Diagnostic Tool Identifies Factors in Well-Productivity Decline
You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.
Wells in deepwater reservoirs show significant rate decline with time as the result of various causes. A diagnostic tool for quantification of factors influencing well-productivity decline is presented in this paper. The diagnostic tool helps identify well-stimulation candidates and potentially can help increase production. The work flow presented provides a tool for monitoring well-productivity changes to identify the main causes of productivity decline and to quantify effects on the normalized productivity index (PI).
Most current and future deepwater reservoirs are in structurally deep, high-pressure environments in which reservoir and rock mechanisms that affect long-term well productivity are poorly understood. PI trends derived from field production and pressure data reflect the composite effects of wellbore damage along with changes caused by multiphase-flow and pressure-depletion effects on fluid properties and permeability. However, only PI decline caused by wellbore damage should drive well-stimulation decisions because only stimulation can improve the permeability of the near-wellbore region. In addition, reservoir simulation input PI multipliers used to match well performance should be adjusted appropriately to ensure that the resulting simulated output PI trends are consistent and reliable and to avoid duplication of reservoir and fluid effects that are already captured in the simulation model.
Effect of Reference Pressure on Well PI Calculations
PI theoretically is estimated with the average reservoir pressure; however, average reservoir pressure may not be estimated accurately for fields where reservoir size and shape are highly uncertain. In cases where the average reservoir pressure is not available, initial reservoir pressure, or the buildup pressure at 1 hour after shut-in, have been used by different operators to estimate well PI. A need exists to understand how these different reference pressures can affect the reliability of the estimated well PI. Equations used to derive well PI, including under pseudosteady-state (PSS) flow conditions, are presented in the complete paper.
Effect of Using Different Reference Pressures on PI Estimation: Synthetic Cases
A numerical synthetic case is generated to investigate the effect of the reference pressure on well PI. The case represents a circular reservoir 50,000 ft in radius and 30 ft thick. The fluid viscosity is 3.6 cp, and the formation volume factor is 1.1 RB/STB. The reservoir porosity is 10%, and the permeability is 300 md. The skin is increasing with time. The flow rate is constant at 1,000 STB/D, and the well is shut in for 21 periods. PI is calculated at each buildup using a reference pressure from each recognized flow regime. PI using the initial reservoir pressure was also estimated. The first reference pressure is chosen from the wellbore storage flow regime period at 0.01 hours, while other reference pressures are chosen from the infinite acting radial flow (IARF) regime. The PI is calculated finally using the average reservoir pressure during transient and PSS conditions.
The synthetic case results show that using a reference pressure during the wellbore storage flow regime would provide a misleading PI trend. PI should be declining from Buildup 1 to Buildup 21 as the skin increases over time; however, the PI estimated using 0.01 hours shows an increase in PI at 500 hrs before the start of the PSS followed by a flat trend toward 5,000 hours and then ended by a gentle decline compared with the true decline using the average reservoir pressure. Although using any reference pressure during IARF should result in the same PI trend as would be seen when using the average reservoir pressure, the end of the IARF reference pressure provides the most-accurate estimate of PI. On the other hand, the initial reservoir pressure provides a pessimistic PI trend during PSS conditions because of reservoir depletion.
The results of the presented study provide more understanding of which reference pressure should be used to estimate PI. The complete paper discusses in detail a work flow for removing fluid-properties changes and multiphase and rock-compaction effects from PI calculations.
Validation of the Proposed Method Using Synthetic Data
A two-phase (oil/water) numerical synthetic case was generated to test the validity of the proposed work flow. The reservoir is circular with a 3,000-ft radius and is 100 ft thick. The reservoir features two wells, the oil producer at the center of the reservoir and the water injector 1,000 ft from the producer. The oil producer began 3 years before the water injector. The reservoir porosity is 23%, and the permeability is 400 md. The permeability compaction correlation is implemented in the reservoir simulation. The skin in the oil producer is increasing with time; there are eight buildups during the production history of the oil producer.
The end of the radial flow regime is within 12 hours on the basis of reservoir and fluid properties. Therefore, the pressure at 12 hours after shut-in is chosen as the reference pressure for the observed PI calculations. Given the reservoir size and shape, with the fluid and rock properties, the analytical normalized PI is estimated at each buildup using the average reservoir pressure.
