Policies for Building the Technology Performance Insurance Market

Decarbonizing the global economy will require substantial investment in new industrial-scale climate technologies. A key step in this process is moving from prototype to full-scale demonstration and deployment. Through both government and private-sector investment, these technologies attempt to cross this “valley of death,” often involving considerable expense and risk associated with building at much larger scale. A particular challenge for first-of-a-kind (FOAK) to nth-of-a-kind (NOAK) projects is technological performance risk, where a lack of existing history leads to uncertainty about whether the finished facility will operate as intended. Without a way to guarantee performance, these projects can struggle to find adequate contracted offtake and financing.

One compelling solution lies with insurance, which can shift technological risk from companies or project owners to an insurer at a cost to the insured. Often, the perceived risk of an emerging technology is much higher than the actual risk, leading to a high cost of capital. By accurately assessing and then assuming this risk, insurers can play a key role in helping project developers obtain lower-cost financing. While private insurance providers have historically offered this type of technology performance insurance (TPI), it is often only sparsely available and at a significant premium. At the same time, existing government solutions, primarily in the form of loan guarantees, may be inadequate due to challenges with the application process and less willingness to take on technology risk for projects earlier in the commercialization process. These loan guarantees also do not address issues with securing offtake contracts.

Historically, markets for commercial-scale clean technologies took decades to coalesce, but achieving 2050 climate goals necessitates a much faster pace of deployment. With time of the essence, governments may need to step in to help build the performance insurance market. There are several ways that the government might play a role in increasing the availability of technology performance insurance. First is through an insurance backstop, which incentivizes private insurers to provide insurance up to a certain level of losses, at which point the government steps in. The second is through a federally operated insurance program, similar to the one offered through the Department of Energy’s (DOE) Loan Guarantee Programs Office (LPO), which would draw on the full technical expertise of the government and offer an insurance policy to project developers. Finally, we also consider opportunities for DOE to facilitate greater collaboration between project developers, insurers, and policymakers.

In this report, we provide background on technology performance risk, explore the pros and cons of each of these policy options, and discuss potential design and implementation questions.

Traditionally, in the development of new climate technologies, public funding has supported basic research while private capital covers commercialization of proven technologies. However, capital-intensive FOAK technology pilots and early commercialization projects in between basic research and full commercialization often lack access to capital markets. In this so-called “innovation valley of death” (see Figure 1), technologies may struggle to reach wider adoption without the help of government innovation programs (Hart 2018).

In recent years, the Inflation Reduction Act (IRA) and the Infrastructure Investment and Jobs Act (IIJA) have helped to catalyze some of the investment necessary to cross this valley, including an array of tax credits and tens of billions of dollars earmarked for demonstration and deployment of new technologies. Previous RFF research has explored the potential of other tools for bringing new technologies to market that support demand in decarbonizing industrial processes such as in manufacturing steel and cement (Bergman et al. 2023). Additionally, DOE has funded the Hydrogen Demand Initiative, which has begun to design and implement mechanisms to support demand for its Regional Clean Hydrogen Hubs program (DOE Office of Clean Energy Demonstrations 2024). RFF also has forthcoming research on demand support mechanisms for hydrogen used in heavy-duty trucking and in ports.

While these supports help bridge the valley of death, they are not well-suited to address technology risk. Some new technologies that are modular—that is, where the size of the project or the output grows by additional modules—have costs and performance growth that follow a predictable, linear path. However, the same is not true for scale technologies, where small demonstrations can be disproportionately expensive on a unit cost basis. In both cases and especially for scale technologies, FOAK to NOAK deployment is capital intensive, potentially requiring financing greater than most venture capital, growth capital, or federal grants can provide. At the same time, these technologies lack substantive historical operating data or trends, resulting in start-up, operating, and life-cycle performance risks that are difficult to quantify. Project investors and lenders desire guarantees that production will be sufficient to cover loan payments, deliver on purchasing agreements, or meet other performance targets (Matrix Global). Otherwise, project owners will struggle to secure low-cost financing or need to maintain large cash reserves on their balance sheet, which reflect the cost of repair, replacement, or risk to business continuity. Both of these factors reduce access to capital and stifle growth.

