Engineers Partner With Hawaii To Study Renewable Energy

SOCORRO, N.M. August 25, 2010 – A team of New Mexico Tech scientists, professors and students has embarked on an ambitious project to improve the electrical grid’s capability of integrating renewable power.

Dr. Kevin Wedeward, chair of the electrical engineering department, is leading Tech’s partnership with the University of Hawaii. The two universities officially started the project May 1, with a $500,000 grant from the federal Department of Energy.


  New Mexico Tech students employed at ICASA are providing vital research for a partnership with the University of Hawaii. The project is examining how the existing electrical grid will respond to various inputs of renewable energy.  Photo courtesy of ICASA


“We’re interested in understanding the impact of renewable resources on the energy grid,” Wedeward said. “Wind power is less predictable because the variability of input. How does that affect the grid?”

Tech’s portion of the research involves modeling various scenarios, based at the Institute for Complex Additive Systems Analysis, or ICASA. Staff researcher Michael Smith, two undergrads and two graduate students are gathering historic wind data from Hawaii and modeling a simulated power grid that mimics Hawaii’s infrastructure. Once the models are built, they will run various scenarios on Tech’s supercomputer, Exemplar.

“People don’t fully understand the impact of renewable energy on the existing grid,” Wedeward said. “People can speculate where the problems will lie, but no one knows the magic threshold where something unintended will happen when wind power is put onto the grid. We need to answer these questions before the system exhibits undesirable behavior.”

Amanda Thom, a junior in civil engineering, is gathering information about the electrical grids on each Hawaiian island, which will be used to create a model of those grids.

“I want to know everything about their grid,” she said. “Hawaii is a great example for us to model because it’s an island grid. It’s the best area we can look at because the outside influences are limited.”

Thom said the utility companies are quite guarded about their proprietary information, largely due to security concerns. So, she is gathering data – mostly via the Internet – from the State of Hawaii, the Senate Committee for Energy and Natural Resources, the University of Hawaii, individual utility companies and state commission reports.

Once Thom has assembled as much information as possible about each islands’ grid, the team will select one of the islands to model – the one with the most and the best information. Once the simulated grid is created, the New Mexico Tech team will assemble tens of thousands of scenarios, then model and simulate those cases on Examplar. Each scenario will include differing variables, but will be based on historical wind data, demand cycles and disturbances.

Some scenarios will include heavy usage spikes or outages due to a tree falling on a transmission line.

“We’d like to run a whole suite of scenarios with different variables and unknowns,” Wedeward said. “We don’t have perfect models, but we will incorporate uncertainty and variability in supply, consumers, and grid.”

Exemplar can process 176 separate models at once. Wedeward expects to simulate tens of thousands of distinct scenarios.
Undergraduate student Josh Aragon will be working the closest with Exemplar. A Belen High School graduate, he already earned his bachelor’s in electrical engineering at Tech and is now finishing his bachelor’s in computer science.

Most of Aragon’s tasks revolve around programming, modeling and data reduction. Aragon is setting up the simulated environment in MATLAB and suggesting ways to write code for the supercomputer. In order to understand how disparate variables affect the electrical grid, Aragon is programming N-minus-one scenarios.

“Meaning, you take out one component and see if the system can continue to operate without that component,” he said. “A matrix represents the system mathematically. We’ll write code to choose which parts to take out.”

Natural wind fluctuations add yet another variable to the power generation equation. When wind dies down, coal-fired power plants – or petroleum power plants in Hawaii – have to pick up the slack.

“That’s a big worry,” Wedeward said. “Will adding wind just work the traditional plants that much harder instead of running at a nice constant pace?”

Michael Culp, a master’s student in mathematics, is programming the scenarios for the simulations that will account for the expected fluctuations in wind power, capacity and usage. He also is working on a sensitivity tool to gauge the affect of those changing variables on the grid.

“I’m working on changing values by slight amounts,” he said. “Say our load or power increases – we want to know how much does that affect the grid.”

Basically, Culp said, he is building a tool that Smith and Wedeward can use to analyze the modeling work.

Aragon said he’s excited to work with the parallel processors that are available on Tech’s supercomputer, Exemplar.

“It’s pretty neat to get to work with the Supercomputer,” he said. “The challenge for me is managing the data. We’ll be doing tons of runs, so I’ll be figuring out what’s meaningful in the end results.”

The Tech team is also modeling the control mechanisms that bring wind power into the electrical grid. In many cases (if not all), new infrastructure must be added to incorporate power from new sources.

“One issue is where you bring in that power,” Wedeward said. “If it’s a heavily-used transmission line, maybe the pipe isn’t big enough. So, you may need new lines. You need to put a wind farm where there’s wind, but also where there is capacity.”

Justin Goza, a master’s student in electrical engineering, is working on two tasks: creating a simulation of wind-power generator and “bifurcation,” which means solving the power balance equations that provide guidance as to when a black-out may occur.

Goza, who has already earned a bachelor’s in physics at Tech, said the challenge in modeling the doubly-fed induction generator is defining the conditions of the model.

For the bifurcation, Goza is using higher-order math to converge on an answer.

“When they don’t converge, your system won’t work and the entire grid won’t work,” he said. “I’m inputting all the unknowns into an algebraic equation until no possible solution can be found. You raise the load (power consumption) so high that it causes a blackout.”

This project fits perfectly in Goza’s academic plan and his career goals. His master’s thesis focuses on substation load flows. Plus, a career in electrical power runs in his family. His father works at the Four Corners Power Plant near Farmington; his uncle works at a power plant in Springerville, Ariz; and a cousin works as a lineman for an electrical company.

“This research job opens up all sorts of opportunities,” he said. “I’m working on something real and seeing results. It’s really been research non-stop.”

The State of Hawaii has set a goal of using 70 percent renewable energy by 2030; however, experts surmise that the maximum capacity is 30 percent because renewable energy sources aren’t as reliable as hydrocarbons.

“It’ll be really cool if we can figure out how to do it,” Thom said. “It’s an exciting project.”

The panacea to these issues is energy storage.

“That’s the Holy Grail,” Wedeward said. “If someone can come up with a cost-effective way of storing energy, these issues go away. Wind plus storage looks like a constant power supply. People are working on it, but none of the existing choices are good at storing energy. You need something pretty darned big to store it at those levels.”

– NMT –

By Thomas Guengerich/New Mexico Tech