GAS NATURAL FENOSA Produces Forecasts for the Electrical Market Using MATLAB’s optimization software

MathWorks today announced that GAS NATURAL FENOSA used MATLAB products to develop models that enable the company to project capacity and demand and optimize generation of asset portfolios. In particular, the company develops optimization and forecasting models that incorporate historical usage patterns, weather forecasts, production costs and regulatory constraints, and other operational factors. As a result, GAS NATURAL FENOSA has doubled staff productivity and can adapt more quickly to regulatory changes, reducing its response time from months to one or two weeks.

GAS NATURAL FENOSA engineers used MATLAB to develop a set of core models that analyze available data, forecast results, and optimize generation plans. Each MATLAB model accesses a central database for historical power consumption and price data, weather forecasts, and parameters for each power plant. Optimization Toolbox  was applied to minimize production cost among several plants given a set of constraints, including carbon caps and maximum capacity. The engineers used Statistics Toolbox to develop and assess price simulation scenarios.

Using MATLAB Compiler, the team created standalone programs from each model that run automatically day and night, enabling the developers to more easily manage updates to the models and access to the models for a variety of end users providing improved management of updates and control of access to models. The team also used Simulink to model the behavior of generators in the GAS NATURAL FENOSA infrastructure.

“Our market changes very quickly, so we need to know how to promptly respond to changes in regulatory standards or in the structure of the electricity industry, as well as to other factors, such as increased production of renewable energy,” said Isaac Pérez, head of the Iberian Electricity Markets Technical Office. “We tried using a commercial software package without development and customization capabilities, but it did not address the numerous problems we needed to solve. In our circumstances, closed systems don’t work well. We needed an open platform that would enable us to develop our own algorithms and computations.”

“A company like GAS NATURAL FENOSA, with a generating capacity in Spain of over 15,000 megawatts and a portfolio of assets including a variety of generating technologies, needs to accurately predict the different variables affecting each of the markets where it operates,” said Juan Nasarre, managing director of MathWorks for Spain and Portugal. “MathWorks tools enable GAS NATURAL FENOSA to detect better business opportunities, reduce generating costs, and therefore improve its sales margin.”