COMPARATIVE STUDY OF DECLINE CURVE ANALYSIS METHODS USING A LAB-SCALE GAS RESERVOIR
Open Access
- Graduate Program:
- Petroleum and Mineral Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- June 16, 2016
- Committee Members:
- Luis F. Ayala H., Thesis Advisor
- Zuleima T. Karpyn, Committee Member
- Turgay Ertekin, Committee Member
- Keywords:
- Natural Gas Engineering
- Reserves Estimation
- Reservoir Engineering
- Decline Curve Analysis
- Original Gas in Place
- OGIP
- Lab-scale Reservoir
- Abstract:
- One of the most effective approaches to estimate Original Gas in Place (OGIP) in dry gas volumetric reservoirs is the use of Decline Curve Analysis methods. The strength of these methods is that they rely on the availability of initial reservoir pressure and production data (flow-rate vs time), which are generally abundant, to generate estimates of original gas in place and future production predictions. Decline curve analysis methods are generally validated using field production data or data from computationally reservoir models. Some disadvantages of these validation approaches include the fact that gas reserves cannot be readily obtained from field data, and the accuracy of production predictions from reservoir models is subject to the model reliability. The aim of this study is to investigate the use of a lab-scale gas reservoir to generate reliable production data for a rigorous validation of decline curve analysis methods recently proposed in the literature. The methods under consideration are Flowing Material Balance (Mattar and Anderson, 2003), Ye and Ayala (2012, 2013), Stumpf and Ayala (2016), and Zhang and Ayala (2013, 2014a, 2014b). The lab-scale reservoir was designed, built and tested in a number of experiments, performed at different initial reservoir pressures and confining pressures. The production data obtained were used to estimate OGIP and compared against direct volumetric calculations. The divergence between these two values was called “error”. OGIP estimates showed good agreement with lab data, with variations in performance quality. The decline models proposed by Ye and Ayala (2012, 2013) and Zhang and Ayala (2013, 2014a, 2014b) yielded the most accurate estimations of Original Gas in Place, with an average error of 8.32 % for the first method and 8.67 % for the second. The Flowing Material Balance method was found to underperform for most lab conditions tested, showing an average error of 11.63 %.