All Episodes

[Episode #58] – Solar with Storage

0:00
-1:07:32

Historically, thermal concentrating solar plants were the only type of solar power equipped with storage. But as cheaper PV systems became dominant, thermal solar plants fell into disfavor. Now solar PV systems are beginning to integrate storage based on lithium-ion batteries, and this storage isn't just used to supply power when the sun is down; it is providing grid stabilization services too, which only adds complexity to an already-complicated picture for the future of storage - confounding attempts to model how much storage we’ll need, and of what kind, and when will we need it. Is a large amount of seasonal storage required on a high-RE grid, as some analysts have suggested? Or will other technologies reduce the amount of storage we’ll need? And can we even forecast that need, years or decades in advance? We’ll delve into all those questions and more in this deep dive into combined solar and storage systems.

Geek rating: 7

Guest: Paul Denholm is a member of the Grid Systems Analysis Group in the Strategic Energy Analysis Center at the National Renewable Energy Laboratory. He is a leading researcher in grid applications for energy storage and solar energy. He pioneered a variety of research methods for understanding the technical, economic, and environmental benefits and impacts of the large scale deployment of renewable electricity generation.

On the Web: Paul Denholm’s page at NREL

Recording date: December 1 2017

Air date: December 13, 2017

Chris Nelder: Welcome, Paul, to the Energy Transition Show.

Paul Denholm: Good evening.

Chris Nelder: I think we all understand that adding storage to a solar plant can compensate for the variability of the solar plants output. But what's the primary financial case for systems with integrated solar and storage? I mean is it mainly just that it allows a plant to run close to full time instead of only when the sun is out? Or is it that the storage system smooths out fluctuations in the output of the solar array? Or is it that the storage component allows the system to participate in additional markets by offering for example, grid regulation's services or is it sort of all the above?

Paul Denholm: It's a little bit of all the above but quite frankly the primary driver right now is tax credits. So right now photolvotaics are eligible for the 30 percent investment tax credit which basically roughly takes about 30 percent of the price off the top of the cost of the PV system. Storage is not eligible for the investment tax credit unless it's actually physically coupled to the PV system, and gets most of its energy from federal tax. So under those configurations the storage is considered part of the PV plant and then becomes eligible for that 30 percent investment tax credit.

Chris Nelder: Wow. I didn't even realize that. So okay that's super interesting.

Paul Denholm: Yeah that really gets into one of the things that's interesting about when you start getting into the energy analysis field is the realization that it's not just physics and electrical engineering. For better or for worse if you really want to get into the details you have to start learning things about tax law and financing and things like that. So that can be, sometimes be a pain but it's critically important to understand all of the financial underpinnings of how our industry works.

Chris Nelder: Well we have certainly had a couple of lawyers on the show, but I haven't had any tax experts on the show so maybe that's something we need to do. Well, ok so what if we took that out? I mean that's really good to know, what if we took that element out, what if there were no more ITC ,then to restate my question where would the major value be in having a combined solar and storage system?

Paul Denholm: Sure, so the major value would actually be the reduction in cost. So it turns out that batteries have a bunch of power electronics in them and controls in them and those systems, those power electronics, those controls are actually quite similar to the controls and power electronics that you need in the photovoltaic plant. So it turns out that what you can do is, you can engineer the systems so they share a lot of the common components. In an ideal world you really wouldn't combine photovoltaics and storage. You'd put storage exactly where it's needed and photovoltaics in the best site. So you probably actually would never combine them in most cases. But it turns out that the advantage of reducing the costs means there's some synergies there that can be combined and that can reduce the cost of the combined system. The services that you described kind of at the beginning of your question about whether or not storage can provide all these additional benefits by helping smoothing out the photovoltaics and providing additional services, that's all true, but it actually doesn't need to be coupled to the PV. So you can put a battery in a downtown region, have your photovoltaics in a remote region, and the battery can still active mitigate all the fluctuations and provide additional services. They don't have to be physically co-located or coupled in any manner.

Chris Nelder: Okay that's fair, although the more separated they are on the grid the harder it is for that to happen. I mean then the solar system becomes essentially a very distinct resource from the storage component, right?

Paul Denholm: It does. But with modern controls and communications there actually really isn't any reason why a photovoltaic plant located dozens or even hundreds of miles away from a battery. You know they can talk to each other so that when there's a reduction in output from the PV it's really the system operator who's monitoring not just that one photovoltaics plant, but all the PV plants and the entire system and it's responding to the aggregated output from all of the PV plants, all the wind, and all the demand.

Chris Nelder: Okay. All right that makes sense so going back to your earlier point about how you really get some cost savings from coupling the two systems together. I mean just looking at the capital cost aspect of the combined system it seems that the main way that it reduces the cost over the 2 separate systems is that it let's use one less inverter and inverters are expensive, is that true? I mean that that's the main cost savings.

Paul Denholm: Yeah it is the reduction in the inverter, but there's a bunch of other stuff. I mean one of the things that's so interesting about the decline in photovoltaics costs that have happened over the last decade is for a long time people were fixated on the panels themselves, the PV modules. Turns out that the cost of the PV modules have come down so much that they're now less than half of the total cost, and the costs have come down so much, we're now kind of scraping away pennies from the system and it all adds up, but you have to look at each and every little component. So it's not just the inverter, it's the shared siting and permitting. So instead of instead of getting a permit for two different installations now you only have to get a permit for one. The cost of going to utility and signing up for an interconnection agreement, now you've just got one interconnection agreement. Each individual, are only a small fraction of the system but they do add up. So it's not just the inverter, it's all the other balance of system and soft costs and all the other little factors that add up.

