Production performance forecasting method based on multivariate time series and vector autoregressive machine learning model for waterflooding reservoirs

ZHANG Rui,JIA Hu

Table 1 The result of well group division based on well history analysis.

Well group No. Number of injectors Number of producers No. of main development layer Development period (year-month-year-month) Data volume 1 2 4 27 1975-02—1987-03 146 2 2 5 27 1976-12—1982-09 70 3 3 5 27 1992-01—1998-03 75 4 2 4 33 1992-07—1997-06 60 5 4 6 27 1976-01—1992-08 200 6 3 5 25 2004-09—2018-03 163 7 3 6 27 1998-06—2009-04 131 8 2 4 37 2010-11—2017-02 76 9 3 7 33 2006-04—2018-03 144 10 3 5 27 2003-08—2018-03 176