ESP  Globe Logo

William Guerra, Edeberto Finol, Javier Solorzano, Wilson Benavides – Pacific Stratus Energy; Sandy Williams, Artificial Lift Performance Ltd

This paper was prepared for presentation at the 2015 Society of Petroleum Engineers – Gulf Coast Section Electric Submersible Pump Workshop held in The Woodlands, Texas   April 22-24, 2015.
This paper was selected for presentation by the ESP Workshop Panels (Rotating and Permanent) following review of information contained in an abstract submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the ESP Workshop Panels and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the ESP Workshop or its panel members. The author(s) retain copyright to this paper and have given permission to the ESP Workshop to publish it in proceedings (electronic and hardcopy). Any other electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the author(s) is prohibited.


Optimisation of production, runlife increase and minimisation of downtime results in the lowest $/bbl cost of production, which is the goal for any operator using ESPs. This paper will present the results of a 2 month pilot program to optimise ESPs, using real-time data from a SCADA system and PI historian (called SIGO), pushed real-time to an ESP analysis engine (Pump Checker), which provided automated ESP analysis for every new well test.

The ESP analysis application diagnoses degradation in the ESP due to effects such as wear, deposition, plugging, spinning diffusers or even wrong pump rotation. As well as quantifying the amount of pump degradation, the software determines the amount of lost production (gross and net) and validates well IPR.

Pacific Stratus Energy (PSE) views the results of such analysis in a data room, called the “WOW” room where well and ESP performance can be monitored 24 hours per day. Initially, this project was implemented for a small well population of 3 wells in the Curito field.

This paper will present the results of what was determined to be a very successful project, which has now been implemented across more PSE wells in the Yenac, Mantis, Corcel, Guatiquia and Cubiro fields. The use of an ESP analysis engine across more PSE wells, has had a very positive impact on proactively managing a large quantity of ESPs, resulting in identification of production increase opportunities, early recognition of lost production and the ability to make changes to regain production.


The Curito field is located in the Casanare East Block in the Central Llanos Basin of Colombia as shown in Figure 1. The field was discovered in 2013 and so far three successful wells have been drilled and completed in the field. The wells target the Carbonera C7 formation at a depth of 8300-8500ft.

Production from the 3 wells is approximately 15,000 bfpd (3000 bopd) of 34.5 oAPI crude, with almost no associated gas. As with most Llanos wells, there is a strong underlying aquifer, which results in an early and rapid increase in water cut. The wells have been completed with ESPs sized for a nominal flowrate of 8500 bfpd at 60 Hz.

The cost to complete one of these wells with an ESP is approximately $1.4 million. It is critical to the economic success and future development of this field, that the ESPs have long runlives, early failures are prevented, and that production is optimised.

In order to achieve these outcomes it was decided to implement a full automation pilot program on Curito. Figure 2 provides one representation of what can be achieved through real-time data monitoring using automation. The system implemented by PSE consisted of a SCADA system and data historian linked to the ESP analysis engine.


Figure 1 – Field Location in Colombia

The ESP analysis engine was selected as the preferred application for this project because it goes beyond traditional nodal analysis, as it is specifically for the purpose of ESP diagnosis and troubleshooting.

Early diagnosis of a problem in an ESP and a determination of the lost production can give predictive indication of a catastrophic failure and allows the operator to make proactive interventions on a risk & economic basis. Proactive intervention allows orderly coordination of replacement equipment and a rig, rather than being exposed to the reactive chaos of getting a well back in production when it fails.


Figure 2 – Hierarchy of real time activities[1]


This paper proposes that we can move beyond Nodal Analysis using the ESP analysis engine. The following sections will discuss: the theory of how pump the ESP analysis engine performs analysis and why it is better than Nodal Analysis; the results of the pilot program; details of a wider implementation and; the next steps for the project.

