All Episodes

[Episode #20] – Grid evolution

0:00
-1:24:09

Utilities face a host of rapid changes in a what used to be a staid business: new business models, changing supply and demand forecasts, new distributed architectures, new types of resources, new participants in the power grid that they don't control…yet they still must maintain a highly reliable power grid that operates within fairly narrow parameters.

Meanwhile, difficult questions remain to be solved, about how we’re going to manage our grid power transition, who the winners and losers will be, what destination we’re headed for, what role consumers and “prosumers” will play in the future, and what our reasons are for executing transition the way we do.

We tackle all of these issues in this wide-ranging, very geeky conversation about the “blocks and squiggles” of the grid of the future. Grid power transition, the rebound effect, energy efficiency, utility business models, cutting-edge grid power management considerations, regulation and rate design, electric vehicles as distributed energy resources… they’re all here.

Geek rating: 10

Guest: Eric Gimon is an active researcher and policy adviser on the power sector transformation to a clean, reliable and affordable low-carbon future. His career path has spanned 15 years of researching quantum gravity and high energy physics at some of the world’s top research institutions, to work at Lawrence Berkeley National Laboratory and UC Berkeley, to an AAAS fellowship with the Department of Energy, and finally to a personal transition to climate and energy policy. Eric is currently a Senior Fellow with Energy Innovation: an energy and environmental consulting NGO. His interests and writing cover everything from residential energy management systems to large grids and wholesale electricity markets.

On Twitter: : @EricGimon

On the Web: Eric’s writing on power sector transformation at Energy Innovation

Recording date: May 27, 2016

Air date: June 29, 2016

Chris Nelder: So let's bring him into the conversation. Welcome, Eric, to the Energy Transition Show.

Eric Gimon: Thank you Chris.

Chris Nelder: So you've had a pretty interesting career path which has taken you from researching quantum gravity and high energy physics to a focused study of energy policy. What motivated you to move on to energy transition?

Eric Gimon: Well Chris, it was around 2005 I read a book a very interesting book that a friend of mine loaned me called Thin Ice written by a journalist named Mark Bowen about people who went to these high altitude glaciers in Ecuador and Tibet and places like that and analyzed climate change through these ice core data sets and he knew I was really interested in mountaineering and things like that. And the book is quite interesting because it like for example the Ecuadorian volcanoes give you information about climate conditions in that region and then you can kind of connect that to the archaeological stuff. So I found the book fascinating but as the book went along he kind of gave this message that the scientists were freaking out about climate change. And that if you kind of got them over a beer in a bar or whatever they really actually told you even more disturbing things. So that kind of flagged the issue for me and then not long after that I was in a seminar that John Holdren gave, he's I think currently the president's science adviser. Back then he was at Wood's Hole. And you know I've been aware of climate change back in the 80s and stuff but kind of loosely as this like sea level rise thing. He just put it all in context. I mean one the sea level rise thing was really real. You know showing maps of Florida going under over time. And then some of the broader stuff and putting it together with the book kind of really got me you know activated on the issue. And I'd been doing physics for a while and I was thinking of doing something else and I talked to an old physics teacher of mine when I was an undergrad at Stanford who said you know we need more people in climate that have kind of an analytic and hard sciences background. So that kind of drew me into this space and I was really attracted to the grid and renewable energy. One because I had solar panels on my roof and two because it seemed to be such a proactive thing. You know instead of like feeling bad about what's happening, this is something that works and we can make it work better. And so I gradually shifted my attention from physics to energy and energy policy and started you know where I didn't really understand where power came from in my plug to really getting more of an understanding of the whole system. And then I took a fellowship at the Department of Energy for people who have PhDs to work as advisors in the government and that was the beginning of a big long journey that's been fascinating. I really enjoy the topic and it's nice to be working on something that matters.

Chris Nelder: I remember you told me something about how Steven Chu actually persuaded you to push on into policy.

Eric Gimon: Yeah. So he was my freshman physics professor.

Chris Nelder: Oh wow.

Eric Gimon: It was kind of funny we had this kind of crossing ships thing for a while. I was at Berkeley in physics and then he became the director of the lab. And then I stopped doing physics and was kind of moving into the policy stuff and then he went to D.C. and I went to D.C. a few months later. We talked a few times when I was there in D.C. We didn't overlap too much but he was the one that kind of convinced me that having more analytic people in the space would be very helpful. And I found that that's the case. Though people don't always quite know what to make of you.

Chris Nelder: So let's dig into the fun technical stuff. So we're both interested in figuring out what a low carbon high distributed grid power system might look like. So let's start with that. I guess the obvious first question is how much can we say at this point about what the composition of such a grid power system might look like say 50 years from now? And if grid power at that point is mostly coming from renewables then how do we deal with its variability? I mean a lot of the analysts assert that renewables will have to be backed up 100 percent by either dispatch or conventional generators or battery storage systems. How do you unravel this puzzle?

Eric Gimon: Oh that's a great question. The first thing is nobody quite knows where this is headed. I mean if you get 10 experts in a room you'll get 15 versions of what this grid could look like.

Chris Nelder: Yeah.

