Data-driven analytics for early warnings in reservoir and well performance

Canchucaja R.F. · Society of Petroleum Engineers - SPE Russian Petroleum Technology Conference 2020, RPTC 2020 (2020) · DOI: 10.2118/201930-ms

Resumen

This study is aimed to detect deviations by using an algorithm fast enough to process data and to elicitinformation as raw data is being gathered in order to calculate past and current performance parameters andto calculate data-driven future performance. Data manipulation is normally limited to gathering for historymatching purposes, but it is not usually processed with data mining techniques to detect performance patternchanges, to discriminate between stable and non-stable periods, to detect changes of fluid density and toperform production forecasting. The algorithm is developed in R language due to the easiness to executestatistical operations with a great deal of data. Copyright 2020, Society of Petroleum Engineers.