Case Study - Energy forecast and optimisation

The objective of the project was to increase the net energy yield of a large Power Plant (geothermal and solar).

The Challenge

To increase the energy yield of the power plant, engineers wanted to understand how they could better allocate existing geothermal resources (brine, steam, and NCG flow rates and temperatures) through modelling the ongoing generator efficiencies and correction curves. For optimal performance,all internal and external variables were considered, including energy tariffs for different licences, flow rates and temperatures of geothermal resources,environmental conditions (temperature,humidity, and wind), and generator-specific modeled thermodynamic variables (curves)and technical constraints.

The Solution

First, we built a custom prediction model to predict the Net Power Factor (efficiency) of a generator considering the internal and external parameters. Advanced mathematical tools and machine learning were used for these rigorous models.  

Results

40-percentage point increase in accuracy compared to previous forecasts

Our prediction model achieved an accuracy of 94% compared to the realised production figures for 2022 and 2023, representing a 40-percentage point increase in accuracy compared to previous forecasts (54%). The second phase of the project involves deploying an optimisation model to enhance production profitability through quantified recommendations around the allocation of resources and equipment parameters.

‍

‍