Вы находитесь на странице: 1из 6

Reprinted from:

for AGAR Corporation


Case Studies and Technology Discussion

AGAR probes installed on a desalter in


the AMOCO Mandan Refinery

Minimize Wastewater
Contamination
Source reduction techniques dramatically reduce
hydrocarbons lost to sewer system

Authors:
L. Helm, Amoco Corporation, Mandan, North Dakota;
C. Spencer, Mustang Engineering, Inc., Houston, Texas; and
R. T. van den Dungen, DUBRAE Consultancy b.v. (previously, SHELL INTERNATIONAL b.v.)

Article reprinted from HYDROCARBON PROCESSING® magazine, June 1998 issue, pgs. 91 –100.
F/13M/898 Copyright © 1998 by Gulf Publishing Company, Houston, Texas. All rights reserved. Used with permission.
LOSS PREVENTION/ENVIRONMENT

Minimize wastewater
contamination
Source reduction techniques
dramatically reduce hydrocarbons
lost to sewer system

L. Helm, Amoco Corporation, Mandan, North


Dakota; C. Spencer, Mustang Engineering, Inc.,
Houston, Texas; and R. T. van den Dungen,
DUBRAE Consultancy b.v.

veraged total hydrocarbon releases to refinery

A sewers range from 0.5% to 4.0% of the total


crude charge. Unfortunately, refiners are los-
ing valuable feedstocks to the sewer and must spend
nearly $10/bbl to recover and reprocess lost hydro- Fig. 1. Energy absorption probes system installed on an electrostatic
dehydrator (desalter) for interface control.
carbons.
The desalting system is the largest contributor to
wastewater; improving operating methods on this unit effectively. Many operating companies worldwide are
can reduce shutdowns and maintenance expenses. documenting significant savings through equipment
Equally important, better operating practices can availability, minimized wastewater treatment, lower
decrease organic loading on the wastewater treating recovery/reprocessing costs, while still achieving com-
unit. Several case histories illustrate various source pliance with environmental legislation (such as the
reduction methods to minimize hydrocarbon emulsion U.S. Benzene NESHAPS).
losses to sewer.
Incentives for improved control. The traditional
Reduce losses at the source. Due to a mature and approach to minimize wastewater contamination does
more competitive marketplace, operating companies not focus on solving the problem at the source. Instead,
are re-evaluating fundamental manufacturing strate- wastewater streams are combined into a common sys-
gies. Consequently, the incentive to apply new tech- tem, which is then processed through slop tanks, CPI’s,
nologies and improve operations is more compelling API’s, hydrocyclones, air flotation units and/or ben-
than ever before. New measurement technologies, zene strippers. Although, these processes eventually
such as energy absorption, have steadily entered the remove hydrocarbons from the wastewater to required
marketplace to meet this need. levels, treatment costs can be particularly high.
Source reduction is an area where innovative tech- These costs include: specialty chemicals needed to
nology is being evaluated and used by many major oil break emulsions at the WWTP and the expense to
and petrochemical companies. The most effective way reprocess recovered hydrocarbons. Many refiners oper-
of reducing the hydrocarbon content in final effluent ating their crude units near maximum capacities will
is to avoid contaminant losses at their source. Energy suffer significant lost opportunity costs as fresh crude
absorption (EA) technology has proven to be a useful feed is displaced to process recovered oil. This can fur-
tool when controlling the amount of hydrocarbons sent ther exacerbate already poor industry margins.
to wastewater treatment plants (WWTPs). EA can be Typical costs have been documented in an inde-
applied for monitoring and controlling many separa- pendent study by Wright Killen/Ernst & Young. The
tion processes. Additionally, these new control sys- consultants surveyed a representative portion of U.S.
tems help to relieve the pressure from tighter envi- Gulf Coast refining industry to determine typical
ronmental restrictions by reducing the quantity of industry oily-water treating requirements, crude and
hydrocarbons released to wastewater pretreatment product losses, and associated costs.
systems. Study data show that total hydrocarbon releases
Field applications have verified that EA technol- to plant sewers from all refinery sources normally
ogy enables operators to meet operational goals cost- ranged from 0.5% to 4.0% of total crude charge. Addi-
HYDROCARBON PROCESSING / JUNE 1998
tionally, for an average U.S. Gulf Coast refinery, total
costs for recovering and reprocessing these releases 6

