Application of Nonparametric Regression on Well Histories of Geothermal Production Fields in the Philippines

Abstract

The nonparametric regression ACE (Alternating Conditional Expectation) method was applied to the different production fields of Energy Development Corporation (EDC) in the Philippines. Production chloride chemistry well histories were correlated to injection rate histories based on nonparametric regression to establish well-to-well connectivity. The method of inferring well connectivity by this approach can be conducted with routinely measured production and geochemical data which does not require operational disruption that would be needed with the typical tracer test. Attaining an understanding of the connections between wells is very useful in designing a strategy for brine injection, and predicting where thermal breakthroughs are likely to occur. The comprehensive field wide application of the forecasting method shows that applicability and usefulness of the technique should be calibrated with actual tracer tests data. The equation derived for well connection indices was found to be applicable only to Palinpinon-1 in Southern Negros Geothermal Production Field.