Energy Weekly Review 2026-06-05

Week In Review

The week’s news converged on one stubborn question for the energy transition: how do you serve loads that grow faster than the grid can plan for them? Two pieces from MIT Technology Review and Utility Dive made the point bluntly. A feature on virtual power plants as a data-center resource argued that aggregating millions of small assets — EV chargers, water heaters, batteries — can shave hours off the worst hyperscaler demand peaks faster than any new plant could be built, and Google promptly underwrote that thesis with a first-of-its-kind funding deal for a 100 MW virtual power plant in the PJM grid operated by Voltus. Hyperscaler money is no longer just chasing nuclear PPAs; it is now paying to wring flexibility out of the existing load.

On the supply side, the American Clean Power Association’s first-quarter numbers, reported by PV Magazine USA, marked a quiet historic crossover: utility-scale solar capacity in the United States is now larger than wind for the first time, while a pipeline of more than 50 new utility-scale solar, wind, and storage projects was announced as developers raced to lock in tax credits ahead of looming federal deadlines. The numbers documented a buildout dominated by photovoltaic panels and lithium batteries, with onshore wind reduced to a fraction of its historical share of new additions.

Advanced nuclear had perhaps its busiest single week of the year. The American Nuclear Society’s June industry update collected a half-dozen substantive moves at once: X-energy filed to enter the UK’s Generic Design Assessment for its Xe-100 high-temperature gas-cooled reactor; Oklo and Sweden’s Blykalla expanded a transatlantic partnership to commercialize a 55-MWe lead-cooled fast reactor in the United States; a Utah joint venture committed to building a Holtec SMR-300 at Green River; and a molten-salt reactor test facility opened in Eindhoven. World Nuclear News separately reported partnerships unlocking Deep Fission’s mile-deep “Gravity” borehole reactor and additive-manufactured components for NX Atomics. UK-based First Light Fusion closed a £25 million first tranche of fresh funding anchored by the UK Atomic Energy Authority, a vote of confidence in the projectile-driven inertial confinement approach.

Geothermal and long-duration storage rounded out the week. Fervo Energy’s Cape Station enhanced geothermal project in Utah locked in offtakers for an expanded 500 MW build-out as it approaches first power, and Energy-Storage.News documented a wave of pumped-hydro, high-temperature thermal, and geothermal projects advancing across the US — including Sage Geosystems’ $97 million Series B to take its pressurized-rock storage approach to commercial scale. The week made one trend hard to miss: a year ago, “firm clean power” mostly meant a slide labeled “nuclear coming soon.” This week it meant signed offtake agreements, funded balance sheets, and shovels in dirt for at least four distinct technologies competing for the same demand.

Items

Virtual Power Plants Emerge as a Credible Tool for Data-Center Demand

MIT Technology Review made a careful case this week that virtual power plants — software platforms that aggregate thousands or millions of small flexible loads and batteries — may be one of the only resources that can scale fast enough to absorb hyperscaler demand growth without either firing up retiring coal plants or waiting a decade for new transmission. The argument hinges on a simple asymmetry: large data centers need firm capacity for only a small fraction of hours each year, and existing distributed energy resources, properly orchestrated, can cover most of those peaks at a fraction of the cost of new generation.

The piece walks through pilot results from utilities that have used residential batteries, smart thermostats, and managed EV charging to defer hundreds of megawatts of conventional capacity. It also engages honestly with the limits: most VPPs today are measured in tens of megawatts, not gigawatts, and the contracts, market structures, and metering systems needed to scale them up are still being invented. Where the article breaks from boosterism is in noting that the most promising VPPs are not residential-only but blend behind-the-meter and front-of-meter assets so that a single platform can dispatch both a homeowner’s water heater and a 50 MW grid battery.

For a sector where the dominant policy conversation has been “build more nuclear” versus “build more gas,” the framing is useful. AI training and inference loads are not flat baseload; they are spiky in ways that look more like industrial demand than residential demand. Tools designed for industrial demand response — paying customers to drop or shift load on short notice — translate naturally to AI workloads that can themselves be paused or relocated. The article suggests this is no longer a fringe idea inside utilities; it is the position several major load-serving entities are quietly converging on.