U.S. refinery cash margins, $/bbl


is $9.64/bbl. Therefore, if the refinery throughput is 5
Gulf Coast
100,000 bpd and it is releasing only 1% of this to the
sewers, the recovery and reprocessing costs to the 4

refinery are approximately $10,000/day. 3 East Coast


Midcontinent
Improved approach: controlling the source.
Instead of treating the wastewater after it has been 2

contaminated, it is more efficient to identify the 1 L. A. Basin


stream and use more sophisticated control to prevent
contamination at the source. In a typical crude oil 0
1Q92 2Q92 3Q92 4Q92 1Q93 2Q93 3Q93 4Q93
refinery, contamination contributors to wastewater
Quarter returns
(expressed as a percentage of total oil requiring recov-
ery) can be quantified:
Fig. 2. Refinery cash margins by U.S. locations.
• Desalters 40%
• Storage tanks 20%
• Slop oil recovery 15% especially during startup, shutdown and system
• Other processes 25%. upsets when accurate readings are most critical. Oper-
If improved control methods can be applied to these ator reliance on the accuracy from resultant level read-
processes (minimizing hydrocarbons released into the ings has often led to problems. Externally-mounted
wastewater), then significant operating credits could instruments, such as sight glasses, are equally inef-
be realized. fective because conditions outside of the vessel are
rarely representative of those within. Of the level-
Optimizing the process. An efficient process is the dependent instruments used on separators, capaci-
most profitable. A good example of such a system is tance probes have performed favorably. However, their
the crude oil desalting process. Improved control of effectiveness is limited by conductivity variations,
this unit can facilitate impressive returns such as: “blindness” from coatings and their inability to detect
longer run times, lower equipment fouling and corro- suspended oil in water.
sion, and reduced maintenance requirements between Other solutions. EA technology is more sophisti-
and during shutdowns. The desalting system is a par- cated than level control instrumentation and has been
ticularly good example to demonstrate source reduc- effectively used on source-reduction efforts. These
tion opportunities. The desalter not only has a critical instruments satisfy several basic requirements criti-
impact on operating costs, but also on the wastewa- cal for reduction applications:
ter treatment. The effects on waste treatment facilities c Direct contact with the process (inserted into the
from poor desalter performance become increasingly system)
important as emissions limits are more stringently c Quantitative measurement capability of 0% to
enforced by government legislation. 100% hydrocarbon/water concentration (in both oil-
Many refiners try optimizing the desalter via chem- continuous and water-continuous phases)
ical addition. These chemicals not only have a direct c Local or point specific measurement (avoids errors
price, but their potential effects on downstream pro- due to averaging over a large hydrocarbon/water dis-
cess equipment and some catalysts must be consid- tribution)
ered. But even under the best circumstances, chemical c Minimal affect on measurement from fluid prop-
alternatives should be used only after more efficient erties (specific gravity, pressure, temperature, vis-
control alternatives have been explored. cosity and coating build-ups).
Separation processes. When a specific separa- The EA probes use a transmitter/antenna combina-
tion process is identified for source reduction, the pro- tion intrusive to the process. Instruments are positioned
cess media’s characteristics must be considered. Then to penetrate the tank or vessel so that their antennae
the best ways of controlling separation must be deter- reside at the specific points where measurement is
mined. The interface between hydrocarbon and aque- desired. It uses a high frequency electromagnetic mea-
ous phases is rarely clear cut, especially in desalters. surement to determine volume percentages in two-phase
Historically, level-control instruments such as sight mixtures. The signal from each instrument is expressed
glasses, floats, displacers, differential pressure sen- in units of volume percent of the phases (typically oil
sors and capacitance probes have been used to con- and water) and reflects the fluid content in the imme-
trol interface. Results obtained from these technologies diate vicinity. The system monitors the position of an
have ranged from acceptable (under ideal operating interface, but also can track changes in the size and
conditions) to very poor and misleading (during upset rate of growth of an emulsion or dispersion. They can
conditions). This is not necessarily due to any failure identify the hydrocarbon dispersion in the aqueous
of the specific level control instruments, rather it phase that occurs at and below the bottom of an emul-
reflects a failure to confirm that a true level is pre- sion layer. When used in separators, EA monitors the
sent. percentages of water at various points or levels within
When using specific gravity-dependent control for the system, either controlling or monitoring the qual-
a separation, phase densities must be specified from ity of the separation as a function of the relative con-
product data and possible variations of these densi- tent of the phases.
ties are not always addressed. However, densities vary, Most important to the operator is that by using EA
HYDROCARBON PROCESSING / JUNE 1998
technology, they have eyes into the process. They can
700
identify and overcome problems normally associated Basis: 100,000 bpd
600