Eric Gimon: But if you start from the basics, flexibility is very important in a power system. That was already true before we went on this energy transition of course. You don't always know what people are going to consume and when and so you need resources that can be on standby to kind of adjust so that's the regulation or reserves. And also power plants go down. When I was at the DOE there was an unforced shutdown of a power plant down near Florida, a nuclear power plant. And the people there told me if it hadn't happened on a weekend if it happened during the week they probably would have taken out a big chunk of the southeast part of the Eastern interconnection. So these are always issues that were there. The problem when you have renewables is there are more variables, so you have more variability to deal with. I mean there are benefits to renewables, one shouldn't forget. For example they tend to come in smaller chunks. You don't lose two gigawatts worth of wind turbines at at time or solar panels and so on. But it's definitely more variation and so that's what get people to say things like well the sun doesn't always shine or the wind doesn't always blow. But the truth is that doesn't mean you need one to one backup because the wind may not blow here but it may blow over there. So if you have a large grid you can kind of average over things. And if you have a diversity of resources you can use the wind at night, and the sun during the day and so on. So it's never a one to one thing just like it wasn't for... You know we don't have a backup nuclear plant sitting around in case the other nuclear plant goes down. That is not a very economic way to run a grid. But you do need this flexibility. And so one way we think about the flexibility is the flexibility supply curve. This is a concept that was introduced by a great paper from maybe 2006-2007 from the NREL folks on energy storage. One of the things I learned from that paper is that the stories we do have, most of it is pumped hydro storage, was built in order to accommodate nuclear power. Because nuclear power was producing power at night that nobody really needed and so they had to store that power so that it was available during the day when people needed it more.

Chris Nelder: Wow. I've never heard that before.

Eric Gimon: Yeah there's a nice description of that at the beginning of the paper. Maybe you can link to the paper.

Chris Nelder: Yeah.

Eric Gimon: But there was this diagram in there and it starts on one axis you have kind of price and the other axis is a mount or something I can't quite be sure about it. But there's a lot of latent flexibility in a power system. Just... You haven't organized the rules or you haven't connected things or you know you have different ownership structures and so on. And so you're not taking advantage of that flexibility. An example of that is there was a study done called the Low Carbon Grid study to examine what it would mean for California to go to 50 percent renewable power. And it came on the heels of another study by a consulting group called E-3 for the big IOUs in California. But he came to very different conclusions. And one of the metrics that you look at a high penetration system is the amount of curtailed power. I mean there are different metrics you can use like likelihood that you'll have a blackout. Things like that. But one way that some of these studies work is to look at how much electricity you have to dump. What you set up is as over recurring what you need and then you have some extra...

Chris Nelder: Right.

Eric Gimon: And that E-3 study had much more curtailment than the Low Carbon Grid study. It examine kind of what were some of the factors that led to having less curtailment. And one of them was say California's importing power from Wyoming. And at some point decides it doesn't really need it. Now if that plant in Wyoming has the option to then go on and sell the power to Colorado which at that time might be happy to receive that power especially since it's quite cheap, then the power is no longer wasted and doesn't get curtailed. But that's a function of the kind of contracts and agreements you have has nothing to do with the underlying power system.

Chris Nelder: Right.

Eric Gimon: Similarly there are rules about what needs to be on at what time in a given what's called a local capacity resource basin. So you have basins that are constrained in terms of transmission. So you think of like L.A. basin. There's a certain number of power lines that go in and out of there. And you know over time people kind of create these rule of thumbs about how much local thermal generation is beyond compared to how much you're importing and so on. And if those rules don't adjust then they can become barriers. So that's that the kind of operation or latent. And then there's the demand side which I know is dear to your heart so like electric vehicles can modulate the way they charge or they can charge at night but not during peak times, for example. You could you know use electricity to heat water in water heaters preferably with heat pumps and all kinds of demand management tools that provide you with flexibility. And then I think I don't have all at the top of my head. The thermal fleet you do use, you would like that to be more flexible. So you want to use things like gas for power rather than energy. If you have a lot of supply of cheap plentiful energy from renewables what you need to manage is kind of the gaps in between. And so between demand flexibility operational flexibility and then what's left of your fossil fleet you have a lot that you can manage with. And then finally you get to storage. So storage and back when they wrote this paper was still very expensive. It's still an expensive solution now but storage is starting to become competitive for example with peaker plants, the plants that get used to manage the peak loads in the 20 to 200 busiest hours of the year. And I think over time curtailment will become a strategy for managing the grid. But you have to be very careful. It's a bit of a dirty word for projects because they have an idea of how much power they're going to sell and if you curtail they don't sell as much and then that undermines the project economics. But if you think about the way that technology has been dropping in price over the years we've seen year to year price drops of 5-10 percent sometimes. I mean not every year. So the idea that you might lose 5 percent of your power output because you're just feathering your windfarm or something so that you keep the ability to go back up isn't completely crazy but it needs to be part of the project economics right from the beginning, so that it doesn't undermine the economics of the project going forward and needs to be valued. So at the end of the day there are a lot of ways to manage the flexibility and you shouldn't think of it as having this type of backup power ready to kick in when the sun goes down or the wind stops blowing.

Chris Nelder: So you wouldn't say that a grid actually needs a 100 percent backup power from dispatchable conventional generators or battery storage. But do we really know how much we might need? I mean you know we've made some really heroic attempts to model this stuff out using different assumptions. Studies at NREL, studies at you know the other national labs trying to figure out how much storage you really need. But you know I look at these things and I go OK well that makes sense and those are reasonable assumptions to make, but in the end that's all you're doing is you're modeling around some assumptions and you really don't know if those assumptions are anywhere close to what the reality will be once we get there 40 or 50 years from now and I just keep coming back to the conclusion that we don't really know. We don't really know what the composition of a mostly renewable grid wiill actually look like or how much of the power will actually be provided by renewables or how exactly we're going to manage that grid on an hour by hour basis or anything of that kind. What do you think?

