Paying the Price: The Hidden Costs of Artificial Lift Selection in Unconventional Production
What is the cost you pay in unconventional production operations when you do not select the right artificial lift equipment?
Average electrical submersible pump (ESP) failure costs in Permian Basin are conservatively estimated at $180,000 per incident (Williams and Ortega, World Oil, June 2019). A 3 to 6 month run life for ESPs is begrudgingly considered as acceptable failure frequency in many shale or resource plays, often caused by sand loadings or motor failures due to gas slugging. Selecting the right timing and type of artificial lift for each well at each phase of its life cycle is a herculean task.
Artificial lift life cycle optimization requires continuous monitoring of existing lift diagnostics, adjusting lift parameters to operate them at near-optimum conditions, and screening alternate scenarios to determine the right timing to switch to a better lift type.
Background
With recent unconventional completion trends that are characterized by larger frac jobs (fluid volumes and proppant intensity), there has been a silent, but not-so-secret anymore, issue creeping up to annoy production operators. Premium resin coated and ceramic proppants are replaced by abrasive frac sand. High concentrations of formation and frac sand lead to premature ESP failures. These are further compounded by gas slugging when the ESP’s gas separation efficiency is compromised or when produced gas-liquid ratio (GLR) increases sharply as the flowing pressure falls below saturation pressure either in the reservoir or wellbore.
By now, we all know that unconventional production, in contrast to its conventional cousin, is very dynamic and characterized by its steep decline from early time production rates by as much as 40-80% within the first year. Near-wellbore reservoir pressure depletes faster, and the transient well productivity index declines rapidly. Consequently, when the reservoir pressure rapidly falls below the saturation pressure, it may increase GLR beyond the gas handling ranges for ESP (electrical submersible pump), PCP (progressive cavity pump), or SRP (sucker rod pump).
Unconventional wells not only require artificial lift sooner in their life cycles, but also multiple forms of artificial lift are frequently used in the same well at various stages during production. Indeed, it is common for unconventional wells to employ two or more types of lift within a few months of production. Traditional techniques to select, design and operate artificial lift applications are inefficient for several reasons, or in some cases, have also failed. In terms of total value, artificial lift in an unconventional well is less efficient than conventional well usage due to their higher failure costs and smaller production operating window. The industry has fought back with remarkable improvements in pump design for gas separation, gas handling, and sand management.
However, field profitability relies on proactively using the right sized equipment considering the well’s production profile and operating it as efficient as possible to extend the equipment service life.
Current approaches for artificial lift selection are generally empirical or rule-based that consider target liquid rates, expected GLR, dog leg severity (DLS), anticipated line pressure, solids production, and downhole temperature amongst other factors. Heuristic approaches, combined with anecdotal experience in the field, tend to be the main tool that the production operators can use within their fast decision timescales and large well counts.
A comprehensive approach to the artificial lift life cycle should include present and future well performance, operations, and optimization as an iterative process across the life of the well. This comprises choosing and re-choosing the appropriate artificial lift type, modifying the operational parameters, and scheduling the associated workovers.
This has the potential to bring considerable value in terms of reduction in failures, cost, and downtime, and increase in operational efficiency.
Artificial Lift Timing and Selection
The new integrated artificial lift timing and selection (ALTS) methodology uses a hybrid reservoir model to optimize selection amongst various candidates and operation of the lift system to maximize economic value over a forecasted period (Zalavadia et al. 2023). Most incumbent methods do not include the impact of subsurface performance i.e. reservoir drawdown changes on inflow performance, because of choosing the artificial lift type. The ALTS methodology injects feedback to the reservoir and gets an integrated subsurface and well response implicitly by selecting the artificial lift type and its operational parameters.
The underlying reservoir model is based on identifying a transient well performance that combines the dynamic drainage volume with transient productivity index for estimating dynamic reservoir deliverability (Molinari and Sankaran 2021). This is combined with well deliverability estimation for different lift strategies to perform continuous nodal analysis and forecast phase rates using novel PI-based forecasting techniques (Zalavadia et al. 2022).
The overall workflow can be described in four steps:
1
Dynamic drainage volume (DDV) represents the reservoir depth investigated akin to the concept of distance of investigation in pressure transient analysis. It is the region within the reservoir where pressure depletion has just begun. But how do we calculate DDV rapidly without a 3D numerical model? The first breakthrough comes from using variable field measurements (flowing bottomhole pressure and rates) to calculate a reasonable approximation of DDV in the asymptotic limit.
