Research using predictive modeling to estimate impact of climate change on water and electricity (Purdue University photo/Rebecca McElhoe)

In 2014, residents of Phoenix were asked to plant drought-tolerant plants to conserve water. But the change in vegetation actually made the air feel hotter, leading to residents using more air conditioning powered by water-generated electricity.

The point: Without a better understanding of how climate change affects the connection between energy and water demand, environmental policies could do more harm than good.

Ignoring the interdependence between energy and water systems could bring citywide power outages or shortages, says Roshanak “Roshi” Nateghi (roh-SHAH-nahk nah-TEH-ghee), a Purdue professor of industrial engineering.

Nateghi’s lab develops predictive models to estimate the impact climate change could have on electricity and water use.

“We leverage recent developments in climate science and data science to understand the impact of climate change on engineering infrastructure and communities that rely on their services,” Nateghi said.

The models seek to understand the drivers of change in demand for electricity and water over time. In researching how demand within the energy grid will change, communities will be able to better manage engineered infrastructure systems and even design systems to be more resilient and sustainable.

Nateghi’s research is looking to uncover gaps and assumptions in energy research that have led to underestimating the effects of climate change, like taking into account how electricity and water are connected and used together.

Providing electricity for the average home would require 114 gallons of water per day at a natural gas power plant, Purdue researchers wrote in a Scientific American article.

One of Nateghi’s predictive models projected an increase of electricity and water use due to climate change between 2030 and 2052. The model projected that the Midwest as a whole will be using 19% more electricity and 7% more water during the summer periods. An increase of energy usage could result in supply inadequacies or outages.

Other models predicting the effects of climate change on energy use also have not considered how people react to humidity during the heat of the summer.

Nateghi and her team developed a model that more effectively captures heat stress by taking into account temperature, humidity and pressure. The model moves beyond the typical baseline climate data such as temperature to more accurately predict how climate change could affect home electricity usage during the summer. Understanding how people respond to heat stress would help in the future regulation and policy of energy.

“When it comes to demand forecasting under climate variability, our focus is pretty unique in that we are using different types of algorithms and looking at climate variables more holistically. This approach is to better capture the complex relationship between climate and demand,” Nateghi said.

Using multiple climate-change simulations run by supercomputers and top scientists as input, Nateghi leverages artificial intelligence to predict the impact climate change on demand. The predictive models help determine how demand of electricity and water could change over time due to various climate-change scenarios. Nateghi’s algorithms are trained with years of observational data to “learn” people’s dependence on energy under climate uncertainty.

After dependency is estimated, Nateghi returns to the climate simulation models in order to see how input variables like temperature, humidity and wind speed are projected to change over time.Nateghi’s predictive models are rerun to establish how projections of demand for electricity and water change as the climate changes. In other words, the final projections look at the impact of climate change on electricity and water usage.

Underestimation of energy use is a key factor that threatens the resilience of the energy grid. Past research on energy adequacy planning does not consider the asymmetrical response of demand to temperature variability and is usually based on the average values of demand distributions.

“If you think about the people in the 90th percentile, who are among the higher-intensity users, their sensitivity to climate is actually much more than the average. That sensitivity of high-intensity demands is really getting overlooked in current projections, which could lead to supply shortages with higher frequency and intensity of heat events under climate change,” Nateghi said.

The models that capture dependency between energy demand and climate change are over 90% accurate. The nuanced information provided by the predictive models can assist in better planning and managing of the energy systems and inform regulations and provide evidence for a change in energy policies.

Despite the accuracy, there is still variability when projecting into the future due to an uncertainty associated with climate-change simulation models.

“The wide range of variability in the projections makes planning and management of our engineered systems more challenging under climate change because it means we have to come up with more robust and flexible strategies,” Nateghi said.

About Purdue University

Purdue University is a top public research institution developing practical solutions to today’s toughest challenges. Ranked the No. 6 Most Innovative University in the United States by U.S. News & World Report, Purdue delivers world-changing research and out-of-this-world discovery. Committed to hands-on and online, real-world learning, Purdue offers a transformative education to all. Committed to affordability and accessibility, Purdue has frozen tuition and most fees at 2012-13 levels, enabling more students than ever to graduate debt-free. See how Purdue never stops in the persistent pursuit of the next giant leap at purdue.edu.

Media Contact: Kayla Wiles, wiles5@purdue.edu (working remotely, but will provide immediate response)

Source: Roshanak “Roshi” Nateghi, rnateghi@purdue.edu

A publication-quality photo is available at https://www.purdue.edu/uns/images/2020/nateghi-profile.jpg