Eric Gimon: Well that's part of what makes the subject really exciting for someone me because it's really interesting. You have to kind of reexamine some of your core assumptions. The other thing to think about is it's not like there's an end point and we set the course of the ship towards that end point and then we just kind of leave the tiller there. These are multitrillion dollar investments over years so it's going to be an iteration. And so as we iterate we will learn. Now the question is are we going down a blind alley? Are we going to some crazy place? I think that's where the simulations that work you know renewable energy futures and things like that really helps us give some confidence that this is not a crazy direction to go into. The other thing to keep in mind is I was thinking of this metaphor yesterday. Since you've been thinking about electric vehicles, you know think about hybrid electric vehicle versus a full on electric vehicle. So a hybrid electric vehicle is running off electricity you know for 40 miles and so on and then it deals with your range anxiety and things like that by having an onboard motor that's still taking on gasoline. So you have some of the advantages of the electric and you have some of the advantages of the power density of electric fuel and also the infrastructure. You've got gas stations all over the place. Any time you think you're getting low you know within five minutes most of the time you can go someplace to pump. And you pump the fuel in there quite quickly. And so during that period you have a lot available to you of the old system and the new system. But there is a price you pay. You know it's crazy for a car to carry two kinds of engines you know the electric motor and the engine to create power. It's much more efficient if you kind of go full on to like designing an electric car that's built around being electric like a BMW i3 or a Tesla or something like that. And then you start seeing some of the advantages of electric like having all the weight on the bottom is better for handling things like that but you need the infrastructure like fast charging and so on. And we don't quite know what that will look like for electric vehicles. To me the grid is a bit like that. We're still stuck with some of the choices as we transition. For example we probably don't need an AC grid. Only about 10 percent of the loads are sinked to the AC 60 hertz signal. So most of the stuff that uses electricity in your house either is indifferent like a resistive light bulb or actually has to invest in turning that AC signal into a DC signal. And with the power electronics we have and so on in the long term maybe the grid we need doesn't need AC. So these kind of design decisions that we are stuck with from the old system that maybe we don't need in the future. And so that's like the equivalent of having to carry an old engine around. Now the the beauty of having both for a while is as you have this stock of power plants like in California we have a lot of combined cycle plants that are generating less energy, then they're more available for power. If something is running at 90 percent capacity most of the time then the best it can do is like go up another 5-10 percent and that's assuming there's no outages and so on. It's running at 50 percent of the time. Most of the time. Or if it's not running all the time then it's available to do things for you. So I think there's going to be this kind of intermediate period where we have some of the disadvantages of the old and there are going to be some advantages to having both the old capacity and the new capacity. For example in Texas right now they have very low power prices because they still have a lot of old coal plants that probably won't be around in five or 10 years and they have a lot of new gas plants and wind plants that are generating the same time. And so their customers are taking advantage of low prices right now because of the transition that we're in.

Chris Nelder: That's a really useful analogy. I like that idea of comparing it to a hybrid versus a full electric car. So you know something that often gets left out of these scenarios is the importance and the shifting role I think of energy efficiency. I mean it always seems like an afterthought. We tend to think about supply and demand but cutting down the demand to size before we try to build all the new renewable energy capacity to meet it is obviously the right way to approach this. I mean if I were to try to design a grid right now for the future I'd start with one that needs at least a third less power, because of energy efficiency improvements. And maybe because of you know avoiding losses like you are discussing from converting from AC to DC. And then I go from there. So what do you think the potential for energy efficiency is now and in the future?

Eric Gimon: That's a great question. I used to talk to the people at climate works and they were always trying to quantify how many tons of CO2 they were abating with various kinds of strategies and the kind of renewables and energy efficiency sometimes were at odds with each other you know because by putting energy efficiency there was less need and it messed with their calculus. But in my mind energy efficiency is a really important part of the mix because if you think less about megawatt hours and more about the services you need then if I have you use a third less energy for all the wonderful things you want to do with electricity and then every unit of clean power that I put in is delivering 50 percent more value in a clean way. I got very interested in what was driving CO2 emissions down in the U.S. power sector. So if you go to a EPA thing or the EIA. Sometimes the CO2 data is a little lagging in the power generation data. But if you look at these things you'll see that 2007 was the peak year for electricity generation in the U.S. I think we almost met that year in retail sales in 2014 and we went down again in 2015. So it's been pretty flat since 2007 and emissions have gone down a lot. Now people talk about natural gas switching out coal as the main factor. And that was the question I wanted to investigate. And I started kind of comparing the role of renewables and natural gas fuel switching. And there was an article by Amory Lovins your chief scientist pointing out the role of energy efficiency. And he gave me a real headache. I was like eugh! All of this analysis is already hard enough and now I have to think about this! But I tried to make a solid effort of it. Looked at all the utility programs reported by ACEEE. Looked at some papers in the industry about what some of the codes and standards might have done. And it became very clear to me quite quickly that a) energy efficiency had done a lot more than the other two in getting CO2 out of the system and b) that's something people really aren't talking about enough. You know the fact that load has been flat or almost flat or even declining for the last eight years is a significant issue for utilities, even for planning. When we think about planning power systems it's always this element of well we're going to need to build something new. But if your demand's going away it's a kind of different system to manage. And it's really striking how little this gets talked about. I mean I love solar panels. I've had them on my roof since 2004 but solar panels in 2014 were providing 10 terawatt hours. The distributed PV stuff on the utility scale. 10 terawatt hours to the grid. The grid generates about 4000. Energy efficiency was clearly much more in the 300 terawatt hour range or maybe 200 terawatt hour range. It's hard to quantify exactly but you can't ignore it. If you pretend that it's not there then you're not understanding the data properly.

