Skip to content
Chimera readability score 79 out of 100, Expert reading level.

Dive Brief:
- As large-load customers increasingly look to power their operations with new clean firm technologies like advanced geothermal and small modular nuclear reactors, clean energy tariffs, as well as bilateral customer-utility arrangements, are helping to commercialize and de-risk the deployment of these technologies, the Corporate Energy Buyers Association said in a May 21 report.
- “The goal is to try and get some of these technologies closer to commercialization, and then we'd see costs shift down, like we did with solar and wind,” lead report author Priya Barua, CEBA's senior director of utility partnerships and innovation, told Utility Dive. “I think we're in that phase with that next wave of technologies.”
- Clean energy tariffs “allow large customers to identify, fund, and support specific clean or clean firm resources,” the report said, “within existing regulatory frameworks, pairing customer-driven investment with strong ratepayer protections.”
Dive Insight:
Hyperscalers are spending aggressively on the AI race, including the construction of data centers and the energy needed to power them. The five largest hyperscalers — Google, Meta, Amazon, Oracle and Microsoft — are expected to cross $1 trillion in capital expenditures this year, according to venture capital and private equity firm Bessemer Venture Partners, and $8 trillion total between 2025 and 2031.
Those five companies also “represented 49% of all corporate clean energy [power purchase agreements] globally in 2025,” the CEBA report said, “with a growing share of investments in nuclear, hydropower, and geothermal.”
These investments include a PPA between Microsoft and Constellation Energy, which is driving the restart of Three Mile Island Unit 1, and a 115-MW PPA between Google and advanced geothermal startup Fervo Energy under a Clean Transition Tariff that Google and NV Energy proposed and got approved by the Public Utilities Commission of Nevada.
“In a lot of the partnerships that have been established around some of these technologies, it's really the tech companies that are taking on a lot of the risk,” Barua said. “And so, in addition to essentially safeguarding non-participating customers from a cost perspective, that risk is being taken on, and there's a lot built into these tariffs to really protect the rest of the ratepayers from those risks.”
Small modular reactors are increasingly drawing investments from hyperscalers looking to power future data center operations, thanks in part to their smaller footprint when compared to a traditional nuclear facility, though no SMRs are operating yet in the U.S.
The report identifies three characteristics which the authors associate with “durable” clean energy tariff designs: the allocation of “associated costs exclusively to participating customers”; the embedding of capacity value into rate structures; and “long-term revenue certainty necessary for capital-intensive, clean firm resources.”
The report authors also noted that large load tariffs and clean energy tariffs “are increasingly linked.”
“New state-mandated, large load tariffs establish minimum demand charges, collateral requirements, and long contract terms to protect ratepayers,” the report said. “Clean energy tariffs often build on these foundations, giving customers a pathway to pair load growth with new capacity, improving system economics rather than exacerbating them.”
Barua said she thinks large load customers are “looking at a broader suite of clean firm technologies than they were five years ago, just because of where we are right now, and the recognized need for more capacity. So I do think that as a result we're seeing a lot of innovative partnerships between customers and utilities.”
She pointed to Minnesota, where Google and Xcel Energy utilized the state's Clean Energy Accelerator Charge, a specialized green energy tariff, to bring online a portfolio of resources that includes solar, wind and 300 MW of long duration energy storage, “the biggest long duration energy storage system that that we've seen.”
“Again, that is a new technology, and in that arrangement, Google is taking on some of the risks of bringing that technology onto the system,” Barua said. “But we will see system benefits from it.”

Facts Only

Clean energy tariffs help large customers identify, fund, and support specific clean or clean firm resources within existing regulatory frameworks.
The goal of these arrangements is to move technologies closer to commercialization and shift costs down, similar to solar and wind.
Clean energy tariffs allow pairing customer-driven investment with strong ratepayer protections.
The five largest hyperscalers (Google, Meta, Amazon, Oracle, and Microsoft) represented 49% of all corporate clean energy power purchase agreements globally in 2025.
Investments in nuclear, hydropower, and geothermal are growing among the hyperscalers.
A PPA exists between Microsoft and Constellation Energy, which is driving the restart of Three Mile Island Unit 1.
A 115-MW PPA exists between Google and Fervo Energy under a Clean Transition Tariff.
The partnerships often involve technology companies taking on a lot of the risk.
Durable clean energy tariff designs allocate associated costs exclusively to participating customers.
Large load tariffs and clean energy tariffs are increasingly linked.

Executive Summary

Large load customers are driving the commercialization and de-risking of clean firm technologies like advanced geothermal and small modular nuclear reactors through clean energy tariffs and bilateral customer-utility arrangements. Clean energy tariffs allow large customers to fund and support specific clean or clean firm resources while maintaining ratepayer protections. This dynamic is evident in partnerships between hyperscalers (Google, Meta, Amazon, Oracle, and Microsoft) and energy providers. These investments include Power Purchase Agreements (PPAs) and Clean Transition Tariffs, such as one involving Microsoft and Constellation Energy to restart a nuclear unit, and another involving Google and Fervo Energy for geothermal development. The report suggests that technology companies are taking on significant risk in these partnerships, which is partially mitigated by tariff structures designed to protect other ratepayers. The report also identifies characteristics of durable tariff designs, including allocating associated costs exclusively to participating customers and embedding capacity value into rate structures.

Full Take

The narrative frames the development of clean firm technologies as a successful mechanism for risk transference and cost shifting, driven by the financial power of hyperscalers. The pattern observed is the systematic movement of capital-intensive risk away from the end-user (hyperscalers) and onto the utility and system, shielded by tariff structures. This creates a dependency where "innovation" is intrinsically linked to utility-backed mechanisms.
The implicit assumption is that the regulatory structures—specifically the links between large load tariffs and clean energy tariffs—are robust enough to ensure that cost-shifting benefits the broader ratepayer, rather than simply absorbing costs. This requires scrutiny into whether the "protection" provided by these mechanisms is truly sufficient or if it merely redirects cost burdens. The fact that tech companies are taking on risk while costs are protected suggests a complex, layered negotiation of risk that must be examined for true fairness.
The focus on partnerships where technology companies take on risk raises questions about whose definition of "risk" is being used and who ultimately bears the systemic and financial consequences if these arrangements fail. The emphasis on "long-term revenue certainty" for capital-intensive resources suggests a structural bias favoring long-term, large-scale infrastructure investments over immediate, localized cost management.

Sentinel — Human

Confidence

The text functions as high-quality journalistic synthesis of complex energy finance topics, demonstrating strong contextual understanding and careful sourcing.

Signals Detected
low severity: Sentence length variance is varied; the flow is analytical rather than metronomic.
low severity: The text successfully integrates multiple disparate facts (hyperscalers, SMRs, tariffs, PPA details) into a cohesive narrative.
low severity: Specific data points (e.g., $1 trillion CAPEX, 49% global share, specific PPA examples) are used to anchor the argument, suggesting rooted sourcing.
Human Indicators
The specific citation of named reports (CEBA, Bessemer Venture Partners) and highly specific examples of contracts (Microsoft/Constellation, Google/Fervo) suggest reliance on specific, traceable industry reporting.
The nuanced argument distinguishing between 'large load tariffs' and 'clean energy tariffs' demonstrates a specific, expert-driven analytical framework.
Large-load customers can help commercialize new clean energy technology: CEBA — Arc Codex