Statistical Analysis of Raageshwari Deep Gas Fracturing and Production Logging Data for Understanding Fracture Performance and Future Optimization

Abstract

Raageshwari Deep Gas field situated in southern Barmer Basin of India is a retrograde gas condensate volcanic reservoir. More than 150 fracturing treatments have been pumped in this reservoir to achieve sustained economical production. The paper describes the results of a statistical analysis done to find correlations between production data and fracturing design parameters including petrophysical and geo-mechanical properties of the rock, pre-frac diagnostic tests, and fracturing treatment data including both pumping data and pressure matched parameters. This paper uses the data from multiple production logs to generate stage wise Productivity Index (PI). This PI data was then cross plotted against various parameters and combinations of parameters such as In-situ proppant concentration, formation porosity, net pay, average stress, proppant mass pumped per stage, fracturing fluid recovery rate and percentage, and fracture dimensions. One interesting line of investigation looked at the rate of pressure decline post Step Rate Test (SRT). A method was developed to evaluate the SRT declines even though they were too short to analyze for permeability using post closure analysis. This paper presents the results of these statistical analysis and where reasonable correlations were obtained. It also shows that for this volcanic formation, the rate of pressure decline after the SRT is a better indicator of Reservoir Quality (RQ) and future stage performance than the log derived porosity and permeability. While the use of short term fall-off data is only qualitative, it does appear to be an effective tool for evaluating the potential of a stage just before fracturing which would allow improved onsite treatment optimization. Since the quality of a reservoir generally varies across the areal extent of a field, it is very important to ascertain the same either qualitatively or quantitatively. This paper presents a technique for qualitatively defining RQ, which can be useful to validate the pre-existing workflows used for defining RQ.