This situation creates a chicken-and-egg problem: project developers and technology providers need capital to demonstrate their technology, but institutional investors and lenders are unwilling to invest due to the risk from lack of experience with the technology. This issue is especially pronounced for smaller, less creditworthy companies that do not have the capability of financing capital intensive projects or guaranteeing performance by themselves (Napiorkowski 2023).

Avenues to address technology performance risk exist in both government and the private sector. However, the availability and applicability of these tools are limited, and the cost may be untenable for earlier-stage projects.

3.1. Private Insurance

One mechanism for addressing technology performance risk is through private insurance and reinsurance markets. Technology performance insurance, which was formed over a decade ago as a way to insure emerging solar and wind projects, effectively transfers risk from the project owner to the insurance market (Napiorkowski 2023). In many cases, the perceived risk in capital markets is much higher than the actual technological risk. Insurers who provide this form of insurance rely on technical expertise, including risk engineers, technologists, and industry experts, to assess the technology (Dickson 2021).
Unlike typical insurance contracts, which insure a wide range of individuals or businesses against damage from risks like natural disasters or cyberattacks, TPI is bespoke and specific to a technology type. Traditional insurance relies on historical data, trends, and customer characteristics to assess risk, but these data points are often not available for new technologies. Instead, the insurer must work closely with the client to understand the technology and make up for this lack of operating history. The insurance provider works with a team of technical experts to fully understand the risks involved with the technology so as to structure and price a suitable product.

TPI is able to cover shortfalls in the event of unexpected technology failure, whether due to performance issues at the project level or the underlying equipment for the technology provider. Policies can vary in structure or how much revenue risk is covered (Napiorkowski 2023). The primary types of insurance are either a “performance guarantee backstop,” which provides additional assurance from technology providers that performance expectations and warranties will be honored in case of a technology failure, or “project cover,” which ensures that projects will meet a target performance and generate enough revenue to cover loan repayment. The “project cover” category can be further broken down into coverage of the output or revenues once a project is operational or coverage of any repairs or delays during the startup phase, when performance must be validated (Dritz 2024). TPI is typically focused on mitigating a project’s “Day 1” risk (in other words, the period from when a project is operational) and does not cover the risks associated with project completion or permitting.

While TPI addresses many of the inherent risks of demonstrating new technologies, challenges with timing and risk appetite still remain. To achieve 2050 net zero goals, technologies must be deployed faster than the extended timeline likely needed to build a financing ecosystem for new technologies without government support. TPI played an important role in developing the solar industry, but the supporting ecosystem and technology took decades to coalesce before its widespread availability (Khaykin and Colas 2021). The insurance industry continues to operate under this timeline, requiring expensive and time-consuming efforts to validate performance as well as sufficient scale before offering TPI insurance.

Another key challenge is the industry’s underinvestment in its own capacity building and research and development. Insurers typically only engage with new technologies after projects have been fully designed and are not involved in the demonstration and early deployment phases (Golnaraghi 2024). This creates a significant hurdle, where risk assessments, technical evidence, and other data necessary for underwriting are not readily available. Thus, significant effort is required from not only the developers of the new technology, but also the insurers. Past examples of the industry working with new climate technologies, such as Chubb’s work in methane emissions reduction technologies (Chubb n.d.) and Munich Re’s work on insurance for long duration storage systems (Martin et al. 2023), have required close collaboration with expert organizations like the Environmental Defense Fund and DOE, as well as a substantial buildout of their own internal capabilities.
As a bespoke product, TPI requires significant time and investment on a technology-by-technology basis from insurance providers. Each technology has unique needs and often lacks a pool of projects in which to spread risk in the short term. Our internal conversations with relevant stakeholders indicate that private TPI may only be available for a limited set of technologies and may have a prohibitively high cost.

3.2. Loan Programs Office

DOE’s Loan Programs Office, which offers loan guarantees and loans to projects that are at the full-scale demonstration and commercialization phase, has some capabilities in guaranteeing performance. While it operates several programs, LPO’s Title 17 (section 1703) Innovative Clean Energy Loan Guarantee Program is especially relevant for new technologies.

