SPE 166386

Maximizing Oil Production and Increasing ESP Run Life in a Brownfield

Using Real-Time ESP Monitoring and Optimization Software: Rockies Field
Case Study
Dustin E. Ratcliff, SPE, Marathon Oil Company, and Cesar Gomez, SPE, Ivan Cetkovic, SPE, and Odafe
Madogwe, SPE, Weatherford

Copyright 2013, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Annual Technical Conference and Exhibition held in New Orleans, Louisiana, USA, 30 September2 October 2013.
This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
Marathon Oil operates several brownfields in the Rockies, most of which are approaching a producing life of 100 years. Most
of the fields are on waterflood, with 86% of the oil production coming from the artificial lift method of electrical submersible
pumps (ESPs). This ESP lift method has helped maximize the production of recoverable oil in these fields, achieving higher
drawdowns and larger fluid volume handling.
Prior to 2010, Marathons 770 ESPs were operated with out downhole gauges or sensors and no ESP real time systems were
in use. This made it difficult to properly monitor the well and the equipment, leading to improperly sized ESPs, limited
artificial lift optimization and lost oil production. In 2010 Marathon began installing downhole sensors in all ESP wells that
were pulled for either a failure or a workover and download the data into an ESP real-time monitoring, diagnosis and analysis
software system.
At the time of this paper, Marathon has installed downhole sensors in 430 of the 770 ESP wells. All ESP wells are presently
being monitored via the ESP real-time monitoring and analysis software. During this short period of real-time monitoring,
Marathon has realized production gains of more than 700 barrels of oil per day (BOPD) in one of the fields.
This paper compares previous operations to the results and data from the installation and connection of downhole sensors to
the real-time ESP monitoring program. Monitoring methodologies, analysis and proper ESP optimization are detailed in
several case studies.
Introduction
By some industry estimates, more than 90 percent of all producing U.S. oil wells require some form of artificial lift to bring
hydrocarbon from the reservoir to the surface.i Electrical submersible pumps represent one common form of lift, and consist
of a downhole pump, motor, seal and cables. The pump itself is a centrifugal multistage unit that increases the pressure of the
well fluid running through it and pushes it to the surface. Surface components of an ESP system include a variable-speed
motor controller, surface cables and transformers.ii As implied by its name, an ESP is submerged in the produced fluids it is
pumping to the surface. A main advantage of this type of pump is thatunlike other types of artificial lift systems like
reciprocating rod lift (RRL) or progressing cavity pumping (PCP) unitsit can handle the large volumes of fluid often
associated with waterflooding.
The industrys acceptance and use of ESPs has increased with their reliability, which is tied to improved means of monitoring
and controlling these systems. With these improvements have come enhanced abilities to diagnose and evaluate their
operational efficiency.iii While ESP monitoring and control is a critical component in their management, there are different
levels of monitoring commonly used in the industry. These levels vary in complexity, information provided and amount of
automation.
The first level is to have no downhole monitoring of the ESP operation, which carries significant operational risks and may
prevent the well from producing at optimal rates. The operator will not know that a problem exists until they observe a
decline in production rates or, in the worst case, the artificial lift system fails. At this point, the operator must decide if the
wells remaining production potential warrants intervention. For an onshore well, this requires bringing a pulling unit to the
well for a minimum of two days, typically at a cost of US$20,000. Replacing one or more damaged components of the ESP
system might cost the operator an additional US$50,000 or more. These cost estimates are dramatically higher for ESPrelated workovers in offshore wells, or in remote onshore locations. Workovers in these locations may take one to two weeks

