Blogs

Energy, Security, and Climate

CFR experts examine the science and foreign policy surrounding climate change, energy, and nuclear security.

Latest Post

REUTERS/Amit Dave
REUTERS/Amit Dave

Why We Still Need Innovation in Successful Clean Energy Technologies

Today is my last day at CFR. I’m joining ReNew Power, India’s largest renewable energy firm, as their CTO. I’m excited for a new adventure but sad to leave the Council, which has given me support and autonomy to study the innovations needed for global decarbonization. Read More

Brazil
Can Deforestation be Stopped?
Why has Brazil slashed deforestation over the last decade while Indonesian deforestation has accelerated? The two countries lead the world in deforestation, which, after energy use, is the top source of greenhouse gas emissions. In the last week, each country has released an emissions-cutting plan in anticipation of the Paris climate summit that relies heavily on avoiding deforestation. Figuring out why Brazil has succeeded while Indonesia has lagged can provide insight into how both countries can do more. Earlier this year I gathered a multidisciplinary group to explore the Brazilian experience and extract lessons for climate policy. Some of the highlights are summarized in a short report that we’ve just released. We looked at a wide range of issues, many of which are discussed in the report, but I was particularly intrigued by our discussion of why Brazil and Indonesia turned out so differently from each other. The most obvious reason is that Brazil had an earlier start. Its government has been focused on reducing deforestation for over a decade; the Indonesian government hasn’t started looking at the issue seriously until more recently. This is actually good news, since it’s something that time should overcome. Governance and rule of law also stood out as big factors. Both countries have fairly decentralized governance – a feature that should make controlling deforestation difficult since decisions from the center don’t always translate into action. But Indonesian governance is considerably less centralized, which puts Indonesia at a disadvantage. Similarly, since avoiding deforestation requires effectively enforcing laws, corruption is a big barrier to success. Brazil obviously has its fair share of corruption problems, but Indonesia is arguably worse. Both of these factors make avoiding deforestation more difficult in Indonesia. But there’s a bright side: neither needs to be permanent. Governance structures change over time; countries also reduce corruption and improve the rule of law. None of this is easy, of course, and it’s unlikely to happen just to facilitate avoided deforestation, but it at least offers some promise. The last factor that came out in our discussion, though, augurs much more poorly for Indonesian prospects. Amazonian timber has typically been cleared to create cropland or pastureland, or, more simply, to establish ownership of a given tract of land. The wood itself is typically mostly worthless. In contrast, in Indonesia, the trees that are cut down are usually highly valuable. That means that the economic incentive for deforestation is much stronger in Indonesia – which, in turn, means that policy needs to lean much harder against deforestation in order to succeed. This factor suggests that replicating the Brazilian experience in Indonesia will be more difficult than many would hope. The full report on the workshop dives into a bunch of other issues – including the prospects for private sector led efforts to reduce deforestation, new avenues that public policy might pursue to keep trees standing, and the possibility that the decline in Brazilian deforestation might reverse. Download it here.
Fossil Fuels
Now What’s That Got to Do with the Price of Oil?
