Three professors at the School of Electrical, Computer and Energy Engineering at Arizona State University are partnering with Arizona Public Service (APS) on an ambitious project to research the effects of high penetration of solar energy on an electric grid system. Working with General Electric, National Renewable Energy Laboratory and ViaSol Energy Solutions, the Community Power Project is bringing more renewable energy to the Flagstaff, Ariz. area than ever before. The pilot program also is providing valuable data to guide larger scale distributed generation from solar panels nationwide.

[This research] will lead into much wider adoption of photovoltaic systems as a whole.—Raja Ayyanar, associate professor

In 2010, APS recruited 125 Flagstaff residents to host solar panels on their rooftops and feed into the power grid in return for a fixed rate on their electric bill. The power company owns the equipment and will maintain it over the next 20 years for the course of the study. The Department of Energy provided a $3.3 million grant to the project in 2009.

The solar energy generated by the residential rooftops accounts for one-third of the total 1.5 megawatts of solar energy that APS is producing from this stretch of the power grid. Another third comes from a commercial installation at an elementary school. The final portion comes from a small solar power plant APS built in the neighborhood.

Having this variety of power sources explores the distribution system in a way that one large generating facility could not. Spread out across nine miles, the amount of equipment and the various points where the power travels to and from provides a much more complicated model for the ASU team to simulate and test.

The area power lines currently have a peak load of seven megawatts. Under certain conditions, solar power may account for 30 percent of the total power on them. This is the first time such a high percentage of solar power on the grid will be studied. Ordinarily, solar power on a feeder – or primary distribution line – amounts to well below five percent of the power.

Raja Ayyanar, associate professor, Vijay Vittal, professor and Gerald Heydt, Regents’ Professor, are working together to model the entire system and simulate the behavior of the grid as it responds to the varying levels of energy feeding in from the solar project.

Ayyanar explains that, “We want to understand what can go wrong with the power grid in simulations and field tests in a controlled and safe manner, before similar systems are installed on a larger scale and on several other feeders.” They are producing a simulation using a Geographical Information System that has the location of the 10,000 pieces of individual equipment and conductor segments that APS has installed.

“Our role is modeling the entire feeder, the photovoltaic systems and the protection systems,” he says. “They added the electrical circuit diagram to the GIS, and now the model can show the voltage at any node of the system and the current flowing through each of the segments at any given time.”

With this system, the ASU team is able to predict the behavior of the grid and whether there will be any problems involving the direction of electrical flow or negative effects on the protection system or if there will be any chance of unsafe conditions for workers when repairing the power grid. The team also can quantify the benefits in terms of reduced losses, reduced energy from fossil fuels and better power quality.

“We get a huge amount of measurements from the field and match them to see if there are any discrepancies within our simulation and, if needed, we fine-tune our model,” says Ayyanar. Vittal explains that the system monitors the penetration level at varying times from seconds to minutes, providing a range of data for the team to compare to their model.

“Diversity is an important thing to measure,” Vittal says. The frequency and rate of measurements is another feature of the project that makes it stand out from previous studies.

“In terms of the amount of penetration and how much data we collect and how we use it, this is certainly pioneering work,” says Ayyanar, “It will lead into much wider adoption of PV systems as a whole.”

Professor Heydt notes that the detailed instrumentation will identify any problematic operating conditions.  One area that is being studied in detail is known as “electric power quality” said Heydt.  Data from the study are being used by ASU students in graduate level projects.

Their simulation will help provide answers as to what still needs to be improved with power grids before solar power can be implemented on a grand scale. It will also help the team discover how much farther they can push solar penetration on the grid beyond 30 percent.

With this implementation on such a grand scale, Ayyanar and his associates are beginning to brainstorm future steps to improve similar projects. Microgrids are one idea with which the team is toying. A microgrid is connected to the grid under normal conditions but during a grid disturbance can disconnect and form a small grid of its own supporting local loads with local generation.

Microgrids are one aspect of a “smart grid” which is the effort to bring the electric grid in to the computer age.  With a smart grid, the grid operator has precise and real-time information on the status of every piece of equipment and can respond instantly to failures and disturbances.  Similarly the consumer is empowered with information on their own load and grid conditions to make informed choices about their electricity use patterns.

As part of the Community Power Project, the ASU team will study all aspects of smart grid and make recommendations to APS for future implementations that would benefit APS and its customers.