Pulse Test: Underutilized Technology for Reservoir Management and Connectivity Understanding. Untapping the Technology with Signal Processing Techniques
Abstract Interwell connectivity evaluation is a fundamental component of reservoir management in both mature and undeveloped fields. This is particularly critical in geologically complex settings characterized by faulting, stratigraphic discontinuities, and potential compartmentalization. Pulse testing, which involves generating pressure pulses in a generator well and observing pressure responses in other wells (observation wells), presents a cost-effective method for assessing connectivity because no producers are shut-in (C.R. Johnson; R.A. Greenkorn). However, its practical implementation is often constrained by signal detection challenges-especially in low-permeability formations, reservoirs with high gas saturation, or when wells are widely spaced. In such cases, pressure signals become weak and easily obscured by operational noise or unrelated reservoir dynamics, making conventional interpretation approaches such as transient pressure analysis or numerical simulation less effective. This paper presents a dedicated pulse test workflow developed to overcome these limitations. The workflow integrates advanced signal processing techniques with analytical modelling to ensure both detectability and interpretability of pulse responses in complex and challenging environments. A key part of the workflow, critical for pulse test design and interpretation, is based on signal decomposition, statistical correlation techniques, and analytical models. In the interpretation stage, the method decomposes measured pressure data into signal components attributable to the pulsing well and background noise. By applying diffusion-based mathematical models to isolate and match the response, the workflow quantifies interwell transmissibility and storativity, while also estimating associated uncertainty based on the signal-to-noise ratio. The automated interpretation workflow consists of three main stages: Preprocessing: Removes background trends and noise using detrending, highlighting periodic features in gauge data,Signal Decomposition: Splits pressure data into two parts-one attributed to the pulsing well, the other to unrelated events – allowing a cleaner interpretationParameters Estimation: Calculates interwell transmissibility and storativity, with uncertainty linked to the signal-to-noise ratio and connectivity confidence. The methodology was applied in Field A, a clastic onshore oil reservoir with compartmentalization and an expanding gas cap. The pulse test campaign included three 240-hour cycles involving one injector and three producer-observation wells. The proposed signal decomposition approach enabled the identification of pressure responses with up to 86% correlation to the pulsing sequence, overcoming interference from operational noise and variable reservoir conditions. The derived time lags and amplitudes were used to compute interwell transmissibility, providing a quantified measure of reservoir connectivity across fault blocks. These findings supported decision making related to optimization of the injection strategy and justified new well placement decisions for a planned infill injector. This work demonstrates that automated signal processing significantly enhances the reliability and applicability of pulse test interpretation in reservoirs where the application of traditional methods is challenging. The workflow enables: Interpretation in high gas saturation and compartmentalized reservoirs previously considered unsuitable for pulse testing.Quantitative evaluation of signal presence and strength, offering confidence metrics for interwell connectivity.Operational efficiency, with no need for complex forward modelling or heavy manual curve matching.Scalability, supporting multi-well, multi-cycle pulse test campaigns. Ultimately, this paper provides a methodology and general workflow that extends the utility of pulse testing into more challenging reservoir environments by leveraging signal decomposition and analytical modelling.
Read more