This post was co-written with Peyton Kliefoth, an economics major at Northwestern University and research intern at the Council. Over the weekend, I published a piece in Fortune Magazine explaining a surprising correlation between falling oil prices and tumbling shares of Yieldcos, which are publicly traded holding companies mostly comprising renewable energy assets in the U.S. and Europe (see chart below).   Comparison of oil price and share prices of the top seven Yieldcos, from June to September, 2015 (NYLD: NRG Yield, NEP: Nextera Energy Partners, TERP: TerraForm Power (SunEdison Subsidiary), ABY: Abengoa Yield, BEP: Brookfield Renewable Energy Partners, PEGI: Pattern Energy Group, CAFD: 8Point3 Energy Partners (joint venture between SunPower and First Solar))   Fundamentally, the value of solar and wind projects should not depend on oil prices, since oil is rarely used in the developed world for electricity and therefore doesn’t compete with renewable power generation. It turns out that the cause of falling Yieldco share prices has less to do with what is being traded than who is doing the trading—I write: Few paid attention to an ironic trend: the same investors holding oil and gas assets had also piled into an obscure but crucial class of renewable energy investment vehicles—so-called “Yieldcos”—driving down the financing costs of clean energy. As it turned out, renewable energy prospects hitched to the conventional energy bandwagon hit a bump in the road. In June and July the bottom fell out of the oil market (again), the Fed strongly hinted at interest rate increases, and a number of renewable energy firms sought large sums from public capital markets. Together, these three unrelated developments conspired to spook fossil fuel investors, who dumped renewable energy Yieldco shares and plunged prices into a vicious downward spiral. Now the stakes are high: if Yieldcos fail, renewable energy could lose access to public markets and the low cost of capital necessary to scale up wind and solar. To recover, Yieldcos may have to restructure, seek help from parent developer firms, and hope for constructive public policy to further de-risk renewable energy investments. Whereas the article focuses on the causes of the recent downward spiral in Yieldco share prices and the remedies for stabilizing and lifting prices, in this blog post I’ll assess the underlying renewable energy industry and the long-term prospects for vehicles like Yieldcos. Even before the collapse of Yieldco share prices this summer, doomsayers predicted that Yieldcos were overvalued and went as far as to call the Yieldco model a “Ponzi Scheme.” To those analysts, this summer has vindicated their conviction that a Yieldco is no more than the sum of its parts, and that the stock market had erred in imputing value over and above the constituent renewable energy projects in a Yieldco’s portfolio. I disagree.  This summer certainly proved that Yieldcos, as currently structured, are unstable vehicles whose share prices are liable to spiral upward or downward without much of a change in the performance of the underlying projects. But if they can weather this perfect storm, restructure, and attract a broader investor base, Yieldcos can add considerable value by reducing transaction costs and providing public investors a diversified portfolio of renewable projects. The renewable energy industry as a whole is doing very well right now. The costs of solar and wind projects have consistently fallen, and installed renewable capacity is growing around the world. But extrapolations that solar will account for thirty percent of the global power market by 2050 will not come true without further reductions in the cost of capital and participation from public markets, which can supply the scale of investment needed for renewable energy to rival conventional energy sources. That’s where Yieldcos come in. How Yieldcos Create Value There are three ways a Yieldco creates value over and above the value of the renewable energy projects it comprises. First, it reduces the risk of investing in renewable energy. To accomplish this, renewable energy developers spin off the least risky part of their portfolio—owning and operating renewable energy installations post-construction—creating a Yieldco, an independent, publicly traded entity. Thus, the Yieldco avoids the riskier elements of project development—regulatory approvals, construction, contracting—and only purchases operating or near operational assets from the parent developer that come with guaranteed revenues from long-term power purchase agreements (PPAs) with utilities. Second, Yieldcos offer public market investors—like institutional and retail investors—an easy way to invest in renewable energy; in other words, they reduce the transaction costs that would otherwise block public market capital in a sector dominated by private capital. This is possible because of the way Yieldcos return almost all of the revenue generated by renewable energy projects back to investors. Similar to Master-Limited Partnerships, which are holding companies for oil and gas infrastructure assets, Yieldcos are able to shield shareholders from double taxation, avoiding corporate income tax on renewable project revenues to distribute pre-tax dividends to shareholders. Since solar projects compose a majority of Yieldco assets, and solar panels require next to zero operating and maintenance expenditure, Yieldcos are able to return most (80–90 percent) of their projects’ operating revenue to investors through dividends. The third way that Yieldcos add value—by promising 8–15 percent dividend growth—is what got them in trouble this summer. Yieldcos depend on high share prices to raise equity on public markets and purchase more renewable projects at returns that exceed their cost of capital, driving share prices up further. I call this a “treadmill of equity issuances and dividend payouts.” Unfortunately, the treadmill can overheat,  and when share prices start to drop, they viciously spiral downward. I suggest that relying less on equity and more on debt—up to responsible credit limits—will enable Yieldcos to avoid spirals, though their share values will not be as high as when they were on the treadmill. Still, even if Yieldcos are less aggressive about dividend growth, they add value to renewable energy projects by unlocking