Millions of people already sell electricity from solar panels back into the grid. New battery technology makes it increasingly practical for neighbors to store and share electricity. Energy planners now talk about “citizen utilities” that would generate substantial power to complement the output of traditional utilities.
To facilitate this shift from traditional, centralized power plants to distributed local networks, a trio of Stanford researchers has developed ReMatch, a software tool to analyze the actual patterns of electrical demand and the potential generating capacity of every home in a neighborhood, and then design a cost-effective smart grid for that locale.
ReMatch was developed by Rishee Jain and Ram Rajagopal, assistant professors of civil and environmental engineering, and Junjie Qin, a graduate student at Stanford’s Institute for Computational and Mathematical Engineering.
The software analyzes exactly when each customer in a given market area is using electricity, and then allows planners to design a local infrastructure that matches the right mix of consumers with the right configuration of energy sources. In a paper published by the journal Nature Energy, the researchers tested their methodology by estimating the expected costs and savings of new infrastructure for 10,000 specific homes clustered around the San Francisco Bay Area.
What they found was that the mix of new energy sources suggested by the ReMatch model could reduce the total cost of electricity by about 50% over the next 20 years. The new equipment would cost about $58 million, but it would save about $227 million in operational costs tied to purchases of fossil fuel. That works out to between 10 cents and 13 cents per kilowatt-hour with the new equipment, versus about 23 cents under the existing system.
The researchers say their estimates are conservative. The upfront costs were based on actual current costs of equipment and installation, compounded by a range of interest rates from 3% to 9%. If interest rates soared to 9% for the entire 20 years, which would effectively make the new infrastructure much more expensive, the savings would still be above 40%.
“The most important aspect of this model is that it shows the necessity of having a very granular understanding of consumer demand,” says Rajagopal. “When you start with the observation that different consumers have different patterns, you can generate far more scenarios about how much power you need and where to get it from.”
In an editorial, the journal Nature Energy said the Stanford team had broken important new ground in planning for a new era in power generation.
“The ability to match consumers and infrastructures is the crucial feature of the ReMatch framework,” wrote Angelo Facchini, a professor at the IMT School for Advanced Studies in Lucca, Italy. “Jain and colleagues’ work sheds light on a fascinating perspective for the future of the electricity distribution system.”
Some utilities are already encouraging “distributed electricity resources,” such as solar arrays and battery storage, because they can reduce the strains of peak power demand on traditional power plants.
The big challenge for such efforts, and the key to ReMatch, is to efficiently match up all those potential power sources with the complexities of actual consumer demand.
Solar panels are wonderful at producing cheap electricity on cloudless days when the sun is high, but many residential customers use most of their electricity in the early morning or at night. Other customers may use a lot of electricity during the day but don’t necessarily have enough sunny space for solar panels.
“As a hypothetical example, my wife and I might have room for five solar panels on our roof but we aren’t usually at home when the sun is shining,” says Jain. “My neighbor, on the other hand, might work at home throughout the day but only has space for two solar panels. Wouldn’t it be effective if we could share solar panels, and perhaps share some battery storage, so we could all get the best value out of the equipment? Taking it a step further, we could put together 15 or 20 neighbors and form a ‘virtual utility’ to share the burden.”
By analyzing hourly usage data from smart meters, the researchers identified several basic types of electricity consumers. Some were classic “dual peak” users – people who leave the house for 9-to-5 jobs and use most of their electricity in the earning morning and at night. Others hit peak usage in the middle of the day, which might be typical of people who work at home and use a lot of electronic equipment. Yet another group consists of the night-owls, who use a lot of electricity at night but very little in the morning or afternoon.
“It’s important to note that we aren’t making a judgment about which type of consumer is preferable,” notes Jain. “The reality is that you need every type.”
The new modeling system also analyzes the most cost-effective mix of energy sources, including electricity from traditional power plants, and it can suggest some surprising strategies. Battery storage is a very expensive way to get electricity, for example, but it can be a good investment if it allows sharing between day-people and night-owls, or between homes that can and cannot have solar panels.
To be sure, there will always be uncertainties about future energy prices. Higher oil and gas prices would increase the cost advantage of renewables, and a plunge in prices could have the opposite effect. So the Stanford team built in opportunities to model the impact of alternate scenarios – what Rajagopal calls “segmenting the future.”
“We are not saying that the traditional electricity grid will go away,” Rajagopal says. “This is a system that works in tandem with the existing grid, and it may be that it makes sense for utilities to invest in some of these systems on their own. That’s the kind of thinking we want to open up.”