Chris Nelder: Yeah it's hard to quantify because it's sort of a counterfactual right? To do that. But you know I think not only do people not think about it enough or make it enough of a priority but there's also been you know for those people that really think deeply about this stuff there's also been a debate about the rebound effect. I mean I assume you've looked at some of that literature. Where do you come down on that question?

Eric Gimon: Well it's a complicated question and there was these things about lighting. How lighting got cheaper and more and more lighting gets used and so on. I mean I'm a pragmatic guy Chris. You know I like to look at the data. You look at the data and there's just a sharp kink around 2007 in the amount of electricity consumed and you know there's a dip for the recession but it kind of recovers. It's not the recession. The recession had an effect for sure. It's close in magnitude to energy efficiency. But there's ways you can kind of look at it. There's the economic engines of the EIA and you know what predictions are they making what actually really happened and so on. And there's no way you can jig the data around and not see the effect of energy efficiency in there. So I'm sure there's some rebound. I'm sure it's hard to kind of completely quantify these things but really it's had a major effect on our power grid in terms of emissions and it must really be hurting the utilities to not be seeing growing loads. I talked to a utility guy in Texas. I think he was with Encore. He said he went a conference of utility executives and somebody asked Who here has a growing load? And it was like a room of like 200-300 people and like four or five people raised their hands.

Chris Nelder: Yeah. No it's tough.

Eric Gimon: If you're an S&P 500 company has to grow your profits every year. That's a big paradigm shift.

Chris Nelder: Yeah yeah. Well it's something that I've actually dealt with a lot in my most recent research you know looking at how these demand curves have remained quite flat or even declined in a lot of the country and what the implications are of that for future grid planning.

Eric Gimon: Yeah RMI has really been at the forefront of these things so you're in with the right folks. The other thing I should say is, there's kind of a new kind of EEM on the horizon. As people I meet that are trying to be more results oriented with energy efficiency instead of kind of these programs and getting into kind of counterfactuals and crazy measurement and evaluation they're kind of more using all this data that we're getting from smart meters and so on to just the something kind of like what I was doing. Like is consumption going down? I mean you have to factor in some other factors like temperatures and so on. But there are ways now with that type of analytics that we have to kind of start extracting whether something is happening. So that's one reason for optimism I think because that will open the door to more creativity in getting to the energy efficiency and energy efficiency is really the silver buckshot type thing. It's just so many different ways you can accomplish it and every building is so different. The other thing that's an interesting development with energy efficiency... My friend Sam Borgeson wrote his thesis on using smart grid data in California. And he shows that you can shape energy efficiency instead of thinking of it as a kind of you know straight haircut of consumption you can think about energy efficiency measures that might target peak load or you know behave differently one geographic region to another or a target different types of customers and so on. So I think the line between energy efficiency and demand responsive demand management is going to start getting blurred too.

Chris Nelder: Yeah I would agree. So as we get into this more highly renewable grid how much of a target do you think it would make sense to set for you know how much load reduction we can do through efficiency, let's say 50 years from now from what would have been?

Eric Gimon: Yeah not as much of an expert on these things. Partially I think it's my analytical data driven side gets just thrown off by the sheer complexity and kind of obscurity of this area. So if you ask me to take a guess I think your one third figure is right up there. I mean I like the ideas in the renewed paradigm of Reinventing Fire. And we were talking about if you reinvented a grid where you're at... You know the people at Lawrence Berkeley labs are talking about creating these systems for people in developing worlds where they can be running a refrigerator a TV a fan a few lights and a phone charger or something for 170 watts. Very efficient DC appliances.

Chris Nelder: Yeah we talked about that in episode 12.

Eric Gimon: That's right. You talked with Justin, yeah. So he gives you a sense of what what can be accomplished. I mean that's one big one the AC DC thing. It will be interesting to see how that plays out. Sometimes I wonder if these new USB power standards will be kind of part of the route that takes us there. Because I've seen houses now like my in-laws' in Australia where all the sockets have a USB thing. I assume they're still like converting AC power back to DC but I imagine houses at some point might just wire up for for DC and provide USB sockets all over the place.

Chris Nelder: That's an interesting idea. So shifting back just briefly here to EVs, I think it would be helpful here if we dug down a little deeper into the concept of flexible load. So you know one of the key elements to the future composition of the grid is what you can do with large flexible loads to shift them around and adjust for the rest of the system. And I recently wrote a paper for the Rocky Mountain Institute about the value that electric vehicles can deliver to the grid if they're charged at the right times and in the right places. You've also written about that subject in an article for Green Tech Media and I'll going to both those pieces in the show notes. But you note that distributed resources like EVs and also things like rooftop solar, distributed storage, and other responsive loads can actually provide a wide range of services to the bulk power market, to the transmission grid, including peaking capacity and ramping and voltage support and improved system efficiency. And based on my own research I totally agree with that. But in order to deliver those services to the transmission grid, the distribution system has to cooperate. It has to draw power at the right time and place. And you framed this issue through three different lenses, a physical lens that takes into account the physical properties and locations of distributed resources, an economic lens which considers the economics for utilities and energy vendors, and an information and control lens which is how the physical system is monitored and managed. So let's just use EVs as an example and explore these ideas a bit. How do these three lenses help us understand for example the value of managing EVs on the grid?

Eric Gimon: Okay thats a great question. Lets take these for a second. I want to flag something. So I switch to EV about two and a half years ago and you know the first thing I noticed is my consumption for the house went up by 50 percent. I mean I guess I was covering a lot of it with the solar and so on. But that focuses your attention.