public capital markets through reduced risk and lower transaction costs. By focusing on developing a strong asset portfolio, Yieldcos can still play an important role in scaling up renewable energy. Four Questions Underlying the Long-Term Success of Yieldcos As Yieldcos mature and find ways to avoid short-term share price volatility, their long-term prospects will depend on macroeconomic fundamentals and the health of the renewable energy industry—these are far more logical factors to drive Yieldco value than the price of oil. There are four threshold questions that require affirmative answers for Yieldcos to succeed long-term: Question 1: Will solar remain economical after the imminent expiration of the solar Investment Tax Credit (ITC)? Yes. Although important in the near term to U.S.  solar project economics,  the ITC is not crucial to their long-term viability. Currently, the ITC offers developers of U.S. projects 30 percent of the project value in tax credits through 2016 and 10 percent thereafter. The ITC was crucial to incentivize domestic deployment when solar economics were not so favorable, but today, installed solar costs are on a sufficient downward trajectory to make many projects viable on their own, without the tax credit. The expiration of the ITC will likely cause a drop off in project development in 2017, but falling costs will enable project growth thereafter. First Solar, a leading panel manufacturer, has projected highly competitive costs of $1 per installed Watt in 2017, without the ITC, and historically low bids in recent solar PPA auctions below 5 cents per kWh suggest that solar will be competitive in wholesale power markets even after the ITC step-down. Yieldcos comprising solar assets should be able to weather this short-term storm, especially because many are amassing a global portfolio of assets, diversifying their exposure outside of the U.S. market. Question 2: Can Yieldcos survive rising interest rates?  Probably. Rising interest rates will tarnish Yieldcos’ attractiveness as a low-risk, comparatively high return investment, but rates will have to rise considerably to really damage the Yieldco value proposition. Yieldcos can be attractive because of the spread between the market’s “risk-free rate,” often defined as the yield on a ten-year Treasury Bill (about 2.13 percent today), and the Yieldco dividend yield, currently between 5–6  percent. However, the Federal Reserve has signaled that an interest rate hike is likely by the end of the year, possibly marking the end of a historically low interest rate era. Combined with the falling oil price, the Fed’s hints contributed to plunging Yieldco prices over the summer. Still, rate hikes of a magnitude required to wipe out Yieldcos’ return over the risk-free rate are only distantly on the horizon. Michael Liebreich of Bloomberg New Energy Finance warns that if rates return to their 2007 level of 5.3 percent, Yieldco competitiveness as a low-risk investment would fall. Still, for the foreseeable future, modest interest rate hikes will likely not spell doom for Yieldcos. Question 3: Is there room for growth? Yes, resoundingly. The growth potential of renewable energy in the United States and the world is so high that Yieldcos will not run out of projects to acquire anytime soon. A useful point of comparison is the Yieldco’s cousin, the oil and gas asset MLP. MLPs support around 10 percent of the $1.1 trillion U.S. oil and gas sector and have posted an annualized 27 percent growth in market cap over the last 24 years. By contrast, Yieldcos represent less than 1 percent of the burgeoning renewable energy project finance sector, and Yieldco dividend growth targets are considerably less ambitious at 8–15 percent. Fundamentally, there is certainly room for growth. Question 4: Is a Yieldco all that different from a Ponzi Scheme? Yes. Although a Yieldco does depend on continually raising equity to acquire projects and pay shareholders, it is not a Ponzi scheme, because it comprises real, income generating assets just as do other established financial vehicles. However, it is unclear if the accounting practices that enable Yieldcos to distribute high dividends are sustainable. Unlike a Ponzi scheme, in which new cash is raised to pay existing investors, a Yieldco actually invests new cash in income-generating assets en route to paying dividends. Some may quibble that the difference with a Ponzi scheme is semantic, but the Yieldco model is akin to MLPs and Real Estate Investment Trusts (REITs), both of which are considered established, sound investment vehicles. To call one a Ponzi scheme would be to indict all three models. However, questions remain unanswered about the details of Yieldco accounting. In particular, Yieldcos assume a very low rate of depreciation oftheir operating assets, of which solar installations are often the majority. Since the installations have historically proven to be long-lived, in some cases twice as long as the standard twenty-year PPA contract signed with utilities, Yieldcos only deduct a small “Maintenance” sum from their operating revenues before distributing dividends to shareholders. If this assumption is wrong, however, then Yieldcos will have failed to accurately depreciate their assets, so that over time their asset base shrinks because of inadequate reinvestment and excessive dividend distributions. It will be years and perhaps decades before Yieldco claims of asset life are vindicated or disproven. In the meantime, critics will continue to accuse Yieldcos of hiding the need to reinvest in capital expenditure in order to reward shareholders. But the historical record of long-lived and productive renewable energy projects is on the Yieldcos’ side. In summary, Yieldcos do add value to the projects that they bundle together, and the health of the renewable energy sector can underpin Yieldcos’ long-term success. That means that the conclusion I wrote to the short-term story of Yieldco prices tumbling alongside oil prices applies equally well to the long-term story of how the future of renewable energy may depend on the success of Yieldcos: Renewable energy is on the cusp of becoming a mainstream alternative to fossil fuels—getting there requires a mainstream financing tool. Although the Yieldco model must improve after derailing this summer, getting it back on track is in everyone’s best interest.