Potential savings,
with separation operations. This insight allows opera-

thousand $/yr
500
tors to anticipate and respond to potential upsets, receive
400
advanced warning of an upset and optimize the system
300
to correct and avoid such upsets.
200
Reducing oil loss to the wastewater treatment sys-
100
tems lessens slop oil recovery demand and minimizes
0
the pollutant content of the final effluent. Applying 0.025 0.05 0.075 0.1 0.125 0.15
source-reduction techniques help plants meet their Desalter oil under carry reduction, %
local discharge regulations, while reducing recovery
work required. Fig. 3. Potential savings due to reducing hydrocarbon losses in the
desalting process.
CASE STUDIES
Amoco’s Mandan refinery needed to reduce its total designed and sized segmented probe so that the seg-
plant benzene discharge below 10 metric tons/year ments (each several feet in length) could be sequentially
(mtpy). This decreased level would allow the refinery installed in an area with very limited ground clearance.
to be reclassified under existing environmental leg- Three years later during a scheduled turnaround, the
islation. Using the traditional approach of re-design- primary control probes were removed and replaced with
ing and expanding the WWTP to reduce benzene lev- single-piece probes. The segmented probe had operated
els, costs were estimated to exceed $70 million. flawlessly, however, the new single-piece probe was
However, combining source reduction with several installed for improved convenience and to avoid using
other water reuse initiatives, the refinery’s total cap- a multi-segment probe (with larger cross-sectional area)
ital investment was less than $4 million. The source in a high-pressure vessel.
reduction efforts subsequently reduced annual ben- Run length, debugging, optimization and pro-
zene discharge from an original level of 17 mtpy to cess control. The Mandan refinery’s primary objec-
only 3 mtpy (an 80% reduction). This reduction not tives were: enhanced run length and improved pro-
only exceeded required compliance with EPA regula- cess control and unit optimization. Run lengths
tions, but dramatically reduced emission levels, thus improved dramatically; the probes proved to be imper-
enabling the refinery to be reclassified. vious to varying operating conditions and fouling that
Tank farm. Amoco’s difficulties in dewatering float- affected the reliability and operability. The EA sys-
ing-roof tanks with minimal hydrocarbon losses had tems have been operating in the tank farm since 1991
been quite complicated, especially during rainy weather. and in the desalters since 1992, both without any
The refinery’s engineers selected EA for this task after interruption for cleaning or other maintenance.
testing several other technologies. The alternative As with any new technology, training and familiar-
methods were found to be very limited, unreliable and ization is needed. The EA system was a learning expe-
difficult to operate even in good weather conditions. rience for the Amoco operators. At the tank farm, the
One method previously used by Amoco involved a optimum location for the probes on each tank type
floating device with a swing arm (requiring manual (crude or finished product) was determined after a few
adjustment) for sample collection. Mandan found this months of field experience. On the desalters, the probes
approach to be cumbersome and not particularly accu- made their measurements within the desalter; thus,
rate. In addition, because the task was labor-inten- operators were using this new instrumentation without
sive, the sampling frequency was not sufficient to being able to physically see the measurement points.
deliver the required information in a timely manner. Verification was required to instill confidence in the
Replacing the older method with EA, provided con- readings; new sampling techniques and a procedure
