Sustainable energy sources and technologies are the future, but how do we best integrate them into our present? C. Lindsay Anderson, Biological and Environmental Engineering, works on answering this question. Anderson’s goal is to come up with ways to add renewable energy sources such as wind, solar, and biofuels into today’s energy systems with the help of tools from applied mathematics, like systems modeling and optimization.
The challenge of the work is in the uncertainty of renewable energy sources available today. There’s logistical uncertainty: Wind or solar energy, on a given day, may not generate enough energy. Then, there’s financial uncertainty: It’s unclear how much biofuels are worth and how to sell them. At some point, technology will improve and the uncertainties will be less of a problem, but for now, Anderson is concerned with taking what’s available and making it work.
“Why wouldn’t we try to use all the tools that we have at our disposal to get the most out of what we have,” she asks. “The technology has constraints, but we can make the best decisions possible and extract value—that can be financial value, economic value, and getting the same products at the same price with a lot less environmental impact.”
Examining the Details of Renewable Energy Integration
Energy systems are not, however, plug and play, says Anderson. It doesn’t make sense to just add new technology to an energy grid without a strategy in place. Researchers and industry have to examine each system, factoring in the renewable energy, and come up with strategies and models that will address potential problems. Anderson and her lab are looking for flexible operations strategies that manage for various kinds of risk, be it financial risk or reliability risk, such as potential blackouts.
One of Anderson’s graduate students, Maureen Wanjiku Murage, is examining the Lake Turkana Wind Power project in Kenya. The wind farm will be fully operational in 2016 and will provide the country with 20 percent of its electricity. The challenge with generating so much electricity from a wind farm is that it’s hard to forecast how much energy it will produce on a daily basis.
For the research, Murage visited Kenya Power and Light, the power system operators in the region. She simulated the wind farm and found that daily wind patterns in the region did not match up with daily energy loads. To increase reliability of the system, Murage in a May 2013 paper in Renewable Energy suggests the implementation of pumped hydro storage as a way to reduce the wind farm’s power output shortage. This approach would not only increase energy reliability but also reduce economic burden as it has the potential to increase the wind farm’s daily revenue.
Because there is no current economically feasible way to store energy, it’s necessary to come up with models and operational strategies that work on an instantaneous timescale, down to an hour-by-hour basis. Take New York State’s power system, which has more than 500 generators in its network. If wind power is added, the system needs to balance what’s going in with what’s going out.
“We are trying to come up with better ways to model the power system decisions so we can make very detailed decisions and react for flexible management,” says Anderson. The conservative method is to let generators spin and overproduce energy, because nobody ever wants to have a blackout. “We’re saying we can do better than that. We can look at all of the possible outcomes and make sure that we have the best plan to manage any of those outcomes.”
That said, it’s not feasible to look at every single possible outcome, as there are infinite possibilities. Anderson’s graduate student Laura Tupper, Statistical Science, is working on characterizing the most important and most representative uncertainties of wind power. Tupper is developing models that can identify the specific set of possibilities—the goal being to arrive at the smallest set without losing information.
Because there is no current economically feasible way to store energy, it’s necessary to come up with models and operational strategies that work on an instantaneous timescale.
The work fits squarely into what Anderson is all about. “It has nothing to do with engineering a product at all, but doing a good job in this modeling part means that you can make better decisions about how to implement or use the product,” she says.
A Convergence of Knowledge
Anderson says she’s especially excited about the convergence of expertise taking place in power systems studies today. “We are now getting to a space where everyone understands the importance of the problem, they understand the challenges of the problem. We have people in power systems, operations research, and mathematics working together,” Anderson explains. “You need people who are experts in all of the different systems to make the whole system operate better.”
The convergence of experts is especially crucial to developing faster computational algorithms for industry use, says Anderson. To understand the best decision for a range of possible outcomes is one thing, but to be able to have an algorithm solve the problem quickly, on the scale of an hour, will have a huge impact.
Anderson says that working with her colleagues at Cornell is a boost to tying research with industry. Specifically, she credits the Engineering and Economics of Electricity Research Group (E3RG), which consists of more than a dozen professors across disciplines. “It was developmentally important for me to really integrate into the Cornell research landscape, as well as being super critical in integrating research with industry,” says Anderson. “That was the key group that really accelerated my ability to start getting things done here.”