup. If scale is pres-
ent, the thickness can be m e a s u r e d or the scale can be scraped off and weighed. In order
to be useful, the spool must be easily removable f r o m the system. In some cases it is rela-
tively simple to depressure the system and r e m o v e the spool. In other cases the system
cannot be easily depressured and a bypass around the spool should be installed so that op-
eration can continue while the spool is being removed.


Scale C o u p o n s

Scale coupons m a y be used to detect scale buildup. They are similar to corrosion coupons
except that they are perforated with a m i n i m u m of 6 to 8 holes. T h e c o u p o n is placed in
the system with the flat side facing the direction of flow. This induces turbulence which
accentuates any scaling tendencies. The coupon is weighed before and after exposure to
the fluid and the amount of scale determined by the difference in the two weights.
Visual Inspection

Although not too quantitative, this method is quite reliable. Look inside tanks, break open
lines, look inside valves, look f o r spots where you might expect scale and inspect them
Infrared T h e r m o g r a p h y
Infrared t h e r m o g r a p h y has been used to detect and monitor internal scale accumulations in
North Sea production facilities. T h i s technique is used to measure the surface temperature
of piping and vessels which handle hot p r o d u c e d fluids. Internal scale accumulations will
cause the external temperature to be cooler than normal due to the fact that scales have a
lower themial conductivity than steel.
Increased Pressure Drops in the System

The flow rate must also be measured.

Decreased Injectivity

Injectivity decline is evidenced by increasing injection pressures to maintain injection rate.

This indicates formation p l u g g i n g which m a y be due to scale particles. It also may be due
to other causes.

Non-Chemical Scale Control Devices

Various devices are available which are claimed to prevent scale deposition. Water is passed
through the device, and scale is prevented by rather vaguely described mechanisms. There are three
m a j o r categories of devices:

Magnetic Treatment Devices

It has been claimed that a magnetic field (imposed by either a permanent m a g n e t or an electro-
magnet) causes alterations in the structure of w a t e r or the dissolved ions, thus preventing sca\&.(3J7'3-40>
A more realistic m e c h a n i s m is p r o p o s e d ^ ' ' ' whereby the magnetic field causes increased collisions be-
tween the oppositely charged ions, leading to precipitation of the scaling salts. However, since the



Scale Identification
The success in removing scale accumulations will be determined to a great extent by your knowl-
edge of the scale composition. Identify the p r o b l e m correctly and you have a much greater probability
of solving it.

We have already covered methods of predicting which types of scales might be expected for a
particular water under certain conditions. However, the only reliable method of assessing the problem
is to obtain a sample of the scale or p l u g g i n g solids and analyze it.

If scale is occurring in surface facilities, a sample can be obtained by opening the system and
scraping out some of the scale. Samples can be obtained f r o m an injection well by b a c k f l o w i n g the
well or by running a bailer.

T h e general scheme for identifying the components of a scale sample is much the same whether
performed in a commercial laboratory or in the field. T h e primary difference is that a laboratory anal-
ysis will usually give the amount of each component present, where a field analysis is very qualitative.
A quantitative scale analysis procedure is given by O s t r o f f ^ ^.

Field Scale Analysis

In a field scalc analysis we attempt to guess the composition of the solid by doing the following:
Soak the sample in a solvent to dissolve any hydrocarbons. Watch to see if the solvent
gets darker in color.
C h e c k to see if the sample is magnetic. If it is strongly magnetic, it probably contains a
m a j o r amount of F e j O ^ (magnetic iron oxide). If it is weakly magnetic, it may contain a
small amount of FejO-t, or it m a y be iron sulfide.
Place the sample in 15% HC1. N o t e whether or not a violent reaction occurs. Note any
smell. ( H i S indicates FeS.) N o t e the color of the acid. If it turns yellow, an iron com-
pound is indicated.

It should be noted that w h e n iron sulfide is placed in contact with air, it will oxidize and
convert to an iron oxide. This m e a n s that the composition of scales or deposits originally

TABLE 3.10
Qualitative Analysis of Scales

Soluble Soluble in HC1 Soluble

Component in Solvent Magnetic Reaction Acid Color Smell in Water
Hydrocarbon Yes
CaCOj Violent
CaS0 4
BaS0 4
SrS0 4
FeS Weak Strong Yellow Stinks !(H : S)
Fe 2 0 3 Weak Yellow
Fe 3 0 4 Strong Weak Yellow
FcC0 3 Very strong Yellow
NaCl Yes
Sand, Silt, Clay



containing iron sulfide is a function of the length of time the deposit has been exposed to
air. Analyses performed several days or weeks after the sample has been removed from
the system will frequently find predominantly iron oxide and little or no iron sulfide.
C h e c k the solubility in water. N a C l is water soluble.

Qualitative identification of the components present can be summarized in terms of the preceding
properties in Table 3.10.

Note: Neither sulfates nor sand, silt or clay reacts with anything. A m a g n i f y i n g glass may help
to recognize sand grains or spot sulfate crystals. However, if nothing happens to the sample w h e n you
run a field analysis, send a fresh sample (not the one you worked on) to a commercial laboratory.

A f t e r the sample has been identified a s c h e m e to dissolve or remove the scale or deposit can be
formulated. Chemical removal of scale follows the s a m e principles as scale analysis. Basically, the
problem is to find something that will dissolve the various components of the deposit. If it consists
primarily of a single component, such as calcium carbonatc, then a single solvent (hydrochloric acid in
this case) is sufficient. If it contains more than one component, a series or combination of treating
chemicals m a y be dictated. In addition, mechanical removal methods m a y be used, either alone or in
conjunction with chemicals.

Scale Removal Chemicals

Although hardly a scale, hydrocarbons are often present and can greatly interfere with the action
of acid or other scale removal chemicals. Acid will not react with oil-coated scale! A hydrocarbon
solvent is needed to remove any oil, paraffin or asphaltic materials on the scale so that the chemical of
your choice can get at the solid scale.

Selection of a solvent for hydrocarbons is usually based on trial and error. However, we do
know that the more asphaltic the oil or deposit, the more aromatic the solvent should be. The best ap-
proach is to consult your chemical supplier.

Hydrocarbon solvents can be applied as a pre-wash. They have also been emulsified in hydro-
chloric acid to improve its e f f e c t i v e n e s s . ^ ' ^

In addition to simple organic and inorganic solvents, various complex proprietary solvents are
available which can be used alone to remove 0 \\< 3 - 47S - 48 > or can be miscibly mixed with hydrochloric
acid to improve its effectiveness in penetrating h y d r o c a r b o n s / - 1 ^ ' T h e s e materials range f r o m mix-
tures of alcohols to micellar solutions.

Calcium Carbonate

Hydrochloric Acid
Hydrochloric acid is recommended as the cheapest and easiest w a y to dissolve C a C 0 3 scale un-
der most conditions. Concentrations of 5, 10, or 15% HC1 are normally used.

A corrosion inhibitor must be added to the acid to keep it from dissolving the pipe. You should
think of the useful lifetime of an acid inhibitor in terms of hours rather than days or weeks. O n c e the
inhibitor ceases to be effective, the pipe will be subject to extremely severe p i n i n g which often results
n perforation of the pipe wall. Therefore, it is essential that all acid be immediately flushed f r o m the
system once the cleanout is completed.


A surfactant is often added to help remove any oil film f r o m the scale, but a pre-wash with a sol-
vent or addition of one of the previously mentioned proprietary solvents is_preferred unless the amount
of oil is extremely small. Acid will either not react at all or it will react veiy slowly if the scale is

Figure 3.20 shows the effect of a coating of light oil on the reaction rate of 10% HC1 with
CaCO.} scale. The scale w a s 1/8" thick on the internal surface of a lA" x 1" nipple.


0) I- Water Wet Scale

Q. 80

E Oil Wet Scale

// /
c 60
CD 1
a: 40
8 20
0 / I I I I i I I
c 10 20 30 40 50
Time (min.)

Figure 3.20 Reaction Rate of HCI with C a C 0 3 Scale' 3 - 5 '

Table 3.11 gives the amount of CaCO^ which will be dissolved by different concentrations of

TABLE 3.11
Reaction of HCI with Calcium Carbonate
CaCOj Dissolved
% HCI (lbs/gal of acid)
5 0.6
7.5 0.9
10 1.2
15 1.8
28 3.7

Figure 3.21 is helpful in calculating the amount of 15% HCI required to dissolve various acid
soluble scales.

Chelating Agents
It is possible to dissolve calcium carbonate scale with chelating agents, such as EDTA
(ethylenediaminetetraacetic acid). They function by reacting with the calcium ion to f o r m a water sol-
uble complex, thereby effectively r e m o v i n g the ion from solution.

W h e n calcium carbonate scale is in equilibrium with a water which contains calcium ions and
carbonate ions, then the equilibrium can be described as:

CaCO} Ca++ + CO~



CaCCX 95
For Fe-Og use 318 gal 15% HCI/ft of scale
FeS 180






Z 0.008
= 0.006


0 2 4 6 8 10 12 14 16 18 20 22 24 26
Pipe Diameter - ID, in.

Scale on Inside of Pipe


6 8 10 12 14 16 18 20 22 24 26 28
Pipe Diameter - ID, in.

F i g u r e 3.21 Determination of Scale Volume Inside Pipe(3S0)


If a chelating agent is added to the water, s o m e of the calcium ions will be chelated, resulting in
a decrease in the C a 1 ' ion concentration. When this occurs, s o m e of the C a C O i scale will dissolve and
replace the Ca"1"1" ions lost to the chelating agent in order to maintain the equilibrium of the reaction.
Chelation can also occur at the solid surface of the calcium carbonate scale, taking calcium ions di-
rectly f r o m the crystal lattice.

The basic disadvantages to the use of chelants to dissolve calcium carbonate scale are:

Chelating agents are not efficient "solvents." For example, it takes 7.4 ppm EDTA to che-
late 1.0 ppm C a ' V
The chemicals are relatively expensive.

Chelating agents are seldom used to r e m o v e calcium carbonate scale in nomial oilfield opera-
tions. However, their use has been reported in Prudhoe Bay/3 "1'

Calcium Sulfate (Gypsum)

Hydrochloric acid is not a good solvent for CaS04. T h e m a x i m u m solubility of C a S 0 4 in HCI is

only 1.8 wt % at 25C and atmospheric p r e s s u r e / - 5 5 - '


Inorganic Converters

Inorganic converters are usually carbonates or hydroxides which react with calcium sulfate and
" c o n v e r t " it to acid soluble calcium carbonate or calcium hydroxide. The conversion treatment is then
followed by a hydrochloric acid treatment to dissolve the resulting calcium carbonate or calcium hy-

A good example of this type of chemical is a m m o n i u m carbonate which has been marketed under
several trade names. T h e reaction is:

CaS04 + (NH4 )2 CO, (NH4 )2 S04 + CaCO,


The calcium carbonate is then dissolved with HCI:

CaCO3 4- 2 HCI H20+C02 T + CaCl2

The COt liberated by the acid reaction h e l p s to mechanically dislodge any remaining deposit. In-
organic converters are not r e c o m m e n d e d for dense deposits and are seldom used.

Organic Converters

Organic converters such as sodium citrate, potassium glycollate, and potassium acetate are also
u s e d / 1 5 - 5 ' T h e s e materials react with calcium sulfate deposits causing them to swell and b e c o m e soft
so they m a y be removed easily by Hushing with water. T h e s e chemicals are expensive, require several
hours contact time to work on thick deposits, and should be checked on an actual sample of the scale
before application, if possible.



Chelating Agents
The same c o m m e n t s apply here as for use to r e m o v e calcium carbonate.

Sodium Hydroxide (Caustic)

A 10% solution of N a O H will dissolve up to 12.5% of its weight of g y p s u m scale. Be careful
caustic burns skin!

Salt Water
Water containing 55 000 m g / L of NaCl will dissolve three times as m u c h g y p s u m at 40C
[ 104F] as will fresh water at the same temperature.

Barium Sulfate
Dense barium sulfate is almost impossible to chemically remove. However, several proprietary
"scale dissolvers" are available which have been used to dissolve thin layers of barium sulfate scale.

These materials are chelating agents rather than true solvents. B a r i u m sulfate has a "solubility"
of approximately 120 g/L [0.5 lb/gal] of chemical in most of the c o m m o n l y used scale dissolvers. This
is rather low for a " s o l v e n t " by most standards. In addition, their reaction rates are o f t e n slow and
their cost is relatively high.

However, they are the only chemicals which have been used successfully for the removal of
d o w n h o l e barium sulfate. T h e i r disadvantages limit their application to the removal of thin layers of
scale in systems where extremely large amounts of revenue are at risk, such as the North Sea.

Iron Compounds
Hydrochloric acid is normally used to dissolve iron compounds. O n c e again, it m u s t contain a
corrosion inhibitor to prevent pipe corrosion. In addition, it should contain an iron-stabilizing agent to
prevent precipitation of iron c o m p o u n d s once the acid spends.

Iron dissolved during acidizing can exist as either ferric (Fe" + + ) or ferrous ( F e + + ) ions. As the
acid spends, the pH will rise, and ferric ions will begin to precipitate as Fe(OH)3 ( f e m e hydroxide) at a
(3 54>
pH of about 2.2. - Once the pH reaches a pH of 3.2, essentially all of the ferric iron will have pre-
cipitated. Iron stabilizing agents are employed to prevent precipitation of f e m e hydroxide.

There are two primary types of iron stabilizers:


Chelating or Sequestering Agents

These materials react with ferric ions to f o r m a water soluble complex, thus preventing
precipitation. C o m m o n l y used chelants are citric acid, acetic acid, mixtures of citric and
acetic acids, ethylenediaminetetraacetic acid (EDTA), and nitrilotriacetic acid (NTA).
Reducing Agents

Reducing agents convert f e m e ions to the m o r e soluble ferrous ions. Erythorbic acid has
been used for this purpose, and is capable of solubilizing up to nine times as m u c h ferric
iron as citric acid.' 5 " 5 5 '
In contrast, ferrous ions are quite soluble at pH values up to 7 and above. Since spent acid usu-
ally reaches a m a x i m u m pH of about 5.3, precipitation of ferrous iron is seldom a problem.


If iron sulfide is to be dissolved, care is suggested. The reaction between FeS and HCI is:

FeS + HCI -> FeCl, + H:S T

H i S is extremely poisonous and a f e w p p m in the air can kill you. A toxicity guide is given in
A p p e n d i x 18 (page 406). Fresh air masks should be on location and in use if there is any possibility
that personnel m a y breathe any H i S .

Salt (NaCI)
A fresh water wash is the best medicine for salt deposits.

Sand, Silt, Clay

These are usually present as occluded particles in scale deposits. Once the bulk scale material is
dissolved, these materials may be washed out.

Scale Removal from Surface Lines

The Use of Line Scrapers

Scale removal f r o m surface lines is usually accomplished by the combined use of chemicals and
line scrapers, or "pigs." The basic types of pigs available are shown in Figure 3.22 J 3 S6>

Foam Pigs
F o a m pigs are normally used for cleanout of lines which are not routinely pigged. They are
m a d e of an open-cell f o a m with a hard rubber or plastic wrapper. They are also made in brush form.

T h e s e pigs can be p u m p e d through a series of lines of different diameters, and the b o d y will de-
f o r m sufficiently (within limits) to permit the pig to continue to do an effective scraping j o b .

Figure 3.22 Types of Pigs



Foam pigs are available in various lengths and styles. However, the length is usually twice the
diameter. Foam pigs f o r m a seal against the inside of the pipe by compressing the foam and do not
seal f r o m pressure differential as do the other types of pigs.

Brush or scraper pigs have holes in the pig to allow for bypass. Fluid passing through the holes
in the pig will disperse the scraped solids in front of the pig. This will prevent a build-up of material
in front of the pig which could create a plug and cause the pig to become stuck in the line.

Foam pigs with brushes have a built-in bypass. T h e brushes are short and fixed to the pig body.
T h e blow-by occurs through the brush and helps to keep the brushes clean. If additional bypass area is
necessary, the r u b b e r or plastic covering on the nose of the pig can be cut out allowing passage of liq-
uid through the open-cell foam body. (Figure 3.23/-*

Figure 3.23 Cleaning Pigs

A typical flowline cleanout to r e m o v e oil coated calcium carbonate f r o m a line might consist of
the following steps:
1. Slug of solvent followed by a pig.
2. Slug of inhibited HC1 followed by a pig.
3. Either a neutralizing solution (high pH water) or a thorough water wash to remove all acid.
Acid inhibitors break down with time and all acid, spent or otheiwise, must be flushed
f r o m the line or severe pitting corrosion may result.

Gelled Plugs
Gelled fluids can be injected into a line ahead of a conventional line scraper to aid in the trans-
port ol large volumes of solids. For example, a gelled-water and scraper system was used to remove


approximately 350 tons of mill scale and other debris f r o m a 280 mile-long. 36-in. OD gas line in the
North Sea.'-5 5 6 1

Batches of gelled fluids (Kelzan XC p o l y m e r in water) and pipeline scrapers were used to re-
m o v e loose and loosely adherent rust, silt, welding rods, weld slag, and other debris f r o m the pipeline.

A wide variety of water-base gels, including hydrochloric acid, as well as hydrocarbon-base gels
can be formulated to satisfy the specific requirements of a cleaning operation.

W h e n a gelled fluid is in laminar f l o w in a pipeline, the central portion of the fluid Hows as a
plug with little exchange between the plug and the annular fluid. The velocity of the central plug can
be significantly higher than the mean flow velocity.

A scraper behind a gel forces the fluid in the vicinity of the pipe wall, which contains entrained
solids, to m o v e into the central plug-flow region w h e r e it is e a r n e d forward through the length of the
gel batch.

T h e debris that would remain in front of a mechanical pig in a conventional operation is picked
up, c a n i e d forward, and eventually circulated and evenly distributed over a relatively long distance.
This greatly reduces or eliminates the possibility of sticking the mechanical pig by debris build-up in

Downhole Cleanout
D o w n h o l e cleanout usually consists of r e m o v i n g scale from the tubing, the perforations, or for-
mation face (open-hole completions) and sometimes f r o m the pore spaces or fractures in the f o n n a t i o n

R e m o v i n g scale f r o m tubing is basically the same as removal f r o m surface lines, although pigs
cannot be used. Various mechanical approaches have been used to r e m o v e scale, including the use of
special boring tools in operations such as reaming, milling, drilling and jetting. Successful mechani-
cal scale removal has been achieved and in one operation in the Norwegian sector of the North
Sea'-1,-SK explosive charges were also used. Otherwise, chemical cleaning m a y be considered Obvi-
ously, if the tubing is badly scaled or plugged it m a y be necessary to pull the tubing and clean it on the

Removal of scale f r o m the formation face or perforations usually involves a simple soaking oper-
ation with one or m o r e solvents. However, if scale has formed in the perforation tunnels and pore
spaces back into the f o n n a t i o n for several inches, chemical soaking m a y not be successful. Here spe-
cial downhole tools such as casing brushes, surge/swab tools, suction washers, circulation washers or
retrievable bridge plugs and squeeze packers m a y be required in conjunction with a solvent to achieve
removal of the scale.

In extreme cases, hydraulic fracturing m a y be the only way to penetrate the scaled interval.

Downhole scale problems are more prevalent in p r o d u c i n g wells than in injection wells. Hence,
most of the complex scale cleanout procedures are aimed at producers. It is possible to prevent scaling
problems in injection wells. D o w n h o l e cleanout should not be a frequent p r o b l e m in injection systems
if they are being monitored properly. It is m u c h easier and cheaper to prevent or treat the problem on
the surface than to tiy to r e m o v e downhole deposits after they have formed.

If injection pressures begin to rise and injection well plugging or scaling is indicated, the follow-
ing is suggested:



1. Try to b a c k f l o w the well. If suspended particles have simply filtered out the formation
face, this alone may solve die problem. If you can backflow the well, be sure and try to
catch samples of any solids which come to the surface. If gas is available, a gas-lift valve
can be installed in the injection tubing to assist in the periodic backflow of an injection or
disposal well.
2. If backflowing is not successful, go to a chemical cleanout procedure.

Technically, it is advisable to swab the solvents back to the surface or backflow the well after
treatment. Solvents m a y become supersaturated and re-precipitate the material just dissolved. If
re-precipitation occurs in the wellbore area, serious plugging can result. Thus, cleaning solutions
should be brought back to the surface rather than displaced into the formation w h e n e v e r possible.

Soap Deposits (Napthenates and Carboxylates)

So far in this chapter we have referred to scales which have been related only to the water c h e m -
istry. Two other types of deposit w h i c h can occur in production systems are dependent upon the chem-
istries of both the water and oil phases. These are often referred to as "soap deposits", or more specifi-
cally as calcium naphthenates and sodium carboxylates.

Deposits of soap solids occur when metal ions in the produced water react with naphthenates (un-
saturated cyclic acids f r o m the oil) giving hard deposits, or carboxylates (unsaturated linear acids f r o m
the oil) to produce emulsion soaps. It has been shown that calcium is the predominant cation in
naphthenate deposits, with sodium being the main one in carboxylate emulsions/- 1 - 9> A m a j o r driving
force in the formation of these soap solids is the rise in pH as the fluids pass f r o m the reservoir
through the production system. As carbon dioxide is Lost f r o m solution and the pH rises, these organic
acids tend to dissociate and the carboxylate anion becomes available to react with available metal ions.
The deposits tend to occur once the pH rises a b o v e about 6 . 0 - 6 . 2 / - ' 0 9 '

R COOH > R COO~ + H+

2 R COO~ + Ca++( from the water) -> R COO(Napliiheiwte)

Where: R = unsaturated linear or cyclic hydrocarbon radical

The amount of deposits formed is related to the quantity and type of these acids naturally occur-
ring within the crude oil. Although current thinking suggests that it is not a simple relationship associ-
ated with the total acid n u m b e r (TAN), most problems appear to be with heavier crude oils (low API
gravity) with high TAN. Various physical parameters also affect the severity of deposition, including
pressure, temperature, shear forces, electrostatic fields and the amount of water present in the produced
fluids. Further details on the specific effects of these parameters can be seen in the reference by
Turner and S m i t h / - * w Chemical treatments are usually required to mitigate soap deposit problems, al-
though effects can be reduced by the control of physical design and operating parameters.

Naphthenate deposits are initially soft and sticky, which act as a binder for other solids in the
system (such as sand and corrosion products). (Note: the deposits harden within minutes upon expo-
sure to air, for example w h e n the system is opened for maintenance). M o r e detailed analyses lead
Baugh, et a l . / - J < w to propose that a high molecular weight acid is the m a j o r constituent of naphthenate
deposits. These deposits can lead to plugging problems both in the oil and water sides of the process
system and result in d o w n t i m e for deposit removal and the maintenance of efficient operations. In one
North Sea field where the oil contained about 2% naphthenic acids, deposits occurred in the main sep-
arator, the water treatment system, in the closed drain and flare knock-out drum, as well as in the sand
cleaning system.


The sodium carboxylate soaps appear as very stable viscous emulsions, which often incorporate
other solids such as paraffin waxes. The soap itself will also stabilize emulsions. These deposits tend
to form at the interface between the water and oil layers in separator vessels.

Adding acid to the system causes the "naphthenic" and "carboxylic" species to remain in their
acid f o r m s and thereby prevents formation of the soap solids. Acetic acid is widely used for this pur-
pose. However, the acetic acid can then significantly increase the corrosion r a t e / * 6 7 ' Surfactant based
naphthenic acid inhibitors (used at up to 2 0 0 p p m ) are commercially available and these minimise the
need for additional acid/- 3 59- 3 6 l ) T h e amount of acid used can be reduced by up to 80%, w h e n using
a surfactant naphthenate inhibitor/* 5 *''

3.1 Ostroff, A. G.: Introduction to Oilfield Water Technology', National Association of Coirosion Engineers,
Houston, Texas (1979).
3.2 NACE and API: Corrosion of Oil and Gas Well Equipment (1958).
3.3 Miller, J. P.: "A Portion of the System CaC0 3 - C 0 2 - H 2 0 with Geological Implications," American Jour-
nal of Science, Vol. 250 (March 1952) 161.
3.4 Oddo, J. E. and Tomson, M. B.: "Why Scale Forms and How to Predict It," SPE Production & Facilities.
(Feb. 1994) 47.
3.5 Cowan, J.C. and Weintritt, D.J.: Water-Formed Scale Deposits, Gulf Publishing Co., Houston, Texas
3.6 Carlberg, B. L. and Matches, R. R.: "Solubility of Calcium Sulfate in Brine," SPE Paper 4353 presented
at the Oilfield Chemistry Symposium of the Society of Petroleum Engineers of AIME. Denver. CO (May
24-25, 1973).
3.7 Vetter, O.J.G. and Phillips. R.C.: "Prediction of Deposition of Calcium Sulfate Scale Under Down-Hole
Conditions," J. Petr. Tech. (Oct. 1970) 1299.
3.8 Fulford, R. S.: "Effects of Brine Concentration and Pressure Drop on Gypsum Scaling in Oil Wells," SPE
Paper No. 1830, 42nd Annual Fall Meeting of the Society of Petroleum Engineers of AIME. Houston. TX
(Oct. 1-4, 1967).
3.9 Dickson, F. W.. Blount, C. W., and Tunell, G.: "Use of Hydro-Thermal Solution Equipment to Determine
the Solubility of Anhydrite in Waters from 100' C to 275C and from 1 bar to 1000 Bars Pressure," Am. J.
ofSci. (Vol. 261, 1963) 61.
3.10 Templeton, C. C.: "Solubility of Barium Sulfate in Sodium Chloride Solutions from 25C'C to 95C, J.
Chem. and Eng. Data, Vol. 5 (Oct. 1960) 514.
3.11 Blount, C. W.: "Barite Solubilities and Thermodynamic Quantities up to 3 0 0 X and 1400 Bars," Ameri-
can Mineralogist" (Vol. 62, 1977) 942.
3.12 Vetter, O.J.G.. Vandenbroek, I. and Nayberg, J.: "SrS0 4 : The Basic Solubility Data," SPE Paper 11803
presented at the International Symposium on Oil-field and Geothennal Chemistry, Denver, CO (June 1-3,
3.13 Jacques, D. F. and Bourland, B. 1.: "A Study of Solubility of Strontium Sulfate," J. Pet. Tech. (April
3.14 Fletcher, G. E., French, T. R and Collins, A. G.: "A Method for Calculating Strontium Sulfate Solubil-
ity," U.S. DOE/BETC/RI-8O/IO (April 1981).
3.15 Jacques, I). F., Gollberg, J. B. and Whiteside, W. C.: "Strontium Sulfate Solubility and the Effects of
Scale Inhibitors," paper presented at the NACE Middle East Corrosion Conference/79, Bahrain (April
15-17, 1979).
3.16 Davis, J. W. and Collins. A. G.: "Solubility of Barium and Strontium Sulfates in Strong Electrolyte Solu-
tions," Environmental Science and Technology, Vol. 5 (Oct. 1971) 1039.


S olubility is defined as the limiting amount of a solute which can be dissolved in a solvent under a
given set of physical conditions. The chemical species of interest to us are present in an aqueous
solution as ions. Certain combinations of these ions are c o m p o u n d s which have very little solubility in
water. The water has a limited capacity for maintaining these c o m p o u n d s in solution, and once this ca-
pacity, or solubility, is exceeded the c o m p o u n d s precipitate f r o m solution as solids. Therefore, precipi-
tation of solid materials which may f o r m scale can occur if both o f the following conditions are
1. T h e water contains ions which are capable of forming c o m p o u n d s of limited solubility.
2. There is a change in physical conditions or water composition which lowers the solubility
below the concentrations present.
Solid precipitates m a y either stay in suspension in the water, or they m a y f o r m a coherent scale
on a surface such as a pipe wall. Formation plugging may occur by filtration of suspended particles
from the water. Or, a solid scale may form on the formation face. Either is undesirable. T h e diffi-
culty of removal varies with the type of plugging w h i c h has occurred.

Scale formation frequently restricts flow through injection and flow lines, and tubing strings. It
causes pump w e a r or plugging and creates additional rod loads w h e n it f o r m s on sucker rods. Fire
tubes in all types of heaters fail prematurely when scale formation results in overheating. Corrosion is
often more severe under a scale deposit.

Water f o r m e d scales are responsible for many production problems and their effective control
should be one of the primary objectives of any efficient water handling operation.

Of the m a n y possible water formed scales, only a few are c o m m o n l y found in oilfield waters,
rhese scales are listed in Table 3.1 along with the primary variables which affect their solubility.

Calcium Carbonate
t alcium carbonate scale can be formed by the combination of calcium ion with either carbonate
or bicarbonate ions as follows:

Ca++ + CO3 - Ca C0 3 I (3.1)


Common Oilfield Scales
Name Chemical Formula Primary Variables
Mineral Scale Deposits
Calcium Carbonate (Calcite) CaC0 3 Partial pressure of C 0 2 , pH. temperature, total
pressure, total dissolved solids.
Calcium Sulfate Temperature, total dissolved solids, pressure.
Gypsum (Most Common) CaS0 4 2 H ; 0
Hemi-Hydrate CaS0 4 Vi H 2 0
Anhydrite CaS0 4
Barium Sulfate BaS0 4 Temperature, total dissolved solids, pressure.
Strontium Sulfate SrS0 4 Temperature, total dissolved solids, pressure.
Corrosion By-Products
Iron Compounds Dissolved gases, corrosion. Temperature, pressure, pH
Ferrous Carbonate FeCO., . .'uj Av> j j \ JW 1
1 - - - .
Ferrous Sulfide FeS
Ferrous Hydroxide Fe(OH) 2
F e m e Hydroxide Fe(OH) 3
Ferric Oxide Fe 2 0 3

Ca++ + 2 ( H C O ; ) - CaCO, I +CO : + H20 (3.2)

W h e n found in nature calcium carbonate might be called calcite. or vaterite, so you m a y see
these terms used in other texts.

Effect of C0 2 Partial Pressure

T h e presence of CO? increases the solubility of CaCO.i in water, since it is converted to the solu-
ble calcium bicarbonate. W h e n carbon dioxide dissolves in water, it f o r m s carbonic acid, which ion-
izes according to the following series of equations:

C02 + H20 <h> H2CO, (3.3)

H2C03 <-> H* + HCO; (3.4)

HCO;*H++CO; (3.5)
Only a small percentage of the bicarbonate ions dissociate at the pH values f o u n d in most injec-
tion waters to f o r m H + a n d C O j = , as shown in Figure 3.1. Bicarbonate ions vastly o u t n u m b e r the n u m -
ber of carbonate ions present under normal circumstances. Therefore, it is thought that Equation 3.2 is
the m o r e accurate expression for the precipitation of calcium carbonate because CO-, is lost f r o m solu-

As the concentration of CO2 in solution is increased, the reaction shifts to the left, resulting in
less C a C 0 3 precipitation. T h e water also becomes more acidic (the pH decreases) with the addition of
CO2 to the water.

T h e amount of CO2 that will dissolve in water is proportional to the partial pressure of C 0 2 in
the gas in contact with the water:



Figure 3.1 Ionization of Carbonic Acid as a Function of pH

Partial Pressure of CO-, - (Mole Fraction of CO, in Gas)x (Total Pressure) (3.6)

Mole Fraction of C02 = %C02 in Gas + 100 (3.7)

Hence, if a two-phase (gas + water) system is operating at 690 kPa [100 psia] and the associated
gas contains 10 m o l % CO^ (1 mole of C 0 2 per 10 moles of gas), the partial pressure of C 0 2 in the
gas is:
( 0 . 1 ) ( 1 0 0 ) = 10psia [69 kPa]
So, if either the system pressure or the percentage of C 0 2 in the gas were to increase, the amount
o f C 0 2 dissolved in the w a t e r also w o u l d increase.

Figure 3.2 illustrates the effect of C 0 2 partial pressure on the pH of water containing little or no
dissolved minerals. This data should not be applied to brines since the p r e s e n c e of dissolved minerals
changes the relationship between pH and the amount of dissolved C 0 2 .

The effect of C 0 2 pressure on the solubility of C a C 0 3 in pure water is shown in Figure 3.3.

This data illustrates that C a C 0 3 solubility increases with increased CO2 partial pressures. The ef-
fect becomes less pronounced as the temperature increases.

The reverse is also true. It is one of the m a j o r causes of CaCO^ scale deposition. At any point in
the system w h e r e a pressure drop is taken, the partial pressure of C 0 2 in the gas phase decreases, CO?
comes out of solution, and the pH of the water rises. This shifts Reaction 3.2 to the right and m a y
cause CaCO} precipitation.

Effect of pH
The amount of C 0 2 present in the water affects the water pH and the solubility of calcium car-
bonate. However, it really does not matter what causes the acidity or alkalinity of the water. The
higher the pH. the more likely that precipitation will occur.


Partial Pressure C 0 2 (kPa)

Figure 3.2 Effect of C 0 2 Partial Pressure on the pH of Water W

1 m P a = 1000 kPa = 145 psia


1250 'go

750 CD

500 o

1 1 1 1 1 1 1
Q , r - ' " ' ' o
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Solubility (g C a C C V I )

Figure 3.3 Effect of C 0 2 Pressure on Calcium Carbonate Solubility < 33)


Effect of Total Pressure

The solubility of calcium carbonate in a two phase system increases with increased pressure for
two reasons:

Increased pressure increases the partial pressure of CO2 and increases the solubility of
C a C O i in water as previously explained.
Increased pressure also increases the solubility due to thermodynamic considerations which
will not be discusscd here.
Pressure drops are one of the primary causes of calcium carbonate scale deposition in pro-
duction systems. In addition to decreasing the solubility of CaCC>3 due to the loss of CO2
and the thermodynamic pressure effect, pressure drops across chokes and valves induce
turbulence in the water which helps to overcome supersaturation effects and initiate
In single phase (all water) systems, such as a water injection system, increased pressure increases
the solubility of calcium carbonate solely due to thermodynamic considerations.

Effect of Temperature
Contrary to the behavior of most materials, calcium carbonate becomes less soluble as tempera-
ture increases the hotter the water gets, the more likely C a C 0 3 scale will form.

Hence, a water w h i c h is non-scaling at the surface may result in scale formation in an injection
well if the downhole temperature is sufficiently high. This is also the reason that CaCC>3 scale is often
found on the fire-tubes of heating equipment.

The solubility of C a C 0 3 in pure water at 1 atmosphere C 0 2 partial pressure as a function of tem-

perature is shown in Figure 3.4.

Methods for calculating the temperature at which CaCC>3 scale m a y be anticipated are given in a
later section.

Temperature, C

Figure 3.4 Calcium Carbonate Solubility' 3 ^


Effect of Dissolved Salts

Calcium carbonate solubility increases as the salt content of the water increases. For instance,
adding 200 000 m g / L NaCl to distilled water increases the CaCO? solubility f r o m 100 mg/L to 250

Actually, the higher the total dissolved solids (not counting calcium or carbonate ions), the
greater the solubility of CaCO? in the water and the lower the scaling tendency up to a m a x i m u m of
about 200 000 mg/L.

In Summary, the Likelihood of Forming Calcium Carbonate Scale

Increases as temperature increases
Increases as partial pressure of C 0 2 decreases
Increases as the pH increases
Increases as total dissolved salts decreases
Increases as the total pressure decreases

Calcium Sulfate
The precipitation of calcium sulfate f r o m water results f r o m the reaction:

Ca++ + SO; -> CaS041

Although calcium sulfate occurs "naturally" as a result of temperature and pressure changes
which o c c u r during production operations, the m o s t significant cause is probably due to the m i x i n g of
incompatible waters.

Forms of Calcium Sulfate

Most calcium sulfate deposits found in the oilfield are gypsum ( C a S 0 4 2 H 2 0 ) . According to
O d d o and Tomson, the most likely scale to f o r m f r o m brines will be gypsum at temperatures less than
80C [ I 7 6 F ] / - i v ;

Between 80C [176F] and 121C [250F], any of the three types of calcium sulfate may form,
with g y p s u m being m o r e likely at the low end of the temperature range and anhydrite ( C a S 0 4 ) more
likely at the high end. Hemi-hydrate ( C a S 0 4 Vi H : 0 ) is c o m m o n l y found in this temperature range in
non-turbulent systems with high ionic strengths.

Above 121C [250F], any calcium sulfate scale formed will almost certainly be a n h y d r i t e / - ^ '

Effect of Temperature
G y p s u m solubility in pure water increases with temperature up to about 38C [100F], then de-
creases with temperature as shown in Figure 3.5.

This is quite different f r o m the temperature-solubility behavior of C a C 0 3 . First, g y p s u m is con-

siderably more soluble than C a C 0 3 in the normal temperature range of interest. Second, the m a x i m u m
in the gypsum curve tells us that an increase in temperature could either increase or decrease the solu-
bility of g y p s u m depending on which part of the temperature curve w e ' r e concerned with. This is de-
cidedly different f r o m C a C 0 3 where an increase in temperature always decreases the solubility.


Temperature, UF
40 80 120 160 200 240 280 320

0 20 40 60 60 100 120 140 160

Temperature, C

Figure 3.5 Solubility of Calcium Sulfate in Pure Water' 3

N o t e that above about 38C [100 C 'F], anhydrite becomes less soluble than gypsum, so it could
reasonably be expected that anhydrite might be the preferred f o r m of C a S 0 4 in deeper, hotter wells.
Actually, the temperature at which the scale changes f r o m g y p s u m to anhydrite or hemi-hydrate is a
function of many factors, including pressure, dissolved salt content, flow conditions, and the speed at
which different f o r m s of C a S 0 4 can precipitate f r o m solution.

Predicting which f o r m of calcium sulfate will precipitate under a given set of conditions is very
difficult. Even though anhydrite would be expected above 38C [100F] in preference to gypsum due
to its lower solubility, gypsum m a y be found at temperatures up to 100C [212 FJ.

It is often difficult to precipitate anhydrite directly f r o m solution. However, with the passage of
time, g y p s u m can dehydrate to form anhydrite.

A b o v e 100C [212F], anhydrite will precipitate directly in a stirred or f l o w i n g system. If the sys-
tem is quiescent, hemi-hydrate solubility b e c o m e s l i m i t i n g / 3 - ^ Conversion to anhydrite could be ex-
pected with time.

Effect of Dissolved Salts

The presence of NaCl or dissolved salts other than calcium or sulfate ions increases the solubility
of gypsum or anhydrite just as it does for C a C 0 3 up to a salt concentration of about 150 000 mg/L.
Further increases in salt content decrease C a S 0 4 solubility. (Figure 3.6)

Effect of Pressure
Increased pressure increases the solubility of all forms of calcium sulfate due to thermodynamic
considerations which will not be discussed h e r e / 5 7 J W

Pressure drops are one of the primary causes of calcium sulfate scale deposition in production
systems. In addition to decreasing the solubility due to thermodynamic considerations, pressure drops


Figure 3.6 Solubility of Gypsum in NaCl Brines' 3 -^

across chokes and valves induce turbulence in the water which helps to overcome supersaturation ef-
fects and initiate precipitation.

The effect of pressure and temperature on anhydrite solubility is shown in Figure 3.7. Note that
the pressure effect decreases as temperature increases.

Pressure, psia
1450 7250 14,500

*n 0.15


Pressure, mPa

Figure 3.7 Effect of Pressure and Temperature on C a S 0 4 (Anhydrite) Solubility' 3 ^


Effect of pH
pH has little or no effect on the solubility of calcium sulfate.

In Summary, The Likelihood of Forming Calcium Sulfate Scale

Increases as temperature increases
Increases as T D S decreases
Increases w h e n incompatible waters are mixed

Barium Sulfate
Barium sulfate (barite) is the least soluble of the scales we have discussed thus far.

Ba++ + SO; BaSO, i

Table 3.2 compares the solubility of the four
c o m m o n scales, in distilled water at 25C [77F],
Comparative S o ubilities at 25C
T h e extreme insolubility of B a S 0 4 makes it Scale Solubility (mg/L)
very likely that scaling will occur if both Ba + + and Gypsum 2080
S 0 4 = ions are present in a water. For practical pur- Calcium Carbonate 53
poses barium sulfate is best regarded as insoluble. Barium Sulfate 2.3
Strontium Sulfate 129
Most barium sulfate scales also contain some
strontium sulfate.

Effect of Temperature
Barium sulfate solubility increases with temperature up to 100C [212F], T h e solubility in dis-
tilled water increases f r o m 2.3 mg/L at 25C [77F] to 3.9 mg/L at 95C [203F] as s h o w n in Figure
3 g p . 10) i n c r e a s e j s f a i r i y substantial percentage wise, but barium sulfate is still quite insoluble
even at this higher temperature.

A b o v e 100C [212F] the solubility decreases with temperature in waters with T D S values less
than about 50 000 mg/L. T h e solubility in higher salinity waters shows normal solubility behavior and
increases with temperature as s h o w n in Figure 3.9.<3 U)

Because of the decrease in the solubility of barium sulfate with decreasing temperature, there is a
greater risk of barium sulfate scale deposition in production wells due to the (usual) temperature drop
in the fluids as the well is produced. Very often barium sulfate is also a m a j o r risk in oilfields w h e r e
barium is present in the connate water (reservoir brine) and w h e n w a t e r containing sulfate ions is used
lor water injection purposes. The worst case is probably w h e n seawater is used, but m a n y water sup-
ply wells can contain sufficient sulfate to cause problems

Effect of Dissolved Salts

1 The solubility of barium sulfate in water is increased by foreign dissolved salts just as in the case
ot calcium carbonate and calcium sulfate. The addition of 100 000 mg/L of N a C l to distilled water in-
creases the solubility of B a S 0 4 f r o m 2.3 mg/L to 30 mg/L at 25C [77F], Maintaining 100 000 mg/L
I and increasing the temperature to 95C [203FJ.wiIl increase the B a S 0 4 solubility to about 65


Ionic Strength

Figure 3.8
Barium Sulfate Solubility in NaCl Solutions Calculated from Templeton's Data' 3 10)


220 000 ppm NaCl


BaS0 4
Solubility IQ
0 ppm NaCl

50 100 150 200 250
Temperature, C

Figure 3.9 E f f e c t ^ T e m p e r a t u r e and Salinity on Barium Sulfate Solubility at A t m o s " ^ 7


Figures 3.8 and 3.9 illustrates the effect of salinity on B a S 0 4 solubility over a w i d e range of tem-
peratures. T h e data in Figure 3.8 is plotted as a function of molar ionic strength rather than sodium
chloride concentration. An ionic strength of 1.0 is equivalent to a sodium chloride concentration of
approximately 60 000 mg/L. Ionic strength is defined later in this chapter and an example calculation
is given in Appendix 14 (page 400).

As a rule of thumb, you can assume that B a S 0 4 solubility will double as the temperature is raised
f r o m 25C [77F] to 95C [203F] regardless of the dissolved salt concentration. T h e effect of
dissolved salts is much more pronounced, as demonstrated by the 13-fold increase brought about by
the addition of 100 000 mg/L N a C l with no change in temperature.

Effect of Pressure
Increased pressure increases the solubility of barium sulfate due to thermodynamic considerations
which will not be discussed here.

Pressure drops are one of the primary causes of b a r i u m sulfate scale deposition in production sys-
tems. In addition to decreasing the solubility due to thermodynamic considerations, pressure drops
across chokes and valves induce turbulence in the water which helps to o v e r c o m e supersaturation ef-
fects and initiate precipitation.

Increased pressure increases the solubility of B a S 0 4 . Figure 3.10 shows the effect of pressure and
temperature on barium sulfate solubility in pure water.

Temperature, F
100 200 300 400 500

0 50 100 150 200 250 300

Temperature, C

Figure 3.10 Effect of Pressure and Temperature on Barium Sulfate Solubility in Pure Water'3-11>


Effect of pH
pH has little or no effect on the solubility of barium sulfate.

In Summary, the Likelihood of Forming Barium Sulfate Scale

1 ncreases as temperature decreases
Increases as T D S decreases
Increases as pressure decreases
Increases when incompatible waters are mixed

Strontium Sulfate
Strontium sulfate (celestite) is considerably more soluble than barium sulfate, with a solubility of
129 m g / L in pure water at a temperature of 25C [1TV].(312)

Sr + SO; SrSO, I

Effect of Temperature
Strontium sulfate solubility decreases with temperature.(3.12-3.14) jn p u r e w a t e r tjie solubility de-
creases to 68 m g / L at 125C [257F]/- u ->

Effect of Dissolved Solids

T h e solubility of SrSC>4 in water increases as the NaCJ content of the water increases up to a
m a x i m u m of approximately 1 75 000 mg/L. Further increases in salinity result in decreased solubil-
ity y.12-3.16) j n b r j n e s containing calcium or magnesium, the apparent solubility of S r S 0 4 is greater
than a NaC'l brine of equivalent ionic strength.



X of
/ S
E 700
/ oG

// /

< 400
f /


( 1 2 3 4 5
Ionic Strength

Figure 3.11 Strontium Sulfate Solubility in NaCI Solutions^3' 15>



Effect of Pressure
Strontium sulfate solubility in NaCl brines increases with pressure. However, solubility data
measured by Jacques and Bourland show that the increase is small up to 3000 p s i g / * 7 ^

Pressure drops are one of the primary causes of strontium sulfate scale deposition in production
systems. In addition to decreasing the solubility due to thermodynamic considerations, pressure drops
across chokes and valves induce turbulence in the water which helps to overcome supersaturation ef-
fects and initiate precipitation.

Until the advent of seawater injection in the Middle East, pure S r S 0 4 scale was seldom observed
and w a s not considered a m a j o r problem in water injection operations. However, serious S r S 0 4 scale
problems have occurred in producing wells in a n u m b e r of Middle East fields after breakthrough of
seawater, clue to mixing of the sulfate-bearing seawater and the strontium in the formation waters in
the producing w e l l b o r e s / - 5 , ? - 3 J 9 >

In the majority of cases, however, strontium co-precipitates with barium to f o r m (Ba,Sr) SO4
scale. A study of several barium sulfate scales by Weintritt and Cowan ( 3 , ~ w showed strontium sulfate
contents of 1.2 to 15.9 percent. Barium sulfate concentrations varied f r o m 63.7 to 97.5 percent.

Effect of pH
pH has little or no effect on the solubility of strontium sulfate.

In Summary, the Likelihood of Forming Strontium Sulfate Scale

Increases as temperature increases
Increases as T D S decreases
To a small extent, increases as pressure decreases
Increases w h e n incompatible waters are mixed

Iron Compounds
Iron ions present in a water m a y be either naturally present in the water or the result of corrosion.
Fonnation waters normally contain only a few mg/L of natural iron and values as high as 100 m g / L are
rare. Higher iron contents are invariably the result of corrosion. Precipitated iron c o m p o u n d s are a
common cause of deposit formation and injection well plugging, as well as being indicative of the
equally serious problem of corrosion.

Corrosion is usually the result of CO?. H2S or oxygen dissolved in the water. Most of the scales
containing iron are corrosion products. However, iron c o m p o u n d s can also f o r m by precipitation of
natural formation iron even if corrosion is relatively mild.

Carbon dioxideforms carbonic acid in water which can react with iron to form iron carbonate
scale. Whether or not scale actually forms will depend on the pH of the system. Scale formation is
much more likely above pH 7. Figure 3.12 illustrates the solubility of iron carbonate in fresh water.

Hydrogen sulfide will form iron sulfide as a corrosion product which is quite insoluble and usu-
ally iorms a thin, adherent scale. Suspended iron sulfide is the cause of "black water."

The iron sulfide diagram in Figure 3.13 illustrates the concentration of Fe + + ferrous ion) which
1 stay in solution at various pH values and H--S concentrations in fresh water. Fe + + + (ferric ion) is
seldom found at pH values above 3.0.



Figure 3.12 Iron Carbonate Stability Diagram

Figure 3.13 Iron Sulfide Stability Diagram

Oxygen combines to f o r m several compounds. Ferrous hydroxide, Fe(OH)?, ferric hydroxide,

F e ( O H ) j , and ferric oxide, Fe^O?, are c o m m o n scales resulting f r o m contact with air. For example,
when air-free water containing dissolved ferrous iron ions and bicarbonate ions is contacted by air, fer-
ric hydroxide can be f o r m e d .

2Fe++ + AHCOl + Hi + 2Fe(OH)i i+4C02

Ferrous iron (Fe + + ) is oxidized by the presence of air to give Fe + + + , and ferric hydroxide results.
This is practically insoluble a b o v e pH 4. As shown in Figure 3.14, if oxygen can be excluded, 100
p p m o f F e ( O H ) , (ferrous hydroxide) will still be in solution at pH 8.5.



Figure 3.14 Iron Hydroxide Stability Diagram

" R e d water" is the result of suspended particles of Fe^O.;, another product of oxygen and iron.

Iron c o m p o u n d s can also result f r o m the action of certain bacteria (gallionella ferruginea) which
live in water in the presence of air. T h e s e bacteria take Fe"1"1" ions f r o m the water and deposit ferric

In summary, the chemistry of iron c o m p o u n d s is much more c o m p l e x than that of previously dis-
cussed compounds. This is due primarily to the fact that iron c o m m o n l y exists in two oxidation states
in water, Fe + + (ferrous) and Fe + + + (ferric). These two ions f o r m c o m p o u n d s with the s a m e anions that
possess very different solubilities. It is difficult to quantitatively predict the behavior of iron com-
pounds. It is far more important to prevent their formation a topic which will be covered in a later

Silica Deposits
Silica deposits are not c o m m o n in produced waters or in injection projects. However, they can
be quite serious in boilers and steam generators, and are sometimes a p r o b l e m in cooling waters.

Silica is thought to occur in both the colloidal state as S K X or it can combine with magnesium,
sodium or aluminum ions to form silicate scales. It can also vaporize in steam boilers and then precip-
itate on turbine blades. Because of the complexity of this behavior, it is difficult to predict the condi-
tions under which silica can be kept in solution.


Naturally Occurring Radioactive Materials ( N O R M ) are found in m a n y production and injection
systems. Of interest are scales which are radioactive due to the presence of radium (Ra-226 and
Ra-228, with half-lives of 1608 years and about 6 years respectively) ions which have co-precipitated
f r o m produced water along with scale-forming cations during scale formation. T h e radium ions are
trapped within the crystal lattice of the material. R a d i u m emits alpha ( a ) radiation (short range, but
the most h a r m f u l ) and although it will not penetrate the skin, it still poses a serious risk to health. Ra-
dium also emits beta and gamma (greater penetration, but considered less h a r m f u l ) radiation. R a d i u m
can enter the body by ingestion (for example, as airborne dust or f r o m contamination on the hands),
where it collects and remains in the bones. An easily understandable and concise p a m p h l e t ' 3 - 7 ' con-
cerning the hazards of radioactivity, its measurement and precautions, is issued by the National Radio-
logical Protection Board ( N R P B ) in the UK.

N O R M is also k n o w n as Low Specific Activity ( L S A ) scale, the "activity" being measured in

"becquerels per gram". T h e actual level of radioactivity varies f r o m one deposit to another, depending
upon the amount of radium that has been incorporated. Although this activity is greater than the back-
ground radiation of ordinary rocks, it is still low when compared to m a n - m a d e sources of radioactivity.

