ADVERTISEMENT

Huff ‘n’ Puff EOR Proves Effective in Gas-Condensate Reservoirs

You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.

To ensure continued access to JPT's content, please Sign In, JOIN SPE, or Subscribe to JPT

As the pressure drops below the dewpoint in an unconventional gas-condensate reservoir, the liquid drops out of the gas phase and forms an oil phase in the matrix and fracture. The volume of the oil phase formed in the matrix mostly stays below the residual oil saturation. The gas huff ‘n’ puff process has demonstrated potential in improving recovery from tight oil reservoirs. The objective of the study described in this paper was to investigate the feasibility of huff ‘n’ puff enhanced oil recovery (EOR) in a gas-condensate reservoir.

Introduction

Compositional analyses of fluid samples, taken from early production of three wells located a few miles apart from each other in Eagle Ford, were used to build the black-oil, volatile-oil, and gas-condensate fluid models. The produced field gas was selected as the only viable option for injection.

Phase-Behavior Models

The phase-behavior models were built for black-oil, volatile-oil, and gas-condensate wells on the basis of pressure/volume/temperature reports from bottomhole samples. The reservoir fluids were built by combining mathematically the oil and gas fluids from separators. The pseudocomponents were characterized from the detailed compositional laboratory analysis. The Peng-Robinson equation of state with the temperature-independent volume correction was used for phase-behavior modeling. The Lohrenz-Bray-Clark viscosity model was applied in calculating the phase viscosities. The condensate/gas ratio for the gas condensate fluid is 147 STB/MMscf, which is a rich gas-condensate fluid.

Comparative Study of Gas Injection

To compare the performance of huff ‘n’ puff gas injection in different fluid types, single-well simulation models were built for each fluid type. Several well models of 4,500 ft in length with a fracture spacing of 50 ft were built using a compositional simulator. The reservoir thickness was 130 ft. The hydraulic fractures were set up using the embedded discrete fracture model (EDFM) preprocessor. Each single-well model was history matched using the production data. Because the history-matching solutions were nonunique, the rate and pressure transient analysis for each well were used to limit the results to the realistic solution spaces. The history-matched fracture half-length varied between 50 and 175 ft and the fracture height varied between 50 and 90 ft. The fracture conductivity ranged between 3–30 md-ft.

The history-matched models were used to investigate the performance of gas-injection EOR for all fluid types. The field gas was injected for 10 cycles, and the cumulative produced oil using huff ‘n’ puff was compared to one without gas injection. The incremental oil recovery for the nine studied black-oil cases varied between 30 and 40%. The incremental oil recovery was less than 10% for three volatile-cases and 15 to 20% for 13 gas-condensate ones.

The black-oil single-well models showed the highest hypothetical incremental oil recovery after 10 cycles of injecting produced gas. The cyclic field-gas injection in the volatile-oil wells showed the weakest performance among the three fluid types. The reservoir fluid system gradually shifted from a volatile-oil to a gas-condensate system by injecting the produced gas. The phase envelope expanded as the reservoir fluid mixed with the injected gas, and the bubblepoint pressure increased as the ratio of injected gas increased in the fluid mixture. The miscibility of the produced gas and the volatile oil was dominated by vaporization. However, the low percentage of intermediate components in the produced gas highly reduced the efficiency of the vaporization process.

The hydrocarbon fluid systems were initially above the dewpoint pressure in gas-condensate reservoirs. During the depletion in the gas-condensate reservoirs, the pressure gradually dropped below the dewpoint, and the heavier components condensed in an oleic phase. Because the condensate saturation barely built up above 10%, huge amounts of intermediate and heavy component remain irrecoverable in the matrix. These heavier components will revaporize into the gas phase and become mobile if the produced gas is injected into the reservoir.

Fracture Modeling With EDFM

EDFM provides efficient solutions for modeling complex fracture geometries in terms of reliability, flexibility, and simulation runtime. The main advantage of modeling fractures using EDFM compared with the local-grid-refinement method is the computational efficiency. The EDFM method is capable of modeling fractures with complex geometry independent from the shape and size of gridblocks.

Modeling Fracture Interference With a Multiwell Sector Model

The authors present a general work flow to validate the simulation results with the production data and to interpret interwell fractures. The process begins with construction of a sector model from the full-field data. Then, the EDFM preprocessor is used to generate the hydraulic fractures and long fracture hits. The fracture hits are diagnosed by analyzing the production data and well events. Next, the static model is coupled with the compositional numerical simulator to perform a dynamic characterization of the sector model. A rigorous history-matching process is performed to tune the fracture properties such as fracture half-lengths, fracture heights, and fracture conductivities. Once the field history is matched, the sector model is used to forecast the scenarios of gas EOR. The process can be iterated when the injection and production data of each field huff ‘n’ puff cycle are received. Hence, the geometry and properties of the long interwell fractures can be recalibrated. A gas-condensate sector model was built on the basis of available field data to match the field history and forecast the gas huff ‘n’ puff performance. The sector model corresponds to 12.5% of the full-field model that would reduce the computational cost for sensitivity analysis.