Fossil Fuels
Guest Post: Cleaning Up the Mess at the Nigeria National Petroleum Corporation
This was originally posted by my colleague John Campbell on his Africa in Transition blog. John was formerly U.S. Ambassador to Nigeria and is currently the Ralph Bunche senior fellow at the Council on Foreign Relations. The Natural Resource Governance Institute, a New York-based think tank and advocacy organization, has issued a must-read report, Inside NNPC Oil Sales: A Case for Reform in Nigeria. The authors are Aaron Sayne, Alexandra Gilles, and Christina Katsouris. The Nigeria National Petroleum Corporation (NNPC) sells about half of Nigeria’s oil, worth an estimated $41 billion in 2013. The report concludes that NNPC’s approach to oil sales “suffers from high corruption risks and fails to maximize returns for the nation.” The report is detailed—it runs to seventy-one pages with additional annexes. It is thoroughly convincing and offers specific recommendations. It notes that “the bad practices that undermine NNPC oil sale performance all have political interference at their root.” The report also argues that the new presidential administration of Muhammadu Buhari has a unique opportunity to tackle the problems at NNPC, which have long been ignored. At almost the same time the Natural Resource Governance Institute issued its report, President Buhari announced a wholesale sacking of the directors and senior management at NNPC. The president relieved the group managing director and the executive vice chairman, who had been appointed by former president Goodluck Jonathan. The following day, he fired the nine NNPC executive directors. Buhari’s choice of group managing director is Emmanuel Kachikwe. He has been executive vice chairman and general counsel of Exxon-Mobil (Africa). Among other academic attainments, he holds masters and doctorate degrees from the Harvard Law School,according to Nigerian media. He has also worked for Texaco Nigeria. Nigerian media reports that he intends to reduce the number of group managing directors from nine to four. As with his appointment of new military service chiefs, Buhari’s choices indicate that he focuses on expertise and experience, rather than on political connection. Inside NNPC argues that reform of NNPC does not require omnibus legislation, but rather a bold agenda with a short timeline. Buhari’s personnel choices fit that prescription.
  • Climate Change
    Five Takeaways on the EPA’s Clean Power Plan
    The final version of President Obama’s Clean Power Plan (his carbon dioxide regulations for new and existing power plants) will be released later today by the Environmental Protection Agency (EPA). Many details are already online. The new rules are an important step forward but certainly not without their flaws. Here are five important things, good and bad, that today’s dueling press releases might not tell you. This is an impressively creative “save” given legal and political realities – particularly on the international front The President’s proposed regulations on existing power plants (released last year; today’s rules are the final version) were central to meeting the U.S. target of a 17 percent emissions cut below 2005 levels by 2020. But the EPA received extensive feedback from industry and analysts claiming that meeting the 2020 goals proposed in the draft rule would require a dangerously rapid transformation of the U.S. electricity system late this decade. This apparently led the EPA to delay the first year during which states must comply with the new rule from 2020 to 2022. But that left a big problem: the United States would be unlikely to deliver on its international commitment to cut emissions 17 percent by 2020. So the administration came up with a clever save: a proposed “Clean Energy Incentive Program” that will reward states that cut emissions in 2020 and 2021 by in essence giving them weaker targets from 2022 to 2030. In principle, if states fully utilize that program, the United States will still be in the neighborhood of meeting its economy-wide 2020 target. (“In the neighborhood” because there’s enough uncertainty in the U.S. energy system to make 2020 emissions unpredictable even with the best policies possible.) At the same time, the EPA strengthened its emissions targets for 2030, offsetting the new headroom created by the incentive program and keeping projected 2025 emissions similar to those in the draft plan. None of this matters much when it comes to aggregate emissions. But it matters a lot for how the United States is seen internationally. On that count, the administration deserves applause. No one really knows whether the United States will meet its 2020 target The biggest weakness in the draft Clean Power Plan was its vulnerability to litigation. In particular, its emissions targets were determined in part by calculating how much emissions could be reduced through improved energy efficiency, a tactic that made the whole rule vulnerable to being struck down by the courts. (This is ostensibly a rule governing power plants, but power plants can’t improve consumer efficiency.) The administration wisely ditched that element, leaving the plan on much firmer legal ground. But reliance on the Clean Energy Incentive Program introduces a substantial new source of uncertainty. To conclude that the United States will still meet its 2020 goals, the administration is assuming that states will fully utilize the opportunities created by the incentive program. The program gives special credit for electricity generation from new wind and solar installations and from new energy efficiency in low income communities. But it’s