A $67 billion NextEra-Dominion merger announcement, a 220 GW PJM Cycle 1 intake with 17.9 GW of nuclear inside it, a 500 MW Hydrostor A-CAES entry into Ontario’s long-duration RFP, FERC holding the at-risk readiness deposit line, a $450 million Spearmint BESS close in three months to COD, a world-first second-life BMS safety certification at Moment Energy, and an independent finding of hexavalent chromium and arsenic in Tesla’s Robstown refinery discharge. Four verticals lit up in seven days; the cross-vertical thread is AI-demand routing through PJM and through IRA-era policy plumbing at the same time.

The lede

Last week’s digest closed with the rebalance read consensus across the lithium operator set and the next analytical edge sitting on the risk side, not the demand side. This week the cycle of attention rotated. Lithium operator earnings were quiet (SQM’s Q1 lands next week and remains the cleanest single read on whether the four-leg rebalance gets a fifth confirming print). What activated instead, all in seven days, was the four other verticals: nuclear and grid through the same PJM channel, solar through the FERC enforcement of queue discipline, critical minerals through a clear thesis-risk data point at Tesla’s Robstown refinery, and storage through both a long-duration capital event (Hydrostor, Ontario) and an ERCOT financial close (Spearmint, Red Egret) that captures the IRA transferability stack at its most mature form.

The thread that ties the week together is hyperscaler load showing up simultaneously inside three different verticals via the same RTO. NextEra agreed to acquire Dominion at $67 billion to consolidate the generation footprint behind Northern Virginia. PJM’s Cycle 1 reformed-queue intake brought in 220 GW of new applications, with the nuclear share alone at 17.9 GW, a figure that prior cycles did not approach. FERC affirmed PJM’s at-risk Readiness Deposit framework, the mechanism that keeps cost discipline inside the cleaned-up Transition Cycle. None of those three events makes sense without AI-driven hyperscale load. All three landed in five business days.

The framework does not bend. It adds load-bearing data points across three verticals at once.

Top stories by vertical

Nuclear: NextEra to acquire Dominion at $67 billion, second-largest US nuclear operator on closing

NextEra Energy and Dominion Energy announced on May 18 an all-stock acquisition at roughly $67 billion, a 23 percent premium on Dominion’s prior market capitalization. Both boards approved the deal. Management guided a closing window of 12 to 18 months, pending FERC, NRC, and three state PUC clearances (Virginia, North Carolina, South Carolina). On closing, the combined entity becomes the largest US regulated electric utility by customer count (about 10 million) and the second-largest US nuclear operator behind Constellation. Dominion currently generates more than 40 percent of its electricity from nuclear, with operating stakes at Millstone, North Anna, Surry, and VC Summer. NextEra adds Seabrook, Point Beach, and Duane Arnold (the latter being restarted under a Google PPA).

The strategic logic is data center load. Dominion is the incumbent utility for Northern Virginia, the densest concentration of hyperscaler campuses in the world, and it sits inside PJM. NextEra brings the largest US renewables and storage development pipeline. The companies cite a combined 130 GW large-load project pipeline. CEO John Ketchum’s framing to investors: “The demand for electricity is increasing unlike anything we’ve seen in generations.” (See: news/2026-05-20-nextera-dominion-67b-merger.)

Two structural reads against the nuclear thesis frame. First, a single merger would put roughly a quarter of US operating nuclear capacity under two corporate balance sheets (Constellation plus NewCo), which changes the negotiating dynamic for hyperscaler PPAs, restart financing, and SMR siting decisions inside the nuclear host states. The “renaissance” the published frame names is now concentrating supply-side leverage at the same time it is concentrating demand-side procurement, and that combination raises rather than lowers the price-discovery story over the next 24 months. Second, state utility commission approvals are the gating item, not federal. Virginia SCC and the two Carolina commissions all sit between announcement and close, and consolidation reviews routinely overrun the 12-to-18-month guidance. The deal advances the thesis but does not de-risk the timeline.