SPE 166386

to complete and cost upwards of US$300,000. This estimate does not include the additional cost associated with deferred or
lost production while the well is offline.
The next ESP monitoring level involves measuring the current from the motor using traditional circular amperage charts.
These charts might be generated on a weekly or monthly basis at the well site. Often a technician changes out the charts and
brings them to the field office for interpretation. This monitoring method presents several practical problems, starting with
the tendency for field offices to accumulate these charts but not regularly review them. In addition, the method of manual
retrieval of these charts in the field does not allow for rapid response to a problem. A malfunctioning ESP might not be
detected for several weeks using this method of amp chart retrieval and analysis, which increases the risk of production
declines and the need for a costly well intervention.
A third monitoring level comes from installing downhole ESP sensors and connecting them to remote terminal units (RTUs).
These downhole sensors work by bleeding current to the ground; one can determine the status of the ESP based on the
amount of current bled to the ground. The sensor contains variable resistors that change value with changes in temperature or
pressure, which in turn alters the amount of voltage required to push a preset DC current through the resistors. Therefore, by
monitoring the voltage of the DC generator, one can infer the resistance of the downhole sensor, and determine not only
current, but also the temperature, pressure and vibrations in the pump. While downhole sensors add cost and complexity to
the ESP system, they provide several operational benefits, including longer ESP run life, better ESP system diagnostics,
valuable reservoir pressure data and oil production optimization.
The most basic downhole sensor configuration still requires a technician to download the sensor data at the well (the RTU
can generally store one months worth of data) and bring it back to the office for analysis. While this option provides more
data than a simple amp chart, it is still limited in that the information is not available in a timely manner. Analyzing this data
requires an experienced technician or engineer who is well versed in ESP operations. These employees typically may not
have the time to complete a thorough analysis, given their other job responsibilities.
An increased level of monitoring comes by connecting the downhole sensor to a SCADA (supervisory control and data
acquisition) system. The system continuously monitors well behavior to provide a basic level of automated control. SCADA
systems have been used in the utilities industry in the U.S. since the 1960s, and have found more widespread use in the oil
and gas industry in the last decade.iv The simplest systems cover the fundamentals of well monitoring, including an indication
of whether the well is actually producing and if the ESP is running, and transmits this information back to the field office
without the need for a dedicated technician to go to the well. SCADA systems can also notify field technicians of a problem
by generating an alarm, which allows them to visit the well much less frequently or only when a repair is required.
Bringing Intelligence to ESPs
Weatherford has taken this level of automation a step further by integrating SCADA systems with the LOWIS (Life of Well
Information Software) platform. LOWIS is a web-based, real-time management suite that provides comprehensive analysis
to keep ESP pumps running at optimal levels for longer periods of time. The platform is designed to perform automated
diagnostics, in which the software combines the various signals coming from the field into a mathematical algorithm. This
algorithm tracks trends in the real-time data and compares it to historical data sets.
When one examines the historical trend of ESP failures, the majority of them are repetitive.v The algorithm allows for the
development of an agreed-upon, mathematically tested pattern to all of the common ESP failures. If the current well data
shows a trend that matches the pattern for a historic failure, the algorithm generates a clear alarm for the specific problem
and delivers it to the field technicians via email or a mobile device, at any time of day. The technician can then quickly
pinpoint the abnormal ESP operation, prior to a shutdown or failure, and take appropriate corrective actions. Quickly
returning the system to normal operations minimizes downtime and associated production deferment. The time saved in
surveillance and diagnostics can be redirected to problem remediation and other field activities that further advance
production optimization.
The intelligent monitoring and alarm capabilities of the system not only limits downtime and deferred production, but also
prevents premature failures and unnecessary cycling of the ESP. The number of start/stop cycles of an ESP directly impacts
its run life: the more frequently a pump is stopped and started, the shorter its run life. By monitoring operating efficiency
with LOWIS software, an operator may add years to the ESPs operating life.
The software platform comes with a set of 16 intelligent alarms, which are customized to a fields particular monitoring
needs. These alarms allow for fast and automated operational corrections to maintain optimal pump operation. A typical
correction might include, decreasing or increasing pump speed, or injecting water through the casing to lower the gas fraction
in the pump and prevent gas lock. Gas ingress to the pump is a major cause of failure; the system can accurately measure the
volume of gas entering the ESP and alert the operator if this volume reaches critical levels. To prevent prolonged exposure to
this gas and severe damage to the pump and motor, the system gives the operator the option to shut the well down remotely,
rather than send a crew to the wellsite. The operator can then review how much gas is moving through the different pump
stages and decide if a more permanent solution, such as the addition of a gas separator, is required prior to restarting the well.
To be truly effective as a production optimization tool, ESP monitoring and the establishment of alarm set points should not
be considered a one-and-done process. Initial set points are most often not acceptable throughout the wells operating life.
Alarm set points should be periodically reviewed and maintained under a management-of-change process. A review of the set
points should be completed and updated in the LOWIS software when there is a change in well status, if a well is worked