To understand w h y the radioactive species concentrate within scale deposits, we n e e d to look at a

group f r o m the Periodic Table of the Elements. In atomic size order, G r o u p 2 contains:

1 Be Beryllium
1 Mg Magnesium
Ca Calcium
Sr Strontium
Ba Barium
| Ra Radium

Hence, being in the same chemical group as calcium, strontium and barium, radium will co-pre-
cipitate w h e n sulfate scales are formed. Likewise, if we prevent the calcium, barium and strontium
scales f r o m forming, we will also prevent radioactive scale.

N O R M has been c o m m o n l y associated with barium sulfate scales. However, in practice, any
scale or corrosion product may contain co-precipitated radium ions and be radioactive. In m a n y coun-
tries it is necessary to notify the authorities when N O R M is detected. For everyone's protection there
are special requirements for the handling, transport, processing, storage and eventual disposal of any
samples of N O R M . Hence, production of this material can be costly. In the U S A , for example, the
producer of the N O R M apparently retains long term liability for the disposed material, despite having
to pay for that disposal.

Factors Affecting Scale Deposition

Scales deposit f r o m production fluids because the equilibrium which has existed in the reservoir
for millions of years, has been disturbed. The factors which affect scale deposition can be grouped as




Gauges \

. . .. a a ^ ^ ^ H

" M i - -

Figure 3.15 Pressure Drop Across a Flow Meter

In the previous discussion of the four main scales, we have already covered the physical factors
which include temperature, pressure, pH, ionic strength (total dissolved solids) and the mixing of in-
compatible waters.

System factors include the shape and geometry of the production equipment and the affect that
these have on the produced fluids. An example is s h o w n in Figure 3.15, where a pressure drop is in-
duced by a venturi meter.

Other system factors would include designs where fluids f r o m wells producing different water
chemistries are mixed at pipe connections and manifolds.

The rate factors include:

Degree of supersatitration of the water
Rate of nucleation
Rate of scale crystal growth
Adherence of scale to the metal surfaces.

Since these factors are governed by the laws of chemistry and physics, to some extent they are
outside of your control. However, scale inhibitors can m o d i f y the rate factors and delay onset of scal-
ing. This is usually for a long enough period of time to prevent scale depositing within the system.
Also, if the scale forms but does not adhere to surfaces, it may not cause any m a j o r problems.


The Value of Solubility Calculations

Solubility calculations or scaling indexes m a y be used to predict the f o r m a t i o n of certain types of
scales. The values obtained f r o m these calculation procedures should be taken merely as guidelines.
I hey indicate the degree of "scaling tendency," or the likelihood of scale formation. Simplifying as-
sumptions have been made in the derivation of each equation, since solubility in naturally occurring
waters is a complex phenomenon.

It should be emphasized that if scale f o r m a t i o n is indicated by calculation, it serves as an alarm.

/ y o u are looking at a possible water source, you should avoid those which show scaling tendencies or
make provision for treatment in your planning. Similarly, one should avoid mixing waters which
Would result in a composite analysis which exhibits scaling tendencies under system conditions.


A calculated scaling tendency in an existing system should focus attention on the fact that scale
formation is likely and you should begin inspecting the system for signs of scale formation immedi-
ately. Solubility calculations are an extremely valuable tool, but like nearly everything else, their ab-
solute value will be greatly influenced by the u s e r ' s experience and j u d g m e n t .

N o w a d a y s there are numerous commercial and in-house computer based software p r o g r a m s

which remove the tedious chore of the scaling tendency calculations. This means that a lot of data can
be processed quickly and accurately. However, the validity of the results generated still depends u p o n
the assumptions m a d e when building the computer model. The limitations on computer models in
terms of running t h e m within the appropriate temperature, pressure, total dissolved solids and pH "en-
velopes," must be understood and appreciated.

M o s t of these computer models predict the amount of scale w h i c h is likely to precipitate, as well
as the actual scaling tendency. It is imperative that the user investigates, evaluates and c o n f i r m s the
suitability and reliability of whichever software is selected. T h e solubility calculations still remain an
essential tool to be used within the scale control management strategy.

It is important to keep in mind that the value of the output f r o m any computer model is still
heavily dependent on the reliability of data used as input. Therefore, the water sample must be repre-
sentative of that part of the system being investigated. This is often difficult and sometimes impossi-
ble to achieve. Also the water analysis must be accurate and reliable which means a n u m b e r of param-
eters must be measured on site. This aspect is discussed in Chapter 2, Water Sampling and Analysis.

The f o l l o w i n g section is a s u m m a r y of s o m e of the published scale prediction equations which

have been used for various waters. No attempt has been made to include all of the published solubility
information. T h e s e basic equations are shown to g i v e an understanding of the underlying principles
used within the computer models.

The Basis for Solubility Calculations

All scale prediction methods are based on laboratory measurements of the solubility of a specific
compound at equilibrium conditions. Normally, solubilities are measured in synthesized waters over a
range of temperatures at atmospheric pressure.

Solubility Product Principle

When a sparingly soluble salt is added to water, cations and anions f r o m the crystal lattice of the
solid pass into solution until the solution becomes saturated. In the saturated solution an equilibrium
exists between the ions in solution and ions present in the solid crystal lattice.

Using barium sulfate ( B a S 0 4 ) as an example:

BaSO,4 <r*Ba++ + SO=J


Solid Solution

At a given temperature and pressure the product of the activities of the ions in the saturated solu-
tion is constant and is called the thermodynamic solubility product constant, K s p .

Where: a Ba = barium ion activity
a = sulfate ion activity



The activity of an ion is defined as the product of the ion concentration and the activity coeffi-

So; (3.10)
Where: C,Bo = Ba concentration, moles/L
Cso, = S 0 4 " concentration, moles/L

The activity coefficient is a function of temperature, pressure and ionic strength.

It is often convenient to include the activity coefficients in the solubility constant term so that
the concentrations of the ions can be used in the equations rather than the activities. The resulting sol-
ubility product is then defined as the ion product constant or conditional solubility product constant,
Kc, and is defined in Equation 3.11.

a; = = ( < V )(c s 0 : ) saturation (3.11)

' Ba + * V .SO;

Where: Kc = conditional solubility product constant, molar units

Based on this principle, it is possible to evaluate a solution with respect to the possibility of the
precipitation of a given salt at a given temperature and pressure by comparing the value of the ion
product constant at those conditions with the product of the measured concentrations of the ions in the

If we have a solution which contains a given amount of dissolved B a S 0 4 , and the measured con-
centrations o f Ba~ + and SC>4= are C ^ and C $ ( ) . respectively, the possibilities are a s follows:

1. T h e solution is saturated with B a S 0 4 .

(CBl^)(Cso:) = K< (3-12)

2. The solution is undersaturated with B a S 0 4 . Precipitation cannot occur.

3. The solution is supersaturated with B a S 0 4 . Precipitation can occur.

(CstrXCsl)>K< {3A4)
Thus, precipitation can occur only in the last ease. However, it may not occur in practice diie to
the tact that solutions often remain supersaturated until sufficient energy is available to initiate nucle-
ation of the crystalline solid f r o m solution.


Saturation Ratio
Continuing with the example of BaSC>4 dissolved in water, the saturation ratio (sometimes called
the supersaturation ratio) is defined as the ratio of the ion product to the ion product constant:
TP C .. x C
= = (3.15)
Kc Kc

Where: IP = ion product

SR = saturation ratio
CB-a and C . , - = measured concentration of Ba ++ and SO a' in solution

Thus, it is also possible to express the conditions necessary for precipitation in terms of the satu-
ration ratio:
1. SR = 1 T h e solution is saturated with B a S 0 4 .
2. SR < 1 T h e solution is undersaturated with B a S 0 4 .
Precipitation cannot occur.
3. SR > 1 The solution is supersaturated with B a S 0 4 .
Precipitation can occur.

Calcium Carbonate Scaling Calculations

Calcium Carbonate Saturation Indexes
It is c o m m o n to express the degree of supersaturation. and hence the likelihood of precipitation
of C a C 0 3 f r o m a solution in terms of the saturation index which is defined as follows:

Saturation Index = log 10 (3.16)
Although there is agreement as to the definition of the saturation index, different investigators
have used different nomenclature to describe the term.

Calcium Carbonate Saturation Indexes
Investigator Index Name Abbreviation
Langelier' -" Saturation Index SI
Ryznar'-'--' Stability Index SI
Stiff & Davis'- ' Stability Index SI
i V
Oddo & Tomson' - ' Saturation Index Is

T h e saturation index, which we will refer to as SI, is a measure of the degree of supersaturation,
and thus the driving force available to cause precipitation. The larger the value of SI, the greater the
likelihood that scale will occur. It does not predict the amount of scale which will precipitate.

Figure 3.16 illustrates the relationship between the supersaturation ratio and the saturation index.





2 10

1 0.1


3 - 2 - 1 0 1 2 3

Figure 3.16 Saturation Index (SI) vs Saturation Ratio (SR)

Langelier Saturation Index

This well k n o w n index was developed to predict whether a fresh water saturated with dissolved
oxygen would f o r m calcium carbonate scale or be corrosive.

T h e Langelier saturation index' 5 is calculated from the following empirical equation:

SI = pH - pHs (3.17)
Saturation Actual pH pH at w h i c h
lnde\ o f Water water would be
staturated with

pHs = pCa + pAlkM + C, (3.18)

pCa = l o g (3.19)
Moles Ca /Liter

pAlk M = log (3.20)

Equivalents M Alkalinity / Liter

M Alkalinity = Total Alkalinity = CO; + HCO] , equivalents / L (3.21)

Where: C, = a constant which is a function of the total dissolved solids and temperature

As previously explained, if SI > 0, precipitation of CaCO^ is indicated. A negative value of SI

indicates that the water is corrosive if dissolved oxygen is present. This index indicates the tendency
oi a water to precipitate calcium carbonate, but it does not indicate the amount of precipitate.

Values of the empirical constants used to solve these equations are given elsewhere for molal
ionic strengths of 0-0.02 [0-800 ppm T D S ] and temperatures of 0-90C [ 3 2 - l 9 4 F ] at atmospheric
pressure , a 2 2 )

The nomograph developed by C a p l a n ^ - 1 ^ is reproduced in Appendix 11 (page 397) and permits

determination of the Langelier Index.


Ryznar Stability Index

Ryznar developed an empirical equation for calculating the "stability index"' 0 '-"" of fresh water at
atmospheric pressure. This index not only indicates the tendency of a water to precipitate calcium car-
bonate (or be con-osive if it is saturated with oxygen), it also gives a semi-quantitative estimate of the
amount of scale which will f o r m or the seriousness of the corrosion.

SI = 2pHs - pH (3.22)
Ryznar's stability index values always are positive. They can be interpreted as follows:
Stability Index < 6.5 indicates CaCO.? scale formation. The smaller the index, the larger
the amount of scale indicated.
Stability Index > 6.5 indicates corrosion if dissolved oxygen is present. T h e larger the in-
dex, the more severe the anticipated corrosion.
Caplan's n o m o g r a p h in Appendix 11 can be used to d e t e n n i n e the Ryznar Index.

Stiff and Davis Method

Stiff and D a v i s ^ - ^ empirically extended the Langelier method to apply to oilfield brines. Their
equation is as follows:

SI = pH - pHs (3.17)
Stability Actual p H pH al which
Index of Water water would be
s a t u r a t e d with

pHs = K + pCci + pAlk (3.24)


SI = pH - K - pCa- pAlk <3.25)

Where: SI = Stability Index. If SI is negative, the water is undersaturated with CaC0 3 and
scale formation is unlikely. If SI is positive, scale is likely to form,
pH = actual pH of the water
K = A constant which is a function of salinity, composition and water temperature.
Values of K are obtained from a graphical correlation with ionic strength and the
temperature of the water.

pCa - log (3.26)
Moles Ca++ (Liter

pAlk M = log- (3.27)

Equivalents M AlkalinityjLiter

M A Ikalinity = Total A Ikalinity = CO; + HCO; , mg / L (3.28)

The ionic strength is:

p =UciZf+C2Z;+C}Z;+...+CllZ;) (3.29)

Where: C = concentration of the ion in moles/liter.

Z = valence of the ion



In order to calculate SI we must k n o w the temperature, pH, and the H C O 3 and CO3 concentra-
tions. In addition, a complete w a t e r analysis is necessary to enable calculation of the ionic strength.

It is essential that values of pH, H C O ^ ' a n c l CC>3 = be measured in the field immediately after sam-
pling, since these parameters change very quickly once the sample is removed f r o m a pressurized sys-
tem. Valid calculations cannot be made f r o m laboratory analyses.

Unfortunately, even field measurements of pH values will not suffice when attempting to apply
this method to downhole conditions in producing or injection wells. The pH must be calculated.
Equations are given in the following section which enable estimation of pH values at elevated
pressures and temperatures.

Values of K as a function of ionic strength are given in A p p e n d i x 12 (page 398). These curves
are based on experimental data measured over the f o l l o w i n g range:
Molar Ionic Strength: 0-3.6
Temperature: 0, 30 and 50C [32, 86 and 122F]
Pressure: 101.3 kPa [1 atmosphere]
All curves outside of this data range were extrapolated.

A chart for the determination of pCa and pAlk is given in Appendix 13 (page 399).

T h e results of the calculation may be summarized in Table 3.4.

Interpretation of Stiff & Davis Index
Case Conclusion
SI is negative The water is undersaturated with CaC0 3 .
Scale formation is unlikely.
SI is positive The water is supersaturated with CaC0 3 .
Scale formation is indicated.
SI = 0 The water is saturated with CaCOj.

An example calculation is given in Appendix 14.

Oddo and Tomson Method

The equations developed by O d d o and T o m s o n ^ * - 1 5 , S 2 7 ) enable the calculation of the saturation

index, l s , and considers the effect of total pressure as well as varying CO? partial pressures. Its mean-
ing is analogous to that of the Stiff and Davis Index. They also developed equations w h i c h permit the
calculation of pH.

Their initial work was published in 1982. T h e method was subsequently modified to include the
elfects of additional parameters. The equations given here were published in 1994.


2. Calculate y g , the mole or v o l u m e fraction of CO2 in the gas phase at the specified T and P.
Given that y t is the m o l e fraction CO2 in the gas at the surface.

y* Pfg (5.0 bwpd +10.0 bopd) x 10


1.0 +
~~ MMscf{T + 460)

Where: bopd = barrels of oil per day

bwpd = barrels of water per day
MMscf = million standard cubic feet of gas per day

3. Calculate the molar ionic strength.

]i(Moles/L) = 10~5 (2.2 x Na* +5.0 xCa++ + S.2xMg++ +\.5x Ba++ (3.38)
+2.3 x Sr + 1 . 4 x CV" + 2 . 1 x 5 0 ; + 0 . 8 xHCO;)
Where all ion concentrations are in mg/L and are obtained f r o m a water analysis.

4. Calculate I s or the pH as desired.

Is = log + 5.85 + 1 5 . 1 9 x l 0 " 3 r - 1 . 6 4 x l 0 " 6 7 ' :


-5.27 x 10"5 P - 3 . 3 3 4 ^ / J I + 1 . 4 3

pH = log
+ 8.60 + 5.31x 10"3r - 2 . 2 5 3 x l 0 _ 6 7 ' 2
- 2 . 2 3 7 x 10~5P -0.99-^/jT + 0.658|i

This method is said to be valid over the following data range:

Molar Ionic Strength: 0-4.0
Temperature: 32-3 92F [0-200C]
Pressure: 0 - 2 0 000 psig

Effect of Carboxylic Acids on CaC03 Saturation Index Calculations

Generally, alkalinity titrated to a pH of approximately 4.5 is a s s u m e d to be equal to the bicarbon-

ate ion concentration. In produced waters, a substantial concentration of carboxylic acids (often called
volatile fatty acids or "VFA's") is often present. As explained in Chapter 2, the acids are ionized in
water and the species w h i c h are actually present are the VFA anions, w h i c h are typically dominated by
the acetate ion. The VFA anions contribute to alkalinity.

The total alkalinity is as follows:

Total Alkalinity =HCOJ + CH}COO~ + CH}CHzCOCr + CH}CH2CH2COO~

Acetate Ion Propionate Ion Butyrate Ion
In many cases, the acetate ion concentration is far larger than the concentration of the other VFA
ons. so their presence can be ignored, and the alkalinity can be assumed to simply be the sum of the
bicarbonate and acetate ion concentrations.


Taking the negative root of the resulting quadratic equation:

m + a--y](m -a)2 + 4Kc

Let: G = m + a, and
X = m - a


P m o - ^ * K , m g U s j L

Given the following relationship:

mg/L = moles/L x MW x 1000


P = 500 x MW x [g - yjx2 + 4Kc 1 mg/L (3.42)

Where: P = maximum amount of scale which can precipitate, mg/L

MW = molecular weight of scale compound

This equation can be used to calculate the m a x i m u m amount of any scale which might precipi-

Stiff & Davis

In order to solve this equation for the m a x i m u m amount of C a C O 3 which could precipitate using
data from the Stiff and Davis SI calculation, the following substitution is required for the value of IQ
as pointed out by Valone and S k i l l e r n : ^ - ^

Kc =l0k~pH (3.43)

Where: k = Stiff & Davis "k"

pH = actual system pH

Thus, for CaCC>3, Equation 3.42 becomes:

P =50000^G - ^X2 +4xlOk-p" ] (3.44)

In certain cases, it m a y be desirable to express the results in lb/1000 bbl.

PTB = =0 3 5 ^ (3.45)
1000 bbl L
And Equation 3.24 becomes:

PTB = 1 7 5 0 0 [ G - V > Y +4X10*-"" ] (3.46)

Where: G = Ca ++ + HC0 3 ". moles/L

X = C a ~ - H C O r , moles/L


Valone and Skillern' 1 : ' s> further state that based on Texaco's experience. Table 3.5 can be used to
assess the anticipated severity of calcium carbonate scale deposition.

Calcium Carbonate Scaling Severity
PTB Value Description
PTB < 0 No scale
0 < PTB < 100 Few scaling problems
100 < PTB < 250 Moderate scaling difficulties
PTB > 250 Severe scaling occurs

Oddo & Tomson

In the case of O d d o and Tomson, Equation 3.47 can be utilized to calculate Kc.
K = (3.47)


pKc = pH-2.76 + 9 . 8 8 x 1 0 " 3 r + 0 . 6 1 x 1 0 _ 6 r 2 = 3 . 0 3 x 1 0 ~ 5 / > - 2 . 3 4 8 ^ / j J + 0.77p (3.48)

T h e pH value in Equation 3.48 can be measured or calculated. Substituting into Equation 3.42
w e obtain:

P= 2
50 0 0 0 G-JX + (3.49)

Or when expressed in PTB:

P = 17 5 0 0 G-JX- + (3.50)

Sulfate Scaling Calculations

Traditional Approach to Sulfate Scale Solubilities

Solubility values for C a S 0 4 , B a S 0 4 or S r S 0 4 , can be calculated using the following equation,
providing values of conditional solubility product, K c , are k n o w n for each compound:

Solubility (meq/L) = 1000 ^ X + 4 AT, - x ] (3.51)

The derivation of this equation follows.

The Common Ion Effect

T h e m a x i m u m solubility of a slightly soluble salt is obtained when the concentrations of the cat-
ion and anion are equal.



For example: Given a saturated solution of CaSC>4 in water, w h e r e the C a + + and SO4 concentra-
tions are equal:

Ca++ concentration = S moles) liter

SO4 concentration = S moles)liter



T h e solubility of calcium sulfate is equal to S moles/liter, which is simply the square root of the
molar ion product constant.

However, in most natural waters, the cation and anion concentrations are not equal. In this case,
the difference between the two concentrations is callcd the excess common ioit concentration. T h e
presence of excess c o m m o n ion reduces the solubility of the salt unless a complex ion or ion pair,
such as neutral magnesium sulfate, f o r m s and offsets the effect/-* 6 )

For example, consider a saturated solution of CaSC>4 in water, where the C a + + and SC>4= concen-
trations are unequal:

Ca++ concentration = 5' moles)liter

SOI concentration = S + X moles) liter

In this case, we have arbitrarily selected the sulfate ion concentration as the larger of the two.
The amount by w h i c h the sulfate ion concentration exceeds the calcium concentration, X, is the excess
common ion concentration.

Kc = (S)(S x X) = S2 + SX
The calcium sulfate solubility is equal to S moles/liter, since that is the m a x i m u m amount of cal-
cium sulfate which can be f o r m e d by combining S moles of C a + " with (S + X ) m o l e s of S 0 4 = .


S2 + SX-Kc =0
Taking the positive root of the quadratic equation:

o _ + ^jx + 4Kc

For a divalent ion such as Ca"1"1", Ba + + or Sr + T :

S (equivalents)L) = + 4 K c -X (3.53)


5 (meq/L) = 1000 ^ J X 2
+ 4Kc - (3.54)

The "actual concentration" of C a S 0 4 in solution is equal to the smaller of the C a + + or S 0 4 = con-

centrations (expressed in meq/liter) in the water of interest, since the smaller concentration controls the
unt of calcium sulfate which can be f o r m e d .

ER 3 85

T h e calculated calcium sulfate solubility, S (meq/liter), is compared with the actual concentration
to determine if scale formation is likely.

Interpretation of Sulfate Scale Calculations
Case Conclusion
S = Actual The water is saturated with CaS0 4 .
S > Actual The water is undersaturated with CaS0 4 .
Scale is unlikely.
S < Actual The water is supersaturated with CaS0 4 .
Scale is likely.

This f o r m u l a can be used to calculate the solubility of any divalent salt such as C a S 0 4 , B a S 0 4 ,
or S r S 0 4 .

Calcium Sulfate (Gypsum) Solubility Calculations

T h e data measured by Skillman, M c D o n a l d and S t i f f 5 - 9 ' has been widely used to estimate the
solubility of g y p s u m in oilfield brines. T h e y measured ion product constants in simulated oilfield
brines over the following range:

Temperature: 10, 35, 50 and 80C [50, 95, 122 and 176F]
Ionic Strength: 0-6.0 moles/L
Pressure: 101.3 kPa [1 atmosphere]
Their data are presented in Appendix 15 (page 401).

T h e f o l l o w i n g procedure is r e c o m m e n d e d to assess the possibility of g y p s u m precipitation f r o m a

given brine:

1. Calculate the molar ionic strength using Equation 3.38.

2. Obtain the appropriate value of K c for the temperature of interest f r o m A p p e n d i x 15.
3. Determine the excess c o m m o n ion concentration, X, in moles/liter. This is simply the dif-
ference between the calcium concentration and the sulfate concentration.
4. Calculate the solubility of g y p s u m in meq/liter by solving Equation 3.54.
5. Calculate the "actual concentration" of g y p s u m in the water, which is equal to the smaller
of the C a f ~ or S 0 4 = concentrations expressed in meq/liter.
6. C o m p a r e the calculated solubility with the actual concentration to determine if precipita-
tion of g y p s u m is likely.
An example calculation is given in Appendix 14.

Other c o m m o n l y used methods for predicting calcium sulfate solubility which will not be pre-
sented here include those of Metier and O s t r o f f 3 ^ , and Carlberg and M a t t h e w s / 5 The latter
method can be used for temperatures up to 125C [257F]. It assumes hemihydrate solubility to be the
limiting solubility a b o v e 100C [212F],

Barium Sulfate Solubility Calculations

It is possible to estimate the solubility of barium sulfate for waters which contain predominately
sodium and chloride ions and very little magnesium or calcium ions using the solubility data measured



by T e m p l e t o n ' J , / w for sodium chloride solutions. M o l a r conditional solubility product constants calcu-
lated f r o m his data are presented in Appendix 16 (page 403) over the following range:

Temperature: 25, 35, 50, 65, 80 and 95C [77, 95, 122, 149, 176 and 203F]

Ionic Strength: 0-4.15 moles/L

Pressure: 101.3 kPa [I atmosphere]

The solubility can be calculated using Equation 3.54, and the probability of BaSO.} precipitation
evaluated using the same procedure as previously outlined for gypsum.

Because BaSC>4 has such limited solubility, the appearance of Ba + + and SC>4= ions in any water in-
dicates a danger of scale formation.

Strontium Sulfate Solubility Calculations

The solubility of strontium sulfate in sodium chloride solutions can be calculated in the s a m e
maimer as the solubilities of the other sulfate scales.

Values o f K c have been measured b y several i n v e s t i g a t o r s . H o w e v e r , the values m e a -

sured by Jacques and B o u r l a n d ( i I S ) cover the widest range of conditions thus published, and include
the effect of pressure. The following equation is based on their data and can be used to estimate val-
ues of K c :


Where: Kt = conditional solubility product constant, molar units

X = 1/T

R = A + BX + C$l + Dji + EP2 + FXP + G^LP (3.56)

The units are: T = K = C + 273

P = total pressure, psig
p = ionic strength, moles/L

T h e coefficients of the equation are:

A = 0.266 498 x 10" J E = - 1 . 3 8 3 x 10- 1 2

B = - 2 4 4 . 8 2 8 x 10~3 F = 1.103 323 x 10- 6
C = - 0 . 1 9 1 065 x 10" 3 G = -0.509 x icr9
D = 53.543 x 10- 6

This equation applies over the following range:

Temperature: 38-L49C [100-300F]

Ionic Strength: 0-3.45 moles/L
Pressure: 100-3000 psig


K c values measured by Fletcher, French and C o l l i n s ^ '^ for strontium sulfate in sodium chloride
solutions at atmospheric pressure are given in Appendix 17 (page 404). Their data was measured over
the following range:

Temperature: 10, 35, 50 and 69C [50, 75, 122 and 156F]

Ionic Strength: 0.1-5.25 moles/L

Pressure: 101.3 kPa [1 atmosphere]

The Effect of Ion Pairing on Sulfate Scale Solubility

The sulfate ion will f o r m neutral ion pairs or complexes with magnesium, calcium, barium and
strontium ions. The fraction of these ions which are tied up as ion pairs depends on the temperature,
pressure, ionic strength, and total concentrations of the ions in the solution.

The importance of ion pairing in scale prediction centers on the fact that ions occupied in ion
pairs are not available to f o r m scale.

For m a x i m u m accuracy, the available or " f r e e " sulfate ion concentration should be used in sulfate
scaling calculations. It can be calculated using the method of O d d o and Tomson presented in a follow-
ing section.

Estimation of the Amount of Sulfate Scale Precipitated

W h e n Equation 3.16 is used to calculate the solubility of the sulfate scales, the following equa-
tion can be used to calculate the m a x i m u m amount of scale which can precipitate:

P =(EW)( Actual -S) (3-57)

Where: P = maximum amount of scale which can precipitate (mg/L)

EW = gram equivalent weight of compound which precipitates
Actual = actual concentration of compound in solution, meq/L
S = calculated solubility of compound, meq/L

Oddo & Tomson Sulfate Scale Calculations 1994

T h e equations published by O d d o and Tomson in \994<3 ~" p e n n i t the calculation of the Satura-
tion Index, I s , for five sulfate scales: Calcium sulfate (gypsum, hemi-hydrate, anhydrite), barium sul-
fate and strontium sulfate. T h e equations account for pressure, temperature, ionic strength and ion

Saturation index calculations h a v e traditionally been reserved for calcium carbonate scale. The
saturation index is defined in Equation 3.16.

Although O d d o and Tomson designated the saturation index as "I s ", we have chosen to refer to it
as SI in this book and to identify the particular sulfate scale c o m p o u n d with appropriate subscripts.



Mass Action Stability Constant

The degree of ion pairing can be predicted if the mass action stability constant, K s t , is k n o w n for
each ion pair. In the case of the calcium and sulfate ions, the equation is given as:

j g ^ L

Where: [CaSOl ] = concentration of ion pair, moles/L

[Ca++ ] = concentration of " f r e e " calcium, moles/L
[SOj ] = concentration of " f r e e " sulfate, moles/L

It is a s s u m e d that K s t values for the four sulfate/metal ion complexes are equal.

Free Ion Concentrations

T h e total concentration of a given ion in solution is measured in a water analysis. It is the sum
of the paired ions and the " f r e e " ions. Only the free ions are available to f o r m scale.

In the case of sulfate, a material balance can be expressed as follows:

CSOj = [CaSOA} + [MgSOi;] + [SrSO^] + [BaSO] + [SO?] (3.59)

Measured Concentration Paired slfa'e Fre
Sul) aK

Similar equations can be written for m a g n e s i u m , calcium, barium and strontium ions.

CCa = [Ca++]+ = [Mg+] + [MgSOl] (3.60)

Cs,. = [Sr"]+[SrSO4] Cg<( = [Ba ] + [BaSO4]

Sulfate Scale Calculations

1. Calculate the molar ionic strength.

H {moles(L) = 10" 5 (2.2xNa* + 5.0x Ca++ + 8.2xMg + + +1.5xBa++ +2.3x Sr++) (3 6|)

+1.4 x cr + 2 . 1 x 5 0 ; +o.8 xhco;)

W h e r e all ion concentrations are in mg/L and are obtained f r o m a water analysis.

2. Calculate K s t

K = \0U*K- (3.62)
logATs, = 1 . 8 6 + 4 . 5 x l 0 " 3 r - 1 . 2 x l 0 _ f , r - + 10.7x 10 F
-2.38-^/jT + 0 . 5 8 j i - I S x l O ' ^ T
"d: T - temperature (F)
P = total pressure (psia)
p = ionic strength (moles/L)


3. Convert the ion concentrations from mg/L to moles/L by dividing the ion concentrations
obtained f r o m the water analysis by the appropriate conversion factor given in Table 3.7.

Conversion Factors: mg/L to moles/L
Ion Concentration To Obtain Total
(mg/L) F r o m Water Concentration
Analysis Divide by (moles/L)
S04= 96.060 Cso 4
Ca f + 40.080 cCa
Mg + + 24.305 C\ig
Sr ++ 87.620 Csr
Ba^ 137.330 CBa

4. Calculate CatSum. This is the sum of the four measured metal cation concentrations

CatSum = CCa + CMg + CSr + CBa (3.64)

5. Calculate the free sulfate concentration (moles/L).

- {l + K J CatSum - CSUj)} + {[l + K (CatSum -C^ )T +4Ks,CSOi V

\SO;] = ^ (3.65)

6. Calculate the f r e e metal-ion concentrations.

C ,Mg C,Ca
[Mg++) =
i + KAso;] + K,[SO=4]
Cs, CB(,
[Sr**] =
+ +
[ B a ]
I + KJSO;] + KSI[SO=4]

7. Calculate the Saturation Index values for the sulfate scales.

Calcium Sulfate

Equations are given for the calculation of the Saturation Index (SI) for three f o r m s of calcium
sulfate scale: gypsum, hemi-hydrate and anhydrite. According to O d d o and Tomson, the m o s t likely
scale to f o r m f r o m brines will be g y p s u m at temperatures less than 80C [176F],

Between 80C [176F] and I21C [250F]. any of the three types of calcium sulfate m a y form,
with g y p s u m being more likely at the low end of the temperature range and anhydrite more likely at
the high end. Hemihydrate is c o m m o n l y found in this temperature range in non-turbulent systems with
high ionic strengths.

A b o v e 121 C [250F], any calcium sulfate scale formed will almost certainly be anhydrite/ - 5 , A )



T h e saturation index for the three forms of calcium sulfate follow.


SIGyp = log10{[Cfl++][5O4=]} + 3.47 + 1 . 8 0 x l 0 - 3 r + 2 . 5 x l 0 " 6 r 2

5 3
- 5 . 9 x 10" / - + 0 . 4 9 f i - 0 . 6 6 x IQ~* J\iT


SIHemi = log K 1 { [ 0 / + + ] [ 5 0 ; ] } + 4 . 0 4 - 1 . 9 x l O' 3 T + 11.9 x 10" 6 r 2


- 6 . 9 x l O " 5 P - 1 . 6 6 ^ +0.49(i -0.66x10

SIAnhy = l o g 1 0 { [ C a + + ] [ 5 O ; ] } + 2 . 5 2 + 9 . 9 8 X10" 3 r + 0 . 9 7 X 1 0 " 6 r :
5 3
- 3 . 0 7 x 10" P - 1 . 0 9 7 ) 7 + 0.50|i - 3 . 3 x l O " 7 l T

Barium Sulfate

SIBa = logm{[5a++][SO;]} + 10.03-4.8xl0"3r + 11.4xl0"t,r2

5 3
- 4.8 x 1 0 " P - 2 . 6 2 - 7 ( 1 + 0 . 8 9 \ i - 2 . 0 x 1 0 ~ ^T

Strontium Sulfate

SISr = log | c {[5r ++ ] [50; ]} + 6J1 + 2.0 x 10"3 T + 6.4 x 10~b T2


- 4.6x 10" 5 P 1.89^ + 0.67fi -1.9xl0" 3 T


One of the primary causes of scale formation and injection well plugging is mixing t w o or more
waters which are incompatible. The individual waters m a y be quite stable at all system conditions and
present no scale problems. However, once they are mixed, reactions between ions dissolved in the in-
dividual waters m a y f o r m insoluble products. W h e n this occurs, the waters are said to be

For example, it would be a mistake to mix a water which contained a large quantity of barium
(Ba + + ) with a water containing a large amount of sulfate (SO4"), since the formation of B a S 0 4 w o u l d
be almost ccrtain.

Two waters might be described qualitatively as s h o w n in Table 3.8.

Depending on the amounts of each constituent present, the pH, temperature and the ratio in
which the two waters are mixed, you might expect any or all of the following precipitates to result:
calcium carbonate, calcium sulfate, barium sulfate, strontium sulfate, or iron sulfides.

The situation is further complicated if m o r e than two waters are mixed.

^ ^ ^ M i x i n g two or more incompatible waters on the surface for subsurface injection is obviously un-

Two Qualitative Water Analyses
Component Water " A " Water " B "
Ca Present Absent
HCO 3 - Absent Present
S0 4 = Absent Present
Ba Present Absent
++ +++
Fe or Fe Absent Present
H2S Present Absent

Incompatibility Between Injection and Formation Waters

The second area where problems may arise f r o m incompatibility is when the injection water is
not compatible with the natural formation water (connate water) in the zone where the water is being
injected. Surprisingly, very f e w plugging problems due to incompatibility occur in the injection well.

W h e n water is injected into a well, the injection water miscibly displaces the connate water. Me-
chanical mixing occurs at the boundary between the two liquids. This mixing p h e n o m e n o n is called
hydrodynamic dispersion. N o n - h o m o g e n e o u s permeability distribution has been suggested as the pri-
mary cause of dispersion. In addition, dissolved ions migrate across the boundary between the two
waters as a result of diffusion.

The relative contributions of hydrodynamic dispersion and ionic diffusion to the total mixing pro-
cess are a function of the velocity with which the boundary between the two waters m o v e s through the
reservoir rock. In most injection operations, the injection rates are such that dispersion is the predomi-
nate cause of mixing. However, in most laboratory studies, the individual contributions of dispersion
and diffusion are not separated, and the amount of mixing is described by an apparent coefficient of
dispersion, which includes both effects.

Warner^'-' 7 ' gives equations for the width and v o l u m e of the mixing zone:

w-4.69JWr (3.72)

Vm=5.\l H V D V 7 (3.73)

w = width of mixing zone, ft

vm = volume of mixing zone, bbl
D' = D/v = Coefficient of dispersion, ft
D = coefficient of dispersion, ft2/sec
v = average velocity of fluid flow, ft/sec
r = radial distance from the wellbore to the center of the mixing zone, ft
h = thickness of injection zone, ft
0 = porosity of reservoir rock, expressed as a fraction

Figure 3.17 illustrates the growth of the mixing zone as it m o v e s away f r o m the injection




LL 25 -

> 20

15 -

10 -

' (Assumes D' = 1 cm)
2 5
0 200 400 600 800 1000
Average Radial Distance From Wellbore, Ft.

Figure 3.17 Mixing Zone Width as a Function of Distance

The radial distance f r o m the injection wellbore to the center of the mixing zone is given as fol-

Where: V; = V o l u m e of water injected, bbl

The radial distance to the leading edge of the mixing zone f r o m the wellbore is equal to r +
0.5 w. Therefore, the total cumulative pore v o l u m e w h i c h is contacted by the m i x i n g zone f r o m the
time it leaves the wellbore until it reaches a radius, r, is:

V c =0.56A<l>(r + 0 . 5 wf (3.75)

Where: V c = V o l u m e of pore space contacted by leading edge of mixing zone, bbl

If one assumes the worst case, that two waters are completely mixed within the mixing zone
(which they are not), it is possible to calculate the m a x i m u m v o l u m e of scale which could precipitate
for a given set of conditions.
V = 2 . 9 x 10~ 6 '" (3.76)
Where: V p = volume of scale precipitated, bbl
PPT = weight of scale precipitated, lb/1000 bbl
s.g. = specific gravity of the scale formed

Assume the following:
1. T h e two waters are completely mixed within the m i x i n g zone.
2. Barium sulfate is being formed in the mixing zone. T h e specific gravity of BaSO.j is 4.2.
3. The m a x i m u m amount of scale which can precipitate is 100 mg/L (35 lb/1000 bbl).
4. The injection zone is 10 ft thick.

APTER 3 93

5. The fractional porosity of the injection zone is 0.20.

6. D' = 1.0 cm = 0.033 ft.
It is now possible to calculate the mixing zone volume, the total v o l u m e of reservoir pore space
contacted, and the v o l u m e of precipitate formed, as a function of the volume of water injected. The
results for the conditions stated in the example are presented in the Table 3.9.

Reservoir Mixing Calculations
V, (bbl) r (ft) W (ft) V m (bbl) Vc (bbl) VP (bbl)
1 1 0.8 2 2 4.1 x 10"5
10 3 1.5 10 15 2.3 x 10-4
10- 9 2.6 54 126 1.3 x lO"3
103 30 4.6 306 1151 7.3 x 10"3
104 95 8.1 1733 10 833 4.1 x 10"2
10 299 14.5 9675 104 736 2.3 x 10"'
10 945 25.7 54 364 I 027 617 1.29

The precipitate thus formed is distributed throughout the total volume of pore space contacted by
the mixing zone, V c , and occupies a negligible fraction of that space. In addition, once the mixing
zone has passed a given point in the reservoir, any precipitate is exposed solely to the injection water,
and will often be at least partially r e d i s s o l v e d / - 5 5 ^ Therefore, one would not anticipate any significant
reduction in injectivity due to incompatibility.

Several investigators have reported laboratory experiments which c o n f i r m this conclusion, both in
rocks containing 100% water* 3 - 31 - 3 - 3 ^ and w h e r e an oil saturation is present/ 5 - 5 -' Injection well plug-
ging due to this phenomena is considered extremely unlikely.

If you wish to avoid completely any possibility of permeability reduction, it is possible to inject a
buffer zone of a third water which is mutually compatible with both the injection water and the con-
nate water. The volume required to totally eliminate m i x i n g to any given radius is equal to the volume
of the mixing zone, V m , and can be calculated f r o m Equation 3.73.

It is also possible to treat the injection water with a scale preventive chemical to prevent precipi-
tation in the immediate area of the injection wellbore. However, this approach is unlikely to be effec-
tive farther out in the reservoir since most scale inhibitors have a strong tendency to adsorb on solid
surfaces. The large amount of surface area presented by the reservoir rock makes it extremely likely
that all inhibitor will be adsorbed by the rock within a veiy short distance f r o m the wellbore.

So the good news is that injection well plugging due to incompatibility between the injection w a -
ter and the connate water is an extremely unlikely event.

However, there is some a c c o m p a n y i n g bad news: Severe scale formation can occur in the pro-
ducing wells after injection-water breakthrough.

This can occur because there is ample opportunity for mixing of large volumes of injection water
and connate water in the producing wells after breakthrough of injection water. W h e n the producing
z o n e is comprised of several layers of different permeability (which is usually the case), the injection
water will break through earlier in some layers than in others. T h e result is the simultaneous produc-
tion of both injection and connate water. Simultaneous production of both waters can also occur due
to differing breakthrough times for water f r o m several injectors into a single producer.



Even in relatively h o m o g e n e o u s reservoirs, simultaneous production of the two waters will occur
since the reservoir v o l u m e swept by the injection water increases with time. Both injection water from
the swept area and connate water f r o m the unswept area will enter the producing wellbore, and the rel-
ative amounts will change over the life of the flood.

Evaluation of Incompatibility
Compatibility of water mixtures is assessed either by solubility calculations or by experimental

Solubility Calculations
If solubility calculations are to be performed, the following procedure is recommended:
1. Obtain analyses of the waters to be mixed.
2. Calculate a composite analysis for the waters at various mixing ratios of interest.
3. Calculate the average pH values for the various mixtures
4. Calculate the scaling tendencies of the mixtures.

Calculating the pH of a Mixture

pH = - log[/T ]; therefore. [//+ ] = 10"pw (3.77)

pHaxt = - l o g { F , [ / / + ] , + F2[H+]2+F}[H+]i + ... + Fn[_H+]n} (3.78)

Where: [H + ] = H
Fi, F:-. f"3v> Fn F
= H

Where: [H+] = hydrogen ion concentration, moles/liter

F|, Fi, F3,..., F = fraction of each water in a given mixture
IH + ]|, [H + ] 2 , [H + jj,..., [H+] = hydrogen ion concentration of each water

Compatibility Testing
Fresh samples of the waters to be mixed are taken, filtered to r e m o v e any suspended solids and
then mixed in several ratios. T h e various sample mixtures are set aside and observed to see if any pre-
cipitation occurs.

This procedure is quite useful for the detection of potential calcium, barium, or strontium sulfate
precipitation problems. However, it cannot be relied upon to give an accurate indication of the precipi-
tation of calcium carbonate or iron compounds because both are quite sensitive to pH changes. The
solubility of sulfate compounds is largely insensitive to changes in pH.


Avoid Mixing Incompatible Waters

! importance of staying away f r o m compatibility problems should be obvious f r o m the preced-
>ng discussion. A l w a y s be extremely careful w h e n you are considering mixing waters.


If you have two incompatible waters on the surface, and you wish to inject them both, there are
two methods of doing so without incompatibility problems.

1. Split System

Inject the two waters through separate injection systems into two different groups of wells.
2. Sequential Injection

Store the t w o waters in two sets of tanks and inject them alternately. The zone of mixing in
the injection line is small, and the v o l u m e of precipitate generated is insufficient to cause
plugging in the injection wells.

Water Dilution

This is just the reverse of the preceding problem. An injection water that normally would be
scale forming may be diluted with another water so that the resulting mixture is stable at system condi-
tions. However, test work should be carried out to ensure that this will be the case. It should be kept
in mind that reducing the overall T.D.S. of a water can sometimes induce scale f o n n a t i o n , particularly
barium sulfate

pH Control

L o w e r i n g the pH will increase the solubility of iron c o m p o u n d s and carbonate scales. However,
it will also tend to make the water more corrosive, and m a y create corrosion problems. pH has little, if
any, effect on the solubility of sulfate scales.

This is not a widely used method of scale control. It is usually practical only if a small shift in
pH is necessary to prevent precipitation of insoluble compounds. Accurate pH control is also neces-
sary and this is often difficult in n o m i a l oilfield operations.

Removal of Scale Forming Constituents

Dissolved Gas Removal

Dissolved gases such as H>S, CCK and O2 can be removed f r o m the water by chemical and/or
mechanical means. This removal can reduce the formation of insoluble iron c o m p o u n d s , by reducing
the corrosion that would have been caused by these gases in solution. It should be remembered that
these insoluble iron c o m p o u n d s are more correctly callcd corrosion deposits, rather than scale.

Simply r e m o v i n g CO2 f r o m water will actually increase the severity of scale deposition. H o w -
ever, the pH can be lowered sufficiently to convert all carbonates and bicarbonates to CCX Then re-
moval of CO2 will prevent the formation of carbonate scales.

Dissolved gas removal is discussed in Chapter 7.

Water Softening
Processes such as ion-exchange, precipitation softening or distillation are seldom used to prevent
scale deposition f r o m injection waters. T h e s e processes remove scale forming cations such as C a + +
and M g + + and m a y be used alone or in combination.



The problem with using these processes with oilfield brines centers on the cost involved to re-
move the inherently large quantities involved. It is nearly always far cheaper to use other methods of
scale control.

Of the processes listed, ion-exchange is probably the most generally applied in oilfield-related
operations. Ion-exchange units have been widely used to soften water for use as boiler feed water in
gas processing plants and for steam generators used in thennal recovery operations. They also are
used to soften w a t e r in several of the enhanced oil recovery processes.

Ion exchange and other ion removal processes are discussed in Chapter 8.

Nanofiltration is a membrane process that treats water between the reverse osmosis and
ultrafiltration levels in the filtration / separation s p e c t r u m / 5 " ^ It can r e m o v e particles in the 300 to
1000 molecular weight range, such as h u m i c acids and organic color f o u n d in water. Nanofiltration re-
moves minute particles f r o m water using a filter m e m b r a n e with extremely small pores (10" 3 microns
in diameter).

Nanofiltration m e m b r a n e s have been developed which selectively r e m o v e divalent ions such as

sulfate, m a g n e s i u m and calcium. Monovalent ions such as sodium and chloride pass through the

This technology has been applied in the North Sea to reduce the sulfate concentration of the in-
jected seawater as a m e a n s of controlling barium sulfate scale formation in the producing wells result-
3 34,3 35
ing f r o m incompatibility with the f o n n a t i o n water/ - - * Sulfate removal efficiency is reported to be
as high as 9 8 % at a recoveiy of 75%.

A recovery of 7 5 % m e a n s that 75% of the incoming water will be processed for injection. The
remaining 2 5 % is the concentrated reject stream which contains the ions removed f r o m the injection
water. It is discharged overboard. More information on m e m b r a n e filtration is provided in Chapter 8.

Scale Control Chemicals

Scale inhibitors are chemicals which will delay, reduce or prevent scale formation when added in
small amounts to a normally scaling water. Most of the scale inhibitors used in the oilfield function by
one or both of the following mechanisms:
W h e n scales first begin to forni, very tiny crystals precipitate f r o m the water. At this
point, the scale inhibitor adsorbs onto the surface of the crystals while they are still very
tiny and prevents further growth. This is thought to be the primary m e c h a n i s m by which
most oilfield scale inhibitors work.
In some cases scale inhibitors prevent the precipitated scale crystals from adhering to solid
surfaces such as piping or vessels.
It seems clear that a scale inhibitor which prevents precipitation is required for water injection
operations. If precipitation occurs the scale particles can still f o r m deposits and contribute to plugging
even if they are prevented f r o m forming adherent scales.

The tact that most c o m m o n l y used scale inhibitors function by inhibiting the growth of the scale
crystals means that the inhibitor must be present in the water at the point where the crystals begin to
JaSSL^ Providing an effective chemical has been chosen, two niles must be observed if scale inhibition
s to be successful. *


The inhibitor must be present in the water at the point where scale crystals begin to form
in order to have m a x i m u m effectiveness in inhibiting further growth. This m e a n s that the
inhibitor m u s t be applied upstream of the p r o b l e m area.
T h e inhibitor must be present in the water on a continuous basis so that it is available to
inhibit the growth of each scale crystal as it precipitates f r o m the water. Therefore, contin-
uous injection of scale inhibitor is required.

All c o m m o n l y used scale inhibitors are organic compounds. The most c o m m o n types are:

Phosphate Esters

Phosphate esters are not r e c o m m e n d e d for application above approximately 79C [175F]
as there is a danger of hydrolysis (reaction with F b O ) resulting in a loss of effectiveness.


Phosphonatcs are more temperature stable than the esters and should provide some protec-
tion up to 175C [350F],


Polymers such as the acrylics are primarily used for high temperature applications. They
can be used up to 175C [350F],

Selection and Evaluation of Scale Inhibitors

Like any other operation in the oilfield, scale inhibitors should be selected on a cost/effectiveness
basis. It is very tempting to choose the cheapest chemical available. This is very often a tragic en-or.
Sometime the most expensive material on a cost per unit volume of chemical basis will be the cheap-
est to use based on cost per volume of water treated. Selection should take into account the f o l l o w i n g

1. Chemical Composition of the Scale

Certain c o m p o u n d s are more effective for specific scales.

2. Severity of Scaling

T h e efficiency of scale inhibitors is primarily dictated by the level of supersaturation. The

higher the saturation index, the more difficult it is to inhibit scale foimation.

3. S y s t e m Temperature

Scale inhibitors generally b e c o m e less effective as the temperature increases. Each inhibi-
tor has an upper temperature limit above which they b e c o m e progressively less effective.

4. Laboratory Scale Inhibitor P e r f o r m a n c e Tests

Laboratory scale inhibitor p e r f o r m a n c e tests are comparative tests which d e t e n n i n e percent

inhibition as a function of scale inhibitor concentration. T h e data f r o m this type of testing
is intended to measure relative inhibitor effectiveness. It does not provide a recommended
scale inhibitor treatment concentration for field use.

5. Effect of Suspended Solids

If substantial amounts of suspended solids are present, higher inhibitor concentrations will
be necessary. T h e reason is that scale inhibitors will adsorb onto the surface of solids in
the water, thereby reducing the amount available to inhibit scale fonnation.



6. Compatibility with Other Chemicals

T h e scale inhibitor should be compatible with any other chemicals being added to the sys-
tem, such as o x y g e n scavengers, corrosion inhibitors, or biocicles. It is important that there
is no reaction or interference a m o n g the different chemicals being used in a system.
7. Compatibility with the Water

S o m e scale inhibitors will react with calcium, m a g n e s i u m or barium ions to f o r m rather in-
soluble c o m p o u n d s . Precipitation of these c o m p o u n d s can result in the formation of a
scale, thus creating a n e w problem. Generally, phosphonates and polymers are m o r e prone
to give this problem than the esters. <3 - 36 >

If high concentrations of these ions are present in the water the scale inhibitor must be se-
lected with care to avoid this problem.
8. Effect of Dissolved Iron on Calcium Sulfate Inhibition

Most scale inhibitors are less effective in inhibiting calcium sulfate scale f o r m a t i o n when
dissolved iron is present in the water. T h e effect is severe in the case of phosphate esters
and phosphonates, while most polymers show only slight decreases in performance.

While it is possible to compensate for this effect by increasing the scale inhibitor concen-
tration, the additional cost can be substantial.

Soluble iron concentrations are typically quite Low in sour waters and in waters containing
dissolved oxygen, since most of the iron will precipitate as iron sulfide or ferric hydroxide.
However, concentrations can be substantial in sweet, air-free systems.

Since corrosion is the most c o m m o n source of soluble iron in oilfield waters, an effective
corrosion control program is essential for cost-effective scale inhibition. If the source of
iron is the producing formation, the use of polymcr-typc scalc inhibitors should be
9. Effect of pH

Some scale inhibitors become less effective against B a S 0 4 and C a S 0 4 scales at pH values
below 6. Since values less than 6 are quite c o m m o n in oilfield waters, this can be an ex-
tremely important consideration in chemical selection.
10. Weather

Is it a cold weather operation? Be sure to determine the pour point of any scale inhibitors
and design the chemical injection system to avoid freezing problems.

