Jigar Shah is in a race against time, his progress measured in part by a map on his office wall. Each point on the map represents a project financed by the Department of Energy’s Loan Programs Office, where Shah is the director. The 60 active LPO-financed initiatives include a factory for EV batteries or sustainable aviation fuel, a grid upgrade, and a next-gen geothermal plant.
Shah’s obscure but powerful office is now busy reviewing paperwork for billions in new projects, with the goal of accelerating clean energy, making it affordable, onshoring its supply chain of minerals and manufacturing, and creating millions of jobs, especially in struggling communities. The office has about $74 billion remaining for innovative technologies and about $60 billion for energy infrastructure reinvestment (that is, repurposing or replacing defunct energy infrastructure like old coal plants). But lending out that cash isn’t easy. It can take years to get a deal out the door for companies that apply—and those applications are no sure thing, given how many CEOs are hesitant to step outside of traditional financing and take a big government loan.
It helps that Shah is well-known in clean tech circles as a pioneer entrepreneur and investor (he founded solar developer SunEdison and financier Generate Capital) and, arguably just as importantly, as cohost of the popular podcast The Energy Gang. (He’s also a former Fast Company columnist.) Shah’s robust rolodex has helped him to connect directly with companies and investors, and to double the LPO’s staff by recruiting over 100 people from the private sector.
The office is currently managing a pipeline of 203 applications totaling $263.8 billion in loans across 48 states, according to a spokesperson, but so far their slow, painstaking work has resulted in only four loan approvals. The caution and care is intended to balance out the inherent risk that’s built into the LPO’s mission: It has a mandate from Congress to provide backing to new technologies that might otherwise struggle to find investors. It provided a vital $465 million loan in 2010 to a then-struggling Tesla, and backed some of the country’s first large solar farms more than a decade ago. Infamously, the office also lost more than a half-billion dollars on Solyndra, the solar manufacturer whose bankruptcy and resulting Congressional investigations cast a pall over greentech. (From then until 2021, LPO approved only three loan guarantees, for a single nuclear power plant in Georgia.)
Shah, who was hauled before a tense Republican hearing on climate spending in October, says the risk around Solyndra was poorly managed, and points to safeguards made by his office. And he emphasizes that, despite the occasional misfire, the office actually makes money for the government. It has a loan-loss rate on par with commercial banks, and has earned the U.S. $4.9 billion in interest payments since the program’s inception, including $484 million last year, according to the Energy Dept. Shah already sees proof of the office’s impact in “huge” multiyear power purchase agreements that customers have signed for upcoming LPO-supported geothermal, hydrogen, and nuclear projects, a mechanism that helps provide further certainty to developers and investors.
For Shah and his office, the stakes are nothing less than the future of the U.S. energy economy, and the clock is ticking. Apart from the 2026 deadline included in the climate bill, the U.S. is aiming to cut carbon emissions to half of 2005 levels by the end of the decade. At the same time, energy demand is spiking faster than clean energy can keep up, prompted in part by the power-hungry data centers used to train and operate generative AI systems like ChatGPT.
Amid those challenges, his office finds itself on the precipice of a new existential threat: Should Republicans win control of Congress in November, there’s reason to believe they’ll claw back funding, if not pull the plug on the operation entirely. Shah spoke with Fast Company about how his office is getting billions in loans out the door—and what might happen if the entire program ceased to exist. The interview has been edited for length and clarity.
How do you tell the story of clean tech financing in this country, and where does the office fit into that narrative?
We have had a robust ecosystem of clean tech financing in this country for many years. And when you look at how many companies got financing in ’21 and ’22, it was just extraordinary. I don’t think that the A round and B round and C round are the difficult parts of raising money. Obviously, some people win, some people don’t win, and there’s lots of factors that go into that, but on the aggregate, there’s a lot of money going into these earlier stage rounds.
But then, when you get into a later stage, what you find is, in this infrastructure business, a lot of companies need to deploy a project, a commercialization strategy, that amounts to a billion dollars. Whether that’s next-generation low-carbon steel or low-carbon cement or virtual power plants or car sharing. Your overhead for a car sharing business, when you think about software and the people who work at the company, could be $5 to $10 million a year, right? And so in order to earn $5 to $10 million a year to break even, you need $500 million worth of cars, or maybe a billion dollars’ worth of cars, which, frankly—if you take a billion dollars and divide it by $40,000 per car—it’s not that many cars.