SPE 166386

over or recompleted, or when repeated alarms are occurring for a specific well or wells.
Monitoring ESPs in the fieldMarathon Oil
Marathon Oil has more than 100 years of exploration and development experience in the Rocky Mountain (Rockies) region
of the U.S. Almost all of Marathons oil wells in this region require some form of artificial lift. As Fig. 1 shows,
approximately 86% of oil production comes from ESP lift and 14% from rod lift. Prior to 2010, most ESP wells were being
monitored utilizing weekly amp charts.

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

86%
ESPLift

14%
RodLift
PercentageofOilProductionBasedonLiftType

Fig. 1. Marathons percentage of oil production based on lift type in the U.S. Rockies.

Starting in 2010, Weatherford and Marathon Oil embarked on a program to enhance the monitoring and operation of
Marathon ESP-supported oil wells. Marathon operates a total of 770 wells with ESPs, located in a number of Wyoming oil
fields.
Table 1a shows the fields where the wells were located and the major oil reservoirs in each. Table 1b highlights the general
operating ranges of the wells across all fields. Given the relatively high liquid rates (from 600 to 11,000 stb/d) and very high
water cuts (98-99%) for these wells, applicability of ESP artificial lift is apparent.
Table 1a. Marathon fields and oil reservoirs using ESPs
Name

Major Oil Reservoirs

Byron/Garland

Phosphoria, Tensleep, Madison

Oregon Basin

Phosphoria, Tensleep, Madison

Spring Creek

Phosphoria, Tensleep

Pitchfork

Phosphoria, Tensleep

Grass Creek

Frontier, Curtis, Phosphoria, Tensleep, Madison

Steamboat Butte

Nugget, Phosphoria, Tensleep

Maverick Springs/Chatterton

Phosphoria, Tensleep

Circle Ridge

Phosphoria, Tensleep

SPE 166386

Table 1b. Range of Operating Conditions, ESP Wells

ESP installation depth
Oil API gravity
Water cut
Liquid rates
Intake pressure
Reservoir pressures
Well temperatures
Gas-to-oil ratio
Oil daily production
Water daily production

800 to 7,000 ft
15 to 28
98 to 99%
600 to 11,000 stb/d
75 psi
200 to 1,500 psi
80 to 140 F
3 to100 scf/stb
19,500 stb/d
1,900,000 stb/d

Case study 1Undersized ESP

A wells initial production rate of about 1500 stb/d in 2005 declined to about 1300 stb/day in 2010. Due to the continuous
decline in production rates, a decision was made to pull the ESP and install a downhole sensor and monitor the data via
LOWIS software. While monitoring the data, it was observed that there was consistently high pump intake pressure during
production (Fig. 2).
Further analysis of the pump data recorded by the monitoring system suggested that the fluid level over the pump was
excessively high, indicating the ESP was undersized for the wells productivity. The operator used this information to select a
larger pump that was installed in 2011 (red line on Fig. 3). The downhole sensor immediately recorded a lower head pressure
on the pump, which resulted in a dramatic rise in production levels (Fig. 3).
The increased production was substantiated by the operating analysis chart in the LOWIS software (Fig. 4). This chart shows
the pump curve and the well curve with the minimum and maximum recommended ranges of the pump also indicated. The
red circle indicates the operating point or rate while the blue triangle indicates the calculated ideal rate. The operating point is
in the middle of the operating range of the pump, showing it is correctly sized for this well. Over 60 BOPD of increased
production occurred from this work.