Inhibitor Application
In production systems scale formation m a y occur at any point where a pressure drop occurs, or
anywhere the produced water is heated. Hence, the scale inhibitor injection point will be dependent
upon where the problems occur within a system (see Figure 3.18).

During the design stage, access points for chemical injection facilities should be provided at rele-
vant points throughout the system in order to cater for all circumstances. Including such points at the
outset is a minor cost in comparison to installing t h e m once production has c o m m e n c e d .

Items to consider for the protection of surface facilities include:

Continuous injection of inhibitor should be applied as near to the well head as possible
Additional injection points may be required in areas of severe problems


Within today's business and political environment, it is usual to see good corrosion management
as a m a j o r goal of most oil and gas production operating units. The concern is threefold, taking into
account financial costs, health and safety issues and a strong desire to ensure the protection of the en-
vironment (or at least to be seen to be making such efforts).

It is difficult to find reliable estimates for the overall cost of corrosion worldwide, but we do
have s o m e reasonable data for the USA. It has been estimated that corrosion costs the United States
approximately $276 billion per year. For the US Oil and Gas production industry, the figure is US
$1,372 million a year/4 -"

It is generally considered that about one-third of these costs could be saved by the proper appli-
cation of existing technologies. For example, 2 0 - 2 5 % of US pipeline failures are attributed to corro-
sion. With respect to external pitting corrosion, it is usually due to coating defects (holidays, coating
disbondment, or moisture penneation), inadequate cathodic protection (insufficient current due to poor
design, hot spots, or localized shorting), or both.

With respect to health and safety issues, it is the leaks or other mechanical failures related to cor-
rosion reactions that are of concern. T h e subsequent release of hydrocarbons or associated chemicals
such as hydrogen sulfide, bring the risks of fire, explosion, or toxic atmospheres. Such hazards can
lead to the injury or death of employees or neighboring residents and these will incur both direct oper-
ating costs and potential compensation claims. If the incident is later judged by the authorities as hav-
ing been preventable, then the operator m a y face significant financial penalties in the form of fines and
other legal costs.

In most countries environmental concerns can no longer be ignored. This is especially true for
arge multi-national companies which must avoid upsetting their "neighbors". Even if an oil and gas
production company has the best of intentions with regard to protecting the environment, various
green" organizations around the world act as self-appointed watch-dogs keeping a vigilant eye on our
industry. Pollution of the environment does not just incur the cost of cleanup. Even the financial pen-
alties from local and national regulatory bodies may be regarded as insignificant compared to the ad-
v e r s e political effects and deterioration in the relationship with the local community.


An important aspect in attempting to minimize the adverse effects noted above, is having the
k n o w l e d g e and understanding of w h y and how the corrosion takes place. It is then possible to deter-
mine w h a t might be the best approach to reduce that corrosion.

Let us start with the N A C E (National Association of Corrosion Engineers) International defini-

"Corrosion is the deterioration of a substance (usually a metal) or its properties, because of a

reaction with its environment."

In the oil and gas industry the main construction material is carbon (mild) steel. Various stainless
steels and other ferrous and non-ferrous alloys are used to a lesser extent, as well as some man-made
non-metallic materials.

Most metals are f o u n d in nature as metallic oxides or salts, such as the iron oxide. F e i O j . Refin-
ing to produce pure metal requires a large energy input. This energy is "stored" and is available to
supply the necessary driving force to return the metal to its original state an oxide or salt. This
means that metals are unstable with respect to most environments and have a natural tendency to return
to their original lower-energy state, or " c o r r o d e . "

Iron Ore (Fe2Oi)+ Heat Iron (Fe) -> +C Carbon Steel

{ Low Energy Stale I (High Energy Stale)
Corrosion is an electrochemical process. This means that electrical current flows during the cor-
rosion process. In order for current to flow, there must be a driving force, or a voltage source, and a
complete electrical circuit. A cathode reactant, or oxidizing agent, is also needed.

Voltage Source

T h e source of voltage in the corrosion process is the energy stored in the metal by the refining
process. Different metals require different amounts of energy for refining and therefore have different
tendencies to corrode. This is illustrated in Table 4.1.

Electromotive Force Series tor Metals
Energy to Potential Tendency to
Refine Metal (Volts)* Corrode
Most Energy Magnesium -2.37 Most Eager
Aluminum -1.66
Zinc -0.76
Iron -0.44
Lead -0.13
Copper +0.34 to +0.52
Silver +0.80
Least Energy Gold + 1.58 to +1.68 Least Eager
* Half-cell potential in solution of own salts, measured with respect to hydrogen
reference electrode/"*-J


i n t u n i wr w n n u o i u i i

The magnitude of the driving voltage generated by a metal w h e n it is placed in an aqueous solu-
tion is called the potential of the metal. It is related to the energy that is released when the metal cor-
rodes. Potential values are a function of both the metal and the chemical and physical characteristics
of the water. T h e absolute value of the potential of a given metal is influenced by the water composi-
tion, temperature, velocity and many other factors. However, their relative values remain about the
same in m o s t solutions of interest. Therefore, m a g n e s i u m usually has a greater tendency to corrode
than does iron in the environments of interest.

The Electrical Circuit

In addition to a source of voltage, there must be a complete electrical circuit. T h e electrical cir-
cuit of the corrosion process consists of four parts.

The anode is that portion of the metal surface that is corroded. It is the point at which metal dis-
solves, or goes into solution. When metal dissolves, the metal atom loses electrons and goes into solu-
tion as an ion. Since atoms contain equal n u m b e r s of protons (positively charged particles) and elec-
trons (negatively charged particles), the loss of electrons leaves an excess of positive charges, and the
resulting ion is positively charged. T h e chemical reaction for iron is:
Fe - Fe++ + 2e
Iron Atom Iron Ion Electrons

This loss of electrons is called oxidation. The iron ion goes into solution and the two electrons
are left behind in the metal.

Anode summary:
metal loss
metal ions ( F e : + ) go into solution
loss of electrons = oxidation

The cathode is that portion of the metal surface which does not dissolve, but which is the site of
another chemical reaction(s) necessary to the corrosion process. The electrons left behind by the
oxidation of iron at the a n o d e travel through the metal to the cathodic surface area where they are con-
sumed by reaction with ions present in the water. T h e consumption of electrons is called a reduction
Typical cathodic reactions are s h o w n below. Acid solutions, w h e n o x y g e n is absent:
2H + 2e H2
Hydrogen Ions Electrons Hydrogen Gas

Or, if oxygen is present two other reactions are also possible:

0, +4H + 4<?-> 2H20
Acid Solutions

O, + 2H20 + 4e^- 40H-

Neutral and Alkaline Solutions


Thus, the reaction at the anode areas produces electrons and the reaction(s) at the cathode areas
consumes the electrons. This is the essential feature of an electrochemical reaction: electrons are gen-
erated by a chemical reaction at one point, and then travel to another point where they are consumed
by another reaction. As you may recall, electrical current flow is actually the passage of electrons
from one point to another. Convention says that the electrical current flows in the opposite direction
of electron travel. Thus, as electrons flow from the anode area to the cathode area, electrical current
flows through the metal in the opposite direction, from the cathode to the anode.

Cathode summary:

no corrosion
no metal loss
gain/use of electrons = reduction

In order to support the reactions previously listed and to complete the electrical circuit, the metal
surface (both the anode and the cathode) must be covered with an electrically conductive solution.
Such a solution is called an electrolyte. The electrolyte conducts current from the anode to the cath-
ode. When water contains dissolved ions it acts as an electrolyte. Its electrical conductivity increases
as the amount of dissolved salts or ions increases. In practical terms, pure water (no dissolved salts,
ions or gases) is not coiTOsive. An oxidizing agent must be present to act as the cathode reactant.
Apart from mineral acid use in well stimulation work, dissolved gases are the main corrosive species in
oil and gas production systems.

Hence, liquid water must be present for corrosion to occur at the normal temperatures existing in
oil and gas production systems. As a result, this process is often tenned "aqueous corrosion".

At higher temperatures (say 200+C [390+F]) corrosion reactions differ in that liquid water is
not present. However, an electtochemical process can still take place due to the presence of other liq-
uid electrolytes. An example is molten vanadium pentoxidc (occurring on surfaces where vanadium
bearing crude oils are burned), which acts as both the electrolyte and the oxidizing agent. However,
high temperature corrosion will not be considered within this book.

Electron Conductor
The anode and the cathode must be connected by something which will conduct electrons (elec-
trical current) in order to complete the circuit and provide a path for current to flow from the cathode
back to the anode. In the case of a corroding metal surface, the metal itself is the electron conductor.

This combination of anode, cathode, electrolyte, and electron conductor is called a corrosion
cell. A schematic of the corrosion process for iron is shown in Figure 4.1. This sketch is merely an il-
lustration. Metal atoms do not necessarily dissolve at a single point on a metal surface, nor are cath-
ode areas restricted to one area on the surface. In the case of localized corrosion, such as pitting, these
processes are limited to localized areas. However, in the case of general corrosion the reactions occur
randomly over the metal surface.

At this point in the study of con'osion a key question arises, "Why do certain areas of the metal
surface act as anodes?" The answer is not a simple one. However, in most cases the reason centers on
inhomogeneities in the metal surface, in the electrolyte, or both.


Fe Fe + 2e

Electron Flow

2H + + 2e > H2 T


Figure 4.1 The Corrosion Cell


Commercial metals are not h o m o g e n e o u s , but contain inclusions, precipitates, and perhaps sev-
eral different phases. W h e n the metal is placed in an electrolyte, potential differences exist between
these phases, resulting in corrosion cells on the metal surface.

For e x a m p l e , steel is primarily an alloy of iron and carbon. Pure iron is a relatively weak, duc-
tile material. W h e n it is alloyed with small amounts of carbon (usually 0.2 to l.O percent), a m u c h
stronger material is created. However, as a result of reacting part of the iron with carbon, we n o w
have a metal composed of t w o materials: pure iron and iron carbide ( F e j C ) , the product of the
iron-carbon reaction. T h e iron carbide is distributed within the iron as tiny microscopic islands, and
these islands of iron carbide have a lower tendency to corrode than docs the pure iron. T h e F e j C and
the pure iron are in intimate contact (allowing electron flow), so w h e n the steel is placed in water (an
electrolyte), the electrical circuit is complete and current flows through thousands of tiny microcclls on
the steel surface. T h e pure iron acts as the anode and corrodes, while tiie Fe^C acts as the cathode.

If we were to look through a microscope and concentrate on two adjacent grains of Fe and F e j C
on the surface of a piece of steel immersed in water, it might look like Figure 4.2.

Current . _

Figure 4.2 Fe-Fe 3 C Corrosion Cell


This condition may change with time. As corrosion products accumulate, the potential distribu-
tion on the metal surface may change, shifting the anodic areas.

Intergranular attack is caused or accelerated by potential differences between the grains and grain
boundaries. Local heating can result in changes of the nature of phases or their compositions creating
differences in potential, such as in weldline c o r r o s i o n / * J J

Thus, metals are inherently inhomogeneous materials, and potential differences on the metal sur-
face are a natural result. These differences are one of the primary causes of corrosion.

The important thing to realize is that direct current flows during the corrosion process, and the
amount of current flowing is a measure of the severity of the corrosion. That is, the greater the cur-
rent flow, the larger the amount of metal loss. It is dependent on both the nature of the metal and the
type of environment, or electrolyte. As a "rule of thumb", one ampere of corrosion current flowing for
one year represents the loss of approximately 8.2 kg [18 lbs] of iron f r o m a corroding structure.


As previously stated, different metals have different tendencies to con-ode. However, a discus-
sion of a wide range of metals is not really warranted here. Steel is the primary metal used in the oil-
field and so we will stick pretty closely to the subject of h o w and w h y steel corrodes in aqueous (water
based) electrolytes.

There are, of course, m a n y "steels." Simple low carbon steel is universally used in production
operations for lines, tubing, tanks and treaters. However, some steels containing chromium and nickel
(which add corrosion resistance), and some non-ferrous alloys (contain no iron) are used in a f e w items
such as p u m p s and valves.

N A C E Standard RP0475-98,"Selection of Metallic Materials to be Used in All Phases of Water

Handling for Injection into Oil Bearing Formations, " is an extremely useful guide to materials selec-
tion for water injection s y s t e m s / * ^ A useful guide for production systems is the N A C E Standard
MRO175-2003, "Metals for Sulfide Stress Cracking and Stress Corrosion Cracking Resistance in Sour
Oilfield Environments. "(45}


The speed (rate) of the corrosion reaction is affected by the following parameters.
Water conductivity (due to dissolved ions)
Water pH
Dissolved gas concentrations
Water temperature
Gas pressure and the partial pressures of individual gases
Water stream velocity.

Conductivity of Water
Pure water (no dissolved ions or gases) has virtually no conductivity and hence does not support
corrosion reactions. However, as the concentration of ions increases in the water, its conductivity



increases along with its ability to support the electrochemical corrosion reactions. Hence, when an ox-
idizing agent is present in the water its corrosivity increases as that water gets "saltier".

There is a complicating factor in that the corrosion rate can decrease as T D S concentration in-
creases, due to decreased gas solubility and diffusion rate. Also, increased T D S can lead to the precip-
itation of protective scales on the metal surface.

pH of Water
T h e corrosivity of the w a t e r usually increases as its pH decreases (becomes more acidic). At
higher pH values protective scales (iron hydroxide, or carbonate scale) may f o r m on the steel surface
and prevent or slow d o w n further corrosion. Oilfield water systems are typically in the range of pH
4.0 to 8.3 (see Figure 4.3), apart f r o m in localized areas w h e r e mineral acid well stimulation work is
being conducted.

/ PH

Figure 4.3 Effect of pH on Corrosion Rate

Dissolved Gases

Oxygen, carbon dioxide or hydrogen sulfide dissolved in water causes it to be corrosive. In fact,
dissolved gases are the primary cause of m o s t "oil and gas system" corrosion problems. If they could
be excluded and the water maintained at a neutral pH or higher, m o s t water systems would have very
few corrosion problems.

Dissolved Oxygen
Of the three dissolved gases mentioned, oxygen is by far the most corrosive. Figure 4.4 illus-
trates the effect of oxygen concentration and temperature on the corrosion rate of low carbon steel in
tap water. While oxygen should normally not be present in oil production systems, it is sometimes


tE, >

CD 05
cc 15
DC 4.0 160 CC
c c
o o
'55 "c/>
c o

0 1 2 3 4 5 6 7 8 9 1 0 1 1
Oxygen (ppm)

Figure 4.4 Corrosion of Steel in Tap Water

found as a result of leaks ( O i diffusion into the system). However, it can be a significant problem in
secondary recovery water injection systems.

It is possible to estimate the corrosion rate of mild steel in fresh water using an equation devel-
oped by Pisigan and Singley/ 4 - ; The equation is empirical, and is based on a correlation between the
coiTosion rate of mild steel and several analytical parameters.

mpy = r^ (4.1)
(io s ') a87 W 373
Where: mpy = mils (thousandths of an inch) per year
TDS = total dissolved solids, mg/L
DO = dissolved oxygen concentration, mg/L
SI = Langelier Saturation Index (See Chapter 3)
Day = coupon exposure period, days

This equation applies only to the experimental conditions e m p l o y e d in this study, which are sum-
marized in Table 4.2. T h e corrosion rates were calculated f r o m the weight loss of mild steel coupons
in jars which were stirred at 30 rpm. A total of seventeen waters w e r e used in the study, which w a s
conductcd at a temperature of 20 3C. This equation does not consider temperature or velocity as



Means and Ranges of Data

Parameter Mean Range

TDS, mg/L 473 116-791
Ca ++ , mg/L 68 10-120
M g ^ , mg/L 16 2-52
Na , mg/L 64 7-130
CI", mg/L 119 18-213
S 0 4 = , mg/L 79 16-208
Alkalinity, mg/L as CaCOj 116 21 -232
pH, nnits 7.77 6.70-8.95
pHs, units 7.79 7.60-9.07
Langelier Index, units -0.02 -2.12-1.90
Ionic Strength, moles/L 0.0081 0.0001-0.0198
Dissolved Oxygen, ppm 5.1 1.4-9.1
Corrosion Rate, mpy 11.9 2.1-36.6
Corrosion Rate, mm/y 0.301 0.053-0.930

Dissolved oxygen can cause severe corrosion at extremely low concentrations. Figure 4.5 shows
measured corrosion rates of mild steel as a function of dissolved oxygen concentration in sulfide-free
produced water being injected at Huntington Beach, C a l i f o r n i a / *
If dissolved HTS and/or C O : present, even trace quantities of dissolved oxygen will drasti-
cally increase their corrosivity.

O x y g e n is not present in produced water, but often c o m e s in contact with it as or after it is

brought to the surface. Water f r o m lakes or streams will be saturated with oxygen, and water f r o m
shallow wells may contain some oxygen. Whenever possible, it should be rigorously excluded.

T h e solubility of o x y g e n in water is a function of pressure, temperature, and the chloride content.

Oxygen is less ^oluble in salt water than in f r e s h water. (See Figure 6.5)

O x y g e n corrosion occurs as follows:

A n o d e Reaction

FeFe++ +2e
Cathode Reaction (in neutral or alkaline waters)

O, +2H20 + 4e-+ AOir

or, c o m b i n i n g the two

4Fe++ + 6H20 + 3 0 , -> 4 F e ( O H ) }

If the pH is a b o v e 4, ferric hydroxide is insoluble and precipitates.

O x y g e n accelerates corrosion drastically under most circumstances. It does so primarily by act-

ing as a "cathodic depolarizer."' This m e a n s that it will easily combine with electrons at the cathode
and allow the corrosion reaction to proceed at a rate limited primarily by the rate oxygen can d i f f u s e to
the cathode. Without oxygen, the corrosion reaction is limited by the rate it can dispose of electrons
by reaction with other species in the water such as hydrogen ions. W h e n oxygen is present, it is an


Corrosion Rate (pm/y)

0 50 100 150 200 250

2 4 6 8 10
Corrosion Rate (mpy)

Figure 4.5 Effect of Oxygen Concentration on Corrosion Rate

additional and eager consumer of electrons at the cathode surface, w h i c h results in accelerated corro-

Oxygen attack is usually pitting in nature.

It is important to realize that the corrosion of mild steel increases as the dissolved oxygen content
of the water increases up to a point. If there is sufficient oxygen in the water, the Fe"h+ F e + + + ox-
idation may occur very rapidly, b e f o r e the Fe~ ions have a chance to diffuse away from the metal sur-
face. In this case F e ( O H ) 3 can form on the metal surface and become p r o t e c t i v e / * 6 ' However, in
most oilfield waters sufficient chloride ions are present to interfere with the fonnation of a protective
film and corrosion rates continue to increase with oxygen concentration.

Concentration cells, or differential aeration cells, can cause preferential attack or pitting. A n y
time there is a difference in the oxygen content of water in two adjacent areas in a system, attack will
take place preferentially in the area exposed to the lowest oxygen concentration. Typical examples are
water-air interfaces, crevices and "oxygen tubercles" in water systems.

Dissolved Carbon Dioxide

When carbon dioxide dissolves in water it f o r m s carbonic acid, f r o m which the H + ions decrease
the pH of the water and increase its corcosivity. It is not as c o n o s i v e as oxygen. In s o m e circum-



stances it gives rise to uniformly thinned surfaces, but usually results in pitting. Corrosion caused by
carbon dioxide dissolved in water is called " s w e e t " corrosion.

CO, + H2O <-> H2COy H2C03 <-> 2H+ + CO;

Fe + H2CO3 FeCOs +H2 Fe + 2H+ + CO; - Fe++ + CO; + H2
T h e solubility of CO2 in water, like all gases, is a function of the partial pressure of C O i in the
atmosphere above the water. The greater the partial pressure, the greater the solubility. Therefore, in
two-phase systems (gas + water), corrosion rates go up as the partial pressure of CO2 increases. T h e
effect of C O i partial pressure on the corrosion rate of mild steel in distilled water is illustrated in Fig-
ure 4 . 6 / ^

Partial Pressure of COa (kPa)

1000 2000 3000
1600 I i 64

1200 48 E
13 13
GC 800 - / - 32 IE
0 0
(0 03
c 400 c
0 / 16 k0_
0 0

0 / I I I 0
150 300 450
Partial Pressure of COz (psia)

Figure 4.6 Effect of C 0 2 Partial Pressure on Corrosion R a t e ^

DeWaard and M i l l i a m s ^ ^ measured the corrosion rate of X-52 carbon steel in an oxygen-free
solution of 0.1% (1000 ppm) NaCl saturated with CO2 at partial pressures from 5 - 1 0 3 kPa [0.7-15 psi]
over a temperature range of 5 - 8 0 C [41-176F]. Both polished and grit-blasted stee! specimens w e r e
used. The solution was stirred to achieve a velocity of about 0.9 m/'s [3 ft/sec].

Based on the measured corrosion rates, they developed an equation to predict "worst case" uni-
form corrosion rates. T h e equation was subsequently modified to read as f o l l o w s / * u >

logCK = 5 . 8 - + 0.67 log(/*%>,) (4.2)

Where: CR = corrosion rate, rnni/y

T = temperature, IC (K = C + 273)
pC0 2 = partial pressure of C 0 2 , bar

This equation is presented in the f o r m of a n o m o g r a p h in Figure 4.7. A scale factor of 1 is sug-

gested for brines.

A "rule-of-thumb" guide to whether corrosion is likely to occur based on the partial pressure of
carbon dioxide, is:

Above 30 psi Corrosion is most likely to occur

3 to 30 psi Corrosion might occur
Less than 3 psi > Probably non-conosive


CO, Pressure
"bar T 1 0

Corrosion Rate
" 1 . 0

: : 0.1

-1- 0 . 0 1

Figure 4.7 C 0 2 Corrosion Rate P r e d i c t i o n ^ " '

The corrosion rates measured at high CO2 partial pressures are extremely high. It is likely that in
most cases these high rates will not persist v e i y long in operating systems due to the formation of a
layer of corrosion product on the metal surface. As a corrosion product layer forms, uniform corrosion
decreases, but pitting can b e c o m e an extremely severe problem.

Equation 4.2 can be used for low T D S waters in air-free, sweet, two-phase (gas + water) systems
to estimate the m a x i m u m uniform corrosion rates of a bare-steel surface in the absence of an iron car-
bonate corrosion product film.

As temperatures increase, a protective layer of iron carbonate may f o r m on the surface and slow
down the rate of corrosion. It is extremely difficult to predict whether or not a protective iron carbon-
ate scale will f o r m , and if so at what temperature, because it is a function of many factors including
CO? partial pressure, velocity, and water composition. In many cases, the iron carbonate scale is not
protective, and accelerated corrosion is observed beneath the s c a l e / * n " * 1 3 )

In addition, the presence of liquid hydrocarbons may reduce corrosion under certain condi-

In water systems containing bicarbonate alkalinity, the amount of CO? present to cause con-osion
is a function of pH due to the CO^-bicarbonate-carbonate equilibria discussed in Chapter 3.

As previously mentioned, the presence of any oxygen increases the corrosivity of C O i .

Dissolved Hydrogen Sulfide

Hydrogen sulfide is very soluble in water and when dissolved behaves as a weak acid. For acids,
the smaller the n u m b e r for the ionization constant K a , the less it is ionized. Consider the following

H2S + H20 <r+ HS~ + / / 3 0 + (hydroxonium ion: H 3 0 + <-> IF" + H , 0 )

A small K a value m e a n s that the reaction is very m u c h to the left hand side. For H 2 S K a = 8.9 x
10" mol/dm. C o m p a r e this to acetic acid (K a = 1.7 x 10" 5 ) and is only about 1% ionized. Therefore,



H 2 S is a very weak acid. T h e degree of dissociation and hence the relative concentrations of
H?S/HS~/S 2 ~ are a function of the pH of the solution, as illustrated in Figure 4.8. As the pH of a solu-
tion decreases (towards acid) the H 2 S concentration increases relative to the ionized f o r m s (HS~ and
S 2 _ ). At normally encountered pH values, sour waters will contain a mixture of H?S and HS~. (HS~ is
the bisulfide or hydrosulfide ion). Temperature, salinity of the water and its dissolved oxygen content,
also affect the amount of H 2 S w h i c h ionizes.

Figure 4.8 Ionization of H 2 S in Water as a Function of pH

Systems are termed:

Sweet with < 5 mg H 2 S/liter

Sour with > 5 mg H 2 S/liter

Corrosion caused by H 2 S dissolved in water is called "'sour" corrosion and is usually pitting in
nature. The direct corrosion reaction for steel might be s h o w n as:

Fe + H2S + H20 -> FeS + H2+ H20

However, this reaction is kinetically slow. Since CO? is virtually always present in oil and gas
production systems, it is more likely that the carbonic acid first attacks the steel. T h e H?S then com-
bines with the metal ions (Fe 2 + ) released by that corrosion reaction on the steel:
Fe2+ + H2S ~> FeS + 2H+
The combination of H 2 S and CO? is more aggressive than H?S alone and is frequently found in
oilfield environments. Once again, the presence of even minute quantities of oxygen can be disastrous.

The iron sulfide corrosion product is extremely insoluble and usually adheres to the steel surface
as a scale (not to be c o n f u s e d with mineral scale). It is an excellent electron conductor and is cathodic
to the underlying steel. The steel and iron sulfide thus f o r m a galvanic couple w h i c h tends to acceler-
ate the corrosion at defects in the scale. Deep pitting is the normal result.

Other types of corrosion damage can also occur in sour systems: sulfide stress cracking (SSC).
hydrogen blistering, and a specialized form of hydrogen d a m a g e called hydrogen induced cracking
t H l C ) . T h e s e forms of corrosion are discussed later in this chapter.


Hydrogen sulfide can also be generated in water systems by sulfate reducing bacteria. This will
be discussed in greater detail in Chapter 5.

Organic Acids
Produced waters usually contain varying amounts of volatile organic acids, with acetic and
propionic acids being the most c o m m o n . In low concentrations acetic acid is extremely corrosive and
its attack on steel is approximately as severe as that of hydrochloric or sulfuric acid. Experimental
&afa ( 4 1 5 ) has shown that a concentration of 0.1% (1000 p p m ) acetic acid produced a corrosion rate of
about 50 m p y (mils per year) on a steel surface.

In conjunction with the CO? present in production systems, acetate and propionate increase the
coirosion rate of mild steel. That increase is reduced in the presence of H?S. It has been shown in
field applications that the presence of these carboxylic acids has no an adverse effect on corrosion in-
hibitor p e r f o r m a n c e / " * 9 ^

Measuring Corrosion Rates

Traditionally corrosion rates have b e e n measured in 'mils per year', where a mil is one-thousandth
of an inch (1 mil = 0.001 inch). With the m o v e to SI metric units we now o f t e n see the term " m m p y "
(millimeters per year). It is important not to get these units confused since 1 m p y is regarded as a low
corrosion rate, whereas 1 m m p y is quite severe corrosion. Since,
1 in = 1000 mils =25.4 mm

or 1 mm - = 29.31 mils (say 40)

Therefore 1 m m p y is approximately equal to 40 mpy.

Physical Variables

Water Temperature
Corrosion rates usually increase as temperature increases because all of the chemical reactions in-
volved speed up.

In a system open to the atmosphere, corrosion rates m a y increase at first and then with a further
temperature increase, the corrosion rate m a y drop off due to dissolved gases coming out of solution. If
the system is closed, the corrosion rate will continue to increase with temperature because the dis-
solved gases have no place to go.

Figure 4.9 illustrates the effect of temperature on the coirosion rate of mild steel in water con-
taining dissolved oxygen.

W h e n the water contains bicarbonates, increases in temperature will promote scale formation
which may slow down the corrosion reaction. H o w e v e r , if the temperature is high enough, it can also
lead to decomposition of bicarbonates and produce additional CO2.

System Pressure
Pressure also has an effect on chemical reactions. In the case of two-phase water systems (wa-
ter + gas), the primary importance of pressure is its effect on the solubility of dissolved gases. More



Temperature, C
80 120 160
750 30

a} 500 20
oc ccs
c e9
o 3
5 250 10
E o
o O

32 1 00 1 50 200 250 300 350

Temperature, F

(4 16)
Figure 4.9 Effect of Temperature on Corrosion Rate -

gas goes into solution as the pressure is increased resulting in increased corrosion rates. Two gas laws
support this statement. Dalton's Law states that the total pressure exerted by a gas mixture is the sum
of the partial pressures exerted by each component. Henry's Law states that the amount of a gas which
is dissolved in a liquid is directly proportional to its partial pressure in the atmosphere a b o v e that

Water Velocity
Stagnant or low velocity water usually gives a low general corrosion rate, but pitting is more
likely. The fluid velocity should be high enough to keep the majority of any solids entrained in the
fluid and prevent excessive deposition in the system. The obvious consequences of solids deposition
are the creation of sites for concentration cells and bacterial growth beneath the deposits. Even if sol-
ids deposition is minimal, bacterial activity is always more likely in low velocity areas such as tanks
and f i l t e r s / * ' 7 '

Corrosion rates usually increase with velocity. This is due to the fact that increased fluid velocity
speeds up the transport of diffusion-limited species to and f r o m the metal s u r f a c e / * 7 5 ' For example,
the m o v e m e n t of dissolved oxygen to the metal surface to participate in the corrosion reaction is con-
trolled by diffusion, and increased velocity accelerates con-osion caused by dissolved oxygen.

Figure 4.10 shows the effect of velocity on the corrosion rate of mild steel in seawater at ambient
.temperature exposed for 38 days.

If the velocity is too high, erosion, erosion-corrosion, or impingement m a y result. These prob-
lems can be especially severe if the water contains large amounts of suspended solids, and are usually
worst at changes in direction, such as elbows. Partially protective corrosion product f d m s are con-
stantly removed or eroded away, leaving a bare metal surface which eagerly corrodes.

Erosional Velocity

The erosional velocity> above which erosion m a y occur can be estimated from an empirical equa-
tion presented in API R P 14E. The equation has no demonstrated relationship with corrosion. How-
ever, if the fluid is corrosive, the onset of erosion has a catastrophic effect on the rate at which steel


Velocity (m/s)

Velocity (ft/s)

Figure 4.10 Effect of Velocity on Seawater Corrosion <4-19>

Where: Ve =
maximum allowable velocity, ft/sec, above which erosion would be expected to
occur for a clean, solids-free fluid
C ~ a constant, typically 100-125
p = density of fluid at flowing pressure and temperature, lb/ft 3

This equation was originally developed by the United States Navy during World War II. Based
upon tests with solids-free fluids they arrived at a value of C = 160. However, the API r e c o m m e n d s C
values of 125 for intermittent flow and 100 for continuous flow. This equation is used for gas/liquid
mixtures as well as liquids.

If the fluid contains significant amounts of suspended solids, many operators use values of C less
than 100.

For water, the equation becomes:

V - (4.4)
J62.4 x S.G.

A s s u m i n g that the water has a specific gravity of 1.0, we can solve the equation for several as-
s u m e d values of C to examine the range of probable erosional velocities as shown in Table 4.3.

Erosional Velocities
c m/s ft/sec
160 6.2 20.3
125 4.8 15.8
100 3.9 12.7
80 3.1 10.1



Thus, based on Equation 4.4, it is unlikely that pure erosion would constitute a major problem in
most water injection systems, regardless of the value of C which is chosen.

High velocities are usually not an issue in systems where the fluid must be p u m p e d considerable
distances, as the pump horsepower required to o v e r c o m e friction losses at higher velocities tends to be
excessive. However, high velocities are c o m m o n in some compact offshore water injection systems
where distances are short and the volumes injected are large.


Experience gained in the United States in small diameter piping, typically in the range of 5-20
c m [2-8 in.], indicates that typical velocities of 1-1.5 m/s [3-5 ft/sec] are acceptable for water injection
s y s t e m s / 4 ' 0 ' and at least one author advocates a range of 0.6-2.1 m/s [2-7 f t / s e c ] / * - "

The advent of very large seawater injection systems in the Middle East and elsewhere has made
it very difficult to economically design an injection system within these velocity limits, and velocities
in the range of 2.1-3.6 m/s [7-12 ft/sec] are c o m m o n . ^ " ' 1

In most systems, the primary problem is that the injection lines are oversized in certain parts of
the system, resulting in very low velocities. A m i n i m u m velocity of 1 m/s [3 ft/sec] is recommended
for water injection lines.

Velocities high enough to cause erosion problems occur m u c h less frequently.


Velocities are typically low in vessels, making solids accumulation a virtual certainty. This inevi-
tably results in increased corrosion rates and bacterial activity beneath the solids. To make matters
worse, chemicals such as corrosion inhibitors and biocides have limited effectiveness because the accu-
mulated solids prevent them from reaching the metal surface.

All vessels should be designed to facilitate solids removal. For example, cone-bottomed tanks
equipped with an internal jetting system and center drain permit regular solids removal without a sys-
tem shut-down or vessel entry.

High-Low Velocity Areas

Systems may contain adjacent areas which are exposed to different water velocities. If no oxy-
gen is present, the high-velocity area is anodic to the low velocity area and corrodes. If oxygen is
present, the low-velocity area receives less oxygen, acts as the anode, and corrodes.

Increased velocity in an oxygenated system can cause an initial increase in corrosion rate (sup-
plies more oxygen), then the reaction slows d o w n as the velocity is further increased due to formation
of Fe(OH) 3 on the metal surface. A further increase may rip the protective film off the surface, again
leading to a higher corrosion rate.
Corrosion may be uniform in nature, resulting in u n i f o r m thinning, or it m a y be localized in the
form of discreet pits or larger localized areas. Simply put, uniform corrosion is w h e r e the corrosion
takes place evenly over the whole metal surface. Normally it is accounted for during the design stage
in the f o r m of a corrosion allowance. An additional thickness of metal is included to allow for the an-
ticipated uniform surface corrosion over the expected life of the structure. Hence, system failures do


not normally occur as a result of u n i f o r m corrosion, even though it accounts for the greatest loss of
metal on a tonnage basis.

Localized corrosion is usually the most disastrous and the toughest to detect. It typically gives
some f o r m of pitting and the pitting corrosion rates can be 10-50 times the general coirosion rate.
Hence, holes in the metal can appear within times m u c h shorter than the system's design life.

Nearly all oil and gas system corrosion failures are a result of localized attack, which taKes sev-
eral different forms and is caused by a variety of different situations.

Metals have a granular structure, more easily seen under a microscope. In circumstances where
the grain boundaries b e c o m e anodic with respect to the grains themselves, the corrosion reaction will
preferentially occur along the lines of those boundaries. This "inter-granular" corrosion will give a
line of w e a k n e s s that affects the strength or integrity of the metal. By comparison, trans-granular cor-
rosion occurs across the b o d y of the metal grains.

Galvanic Corrosion
When two different metals are placed in contact in an electrolyte, the coirosion rate of the more
reactive metal will usually increase and the corrosion rate of the less reactive metal will decrease. This
f o r m of corrosion is referred to as bimetallic or galvanic coirosion.

In a galvanic couple, the more active metal is referred to as the anode and the less reactive metal
is referred to as the cathode. O n e of the most important factors in galvanic coirosion is the ratio of the
exposed area of the a n o d e to the exposed area of the cathode. This is referred to as the "area effect."
T h e highest coirosion rate occurs when there is a large cathode connected to a small anode. The cor-
rosion rate also depends on the size of the potential difference between the two metals.

If there is a marked tendency of one metal to corrode in preference to another, such as iron and
copper exposed in aerated salt water, the less resistant metal suffers the entire coirosion. Thus, steel
rivets in Monel or copper sheet corrode very rapidly; Monel rivets in steel plate cause little damage

Accelerated corrosion due to galvanic effects is usually greatest near the junction of the two met-

Mill scale is less reactive than steel, so if pipe is covered with mill scale, which is subsequently
partially removed or knocked off, the mill scale will act as the cathode and the bare steel will act as
the anode and corrode.

A similar p h e n o m e n o n often occurs when n e w pipe is connected to old pipe, especially in exter-
nal coirosion. T h e new pipe acts as the anode and corrodes preferentially.

This principle is utilized in a beneficial w a y in cathodic protection. Steel is connected to a more

reactive metal, such as magnesium, and is " p r o t e c t e d " (does not coirode). The steel has b e c o m e a
cathode and the more reactive metal the anode. Cathodic protection is discussed in more detail later in
this Chapter.

Preferential Weld Corrosion

Galvanic corrosion frequently occurs in or adjacent to welds. Accelerated corrosion has been ob-
served both in the weld metal and in the heat affected z o n e ( H A Z ) adjacent to the weld. M i n o r differ-
ences in chemistry between the weld material and the line pipe are considered the primary cause of this
p r o b l e m / 4 D e p e n d i n g upon the weld metal constitution, the H A Z can b e c o m e cathodic to the weld.



In such cases the corrosion takes place in the weld itself. Improper po.st-weld heat treatment can also
be responsible for accelerated corrosion in the HAZ. T h e d a m a g e often appears as a corroded
groove/channel all the way around the internal circumference of the pipe.

Concentration Cells
Localized corrosion can also be created by differences in electrolyte composition at two adjacent
points on the metal surface. The main forms of concentration cells are n o w discussed.

Under-Deposit (Scales and Sludges) Corrosion

The deposition of any solid on a metal surface which is not sufficiently tight and non-porous to
completely protect the metal surface can cause increased coiTosion under the deposit due to differential
concentration. Even tight, adherent scales can create problems if they form only in spots rather than
uniformly over the metal surface. Sulfate reducing bacteria thrive under scales and sludges, creating
H 2 S and causing localized pitting. A m a j o r p r o b l e m is where the deposit is only partial. T h e junction
of the deposit and the metal surface is effectively a crevice and the resultant pitting attack resembles
that of crevice corrosion (see Figure 4.11).

Figure 4.11 Under-Deposit or Crevice Corrosion

Crevice Corrosion
Crevices promote the formation of concentration cells. This is especially serious in oxygenated
systems where the oxygen in the crevice may be c o n s u m e d more rapidly than fresh oxygen can diffuse
into the crevice. This causes the pH in the crevice to decrease, resulting in a more acidic environment,
which accelerates corrosion.


Air-Water Interface
The water at the surface contains more oxygen than the water slightly below the surface. The
difference in concentrations will cause preferential attack at the water line.

Oxygen Tubercles
This is a f o r m of pitting which results f r o m the same type of m e c h a n i s m as crevice corrosion. It
is encouraged by the formation of a porous layer of iron oxide or hydroxide which partially shields the
steel surface.

Erosion-Corrosion and Impingement

T h e formation of a layer of corrosion product on the surface of a con-oding metal usually results
in decreased coirosion rates. The corrosion product acts as a rather poor coating and partially protects
the surface.

Removal of the coirosion product by erosion due to high velocities, turbulence, or the abrasive
action of suspended solids can result in increased corrosion rates by continually exposing fresh metal
to the electrolyte. This f o r m of corrosion is called erosion-corrosion and is a c o m m o n cause of failure.
Carbon steels are particularly susceptible to this f o r m of attack when the corrosion products which
form are soft and easily removed, such as iron carbonate.

Erosion-corrosion normally occurs in localized areas such as b e n d s or elbows. However, it can

be fairly widespread in a system if significant amounts of suspended solids are present.

Several equations h a v e been proposed for the calculation of the minimum velocity for the onset
of e r o s i o n - c o r r o s i o n / 4 : "' ) Corrosion rates are generally m u c h more severe w h e n flow rates are in ex-
cess of the erosional velocity.

A p h e n o m e n o n similar to erosion-corrosion, but even more localized is k n o w n as impingement.

This occurs when a stream impinges upon a metal's surface and breaks down protective films at very
small areas, resulting in pits or holes at the points of impingement.

Cavitation is caused by the repeated formation and subsequent collapsc of vapor bubbles in a liq-
uid on a metal surface. Cavitation damage is a frequent problem in centrifugal p u m p impellers, the
fluid ends of plunger p u m p s and downstream of throttling valves.

In order for cavitation to occur, the pressure must first drop sufficiently to allow the water to va-
porize. The pressure must then increase sufficiently to collapse the bubbles of water vapor. Repeating
this process at high speed, such as in the case of a pump, vapor bubbles form and collapse rapidly.
Calculations have shown that rapidly collapsing v a p o r bubbles produce shock waves with pressures as
high as 415 000 kPa [60 000 psi]. Forces this high can cause damage in two ways:
Any protective corrosion product films are destroyed at the point of b u b b l e collapse result-
ing in increased corrosion.
The force is sufficient to physically tear metal particles f r o m the surface.
Thus, in a corrosive fluid cavitation d a m a g e is the result of both corrosion and mechanical ef-



' Entrance
^ Friction Increasing 1
V Pressure
Dim to

Friction, Entrance
Loss at Vane Tips
Point of Low 3st Pressure ^
Where Vaporization Starts

Points Along Liquid Path

Figure 4.12 Relative Pressures in the Entrance Section of a Centrifugal Pump

Figure 4.12 shows a pressure profile in a centrifugal pump. If the m i n i m u m pressure is low
enough to permit the formation of vapor bubbles and the impeller creates sufficient pressure to col-
lapse them, cavitation will occur.

Stated differently, the available net positive suction head ( N P S H ) at the suction flange of the
pump must be greater than the N P S H required by the pump to prevent any vaporization of the liquid as
it flows from the p u m p suction flange into the impeller inlet vanes.

T h e available N P S H can be calculated as outlined by Abbott/*--^ V l a m i n g ' * - ^ presents a

method for estimating the N P S H required f o r centrifugal pumps, and Miller'*- ' discusses N P S H ef-
fects on reciprocating pumps.

Cavitation can also occur immediately d o w n s t r e a m of control valves. W h e n fluid passes through
a restriction, it accelerates and the pressure drops as shown in Figure 4 . 1 3 / * - w It is possible for the

Figure 4.13 Pressure and Velocity Profiles


pressure at the vena contracta (point of m i n i m u m fluid stream cross section) to drop below the vapor
pressure of the liquid, allowing vapor bubbles to form. O n c e past the vena contracta the liquid decel-
erates, and if the downstream pressure increases to exceed the v a p o r pressure, the bubbles will col-
lapse, and cavitation will occur.

This type of problem is a c o m m o n occurrence in water injection systems where a portion of the
output f r o m high pressure injection p u m p s is bypassed through a control valve back to an atmospheric
suction tank.

"Anti-cavitation'" valves can be employed to prevent this p r o b l e m / 4 - ^ Valves of this type re-
duce the pressure in a series of staged pressure drops. This technique maintains the flowing pressure
above the vapor pressure of the liquid, and vapor bubbles cannot f o r m , thus preventing cavitation.

Hydrogen Damage
H y d r o g e n atoms are formed at cathodic areas on a steel surface during coirosion by acid gases as
previously mentioned. In sweet, air-free systems, most of the hydrogen atoms combine in a very fast
reaction on the metal surface to form hydrogen gas molecules which subsequently evolve f r o m the
metal surface into the water. A small amount of hydrogen atoms may enter the metal lattice, but the
main cathode reactions are:

H+ + e H and H + H" -> H,

(very fast)

In sour systems, sulfide ions slow down the rate at which hydrogen atoms combine to f o r m hy-
drogen gas molecules. This results in an increased n u m b e r of hydrogen atoms on the metal surface,
which provides the driving force for diffusion of hydrogen atoms into the steel. Hydrogen atoms are
extremely small and can diffuse through most metals.

There are three types of damage which can be caused by the diffusion of atomic hydrogen into
steel: Hydrogen Blistering, Sulfide Stress Cracking (SSC) and H y d r o g e n Induced Cracking (HIC).

Hydrogen Blistering
During sour corrosion, atomic hydrogen diffusing through steel can combine to f o r m molecular
hydrogen within structural defects such as voids, laminations, microcracks, or discontinuities around
i n c l u s i o n s / 4 " ^ Once molecular hydrogen has formed within a defect, it is trapped, because hydrogen
gas molecules are too large to diffuse out through the metal lattice. As coirosion continues, hydrogen

H H+
Fe" J- i
H - Hz H

H -> H2 H

Figure 4.14 Illustration of the Mechanism of Hydrogen B l i s t e r i n g ^



atoms continue to diffuse into the defect creating additional hydrogen molecules, thus increasing the
pressure in the defect. Increased pressure causes growth and further separation of the flaw.

This process is illustrated in Figure 4.14. Corrosion is occurring on the upper surface of the
metal, and three things are happening to the hydrogen atoms:
1. A portion are combining to f o r m hydrogen gas molecules on the corroding surface.
2. S o m e atoms d i f f u s e completely through the metal to combine on the external surface.
3. Atoms diffuse into the void where they form hydrogen gas. As the pressure builds up it
will eventually d e f o r m the metal and cause a blister in "soft" steels.
Blistering is usually confined to the lower-strength grades of steel (70 000 to 80 000 psi tensile
strength and a hardness of Rc 22 or less - Rockwell Hardness scale). Hardened or high-strength steels
will normally crack rather than blister.

Hydrogen blistering is primarily a function of the steel "cleanliness," which is related to the im-
purities in the steel and the method of manufacture.

Corroding Surface
(Sour Corrosion)

Hydrogen Blisters

Figure 4.15 Schematic Representation of Hydrogen Induced Cracking

Hydrogen Induced Cracking (HIC)

Sometimes called "stepwise cracking" or "blister cracking," hydrogen induced cracking is a spe-
cialized case of hydrogen blistering which can occur during sour coiTOsion/*- 5 ^ HIC can occur w h e n
a steel contains n u m e r o u s elongated defects which are parallel to the metal surface. Hydrogen can
gather along the defects and create miniature blisters or c r a c k s / " 0 2 ' Cracks on one plane tend to link
up with cracks on adjacent planes to form " s t e p s " across the thickness as shown in Figure 4.15. T h e
cracks can reduce the effective wall thickness until the pipe is over-stressed and ruptures. Cracking is
sometimes accompanied by surface blistering.

H I C is the most c o m m o n f o r m of pipeline cracking due to H^S environments and it is potentially

more damaging than hydrogen blistering.

W h e n residual or applied tensile stress exists in the metal, the resulting cracks align themselves
perpendicular to the stress. This p h e n o m e n o n is called "stress orientated hydrogen induced cracking
(SOHIC)." This cracking tends to be intergranular, starting where the surface is subjected to the
highest tensile stress.


H y d r o g e n stress cracking may also occur due to the hydrogen generated by cathodic protection,
where H atoms absorb into hard spots of a pipe or vessel. Susceptibility to this type of damage is a
function of the steel's hardness.

Sulfide Stress Cracking (SSC)

Hydrogen embrittlement is the brittle failure of metallic materials at stress levels b e l o w their
yield strength as a result of exposure to atomic hydrogen. Sulfide stress cracking ( S S C ) is a particular
form of hydrogen embrittlement which occurs because of the entry of atomic hydrogen into steel as a
result of sour coirosion. The exact mechanism by which the hydrogen causes a loss of ductility is a
matter of technical debate. <4-33>

In order for sulfide stress cracking to occur, all of the following are required:
1. H y d r o g e n sulfide
2. Water (liquid); even a trace of moisture is sufficient.
3. A "high strength" steel. T h e exact strength level varies with the composition and
microstructure of the steel. (Although other materials are susceptible to sulfide cracking,
we will confine this discussion to steel.)
4. The steel must be under tensile stress or loading. The stress m a y be residual or applied.

If all of these conditions are present, sulfide cracking m a y occur after some period of time. It is
important to realize that sulfide cracking usually does not occur immediately after exposure to the sour
environment but m a y take place after hours, days, or years of s e r v i c e / 4 3 4 > T h e susceptibility of a ma-
terial to failure by this m e c h a n i s m is primarily determined by the variables discussed in the following

Yield Strength or Hardness

" P l a i n " carbon steels with yield strengths b e l o w 620 000-690 000 kPa [90 000-100 000 psi] are
generally considered to be i m m u n e to sulfide cracking. This corresponds to a hardness of about Rc 22.
Strengths above this level are susceptible to cracking. The higher the strength, the shorter the time to
failure. This is illustrated in Figure 4.16 which shows a correlation of failure time with hardness and
applied stress for carbon steel exposed to 3000 p p m H 2 S in water containing 5% NaCl.

If steel is alloyed with other materials, such as nickel, failure can occur at hardness levels less
than Rc 22. Conversely, certain heat treatments can raise the m a x i m u m permissible hardness level
a b o v e this value.

Stress Level

The stress may be either residual or applied. T h e time to failure decreases as the stress level in-
creases (Figure 4.16). In most cases the stress results f r o m a tensile load or f r o m the application of
pressure, or both. However, residual stresses and hard spots can be created by welding, or by
cold-working the material (cold bending, wrench marks, etc.).

Hydrogen Sulfide Concentration

As the H 2 S concentration increases, the time to failure decreases. Figure 4.17 shows a correla-
tion between failure time and hardness for carbon steel in 5% NaCl solution containing various con-
centrations of H 2 S. Delayed failures can occur at very low concentrations of H 2 S in water (0.1 ppm)
and at partial pressures as low as 0.1 kPa [0.001 atm] although the time to failure b e c o m e s very






Day Week Month
_i i i 1111 I i 1111
1 I I I I II 11
.5 1 5 10 50 100 500 1000
Time to Failure, Hours

Figure 4.16 Failure Time as a Function of Hardness and Applied Stress^-35-'

40 I | I I IT'



15 - '60,000.
Stress Level -130% YD
ay Week Month ~Yea
10 i i linil I i ml i i i ..,.i i . i . 1 1
.5 1 5 10 50 100 500 1000 5000 10,000
Time to Failure, Hours

Figure 4.17 Correlation of Failure Time with Hardness and H 2 S Concentration^ 4 - 35 '

pH of Solution

The cracking tendency of carbon steel increases as the pH decreases in H 2 S containing brines as
shown in Figure 4.18. Failures can be drastically reduced if the pH of the solution is maintained
above 9.0.


|llll| I 1 I | I I I l| 1 1 1 II1II1 1 1 1 1 II1II 1 1 1 1 1111

11 1330 Hr 0-
10 O

8 O ^ O o

6 o


4 o

3 o o
2 -y} o 1700-1900 ppm Total Sulfide in 5% NaCI

Ail Rings Rc 33 1 Stressed to 115% YD ~~
0 1 1 1 1 11 III 1 I i 1 I ml i i i 1 i ml i i i 1 i in
0.04 0.1 0.5 1 5 10 50 100 500 1000
Time to Failure, Hours

<4 35
Figure 4.18 Correlation of Failure Time with pH - '


The cracking susceptibility decreases as the temperature increases. This is illustrated for com-
mon grades of tubing and casing in Figure 4.19. Materials for application in environments containing
EhS should be selected according to N A C E Standard M R . 0 1 7 5 ^ 5 ' (latest revision), which is discussed
later in this chapter.

