All Episodes

[Episode #2] – Limits on the Grid – Part 1


What the modeling work of our national renewable energy lab tells us about how far renewables can go on the grid under various scenarios, and their real technical limits.

Guest: Mackay Miller is Principal Analyst in the US Strategy division of National Grid.

On Twitter: @mackaymiller

Recording date: August 19, 2015

Air date: September 30, 2015

Geek rating: 7

Chris Nelder: So let's bring Mackay into the conversation now and explore some of these questions, welcome Mackay.

Mackay Miller: Thanks for having me, nice to speak with you.

Chris Nelder: So Mckay, why don't you go ahead and introduce yourself and tell us what you do over there at NREL.

Mackay Miller: So my name is Mckay Miller, I'm a senior analyst at the National Renewable Energy Lab, NREL since 2010. I should say that I'm speaking in a personal capacity today and any errors are mine, and I'm surrounded by very intelligent smart colleagues in general. Happy to work there.

Chris Nelder: Defenders of the status quo have claimed for decades that the grid can't support more than 3 percent or 5 percent or 20 percent, it's a moving target, of variable renewables without risking blackouts and insisted that so-called baseload power generators which run nearly full time will always be needed to maintain reliability. They've also claimed that because wind and solar are variable, they might have to be backed up by enough baseload power to meet 100 percent of demand. Is that right or is that wrong?

Mackay Miller: Well there are several different claims in there, so we can unpack each of them separately. I think that the natural progression over time of wondering what fraction of renewable energy would be feasible in the system is sort of a natural thing for power system engineers to wonder about. Indeed 20 years ago, even as recently as 10 years ago there was a lot of trepidation around sort of a five per cent limit. I think that that has been a moving target that has been pushed back by a couple of different things. One is just system operators successfully managing their grids with far more than 5 percent. And now in the contiguous United States we've got some systems operating at near 15 percent and around the world we've got systems operating at or above 20 percent wind penetration. That's I think been the dominant argument against these conceptions that we've got some sort of fixed limit out there. The other force that has really helped people think through what these limits might be I think is is the advance in modeling and simulation of power systems that allows us to look at very high renewable energy penetrations in greater detail than we used to be able to do. This comes both from higher quality datasets both in terms of the temporal resolution of wind and solar data that we have for the United States and other countries, and also the spatial resolution has gotten much better so we can now see average annual solar and wind resource at very fine geographic resolution but we can also in many cases have a good sense of what that specific resource is likely to be doing for every four or eight minutes of the year. When you've got that level of granularity of data then you can really start to test how the system is functioning on its normal time scales instead of just making some simplifications about, well we we think that most days the grid operated just fine. Well now you can start to look at you know, is the grid actually meeting reliability constraints at all moments of the year. So a lot of the modeling that the National Renewable Energy Lab has been doing and and others in the field like Enernex and GE Energy with many stakeholders has been to sort of push the envelope on how sophisticated this modeling can get so that we can probe future renewable energy penetrations with some level of credibility. And I think there's been just a lot of brilliant people throwing themselves against this problem now for about 20 years and we've made pretty impressive strides. And so I think that's helped move that target back from those very low levels where people were thinking 20 years ago.

Chris Nelder: Yeah absolutely. So what about the claim though that baseload power is going to have to be there to meet 100 percent of demand?

Mackay Miller: Well it's so far off I think from this question that that we're dealing with today. Typically at NREL we're working with system operators who are still dominated by thermal and conventional hydro power. So on the one hand we don't spend a lot of time worrying about zero percent baseload power it's just not a near-term problem. If you do really think all the way out to what that grid might look like in 2050 or 2060 then you do start to wonder what the role is for different generator types. Now there is, you know, some physics that needs to be taken into consideration here. When you are operating an alternating current grid you know the overall system really does depend on a level of inertia that all the, all the spinning turbines are contributing to that system. So as inertia drops out of the system, and of course wind power provides fairly weak inertia relative to its conventional counterparts and solar power provides essentially zero inertia to the system. You do start to see in a conventional AC system that there is a role for for very large spinning generators. Now those could be coming of course from renewable sources whether they be hydro or geothermal or biomass or even concentrated solar power. So you could imagine a system that has this sort of spinning generator conventional characteristic but is fueled not from fossil fuels. Were the costs of those technologies to come down within reason then of course it would be quite simple to see a 100 percent renewable energy system. Really though,the wildcards there are cost of course. And the question that we've been looking at more specifically in recent years which is, what are the technical and economic limits of the two most economically available technologies.

