A fascinating exchange inside the Energy Changemakers community between John A. “Skip” Laitner, William Prindle, and Lorenzo Kristov begins with an explanation of why virtual power plants typically act as a demand response resource and not as supply injected onto the grid—which would significantly increase their value but faces technical challenges. The conversation culminates with Kristov describing why we need to plan energy at the community level, how incumbent utilities prevent that, and what is likely to be the most significant benefit of distributed energy resources.
You can also listen to the 12-minute excerpt. This is part of a longer video that members can access here.
Laitner
I began working with the idea of virtual power plants, as I’ve shared with a couple of you, back in the nineties when I was with the EPA. And Bill, ironically, I was working with ICF on an Integrated Planning Model. We began talking about the idea of virtual power plants in the nineties.
I’ve got an idea of what virtual power plants are from a megawatts and kilowatt hours equivalent. But as you guys have referenced virtual power plants — is there a policy definition or current code definition? I have kind of lost track of that, so I’d love to get a sense of that.
Prindle
Good question. If there were a definition, it would not be applicable across all 50 states. I’m virtually certain of that. Well, California’s gone further than most states and actually deploying pilots and so on.
Kristov
Let me just add a little bit on that, if I may. Virtual power plant, just as a broad concept, basically involves aggregating multiple resources over multiple locations and coordinating their behavior so that they perform in a certain way to provide something that the system needs. Now in FERC jargon, FERC passed order 2222, a couple of years back, which was to enable what they call DER aggregations to participate in the wholesale markets of ISOs and RTOs. Basically, a DER aggregation is the same thing as a virtual power plant. I don’t know why they need a new name for it, but it’s the same thing.
Although in principle, a virtual power plant doesn’t necessarily need to be on the distribution system. You could have grid transmission connected resources whose behavior is coordinated in a way to have them collectively produce a certain result. But the idea of the FERC order was we can do this on distribution side and we can use these resources and have them participate in the wholesale market.
Now, the stumbling block there…I worked for California ISO, and in 2016 we actually created the capability to do DER aggregations and have them participate in the wholesale market. As of today, nobody is using that capability. So it’s been around for a lot of years, in many ways was the model for what FERC did in 2222, but it’s not being used.
Why? Well, because there’s a really crucial distinction between placing distributed resources behind the meter where they’re only modifying load versus having resources that are actually injecting power into the system.
Prindle
Exactly.
Kristov
And the injection model is what creates complications, costs, uncertainties — mostly regulatory related — that make it difficult to stand up a VPP that’s actually going to inject power.
So what we saw in California is the developers of VPPs or DER aggregations would take what is the ready-made and simpler pathway, and that is just work with devices, mostly batteries or coordinated batteries and solar behind the meter. All they’re doing is modifying the load. They’re not injecting power except perhaps inadvertently at times, but they’re not injecting power to sell into the wholesale market…It’s an aggregated demand response model. So VPPs in practice are mostly aggregated demand response because of the challenges.
Where we see for 2222 to work, it really requires the collaboration of the distribution company, the electric distribution company, because obviously the power is flowing over distribution wires. And once you get into injecting power, the engineering studies become more complicated. The interconnection process becomes more complicated and more costly. So sorry for going on so much about this, but I thought there may be some things here that not everybody knows and understands. So I think the practical implication for VPPs still needs to break this hurdle of how do we create an injecting resource that’s compatible with the distribution utility that doesn’t create uncertainty or high costs around interconnection and around operation.
Laitner
That’s a really good clarification, and I just want to ask one further question because you’re talking about aggregated demand response, and yet VPPs are a form of supply trying to meet that demand. Is there a connection between those two, the supply side and the demand side?
Kristov
Well, there is in the sense that if you can drop demand in response to a dispatch signal, then it actually works — has the impact of adding a supply resource. So in that sense, it’s equivalent. And many years ago, FERC in orders 745 and 719, they said that aggregated demand response, if it’s dispatched by the wholesale market, it should get paid the same price as a generator. It should get paid based on the locational marginal prices if it’s a generator. So in that sense, aggregated demand response can be viewed as a generator from the wholesale market perspective, but it still is not injecting power into the system.
