Form Energy's iron-air system targets $20/kWh for 100-hour discharge — the most credible cost figure in LDES — but roundtrip efficiency sits around 40%, and the bifunctional air electrode durability over thousands of cycles remains a materials science challenge. EOS Energy pursues a zinc-air variant on the same electrochemical logic.

Vanadium redox flow systems (Invinity, VRB Energy) decouple power and energy, enabling arbitrary duration at 65–75% efficiency. The bottleneck is vanadium cost and supply concentration; the electrolyte can be reused indefinitely but upfront CAPEX remains 2–4× lithium-ion per kWh.

Antora Energy and Rondo Energy store cheap renewable electricity as industrial heat (graphite or refractory bricks), then dispatch it as process heat or steam. Round-trip efficiency to electricity is low, but if the end-use is heat anyway, the conversion loss is irrelevant — narrowing the addressable market to industry.

Commonwealth Fusion Systems' SPARC tokamak, now 75% complete, targets first plasma in 2027 and Q>1 shortly after using HTS magnets that achieve field strengths previously impossible at manageable magnet size. TAE Technologies pursues a field-reversed configuration (FRC) approach at a smaller scale.

Helion Energy's Polaris machine uses field-reversed configurations driven by powerful pulsed magnets, targeting net electricity output by 2028 under a Microsoft PPA. Pacific Fusion is developing a magnetized inertial fusion approach with Livermore pedigree.

Zap Energy uses shear-flow-stabilized Z-pinch — a radically simpler geometry with no superconducting magnets — trading engineering elegance for uncertain plasma stability at higher compression. Type One Energy is pursuing stellarator designs with optimized coil geometry to eliminate the disruption problem entirely.

FERC's Order 2023 (cluster-based interconnection studies) and Order 1920 (regional long-term transmission planning) mandate structural changes to how PJM, MISO, and other RTOs process requests. PJM is processing 63,000 MW of requests through 2026 under the new rules; outcomes will determine whether the reform actually clears the backlog or merely reshuffles it.

Dynamic line ratings, advanced power flow control, and reconductoring with high-temperature low-sag wire can increase existing line capacity 20–100% without new right-of-way. FERC Order 1920 requires utilities to evaluate these alternatives, but utilities have economic incentives to build new iron rather than sweat existing assets.

Hyperscalers (Microsoft, Google, Amazon) are bypassing the queue entirely by co-locating generation with load at data center campuses — signing direct PPAs with nuclear plants (Kairos/Google) or building dedicated solar-plus-storage facilities. This solves the data center problem but does nothing for general grid capacity.

QuantumScape's Eagle Line pilot (inaugurated February 2026 in San Jose) uses its proprietary Cobra process to produce QSE-5 cells with a ceramic separator and lithium metal anode. The company licenses to Volkswagen and others rather than building its own gigafactory, betting that the separator process is the moat.

Solid Power (backed by BMW and Ford) and Samsung SDI are pursuing sulfide electrolytes, which are more ionically conductive than oxides but react violently with moisture, requiring dry-room manufacturing at scales that add significant facility cost. Sila Nanotechnologies takes a different path — silicon-dominant anodes in conventional liquid cells as a stepping stone.

Group14 Technologies supplies silicon-carbon anode material (SCC55) to cell manufacturers as a drop-in improvement on graphite anodes, offering 50–60% energy density gains without the full solid-state stack. This approach is in commercial production today and does not require solving the electrolyte deposition problem.

Electric Hydrogen has developed integrated PEM stacks and power electronics targeting factory-gate costs below $300/kW by 2027, claiming a 3× cost reduction on balance-of-plant. Nel Hydrogen and ITM Power are building electrolyzer gigafactories in Norway and the UK respectively, betting that volume production drives down stack cost on the solar-panel learning curve.

Alkaline electrolyzers (Nel, ThyssenKrupp Nucera) cost $500–800/kW in 2026, use no platinum group metals, and are proven at 100+ MW scale. The challenge is dynamic operation — alkaline stacks prefer steady-state power, making them poorly matched to variable renewable supply without buffering.

Pairing dedicated hydrogen production with nuclear baseload (Kairos Power, X-energy, or existing LWR plants) eliminates the variable-supply problem and qualifies for DOE hydrogen hubs funding. The DOE's 45V clean hydrogen tax credit effectively requires near-zero carbon intensity, which nuclear satisfies, but construction timelines make nuclear hydrogen a 2030+ play.

Fervo Energy adapted horizontal drilling and fiber optic monitoring from oil and gas shale to EGS, demonstrating commercial viability at its Cape Station project in Utah — the first 100 MW commercial-scale EGS project, on track to deliver power to the grid in 2026. Fervo went public in May 2026 (Nasdaq: FRVO) and has 500 MW in total build-out at Cape Station.

Quaise Energy is developing millimeter-wave directed energy (gyrotron) drilling to vaporize rock without mechanical wear, targeting ultra-deep high-temperature wells an order of magnitude cheaper than rotary drilling. Sage Geosystems focuses on fracture network optimization to maximize per-well production from shallower horizons.

Oxford PV shipped the first commercial perovskite tandem panels to a US utility-scale project in September 2024 at 24.5% efficiency, with limited warranty coverage. First Solar is exploring perovskite alongside its CuRe cadmium-telluride roadmap but has not committed to commercial perovskite manufacturing timelines.

The dominant technical camp focuses on encapsulant layers, 2D perovskite surface passivation, and moisture-barrier coatings to extend operational lifetime. Saule Technologies (Poland) and several NREL spinouts are pursuing accelerated aging protocols to demonstrate 25-year equivalent stability in compressed lab tests.

Research groups at KAUST, MIT, and Monash are stacking multiple perovskite sub-cells to push efficiency above 40% in theory. This approach sidesteps silicon's cost but introduces new complexity in current matching between sub-cells and compounds the longevity challenge at every interface.