Chris Nelder: Oh, that's interesting. So about oh geeze more than a decade ago when I was actually designing and selling solar systems, as I recall, and it was mainly you know residential, small commercial that kind of thing. As I recall the inverter was typically around 15 percent of the total installed system cost at that point, what is it now?

Paul Denholm: It's probably about 10 percent, and just like photovoltaics have come down in price inverters have come down in price too. They're now pretty much completely commoditized and it's probably unlikely that the inverters will come down too much more in cost.

Chris Nelder: Interesting. So correct me if I'm wrong, but when you're combining solar and a storage system now, especially like utility size one I think, the inverter is typically sized for the solar array only rather than the combined output of the two. Why is that since it would mean the total output of the two systems would be reduced from what they could produce together.

Paul Denholm: That's right and that's kind of the geeky key details of our study that we just recently published, and you're really kind of getting in the weeds but it is kind of important to understand.

Chris Nelder: Hey look, I read that study or at least part of it and getting in the weeds is exactly what we do on this show so go for it.

Paul Denholm: Okay. Well basically, yeah, when you're sharing the inverter it means that when the PV is producing its full output, you can't actually use the storage at its full output. And when we started the study kind of the hypothesis going in was, gosh this is going to really reduce the value of the system because we've got you know again the PV is producing at full output and we can't use the storage, and during periods of peak demand it would be great if we could produce both PV output and storage output. And now we can't do that. Well it turns out that the peak demand for electricity is around 5 to 7 pm on a hot summer afternoon that's when the prices get really high. Well the sun is still shining during those hot summer afternoons at 6 to 7 pm, but not that much. Right now we say that the reliable capacity credit or the fraction of the PV plant that can really contribute to producing reliable electricity is about 50 percent. So that basically says that in a place like California if you build 100 megawatt PV plant, the local utility would really only count 50 percent of that, or 50 megawatts, towards its resource adequacy contribution, or you know how much you could rely on the PV during those peak summer afternoons. And that leaves headroom for the storage plant. Now here's where it gets really fun and tricky. One of the interesting challenges of deploying increased amounts of photovoltaics is that PV right now contributes to that peak generation but give it another few more years as more people put PV onto their rooftops and more utilities install PV, PV will effectively shift the peak demand for electricity. PV will satisfy that demand at 4 pm or so, but now that peak demand for electricity will now be 7 or 8 or even 9 pm in the evening when PV is no longer producing large amounts of output. So when we look at a combined PV storage system in the not too distant future, the PV will produce in the 11 am, noon until 4:00 or 5:00 pm, and then the storage system will take over and produce peak output during 6, 7, 8, 9 pm. So that's where there really is a nice synergy. PV produces during the middle of the day, the storage will produce in the evening and it will no longer be a burden to share these components.

Chris Nelder: How interesting. So combining this basically changes the supply peak on the solar side, and that's actually reminding me of kind of the opposite dynamic which is that as we get into more and more deep electrification through heating and so on that we might actually be flipping certain grids from like summer peaking to winter peaking. And so everything that we knew before is now changing. It's also made me think about the fact that 10 years ago the time of use rate in California for example always had the highest priced hours of the day between like noon and 2, and now that's the peak price because we have so much solar during that time we actually have prices going negative, I mean, wow. Everything's changing under our feet.

Paul Denholm: That's right, and that dynamic is one of the key elements of integrating large amounts of wind and solar into the power system. Exactly what you just described. We're going to need to incentivize people to use electricity when it's most available in the cheapest. And it used to be that we have these kind of simple ideas of on and off peak, and those simple ideas are no longer going to be the case as we put more renewables onto the grid.

Chris Nelder: And I love the fact that San Diego Gas and Electric experimental new tariff, the GIR for electric vehicles is going to be posting a day ahead, hourly varying wholesale prices, I guess it would be retail prices for electricity. So that's really interesting. Then you're getting out of this kind of rough banded approach to time of use and you're actually getting into a very dynamic approach to that. Do you think something like that would happen or would become relevant in storage as well?

Paul Denholm: Yes I think so. You know storage is really just one of the pieces of the puzzle we're going to need to integrate large amounts of renewables as we look at kind of decarbonizing the electric sector with renewables, a lot of things are going to change including how we incentivize consumer behavior and how system operators and individuals will interact with the power system. One of the real key challenges is going to be making sure that people with behind the meter storage, if you've got a Tesla Powerwall or if you've got an electric vehicle or if you've got distribution sided storage, we need to make sure that we incentivize the proper behavior. And I don't mean you know hardheaded regulation here, I mean this will all be done through the markets, but basically making sure that the way that you buy and sell electricity to your grid is appropriately incentivized with the proper price signals. So when electricity is cheap or even free the utility will send you signals and you can charge electricity, you can charge that vehicle at very low cost, and then when the prices are very high you can discharge from your storage system or make sure that your vehicle stops charging.

Chris Nelder: So at what point might we expect solar plants with storage to become competitive with for example natural gas fired peaker plant? How competitive are they now or how far does the cost of a combined solar and storage system need to fall still to become competitive with gas peakers?