The pressure/volume/temperature (PVT) and multiphase effects are removed from the observed normalized PI at 12 hours. A good match is achieved between the normalized PI using the end of the IARF at 12 hours and the one estimated using the average reservoir pressure. The effect of oil viscosity alone before water injection improves the well PI as the oil viscosity decreases with depletion. That uplift in PI could be interpreted mistakenly from field data as a reduction in wellbore skin. After the reservoir pressure increases because of water injection, the combined effects of PVT and multiphase effect reduce the well PI at almost 1,200 hours. Similarly, the compaction effects are also removed from the observed normalized PI at 12 hours after PVT and multiphase effects are removed (Fig. 1).
Dynamic PI Diagnosis. Three main productivity-decline causes affect the oil-producer PI: skin, PVT and multiphase effects, and compaction. The proposed work flow is used to quantify the contribution percentage of each cause. The normalized PI decline resulting from each cause is quantified, and its percentage out of the total decline is plotted vs. time at each buildup. PVT changes affect PI positively when a single phase (oil) is produced as the oil viscosity decreases with depletion, and the reverse is happening when the average reservoir pressure is increasing because of injection. On the other hand, PVT and multiphase affect PI negatively because of relative-permeability and oil-viscosity changes with pressure.
The work flow is applied in a deepwater Gulf of Mexico field at water depths of more than 5,000 ft. The subject well is a horizontal producer completed in two layers of an unconsolidated sandstone. The reservoir has an average permeability of approximately 700 md and a total net thickness of 225 ft. The initial reservoir fluid viscosity is 8 cp, and the formation volume factor is 1.15 RB/STB. The well gradually was ramped up over a long period and achieved a peak production rate of 5,000 STB. Water breakthrough occurred in the eighth month after first production. Significant productivity decline is observed throughout the producing life of the well.
A buildup of a minimum of 30 hours is needed to estimate reservoir horizontal permeability from the late radial-flow regime. The estimated horizontal permeability is then used to estimate the vertical permeability using the early radial-flow-regime results. Three main buildups are identified to be used in the work flow to determine the effect of PVT, multiphase, and compaction on well productivity.
The diagnostic plot of the three main buildups shows negligible change of the horizontal permeability deep into the formation away from the wellbore (all buildups are on top of each other at the late radial-flow regime). However, the early time of the three main buildups is shifted upward with time, indicating either that the effective horizontal length was decreasing or the effective permeability around the wellbore was decreasing with time. The diagnostic and history plots of the three buildups are matched using an increase in skin, a reduction in effective well length, and a reduction in the effective permeability around the horizontal well. The PVT, multiphase, and compaction effects are removed using the proposed work flow. PVT and multiphase factors affect the well productivity decline by up to 20%. Rock-compaction effects result in an additional well-productivity decline of up to 51%.
- The developed work flow can be used to identify major well-productivity decline causes and quantify their contribution over time.
- The work flow is applied to field cases, including vertical and horizontal wells, to improve well-productivity monitoring and forecasting.
- The compaction effect on the field-case well-productivity decline is well-aligned with the field reservoir simulation study using a laboratory rock-compaction table.
- The resulting PI trend obtained from the developed work flow is incorporated in the field reservoir simulation to ensure that the PI multiplier is not duplicating reservoir and fluid effects.
Diagnostic Tool Identifies Factors in Well-Productivity Decline
01 February 2020
Transocean's Backlog Signals A Deepwater Revival
Transocean reported a long list of reasons why the deepwater drilling business is coming back strong. Making money is still not one of them.
Eni Makes Another Discovery Offshore Mexico, Expects Several More
The Italian oil and gas company has found its second big well offshore Mexico and is setting its sights on several more in the near future.
One of the frustrating aspects of well-productivity analysis is identifying the causes of lower-than-expected production/injection during initial well lifetime. Our task is to evaluate the multivariate aspects of well design.
No editorial available
Don't miss out on the latest technology delivered to your email weekly. Sign up for the JPT newsletter. If you are not logged in, you will receive a confirmation email that you will need to click on to confirm you want to receive the newsletter.
11 February 2020
06 February 2020
12 February 2020
07 February 2020
13 February 2020