The Efficiency of the Novel Technology Reservoir Pressure Reconstruction Without Well Shut-in and without Production Losses

Belyakov, A. A.; Gulyaev, Danila; Krichevskiy, Vladimir; Nikonorova, Anastasia Nikolaevna; Iskibaev, Roman Edisonovich

doi:10.2118/206490-ms

Cited by 2 publications

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Injection Optimization for Oil Production Increase by Quantitative Assessment of Crosswell Interference Based on Machine Learning

Kim,

Gulyaev,

Voron

et al. 2023

Day 3 Thu, November 23, 2023

View full text Add to dashboard Cite

Understanding reservoir pressure dynamics and crosswell interference is crucial for optimizing pressure maintenance systems in heterogeneous reservoirs with complex geology. This paper presents a case study from Eastern Siberia, highlighting the positive impact of applying Multi-Well Retrospective Test (MRT) machine learning technology on production enhancement. MRT technology relies on mathematical algorithms for annualizing long-term records of bottomhole pressure and surface rates from a group of wells through multiwell deconvolution. It requires historical data of bottomhole pressures for the tested well and flow rate history for all wells under study. Multiwell deconvolution involves a fully or semi-automated search for initial pressure and Unit-rate Transient Response (UTR) for tested wells and cross-well intervals, aligning actual pressure records with total sandface flow rate variation history. It also quantifies the accumulated pressure impact of surrounding wells on the tested well. The study area featured nine wells with declining production rates, including six producers and three injectors. The primary objective was to assess production enhancement potential, primarily through injection optimization. The seven-year dataset encompassed flow rate and pressure variations during production. Before employing machine learning, data were preprocessed to reduce the number of analysis points, synchronize flow rate and pressure change timings, and remove outliers. Mathematical deconvolution procedures were then applied to derive UTRs, with UTR convolution providing crosswell pressure impact. Two injection wells were found to have a significant cumulative pressure impact on production wells. Mathematical well shut-ins yielded reservoir pressure and well productivity index. UTR interpretation via pressure transient analysis algorithms offered insights into reservoir transmissibility, well skin, and interference-free drainage areas. Machine learning algorithms generated pressure/rate forecasts for different well targets, indicating that the optimal production increase could be achieved through a 1.5x increase in injection rate for one well and a 2.7x increase for another well, resulting in a twofold oil production increase with constant water cut. Field implementation demonstrated that MRT technology is a powerful tool for optimizing injection targets and increasing oil production. Additionally, MRT provides reservoir pressure data without well shut-ins, enabling the operator company to gather information for reservoir pressure mapping without production deferment, resulting in a significant increase in Net Present Value (NPV).

show abstract

Injection Optimization for Oil Production Increase by Quantitative Assessment of Crosswell Interference Based on Machine Learning

Kim,

Gulyaev,

Voron

et al. 2023

Day 3 Thu, November 23, 2023

View full text Add to dashboard Cite

show abstract

Method for assessing well interference at under-gas cap zone using CRM material balance model

Bekman,

Ruchkin

2024

TSU Herald. Phys Math Model. Oil, Gas, Energy

View full text Add to dashboard Cite

Optimization of injection operation conditions is a primary task when designing the development of mature oil fields. To select optimal injectivities, solutions to the optimization problem based on a CRM (capacitance resistance model) analytical model are used. CRM models based on analytical solutions of the material balance equations of weakly compressible fluids due to their speed can be used as an alternative to flow simulation models in solving a number of problems to support oil field development. The main task of CRM models is to determine the well interference factor, i.e. the share of fluid produced due to a particular injection well. These factors can be used to analyze waterflooding and develop solutions for waterflood optimization. At the same time, in some cases reservoir fluids may have high gas content, which is a classical limitation of a CRM model and leads to underestimation of the interference factors when estimating the injection performance. In this paper, the authors propose an approach to solve the problems of distortion of such factors, and for the first time in CRM application practice a model has been adapted for use in the conditions of high GORs and has a potential for being applied for under-gas cap zones. This is achieved by including in the equations the physical properties of gas and their behavior under reservoir conditions. The complication of algorithms did not significantly affect the model run time, but allowed to match the model both separately and jointly for liquid and gas phases. Thus, the improved classical CRM model has significantly expanded the scope of application of operational waterflood analysis tools for assessing the current situation. No less complicated and urgent is the problem of forecasting the production of gas liberated in the conditions of the previously mentioned under-gas-cap zones, because it implies the need to estimate the future reservoir pressure. The authors intend to devote the next paper to finding a solution to this issue, taking into account the proposed new method of reservoir pressure estimation using a CRM model.

show abstract

The Efficiency of the Novel Technology Reservoir Pressure Reconstruction Without Well Shut-in and without Production Losses

Cited by 2 publications

References 10 publications

Injection Optimization for Oil Production Increase by Quantitative Assessment of Crosswell Interference Based on Machine Learning

Injection Optimization for Oil Production Increase by Quantitative Assessment of Crosswell Interference Based on Machine Learning

Method for assessing well interference at under-gas cap zone using CRM material balance model

Contact Info

Product

Resources

About