Chris Nelder: Yeah.

Eric Gimon: And I had installed all these inductive monitoring devices on some of the circuits in my house so I could see what was doing what and so on and I was just not paying attention to it and I was not putting that stuff into Like rate schedules and by the way I knew I couldn't find it like an Excel spreadsheet with my rate schedule on it. I had to kind of type everything in by hand. So you know I've even though I had all this stuff and I was an energy geek I wasn't really looking at my rates and things like that. But as soon as I had the electric car and it was consuming all this electricity. Well now I really looked at my rates and thought about it like when am I going to run my car? When am I going to charge it and setting the charging time. As a somewhat trivial side, the metering stuff was kind of nice in that I have a sense of like when stuff breaks down in my house. And I understand better like what's going on you know oh this fridge is clearly dead. It's been doing this for a few days. I think that's a value stream people underestimate in energy management systems.

Chris Nelder: Just the knowledge, yeah.

Eric Gimon: So back to your question. So imagine imagine I want to get regulation out of EVs. I convince BMW or Tesla or Nissan or whatever to kind of manage their whole fleet of things out there and you know sign up their customers and so on. That could be a great resource for the grid. You know imagine we get to those million EVs and they're eight kilowatt units or something and you know that's like eight gigawatts. I mean they're not all at once and so on but you're talking in the gigawatt scale of adjustable load. That's huge. That's a whole nuclear power plant right there. But now think about it from the distribution circuit. Suppose like my whole electric car fanatic club like move to my neighborhood and we're all living here and we all decide hey you know 11 o'clock is the best time to start charging, right? Not 11:01, not 11:02, not 10:59. Well what happens on my local circuit? You know we all start charging at eight kilowatts you know which is like three or four times a normal household load all at once you know all in my pocket, well things are going to go crazy on my local circuit. And that's going to create big voltage spikes and so on. And the California SO doesn't care. We could be, you know one car could be in Bakersfield, one car could be in San Francisco, one in L.A. and you know it might be happy to send signals out through third parties or directly to like start charging exactly at 11:00.

Chris Nelder: But if that all happens in your neighborhood you're going to probably blow distribution transformer.

Eric Gimon: Right exactly. So part of the physical lens is kind of understanding like where things are and how they affect things. You know in practice most of what a local distribution network is about is how often they have to move their tap changers on their local transformers. But they're not used to thinking about things too dynamically. They have kind of these templates and that that drives the investment and then mostly they I think overinvest. That's that's a loaded word. They you know overengineer to make sure that things work. And so there's still... I mean I think we can do a lot of stuff today with DERs and so on and we probably won't hit some of these physical barriers but I've talked to like a distribution engineer. We had one call... talking to distribution engineers. So we talked to a guy in Germany and he mentioned how they have these systems where they heat bricks at night with electricity and then that provides heat during the day. And he was saying if all these bricks turn on at once they created problems for them. And so they developed this program where they would kind of rotate through houses so that it would kind of smooth it all out. And I hope that gives you a sense a more visceral sense of kind of what some of the physical challenges might be.

Chris Nelder: Sure.

Eric Gimon: And so I think it's important for advocates and people like that not to completely blow off the utility when they talk about some of the challenges that they have. The issue is often incumbents use reliability or issues of that ilk to defend the status quo. So there really needs to build up over time a trust between people who want change or people who want things not to break. That we're all on a common page and kinda understand the issues. And that brings me to what I call the economic lens which is if there is no incentive for the utility to have DERs succeed, it's going to be really hard to build that type of trust. And they're not going to put in the investments necessary to make it work or they're going to see it as it will this is a great opportunity to gold plate and it just kind of put the investment everywhere as long as it gets put in the rate base and not necessarily maximize the utility for the customers and keep the costs down.

Chris Nelder: Right.

Eric Gimon: And that's I think... In the transition, in the energy transition of electric grid we've got to make sure that it never gets too expensive. Overall it would just be bad for everyone in the pioneer states in the pioneer places. It would be terrible. It would slow things down because people would just turn and say look it's very expensive you can't do this stuff. So we really have to keep an eagle eye on kind of what happens to cost and how costs are justified. And then you know the third element is that the independent system operator that manages the transmission grid and even the utility they're not used to thinking of millions and millions of devices that they need to talk to. So you need the command and control part of things and you need communications that give you information about how much people consume or how much people provide so you can pay people adequately. So that whole communication layer is going to become important. And you know one of the things I brought up is one way to gold plate is to kind of build all your own modems all your own fiber all your own stuff as a utility but that's not necessarily you know maximizing for society. And what I proposed and one of the articles is what I would call borrowing the words from Ronald Reagan, trust but verify. If you have ways through the Internet or other telecom assets that already exist to really give a fine grain signal and it's going through some kind of third party aggregator and so on, all they really need to do is to kind of audit that person and make sure they're doing it right. You know and as long as that person has an incentive to keep you know being honest with that stuff. I mean there might be some resilience issues and so on. But as long as that person knows that there's these audits going on then they have an incentive to make it work. And an example of that is I saw a few years back ... who Ran ARPA-E at DOE and then was a vice president for energy at Google was talking about this experiment at Google where they had several electric cars in the parking lot that were responding to what's called a regulation signal, so a signal that helps you balance the grid on a five minute by five minute basis. They were responding to a signal perfectly. You know here was the black line that was like what the system operator wants and here was the red line of like what all these charges are doing and they're like right on top of each other. And then he shows a line of like what a big generator was doing at the same time to respond to the regulation signal. And it was you know like in the neighborhood but was often off and it would kind of swing up and down. It wasn't you know an exact response but they said look that guy can get paid to provide regulation even though he's doing a so-so job because he's got the revenue grade meter. But we don't have these individual revenue grade meters at all these cars. But look at what our aggregate of managed chargers can do, it's a much higher quality thing. So I think that's something we really have to look at is how does the you know the payment system, the auditing system, and the control system work without having to kind of build a whole parallel architecture.

