Abstract

In 2023, Chesapeake Energy and ResFrac created the first-ever framework for hyper-calibration that includes time-lapse interference testing data in addition to using industry-standard data. The project aimed to develop a set of model parameters for predicting fracture and production characteristics within Chesapeake’s Haynesville asset. Rate Transient Analysis (RTA) and analytical models were insufficiently detailed to explain productivity changes over time and required non-generalizable corrections. A fully coupled hydraulic fracture, reservoir, and geomechanical numerical modeling approach successfully described the observed historical production across a defined region. Data used for modeling included time-lapse interference tests, treating pressures, perforation imaging data, laboratory data, and production histories. The project was set up in three primary phases, which can be described as follows:

• Phase 1 Calibration − Initial Base Model: Stress profile calibration and parent well history match

• Phase 2 Calibration − Time-Lapse Interference Test: Measured and matched conductivity loss using Devon Quantification of Interference (DQI)

• Phase 3 Calibration − Model Validation Using Child Wells: Applied calibrated fracture and reservoir model to offsetting wells

Investigation in Phase 1 indicated that individual mechanisms for Stimulated Rock Volume (SRV) degradation could account for some of the observed production trends. However, additional investigation with Phase 2 was necessary to find the unique combination of stress and time-dependent conductivity degradation, proppant conductivity characteristics, and matrix compaction curves necessary to capture the measured conductivity loss and production responses. In Phase 3, model parameters replicating interference testing results successfully predicted the performance of adjacent pads with differing completion designs in the area of interest, establishing high confidence in model predictivity.