Accelerating Energy Technology Innovation - A Panel Discussion at Harvard
Last Friday, I moderated a panel discussion called “Accelerating Energy Technology Innovation” at Harvard’s Climate Leadership Summit. I had a great time hosting and chatting with industry leaders who represent important parts of the energy technology value chain: Ted Wiley (Co-Founder, President & COO at Form Energy), Kiran Bhatraju (CEO at Arcadia), and Jessica Harrison (Executive Director at VEIR).
Key takeaways:
Implementing energy technologies projects has a very limited number of shots with the utilities, so it requires strong executive commitments from the utilities, clear demonstration of value-add, and de-risking pathways for downstream impacts.
While first-of-a-kind projects still go through the “valley of death”, the current environment is much more favorable to deployment because of the financial and resource support from many utilities, states, and the federal government.
We need to invest in both innovation and deployment to meet the urgency of electrification and decarbonization while meeting the differences in needs and complexity across the regions. Both work together and are necessary to solve the challenges that we have ahead.
AI will demand terawatts of energy generation, in addition to the 2-3x of global generation capacity needed to electrify and decarbonize the economy. AI compute will be the most valuable commodity, pushing the AI companies to secure more energy generation plants for themselves and the state actors to invest more resources into AI compute for national security reasons.
With a potential Republican sweep and/or Trump presidency, pull-back is expected on currently uneconomical technologies. However, we will still keep many policies that produce economic value and continue to make progress, especially since the majority of government support on energy transition has been allocated to the Republican districts.
Below is a summary of what we learned, with the responses paraphrased.
1. You each represent an important segment of the energy value chain (grid, storage, and data). Can you help describe your product/technology and what critical problem(s) you solve for energy transition?
Ted: Form Energy is developing iron-air batteries, which are batteries the size of a football stadium that provide multi-day reliability and resilience to the grid (e.g. something that functions like pumped hydro but could be sited anywhere) complementary to the lithium-ion batteries that are very cost-efficient in providing a burst of energy for up to four hours.
Kiran: Arcadia is a data platform that unlocks and normalizes utility data and energy market data for every segment of distributed generation (DG), because utilities have very little incentive to provide this data (think of us being the Plaid for the energy market).
Jessica: VEIR is working on superconducting transmission lines that unlock 5-10x the capacity of conventional lines, which enables us to reduce the physical footprint of transmission and expand capacity in time to meet our energy transition goals.
2. Perhaps one of the most important stakeholders you interact with is the utilities; given the nature of the utilities business (regional monopoly, risk-averse by design, capex-seeking, etc.), what approaches to working with utilities have you found most beneficial in accelerating technology development and deployment? Are they on your side or against you?
Kiran: We believe that the market has to evolve and how utilities operate needs to be much more than just being distribution entities. There are a few progressive utilities that may care about the data for distributed generation and energy markets, but generally, utilities’ incentives are to get a rate of return from capex projects like long-duration energy storage or transmission lines. I can't say that utilities are necessarily active partners for the work that we do for our customers, I wish the opposite were true!
Ted: We think about five core things when engaging with utilities: 1) connection with and commitment from utilities’ top executives, 2) progressive and proactive policy environment, 3) value-add business case to the utilities, 4) buy-in from engineers responsible for testing and implementation, and 5) terms of procurement that allow sustained success. We select our customers carefully because the projects are big, they take a long time, and they need to work as advertised the first time. In that sense, your customer needs to be your partner.
Jessica: It’s important to get input upfront and confirm the product-market fit. We are pre-commercial but have been working with National Grid for 1.5 years now and have another 3.5 years of collaboration agreement. We are working with them on the product requirements, as well as defining the qualifications and testing to help ease their concerns about performance and risk. Also, when building an initial large-infrastructure project with new technology, you need to have a clear idea of where it’s headed and what it’s trying to prove out. The goal is to de-risk it technically and commercially.
3. Given the combined technology risk and commercial risk, many energy technology companies with capital-intensive products face the challenge of “valley of death” at the first-of-its-kind scaleup phase. For Ted and Jessica, how do/did you approach this, and for Kiran, how does data and software help bridge the gap?
Ted: First-of-a-kind (FOAK) projects are really expensive because of the overhead cost to carry the factory. But there's an immense amount of help right now. Some utilities are willing to support pilot projects with their balance sheets. States such as Massachusetts are introducing bond bills. The DoE’s non-dilutive funding from Bipartisan Infrastructure Law is a once-in-a-decade opportunity. And the IRA provides incentives to build manufacturing capacity in the US. Without the confluence of these, we would not be able to go as fast as we have. For the next generation of technologies to pass through the valley of death, we will need an ongoing mechanism to support the demonstration of those technologies as well.
Kiran: We work with about 400 different DG companies and a lot of their valley of death is around soft costs. As an example, if you install rooftop solar in the US today, it's roughly between $3-$4/Watt, whereas in Australia it’s $1-$1.50/Watt. All of that cost difference is not from capex but from understanding the customer's data, the local energy tariffs, permitting, creating proposals, and acquiring customers. It's this pain that we solve, where our platform enables DG companies to shorten that lifecycle and gets companies through that valley.
Jessica: First, I would note that for the FOAK and pilot projects, it's not just the project but also the risk in insurance of that project. Federal funding is incredibly helpful but I’d also say that some states have been doing great work here. For example, NYSERDA has done a fantastic job of thinking through how to fill the gap where banks, equity, or venture dollars are limited. Massachusetts has done a lot of great work here as well. I see that people are starting to truly understand the problem and investing more.