Paul Denholm: Yes so this is another thing that's been really interesting to watch over the last 10 years. I've been working on storage and solar for the better part of 15 years now, and we saw solar prices come down quite rapidly over the last decade, and storage plant prices have only in the last few years started to kind of follow that trajectory. When you look at the projections for storage, and there's a lot of uncertainty about how cheap storage really is just because a lot of the pieces of storage aren't completely commoditized. And what I mean by that is you know storage right now, utility scale storage hasn't reached the point where you can really get a turnkey storage installation. All the individual pieces still kind of have to be cobbled together a little bit. You've got the battery, you've got the charge controller, you've got the air conditioning HVAC system, all the software, it's coming together but you still kind of have to put those pieces together. And whenever you have to kind of make your own system that adds costs. And so over the next couple of years what we're going to see is we're going to continue seeing the cost of the batteries themselves come down but we're also going to see more developers and integrators offering a complete and total package. When we look at the cost projections from a variety of vendors my best guess, my personal belief is that right around the time of 2020 I really think that storage will become cost competitive with new peaking resources. And given the low cost of PV, you know PV plus storage will also be cost competitive. Now in some places it's already cost competitive now. If you've got transmission constraints in an urban area where you can't install or it's very expensive to install peaking generation, a storage solution makes sense right now. But in kind of larger areas of the U.S., when we do the math on what we think natural gas prices will be are on the 2020 timeframe, I really do believe that a neutral investor, a system plan or a utility that needs a new peaking resource, I think they'll be able to look at storage in much of the country in 2020 and say over its lifecycle cost when we do our full financial calculations dealing with all of the taxes and financing and things like that, right around 2020 I think you'd see this kind of tipping point of breakeven. Now I'd say you know that's plus or minus a few years, again people would argue with me and say that it's cost competitive now. Others would say I'm being a little bit more aggressive, but I think 2020 is a reasonable benchmark for that break even point.

Chris Nelder: Well I'm glad you mentioned the fact that utility planning horizons come into play here because well I watched the utility sector pretty closely, and I do not see anybody talking about procuring solar and storage as an alternative to gas peaking plants right now. 2020 is right around the corner.

Paul Denholm: It is, well there's a couple reasons for that. One is, most utilities aren't procuring much of anything for peaking capacity. We're pretty much set for capacity right now in much of the country. So we're going to have to wait for some of these older plants to retire, and they do, I mean power plants wear out so there's kind of a continuous retirement cycle and it will take a few more years of retirements before we start really needing new capacity. And yeah if you're doing the math right now I'd say in much of the country if I was doing the pure economics on storage versus new combustion turbine, a new combustion turbine would probably win most of the time. Again there'd be some spots you know in Southern California or in urban areas where the stored solution would probably beat out the combustion turbine. But if I'm in a place like right where I am right now in Colorado, I'd probably go with a combustion turbine solution. But again let's just keep watching these prices of storage come down. And the other thing that we have to do is we have to get these utilities comfortable with storage. Utilities are conservative and that's a good thing, because we depend on them for reliable electricity, so we have to wait till a few more of these first movers, these storage plants like the one that has been installed in Aliso Canyon, the new plants in Australia, we need to let a few of those plants get a couple of years under their belt, that data will get disseminated, utilities talk to each other when they meet up at conferences and trade shows and just general meetings. That information will go out, the folks will talk to each other, they'll get comfortable with the storage technology and all of that will converge. It's just the combination of the need, the price, and the comfortableness with the technology.

Chris Nelder: So you know I wasn't planning to actually go down this rabbit hole but now that you mention it, I really want to just take one minute to talk about LCOE data because you know this latest Lazard report generated an unusual amount of attention just because it finally said look renewables have now won. They're beating gas, they're beating everything in the market now, they're beating coal. And you know that's good news and I'm glad the folks are finally realizing that but when I looked at the LCOE data pretty closely as I always do when the Lazard report comes out, once again I'm seeing the official low end figures for utility scale PV plant you know in the whatever it was I don't remember now 45 dollars to 55 dollars kind of a megawatt hour range, something like that. But when you look at the actual PPAs that are being struck around the world not so much in the U.S., but you know Chile, Abu Dhabi, Saudi Arabia, I think a few other places, we're seeing PPAs being struck at about half that low end range of the Lazard data. You know and 25 dollar or 20 dollar megawatt hour range. So, what's your sense of what's going on here. Why is it that the Lazard data or anybodies LCOE data, I don't mean to just pick on them, lags so much behind the actual PPA prices?

Paul Denholm: Well we do have a little bit careful here. Again I hate to keep coming back to this financing question but yeah there's the technical reasons and there's all the non-technical reasons. So if we're talking like a place like Chile, Chile has got amazing solar resource. I mean the U.S. has a really really really good solar resource, but Chile is one of the few places in the world that actually has a better solar resource than the U.S., it's truly amazing. It's a high high dry desert, the Atacama Desert. You know it's famous for like no recorded rainfall ever type of thing. So you can count for a little bit of that in Chile for that reason. But the big function of the other difference in costs between the PPAs is that we're seeing in the U.S. and the PPAs we're seeing in other countries is often financing. Local governments giving access to cheap money, subsidized land and other subsidies. The other challenge of course when we look at some of those PPAs is whether or not the escalation factors or other things, so sometimes just to get the headlines, somebody will offer a PPA that will be really really low actually below cost and then will have an escalation factor over 20 years to recover that cost. So it's messy.

Chris Nelder: Yeah, well I mean just let me restate the question then. So from a sort of policy guidance standpoint, like if you're trying to really usefully inform utility procurement or even you know regulators looking at what are the trends and what kind of things should they be paying attention to, is the Lazard data really giving them good guidance?

Paul Denholm: You know I haven't dug into it enough to really comment on that. I'd say that for solar, the tracking the sun data that comes out of Lawrence Berkeley National Lab, you know it'll give a wider snapshot of the data. It'll show you know the longer term trends but it also slow the variability across regions and across things. So you know if you're doing a planning study you're not going to rely on a single data source, you're going to do the math yourself. But I'd just say that useful as our data is one data point and then also compare it to the other data sets that are out there from like I said Lawrence Berkeley on the tracking the Sun and also the wind markets report. Both those studies do a really good job of summarizing what's going on in the U.S., so I would tend to rely on those datasets, you know shout out to my brethren at the other national labs as well.

