Empirical methods of decline-curve analysis for shale gas reservoirs: Review, evaluation, and application

Abstract

Shale gas reservoirs are accessed using horizontal wells with fracturing technology. To improve production, the bottom-hole pressure is usually controlled. However, predicting the estimated ultimate recovery of the well becomes difficult with this production mode because the decline curves have a long tail and the production data changes dramatically. Comparing six decline-curve analysis methods shows that the empirical method is easiest. Therefore, 24 empirical models are summarized, and 18 basic models are further analyzed by using actual production data. The results show that most empirical models are based on exponential or power functions, especially using the exponential function widely in recent years. Forecasting production in these models does not always appear as a monotonic decline with the influence of model structures. Some models can transform into others under certain conditions. The dimensions of some models are not uniform because the units of their time exponents are always neglected. Empirical models are influenced by data, and none of the models can be applied in every case because of varying data. The number of reliable models increases with an increase in the amount of data and a decrease in fluctuation. The relationship between empirical models and data characteristics rather than flow regimes should be noticed. General methods of noise reduction are not effective with the drastic fluctuations of shale gas wells, so improving production data preprocessing is important. Although model recommendations are given in this paper, a detailed selection principle should be further studied based on more data from different shale gas reservoirs.