And that to me is ultimately where a lot more of the value of VPPs is going to be realized — when they can actually be injecting power. And it goes back to this question about shared resources and the explanation about the Hunter’s Point where the houses were producing more power than they need and it was just going into the grid to be compensated at probably some low rate by the utility.
Where we want to go, I believe, with local energy is plan at the community level to have shared resources to oversize solar panels wherever possible, especially on large facilities like community, community centers, warehouses, and those things. And then be able to use that excess to supply the neighborhood so that especially in retrofit neighborhoods where you have tree canopy, you don’t want to cut down trees in order to install solar panels, but you want to take maximum advantage.
So this is where you have a capability of sharing resources that’s much more viable if you’re under a municipal utility or a rural electric co-op and not regulated by a public utilities commission whose job it is to protect the utility monopoly. That’s essentially the conflict that you run into. And it’s that utility franchise agreement that stands in the way of being able to share resources.
California has got a proceeding right now on community solar. We’ll see where they go with that. There’s a good proposal in front of them, but the utilities want to basically continue business as usual, which has not been very productive in terms of shared community projects. So where I see we want to go is basically integrate energy planning with urban planning and city planning and start building facilities locally that can supply energy to their neighborhoods that can supply energy to local customers. But that’s going to require statutory changes.
Laitner
Lorenzo, you’ve given me really helpful feedback. Thank you. This is really helpful. But damn, you, I’ve got more to think through and puzzle now.
Prindle
There are analytical challenges too because when you can treat everything as a load decrement, be it demand response, energy efficiency, or even PV…[using] a model like IPM [Integrated Planning Model] that I know well, you just change the load curve and then you just build the system at an aggregate level to that new load curve. But when you’re injecting, you’re changing the dynamics of the distribution system. You have to have voltage and frequency and other characteristics that are maintained down at the feeder and circuit level. So, eh, eh, it becomes complicated from an analytical point of view. You can in theory, value the locational and temporal benefits a lot better, but man, it takes a ton of data and a ton of analytics.
ICFs done some of that for some utilities including PG&E, but it gets really granular and expensive really quickly. And there’s some challenges there that as I’m seeing utilities asked to put forward grid modernization plans, which are really distribution utility plans, not vertically integrated utility resource plans. It’s a whole other beast. And few to date are really doing the really granular analytics to say, well, for this substation, that feeder, we could probably accommodate this many kWs or megawatts of distributed resources as long as they have these particular characteristics. And Lorenzo referred to the engineering studies, like the utility needs to know everything about that power source.
Laitner
Bill, you just reminded me of Stephen Colbert when he goes “eh, eh.” What you just added, actually, this is really a critical point, building on Lorenzo’s comments as well. We’ve definitely got to think well beyond business as usual, and that’s where we’re stuck, I think, in a lot of ways.
Prindle
Well, you mentioned Steven Colbert. I have to recall the incident where Jim Rogers, when he was chairman of Duke, went on Colbert. They got to talking about why was Duke energy promoting energy efficiency? And of course Colbert holds up a compact fluorescent light bulb and says, well, don’t these have mercury in ’em? And Roger says, well, you have to smash ’em on the floor and lick ’em to get ill health effects. And Colbert says, well, if they didn’t want us to lick them, why did they make them look like soft serve ice cream?
Kristov
No, but all the technical issues that Bill is raising are of course valid. But on the other side of this, to me, the bigger picture is that for 100 years we’ve have a system that’s planned from the top down. It’s not really connected to the end uses of energy or the local needs or what’s happening locally on the ground. We have this demand side and we measure that and then we throw all the rest of that demand into a capacity expansion model. So we can build big stuff. And I think the technical potentials of distributed resources need to overturn that planning paradigm and make more bottom up so that after you look at what you put on the individual buildings and houses, now you look at the surrounding community and say, how can we maximize local production to meet the needs of the community? So the bulk power system, when you get up to that level, becomes the residual supply and not the primary supply.
And I think that’s the potential when I actually think about playing out 5, 10, 15 years into the future, the value of distributed resources is really going to be avoiding building big infrastructure. And I think that’s a lot of where the battle lines are right now because there are so many incumbents that make money building big infrastructure that they don’t want to see proliferate, something that’s going to ultimately substitute for big infrastructure.
Prindle
Bingo.
Laitner
Agreed.
Are you an energy changemaker? Join the community!
