Onugha, Ifeyinwa Ngozi
The Impact of Production Data Uncertainty on Assisted History Matching and Performance Prediction
In assisted history matching, the focus is usually to generate multiple fitting models to the observed production data for the purpose of forecasting the future performance of the reservoir. The historical production data is often considered as being the truth case and is used as the basis to match these models. In reality, these data are prone to measurement error. Measurement error in recorded production data are often due to well mechanical problems and database artefact errors during production data entry.
The research methodology essentially involves data analyses of observed field and well historical fluid production and fluid injection profiles, well bottom hole pressures and fluid injection pressures to account for model inadequacy resulting from database artefacts and wellbore mechanical problems. The “misfit” term the basis of which reservoir forecasting is made is then redefined to include model inadequacy resulting from production data uncertainty.
Prof. Mike Christie, Dr Dan Arnold