Chris Nelder: Right. So that brings up the important question of the economic balance versus the tactical balance. I hear most proponents of renewable energy talking about a grid that is 100 percent powered by renewables, but I wonder if that's really the right goal. I mean put another way, why did the RE Futures study focus on 80 percent renewables and only modeled 90 percent at the high-end instead of 100 percent?

Mackay Miller: Right. The renewable energy future study was actually, and this is a little bit anecdotal, but the origin of that study, it was really supposed to have a sibling study which was looking at very high penetrations of nuclear in the system. And so it was meant almost as a portfolio of studies to look at dramatically different electricity sector futures in the United States.

Chris Nelder: So there was a little segment set aside there for nuclear in that conception of 80 percent or 90 percent?

Mackay Miller: Well it was actually intended I believe to be an entirely different study, so that there could have been renewable electricity future study and the nuclear futures study. So the Department of Energy and the US is really tasked was thinking quite broadly about different pathways for the future, the quadrennial energy review that just came out I think is a testament to that. They of course look across all technologies, and grid technologies as well. But the 100 percent renewable energy futures idea was not really, I don't think, in common currency at that time when the REF study was being designed. At that time like you mentioned earlier. Really the question is Where can we get past 20 percent. And so as a lot of the studies that the National Renewable Energy Lab and others have done over the past 10 years have really been directed at getting past those what we now know are artificially low limits and try to really do high quality engineering studies to explore what the likely technical and economic limits of the grid are. If you go into the REF study it's not just an 80 percent target. There's actually quite a few sensitivities around looking at 7 percent or 90 percent different mixes of base sort of dispatch base baseload renewables whether that's biomass or geothermal looking at highly constrained transmission build outs, looking at slow cost trajectories fast cost reduction trajectory. So really the power of that study is not just in suggesting that yes an 80 percent renewables system is possible but also the power of that study I think is exploring a lot of these sensitivities and looking at how pushing one lever might affect the entire system's outcome.

Chris Nelder: Right. So the whole concept of trying to model a hundred percent renewable system wasn't really on the table there.

Mackay Miller: It wasn't on the table, and I also think that you know the 100 per cent studies that have come out have not been modeled at the level of rigor that the REF study was modeled at, so we use off the shelf industry standard production cost model Grid View to do that study which really is widely used in the utility and ISO sector. It allows for sub-hourly modeling of the entire system for every hour of the year and sort of gives you that at least credible level of confidence level around supply and demand being balanced at all parts of the year. I'm not familiar enough with the 100 percent work of Jacobson and others to know whether they've they've sort of done it in all that detail but I'm quite certain that they haven't pulled in the data that we pulled in for REF or done the level of modeling or certainly the sensitivity analysis that would be sort of standard if you've got a technical review committee of utilities and system operators looking over your shoulder sort of assessing the work. And that last part is really a standard operating procedure at NREL is to put a technical review committee that's large and diverse, stand it up for each project to make sure that these are credible studies.

Chris Nelder: And very rigorous. Yeah, I think that's right. I don't think the Jacobson or Delucchi studies have come down to the resolution level that you guys have. So I wonder what are the real technical limits if any to 100 percent renewable grid. I mean only about 13 percent of US grid power is currently provided by renewables.

Mackay Miller: Variable renewables.

Chris Nelder: You're right, and about 7 percent by non-hydro renewables. So how much confidence can we have today given the limits of our current modeling about what the real technical limits might be in the future?

Mackay Miller: Right, well probably worth just defining our terms a bit. Iceland will have quite an easy time getting to 100 percent renewable grid because they've got this sort of wondrous geothermal resource and geothermal just behaves nicely in the power system, or a lot more like the power systems wWe've all grown up to expect.

Chris Nelder: Yeah, it's essentially baseload power.

