Вы находитесь на странице: 1из 22

In general* electric submersible pumps (ESP) do not like gas.

Regardless of
the source or nature of gas inclusions in the flowstream* problems can occur
when the gas reaches the ESP. Recent experience has revealed that selecting
the most appropriate gas handling solution is not a simple task. The reason is
that gas inclusions can manifest themselves in many ways from massive
slugs to entrained bubbles* at a broad spectrum of flow rates and pressures.
Different well geometries and completion styles may affect the way gas flows
or agglomerates in the flowstream. Accordingly* each situation must be
evaluated on its own merit.
Because the gas flow regime can be variable* real-time monitoring of flow*
pressure and temperatures provide valuable information in deciding the best
method of gas handling* but it also has significant long-term benefits in
detecting changes in the operation that could lead to premature pump
failure.
ESPs generally have a low tolerance for gas* usually no more than 10% to
20% gas volume fraction (GVF). Because ESPs rely on centrifugal force to
move fluid* generally radial or a combination of radial and axial* high- and
low-pressure areas are created in the pump stage vanes as they rotate. As
gas enters the pump* the lighter gas phase will slip apart from the heavier
fluid phase and accumulate on the low pressure side of the blade until it
finally blocks the passage of the entire vane/vanes in the pump stage (Figure
1).
In this way* as gas enters the intake stage of an ESP* it will eventually cause
the pump to gas lock. Surface equipment used while producing the well can
help alleviate this problem. In the case of a switchboard application fixed
speed the options are limited. However* in a variable speed drive (VSD)*
the speed of the pump can be changed as the pump load varies due to gas
slugging* thus allowing the pump to ride through the slug. In the case
discussed later in this article* the VSD was set up in current mode where the
drive followed a preset current point to track the operation of the pump in
lieu of just shutting the system down. The use of a VSD has its limitations as
well. If the GVF is high enough* it will eventually lead to a gas locking
condition. When this happens* the flow of liquid stops and the result is a
sharp increase in motor winding temperature and a sharp decrease in load
on the motor. If the temperature spike is not detected and steps are not
taken to immediately remediate it* the ESP can be destroyed. If the flow
regime includes gas slugging* the damage can be even more precipitous
(Figure 2).
Dealing with gas

Over the years* operators and their service partners have developed
methods to deal with gas in pumping wells. These techniques vary from
simple to sophisticated. In some cases* the system can be equipped with a
shroud that displaces the pump assembly intake point down below the
source of gas influx (usually the perforations). Free gas entering the well will
experience simple gravity separation and bubble up into the annulus*
allowing the liquid phase to enter into the shroud and flow up to the ESP
intake.

This technique is called gas avoidance* as the shroud minimizes the amount
of gas reaching the pumps intake altogether. In deviated wells* gas entry
can be avoided by using a weighted self-orienting intake that always draws
from the low side of the casing* thus avoiding gas flow that is percolating up
the high side due to gravity separation (Figure 3).

In some cases* gas can be separated in advance of the pump intake either
by forcing all production to take a tortuous path to the intake ports or by
creating a vortex in the flow stream. For higher GVFs* various rotary
techniques can be employed. These add energy to the flowstream to
stimulate the separation process. Centrifugal separators assist in separating
the gas from the liquid and disperse the gas through discharge ports into the
casing to be produced up the annulus.
Advanced gas handlers (AGH) operate on the principle of homogenizing the
produced mixture as it is being passed through the pump* allowing for
increased GVF to be handled in a way that surpasses the effectiveness of the
centrifugal separator. The AGH homogenizes the gas and liquid phases*
compresses a portion of the gas back into solution and induces a gas lift
effect in the tubing above the pump. This enables the pump to produce the
mixture with a limit of approximately 45% GVF without gas locking* reducing
the load on the pump and improving its overall lifting efficiency.
The most effective technique* in terms of gas handling* employs axial flow
technology* which has the capacity to effectively handle GVFs up to 75%.
Developed jointly by Institut Franais du Ptrole (IFP)* Total* and Statoil* the

