Abstract:
Uncertainty quantification of power (MWe) prediction from a geothermal numerical model is time-consuming. Existing techniques and workflows are still being tested. Our study focuses on using the Experimental Design (ED) and Response Surface Methodology (RSM) framework as an alternative method to facilitate probabilistic MWe estimation. This study involves two types of Experimental Design: the two-level Full Factorial and Plackett Burman, and the investigation is divided into two parts. The first part investigated the difference in the prediction of the Wairakei geothermal field power potential using the polynomial model built from the 64 simulation runs based on the two-level Full Factorial design and the 12 simulation runs based on the Plackett-Burman configuration. The second section compared the MWe predictions using the Plackett-Burman design and the volumetric stored heat method. The simulated MWe potential of Wairakei for 50 years is comparable for both the two-level Full Factorial and Plackett-Burman designs, indicating the latter as a practical design for building a proxy model of a numerical model. Furthermore, the MWe prediction of the polynomial model created based on the ED Framework closely resembles the actual production capacity of Wairakei compared to the volumetric stored heat method. The results based on the proxy modelling further indicate that the 353 MWe production generation could be sustainable at least for another 40 years.