Mackay Miller: Yeah. So I don't think there's much question that if you have that sort of endowment a 100 percent renewable system is possible, but 99 percent of the world doesn't have that endowment. We've got some mix of various resources and so I think the real question facing the U.S. but also other countries is you know what's the optimal mix of renewables in our country, if it turns out that the mix happens to be pretty light on that sort of dispatchable renewable energy then we'll have to lean a bit harder on the variable wind and solar. Lucky us that happens to be pretty cheap right now and costs just continue to fall, but we will in all likelihood have to operate our grid differently, plan it differently, think about transmission and generation capacity additions in a different way. And so in the U.S. you know the results that we've looked at in the Renewable Electricity Future Study are built upon subnational assessments of what those resource mixes are, and so if you get into that study the United States is sort of divided up into several grid regions and it really looks at how much biomass is there and how much wind and solar and how much hydro and geothermal can we get. And there's a capacity expansion planning part of that modeling problem which really looks at what are the economical options in each region, and we put it all together and of course the outcome there is that of that 80 percent a very large chunk of it is dispatchable renewables in the REF study. So you're looking at a lot of biomass, a lot of geothermal, a lot of small hydro. So when you get down to the actual fraction that is wind and solar, it's in the 40 to 45 percent of most of those scenario runs. So you're not seeing wind and solar providing you know 80 percent of the power, it's actually a very large fraction of this let's call it well-behaved renewable energy.

Chris Nelder: Right, so if you're taking that kind of an approach where you're trying to figure out what's really practical and you're doing a bottom up analysis based on that, then maybe the question isn't so much how much confidence we can have about the technical limits but how far can we build this up and be relatively confident about how this system behaves. Is that fair to say?

Mackay Miller: Yeah and luckily this is power system evolution. It's not a basketball game where things move really quickly. People sometimes I think forget that this stuff happens on decadal time scales, not over the course of 10 minutes. And there is a professional class of engineers and regulators and policy makers who you know apply technical rigor to these decisions and economic cost benefit assessment. And a lot of engineering judgment. So the good thing we've got going for us is that I don't think we're going to wake up one day and find that we've built a system that just can't keep the lights on. These things tend to, for better for worse, they evolve pretty methodically.

Chris Nelder: Well, I think for most people that would actually be a source of comfort to know that we're not just plunging headlong into some unknown future and seeing how it works out you know.

Mackay Miller: That's a little bit of a hype cycle that gets maybe put up in the blog post and in the media.

Chris Nelder: Yeah. Well absolutely. Got to get those clicks. So the RE Futures study used cost estimates that were available in 2010. And that's actually quite old data now considering how rapidly the costs of wind and solar have fallen. If the RE Futures study were redone using 2015 cost data I mean how might its conclusions change or maybe the right question is what is the actual long term cost of energy transition on the grid likely to be, and how much confidence can we have about today's long term cost projections given how rapidly things are changing?

Mackay Miller: Yeah, I do think if you ran that study again today you would find markedly different results, either in perhaps a quicker timeline to the same penetrations or just a cheaper overall build out. But certainly the cost question is something that's been front of mind for Department of Energy. They've essentially commissioned NREL over the past year and a half to launch the first edition of something called the Annual Technology Baseline and this will be an annual product but just came out about a month and a half ago, it provides 2014 cost estimates for the full gamut of renewable and conventional technologies. And on the renewable side also provides peer-reviewed sort of vetted cost estimates with error bars out to 2050 by different resource class areas in the US, so really looking at you know how much a wind turbine will cost in a Class 1 area versus a Class 5 area. So this is all now in the public domain. Its sort of an energy modelers Wonderland. Just a tremendous resource if you're looking for both credible and highly detailed cost estimate data. And so really while we haven't redone the REF study, I think perhaps DOE and NREL have done something slightly more important which is provide what we think is a really strong basis for other people to do good modeling and to use really robust data. Just to go into one of the data types, I think its pretty interesting to note that if you look at wind for example in this annual technology baseline, you dont see dramatic cost declines over the next 30-40 years for most of the wind resource types you're looking at fairly flat costs from a dollar per megawatt capital cost point of view. What you are seeing though is that for that same dollar you're just getting pretty steadily increasing capacity factors all the way out through 2050.

Chris Nelder: So actually you've got a flat cost per megawatt but a falling cost per megawatt hour.

Mackay Miller: Right. So you're getting more for your money but you're also finding that in what is today a relatively low quality wind resource region, 10 years from now, 20 years from now is going to be a very decent wind resource region both because wind turbines have gotten taller, the blades have gotten longer, everything has gotten slightly easier to transport perhaps because it's been lightweighted a bit.

Chris Nelder: Right, and reliability seems to have improved as well.

Mackay Miller: Yeah, that's right. So so much bigger swaths of the country at that point are potentially a sort of investment grade wind farm site where today they aren't. So that's one of the unintuitive things that has come out of this annual technology baseline is that it's not all about cost, it's really about this sort of mix of cost and capacity factor and access to resources around the country.

