IE 11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

Boston Uses Big Data to Make Electric Grid Predictable, May Apply Model in Other Cities

By predicting how much power each building in Boston will need during any given hour, the city hopes to transform its electrical grid.

Without data, America’s energy infrastructure is a tense minute-to-minute battle to provide as much electricity as people need at any given moment, where a slip-up could mean a blackout. With data, it can become an efficient machine that predicts usage, manipulates demand and generates electricity from local, renewable sources.

The Massachusetts Institute of Technology (MIT) believes it’s found a way to make that data reachable.

It takes an enormous amount of data to paint that futuristic picture. Even in a medium-sized city like Boston — roughly half the size of Philadelphia — there are nearly 100,000 buildings. And depending on local laws and regulations in any jurisdiction across the U.S., the whole-building or tenant-specific energy usage data for each of those buildings might be hidden from public view.

So Christoph Reinhart took a shortcut. Rather than pulling untold fathoms of rough data on every building in Boston, the MIT researcher gathered enough data to model energy use in those buildings — for every hour of every day of any given year. By identifying a number of building profiles and then calibrating them with actual usage data, Reinhart and his team achieved 94 percent accuracy when predicting how much power the grid would need at any given time.

That’s huge, Reinhart said. It flips the tense, minute-to-minute battle of supplying power to the grid on its head.

“You can start seeing the future and predicting trends, and then you can influence the future,” he said.

Over time, that could save a lot of money for energy companies, customers and the city. That’s because reducing the need for energy means reducing the need to pour money into new infrastructure such as substations, transmission lines and power plants.

Though the specifics will change depending on how municipal officials decide to prioritize its energy goals, the Boston Community Energy Study that accompanied MIT’s research estimates that Boston could save between $600 million and $1.7 billion during the next 25 years while transforming its power grid.

All that while providing business opportunities for local companies and reducing dependence on fuels that Massachusetts buys from out of state.

“If you’re talking about putting insulation into a building, or replacing the lighting, or changing out the heating and cooling systems, that’s work that has to be done in that building, and it’s going to be typically done by people living in the town or near the town,” said Cliff Majersik, executive director of the Institute of Market Transformation in Washington, D.C.

On top of that, it could also make a significant dent in its carbon emissions. Boston has a goal of reducing greenhouse gas emissions 25 percent by 2020 and 80 percent by 2050. The MIT study suggests that the city could, through the methods outlined in the report alone, cut its emissions 30 percent.

And that’s just Boston. Should other cities begin using models like the ones developed at MIT, Majersik said, they would have their own priorities. They might come up with different ways to use them too.

“It just provides a whole new vision," he said, "and we’ve only scratched the surface in this whole new wealth of ways we can use this data."

Portrait of a new system

The whole energy system — from the places where the fuel comes out of the ground, to the plants that turn the fuel into energy, to the wires and substations that deliver that energy into buildings — is built for one day.

That day is the hottest day of the summer, when everybody turns on their air conditioning. It’s the day when any given city will use more electricity than any other time in the year. All the infrastructure that puts power on the grid needs to be able to meet that demand — even if it doesn’t have to meet that level of demand for the rest of the year.

“That’s kind of crazy, to invest billions of dollars in infrastructure that will be used so [rarely],” Majersik said.

But the grid doesn’t have to be built that way. Majersik, an advocate for public benchmarking energy usage in buildings, sees a different picture emerging.

With highly granular information like the models MIT describes in the study, cities can build a system that handles hot days much better — more predictably. If one knows not only what the peak energy demand is, but which buildings are driving that demand, they could target those buildings to push down the peak.

So in the future, the grid won’t need to build as many new power plants, wires and substations to supply energy on the hottest day of the year.

“It’s a lot cheaper to have people do this at MIT with data than it is to pay people to use cranes to put in [infrastructure],” Majersik said.

There are many different ways the city could encourage its residents to get on board with energy-saving efforts, and Boston hasn’t yet settled on any methods in particular. City Chief of Environment, Energy and Open Space Austin Blackmon said that’s because Boston is working to connect the findings of the study to the other aspects of a city-wide planning initiative.

For example, one option outlined in the study is community energy districts. That involves putting pipes underground that can share steam between buildings, effectively helping buildings heat each other. Laying those pipes means digging up road, so it would be much better for the city to establish those districts while it’s doing excavation for other projects. That means connecting the energy plan with the transportation plan.

There are similar connections when it comes to public health, emergency management, affordable housing and city planning (see sidebar).

Energy Plan Connections

In revamping Boston's energy infrastructure, many other city systems would be affected, like transportation, public health and emergency management, to name a few. Here's a look at how.