So what you find is that once these companies have proven their concept and gotten through the demonstration phase, they often need to get to a billion dollars. And so that next round of financing that they’re going after is hard to raise. Investors are like, Am I really gonna give you a billion dollars? And I need a rate of return on that billion dollars because I need it to be a venture capital return or a private equity return, or whatever it is.
So people’s belief that the Loan Programs Office actually is in business means that they’re more likely to get that billion-dollar project done—where they only have to raise $300 million of equity, and we bring in $700 million of debt. Now, the weighted average cost of capital isn’t 20%, it’s 12%. And [as an energy developer], you’re like, “Well, actually, I can make those numbers work. I can provide a service that actually is affordable for consumers and meets everyone’s needs if my cost of capital’s 12%, instead of 20 plus percent.”
And what do you say to those who think that the government shouldn’t be in this kind of business at all? Or maybe another way to put it is, What would happen if we didn’t have the Loan Programs Office?
Well, we know what would happen. I mean, from 2011 to 2021, which is when the Loan Programs Office was not really active, nobody really succeeded at scale except for solar, wind, battery storage, and EVs—all four sectors we supported in LPO 1.0. So the folks who got financing in LPO 1.0 got over that bridge to bankability, and the other sectors basically went sideways for 10 years. Think geothermal, long duration energy storage, think electric trucks. All of those sectors basically went sideways. They did demonstration projects, but they never got to the billion-dollar scale.
As a result, there was a backup of a bunch of these companies that were ready for LPO to be a merchant again. And I would suggest to you that there’s a bunch of additional companies that received C round financing because those investors believe that five years from now, they might be able to get money out of LPO. I think so many [financiers] believe that LPO is gonna be around five years from now, that they’re starting to take risks again. They’re starting to help companies across the spectrum.
Well, some in Washington would like to power down the LPO. In terms of the stakes here, I guess why should the average American want the office to flourish?
[Laughs] Because the business model has changed. In the early days of clean tech, the business model was to perfect the technology and then license it or manufacture it in China or Europe, right? And so we didn’t need to provide C and D rounds, and you didn’t need LPO because companies were saying, “Look, once your solar technology works, go to China or go to Malaysia or go to Europe.” Today I think everyone, from the average American to our sitting president, has said “We have 45 years of history of inventing everything in the world here in the United States, and we don’t want you to send it overseas to commercialize it anymore. We want you to commercialize it here. Tie it to the American worker, create jobs here.” I think that is a full-throated message we’ve received from voters in the 2016 election and the 2020 election. And we passed the bipartisan infrastructure law and the Inflation Reduction Act so we can commercialize the technology here.
It’s part of a giant buildup of infrastructure, too, part of a bigger vision under the Biden administration. Are there other factors this time around that are front and center for you or for the office that maybe weren’t as important last time? Are there important lessons from 1.0 you’re thinking about now?
One of the biggest lessons we learned from 1.0 is that we should never take real technology risk. So we take perceived technology risk all day, but when you look at all of the failures we’ve had in 1.0, the only one we got beat up on is Solyndra, because Solyndra had real technology risk, and in the end, the technology didn’t work. But we didn’t really get beat up on Fisker Automotive or Unbound Solar or Tonopah Solar or Ivanpah or KEPCO Solar. Like we’ve had other losses in the portfolio, but people thought that we underwrote the risks properly there. Technology risk should be taken by our partner offices within DOE that do demonstration work. We should not be taking real technology risk; we should be taking perceived technology risk and scale up risk.
I think the second one that we learned is that, for certain sectors we need other policy, not just LPO. So when you think about the battery supply chain, not only do you have our office and the Critical Minerals [program], but you also have this 37-50 tax credit for electric vehicles that use domestically sourced or processed critical minerals. So that now provides belts and suspenders. Not only are we providing the financing for the critical minerals, but if there’s dumping going on by other countries around the world, where they’re selling their critical minerals for less than the cost to make them, or to process them, we’re protecting our industries by having this domestic content requirement, either for solar panels or for critical minerals.
And so when you think about the way that the [Inflation Reduction Act] was structured, there’s a set of policies, not just LPO. We’re not on our own island. We partner with the 45-V [tax credit] program for hydrogen, or with the 45-U program for nuclear, or with the 30-D program for battery materials. We’re partnered with these other policies to provide a clear signal to the equity investors that, “Hey, we are open for business and we want these projects to succeed.”
Still, it can be challenging to break across silos in Washington, not just between other agencies but within your own. I wonder how big a challenge that is for you and for the Loan Programs Office when approaching this outlay. How do you work with the Environmental Protection Agency and others?