Chris Nelder: Of course. Ok. Alright well I don't want to get too far down the rabbit hole but I did want to raise that question. Alright, so back to your most recent paper on coupled solar and storage systems, which is actually just published in August this year, you did a cost benefit analysis of the various ways of coupling the systems, and that looked like a pretty technical paper and I confess that you know I've checked it out but I didn't really study it in depth, but the main conclusion that I took away from it was basically that the net benefits are greatest when the solar PV and the battery systems are coupled on the DC side of the inverter, and that those benefits are greater than for a PV system alone. Is that about right?

Paul Denholm: Yep. So it's basically taking advantage of the shared inverter. It' basically that reduction in cost exceeds the benefits of PV alone because of the additional direction costs of shared inverter plus all the additional benefits that you can get from adding storage to the system.

Chris Nelder: Okay. So again what do we do with this in terms of making it useful to policymakers? Is that basically the main idea is just go ahead and make sure that your solar and PV are coupled on the DC side of the inverter, is that the main takeaway?

Paul Denholm: Well I think part of it is and one of the things that if you talk to anybody in the electricity storage or energy storage orbit sphere is that they want to make sure that their technologies are appropriately valued. So a big part of my job is to figure out what kind of performance metrics and valuation methods do we need to make sure that storage gets its fair shake. And the challenge has always been in the storage community. And this gets into this whole idea of stacked benefits is is storage going to be fairly compared to a conventional combustion turbine or another type of gas plant or even photovoltaics. And what the storage industry and storage folks have been fighting against all these years is modeling tools and planning tools that don't actually value storage to its fullest capability. And so a little part of my job is helping the storage folks. I'm technology neutral, you know I don't really care. But I just want to make sure that all technologies kind of get their fair shake. So part of my job is to kind of look at appropriate valuation of storage, use really advance state of the art modeling tools and simulation tools that quite frankly a lot of system planners don't even have access to and do those studies, and so we can say hey look at this solar storage plant appears to make sense. Hopefully that will help system planners take a look at solar plus storage and do that valuation. m\Maybe consider those combinations that they hadn't before. Maybe just make sure that solar plus storage is appropriately evaluated so that when they do need new capacity it's just not oh another combustion turbine just boom.

Chris Nelder: Yeah, exactly. I'm glad you brought that up. Actually we discussed this issue away back in episode eight, almost two years ago, which is our first episode on storage. We had Jason Burwen of the Energy Storage Association on the show, and you know we're talking about how the opportunity for systems equipped with storage was limited because as you say the current market rules don't allow storage systems to sell multiple services and benefit from this value stacking, which would help them compensate for the fact that they cost more than just the generator alone. So for example just to kind of explore that topic briefly here, they can bid into the day ahead or hour ahead or the spot markets, or they could do power factor correction or other ancillary services and so on. But under the current market design they can't do all those things simultaneously or bid for all those things simultaneously or just sell whatever service is needed at the moment. And you know Jason and I spent some time in that episode speculating about you know when might the market rules change or when might regulators make it possible for this kind of value stacking to really work for storage, so what's your thought on that? Do you see any forthcoming regulatory actions that might change that?

Paul Denholm: Sure. The biggest one is the Federal Energy Regulatory Commission issued a NPRM or a Notice of Proposed Rulemaking, and that's kind of the big exciting action in the storage community right now. Essentially FERC has recognized that current market rules, current market designs, do not appropriately compensate storage. There have been some attempts in the past to change individual rules and I think the best example of that is a rule that came out about seven or eight years ago which compensates storage for its ability to move really fastly, or react really fastly. They basically said that hey storage, when somebody flips on a light or a big steel mill comes online really quickly and you need a big jolt of electricity right away, certain storage technologies can do that better than any other plant. They can ramp faster than a coal or combined cycle plant and they should be compensated for that. So they basically said to the independent system operators and regional transmission organizations hey you need to change the rules to compensate storage for that, and they have changed those rules. But that's just one piece. That's just one single service. So FERC is now looking at it in a more comprehensive manner to understand how can we better compensate storage for all the services that it can provide. Now it will never be able to be compensated completely just because there's kind of fundamental market issues. You gave an example of basically providing voltage support. A lot of the services have traditionally provided on the distribution side of the network and there's certainly a lot of benefits that storage can provide on the distribution side. We're probably never going to have an integrated market where all of the services that occur on the wholesale side are simultaneously compensated on the distribution side. There are some proposals for that but they're way off in the future. So so storage folks, they can definitely improve the situation by getting new market rules but folks are going to have to get used to the fact that they're never going to be able to be compensated for all of the things they can do. But the good news is, again this is this kind of 2020 timeframe thing, it looks to be that the point is that if you can get compensated for most of those we can reach that tipping point and storage can compete on just the number of mechanisms it can get compensated for in a more complete market. So that's the good news for storage.

Chris Nelder: Fantastic. I've got to say it just bugs me so much that concentrating solar thermal plant like Ivanpah in California's Mojave Desert wasn't equipped with storage that it could function as around the clock generator instead of only running when the sun was up. I mean the plant's developer BrightSource certainly had the storage technology when they built that plant and nationally I think we could have built a lot more concentrating solar plants equipped with storage, but I mean the technology's been around and the costs are known and it's all been well proved out and that's all been de-risked and well understood but we just haven't done it and I think that's simply again because there wasn't a regulatory incentive or a requirement to do so. So now that we're getting to higher penetrations of variable renewables on the grid and storage is becoming more relevant, do you think that will change? Do you think we're going to start requiring storage or at least offering some incentive that might bring CSP back into the game here.

