Improving deconvolution reliability in well test analysis with an automated objective function
Mina S. Khalaf*
Deep Resources Engineering. 2026, 3(1): 100231. 
doi.org10.1016j.deepre.2025.100231.pdf
Abstract: Deconvolution is a mathematical technique that eliminates the effects of production rate variations in pressure transient data, enabling the recovery of a smoother pressure signal extended across both drawdown and buildup periods. This approach has distinct advantages over conventional analysis methods, including enhanced radius of investigation, improved derivative quality, and simplified use of existing interpretation techniques. Since the introduction of von Schroeter deconvolution method, as the first stable algorithm, in the early 2000s, several researchers have advanced the methodology by refining algorithms and improving the objective functions that govern their performance. Despite these developments, the definition of a robust and reliable objective function remains a challenge in achieving accurate deconvolution results. This study introduces a new objective function (NOF) for well test deconvolution that simplifies weighting schemes, eliminates subjective parameter tuning, and improves robustness under noisy conditions. Unlike earlier methods, the NOF assigns automated weights to pressure and rate measurements based on gauge accuracy, ensuring objective normalization of errors and removing the need for the unresolved flow rate regularization parameter. This streamlined approach not only reduces complexity but also enhances accuracy. The proposed noise-oriented NOF consistently outperformed traditional formulations across all simulated and field cases. In four synthetic examples with varying reservoir conditions and noise levels, the NOF successfully recovered smooth pressure responses and accurate flow rates, while the original von Schroeter and Levitan objective functions produced distorted or shifted signals and failed to adjust rates. Quantitative comparisons showed that this study produces a marked reduction in average deconvolution errors. This study enhances the stability and accuracy of deconvolution and offers a practical and superior tool for reservoir characterization, making it easier for engineers to extract meaningful reservoir properties, extend the radius of investigation, and reduce interpretation errors in real field applications.
Highlights:
• Introduces a new objective function (NOF) for well test deconvolution.
• Automates weighting of pressure and rate data using gauge accuracy.
• Eliminates subjective tuning and the unresolved flow rate parameter.
• Consistently improves accuracy and robustness under noisy conditions.
• Validated on synthetic and field cases with reduced deconvolution errors.
Keywords: Pressure transient analysis; Deconvolution; Objective function optimization; Reservoir characterization; Measurements uncertainty
Cite: Khalaf, M.S., Improving deconvolution reliability in well test analysis with an automated objective function, Deep Resources Engineering 2026, 3 (1): 100231. https://doi.org/10.1016/j.deepre.2025.100231
Citations: ……
