Perpustakaan Digital ITB

This study evaluates several commonly used Decline Curve Analysis (DCA) methods, namely Arps, Power Law Exponential (PLE), Stretched Exponential Production Decline (SEPD), and Duong, to analyze production decline behavior in unconventional gas reservoirs using production data from the Boggess Field. The results show that traditional DCA models have limitations because they generally require sufficient production history to calibrate model constants that are specific to each well or field. Therefore, this study develops a new hybrid Duong-Arps DCA model that only requires the peak production rate as the main input and uses fixed catalog-based constants derived from the Boggess field. The proposed model applies a sigmoid function as a time-dependent weighting factor to combine the early-time decline behavior of the Duong model with the smoother late-time decline behavior of the Arps model, expressed as ????(????) = ????(????)???????????????????????? +(1 ?????(????))????????????????????. The proposed model produces a forecast with an average error of 10% on the main dataset. This indicates that the hybrid Duong-Arps model can be used as an alternative earlystage forecasting method for unconventional gas wells when production data are still limited. However, once sufficient production history is available, calibrated traditional DCA methods, especially the Duong model with well-specific parameters, are still recommended to obtain a more accurate forecast.