entirely plausible that the special credit won’t be enough to get states to make power plants install all the wind and solar that the EPA is assuming they will. (It’s also fair to say that the odds of every state fully utilizing its opportunities is close to zero.) And, the farther states end up away from fully utilizing the incentive program, the farther the United States will be from its broader 2020 target without other policies. Which leads to... The United States may need additional policies to deliver on the 2020 target The administration is now relying heavily on wind and solar to meet its 2020 target. But if the Clean Energy Incentive Program isn’t enough to incentivize investment there, the U.S. government will need additional policies on that front. The most obvious place to look is an extension of the Production Tax Credit (PTC) for wind and Investment Tax Credit (ITC) for solar. But these are going to get very expensive (and perhaps politically unsustainable) if they remain in their current forms while investment ramps up to the level that the EPA envisions. Expect renewed administration focus on crafting some sort of legislative deal that would reform and extend the PTC and ITC, perhaps with a sunset around 2020-2021 as the Clean Power Plan phases in. No one knows what mix of renewables, natural gas, and efficiency will result from the plan There is a lot of reporting, including by many who should know better, claiming that the plan will result in massive amounts of renewable generation and no increase in natural gas above business as usual in the long run (2030ish in this case). This reporting is based on two things. First, the EPA, in developing its targets, uses “building blocks” – new renewables, improved coal plant efficiency, and extra coal-to-gas switching – and the new rule reportedly relies heavily on the renewables block. Many are concluding from this that states will be requires to massively increase renewables use. But – and this is really important – the “building blocks” tell you nothing about what measures states will actually use to comply. Once the building blocks are used to determine state targets, the states decide how to meet those targets. At that point, it’s as if the building blocks never existed. If a state wants to use only solar to meet its targets, it can do that. If it wants to use only natural gas or nuclear, it can do that too. The second reason you’re hearing that the final plan will rely largely on efficiency and renewables is that when the EPA models the real-world impact of the rule, it reportedly foresees lots of new efficiency and renewable energy, and not much new coal to gas switching. But this is a feature of the EPA model, not something that the rule requires. In particular, the EPA model is well known to predict huge increases in efficiency. If, as many experts assume, it is substantially overestimating the efficiency response, you’ll see more coal-to-gas switching (and more renewables investment) in the real world response. Something similar applies to misestimates of renewables investment, though it’s not as clear there what the weaknesses of the EPA model may be. This plan, while good, is far from perfect – but much of that is simply a reflection of political reality The EPA estimates benefits well in excess of costs for the plan. Even if they’re way off, it’s likely that benefits will still exceed costs, making the new rules an important step forward. That said, this is certainly not the best of all worlds. There’s no economic and little environmental rationale for restricting the new Clean Energy Incentive Program to wind and solar rather than including nuclear, coal with carbon capture, or coal-to-gas switching (with reduced credit to account for the carbon content of gas) – all the restriction really does is increase the cost of delivering the targeted emissions cuts. It also increases the risk that the United States won’t meet its international commitments for 2020. This one is an own goal – the EPA could have taken a more expansive approach to early compliance it if wanted to. Beyond that, though, theoretically superior tools were basically out of political reach. Economy-wide carbon pricing legislation could have gone further in creating nationally uniform incentives for emissions reductions. (Of course, in the real world, economy-wide legislation would have had its own myriad carve-outs and distortions for various preferred technologies and industries.) It could have created incentives that cut across sectors (e.g. electric power and industry). It also could have generated revenues to reduce the federal deficit, help low income consumers, and assist industry in transitioning. The Clean Power Plan, in contrast, generates no revenues for any of this, though individual states might still raise money through their own implementation plans. But economy-wide legislation, whether cap-and-trade or a carbon tax, has been a political non-starter for years; one can’t fault the administration for not doing something that was politically impossible. Bottom line: Politics has greatly constrained the realm of the possible for emissions cutting policy. A fundamental shift in U.S. politics could in principle yield something substantially better – but that isn’t the universe we’re living in. For the time being, the principal alternatives to the Clean Power Plan as it stands are inaction; a different set of EPA regulations that’s far less flexible (and hence less economically sound) or far weaker; or, potentially, large subsidies to a range of zero-carbon energy generators. The Clean Power Plan is a vastly superior way forward.