tinuous operation and required comparatively little were developed that also followed a learning curve.
operator involvement. Environmental and operational benefits. The
The desalters. In the more demanding environ- Mandan refinery substantially reduced hydrocarbons
ment of the desalters, Amoco had developed a proce- lost to the wastewater system. Equally important, the
dure of purging the emulsion layer from the vessel dur- refinery minimized final effluent benzene by over
ing severe upset periods. Operations staff prided 80%. The EA system has proven to be very accurate
themselves on performing this function; they developed and easy to use. Because the new technology operates
the determination and execution of purging to an art. on a continual basis, the total effectiveness of the refin-
Using EA methods for this application, however, con- ery’s operation has improved. The probes have
verted this art into a science. The EA probes were pre- enhanced Amoco’s sampling techniques, with most of
set to continuously indicate the percentage of water in the sampling no longer lost to the sewer.
the emulsion layer and then automatically determine Improved manpower utilization. Amoco has
the optimum time to purge. noted multiple benefits in the tank farm water drainage
Engineering design, construction and startup. operations. When operators are draining water from a
Installing the probes in the desalters was quite simple, tank, they simply set the probe to close the drain valve
but more significantly did not require any downtime when it detects hydrocarbon approaching the drain
(Fig. 1). The probe entries were hot-tapped onto the ves- line. The operator can leave the area to complete other
sel at normal operating conditions. The hot tapping was tasks with confidence, since the probe can be adjusted
done through a vertical manway, with a specifically to automatically shutdown at the desired hydrocarbon
HYDROCARBON PROCESSING / JUNE 1998
concentration. The system has also been set-up to acti-
vate a control room alarm and/or local signal to let the
$/bbl
operator know the operation is complete. Besides
adding flexibility and efficiency to the operator work
force, the potential for either under- or over-draining Oppportunity cost 1.45
the tanks has been virtually eliminated.
Better control. Other refiners who have retrofitted Processing cost
their desalters with EA technology have observed a
• Fuel 0.67
dramatic shift in operator understanding of desalter
operations. Before using EA methods, desalter upsets • Power 0.21
often remained undetected until they manifested them-
• Cooling water 0.09
selves in downstream equipment disruptions. In
Amoco’s 75,000-bpd refinery, crude cuts of 10,000 bpd • Catalysts/chemicals 0.22
to 15,000 bpd were not uncommon during severe
desalter upsets. Following installation of the improved
Total 2.64
control system, operators reported a greatly enhanced
sense of control over the unit from the monitoring meth-
ods that allowed them to see and respond to potential WWTP/Emulsion breaking 35.00
upsets in a systematic and effective fashion. While
Lost opportunity and reprocessing costs for
desalter operations and tank-water drawing represent typical Gulf coast refinery are $9.64/bbl
about 75% of the source reduction opportunities in a
refinery, similar environmental, profitability and oper- Fig. 4. Reprocessing costs for a typical U.S. Gulf Coast refinery.
ational benefits are possible by applying EA technol-
ogy in any refinery process that involves controlling
hydrocarbon/water separation. monitor the function and output of the probes.) The
most critical factor for Shell was acceptance of the new
Case history two—the Reichstett-Vendenheim instrumentation. Reliance on accuracy was emphasized
Refinery, France. The selection process for new inter- and periodic sampling for verification was encouraged.
