A correlation for reservoir characterization using recorded real-time surface drilling parameters and well log data

Recent economic developments of the US gas market and enhanced technological improvements lead towards an increase of future operations in the sector of shale gas plays. The Eagle Ford field in Texas, being amongst the youngest US shale plays, serves as a good example of how correlating recorded real-time surface drilling parameters and well log data can be used to improve reservoir characterization. Variations of properties occurring horizontally and vertically, across the entire play or even along the wellbore are regarded as a major challenge directly affecting the economic development of shale gas reservoirs. An enormous amount of data is collected at present but not analyzed and evaluated in detail. Instead the trend is evolving that more data is generated, resulting in the incapability to integrate the data. Regression analysis is used to determine quantitative relationships between a real-time surface drilling parameter and petrophysical logging data for wells located in the same geographic and geologic area. This research describes how the rate of penetration correlates with the gamma and acoustic log (slowness of elastic waves) for the predominant shale section of each well and how the regression outputs contribute to optimize reservoir characterization. Within the shale formation, the gamma log (GR) shows a good correlation with the rate of penetration. Information from mudlogs and daily drilling reports is used to identify possible reasons for mists between the actual and the calculated rate of penetration. Studying a defined set of data in depth has proven to be a reliable indicator for comparing and categorizing wells. The results depict similarities and differences amongst the wells based on the properties of the formation they were drilled in. It is expected that additional real-time surface drilling parameters besides the rate of penetration are useful to obtain improved results. They can be used to normalize the rate of penetration to optimize the comparison between wells and to detect mists between the regression output and the actual rate of penetration measured on a real-time base.