Grid & Transmission: PJM Cycle 1 closes at 220 GW, nuclear posts 17.9 GW of queue entry; FERC holds the at-risk readiness deposit line

PJM disclosed on April 29 (and the implications got walked through this week) that Cycle 1 of its reformed interconnection process drew 811 generation projects totaling 220 GW. The application window closed April 27. A 91-day review runs through July 27, model posting is scheduled for June 26, and Phase I studies begin July 28. Projects under the new cadence are expected to clear in one to two years rather than four-plus.

The fuel mix is the read worth holding onto:

  • Natural gas: 105.8 GW across 157 projects. Dispatchable capacity chasing data-center load.
  • Storage: 66.5 GW across 349 projects. Largest single project count in the cycle; firming + capacity-market arbitrage.
  • Nuclear: 17.9 GW. Prior PJM cycles had nuclear in the low single-digit gigawatts or absent entirely. Some portion of the 17.9 GW maps to the announced hyperscaler restart and small-reactor megadeals (Constellation at Three Mile Island, Vistra, X-energy, Kairos, TerraPower, Oklo); the remainder is utility-led or merchant.
  • Solar: 14.8 GW. Notably lower than the speculative pre-reform pipeline; survivors of the readiness screen.
  • Solar-storage hybrids 8.9 GW, wind 4.7 GW, hydro 0.15 GW, other 0.5 GW.

Two days later, on May 21, FERC denied a section 206 complaint by Gaston Green Acres Solar and Bethel NC Hwy 11 Solar challenging PJM’s at-risk Readiness Deposit framework. The complaint argued the framework was unjust because it does not permit Transition Cycle 1 customers to withdraw without forfeiting deposits when network upgrade costs jump materially between Phase III and the final retool study. FERC sided with PJM, holding the framework intact. Readiness Deposit 1 stays at $4,000 per MW of project size; Readiness Deposit 2 stays at 10 percent of allocated Phase I upgrade cost minus Deposit 1. Walking at retool costs both deposits. (See: news/2026-05-22-ferc-affirms-pjm-readiness-deposit, news/2026-05-23-pjm-cycle-1-220gw-nuclear-storage-signal.)

Against the published grid thesis (transmission is the binding constraint; the queue is where the constraint shows up; FERC Orders 1920 and 2023 are the reform vehicles whose implementation is the most important infrastructure story of the decade), the week confirms in both directions. PJM Cycle 1 is the first clean post-reform read on what developers actually want to build under the new rules: dispatchable + storage + nuclear, with solar surviving at lower-but-more-real volumes. FERC’s denial is the enforcement gear that keeps the cleaned-up queue clean. The two events together read as institutional follow-through, not policy slippage.

Storage: Hydrostor enters 500 MW A-CAES into Ontario long-duration RFP; risk #4 now resting on three institutional capital events in eight days

Hydrostor announced on May 13 the Quinte Energy Storage Centre, a 500 MW / 8,000 MWh (16-hour duration) adiabatic compressed-air project in Greater Napanee, Ontario, sited adjacent to OPG’s Lennox Generating Station. The first build phase is 4 GWh, scaling to 8 GWh and 16 GWh. First-phase commercial operation targeted for the early 2030s. The project is being submitted into IESO’s long lead-time RFP, structured for 40-year contracts and targeting up to 800 MW of long-duration storage. Equity partners include the Mohawks of the Bay of Quinte; Baker Hughes is equipment partner; Canada Growth Fund is supporting via convertible loan. (See: news/2026-05-18-hydrostor-quinte-acaes-ontario-ldes.)

A-CAES is non-lithium by construction: compressed air in a purpose-built cavern, heat-of-compression captured separately, recombined on discharge. The capital intensity sits in the cavern, compressor, and heat exchanger; the duration curve sits where the cavern volume sits, which is why 16-hour scale is the natural deployment unit rather than the 2-to-4-hour band that defines lithium storage today.