Chris Nelder: Well I'm glad to hear that that new data has become available because you know frankly one of the things that bugged me when I was plumbing the NREL data that was available some years ago on the LCOE, that's the levelized cost of energy studies from technology to technology. Some of that data was just really really old. And so these error bars for solar and wind especially were just huge, and none of them corresponded to the actual prices that were being cut for power purchase agreements, not in the field today. And so if you just look at those LCOE charts you go, 'wow you know nuclear is cheaper than solar.' But then when you look at the actual cost of a new nuclear plant being built today in say the U.K. versus a new wind farm, wind was way cheaper. And so it's really important I think for us to have more accurate, more current more nuanced data like that that we can look at.

Mackay Miller: Yeah. I think that those are really important points. I wish this dataset maybe got a little bit more into that because just on the nuclear side I'm not sure how hard you could push on these cost estimates in this annual technology baseline for nuclear.

Chris Nelder: Well we haven't built that many nuclear plants in recent years so who knows what the real costs are. In a lot of cases we have to just take developer estimates and those are notoriously historically wrong.

Mackay Miller: Right as Michael Liebreich has said, the nuclear industry has a really big challenge ahead of it which is to get serious about what it really costs and to help policymakers really understand. I think there's a rational suspicion of what nuclear truly costs that that the industry could do a lot to help resolve. Of course it would be helpful if 100 of the things were under construction and we could have empirical data to look at. But very hard to push on. I have heard anecdotally that commonly in the U.S. a wind farm developer is going to go to a bank with a proposal to do a wind farm and they like all power plant developers have to have a loss reserve set aside for maybe construction overruns or any sort of project delays.

Chris Nelder: Acts Of God, et cetera.

Mackay Miller: Yes I've heard that that is typically below 5 percent of the total project cost and even you know sort of in the 3 percent range of total construction cost for wind farms.

Chris Nelder: That's a really small margin.

Mackay Miller: Very small margin because these are not complex projects. At the end of the day yeah it takes a lot of cranes and whatnot but it's not building a nuke plant.They're kind of turnkey a little bit. I have heard that you know also anecdotally the same loss reserve margin, cost overruns et cetera for a nuclear power plant is typically in the 50 percent range.

Chris Nelder: Right. And I would venture that that's probably understated because that's assuming that you've got federal backstopping for all sorts of risk.

Mackay Miller: Yeah, I've gotten in some debates on this question about whether those insurance backstops are really a subsidy. And maybe you know more than I do but at least some of the folks I've talked to have said that most of that backstop fund is paid into by the nuclear developers and private companies. I'm not sure I haven't gone out and done my homework.

Chris Nelder: Well actually there's a whole sequence of different sorts of risks right there, so you can have the risk of nonperformance, you can have the risk of construction delays, you can have the risk of you know obviously some sort of a meltdown or something like that. And then of course there's all the risk involved in actually managing the waste after the machines are shut down and all of that is federally backstopped as well so...

Mackay Miller: The decommissioning of San Onofre is supposed to be like $4.4 billion. Do you know where that cost falls like who's shouldering that? Was that built into the rate case zero when it...

Chris Nelder: Well nominally as I understand it, it is but there's also an ongoing contentious debate about whether or not the money that's been set aside is sufficient and if it isn't who winds up holding the bag for that. Let's not go too far down that rabbit hole today.

Mackay Miller: Costs are complicated.

Chris Nelder: But just kind of taking us back on track here. You know the RE Futures study only looked at technologies that were really commercially viable in 2010 as well as the cost, so it left out some of the things we talked about earlier. Next generation technologies like enhanced geothermal, marine energy, floating offshore wind turbines. And of course that was a very defensible choice for a study that wants to be empirically rigorous but being of a scientific mind I don't suppose you want to speculate about that too much but just suppose 20 years from now which would of course be well before the 2050 scenario is contemplated in the RE Future study that enhanced geothermal and some marine technologies had reached the point of commercial grid cost parity because those technologies could provide 24/7 power with very little variability as we said earlier basically functioning as baseload. How might they change the outcomes of that study if they were included?

Mackay Miller: Yeah, that would be a dramatic change, capped really only by the technical availability of those resources. For every renewable resource there is a technically recoverable potential you know how much is actually out there and then there's the economically recoverable potential. Much like in the oil and gas fields so the total likely recoverable geothermal in the U.S. if memory serves would not be enough to service all demand in the country but it is pretty large I think it's in the sort of 200 gigawatt range. The U.S. has roughly a 1,000 gigawatt system although we don't ever have a peak demand of a thousand gigawatts.