4. Some claim that the currently available technologies are enough to meet the Paris Agreement goals, marked with the urgency from 2024 crossing 1.5 degrees above the preindustrial period; where on the spectrum do you fall on investing in innovative technology versus investing in scale-up/deployment of relatively mature technology?
Ted: We need both and I don't think that one outweighs the other. Innovation has a major role in making things better, especially those we do not have yet. Then, deployment gets the scaling going. For example, although Moore's Law famously described the innovation in the chip industry, Wright’s Law also described the scale of production and cost reduction in the industry. Solar looked unviable as a grid energy technology in the 1980s when it was primarily an asset on satellites but now it's 99.9% cheaper. This is because of both innovation and the production scale it has achieved, which means both work together and are necessary to solve the challenges that we have ahead.
Kiran: Some new technologies have a reasonable ability to be deployed at scale, which is the sweet spot. But technologies like small modular reactors (SMRs) or hydrogen infrastructure are far from mass deployment, which might not be where investment dollars need to be. We also need to factor in the complexity of deploying assets and the long sales cycle, let alone the technology development.
Jessica: The industry itself is ripe for innovation. If you think about the change in the grid industry right now, all three sides (supply, demand, and transmission) are changing at the same time. While the industry has had experience with change and disruption before, we expect a significant challenge - by some estimates, we’ll need to double or triple our grid capacity in the coming decades while reinvesting to maintain the existing infrastructure. On top of that, there’s additional complexity because of regional differences and respective needs. Several regions are in different places in their energy transition and their end states may differ too. Because of this, we need a suite of technology options, of which not all are there currently, especially when we need to deploy at a pace that we've never done before.
5. The number one buzzword for 2024 might be AI, not only the generative AI for texts and images but also the industrial AI applications that accelerate energy technology innovation. How has AI helped your company, if any, and how do you envision it helping you in the future? What are your worries and doubts about AI?
Kiran: We save millions of dollars for a lot of the massive hyperscalers (e.g. data centers) by helping them understand when to shift compute power according to energy tariffs. We also have a GenAI product where we can point at a building and understand the load profile without even having to access the meter. But it’s a tale of two speeds: while the energy industry moves slowly by design, software giants move at radical speed with an insatiable amount of computer power they need now. For example, Amazon just recently bought a nuclear facility in Pennsylvania to build data centers around to get their clean baseload. All of these Big Tech companies will attempt to build utility-scale distributed resources because utilities aren't going to do it for them in time. And I worry that at some point, these hyperscalers may not care as much about clean electrons in their path to get to AGI, and they may make fossil fuel-based plants economic again.
Jessica: We’re predicting doubling the power capacity growth, even before we talk about hyperscalers that increase the challenge. It's going to spur the conversations we haven't had before, which is possibly being in a position of telling people that billions of dollars of economic value need to be deferred because we can't interconnect.
Ted: I hear estimates that are in the terawatts of new dedicated demand for energy generation just for AI when we have already been talking about doubling or tripling global generation capacity by 2040 just to solve climate change under the hypothesis of electrifying everything and decarbonizing the grid. I was also listening to a podcast in which Sam Altman said that for the next couple of decades, compute will be the most valuable commodity that countries will have. And when AI starts to cross that threshold from being something very useful in select use cases to a core building block of national security, you can imagine that there will be no price that any national state actor is not willing to pay to maintain their position in the world order.
6. Looking ahead to the November election potentially affecting the future funding and policy landscape, how do you foresee your company and the industry at large affected, and perhaps more importantly, not affected? How will the ecosystem kind of evolve? Perhaps more collaborators, more competitors?
Jessica: The genie is out of the bottle: solar is cost-effective, people are already gearing up for investments, and the money that's in that bill is distributed over so many areas including both blue and red states. Another element is that we have a lot of state-level activities where even if the federal government pulls back on their commitments and their interests, a lot of states will take over and continue to push for it. Also, specific to VEIR, we didn't get the tax credit, for better or for worse, so the elections probably won’t have a significant influence on our ability to move forward.
Kiran: With the potential Trump administration, I could see the pull-back on uneconomic technologies, such as carbon credits and hydrogen. But the technologies that are already economic will continue to grow because we need them. Solar is cheaper than your local utilities, storage can make it more resilient, and EVs are more fun to drive. On the state front, DGs like community solar are and will continue to be one of the fastest-growing competitive energy trends. For example, we went from 1 state 6 years ago to now with 15 states, especially in California this year. Lastly, there will be more states and utilities looking at large nuclear installations - the three reactors in Georgia were way over budget and way late but the third project was ~60% cheaper than the first.
Ted: We will continue to move forward and make meaningful strides regardless of the outcome. It’s worth noting that the majority of the benefits of the manufacturing tax credits since 2021 have been accruing to the Republican districts because that is where the majority of new manufacturing facilities have been announced.
For any inquiries, suggestions, or collaboration requests – feel free to contact us at energytransitionnotes@gmail.com.
Title image generated with DALL-E on ChatGPT using the prompt: “Draw me a photorealistic image of a panel discussion session titled “Accelerating Energy Technology Innovation” at Harvard Climate Leadership Summit, where a diverse group of panel speakers discuss commercial opportunities, technical constraints, policy support, and future pathways of batteries, grid, and software.”
About the author(s)
Dennis Cha is an MS/MBA candidate at Harvard Business School focused on energy and climate. Prior to Harvard, he worked at Google on hardware supply chain, with PG&E on electric infrastructure operations and wildfire analytics, with SoCalGas on asset decarbonization, and at Taslimi Construction on sustainable construction management. He is trained in civil engineering and data science, and began his career in structural engineering. Outside of work, he enjoys scuba diving, youth mentoring, and traveling.