Inside the published storage thesis (lithium is the substrate, alt chemistries are a tracked risk to grid share), this is the third institutional capital event on risk #4 inside eight days, after Alsym + Juniper’s 500 MWh sodium-ion procurement (May 12) and the Cerberus + Eos Frontier Power USA zinc-bromide IPP structure (May 13), both covered in last week’s digest. Hydrostor lands at the longest duration of the three (16 hours, against Frontier’s 4-to-16-hour band and Alsym’s 4-hour profile) and uses Ontario’s IESO LLT contract structure (40-year, capacity-payment-based, single technical envelope) as the procurement template. The IESO LLT round is now the cleanest existing analogue for what a US RTO LDES procurement designed for non-lithium 16-hour storage would look like.

Separately, Spearmint Energy reached financial close on Red Egret, a 300 MW / 600 MWh ERCOT standalone BESS in Texas City, with commercial operations targeted for August 31, 2026. The stack: $225 million construction debt (First Citizens + Investec coordinating, Nord/LB and East West also in), $126 million ITC transfer commitments, and $96 million Nuveen preferred equity. Equipment is 149 Sungrow PowerTitan 2.0 enclosures. Three months from close to targeted COD. (See: news/2026-05-24-spearmint-red-egret-450m-itc-transfer.)

Red Egret is now a recognizable template: IRA transferability sized at roughly half the debt, infrastructure-credit preferred equity rounding out the stack, project clears in months rather than years once interconnection is in hand. The Sungrow content sits inside the open FEOC question that any ERCOT BESS using non-trivial Chinese equipment carries through 2027 and beyond as the material-assistance thresholds tighten; the article does not disclose the project’s specific compliance pathway, and that is the line worth tracking across subsequent Texas closings.

Framework read on storage: lithium share in the 2-to-4-hour band remains the modal case and the thesis substrate. Risk #4 stays weighted at “named and partially funded,” with the underlying capital and procurement-design evidence now resting on three events rather than two. The 16-hour duration band where non-lithium is now genuinely contested is the same band where lithium economics were always weakest. Mark the change without overcorrecting.

Critical Minerals: hexavalent chromium and arsenic found in Tesla’s Robstown lithium refinery discharge

An independent water test commissioned by a Nueces County drainage district, analyzed by Eurofins Environment Testing in April 2026, found hexavalent chromium (0.0104 mg/L), arsenic (0.0025 mg/L), and elevated concentrations of strontium, vanadium, lithium, and ammonia in discharge from Tesla’s lithium refinery near Robstown, Texas. None of the substances are listed in Tesla’s TCEQ discharge permit, and TCEQ’s February 2026 investigation tested only for the standard pollutants listed in the permit. The drainage district that manages the receiving ditches was not notified when TCEQ issued the original permit. (See: news/2026-05-19-tesla-robstown-lithium-discharge.)

The Robstown facility cost approximately $1 billion and began operations in December 2024 under a TCEQ permit authorizing up to 231,000 gallons per day of treated wastewater discharge. Tesla had marketed the refinery as “acid-free clean” lithium processing, positioning its novel extraction method as an improvement over conventional acid-wash approaches.

This sits squarely against the critical-minerals thesis frame. The published frame names US-domestic refining as the supply-side gap with the fewest credible options to fill quickly, with FEOC compliance dependent on battery-grade lithium processed outside China. Robstown is the most prominent US-domiciled lithium hydroxide facility currently in operation. Operational and permitting credibility there matters in both directions: a facility that delivers clean reliable throughput at commercial scale is a genuine step toward a domestic lithium hydroxide supply chain, and a facility generating contested discharge, state regulatory gaps, and potential legal exposure from downstream drainage districts is a risk to that timeline and to Tesla’s own FEOC-compliant battery supply math.

The TCEQ’s failure to test for heavy metals or lithium in either the permit or the investigation is a separate structural issue. It suggests the state’s permitting framework for novel lithium processing chemistry has not caught up to what the process actually outputs. That gap is the kind of regulatory drag that bends rather than breaks the thesis: it does not invalidate the US-domestic refining build, but it lengthens the cycle from “facility commissioned” to “facility trusted as a reliable node in IRA-eligible supply.” Watch TCEQ’s response (whether the agency opens a revised investigation including the Eurofins-identified parameters), downstream legal standing (the drainage district was not notified), and any rerouting of Tesla’s FEOC-eligible cell supply math if Robstown throughput is constrained by regulatory action.