Fig. 2. LOWIS analog charts showing pump intake pressure, current and motor temperature readings.

SPE 166386

330

4500

300

4000

270
240
210

3500
3000

180

2500

150

2000

120
90
60
30
0

1500

WaterBbls /day

OilBbls /day

1000
500
0

Oil
Water

Fig. 3. Oil and water production rates for a Marathon well with an undersized ESP. Upon the installation of a larger ESP in mid 2011,
production rates rose dramatically.

Fig. 4. Pump curve with well curve overlaid. Pumps operate most efficiently where the pump operates at or close to the best
efficiency point. The initial ESP was too small for the fluid rates in the well, which resulted in the pump head pressure being too
high.

Case study 2Achieving proper frequency control

In another well, the downhole sensor and automated monitoring system indicated that there were large fluctuations in oil
production rates with only small changes in PIP (Pump Intake Pressure) (Fig. 5). This prompted the operator to install a
variable speed drive which was set to operate in PID (Proportional-Intergral-Derivative) mode following PIP. In this case
PID mode automatically adjusted the motor frequency according to the changes in pump intake pressure. This enables proper
control of the rates resulting in better stabilized production from the well. An operating analysis chart generated in LOWIS
software provided minimum and maximum frequency limits for the VSD to ensure optimal frequency control (Fig. 6). Based
on the optimization for this well, approximately 15 BOPD of increased production was achieved.

SPE 166386

40
38
36
34
32
30
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0

250
230
210
190
170
150
130
110
90
70
50

PumpIntakePressure(PSI)

OilBbls/day

30
10
10
Oil
PumpIntake
Pressure

Fig. 5. Large fluctuations in oil production with only small change in PIP.

Fig. 6. Multifrequency curve for ESP experiencing large fluctuations in oil production after installing a variable speed drive.

Case study 3Identifying a tubing leak

A LOWIS system tracking pump pressure and temperature in an oil well identified both high motor temperature and high
pump intake pressure, as indicated in an analog trend chart (Fig. 7). The system generated an alarm prompting Marathon
technicians to check the well. The well was not surfacing fluid, which indicated that production was being diverted downhole
through a tubing leak. This would further account for the high motor temperature and PIP readings.
The technicians pulled the well and found a tubing leak located 30 joints from the surface. Thanks to the prompt notification
provided by LOWIS software, the operator was able to quickly pinpoint the problem, fix the leak, replace the ESP seal and

SPE 166386

re-run the same equipment. Production resumed and the ESP continued operating within safe temperature and PIP levels.
This case indicates the value of automatic monitoring to limit ESP equipment failures and deferred production.

Fig. 7: LOWIS charts showing the PIP and motor temperature trends.

Fig. 8. LOWIS analog status screen, showing the motor-temperature and PIP analogs in red to indicate the alarm.

Case study 4Identifying plugged intake

While monitoring ESP performance on a remote well, the LOWIS status screen showed a rise in PIP levels (Fig. 9). Because
the well had a history of scaling issues, Marathon engineers believed the PIP increase and the drop in liquid rate (Fig. 10)
were a result of scale-induced plugging of the ESPs intake. The monitoring system allowed the engineers to shut down the
well remotely from the office. This eliminated the need to send personnel to the well for a manual shut-down. A shut-down
was performed in an attempt to clear the pump intake.

SPE 166386

WaterBbls/day

OilBbls/day

Fig. 9. LOWIS status screen showing a dangerous rise in PIP values for the ESP, which prompted Marathon to shut down the well
remotely.

Fig. 10. Production chart with highlighted area (circle) showing when the ESP was operating with a plugged intake.

Case study 5Identifying an underload situation

An ESP well was cycling on underload shutdown every two hours (Fig. 11). The information from LOWIS software was
used to investigate the root cause. PIP and fluid production data indicated that there was sufficient fluid to produce (Fig. 12).
This suggested a problem of high gas rates through the pump, which might lead to gas lock and trigger the shutdown.
Marathon pulled the well and installed a gas separator, re-running the same size of pump equipment. Upon startup, the well
experienced a production increase of roughly 17 BOPD, with no further shutdowns of the ESP.