Yield Strength, kpsi

40 60 80 100 120 140 160 180
I*- N-80| SOO-
J-55 ->l L-80 K - r k - P-110"-W-140-*I
-IC-75 l- I*s00-->I<-V-I50 -
150 SOO-95 -(25 ^ _ 300
/ V-150

SOO-125 SOQ-140

a 200 ffl
a CL
a> 0}
C-75 N-80
SOO-95 100
N.N & T , / * *
A MOD N-80
J-55 *
300 600 900 1200
Yield Strength, MPa

Figure 4.19 Effect of Temperature on Sulfide Stress Cracking Resistance** 3 ^



Some Key Points Regarding SSC:

The higher the H?S concentration, the shorter the time to failure
No water > no corrosion > no hydrogen d a m a g e
Steel below Rc 22 is considered i m m u n e to SSC
Mild steel cracking tendency increases as solution pH decreases
Cracking tendency decreases as the temperature increases.

Stress Corrosion Cracking (SCC)

S C C is the brittle failure of an otherwise ductile metal, due to the synergistic action of a specific
corrosion environment and tensile stress (applied or residual). The combined effect of the two is
greater than the sum of the two single effects and it only occurs in alloys. Examples include:

cracking of austenitic stainless steels such as 304 and 316, in the presence of chloride ions
(in seawater at 66C [150F]). Such cracking can occur in wet sour gas, w h e r e seemingly
insignificant amounts of chloride ions are present and water can c o n d e n s e / 4 - ' 7 ' It appears
that synergistic action exists between the chloride and frbS.
carbon steel in strong caustic ( N a O H ) solutions
copper alloys in ammonia containing environments.

Typically there is little metal loss and the cracking often occurs without any other signs of corro-
sion damage. As a result there is no forewarning of failure.

SCC can be prevented by removing either the stress or the corrosive environment.

Corrosion Fatigue

Metal fatigue is the cracking failure which occurs as a result of repeated cyclic stressing of a
piece of metal. Corrosion fatigue is a similar cracking failure, but it occurs in a much shorter time due
to the presence of a corrosive environment (and the cyclic stressing). M o s t sucker rod breaks and drill
pipe failures occur as a result of coirosion fatigue. It can also be the cause of p u m p failures. It is pos-
sible on any part that is subjected to a cyclic stress in a corrosive environment. It is a brittle failure
which can occur at stress levels well below the yield strength/* 3 S >

The failure of sucker rods due to corrosion fatigue typically shows four areas of damage. A cor-
rosion pit on the s u r f a c e of the rod initiates the failure. Due to compression on the down-stroke of the
rod and tension on the up-stroke, over many cycles the pit develops into a crack that grows. T h e h a m -
mering together of the two sides of the crack during this cyclic operation, result in a very smooth
half-moon shaped surface. Eventually the growth rate of the crack becomes so rapid that this 'smooth-
ing' effect has not taken place and the surface of this part of the rod cross-section appears grainy, or
crystallized. Finally there is so little metal holding the rod in one piece that the tensile strength of that
section is exceeded and an elastic stretching and shearing occurs (See Figure 4.20).

N o t e : T h e various forms of corrosion cracking d a m a g e mentioned above, that is, hydrogen

embrittlement (HE), blistering, HIC, S O H I C , SSC, S C C and coirosion fatigue, are n o w k n o w n collec-
tively as environmental cracking. It is important to remember that the overall amount of metal loss
(due to corrosion) may be quite small w h e n these types of failure occur.


Crack Initial" Smooth Area

Fibrous Area

Half Moon
(Line between
smooth and Stretched and Torn
fibrous areas) Steel at Final Fracture

Figure 4.20 Sucker Rod Failure Due to Corrosion Fatigue

Microbiologically Influenced Corrosion (MIC)

Corrosion can be caused by bacterial activity. Internal damage can result f r o m bacterial activity
in injection systems. External d a m a g e of buried structures can result f r o m bacterial activity in soils.
This subject is discussed in Chapter 5.

Fretting Corrosion

Fretting corrosion occurs when t w o metal surfaces are in contact with each other in a corrosive
environment and are subjected to vibration or other relative motion. Accelerated attack results from
the continuous removal of any protective films f r o m the metal surfaces. It occurs mainly on machine
parts such as ball and roller bearings, shafts and gears.

Electrolytic Corrosion

Electrolytic corrosion is quite similar to galvanic corrosion except that the direct current causing
the corrosion originates f r o m an external source. A typical example of electrolytic corrosion is the cor-
rosion of foreign structures by stray currents f r o m impressed current cathodic protection systems.

Dealloying / Selective Leaching

Selective leaching is the removal of one element f r o m a solid alloy by corrosion processes. Two
c o m m o n l y encountered examples are the selective removal of zinc in brass alloys (dezincification) and
the selective removal of iron f r o m gray cast iron (graphitization). These reactions m a y be considered
to be f o r m s of galvanic corrosion.

This process typically occurs in soft water with a high C 0 2 c o n t e n t / * ^ Other factors that en-
hance this f o r m of coirosion include high temperature and chloride concentration, low water velocity
and the presence of crevices or deposits. If it takes place over the entire surface, as is typical in high
zinc content brasses, the dealloying is termed uniform, or layer dezincification.



The c o m m o n l y accepted mechanism consists of three steps: (1) the brass dissolves, (2) the zinc
ions stay in solution, and (3) the copper plates out of solution and is redeposited as a spongy, porous
replica of the original part. Examples of this type of corrosion have been observed in Admiralty brass
exchanger tubes.

Gray cast irons are alloys of iron, carbon (2-4%) and silica (1-3%). T h e carbon exists as flakes
of graphite within the metal. T h e graphite gives the metal a gray color on a fractured surface, hence
the name gray cast iron.

W h e n graphitization occurs, the iron matrix is selectively dissolved, leaving the graphite network
behind. Dimensional changes do not occur, but the cast iron loses strength and its mechanical proper-
ties, resulting in premature failure. It may occur in salt waters, acids and w h e n sulfate reducing bacte-
ria are a c t i v e / * 5 ^

Corrosion Due to Mercury

In most production systems the presence of mercury w a s not previously measured. However,
specific corrosion problems were encountered in a l u m i n u m heat exchangers used within gas and gas
condensate production systems. Analysis revealed that mercury is produced as a very small fraction of
the total v o l u m e of gas, so m u c h so that it is measured in micro-grams per cubic meter (pg/m 3 ).
[Note: 1 p g / m 3 is 1 part of Hg in about 10 1 0 parts of gas]. Some typical levels of mercury concentra-
tions in produced gas include:

< 0 . 0 1 p g / m 3 in the US and A f r i c a

0.01 to 180 p g / m 3 in Groningen, Netherlands
1 to 9 p g / m 3 in the Middle East
69 to 119 | i g / m 3 in S. America
Up to 4 5 0 0 p g / m 3 in N. G e r m a n y

Initially the mercury reacts with the a l u m i n u m to f o r m an amalgam:

Hg + Al -> Hg(Al)
In the presence of water the overall process of a m a l g a m corrosion will occur, where the amal-
gam goes on to react with water;

2Hg(Al) + 6H20-> Al202 -3H20 + 3H2 +2 Hg

W h e n water is not present the process of liquid metal embrittlement ( L M E ) occurs. This can
give rise to rapid brittle failure of the aluminum structure. The mercury diffuses to grain boundaries in
the aluminum body and f o r m s a liquid a m a l g a m (as in the first equation above). Cracks then initiate
and propagate via the grain boundary. This reaction is also possible in Al alloys, Cu alloys and s o m e


Approaches to Corrosion Control

There are many methods of controlling or preventing corrosion in water handling systems. They
can be classified into six general categories:


Chemical Inhibitors
Chemical inhibitors can be viewed as a "coating," since most corrosion inhibitors used in oil and
gas production systems are organic film f o r m i n g materials.

Isolate the Metal from the Electrolyte

Plastic coatings and linings are very popular in tubular g o o d s and tanks. Cement linings have
frequently been used in tubular goods.

Cathodic Protection
The application of sufficient electrical current to a metal immersed in water will stop corrosion.
Current can be applied with sacrificial anodes or with impressed current if a source of p o w e r is avail-

Alter Material Composition

A corrosion resistant alloy steel or a non-ferrous metal m a y be substituted for carbon steel. The
general t e r m for these materials is corrosion resistant alloy ( C R A ) . Corrosion resistant metals typically
rely on a passive oxide protective f \ \ m ( 4 4 0 ) resulting f r o m the addition of mainly chromium and nickel
to the alloy fonnulation. These alloys are m o r e expensive than carbon steel. It is often necessary to
sacrifice some mechanical properties in order to achieve the necessary coirosion r e s i s t a n c e / 4 ^ This
is particularly true when alloys are susceptible to stress corrosion.

Plastic materials m a y be used but they are usually subject to m o r e severe pressure and tempera-
ture limitations than metal.

Physical Alteration of the System

Typical examples would include increase of N P S H to prevent cavitation, increased or decreased
velocities by changing the diameter of tubular g o o d s or line pipe, and institution of regular line pig-
ging and vessel clean-out to reduce solids deposition. Examples of other system alterations, which
might be considered at the design stage or for during a system workover, include:
ensuring that there are no crevices as a result of the construction methods
minimizing dead-end pipe legs
minimizing stagnant areas in vessels and pipelines
installing large radius bends in piping, rather than sharper angles which lead to impinge-

Alter Electrolyte Composition

The water composition m a y be altered by changing the p H , chemically or mechanically r e m o v i n g
dissolved gases, or by mixing with another water. Other techniques are used, but these are most com-
mon in oilfield operations.

Method Selection
The selection of the method or combination of methods to be used in a particular application
should be based on minimizing the cost of achieving the following objectives:
1. Obtain an acceptable service life for the equipment.



2. Prevent pollution of the environment.

3. Minimize the generation of suspended solids. This is especially important in water injec-
tion systems. Injection-wellbore plugging can be expensive.

Cost of Materials
The initial cost of materials, plus the cost of any additional materials required during the project

Labor Costs
Initial installation costs plus maintenance costs must be considered.

The Present Value of Money

Remember that the cost of expenditures to be made in future years must be discounted back to
their present value to obtain the "true cost." Deferred dollars do cost less; however, the "put it off un-
til tomorrow" philosophy can be dangerous. Both material and labor costs are rising rapidly and even
deferred dollars can add up to a pile of cash. The tendency to defer investment has to be balanced
against the fact that it is often far cheaper to go ahead and spend the money on a good system in the
beginning. If corrosion gets out of control, the remedy can be extremely expensive. On the other
hand, it may be cheaper to let the system corrode and replace it when the time comes. This is seldom
the case, however.
The severity and nature of any corrosion which may occur must be estimated during the design
of the system so that adequate control measures can be instituted at the beginning. The system must
then be monitored to measure the corrosiveness of the system so that the corrosion control program
can be adjusted as necessary. The measurement of corrosion rates is extremely critical and will be
covered later in this chapter.

Corrosion inhibitors are selected for water systems in a variety of ways. Some of the more popu-
lar are:
Wild guess
Experience in the area
Advice from chemical salesman
Laboratory tests
Field tests
Usually a combination of these techniques is employed. The following steps arc recommended
when an inhibitor must be selected:
1. Find out what is causing the corrosion. This is actually the first step in selecting any con-
trol method, and it is very important in selecting an inhibitor.
Determining the cause may involve determining the amount and kinds of dissolved gases
present, or analyzing the corrosion product. The type of attack also must be determined.
2. Make an initial inhibitor selection with the help of technical personnel in your own com-
pany and in consultation with the chemical company of your choice. Major chemical com-


parries maintain large technical staffs and can give you excellent advice on which of their
products stands the best chance of working once you have defined the problem.
Initial chemical selection can also be made using laboratory or field tests. Good quality
laboratory testing is difficult at best because of the difficulty of transporting fluids to the
lab and running the tests without contaminating or changing the water in some way. How-
ever, comparative data can be generated using standard test waters.
Both linear polarization and potentiodynamic polarization^ 4I-4-42> methods, which are dis-
cussed later in this chapter, are extremely useful in selecting corrosion inhibitors for water
injection systems. A portable side-stream apparatus can be used for initial chemical
screening, as shown in Figure 4.21.

Figure 4.21 Side S t r e a m Testing S e t u p

Water is flowed directly from the system through the test cell. Inhibitor is injected into the
side-stream and the corrosivity of the inhibited water is monitored with one or both of the
The advantages of side-stream testing are many:
Chemical evaluation is made on-site using uncontaminated water directly from the
injection system.
Testing can be carried out quickly. Several inhibitors can be evaluated in a few
Linear polarization meters measure the instantaneous general corrosion rate.
Changes in corrosion rate can be observed as they occur.
A portable potentiodynamic analyzer is the only monitoring tool available which
permits rapid measurement of both the general corrosion rate and the pitting corro-
sion rate.
3. Once the initial chemical selection is made, field testing should then be carried out with
diligent monitoring to determine how good a job you are doing. You may want to test one
chemical or concentration in one part of the system and another in a different part of the
Once your final selection of inhibitor and concentration is made, monitoring must be con-
tinued. Systems can change in corrosivity with time.



Guidelines for Inhibitor Selection

Inhibitor Solubility
Water-soluble inhibitors are normally used in water systems/**" Water dispersible inhibitors
may also be used and often afford better protection than the truly soluble inhibitors. However, there is
usually concern over the possibility of "formation plugging" by inhibitors that are not completely solu-

Dissolved Oxygen
Normal" organic corrosion inhibitors do not effectively inhibit corrosion caused by dissolved
oxygen. There are some corrosion inhibitors on the market which are somewhat effective against oxy-
gen. However, because of their chemical nature these tend to be both expensive and have an unpleas-
ant smell.

If other chemicals are being injected into the system (scale control, oxygen scavenger or
bactericide), compatibility of the different chemicals should be checked. If they react with each other,
their effectiveness may be reduced or destroyed.

Inhibitor Detergency
Most corrosion inhibitors possess "detergency" to help keep the system clean. This is desirable
since the inhibitor must be able to reach the metal surface to do its job. If sludges accumulate on the
surface, they can interfere with the inhibitor, so a certain amount of detergency is desirable to maintain
a clean system. Problems can arise when inhibitor injection is first initiated into a system if the inhibi-
tor has too much detergency. Large quantities of solids can be loosened and cause system or fonnation
It is advisable to inspect surface facilities prior to initiating injection. If they are extremely dirty,
it may be advisable to clean the system prior to starting inhibition. Otherwise, it is advisable to con-
sider installation of cartridge filters at appropriate points to keep dislodged solids out of the fonnation.
It is an operational headache until the system cleans up, since the cartridges must be changed fre-
quently. However, it may be worth it in the long run if formation damage can be avoided.
The detergency property of some inhibitor formulations can lead to foaming problems. Hence,
selected inhibitors must be checked to ensure that they do not have an adverse foaming tendency in the

Inhibitor Cost
Inhibitor cost is listed last to emphasize a very critical point. Never make an inhibitor selection
based on which inhibitor costs the fewest dollars per gallon. Sometimes the most expensive inhibitor
($/gal) will be the most effective (take fewer gallons) with the net result that the total cost of inhibition
is lower than if a cheaper chemical is used.
Obviously, the most expensive inhibitor is not always the best. Inhibitor effectiveness can only be
established by evaluation. The primary point to be made is that the "cheapest" inhibition program is
often not with the least expensive inhibitor.


Summary of Corrosion Inhibitor (CI) Requirements

The optimum selection should be:
The most cost effective CI (compared to other inhibitors)
An inhibitor which readily transports to the system surfaces which need protection (includ-
ing adequate detergency to penetrate deposits)
That which provides an acceptable reduction in coirosion rate
The product which has no adverse properties re foaming or emulsion forming tendencies
Compatible with other chemicals in the system
Cost effective compared to other available corrosion prevention methods (e.g. linings or

Inhibitor Application
Organic inhibitors are supplied in liquid form and are injected with chemical pumps. The chemi-
cal is often diluted for ease of application. In cold climates the chemical must be winterized (usually
with alcohol) or kept in a heated building. If it is to be winterized, the chemical must be compatible
with alcohol. Alcohol will cause some inhibitors to precipitate from solution.
Inhibitors are injected into water systems on a continuous basis. A good inhibitor should be ef-
fective at a concentration of 5-25 ppm. As a "rule-of-thumb" the erosion velocity for stripping inhibi-
tor film from surfaces is commonly taken as about 9 m/s [-30 ft/sec]. Such velocities are typically
only encountered in gas systems.


Since the technology of coatings and linings is both wide and complex, this book provides only a
review of the key issues.
Coatings and linings form a relatively thin barrier between the metal surface and the corrosive
environment. To be effective they must be water resistant and form a continuous layer, free of
holes/*4*' Unfortunately the perfect coating does not exist and those that come close to perfection
tend to be expensive. Most coatings will have some holes or imperfections (called holidays), even if it
is a small number.
An effective coating needs to be well bonded to the metal substrate, have high electrical resis-
tance and resist water transfer. These aspects are considered in deciding which coating components
will be included within the final formulation. These components and their main properties may be
summarized as shown in the widely used reference by Munger and Vincent/*44'
bond strongly to the substrate
form a strong layer
resist corrosion and other chemicals
create a strong bond with the intermediate coat
have the appropriate flexibility to cover the whole surface and withstand expansion
and contraction of the substrate



Intermediate (Body) Coats

provide thickness (body) to the coating
exhibit good chemical resistance
resist transfer of water vapor or liquid
increase the electrical resistance of the coating
form strong bonds with both the primer and topcoat
form a resistant seal for the whole coating
act as the first barrier to the environment
provide water, chemical and weather resistance as needed
produce a surface which is tough and wear-resistant
exhibit the pleasing or attractive appearance.
In order to provide effective con'osion protection to the underlying metal, the finished coating
system should exhibit the following properties/*- 4
effective electrical insulation
inhibition of moisture transfer
ease of application
resistance to holidays forming as time passes
strong adhesion to the metal surface
resistance to coating disbondment
ease of repair
maintenance of a constant electrical resistivity with time
optimum thickness and chemical resistance
UV degradation stability while handling and installing
no toxic interaction with the environment.
As can be seen, these are a combination of the individual properties of each layer in the coating.

Types of Coatings
Wax was used as a protective coating in ancient Egypt. Its modern equivalent is micro-crystal-
line wax applications/ 4 ^ The coal tar enamels (CTE) and extruded asphalt mastic coatings of the
early 20th century are now being replaced by more effective systems. Fabric reinforced petrola-
tum-coated tapes^-' were the first wrapped tape fonn of coating. These were followed by
polyethylene tapes, then more recently the fused multi-layer type systems. More recent coating devel-
opments include:
extruded poly olefin systems
fusion bonded epoxy
epoxy coal tars and urethanes
multi-layer epoxy/extruded polyolefin mixed systems
Detailed information on the types of coating materials available is given in Munger and
Vincent,'"'-44' with further practical material in Peabody/*^


Types of Plastics
There are two basic types of plastics used for coatings and linings: thermoplastics and thermo-
setting plastics.
Thermoplastics become softer when heated, but they can then be cooled and regain their original
physical properties. Of course, if they get too hot, the molecules may break down, but this is not a
problem in normal applications. Examples of thermoplastics are PVC and polyethylene.
Thermosetting plastics become harder and more brittle as they are heated. Cooling will not re-
store them to their original state. Examples of thermosetting plastics are phenolics and epoxies.

Application of Coating
In order to achieve the desired coating properties/4'4 ^ both the surface preparation and the appli-
cation of coating layers should be done under optimum environmental conditions. There are five im-
portant considerations:
air temperature
substrate surface temperature
the relative humidity
dewpoint temperature
the difference between the surface and dewpoint temperatures.
For example, most coatings will not diy properly at low temperatures and at high relative humid-
ity. Also, condensation on the surface could lead to trapped water and subsequent coating failure.
This necessitates the measurement of the dewpoint temperature before, during and after applying the
coating. A major cause (-70%) of coating failure is improper application.

Coatings are applied in two ways:

Liquid Applied Coatings
Most coatings are applied as a liquid in a series of thin coats to achieve the desired thick-
ness, and then baked.
Fusion Bonded Coatings
This type of coating is applied as a powder to a heated pipe, which bonds and cures the
coating simultaneously. A popular example is fused epoxy.
Some typical examples of coatings that have been recommended for water injection service are
given in Table 4.4.

Coating Thickness
Coatings can fail because they are too thick, as well as when they are too t h i n / 4 ^ The correct
film thickness of each coat will lead to the optimum performance and lifetime of a coating system. It
was once thought that the primer should fill the microscopic contours of a metal surface in order to
provide optimum performance. It is now known that a well-formulated and properly applied primer
will follow the contours of the surface anchor-pattern and does not necessarily need to exceed the
depth of that profile/4 45 ' Hence, it is advisable to follow the coating manufacturer's recommended diy
film thickness (DFT). Film quality is also important and it should be uniform, smooth, wet and flow-
ing, when applied.



Internal Coatings for Water Injection Service
Thickness Max Temp
Type mils mm F C
Epoxy-polyamide 12-20 0.3-0.5 150 66
Fused Epoxy 10-20 0.25-0.5 250 121
Modified epoxy 8-15 0.2-0.4 300 149
Epoxy-phenolic 5-9 0.13-0.23 250 121
Phenolic 5-9 0.13-0.23 400 204
Nylon 10-30 0.25-0.76 225+ 107+

Plastic coatings are also classified according to thickness. Coatings with a thickness of 127-178
pm [5-9 rails] are classified as thin-film, while coatings in excess of 178 pm [9 mils] are usually called
thick-film coatings.

Thick-film coatings are typically recommended for water injection service, although thin-film
coatings are also in common use. The reason is that it is much easier to achieve a 100% holiday-free
coating with the thick-film materials and most vendors will guarantee thick-film coatings to be holi-

In contrast, the specification for standard thin-film coatings quoted by most vendors is that a
minimum of 85-90% of the lengths in a given order (depending on range and diameter) will be holi-
day-free. The remaining lengths can contain 5-9 holidays each, again depending on pipe size.

Holiday-free thin-film coatings can be obtained at a premium price. However, the extra cost is a
small price to pay for the increased quality and can be justified in most cases.

Example Coatings
The most popular primer for many years has been inorganic zinc (IOZ). When applied at 64-89
pm (2.5 to 3.5 mils) dry film thickness (DFT), it protects steel better than other air-dry protective
primers. A high-build polyamide epoxy can be used at 102-152 pm (4-6 mils) as topcoat, which adds
chemical resistance. This coat reduces the sacrificial action of the zinc, but extends its long-term func-
tion of preventing undercutting. However, in exterior service the epoxies tend to chalk and fade fairly
rapidly due to sunlight. The choice is to add refresher coats every few years, or use a third coat of
urethane acrylic at 38-64 pm (1.5-2.5 mils) DFT. This urethane outer coat gives outstanding gloss
and color retention, as well as chemical and abrasion resistance. They have low chalking rates and
prevent chalking of the underlying epoxy. This combination coating can last 10-15 years before a re-
fresher topcoat is needed/*^
Nowadays, especially with offshore pipelines and structures, multilayer fusion bonded epoxy
(F3E) resins, polyethylene and polypropylene as multiple layers and in some cases combined with
FBE, are in common use.

Pipe Preparation
Pipe or tubing is usually pickled to remove mill scale and acid soluble material, then either sand
or flintblasted to white metal. This is an extremely critical step. If the surface preparation is poor,
disbonding and subsequent coating failure is likely.


Coating Inspection
The coating should be visually inspected between coats, and the final cured film thickness mea-
sured. It should then be "holiday checked." This is accomplished by applying an AC potential of
about 2500 volts across the coating with either a wet sponge or a dry spark for thick-film coatings. A
67.5 volt DC potential with a wet sponge is used for thin-film coating inspection. If there is a thin
spot in the coating, or if the coating is missing in spots, current will flow through the thin or bare spot
and an alarm will sound. The joint should be rejected.
If you are really concerned about quality control, you should hire an independent inspector to as-
sure quality control. This is not to infer that coating companies will intentionally do a poor job. Quite
to the contrary, they usually do a very good job. But if the application is really critical, it is in your
own best interest to double check.

Tubing Couplings
One of the most critical areas for coatings and linings in tubing is the pin end, around the
threads, and the exposed section of the coupling. A number of systems are available to eliminate or re-
duce holidays in this area, including various types of rubber and plastic compression rings/*49>

Pipe Connections
There are a number of systems available for joining internally coated line pipe.

Threaded Pipe
Internally coated threaded line pipe is used primarily in small systems where fittings are used
such as at injection well tie-ins/"' It is also used for tie-ins on fiberglass injection systems.

Externally Grooved Connection

V-Thread Coupling Coat

This connection can be used for low pres-
Center 10-14 Threads sure injection lines. Non-metallic seals are used
in a clamp to connect this joint. Seals that allow
the ends of the coated pipe to seal against the
rubber should be used/*JW

Coating Stops

At This Point

Figure 4.22 Coating Detail for T h r e a d e d Line Figure 4.23 Coating Detail for Grooved
Pipe Connection^- 5 0 > Connection^50;



Flanged Connections
Flanged connections can be used for high pressure tie-ins. The flanged sections should be pre-
fabricated and shop coated if possible.

Weldable Couplings
Proprietary weldable couplings such as the AMF Thru-Kote1 is in common use.

Figure 4.24 S c h e m a t i c of T h r u - K o t e Joint

The Thru-K-Ote system utilizes mechanically belled pipe and an internal metal sleeve machined
from sections of pipe. The sleeve is internally coated, and is wrapped with a heat-resistant material to
prevent damage to the coating during the welding process. O-rings prevent welding gas damage to the
epoxy sealant which is used to seal the sleeves to the bell-joint annulus and provide a smooth contour
to eliminate turbulence. A similar connection for unbelled coated line pipe is also available.

Mechanical Interference Fit Connections

Proprietary mechanical connections which eliminate the welding process are also in common use.
One basic type of joining system is the bell and pin joint such as Zap-Lok2/4
In the Zap-Lok system, shown in Figure 4.25, the ends of each joint of line pipe are specially
modified by cold working to create the joint surfaces. The bell end is expanded to a diameter slightly

Belled End Epoxy Secondary

Preparation Resilent Seal

Figure 4.25 S c h e m a t i c of Z a p - L o k Joint

Thru-Kote1 is a registered service mark of AMF Tuboscope, Inc.. Houston. TX.

Zap-Lok is a registered service mark of Pipelines Systems, Inc.. Houston. TX.


less than the outside diameter of the pipe. The pin end has a groove rolled in the circumference with
the end wedged slightly inward to allow insertion into the belled end.
Epoxy resin is applied to the bell and pin of each joint and the two ends are forced together using
a portable hydraulic press, resulting in a controlled interference fit. The epoxy serves both as a sealant
and as a lubricant to aid in making up the connection.

Field Coating of Weld Area

In pipe diameters greater than 6 inches, it is possible to weld internally-coated pipe in a normal
manner, and then abrasive blast the weld area and coat it with a fusion bonded epoxy in the field. The
abrasive blasting and coating is done by a special crawler machine/*5-7

Handling Plastic Coated Pipe

One of the many keys to successful service from internally coaled tubular goods is careful han-
dling during installation. If you have done a perfect coating job and then damage the coating during
installation, you have just assured premature failure.
Some handling recommendations:
1. Don't use hooks or metal bars inside the ends of the pipe when unloading.
2. Don't run metal rabbits.
3. Leave the thread protectors on the pipe until you are ready to stab it. Use a stabbing guide
if possible.
4. Don't nin wire line tools in coated tubing unless you absolutely have to.

Cement Linings

Cement-lined'*-^"*-^ tubing and line pipe has been widely used in injection systems. The most
widely used lining composition consists of Portland cement and sand. The pipe is lined by a centrifu-
gal casting method to an average thickness of 6.35-9.5 nun [14-3/8 in.] depending on pipe size.
The Portland cement used should contain less than 3% tricalcium aluminate so that it will be re-
sistant to attack by any sulfate ions in the water. Reaction between tricalcium aluminate and sulfate
ions causes the cement to swell and crack. ASTM Type III or API Class B cements are resistant to sul-
fate attack.

Advantages of Cement Linings

1. Low cost.
2. They can be applied to new or used tubular goods.
3. A certain amount of cracking can be tolerated without loss of con-osion protection. The
pH at the base of the crack (the pipe wall) is usually high enough to prevent appreciable
attack. However, con'osion can occur under some conditions.

Disadvantages of Cement Linings

1. Cement linings add weight and reduce the I.D. of the pipe.
2. It is not advisable to acidize through cement lined tubular goods. Acid dissolves cement.



3. It is difficult to get a good welded joint. Several methods have been used with varying de-
grees of success. Asbestos gaskets and hydraulic cemcnts have been very popular. The
gaskets have apparently been most successful.
4. Excessive pressure surging can cause extensive cracking, and even failure.

Handling Practices
1. Don't bend cement-lined pipe unless you have to. Suggested minimum bend radii are
shown in Table 4.5. For sharp bends, use cement lined fittings.
2. Be careful during loading and unloading. Don't drop the pipe, pick it up by the ends and
let it sag in the middle, or run over it with a truck. Avoid impact loads.
3. Apply the same general handling practices as suggested for plastic coated tubing.
Minimum Bend Radii for C e m e n t Lined Pipe
Pipe Diameter Minimum Bend
(in.) Radius (ft.)
2 3/8 150
3 A 220
4 VI 275
6 5/8 300
8 5/8 485
10 % 575

Plastic Liners
Plastic pipe can be inserted into steel pipe or tubing to provide internal corrosion protection.
Three basic types of lining systems are presently in common use in oilfield operations:

Grouted Liners
Plastic pipe is inserted into individual joints of tubing or line pipe, and Portland cement is
pumped into the annular space between the two pieces of pipe. Typical liners include
polyvinylchloride (PVC) and glass fiber reinforced epoxy (FRE). The PVC is recommended for appli-
cation up to about 71C [160F] while the FRE is usually selected for temperature in excess of 7lC
[160F] up to a maximum of about 121C [250F].
Longer lengths of plastic pipe are inserted into pipelines and grouted in place with Portland ce-
ment. Thermoplastics such as PVC or polyethylene in lengths of about 750 to 1500 meters
[2000-3000 ft] are commonly used. This procedure is usually employed to repair badly corroded lines.
Shorter lengths of fiberglass pipe have also been used as liners. The glass fibers reinforce the compos-
ite fiberglass/grout/steel system and increase its burst pressure/4 57'

Expanded Liners
Although several plastics have been used, liners manufactured from ultra high molecular weight,
high-density polyethylene (UHMWHDPE) plastic with much higher toughness and ductility than con-
ventional extruded polyethylenes is widely used in pipelines/ -0 The liner is inserted into the pipe-
line in lengths that normally range from 2500 to 5000 ft, and the lined sections are connected with


special flanges. The liner is then pressurized to expand the plastic liner until it compressively sets
against the internal wall of the steel line. The maximum operating temperature is said to be approxi-
mately 82C [180F],

Bonded Liners
Line pipe or tubing can be lined with polyvinylchloride (PVC) which is bonded to the pipe with
a thermoset bonding agent. The bonding agent is applied to the internal tubing surface and to the ex-
ternal surface of the PVC liner, and allowed to dry. The liner is then inserted into the pipe, heated,
and pressurized. The heat-softened plastic expands and bonds to the inner wall of the tubing. Regular
line pipe can also be lined and joined with a mechanical interference joint as shown in Figure 4.26.



Coating of Water Storage Tanks and Other Vessels

It is generally advisable to completely coat the internal surface of water storage tanks to protect
against corrosion in three areas:
1. The underside of the roof and the walls above the water line. Only a coating can protect
this area against corrosion resulting from condensation of water vapor. In water storage
tanks, corrosion in this area is frequently aggravated by oxygen entry.
2. The tank walls below the water line. Con-osion inhibitors can control attack in this area;
however, coating is recommended if the remainder of the tank is coatcd.
3. The tank bottom. Due to the frequent occurrence of deposits on tank bottoms, it is a fa-
vorite location for bacterial growth and under-deposit corrosion. Biocides or con-osion in-
hibitors are seldom effective in this situation because they are shielded from the metal
If possible, small tanks should be coated in a coating plant or at the fabrication yard. Better con-
trol can be exercised over surface preparation and coating operations, and the results are usually better
than a field coating job.

Filters should also be coated, and deaerators must be coated. Coating application should be car-
ried out in a workshop.



Field Application
If a tank must be coated in the field after it has already been in service, adequate surface prepara-
tion is the biggest obstacle to a successful coating job. All grease or oil should be removed from the
surface with solvents. The surface should then be dry sandblasted to bright metal, or machine cleaned
with wire brushes. All dust should then be removed from the surface by blowing with dry air or use
of an industrial vacuum cleaner. The first coat should then be applied immediately before the surface
rusts or is contaminated in some other way. Never finish the surface preparation and then wait until
the next morning to start the coating operation. Application in high humidity should be avoided if at
all possible.

Rigorous inspection during both the surface preparation and the actual coating job is highly rec-

Coating Selection
There are many coatings available which can be used for internal tank coatings. Although the
most widely used oilfield tank coatings have historically been the coal tar epoxies, other widely used
coatings include the epoxy, epoxy-phenolic, and flake-filled polyester coatings/4,61-4-62)
Flake-filled polyester coatings have been used very successfully for internal coating of filters,
deaerators, and other vessels handling aerated seawater.

Cathodic Protection

It is almost impossible to get a perfect coating job. There will always be a few imperfections, or
"holidays," and corrosion at these points can lead to coating failure. Therefore, it is strongly recom-
mended that cathodic protection be applied in conjunction with vessel and pipeline coatings. In the
USA it is now mandatory (by both federal and state regulations) to apply cathodic protection to under-
ground transmission pipelines, gas distribution systems and underground storage tanks for petroleum
The key aspects of cathodic protection (CP) are summarized below. Detailed discussions of CP
theory and application are available in a number of References 4.2, 4.63-4.65.

Earlier in this chapter it was pointed out that metal loss occurs at anodic areas on steel. These
are areas where current is leaving the metal surface to flow through the electrolyte to the cathodic ar-
eas. When current leaves a metal surface it takes metal with it.
1 he principle of cathodic protection is to convert the whole metal surface into a cathode. This is
accomplished by forcing sufficient current to flow to the structure to be protected so that there is no
net flow of current from any point on the metal surface.
When external current is applied to a corroding surface, part of it goes to the anodic areas and
part of it goes to the cathodic areas as shown in Figure 4.27.
The secret to success is to apply sufficient current to the structure so that ipa > ia. In other words,
there must be enough current going to the anodic areas to "cancel out" the natural corrosion current.
There must be no net current leaving the surface if corrosion is to be stopped.


corrosion current flowing from the anodic,

or corroding areas
corrosion current flowing to the cathodic
ia = ic
total amount of external protection current
applied to the structure
portion of the protection current going to
the cathodic areas
portion of the protection current going to
the anodic, or corroding areas
ip = ipc + ipa

Figure 4.27 Illustration of Cathodic Protection

To complete the CP system the current must be discharged from an anode connection to earth.
This is called a ground bed/*-J The anode / ground bed is consumed by the oxidation side of the cor-
rosion reaction. There are two basic forms of anode; galvanic (or sacrificial) anodes and impressed
current anodes.
A number of questions need to be answered in the planning and design of a CP system/*-' *'53"'55'.
These include:
What is the soil resistivity? A measure of the ease with which current flows through the
soil; i.e. the lower the resistivity, the easier the flow.
How much current is required to achieve protection?
Will galvanic anodes or impressed current be the system of choice?
How many anodes/ground beds are needed?
Where will the anodes/ground beds be located?
Is the structure being protected bare metal or coated?
If coated, what is the coating's quality re current requirements?
What is the size and complexity of the structure?
Are there other structures in the vicinity which might lead to current interference (stray
current corrosion)?

Let us consider the choice of current source.

Current Sources

Galvanic (Sacrificial) Anodes

Galvanic anodes are made of special high purity alloys of magnesium, aluminum, or zinc. All of
these materials are more active (eager to corrode) than steel in most soils and waters. Therefore, if we
connect one of these more active metals to a piece of steel and immerse them in the same electrolyte,
the more active metal will tend to corrode (become the anode) and the steel will stop conoding (be-
come the cathode) as shown in Figure 4.28.
A summary of the properties of anode materials commonly used in water is given in Table 4.6.



Figure 4.28 G a l v a n i c Cell

E n e r g y C a p a b i l i t i e s a n d C o n s u m p t i o n R a t e s of G a l v a n i c Alloys in S e a w a t e r ' 4
E n e r g y Capability Consumption Rate
Galvanic Anode Alloy (amp-hr/lb) (lb/amp-yr) Potential (-volts*)
Aluminum-Zinc-Mercury 1250-1290 7.0-6.8 1.0-1.05
Aluminum-Zinc-Indium 1040-1180 8.4-7.4 1.05-1.10
Aluminum-Zinc-Tin 420-1180 20.8-7.4 1.0-1.05
Zinc (MIL-A-18001H) 354-370 24.8-23.7 1.0-1.05
Magnesium (H-l Alloy) 500 17.5 1.4-1.6
* Reference Electrode: Ag/AgCl

Magnesium anodes have the highest driving potential of the common sacrificial anode materials,
making it the choice for buried structures in soils of resistivity up to about 2000 ohm-cm and high-re-
sistance waters. It is often necessary to use an external resistance to obtain reasonable anode life in

Zinc can only be used in low resistivity soils (below about 1000 ohm-cm) and waters. The most
commonly used zinc alloy, ASTM B418-80 Type 1, should be restricted to temperatures below 50C
[120F]/*tf<w At temperatures above about 66C [150F], zinc and steel tend to reverse polarity. The
steel becomes the anode and corrodes in preference to the zinc.

Aluminum is the preferred anode material for most oilfield brine systems because it has a high
current capacity relative to zinc and magnesium. However, the current capacity begins to decrease at


about 55C [130FJ. falling from 1100-1200 amp-hr/lb to values as low as 440 amp-hr/lb at a tempera-
Uire of 75C [165F], depending on anode composition.'4
Advantages and Disadvantages of Galvanic Anodes' / ' '^

Advantages Disadvantages
I No external power source Low driving voltage
Relatively easy installation Only for use in low resistivity soils
Low installation costs Limited current outputs
Low maintenance Not economical for high CP current
Stray current is less likely Little control if stray current does occur
More economic with small current needs Not suitable for large diameter bare pipe |

Impressed Current Systems

DC current can also be applied through an inert anode such as graphite if power is readily avail-
able. A rectifier is necessary to convert AC current to DC current. A schematic diagram of an im-
pressed current cathodic protection (ICCP) system is shown in Figure 4.29.
One application of impressed current cathodic protection (ICCP) where "rule of thumb" design
has often been used, is the protection of the soil-side bottoms of above ground storage tanks. Coatings
and CP are used in mitigating tank-bottoms con osion. Coatings reduce the amount of current required,
as high as 99.8% for a very good coating, but do not eliminated the need for CP. Demoz and
Friesen'"4'5^ have provided equations for calculating the number of anode strings, their individual
length and spacing and the overall anode bed resistance.

Figure 4.29 I m p r e s s e d C u r r e n t C a t h o d i c Protection S y s t e m



Advantages and Disadvantages of Impressed Current Anodes^ 6j '

Advantages Disadvantages
Effective in high resistivity soils Increased maintenance
Current output is controllable Higher operating costs
Higher driving voltages Can cause stray current interferences
Flexible in application Higher installation costs
Higher current outputs Greater hydrogen production at the
Use on bare or poorly coated structures anode/ground bed
Protects larger and more complex Hi production at coating holidays can
structures lead to disbondment

Stray Current Corrosion

Especially when using ICCP, it is possible that some of the applied current will contact structures
other than the target. In such circumstances, corrosion will occur in those structures where the current
leaves and returns to the CP system via the soil. Figures 4.30 and 4.31 demonstrate this effect and a
means of preventing its occurrence.

Figure 4.30 Stray Current Corrosion


Figure 4.31 Stray Current Corrosion

Criteria for Cathodic Protection

The most important criterion for protection is to supply sufficient current to the vessel to achieve
a vessel-to-water potential of-0.85 volts with respect to a Cu/CuS04 half-cell, or -0.80 volts with re-
spect to a Ag/AgCl reference electrode.
Current densities (the total amount of current applied divided by the area to which the current is
applied) are often used as design criteria. Commonly used rules of thumb are:
Coated Vessels:
5 - 2 2 wA/w2 [ 0 . 5 - 2 . 0 mAj ft2 ]

Uncoated Vessels:
5 4 - 1 0 8 mAj m2 [5.0-10.0 mAj ft2 J

Where: 1mA = 1/1000 ampere

There is a considerable range in requirements due to differences in water composition and con-
ductivity. Potential measurements should be used to determine if protection is being achieved.
Overprotcction (too much current) can cause coating disbonding (or disbondment).
Regarding storage tanks, it is also important that the anode (or anodes) be uniformly distributed
within the tank to give uniform current distribution (see Figure 4.32). The anodes should not be al-
lowed to lie on the bottom of the vessel or to touch its sides/* 70 '
Cathodic protection is a rather complex art and you should enlist the aid of your own company
cathodic protection expert or a good supplier to help you with the details of an installation. You
should then leam how to monitor performance by taking current and potential readings to ensure suc-



Figure 4.32 Cathodic Protection of a W a t e r S t o r a g e Tank



There are three types of non-metallic tubular goods in common use in oilfield systems^ 7'

Extruded Thermoplastic Pipe

Typical examples suitable for water service are:
1. Polyvinylchloride (PVC)
2. Chlorinated Polyvinylchloride (CPVC)
3. Polyethylene
4. Polypropylene
5. Acrylonitrile-Butadiene-Styrene (ABS)
Of these, PVC is probably the most widely used.

Glass Reinforced Pipe (GRP)

GRP was previously known as "Glass Fiber Reinforced Thermoset Pipe (FRP)". There are two
major classes in common usage:
Fiberglass reinforced epoxy
Fiberglass reinforced polyester


The reinforced epoxy is most commonly used in the oilfield. It is the strongest of the non-me-
tallic tubular goods, and the most expensive. It has one peculiarity which is not found in other
non-metallic piping. It will "weep" before it will fracture or burst. (Weeping is the seeping of fluid
through the pipe walls by flowing along the glass fibers until it reaches the outer pipe surface.) Be-
cause of its strength and relatively high temperature rating (up to ~150C [300F]), fiberglass rein-
forced epoxy has been used for high pressure water injection systems and for tubing in injection and
disposal wells. This type of pipe is made by applying epoxy to glass filaments which are wound
around a removable mandrel or by a centrifugal casting process.
Fiber reinforced pipe should be purchased to API Specifications:
API 15LR Low Pressure Linepipe (< 1000 psi; < -6900 kPa)
API 15HR High Pressure Linepipe (> 1000 psi; > -6900 kPa)

Composite (or flexible) pipes are generally made up of non-metallic materials (plastic and syn-
thetic rubbers) and are often reinforced with steel strand (wire) or intcr-locking steel sheeting. Appli-
cations, vary between short connecting lines from wellheads, to flow-lines from sub-sea wellheads to
surface production facilities. Pressure ratings as high as API 10,000 have been built and deployed.
Sour service capability is a key selection factor for many composite type systems.
A class of pipe previously used in oilfield applications was cement-asbestos. Made of Portland
cement, asbestos fibers and silica, it was relatively fragile (necessitating careful handling) and had a
maximum working pressure of about 1035 kPa [150 psi]/* :>

Joining Methods
The methods used to join common types of non-metallic piping are summarized in Table 4.7.

Heat Welding
A heating element is used to soften the ends of the joints which are then butted together and held
in position until the joint cools. A slight amount of upsetting occurs.
Joining M e t h o d s for Non-Metallic P i p e
Joining Methods
Material Heat Solvent Thread
Fiberglass Reinforced Epoxy
Fiberglass Reinforced Polyester
* A rubber ring seal is also used.



Solvent Welding
Solvent is applied to both the pin and collar ends, then glue is applied over the solvent. The
joints arc quickly pushed together allowing the pin end to fuse inside the collar.

Threaded Joints
Threaded joints can be used where the wall thickness is sufficient. In threaded and coupled ther-
moplastic pipe the working pressure rating is reduced by 50%. Threads in fiberglass reinforced epoxy
pipe are either cut in an upset end or molded onto the standard pipe so there is no reduction in wall

Advantages of Non-Metallic Piping

1. Non-metallic materials are immune to aqueous corrosion.
2. They are light weight, resulting in lower freight costs and easy handling.
3. Non-metallic pipe is quickly and easily joined and installed.
4. No external corrosion protection, such as coating, wrapping, or cathodic protection, is
5. The smooth internal surface results in lower fluid friction loss.

Disadvantages of Non-Metallic Piping

1. Non-metallic pipe has rather low working temperature and pressure limits.
2. These limits are difficult to predict with assurance. The pressure and temperature limita-
tions are often interrelated (i.e., the higher the temperature, the tower the permissible work-
ing pressure, as shown in Figure 4.33). Furthermore, the physical properties change with
3. Careful handling is required in loading, unloading and installation.

i r h k.
-(GRE) & (GRP)
Asbestos Cement
3 fin
X 70
c 60
3 40
5 30

73.4 80 90 100 110 120 130 140 150 160 170 180
Temperature (F)

Figure 4.33 T e m p e r a t u r e Limitations for Non-Metallic Piping' 4 71>


4. Plastic pipe should be buried to protect it from sunlight (which will cause deterioration
with time), mechanical damage, freezing and fire.
5. Weeping can occur in reinforced epoxy pipe.
6. Non-metallic pipe has very low resistance to vibration and pressure surging. Both rein-
forced epoxy and PVC piping is commonly downrated 30-50% from the manufacturer's
recommended maximum working pressure to allow for pressure surging.


Dissolved gas removal is covered in Chapter 7.


Many installations are designed in such a way that future failures are inevitable. Following are
some recommendations which should be considered during the design of any oil or gas production sys-

Avoid Crevices
Crevice corrosion has already been discussed and we know that crevices are bad. Some exam-
ples of man-made crevices are shown in Figure 4.34^-

(a) Welds




(b) Fasteners

Figure 4.34 Welds and Fasteners



Avoid Low-Flow Areas

Solids or sludge will drop out and encourage the formation of concentration cells as well as cre-
ating ideal spots for bacteria to grow. The minimum velocity in pipelines should be 1 m/s [3 ft/sec].

Avoid High-Flow Areas

Turbulent flow, especially when suspended solids or gas bubbles are present, leads to accelerated
coirosion rates. Impingement or erosion corrosion can also result.
A long radius bend is sometimes used in place of an ell (the bend radius should be at least 5
times the pipe diameter), and the substitution of two 45 bends for a single 90 bend is quite effective
in reducing problems inherent with sudden changes in fluid direction. Another popular and effective
method is the substitution of a "blind" tee for an ell as shown in Figure 4.35.

Flow / Plug or Cap

Figure 4.35 Blind Tee

Where economically feasible, it is recommended that fluid velocities be maintained in the range
of 0.9-2.1 m/s [3-7 ft/sec].

Adequate Net Positive Suction Head

Pumps must be supplied with water at a pressure sufficient to prevent cavitation. It will also help
to prevent oxygen entry.

Select linings, coatings or corrosion resistant materials to withstand the environment if that seems
to be the most economical way to go. Use NACE Standards such as NACE Standard RP0475^^; and
NACE Standard MR-0175^ as guides during the design stage.
The selection of corrosion resistant alloys (CRAs) is a complex procedure/*73> since an incorrect
approach can lead to corrosion failures or the use of an over-expensive alloy. Bear in mind that the
corrosion resistance of many CRAs differ from one temperature to another and can vary with changes
in the concentration of critical environmental components. For example, the presence of elemental sul-
fur can have a profound impact on an alloy's resistance to stress corrosion cracking.
While choosing a CRA will tend to give high capital expenditure (CAPEX), low operating expen-
diture (OPEX) and good corrosion resistance, the overall data needs to be compared with use of carbon
steel and forms of inhibiting the corrosion. The latter theoretically offer low CAPEX, but higher
OPEX and the need for corrosion monitoring and management. The guiding principle should be to
check the data and don't make assumptions.


It is suggested that CRA selection should start with a literature review,'' comparing anticipated
field conditions, in order to narrow the field of products for testing. This review would obviously in-
clude industry standards and recommended practices. At the same time consultation with specialists
would enable the list of potential materials to be further reduced. The list of preliminary candidate
CRAs could then be tested under simulated field conditions.
Following are some points worth noting.
Avoid coupling dissimilar metals so that galvanic coirosion will not be a problem.
If hydrogen sulfide is present, or if there is reason to believe that it will be present at some future
date due to bacterial activity, then the possibility of sulfide stress cracking must be considered.
The only completely effective method of preventing this form of failure is through selection of
materials which will not crack. However, it should be noted that many materials which are immune to
sulfide stress cracking, such as the low strength carbon steels, are still susceptible to weight loss corro-
sion. Normal corrosion control procedures are required to preserve the integrity of these types of ma-
The prudent operator would be well advised to select materials in conformance with NACE Stan-
dard MRO175 for any portion of a system which falls in the sulfide stress cracking region as defined in
Figure 4.36.

Grains H 2 S perlOOscf
10 100 1000

Sulfide Stress Cracking Region

(Oil, Gas and Water)

265 psia Total Pressure

Sulfide Stress Cracking Region

Safe Region (Gas and Water)

0.0001 0.001 10
I I i Mol % H 2 S In Gas | I
1 10 1 000 10 000 100 000
ppm H2S in Gas

Figure 4.36 Sulfide S t r e s s Cracking R e g i o n s



This diagram is divided into three regions:

1. The "safe region" for all fluids. No sulfide cracking problems would be anticipated, re-
gardless of the fluids being handled or the materials employed.
2. Sulfide stress cracking problems would be anticipated for systems handling sour gas and
water. Although intended for sour gas systems, it would also be applicable to sour water
3. Sulfide stress cracking problems would be anticipated for multiphase systems handling
sour crude oil, gas and water.
There are three situations where sulfide stress cracking problems might be encountered in water
1. Naturally sour water is produced and reinjected.
2. A naturally sweet injection system becomes sour due to sulfate reducing bacteria activity.
3. A sweet oil reservoir becomes sour due to bacterial activity, resulting in sour produced flu-
ids following breakthrough of injection water. This has occurred in several fields being
flooded with seawater in the North Sea and in California.
If the produced fluids are disposed of and are not reinjected, as they typically are in offshore sea-
water floods, then potential cracking problems are confined to the production facilities. However, if
they are reinjected, then the injection system must also be considered.
Many water injection systems will not be subject to sulfide stress cracking problems due to low
pressures and/or low levels of hydrogen sulfide. However, there are an increasing number of high
pressure injection systems, and the possibility of reservoir souring by bacterial activity makes the pos-
sibility of sulfide stress cracking a very real concern in both the injection and production systems.
Consider the following examples of specific alloy use. Generally it has been found'4 3> that 13%
chromium steel (often shortened to "13 Cr") is acceptable for down-hole applications where there is in-
sufficient oxygen to cause problems. It is not considered suitable for surface equipment. The presence
of H2S may lead to accclerated pitting and/or cracking of 13 Cr steel. Duplex steels such as 22 Cr and
25 Cr are more resistant, but are susceptible to localized coirosion with small amounts of O? and H2S.
Alloy 28 has been successfully used for down-hole tubing and casing liners. It is resistant to H2S en-
vironments, but has limited resistance to SCC by elemental sulfur.