By assuming the debt obligation of the investor through its loan guarantees, LPO’s Title 17 (section 1703) program effectively assumes project performance risk. The program requires a comprehensive review of each applicant, which determines a project’s eligibility, technical feasibility, and commercial viability (Jackson 2019). This process is conducted by LPO staff, which includes engineers and finance and legal experts, as well as outside advisors. Passing this review is itself an indication of a project’s commercial viability and its pathway toward offtake and creditworthiness.

For each applicant, LPO determines a credit subsidy cost, which is the expected cost to the government of its loan guarantees considering the time value of money; potential risk factors, such as default; and the estimated recovery rate (see Figure 2). Historically, approved projects were required to pay this cost, which represents one of the most significant financial costs to the applicant (Loan Programs Office n.d.). In the near term, $3.6 billion in credit subsidy funding was appropriated in the IRA, expected to cover the subsidy costs for up to $40 billion in loan authority. However, this funding is only available for projects through September 2026 or until funds run out (Loan Programs Office 2023a).

Several factors limit the applicability of LPO’s work to the commercialization of innovative technologies more broadly:

• The application process is both time and resource intensive, taking up to a year to complete.
• When credit subsidy funding is exhausted, accepted applicants need to pay for their own risk through the credit subsidy on top of an expected 30–50 percent equity commitment, which can discourage potential applicants from applying.
• The riskiness of newer technologies may be mismatched with the risk appetite of LPO. The level of derisking that LPO’s application process requires may be a bad fit for newer, inherently unpredictable technologies. Additionally, a project delay or failure to reach a commitment from LPO is often a negative market signal that impedes a project’s ability to raise capital elsewhere.
• During the diligence process, LPO requires applicants to provide Engineering, Procurement, and Construction (EPC) contracts (Loan Programs Office 2023b). These contracts are designed to shift project risk from the project owner to the contractor in charge of construction. While the ideal form of this contract (and the version preferred by LPO) is “fully wrapped,” where the contractor guarantees a fixed price, fixed date, and a guarantee of project performance (Blackridge Research & Consulting 2023), new technologies may struggle to obtain these terms. More likely, these contracts may not shift an adequate amount of startup risk to the EPC contractor, leaving the door open to project delays or cancellations.

With the limited availability and applicability of existing tools, the government has an opportunity to take a role in mitigating technology performance risk.

4.1. Government Insurance Backstop Program

A public–private partnership between the private insurers and the federal government in the form of a backstop program could reduce insurers’ risk exposure while still giving them skin in the game. Insurers who participate in a backstop program would need to cover up to a certain amount of potential liability for a new technology. The government would pay any additional losses that exceed the covered liability.

Backstop programs have historically allowed insurers to provide insurance even when modeling and pricing risk exposure is difficult. One example is the Price–Anderson Act of 1957, which helps cover liability claims by property owners and members of the public resulting from nuclear power plant accidents. The law places a cap on the liability each plant is responsible for, which encourages private investment in the sector. Under the law, each plant pays an annual premium for $500 million in private insurance. On top of that, a second tier triggers in the event of a nuclear accident with damages and claims in excess of $500 million, where each plant is assessed a prorated share of the excess, up to $158 million per reactor (or $15 billion total) (US Nuclear Regulatory Commission Office of Public Affairs 2024).

Another example is the Terrorism Risk Insurance Act (TRIA), which created an insurance backstop after September 11, when terrorism coverage became prohibitively expensive or unavailable in the market. The law helped to bring reinsurers back to the market by capping their risk exposure in the case of large loss events (National Association of Insurance Commissioners 2023).