Poseidon gas handling system is a multiphase helicoaxial pump that fits


between the ESP intake ports and the ESP itself (Fig. 3). The system is
designed for ultra high GVF oil wells or high rate gas wells that require
dewatering and are produced by ESPs. It has been characterized as a
multiphase gas handler because it can deal with a very broad spectrum of
gas flow regimes. It can be installed above a gas separator so separated gas
can be vented to the casing annulus* or directly between the pump intake
and the pump.
The device adds axial velocity to the fluid to the point where the centrifugal
pump can handle it. It essentially charges the first three impellers of the ESP
so that the gas does not accumulate in the vanes and block the passage.
Once charged* the pump can efficiently move the resulting liquid/gas
mixture without losing its prime. Even if a large slug of gas appears* the
pump can usually handle it.
Experience has shown that companies faced with gassy wells should follow
best practices when selecting and implementing gas handling technology.
Users recommend that downhole monitoring of pump performance should be
employed on all ESPs used in high GVF wells. It is possible to monitor
multiple sets of parameters simultaneously using sensors in the pump and
motor* transmitting the resulting data uphole multiplexed on the power
cable. One set of sensors measures pump/motor parameters such as intake
and discharge pressure* motor winding temperature* vibration and current
leakage. The VSD measurements such as current* motor-speed and
frequency are measured and combined on the same set of data for
monitoring and trending. Pump performance can be fine-tuned using the
combination of real-time data monitoring and the use of a variable speed
drive (VSD) at the surface. The VSD provides the capability to control and
manage the ESP during all phases including startup and initial production.
Flow rate* dynamic head and drawdown can be optimized and* in many
cases* gas slugging effects can be limited.
One problem with automated systems is frequent shutdowns and automatic
re-starts. Each time an ESP re-starts its life expectancy is reduced.
Accordingly* it is most desirable to tune the pump so it operates continually*
with re-starts eliminated or at least minimized.
A case in point
Chesapeake Energy faced several gas-related challenges in one of its US
onshore fields. The field includes 26 horizontal producing wells with liquid

production rates varying from 100 b/d to 2000 b/d. Gas/liquid ratios (GLR)
varied as well from as little as 150 scf/stb to 15*000 scf/stb. Clearly* a onesize-fits-all solution was inappropriate each well had to be evaluated
individually and the optimum solution identified. Wells experienced frequent
gas-locking* resulting in lost revenue and production interruptions
Shutdowns were also experienced because of power outages or surges
resulting in field failures. In addition* start up was difficult and getting the
well stabilized was even more of a challenge.
Read more: http://www.egpet.net/vb/threads/47503-Electric-submersible-pumps-fine-tunedfor-gassy-wells#ixzz4U59uUwnB

INTRODUCTION
Static and producing bottom hole pressures are important parameters for
investigating the inflow and vertical lift performances of an oil well. There are
different methods which can be used for determining bottom hole pressures of
electric submersible pumped (ESP) wells:

1) Conventional pressure bomb: used if the ESP completion is equipped with Ytool (by-pass system)* usually the Y-tool is not used due to it's required large
casing (not less than 9 5/8")* and the Y-tool plug can cause leakage and/or
stuck. Generally* this method requires the following steps:
a. Switch-off the ESP*
b. Retrieve the plug of the Y-tool*
c. Install the pressure bomb at the bottom hole of the well*
d. Install the plug of the Y-tool above the pressure bomb*
e. Re-run the ESP and operate it for enough time*
f. Switch-off the ESP*
g. Retrieve the plug of the Y-tool*
h. Retrieve the pressure bomb*

i. Install the plug of the Y-tool* and


j. Re-run the ESP for operating the well.

It is clear enough* at this point* that the previously described method of BHFP
measurement takes more time and it is economically infeasible (loss in oil
production due to shut-in times). In addition* many expected problems may
occur due to many wire-line jobs (six runs).

2) Pressure sensors: which may be installed with the down-hole pump assembly
so that accurate pressure readings may be obtained whenever required* but the
cost would be high if they were used in every pumped well in a field or area.

3) Acoustic well sounding technique: which records fluid level in the annulus*
the bottom hole pressures are usually determined by this method. But this
method can not be used if the annulus is to be packed off (a packer used).

It is clear enough that all the previously described methods of BHFP


measurement require a device. Then* require added money.

Gibbs and Nolen3 and Podio et al4 introduced a well Analyzer's computer and
A/D converter* that can be used in conjunction with an acoustic gas gun and
microphone. The gas gun generates an acoustic pulse in the well-bore and the
microphone converts the reflected acoustic pressure pulses to electrical signals*
which are digitized by the analog to digital A/D converter and stored in the
computer. The computer displays these signals and processes the data as
introduced by software to automatically determine fluid level depth. To calculate
the producing bottom-hole pressure* the casing pressure is measured at the
time of fluid level determination. When liquid is present above the formation
and gas is flowing upward in the casing annulus* the casing vent valve is closed
and sequential measurements of casing-head pressure are made for
approximately 10-15 minutes so that an accurate casing pressure build-up rate
can be obtained. The program uses this rate and the annulus void volume to
calculate the casing annulus gas flow rate. This allows determination of the
gaseous liquid column gradient from empirical correlations. It is noted that the

tubing joints should have the same length because the depth to the fluid level is
computed by estimating the total number of collars from the surface to the fluid
level.