Nodal Analysis vs. ESP Analysis Engine

Nodal analysis is a very useful tool for performing artificial lift design and sensitivities to predict future rates and is used in all ESP design software. The problem with nodal analysis and pumps is that when the well production rate does not match what is predicted by design software, we do not know whether this is because the reservoir performance is different, whether there is a problem in the pump, or whether the operating data is incorrect.


The ESP analysis engine uses a technique that goes beyond traditional nodal analysis to:

  • Validate ESP performance;
  • Quantify degradation in the pump resulting in less production;
  • Determine the amount of lost production (gross and net) and;
  • Calculate a real bottomhole flowing pressure (Pwf) for use in future design work.

This technique is specifically for troubleshooting and diagnosis of ESPs[2,3] and is a workflow or methodology that does not exist in most ESP design software or nodal analysis programs.

The process used by the ESP analysis engine is as follows:

  1. Pressure and depth are calculated in the wellbore starting with the wellhead pressure and temperature and performing a top down calculation using standard industry PVT and multiphase flow correlations
  2. The discharge pressure of the pump is determined (or real value used, if a discharge pressure sensor is available) and the measured pump intake pressure is used to determine the actual pressure differential (dP) and TDH across the pump.
  3. The production rates measured at surface are converted to downhole flowrates using PVT and multiphase flow correlations
  4. Using the pump curve (composite if using taper pumps), the theoretical TDH that the pump should produce for the downhole flowrate is determined (see Figure 3)
  5. Theoretical and actual TDH and Pump dP (pressure differential = TDH x density of fluid in pump) are determined and plotted on the gradient traverse plot and the pump curve. When a pump has a problem resulting in a production loss, the measured intake pressure will always be higher than the theoretical intake pressure. See Figure 3 and Figure 4.
  6. Where a difference in theoretical TDH vs. actual TDH exists, this value is quantified in terms of pump degradation and the consequent lost production is determined. See Figure 5 and Table 2.
  7. The actual Pwf is calculated based on the measured intake pressure and a productivity index (PI) value determined for the test. Bad test data (rate) results in a PI determination outwith the normal range and is indicative of a bad test; this can be used to validate well tests.
  8. The ESP analysis engine then takes all the wells in the field and provides a summary of the analysis of the last test, ranked by lost oil. This table provides a prioritised summary of the ESPs with a problem that are resulting in lost production. See Table 3.


Figure 3 – Using the downhole rate and pump curve the theoretical TDH is determined


Figure 4 – What the pump should do vs. actual performance


Figure 5 – Key aspects of the gradient traverse plot


Table 2 – Pump degradation and consequent lost oil quantified

Armed with the knowledge of the ESPs with problems, this information can be used to evolve the specification of future ESP systems and/or allows an economic evaluation to determine whether a proactive intervention is required.


Table 3 – Ranked list of ESPs with degradation resulting in lost production

Automation Process

One of the principle goals behind the pilot was to have all the information loaded automatically into the ESP analysis engine. Figure 6 shows the process that was developed for upload of information, this process caters for automatic upload of data from SIGO, whilst still allowing multiple well and test upload for wells that do not have SIGO and manual entry/edit of data within the ESP analysis engine.


Figure 6 – Process to automatically load data into the analysis engine

Once data is uploaded a series of data quality checks are run to check that the data set is complete and that the values conform to the expected range. Well configuration and test data that fails the checks is flagged in the header of all views so that they are very ‘obvious’, the failed wells and failed tests that fail the data check are highlighted so that the vales can easily be corrected at source.

The ESP analysis engine then performs this analysis on every well automatically, whenever a new well test is available. In Curito, where there were only three wells, performing such analysis manually every time there is a new test is feasible. However, as the number of wells increases, automatic analysis of the data becomes critical.

Automated analysis frees up engineers from manual data entry and processing of data, to allow them to focus on investigating problems, and taking action to restore lost production. See Figure 7 and Figure 8.



