by Jeff St. John
August 02, 2019

In the past month, Pacific Northwest utility Portland General Electric has unveiled two major plans for cutting its carbon emissions by 20 percent below 1990 levels by 2025, and 80 percent by 2050 — one from the bulk grid resource level on down, and the other building from the individual customer on up. 

The grand-scale vision comes from PGE’s recently filed integrated resource plan (IRP), its roadmap for securing the generation and demand-side resources it will need through 2025. As the first IRP to include PGE’s decarbonization goals, it includes retiring its last coal-fired power plant by 2020, and adding significant amounts of wind, solar and grid-scale energy storage.

PGE is already pursuing this combination via its Wheatridge project with NextEra, with 300 megawatts of wind, 50 megawatts of solar and 30 megawatts of batteries set to come online in 2021. 

But the IRP also foresees a sizable portion of PGE's future resource needs — about 200 megawatts by 2025 — being filled by “distributed flexibility.” This new term encompasses the range of energy shifting, shaping and generation that can come from well-integrated “technologies and energy behaviors of our customers” both residential and business, “to provide the same services and value that power plants and grid investments provide.” 

Distributed flexibility doesn’t fit into the traditional categories found in IRPs, such as energy efficiency, demand response, distributed generation or dispatchable energy storage. And while it’s partly under utility control and direction, much of its value to the grid will rely on individual customer decisions on how to invest and behave around their energy usage — or, perhaps, how energy services providers and aggregators can help them do so. 

Utilities around the world are struggling with how to integrate the behind-the-meter energy capability and potential created by distributed energy resources. In the U.S., states like California and New York are taking a lead in developing regulations that ask utilities to include customer-owned DERs as part and parcel of their distribution grid investments and long-range planning needs. 

But for PGE, the linkage between customer-based distributed flexibility and its grand-scale clean energy goals is closer than for most investor-owned utilities. With about 887,000 customers, PGE is Oregon’s largest utility, but relatively small compared to its U.S. investor-owned brethren. And a large portion of its customers are residential or small and medium-sized businesses, which have traditionally been harder for utilities to tap for energy services than larger commercial and industrial customers, like those PGE has enrolled in existing demand response programs.

Seeding neighborhoods for distributed flexibility 

That brings us to PGE’s second announcement last month, the launch of its Smart Grid Test Bed pilot. Over the next five years, PGE will work with three Oregon cities with a total of 20,000 customers to provide them with the technologies and utility signals required to make money by shifting energy consumption at times the grid needs it most. 

This will include traditional demand response, such as mass-market residential smart thermostats to ease summer peaks, and utility-controllable electric water heaters to shift winter peaks. But it also includes tapping new resources, like the smart EV chargers expected to make up an increasing share of PGE’s load, or the backup power and battery storage systems increasingly accompanying rooftop solar installations. 

However, “not all distributed energy resources are created equal,” Andy Macklin, director of grid products and integration, said in an interview. Utilities like to measure grid resources in terms of the megawatt-hours of energy they provide, the megawatts of capacity they can be relied on to cover the grid’s most-stressed conditions, and the flexibility they have to shape and shift those qualities within different time scales.  

PGE proposes that eventually, it will be able to orchestrate the combination of customer-sited resources to serve as a virtual power plant. That’s an oft-used, but broadly defined, term for aggregating and controlling DERs from industrial combined-heat-and-power systems to household HVAC units in ways that can mimic the responsiveness of utility-dispatched generators — or, more accurately, mimic the key discrete services that those generators provide. This can range from predictable daily load reductions from smart thermostats and peak-time rebate signals, to instantly dispatchable batteries to provide frequency regulation and voltage support. 

“If we assume kind of a broad diversity of resources, spread across our service territory, that all have certain defined characteristics — up, down, duration, frequency of ability to call — the virtual power plant brings those together," Macklin said. "It relies on coordinating technology and real-time understanding of the system.”

But once that’s in place, “it should be able to essentially dispatch that portfolio.”

That’s the end goal, at least.

“The test bed is meant to bring that future state forward, to work directly with customers, to demonstrate how the future smart grid can work at relative scale, and on up — and how we can use that to help integrate those intermittent resources,” he said.  

From peak-time rebates to virtual power plant

The three communities involved in the project, Hillsboro, Milwaukie and the north Portland neighborhood of Overlook, have been closely involved in the final shape of the program, Macklin said. “The neighborhoods were chosen very purposefully to get broad representation of our customer base — geographical diversity, customer diversity, system diversity.” 