Solar: FERC PJM ruling sharpens the cycle; queue discipline holds

The same FERC ruling that affirmed PJM’s at-risk Readiness Deposit framework specifically resolved a solar-developer challenge. Two North Carolina solar projects (Gaston Green Acres, Bethel NC Hwy 11) tried to walk Transition Cycle 1 without forfeiting deposits when network upgrade cost allocations jumped between Phase III and retool. FERC held the line. The mechanic that gets affirmed: developers who walk at retool walk away from both readiness deposits, full stop. (See: news/2026-05-22-ferc-affirms-pjm-readiness-deposit.)

Two reads against the solar thesis frame. First, network upgrade cost surprises are real and getting bigger: Phase III-to-retool deltas have been wide enough across the cycle that two North Carolina solar projects pursued a tariff change rather than absorb the cost. The Commission held the line, which means the deltas continue to be the developer’s risk to underwrite. Second, the marginal queue project gets cleaner: by holding the deposit framework, FERC reduces the cost-shift to remaining projects when developers exit, which makes the cycle’s late-stage economics more predictable for the survivors. Constructive signal for developers with conservative cost frames, negative signal for those that priced offtake on the lower bound of Phase III estimates.

The PJM Cycle 1 readout (solar at 14.8 GW with the readiness screen filtering pre-reform pipeline) and the FERC denial fit together as the same enforcement vector: solar pipeline economics in the eastern interconnect now run through cost discipline rather than open-door speculation. The published frame holds (interconnection is the constraint), and the implementation continues to route capital toward developers who can underwrite the cost.

Critical Minerals / Storage crossover: Moment Energy notches a world-first UL safety certification for a second-life BMS

Moment Energy announced on May 19 that its battery management system for second-life energy storage received a UL functional safety certification, on top of the UL 1974 second-life cell certification it has held since 2023. Energy-Storage.News reported the BMS cert as a world first for a repurposed-cell system. The operational answer to a real question: can a system built around cells already cycled inside an EV, then pulled, sorted, and rebuilt into a stationary pack, meet the same safety bar as a first-life lithium-ion BESS. The cert says yes for Moment’s architecture. (See: news/2026-05-21-moment-energy-second-life-feoc-ul-cert.)

The reason this matters now is the prohibited-foreign-entity sourcing regime under IRS Notice 2026-15. The §45Y, §48E, and §45X tax credits now carry a material-assistance cost ratio of 40 percent in 2026, rising 5 points per year to 60 percent by 2029. Cells alone account for roughly 52 percent of project bill-of-materials in a typical utility-scale BESS, which means any Chinese-origin cell content is structurally incompatible with the credit at this threshold. Moment’s claim is that a cell already inside the United States, harvested from a retired EV pack, is no longer foreign-sourced under FEOC treatment.

This is a regulatory-hinge data point, not a primary-supply-lane one. The US retired-EV cell pool is real but small relative to 24 GW of 2026 utility-scale battery additions. What it does mark is the FEOC compliance regime forcing the storage industry to underwrite every legitimate non-Chinese cell source, including ones the market had previously deprioritized on safety grounds. Watch two follow-on signals: which IPPs or utility offtakers cite second-life as a FEOC compliance path inside the next two quarters, and whether IRS or Treasury issues clarifying guidance on the treatment of repurposed cells under the material-assistance test. The first tells us whether the market believes the workaround; the second tells us whether it survives review.

Framework check

Storage / Lithium (risk #4: alt-storage capturing grid share). Re-weighted up one notch last week from “named but unfunded” to “named and partially funded.” Hydrostor’s Quinte entry is the third institutional capital event on this vector inside eight days. Weight stays at “named and partially funded”; cadence tightens. The contest is at 6-to-16-hour duration; lithium economics in the 2-to-4-hour band remain unchanged.