Paul Denholm: Yeah, so the first thing is, and I don't think I'm saying anything controversial when I say this, we'll never have another CSP plant in this country built without storage. The Ivanpah plant was developed under times where storage wasn't valued and utilities under RPS requirements were simply required to acquire kilowatt hours. And they didn't really care when they happened they just had to acquire kilowatt hours. Things have changed now where utilities are more interested in okay why are those kilowatt hours being developed. And so that with a combined of the drop in the cost of PV means that CSP without storage simply is not cost competitive against PV. So we're not going to see any more CSP built without storage. So the next generation CSP plant that got built in the early stages of operation is the Crescent Dunes power plant in Nevada and that's a really interesting plant because it has several advancements, it has integrated storage, it uses a technology called Direct Storage, and there's a bunch of advantages of that, and it can provide energy for 12 hours after sunset. So it's not quite a complete around the clock but it'll meet that peak demand well into the evening. It will help keep the lights on in Las Vegas well into the evening. It won't be around the clock but that's fine at 4:00 am you know there's wind and there's other resources and things like that. So the Crescent Dunes plant along with other proposed plants that's kind of the future look out for CSP. CSP is in a fight with the costs of batteries coming down and the cost of PV coming down. CSP is in a challenging position where it needs to further reduce costs and there are a number of technologies under development here at the National Renewable Energy Lab we were working on a bunch of advanced power cycles using new heat and storage mechanisms and new power cycles and things like that to get the price of CSP again cost competitive so it can provide dispatchable energy over many hours after sunset. So stay tuned for cost reductions for CSP, but hopefully CSP will stay in the mix because it offers a bunch of other technology advantages that could be valuable to the grid as well.

Chris Nelder: Are you talking about inertia.

Paul Denholm: I am talking about inertia. Inertia, primary frequency response, all the wonderful things that we get when we've got a spinning machine. Again we are looking at how to get those services without spinning machines, and we think we know how to do that.

Chris Nelder: Yeah I've got a question queued up for about that, but we dont have to go there quite yet. So you know just kind of getting back to your earlier point about how you don't really need to co-locate a solar and storage system for those systems to provide the good benefits that you would look for or the difference between sort of being on the distribution side of the grid as opposed to the wholesale grid or whatever. How does the operator of a solar or a storage system decide how to deploy it, and which services to sell, especially since they can't devalue stacking on the storage side. And how are these various services remunerated and does that change their choices?

Paul Denholm: Yeah that's a really tough question. You know it really depends. Right now a lot of the storage plants that are going in are small and they're, I don't want to use the word experimental but they're kind of the early stages, and they're providing these single services under a long term power purchase agreement. So we don't have a whole lot of market exposure for storage just yet. So I guess what I'm saying is storage developers haven't really needed to think through which services they're going to provide because the utilities are asking for specific things. Again the Aliso Canyon plant is a perfect example of this. That plant is there to provide firm capacity. That plant is there to simply make sure that when demand peaks between 4 pm and 8 pm, it can deliver energy. So it's not really thinking about okay now I sell regulation and spin during this hour and voltage support during these hours. It's simply charging in the morning and discharging it for the evening to make sure the lights stay on. As we move forward there will be need to be more of that kind of careful consideration of what services are going to need to be provided. And that's going to be a real challenge for storage providers to kind of think about how they interact with the marketplace. But again I would expect to see the results of something like this FERC to actually better better set rules so that they get a fair shake and can be appropriately compensated in the marketplace through kind of establishing of a standard marketplace for energy stores as well as other technologies.

Chris Nelder: Yeah, okay. Fair enough. And what about the regulatory context? I mean since they can participate in more markets than simply selling energy, I wonder if integrated solar and storage systems might be more desirable or valuable in a deregulated market than they would be to let's say a vertically integrated utility. Is there any sort of interesting or useful distinction there?

Paul Denholm: Yeah, and that is interesting. I mean it's funny to talk about this because markets in the U.S. for electricity have been really successful. They've made the markets more efficient. We've seen better ability to share energy over larger regions. So there's been a huge number of advantages of expanding market footprints and we continue to see that as all of the markets have expanded in size as we've introduced the energy and balance market in the West, as there's proposals to expand the Southwest Power Pool extending that across the Rocky Mountains here into Colorado and northern New Mexico. So we're seeing this continual march forward into expansion of electricity markets and you know I'm an American, I love markets, but one of the challenges of course is that yes in a completely classic vertically integrated utility it is a lot easier for them to actually stack these values and say hey a storage plant can do this this and this and since they're not burdened by this market can compensate for that and that went for this, it actually can be a little bit easier for a vertically integrated utility to realize the benefits of energy storage. That said they need to make sure that they've got the tools to appropriately plan that and you know there's been some lag among vertically integrated utilities to adopt energy storage just because again, their planning tools, their way of evaluating new generation technologies might now really value that. So, I wouldn't call it a chicken and egg but we've got to kind of competing elements here. One is that markets might not be able to compensate for all these services. On the other hand vertically integrated utilities need to actually know how to value the services in the first place.