face control instrumentation was part of a desalter Debugging, optimization and process control.
system redesign for Shell-advised refineries. The inad- The first installed probes were closely followed to
equacy of existing density-based instruments had been evaluate performance. The immediate success of the
confirmed by failure reports from many sites. During control applications, however, nearly rendered proper
this process, Shell determined that sharp interfaces evaluation impossible. Operations managers did not
required by traditional level instrumentation were allow the probes to be touched for recalibration or set-
not realistic. The problems were aggravated when the tings alterations. The probes were regarded as (and
displacers were mounted in external chambers. Shell’s remain perceived as) “pieces of magic fit for the job.”
multidisciplinary group of specialists decided to eval- Because they required no maintenance, instrument
uate EA, with controlling percentage oil-in-water as engineers did not have the opportunity to become
the focus objective. familiar with them. Therefore, it was advised that the
Technology choice. Shell’s instrument engineers engineers be allowed to experiment with probes des-
initially reacted by simply comparing the cost of EA tined for new installations in workshop configurations
probes with other density-based units. However, the so they could become familiar with the probe behavior
company’s process engineers provided data and refer- before installation.
ences from exploration and production operations. This Improved control. The reliability of the resultant
practical experience documented successful application interface control greatly contributed to the total suc-
of the new technology. The decision-making team began cess of the re-design project. Savings in chemical con-
evaluating all alternatives according to pre-established sumption and reduced oil in the effluent water streams
criteria, among them oil-in-water detection. Capacitance readily justified selecting the new technology. Reduced
technology also had some track record in the field and overhead corrosion and prolonged catalyst life also rep-
has not improved performance of these systems. Even resented substantial savings, but required greater study
lesser-known technologies such as time domain reflec- to quantify. More important to Shell, however, was the
tometry (TDR) were reviewed, but were found to be contribution of the new control systems to improve-
unsuitable for process conditions. The team concluded ments in plant availability, unit throughput, equip-
that none of the available level measurement instru- ment reliability and total quality of control.
ments would accomplish the objectives. Consequently, The EA instruments have since been selected as
the EA oil-water measurement instruments were essential and standard elements for all future re-
selected for testing to confirm performance figures and design projects. For example, the Shell Pernis Refin-
reliability in demanding conditions. ery (Netherlands) became one of the many group-
Engineering design, construction. Because the advised refineries to follow the Reichstett lead, and
standards are not yet equal on all continents, the design it has reported very positive results.
of EA probes was adapted to conform to Shell’s flanged
connection requirements for the Reichstett refinery. Economic benefits. A Hawaiian refinery has used
(For example, most refineries later opted for placing EA technology to reduce their desalting process chem-
sample valves on the connections to enable them to ical feed by an average of 25% to 33% (resulting in a
HYDROCARBON PROCESSING / JUNE 1998
the same study is that refiners using the EA methods
12 to monitor desalter emulsion characteristics better con-
trolled their chemical injection programs. The potential
10 savings from reducing lost opportunity costs, emulsion
Potential savings, million $/yr