This automated ESP performance analysis has the potential to add huge business value for PSE, where data from 200 wells has to be handled. The use of the ESP analysis engine allows the management of ESPs to move along the maturity level from being reactive to proactive or managed as shown in Figure 9.


The engineers using the ESP analysis engine found that they needed to see multiple views and screens at one time and implemented a collaborative workspace (WOW room) with multiple screens to help them view their well data. See Figure 10. The engineers meet in this room to review their wells.


Figure 10 – WOW room: used for well review process

Pilot Project Results

Prior to implementing the ESP analysis engine, other software had been used to perform analysis of PSE wells. One of the first steps performed in evaluating The ESP analysis engine was to analyse tests using the incumbent software and The ESP analysis engine. The results from The ESP analysis engine were found to be within 3 psi of the results from other software, which verified the accuracy of the ESP analysis engine calculations.

For the engineers using the ESP analysis engine they were able to save significant time analysing well production test results. Prior to The ESP analysis engine, analysis of a well in other software required manual data entry and could take up to two hours per test. The ESP analysis engine provides the results automatically, in a format that is easy to understand, saving engineers time. Both field and office personnel found the tool very easy to use.

After completion of the pilot on Curito a comprehensive review was conducted to evaluate the business benefit for PSE.

Curito wells are currently production constrained due to facility limitations. It is difficult to perform accurate well tests as the ability to handle/measure the water production during testing. All three wells have high intake pressure and very little drawdown, higher production is possible, as and when improved fluid handling can be implemented at surface.

All three ESPs have evidence of pump degradation, believed to be due to wear, because of sand production. Solids production levels of up to 80-400 PTB, with occasional peaks of 1100 PTB, have been measured from surface sampling on all three wells.  These levels of solids are high and expose the ESPs to operating conditions, which are categorized as ‘SEVERE’ per Table 4[4]

Quantification of the amount of degradation allows the degree of wear to be tracked with time and can be used to proactively indicate impending failure of the ESP system.

From automated analysis of all the historic production tests it was possible to identify incorrect tests, using the determination of PI for each test. See PI scatter vs. trend in Figure 11, wells lying away from the trend are incorrect test measurements. As a result of this process, it was decided to develop an automatic well test validation routine, to identify bad tests proactively and provide the ability to exclude bad tests.Table 4 – Effect of sand concentration on pump wear


Figure 11 – Scatter of PI determination vs. trend can be used to recognise good vs. bad well test data

On Curito-1 the pump degradation was determined to be 23%, resulting in a higher intake pressure and lost production. The amount of lost oil due to such degradation was quantified to be what is now 93 bopd (previously 253 bopd during the pilot). Curito 2 has 14 % degradation and Curito 3 has 11% degradation, the degradation levels in each well are congruent with the amount of time that the ESP in each well has been producing.

As the pump has worn, the field personnel have slowly increased the frequency (see Figure 13) to offset the pump wear and maintain production rates. As frequency is increased, more HP is required and at some point in time surface power limitations will result in being unable to increase the frequency further and production will decline. An additional concern is that pump wear due to solids is proportional to the cube of frequency; speeding up the pump is actually increasing pump wear.


Figure 12 – Results of analysis of Curito 1


Figure 13 – Frequency increase to compensate for pump wear and maintain production

The main benefit for PSE in Curito is that it was possible to get an early determination of pump wear from the first well in this new field. The implication is that future ESPs in this field should:

  • have enhanced stabilisation (1:1) using tungsten carbide or silicon carbide
  • use stage coatings or enhanced metallurgy resistant to wear
  • should be over-staged and designed to operate at low frequency

This knowledge allows PSE to start talking to equipment providers and ensure that they will have such equipment in Colombia by the time the next Curito well is completed, or a workover on one of the first 3 wells is required.

Since a workover on these wells costs $1.4MM, knowing proactively that solids resistant equipment is required, has significant business value in terms of longevity of the ESPs and maximising production in the presence of solids. These upgrades to the pump specification added roughly 30% to the cost of the pump section.