2
By calculating the average reservoir pressure within the DDV using a modified material balance approach, this second novelty provides the basis to calculate a transient productivity index. With average reservoir pressure and productivity index available at any point in time, the well’s instantaneous inflow performance (IPR) is adequately characterized.
3
The final innovation is to predict how the productivity index will behave in the future using a data-driven model based on cumulative production. This last PI-based forecasting step completes the list of ingredients to predict how different operating strategies will determine the pressure depletion and productivity index decline simultaneously to provide a future IPR.
4
The availability of current and future IPR opens several possibilities including artificial lift life cycle optimization. For a producing well, we could rapidly screen alternate lift types by selecting representative candidates. It is common to consider variety in both lift type (ESP, gas lift, SRP etc.) and their nominal design specifications. Nodal analysis can now calculate the expected flow rates for each day, while the hybrid reservoir model described above can estimate the future IPR because of the new operating strategy.
Each lift type may have different life spans or expected failure rates, which could be incorporated as part of the forecast based on field knowledge. Capital and operating costs may differ between each lift type or even pump vendors. Sensitivities to economic prices or lift operating parameters can provide further insights to make robust artificial lift decisions. The impact of selecting a lift type today can also be seen in the forecasted performance with the lift diagnostics. For instance, even if a well is a good ESP candidate today but will soon be a better candidate for SRP, then the additional workover costs may not support its conversion today. Similarly, even if selecting a large ESP makes sense today but the production will soon decline, so the pump will mostly operate under-thrust for its useful life and cause premature failures, it should be considered in decision-making.
Case Study
The following example presents a well from a major liquids-rich US shale play, where the well has been producing for 3 months. After the first month, when oil production declined 60% from its initial peak production, a medium-sized ESP was installed to arrest the production decline. While the production decline was arrested, no production uplift was observed over the next few weeks. It is desirable to evaluate the upside potential of the well with a larger ESP or complete the well with gas lift in the future.
The hybrid reservoir model estimated that the existing pump capacity exerted a conservative drawdown that limited the well production. Pump replacement costs for each lift type after its expected life expectancy is included as part of the economic analysis. The ALTS model predicted a slight upside of $0.4 MM incremental NPV over a 2 year forecasted period with a larger sized ESP. The size of the prize can now be evaluated against field observations of higher drawdown induced pump failure rates for a proper economic analysis. Whereas, converting to gas lift has a high upside potential of $2.3 MM incremental NPV over the same period. When economy of scale is achieved across the field with several wells favoring gas lift for sufficient capital expansion, it can become a viable option to compete with ESPs.
Conclusion
The new ALTS workflow provides a pragmatic approach for operators to manage production operations. Avoiding guesswork or heuristics, each well’s past performance can be used to calibrate and predict its future performance in developing a customized AL strategy to maximize economics over the full well life-cycle.
The key innovation lies in the incorporation of a reservoir model enabling evaluation of subsurface performance under varying field operations. This empowers operators to make informed decisions regarding the timing and selection of artificial lift systems, leading to enhanced production rates, reduced deferred production, and extended equipment life.
It is time to embrace the future of artificial lift optimization. By shifting from heuristic-based practices to well-specific analysis, we can drive efficiencies, mitigate risks, and unlock the full potential of our reservoirs.
References
Molinari, Diego and Sankaran, Sathish. 2021. “A Reduced Physics Modeling Approach to Understand Multiphase Well Production Performance for Unconventional Reservoirs”. Paper presented at the SPE/AAPG/SEG Unconventional Resources Technology Conference, Houston, Texas, USA, July 2021. doi: https://doi.org/10.15530/urtec-2021-5023
Williams, Jason and Ortega, Miguel. 2019. “Extending ESP service life in unconventional wells”. World Oil. url: https://www.worldoil.com/magazine/2019/june-2019/special-focus/extending-esp-service-life-in-unconventional-wells
Zalavadia, Hardikkumar, Sinha, Utkarsh, and Sathish Sankaran. “An Improved Method for GOR Forecasting in Unconventionals.” Paper presented at the SPE/AAPG/SEG Unconventional Resources Technology Conference, Houston, Texas, USA, June 2022. doi: https://doi.org/10.15530/urtec-2022-3723189
Zalavadia, Hardikkumar, Gokdemir, Metin, Sinha, Utkarsh, Singh, Prithvi, and Sathish Sankaran. “Real Time Artificial Lift Timing and Selection Using Hybrid Data-Driven and Physics Models.” Paper presented at the SPE Western Regional Meeting, Anchorage, Alaska, USA, May 2023. doi: https://doi.org/10.2118/213040-MS
Leave a Reply