Transportation: Part of Boston's plan is to share heat and energy between nearby buildings. That means laying down new pipes underground, which would become more cost-effective if done when the city is excavating those streets for other reasons.

Public health: Greenhouse gases can worsen existing breathing problems and have been linked to thousands of deaths worldwide each year. Reducing Boston’s need for energy means burning fewer fossil fuels, which puts fewer pollutants into the air.

Emergency management: Some of the suggestions included in the Boston Community Energy Study, most notably “microgrids,” could help buildings keep the power running even during blackouts caused by major storms such as Sandy. The plan calls on targeting those efforts in the places that most need power at all times, such as hospitals.

Affordable housing: Because energy can be a significant cost burden on low-income residents, the study suggests building “energy justice” projects that specifically help to make affordable housing more energy efficient. Affordable housing is also more likely to shelter people who would be vulnerable during a power outage, such as the elderly and those with health problems, so microgrids would be important in those areas for keeping the power running during a storm.

City planning: By connecting the energy study with a broader city-wide planning effort, considerations about energy use could be built into decisions about what public projects Boston should prioritize. Requirements for developers to keep energy efficiency in mind, or even to provide hookups for heat and power-sharing, could be built into zoning rules.

Getting to the future

When all is said and done in the planning process, there are several tools Boston could use to accomplish its energy goals. The building-level data available through MIT’s model could mean targeted incentives. The city could craft programs specifically designed to get building owners in certain parts of town to, for example, improve insulation to reduce the need for heating and air conditioning.

Usually, incentives are more first-come-first-serve than tailor-made for the places they’d be most useful.

“[This] is a very new idea that is very rarely happening,” Majersik said.

Alternatively, the city could target efforts to install batteries that could smooth out peak demand by drawing in grid power when it’s cheap and supplying power when the grid is expensive. The city could also work to establish heat-sharing networks between buildings, create “microgrids” and share solar-generated electricity within neighborhoods.

“We’re moving away from a centralized system to one that’s much more distributed,” Blackmon said.

Another potent tool for the city, Blackmon said, is zoning. As developers plan out new projects, the city could require them to establish the sorts of connections that would make heat- and power-sharing easy.

That could be a particularly powerful mechanism in the near future, he said, because Boston is going through major expansion at the moment. In fact, he added that there have only been two periods in the city’s history when it was expanding faster than it is now.

“In terms of the development that’s in the pipeline and the impact that [those projects] have, having these ongoing planning efforts — and the community energy study is one of those — it’s a really excellent time [to encourage energy innovations],” he said.

What’s next

The level of detail in MIT’s models is unprecedented, Reinhart said. Even California’s new building energy benchmarking law doesn’t provide hour-by-hour statistics for every single building.

But as cutting-edge as MIT’s work is, there is still room for growth. For instance, it doesn’t yet take into account factors such as hyperlocal weather and urban heat island effects.

And it is, by its very nature, limited. The building profiles Reinhart built were based on the architecture and consumption patterns in Boston. So their usefulness, for now, is regional.

“If we go from Boston, say, to Providence, we can actually say that there is a similar building stock as in Boston, similar patterns, so we can largely use the same … library,” Reinhart said.

All those limitations can be overcome with time and work, Reinhart said. MIT has models that could be used to build in urban heat island effects, and he definitely wants to start looking for partners in other cities to help put the simulations into use outside Boston. He’s already done so abroad, in places like Riyadh, Saudi Arabia, and Lisbon, Portugal. In the U.S., he said some possible places to apply the models might be on the West Coast.

That’s partially because many West Coast cities have carbon reduction goals that make the tool more immediately useful, and also because those places are beginning to build some robust energy benchmarking programs. Last year, California became the first state in the nation to pass a law that requires multi-tenant buildings to disclose their whole-building energy usage statistics — a step Reinhart said would make it much easier to tweak Boston’s model to fit a place like Los Angeles or San Francisco. Seattle has its own benchmarking program, and Washington state has toyed with the idea of passing a law similar to California’s.

Wherever the MIT team decides to go next, Reinhart said he will be looking for a few things in particular. One is places where the grid has trouble meeting demand. Another is local buy-in.

“For me, a [criterion] of where I want to go is how much engagement we get from local cities,” he said, adding that eventually he wants to see similar practices in place across the country. “I envision that these model should just become part of the data that cities have."

Ben Miller is the associate editor of data and business for Government Technology. His reporting experience includes breaking news, business, community features and technical subjects. He holds a Bachelor’s degree in journalism from the Reynolds School of Journalism at the University of Nevada, Reno, and lives in Sacramento, Calif.