Common Sense
In short, a knowledge of the things which cause problems can save many dollars if applied dur-
ing the design of a system. There is no substitute for common sense.


Corrosion monitoring is used to detect damage before it becomes serious and hence pennit action
to be taken to control that coirosion. Monitoring also allows good management of the system by pro-
viding data to enable optimization of the corrosion control program. NACE International recommends
never to rely on any single method of monitoring corrosion, to use several techniques simultaneously
and to keep full records of the data collected.
The following corrosion monitoring techniques are not meant to be an exhaustive list, but simply
an effort to cover the main methods that have been used in oil and gas production systems. Some have


been used for a long time and are well proven, while others are still to gain wide acceptance. Those to
be considered include:
Weight-loss coupons
Linear Polarization Resistance (LPR) probes
Electrical Resistance (ER) probes
Galvanic probes
Hydrogen monitors
Field Signature Method (FSM)
Dissolved gas analysis
Corrosion product analysis
Iron counts (iron concentration analysis)
Con'osion inhibitor concentration analysis
Bacterial enumeration and generic identification
Down-hole and piping examination
Inspection methods:
- Ultrasonic measurement
- Radiography
- Visual inspection
- Failure analysis
An important consideration for any monitoring device that must be inserted into the production
system, is that it needs to be in a location where con-osion is likely to occur. Similarly, non-intrusive
measurement devices such as ultrasonic instruments, need to be positioned over corroded areas before
they can detect the damage.

Weight-Loss Coupons
Coupons are simply pieces of metal which are inserted into a system to evaluate one or more of
the following:
1. System corrosivity
2. Material performance
3. Inhibitor performance
The coupons consist of the same metallurgy as the construction material of the system under in-
vestigation. They are weighed before being inserted into the system and after removal from the system
to detennine the amount of metal lost. More detail is given in NACE Standard RP-0497, "Field Cor-
rosion Evaluation Using Metallic Test Specimens" /4?4>

A coupon may have several geometric shapes:

Flat Plates
These are commonly used in many systems. They are often mounted in pairs so that duplicate
measurements can be made.



Metal rods, 6.4 mm [0.25 in.] in diameter, are in common use.

Disc-shaped coupons are mounted flush with the pipe wall. They are mounted on the bottom of
the line, and are generally used in conjunction with the high pressure access systems discussed in the
next section.

Short sections of pipe are mounted in the system in such a way that they can be removed for
evaluation. They are commonly threaded or flanged on both ends.

Holders and Insulators

Coupon holders are designed so that the coupon is isolated from the structure (see Figure 4.37).
If the coupon and the structure are electrically connected, the corrosion coupon will nonnally corrode
faster than the structure of interest.
Holders should either be mounted in such a way that they can be isolated from the flow stream
by valving for removal, or a special retractable holder may be used which can be withdrawn while the
system is under full operating pressure.

Stainless Nuts



Flat Insulation Plate

Steel Bolts

Figure 4.37 Electrical Isolation of C o u p o n s on t h e Holder


Coupon Retrieval Systems

Two types of systems are available for the insertion and removal of coupons without system shut-
down or depressurization as listed in Table 4.8.
T A B L E 4.8
C o u p o n Retrieval S y s t e m s
Maximum Working
Description Pressure
Inserted and retrieved through stuffing
10 300 kPa [1500 psi]
box and packing gland with screw jack.
Pressure-balanced retriever 41 400 kPa [6000 psi]

Low-pressure systems are made and marketed by several suppliers. Each system consists of a
full opening valve (usually a ball valve) permanently mounted on the system, a stuffing box, a packing
gland, and a lock nut to hold the mandrel in place against pressure.

The Rohrback Cosasco system is shown in Figure 4.38.

The high pressure systems consist of a permanently mounted access fitting, a removable carrier
plug holding the coupon, and the pressure balanced retrieval system. Retrieval or installation is ac-
complished as shown in Figure 4.39.

The unique feature of these systems is the ability to equalize or "balance" the pressure inside the
retrieval tool with the system pressure. The pressure is the same on both sides of the carrier plug, so
the system pressure exerts no force on the plug during retrieval or installation.

Coupon Handling

Try not to touch the coupon with your bare fingers when mounting it in a holder. Contact with a
finger may result in increased or decreased corrosion at the point of contact in the shape of a finger
print. Whether the corrosion is increased or decreased depends on whether your finger is salty or oily.
Disposable plastic gloves are used to prevent this problem.

Coupon Location and Orientation

The coupon must be placed in a representative location, and oriented in such a way that it will
not induce turbulence in the flow stream.

If a pitting attack is taking place along the bottom of the pipe, you may not detect it with a cou-
pon located in the center of the stream. A pipe spool or a flush mounted disc coupon would be a
better choice.

Coupon Evaluation

Coupons are supplied which have been sandblasted or surface ground and pre-weighed. When
you remove them from the system, they should be visually examined for any pits or scale formation
before sending them off to a laboratory for evaluation. You may want to request an analysis of the


1. Access Fitting in service,
externals removed
2. Service Valve installed,
retriever goes in.
3. Retrieve- is attached to
4. Retriever extended, plug
past gate, valve closed.
Retriever 5. O
Plug removed. |: ji p|

Service Valve z
Access Fitting. m
Solid Plug with O
corrosion c o u p o n s X
Cut-away ol an
Access Fitting
showing a typical
solid plug
with corrosion
coupons attached.

Figure 4.38 Low P r e s s u r e C o u p o n Figure 4.39 Rohrback C o s a s c o High P r e s s u r e Retrieval S y s t e m

Retrieval S y s t e m
(Courtesy Cortest)
- jJl

Calculation and Interpretation of Coupon Results

Corrosion product is cleaned off the coupons using a special procedure in a laboratory and then
they are re-weighed. The amount of weight loss is treated as though it were uniformly removed from
the total surface area of the coupon and converted into an average penetration rate. Coirosion rates are
commonly expressed in mils of thickness lost per year (mpy) or millimeters of thickness lost per year
Weight Loss
Coirosion Rate =
Exposure Time x Coupon Surface Area x Metal Density

mpy =365000
Where: mpy = mils per year
W = weight loss, g
A = coupon surface area, in 2
D = metal density, g/in 3
T = exposure time, days

For steel, which has a density of 128.8 g/in3, the formula reduces to:
mpy =2834-
The calculated con-osion rate should be rounded off to the nearest tenth of an mpy.
In SI units, the formula for steel is:
mm / V = 465
Where: ram/y = millimeters per year
W = weight loss, g
A = coupon surface area, cm 2
T = exposure time, days

This corrosion rate is significant in systems where unifonn corrosion occurs. However, it is rela-
tively meaningless in pitting systems, because all of the weight loss is occurring in a few isolated spots
on the coupon. The number, depth, and diameter of the pits should be noted.
Coupon exposure times vary depending on the corrosivity of the system. Corrosion rates usually
start out high on the fresh metal surface of the coupon and very short exposures may give unrealisti-
cally high rates. Exposure periods of 30-90 days are common.
Often, pitting corrosion will not occur on a coupon even though coirosion is occun-ing in the sys-
tem. Although this can be due to many factors, one common cause is that the coupon exposure period
is too short. Exposure periods of 6 months or more are often necessary to detect pitting on a coupon.
This is due to the fact that pitting often doesn't begin immediately after a piece of metal is placed in
water. There is usually a delay before pitting begins. However, once a pit is initiated, it usually deep-
ens very quickly.
This means that it may be desirable to maintain two sets of coupons in a system: one set exposed
for 30-90 day periods to measure changes in general coirosion rates, and a second set exposed for peri-
ods of 180 days or more to monitor pitting coirosion rates.



Very low corrosion rates are drastically influenced by cleaning loss. Cleaning loss is the amount
of metal which is removed from the coupon during the cleaning operation. Depending on the cleaning
method, this can easily amount to 0.5 mpy. Cleaning loss is not subtracted by most commercial labo-
ratories when they report results. Thus, if a corrosion rate of 1.0 mpy is reported, the actual weight
loss due to corrosion may be only 1.0 - 0.5 = 0.5 mpy.
Of course, you are usually interested in changes in con-osion rate rather than the absolute value,
and cleaning loss should remain reasonably constant, so it is not normally critical.

Evaluation of Pipe Spools

Spools are usually too heavy to weigh, and are split open for visual evaluation and calipered to
determine any reduction in wall thickness.
The following evaluation procedure is recommended:
1. M a r k t h e t o p o f the s p o o l a n d t h e d i r e c t i o n o f flow.
2. Remove the spool from the system and replace it with a new spool.
3. Split the spool longitudinally in the horizontal plane with a saw or welding torch.
4. Photograph the internal surfaces of the top and bottom sections.
5. Remove samples of any scale or deposited solids from both the top and bottom segments
and send them to a laboratory for analysis.
6. Clean the internal surfaces thoroughly with a wire brush and photograph again.
7. Note the general condition of both halves of the spool.
8. Count the number of pits and measure their depths. Calculate the pitting rates based on
the exposure period.
The primary advantage of a spool over a coupon is that it is an actual piece of the system. The
true condition of the pipe surface can be determined. The type of attack and its location on the cir-
cumference of the pipe can be ascertained. This can be very important, as corrosion is often confined
to a specific segment such as the bottom 90 quadrant.

Use of Coupons to Monitor Inhibitor Performance

When monitoring the performance of an inhibitor it is advisable to first determine the uninhibited
corrosion rate in the system as a base. Once this has been established, corrosion inhibition is initiated
and coupons are mounted in the system. Inhibitor performance is usually measured in terms of Percent
Uninhibited Rate - Inhibited Rate
%Protection = xlOO
Uninhibited Rate

Some Advantages and Disadvantages of Weight Loss Coupons

Advantages Disadvantages
Simple Slow (1 month or more)
Inexpensive (relatively) Small quality of data
Measures local and general corrosion Laboratory processing needed
rates Only gives an average corrosion rate


Linear Polarization Meters Instantaneous Corrosion Rates

Coupons give a direct measurement of metal lost over some period of time. Linear polarization
measurements give a reading of the corrosion rate occurring on a coupon or "electrode" at the moment
the measurement is being made.

A source of current is connected between two metal electrodes and a small amount of current is
made to flow between the two electrodes. An electrolyte, such as water, must be present to conduct
current between the probes. This current flow changes the potential of the "test" electrode. It has
been shown that the corrosion rate (or more precisely the corrosion current) is proportional to the test
current divided by the change in potential when the potential change is small (20 mv or less). So we
apply sufficient cun'ent to the test electrode to change its potential by a specific amount (say 10 mv),
and measure the amount of current required to effect the change.

Corrosion Rate = K
Where: Al = current necessary to change potential of test electrode by 10 mv
AE = 10 mv
K = calibration factor which is a function of the size of the electrode and the type of

Types of Instruments Available

Linear polarization meters are available in either 2-electrode or 3-electrode models.
When a two-electrode probe is used in fresh water, the reading must be corrected for solution re-
sistivity. No correction is required in most salt waters and the con-osion rate may be read directly from
the instmment. The potential change of the test electrode is measured with reference to the second, or
"working" electrode. (Figure 4.40)

Figure 4.40 S c h e m a t i c of Two-Electrode Linear Polarization Meter



A three-electrode probe requires no correction for solution resistivity and a direct readout of cor-
rosion rate in mpy is possible. In the three-electrode model, the third electrode is allowed to corrode
freely (no current is applied to it), and it is used as the reference electrode. (Figure 4.41).

DC Current


Figure 4.41 S c h e m a t i c of T h r e e - E l e c t r o d e Linear Polarization Meter

Both two and three electrode probes are available with replaceable probe elements which may be
utilized as weight loss coupons. Probes are also available in the form of flush mounted probes de-
signed primarily for use in pipelines. Either single-reading meters or recording meters are available.
The recording meters are available in either single-channel or multi-channel models.

Taking Readings
Single readings can be taken with a small battery-powered meter which can be carried around to
various probe locations. In many cases this is adequate. In other cases, a recording meter should be
left hooked up to the electrode for an extended time period. In all cases, probes should be mounted in
the system and left there long enough to "age" or become accustomed to the environment so that repre-
sentative corrosion rates may be measured.
The use of recording meters has proven to be extremely valuable, especially in systems where the
corrosivity seems to change without reason. Many systems are corrosive only on a part-time basis.
When ccrtain conditions are present the system may suck in air, for example. A spasmodic problem is
very difficult to solve without a continuous record of corrosion rate, which can then be correlated with
other events to find the cause of the problem.

One of the problems with linear-polarization readings is that they do not measure pitting rates.
The tendency to pit can sometimes be inferred by changes in the natural potential difference between
two electrodes without any applied current. This approach is not very reliable.

Probe Removal
Probes should always be mounted in such a way that they can be removed from the system, ei-
ther by providing isolation valving or by using retractable probes. Some typical installations are
shown in Figure 4.42. Probes should be removed periodically, cleaned, and inspected visually for
signs of pitting.


\ Electrical

Heavy Duty



Pair Probe

A: Standard Linear Polarization Probe. B: Linear Polarization Probe.

System must be depressurized for Mounted on high pressure
installation and removal. a c c e s s fitting. Can be retrieved
without system depressurization.

30' Overall

C: Linear Polarization Probe. Typical low pressure retractable probe installation.

Figure 4.42 Linear Polarization P r o b e s a n d P r o b e Installations (Courtesy Cortest)



Probes sometimes short out due to accumulation of iron sulfide between the electrodes. This re-
quires removal for cleaning.

The probe elements, or in some cases the probe assembly, must be replaced periodically, which
necessitates removal. If they are pre-weighed prior to installation, they can also be used as corrosion

Some Advantages and Disadvantages of Linear Polarization

Advantages Disadvantages
Measures instantaneous corrosion rate Restricted to electrolyte (must have
Measures general rate water)
Can use continuously Can "short" due to FeS

Potentiodynamic Polarization Studies

Potentiodynamic polarization studies performed in the field provide a rapid means of estimating
the pitting corrosion rate. It is the only method available capable of this feat, and is an invaluable tool
to evaluate the effectiveness of inhibitors to control pitting. It is normally used on a periodic basis
rather than as a continuous monitor.

General corrosion rates can also be estimated by this method. However, these rates can be mea-
sured more easily by linear polarization.

A potentiodynamic instrument is an electronic device that can control the electrochemical poten-
tial of a small piece of metal immersed in water (an "electrode") by the application of DC current.

When the electrode potential is changed over a range of values, the clectrode is said to have been
"polarized." A plot of the logarithm of current density (amount of applied current/electrode surface
area) versus the electrode potential is called a polarization curve. A curve generated by a continuous
change of the potential at a constant rate is called a potentiodynamic polarization curve.

In order to study coirosion phenomena it is usually necessary to run two separate polarization
curves, or "scans." The cathodic curve is generated by applying current to the electrode, resulting in
successively more negative potentials. The anodic curve is the result of drawing current from the elec-
trode, giving increasingly positive potential values.

The technique as described by Martin^-75' is summarized as follows:

1. Insert a three-electrode probe into the system and allow it to come to equilibrium with the
system. One electrode is used for the cathodic curve, one for the anodic curve, and the
third is used as a reference electrode.
2. Perform a cathodic potentiodynamic scan.
3. Execute an anodic scan on a different electrode, this time reversing the scan direction at a
current density of about 5 mA/cnr. If this back anodic scan shows positive hysteresis, its
intersection with the cathodic curve gives a current density proportional to pit growth in
the system at the time of the test.
4. Remove the anodic test electrode from the system and determine the ratio of the unpitted
area to the pitted area. ( A typical value in many systems is 2 0 / 1 . )


5. Calculate the pitting rate using the following equation:

Pitting Rate (mpy) = 455 x PCD x PAR

Where: PCD = Pit current density, m A / c n r . The intersection of the anodic back scan with the
cathodic curve.
PAR = Pitted area ratio. The ratio of the unpitted area to the pitted area.

An idealized plot of the polarization curves for carbon steel in an oilfield brine is shown in Fig-
ure 4.43. Calculation of the pitting rate is as follows:

PCD = 0.005 mAj cm-; PAR = 20

Pitting Rate = (455)(20)(0.005) = 4 5 . 5 mpy

Scan Re verse


PIT Current De nsity

i i i
0.001 0.01 0.1 1.0 10.1
Current Density (mA/cm )

Figure 4.43 Typical P o t e n t i o d y n a m i c Polarization Curve 1 ' 4 75'1



Electrical Resistance Probes

The most widely used resistance probes utilize a loop of wire which is inserted into the system
and allowed to corrode. The electrical resistance of the wire is inversely proportional to its cross-sec-
tional area.

Where: R = resistance of wire

p = resistivity of the wire material
L = length of wire
A = cross-sectional area of the wire
Therefore, as the wire corrodes away, its cross-sectional area is reduced and the resistance in-
An instrument is attached to the probe which measures the ratio of the resistance of the wire that
is exposed to corrosion to the resistance of an internal reference wire that is protected from corrosion.
(Figure 4.44) The resistance of the exposed wire is affected by temperature as well as wire diameter.
By measuring the ratio of the two resistances, temperature changes that equally affect both wires are
cancelled out and the readings are directly proportional to the diameter of the exposed wire.

Exposed A

Element Probe

\ Check

Cable up to
1000 ft. for

I Instrument

F i g u r e 4.44 S c h e m a t i c of C o r r o s o m e t e r Electrical R e s i s t a n c e P r o b e (Courtesy Rohrback Cosasco)


The instrument readings are plotted versus time, and the slope of any straight line portion of the
curve may be converted to a corrosion rate, as shown in Figure 4.45. It does not provide instantaneous
data like linear polarization measurements.


| 160
c t* *

I tr31

120 A
B 100 -A
E 80 V
o P eriod Corr. Rate

8 60 s A 185 pm/\/
5 40 B 25 pm/\/
C 927 pm/\j
0 5 10 15 20 25 30 35 40
Time (Days)

Figure 4.45 Typical Corrosometer Data (Courtesy Rohrback Cosasco)

Probe elements other than wires arc also available as illustrated in Figure 4.46. Small, strip-type
elements can be mounted flush to a pipe wall. Thin-walled, tube-type elements are used when high
sensitivity is required for measurement of low corrosion rates.

Figure 4.46 Corrosometer Probe Element Styles (Courtesy Rohrback Cosasco)

Resistance probes do not require a conductive electrolyte and are often used in gas systems as
well as in aqueous systems.
Detection of pitting is also a problem with resistance probes. Visual inspection is still the only
reliable way to determine the type of attack.
Accumulation of conductive corrosion product (like iron sulfide) on the probe wire can cause a
decrease in resistance, resulting in an apparent decrease in coirosion rate. Shorting can also occur.
Once again, the probes should be removed periodically, cleaned and examined.
A more recent development related to ER probes are Inductive Resistance probes. They are simi-
lar to ER probes but have significantly improved sensitivity.



Some Advantages and Disadvantages of Electrical Resistance Probes

Advantages Disadvantages
Simple Give average rates between readings
Relatively inexpensive Erroneous readings with FeS and high
Continuous monitoring conductivity environment

Galvanic Probes
When two dissimilar metals are immersed in water and connected together with a wire, current
will flow between them because of their natural potential difference. The amount of current flow is a
direct function of the corrosivity of the system.
Commercial probes typically utilize a brass probe and a steel probe connected to a microammeter
to measure the current flow as shown in Figure 4.47.

Figure 4.47 G a l v a n i c P r o b e Assembly*" 4 - 7 ^

Recording microammeters are most commonly used with this type of probe. It is quite inexpen-
sive compared to other monitoring instruments. However, it does not give a readout in mpy, and its
primary utility is in detecting changes in the corrosivity of the system. Therefore, the use of a record-
ing model is highly recommended. Anything which will cause a change, such as velocity, pH or dis-
solved oxygen, will cause a change in current. However, they are extremely responsive to the presence
ol dissolved oxygen and are most commonly used as "oxygen detectors."
Figure 4.48 compares the response of a galvanic probe and an oxygen meter to very small quan-
tities of dissolved oxygen in a specific water. The scale for the oxygen meter is 0-0.84 mg/L and the
galvanic probe scale is 0-400 \lA.<4-76)


F i g u r e 4.48 Comparison of Galvanic P r o b e and Oxygen Meter R e s p o n s e ' 4 - 7 ^

Hydrogen Permeation Rate Measurement

As explained previously, many of the hydrogen atoms generated by corrosion in sour systems en-
ter the steel. The rate at which atomic hydrogen permeates the pipe wall is an indirect indication of
corrosion rate in sour systems.

Pressure Hydrogen Probe

The simplest device for the detection of hydrogen penneation is a hydrogen probe. It is a hol-
low, thin walled, steel tube which is sealed 011 one end and equipped with a pressure gauge on the
other end. The tube is inserted into the system and corrosion occurs on the external surface of the
tube. Hydrogen generated by the coirosion reaction permeates through the wall of the tube into the
hollow interior, causing the pressure to increase. The rate of pressure increase is proportional to
the corrosion rate.<4J7-47S>

Electrochemical Hydrogen Probe

Another method of measuring hydrogen penneation rates utilizes an electrochemical cell which is
attached to the outside surface of a pipe or vessel and refeired to as a "patch probe". The hydrogen
generated by coi rosion on the internal surface of the equipment travels through the steel wall into the
cell where the hydrogen is detected and "counted."^-77'4-79)



Two different types of hydrogen probes are illustrated in Figure 4.49.

Pressure Hydrogen Probe

H"1" + e

1 ' - : T -
Palladium Foil
(Working Electrode)
Electrochemical Hydrogen Probe
Auxilliary. Reference
Electrode Electrode

Figure 4.49 Hydrogen P r o b e s

Field Signature Method (FSM)

This method is non-intrusive, in that its measuring probes are externally attached/ 45 ^ FSM
detects corrosion damage over a relatively large area of the structure, compared to the other methods.
Sensing pins positioned apart at a distance of 2-3 times the wall thickness, measure the voltage created
in response to an induced current. The change in voltage distribution, or electrical field pattern, is re-
lated to the change in wall thickness. The procedure is to compare successive measurements and
hence the change in level of corrosion damage.
Additional Electrochemical Methods
Other techniques which have been used in studying corrosion, some just in the laboratory, in-
Electrochemical Noise (EN) Technique
Acoustic emission (AE)
Electrochemical Impedance Spectroscopy (EIS)
Harmonic analysis
Thin Layer Activation (TLA)
Zero Resistance Ammetry (ZRA)
further information about these methods can be gained from the Corrosion Doctors website/4*0'


Dissolved Gas Analysis

Since we know that the presence of even very small amounts of dissolved oxygen can cause se-
vere coirosion, measurement of dissolved oxygen concentrations using CHEMets (see Chapter 2) or an
oxygen meter is extremely important in systems which are supposed to be airfree.
Dissolved ILS concentrations are routinely measured in normally sweet systems as a means of
detecting the presence of sulfate reducing bacteria.

Corrosion Product Analysis

One area of monitoring sometimes overlooked, is the wealth of infonnation that can accrue from
an analysis of the corrosion product. For example, the appearance of iron sulfide in a normally sweet
system would indicate the probable presence of sulfate reducing bacteria. Iron sulfide will oxidize
when exposed to air and convert to iron oxide. Therefore, a laboratory analysis may show iron oxide
when in reality the con-osion product was originally iron sulfide.
However, both good sampling and analysis of the corrosion product as well as reasoned interpre-
tation, is necessary to obtain meaningful results. This is where the chemist goes beyond the simple
corrosion equations and formulae which we nonnally use for convenience. In order to do this we also
need to use more complex analytical equipment such as x-ray diffraction (XRD), EDS (energy
dispersive spectroscopy) and SEM (scanning electron microscope). Since these need to be operated in
a specialist laboratory, the stability of the samples and procedures to minimize their change during
transportation, become crucial considerations. Typically samples can undergo oxidation on exposure to
the air.(4-81)

However, the results of these analyses can give a guide to the cause(s) of the corrosion.'^ SL4-S:i
Corrosion product layers may well display top to bottom variation in composition, which gives impor-
tant clues re the mechanism that has occurred/*5^
For example, the predominant reaction in wet C0 2 environments up to about 80C [176F] is;
Fe + H2CO} -> FeC03 + H2

Above that temperature Fe304 (magnetite) is likely to be the corrosion product.

With FLS the situation is more complex. In water it does not form a simple acid and it gives rise
to several iron sulfide species/**-' Up to about 90C [194F] in most oil and gas conditions, the fonn
of iron sulfide produced is inackinawite, with a non-stoichiometric formula of FeS(i.x). This is often
depicted as FegSs- At temperatures higher than 90C [194F] and moderately high H2S concentrations,
troilite (FeS) and pyrrhotite or greigite (Fei.xS) predominate. Formation of pyrite and marcasite (FeS2)
requires the presence of elemental sulfur (for example if oxygen ingress has occurred and converted
some H>S to S)/48"
Reaction kinetics favor mackinawite formation, so it is often the 1st layer on the metal sur-
factJ 4 - 8,) It fonns quickly and is typical when low H2S partial pressures exist.
When both CO2 and H2S are present, the con-osion product is a function of the two partial pres-
The fonn of rust produced when dissolved oxygen is present, depends on both the concentrations
of dissolved oxygen and chloride ions. Also, when oxygen enters into a system it may well react with



any corrosion products already present. Subsequent interpretation of the analytical results then pres-
ents further problems.
The effect of dilute acid and a measure of the magnetic properties of a deposit have commonly
been used as site tests. Unfortunately these can be misleading/*5-' Effervescence could be due to
H2S evolution, not just CO2. Also, some forms of iron sulfide are not soluble in acid. Likewise, when
a deposit was seen to be magnetic, it was typically assumed to be magnetite. However, iron sulfide in
the form of greigite (Fe3S4) is also magnetic.
Hence, in order to make a sensible appraisal of the corrosion process from the product layers, it
is necessary to use all appropriate analytical tools. Then carefully assess the results using a good
knowledge of corrosion chemistryJ4S1>

Iron Counts
Iron counts (for details see NACE Standard RPO192)(4S3), or more precisely, the dissolved iron
concentration in the water can be a good measure of the amount of corrosion occurring in a sweet sys-
tem. However, several precautions should be taken to assure good results.
1. Natural Iron
Determine if there is any "natural" iron in the water. Occasionally formation waters
contain a few mg/L iron, even when no corrosion is occurring. Even rarer are those wells
producing from iron rich strata that give 50-100 mg/L of dissolved iron. Hence, high iron
counts are usually as a result of corrosion reactions.
2. Precipitated Iron
Of primary concern is the amount of iron dissolved in the water in systems containing little
or no oxygen. This means a single speck of solid corrosion product can give erroneously
high readings. Also, exposure of the water to air will cause the dissolved iron to precipi-
tate as ferric hydroxide. Therefore, it is desirable to filter the sample immediately after
sampling to remove any precipitated iron.
Then even if the iron precipitates from the filtered sample, we know that it was dissolved
at the time of sampling and the sample can be acidified to put the iron back in solution for
3. Oxygenated and Sour Systems
Iron counts in systems which are thoroughly aerated or which contain H2S are of limited,
or no, value. The primary problem is that the iron being dissolved by corrosion is usually
precipitated somewhere in the system and very little stays dissolved in the water.

Inhibitor Concentrations
Inhibitor concentrations in a fluid can be determined in several ways. If you are interested in
measuring inhibitor concentrations, contact the chemical company of your choice, and they will give
you the details.

Bacteria Counts
Detennination of the bacteria content of a water is another indirect method of inferring
corrosivity. It is covered in Chapter 5.


Downhole Inspection

Mechanical-Multifinger Caliper Tools (Tubing and Casing)

Mechanical caliper tools are run in wells on wirelines to detect internal corrosion of tubing or
casing. A continuous profile is measured by spring-loaded mechanical feelers arranged around the cir-
cumference of the tool which are forced against the inner pipe wall. A continuous log is produced
which shows the profile of the internal surface of the tubing or casing in varying degrees of detail.

There are three types of tools available:

1. Maximum Recording
This tool produces a single trace which is a recording of the displacement of the feeler
which has extended the furthest distance from the tool. It does not tell you how many
feelers extended that distance. This means that a pit contacted by a single feeler would
produce the same record as circumferential corrosion contacted by all feelers if both the pit
and the circumferential corrosion were the same depth.
2. Minimum-Maximum Recording
There are two types of tools available which continuously record both minimum and maxi-
mum feeler extension. The simplest tool produces a two-line recording indicating mini-
mum and maximum internal diameter measurements. There is also a tool which records 3
minimum and 3 maximum measurements, one each per 120 section, or 6 maximum cali-
per measurements, one per 60 section.
3. All-Feeler Recording
This tool continuously records the extension of all feelers. The record will consist of as
many lines as there are feelers, making it possible to determine the pattern of attack as
shown in Figure 4.50.
Tubing OD = 2.375 in. Area Reduction = 25.02%
Tubing ID = 1.995 in. Total Area Remaining = 978 sq. in.

15 2



9 8

Mechanical Caliper Inspection Report by J. C. Kinley Co.

F i g u r e 4.50 C r o s s Section of C o r r o d e d Production Tubing'' 4 84>



Electromagnetic Induction (Casing)

There are two types of casing inspection tools which utilize magnetic field induction to detect
both internal and external casing corrosion.

DC Electromagnetic Induction
Typical downhole DC tools consist of one or more circular an'ays of sensor pads, also called
shoes, mounted so as to be in close proximity of the inside casing surface. The tool saturates a section
of casing between the poles with an induced electromagnetic field, setting up symmetrical lines of
magnetic flux as shown in Figure 4.51. If there is an anomaly in the pipe wall such as a pit or hole,
the symmetry of the flux patterns is disturbed, and Jinx leakage occurs. The flux leakage is detected
and measured by the sensor pads as they pass through the area of flux leakage. This measurement will
detect both internal and external corrosion.

* Note disruption in the field caused by the pitting (Publication No. 9541, Report 01/87,2.5M, Dresser Industries)

Figure 4.51 Downhole Electromagnetic Inspection Too\(4 84>

Simultaneously, a high frequency AC current is used to generate a magnetic field which will in-
duce eddy currents in the interior wall of the casing. In most tools, the eddy currents penetrate only
about I mm into the casing wall. An eddy current response is indicative of metal loss on the internal
surface. Defects on the outer surface of the casing are not detected by eddy current testing due to the
shallow depth of investigation.

Comparison of the flux leakage and eddy current responses make it possible to differentiate be-
tween internal and external corrosion damage as shown in Figure 4.52.
This type of instrument typically cannot detect pit depths which are less than 20% of the casing
wall, or less than 6.35 mm [0.25 in] in diameter. Pit-depth accuracy is approximately 15% of the pit
depth under ideal conditions where all casing properties (weight, grade, etc.) are known. (4S6>

A C Electromagnetic Induction
There are two types of tools which fall in this category. In both cases a low frequency AC cur-
rent is passed through a transmitter coil to induce an electromagnetic wave in the casing wall. The


I. D.
Inreasing Casing Thinning
1350 1350

I L_

1400 I 1400
f External

( Pitting
1450 1450


{ 1500
Electronic Internal Casing Thickness Log
Caliper Log

Figure 4.52 Typical R e s p o n s e of DC Electromagnetic / Eddy Current Inspection TooK 4 85^

wave is detected by a receiver coil(s) located 305-610 mm [12-24 in.] away. The amount of phase
shift or change of amplitude of the received signal(s) is used to calculate wall thickness. (Figure 4.53)
Type 1
The measurements made by this type of tool do not account for changes in magnetic prop-
erties or casing grade, and are not considered suitable for quantitative corrosion measure-
ments. Some tools give a qualitative assessment of the internal diameter.
Type 2
These tools measure the magnetic properties of the casing while logging, which allows a
quantitative measurement of wall thickness. A quantitative measurement of casing ID is
made simultaneously. In addition, it is possible to use this type of tool to detect corrosion
damage in multiple casing strings.1{4-sr>
Neither type of tool is recommended for detecting small holes or isolated pitting. Its primary
utility is for the detection of severe general coirosion or large holes/4,S6)

Ultrasonic Scanning (Tubing and Casing)

Some ultrasonic tools measure only the ID of the tubing or casing, while others measure both
casing thickness and casing ID, and can discriminate between internal and external damage. (Figure
4.54) The instrumentation consists of one or more ultrasonic transducers focused perpendicular to the
pipe wall/ 4 , w
This type of tool works in either oil or water. Signal response is attenuated by gas, heavy mud.
and scale buildup. While the caliper reflects reduction in diameter due to scale deposits, the thickness
measurement responds only to metal thickness.
The accuracy of the dimensional measurements is typically 0.6 mm [0.025 in.].



(Courtesy Schlumberger)

Pipeline Inspection Tools

Pump-Through Inspection Tools

Inspection tools are available which can be pumped through a pipeline and will give a continuous
record of pipe wall condition. They are often referred to as "intelligent pigs" or "smart pigs."

Use of these tools require that the pipeline be equipped with full-opening valves, and suitable pig
launching and receiving facilities.

The minimum inspectable diameter is 4-6 inches, depending on the type of tool. Most tools re-
quire a minimum bend radius of 3D, although 5D may be required for the smallest diameters.

Careful cleaning is mandatory prior to any pig runs. This is especially true in produced-water or
wet crude-oil lines, as wax buildup can impede an inspection tool's progress and affect its accuracy.


Magnetic-Flux-Leakage Pigs (MFL)

This type of tool measures wall thickness, and detects both internal and external corrosion.
Some tools can differentiate between internal and external corrosion/*69'
MFL tools can be used in gas lines as well as in liquid-filled lines.
The magnetic flux leakage principle is described in this chapter under the heading of Downhole
Inspection. A simplified drawing of an MFL inspection pig is shown in Figure 4.55, and Figure 4.56
shows the response of the magnetic field to a corrosion pit.

Pipe Wall Battery Pack P r e s s u r e Vessel Odometer Wheel

Leakage-Flux Detector

Rubber C u p s

Figure 4.55 Simplified S c h e m a t i c of MFL Pipeline Inspection Pig' 4 - 90 ^

Figure 4.56 Pit Detection with a n MFL Pipeline Inspection P\q(490>



Ultrasonic Pigs (US)

Ultrasonic tools require that the line be filled with fluid in order to acoustically couple the ultra-
sonic transducers and the pipe wall. The liquid should be homogeneous in order to prevent scatter of
the ultrasonic signal. Gas bubbles, oil droplets, or suspended solid particles can cause problems.
The ultrasonic wall-thickness tool utilizes the pulse-echo concept to detect metal loss. Non-con-
tacting ultrasonic transducers are affixed to the tool body so that they remain a fixed distance from the
pipe wall as the tool travels through the pipeline. (Figure 4.57) An ultrasonic pulse of known velocity
is transmitted through the liquid to the pipe wall. There are two echoes: the first arriving from the in-
side surface of the pipe and the second arriving from the outside surface.

Ultrasonic S e n s o r

S e n s o r Standoff (A)


| 27.0

% 15.0 -Inl e r n a l -
| 9.00
<f External
3.00 t 1 |

200 400 600 800 1 , 0 0 0 1 , 2 0 0 1,400 1 , 6 0 0

D i s t a n c e Travelled, m m

Figure 4.57 Pipeline Wall-Thickness M e a s u r e m e n t with an Ultrasonic P\g(4-9*>

Late amval of the first echo represents loss of thickness from the inside wall (internal corrosion),
while early arrival of the second pulse represents metal loss from the external pipe surface.(4 9n

Wireline Tools
Also available are pipeline inspection tools which are pulled through a tine with a cable or wire-
line. This approach makes smaller diameter tools possible, since the data processing and storage
equipment do not have to be built into the tool. The data is transmitted through the cable to external
data acquisition equipment.
The use of a cable limits inspection length to a little over 2 km [1 mi.] with a copper wire/^59'


Inspection of Surface Facilities

Ultrasonic Inspection
Ultrasonic devices are used to measure metal thickness and to detect defects in the metal.
The basic concept of ultrasonic inspection is to generate high frequency sound waves and trans-
mit them into the material being inspected. These waves travel through the specimen until they en-
counter the opposite surface of the specimen or an internal discontinuity. Either an interface or a dis-
continuity will reflect the sound waves back to their origin where they can be detected (Figure 4.58).
The time interval between transmission of the ultrasonic wave and the arrival of the reflected wave
back at the point of transmission is proportional to metal thickness or the depth of the flaw. This
method is called the pulse-echo technique.

Couplant _
Transmitter Back Wall

Impulse Flaw
< d, >Echo Back Wall Echo

Zero Line

Figure 4.58 P u l s e - E c h o Ultrasonic M e a s u r e m e n t Principle

The sound wave is both transmitted and received by a transducer, which contains two crystals
made of piezoelectric materials such as quartz or a polarized ceramic. Piezoelectric materials will vi-
brate when subjected to a changing electrical field and will also produce an electrical potential when
the material is placed under mechanical stress.
An electrical current is passed through one of the crystals, and it vibrates at a very high fre-
quency (5-15 megahertz). This vibration produces ultrasonic sound waves, which are transmitted into
the metal. This crystal is the transmitter.
The second crystal produces electrical current when struck by the reflected vibrations and acts as
a sound wave receiver. (Figure 4.59)' 4!

Current Applied Current Produced

Transmitter Receiver
Crystal Crystal

Test _


Figure 4.59 Ultrasonic Transducer^ 4



Ultrasonic wall-thickness measurements are quite accurate on smooth, flat plates. (Figure 4.60(a))
However, it should be noted that serious errors can occur when attempting to measure the remaining
wall thickness opposite corrosion pits because the signal is scattered by the irregular surface/*^

Echo Pulse No Echo

0 0.1 0.2
-L 0 0.1 0.2



F i g u r e 4.60 Ultrasonic Measurements^- 9 3 - 1

In the example shown in Figure 4.60(b), the internal roughness scatters the signal such that insuf-
ficient energy returns to the transducer to give a reading.

Radiography is widely used for inspection of welds. It is also used for corrosion detection.
Either X-rays or gamma rays can be used, as they have similar wave lengths. However, gamma
ray radiography is the most widely used method for field inspection. Gamma ray sources (iridium-192
and cobalt-60) are much smaller than X-ray machines and require no electrical power/4'94'
The process of radiography involves passing gamma rays or X-rays from a source, through the
item being inspected, onto a piece of photographic film. When developed, the film is called a radio-
The amount of radiation which passes through a metal in a given length of time is inversely pro-
portional to its thickness. This means that more radiation will pass through a thinned, corroded area
than through a thicker, undamaged area. The photographic film becomes darker as it is exposed to in-
creased amounts of radiation. Therefore, pits or corroded areas show up as dark spots or areas on a

Radiographs are frequently taken in the horizontal plane to give a profile of the pipe walls. They
can also be taken in the vertical plane to establish the pattern of attack on the bottom (or top) of lines
or vessels.
Radiographic inspections can be carried out while equipment is in service.

Visual Inspection
To date there is no substitute for the human eye coupled with a modestly intelligent brain. Al-
ways utilize every opportunity to inspect. Any time a line is open, a pump is down, or a tank is being


cleaned, gel right in the middle of things and look. Use a camera, take notes, measure anything you
can, and make drawings. This is the most reliable way of determining system condition.

Failure Analysis
Any time a failure occurs, go out, get nosy and conduct a postmortem. Use consulting or com-
pany laboratories if necessary and determine the cause of failure. Chances are you can avoid repetition
of the failure if you can detennine its cause.
A failure is your most expensive form of coirosion coupon. Use it wisely. Don't pinch pennies
in analyzing it.

Equipment Records and Failure Reports

One of the most important things you can do is to keep accurate, up-to-date records. Know what
you have done and when you did it. Once again, the idea is to keep track of your past performance
(right or wrong) so that you can profit from it in the future.

Final Philosophy
Never rely on any single method of monitoring corrosion. Use several techniques simulta-
neously, if possible, and keep records with a passion!
A summary of corrosion monitoring techniques is presented in Table 4.9.
C o r r o s i o n Monitoring T e c h n i q u e s

Method Measures Comments

Coupons Average coirosion rate by weight loss. Must be positioned where corrosion
Pitting rate by pit depth measurement. is occurring.
Spools Pattern of attack. Pitting rate by pit Very useful in surface systems.
depth measurement. May be able to Not as easy to remove as a coupon.
Linear Polarization Instantaneous general corrosion rate. Requires conductive fluid (water).
May have problems in sour sys-
Potentiodynamic Polarization Estimate pitting and general coirosion Used primarily for corrosion inhibi-
rates. tor evaluation.
Electrical Resistance Change in electrical resistance of cor- Not normally used in sour systems
roding element. Gives general corro- due to conductivity of iron sulfide.
sion rate.
Galvanic Probe Current generated by bimetallic cou- Primarily used for Oi detection.
Hydrogen Probe Hydrogen generated by con'osion of For sour systems. Must be temper-
probe. Rate of pressure increase is ature compensated.
proportional to corrosion rate.
Hydrogen Patch Probe Hydrogen generated by corrosion of Used only in sour systems.
pipe wall. Gives hydrogen penneation



C o r r o s i o n Monitoring T e c h n i q u e s (Cont'd.)

Method Measures Comments

Dissolved Gas Analysis 0 2 , H 2 S, CO-! Presence of H 2 S in sweet system
indicates sulfate reducing bacteria.
Corrosion Product Analysis Indicates which dissolved gas is re- FeS will oxidize to iron oxide on
sponsible for corrosion. exposure to air.
Dissolved Iron Amount of iron dissolved by corro- Not quantitative in sour or oxygen-
sion. ated systems. Must subtract any
"natural" iron.
Inhibitor Concentration Concentration of inhibitor present in Helpful to determine inhibitor dis-
fluid. tribution in system.
Bacteria Counts Number of bacteria present. Related to corrosion rate.
Mechanical Calipers Internal corrosion in tubing or casing. Scale or corrosion product may
Pitting or general. mask pits.
Electromagnetic Induction Measures wall thickness and ID of Does not detect small holes or iso-
casing. lated pitting.
Ultrasonic Scanning Measures ID in tubing and casing. Response is attenuated by scale
Some tools also measure wall thick- buildup.
Magnetic-Flux-Leakage Pigs Detects both internal and external at- System must be built to accept tool.
tack in pipelines. General or pitting Reserved for large systems due to
corrosion. cost.
Ultrasonic Pigs Measures both ID and OD of pipe- Line must be filled with liquid.
Wireline Pipeline Inspection Measures both ID and OD of pipe- Maximum inspection length is a lit-
Tools lines. tle over a mile.
Ultrasonic Inspection Thickness of metal. Presence of pits Very localized measurement.
or cracks.
Radiography General or pitting corrosion. Particularly useful in locating pit-
ting corrosion in piping and well-
Visual Inspection Pattern and severity of attack. Extremely reliable but often incon-


4.1 vvww.corrosioncost.com
4.2 Peabody, A. W "Control of Pipeline Corrosion," 2nd Ed. (Edited by Biachetti. R. L.), NACE Interna-
tional, Houston Texas, USA. (2001).
4.3 Fontana, M. G. and Greene, N. D.: Corrosion Engineering, McGraw-Hill, New York (1967).
4.4 N A C E Standard RP0475-98; Selection of Metallic Materials to be Used in All Phases of Water Handling
for Injection into Oil Bearing Formations, N A C E International. Houston, Texas (1998).
4.5 NACE Standard MR01 75-2003; Metals for Sulfide Stress cracking and Stress Corrosion Cracking in Sow-
Oilfield Environments, NACE International, Houston, Texas (2003).
4.6 Anon.: Bet: Handbook of Industrial Water Conditioning, Betz, Trevose, Pa.(1962).
4.7 Pisigan, R. A. Jr.and Singley, J. E.: "Evaluation of Water Corrosivity Using the Langelier Index and Rel-
ative Corrosion Rate Models," Materials Performance (April 1985) 26.
4.8 Wheeler, D.: "Treating and Monitoring 450,000 b/d Injection Water," Petroleum Engineer (Nov. 1975)
4.9 Thodes, F. H. and Clark, J. M.: "Corrosion of Metals by Water and Carbon Dioxide Under Pressure,"
hid. Eng. Chem., Vol. 28 (1936) 1078.
4.10 de Waard, C. and Milliams. D. E.: "Prediction of Carbonic Acid Corrosion in Natural Gas Pipelines," In-
dustrial Finishing & Surface Coatings (Nov. 1976) 24.
4.11 de Waard, C., Lotz, U. and Milliams, D. E.: "Predictive Model for C 0 2 Corrosion Engineering in Wet
Natural Gas Pipelines," Corrosion (Dec. 1991) 976.
4.12 De Berry, D. W. and Clark, W. S.: "Corrosion Due to Use of Carbon Dioxide for Enhanced Oil Recov-
ery," CO; Corrosion in Oil and Gas Production, National Association of Corrosion Engineers (1984) 28.
4.13 Crolet, J. and Bonis. M. R.: "Prediction of the Risks of C O : Corrosion in Oil and Gas Wells," SPE Pro-
duction Engineering (Nov. 1991) 449.
4.14 Lotz, U., van Bodegom, L and Ouwehand, C.: "The Effect of Type of Oil or Gas Condensate on Car-
bonic Acid Corrosion," Corrosion (Aug. 1991) 635.
4.15 Nesic, S. et al, "Integrated C 0 2 Corrosion - Multiphase Flow Model". SPE 87555, 1st International Sym-
posium on Oilfield Corrosion, Aberdeen, 28th may, 2004. SPE, Richardson, Tx.
4.16 Speller, F. N.: Corrosion Causes and Prevention, McGraw-Hill Book Co., New York (1951).
4.17 Patton, C.C.: "Practical Considerations in the Design of Oil Field Water Injection Systems," Materials
Performance (Nov. 1977) 9.
4.18 Sheldon. W. D.: "Velocity-Accelerated Corrosion Testing and Predictions," Materials Performance (Sept.
1990) 61.
4.19 Tuthill, A. H. and Schillmoller, C. M.: "Guidelines for Selection of Marine Materials," presented at the
Ocean Science and Engineering Conference - Marine Technology Society, Washington D.C. (June 14-17,
4.20 Hilliard, H. M.: "Designing for Corrosion Control," Proceedings of University of Oklahoma Corrosion
Control Course (Sept. 14-16. 1970) 27/1.
4.21 Anon.: "How Fluid Velocity Affects Corrosion," Canadian Chemical Processing (May 1967) 73.
4.22 Brown. J. S.: "Large Seawater Supply Development for Water Injection," paper presented at the Joint Pe-
troleum Mechanical Engineering and Pressure Vessels and Piping Conference, Mexico City (Sept. 19-24,
4.23 Joosten, G. Payne, "Preferential Corrosion of Steel in C 0 2 Containing Environments," CORROSION/88,
paper no. 211 (Houston, TX: NACE 1988)
4.24 Craig, B.: "Critical Velocity Examined for Effects of Erosion-Corrosion," Oil & Gas J. (May 27, 1985)



I t is estimated^ that 120-210 million barrels water / day are produced along with oil and gas world-
wide, with the water to oil ratio averaging about 3:1. Another estimate^-puts the volume at
300-400 million barrels of water per day (bwpd) and 5-6 barrels of water for every barrel of oil. The
US average is between 7:1 and 9:1 (water to oil), but it can be as high as 50:1. Hence, for the oil and
gas companies the largest volume of fluid that they produce incurs significant costs and gives no finan-
cial return on that investment.

There are certain key components of the produced water as far as environmental considerations
are concerned, which affect decisions on how that water will be utilized or discarded. These are salt
content, hydrocarbon content, toxicity (inorganic, organic and the separate affects of chemical addi-
tives) and NORM. In this chapter we are primarily concerned with the "oil" or hydrocarbon content
and how it might be reduced to an acceptable concentration by suitable water treatment equipment.

While the hydrocarbon content is often referred to as "oil and grease", this quantity is defined by
the analytical procedure used in its measurement. However, as mentioned in Chapter 2 (under "Total
Petroleum Hydrocarbons (TPH) in Water") we must consider both aliphatic and aromatic hydrocar-
bons, whether they be dissolved in, or dispersed throughout the bulk water phase.

The produced water separated in oil and gas processing facilities may contain some residual gas
and one or more of the following oil-in-water mixtures:
Free oil (oil that quickly separates due to gravity, when the mixture is allowed to settle)
Emulsified oil (tiny droplets of dispersed oil that remain as a stable suspension, or separate
only very slowly)
Soluble oil (hydrocarbons or salts of organic acids, dissolved at ppm levels in the water).
Many authorities simply class the "free" and "emulsified" oil together as "dispersed" oil. An
emulsion is a mixture of two immiscible liquids, one of which is dispersed as very small droplets in
the other, this dispersion being stabilized by an emulsifying agent. "Oil-in-water" (O/W) emulsion,
or dispersed oil, is also known as "reverse" emulsion; the water being the bulk or continuous phase
with the oil forming the dispersed phase.


So, why do we de-oil this produced water?

To avoid environmental problems re the discharged water and hence comply with discharge
To prevent fonnation blockage, when injecting into the ground for either secondary recov-
ery or disposal purposes
To recover the oil.
The major issue nowadays is concern for the environment.


The disposal of produced water can be classified into two categories: surface disposal and
subsurface, or deep-well, disposal. The vast majority of surface disposal of produced water occurs off-
shore, where the water is discharged into the ocean. This chapter deals with the surface disposal as-
pects and subsurface disposal will be included within Chapter 7.
Expanding the options for dealing with produced water, we can include:
1. Discharge to sea or water course
2. Rcinjection (secondary recovery or disposal)
3. Beneficial re-use
4. Downhole separation and reinjection
5. Subsea separation
6. Tiy to reduce the amount of water produced from the reservoir.
Typically in the US onshore arena most of the produced water is reinjected (for disposal and en-
hanced oil recovery), whereas most is discharged by the offshore sector of the industry.

Surface Discharge to Sea or Water Course

In most countries, water produced in both onshore and offshore production operations must be
processed to reduce the oil and grease content to levels acceptable to local regulatory agencies. Al-
though the toxicity of the discharged water is of concern in some areas, removal of dispersed oil prior
to discharge is often the sole process applied to this type of water.
There are four oil removal processes commonly used for this purpose:
1. Skim tank
2. Conjugated plate interceptor
3. Induced gas flotation cell
4. Hydrocyclone
These processes may be applied individually, or in combination to reach the desired result. Typi-
cal process designs are shown in Figures 6.1 and 6.2.
Each of these pieces of equipment is described later in the Chapter under the heading of "Sus-
pended Oil Removal."
Centrifuges have been used to a limited degree, but are very costly. Membrane filtration has
been tested, but presents significant operational challenges.