Chris Nelder: Now there I assume you're talking about more or less conventional geothermal or the kind of thing that for example has been deployed at the geysers in California and not this sort of what MIT wrote up some years ago is this sort of universal geothermal where if you drill deep enough anywhere in the Earth's crust you're going to get usable amounts of heat.

Mackay Miller: I don't know what the academic consensus is on if that sort of happened what would the recoverable potential be. What would this look like? This is like fracking in everybody's backyard for baseload power or something.

Chris Nelder: Well I mean again it's hard to say, like oil and gas it only exists when you prove it to the drill bit. Right. And we haven't actually done the drilling with universal geothermal to know. That whole concept I think was sort of stillborn when the risk of earthquakes associated with it arose.

Mackay Miller: Which is funny in itself. Earthquakes can kill one industry but others are impervious.

Chris Nelder: Right. Right.

Mackay Miller: It would be transformative if enhanced geothermal came down to that cost. I mean it's really that's a big deal. It just allows for pushing coal out of the system quite quickly. You know that's sort of one of the things that comes out of all of our studies is that you can really decrease the amount of coal and natural gas that you use in the system using wind and solar. Right. You can save a lot of fuel essentially as you go past 20 and 30 and 40 percent penetrations, you're just running coal and gas units a lot less. And that has a positive CO2 and SOx and NOx impact.

Chris Nelder: I think also the potential implication at least in my mind is that these technologies that could provide 24/7 power from renewables, whether it's marine or geothermal or whatever has the potential to basically replace all of the existing baseload power.

Mackay Miller: It gets you that same CO2 and sort of pollution benefit but it also allows you to fully retire that unit. And that is not something that at least with current technology you can do as quickly with wind and solar. That unit needs to hang around for a couple thousand hours a year.

Chris Nelder: To provide that reliability.

Mackay Miller: Yeah there are week long stretches where it's just...

Chris Nelder: The wind is dead or...

Mackay Miller: Right. So a sort of universally available relatively low cost renewable energy baseload source would allow for those units to be retired. And you wouldn't have this sort of unanswered question around the capacity payment required for those units anymore.

Chris Nelder: Exactly, so yeah that was just the thing I wanted to clarify was that we're talking here about a study that tends to model things out to 2050. It's definitely not out of the realm of possibility that some of these other renewable technologies that could provide some sort of baseload power would in fact become commercially and technically viable. Before 2050 and so...

Mackay Miller: This study deliberately didn't comment on the feasibility of that. It was really looking at operating the system.

Mackay Miller: Right because it's unknowable and you're scientists and you're not trying to speculate about things.

Mackay Miller: Well, those are valid questions though this study didn't answer them or even ask them really. But there are very large parts of DOE and NREL trying to attack those questions of how do you push these things through the labs and out into early stage deployments and then get them into market. You know that was really my introduction into NREL and some of my early work was all of the sort of valley of death questions. Those have not gone away. We've still got very severe valley of death problems for a lot of technologies. I would say stepping back, there is an enormous endowment of relient research that the Department of Energy Lab system has generated that hasn't gotten pulled out to market and people can sort of disagree about why that is but it's there, there's just a huge endowment of patents and IP out there that I think could probably be diversifying our options if we were doing a little bit better job of unclogging that pipeline getting stuff out to market.

Chris Nelder: Yeah. And if we could get the VCs of Silicon Valley and all the rest of the investor community properly organized marching in same direction and working on these problems instead of sort of chasing the next social media doodad or whatever.

Mackay Miller: I'm not fully convinced Silicon Valley is the right audience for that. I don't know. I think we've learned a lot about them and their sort of tolerance for this type of problem in the past 10 years.

Chris Nelder: Well that's a fair point. That's a fair point. I mean who do you think might be able to step up?

Mackay Miller: I wish I had an easy answer for that. But these are 10 to 20 year bets which really disqualifies a lot of Silicon Valley and that's why you've seen them go hunting where the ducks are. They're looking at software, they're looking at storage solutions with very evident consumer pathways right. So I think you're seeing that the sort of patient capital required for pulling energy technologies out of the lab system and into the market.

Chris Nelder: Getting them over the valley of death.