Post Pilot Results

Following what was judged to be a successful pilot project, PSE took the decision to move to a wider deployment across a greater number of their wells. At the time of writing there are now 77 wells using the ESP analysis engine.

During the pilot a number of recommendations were made by PSE engineers (and other operators) using the product that led to additional functionality being developed and included as part of the ESP analysis engine. Such enhancements included;

  • A dashboard to give a snapshot of field / well and optimisation opportunities
  • The ability to predict gross and net oil rates at higher frequency as well as % drawdown
  • A process to validate new well tests and identify bad tests, which can then be ‘disabled’ in the engine
  • An optimisation workflow that results in automated recommendations based on frequency increase and proactive pump change opportunities

These changes were implemented and added additional benefit, from use of the ESP analysis engine to assist in managing by exception and identifying candidates for production increase. The dashboard (Figure 14) gives an immediate indication of:

  • Number of wells on system
  • Gross and net production from all ESP wells
  • Additional oil from speed changes to ESPs
  • Additional oil form workovers
  • How recently well have been tested
  • ESPs operating in and out of range
  • Number of wells and level of degradation
  • Tests that have failed the import process


Figure 14 – ESP Analysis Engine Dashboard

The ESP analysis engine has been coupled with a predictive calculation engine to:

  1. Calculate the production for a degraded pump and catalogue pump at increased frequencies
  2. Use an “Optimisation Opportunities” workflow to generate a recommendations table prioritised by additional oil opportunity which categorises opportunity based on: optimisation of the existing ESP; optimisation opportunities where a pump change is recommended and; wells with limited opportunity.

In January 2015, due to the continued low oil price, a renewed focus was placed on reducing costs and getting more oil from existing wells. Based on the recommendations table, a review was performed on the existing 44 wells on the system at that time, 13 wells provided opportunities (29 % of the population).

Implementing the production increasing opportunities was complicated by the following factors:

  • The ability to handle more produced water on a field by field basis
  • A reluctance to produce wells at higher drawdown because of fear of coning in water if drawing wells down harder
  • Concerns that some of the wells were showing effects of increasing damage.

The changes to the wells have been performed gradually in small steps to monitor effect on watercut. A surprising fact was that several of the wells showed decreased water cut at higher frequency!

Based on the diagnosis from the ESP analysis engine, the premise that some wells were showing increasing damage was questioned and instead it was proposed that reservoir depletion was occurring. This was subsequently confirmed to be correct and validated through pressure transient analysis

The proposed optimisation opportunities are still being implemented but as of March 2015 the changes have resulting in an additional 1052 bopd (see Table 5). Two wells were also identified that can result in large production increases, but workovers will be required to realise these gains.


Even at today’s reduced oil price this is a substantial production and revenue add for the operator and hugely outweighs the cost of the ESP analysis engine.

However, this has not been the only benefit. Most of the PSE fields are characterised by having high water production and the majority of fields were limited by their capability to handle produced water. In order to achieve these production gains wells had to be switched off or their output reduced.  As a result of theses optimisation changes the total water across the fields was reduced by 2103 bwpd reducing fluid treatment and disposal costs.


The information presented in this paper provides a summary of the results of a pilot to test a new ESP engine in combination with real-time ESP operating data and well test information for 3 wells in the Curito field.  The project was successful and the ESP analysis platform has now been implemented on additional fields and wells.