The rollout to these customers started last week, in the form of automatically enrolling them all in PGE’s Peak Time Rebate program, which pays customers for reducing electricity use on peak days, unless they choose to opt out of the program. This use of an “opt-out” approach to enrolling customers, rather than requiring them to opt into a new program, is critical for PGE to reach its goal of hitting its target participation rate of two-thirds of all the customers involved — about 10 times the national average for similar utility programs. 

"We picked [Peak Time Rebates] as the first tool, because it’s the easiest tool for customers to engage,” Macklin said. “There’s no risk of downside for a customer, which makes it a good candidate for an opt-out program,” as compared to, say, programs that might penalize customers for failing to reduce energy use, or expose them to price spikes during peak hours.  

But the Peak Time Rebate program is just the first tool.

PGE’s customers are being shifted to time-of-use rates starting next spring, which will provide further price signals for behavioral or automated energy shifting. “As we move through the test bed project over the next few years, we’ll look for more ways to get customers engaged in their energy use,” Macklin noted. 

As for incentives for customers buying the technologies involved, PGE already offers rebates for smart thermostats and water heaters, and has received state regulatory approval to direct funding for a battery storage pilot program into the test bed. PGE’s IRP notes that it is planning to have about 4 megawatts of customer-sited, utility-controlled energy storage in place by 2025. 

On the utility side, the three test bed communities are first on PGE’s list to undergo a broad-ranging distribution grid technology upgrade it has underway, Macklin said. That includes a new field-area network, new distribution automation and substation automation, and an advanced distribution management system to coordinate grid operations and controls. 

This increased visibility and control over its grid will allow PGE to monitor how the changing energy consumption patterns and growing share of DERs in its three test beds are affecting day-to-day distribution operations, he noted. Utilities in states like Hawaii, California and Arizona have started to face challenges from the growing amount of energy being generated at the ends of certain solar-rich distribution circuits, and many of the earliest utility efforts to control DERs have been aimed at mitigating those problems. 

But PGE is also looking at how grid operators could tap the responsive DERs in the test beds as an active tool in grid operations.

“Today we use our demand response programs during peak events to help shift load” at the systemwide level, Macklin said. “Our future vision is that we would be able to recognize locational value as well as system value — that if there’s a situation for a particular feeder or substation, that we could leverage distributed energy resources to help.” 

Future uses in distribution operations, DER integration

Oregon hasn’t yet instituted a regulatory framework that would allow PGE to link customer DERs with utility distribution grid controls, but one should be on its way by the end of the test bed trial period in 2023, he said.

In March, the Oregon Public Utility Commission opened a docket on distribution system planning, its version of the distribution resource planning proceedings underway in California, Nevada and other states to integrate DERs into how utilities invest in their distribution grids. PGE expects to file its first distribution system plan with the commission late next year. 

PGE is also exploring multiple options for aggregating and managing the array of customers and devices that will emerge in the three test-bed communities.

“We believe our key strength is in being the smart grid platform that links distributed energy resources to the clean, intermittent resources in our fuel supply,” Macklin said.

For now, that means effectively communicating signals such as peak-time rebates and time-of-use pricing peaks to both the customers willing to turn down energy use manually and the devices preset to respond to these energy-saving opportunities. 

But more sophisticated and flexible DERs, like the customer-sited, utility-managed batteries included in the pilot, require more sophisticated and flexible control solutions. “We recognize that others may be better situated to, for example, aggregate all the different energy devices and [internet-of-things] devices that may exist,” he said.

The rising popularity and penetration of home automation platforms from Google and Amazon, for example, have turned them into key partners for utilities seeking to reach a critical mass of automated residential customers, as Wood Mackenzie Power & Renewables has noted.  

“We are absolutely working with other partners to see how we can bring some of that automation to deployment,” Macklin added, though he wouldn’t name the companies involved. “Some of these agreements haven’t been announced yet, but we’re working with other partners.” In turn, he said, “we hope to be able to deliver that two-way value back to customers.” 

The smart grid test bed, like the IRP, is a long-range plan, with a total evaluation period of five years. But most of the marketing and deployment will take place in the first three years, to ensure the high level of participation PGE is seeking to prove out its more ambitious goals for the project. 

“For us, probably one of the biggest things is learning how best to partner with our customers,” he said. “Bottom line in the test bed is, how do we do that effectively, so that they want to participate and they see the value.”