Figure 6.1 Typical O f f s h o r e P r o d u c e d W a t e r T r e a t m e n t





Figure 6.2 O f f s h o r e P r o d u c e d Water T r e a t m e n t with a Hydrocyclone

Many countries adhere to the Barcelona Convention limits of less than an average of 40 mg/L
dispersed oil and grease, with a maximum upset level of 100 mg/L in water discharged to the sea.
However, some countries choose to set lower levels, such as Brazil with 20 mg/L maximum. Obvi-
ously more stringent standards are often set for enclosed sea areas. For example the limit in the Baltic
Sea is 15 mg/L, with the proviso that a limit of 40 mg/L can be used if best available technology
(BAT) is incapable of achieving 15 mg/L. Inland waterways usually have even tighter limits for

In the US no individual discharge is allowed to surface waters unless it has been authorized by a
National Pollutant Discharge Elimination System (NPDES) permit. These permits are issued by the
Environmental Protection Agency (EPA). The discharge limits are based on "best available technol-
ogy" (BAT) being used to protect water quality. Currently BAT is based on gas flotation technology.
As well as an oil and grease concentration (42 mg/L maximum, with 29 mg/L maximum monthly aver-
age) and volumetric rate limits, the discharge limits incorporate best management practices, toxicity
and ecological testing, listing of chemical additives used and provision for extra chemical monitoring.

North Sea operations are governed by the OSPAR (Oslo-Paris) Convention (Convention for the
Protection of the Marine Environment of the Northeast Atlantic), which again incoiporates BAT and
had a limit of 40 mg/L oil and grease concentration. However, from the 1st of January 2006 the lower
limit of 30 mg/L came into force, with zero harmful discharge to be achieved by 2020. A secondary


target is to reduce by 15% the quantity of oil in produced water in 2006, compared to the average fig-
ure for the year 2000 discharges. Within the UK the "Offshore Petroleum Activities (Oil Pollution,
Prevention and Control) Regulations 2004 (OPPC is the short title), are expected to modernize and
strengthen existing offshore discharge regulations'6-'. Under OPPC a variety of offshore activities will
require permits, including produced water reinjection.
OSPAR's overall intention is to eliminate pollution by 2020 and achieve concentrations of natu-
rally occurring substances near to their background levels in the marine environment. For man-made
materials the target concentration is as close to zero as possible.
Norway is pursuing a "zero harmful discharge" goal by 2005. The "Environmental Impact Fac-
tor" (E1F) is the tool being used to assess ecological risk and their aim is to achieve an EIF in 2006
which is 80% reduced from the 2002 figure. The zero discharge applies to added environmentally haz-
ardous chemicals as well as harmful natural compounds. Further details are available in the paper by
Johnsen and Utvik<^J'.

Produced Water Reinjection

Produced water from onshore US fields is most commonly reinjected into an underground forma-
tion/15'^ While some is reinjected simply as a means of disposal, the majority (more than 70%) is used
to either maintain reservoir pressures or to hydraulically drive the oil towards the producing wells. In
the US offshore sector, in common with most offshore operations worldwide, only a small amount of
produced water is reinjected.
In order to meet the increasingly stricter water discharge standards, produced water handling has
become a significant environmental challenge. The costs of developing improved water treatment
equipment, plus the operating and maintenance costs, have caused some operators to consider the via-
bility of produced water reinjection.'6'4' Extended field trials have shown that reinjection can be
achieved with no apparent FUS fonnation or loss in injectivity and no corrosion rate increase. Hence,
with much less produced water being discharged, there would be less adverse effects on the marine en-
Typically the produced water would first pass through a de-oiling unit (for example a
hydrocyclone), then if necessary, a cooler. The water may be filtered and chemicals can be added to
minimize bacterial growth and corrosion/6-4-6--'' Finally the produced water is injected into the selected
formation, either separately or mixed with seawater.

Beneficial Re-use
This includes use in drilling fluids, steam generation for tertiary recovery projects and agricul-
tural irrigation. The last two options have been used successfully in California for many years.
Further afield, experimentation with irrigation has been conducted in Oman/6-6' The incoming
produced water was treated in three stages, gas-flotation, anthracite filtration and then activated carbon
filtration, which reduced oil-in-water levels from 50-300 ppm down to less than 0.5 ppm. Successful
irrigation of three crops was achieved in comparison with fresh tap-water. Although problems of salt
build up in the soil needed to be overcome, the residual oil content was not giving any apparent prob-

Downhole Separation and Reinjection

Downhole Oil / Water Separation (DOWS) is achieved either by hydrocyclone or gravity separa-
tion devices/6'-65-6-7' The separated water, containing less than 500 mg oil/L of water, is then



reinjected into a lower formation. The aim is to reduce surface environmental concerns and at the
same time reduce produced water management costs. However, up till 1998 there were only about 40
applications worldwide. Subsequently the system has been less favored by producers, because of unfa-
vorable economics and the fact that it did not meet performance expectations. The trouble in gaining
acceptance has been recognized by oil and gas executives/*5^ By 2002 only one manufacturer was
still offering DOWS equipment. However, this technology may well be further developed and used in
field applications.

Sub-sea Separation
This approach removes the water management problems from the surface facilities, but it could
be argued that it creates even more technological problems in the design and operation of the under-sea
equipment. From an environmental viewpoint the separated water must still be dealt with; probably by

Reduction in Produced Water Volume

A reduction in the volume of produced of water is achieved by blocking devices such as packers
and plugs, as well as "shut-off chemicals". The aim is to reduce the amount of water which reaches
the production wells.
As might be expected, some oil companies adopt more than one of these approaches in order to
manage their produced water effectively. For example, efforts in Saudi Aramco^-9' were reported to
include improvement in primary separator performance, more extensive reinjection facilities and mini-
mizing water production by both well location planning and water shut-off measures. There has also
been work on developing downhole separation and reinjection of the produced water.

The United States Environmental Protection Agency began requiring biomonitoring of aquatic
toxicity for produced water discharges in the Outer Continental Shelf of the Gulf of Mexico from De-
cember, 1993 onwards. Similar regulations apply in most other parts of the world.
Aquatic toxicity is measured by comparing survival, growth and fecundity (egg production) of
living organisms in either synthetic water or natural seawater with that of the same type of organism
exposed to that same water containing varying percentages of produced water.

Toxic Substance Identification

Toxic effects resulting from the discharge of produced water is the result of naturally occurring
toxic components originating from contact with the produced hydrocarbons and/or the formation rock
and/or treatment chemicals, which end up in the water phase.

Treating Chemicals
It can be determined if treatment chemicals are the primary cause of toxicity by measuring the
toxicity of the produced water prior to chemical addition, and again after chemicals have been added.
If treatment chemicals prove to be the source of the problem, the identity of the specific chemical or
chemicals responsible can be established through a series of toxicity tests using the produced water and
different combinations of treatment chemicals.


Treating chemicals can be classified in three categories:

1. Production Treating Chemicals
Emulsion Breakers Paraffin Compounds Water Treating Chemicals
Detergents/Cleaners Biocides - Coagulants
CoiTOsion Inhibitors Scale Inhibitors - Flocculants
Oxygen Scavengers Antifoamers - Reverse Emulsion
2. Gas Processing Chemicals
Typically methanol and/or glycol for hydrate inhibition, or glycol lor gas dehydration.
3. Stimulation And Workover Chemicals
Production treating chemicals are the largest single source of concern, because these chem-
icals come in contact with produced water.

The Most Toxic?

The following types of compounds appear to offer the greatest potential for toxicity based on
acute toxicity data using varying species and protocols/6 70'
1. Quaternary Ammonium Compounds
"Quats" as they are often called, are commonly used as biocides, corrosion inhibitors, re-
verse emulsion breakers and coagulants. They are quite soluble in water.
2. Amine Salts
Amines are commonly used as corrosion inhibitors and biocides.

Chemical Treatment Issues

1. System Application
Chemicals injected into pipelines or water injection systems will not be discharged over-
board. and so are not of concern. Glycol dehydration of gas is performed in a closed sys-
tem. Workover and stimulation chemicals can be caught and re-used or treated prior to
The primary concern is treating chemicals which are added to the produced fluids and end
up in the water phase.
2. Treatment Method: Batch or Continuous
Of particular interest are chemicals which are typically applied in high concentration
batches in nonnal production operations. These include scale inhibitors (squeezes), corro-
sion inhibitors and biocides. Low level continuous treatment obviously offers a lower po-
tential for toxicity.
3. Concentration
One must consider three concentrations: The actual treatment concentration, the concentra-
tion of the chemical in the water when it is discharged overboard, and the No Observable
Effect Concentration (NOEC).
Some fraction of the chemical injected into the system will be consumed or retained in the
system, which reduces the discharge concentration to a value less than the treatment con-
centration. In addition, if batch treatments can be staggered, the concentration of a



chemical in produced fluids from one well will be diluted by production from other wells
on the platform.
4. Solubility
Biocides and scale inhibitors are the two classes of chemicals which are highly soluble in
water, as are water treatment chemicals such as reverse breakers, coagulants and surfac-
tants. Corrosion inhibitors can be water soluble or oil soluble. Emulsion breakers typi-
cally have very little solubility in water, and most paraffin treating compounds tend to be
more soluble in oil than in water.1(6 W)
Many treating chemicals contain multiple components, and one or more of these may be
soluble in water, while others are not. There is also the question of how a particular com-
ponent partitions between oil and water.
5. Adsorption
Chemicals which are not very soluble in water can adsorb on the surface of solids, such as
corrosion products, which can end up in the water being discharged overboard.

Naturally Occurring Toxic Materials

If toxicity proves to be the result of chemical compounds which are naturally present in the pro-
duced water, the cause of the toxicity is typically identified by subdividing the water into various frac-
tions and experimentally determining the toxicity of each fraction. A series of fractionation proce-
dures, each designed to remove a class of compounds, is used to subdivide the mixture into a number
of samples that can be tested individually. This procedure makes it possible to identify the class of
compounds responsible for the toxicity.

The primary naturally occurring toxicants found in produced waters are:

Dispersed Oil Radionuclides

Soluble Oil(611) 'Sulfides

Heavy Metals: Cd, Cr, Cu, Pb, Ni, Ammonia

Ag, Zn

In addition, many produced waters have extremely high salinities, and all produced waters are es-
sentially free of dissolved oxygen.

While identification of most naturally occurring toxicants is reasonably straight forward, the defi-
nition of soluble oil is somewhat more elusive. The term "soluble oil" is often used to describe the or-
ganic materials which are dissolved in the water. Rabalais'* 5 categorized them in four groups in de
scending order of abundance:
1. Organic Acids: The largest component.
Fatty Acids: Water produced with paraffmic oils often has high concentrations of fatty
acids. However, the fatty acids in produced water are produced as sodium salts of the
aci,j (6.13)

Aromatic Acids: Water produced with asphaltic oils contains notable amounts of naphthen-
ic acids.


2. Saturated Hydrocarbons
3. Volatile Aromatics and Phenols: Benzene and toluene comprise about 80% of these com-
pounds. They are typically more abundant in produced waters from gas condensate opera-
4. Polynuclear Aromatics: Smallest fraction; however, it is also the most toxic.

Toxicity Reduction
If you wish to reduce the toxicity of a given discharge, it is necessary to first identify the
cause(s) of the toxicity.

Treatment Chemicals
If toxicity is due to treatment chemicals, there are several options which might result in reduced
1. Chemical Substitution
The first option would be to substitute a less toxic treatment chemical for the offending
product. For example, a glutaraldehyde might by substituted for a quaternary amine
2. Change Treatment Method
The use of continuous treatment in place of batch treatment would substantially reduce
peak chemical concentrations in the produced fluids. This could significantly reduce toxic-
3. Alter Chemical Composition
The use of a less toxic solvent package, or alteration of the solubility characteristics to re-
duce the concentration in the water phase could reduce toxicity.

Natural Toxicity
Reduction of naturally occurring toxicity is somewhat more problematic due to the cost of addi-
tional equipment, increased operating costs, and space and weight limitations on offshore platforms.
Some of the more common water treatment equipment which could be considered for the purpose
of reducing naturally occurring toxicity are:
1. Gas Flotation
Dispersed gas flotation cells arc commonly used offshore to reduce the oil and grease lev-
els in produced water to acceptable levels. A flotation cell can also tend to lower the con-
centration of suspended solids.
2. Hydrocyclones
Hydrocyclones are commonly used to reduced the concentration of dispersed oil in pro-
duced water. This type of equipment typically has little effect on suspended solids.
3. pH Adjustment
Modest pH adjustment using phosphorous acid has been successfully employed in the Gulf
of Mexico as a means of precipitating "dissolved oil" so that it can be removed with dis-
persed gas flotation.



4. Filtration
Filtration equipment is primarily used to reduce the concentration of suspended solids in
produced water. However, it is also used to reduce the concentration of dispersed oil.
5. Air Stripping

Air stripping using a countercurrent stripping tower(s) could be used to reduce the concen-
tration of volatile fractions. Of course, the impact on air quality would have to be consid-
ered, and some sort of scrubber would probably be required.


Oil carryover can cause many problems in the various options mentioned above for dealing with
produced water. Hence, the oil content of that water should be minimized, particularly prior to surface
water disposal or injection into an underground formation. The following section briefly describes
those oil/water separation technologies most commonly used in oilfield production operations. A sum-
mary of various techniques used in offshore situations has been developed by the OSPAR Commis-

The first step in reducing oil carryover is to make sure that the primary oil/water separa-
tion process is functioning properly and that the demulsifier is doing its job. Remember that the
main objective for the demulsifier is to demulsify (resolve water-in-oil emulsions) and dehydrate the
oil. Wherever possible the second aim is to produce clean water. It is often forgotten that the primary
separation vessels are designed for the separation of gas and water from the oil. They are not specifi-
cally designed as water treatment vessels. Run a thorough check to ensure that the treating tempera-
ture, residence time and demulsifier selection and concentration are acceptable. If the primary separa-
tion does not produce water with the desired oil content, then oily water treatment processes are
needed. One or more oil-in-water separation devices may be required to comply with the oil concen-
tration limit imposed by the final destination of the produced water.

Earlier in this chapter it was pointed out that the "oily" water separated from the hydrocarbons in
the primary separation vessels, consists of small droplets of oil dispersed in the water phase. The
shearing forces encountered by the fluid mixture while passing through the oil and gas production sys-
tem, do not give a dispersion of equally sized droplets. The droplets can range in size from less than I
pm (micron = 10 - 6 meter) up to about 500-600 (im. The larger oil drops (say > 2 0 0 p m ) separate very
easily within the primary separation stage. The remaining oil droplets cover a range of diameters best
summarized as an "average droplet size distribution" termed the djo- This represents the diameter at
which 50% of the oil droplets have bigger diameters than d$o and 50% are smaller than this value.
The droplet size distribution is measured using one of various commercially available instruments,
some of which can be used continuously on-line in the system.

The stability of the droplet dispersion is dependent upon:

the ionic charges developed on the droplets
droplet size and density
pH and ionic strength of the water
the water temperature.
In nature, particles distributed in water tend to attract layers of ions that impart an overall nega-
Uve charge (see Figure 6.3(a)). In simple terms, the size of that charge is dependent upon the nature of


Figure 6.3 Particle Charges in Natural Waters

the particle, or in our case, the oil droplet. Due to this negative charge, as oil droplets approach each
other long range repulsion of these "like charges" (Figure 6.3(b)) occurs and the droplets are prevented
from colliding. Hence coalescence of the oil droplets is hindered.

Increased ionic strength of the water overcomes the ionic charge on the droplets, thereby decreas-
ing emulsion stability. The emulsion stability would tend to decrease with increased temperature. The
effect of pH is complex. At high pH levels materials in the oil can form natural soaps which stabilize
an emulsion. The high pH could also lead to mineral scale forming and stabilizing the droplet disper-
sion. Decreasing the pH would counter these effects, but going too acidic might lead to corrosion

The various pieces of equipment available to separate oil from the produced water rely on a
physical separation of the two phases. This is governed by Stokes' Law:

V = (6.1)

Where: V = oil droplet rising velocity m/s ft/sec
d = diameter of oil droplet m ft
Pw = water density kg/m3 lb/ft3
Po = oil density kg/m3 lb/ft3
H = water viscosity (dynamic) kg/ms lb/ft-s
g = constant 9.81 m/s 32 ft/sec 2

Anything that can improve the velocity of separation of the oil droplets, will ultimately assist in
the physical separation of the oil and water within the residence time limits of the equipment. From
Stokes' Law we see that the velocity will increase if we can:

Increase the gravitational effect

Increase the difference between the oil density and that of the water
Increase the diameter of the droplets
Decrease the viscosity of the water (a marginal effect)



The various designs of oil-in-water treatment equipment incorporate one or more of these param-
eters in order to improve separation efficiency.

As we will see while progressing through this chapter, de-oiling (oil removal) device efficiency
decreases as oil droplet size decreases. Since the diameter is a squared function in the equation it has
the largest impact on velocity, so it is not surprising that much effort is spent in trying to increase oil
droplet size.

Principles of Water Treatment

The overall aim is to merge small oil droplets
in order to form larger droplets which subsequently
separate more easily and in a shorter time.

The first step is overcoming the long range

repulsion between the droplets, by neutralizing the
negative charges. Typically a long-chain cationic
polyelectrolyte coagulant would be used. The next
step, flocculation, is where gentle agitation may be
used to encourage collisions between the droplets
and hence bring them together in clusters (see Fig-
ure 6.4). Once together, the small droplets can co- Figure 6.4 Coagulation and Flocculation by
alesce into larger oil drops. Bridging

Inorganic trivalent metal salts such as ferric sulfate, ferric chloride and alum, have been used for
coagulation in the past. They react in the water to produce voluminous hydrated hydroxide precipi-
tates that incorporate the dispersed oil within their floes. However, the dosage of metal salts required
to neutralize the negative charge on the oil droplets is large compared to the dosage of polymer re-
quired to neutralize the same charge. Also, the floes generated are weak and large amounts of oily
sludge are formed. Sludge formation is a particular problem when H2S is present in the system. For
offshore operations disposal of any sludge is a problem, but this oil bearing sludge is a disposal prob-
lem even in onshore facilities. Hence, the organic polymeric coagulants tend to be used nowadays.

The optimum chemical product and its concentration can be determined by test procedures re-
lated to the method employed within the oil treatment equipment. Three traditional selection test
methods are the:
Bottle test
Gang (or Jar) stirrer test
Bench flotation cell test

The closer the test resembles the field equipment, the more likely the chosen product will per-
fonn adequately in the system. For example, using the bench flotation cell for selecting de-oiling
chemical for gas flotation units, the selected product has often performed equally in the system at
within lppm of the experimentally optimum dosage rate.

Due to environmental considerations, some operators are reluctant to consider the use of chemi-
cals. However, the use of small quantities of chemical can often bring about significant improvements
in oil removal and an overall reduction in adverse materials entering the environment. Over-treatment
of chemicals can give adverse results, as well as increasing the costs. Since some regular demulsifiers,
corrosion inhibitors and bactericides can interfere with the reverse emulsion breaker's performance, or
that of the water treatment equipment, chemical use must certainly be optimized. It is often advanta-
geous to enlist the support of a reputable chemical service company to assist in this optimization.


For the purposes of this book, the mechanical equipment available for removal of oil from water
has been grouped as follows:
Skimmer Vessels
Gas flotation units
Plate separators
Filters / Granular Media coalescers
Hydrocyclones / Centrifuges
More recent developments.

It should be noted that as well as removing rising oil droplets, many of these devices will also
aid in the removal of some of the larger solid particles.

Skimmer Vessels
Skimmer vessels are the simplest and most widely used device for removing dispersed oil from
water. They are simply vessels which provide retention time so that free oil will have time to rise to
the surface where it can be skimmed o f f and collectcd. There are many different designs ranging from
field-modified storage tanks to purpose-built vessels incorporating special inlet and outlet systems such
as in Figure 6.5. Skimmer vessels are used primarily to separate the bulk of the free oil from the water
and are often used upstream from other oil removal equipment.

Figure 6.5 Skimmer Tank (Courtesy Natco)

This category includes skim tanks, vortex tanks, horizontal skimmer vessels, API separators and
skim piles. Performance efficiency should be about 100% oil removal for droplets > 1 5 0 pm, with
varying levels of removal of droplets between about 100-150 pm. Generally they can handle large
fluctuations in influent oil content. However, efficiency is obviously affected if an emulsion layer de-
velops at the oil-water interface in the vessel.



Vertical Downflow Skim Tanks

Most water skim tanks in onshore installations utilize a vertical downflow configuration because
of its simplicity and low space requirements. In the downflow design, the oil droplets must rise up-
ward countercurrent to the downward flow of the water.

Three commonly used downflow geometries are shown in Figure 6.6. in the first two designs,
the emulsion is distributed laterally near the center one-third of the tank by passing through a spreader
bar network, or discharging from the inlet line upwards toward a spreader plate. Some coalescence oc-
curs in the zone between the spreader and the water collector, and the buoyancy of the oil droplets
causes them to rise counter to the water flow. Oil is collected and skimmed o f f the surface.

Residence times can be from hours to days, however, the basic disadvantage of these designs is
channeling. This is often due to solids build up in the vessel bottom causing only a fraction of the
available tank volume to be used. As a result, the actual transit time through the vessel is a small per-
centage of the theoretical value.

Field tests on the designs shown in Figure 6.6(a) and 6.6(b) show volumetric utilization efficien-
cies of 20% or less, while the third design has an efficiency of less than 5%.

Figure 6.6 Common Skim Tank Designs


Vortex Tank
Since the size of skim tanks will increase with the volume of produced water throughput, the
weight and "footprint" of such tanks will limit their use in offshore situations. The vortex tank is an
improved vertical downflow skim-tank design developed by Amoco. (Figure 6.7) This design exhibits
much better volumetric utilization efficiencies (typically 40-80%) and higher oil recoveries than more
conventional configurations. Furthermore, existing vessels can be easily converted to a vortex tank.

To Inj.

Figure 6.7 Vortex Flow System

The vortex design maintains high recoveiy efficiency because of its unique flow pattern which
helps achieve high volumetric utilization. It is considered to be one of the best overall field-proven
skim-tank configurations because of its simplicity, high recovery efficiency, and low space

The design employs a low velocity vortex current to overcome the tendency of oily water to
channel directly from the tank entry to the discharge. The vortex current is created by injecting the
water into the settling tank tangential to the tank wall, a short distance below the oil-water interface.
The water enters the tank at a 60 angle through a horizontal converging-diverging diffuser nozzle. The
nozzle is specially designed to reduce turbulence in the tank near the nozzle exit. The water is dis-
charged near the bottom at the center of the tank.

The incoming fluid causes the fluid in the tank to rotate slowly about the vertical centerline of
the tank. This rotational velocity is much too small to result in any separation by centrifugal force;
however, it is adequate to overcome any tendency for the incoming fluid to channel directly to the tank
outlet before gravity separation of the oil and solids from the water can occur.

Oil rises to the surface where it is continually skimmed by the oil skim line. This fixed line is
normally located as high in the tank as possible. It is a common practice to place it one foot down
from the top of the tank.

Horizontal Skim Vessels

Many offshore installations use horizontal pressure vessels similar to the one shown in Figure 6.8
as skim tanks. Safety concerns often dictate the use of a pressure vessel due to the possibility of
ovcrprcssuring and gas venting. In addition, horizontal vessels are typically more efficient than verti-
cal vessels.



G a s Out

0= IV
W a t e r In Water Water
l t e i Zfl-> W a t e r

- ^ O u t

Oil Out

Figure 6.8 Horizontal Skim Vessel (Courtesy ESI)

API Separator
API separators were originally rectangular concrete basins, but now are available in other materi-
als of construction. An entry box normalizes the flow into the unit and baffles reduce any turbulence
as the oily water enters the main separator section. Here, the quiescent flow allows oil droplets to rise
to the surface, where they are skimmed off by a series of paddles and pushed into an oil collection
trough. Sludge settling on the bottom of the separator is scraped to a removal basin (see Figure 6.9).

Oil Weir

Hopper Water Outlet
Inlet Box

Figure 6.9 API Separator

Due to their size and weight, especially when full of water, API separators are typically only
used onshore.

Skim Piles/Caissons

Skim piles and caissons are devices used in offshore situations. They are typically the final unit
through which the discharged produced water passes into the sea.

The caisson is the simpler of these two devices and simply consists of a large diameter vertical
pipe (closed at the top) which extends down into the sea. Treated produced water feeds into the top of
this "pipe" and as the water slowly makes it way down and out into the sea, remaining free oil droplets


have the opportunity to rise to the top end of the caisson. Any accumulated oil is then skimmed of the
surface and returned to the oil system.

Skim piles use tilted plates within a caisson in order to provide shorter distances for the oil drop-
lets to travel. As the droplets move up the underside of these plates they are collected in vertical riser
pipes that facilitate taking the collected oil to the top of the vessel. Again the oil is removed at the
surface. In essence, the skim pile combines the effects of a tilted-plate separator and a caisson.

Some Advantages and Disadvantages of Skimmer Vessels

Advantages Disadvantages
Skim Tanks Simple design Channeling
Long residence time (in theory) Solids deposition
Copes with slug flows Not for offshore
Separates bulk of free oil Emulsified oil not removed
Horizontal Skim Vessel Can use offshore Upset by wave motion
More efficient than vertical tanks Only removes free oil
Copes with slug flow Solids deposition

Flotation Cells
Gas flotation is a process wherein fine gas bubbles are formed in the water and attach to particles
of oil or solids and help to lift them to the water surface for collection/* 7 4 ' The bubbles give an ap-
parent reduction in oil droplet density, which makes the term (r w - r c ) bigger in Equation 6.1 and
hence increases the separation velocity, V. Coagulants, polyelectrolytes, or reverse demulsifiers
(de-oilers) may be added to produce larger oil droplets and hence, further improve performance.

The essential features of a gas flotation unit are:

Introduction of finely dispersed gas bubbles into the stream to be treated.
A zone of minimum turbulence.
A means to remove the floated material from the water surface.
Commercial gas flotation units may be classified by the way in which gas is introduced into the
water. There are two basic classifications: dissolved gas flotation and induced gas flotation.

Dissolved Gas Flotation

The gas is dissolved in the water under pressure. When the water flows into the flotation vessel,
the pressure is released allowing the gas bubbles to evolve/614~616> (Figure 6.10) These units are sel-
dom used in modern oilfield operations because of their rather ponderous size, high cost and the gener-
ally better results achieved with induced gas flotation/*' 5 )

Induced Gas Flotation

The gas is mechanically dispersed in the water using either a mechanical rotor or an eductor.
1. Rotor-Type Induced-Gas Flotation

Rotation of the rotor generates a liquid vortex resulting in a gas-liquid interface which ex-
tends through the interior of the rotor to a point just above the bottom of the cell. (Figure



Figure 6.10 Full-Stream Dissolved Gas Flotation (Courtesy SPE)

6.11) The gas cavity within the vortex is at sub-atmospheric pressure resulting in the flow
of gas from the vapor space in the cell through the rotor where it is mixed with the water.

Produced gas is used for flotation. A gas pressure of approximately 0.2 kPa [0.5 ozs (0.03
psig)] is maintained in the cell to prevent air entry.

Most rotor type induced gas units contain four separate flotation cells. The oily water
flows through each of the four cells in series, with about one minute residence time per
cell and the oil and solids content is progressively reduced in each cell. (Figure 6.12)
These units can remove free oil and emulsified oil droplets greater than about 10 microns
in diameter. From 90 to 95% oil removal is expected. Experience shows that an effluent
containing 20-40mg/l oil-in-water is typical from influent oil contents of 100-300mg/l.

Figure 6.11 Cross Section of Rotor-Type Dispersed G a s Flotation Cell

(Courtesy Wemco Division, Envirotech Corp.)


Figure 6.12 Four-Stage Dispersed-Gas Flotation Cell (Courtesy Natco)

There are examples of even better performance, especially when suitable chemical
treatments are used
Typical gas water ratios are around 6.25 m 3 /m 3 [35 ft 3 /bbl] at design throughput/ 6 - 1 ' However,
since the total volume of gas dispersed in the water is not normally adjustable, water throughputs less
than design result in substantially higher gas/water ratios.
2. Eductor-Type Induced-Gas Flotation

Clean water is pumped from the discharge of the flotation cell through an eductor in each
cell as shown in Figure 6.13. This creates a localized low pressure area at the nozzle exit
within the eductor. This causes natural gas to flow from the vapor space into the eductor
where it mixes with the water as it leaves the nozzle. It works exactly like the air ejector
shown in Figure 7.9 (see Chapter 7), except that the pressure drop in the nozzle is achieved
by pumping water through it rather than by pulling gas through the nozzle.

This type of unit typically uses less power and less gas than a rotor type unit. Gas/water
ratios are typically less than 1.8 m 3 /m 3 [10 ft 3 /bblj at design throughput/ 6 7 7 ^ The volume
of gas dispersed in the water is not adjustable in most commercial units, so throughputs
less than design result in higher gas/water ratios.

Eductor type induced gas notation units are available which contain one, three, or four
cells. A four cell unit is shown in Figure 6.14.

At least one vendor supplies a unit which is housed in a pressure vessel and has no internal
moving parts (Figure 6.15).
The performance of gas ilotation cells is influenced by many variables including the gas concen-
tration, water salinity, pH, temperature, crude oil type, inlet oil concentration, and chemical flotation

Induced-gas flotation cells are widely used offshore to remove dispersed oil from waters being
discharged overboard. They are used infrequently in onshore water-injection operations. Induced-gas
notation cells are not bothered by suspended solids and are capable of removal efficiencies of > 90% if



Figure 6.13 Cross Section of a Cell in a Multisep Induced-Gas Flotation (Courtesy Monosep, inc.)

Figure 6.14 Nozzle-Type Four Cell Gas Flotation Cell (Courtesy Wemco Division. Envirotech Corp.)


Gas Out

Skim Eductor Gas Return

(Valve Optional)
Secondary Skimmer

Primary Skim Trough

Inspection Hatch

Sight Glass
Purge Gas

Process Outlet
Water Level Control
Outlet Valve Access Hatch

Eductor Supply Pump

Process Inlet
Pressure Switch For
Pump Protection
Primary Skim Out

Eductor Water
Supply Header

Eductor Gas
Supply Header
Eductor Gas
Supply Header

Striker Plate

Slip On Flange

Level Probe^ Sensing Probe

Eductor Nozzle
# Froth

Eductor Nozzle

Water Recirculation

Eductor Assembly Level Controller

Figure 6.15 Induced Static Flotation Cell (Courtesy Envirotech Corp.)



operated properly. They typically require careful attention and do not accept system upsets

Factors which affect the performance of gas flotation units include:

oil droplet rise rate

droplet rise distance
horizontal water flow velocity
design of the inlet, outlet and skimmings removal device.

A generalized comparison of induced gas flotation and dissolved gas flotation follows:

Induced Gas Flotation Dissolved Gas Flotation

Large, fast rising bubbles Small, slow rising bubbles
Turbulent separation Quiescent separation
Thin, wet skimmings (float) - (4-6% of throughput) Thick, "dry" float (< 1% of throughput
Insensitive to temperature Higher temperature > lower efficiency
Smaller "footprint" Larger "footprint"
1 5 minutes total retention time 20-30 minutes total retention time

Some Advantages and Disadvantages of Flotation Cells

Advantages Disadvantages
Achieve discharge limits with correct chemical Moving parts (maintenance)
dosage Scale fonnation
Selected chemical works in system at test dosage Upset by oil and hydraulic slugs
Long use experience Restricted visual monitoring in H ; S systems
Will remove/strip some dissolved and volatile Sensitive to wave motion
organics Relatively heavy/bulky
Skimmed solids returned to oil system

Plate Separators

Plate separators are a form of coalescing device that provide solid surfaces which can be con-
tacted and wet by small oil droplets. The droplets collect on the surface, collide and coalesce into a
continuous oil fdm. As the film becomes thicker, large drops of oil break loose due to gravity and/or
fluid flow, and rise to the water surface.

Since plate separators have been available for some time and have been extensively used, we
cover them separately in this section. The more recent coalescing devices are discussed later in the

Parallel Plate Separator

The parallel plate separator is a gravity separator. As the name implies, the parallel plate separa-
tor consists of a stack of tilted parallel plates approximately 4 to 10 cm apart. The oily water passes
between the plates and free oil particles rise until they reach the underside of a plate where they collect
and coalesce. (Figure 6.16) Any particle heavier than water will fall to the top of the plate below.


The coalesced oil travels upstream alone

the plate until it reaches the water surface where
it is skimmed o f f and collected. Hence, the es-
sential feature of a parallel plate separator is to
reduce the distance an oil particle must travel
before reaching a collection surface to a few

Tilted plate separator (TPS), or interceptor

(TPI), is the generic name for this piece of
equipment. The name is derived from the fact
that the functional plates are always tilted at an
angle, typically at 45. The Parallel Plate Inter-
ceptor (PPI) consists of flat plates forming a se-
ries of chevrons as shown in Figure 6.17.
When the plates are corrugated it is a Corru-
gated Plate Separator (CPS) or Interceptor
(CPI), see Figure 6.1 8. Please note that PPI can
also mean Pressure Plate Interceptor, where the
plates are within a pressurized vessel operating
at the same pressure as the feed header. This
Figure 6.16 Parallel Plate operation avoids any shearing action.

The CPI is more commonly used in oil

production operations than is the Parallel Plate
Interceptor. It is a fairly efficient separator and
occupies a modest amount of space, which
made it rather popular in installations where
space is at a premium, such as in offshore pro-
duction facilities. An apparent advantage of the
CPI is the concentration of oil and solid streams
as shown in Figure 6.19.

Plate coalescing devices use closely

spaced plates to reduce the distance the oil
droplets have to travel. However, because of
the close spacing, they are susceptible to plug-
Figure 6.17 Parallel Plate Interceptor ging with solids.

The water residence time in plate separators is typically 5-15 minutes. A large proportion of oil
droplets > 3 0 |j.m diameter will be removed, with an oil removal efficiency of 100% being claimed for
droplets >35 | i m in diameter. With influent oil concentrations of 500-4000 mg/L, it is expected that
the effluent will range from 50-500 mg oil per liter of water. It is also claimed that slugs of 800 000
mg/L of oil-in-water can be handled. However, unit efficiency is severely reduced when blocking oc-
curs between the plates.



Figure 6.18 Corrugated Plate Interceptor (CPI)

Figure 6.19 Looking Down the CPI Plates

Some Advantages and Disadvantages of Parallel Plate Separators

Advantages Disadvantages
Small compact units Solids (wax, scale, etc.) plug gaps between the
Weight and "footprint" OK for offshore plates
Units can be added as needed Channeling reduces residence time
' No moving parts Chemical often ineffective (due to inadequate time
Apparently low capital and installation costs to function)
High oil and solids removal efficiency


Traditional Media Coalescers and Filters

Granular Media Coalescers

A granular bed coalescer is a vessel which contains a granular material with a large amount of
surface area. As the oily water flows through the coalescing media, the finely dispersed oil droplets
collect on the media surface and coalesce to form larger droplets. Once they become large enough the
f l o w of water will displace them from the media and they will float to the surface of the water for

Both upllow and downfiow granular media coalescers are available. Three typical units are
shown in Figures 6.20-6.22. The nut shell media filters discussed in Chapter 7 are also used as
coalescers and for final "polishing" to remove oil down to very low levels (say < 2 mg/L). In most
oily water applications nut shell filters will retain 98% of droplets and particles of > 2 pm diameter.
With influent oil content of 25-50 ppm an effluent of < 5 ppm would be expected and with < 25 ppm
inlet, the effluent would be about 1 ppm oil-in-water.

These units gradually become saturated with oil, and must then be backwashed. Some operators
dispute the claim that these units are easily cleaned (by backwash). Apparently it becomes increas-
ingly difficult to remove the oil from the nut shells during backwashing, even with the use of cleaning
chemicals. Manufacturers typically recommend having less than about 100 mg/L total suspended sol-
ids and less than about 100 mg/L oil, in the inlet stream. Eventually the media needs to be replaced.

Some Advantages and Disadvantages of Media Filters

Advantages Disadvantages
Can consistently meet oil-in-water limits Limited oil + solids inlet requirement
Large adsorption surface area Runtime dependent on solids loading
Removal of finely dispersed oil droplets Heavy and bulky
Short backwash time High maintenance
Waste stream disposal
Particulate and wax plugging
Periodic chemical cleaning

Oil Coalescing Filter Cartridge

Oil coalescing filter cartridges are also available. In the cartridge shown in Figure 6.23, How is
radial from the inside to the outside. Filter life is primarily determined by the suspended solids con-
centration in the water, as oil will not plug the cartridge.

Adsorption Filter
Adsorption filters typically contain chemically treated cellulose fibers which allow the chemical
adsorption of aliphatic oil components. Droplets and particles > 2 0 pm diameter are removed by the
filter, but regeneration is not possible. Hence, the disadvantages include frequent filter replacement
due to filter clogging and the associated costs.



Figure 6.20 Upflow Colexer Coalescer Figure 6.21 OSX Oil/Solids Downflow
(Courtesy Infilco Degremont. Inc.) Coalescer (Courtesy Natco)


Cross-Sectional View

1. Perforated Polypropylene Core

2. Corrugated Primary Coalescing Medium
3. Plastic Perforated Outer Cage
4. Polymeric Outer Wrap
5. Polypropylene End Cap

Figure 6.23 Oil C o a l e s c i n g Filter Cartridge (Courtesy Pall Well Technology Corp.)

Membrane Filters
Membrane filters can be either ceramic or synthetic, with 0.1-0.2 pm pores. Offshore tests have
revealed that 70-90% dispersed oil removal is possible.

Unfortunately such tests also reveal that frequent shutdown for maintenance might be expected,
with the additional need for relatively intensive supervision.

Some Advantages and Disadvantages of Membrane Filters

Advantages Disadvantages
High efficiency Needs pretreatment
Remove emulsified and dispersed oil Frequent mechanical failure
Membrane elements are simple Membrane fouling
Large recycle flow
High capital cost
Requires chemicals
Sensitive to feedwater quality
Sludge for disposal

Enhanced Gravity Separation

A hydrocyclone is a device which multiplies gravity by as much as 2000 times, thereby accelerat-
ing the oil-water separation process. Although it has been claimed that a hydrocyclone will easily



separate oil with droplet sizes down to 8 (am, an industry survey quotes 98% removal for droplets >
15-30 pm. Below 15 pm the removal efficiency decreases significantly. While chemicals might help
by increasing the droplet size, in many systems insufficient reaction time is available upstream of the
hydrocyclone. Influent oil content should be less than 5000 ppm (20 000 ppm occasional slugs), from
which an effluent oil concentration of 20-250 ppm is achievable (i.e. up to 99% removal efficiency).

A schematic drawing of a typical hydrocyclone designed to remove oil droplets from water is
shown in Figure 6.24. Water is introduced tangentially through a rectangular port into a conically
shaped chamber. The water rotates rapidly around the chamber and moves downward, exiting at the
small end of the cone. As the water moves down the cone, the decreasing diameter causes the rota-
tional velocity of the water to increase. This increases the centrifugal acceleration of the water (the
denser fluid) toward the outer wall of the cone, and the centripetal acceleration of the lighter oil drop-
lets towards the center of the cone. By setting of the outlet pressures at either end of the hydrocyclone
a slender core of oil (typically 1-2 mm in diameter) forms in the center and moves towards the wider
end of the hydrocyclone hibe. This oily stream is called the overflow, concentrate, or reject stream and
represents about 1 to 3% of the feed to the unit. The clean water exits from the bottom of the cone
and is called the underflow.

Clean Water

Oil/Water Inlet
Oil Concentrate Discharge

Figure 6.24 Oil/Water Hydrocyclone' 619 >

Transit time through a hydrocyclone is typically less than one second.

The pressure drop across a hydrocyclone has typically been 345-730 kPa [50-100 psi] per stage.
An additional 345-730 kPa [50-100 psig] back pressure must be applied to the underflow stream to
collapse the air vortex and force the overflow stream to flow out the top. This translates into feed
pressures of about IT00 kPa [155 psi] for a single stage unit. Because of the pressure requirements,
the water must either be separated at high pressure or pumped. If a pump is required, it must be a spe-
cial low-shear pump in order to minimize the creation of small oil droplets, which reduce separation
efficiency/ 6 7 9 '

In addition to units which operate at normal feed pressures there are designs which operate at
feed pressures of about 350 kPa [50 psi]. These units have capacities of 48-159 m 3 /d [300-1000
bbl/day] and offer excellent oil removal efficiency. The low feed pressure should significantly extend
the range of applications of this technology.

'CHAPTER 10 269

Hydrocyclones are used primarily for the removal of oil from water in many offshore production
operations. 1 " They are compact, insensitive to platform motion, and have no moving parts. How-
ever, excessive solids in the feedwater must be removed, otherwise they can block the "reject" stream
outlet which is typically only 1-2 mm in diameter. Although hydrocyclones have seen little, if any, ap-
plication in water-injection systems, they represent an important oil/water separation technology and
are worthy of consideration when economically feasible.

Since the feed water to the hydrocyclone is under pressure, it retains some of the produced gases.
Hence, a degasser vessel is always required downstream of the hydrocyclone. This acts like a flota-
tion device which can enhance dispersed oil removal.

Some Advantages and Disadvantages of Hydrocyclones

Advantages Disadvantages
Relatively small size and weight Dependent upon correct inlet pressures
Not sensitive to platform motion Dependent upon inlet droplet size
Modular build-up as water volume increases Downstream degassing vessel needed
Easy to maintain (no moving pans) "Reject" stream recycled (possible emulsions)
Established "track record" Solids plugging or erosion
Removes smaller oil droplets than previous de-
1 vices
Easy installation
No spare parts to stock

Dynamic/Rotating Hydrocyclone and Centrifuges

The principle of operation of a dynamic hydrocyclone is exactly the same as the static unit, ex-
cept that an external motor is used to rotate the cyclone wall and the inlet vortex cone at 1600 to 3000

As might be anticipated, the dynamic hydrocyclone has shown somewhat improved performance
over a static hydrocyclone due to the increased gravitational forces. This is especially true for oil
droplets with diameters less than 15 pm.

Increased rotational speeds will generate higher centrifugal forces, resulting in better oil-removal
efficiencies at a given flow rate.

The potential for solids fouling is greater in the dynamic device than in the static units because
of the low axial velocity (approx. 0.6 m/s) at the cyclone wall. Water-wet solids will migrate to the ro-
tating wall, accumulate, then move toward the outlet. This may be a problem in waters with high sus-
pended solids concentrations.

In comparison, centrifuges rotate at about 8000 rpm, creating a separation force equivalent to
about 5000 times gravity. Centrifuges are quite expensive and require regular maintenance. Along
with the high running costs (when compared to say hydrocyclones), it means that centrifuges are still
not widely used for high volume throughput produced water de-oiling. However, several are in service
in North Sea gas facilities for separation of oil from water prior to overboard discharge of the water.
Centrifuges can remove oil droplets down to about 2-5 microns/**-" The main advantages and disad-
vantages of centrifuges might be summarized as:



Advantages Disadvantages
Highly efficient oil removal High capital + operating costs
Unaffected by platform moving High energy consumption
Efficiency less related to droplet size Intensive maintenance needs
Does not need pressurized feed or chemicals Can be upset by sand and solids
Very low oil-in-water in effluent stream

More Recent Equipment/Process Developments

For many years it has been accepted that large dispersed oil droplets are efficiently removed in
the various physical separation processes. During the 1990's it became increasingly recognized that at
the lower end of the droplet size spectrum, the size of the oil droplets play a significant effect in the
performance of de-oiling equipment and processes. Oil-in-water analyses in conjunction with
droplet/particle size analyses, showed that the various designs of de-oiling equipment had differing
abilities to remove smaller oil droplets. Table 6.1 gives an approximate comparison of equipment per-
formance against mean droplet size. Obviously any one particular piece of equipment may perfonn
better than shown in this table, dependent upon inlet oil-in-water concentration, use of de-oiler
chemicals and the level of operational maintenance on that process equipment.

In a process system there are obviously many sources of shear. This shearing effect gives rise to
a range of oil droplet sizes and ultimately affects the de-oiling equipment efficiency. To counter this
effect the droplets need to be coalesced into larger droplets, which will then more easily separate in the
available equipment. Chemicals, specially designed physical coalescing devices, or a combination of
these, must be used.

Some of the equipment and processes developed in recent years were specifically aimed at in-
creasing the mean oil droplet size, thereby leading to improved oil removal in the currently installed
separation devices. This has been a fairly typical approach offshore, with the intention of gaining
better hydrocyclone performance. Other equipment/processes are intended as "stand-alone" separation

Comparison of Ec uipment Performance
Minimum Droplet Outlet Oil
Category Type Size Removed (pm) Concentration (ppm)
Simple gravity separation Skim vessels 100-150 <500
API separators
Skim pile (as low as 100)
Plate coalescer Parallel plate 30-50 < 100
Coirugated plate
Gas flotation Dissolved gas 15-20 <30
Induced gas
Enhanced coalescence Various 10-15 < 20 (after hydrocyclone)
Enhanced gravity Hydrocyclones 5-15 < 50 (often < 30)
Centrifuges 2-5 < 10
Combined Coalescing pads > 2 < 10
Filtration Multimedia > 1 < 5
Membrane <1


units, or for the removal of specific components dispersed in the water. Some examples include the fol-
Mare's Tail
Epcon CFU
Versa Flo

This is not intended to be an all inclusive list, but one which gives the reader a flavor of the
practical efforts to improve separation of oil from produced water.

This is a patented extraction process in which 0.5 to 2.0% of liquid condensate (typically from
the scrubber vessel in the gas-compression train) is mixed with the separated produced water, upstream
of the hydrocyclones/ 6 i2 ^(see Figure 6.25) This condensate extraction helps to remove some of the
"oil" components dissolved in the water and enhances dispersed oil removal in the downstream
hydrocyclone, by helping to coalesce the finely dispersed oil droplets. The critical process parameter
is the injection and mixing of the condensate with the water, which is achieved using special in-line in-
jection mixers. Another requirement is that there must be adequate residence time (3-5 seconds) for
the extraction and coalescence to take place. Greater than 90% PAH removal has been achieved, along
with 90% reduction in dispersed oil concentration compared to the normal hydrocyclone operation.




Mixers /] 2nd Stage [\


Water Oil
CTour Standard
Process Process



Figure 6.25 Standard CTour Process (Courtesy of Prosep Technologies, inc.)



Specific process requirements include:

condensate must remain liquid during the injection and extraction stages
any residual condensate should evaporate at atmospheric temperature and pressure'' 6 -^

Field trials in Norway' 7 5 "' achieved a 52% reduction in EIF (environmental-impact factor) and
several full sized units are now being installed on platforms. As well as the process benefits, CTour
offers low weight, height, footprint, CAPEX and OPEX. Apparently, it also reduces the overall chemi-
cal usage.

A development of the basic process which shows a potential for process enhancement is the
"downstream" configuration (see Figures 6.26). Apparently the advantages include:
a significant reduction in the required volume of condensate
much lower oil content in the discharged water
more consistent system perfonnance
One potential problem has been noted. If there is a high BTEX concentration in the condensate,
some of this will transfer to the treated water. However, this problem may be overcome by pretreat-
ment of the condensate stream in a rectifying c o l u m n / 6 - 5 '




200-1000 ppm O-in-W

Downstream Process
Hydrocyclone Water Oil
10-100 ppm
Cond. rate 0.2-0.5%

< 3 ppm O-in-W

Figure 6.26 CTour Downstream Configuration (Courtesy of Prosep Technologies, inc.)

Mare's Tail
As previously stated, hydrocyclones often do not perform efficiently because many of the inlet
oil droplets are too small. The Mare's Tail is a coalescing device consisting of an inline installation of
stranded polypropylene media (its name is indicative of its appearance). Passing the produced water
over this medium upstream of the hydrocyclone, increases the average oil droplet size and hence

'CHAPTER 10 273

improves the oil removal efficiency of the hydrocyclone. PAHs and alkylated phenol components
within the dispersed oil (they were previously thought to be water soluble) were also r e m o v e d / 6 '^ In
one trial the hydrocyclone oil removal efficiency was typically about 40%. This increased to 65%
with the Mare's Tail coalescer installed inline upstream.

This is also a coalescing device, with the media contained in cartridges that are inserted into the
hydrocyclone inlet chamber. The cartridge consists of three modular stages of media (for droplet cap-
ture, droplet growth, and droplet release) and looks like thick discs of structured packing. One field
trial showed mean oil droplet size increasing from a typical 7 pm at the hydrocyclone inlet, to 30 |im
with the PECT-F cartridge in place.

The Hydrollokk system (see Figure 6.27) consists of chemical and mechanical coalescence prior
to a hydrocyclone. Coagulant is added to the water to overcome the electrostatic charges on the oil
droplets. After mixing in the upper section of the vessel, flocculent is added as the water passes into
the lower coalescence section. This produces larger droplets in the feed stream to the hydrocyclone
and tests have shown oil removal efficiency increasing from 75% (without Hydroflokk) to 98%.

Produced Water
from Inlet
Separator \nmm 1 .. Motor YSS^-lh-
Static Mixer I q l

Coagulant Flocculant

Oil Reject < X , Produced Water

to Re-Injection or
Hydrocyclone Overboard

Figure 6.27 Hydroflokk System

The Epcon CFU (Compact Flotation Unit) combines centrifugal force and nitrogen gas flotation
(see Figures 6.28 and 6.29).

Residence time is up to 1 minute and flocculation chemical can enhance performance. No exter-
nal energy is needed and the vertical vessel has a small foot-print. Since there are no moving parts,
maintenance requirements are apparently minimal.

In several field tests on North Sea platforms/ 6 '"' dispersed oil was reduced by 50-75%. Several
key aromatic materials were removed to varying extents and the EIF was reduced by 20%. In ongoing
applications on some North Sea installations, the discharged water oil-in-water levels are varying be-
tween less than 10 ppm up to 15 ppm.




Figure 6.28 Epcon CFU Figure 6.29 Plan View Showing Tangential Inlet

The VersaFlo equipment is a vertical column single-cell induced gas flotation unit used for pol-
ishing treatment of produced water. Its inlet device helps to pre-coalesce oil droplets. Having a low
downward water flow velocity, the unit displays high separation efficiency/ 6 , - 4 ) A two-stage oil skim-
ming system enables the VersaFlo to cope with dynamic conditions (platform movement). With an in-
let concentration of 70-100 ppm oil-in-water from upstream treatment units, VersaFlo can typically
achieve an outlet of 18-25 ppm oil-in-water.

TORR is short for Total Oil Remediation and Recovery and combines filtration, coalescence
and gravity separation. The key to this process is RPA (Reusable Petroleum Absorbent), a patented
highly oleophilic / hydrophobic thermo-sct polymeric backbone material. This is packaged in a series
of absorbent pads followed by recovery compartments.

As oily water passes through the absorbent pads the emulsified oil droplets absorb into the RPA.
More oil absorbs until saturation is achieved. Subsequently the flow regime causes desorbtion of
larger (coalesced) oil droplets from the RPA surface. In the next empty compartment the oil droplets
rise and are recovered. The water then passes through further stages of RPA and recovery compart-

Approximately 15% to 20% of the flow through the TORR is returned to the separator so that
the hydrocarbon can be recovered. The system will remove dispersed oil droplets of 2 microns and
greater diameter, producing treated water with less than 10 ppm non-soluble hydrocarbon content.