Spath et al5 described a technique to exploit the measurements made below


ESP installations to determine well and reservoir properties. Well testing using
pressure gauges below ESPs is not new; the key advantage of the technique
described here is that the well and reservoir properties may be obtained more
accurately with higher resolution and without shutting the well in. In addition*
the properties may be continuously determined for purposes of real-time
production management. The technique is based on changing the flow rate by
using a variable speed device (VSD)* which changes the frequency (Hz) of the
power supplied to the ESP motor* the production rate of the well is perturbed
about a nominal flow rate. The resulting variation in bottom-hole flowing
pressure is measured and modeled using the appropriate theoretical reservoir
response (type curves) and nonlinear regression. Once the reservoir model is
obtained* the well and reservoir parameters are computed from the regression
analysis; variation in the properties (e.g.* skin* reservoir pressure or distance
to fluid interface) can be continuously monitored. Knowledge of the well and
reservoir properties* and their variation over time* from the described
technique* allows operators to optimize production rates and recovery while
minimizing capital investments and operating expenses.

El-noby16 discussed and evaluated the different practices and applications in


concern of testing the naturally non-flowing wells. He recommend some
appropriate methods to be used for monitoring and optimizing the well
production performance. Also* new correlations have been proposed to identify
the completion factors for each reservoir based on the actual productivity index
measurements for better planning and forecast estimations.

The basic objective of this study is to develop a method capable of calculating


the pump intake pressure (PIP)* the bottom hole flowing pressure (BHFP)* the
dynamic fluid level (DFL) & net liquid above pump (NLAP) for ESP wells using
production and completion data only.

PUMP PERFORMANCE CURVES

The performance curves of a submersible electrical pump (Figure 1) represent


the variation of head* horsepower* and efficiency with capacity. Capacity refers
to the volume of the produced fluid rate* which may include free and/or
dissolved gas. These curves are for a fixed power cycle (normally 50 or 60
cycle) and can be changed with variable frequency controllers.

The head (in feet per stage) developed by a centrifugal pump is the same
regardless of the type or specific gravity of the fluid pumped. But when
converting this head to pressure* it must be multiplied by the gradient of the
fluid in question. Therefore* the following can be stated7:

(pressure developed by pump) = (head per stage) (gradient of fluid)


(number of stages) (1)

The total fluid rate (liquid plus gas) at any conditions of pressure and
temperature is* then:

(2)

Where VF is formation volumes factors.

When pumping gas with the liquid* the capacity and* consequently* the head
per stage as well as the gradient vary as the pressure of the fluid is elevated
from the intake value PIP to the discharge value Pdis. Thus* the above equation
can be written as follows:

(3)

Note that parentheses are included to indicate that h and Gf are functions of the
capacity V* which is given by Equation 2.

The gradient of the fluid at any pressure and temperature is given by:
(4)

but*
(5)

where W is the weight of the capacity V at any pressure and temperature*


which is equal to the weight at standard conditions. Hence:
(6)

Substituting Equation 6 into Equation 4 gives:

(7)

fsc is the weight of 1 bbl of liquid plus pumped gas (per 1 bbl of liquid) at
standard conditions* or:

(8)

Substituting Equation 7 into Equation 3 gives:

(9)

The total number of stages is obtained by integrating the above equation


between the intake and the discharge pressures:

(10)

or:
(11)

For each pump* there is a capacity range within which the pump performs at or
near its peak efficiency (see Figure 1). The volume range of the selected rate
between the intake and the discharge pressures should remain within the
efficiency range of the pump. This range* of course* can be changed by using a
variable frequency controller.

HEAD PER STAGE

Calculating head per stage for submersible pumps is based on the head
coefficients (as shown in Table 1) and the volume of the production rate as
follows:

(12)

The production rate for submersible pumps is considered for two cases:
pumping liquid only and pumping liquid & gas.

For liquid only:


(13)

For liquid and gas:


(14)

The head per stage which calculated from Equation 12 is based on fresh water.
Then* the head per stage for viscous liquid need to the following:

Find equivalent water capacity (Qw) as follows:

(15)

Find the viscous head (h) as follows:

(16)

Where Cq & Ch are viscosity correction factors for capacity & head respectively*
as shown in Table 2.

The head per stage (calculated from Equation 12) is based on 60 hertz motor
frequency. Then* if the motor frequency is more or less than 60 hertz* the head
per stage for this frequency need to the following:

Find equivalent capacity at 60 hertz (Q60) as follows:


(17)

Find the head @ motor frequency (hHZ) as follows:


(18)

TOTAL DYNAMIC HEAD

The total dynamic head developed by the pump is considered for two cases:
pumping liquid only and pumping liquid & gas.

