Assessment of Waterflood Efficiency in Low-Permeability Reservoirs Using Production Data

Thesis Abstract

Low-permeability reservoirs pose significant challenges to the effectiveness of waterflood secondary recovery methods, primarily due to their limited pore connectivity, heterogeneity, and capillary trapping phenomena, which often result in suboptimal sweep efficiencies and uneven displacement fronts. This study aims to evaluate the efficiency of waterflooding in such reservoirs through comprehensive analysis of production data, ultimately informing improved reservoir management strategies. The specific objectives include quantifying waterflood performance metrics, identifying factors influencing sweep efficiency, and developing predictive models for waterflood response prediction in low-permeability contexts. A quantitative, empirical research design was adopted, focusing on a case study region comprising three mature low-permeability sandstone reservoirs with a combined production history of 15 years and a total of 120 active producer wells. The population comprised all wells within these reservoirs, from which a stratified random sampling of 60 wells was selected based on production maturity, formation properties, and injection history to ensure representativeness. Data collection involved extracting detailed production records, injection rates, pressure surveys, and reservoir property logs from field databases, complemented by semi-structured interviews with reservoir engineers for contextual insights. To ensure data integrity and validity, production data were cross-verified against operational logs, and calibration was performed using production history matching techniques. Reliability of the data instruments was assessed through consistency checks and comparison with auxiliary data sources. Analytical approaches included multiple linear regression to identify significant predictors of waterflood efficiency, Analysis of Variance (ANOVA) to compare performance across different reservoir zones, and survival analysis to assess well longevity post-waterflooding. The study also incorporated simulation-based sensitivity analysis using the Dykstra-Parsons model and development of a quantitative Waterflood Efficiency Index (WEI) derived from cumulative oil production, water cut, and residual oil saturation data. The expected findings suggest that factors such as reservoir heterogeneity, initial water saturation, injector-producer spacing, and fracture networks significantly influence waterflood performance. The analysis anticipates revealing a moderate correlation between net injection volume and incremental oil production, with diminishing returns beyond certain injection thresholds owing to channeling and capillary trapping. The predictive models are expected to demonstrate robust reliability, offering practical benchmarks for operational decision-making in low-permeability reservoirs. This research contributes to reservoir engineering knowledge by providing a quantitative framework for assessing waterflood efficiency through field data analysis and model development specific to low-permeability formations. It fills the existing literature gap regarding empirical data-driven evaluation techniques in such reservoirs, often dominated by simulations or laboratory-scale studies. Additionally, the study advances understanding of the critical reservoir parameters governing sweep efficiency, thus informing more targeted field-scale waterflood design. The study concludes that optimizing waterflood performance in low-permeability reservoirs necessitates tailored injection strategies that consider heterogeneity, fracture distribution, and capillary effects. It recommends the integration of production data analytics into reservoir management practices, the development of improved heterogeneity-aware simulation models, and the adoption of advanced monitoring technologies such as microseismic and tracer tests for real-time reservoir surveillance. Future research should explore the application of machine learning algorithms for predictive modeling and extend the analysis to the effects of enhanced recovery methods, such as polymer and gas injection, on waterflood efficiency in similar formation types.