Well first of all i'm little bit shocked to hear you think that Tom DeLay might have actually been telling us the truth about the virtues of deregulated markets. But no that's a great distinction. And actually it's kind of the inverse of what I expected, so bravo. I love it when I turn out to be totally wrong about something. Ok, so you know moving on a little bit to you know kind of the bigger question of integrating storage onto the grid. I literally laughed out loud at a slide in your presentation titled 'Do we really need stores to operate the renewable grid of the future', which I've listed in the show notes, that particular slide asks How can we possibly make the grid work with lots of variable generation?, and it offers three answers. Number one, claim it is impossible or it is possible but a lot of storage is needed. Two, build lots of renewables and see what happens. Three, perform actual science, math engineering and analysis. So now I've seen a lot of sort of back of the envelope estimates on how much storage we'll need on a highly renewable grid in the future, but of course most of them make simplifying assumptions that lead you to wildly outsized conclusions. For example they don't take into account things like demand response or load shifting or efficiency gains or larger balancing areas and distributed storage like residential behind the meter storage or solar systems combined with storage and wind farms. Instead they imagine some absurdly large need for storage that would have to be provided by some absurdly large standalone utility scale storage systems particularly if they have some ulterior motive like advocating for base load plants or casting doubt on the feasibility of energy transition at all. So what's the right way to think about how much storage we'll really need a high renewable future?

Paul Denholm: So this is a question that I talk to my colleagues about almost on a daily basis, about what is this hundred percent world look like? Or kind of the 80 to 100 percent world. What we tend to think about is kind of thinking historically, utilities used to worry about 10 to 20 percent electricity from renewables, and there's a bunch of kind of anecdotal conjectural pieces from 15 years ago about OK well this wind thing it might work maybe we can do 10 percent but after that it is just not going to work. And utilities integration studies.

Chris Nelder: Oh I remember those studies. The grid was going to fall over when we got 20 percent renewables, or whatever.

Paul Denholm: Yep. So utilities, really to their credit pioneered along with the Department of Energy, NREL and other entities pioneered a series of studies that looked at what happens at 10, 20 and increasing penetration, and nothing really bad happened at 10 to 20 percent. You had to do a couple things. And then we moved on 35 percent, 50 percent. And then in 2012 we did the renewable exergy future study where we looked at 80 percent renewables. And basically what happens at 80 percent as you start running into these seasonal supply issues. And when people talk about needing these like massive amounts of storage it's really because the amount of wind production peaks in the spring, solar production peaks in early summer and the demand for electricity peaks in kind of the mid to late summer. And so if you kind of look at that you might think well gosh we need to really shift all this energy from spring to late summer and that's going to require this incredible amount of storage. Well the problem with that is yeah we don't really understand how patterns of demand for electricity will change over time. We don't understand how market forces will change all this. And the other thing which is probably a kind of a little bit of a pet peeve of mine is this kind of fixation with 100 percent. It is a highly non-linear relationship between how hard it is to do something and how close to 100 percent we get. We know that 80 percent you know it looks like we kind of know how to do this, but beyond 80 percent it starts getting harder and it really is going to come down to OK what we go from 80 to 81 and 82, it keeps getting harder. And maybe by the time we get to 96 percent it's gotten to the point where it's like you know what, there's better ways to spend our money if we're talking about decarbonizing the electricity system or increasing sustainability. Maybe we need to focus on making sure we get that last little bit of carbon out of the heating demand or aviation or transportation or something like that. So I think there may be a little bit of an unhealthy fixation on 100 percent. Maybe 96.38 is good enough, I don't really know but you need to be really careful because you can get some very very strange numbers if you don't look at the pathway of how we get there and realizing that you know there are markets involved. If you just assume that we just keep using the same technologies that we've used over the last hundred years you're going to get a pretty strange answer.

Chris Nelder: In particular recently I have seen nuclear advocates arguing that that under an 80/20 kind of a solution, that last 20 percent really would have to be provided by nuclear and they round up all their reasons for that, but I just struggle with this because I know how much things change now especially during this transition from year to year. It's incredible the amount of change that goes on and how models that were even just developed recently are now just sort of obsolete. And I wonder you know what do we do about that? And how can we project how much storage we'll need in a higher RE future, or is that even the wrong idea? Should we just not even try to project that?

Paul Denholm: I think it's useful. I mean we want to provide guidance from an R&D and a development standpoint. I mean a big part of what I do is provide feedback to the Department of Energy in terms of what technologies should we be exploring and it does look like if we had a good useful seasonal storage technology it would solve a lot of problems. If we had really cheap hydrogen storage or power to gas or seasonal thermal storage it would really make the problem easier. So it's worth investing the money for the already useful for us to exploit the technologies because if we do need those technologies they would become available. So I don't know that these projections are useless, we just have to be really careful saying we can't do it because of this without making sure that we understand what do we mean when we say we can't do something. And that's really what it comes down to, because when we say you can only integrate a certain amount of renewables onto the grid there is no technical limit to the amount of renewables we can put on the grid. It just comes down to cost. At what point is just throwing more wind and solar onto the grid and not doing anything to compensate, at what point does that become too expensive? That's a relatively low number, but it turns out that gosh all you have to do is improve forecasting. All you have to do is create mechanisms so that you can buy and sell energy to your neighbors. All you have to do is improve rate structure so they appropriately charge electricity for when it costs more or less. You those things and all of a sudden it becomes possible to integrate 35, 50, 60, 70 percent, and the whole idea is at each step, at each point where it starts to get expensive, at each start where you start to find technical challenges, you figure out which one of those technical challenges can be solved with which technologies, how expensive are they, you find the cheapest one and if those technologies don't exist you need to tweak them a little bit then that's when you do the engineering and the R & D and you solve the problem. It's just engineering and we've got really good engineers all around the world that are working on this problem.

