Eskom needed a visually interactive energy flow model, with forecasting, impact simulation and optimisation algorithms to enable planning engineers to model and determine the effects of different demand and supply side options on primary fuel costs and system profile impacts.
Development of the Eskom Energy Flow Simulator has been a highly successful four-year research assignment aimed as a decision-support system in the goal towards keeping the lights on and making Eskom a high performance organisation.
The model started off as a technical Energy Flow Simulator, including modelling for demand side initiatives and renewable generation, but has been expanded to provide visual and interactive feedback on decision effects in terms of the Eskom income statement, Eskom’s financial sustainability and its financial performance. This specifically was achieved regarding fuel costs (from a production cost minimising LP solver), and revenue derived from a sales perspective.
The Eskom Energy Flow Simulator has evolved into a model for diagnostic and predictive decision support on primary energy planning, primary fuel purchasing, maintenance planning, and thus on energy production simulation, where primary costs can run into tens of billions of rands.
This tool has reached its current form as a result of the working group model, and the collaboration and partnership between EOH and Eskom business and technical specialists.
Training about the statistical models contained in the tool and the academic concepts of its functioning is ongoing to assist with optimised decision-making and scenario analysis under constrained operating conditions.
The Energy Flow Simulator enhances decision support and helps Eskom keep the lights on.
Examples of its application include:
- Scenario analysis and investigation, including a weekly optimum fuel planning module, run for 5 years, on a weekly basis, optimising 14 energy generation slots per week, each with 168 hours.
- Scenario planning e.g. power station Energy Availability Factor (EAF) level modelling and national economic impact assessment.
- Sales forecast variation impact analysis on plant utilisation and fuel costs.