Variation in b-sigmoids with flow regime transitions in support of a new 3-segment DCA method: Improved production forecasting for tight oil and gas wells

Abstract

This study uses a commercial reservoir simulator to generate production rate data for shale wells with systematic variations in fracture treatment design parameters using a plausible range of reservoir properties. The synthetic well data is then used to generate diagnostic log-rate log-time plots, which allow for a distinction of four types of flow regimes. Such a distinction of flow regimes is of practical relevance, because of the inverse link with reservoir properties and fracture treatment design parameters. Each flow regime has a characteristic slope, which traditionally has been attributed to typical flow conditions in the reservoir. The synthetic production data from the reservoir simulator are used to constrain b-sigmoid patterns for narrow discrete time increments of production (1 month) throughout the various flow regimes occurring during the economic life of each synthetic well model. Nearly a hundred different models were generated as a basis for our analysis. From the analysis of the DCA parameters for time series, using the difference in inverse of decline rates (loss ratio) for two adjacent time steps, it appears possible to recognize systematic shifts in the instantaneous decline rates and temporal b-values. In particular, the generated b-value patterns (b-sigmoids) appear diagnostic for flow regime changes recognized on diagnostic log-rate log-time plots. The results demonstrate that it is possible, based on well design and reservoir parameters, to establish correlation trends between b-sigmoids and flow regime changes using time series of the DCA parameters. The results are subsequently used to improve the accuracy of a modified 3-segment DCA method based on Arps equation. The proposed method is applied to synthetic production data, generated based on an Eagle Ford shale oil well (with known reservoir and fracture properties), to demonstrate the practical value for production forecasting and reserves estimation. This research paper paves the way for wider, future application of the methods described.