Chris Nelder: Right, but you know from a sort of a system planning standpoint or a procurement or policy standpoint you know I think there's still a lot of people that think about grid assets, generators especially that are going to have a 20, 30 year lifespan at least. And they want to be planning for those things well in advance and they want to know that there's going to be you know enough need for those assets you know that they're going to be useful throughout that life. But you know when I look at like this question of do we need to start planning for big plants, i'm beginning to think now that we might be out of the era of big power plants period because of all this uncertainty, because things are changing so much that maybe instead of building anything that's a gigawatt or bigger we ought to just be moving toward more of an incremental procurement approach at the utilities and buying assets as they become clearly needed on the grid and not because of some projection of future demand which by the way we've got about a decade of failed projections of future demand. You know demand's been flat for about 10 years now or even declining and utilities are still putting out these projections that justify their future procurements of increased growth, you know starting next year. So I mean would you subscribe to that idea that maybe we're out of the era of building big power plants?

Paul Denholm: Yes and No. I was trained classically as a physicist before I became an electrical engineer so I'm kind of a natural affinity for all things nuclear just because of my training. So this isn't really coming from any particular anti-nuclear stance, but it just turns out that the economy of scale rules for things like storage don't really apply on the deployment side. I mean the economy scale is critically important on the manufacturing side. The only reason why we've got cheap solar, wind and increasingly cheap solar is because we make really big factories that can just crank these things out by the gigawatts. But when you actually deploy them you know you can deploy a 10 megawatt solar plant, and a 10 megawatt solar plant you know isn't really that much more expensive than a hundred megawatt solar power. There are definitely economies of scale for solar and wind but they're not on the gigawatt size. I mean when you buy a nuclear power plant your buying an AP1000 or something that's a gigawatt, whereas with solar you can do a 10 or 100 megawatt plant when you can do 100 megawatt plant and the same way with storage. So I wouldn't call it luck but it's just the nature of the scalability of renewables that it turns out that they can be deployed in these kind of smaller increments as needed and not suffer from limited economy of scale. So it's really just the nature of the technology. And again, the nice thing about storage is it can be installed in kind of these modular units and can be sited in places where it's really needed. Another advantage of storage, you can put it in an urban area, you can put it in a transmission constrained area as needed.

Chris Nelder: And you can add to it as the demand becomes necessary or blindingly obvious. You don't have to anticipate that decade in advance.

Paul Denholm: That's right. That's right. The time to site these plants and build them is considerably less.

Chris Nelder: Yeah, well you know speaking of simplistic storage models, I've seen some pretty wild claims about how much seasonal storage we'll need in a high RE future, like enough storage to meet 100 percent of demand for six weeks for the entire grid, or some absurd like that. What I haven't seen yet is a good sophisticated model of exactly how much seasonal storage we would be likely to need on a well managed grid with all the aforementioned non storage options in play, right. And so I wonder you know what's your sense of how much seasonal storage we might actually need on a high RE grid?

Paul Denholm: I don't know.

Chris Nelder: That's a good answer.

Paul Denholm: Look we are a considerable ways away from achieving those levels of penetration. I'd say that in a sense, not to put this too snarkily, right now we've got more important things that we're working on. We're working on making sure that we've got all of our i's dotted and our t's crossed on achieving 30 and 40 and 50 percent renewables. There are a lot of questions about how the markets will be structured, you know making sure that we've got reliability. There's a bunch of just detailed electrical engineering stuff that we're primarily focused on in the near term. We'll get to that question, the collective we, and there are certainly people working on this. I think the Europeans are a little bit ahead of us on some of these seasonal storage issues, but we'll get there, But we need to make sure that we've really nailed down 50 percent renewables and make sure these last little engineering questions have been solved because the last thing we need is to get to those level of penetration and you know we forgot something. So a lot of our time is just running really really detailed engineering models to make sure that we've got these things nailed down and then we'll move on and we'll figure that question. You know come back to me in a year or two and i'll give you a better answer than I don't know.

Chris Nelder: That's a perfectly good answer actually because again I just feel like it's one of these things where why should we have to project how much seasonal storage we're going to need decades in advance? Why can't we just wait until it shows up, but you know your point about the need for developing the markets and allowing them to evolve as well is an important one. And you know this issue of I think it's been generally been called value deflation now, although it's probably more appropriate to call it price deflation where you get more and more solar and wind on the grid and the price falls to the point where you know it's increasingly becoming difficult to even finance as wind and solar plant let alone a conventional alternative. What are your thoughts about that? Like how do you think we're going to deal with this issue of price deflation in the future?

Paul Denholm: I think we need to expose people to the market and the market will solve a lot of these problems. Once you start exposing consumers, not just industrial and large commercial customers but everybody to some kind of flavor of time varying prices. And a lot of people disagree with me. I've had very smart people tell me that you're never gonna expose residential customers to real time prices and they make a lot of compelling arguments. But with an increasingly kind of you know iPhone and you know tech savvy communities, we will get to the point where you can expose people to some kind of time varying prices that are a little bit more sophisticated than just kind of the standard time of use rates. When you expose people to time of use prices or some kind of time varying prices, electricity demand patterns will change and that will be a big part of it. And we'll just figure it out. Again, coming back to your original statement about these models kind of projecting needing all this stuff, there is no market in these models. They just assume that the world of 30 years from now everybody will use electricity in the exact same way, and that's a true statement because basically the way these models work is you take historical electricity demand from five years ago or something and you assume that's going to be the demand pattern that's going to exist in 2050 and then you try and meet that demand with a fairly inflexible set of resources and then voila, there's your answer that says it's going to cost 12 trillion dollars to do all this. But expose people to the markets and expose those markets to the evolving technologies that could be developed and I think that a lot of these problems will solve themselves.