Mackay Miller: Yeah, it's really a little bit of an unsolved problem. I think there's some hope of course that federal and state governments in league with utilities and the private sector could pull together to put in place some sort of test bed ideas that could help do that. And there's of course hope globally that some of the more command and control economies around the world could help pull some of this stuff out of the out of the lab. But to my view I haven't seen anyone crack that code yet.

Chris Nelder: Well personally one of my great hopes is that I mean obviously as the energy transition is going on we're seeing a number of different kinds of market reforms. And so I think one of my sort of distant hopes would be that we would put in place the right sort of market incentives that would incentivize utilities who can put a billion dollars down as patient capital for 20 years to make these things work. Because nobody's expecting them to turn 20 percent plus returns in a year nobody's expecting them to get a three year exit with a 5x multiple or anything like that right. So that really brings into question what the market has to do. We've kind of talked about the technical side of grid integration so far but we also know, or at least speculate based on a number of different studies that renewables bidding into the market at zero cost, the wind suddenly blows, the sun suddenly shines, all of a sudden this installed capital of solar modules and wind turbines starts pumping out power so the zero cost suddenly comes in and disrupts the wholesale spot power markets as they're currently structured. And so there's a lot of discussion about how wholesale markets will need to change as more renewables are integrated. And what economic limits if any there might be to the markets as they're currently structured or as they might be structured in the future.

Mackay Miller: I think this question is most interesting to me because very smart people disagree on it right now. I've met some people who say, look we really are headed for a situation in which we've got enough zero marginal cost energy coming onto the grid within the next 10 to 15 years you're going to start to see some problems arise. And specifically what we're talking about here I think at least if you're putting your climate lens on is that you're going to see perverse outcomes wherein the wrong generators are out of the money first and they're retiring first. And I think a lot of this sort of camp holds up Germany as Exhibit A here and says you're seeing natural gas and nuclear get pushed out and lignite hanging around. And so you're essentially getting this disorderly retirement path. I think really what the hope is that you've got more early retirement path where the dirtiest generators are leaving first. So I think there are some interesting arguments to be made there that absent some market reforms or carbon price, pushing a lot of wind and solar into the system is potentially going to have some unintended consequences. There's also another sort of school of thought that says this is a slightly different thread but in the broadest sense a market can function with a lot of zero marginal cost energy coming into it and Exhibit A here is Scandinavia. You've got 60-70 percent hydro resources and yet market clearing prices are typically at 30-40 euros per megawatt hour.

Chris Nelder: Which ought to be sufficient enough to support the conventional fossil fuel generation.

Mackay Miller: Yeah, or close. You know it's not, I don't think it would make a lot of generators happy but it's at least better than zero. And this school of thought is also saying that there are some very important nuances that energy markets are currently discovering and putting into practice. One of them for example is valuing the opportunity cost of running a generator as energy or keeping it spinning in reserve. So there are moves afoot in ERCOT in Texas to really raise the prices paid to those generators that are generating instead of hanging back as spinning reserve and really that's a move to try to at the end of the day raise the net margin for these generators who are actually in the market right.

Chris Nelder: The term of art there being that they're going to create a capacity market, they're going to pay these guys just to maintain capacity.

Mackay Miller: No it's actually this is the sort of William Hogan classic energy only market crowd saying you don't need to go capacity market. You can tweak your energy only market to better value the operating reserves through operating reserve demand curve. That's one problem historically with energy only markets is that you'll be running along just fine. Pretty low prices and then all of a sudden you're at $700 a megawatt hour. You just spike way high. The preferred market function would be a smoother ramp up there so that you're getting better market outcomes along the way. Getting the right signals sent on an investment time scale for generators to be in there but instead you've got these really steep supply curve dynamics taking place, and at the end of the day what that means is that generators are only recouping costs for 20 hours every five years. They're making all their money in a tiny, tiny window of time. And so you know the mathematicians who are the economists who are really good at this stuff have figured out ways to sort of broaden the slope of that curve a bit so that for more hours of the year these generators who were at the margin are in the money a little bit and they're they're hanging around a bit longer without having to put a full blown capacity market in place.

Chris Nelder: Interesting. So we don't really know how economic limits might come into play in terms of limiting what renewables can actually do on the grid in the future. I mean we really have a lot of market evolution to happen between here and there.