It’s a good question. I don’t know how to answer it per se. I would say that all of our colleagues at the other agencies want to coordinate with us as much as we want to coordinate with them. So the good thing is that there’s no lack of interest. It’s not like I call people up and they say, “Jigar, why am I wasting my time with you?” They’re saying, “I’m so glad you called, we actually were thinking about this the other day and we wanna coordinate with you.”
I think the other thing that I would say is that, because our loan applicants desperately need to coordinate with those other agencies—because if you’re offshore wind, you need to get a BOEM [Bureau of Ocean Energy Management] permit, you need to go to the Department of Commerce and get a national Marine Mammals Life permit for the construction—once someone becomes an applicant, we have the ability to advocate on their behalf across the entire government. So we’re not coordinating on a theoretical basis; we’re coordinating on behalf of our applicants. A couple of our applicants are using the Class VI wells to do carbon sequestration, so we’re coordinating closely with our friends at EPA who are providing the Class VI wells.
And I think that [when regulatory hurdles pop up] our office has really assumed positive intent. We’re not saying, “Oh, they just hate carbon sequestration or they just hate offshore wind.” We’re going over there and saying, “What is stopping you from moving quickly and doing this?” And they actually often have good questions. And then we, on behalf of our applicants, get the answers for them, and we help them through that. And because we are the filter, they trust us. And if LPO has done a bunch of diligence on the applicant, then we’re a more trusted voice across the government. So we can play that facilitation role for most of our applicants.
You have over 200 loan applications in the pipeline for $263 billion in loans, but so far the office has approved only four projects: a $2.5 loan for a lithium-ion battery manufacturing project by Ultium Cells; a $3 billion loan guarantee to a Sunnova solar-and-battery virtual power plant (VPP) project; a $102 million direct loan for Syrah Technologies’ battery factory in Louisiana, and a $504.4 million loan guarantee for ACES, a hydrogen storage facility in Utah. What are the challenges you’re facing as you sort through and choose other applicants?
I think it’s important to remember that we are not picking which applications to support. What I’ve told my team is that every single application that comes in that meets the statutory requirements that Congress laid down from us gets a thumbs up. We’re not determining whether we think hydrogen is better than transmission is better than carbon sequestration is better than whatever. We’re helping all of them equally as long as they qualify under the program.
Now, we do require them to have a high-quality application, which means they have to fill out all of our forms correctly. And I would say that that is a far taller order than you would think. Because many of these applicants are extraordinary innovators, and in some ways artists, but they don’t actually know how to buckle down and actually be a banker, right?
And so, to some applicants I’m saying, ‘I don’t think you can do this. You need to hire a consultant to do this for you, because you’re the artist. You need to find somebody who’s boring who can just fill out my paperwork. Because I am not allowed to give you money and to invest in you unless you fill it out properly. That is just a government rule.’
And it’s not impossible or hard to fill it out properly, but a lot of folks are not used to that. They’re used to just raising equity. So they’re selling a dream. And I’m like, yeah, yeah, yeah—but I’m actually looking at receipts. Like, that’s my job. [Laughs] So you gotta submit the receipts properly. And so that’s one of our biggest challenges. We have 205 active applications, but only about a quarter of them are actually capable of getting into due diligence the first time through. Because they’re actually capable of filling out the paperwork properly, and they’ve got project finance experts on staff, or they’ve hired somebody. And so the other 150, I’m like, “I promise you, we care deeply about you, but you gotta finish this checklist…”
And how does borrowing from the government bank compare with private debt financing?
It’s the same as a commercial bank. On the front end, it’s just, do you qualify? What statute do you qualify under? On the back end, we do a financial model. We say, what do your contracts look like? If you don’t have good contracts, then what’s your cash flow look like? Because if you have good contracts, then maybe we can live with a 1.3 debt service coverage ratio [a measure of the cash flow needed to meet a project’s annual obligations].
Some of the sectors that we’re in, like sustainable aviation fuel, have really good contracts, where airlines are signing 20 year contracts. But renewable diesel has terrible contracts. Folks there say, I’ll buy the fuel from you for the next six months and then after that, I’ll buy it again. So in that case, we’re saying, well, it needs to be a 2.5 debt service coverage ratio. And that’s exactly what a commercial bank would do too.
What’s different between us and a commercial bank is not our underwriting criteria. What’s different is that we’re willing to go first, where a commercial bank often says, I really need to have ten of my friends do the first deal first, and then I’ll do the 11th deal. Like, I don’t wanna be first. Whereas we are paid and told by Congress to go first.
Speaking of which, what are your obligations to Congress?
