With a global push to use more renewable energy sources, wind presents a promising, underutilized resource. Despite numerous technological advancements in upgrading wind-powered systems, a systematic and reliable method of assessing competing technologies has proven difficult. Researchers compared the performance of various wind turbine designs using advanced data science methods and ideas from the social sciences.
With a global push to use more renewable energy sources, wind presents a promising, underutilized resource. Despite numerous technological advancements in upgrading wind-powered systems, a systematic and reliable method of assessing competing technologies has proven difficult.
In a new case study, researchers at Texas A&M University used advanced data science methods and ideas from the social sciences to compare the performance of different wind turbine designs in collaboration with international energy industry partners.
“At the moment, there is no way to validate whether a newly developed technology will increase wind energy production and efficiency by a specific amount,” said Dr. Yu Ding, Mike and Sugar Barnes Professor in the Wm Michael Barnes ’64 Department of Industrial and Systems Engineering. “In this study, we provided a practical solution to a long-standing problem in the wind industry.”
Researchers have used advanced data science methods and ideas from the social sciences to compare the performance of different wind turbine designs.
The results of their study are published in the journal Renewable Energy.
Wind turbines convert the energy transferred from the air hitting their blades into electrical energy. Wind energy will account for approximately 8.4 percent of total electricity production in the United States by 2020. Furthermore, the Department of Energy intends to increase wind energy’s footprint in the electricity sector to 20% over the next decade in order to meet the country’s ambitious climate goals.
In keeping with this goal, there has been a surge of novel technologies, particularly with regard to the wind-rotating blades. These upgrades promise to improve wind turbine performance and, as a result, power production. However, determining whether or not these quantities will rise is difficult.
One of the many factors that makes performance evaluation difficult is the sheer size of wind turbines, which are frequently several hundred feet tall. It is impractical to test the efficiency of these massive machines in a controlled environment, such as a laboratory. Using scaled-down versions of wind turbines that fit into laboratory-housed wind tunnels, on the other hand, produces inaccurate values that do not capture the performance of the actual-size wind turbines. In addition, the researchers noted that replicating the wide range of air and weather conditions that occur in the field is difficult in the laboratory.
As a result, Ding and his colleagues chose to collect data from inland wind farms for their study by collaborating with a wind farm-owning industry. They included 66 wind turbines on a single farm in their analysis. These machines were outfitted with sensors that continuously tracked various items such as turbine power, wind speeds, wind directions, and temperature. The researchers gathered data for four and a half years, during which the turbines received three technological upgrades.
Ding and his colleagues – We outfitted these machines were unable to use standard pre-post intervention analyses, such as those used in clinical trials, to assess the difference in power production and performance before and after the upgrade. In clinical trials, the efficacy of a particular medicine is tested using randomized experiments in which test groups receive the medication and controls do not. The test and control groups are carefully selected to be otherwise comparable so that the effect of the medicine is the only factor distinguishing the groups. However, the wind turbines in their study could not be neatly divided into test and control-like groups, as required for randomized experiments.
“The challenge we have here is that even if we choose ‘test’ and ‘control’ turbines in the same way that clinical trials do, we still cannot guarantee that the input conditions, such as the winds that hit the blades during the recording period, were the same for all the turbines,” Ding explained. “In other words, in addition to the intended upgrades, we have a set of factors that are different pre-and post-upgrade.”
As a result, Ding and his colleagues turned to causal inference, an analytical procedure used by social scientists in natural experiments. Despite the confounding factors, the analysis allows one to infer how much of the observed outcome is caused by the intended action, which in this case was the turbine upgrade.
The researchers included turbines in their causal inference-inspired analysis only after their input conditions were matched. That is, during the recording period, these machines were subjected to similar wind speeds, air densities, and turbulence conditions. The research team then reduced the uncertainty in quantifying if there was an improvement in wind turbine performance by using an advanced data comparison methodology developed by Ding in collaboration with Dr. Rui Tuo, assistant professor in the industrial and systems engineering department.
Despite the fact that the method used in the study necessitates months of data collection, Ding claims that it provides a robust and accurate way of determining the merit of competing technologies. He believes that this information will be useful to wind operators who are deciding whether to invest in a specific turbine technology.
“The federal government still subsidizes wind energy, but this will not last forever, and we need to improve turbine efficiency and cost-effectiveness,” Ding said. “As a result, our tool is critical because it will assist wind operators in identifying best practices for selecting technologies that work and weeding out those that don’t.”
Ding received a Texas A&M Engineering Experiment Station Impact Award in 2018 for data and quality science innovations that have impacted the wind energy industry. Nitesh Kumar, Abhinav Prakash, and Adaiyibo Kio from the industrial and systems engineering department, as well as technical staff from the collaborating wind company, also contributed to the research.
The National Science Foundation and industry are funding this research.