Consumer demand, political demand and improved engineering are moving the world towards a greater use of renewable energy. However, reliability — a critical part of any energy source — remains a limiting factor.
How much solar energy can you generate on a cloudy day, or when snow blankets the solar panels? How many wind turbines can you operate when it’s not windy? These sound like simple problems, but they make renewable energy less reliable than fossil fuels.
You can’t control the sun, the wind or the snow. But using predictive and prescriptive analytics, along with optimization algorithms, you can do the next best thing: Optimize your energy mix.
You can build models that show, for example, what electricity flow you have in renewable energy vs. other sources, and augment that with data streams that project the weather. Exactly how much sun will we get? How much wind? What will our solar farms generally produce? This improves the reliability of renewable energy, and tells you exactly when you need to augment it with fossil fuels.
New technologies in renewables also create unknown or different maintenance practices. How do you know what your maintenance schedules should look like for a new system if you don’t have historical data? There are ways for energy optimization to help here too. By deploying drones to inspect wind turbines using high-res cameras and thermal imaging, you can get a better picture of each asset’s health, digitized and ready to be fed into maintenance optimization algorithms. This is particularly valuable for wind turbines and solar panels in remote places, or places that are difficult and expensive to have humans check. This enables people to make decisions about maintenance, rather than just do inspections.
Two Examples of Energy Optimization
FICO customers in Europe and around the world are solving these kinds of problems with optimization. Here are a couple of examples:Wind power optimization: WindFarm Designs has used FICO® Xpress Optimization to solve a wind farm layout optimization problem that has been studied for decades, which enabled it to create more efficient wind farms and increasing profitability for its clients. The layout optimization problem is how to best place turbines in order to maximize energy yield, while still respecting turbine power load constraints. WindFarm Designs, managed to integrate load constraints into the optimization problem, improving layout and resulting in 2-5 percent improvement in energy yield and profitability for project owners. In fact, WindFarm was a 2016 FICO Decisions Award winner for Analytic Excellence for this work.
Demand forecasting: Statnett, the Norwegian transmission systems operator (TSO) for energy, is using Xpress Optimization to forecast future electricity needs in Norway and optimize the current infrastructure. Statnett forecasts the country’s future electricity needs and market changes through 2040 and beyond using a combination of predictive analytics and optimization algorithms. This is critical work, since energy infrastructure is not fast or cheap to build — every power supplier must determine what kinds of energy to invest in to meet demands.
As someone with a lot of experience in the energy sector, I’m excited about the possibilities energy optimization has to help us move toward an eco-safer future.