SPE 166386

WaterBbls/day

OilBbls/day

Fig. 11. LOWIS screen shot, showing two-hour intervals of underload shutdown for the ESP.

Fig. 12. Data from the downhole sensor showed that there was sufficient fluid left in the well to produce, which enabled Marathon to
make the correct decision and run a gas separator downhole.

Case study 6Low static bottomhole pressure well

A screenshot of the static pump intake pressure in a LOWIS analog chart (Fig. 13) from downhole sensor data showed a
value of approximately 280 psi. A static bottomhole pressure calculation of 290 psi was calculated using LOWIS analysis
workbench (Fig. 14). When the well was started, the operator noticed that the producing bottomhole pressure was only 40 psi
less than the static value.

10

SPE 166386

Fig. 13. Static bottomhole pressure data generated in LOWIS software, using ESP sensor data.

Fig 14. Static bottomhole pressure calculation in LOWIS software.

Prior to running the sensor, erroneous fluid level shots were indicating approximately 180 ft of fluid over the pump, and
therefore the operator believed the system was optimized. After running the sensor and finding this small difference between
static and producing bottomhole pressures, Marathon pulled the original ESP and upsized. This work provided an oil
production increase of roughly 25 BOPD.
Conclusions
At the time of this paper, Marathon has installed downhole sensors in 430 of their 770 total ESP wells. The automated
monitoring and diagnostic capabilities of LOWIS software integrated with these sensors has helped Marathon increase oil
production by approximately 700 bopd (15%) in just one of the fields. By optimizing ESP operations and the drawdown
efficiencies of the wells both total fluid and oil rate were increased while leveling off the WOR (Water to Oil Ratio) trend
(Fig. 15). This production increase would not have been possible without the combined benefit of downhole sensors and the
monitoring system.

SPE 166386

11

Fig. 15. Field-wide production of oil has steadily increased with the introduction of downhole ESP sensors and the LOWIS system
(green curve, top chart). In addition, the water-to-oil ratio, which had been climbing for years, has leveled off (bottom chart).

Marathon plans to install downhole sensors and monitoring systems on other wells in these fields as the wells are taken
offline for workovers. The operator is also investigating the benefit of downhole sensors to better understand the well
behavior during waterflooding. This monitoring pilot has already shown a positive trend in multiwell production response as
water injection increased.
Ultimately, thanks to the combined effect of the downhole sensors and real-time monitoring system, Marathon has enjoyed
increased production rates and lower operating costs with streamlined field operations. Field technicians are working more
efficiently and safely, focusing more time and effort on other projects that will provide further field improvements.
Acknowledgements
The authors would like to thank Marathon Oil Company for agreeing to share their field experiences and data in preparation
of this paper.
i
ii
iii

iv

Dickson, Brian. Advanced Technologies Optimize Artificial Lift, Production Operations, American Oil & Gas Reporter, June 2010.
Lyons, William C., ed. Standard Handbook of Petroleum & Natural Gas Engineering 2, 6th ed. (1996).
Molotkov, R.V., Stephenson, G.B. and Seale, S.R. ESP Monitoring and Diagnosis Enables Operational Efficiency, paper SPE
115340 presented at 2008 SPE Russian Oil & Gas Technical Conference and Exhibition, Moscow, Russia, 28-30 October 2008.
Supervisory Control and Data Acquisition (SCADA) Systems, Technical Information Bulletin 04-1, National Communications
System, Arlington, VA, October 2004.
Al-Muqbali, H., Awaid, A., Al-Bimani, A., Al-Yazeedi, Z., Al-Sukaity, H., Al-Harthy, K. ESP Well Surveillance using Pattern
Recognition AnalysisPDO, 2013 SPEGulf Coast Section Electric Submersible Pump Workshop, The Woodlands, TX, 24-26 April
2013.