For liquid only* the head per stage is constant for all stages. Therefore* the
total head given by:

(19)

For liquid and gas* the head per stage is changing from stage to stage.
Therefore* the total head given by:
(20)

In case of pumping liquid and gas* the head per stage changes from stage to
stage because the pressure changes from stage to stage* which may affect the
production rate (volume capacity ) from stage to stage. The pressure at the last
stage is equal to the discharge pressure of the pump:

(21)

Where the discharge pressure of the pump can be calculated from the twophase flow correlations.
The pressure at the stage before the last stage is equal to the pressure at the
last stage minus the pressure developed by the last stage* and so on:

(22)

(23)
Finally* the pressure at the first stage is equal to the pressure at the second
stage minus the pressure developed by the second stage:
(24)
.

Recent Threads by ali :

Reply

Crystal Ball software


Incident Commander Simulation software for...
...
Standards and codes required for offshore...
Aspen HYSYS 7.3
Reply With Quote

Blog this Post

Thanks

Sponsored Links

08-29-2011, 01:37 PM#3

ali
Active Member

Join Date
Apr 2007
Posts
218
Blog Entries
1
Thanks
10
Thanked 35 Times in 20 Posts

PRESSURE DEVELOPED BY THE STAGE

Calculating pressure per stage for submersible pumps is based on the head per
stage and the average specific gravity of the liquid:

(25)

where:

(26)

TOTAL PRESSURE DEVELOPED BY THE PUMP

The total pressure developed by the pump is considered for two cases: pumping
liquid only and pumping liquid and gas.

For liquid only* the pressure per stage is constant for all stages. Therefore* the
total pressure given by:

(27)

For liquid and gas* the pressure per stage is changing from stage to stage.
Therefore* the total pressure is given by:

(28)

PUMP INTAKE PRESSURE


The pump intake pressure is equal to the discharge pressure of the pump minus
the total pressure developed by the pump:

(29)

Where* as mentioned before* the discharge pressure of the pump can be


calculated from the two-phase flow correlations.

BOTTOM HOLE FLOWING PRESSURE


Knowing the pump intake pressure (PIP)* the bottom hole flowing pressure at
mid of perforation can be calculated from the two-phase flow correlations
assuming the flow pass into the existing casing.

DYNAMIC FLUID LEVEL

The dynamic fluid level in the well annulus can be determined as the following:

(30)

(31)

(32)

Where :
C = 100 for old pipe
= 120 for new pipe
= 150 for plastic pipe

NET LIQUID ABOVE PUMP

The net liquid above the pump can be determined as the following:

(33)
.
Recent Threads by ali :

Reply

Crystal Ball software


Incident Commander Simulation software for...
...
Standards and codes required for offshore...
Aspen HYSYS 7.3
Reply With Quote

08-29-2011, 01:37 PM#4

ali
Active Member

Blog this Post

Thanks

RESULTS AND DISCUSSION

Join Date
Apr 2007
Posts
218
Blog Entries
1
Thanks
10
Thanked 35 Times in 20 Posts

The new developed method is applied and the designed VISUAL BASIC
computer program is used to calculate BHFP* PIP* DFL* and NLAP.

The required data to calculate the PIP* BHFP* DFL & NLAP showed in Tables 3*
4 & 5. The well bore data showed in Table 3 * The down-hole pump data
showed in Table 4* and the production data showed in Table 5.

The Echometer surveys results from actual producing ESP wells showed in Table
6* these wells were selected from the Western Desert area of Egypt covering a
wide range of a variations in their reservoirs* fluid properties* and ESP
assembly.

Table 7 showed comparison between the measured and calculated values of


DFL* NLAP* PIP & BHFP. The error% is calculated as the following:

(34)

Also* Figures 2* 3* 4* and 5 showed comparison between the measured and


calculated values of DFL* NLAP* PIP & BHFP.

The results indicate that* the absolute error percent between the measured and
calculated values of DFL* NLAP* PIP & BHFP is not more than 10%.

CONCLUSIONS

The pump intake pressure (PIP)* the bottom hole flowing pressure (BHFP)* the
dynamic fluid level (DFL)* and net liquid above pump (NLAP) in annulus for
Electric submersible pumped (ESP) wells can be determined accurately by the
developed method using the production and completion data.

The VISUAL BASIC computer language was used to design a computer program
for an easy use.