Congress has been very clear in the legislation. I mean, obviously what they say on the dais is different on different days. But in the legislation, they say: here’s how much loan authority you have, and here’s how much credit subsidy you have. And the credit subsidy by definition is loan loss reserve—how much money you’re allowed to lose.
So, generally speaking, I’d say in LPO 1.0, we set aside $5 of loan loss reserve for every $1 of actual losses we incurred. So we were really good with the taxpayer’s dollars. I think in LPO 2.0, I think that we’re probably gonna be $2 of loan loss reserve for every $1 of losses. Because we’ve learned a lot. And so we’re a little tighter. We don’t have to set aside $5 because that was the first time we were doing it, so we were overly conservative.
But Congress has been very clear that we are allowed to take real risk. We’re allowed to have losses. They want it to be smart risks. They want us to be thoughtful about what risks we’re taking. And in general, we’re taking management-team execution risk. And I’d say that everything that has gone wrong in any of our projects, we actually foresaw that, and wrote it up in the credit, which is where we say, here are the 10 things that could go wrong here. So I think my team is really extraordinary.
Can you briefly talk about what makes your team different this time around and, and the office different this time around and what, what it brings to the table? Apart from deep podcast experience of course.
[Laughs] I’d say the big thing that we have done this time around is we have really built the institution. So an LPO 1.0 era gave us money that had to be obligated by the end of 2011. We don’t have that same pressure to just throw money out the door. And so we’ve been very careful and cautious about building the institution. And as a result, we’ve been able to attract high quality people. I would say that the vast majority of people that we’ve hired have come from the private sector and have said, ‘You know what? I have this enormous body of work, but right now I’m mission-driven and I really want to join the government to meet this moment.’ And so having this extraordinary expertise with people with twenty years of experience joining the government has been really gratifying to see, gratifying to see how many people are willing to make that commitment. And so we’ve had over a hundred people do that from the private sector.
Energy demand is surging thanks in part to the power hunger of generative AI. What role do you see now for nuclear and geothermal?
So in order to meet the President’s goals of decarbonization by 2035—but also electrifying our economy a lot by 2050, to get the full emissions reductions—we’re talking about a lot of electricity load growth. And then on top of that, you have AI. That has come in, and they need a lot of electricity. And so, most of the prognosticators out there are saying that we’re gonna have to double electricity sales by 2050. And so if you build as much solar and wind as you can possibly do—which we are totally for—it doesn’t get you to double the electricity sales by 2050. So then you need more clean, firm generation. Think nuclear, geothermal, hydro. And so it’s not either, it’s both. We need both.
The bottom line is, today [nuclear, geothermal, and hydro] on a new basis will cost, let’s say, $99 a megawatt hour. But they reduce the amount of new transmission distribution you have to build. So a certain amount of it actually is quite cost effective—versus $35 a megawatt hour for solar, where you have to build more transmission distribution and then the load piece. We’ve done a lot of modeling at DOE that shows how, in a transmission constrained environment—which is where we are—having more clean firm generation, even if it’s $99 a megawatt-hour, is more cost effective for the entire grid.
How will the grid need to change?
Building lines like China does is not something we’re gonna do in this country. But we have a lot of unused capacity in our existing grid that can be unlocked with grid enhancing technologies, with smart wires, reconductoring, and other upgrades. For a long time we ran our grid where demand could do whatever it wants, and supply had to modulate itself to meet demand. Today, every single appliance you buy comes with an app on your phone so you have the ability to modulate demand with the same level of dexterity that you can currently only modulate supply. And we’ve tested that technology for 30 years at DOE. And unlocking that potential is 90% cheaper than building new generation and new transmission.
So yes, we have to build a lot more generation and we have to build more transmission, but we can make life easier on ourselves if we also lean into demand flexibility, which includes virtual power plants [networks of production and storage systems that help balance supply and demand] and long duration energy storage [batteries that last hours longer than lithium-ion].
And that includes individual homeowners who are putting solar on their roofs and batteries in their basements that can feed back to the grid.
Batteries help [the whole grid] become more efficient. There’s a lot of people who have backup batteries that they’re putting in their garage or wherever else. So we’re like, ‘Hey, instead of charging it right at this time, why don’t you charge it when there’s excess capacity on the grid? And why don’t you discharge it when there’s a peak, and get paid for it?’ And so I think what we’re saying to everybody is, ‘Look, we are in load growth again, so let’s be smart about how we do this.’ Because you can do it the hard way, which is expensive, or you could do it the easy way, which is using technologies that we’ve been testing for 30 years.