Mackay Miller: I would say that this is going to be you know if I had to place my bets now its going to be one of the biggest areas of research in the coming years. The past decade was really marked by, let's look at the technical feasibility of these high penetrations. I think increasingly people have come around and said, all right you know what 30, 40 even 50 percent variable renewables probably doable. You'll run into some problems. Those are largely solvable through known modifications to operations and planning. But there is this large landscape of questions that have gone relatively less examined and that is, what does wholesale power market behavior look like under those penetrations? Can we anticipate, not so much what dispatch looks like, I think that's fairly easy to anticipate but you know in the aggregate what are the signals, the investment horizon signals that a market is sending at that point. Those are the sort of dual functions that all these wholesale markets are asked to perform. On the one hand organize day to day operations, minute to minute operations, help optimize dispatch and then also help send appropriate signals so that people can make 30 year investments. Those are two things we're asking of wholesale markets to do that are not always easy for them to do and I think the second one, that's going to be that the trickier challenge and a big focus of research for folks coming up. There's actually a new NREL report that came out this year that does look at what the likely generator revenue impacts would be in Texas under sort of a 30 to 40 percent wind scenario. And that report is important both because it starts to give us a sense of what those impacts would be but also because it uncovers a lot of the modeling complexity that that lies underneath. It's really saying look this is not an easy problem to model here. There's just much more complexity there than modeling a power system itself.

Chris Nelder: Well yeah, and in reality the wholesale markets are currently set up to work with and recognize the value of conventional generators. Right. I mean they were not designed to recognize the value of variable renewable generation. So there's a necessary transformation here of the wholesale markets the way the structured how they're valuing what they're valuing. And I guess the takeaway for people should be that not only do we not know how those markets will evolve today such that maybe they'll be able to incentivize these other types of evolutions that we've talked about but also we can't foreclose necessarily, we can't say with any specificity what the limits of renewables might be on the growth, on an economic basis because we don't know how the market design will change.

Mackay Miller: Yeah, I think that's fair to say. And I think that people smarter than me disagree very vehemently about this. They could say that markets with only modest changes are going to be able to handle very high penetrations and then others are saying that 10 years from now we're going to be up a creek.

Chris Nelder: That's a bit of a disconcerting disconnect there isn't it.

Mackay Miller: Well it's disconcerting until you go back and read the history of market debates.

Chris Nelder: Is this just a replay of the same old debate that was happening 10 years ago or...

Mackay Miller: It's a a new verse right.

Chris Nelder: Same song right. So final question. I thought your power systems of the future study was really quite interesting laying out these kind of five main pathways by which we could see more renewables on the grid. This next generation performance based regulation, a clean restructuring pathway, a DSO pathway, a bottom up coordination pathway, and a bundled community energy planning pathway, and without taking this particular episode to go into too much detail on that, I just wondered if you could comment on you know what are the main takeaways from this study and are these different pathways mutually exclusive or might they be able to come into play at different times and places.

Mackay Miller: Yeah. The origin story this report's probably important here. NREL of course does primarily domestic research modeling of the domestic United States power system. We've also, for 20 years or more been working with our international colleagues in Europe, in Asia, and Africa, in Latin America to think about renewable energy questions. And this particular report which came out earlier in 2015 is really trying to look across our major partners and say, look we're not all starting from the same starting point. That should be obvious of course but sometimes in the sort of US centric bubble or the European centric bubble it bears repeating. So we're not starting from the same starting point and yet we're all facing these pretty interesting trends, and these trends by and large are global. So you've got a bunch of power systems starting from a different spot, but 10 trends around the world that are sort of like winds blowing and how's everyone going to set their sails to arrive at the place where they want to be. And so we sort of run through those 10 trends in some detail and then we, I think for the first time tried to lay out where we think, if you look across all the power systems in the world you can sort of think about them in roughly four starting points so you've got those markets that are still typically vertically integrated, and they haven't undergone any market restructuring, they've still got their utility operating the system from generation all the way down to distribution. The second major type is the restructured market generation has been sold off and made more competitive, and there there's typically some sort of independent transmission operator and in some cases there's also retail competition. And then you've got sort of these marginal energy access zones where you've maybe got unreliable access to electricity. Much of the developing world whether it's in South Asia or subsaharan Africa Latin America and then you've of course got sort of the rural completely disconnected world. And what we were trying to do with this report is say starting from any of those points, what are the forces that have been unleashed, you know what opportunities are they going to unleash. What pathways might we see to arrive at a much better place in relatively short time leveraging this sort of technological and business model innovation that's happening right now. So that's where we end up with what we think are sort of five pathways that could be broadly applicable around the world.

