The Use of Conventional Decline Curve Analysis for Predicting Future Performance of Tight-gas Wells
Engineering Honour Degree 2007
University of Adelaide
Decline-Curve Analysis is a method used widely for production forecasting and estimating remaining reserves in case of oil and gas reservoirs. Decline curve analysis (DCA) is based on the assumption that the past performance trends in the production can be characterized mathematically and can be used to predict future performance with a level of certainty. One of the basic assumptions while undertaking DCA is that the past production methods should remain unchanged.
DCA is one of the oldest and the most practical tools used by a reservoir engineer to predict the future performance of a reservoir. The method is deterministic and the estimation of the remaining reserves is based on known historical data. It is one of the most accepted methods used, partially due to the success of the predicted forecasts, and partially because of nonadequate prediction by other methods.  Wells in the same geographical area and producing from similar geological formations are expected to have similar characteristics in their declinecurve parameters. This knowledge is widely used in the industry for asset management by using the same decline-cure parameters to predict the performance of newly-completed wells.
As a part of this research, an Excel-VBA based interface computer program was developed based on the methods used for DCA and was later validated using published data. The developed program was then used to analyse the production data from five Tight-gas wells in the Cooper Basin. By comparing the results obtained using different methods, it is demonstrated that the Non-linear regression technique predicts the future performance of a reservoir closely. It was analysed that linear regression techniques impose more weight on smaller values of production rates, so were not the best estimator amongst the different methods used in the developed program. Another problem discussed was the inability of different linear regression methods to produce equivalent results.
Another issue which was highlighted in the report was the inclusion of transient data for Declinecurve analysis. Usually the production data is available over a long time span, which include both transient and boundary dominated flow data. For decline curve analysis, Arps’ equation should be applied only to data that have been affected by the boundary conditions. Inclusion of the production data for transient phase will result in over-estimated forecasts and is therefore technically incorrect. The report includes an equation which should be used to establish the boundary-dominated production data. The same equation was later used in the case studies to pick out boundary-dominated data for analysis.
But, tight-gas reservoirs present various challenges to the evaluation engineer. Applying classical reservoir engineering techniques to these reservoirs is challenging due to the length of time required to reach pseudo-steady state flow and to establish the drainage area. Using the developed program for tight-gas reservoirs in the Cooper Basin presented another difficulty. The decline exponent ‘b’, for all the tight-gas reservoirs, came out to be more than 1, and which violated the assumptions for the use of Arps’ equation.
Conventional decline curve analysis inherently assumes a single-layer reservoir, the well producing at constant bottomhole pressure and stabilized flow conditions. In addition, it is documented that the decline exponent should range between zero and 1. Although the existing methods provided a close representation of the historical tight-gas production data, but the Arps’ equations resulted in invalid results. Thus the need for the development of new equation is proposed which can be used for the value of decline exponent.