handling costs, and desalter chemical costs for an aver-


8 age U.S. Gulf Coast refinery, is $183,000/yr. (This con-
clusion is based on the conservative assumption of an
6 oil undercarry reduction efficiency of 80%.) Per barrel
costs are detailed in Fig. 4.
4 As the lost opportunity and reprocessing costs for a
U.S. Gulf Coast refiner are $9.64/bbl, achieving an 80%
2 reduction in hydrocarbon losses to the sewer from all
Basis: 100,000 bpd
sources equal to only 0.5% of the total crude can yield
0 potential savings of $1.4 million/yr. Fig. 5 shows how the
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 plant-wide savings can increase with higher loss rates.
Total crude charge, $ Eliminating pollutants at the generating source allows
the facility to meet legislated removal requirements
Fig. 5. Plant-wide savings by reducing hydrocarbon losses to sewer. without large capital outlays for downstream treating
equipment. This approach also reduces the operating
costs of existing wastewater handling facilities.
calculated payback of 2.5 months). Another refinery in Refiners and other petrochemical industry opera-
Baton Rouge reported that the payback for its desalter tors feel the pressure of restricted effluent contamina-
control investment took approximately six months. A tion legislation. Source reduction is an attractive alter-
Louisiana refinery experienced more than an 80% native to costly treatment plant extensions and
reduction in effluent brine oil-in-water content (with upgrades. Improved interface control can deliver high
an annualized savings of $200,000). In South Texas, a return on investment through the additional positive
refinery reduced their desalter brine oil content (pre- impact on throughput and equipment availability and
viously ranging from 0.5% to 1.0%) to a level consis- reliability.
tently less than 500 ppm. A Southern California refin-
ery using EA techniques since 1987 reduced their ACKNOWLEDGMENT
demulsifying chemical consumption by more than 30%, Based on a paper originally presented at the Gulf Publishing
Co./Hydrocarbon Processing Second International Conference and Exhi-
while still minimizing effluent oil-in-water content. bition on Process Optimization, March 23–27, 1998, Houston, Texas.
Finally, a Mississippi refiner has reduced crude oil in its
desalter brine effluent to less than 0.5% from a previ-
ous level of 2% to 4%. In each case, the benefits to the
plant reached beyond a simple improvement in effluent The authors
condition. Reduction in slop-oil recovery and repro- Les Helm has 27 years experience with Amoco
and is presently the Transition Manager of
cessing costs significantly improved facility bottom- Amoco’s Mandan Refinery. At the writing of this
line operating costs as well. article, he was the refinery project representative.
Profitability improvement. From the study by Mr. Helm holds a BS degree in mechanical engi-
Wright Killen /Ernst & Young, improvement in prof- neering from North Dakota State University.
itability can be realized by applying source reduction
technology. The data and conclusions are based on inter-
views with representative refineries in the U.S. cur- Calvin Spencer is a consulting process engi-
rently using EA technology. This data was compared neer with Mustang Engineering Inc., located in
with overall industry margins and operational data Houston, Texas. He has 26 years of experience
(Fig. 2), and final conclusions were extrapolated. related to process design, project management,
environmental regulations, control and treatment
The data indicated that conventional single-stage technologies, and process safety management
desalter design and operations typically limit control in refining, petrochemical and specialty chemical
of oil undercarry in the washwater effluent to no less industries. Mr. Spencer holds a BS degree in
than 0.5% to 1.0%. For some refiners, oil undercarry chemical engineering from the University of Texas
at Austin. He is a licensed professional engineer
runs as high as 3.5% to 4.0%. Expressed in terms of in Alabama, Iowa, Louisiana, Pennsylvania and Texas.
total crude charge to the refinery, oil undercarry losses
from desalters range from 0.025% to 0.15% of feed to
the crude units. Fig. 3 illustrates the potential savings Robert T. van den Dungen worked 25 years for
specific to reduction of losses in the desalting process. Shell International B.V. in a broad range of func-
tions and locations. Before the start of his present
As much as 20% of the oil dropped to the sewer will engagement as director/consultant of Dubrae
reach the WWTU. At this point, the remaining oil is in Consultancy B.V. in Aerdenhout, the Netherlands,
emulsion—making it difficult and costly to recover. he was senior consultant Industrial Automation for
Average recovery costs were reported to be $7.00/ bbl field measurements. Currently, he applies his
experience to control of business processes, as
of emulsion with a 20% oil concentration, or $35.00/bbl TQM consultant and assessor for the European
of recovered oil. Foundation for Quality Management.
Chemical-injection program costs. Also, found in

Вам также может понравиться