Storage / Lithium (risk #1: Chinese refining concentration tightens). Unchanged at the structural level. Moment Energy’s UL cert is a tactical workaround at the cell-source margin, not a refining-side data point. The refining concentration story is still where the published thesis flags the binding non-Chinese constraint.

Storage / Lithium (risk #2: sodium-ion or solid-state cost parity). Unchanged this week. Last week’s Alsym + Juniper datapoint stands; nothing further. Watch cadence tighter, not framework reweight.

Grid (risk #1: Order 1920 watered down by lobbying). Unchanged at the rulemaking level. FERC’s at-risk readiness deposit affirmation is procedurally adjacent: the Commission held one queue-discipline mechanism this week, which is a constructive but indirect read on whether the broader Order 1920 implementation gets defended at compliance.

Grid (risk #3: load growth outpaces grid investment). Sharpened, not bent. PJM Cycle 1’s 220 GW intake is the supply-side response trying to keep pace with the demand growth, and the mix (dispatchable + storage + nuclear) is the response a developer base actually building for AI load would file. Whether intake clears at Phase I (August) is the next read.

Nuclear (thesis frame: renaissance is real, hyperscaler procurement is the institutional confirmation). Strengthened, with a new risk to log. NextEra-Dominion concentrates supply-side leverage in two operators (Constellation + NewCo) at the same time hyperscalers are concentrating demand-side procurement. That is constructive for nuclear capital allocation and price discovery; it is also a new market-structure data point that the published frame had not weighted explicitly. PJM Cycle 1’s 17.9 GW nuclear queue entry is the second institutional confirmation in the same week.

Nuclear (risk #5: hyperscaler demand collapses faster than anticipated). Marked down a notch. Two institutional events in five days (a $67B merger justified by data-center load, plus a 17.9 GW queue entry that reflects the procurement rush) make the collapse scenario harder to underwrite at current information. Not off the page; less weighted.

Critical Minerals (named-risk: western refining permitting and disclosure fails). Bent toward risk. Tesla Robstown’s contested discharge, the TCEQ permit’s missing analytes, and the unnotified downstream drainage district are the kind of cluster the published frame names as the regulatory drag that slows US-domestic refining buildout. Not a thesis break: refining still needs to come on line and the IRA money is still there. But the cycle from “facility commissioned” to “facility trusted” is longer than the prior digest treated it.

Solar (risk #1: interconnection reform stalls). Unchanged. The FERC ruling and PJM Cycle 1 readout both read as institutional follow-through, not reform slippage.

Climate Policy (IRA implementation thesis). Confirmed. Spearmint’s Red Egret close is a textbook execution of IRA transferability at scale. Moment Energy’s regulatory-hinge play is FEOC implementation working exactly as designed: the rule is forcing capital toward non-Chinese supply paths, including ones the market had previously parked. The implementation layer is where the policy thesis lives, and the implementation layer is producing.

Net read: thesis intact across all five active verticals. One risk down a notch (nuclear demand collapse), one risk up a notch on cadence (storage alt-chemistry), one risk newly logged (US refining permitting / disclosure credibility). Demand frame intact. Supply frame intact across nuclear and grid; sharpened toward risk in critical minerals. Policy frame intact and producing.

Cross-vertical thread

The clearest single thread of the week is AI-driven hyperscale load showing up inside the same five business days across three verticals through one RTO. NextEra’s bid for Dominion is unintelligible without Northern Virginia data-center load. PJM’s Cycle 1 nuclear cohort (17.9 GW in a single queue cycle) is unintelligible without the hyperscaler restart and small-reactor procurement deals announced over the prior 18 months. PJM’s gas-plus-storage cohort (105.8 GW gas + 66.5 GW storage) is the same load-following stack developers file when the firm large load is real. Spearmint’s $450 million ERCOT close is the same demand vector showing up inside a different RTO with the IRA transferability stack matured around it. Moment Energy’s UL cert is the FEOC regime forcing the supply side to find legitimate non-Chinese cells fast enough to keep up with that same demand growth.