Chris Nelder: It's those very same issues that I reached that got to the point where I was like OK this is just going to take some more development. We're not really sure where this is going and the best we can do is try to provide some guidance. And you know one of the other interesting questions that comes up there is what's your best market design in terms of who should own this infrastructure? I mean all these services and values can exist whether the charging of the vehicles is being controlled by the user or directly by the utility or via a third party aggregator who can operate a network of charging stations or at least control them in order to provide you know a demand response service or frequency regulation service. But there's differences in terms of how quickly you get there and what the social equity side of things looks like and so on depending on who really owns it. So how do you see the pros and cons of utility ownership or control versus having like a third party aggregator provide these services?

Eric Gimon: Yeah that's a really important question and one that I think a lot of regulators are struggling with. So on the pros you know there's clearly an efficiency advantage having utility manage some of these distributed assets because they can provide an integrated frame. They you know they can do distributed resource planning or some kind of integrated planning effort and try to think about where things should be to optimize. And they have access to capital. They could run auctions to try to keep prices down. Things like that. And also they're more sensitive to the social mandates, so they can make sure that it's a grid that works for everybody like low income customers and so on and not just something where people who have access to capital can take advantage and others can't. You know the problem is back to that economic lens they're not always incentivized to do what's best to optimize the system or you know what's best for their customers. Now you know I'm not saying they're bad people it's just we tend to live through the incentives that we react to. So I think it's important to create a regulatory environment that if you're going to let them own assets that has them have the proper incentive. The other way I think about it is if you take something like storage and you look at any kind of complex system like a tree or rainforest, whatever, the storage tends to be distributed. So it's not like there's just one small size and you see it everywhere more like a kind of a fractal nature. I was a little disappointed that Lorenzo didn't talk more about his idea of fractal grids when he was on your show, where you see kind of a self replicating structure. So you know a little bit of storage here at this level and then where things kind of come together, a slightly bigger stores there, and things... You know the main trunk and another a bit of storage there. And for the customer I think what you will see evolve over time, I'm not sure when exactly this will happen but it's clearly the direction things are going... Some kind of energy box or energy management system. Something that talks to the grid and is kind of like your agent and you don't worry about it much but it thinks about when your fridge is pumping or how to charge your car whatever it manages all of these things for you. And I think for that agent to have some of the storage at its beck and call gives it some leverage with the utility. But if the utility owns all the stuff then you know we're back to kind of full on monopoly and we're not taking advantage of the possibilities of competition and the creativity of different distributed actors.

Chris Nelder: I like that idea of there being sort of a little box with a little storage on board that actually could be a very interesting part of all this. And it brings up the importance of the communications infrastructure. I mean one of the missing pieces in this whole question is the communications for utilities to monitor the performance of various DERs on their systems and other grid conditions. And so far you know we've only got advanced metering infrastructure in about half of the grid's customers in the US. And you've actually suggested that instead of utilities directly connecting DERs to their control systems for example requiring a revenue grade meter to be installed at each residential EV charging station like you are just talking about which is slow and costly that they could actually use existing telecom assets like cell phone towers or the Internet to more efficiently collect real time data on the actual grid conditions and then distributed resource performance. And I guess that kind of leads into this whole Internet of Things concept as well. And that all makes sense to me but I suspect that most utilities would actually consider it too risky or at least that they'd be unwilling to give up total control over the communications and would want the data delivered from equipment that they've already deployed and certified or that has their stamp on it in one way or another and I just wonder what it would take to get them to give up their requirements for revenue grade meters and support these other communications mechanisms.

Eric Gimon: Yeah well that's also a good issue. Yeah well we've seen so far is lots of pilots, right?

Chris Nelder: Right.

Eric Gimon: It's kind of like you know death by pilots and there are a lot of tentativeness from the C-suites... I think you know I think if they had more certainty about kind of what this all means for their business models there would be maybe a little less fearfulness about you know going this way. One of the beautiful things about America is we have so much diversity that I think we'll see different solutions or different approaches take place in different parts of the country so we'll get a chance as part of this iterative energy transition we'll get a chance to see a lot of different ways that things can go. But you know on the plus side of what might motivate utilities you know if they have the right incentives they could find you know much cheaper stuff from the DERs and that might push them in this direction. There's also cyber security risk. So one thing somebody told me, an expert in the kind of energy and communications space told me, is you know they don't want to be in all the internet of things. Because if they're in the internet of things they're responsible for all that cyber security and potential infections and so on. You know all the way into their main system and they really need to kind of firewall themselves from that. So at some level they're better off with a lot of third party aggregators of different people who were kind of exposed to the kind of raw messiness of the Internet of Things and they'd rather have kind of cleaner protocols that talk to kind of aggregate entities. So I think in the long run for utilities getting too far into the weeds isn't necessarily the best way to go.

Chris Nelder: That's a fair point. I mean there's certainly a very justified defensive aspect to their stance for sure. They after all are required to maintain a reliable and safe and unhacked grid, right?