Source: MIT Technology Review

Google Funds a 100 MW Virtual Power Plant in PJM Through Voltus

Within days of the MIT piece, Google announced what it described as a first-of-its-kind deal: it will fund the development of a 100 MW virtual power plant in PJM Interconnection, the largest US grid operator, run by distributed-energy aggregator Voltus. Under the agreement, Voltus will recruit and orchestrate EV chargers, smart thermostats, residential and commercial batteries, and other flexible loads, paying participants to curtail or shift consumption when the PJM grid is stressed.

The deal is structurally novel because Google is not buying the VPP’s output as electricity. Instead, it is providing capital to stand the VPP up in exchange for grid services that effectively counterbalance the load Google itself adds to PJM through its data centers. The arrangement is closer to a corporate offset purchase than a power purchase agreement, but the asset being created is real, measurable, and dispatchable rather than a paper credit.

PJM has been the focal point of the AI-data-center grid crisis: capacity prices in its most recent auction set records, and queue interconnection delays have stretched to five or more years. A 100 MW VPP is small compared to a single hyperscale campus but is one of the few resources that can be assembled in months rather than years. If the model works, the precedent matters more than the megawatts — it gives hyperscalers a template for paying their way out of the load they create, without waiting for new generation.

Source: Utility Dive

US Solar Pipeline Surges as Solar Overtakes Wind in Installed Capacity

The American Clean Power Association’s first-quarter 2026 market report, summarized by PV Magazine USA, recorded 6.4 GW of new utility-scale clean power capacity added in the first three months of the year: 3.6 GW of solar, 2.4 GW of energy storage, and just 415 MW of wind. With the quarter’s additions, the cumulative US fleet of utility-scale solar — counted in nameplate capacity — has now passed the wind fleet for the first time, a milestone that would have seemed implausible a decade ago when solar was a fraction of wind’s size.

The more arresting number is the pipeline. Developers announced more than fifty new utility-scale projects during the quarter, racing to break ground before federal tax credit deadlines that have grown more important as energy policy has shifted. Texas alone now hosts 96.4 GW of installed clean capacity, putting it on track to become the first state to cross 100 GW. Storage continued its run as the fastest-growing category, with developers increasingly pairing batteries with new solar at a roughly one-to-one ratio.

Wind’s collapse to fewer than 0.5 GW of additions in a quarter is the other story embedded in the data. Onshore wind has struggled with siting and supply-chain headwinds, while offshore wind has faced US federal policy reversals. The result is a US clean-energy buildout that, for the foreseeable future, will be dominated by photovoltaic panels and lithium batteries, with wind playing a smaller role than industry forecasts assumed even two years ago.

Source: PV Magazine USA

X-energy Files to Enter UK Generic Design Assessment with Its Xe-100 Reactor

US advanced reactor developer X-energy submitted an application this week to enter the United Kingdom’s Generic Design Assessment process for the Xe-100, a 320 MWth (around 80 MWe per module) high-temperature gas-cooled reactor that uses TRISO-coated particle fuel and helium coolant. GDA is the UK’s pre-licensing review used to clear new reactor designs for site-specific applications. Successfully completing it would make the Xe-100 one of a small handful of advanced reactor designs eligible for UK deployment.

The Xe-100 is one of the more mature non-light-water designs in the West. It is the design Dow Chemical has been working with X-energy to deploy at a Gulf Coast chemical plant under a DOE Advanced Reactor Demonstration Program award, and Energy Northwest has selected it for a project in Washington State. The reactor is designed to operate at outlet temperatures around 750°C, hot enough to drive industrial steam and hydrogen production directly, which is the property that makes it interesting to chemical and industrial customers in addition to utilities.

A GDA filing is mostly a regulatory milestone, not a technical one, but it signals that X-energy now sees a credible commercial path in three jurisdictions — US, Canada (already in the CNSC pre-licensing process), and the UK — and is willing to fund the multi-year regulatory submissions to pursue them in parallel. For Britain, which has set explicit policy goals around SMR deployment but has so far made only modest progress, adding a non-light-water entrant to the GDA queue alongside Rolls-Royce and GE Hitachi widens the design pool meaningfully.

Source: ANS Nuclear Newswire

Deep Fission Locks In Construction Partner for Mile-Deep “Gravity” Reactor

US startup Deep Fission announced that engineering services firm Day & Zimmermann will oversee pre-construction planning and above-ground construction for the company’s Gravity reactor, a 15 MWe pressurized-water reactor designed to be placed at the bottom of a mile-deep, optimised borehole. The concept uses the column of water above the reactor to provide passive containment pressure and a natural radiological barrier, eliminating the massive concrete containment structure that dominates conventional plant designs.