Water Soluble Hydrocarbons

Water soluble hydrocarbons consist of a wide range of organic chemicals, including simple
aliphaties, simple aromatics (BTEX), carboxylic (naphthenic) a c i d s / 6 - - ' polynuclear aromatic hydro-
carbons (PAHs), phenols, alkyl phenols and long-chain fatty a c i d s / 6 - 6 ' These have varying solubility


in water, with the absolute solubility of each component being affected by a complex mix of factors in-
cluding water salinity, pH, temperature and the quantity of free oil in the mixture. However,
carboxylic acids tend to be the predominant component of the water soluble fraction.' 6 --'

The important point to note is that the officially recognized methods of analyses have previously
not accounted for water soluble hydrocarbons. Once we begin to include them in results it gives the
impression that oil discharges have increased. As an example, one field (anonymous) has an average
of 20 ppm of soluble hydrocarbon in its discharged water. This is typically accompanied by 15ppm of
dispersed oil. For a discharge limit of 30 ppm "oil-in-water", the 15 ppm is well within that limit.
However, the combined 35ppm would exceed the limit unless changes are implemented.

For most locations it is neither environmentally desirable, nor politically expedient, to simply de-
clare a higher limit. In line with efforts to reduce overall "oil" discharges, producers are either using
or considering use of methods for removing or reducing the amount of soluble hydrocarbons in the dis-
charged water. Methods available include the following:

1. Acid treatment (sometimes termed "acid springing")

The water pH is reduced, typically with phosphoric acid (or sulfuric), so that the salts of
the fatty acids are converted to their oil-soluble acid form. Unfortunately, this will cause
precipitation of phosphate scales when di-valent ions such as calcium are present at reason-
able concentrations. As might be expected, a side effect of adding these strong acids is
corrosion of steel surfaces.

A more recent development' 6 --' using activated hydrophilic acids, can be cost effective in
transferring carboxylate anion type soluble organics from the water to the oil phase. This
is apparently achieved without corrosion or scale deposition problems.
2. Steam stripping

This process is a high energy consumer and invariably leads to scale formation unless a
scale inhibitor is used. However, it can remove greater than 90% of BTEX components.

3. MPPE (Micro Porous Polymer Extraction)

The water is passed through a bed of porous polymer beads, each of which contains an ex-
traction solvent. A two-bed system is used (see Figure 6.30), alternating between extrac-
tion (about one hour) and then regeneration. The MPPE bed should be replaced about ev-
ery 1 -2 years, to avoid loss of effectiveness.

While the removal of dispersed oil is not the primary aim of the MPPE process, hydrophobic
aliphatic compounds are removed by the extraction system. Hence, the removal efficiency of dispersed
oil is very much dependent upon the aliphatic component mix. Longer hydrocarbons than CSo will
pollute the MPPE material. Otherwise, the MPPE performance is expected to give:

BTEX removal efficiency = 99%

PAH removal efficiency = 99%
(Dispersed aliphatic removal = 50-99%)

The MPPE process is a relatively low energy consumer, although energy is required for raising
steam. It has a small "footprint", produces no waste and does not need chemicals. However, there is a
need to produce demineralized water for feed to the steam plant.

At higher temperatures the extraction performance docs decrease. However, testing at influent
temperatures up to 60C [140F] have not shown any detrimental e f f e c t / * - ^



Water Steam

Extraction Regeneration


> Hydrocarbons
Water + Steam
Hydrocarbons Condensate

Figure 6.30 Micro Porous Polymer Extraction

Related Aspects
Some soluble components (such as phenols) may be wholly or partially removed by other meth-
ods including, nano-filtration, adsorption, biological treatment and oxidation.

It should be noted that the various forms of gas flotation systems do give rise to some gas strip-
ping of the soluble hydrocarbons, particularly those which are more volatile.

Typical Produced Water Treatment Systems

Equipment Selection Criteria

With respect to offshore produced water treatment, especially with the increase in deepwater de-
velopments, process equipment selection is likely to be based on a combination of the following fac-
Small footprint
Compact build
Light weight (diy and "wet")
Not sensitive to wave motion
Efficient oil removal
Modular (re expansion)
Easy to retro-fit
Low shear (does not enhance emulsions)
Low. or no, chemical use
Not sensitive to flow or oil concentration surges
Not adversely affected by solids


Low pressure requirements

Low maintenance
Low waste generation
High turn-down rate (for varying water production rates)
Minimized operating costs

The first four items on this list are unlikely to play a part in onshore equipment selection.

An additional factor to consider is that of interferences in the operation of the various pieces of
equipment and whether they are likely to be present within the system under consideration. Bansal
and Caudle^-- have discussed interferences with respect to some commonly used equipment. Some
of the typical interferences are listed as follows:
Scale formation (due to carbon dioxide coming out of solution)
Solid iron corrosion by-products
Non-produced fluids within the system (e.g. completion and workover fluids, stimulation
High levels of soluble "oil" (not removed by most physical processes)
Significant process flow surges (both in volume and oil concentration)

At the design stage, taking these into consideration along with the desired requirements from the
list above, can facilitate their control in the actual operating system.

System Process Flow Diagrams

The following diagrams (Figures 6.31-6.36) show some typical produced oily water treatment
equipment combinations which are in use around the world.



- * Gas

HP Separator
D Gas

O/W = 800 ppm

LP Separator
-> Oil

H2O *] Electrostatic \
O/W = 300 ppm I Desalter J

O/W = 520 ppm Oil (fiberglass Media-

Coaiescers Candle Type)

Surge Solirts
Drum Removal 0/W= 50 ppm

Open. Skimmed
Settling Tank
Discharge to
0 / W = < 10 ppm

Figure 6.31 Onshore European Oilfield

Disposal Well
O/W 500 ppm

time 50-60 min)

Figure 6.32 Onshore Middle East (Persian Gulf) Oilfield


(Specificalion 40 ppm Oil-in-Water

nr reducing to 30 ppm)

HP Separator
1 System @ 70-90C

645 ppm IP Separator

dso 14 (im
300 ppm LZ*"
Oil-in-Water L LP Separator
d5o 11 urn
- a
- ^ 133 ppm . - Q P E x p o r t Oil
33 ppm dso 8 nm
Oil-in-Water 35 ppm
dso 5-7 (am Oil-in-Water 37 ppm; d 5 o 3 |am
dso 5 pm
f . ^ Degasser

33 ppm; dso 5 n m

Water Discharge
33 ppm; dso 5 |am

Figure 6.33 Typical North Sea Platform

1000 psi

(\ HP |\
\ Separator J

550 psi
> / | Medium l \
Oil + H 2 0 Pressure
V. Separator J Gas

Oil + H 2 0
75 psi > Oil

H 2 0 (Oil-in-Water = 850 ppm)

CPI > Flotation

X /
Oil-in-Water =
400 ppm

(Oil-in-Water = 50 ppm)

Figure 6.34 Typical Gulf of Mexico (GOM) Platform



o HP Separator

D IP Separator

Gas Flotation
1000 p p m ;
LP Separator A

Unit dso 10-15 n m

( ) Degasser
20-50 p p m ; Heating
dso 5-8 p m Medium ir

Effluent Water > Oil
15-25 p p m Oil-in-Water
100-300 p p m Oil-in-Water

Figure 6.35 GOM Deepwater Oilfield

2-Stage Desalting'
60C O-in-W Discharge Limit = 30 ppm
9.4 bar

700 ppm

O-in-W Water
70 ppm

45 ppm

Figure 6.36 West Africa Offshore System


6.1 Veil, John A, "An Overview of Produced Water in the United States", Produced Water Workshop,
Aberdeen, 21st/22nd April 2004, NEL, Glasgow; and,
6.2 Laurie, Angus, "Implementation of the Offshore Petroleum Activities (Oil Pollution. Prevention and Con-
trol) Regulations (OPPC) 2004", Produced Water Workshop, Aberdeen, 21st/22nd April 2004, NEL, Glas-
6.3 Johnsen, Stale and Utvik. Toril Roe, "The Zero Discharge Work - Results and Follow-Up". Produced Wa-
ter Workshop. Aberdeen, 21st/22nd April 2004, NEL, Glasgow.
6.4 Bakke. Siri, et al, "Produced Water Reinjection (PWRI) - Experiences from the Ula Field", in "Produced
Water 2: Environmental Issues and Mitigation Technologies". (1996), edited by Reed and Johnsen. Ple-
num Press, New York.
6.5 OSPAR Commission, "Background Document Concerning Techniques for the Management of Produced
Water from Offshore Installations", (2002), OSPAR Commission, ISBN 0 946956 87 1.
6.6 Hirayama, Akihiko, et al, "Omani Oil Fields Produced Water: Treatment and Utilization", SPE 74413,
SPE International Petroleum Conference, 10-12 February, 2002, Villahermosa, Mexico.
6.7 Veil, John A, "An Overview of Applications of Downhole Oil / Water Separation Systems", Produced
Water Workshop. Aberdeen, 21st/22nd April 2004, NEL, Glasgow.
6.8 Daneshy, Ali and Donnelly, John, "The Funding and Uptake of New Upstream Technology", Journal of
Petroleum Technology, June 2004.
6.9 Hintermaier, R.W., "Production Water Management", SPE 38798, 1997 SPE Annual Technical Confer-
ence, 5-8 October, 1997, San Antonio. Texas.
6.10 Hudgins, C. M. Jr.: "Chemical Treatments and Usage in Offshore Oil and Gas Production Systems," J.
Pet. Tech. (May 1992) 604.
6.11 Place, M. C.: "Dissolved Organic Compounds in Produced Water." SPE Paper 22780 (1991).
6.12 Rabalais, N. N., et. al.: Fate and Effects of Nearshore Discharges ofOCS Produced Waters, OCS Study,
MMS 91-0004-0006 (1991).
6.13 Stephenson, M. T.: "Components of Produced Water: A Compilation of Industry Studies," J. Pet. Tech.
(May 1992) 548.
6.14 Sport, M. C.: "Design and Operation of Dissolved-Gas Flotation Equipment for the Treatment of Oilfield
Produced Brines," J. Petr. Tech. (August 1970) 918.
6.15 Ellis, M. M. and Fischer, P. W.: "Clarifying Oilfield Refinery Waste Waters by Gas Flotation," J. Petr.
Tech. (April 1973) 426.
6.16 Bradley, B. W.: "Flotation Offers Another Water/Oil Separation Alternative," Oil & Gas J. (Dec. 9,
1985) 42.
6.17 Bradley, B. W.: Two Oilfield Water Systems. Steamfiooding. Oil Removal, Robert E. Krieger Publishing
Company, Inc., Melbourne, Florida (1986).
6.18 Strickland. W. T.: "Laboratory Results of Cleaning Produced Water by Gas Flotation," Soc. Petr. Engrs.
J. (June 1980) 175.
6.19 Flanigan, D. A., Stolhand, J. E., Shimoda, E., and Skilbeck, F.: "Use of Low-Shear Pumps and
Hydrocyclones for Improved Perfomiance in the Cleanup of Low-Pressure Water," SPE Production Engi-
neering (August 1992) 295.
6.20 Meldrum, N.: "Hydrocyclones: A Solution to Produced-Water Treatment," SPE Production Engineering
(Nov. 1988) 669.
6.21 Khatib, Zara, "Handling, Treatment and Disposal of Produced Water in the Offshore Oil Industry",
(1998), SPE 48992, SPE, Richardson, Texas.
6.22 Knudsen, B.L., et al, "Meeting the Zero-Discharge Challenge for Produced Water", SPE 86671, Interna-
tional Conference on Health, Safety and Environment in Oil and Gas Exploration and Production, Cal-
gary, 29th-31 st March 2004.



6.23 Sabey, John B., "CTour - Simultaneous Removal of Dispersed and Dissolved Hydrocarbons in Produced
Water", 18th-20th January, 2006, 16th Anual Produced Water Siminar, Produced Water Society, Houston.
6.24 Juniel, Kevin A, Frankiewicz, Ted and Lee, Chang-Ming, "Novel Vertical Flotation Design for Gulf of
Mexico Deepwater Floating Structures", 21st-23rd January, 2004, 14th Annual Produced Water Seminar,
Produced Water Society, Houston.
6.25 Hart, Paul R, "Removal of Water Soluble Organics from Produced Brine without Formation of Scale",
(2003), SPE 80250, Richardson, Texas.
6.26 Bansal, K and Caudle, D, "Interferences With Produced Water Treatment for Dispersed Oil Removal",
(1998), SPE 46576, SPE, Richardson, Texas.
6.27 Adamczyk, Pam and Meijer, D. Th, "Removal of Dissolved Hydrocarbons from Offshore Produced Water
with Micro Porous Polymer Extraction (MPPE)", 21st-23rd January, 2004, 14th Annual Produced Water
Seminar, Produced Water Society, Houston.
6.28 Di Lullo, G and Rae, P, "New Insights into Water Control - A Review of the State of the Art", (2002).
SPE 77963, Richardson, Texas.

Raw water treatment includes any procedure that prepares the source water for its ultimate use.
This might include any or all of the following:
clarification (gross suspended solids removal)
filtration (see Chapter 7)
dissolved gas removal (see Chapter 7)
precipitation softening
ion exchange
membrane processes.
Obviously the level and complexity of treatment will depend on the concentration of the various
"contaminants" in the source water and the final water quality desired for the specific process.

The destabilization and removal of suspended solids is facilitated by coagulation and flocculation
processes and ultimately achieved by physical separation. The coagulation (charge neutralization and
destabilization) and flocculation (agglomeration of fine particles into larger masses) steps may be en-
hanced with the use of either inorganic or polymeric organic chemicals.

Examples of the inorganic materials include alum (various forms of aluminum sulfate), ferric sul-
fate and sodium aluminate. When added to water these form hydrated voluminous floes of either alu-
minum or ferric hydroxides. These floes combine with the suspended solids and are removed as bot-
tom sediments from a separation basin. Details of the chemistry involved in these reactions are
available in several texts/ 4 ' 7 - 5 -'

The polymeric organic primary coagulants are cationic polyelectrolytes, typically based on
polyamine or polyacrylate chemistry. They act by both charge neutralization and bridging of particles.

Water softening is used to treat water prior to certain enhanced recovery processes and for boiler

The process of softening water is one where certain dissolved ions are removed from the water to
e v e n t the formation of insoluble scales or deposits. Removal of calcium, magnesium, carbonate and


bicarbonate ions from water are common examples of softening. Other ions may be removed if

Raw water may be softened by three methods:

1. Precipitation softening
2. Ion exchange softening
3. Distillation
Each of these methods can be used independently or in combination to soften water. The choice
of methods is influenced by the amount of softening required, the volume of water to be softened, and
the installation and operating costs for the process selected.

Precipitation Softening
Precipitation softening refers to the removal of hardness and alkalinity by adding a chemical to
the water which will combine with calcium, magnesium and bicarbonate ions and form an insoluble
precipitate. Thus, hardness and alkalinity are removed from the water.

Lime and Lime-Soda Softening

The most commonly used chemicals are lime (calcium hydroxide) or lime plus soda ash (sodium

When lime is used alone the process is called lime softening. Lime is added to the makeup wa-
ter, causing an increase in pH. Calcium, carbonate, and bicarbonate ions are precipitated as calcium
carbonate. If additional lime is added, magnesium ions will also be precipitated as magnesium hydrox-
ide. Appreciable silica removal may also be realized due to adsorption on the magnesium hydroxide.

When soda is used along with lime, additional calcium is precipitated as calcium carbonate pro-
ducing a lower hardness water. The process is referred to as lime-soda softening and may be classified
as hot or cold depending on the temperature of the water.

Hot process softeners produce lower hardness and alkalinity than the cold process, and the reac-
tions proceed more rapidly/ 5 -'

Hot lime-soda softeners can reduce the hardness of the water down to 15 to 25 ppm, while the
cold process is limited to 30 to 50 p p m / * "

Caustic Softening
Caustic has also been used to reduce the hardness of high TDS waters from about 500 to less
than 20 ppm in steamflooding operations in California/ 8 -^ Sodium hydroxide is added to the water to
raise the pH to about 9.3. Calcium and magnesium ions are precipitated as calcium carbonate and
magnesium hydroxide.

This process offers significant capital and operating cost savings over a conventional weak-acid
ion-exchange system, with an additional benefit of partial silica removal.

Compared with the lime process, it is said to be cost competitive, operationally more reliable, and
offers the additional benefit of producing only about 20% of the solid sludge volumes produced by the
lime process.



Ion Exchange Softening

Ion exchange softening is achieved by flowing the raw water through a vessel filled with small
spheres (called beads) of a material referred to as an ion exchange resin. The resin beads are usually
made from a synthetic plastic and are both permeable and porous. The porous beads contain many
electrically charged sites throughout their structure to which are attached ions or ionizable groups of
opposite charge/^ 7 -'

As the raw water flows through the bed, the ions initially attached to the beads are replaced by
ions from the water which have a higher affinity for the resin, and the ions originally attached to the
resin are expelled into the water. Thus, ions in the raw water are exchanged for ions attached to the
resin beads.

Once the bed is saturated with ions removed from the water, the bed may be regenerated by soak-
ing it in a solution containing a high concentration of the ions originally attached to the resin beads.
Ion exchange occurs once again, restoring the resin to its original condition.

Strong-Acid Ion Exchange Resins

A strong-acid cation exchange resin is usually a sulfonated copolymer of styrene and
divinylbenzene. The resin functions by exchanging sodium ions for calcium and magnesium ions and
this process is often referred to as sodium exchange softening. When water containing calcium and
magnesium ions is passed through a bed of sodium exchange resin, the calcium and magnesium ions
are removed from the water and replaced with sodium ions.

This system works well on oil-free water with T D S < 5000 ppm. When the TDS of the water is
higher, sodium ions in the produced water compete with calcium and magnesium ions for sites on the
resin. This makes it very difficult to reach hardness levels < 1 ppm, which are typically required for
steam generators.

When the beads become saturated with calcium and magnesium ions, it is regenerated by wash-
ing with a strong solution of sodium chloride. The high concentration of sodium in the solution re-
verses the exchange process, replacing the calcium and magnesium ions on the resin beads with so-
dium ions, making them once again ready for use.

Weak-Acid Ion Exchange Resins

A weak-acid cation exchange resin is typically a carboxylic acid group within an acrylic
divinylbenzene matrix. This type of resin exhibits a very strong selectivity for calcium and magne-
sium ions over sodium ions. Hence, it can produce very low hardness levels in brines and is com-
monly used to soften waters with TDS > 5 0 0 0 ppm.

The resin is considerably more costly than the strong acid resin, and requires a two-step regener-
ation. Hydrochloric acid is first used to remove the calcium and magnesium; then it is treated with
caustic soda (sodium hydroxide) to convert the resin back to the sodium form. The cost of this regen-
eration cycle is several times higher than the brine regeneration required for strong acid resins.

The two-step regeneration cycle is necessary to prevent the production of an acidic soft water. If
the resin is placed in service after being regenerated with acid alone, the resin will soften the water by
exchanging hydrogen ions for calcium and magnesium ions, thus lowering the pH. This is avoided by
conditioning the resin with caustic after acid regeneration. Sodium from the caustic exchanges with
the hydrogen in the resin. Then, when the resin is placed in service, sodium ions are exchanged for
calcium and magnesium ions.


Anion Exchange Resins

Other types of ion exchange units will remove anions from water. A weakly basic type will
move sulfate, chloride and nitrate. A strongly basic type will remove these ions plus silica, carbon di
oxide, and sulfides.

Demineralization is accomplished by using a hydrogen exchanger to remove cations and an anion
exchanger to remove anions. It is used to obtain high quality water when the raw water contains small
quantities of dissolved solids.

Both the initial investment and operating costs are high, making it difficult to justify demineral-
ization for most oilfield facilities.

Distillation is the most expensive method of softening and is used primarily to remove dissolved
solids from brackish waters to produce high quality feedwater. It is not widely used in oilfield installa-

Membrane Filtration
The separation of particles from liquids by normal filtration methods has a lower limit of approx-
imately 1-2 microns ( p i ) . For the removal of smaller particlcs, molecules, or ions, it is necessary to
use some fonn of membrane filtration. These membranes act as a barrier and their sub-microscopic
pores allow water and selected materials to pass through, dependent upon the process. A more concen-
trated residual solution is removed from the feed side of the membrane.

Membrane processes have been classified into four groups, loosely defined by the size range of
species removed from solution.' 5 "" As you will note in Table 8.1, there is some overlap between these

Microfiltration generally is only removing suspended materials. The other processes can remove
emulsified oil and dissolved molecules and ions, as shown in Table 8.2J S 4 >

To be more precise, ultrafiltration is capable of removing high molecular weight components.

Nanofiltration goes further and rejects ions with more than one negative charge (e.g. sulfate or phos-
phate), uncharged dissolved materials and some positively charged ions, depending upon their size and
shape. In Reverse Osmosis water should be the only material passing through the membrane, with all

The Four Membrane Processes
Approximate Size Typical Operating
Process Name Range Removed (pm) Pressure (bar)
Microfiltration 0.05-5 < 2
Ultrafiltration (UF) 0.003-0.1 1-10
Nanofiltration (NF) 0.0005-0.007 5-35
Reverse Osmosis (RO) 0.0001-0.001 15-150



Typical Membrane Process Streams
Process Name Treated Stream Permeate Concentrate
RO water solution low salinity water salty water
NF water solution softened water very "hard" water
UF oil emulsion oil free water concentrated oil emulsion
(< 10 ppm oil-in-water)
MF waste water clarified water waste product

dissolved and suspended matter essentially being rejected. l S J > A diagrammatic representation of these
processes is given in Figure 8.1.

Membrane Technology

The bulk of commercially used membranes consist of only a few basic materials. The original
membranes (used for RO, NF and UF applications) consisted of cellulose acetate. While being rela-
tively inexpensive and also being less prone to fouling, it has pH and temperature limitations. Cellu-
lose acetate is also digested by some microorganisms. Various polymeric materials are now used to
make membranes, including polyacryonitrile (PAN). Treatment of this latter material renders it so hy-
drophilic that oils and greases will not blind or foul the membrane/ 5 -'''

In the oil and gas industry membranes are most commonly used in a spiral wound format (Figure
8 . 2 ^ )


J Bacteria +
Fine solids

Microfilter (0.5 - 5 pm)

Oil Emulsions
Ultrafilter ( 0 . 0 0 3 - 0 . 1 pm)
Divalent anions
Nanofilter (0.0005 - 0.007 pm)
.Salts + metal ions

Reverse Osmosis (0.0001 - 0 . 0 0 1 pm)


Figure 8.1 Membrane Processes


Perforated Central Tube

Permeate Collection Material

Feed Channel Spacer
Outer Wrap

Figure 8.2 Spiral Wound Membrane Construction (Courtesy of GE Infrastructure Water & Process Technologies)

Typically these spiral wound elements have been limited to operating temperatures of less than
45C [~115F], which is not a problem for water de-salination. Where necessary, however, membranes
are now available with higher temperature resistance.

Since membranes are primarily designed to remove dissolved materials from the feed stream, per-
formance is adversely affected by significant amounts of suspended solids. There is also the need to
avoid oxidizing chemicals such as sodium hypochlorite which may well have an adverse reaction with
the membrane. Hence, optimum performance depends on adequate pre-treatment of the feed water.
This usually consists of one or more of the following steps:
suspended solids removal (pre-filtering)
removal of any oxidizing chemicals (e.g. chlorine)
actions to prevent solids precipitating within the membrane process.

Specific Oil and Gas Membrane Technology Applications

Where seawater injection is required for secondary recovery or pressure maintenance, the pres-
ence of significant levels of calcium, barium, or strontium ions in the fonnation water can lead to min-
eral scale deposition. This is due to the levels of sulfate ions present in the seawater. In one North
Sea situation the formation water contained v e i y high levels of barium ions ( 2 6 5 0 mg/L), giving a high
potential for barium sulfate scale precipitation/ 5 ^ Due to the level of scaling anticipated, conven-
tional chemical scale inhibition was expected to be both expensive and difficult to ensure.
Nanofiltration was applied to significantly reduce the level of sulfate ions (to ~ 8 0 mg/L) passing into
the formation/ 5 -^ At this reduced level of sulfate, conventional scale inhibition was then feasible.

A similar situation existed in an offshore Angola field, where the fonnation brine contained about
2 3 0 mg/L barium and about 225 mg/L strontium. Since the sea water contained about 2860 mg/L of



sulfate a major scaling problem was anticipated. Nanofiltration of the seawater produced a permeate
for injection which contained less than 40 mg/L sulfate. In the Campos Basin offshore Brazil, there
are formation waters with 40-200 mg/L barium and 200-600 mg/L strontium. Hence, several of the de-
veloping fields will use nanofiltration sulphate removal plants. Even so, to prevent sulfate scaling the
level of sulfate in the LSSW (low sulfate sea water) must be reduced down to about 20 mg/L/ 8 - 7 '

With increasing offshore developments in deep waters, the long tie-backs to production facilities
and practical difficulties with respect to any remedial work, have resulted in an emphasis on flow as-
surance. Therefore, it is likely that we will see increased use of nanofiltration technology.

Reverse Osmosis
The process of osmosis takes place throughout nature, whereby water can flow across a semi-per-
meable membrane that otherwise will not allow the passage of dissolved materials. Where a semi-per-
rncable membrane separates two solutions, water will pass from the less concentrated side to the more
concentrated solution, in an effort to balance the system. In Figure 8.3 we have a salt solution on one
side of the membrane and distilled water on the other side. Due to osmosis water passes through the
membrane and dilutes the salt solution. Water transfer stops when the osmotic pressure is balanced by
the hydrostatic head of pressure.


Solution Membrane

^ U&Jtr

Figure 8.3 Osmosis and Osmotic Pressure Experiment

Note: if only water was present in the system represented by Figure 8.3, then the fluid level would be
the same in both tubes.

From the van't Hoff formula, osmotic pressure (P 0 ) is:

Po = CRT = p h
Where: C = molar salt concentration
R = the gas constant
T = the absolute pressure
p = specific gravity of water
h = the difference in water levels (see Figure 8.3)


Osmosis is a reversible process. By applying a pressure that exceeds the osmotic pressure, water
is made to pass through the membrane to the less concentrated solution. This process is called reverse
osmosis. When seawater is pumped under pressure through a semipermeable membrane, a desalinated
"permeate" is produced. A higher concentration of salts remains in the rejected seawater side of ibe
process (Figure 8.4).

Figure 8.4 Reverse Osmosis

In Figure 8.4, P p is the applied pressure from a pump. The pressure applied has to overcome the
osmotic pressure (26-27 bar for seawater) and supply the overpressure needed to push water through
the membrane. In our diagram;


Where: hr = the difference in water levels in Figure 8.4

The ratio of volume of desalinated water to the volume of feed seawater is the recovery ratio:

Recovery Ratio =

Quite often in seawater desalination processes a recovery rate of 0.5 is used, with half of the feed
becoming permeate. However, this may be achieved via several stages of membrane separation.

Semi-permeable membranes provide optimum performance with continuous water flow and stable
operating pressure/ 6 '^ Varying pressures and intermittent flow tend to stress the membrane and
shorten its operational life. Also, membranes are not perfect and a fractional amount of some small
ions may penetrate with the water/ 5 S >

Prior to the reverse osmosis plant, seawater is typically screened to remove macro-organisms and
solids, then chlorinated (to about 0.5 ppm residual chlorine) to kill micro-organisms. Subsequently the
water is filtered through a dual media of activated carbon and garnet. The activated carbon removes
the residual chlorine, thereby preventing any membrane damage.



Potable (Drinking) Water

Potable water (drinking water) and its preparation is not a feature of this book. However, oil and
gas production is often located in remote areas (including offshore facilities) and the provision of
drinking water will be necessary to enable daily operations to continue. Suffice to say that some of the
processes mentioned in this book are utilized in the processing of raw water to make potable water
supplies. Such processes, depending on the raw water supply, include filtration, softening, reverse
osmosis and chlorination.

While some countries have their own specific potable water quality standards, many states and
production companies simply follow the guidance provided by the World Health Organization
( W H O ) / 5 p> Often of interest is a measure of a water's hardness. A commonly used classification is
shown in Table 8.3.

Water Hardness Categories

Hardness (mg/L as CaCOj)
Sort 0-17
Slightly Hard 17-51
Moderately Hard 51-120
Hard 120-180
Very Hard > 180

8.1 Betz: "Betz Handbook of Industrial Water Conditioning." 9th Ed., Betz, Laboratories, Inc., Trevose.
Pennsylvania. USA (1991).
8.2 Kemmer, F. N Editor: The Nalco Water Handbook, McGraw Hill, New York (1979).
8.3 Jan, R. J. and T. G. Reed. Jr.: "A New Caustic Process for Softening Produced Water for Steam Genera-
tion," SPE Prod. Engr. (May 1992) 199.
8.4 Wagner, Jorgen, "Membrane Filtration Handbook - Practical Tips and Hints," 2nd Ed. (Nov. 2001),
Osmonics Inc., Minnetonka, Minnesota, USA.
8.5 GE Water Technologies, "Membrane Filtration of Oil and Gas Field Produced Water" at
8.6 O'Donnell, Keith, "Desulfating Injection Water - Membrane Technology Works on North Sea Platfonn,"
Oil & Gas J., PennWell Corporation, Vol. 94, Tulsa, Oklahoma. USA (Dec 2, 1996). p. 58-63.
8.7 Simpson, et al., "Sulphate Removal for Barium Sulphate Scale Mitigation - Kinetic vs. Thermodynamic
Controls in Mildly Oversaturated Conditions," SPE 95082, SPE Inc. (2005), Richardson. Texas, USA.
8.8 Lachish, Uri, "Optimizing the Efficiency of Reverse Osmosis Seawater Desalination," (Sept. 2002), at
8-9 WHO, "Guidelines for Drinking Water Quality," 3rd Ed. (2004), WHO. Geneva. ISBN 92 4 154638 7
Also available at: http://www.who.int/water_sanitation_health/dvvq/gdwq3/en


Estimated Water Usage for EOR P r o c e s s e s ^
Water Usage
Process (bbl water/bbl oil)
Polymer 16-50
Surfactant Polymer 10-15
Alkaline (Caustic) 22-33
Carbon Dioxide 1-3
In-situ Combustion (Wet) 0.5-1
Steam 2-5

EOR Oil Production

Various estimates have been reported^9 regarding the volume of oil production from EOR
projects. The variations in numbers between these texts can be explained by the way the authors have
grouped the different types of EOR methodology. In some reports, oil derived from processes produc-
ing relatively small volumes, appears to have been omitted from the totals. Some of these estimates
for oil production are summarized in Table 9.2, (where bopd means "barrels oil per day").

Estimated Oil Production from EOR Projects
1992 (bopd)"-'^ 1998 (bopd)' 9 2002 (bopd)' ?2<;
World World World
Process Total US Total US Total US
Thermal 777 000 454 000 n/p 419 349 n/p 317 000
Miscible 702 000 191 000 n/p 281 000 n/p 298 000
Chemical 80 600 11 900 n/p n/p n/p n/p
Total 1574 600 656 900 > 2000 000 | 759 965 1530 000 669 000
Note: n/p = not provided

However, due to its consistency of approach, data from the "Oil and Gas Journal" biennial sur-
v e y s of worldwide EOR activities is probably the best for comparing trends. A recent survey' 9 " 1 '
showed that the highest number of EOR projects were in U S A , Venezuela, Canada, China and Trini-
dad. The largest EOR project, using steam, produces about 220 000 bopd in Indonesia, whereas the
largest US application is in California and produces about 100 000 bopd. However, these should be
compared with the data in general.

(9 J6>
Figures released in 2002 for the "rest o f the world" showed oil production of 861 000 bopd
coming from about 165 EOR projects. Equivalent US figures showed 669 000 bopd from 147 pro-
jects, of which only 17 produced more than 10 000 bopd. In fact, those 17 systems produced about
498 000 bopd. Hence, this data pointedly shows that the majority of EOR projects are relatively small
oil producers.

In the U S A the majority of steam floods are in California, where heavier oils are produced.
Likewise, the main zone for COT floods is N e w Mexico, Wyoming and the western areas of Texas and
Oklahoma, where they have access to adequate sources of carbon dioxide. Polymer related E O R has
almost disappeared in the U S A , but it still represents about 2 0 % of projects throughout the rest of the



One author believes that "EOR is producing much less oil than originally anticipated, because of
scientific, technical and economic constraints"/ 9 -^ Also, such projects can apparently cost as high as
$20-30 per barrel of oil produced.' 9 - 5 '

A polymer flood is earned out by injecting an aqueous polymer solution into the reservoir to dis-
place the oil from the rock into producing wells. The polymer increases the viscosity of the solution
which increases its ability to displace viscous crude oils. Other mechanisms are also involved, but
they will not be discussed.

The composition and properties of the water which is used to make up the polymer solution are
quite important, as they affect the viscosity of the solution. Also important are the agents which cause
polymer degradation and decreases in solution viscosity, which can cause a reduction in displacement

Two basic types of polymers are currently the most widely used: polysaccharides and

The polysaccharide typically used in enhanced oil recovery processes is xanthan gum and is often
referred to as a "biopolymer." It has a molecular weight of about 5 million.

Xanthan gunt is produced by the microbial action of xanthomonas campestris on a carbohydrate

substrate. The biopolymer is an extracellular slime which forms on the surface of the bacterial cell.
The fermentation broth is pasteurized to kill the xanthomonas campestris, and the polymer is precipi-
tated from the broth by a suitable alcohol.' 9 5 '

Xanthan gum solutions exhibit relatively stable viscosity properties as a function of salt concen-
tration, pH, and temperature/ 9 ^

Microbiological Problems
Once a biopolymer is dissolved in water, it is highly susceptible to bacterial attack by both aero-
bic and anaerobic microorganisms, resulting in a significant reduction in solution v i s c o s i t y / 9 - - * - 9 , I n
addition, some aerobic slime forming bacteria produce biological masses which contribute significantly
to formation plugging.' 9 -"

Water Salinity Effects

The viscosity of xanthan solutions increases with increasing brine salinity. The presence of
multivalent cations has no additional effect on viscosity beyond their contribution to total salinity/ 9 - 15 '

Dissolved Oxygen
At room temperature, dissolved oxygen has no apparent effect on the viscosity of biopolymer so-
lutions. However, at higher temperatures, enormous losses in viscosity are observed. (9- '

PH Effects
The maximum fresh water solution viscosity has been observed at a pH of 5.5. Polymer viscos-
% is highest in high or low pH solutions/ 9 *'


The polyacrylamide molecule is a long chain made up of acrylamide monomer molecules. It hi
a basic structure as follows:

- CH - CH 2 CH CH 2 - CH-
i1 i1 i1
C o c = o c = o
11 11 1
NH 2 NH 2 NH2

Figure 9.2 Basic Structure of Polyacrylamide

There are several mechanisms involved in the formation of the polymer molecule, which results
in a wide range of chain lengths. The average molecular weight of commercial polyacrylamides typi-
cally ranges from 1 to 10 million.

Polyacrylamide is usually modified by chemically replacing some of the amide groups with car-
boxyl groups. This process is called hydrolysis, and the percentage of amide groups which have been
replaced by carboxyl groups is referred to as the percent hydrolysis. The percent hydrolysis typically
ranges from 0 to 30%. The primary reason for carrying out partial hydrolysis is to increase the viscos-
ity of the polymer solution. The apparent solution viscosity of a partially hydrolyzed polyacrylamide
dissolved in fresh water is much greater than that of a corresponding unhydrolyzed polyacrylamide/ 9 -^

The carboxyl group ionizes in water, leaving a negative charge (Figure 9.3). The net result is a
series of negative charges along the polymer chain, creating a negatively charged polymer/ 9 - 9,1 This
type of polymer is often referred to as an anionic polyelectrolyte.

i1 I1 i1
c = o c = o C = 0
11 11 11
NH2 0 - NH2

Figure 9.3 Partially Hydrolyzed Polyacrylamide

Polyacrylamides can be manufactured in the field using small portable plants.

Effect of Sodium, Calcium and Magnesium Ions

Increasing concentrations of sodium ions result in decreased viscosity of partially hydrolyzed

polyacrylamide solutions. This is illustrated for a 0.25% solution of partially hydrolized
polyacrylamide in Figure 9.4.

This occurs because of neutralization of the electrical charges within the polymer molecules by
the oppositely charged sodium ion. As the net charge on the polymer molecules decreases, the repel-
lent force between molecules diminishes. The polymer chains coil up into a ball, thus causing a de-
crease in solution viscosity/ 9 - 9 ' 9 " 9,9 "^''



2 NaCI =
10 ppm 3
100 ppm *
1,000 ppm

- 5,000 ppm
10,000 p p m
20,000 ppm

1 1 1 1 III! 1 1 1 1 llll
1 10 10 2 103 104
Shear Rate, sec

Figure 9.4 Effect of NaCI Concentration on Polymer Solution V i s c o s i t y ^

Both calcium and magnesium ions cause larger decreases in viscosity than sodium i o n s / 9 p ' Cal-
cium ions cause a greater reduction in viscosity than an equivalent amount of magnesium ions, and
mixtures of calcium and magnesium ions cause even greater reductions than either ion alone. The ef-
fect of salts on the viscosity of a 2 5 0 0 ppm solution of partially hydrolyzed polyacrylamide is shown
in Figure 9.5.

Figure 9.5 Effect of Salts of the Viscosity of a Polymer Solution


Effect of Oxygen
The presence of dissolved oxygen has a number of negative effects. It can cause:
Dramatic increases in the corrosion rate of any carbon steel in the system.
Precipitation of dissolved iron.
Support the growth of aerobic microorganisms.
Significant polymer degradation. Partially hydrolyzed polyacrylamides are very sensitive
to oxidation, especially in higher temperature reservoirs/ 9 - 9 - 9 - 7 -^

Microbiological Problems
Polyacrylamides are relatively immune to bacterial attack/ 9 -^ However, the presence of aerobic
microorganisms can contribute significantly to corrosion and plugging problems, and sulfate reducinn
bacteria have been observed as a contributor to corrosion problems in polymer injection systems/ 9 - 1 S )

pH Effects
Lowering the pH of the polymer solution by adding a small amount of acid lowers the viscosity
of the solution. Acid exerts a greater influence on the shape of the molecules than an equivalent
amount of a neutral salt. The hydrogen ions convert the ionic carboxylate groups in the molecules to
uncharged carboxylic acid groups/ 9 , l(>>

Effect of Oxygen Entry

If oxygen is excluded, the presence of dissolved iron ions appears to have only a slight effect on
solution viscosity. However, if oxygen entry is permitted, a rapid decline in viscosity o c c u r s / 9 7 ^

Oxygen entiy also results in the precipitation of iron ions from solution as insoluble ferric hy-
droxide. It is an excellent plugging agent by itself, but in the presence of high molecular weight poly-
mers, the problem is considerably more severe.

Water Treatment for Polymer Flooding

The water used to make up polymer solutions should be as clean as possible.

Polymers tend to adhere to solids and agglomerate. When produced water is used, oil canyover
is not generally considered an acute problem, but should be minimized because of relative permeability
considerations and the fact that oil also helps to agglomerate solids and aggravate plugging.

Filtration and Oil Removal

Filtration is frequently necessary, and dispersed oil removal is nearly always necessary when pro-
duced waters are used.

Permissible levels of suspended solids and oil are subject to discussion. One prominent chemical
supplier advocated 1 ppm suspended solids and 1 ppm o i l / 9 ^ A major oil producer in West Texas
quoted polymer supplier recommendations that the water should have less than 20 ppm of hydrocar-
bons, and no solid particles over 2 microns/ 9 - 1 6 )

It may also be necessary to filter the polymer solution to remove any gelled agglomeration of
high concentration polymer that might result in injection well plugging.



Plastic piping and plastic or coated steel vessels and equipment should be used to prevent corro-
sion and the generation of iron compounds.

Bacterial Control
The water should normally be treated with a biocide prior to mixing with the polymer/ 9 - 7 J ) Lab-
oratory testing should be conducted to ensure that the biocide will not degrade the polymer/ 9 7 7 '

The use of an effective biocide is absolutely essential for polysaccharide polymer solutions and is
strongly recommended for polyacrylamide solutions.

Dissolved oxygen should be excluded to prevent the growth of aerobic bacteria.

Dissolved Oxygen Removal

Oxygen must be excluded or removed from polyacrylamide solutions to prevent degradation. It
is best to eliminate all oxygen before the polymer is added, and then add the polymer in such a way
that oxygen is not introduced. This minimizes the need for oxygen scavengers in the polymer solution.

Oxygen exclusion is also recommended for polysaccharide polymer solutions, although degrada-
tion only occurs in high temperature applications.

Oxygen also accelerates the corrosion of carbon steel, causes the precipitation of soluble iron
from solution, and facilitates the growth of aerobic bacteria.

The use of oxygen scavengers in polymer solutions presents a number o f unusual problems, and
they must be used with care. It is usually best to remove dissolved oxygen from the water before the
polymer is added, and then try to prevent air entry during polymer mixing.

One of the problems with the use of conventional sulfite oxygen scavengers is that, given a
choice between reacting with sulfite and ferrous ions, oxygen will first oxidize the ferrous iron to in-
soluble ferric iron. However, oxygen reacts with sodium hydrosulfite in preference to the ferrous ion.
In addition, it is such a powerful reducing agent that it is able to re-solubilize ferric hydroxide which
has already formed in the system.

For this reason, many early polyacrylamide polymer floods used sodium hydrosulfite (sodium di-
thionate) as an oxygen scavenger/ 9 - , 3 > The problem with this material is that while stable in powder
form, it deteriorates within a few hours when dissolved in water/ 9 77 '

Because of the difficulties in applying a powdered product in the field, it is more common to
control the ferrous iron content of the water and use sodium or ammonium bisulfite oxygen scaven-
gers. The problem in their use arises from the fact that they usually must be catalyzed with divalent
metal ions, which can contribute to polymer degradation/ 9 -' '

Although dissolved oxygen does not cause degradation of polysaccharides at normal ambient
temperatures, it does result in serious degradation at elevated temperatures. Unfortunately, neither sul-
fite or sodium hydrosulfite oxygen scavengers can be used to remove dissolved oxygen from
polysaccharide polymer solutions, because they produce free radicals as intermediate products that can
initiate biopolymer degradation. Thus oxygen removal prior to mixing is necessary. The use of anti-
oxidants may also be required./ 9 7 '

All vessels should be blanketed with nitrogen.


Many waters contain very small amounts of natural iron. When this is the case, oxygen exclusion
and the use of plastic piping, and plastic or coated steel vessels and equipment to prevent the genera-
tion of iron due to corrosion, will minimize any problems due to iron.

If high iron levels are naturally present, dissolved iron removal and oxygen exclusion will be

Scale Control
The scaling tendencies of all waters should be evaluated. Because the pH of polymer solutions is
frequently basic, calcium carbonate scale may be a problem. Scale inhibitors must be selected which
are compatible with the polymer.

Sodium, Calcium and Magnesium Ions

Since the viscosity of polysaccharide polymer solutions increases with increasing salinity, it may
be desirable to increase the salinity of the water when using biopolymers.

The presence of sodium, calcium and magnesium ions has the opposite effect on partially hydro-
lyzed polyacrylamide solutions, and their concentrations must be limited for maximum viscosity in-
crease. For that reason fresh waters are normally used where available. Softening may be employed
to remove calcium and magnesium ions if desired.

In many cases, however, brines are used with polyacrylamides and the negative effects of these
ions are ignored if a sufficient increase in viscosity can be achieved to materially increase oil recovery.

In this process, oil is displaced from the reservoir rock as shown in Figure 9.6.

In most cases, the surfactant slug ahead of the polymer is a micellar solution, sometimes referred
to in the literature as a microemulsion, soluble oils, or swollen m i c e l l e s / 9 ^ Its primary purpose is to
lower the interfacial tension and miscibly displace oil that cannot be displaced by water. I Ip to 28%
additional recovery of reservoir oil has apparently been achieved, but in some cases the economics
have not been favorable.


Figure 9.6 Schematic Diagram of the Surfactant-Polymer Process



Micellar solutions contain at least three components: hydrocarbon, surfactant, and water. The
hydrocarbon can have a broad range of compositions, and the surfactant is usually a petroleum
sulfonate. The water must be fresh.

Alcohol can be added as a fourth component. It serves as a co-surfactant and increases the lati-
tude in the composition of micellar solutions.

If distilled or deionized water is used, salts may be added. They tend to lower the viscosity of
the micellar solution and increase phase stability/ 9 7'^

Additives such as biocides may be incorporated into the micellar solution.

Surfactant slugs with compositions other than micellar solutions are also u s e d / 9 / ^

Water Treatment

Exceptionally clean water of carefully controlled composition is used to make up the surfactant
slug. Requirements vary with the specific surfactants used.


When crude oil is displaced by water, the displacement efficiency can be improved by lowering
the interfacial tension between the two fluids. In many cases this can be accomplished by the addition
of a strong base to water. Sodium hydroxide (caustic) and sodium orthosilicate have been widely used
due to cost considerations.

At typical use concentrations, the pH of the alkaline solution is in the range of 12 to 13.

Effect of Divalent Ions

At these very high pH values, the precipitation of divalent ions such as calcium, magnesium and
iron as hydroxides can be anticipated. Hence, these ions must not be present, and softening is nor-
mally required to reduce the total hardness of the water used to make up the alkaline solution to less
than 1 mg/L as C a C 0 3 / 9 / w

Effect of Salinity

It usually requires a lower concentration of caustic or sodium orthosilicate to achieve the required
interfacial tension when some NaCl is present/ 9 ' 1 9 ) The required level is established experimentally
lor the specific crude oil in question.

Water Treatment

Suspended solids and dispersed oil must be removed. Additional water treatment usually consists
of adjustment of salinity to the desired value by the addition of NaCl. or by blending with another wa-
ter. followed by softening to remove the divalent metal ions.

'CHAPTER 10 353


When carbon dioxide mixes with crude oil, under certain conditions a single-phase fluid is
formed which is miscible with the reservoir oil. The objective of C 0 2 injection is to form a miscible
bank to displace the oil, and then to displace the miscible bank with gas or water. An additional
10-15% oil recovery might be anticipated.

Of the several methods of CO: injection, three also utilize water injection:' 9 - 0 '
Injection of a slug of CO: followed by water.
Alternate injection of slugs of CO2 and water. Also known as the WAG (Water Alternating
Gas) process.
Simultaneous injection of C O 2 and water.

It is the purchase and processing of the carbon dioxide which is the most expensive part of oper-
ating a C O , f l o o d / 9 -9'1

Water Treatment
The water which is injected is usually produced water, and it should be processed in exactly the
same way it would be for normal water injection.

Corrosion Control
The only additional problem is the increased corrosiveness when the water and carbon dioxide
mix. Mixing occurs only once in the first process: at the interface between the CO: and water. The
effect of the increased corrosion rates during this period on the overall life of the system should be

In the WAG process water and CO: a r e transported to the injection wells through separate lines.
Mixing of the two fluids occurs in the injection wells at the end of each cycle, and can result in severe
corrosion/ 9 - 0 '

Simultaneous injection of the two fluids presents the largest amount of mixing of the two fluids,
as well as the most severe corrosion problems.

The primary methods of corrosion control in llie areas where mixing occurs are corrosion resis-
tant materials (both metallic and non-metallic), coatings and l i n i n g s / 9 : 0 >

In-situ combustion involves the injection of air or oxygen into the reservoir. Spontaneous igni-
tion occurs in the vast majority of reservoirs, and the combustion zone moves away from the injection
well consuming part of the reservoir oil as fuel. The result is an extremely complex displacement pro-
cess, as shown in Figure 9.7.

Water may also be injected along with the air. This process is referred to as a Combination of
Forward Combustion and Waterflooding (COFCAW). The water is heated by the residual heat in the
rock behind the combustion zone, flashes into superheated steam and enhances the oil displacement
process. However, the major benefit of simultaneous water injection is a reduction in the amount of
air required to recover a ban-el of oil by a factor of three or more.'9-2"



Injection Well Production Well

Air *- Combustion Gas -=ps>--Oil and


Burning Front Hot Water Path of Burning Front

Figure 9.7 Forward In-Situ Combustion Process' 93 '

The water and air are normally kept separate by injecting one fluid down the tubing and the other
down the annul us in order to minimize oxygen corrosion. A joint of Inconel tubing on the bottom of
the tubing string will effectively control con-osion in the area of m i x i n g / 9 " '

There seems to be no special requirement for water used in the COFCAW process other than it
should be of good quality to avoid downhole plugging problems/ 9 --1' Hence, normal treatment proce-
dures should be adequate.

Steam-flooding has been by far the major enhanced recoveiy process for many years, especially
in the USA. At one stage it represented over 75% of all enhanced production/ 9 7 ' However, an Oil
and Gas Journal survey*'9- ' showed a steady decline in the number of US steam EOR projects since
about 1990; from 137 down to 46 in 2004. Although it remains the highest producer, at about 51%
(2004 data) of oil from US EOR, the number of such projects is now outnumbered by COi miscible
floods (70). Steam EOR still accounts for the most projects throughout the rest of the world/ 9 - 7 ''

The purpose of injecting steam into a reservoir is to heat the crude oil and reduce its viscosity.
Since the flow rate of oil through a reservoir rock is inversely proportional to its viscosity, consider-
able increases in production rates can be achieved.

There are two steam recovery t e c h n i q u e s / 9 ^

Steam Stimulation

Sometimes known as cyclic steam injection, steam soak, or huff and puff. Steam is in-
jected into a producing well for a period of several weeks. Steam injection is halted, the
well is shut-in for a few days, and then placed on production. This cycle is then repeated
as many times as desired.

Steamflooding is similar to waterflooding in that steam is injected into a number of injec-

tion wells and oil is produced from adjacent producing wells in a pattern. It is the princi-
pal steam injection method.
In either case, the injected steam is produced by a boiler which is usually referred to as a steam
generator in oilfield operations. It is a once-through system which converts water to a mixture of ap-
proximately 80 percent steam and 20 percent water, normally referred to as 80% quality steam.


The steam generator feedwater must be treated to prevent corrosion and/or scale formation on the
boiler tubes, as well as to prevent corrosion and the formation of scales or precipitates in the water
phase. It should be noted that all of the dissolved solids in the leedwater will be concentrated in the
water phase, which is injected along with the steam. Consider these points for the system shown in the
process flow diagram given as Figure 9.8.

Surge Wash Tanks



Production 50-350 ppm
Header Softeners
rH H Steam-Flood

-H K
O-in-W 0-2 ppm Surface
25-175 ppm -H K
H H Filters

Figure 9.8 Example of Produced Water Used for Steam EOR

Corrosion Control

If the feedwater contains dissolved oxygen, it must be removed. This prevents boiler tube corro-
sion as well as the precipitation of soluble iron ions as iron hydroxide. Otherwise, no treatment is re-

The feedwater supply system should be constructed from non-metallic materials, or coated or
lined steel. This prevents con-osion and the generation of iron.

Scale Control

All steam generator leedwaters must be softened to remove calcium and magnesium ions and pre-
vent scale. The softened water should contain zero hardness (Total Hardness < 1 mg/L as CaC03).

Waters with less than 5000 mg/L TDS are softened using sodium ion exchangers. This process
uses a strong-acid resin which is regenerated with a NaCl solution.