The area of interest has 13 producers, including the three parent wells (W01, W02, W03) and 10 child wells (W04 through W13). The parent wells have been producing for approximately 8 years, and the newest child wells have been producing for 4 years. The well spacing is 550 ft, and each well has multiple planar hydraulic fractures with a different fracture configuration. The cluster spacing varies from 22 to 49 ft, with the smaller cluster spacing for the newer wells. The reported production data and pressure response were used to achieve a robust history match for the sector model. The investigated parameters were the fracture conductivity, fracture half-length, and the fracture height. The well constraint for history matching was set to be the oil rate for each producer. The response parameters to be matched in this study were the bottomhole pressure and gas production for all wells. A well-­history-matched model would provide a reliable base case to begin forecasting huff ‘n’ puff performance. The connectivity between parent and child wells through long interwell fractures was diagnosed by analysis of the production data and pressure response. As a result, eight long fractures were added to the sector model by EDFM. Analysis shows that the interwell fractures can be as long as 2,500 ft.

Huff ‘n’ Puff Design

The gas EOR project was modeled in the 13-well sector model calibrated by history matching. Several scenarios were considered regarding the number and location of the gas injectors, the period and rate of injection, and the period of production before the next injection cycle. Fig. 1 shows that Well W08 was selected for the first cycle of injection. Four offset wells were shut in during the injection to guarantee pressure containment and avoid gas escape through the long connecting fractures. Furthermore, clean baselines were established by shut in for those offset wells, enabling the operator to detect the pressure communication between the injector and the offsets. The maximum injection pressure and gas injection rate were set to 7,500 psi and 16 MMscf/D. Approximately 400 MMscf of gas was injected in 32 days.

Fig. 1—The sector model, including the long interwell fractures.

 

The model perfectly matches the bottomhole pressure observed in the field. At the end of injection, the bottomhole pressure built up to 7,500 psi, which is higher than the initial pressure of 6,000 psi. After reopening the well for production, the pressure began to decline gradually, and it is expected that it will take a few months for the pressure to return to the condition existing before gas injection. Then, the model was run for 10 injection cycles to forecast the cumulative oil uplift. The cumulative oil can increase approximately 19% by injecting field gas for 10 cycles. The most-challenging part of sector modeling was characterizing the pressure communication and gas escape through long fractures caused by fracture reactivation.

Conclusions

The modeling work flow in this study employed the nonintrusive EDFM method in conjunction with a commercial compositional simulation to perform robust history matching. The following conclusions can be drawn from this study:

  • Injection of field gas causes higher incremental oil recovery in a gas-condensate well compared with a volatile-oil well. The hypothetical incremental oil recovery by injecting produced gas in a gas-condensate well is approximately 15–20% after 10 cycles of injection.
  • The fracture interference in the gas-condensate area is not detrimental to gas containment and pressure buildup, because the well and fracture spacing are less dense compared with the black-oil area.
  • Multiwell modeling and simulation with several fractures can be much faster if the fracture modeling is performed using EDFM rather than with local grid refinement.

The work flow used in this study reveals opportunities for EOR in different shale fluid types. Huff ‘n’ puff using available field gas in the area of interest shows the best performance in black-oil and gas-condensate reservoirs. This study demonstrates that the best fluid-type candidate for EOR operation in each region will depend on the miscibility between the components of the in-situ fluid and the ones in the available injectant.

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper URTEC-2019-987-MS, “Gas-Injection EOR in Eagle Ford Shale Gas-Condensate Reservoirs,” by Reza Ganjdanesh, SPE, Wei Yu, and Mauricio Xavier Fiallos, SPE, The University of Texas at Austin; Erich Kerr, SPE, EP Energy; Kamy Sepehrnoori, SPE, The University of Texas at Austin; and Raymond Ambrose, SPE, EP Energy, prepared for the 2019 Unconventional Resources Technology Conference, Denver, 22–24 July. The paper has not been peer reviewed. Copyright 2019 Unconventional Resources Technology Conference. Reproduced by permission.

Huff ‘n’ Puff EOR Proves Effective in Gas-Condensate Reservoirs

01 November 2019

Volume: 71 | Issue: 11

No editorial available

ADVERTISEMENT


STAY CONNECTED

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.

 

ADVERTISEMENT

No editorial available

ADVERTISEMENT

ADVERTISEMENT

ADVERTISEMENT