  • Technology and Innovation
    Five Things I Learned About the Future of Solar Power and the Electricity Grid
    Nestled in the foothills of the Rockies in Golden, Colorado, the Energy Department’s  National Renewable Energy Laboratory (NREL) was established in 1977 to help bring new energy technologies to market. Today it is one of seventeen national laboratories overseen by the Energy Department and the only one whose sole focus is renewable energy and energy efficiency research and development. I spent a full day touring the facilities and interviewing researchers working on a range of solar photovoltaic (PV) technologies and on integration of clean energy into the electricity grids of the future. Here’s what I learned: 1. NREL is unique in the solar research ecosystem—that’s a bad thing. Originally christened the Solar Energy Research Institute, NREL is best known as the gold standard of solar technology. One researcher remarked to me, matter-of-factly, “We are the best in solar PV there is.” It is easy to see why. Cutting-edge research on myriad solar technologies is co-located on one campus, and basic science, economic modeling, manufacturing development, and systems integration are all neighbors. Around the world, just two institutions (Germany’s Fraunhofer Institute for Solar Energy Systems and Japan’s National Institute for Advanced Industrial Science and Technology) come close to NREL’s breadth of solar activities... Unfortunately, limited resources constrain NREL’s ability to leverage its integrated research capabilities to commercialize promising technologies. For example, take NREL’s work on an upstart technology I’ve written about elsewhere—solar perovskites. In contrast to academic researchers’ obsession with making fingernail-sized devices that are highly efficient under perfect lab conditions, the researcher leading NREL’s perovskite work wants to scale up manufacturing of larger areas of solar perovskite coatings and achieve long-term stability in the real world. Those goals are ambitious and sorely needed, but with only three and a half researchers supporting them, they will be tough to achieve. Some major research universities (e.g., MIT) also house integrated research programs that help researchers fill the gaps between basic research and commercial success. But many more such centers are needed to institutionalize energy technology development. Prototype printer heads for inkjet printing perovskite solar coatings in a scalable manufacturing process. The process is contained inside a “glovebox,” into which researchers can reach, through the rubber arms, to interact with the process under a controlled atmosphere (Varun Sivaram). 2. Rapid solar PV market evolution means moving targets for researchers. Earlier this year, First Solar (the lone American panel maker in a Chinese-dominated industry) stunningly projected $1 per Watt fully installed cost for utility-scale solar installations by 2017 (this includes a major step-down in tax credit subsidies to solar power). If they achieve this cost, some of the research projects I saw at NREL may have to adjust their targets even lower. For example, an ingenious reactor to mass-produce NREL’s record efficiency solar cells has a long-term panel cost target of $0.70 per Watt. Because the panels are highly efficient, the remainder of the costs to complete the installation should be roughly 33 percent lower than with existing panels; still, even if this reactor were scaled up to produce solar panels in 2017, the fully installed panels would cost $1.10/W, already higher than the industry roadmap. As the leader in solar innovation, researchers around the world will look to NREL to clearly articulate the value of new solar technologies and why a combination of low cost, superior performance, and new applications is preferable to today’s race-to-the-bottom solar commodity market. The reactor (left panel) in which researchers created the world-record efficiency solar cell (46 percent efficient under concentrated sunlight) (Varun Sivaram). For four decades, NREL has compiled the record efficiencies of solar cells and published them in a chart (right panel) that is consulted around the world (U.S. Department of Energy). 3. Reliability in the real world makes and breaks solar PV technology. At NREL’s Outdoor Test Facility (OTF), rows and rows of solar panels endure the rain, snow, and even hail that Golden, CO hurls at them. Out in the field, all sorts of unexpected things can go wrong, and NREL partners with companies who want to learn about the failure modes that could plague their technology. My tour guide pointed out the rooftop shingles coated with a flexible solar panel—although the panels still appear to work, the ensuing leaks from poking wires through the roof had doomed the product and the company. Later, I saw panels which had worked fine for the first year but whose sealing material had gradually given way to attacks by moisture, which now eroded the power-producing material itself. The take-home lesson was that exciting technologies in the lab still have a long way to go to demonstrate the twenty-year reliability that the market demands. As we