Eric Gimon: Right. But I guess why I was bringing the point up is of course you could take it as well let's just not do the DER stuff then we won't have any of these cybersecurity problems. But say granted this stuff is happening, some people are in the kind of you know... Utilities are like the Rockefeller octopus and they want to control everything. Well that's not really true. And there are some areas that don't quite make sense for them to go into and they'll be happy to have an ecosystem there. And so if the regulators and the society gives them the right incentives and rewards them well and gives them like visibility towards good financial health around these areas I think we will see them more open to having certain parts participating.

Chris Nelder: Well you know this does kind of leads into a broader question about how well distribution utilities are adapting to the new reality of DERs or thinking about how to leverage them in their systems in order to provide value to the transmission grid. And you know that that leads to another area that I've been researching and that's how to get utilities to prioritize distributed energy resources over big centralized conventional grid assets like big power plants and big substations or at least give DERs a chance to compete on an equal footing. And this is an area that I think most utilities aren't quite yet comfortable and familiar with. Perhaps they don't know that a portfolio of DERs can provide the same services as a big power plant or perhaps they consider it too risky and they don't really want to try it over something that's tried and true like a big gas fired power plant. What do you think the big challenges are in getting utilities to reorient their capex planning around DERs?

Eric Gimon: Well the first one is incentives. Utilities are engines for taking capital and getting nice rates of returns on it. So if you're going to be able to deploy more capital and your regulators aren't going to complain by going the old powerplant way why would you go the other way? Then there's experience. They just don't have enough experience with these things and they're naturally very conservative entities. Some of them are in the power markets where are you have incumbents like generators and so on who are actively going to resist competition. You know if I was a peaker plant I wouldn't be so happy about a demand response product that's at half the price or batteries coming in and eating up my market. There's a certain inertia and then at a deeper level I think there's two things. One is a lot of this takes a shift to what I would call a more statistical thinking. Like let's go back to that Google car example. They don't really know who's going to be parked there all the time. Somebody might come in, you know, leave early. Or maybe there's 301 cars one day and 299 cars the other day. So this is kind of weird kind of certainty issue. Who exactly is providing me with this regulation? Imagine if you're now... you're Like Nissan or BMW and it's like a distributed fleet. I don't even know where some of these things are. So if you're thinking in terms of this kind of very deterministic model like I know where every asset is and what its current state is and how it will behave then all this stuff is very fear inducing. But if you have this kind of more statistically based mentality, well yeah maybe that car will be there or not that car. But you know what? 300 or 1000 cars, some of them are some and not there is actually a more reliable aggregate asset than a power plant. It's going to be down less often. And it's never going to be all the way down. There's going to be very few days it will be like zero cars. I don't know maybe New Year's or something. And those would be fairly predictable. So I think there's that deeper level becoming more comfortable with that kind of statistical nature of things. And then you still need the ecosystem to exist. You know you still need the vendors and the equipment and the kind of people you can talk to that makes sense and have and kind of package solutions the way that makes sense for the people and utilities. And then there's also a legal dimension to these things. For example I was talking to somebody at EDF about what does it take to have non transmission alternatives be looked at the same way as like new generation or new transmission or something like the PJM market. Well you know transmission lines are clearly like interesting things. Make sense for FERC to regulate them. The wholesale power market's been given over to these markets but they don't quite know how to get into whatever vague notions one might have about non transmission alternatives even if they're much cheaper. They don't have an easy way to factor them in and then sometimes the rules aren't just written right. Sometimes the rules... Who was I was talking to about a storage project that we're just not performing and PJM and part of the reason was it was basically having to go to full power right away because it was required to do as best it can to be at I don't know 800 kilowatts or whatever it was. And so it was just draining itself right away and not particularly useful. And then when they figured out like that rule doesn't make sense, then they had a really nice asset that was responding well to regulation signals. And so we often see these kind of institutional barriers. You know something needs to stay on for four hours, but does it really need to stay on for four hours or could you have a rotation in our relay team of assets? And so often that's the problem with these kind of multi... the non vertically integrated utilities.

Chris Nelder: Yeah I think in a lot of ways it really does sort of begin with the regulatory framework doesn't it? Because that's what the utilities are working within. So what are some of the things that regulators should be thinking about and doing at this point to encourage a transition?

Eric Gimon: Well one thing a lot of my colleagues are very excited about is what's called performance based regulation. The idea that you should try to pay for what you want and not for what you got. Or another way is pay for what you want and not for what you used to get or what you used to need. And so with this approach there's a whole variety of different approaches but you're trying to focus instead of telling the utility oh this was an OK investment this wasn't a good investment or this will be an okay investment this is not... Focus more on I'd like to get peak use for Manhattan down by this much every year. You know I'd like to use my transmission lines more efficiently every year and things like that and then reward the utilities around how well they perform. And now that's a tricky game because if you still leave on the table huge amounts of money from just great recovery on capital deployed they may not shift in the way you want and then the big issue becomes kind of information asymmetry. How well does a regulator know what a utility might or might not do when they're structuring incentives because you know if they end up giving the utility too big an incentive and the utility actually can perform quite easily then the public's going to get mad at them or consumer advocates are going get mad at them. Why do you give away so much money? They had this solution and they did it. Why don't you just mandate that they do this solution right off the bat? I think from the utility's point of view there's an advantage to keeping the regulators more in the know. You know providing them with cleaner and better information in the long term because it creates a tighter coupling between the two and more of an opportunity to kind of seek a win win for the utility and for the customers instead of kind of big swings where maybe they have a big payoff on this accomplishment but then there's a feedback mechanism where everybody's angry and then there they don't find another way to kind of find value. Another issue in the PBR world that is hard to deal with when you're talking utility is could there be a world where utility is more profitable but smaller? At some level for investors more profitable is what they care about. But if you're running a company typically your goal isn't to make it smaller, so that I think will be another challenge with thinking about performance, especially like we were talking about you know if loads are going down how are you going to provide you know 5-6 percent growth or whatever on your invested capital?