These programs might serve as a framework for a technology performance risk backstop. Key questions include:

• Catastrophic vs. Bespoke Risks: Both Price–Anderson and TRIA cover catastrophic risks that are low probability but potentially very high liability. TPI differs in that it is bespoke. TPI represents a market opportunity for insurers, rather than solely an instance where the insurance market is unable to underwrite risk. Insurers should have an incentive to participate in a backstop because it can lead to stronger revenue growth and earnings. In a market-sizing analysis, an estimated $800 billion in capital expenditures for climate technologies in 2030 would imply roughly $10 to $15 billion in associated insurance premiums (Javanmardian et al. 2022).
• Learning, Innovation, and an Off-Ramp: Both Price–Anderson (Rosenthal 1990) and TRIA (National Association of Insurance Commissioners 2023) were originally designed with a finite timeline to overcome an immediate problem. However, both have persisted long after their initial mandate. Because Price–Anderson has been extended into the present day, critics argue that it continues to “[isolate] the nuclear industry from normal market risks while imposing those risks on the public” and reduces incentives for improving safety over time (Rosenthal 1990). On the other hand, proponents believe that the measure maintains the viability of the industry and works well to ensure that potential victims have a guaranteed source of compensation (Holt 2024). For a TPI backstop, the goal should be to spur further innovation in the insurance industry and the underlying technology to reduce potential risks. Designing a backstop program for TPI requires consideration for how technologies “graduate” from the program. The purpose should be to address risk in FOAK to NOAK deployments, not mature technologies.
• Pooling Risk: For insurance to be efficient, risk could be spread over a large number of projects. A backstop is particularly useful when the number of projects is limited, but less necessary when a sufficient pool is created. Thus, another key goal of a TPI backstop would be to catalyze the formation of pools of projects and even pools of technologies across participating insurers and reinsurers. Doing so could reduce the cost of insurance and accelerate the technology’s pathway to commercialization (Golnaraghi 2024).

With these considerations in mind, below (and Figure 3) are some of the key design elements a backstop program must consider:

• Threshold for federal assistance eligibility: For some backstops like the TRIA, only losses past a certain threshold are eligible for federal assistance. As new technologies commercialize, measured by metrics such as number of deployed sites globally or hours of operation, this threshold could be raised over time to gradually shift risk back to the private sector.
• Insurer deductible: Insurers cannot feasibly cover all performance losses without ballooning the cost of insurance. One solution is that a backstop program should include a deductible before insurance steps in. Historically for TPI, this could range from 10–40 percent of the expected baseline performance depending on the specific project or technology characteristics (Napiorkowski 2023).
• Share or cap above deductible covered by government: Once insurance kicks in, one must determine how much of the loss could be covered by the government. The government can either fully subsidize a share of all losses (as is the case with TRIA) or cap an insurer’s responsibility at a certain threshold. This structure—and the associated liability share or cap—is important for determining the program’s potential cost to the government.

4.2. Government-Operated Insurance Program

Instead of a government backstop, the federal government might also provide its own TPI program, akin to the loan guarantees from LPO. Box 1 below provides an overview on the design of LPO’s credit subsidy, which could be a model for a federal insurance program. In both cases, the goal is to protect the project loan issuers (or guarantor in the case of LPO) against the risk of default. Additionally, the government may have more built-in capacity than private insurers to assess emerging climate technologies. Such a program would likely require the government to either commit significant technical expertise to accurately assess and price the risk of new technologies or offer a subsidized premium to project developers.

Many examples of government-run insurance programs exist, and perhaps the most relevant policy example is the Multilateral Investment Guarantee Agency (MIGA). MIGA is a member of the World Bank Group, sponsored by both developing and industrialized countries to encourage foreign direct project investments in developing countries. The agency offers political risk insurance products that cover five key areas: currency risk; government expropriation; war and terrorism; breaches of contract; and the non-honoring of financial obligations. MIGA’s guarantees allow investors and lenders to commit capital to riskier projects in developing countries.

Any insurance product introduces moral hazard, where policyholders are less likely to address issues and risks than they otherwise would be. For government-operated insurance, addressing these issues is imperative to ensure that programs are effective. In the case of TPI, moral hazard could complicate the pathway to commercialization, as manufacturers and project developers lack incentive to address risks that are being subsidized (International Finance Corporation and Multilateral Investment Guarantee Agency 2022).