Chris Nelder: Well these are all very different markets and different environments obviously.

Mackay Miller: Yeah, and to your point, you know you asked this question to what extent might they be mixed and matched right. We had this debate a lot in writing the paper. I think there's a lot of room for mashing up and yet I do think there's these starting points and the inherent path dependency of not just energy systems but also of the institutional backdrop in countries is going to persist. I think we're going to continue to see path dependence. I'm a little cautious about ideas that falling solar prices are going to sort of unleash orders of magnitude faster economic development anywhere. I think that's maybe giving a little too much credit to what a solar panel can do. So I think that you're going to continue to see a relatively high level of path dependence. So I'm not entirely sure that these five pathways are universally applicable to everywhere.

Chris Nelder: Right. I should probably even hesitate to say this because I haven't really read the paper in depth like I like to do but you know it occurred to me for example that even in the U.S. which is probably not the scenario envisioned for this particular pathway you could have under the bottom up coordination concept communities solar grids OK which are designed to just support that community whether you're imagining a housing development or a rural community or whatever interfacing with let's say the clean restructuring pathway in which power markets are deliberately designed to encourage more clean energy integration. And so you've got a variety of pathways as it's put here coming into play sort of at the same time maybe even in unexpected ways.

Mackay Miller: Well that is I would say largely true I wouldn't disagree with that. Part of that is just a feature of the US being so crazy and having so many different setups. So I think that the REV proceeding and the CALISO stuff around EG really are signals in that direction.

Chris Nelder: I mean this is one of the things that just makes me mental, when you look at the countries that have really succeeded in energy transition, you know Germany, Denmark, a few others, you really have a much more unified market and this is something that people who don't really look at the rest of the world understand you know that it's not this massively balkanized system that we have here in the States where you've got 50 different states, more than 50 different grids in some cases where you've got all these different levels of oversight and management from FERC at the top and then you've got your ISOs and your RTOs and all these different layers of regulation and oversight and then you've got this whole massively discombobulated aggregation of generation entities from municipals to ISOs and so on. It's just enormously complicated, and if you really want to execute an energy transition on the grid you have to deal with all that complexity.

Mackay Miller: I think we must like it that way. I don't know what the answer is but this is kind of how the United States runs everything right.

Chris Nelder: We're a big fan of Dtates rights we love States rights. We are a republic after all.

Mackay Miller: I do think, I'm I don't know I'm not so pessimistic. I think the US tends to maybe pull stuff off a little bit more by the seat of our pants. But I am holding out hope that we're going to pull off some pretty interesting stuff. We are certainly not Denmark. We're certainly not Germany, and at the other end we're definitely not China and I don't know I think I'm not so pessimistic about this Balkanization. I think it's certainly got a lot of cons but there are huge pros, I think we've got an interesting legacy of innovation and institutional innovation that I think a lot of people admire. California Energy Imbalance Market to me is kind of interesting because for as long as anyone can remember the West in the U.S. has been just completely anathema to the idea of organized wholesale power markets right. And yet you know what you're seeing right now I think is folks innovating ways to get past that sort of institutional reluctance and perform many of the same valuable functions of a wholesale power market without creating a wholesale power market in sort of a virtual quasi power market in the West. That is going to reduce the cost of adding more wind and solar it's going to make dispatch more efficient and it's this sort of institutional flexibility, and institutional innovation that I think is an underexamind success story. People tend to really look at the technical innovation in Silicon Valley and all that but institutional innovation in my mind is going to be one of the critical enablers of the energy transition and our messiness makes us actually kind of an interesting laboratory for that.

Chris Nelder: That's the wonderfully optimistic note on which to wrap this up. Well listen Mackay, this has just been a really interesting conversation for me. It's such a pleasure to talk to somebody who understands this stuff as well as you do and I really appreciate you taking the time to be on the show.

Mackay Miller: Chris I wish you the best of luck and I hope that I haven't offended your listeners enough that maybe you'll have me back on sometime.

Chris Nelder: I doubt you've offended anyone. And I definitely would like to have you back on. NREL puts out so much great material and it's almost just a shame just to sort of skim along the surface of it here because all that stuff is really deep and thorough. So yeah definitely would like to do that.

Mackay Miller: I love your research too Chris. I'm really glad that that we get to see each other on Twitter so much.

Chris Nelder: Yeah yeah. Cool. All right thanks buddy.