Chris Nelder: Well in addition to performance based regulation don't we also need some very deliberate signals that we really want to build a renewable future?

Eric Gimon: Yes.

Chris Nelder: And so for regulators what does that look like?

Eric Gimon: I mean I think at some level they need the backup of legislatures or governors. You know so California we have these clear long term goals for 2030 now. That organizes the mind and helps the utility and the regulator be on the same page. In New York they have the renewable energy vision. That's very exciting. I'm not sure... And you guys at RMI have been very involved I know in helping them frame that. I think it will be interesting to see how that plays out. So I think if you have this kind of energy it really helps. Now it's not a coincidence that this is happening in New York and California. These are big states with lots of regulatory capacity. They have big staff. Hawaii is having to do some of the same things that New York is thinking about with 10 percent or 5 percent of the staff. So I think clear signals you know clear leadership all these things really matter. And at some level both utility executives and regulators have to learn to be brave. You know they have to step out from the pack. And there will be rewards I think I mean one way I all I've thought to talk to state decision makers and stakeholders is some of these utilities will be the big winners. They'll learn the energy transition, they'll make a good transition and then they'll be eating up utilities in other states or whatever they didn't learn. You know the FirstEnergys or whatever will not be the big companies of the future I don't think.

Chris Nelder: Yeah. Well I think you've made a good point there that there's a role for regulators to play but we shouldn't expect regulators to carry all the water. Right? There's a role for legislators as well. You know there's a role for elected officials to just set a vision. I mean I think actually Governor Jerry Brown has been brilliant in that way. So we should probably dig a little deeper to the market side of this transition and how we used markets to get from where we are now to a mostly renewable grid. I think we all understand that that time of use rates for example are a good start and that's a place where regulators have played an important role in putting those kind of rates in to help the utility manage its load shape. But I think if we want to end up with an optimal design we're ultimately going to need nearly real time rates for all loads in all places and get away from things like fixed charges and demand charges or even interval based time of use charges which are currently being attempted as essentially stopgap measures to help utilities survive the transition. What do you think about that idea of just kind of moving the whole grid and all of the assets on it to kind of a real time pricing model?

Eric Gimon: What do I think of that idea? So initially I was kind of in favor. You know I think on paper it sounds great. And obviously you wouldn't want like you and I would be like responding to real time prices or whatever right. This energy box I talked about before it would be talking to these things and this would be like I don't know like that machine language or like you know the way the Internet has all kinds of protocols that you and I never have any dream of what they are.

Chris Nelder: Right the most we would do is set some parameters on it, the range we want it to operate within.

Eric Gimon: Right. I came home today my house was a little too hot. I didn't like it. You know whatever you do, like fix that. But then you know if you dig deeper it's a little bit like the energy only market approach in Texas for the wholesale market. I think on the wholesale level nobody quite knows how the markets will evolve but I suspect that the energy will kind of decouple from real time stuff. That energy will become more and more of a kind of a bulk commodity that's not that time sensitive. And then there will be a more active m,arket and kind of energy shifting and energy management and things like that. At the residential level I wrote a chapter in a book called The Future of Utility, Utility of the Future" that was called the customer centric approach to the grid. It's like if you thought about the grid from the beginning not what you have today but kind of first principles. What are all the things the grid does for you? I was surprised at the end of day being my list ended up being like 20 or 25 different things long. A lot of them you just don't think about. You start thinking about them more when you design like a battery plus PV system out in the woods or something. You know like like the size of the pipe you know like you consume a whole bunch of power some months and then you go on vacation for a month and you hardly consume any, you don't really incur any penalty for that. So there are a lot of benefits that accrue from connecting to the grid and it's unclear to me that we should think about all these benefits tied to kind of just energy provision. Maybe like you say real time becomes is very granular approach that allows you to capture all kinds of different value streams through one signal but even if that was true there would have to be overlays of different kind of risk management products or reconfiguration products that would kind of disconnect you from that. I mean last year at Texas the wholesale prices dropped almost 40 percent and the average retail price dropped 3 percent. You know that's because there are these like buffering mechanisms between the wholesale and retail and something like that would have to exist even you know with this fancy energy box. So I don't think that the real time pricing is a panacea. And I think they'll be different approaches. For example an approach would be your energy box negotiates with the grid and either buys a complete block of energy just a fixed amount you know like I'm going to consume 1 kilowatt per hour or something like that. Or it could have a block where it says I'm going to consume between minus 1 and plus 1 kilowatts at my whim and you're going to just track that for me. You're going to provide squiggles. Somebody told me the other day that utilities are companies that take blocks and provide squiggles. So that would be a squiggle provision and then a third block could be the opposite where you as the prosumer or whatever are providing a squiggle back to the utility, where you say as long as you're within this minus 1 to plus 1 kilowatts or whatever I'll just follow your signal. And my energy box is going to manage my internal storage or my fridge or whatever it has to do back here. And you can have a combination of those blocks. So that would be kind of a market system where there's different elements of real time and other things going on. But we're still a long way to that. I think we probably don't want to get into blunt instruments like demand charges and fixed charges, I think because it's going to become distortive. You know in California the biggest market now for storage