The negative-space version of the thread is also useful. Lithium operator earnings were quiet this week, which is what we should expect between the LAR/SGML/LAC cluster (May 12-15) and SQM (around May 26-27). The absence of lithium-side news did not stop the verticals around lithium from moving. That is itself a confirmation of the published frame: lithium is the substrate, not the only signal worth tracking. The infrastructure verticals (nuclear, grid, storage capital structures) move on their own clocks, often at higher dollar amounts and with more direct policy and procurement consequences than the underlying mineral cycle.

Watch list: week of May 25-31, 2026

  1. SQM Q1 print (around May 26-27). The fifth cost-frame leg from the operator-confirmed rebalance read. Positive surprise closes the early-call window on the rebalance for the consensus-stage read. A miss on volume or contract pricing would reopen the cycle-call debate at the consensus stage, which would itself be a bigger story than confirmation. (Lithium / Storage substrate.)
  2. NextEra-Dominion FERC 203 filing and HSR window opening. First procedural milestones for the merger. Worth tracking for any early signal on state PUC posture, Virginia SCC docket initiation, and HSR scrutiny on horizontal market overlap in PJM and MISO. (Nuclear / Grid.)
  3. TCEQ response on the Robstown Eurofins findings. Whether the agency opens a revised investigation including hexavalent chromium, arsenic, and lithium, and whether Tesla’s permit gets modified to require disclosure and testing for those substances. (Critical Minerals.)
  4. PJM Cycle 1 model posting (June 26) and the run-up to Phase I (July 28). No big drop expected inside this week, but any preliminary PJM disclosure on cycle composition by state or zone, or on early withdrawals, is the first procedural read on whether the 220 GW intake holds together. (Grid.)
  5. Any second meaningful US sodium-ion or non-lithium long-duration procurement disclosure. Last week’s watch item carries forward. A second sodium-ion deal or a Form Energy / ESS Inc / Energy Vault US procurement event inside the next 90 days would be the fourth marker on risk #4 and force a more material reweight than today’s data points alone justify. (Storage.)

Clean Power Press is editorial, not advisory. Nothing here is a recommendation. Positions, prices, and projects move; we cover how to think about them.

Sourcing log

  • NextEra-Dominion merger: Utility Dive, Fortune, American Nuclear Society coverage (May 18-20 2026).
  • PJM Cycle 1 intake: PJM Inside Lines news release, April 29 2026.
  • FERC Docket EL26-39-000 (PJM at-risk Readiness Deposit affirmation): FERC May 21 2026 Commission meeting summary.
  • Hydrostor Quinte Energy Storage Centre: Energy-Storage.News, May 13 2026.
  • Spearmint Energy Red Egret financial close: Energy-Storage.News, May 22 2026.
  • Moment Energy UL safety certification: Energy-Storage.News, May 20 2026; Moment Energy company announcement, May 19 2026.
  • Tesla Robstown lithium refinery discharge: Autonocion, May 2026, citing Eurofins Environment Testing analysis (April 2026) and the Nueces County drainage district commissioned sample.
  • Prior digest framework, watch list, and rebalance read: posts/weekly-2026-05-17.md.
  • Vertical thesis frames: /projects/cpp/files/vertical-thesis-frames.
  • In-period news flashes: news/2026-05-18-hydrostor-quinte-acaes-ontario-ldes, news/2026-05-19-tesla-robstown-lithium-discharge, news/2026-05-20-nextera-dominion-67b-merger, news/2026-05-21-moment-energy-second-life-feoc-ul-cert, news/2026-05-22-ferc-affirms-pjm-readiness-deposit, news/2026-05-23-pjm-cycle-1-220gw-nuclear-storage-signal, news/2026-05-24-spearmint-red-egret-450m-itc-transfer.
weekly-digestpjmnuclearai-demanddata-centersinterconnectionlong-duration-storagefeocirathesis-confirmthesis-risknexteradominionhydrostorspearminttesla-robstownmoment-energy