The premise is unusual enough to be worth stating plainly: instead of building a containment dome strong enough to contain a worst-case accident, the reactor sits inside a sealed, kilometre-deep shaft drilled into bedrock, with the surrounding rock and water doing the work. The approach trades unfamiliar drilling and retrieval challenges for radically simplified surface infrastructure, and Deep Fission argues the resulting plant footprint can be small enough to colocate with industrial loads or data centers without the typical exclusion zone.

The partnership with Day & Zimmermann is notable because it represents the first time a major established US energy construction firm has committed to a borehole reactor concept. Engineering credibility is the bottleneck many advanced reactor startups hit just before licensing, when regulators and customers want to know who will actually build the plant. The deal moves Gravity from a paper design closer to a deployable system, though licensing through the NRC remains the more difficult hurdle ahead.

Source: World Nuclear News

Oklo and Blykalla Expand Transatlantic Partnership on Lead-Cooled Fast Reactors

California-based Oklo and Sweden’s Blykalla widened a previously announced strategic partnership this week, formalizing a plan to bring Blykalla’s 55 MWe SEALER lead-cooled fast reactor to the US market alongside Oklo’s existing sodium-cooled Aurora design. Under the expanded agreement, the two companies will share supply-chain development, regulatory engagement, and project pipeline work, with Oklo positioned to support US deployment of SEALER as a complement to Aurora rather than a competitor to it.

Lead-cooled fast reactors are an old idea — derived from Soviet submarine reactor technology — that has been revived because lead has favourable properties for passive safety: it is opaque to gamma radiation, has a very high boiling point, and is chemically inert with water and air, which simplifies emergency response compared with sodium. Blykalla has been developing SEALER as a small, factory-built unit intended for industrial customers and remote sites, and the company has a multi-year roadmap toward a Swedish demonstration plant.

The expanded Oklo partnership is strategically interesting because it groups two distinct fast-reactor coolant approaches under a single commercial wrapper, hedging the technology risk for customers who care more about reliable low-carbon heat than about coolant chemistry. With the US Nuclear Regulatory Commission’s new Part 53 framework now finalized for advanced reactors, the practical question for both designs is no longer whether they are licensable but whether they can be built at a cost that justifies the engineering effort. Pooling commercial resources is one way to attack that problem.

Source: ANS Nuclear Newswire

First Light Fusion Closes £25 Million Round Anchored by UK Atomic Energy Authority

UK fusion startup First Light Fusion announced the successful first close of a £25 million funding round this week, led by East X Ventures’ fusion fund Starmaker One and including a strategic investment from the UK Atomic Energy Authority. First Light pursues an unusual fusion approach called projectile-driven inertial confinement: instead of imploding a fuel pellet with lasers, the company fires a high-velocity projectile at a target that focuses the resulting shock wave to compress and heat the fuel.

The technology lineage traces back to “amplifier” work originally developed for shock-physics research. The pitch is that mechanical drivers are far cheaper, more efficient, and easier to scale than the gigantic laser systems used at facilities like the National Ignition Facility, and that a commercial plant could in principle be built with off-the-shelf industrial components rather than bespoke optics. The trade-off is that projectile drivers produce slower implosions, which has historically been a problem for reaching the conditions needed for net energy gain.

The UK Atomic Energy Authority’s involvement is the more strategically important piece of the announcement. UKAEA is the public body running the JET tokamak and the planned STEP fusion demonstration plant, and its imprimatur on a non-tokamak approach signals that the UK is hedging across multiple confinement schemes rather than betting solely on magnetic fusion. The funding is modest by the standards of Commonwealth Fusion Systems or Helion, but for a relatively contrarian technical approach it is enough to keep First Light’s experimental program running through its next set of compression milestones.

Source: Nuclear Industry Association

Fervo Energy Locks In 500 MW of Offtake for Cape Station Enhanced Geothermal

Fervo Energy expanded its Cape Station enhanced geothermal project in Utah to 500 MW and locked in long-term offtake agreements covering Phase 2 capacity, in addition to the Phase 1 contracts already in place with Southern California Edison and Clean Power Alliance. Phase 1 of the project — the first 100 MW — is on track to begin delivering power to the grid in late 2026, which would make Cape Station the first commercial-scale enhanced geothermal system to come online anywhere in the world.