Chris Nelder: Interesting. I like that idea. I mean after all having negative prices is kind of a good problem to have, right. Like you would think that we would have plenty of ideas about how to stir some loads in that direction, given time.

Paul Denholm: Yeah, you know I've got a button on my washing machine that says start now, start in two hours, start in four hours, start in six hours. There's no incentive for me to hit any other buttons. I mean I do just because I live in Colorado and I think there might be some off peak wind available at 4 in the morning and I could actually use curtailed wind from my washing machine.

Chris Nelder: Yeah, but you don't even have an interesting TOU rate under Excel either.

Paul Denholm: That's right. And I don't know, if I knew that I was going to be able to wash my dishwasher and Excel would pay me to do that, i'd hit that button.

Chris Nelder: Yeah, yeah, yeah. Or even better, you might have some little black box that's negotiating with your utility on behalf of all the appliances in your house and it's aware of the rate and even if that's a time varying rate that's you know very dynamic it's able to keep up with that and do some negotiations and figure out when to turn things on and off.

Paul Denholm: That's right.

Chris Nelder: Okay. So I want to go back to this issue of how we manage the inertia of the grid as we move toward a high RE future. So in episode 55 just a few episodes ago we talked about how virtual synchronous machines would allow inverter based generators like solar plants to perform frequency regulation and other grid stabilization services, functions that in the past have always been provided by conventional generators where you've got some big rotating mass inside of the generator providing the inertia. But battery storage systems can actually respond far more quickly as you pointed out a moment ago then spinning reserve generators if we had the right ways to control and dispatch them. So do you think that innovations like virtual synchronous machines can eventually help us replace conventional generators and support a larger share of renewables on the grid?

Paul Denholm: Absolutely. That's a really easy question. I think the emergence of things like virtual inertia really represent exactly what I'm talking about when I talk about engineering solutions. You've got a problem, you need to replace synchronous machines, and you find a solution and those solutions don't even actually cost that much money. It's software, it's controls, it's intelligent people and it's all going to come together. So I think it's a beautiful engineering story that I think we'll look back in 30 years and there'll be all kinds of interesting articles written in iEEE journals about this transition from synchronous machines to inverter based machines and successes along the way and it's just a great story. As an engineer, it's kind of one of my favorite stories about how we figured out how to make this all work.

Chris Nelder: Have you thought about it in kind of a blue sky way like you know just sort of letting your mind wander and fantasize about what we could do in the future? Is there anything interesting you'd want to share along that line?

Paul Denholm: I do think it just comes down to lighting the combination of good engineering, good science and markets kind of solve these problems. What grid integration of renewables comes down to is just a lot of hard work identifying all the possible ways that renewables can and cannot replace conventional generation. And when you can't, when you break the grid, when you're simulations fail, or you know when people say you can't do that it's just putting your head down and figuring out ways to do it and making sure that you know it's cost effective right. I mean at the end of the day I'm allowed to think blue sky but with a lot of constraints. You know I can think about 100 percent renewable future but I can't think that I've got an infinite amount of money to do that. And so having those constraints that's what makes the job actually really interesting and fun because if you could just say I could just build an infinite amount of wind and solar, well that's kind of boring. But I can build a certain amount of wind and solar but now I've got all these constraints of you know how I can integrate that into the power system. That's what makes the job interesting. But that's why we're all here and that's why we're gonna figure this out.

Chris Nelder: Alright, well so far we've only really talked about storage on the grid and solar a little bit, but you've actually done a lot of interesting work on the interactions of EVs with the grid as well. And since that's my wheel house, that's what I work on mainly at RMI, I can't let you go without asking you at least one question about that. So, some people are quite skeptical that electrifying personal vehicles is even a good idea and they're not convinced for example that it would reduce overall net emissions or maybe they've seen some research suggesting that the environmental footprint of mining lithium outweighs the benefits of EVs or maybe they think that maybe there might be a limit to the production of rare earth metals that will somehow limit how many EVs we can produce at an acceptable cost and so on. So given all the research that you've done on EVs, what's your main conclusion about whether EVs are a good idea for society as a whole?

Paul Denholm: I like electric vehicles. We ultimately need a solution to decarbonising the transportation sector. And given the challenges of production of biofuels, given the challenges of hydrogen production, I like battery electric vehicles as a flexible source to charge during overnight periods of high wind, charge during the middle of the day during periods of over generation and solve the duck curve problem, I think there's just kind of a natural synergy between battery electric vehicles and some of these supply demand mismatches. So I'm a fan. Don't have one yet. I don't quite drive enough to really justify electric vehicle but I enjoy watching the prices come down, I enjoy seeing the projections and I enjoy thinking about electric vehicles as being potentially a significant part of the solution to integrating large amounts of renewables onto the grid.

Chris Nelder: Are you at all worried about this issue of kind of the relative environmental footprint of emissions or on the lithium production side or any of that?

Paul Denholm: So I'm not an expert on lifecycle assessment in terms of mining. I do know that from some of the early work that was done, basically saying if you charge with off peak coal generation that will actually increase the carbon footprint. And I think those numbers are probably right but that's not really what we're talking about in the future. We're not really talking about charging with off peak coal, we're talking about charging with off peak wind or off peak solar. So when you start charging with off peak wind and solar it basically becomes a no brainer.

Chris Nelder: Yeah. Well, fortunately our very next episode, the one right after yours is an interview with your colleague Garvin Heath about LCA, so we'll dive right into that topic in that episode. Well this has really been a pleasure Paul. That hour zipped right by, so thank you so much for taking the time to be on the show. I appreciate it.

Paul Denholm: Thank you. Happy to chat.