Comparing the results obtained by the new method and the measured values of
pump intake pressure (PIP)* bottom hole flowing pressure (BHFP)* dynamic
fluid level (DFL) and net liquid above pump (NLAP) indicated a good agreement
and shows that the new method is reliable and dependable to be used.

ACKNOWLEDGMENT

The authors would like to express their sincere thanks and appreciation to the
Petroleum Engineering Department* Suez Canal University* for encouragement
to publish this paper* and for the staff of the Khalda Petroleum Company (KPC)
for their cooperation and providing field data.

NOMENCLATURE

Bw = the water formation volume factor* bbl/stb


Bo = the oil formation volume factor* bbl/stb
Bg = the gas formation volume factor* bbl/scf
C1* C2* C3* C4* C5 & C6 are head coefficients
DP = pump depth* ft
DFL = dynamic fluid level* ft
dP = the differential pressure developed by the pump* psi
d(St) = the differential number of stages
FT = tubing friction loss* ft
Gf = the gradient of the pumped fluid* psi/ft
GOR = produced gas oil ratio* scf/stb
h = the head per stage* ft/stage
hw60 = head per stage @ 60 hertz* ft (for fresh water)
HZ = the motor frequency* hertz
h60 = the head per stage @ 60 hertz motor frequency* ft
hHZ = the head per stage @ hertz more or less 60 hertz* ft
hst = the head per stage* ft
ht = the total dynamic head* ft
PC = casing pressure* psi
Pn = the pressure at the last stage (stage number n)* psi
Pdis = the pump discharge pressure* psi
Pn-1 = the pressure at the stage before the last stage (stage number n-1)* psi
Pn-2 = the pressure at the stage before the before last stage (number n-2)* psi

P1 = the pressure at the first stage* psi


P2 = the pressure at the second stage* psi
PIP = pump intake pressure* psi
Pdis = pump discharge pressure* psi
Qw60 = the production rate @ 60 hertz* bbl/day (for fresh water)
Q = the production rate @ pump conditions (pressure & temperature)* bbl/day
Qsc = the production rate @ standard conditions* stb/day
Q60 = the production rate @ 60 hertz motor frequency* stb/day
QHZ = the production rate @ hertz more or less 60 hertz* stb/day
Qo = oil production rate* stb/day
Qw = water production rate* stb/day
Rs = the solution gas oil ratio* scf/stb
st = the number of stages
TID = tubing inside diameter* in
WHPH = the well head flowing pressure* ft
WHP = well head flowing pressure* psi

Symbols

w = water specific gravity


avg = average specific gravity of the liquid
g = gas specific gravity

o = oil specific gravity


gsc = the density of gas (in lb/scf) at standard conditions.
Pp = the total pressure developed by the pump* psi
P2 = the pressure developed by the second stage* psi
Pst = the pressure developed by one stage (pressure per stage)* psi
Pn = the pressure developed by the last stage* psi
Pn-1 = the pressure developed by the stage before last one* psi

REFERENCES

1. Hagedorn* A. R.* and Brown* K. E.: Experimental Study of Pressure


Gradients Occurring During Continuous Two-Phase Flow in Small-Diameter
Vertical Conduits* JPT* 475-484* April 1965.

2. Beggs* H. D.* and Brill* J. P.: A Study of Two-Phase Flow in Inclined Pipes*
JPT* 607-617* May 1973.

3. Gibbs* S. G.* and Nolen* K. B.: "Wellsite Diagnosis of Pumping Problems


using Minicomputers*" J.P.T.* PP. 1319-1323* Nov. 1973.

4. Podio* A. L.* McCoy* J. N.* and Dieter Becker: "Integrated Well Performance
and Analysis*" SPE 24060* Western Regional Meeting* Bakersfield* California*
March 30 April 1* 1992.

5. Spath* J. and Martinez* A. D.: "Pressure Transient Technique Adds Value to


ESP Monitoring*" Paper SPE 54306 Prepared for Presentation at the 1999 SPE
Asia Pacific Oil & Gas Conference and Exhibition Held in Jakarta* Indonesia* 20-

22 April 1999.

6. Elnoby* M. G.: "Testing of The Naturally Non-flowing Wells*" Ms. thesis*


Submitted to Petroleum Engineering Department* Faculty of Petroleum and
Mining Engineering* Suez Canal University* Suez* Egypt* 2003.

7. Brown* K. E.* et al.: The Technology of Artificial Lift Methods* Vol 2b.
Chapter 4. Tulsa* Oklahoma: The Petroleum Publishing Co.* 1980.
Read more: http://www.egpet.net/vb/threads/50249-Production-Data-Analysis-for-ElectricSubmersible-Pumped-%28ESP%29-Wells#ixzz4U5AK582I