In order to achieve zero hardness with waters containing more than 5000 mg/L TDS, it is neces-
sary to use weak-acid resins. Regeneration requires an acid followed by a base (usually HC1 and



Resin Fouling
Steam generator feedwater entering the ion exchange units must be quite clean. Dispersed oil,
suspended solids, and precipitated iron compounds must be eliminated. Otherwise the resin will be-
come fouled and its capacity for ion exchange will be reduced.

Other Precipitates
There is always a possibility of the formation of silica or sodium salt deposits when high TDS
waters are used. However, this is usually controlled by limiting steam quality.' 9 231

Microbiol EOR
The micro-organisms involved in MEOR produce the required chemicals in-situ. They contribute
to EOR by producing biopolymers and surfactants, as well as gases within the reservoir. To some ex-
tent the organisms also help to plug high-permeability parts of the reservoir, thereby helping to im-
prove sweep efficiency.

Further technical information on EOR is available with the References 9.30-9.31.

9.1 Leonard, J.: "EOR Set to Make Significant Contribution," Oil & Gas J. (April 2 1984) 83.
9.2 Donaldson, E. C.. Chilingarian, G. V. and Yen, T. F.. Editors: Enhanced Oil Recoveiy, I. Developments in
Petroleum Science, 17A, Science Publishers B.V., Amsterdam (1985). Chapter 6. Collins, A. G. and
Wright, C. C.: "Enhanced Oil Recovery Injection Waters."
9.3 van Poollen, H. K. and Associates, Inc.: Enhanced Oil Recoveiy, Pennwell Books, Tulsa, Okla. (1980).
9.4 Salamone, J.C., Clough, S.B., Salamone, A.B., Reid, K.I.G., and Jamison, D. E.: "Xanthan Gum-A
Lyotropic. Liquid Crystalline Polymer and Its Properties as a Suspending Agent," Soc. Petr. Engrs. J.
(Aug. 1982) 555.
9.5 O'Leary, W. B Boivin, J. W., Dasinger, B. L Beck, D, Goldman, I. M and Wemau, W. C.: "Biocide
Evaluation Against Sessile Xanthan Polymer-Degrading Bacteria," SPE Paper 13588, presented at Interna-
tional Symposium on Oilfield and Geothermal Chemistry, Phoenix, AZ, (Apr. 9-11,1985).
9.6 Auerbach, M. H.: "Prediction of Viscosity of Xanthan Solutions in Brines," SPE Paper 13591, presented
at International Symposium on Oilfield and Geothermal Chemistry, Phoenix, Arizona, (April 9-11, 1985).
9.7 Wellington, S. L.: "Biopolymer Solution Viscosity Stabilization-Polymer Degradation and Antioxidant
Use," Soc. Petr. Engrs. J. (Dec. 1983) 901.
9.8 Chatterji, J. and Borchardt, J. K.: "Applications of Water-Soluble Polymers in the Oil Field," J Petr.
Tech. (Nov. 1981) 2042.
9.9 Mungan, N.: "Enhanced Oil Recovery Using Water As A Driving Fluid," World OH (Feb. 1, 1982) 95.
9.10 Mungan, N.. Smith, F.W. and Thompson, J.L.: "Some Aspects of Polymer Floods." Trans. AIME (1966)
9.11 Ryles, R. G.: "Chemical Stability Limits of Water-Soluble Polymers Used in Oil Recovery Processes,"
SPE Paper 135S5, presented at International Symposium on Oilfield and Geothermal Chemistry, Phoenix,
Arizona, (Apr. 9-11, 1985.)
9.12 Knight, B. L.: "Reservoir Stability of Polymer Solutions," J. Petr. Tech. (May 1973) 618.
9.13 Bradford. R.A., Compton. J.D. and Hollis. P.R.: "Operational Problems in North Burbank Unit
Surfactant/Polymer Project." J. Petr. Tech. (Jan. 1980) 11.
9.14 Shupe, R. D.: "Chemical Stability of Polyacrylamide Polymers," J. Petr. Tech. (Aug. 1981) 1513.

C ooling water is circulated through equipment to absorb and carry away heat. The three basic types
of cooling water systems are shown in Figure 10.1. The mechanical design of cooling systems is
beyond the scope of this book. However, further details are available within the References 10.2-10.5.

The open recirculating cooling water system is by far the most common where plentiful supplies
of easily accessible water are available. This type is used in most oil and gas processing plants to cool
compressed gas and overhead streams from fractionating towers.

Closed recirculating systems usually involve much smaller volumes of water and are often used
to cool natural gas engine jackets and compressor cylinders.

Once-through cooling systems are seldom used because of the large volumes of water required
and potential thermal pollution problems when the water is discharged back to the environment.


In open recirculating cooling systems, the cooling water is pumped through a heat exchanger
where the process stream is cooled and the water is heated. The warm water then passes to a cooling
tower or spray basin where it is cooled by contact with air before circulation back to the heat
exchanger. The simplest towers use natural draft, relying on the density difference between the warm
moist air inside the tower and that of the cooler "dry" air outside the tower. Forced or induced draft
towers use fans to enhance cooling capacity.

The water is cooled in the tower or spray pond by evaporation o f a portion o f the circulating wa-
ter. This evaporation loss results in an increase in the concentration of the dissolved solids in the cir-
culating water since the evaporated water contains no dissolved solids. The water lost by evaporation
must be replaced by makeup water.

Water is also lost by entrainment of droplets in the circulating air and carried away from the
cooling tower. This is called drift or windage loss. These droplets of circulating water contain the
same concentration of dissolved solids as the rest of the circulating water. Therefore, their loss does
contribute to the concentration of dissolved solids in the system as does evaporation loss. How-
ever, windage losses must also be replaced by makeup water.

Water lost by inadvertent leakage must also be replaced.

'CHAPTER 10 359


Cooling Exchanger



Examples Characteristics Common Problems

Spray Ponds Average Temperature Change - Corrosion
Cooling Towers say 10 to 17C [20 to 30F] Fouling
Evaporative Condenser Amount of Water Used - Scale
Moderate Microbiological
Wood Decay


Examples Characteristics Common Problems

Diesel-Engine Jackets Average Temperature Change - Corrosion
Automobile Radiators say 5.5 to 8.5C [10 to 20F] Fouling
Chilled-Water System Amount of Water Used -


Heat Exchanger

Examples Characteristics Common Problems

Potable Water System Average Temperature Change - Corrosion
Process Water say 3 to 5.5C [5 to 10F] Fouling
General Service Amount of Water Used - Scale
Extremely Large Microbiological

Figure 10.1 Cooling System Characteristics



Evaporation Losses
It requires approximately 1000 Btu's to evaporate one pound of water. This is equal to the
amount of heat required to raise the temperature of 100 pounds of water 5.5C [10FJ. Therefore, for
each drop of 5.5C [10F] through the tower, one pound of water will be lost by evaporation per 100
pounds of water cooled. This is the equivalent of 1% of the water circulating. Therefore, the evapora-
tion rate can be calculated as follows:

Evaporation Rate (spm ) = ~x - ~~ f 10.1)

10 100

Windage/Drift Loss
Windage loss for induced draft cooling towers may be estimated as from 0.10 to 0.20 percent of
water circulation. Reference to the cooling tower manufacturer's specification sheet will often estab-
lish this figure for the installation in question. Where 0.2 percent is guaranteed, use 0.15 percent, and
where 0.1 percent is guaranteed, use 0.07 percent.

A much wider variation (0.3 to 1.0 percent) in windage losses occurs in natural draft towers, and
most calculations can be based on total mechanical loss; i.e., blowdown plus windage. Windage loss
in spray pond systems is typically 1 to 5 percent of the circulation rate.

Blowdown is the continuous or intermittent removal of some of the circulating water.

Evaporation results in increased concentration of the dissolved solids in the circulating water.
The concentration is decreased by removing a portion of the circulating water and replacing it with
makeup water containing a lower concentration of dissolved solids. Thus, the circulating water is di-
luted to control the buildup of dissolved solids in the circulating water.

Other Losses
The only other water loss that may need to be considered is that due to leakage. In well main-
tained and managed systems leaks will be insignificant. Where there is a leakage problem an estimate
needs to be made regarding the volume lost and this value can be included within the calculations.

Cycles Of Concentration
Cycles of concentration, or concentration ratio (also known as concentration factor), is an expres-
sion of the degree to which the dissolved solids concentration in the original circulating water has been
increased or concentrated by evaporation losses. Since the water originally introduced into the circu-
lating system was "makeup water," the degree of concentration, or cycles of concentration, can be ex-
pressed a s f o l l o w s : "

^ , ppm chlorides in circulating

5 water
Cycles = (10.2)
ppm chloride in makeup water
The chloride ion is normally used because it is quite soluble and seldom lost from water by pre-

CHAPTER 10 361

Experience shows that optimum cost savings are achieved at about 4-6 cycles of concentratio
At high (8-10) cycles, water savings are often outweighed by the increased difficulty and cost of effec-
tive water treatment.

Makeup Water

The amount of makeup water required is equal to the sum of all losses of water from the circulat-
ing system.

Makeup = Blowdow v + Windage + Evaporation (10.3)

Wherever possible the raw water supply should be fresh water.

Material Balance Calculations

In previous sections we have defined evaporation and windage losses, blowdown, cycles of con-
centration, and makeup. It is now possible to derive an equation for the calculation of blowdown rate
by making a material balance based on chloride ion concentrations.

At equilibrium, the chlorides entering the system must equal the chlorides leaving the system.
Furthermore, the chloride concentration in the blowdown is the same as in the circulating water.

Consider the following four equations:

MxC/m = (B+W)XCIR (10.4)

Cl B = CIR (10.5)

M = B+W+E (10.6)

C= (10.7)
CI u

Where: R circulation rate

B blowdown rate
M makeup rate
W windage rate
E evaporation rate
C cycles of concentration
C1M chloride concentration in the makeup water
C1B chloride concentration in the blowdown water
CIR chloride concentration in the circulating water

Substituting Equations 10.5 and 10.6 into Equation 10.4, we obtain:

(B + W + E) x CLU ~(B + W)x ClR

Rearranging this expression into the fonn of Equation 10.7:

CL B + W+E
C= (10.8)



Equation 10.8 can be rearranged and solved for B to yield the following expression:
B= (10.9)

Example 10.1: Given:

Circulation Rate = 10 000 gpm

Water to tower = 100F
Water from tower = 80F
Guaranteed maximum windage loss = 0.2%
Chlorides in tower makeup = 100 ppm
Chlorides in circulating water = 500 ppm

Windage loss rate

Evaporation rate
Cycles of concentration
B l o w d o w n rate
Total makeup rate

1. W - 0.0015 x 10 000 gpm = 15 gpm

(Assumes that actual losses were only 75% of the guaranteed maximum.)

o A T R 100-80 10000 _
2. Er = x x = 2 0 0 gpm
10 100 10 100

3. C.P- 500
Clu 100

4. g _M^x(C--p]_20C>-[l5*(5-l)]_
35 gpm
C-l 5-1

5. M = B + W +E = 35 + \5+200 = 250gpm



Cooling waters are treated to prevent four types of problems:
1. Scale formation on the heat transfer surfaces.
2. Corrosion of metal piping and heat exchangers by the cooling waters.
3. Fouling of the heat exchangers and the cooling tower.
4. Deterioration of wooden cooling towers.

Scale Prevention

Calcium Carbonate Scale

Calcium carbonate scale is extremely common in cooling waters, especially in the heat
exchangers where the water is hottest.

The Ryznar Index

The Ryznar Stability Index is commonly used to estimate calcium carbonate scaling tendencies
for cooling waters. It should be calculated for both the makeup water and the circulating water. The
most common methods of preventing calcium carbonate scale are:
1. If the makeup water is not scaling, control the number of cycles of concentration so that
the Ryznar Stability Index will not exceed 6.0.
2. Add sufficient acid to the makeup water to maintain the pH of the circulating water at a
value slightly less than pHs (see pages 77 and 78).

Sulfuric acid is normally used for pH control since it is usually the most economical.
However, caution must be exercised not to precipitate calcium sulfate scale. This is most
likely to occur when the makeup water is high in sulfates and alkalinity.

The rule of thumb for avoiding calcium sulfate scale is to limit the concentration of cal-
cium sulfate in the circulating water to a maximum of 1700 ppm. Another guide is that the
calcium concentration multiplied by the sulfate concentration (when both are expressed in
ppm) should not exceed 500 000.

If calcium sulfate scale becomes a problem, HC1 can be substituted for H2SO4.
3. Use inorganic polyphosphates (sometimes in conjunction with pH control) or organic scale

Magnesium Scales
Magnesium hydroxide will precipitate only in veiy alkaline waters and is not a widespread prob-
lem. Magnesium carbonate is usually controlled in the same manner as calcium carbonate.

Prevention of silicate scale deposition is normally achieved by limiting the silica concentration in
the cooling water to a maximum value of 175 ppm (preferably 150 ppm).

The cycles of concentration based on the silica ratio should not be allowed to fall appreciably be-
low the cyclcs based on chlorides and hardness, as this would indicate silica deposition. If both silica
and magnesium cycle at a lower figure than chlorides, this indicates deposition of magnesium silicate.



Pretreatment of Makeup Water

The formation of calcium and magnesium scales can be prevented by removing the calcium and
magnesium ions (hardness) and alkalinity from the water by a softening process. The cold lime pro-
cess (see Chapter 8) is commonly used for cooling water.

Corrosion Control

Inorganic corrosion inhibitors such as chromates, polyphosphates, or a combination of the two,

were used for corrosion control for many years. Zinc was also used in combination with either
chromates, polyphospates, or both, to improve inhibition.

Since many of these materials are toxic to living organisms, disposal of blowdown into surface
waters posed a serious pollution hazard. In fact, chromates are now seldom (if ever) used because of
the handling and disposal problems. Modern con-osion inhibition treatments are often based on careful
pH control and the use of phosphates, hence the term alkaline phosphate programs.

Organic inhibitors have generally been less effective in cooling waters than the inorganics.

Microbiological Control

Algae, fungi, and bacteria can all contribute to problems in open cooling water systems. They
can encourage corrosion, cause fouling and cause deterioration of the wood in cooling towers.

Chlorine is most widely used to control the growth of microorganisms/ 7 0 -' A free chlorine re-
sidual of 0.3 to 1.0 ppm is usually adequate. Chlorine residuals of greater than 1.0 ppm for long peri-
ods can cause deterioration of tower wood, and should be avoided.

A common procedure is to chlorinate to a residual of 1.0 ppm for four hours each day. Continu-
ous chlorination to maintain a residual of 0.2-0.3 ppm is also used, but this may necessitate a periodic
slug treatment with another biocide to kill resistant growths.


Fouling of heat exchanger tubes and cooling towers is usually defined as the deposition of
non-scale-fonning materials such zs:(10-3>

1. Suspended solids in the makeup water

2. Particulate matter from the atmosphere
3. Biological masses
4. Corrosion products
5. Hydrocarbon contamination due to process leaks

Filtration and/or coagulation and settling may be employed to remove suspended solids from the
makeup water. Side stream filtration and/or silt dispersants may be employed in the circulating system
to reduce solids deposition. Biocides and corrosion inhibitors are used to prevent the generation of
corrosion products and biological masses.

CHAPTER 10 365

Tower Wood Deterioration

Tower w o o d deterioration can usually be avoided by following three rules:
1. Maintain an effective biocide treating program from start-up.
2. Avoid excessive chlorination. (1.0 ppm residual for long periods)
3. Maintain the pH of the water at less than 8.3.

Some Advantages and Disadvantages of Open Recirculating Cooling


Advantages Disadvantages
Less water is needed Higher capital cost than once-through systems
Enhanced corrosion control Large cooling towers
Enhanced scale control Blowdown may give environmental problems
Highest temperature change ~10-17C [~20-30F]' " Air saturation of the water
Ingress of dust, bacteria, algae, fungi, and other
airborne materials


A closed recirculating cooling system is one in which the water is circulated in a closed loop.
The heat absorbed in the cooling operation is dissipated in some type of heat exchanger. Air-fan cool-
ers are frequently used.

Since the cooling system is closed, there is no evaporation, and therefore no change in water
composition. Excellent quality makeup water is generally used for best system operation. Makeup
water is only necessary to account for any leaks, or if the system has been drained. Leaks, as well as
oxygen entry, can occur at faulty seals or packing in pumps.

High chemical treatment levels are often required, but since water losses are negligible, these lev-
els are economical. While chromates are extremely effective, toxicity and environmental issues cause
such products to be unacceptable. Closed system inhibitors tend to be based on molybdates (very ef-
fective), nitrites and borates, or combinations of these chemistries.

The advantages of a closed system lie in improved control of cooling water temperature through
the heat producing unit, and in the relative freedom from the problems of operating an open system.
The disadvantages include higher capital cost and the cooling ultimately being limited by the ambient
air temperature. Only medium temperature changes of about 5.5-8.5C [10-20F] are achievable.


Once-through designs are usually limited to those locations where there are such large volumes
of water (typically near an open stretch of sea), that the temperature rise is minimized in the outlet
stream. Typically the average temperature change might be limited to 3.5-5.5C [ 5 - 1 0 F ] / m ^ This
reduces the possibility of upsetting the local ecosystem at and close to the water discharge point.
Evaporation should be negligible, so the water chemistry usually remains unaltered.

The obvious advantages of a once-through system are:



no cooling tower and associated equipment

no water treatment (or minimal, typically limited to screening).

Unfortunately the disadvantages of corrosion, fouling (sedimentation of entrained solids, scaling

and biological growth) and thermal pollution must all be taken into account. Oxygen corrosion will
produce hydrated iron oxides which may deposit within the system and impede heat transfer. Scale de-
posits are possible when sufficient calcium and magnesium hardness salts, or silica, are present.

It is possible to minimize both the corrosion and the scale deposition using polyphosphates. An
alternative is organic polymers for the scale control. However, with once-through systems it is eco-
nomically unsound to attempt chemical treatment on the large volumes of water used. When circum-
stances do allow once-through, corrosion resistant alloys are usually employed to overcome corrosion
problems. The design should also prevent ingress of solids and minimize the possibility of scale depo-
sition within the system.

10.1 Ostroff, A. G.: Introduction to Oilfield Water Technology-. Second Edition, National Association of Cor-
rosion Engineers, Houston, Texas (1979).
10.2 Kemmer, Frank. N., Editor: The Nalco Water Handbook, McGraw Hill. New York (1979).
10.3 Betz: Betr Handbook of Industrial Water Conditioning. 9th Ed., Bete Laboratories, Inc., Trevose, Penn-
sylvania, USA (1991).
10.4 GPSA, "Section 11: Cooling Towers", Gas Processors Suppliers Association Engineering Data Book,
GPSA, Tulsa. Oklahoma, USA (2004), p. 11.1-11.14.
10.5 Definitions, Cooling Technology Institute (CTI), Houston. Texas, USA, at http://www.cti.org.

'CHAPTER 10 10
yV s in other industrial areas, steam is used in the oil and gas production and refining sector for:
1. Turbine drives for blowers, pumps and generating electricity
2. Process heating and within processes (e.g. distillation, tertiary oil recovery)
3. Strippers (e.g. deaerators)
4. Vacuum eductors
5. Space heating.

Steam is used because it has a high heat capacity' 77 "" and is able to c a n y a significant amount of
energy on a unit mass basis (mostly as latent heat). It is generated in what we commonly refer to as a

Although there are many different types of boilers, they are all essentially devices for transform-
ing liquid water to steam by heating the water. There are two main categories of boiler (See Figure
Fire-tube; combustion gases pass through tubes which are surrounded by liquid water
Water-tube; water passes through the lubes which are heated from the outside by the com-
bustion gases.

Within any boiler, chemical energy in the fuel is converted into heat' 77 57 and transferred to the
boiler feed water (BFW) as efficiently as possible. This heat transfer occurs by radiation, conduction
and convection and the boiler should be designed to achieve high quality steam. Steam quality is a
measure of the amount of liquid water in the steam:

Steam Quality (%) = (1 -Fraction of Liquid Water) x 100

When the water evaporates to form steam, the dissolved solids from that water are left behind.
As they concentrate in the remaining water they frequently fonn scales and deposits. Deposition on
heat transfer surfaces can cause the metal to overheat, leading to premature failure.

Corrosion can also be a serious problem in boilers. Dissolved gases such as oxygen, carbon di-
oxide or hydrogen sulfide pass with the steam, causing corrosion of steam and condensate lines. Low
pH values also contribute to corrosion.

Boilers are usually classified in terms of their operating pressure, but it is the corresponding op-
erating temperature which dictates the degree of treatment required for boiler feedwaters. The higher
the operating pressure (and temperature), the more stringent the treatment requirements.

'CHAPTER 10 369



Fire-Tube Boiler

-Tube Boiler

Figure 11.1 Integral Furnace Boiler and Configuration of HRT Fire-Tube Boiler

An arbitrary classification of boilers is given in Table 11.1.

TABLE 11.1
Boiler Classification
Operating Pressure Classification
Less than 200 psi Low Pressure
200-500 psi Intemiediate Pressure
500-2000 psi High Pressure

Most boilers used in the oilfield are of the low or intermediate pressure class. Medium pressure
boilers are generally used for gas plant and refinery operations.

The overall aim of any utility operation should be to minimize energy costs, but at the same time
maintain boiler system efficiency. Water treatment can play a part in this e f f o r t . ' The purpose
of boiler feedwater treatment is threefold: 17

1. Prevent the formation of scalcs and deposits.

2. Minimize or prevent corrosion in the boiler and steam systems.
3. Maintain steam purity. The primary causes of decreased purity are c a n y over of water
droplets, and silica dissolved in the steam.

Boiler feedwater treatment is classified as external or internal treatment. External treatment re-
duces the level of contaminants in the water before it enters the boiler. Typical external treatment pro-
cesses i n c l u d e r / / j ; / , w clarification, filtration, softening, dealkalization, demineralization, deaeration,
iron and manganese removal, or some combination of these processes.

Internal treatment refers to chemicals added to the boiler feed water which minimize scale, corro-
sion and foaming inside the boiler system.



Boiler feedwater should have the lowest levels of hardness, alkalinity, sulfate, silica, iron, copper,
manganese, dissolved gases and suspended solids, that can be economically afforded.

Maximum permissible concentrations for each constituent in the feedwater cannot be set. How-
ever, a preliminary estimate can be made from the recommended boiler water composition limits pub-
lished by the American Boiler Manufacturers Association ( A B M A ) as shown in Table 11.2. Similar
tables of chemical concentration limits r / / 5 / 7 " w are recommended by CIBO (Council of Industrial
Boiler Owners) and A S M E (American Society of Mechanical Engineers).

TABLE 11.2
Recommended Maximum Limits in the Boiler
Boiler Pressure Dissolved Solids Total Alkalinity Suspended Solids Silica
(psig) (ppm) (ppm) (ppm) (ppm)
0-300 700-3500 140-700 150 125
301-450 600-3000 120-600 10 90
451-600 500-2500 100-500 8 50
601-750 200-2000 40-400 3 35
751-900 150-1500 30-300 2 20
901-1000 125-1250 25-250 1 8
1001-1800 100 20-150 1 1
1801-2350 50 15-100 1 0.5
*Silica limits based on limiting silica in steam to 0.02-0.03 ppm

Iron and oxygen in the feedwater should be near zero. The oxygen content is reduced to between
5 - 4 0 ppb in a deaerator (typically by steam stripping). Any further reduction is achieved using chemi-
cal oxygen scavengers.

Continuous blowdown can be employed to limit the concentration of the various constituents in
the boiler shell. However, if the feedwater composition is such that very high blowdown rates are re-
quired, then it is generally more economical to remove offending constituents by external treatment of
the feedwater.

Three basic types of processes may be carried out in the external treatment of boiler feedwater:

1. Suspended solids removal

2. Dissolved gas removal
3. Water softening

Softeners can remove hardness, which might otherwise lead to scaling and foaming problems,
both of which adversely affect boiler efficiency. Scale deposition on heat transfer surfaces can lead to
hot spots and subsequent tube failures.

Silica removal is also important since it is soluble in steam and can be carried over into the dis-
tribution system. There it may condense on steam turbine blades as a smooth glass-like deposit/' 7 ^ or
erode surfaces.

These processes are discussed in Chapters 7 and 8.

CHAPTER 11 371


Internal chemical treatment is usually necessary to provide insurance against scale and corrosion
even if the water has been externally treated. When boilers are operated at low to moderate pressures
and the raw feedwater is of good quality (low hardness, silica, and turbidity), internal treatment alone
will often be adequate.

Scale Control
What is the problem with scale deposition in boilers? Typically it occurs on heat transfer sur-
faces and gives rise to:
Loss of heat transfer efficiency ( lower boiler efficiency)
Hot spot failures
Greater heat loss via the stack gases.
The most common types of scale include*77,97 calcium carbonate, magnesium hydroxide, magne-
sium silicate, calcium and magnesium phosphates and calcium sulfate. A more comprehensive list of
the scales and sludges that can occur in boilers is available in Reference 11.7

Scale formation within a boiler is controlled by precipitation of calcium and magnesium as a

non-adherent sludge, or by chelation of the calcium and magnesium ions.

Phosphate Treatment
A soluble sodium phosphate is added which reacts with calcium to fonn an insoluble precipitate,
calcium phosphate. Magnesium and silica are precipitated as magnesium hydroxide, magnesium sili-
cate, or calcium silicate. The alkalinity of the makeup is usually adequate to produce the necessary
OH - for the magnesium precipitation, although in some cases caustic must be added. The pH should
be maintained above 9.5 both for magnesium precipitation and to ensure the formation of less adherent
precipitates. A reserve or excess of 40 ppm P 0 4 = should be maintained. A review of the various ap-
proaches to phosphate treatment since its introduction in about 1943, is given in the Reference 11.10.

Sludge dispersants may be used in conjunction with phosphates to prevent adherence of the pre-
cipitated particles and to maintain the resulting "sludge" as a non-adherent sluiry. Sludge dispersants
coat the finely divided particles as they are formed so that they will not fonn large crystalline precipi-
tates. Smaller particles will remain dispersed at the velocities encountered in most boilers, enabling ef-
ficient removal during blowdown.

An effective blowdown schedule must be implemented in order to ensure that this sludge is re-
moved from the boiler.

Typical examples of sludge dispersants are tannins, lignins, starches and certain synthetic poly-
mers, such as the polyacrylates.

Soda Ash or Caustic Treatment

In this process, sodium carbonate, sodium hydroxide, or both are added to the boiler water to
supplement the alkalinity supplied by the makeup water, which is not softened. The sodium carbonate
results in the precipitation of calcium ions as calcium carbonate. The addition of caustic elevates the
pH, causing the precipitation of magnesium and silica as magnesium hydroxide and magnesium sili-



Sludge dispersants are often used in this process to prevent deposit formation.

This method of treatment is only used with boilers operating below 250 psi which utilize high
hardness feedwaters.

The main problem attributed to caustic treatments is the higher risk of damage by caustic goug-
j n g (li.in ( s e e c o r r o s j o n section)

Chelating Agents
Chelating agents are materials such as EDTA which form soluble, complex ions with calcium and
magnesium. Thus, precipitation is prevented and the calcium and magnesium ions are effectively
maintained in solution.

The cost of chelating agents limits their use to very low hardness feedwaters (usually 5 ppm).

Over treatment by even 1-2 ppm in the feedwater can result in a buildup of chelant concentration
to very high levels in the boiler. Over a period of several months the chelant will attack metal result-
ing in general thinning coirosion, especially in stressed areas. Therefore, addition of chelating agents
must be very carefully controlled.

The use of chelating agents may prove attractive for some low pressure systems, but they have
not been successfully used in high pressure boilers/ 7 7 '

Corrosion Control
Pitting corrosion in boilers is usually caused by dissolved oxygen. Caustic embrittlement can
also occur which can result in catastrophic failure.

Oxygen Corrosion
Some oxygen will remain in the boiler feedwater, even after deaeration. Oxygen scavengers are
used to remove this dissolved oxygen and prevent corrosion due to its presence. While catalyzed so-
dium sulfite is still a commonly used chemical oxygen scavenger, hydrazine is tending not to be used
due to its associated safety hazards. Other possible scavengers^ 7J 7; '' s 77 7 -' , include tannin based mate-
rials, carbohydrazide and diethyl hydroxylamine (DEHA). Scavengers are discussed Chapter 7 and the
References 11.3, 11.7, and 11.12.

Where it is used, unreacted hydrazine can decompose at elevated temperatures yielding ammonia
and nitrogen. Ammonia in the presence of small amounts of oxygen will result in the corrosion of
copper alloys. For this reason, it is important not to maintain excessive hydrazine residuals.

Caustic Embrittlement
When ordinary carbon steels or austenitic stainless steels are stressed and placed in contact with
hot water containing a high concentration of hydroxyl ions, intergranular cracking can occur, resulting
in failure. This type of failure is referred to as caustic embrittlement and is a form of stress corrosion

In order for caustic embrittlement to occur, four conditions are necessary:

1. Hydroxides must be present in the boiler water. The use of caustic or sodium carbonate
(which can decompose at elevated temperatures to yield hydroxyl ions) is a major

CHAPTER 11 373

2. A joint or seam into which the water can leak must be present.
3. Concentration of the boiler water must occur within this seam.
4. The steel exposed to this concentrated solution must be stressed.
The presence of cracks or seams and large residual stresses can be reduced by proper design.
However, it can still be a problem, especially at tube ends.

All boilers should be tested using an Embrittlement Detector such as the one designed by U. S.
Bureau of Mines. If cracking tendencies are detected, treatment utilizing phosphates or sodium nitrite
should be initiated. Another form of corrosion damage caused by caustic is caustic gouging,1Oi s.u.in
which occurs under heavy porous deposits in areas of high heat flux.

Foaming results in water being carried from the boiler with the steam. It is influenced by several
factors, including:
1. Water level in the boiler
2. Dissolved solids concentration
3. Suspended solids in the water

Foaming can be prevented or reduced by boiler blowdown, a complete change of boiler water, or
by the addition of antifoamers.

Use of antifoamers as a routine part of internal treatment is good practice. However, the use of
excessive amounts can actually cause foaming and should be avoided.

When steam is generated, essentially pure water vapor is discharged from the boiler, leaving all
dissolved and suspended solids behind in the boiler. Therefore, the solids concentration in the liquid
water in the boiler continuously increases as the water evaporates. In order to maintain the dissolved
solids content below the allowable maximum, some of the concentrated water must be removed from
the boiler and replaced with water containing a lower amount of dissolved solids. The water removed
from the boiler is called blowdown.

In order to hold the dissolved solids content at an acceptable level the amount of solids removed
by blowdown must equal the amount entering the boiler with the feedwater. This is illustrated in Fig-
ure 11.2.

Typically, continuous blowdown from the steam drum is used to control the level of dissolved
solids in the boiler water. Periodic blowdown from the "Mud drum" is used to control the level of sus-
pended solids within the boiler. Blowdown rates must be optimized, since;

* too little > deposits or carryover

* too much wasted water, heat and chemicals.

By using oxygen scavengers that do not contribute to total dissolved solids (TDS), cycles of con-
centration can increase marginally and reduce the feedwater requirement. Although this may seem a
small amount, the annual savings may prove to be worth the effort

Guidelines for maximum levels of key constituents in boiler waters recommended by the A B M A
were given in Table 11.2. The desired maximum levels can be maintained by continuous blowdown.



Figure 11.2 Control of Boiler Water Solids by Blowdown (Courtesy Nalco Chemical Co.)

The required amount can be calculated from the following formula:

xlOO (11.1)
Where: B = blowdown, % of feedwater
TDSf = total dissolved solids in feedwater, ppm
TDS max = Maximum allowable dissolved solids in boiler water, ppm

The total dissolved solids can be estimated from conductivity measurements. Chloride concentra-
tions are often used in place of TDS values when their concentration in the feedwater is relatively low
and constant. Silica content can also be used if it is the limiting factor (e.g. in steam turbine systems
with pressures over 900 psi). Day-to-day control via conductivity, TDS, etc., is limited in high purity
water due to test interferences. An inert fluorescent tracer'77,5'1 can be used to accurately measure
boiler cycles. The greater accuracy can allow increased boiler cycles, with resultant savings


Once steam leaves the boiler and is used for its intended purpose, heat will be lost and the steam
will condense and form hot water. This condensed water is called condensate, and all or a portion of it
is usually returned to the boiler as feedwater. Boiler efficiency relies heavily on the ability to
maximize the return of condensed steam' 77,5> to the boiler cycle.

Although condensate is essentially distilled water, it is usually quite corrosive due to the presence
of carbon dioxide. Oxygen may also be present due to leaks in the condensate lines, poor oxygen re-
moval from the original feedwater, or exposure to the atmosphere in open condensate receiving tanks.

The primary source of carbon dioxide in steam systems is the thermal decomposition of carbon-
ate and bicarbonate ions in the feedwater.

Carbonate ions: CO" + H20+Heat 2OH- + C 0 2

Bicarbonate ions: 2HCO; + Heal-* CO' +H:0+C0_
CO; +H :0 + Heat 2 OH'+CO,

CHAPTER 11 375

Only about 80% of the carbonate ions convert to hydroxyl ions and carbon dioxide. Therefore 1
ppm of C 0 3 = will produce 0.58 ppm of CO : .

The bicarbonate decomposition to form carbonate ions, water and carbon dioxide is 100% com-
plete. This means that 1 ppm of bicarbonate ion will produce 0.36 ppm of CO? plus 0.49 ppm carbon-
ate ion. When the carbonate ion decomposes (assuming 80% completion), an additional 0.29 ppm CO-,
is created, giving a total of 0.65 ppm COj for every ppm of bicarbonate ion which decomposes.

When alkalinities are expressed as equivalent C a C 0 3 , the following factors apply:

1. One ppm of C 0 3 = (as C a C 0 3 ) will form 0.35 ppm C 0 2 .
2. One ppm of H C 0 3 " (as C a C 0 3 ) will form a total of 0.79 ppm C 0 2 .

This assumes that the carbonate decomposition is 80% complete.

When the steam condenses, the carbon dioxide dissolves in the condensed water, increasing its

Condensate return systems can be chemically treated to prevent corrosion by dissolved carbon di-
oxide by using neutralizing or filming amines. The pH of the condensate is raised to at least 8.5 and
preferably between 8.8 and 9.2, with blends of volatile neutralizing amines (e.g. morpholine,
cyclohexylamine, etc). Ammonia can also be used, but it has the disadvantage of attacking copper or

When a large portion of the steam condensate is returned to the boiler feedwater, the condensate
must be deaerated (degassed) to remove the carbon dioxide and keep it from accumulating in the sys-
tem. Any oxygen present due to ingress via leaks in the condensate system, is also removed in the

Oil Contamination
Oil is hazardous in boilers because it forms a heat insulating film. There are normally two major
sources of oil contamination in plants and facilities:
1. Reciprocating pump discharge.
2. Condensation and recovery of the still stripping steam.

Emulsified oil can be removed with DE filters and a variety of other methods. Free oil can be re-
moved in a skimming tank.

Boiler Efficiency
Some relatively straightforward approaches to calculating the cost of steam generation* 77 and
or looking at the improvement in boiler efficiency' 7 7 , 4 > can be found at the US Department of Energy
site on the Internet. Steam production is an energy conversion process where the energy in the fuel is
transferred to the steam, with the most significant energy loss being via the flue gas. Steam costs are
dependent upon what happens to the steam within the generation and distribution system. Hence, an
accurate determination is quite involved and it is recommended that the correct value be reached via
the development of a heat and material balance simulation model of the system. Otherwise, incorrect
steam pricing will lead to bad decisions. Simply put,

. energy contained in the produced steam

Boiler Efficiency =
energy supplied by the fuel



11.1 Ostroff, A. G.: Introduction to Oilfield Water Technology, Second Edition, National Association of Cor-
rosion Engineers, Houston, Texas (1979).
11.2 Kemmer, Frank N., Editor: The Nalco Water Handbook, McGraw Hill, New York (1979).
11.3 BetzDearborn: "BetzDearbom Handbook of Industrial Water Conditioning," 7th Ed., Betz Laboratories,
Inc., Trevose, Pennsylvania, USA (1991).
11.4 U.S. Dept. of Energy, "Improving Steam System Perfonnance," at:
11.5 NEM Business Solutions, "An Introduction to Steam Boilers and Steam Raising," at:
11.6 Frasca, Michael, "Boiler System Efficiency Improves with Effective Water Treatment," at:
11.7 GPSA Data Book (2004), "Utilities," Section 18.
11.8 CIBO, "Energy Efficiency Handbook" at: www.oit.doe.gov/bestpractices/steam/pdfs/steamhandbook.pdi
11.9 Cleaver Brooks, "Solving Boiler/Cooling System Water Problems," at:
www.cleaver-brooks.com/Servicel l.html
11.10 Stodola. Jan and Marvin D. Silbert, "Enhanced Phosphate Treatment for Drum-Recirculating Boilers," In-
dustrial Water Treatment, Vol.28. No. I (Jan./Feb. 1996), p. 41-46. Also at:
11.11 Moore, Mark A, "A Brief Introduction to Water Chemistry for Cogeneration Plants," 5th Conference &
Exhibition on Chemistry in Industry. Manama, Bahrain (Oct. 14-16, 2002).
11.12 Accepta, "Boiler Water Oxygen Scavengers: sulfite. Tannin, DEHA, C'arbohydrazide...", at:
www.accepta.com/Industry VVater Treatnient/boiler water oxygen scavengers.asp
11.13 "How to Calculate the True Cost of Steam," (Sept. 2003), at:
11.14 Harrell, Greg, "Steam System Survey Guide", May 2002, Oak Ridge National Laboratory, at:

CHAPTER 11 377
A lthough the emphasis in this book is on the water chemistry, much of the information presented
applies to surface facilities. What follows is a summary of some of the most important consider-
ations. Further details are to be found in Volumes 1, 2 and 4 of Gas Conditioning and Processing,
companion books in the Campbell Petroleum Series.

With surface facilities the presence of water can result in one or more of the following:
Water in oil separation problems (emulsions)
Scale formation
Erosion/Corrosion, including
stress corrosion cracking (e.g. in stainless steels)
sulfide stress cracking (with sour systems)
Erosion and corrosion are commonly combined because in piping and vessels it is sometimes dif-
ficult to distinguish between the two. Control of velocity and impingement are important mechanical
design details.

All except water-in-oil emulsions have been discussed previously, but there are some considera-
tions unique to surface facilities.


In separating water from oil in the surface production equipment, there will inevitably be a small
amount of oil entrained with that water. The operation of downstream water de-oiling devices is de-
pendent upon both the efficiency of this separation and on any chemicals used within the system.
Hence, when addressing any problems associated with water treatment, the starting point is to see that
the primary separation equipment is functioning correctly.

It needs to be established whether the regular demulsifier is still performing at its optimum level.
Is this demulsifier being injected and at the correct concentration? Is the current product no longer
suitable because of changes which have occurred in the system?

Some of the chemistries that have been tried as demulsifiers in the past have done a very efficient
job in that respect, but shown adverse side effects. These adverse effects have either been on the
downstream reverse-demulsifier, or have shown up as direct interference with equipment operation.

'CHAPTER 10 379

An example of the latter was the "killing" of the froth in gas flotation units, thereby preventing effi-
cient separation.

Once it is established that the primary separation is functioning well, a sensible investigation of
the water treatment processes is possible.


Water cooled heat exchangers are used commonly for the cooling of gas, dehydration and sweet-
ening liquids, and hot hydrocarbon liquids prior to storage. The discussion on cooling water treatment
in Chapter 10 applies to heat exchangers using such water. Water from lakes, rivers and the oceans
also are used extensively for cooling.

An exchanger cannot be properly specified and designed until the cooling water properties have
been determined reliably. What kind of exchanger is feasible with the kind of water available? It of-
ten is not practical (mechanically or economically) to eliminate all of the problems listed above.

The combination of process requirements and water quality may dictate that scale formation can-
not be eliminated, only its rate of formation can be controlled. Reduction of pH to prevent calcium
carbonate scale may introduce metallurgical problems. Many questions are pertinent. What kind of
scales are anticipated? Can these scales be removed by available chemical cleaners or is mechanical
scraping required? How often will cleaning be required? Is chemical inhibition of the scale a viable
option? The final choice of heat exchanger type should represent an intelligent answer to questions like
these as well as those raised by process concerns.

If a tube-and-shell exchanger is considered, should the water be placed on the tube side or the
shell side? The tube side is easier to clean. Final process calculations cannot proceed until such deci-
sions are made. A plate type exchanger might be preferred because it can be disassembled for mechan-
ical cleaning. That said, re-assembly of plate-exchangers without incurring any leaks is apparently a
skill in itself.

Erosion/corrosion can dictate the use of metals that are poor conductors of heat or are difficult to
fabricate. Stainless steels, for example, are poor conductors of heat. Their use reduces thermal effi-
ciency. Titanium is a good heat conductor but is expensive to fabricate. Fortunately, there are applica-
tions where titanium may be used economically in plate exchangers.

As referred to earlier in this book, chloride cracking (a form of stress corrosion cracking in stain-
less steels) can be a problem when the chloride ion is present in the water. Above about 65C [150F],
a commonly used material like Type 316 stainless steel is not satisfactory for cooling with sea water.
Titanium in a plate exchanger is a more satisfactory choice near, or above, this temperature. Of
course, in a typical sea water, temperature in excess of 65C [150F] are usually well above the cal-
cium carbonate scaling temperature. Usually it is not economical to treat sea water to prevent scale
formation in offshore process applications if the air is cold enough for aerial cooling. With compres-
sors or other "hot" equipment, aerial cooling is most always the choice offshore in temperate climates.
This eliminates concern about chloride cracking of stainless steel equipment and scaling in heat

In tropical climates where the air is warm and/or applications where low temperatures are desired
(like dehydration), water cooling may be the preferred choice.

A "once-through" system using raw sea water is veiy satisfactory if temperature and metallurgi-
cal constraints are adhered to. At least minimal filtration is necessary to prevent the ingress of marine



organisms, even though initial tests may indicate their absence. The marine structure acts as an artifi-
cial reef for their development and they could become a problem. However, process tubing and piping
is not as prone to plug as a reservoir rock. In some cases a course screen to keep out large creatures,
with no chlorination, may be satisfactory. If a filter is used, chlorination is necessary to control bacte-
rial growth and fouling within the filtration system.

Deoxygenation of sea water is not always carried out when the only use is process cooling, since
corrosion can be economically controlled by utilizing corrosion resistant materials.

The use of saline underground water for cooling involves the same considerations if it is not
"sour" with sulfur compounds. If the latter are present, the later discussion for sour systems applies.

Once-through cooling from fresh water sources is largely a scale prevention problem.

Indirect Heaters
This is any device where heat from the burner and combustion gas transfers to a liquid bath
which, in turn, transfers heat to another medium. A water bath heater is used commonly to heat gas or
oil. In a typical installation the temperature of the surface of the burner tube in contact with the water
is over 93C' [200F], well above the scaling temperature of many salts present in water.

Check the water used to fill the heater. Raw water is seldom good enough. This is a batch pro-
cess so proper treatment is inexpensive. However, like your automobile radiator, it is a good idea to
change the treated water periodically. With most heaters of this kind it is convenient to check periodi-
cally for scale formation on the fire tube.

Sour systems require considerations discussed earlier in this book. Hydrogen sulfide (H2S) is not
corrosive in the absence of liquid water until the system temperature reaches about 288C [550F],
Thus, if liquid water can be eliminated or inhibited, only heaters represent a common problem.

The use o f dehydration to eliminate liquid water is common. Aqueous solutions o f chemicals
like the amines can effectively remove the sulfur compounds. Figure 12.1 summarizes some of the
problems encountered in these types of units as well as others commonly used with sour systems. Dis-
cussions about corrosion testing, corrosion inhibitors, etc., in early chapters apply equally well for
these systems.

There are several concerns that apply for all sour process systems.

Normal liquid velocity in process lines will be 8-10 ft/sec. Reducing this velocity to 3.0-3.5
ft/sec in a sour system will minimize erosion/corrosion and facilitate the effectiveness of a filming-type
corrosion inhibitor. Long radius ells are preferred to minimize impingement effects. Stress relieved,
butt- welded pipe joints are preferred to threaded joints. Avoid the use of dissimilar metals to mini-
mize bimetallic corrosion.

Control valves should have "soft seats" to compensate for embrittlement A Rockwell hardness
of 18-20 usually is preferred.

CHAPTER 12 381

Some Treating Problems

Factors Affecting Corrosion

Glycol Regeneration Caustic S t e a m S t r i p p i n g Still

Condensation of Regeneration EX
" acidic water

Add formation by excessive
heal andfor air contamination
M jCooterD
Release o f s e l e c t ^ S o i i d " B e d Regeneration

Excessrve heat, condensation,

J J)
add conditions, and erosion Hydrogen sulfide at
lower temperatures.
RIcti-Lean Critic^ vskxaties. (Hydrogen penetration)
Exchanger caustj<; smMtlement
erasing, saSum sulfide

Fired Heaters A m i n e Treating Fractionation

Low oil flow

High velocity
caused by

Excessive heat
at tube fins Firing Conditions

Sulfur Corrosion Flame Impingement
s s ~sfr
Critical solution

Figure 12.1 Summary of Problem Points in Sour Gas Treating




With proper attention to mechanical details like those above and proper stress relief o f all welded
parts, carbon steel plants have proven satisfactory where initial cost was an important factor. Many
plants use Type 304 and 316 stainless steel at critical points. Metal-clad and plastic coated parts also
are being used. Type 7072 aluminum-clad tubing has been employed. But, regardless of the metal-
lurgy chosen, it is recommended that you limit velocities, impingement, etc.


The use of a proper inhibitor is almost always desirable in sour production/processing systems.
Normally it will be of the filming type. The usual goal is to reduce coirosion as far as possible, since
its elimination is not practically achievable.

Figure 12.2 shows the combined effect of cooling and adding a corrosion inhibitor to one
triethylene glycol (TEG) dehydration unit.

This chapter and the ones which immediately precede it give a brief overview of various applica-
tions which require a reasonable knowledge about how water behaves in petroleum systems and how
to handle it effectively. Water occurs in all petroleum reservoirs. Ignoring it, or handling it improp-
erly, will give rise to significant further costs. Greater efficiency, lower overhead and greater profit are
the end result of improved water handling. This book is dedicated to that proposition.

Figure 12.2 Possible Effect of Corrosion Inhibitors in TEG Systems

Appendix 11
Nomograph for Determination of Ryznar and Langelier Scaling Indexes

Given the following data, find the
values of both the Langelier and
the Ryznar Indexes: -0.6
pH = 6.9 -0.7
Temp. = 70F
TDS = 72 ppm
Calcium hardness as
CaCO., = 34 ppm 3
Methyl orange alkalinity as -4
CaCO;, = 47 ppm (5) ^ 7.0 -5
Solution: B
1. Reading at the bottom of the 7.5
left-hand side, find TDS = 72
and note the intersection of -20
this reading with the curved <8.0 -30 I
70F line. -40

Carry this intersection hori- -50

zontally to pivot line 2. 8.5 -60
70 m
-80 O
Connect that point with Ca
hardness = 34 on the right- 9.0
hand scale. -200

Note the intersection with

xa 9.5- 380
pivot line 3. -400
5. Connect that point with alka- -660
linity = 47 on the left-hand -700
scale. ^1,000
Note the intersection on pivot
line 4. Connect this inter-sec-
tion to pH = 6.9.
7. The Langelier Index = -1.8
TDS = total diss, solids, ppm
and the Ryzner Index = 10.5.
This water would be said to
be very corrosive.

Ref: Cap Ian. F "Is Your Water Scaling or CorTosive?." Chemical Engineering (Sepi I. 1975) 129

Appendix 12
Values of Stiff & Davis "K" for C a C 0 3 Scale Calculation

0 0.2 0.4 0.6 0.8 1J 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0

Molar Ionic Strength QA)


Appendix 13
Conversion of mg/L Calcium and Alkalinity into pCa and pAlk
Appendix 14
Example Scale Calculations

Ionic Strength, CaCO.i Scaling Index and C a S 0 4 Solubility

Water Analysis Ionic Strength Calculation

Ion (mg/L) Conversion Factor
Na + 7559 2.2 x 10""5 -
16 630 x 10"5
Ca""" 877 5.0 x 10"5 = 4385 x 10"5
Mg~ 239 8.2 x 10-s - 1960 x 10"5
cr 12 294 1.4 x 10~5
17 212 x 10"5
co3= 0 3.3 x I0~5 = 0 x 10"5
HCO 3 556 0.82 x 10~5 =
456 x 10~5
so46 1740 2.1 x 10~5 = 3654 x 10"5
Total Ionic Strength = = 44 297 x 1(T5
= 0.44

Temperature = 60C ; pH = 7.04

Calcium Carbonate Scaling Index Calculation

K = 2.24 (from Appendix 12)

pCa = 1.67 (from Appendix 13)
pAlk = 2.05 (from Appendix 13)

SI = pH - (K + pCa + pAlk) = 7.04 - (2.24 + 1.67 + 2.05) = +1.08

SI > 0 , so CaCO.} scale is likely

Calcium Sulfate Solubility Calculation

Kc = 9.2 x 10 - 4 (from Appendix 15)

4 Kc = 36.8 x 10"

Concentration Conversion
Ion (mg/L) Factor M (moles/L)
Ca ++ 877 2.5 x 10 5 2192.5 x 10-5
so4= 1740 1.04 x 10- 1809. x 10"-
X = AM = 382.9 x 10"5

S = 1000 j ( 0 . 1 4 6 + 3 6 . 8 ) x 10 - 4 J -382.9x 10'5!=57meq/L

Concentration Equivalent Concentration

Ion (mg/L) Weight meq/L)
Ca++ 877 43.9

S04= 1740 36.3

Actual C a S 0 4 Concentration = 36.3 meq/L

S > Actual, so C a S 0 4 scale is unlikely.

Appendix 15
Calcium Sulfate (Gypsum) Conditional Solubility Product Constants
Appendix 16
Barium Sulfate Conditional Solubility Product Constants

95C [203 S F]

80C [177F]

65C [149F]

50C [122F]

35C [95F]

25C [77'F]

0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Molar Ionic Strength <ju)

Appendix 17
Strontium Sulfate Conditional Solubility Product Constants

Temperature, F

Appendix 18
Toxicity of Hydrogen Sulfide Gas
Concentration Result
10 ppm Can be smelled. Safe for 8 hours of exposure.
100 ppm Kills smell in 3 to 15 minutes. May sting eyes
and throat.
200 ppm Kills smell quickly. Stings eyes and throat.
500 ppm Loss of sense of reasoning and balance. Respira-
tory paralysis in 2 to 15 minutes. Prompt artifi-
cial resuscitation required.
700 ppm Unconsciousness within 15 minutes (maximum).
Cessation of breathing and death will result with-
out immediate artificial resuscitation.
1000 ppm Immediate unconsciousness. Permanent brain
damage may result unless rescued promptly.
May be fatal within 30 minutes.


Похожие интересы