Finally, a government-operated program has very different goals from those of a private insurer. Where government has social objectives, private insurers are profit driven. A federal TPI, especially one that subsidizes premiums, could make it difficult for private insurers to compete or build their internal risk engineering capacity for new technologies. If the ultimate goal of the program is for new technologies to commercialize without federal assistance, then insurance companies will still need to play a role in the market.

4.3. Facilitate Engagement among Insurers, DOE, and Industry

Much of the hesitancy around insuring new technologies revolves around a lack of available historical data to write and price policies. Insurers are typically not involved until projects have been fully conceptualized. DOE could potentially facilitate collaboration among project developers, manufacturers, insurance providers, and policymakers to enable earlier engagement on technology risk. Collaboration offers three key benefits:

• Benefit to Insurers: In most cases, as companies develop decarbonization technologies, they have interacted with DOE offices, whether through grants, prizes, loans, or other financing opportunities. The agency is well-positioned to provide data, risk factors, and assistance in identifying existing projects from early in the demonstration stages. This information is valuable to insurers as they assess potential risks. It could also allow insurers time to proactively, rather than reactively, build up risk engineering assessment capacity for technologies reaching deployment.
• Benefit to Project Developers: Insurance risk engineering teams can also play an important role earlier in the project design process, rather than after a project is fully conceptualized. Earlier involvement allows insurers to advise on project risks, including climate, equipment, construction, and operation. Moreover, insurers have also historically played an important role in setting risk management standards and codes of practice for new technologies. For example, in the offshore wind industry, insurers helped to identify and categorize risks as a set of best practices and standards for replication for the industry (Golnaraghi 2024).
• Benefit to Policymakers: Policymakers at DOE and elsewhere in the federal government also stand to benefit from this collaboration. Identification and assessment of risks by insurers and project developers can inform the development of federal standards and regulation, especially around the safety of new technologies.

In addition to what role the government might play in addressing technological performance risk, other important considerations are cost to the government, risk management participation by other stakeholders, and feasibility.

• Cost to the government is difficult to estimate for either a backstop or government-run program without details on which emerging technologies are included and additional data on premiums or risk of default for each technology. However, a backstop reduces the cost to the government somewhat because private insurers still have some, albeit more limited, participation. In contrast, a government-run program is exposed to much larger costs, especially when subsidized premiums fall below actual risk—as is the case with the National Flood Insurance Program. One model to consider in estimating the cost to government is the LPO’s credit subsidy calculation (see Figure 1 and Box 1).
• Other stakeholders outside of the technology developer, project owner, and insurer will be necessary to reduce project risk and accelerate commercialization. While insurance can play a key role in derisking technology, other risks, such as construction and permitting, could still affect a project. For example, EPCs that are not fully wrapped may instead only offer time and materials contracts, which pay the contractor based on the time and materials needed, without a cap on either. In this case, project owners and investors could be exposed to significant cost overruns. It is important to ensure that EPCs and other stakeholders, such as offtakers and feedstock providers, have skin in the game and an incentive for the project to meet cost and schedule demands. Permitting delays are also a key obstacle to commercialization and can expose projects to significant cost overruns and affect the project’s economic viability. A suite of solutions may be necessary and would require both additive policies (such as insurance) and reform (such as permitting).
• Political/institutional feasibility is also an important consideration for any of these potential policy pathways. Both a backstop and a government-operated program would require new legislation, including appropriations. Historically, the government taking on levels of risk common in the private sector has led to political blowback. Failures and payouts are to be expected—even desired—as an indication of a functioning system. However, the lack of risk appetite among those implementing the program as well as in Congress’s criticism when inevitable failures occur are both potential impediments to a successful program.

While new legislation like the IRA and IIJA have laid out opportunities to commercialize new technologies and “bridge the valley of death,” more work can be done to improve the financing environment. Risk management is an essential part of financing new projects, and the ability to mitigate these risks, potentially through insurance, will play an important role in the pathway to commercialization. In this paper, we have outlined several paths by which the federal government can facilitate the market for technology performance insurance and help drive the deployment of new clean technologies. Much work remains to be done to understand the implementation and implications of this policy idea. We look forward to testing these ideas with additional stakeholders and elaborating the most promising pathways.