The Pacific Northwest Smart Grid demonstration, one of the largest in the United States, has concluded with what are some important lessons for moving forward with smart grid deployments both in the US and elsewhere.
The 5-year, US$178-million research project included 11 utilities, the Bonneville Power Administration, two universities and multiple technology companies in the five Northwest states of Washington, Oregon, Idaho, Montana and Wyoming. It was supported with funding through the American Reinvestment and Recovery Act.
“The Pacific Northwest Smart Grid Demo experienced much success, while also identifying many opportunities for growth,” said project director Ron Melton of Battelle, which led the project. “The knowledge gained through this project will help prepare the region and nation for a bright energy future that strengthens our economy, protects our environment and enhances our quality of life.”
Smart technologies evaluated
The project evaluated 55 different technologies, many of which showed they can reduce energy use and possibly also cut power bills. The degree of savings varied with the technology and test location. Among examples, smart meters with remote capabilities enabled Avista Utilities to immediately start and stop electric service in Pullman, Washington, which the utility reported saved time for its customers and could both eliminate 2,714 service calls in a year and save about US$235,000 annually.
A 5MW/1.25MWh battery installed in Salem, Oregon, could save Portland General Electric up to US$146,000 annually by providing an alternative power source during periods of extreme peak power demand up to 300 times a year.
Voltage controls reduced Avista's distribution system voltage by 2.1%, which is expected to translate into about 7.8GWh of annual energy savings and US$500,000 in reduced annual costs for its Pullman feeder distribution power lines.
Tests also showed some technologies can improve reliability. For example, NorthWestern Energy determined the fault detection, isolation and restoration system it installed significantly reduced two power outages. A June 2013 outage in Helena, Montana, lasted just 51 seconds for customers served by a power feeder line connected to the fault detection system, while customers served by another line were without power for 119 minutes.
Avista reported the fault detection, isolation and restoration system and other reliability enhancements it installed led to an annual average of 17% fewer outages and more than 12% shorter outages, as well as 353,336 avoided outage minutes between August 2013 and December 2014.
Transactive control works
The project also demonstrated that the concept of transactive control works and potentially provides many benefits on a regional power grid.
The project used transactive signals that represented the predicted price and availability of power in the present and several days into the future. These signals were updated every five minutes and sent to participating utilities. When transactive signals predicted peak power demand, and therefore also high costs, the project's smart grid technologies were designed to reduce power use.
To help test the transactive control technology, Alstom Grid built a model of the regional grid, which ran in parallel with the actual grid while using both real data and estimations. Analysis showed the transactive signals would have correctly advised smart grid equipment to alter their operations during two critical moments on the actual regional grid. These were an unexpected outage at a nuclear power plant in Washington state on Feb. 5, 2014, when the plant dropped to less than half its normal generating capacity, and a sudden increase in winds on Feb. 15, 2014, which peaked wind power generation at 2,884MW.
To evaluate the potential impact of transactive control beyond the project, IBM created a model that rapidly simulated different scenarios on the regional grid. Tests run on that model showed the Northwest's peak power demand could be reduced about 7.8% if 30% of the regional electric grid used transactive, demand responsive equipment. The modeling also showed transactive energy approaches can lower the Northwest's overall power costs by taking advantage of wind energy when it's abundant and inexpensive.
Lessons learned in the demonstration
The key lessons learned are:
- Better tools are needed to ensure smart grid data is of high quality and the equipment generating that data is working correctly. Many participants were not prepared to deal with the onslaught of data and sometimes mislabeled data with incorrect units or times.
- Smart grid technologies should be designed to work together and smart grid technology standards should be further developed. This would have reduced the efforts required of project participants to make equipment from various vendors work together.
- Smart grid technologies and their markets need to mature and stabilize for smart grid deployment to succeed. Some manufacturers went out of business or stopped servicing their products, while some equipment simply failed.
- Public involvement is the key to a successful smart grid deployment, though there is no one-size-fits-all approach.
Regional and national smart grid efforts are ongoing and several of the utilities are continuing smart grid programmes on their own. For example, Idaho Falls Power plans to implement conservation voltage reduction throughout its entire power distribution system. Flathead Electric will install 1,000 residential and small business water heater load controls annually during the next five years. Avista Utilities will install voltage controls and fault detection, isolation and restoration technologies throughout its service territory and start installing smart meters for its Washington customers in 2016.
The project's main participants included Avista Utilities, Benton PUD, City of Ellensburg, Peninsula Light Company and University of Washington in Washington, Flathead Electric Cooperative and NorthWestern Energy in Montana, Idaho Falls Power in Idaho, Lower Valley Energy in Wyoming, and City of Milton-Freewater and Portland General Electric in Oregon. The technology infrastructure partners were Alstom Grid, IBM, Netezza (now part of IBM), QualityLogic, Spirae and Vaisala (formerly 3TIER).