The key conclusions from this project are:

  1. The ESP analysis engine was tested against other well modelling software and was accurate, giving confidence in the accuracy of its calculations.
  2. PSE engineers found the ESP analysis engine intuitive, easy to use and saved lots of time by having all well tests automatically analysed
  3. Using the ESP analysis engine, the degree of degradation in the pump due to wear is quantifiable and progression of degradation can be tracked with time as well as lost oil.
  4. Using the results of the ESP analysis engine analysis, PSE have a basis for a revised equipment specification that can be used to ensure that vendors have appropriate equipment in Colombia for Curito wells.
  5. The ESP analysis provides a Validated well IPR for future ESP design.
  6. The ESP analysis engine was a technical success as a pilot, following agreement of commercial terms for deployment across more wells, the system is being implemented on more PSE fields.
  7. During the pilot a number of requests for additional functionality were submitted. This additional functionality has been implemented notably the “Optimisation Opportunities” workflow, which was instrumental in identifying the opportunities that led to 1052 additional barrels of oil per day
  8. It is believed that the ESP analysis engine is an immensely valuable tool for ‘management by exception’ of a large population of ESPs and can provide benefit to other business units.


  1. Camilleri, L.A.P. and MacDonald, J. 2010. How 24/7 Real Time Surveillance Increases ESP Run Life and Uptime. Paper SPE 134702 presented at the SPE Annual Technical Conference and Exhibition, Florence, Italy, 19–22 September.
  2. “Demystifying ESPs: A Technique to Make you ESP Talk to You” by A. J. (Sandy) Williams. Presented at ESP Workshop Houston 2000.
  3. ESP Monitoring – Where’s your speedometer? A.J. (Sandy) Williams*, Julian Cudmore, Stephen Beattie (Phoenix Petroleum Services) Presented at ESP Workshop Houston 2000.
  4. Takacs, G., Electrical Submersible Pumps Manual, 2009 Elsevier Inc.


William Guerra is a Petroleum engineer with over 14 years of experience in production optimization operations, artificial lift, well testing, water disposal injection wells and reservoir simulation. He has also led the planning and execution of oilfield exploration and development projects.  He has worked for ECOPETROL S.A., Petrominerales Colombia Ltd and currently is currently Production Engineering Team Leader with Pacific Stratus Energy.

Edeberto Finol is a petroleum engineer with over 22 years experience in the oil industry, including: logistics and operations Well Service and Workover, planning and procurement of artificial lift systems, selection and stimulation steam, diagnosis and optimization production wells, optimization of artificial lift systems, selection and design of artificial lift systems, facilities and operations of oil and gas production, production Engineering management, among others. I have worked with PDVSA, Repsol YPF, Mansarovar Energy Colombia LTD and is currently working with Pacific Rubiales Energy.

Javier Solorzano is a Petroleum Engineer with 7 years experience in short and long term well test testing. For more than three years, has served as the Engineer for Control and Optimization of Production, conducting the review of the conditions of the wells using methodologies such as nodal analysis of the well to improve production and / or prolong the run life of the equipment. Javier also participates in the interdisciplinary group of special projects (which includes areas such as Production, Geology, Reservoir, Drilling and Completion) which aims to create strategies to increase oil production or reduce water production. Additionally Javier was the focal point responsible for coordinating the review of the 44 well that allowed a reduction in costs associated with reduced water handling and increasing production by of over 1,000 bopd with zero capital investment.

Wilson Benavides Has a Bachelor´s degree in petroleum engineering from the Surcolombiana University. He joined the oil and gas industry in 2013 as a junior production engineer for Petrominerales. One year later he transferred to Pacific Rubiales Energy and began his training process in the production optimization area. Wilson has worked in production optimization projects, including Pump Checker and Pump Analytics pilot projects, helping to optimize the oil rates in the wells in the short and middle terms. He currently lives in Bogotá, Colombia.

Sandy Williams has worked in the Petroleum Industry for 25 years. He is the founder of Artificial Lift Performance Ltd (ALP), a company which specializes in helping operators get more oil – a skill which he explains to the layman as being ‘like steroids for oil wells’. Sandy worked 9 years for Amoco, then Phoenix Schlumberger Artificial Lift before becoming a consultant focused on artificial lift and production optimization. He has worked and lived in the USA, Oman, Venezuela, Ecuador, Colombia and has taught over 150 courses related to production optimization and artificial lift and is fluent in Spanish.