Enhanced geothermal systems (EGS) are an attempt to make geothermal power site-agnostic. Conventional geothermal plants depend on rare natural reservoirs of hot, permeable rock and fluid; EGS uses horizontal drilling and hydraulic stimulation — techniques borrowed from the oil and gas industry — to create artificial reservoirs in dry hot rock that occurs nearly everywhere at sufficient depth. Fervo’s Cape Station has been the field test of whether those techniques can be combined economically at utility scale. Well tests last year established Cape as the most productive enhanced geothermal generating station in history, with the field achieving flow rates and temperatures that the National Renewable Energy Laboratory had not expected to be reached until 2035.

The new offtake agreements are the commercial flip side of the technical milestone. Geothermal has long been able to make a compelling story about firm, 24/7 clean power but has struggled to put steel in the ground at scale because the upfront drilling risk has scared off financing. Cape Station’s 500 MW pipeline, fully contracted to investment-grade buyers, demonstrates that the financing model now exists for EGS, not just the geology.

Source: Power Engineering

Pumped Hydro, Thermal Storage, and Geothermal LDES Projects Advance Across the US

Energy-Storage.News this week catalogued a notable cluster of long-duration energy storage (LDES) projects moving forward simultaneously across different technologies in the United States. The list includes a pumped-hydro project advancing toward construction in the western US, a new high-temperature thermal energy storage installation commissioned by Electrified Thermal Solutions at Southwest Research Institute in Texas using ceramic-brick “Joule-Hive” modules, and several enhanced geothermal projects securing site control and financing.

The pattern matters because lithium-ion batteries — the dominant grid-storage technology of the past decade — saturate economically at durations of around four to six hours. Above that, the cost of additional cells outweighs the value of the additional energy stored. Closing the gap between four-hour batteries and seasonal storage requires technologies that store energy more cheaply per kilowatt-hour but with slower charge/discharge cycles, which is exactly the niche pumped hydro, thermal storage, and geothermal-as-storage are designed to occupy.

The roundup notes that all three approaches are now seeing real revenue commitments rather than just R&D grants. Pumped hydro is benefiting from new permitting reforms, thermal storage is finding customers in industrial decarbonization, and geothermal is increasingly being marketed as both a generator and a long-duration storage asset, since flow rates can be modulated against demand. Together, they sketch the outlines of a grid where LDES is no longer the missing layer of the stack but a contested market with multiple credible bidders.

Source: Energy-Storage.News

Sage Geosystems Closes $97 Million Series B to Commercialize Pressurized-Rock Storage

Geothermal and energy storage startup Sage Geosystems closed a $97 million Series B funding round this week, led by Ormat Technologies and Carbon Direct Capital, according to an Energy-Storage.News roundup. Sage occupies an unusual niche: rather than using geothermal heat to make electricity, its core technology stores electricity as pressurized water injected into deep hot rock formations, which is later released through a turbine to generate power on demand. The company calls the approach “geopressurized geothermal,” and it functions as a long-duration storage technology rather than a pure generation technology.

The pitch is conceptually similar to pumped hydro but does not require a mountain reservoir. Instead, the formation pressure of deep rock acts as the “upper reservoir,” storing energy as elastic strain in the rock and pressure in the trapped water. Round-trip efficiencies are lower than lithium-ion but multi-day discharge durations are achievable, and the capital cost per kilowatt-hour of storage is projected to fall well below batteries at scale. Sage has been running pilot tests in Texas alongside utility partners, including a contract with Meta for an early commercial deployment to support a data center.

Ormat’s involvement as lead investor is the part of the announcement most worth attention. Ormat is the largest pure-play geothermal company in the West and has decades of operating experience in conventional flash and binary plants. Its participation signals that incumbent geothermal expertise is willing to underwrite a fundamentally different application of the same drilling and reservoir engineering. If Sage’s approach proves out at commercial scale, it would give the grid a long-duration storage asset that uses an industrial workforce and supply chain already developed for fossil energy — which is a non-trivial advantage in a buildout where everything else is supply-constrained.

Source: Energy-Storage.News