snaked around the rows of the OTF, two words came to mind: “testbed” and “graveyard.” One First Solar test setup had been in operation for over two decades and still produced 80 percent of its original output—the data from this test had emboldened First Solar’s investors and helped the company achieve its current success. But I also passed dozens of failed relics, sober lessons from the heady days when capital poured into new solar startups that have since expired. NREL’s Outdoor Test Facility (OTF) hosts solar panel test setups from industry partners for multi-decade reliability studies (Varun Sivaram) 4. Standards, not new technology, are crucial for integrating clean energy into electricity grids. At NREL’s Energy Systems Integration Facility (ESIF), researchers shared their views about the challenges and opportunities from modernizing the U.S. electricity grid to integrate new energy technologies. Specifically, ESIF is interested in integrating distributed energy resources (DERs), which include solar panels but also other decentralized ways to make or save energy (e.g., fuel cells, batteries, efficient appliances). In theory, DERs can improve the efficiency of the electricity grid, reduce electricity consumption, and save ratepayers money while also maintaining grid reliability. But in practice, this is complicated by the proliferation of DERs made by different vendors to different specifications and operated without much regard for their effects on the grid, positive or negative. The solution could come from emulating the successful IT industry. There, a robust set of standards have enabled different pieces of hardware to operate seamlessly with software applications, a design feature known as “interoperability.” In much the same way as a computer language employs a concept called “abstraction” to send generalized instructions to diverse hardware, the electricity distribution grid is in need of a standard language, or “common semantics” to coordinate the diverse DERs that will connect to the grid in the coming years. ESIF hopes to create a set of standards that enables such a language and ensures DER interoperability; more federal support would be helpful to accelerate this work. Indeed, the gains from system-level innovation, according to several ESIF researchers, dwarf the expected gains from new energy technology—“we have all the technology we need” was a refrain I heard often. Two of the “Smart Homes” that ESIF has set up to simulate the integration of homes packed with internet-connected, energy-efficient appliances into the electricity grid (Varun Sivaram) 5. A decentralized grid poses serious cybersecurity threats that require immediate attention. As grids integrate more DERs, shifting from a centralized to a decentralized model, there are two opposing effects on grid resilience. The physical resilience of the grid to failure improves, because the strategic value of central power stations and bulk transmission lines decreases as more power is generated and saved closer to the customer. However, the cybersecurity risk actually increases with decentralization, because access points for malicious hackers proliferate—imagine a hacker accessing a mobile phone to breach a home energy system to attack a utility distribution substation, etc. The root cause of the opportunity to efficiently coordinate DERs—their increasing connectivity via the Internet—is also the source of increased cybersecurity risks. Fortunately, these are not new problems, and their solutions are well catalogued. Enterprise IT best practices have long incorporated “role-based access” protocols, in which a proliferation of users on a network does not compromise the network’s integrity, because of walls that isolate decentralized users from the rest of the system. I learned from researchers at ESIF that electricity utilities are far behind other enterprises in their IT practices, and that an immediate culture shift is imperative if grid decentralization is to reduce, rather than enlarge, resilience risk. The contrast between ESIF and the solar research facility was striking to me. ESIF, by decades the younger of the two, was manned by researchers intent on modernizing the century-old utility business model. On the other hand, the solar researchers I met brought decades of experience in solar PV and had a long-term research orientation, at odds with the quarterly target obsessions of a solar industry that is rapidly reducing its costs. But to label the two facets of NREL as its future and past would be a mistake. New solar technologies will be essential to displace fossil fuels, and NREL plays a crucial role in advancing solar PV research and methodically preparing new technologies for the field. Coupled with the next-generation grid that ESIF envisions, tomorrow’s energy systems may look fundamentally different from today’s. I am grateful to the staff of the National Renewable Energy Laboratory for their openness and hospitality, including: Bryan Hannegan, Greg Wilson, Jen Liebold, Tami Reynolds, Jim Cale, Martha Symko-Davies, Erfan Ibrahim, John Geisz, John Simon, Matt Beard, Joe Berry, John Wohlgemuth, Jao van de Lagemaat, and Paul Basore.