German battery revenues fell 8% in July to €205k/MW/yr, but that headline hides the real story. A year ago aFRR capacity supplied 91% of revenue. It's now 52%. Prices compressed as more batteries prequalified, so the optimiser rebuilt the mix around FCR and day-ahead arbitrage instead.

PJM's coincident peak steps up 17.1 GW between 2029 and 2030, taking it from 173.7 to 190.8 GW. That's 8.8x the median annual growth across the forecast, almost entirely data centres. But how much connects on schedule hinges on the unresolved FERC fight over co-located load.

ERCOT July TB1 day-ahead spreads averaged $45/MWh, up 59% from June but still 36% below July 2025. Two hot days on the 21st and 22nd drove most of the recovery. Real-time spreads fell 46% year-over-year to $40/MWh.

🎧 Transmission: Battery storage costs have fallen two-thirds since 2022, but that trend could reverse. Marek Kubik on why CATL's market share dropped from 32% to 20%, why sodium-ion could repeat LFP's takeover of grid-scale storage, and what breaks the duration ceiling.

NEM battery revenues rose 25% in July to $44k/MW/year, with Victoria taking the top spot for the first time since May 2025. Two evening price events on the 8th and 30th added $9k/MW/year between them. Still, that leaves revenues at roughly half the 12-month average.

U.S. BESS transaction volume rose 38% QoQ in Q2 to 3.8 GW and 8.5 GWh, even as deal count fell to 12. Three solar-plus-storage hybrids drove growth: Darden (1.15 GW/4.6 GWh), Steel River (1.63 GW/1.9 GWh), and CO Bar (1.2 GW/4.0 GWh).

WEM battery revenues fell 9% to $38k/MW/year in July, and the entire figure came from capacity payments. The merchant component sat at -$6k/MW/year despite a wide $126/MWh 2-hour spread. Synergy's four batteries ran at a loss on 25 of 31 days. Neoen's Collie pair stayed positive.

The NEM data centre pipeline sits above 26 GW of nameplate IT load across 74 projects, but only 1.7 GW is under construction for 2026 and 2027. Everything past 2028 is planning-stage. The near-term demand is credible. The rest is a wish list until developers commit.

Great Britain added just 209 MW of battery capacity in Q2 2026, the lowest quarterly figure since early 2024, taking the fleet to 7.6 GW. Worth noting the buildout runs lumpy though: a couple of large sites energising inside the window can swing the headline number either way.

PJM: 82 batteries cleared Phase 2 of Transition Cycle 2, totaling 8.5 GW of energy interconnection rights. Virginia leads with 3.2 GW, tracking data-center load growth. But PJM's two new 2025 capacity mechanisms have mostly expedited solar and gas, not storage.

NSW: batteries set the energy price in roughly 1 in 4 to 5 dispatch intervals by March 2026, up sharply from a year earlier. The step-change from January tracks new BESS capacity coming online in late 2025, adding competition at the margin and shifting spot price formation.

PJM: six months after the October 2025 regulation redesign, regulation makes up 77% of a 4-hour battery's revenue stack, about $56 of $73/kW-month. It cleared $97/MWh in May, down from April but 3.4x the year before. Bid it first and continuously; arbitrage is weather-contingent.

🎧 Transmission: GB Energy has a queue of 50GW+ in unsolicited co-investment enquiries, which says a lot about private appetite for state-backed risk-sharing.

ERCOT BESS additions doubled annually for years. Growth ran near 100% through 2024, eased to 76% in 2025, and is tracking closer to 40% this year. 2026 is on pace for 5.5 to 6 GW, roughly matching last year rather than beating it. With revenues at $29/kW-year, the lag has turned.

What does the average generation look like in Italy, out to 2040?

Poland procures only 400 to 500 MW of aFRR in each direction, and that's now the most lucrative stream for batteries there. With a fast-growing fleet chasing a small requirement, ancillary revenues are set to saturate quickly. The capacity leg still pays more than activation.

CAISO batteries earn most of their return from Resource Adequacy contracts, not merchant energy. That's the structural split from ERCOT. Now Slice-of-Day reform forces storage to bid capacity into the evening ramp rather than sit available all day, keeping demand for evening capacity firm.

Australia's NEM: the same 100 MW, 4-hour battery earns a different return at every connection point. Across NSW's renewable energy zones, the best-connected node still runs 0.1 to 1.7 IRR points below the state reference. In north Queensland the gap reaches 3.3 points.

Spain recorded nearly 700 hours of negative day-ahead prices in the year to 22 July, almost double Germany's 348. But depth tells a different story: Spain bottomed at -€9.8/MWh, Germany at -€499. Spanish renewables sell merchant, so they stop bidding at zero.

NEM small-scale battery fleet reached 7.8 GW by end of June, up from 1.4 GW a year earlier. These sit behind the meter and reshape residential demand rather than bidding in, but that shift is enough to add to the spread compression grid-scale BESS is already feeling.

NEM battery offtakes: physical tolls remain the backbone, accumulating to around 2.3 GW by 2029. But PPAs are the fastest-growing structure from 2027, essentially absent until 2025 then climbing to 1.7 GW through hyperscaler and industrial deals with Amazon, Rio Tinto and BHP.

Germany's aFRR capacity prices have averaged €15/MW/h so far in 2026, down 35% year on year as more prequalified batteries compete. But the average tells you little: half of negative blocks clear below €5/MW/h while single blocks have cleared above €2,000/MW/h. And it settles pay-as-bid.

If you’re passionate about analytics and battery storage in the US, this is a damn cool job job-boards.eu.greenhouse.io/modoenergy/j...

🎧 Transmission: Germany has nearly 3 GW of batteries live, but the grid connection queue has grown so large that regulators are now filtering out projects that won't actually get built. Green Flexibility's team on what building there really looks like once the slide meets the grid.

Germany: the main value of a behind-the-meter battery isn't wholesale arbitrage, it's grid fees. Large industrials pay up to 90% less in network charges for a flat load profile, and a battery flattens it. With grid fees at 40-50% of costs, that dwarfs any spread capture.

Great Britain: NESO is weighing a rule (option 3A) to stop storage being bid down in the Balancing Mechanism during constraints. It targets £180m in replacement energy costs. But Scottish batteries have saved consumers around £30m once all effects are counted. The ban would leave just £11m.

Spain: modelling single-axis trackers instead of fixed-tilt lifts the unclipped solar capture price by about €6/MWh in 2028. Trackers hold a plateau through midday and generate later into the evening, pushing more output into the sunrise and sunset hours that move prices most.

PJM: A distribution-connected battery under 20 MW can skip the interconnection queue with a WMPA and reach wholesale markets in months rather than two years. But regulation is 47 to 66% of that revenue, and PJM's regulation market is only 750 MW. That erodes as supply grows.

SPP added eight batteries totaling 820 MW in Q2, lifting the fleet to 1.4 GW across 15 projects. The near-term pipeline holds another 1.3 GW, concentrated in Q4 2026. Most of that activity runs through the Surplus process, now the fastest route to market at 11.8 GW queued.

Australia's NEM has seen 7 GW of new BESS come online since the last major volatility event in 2022. The open question is whether those extreme price spreads reflected system under-capacity, now being filled, or a durable feature of how the grid handles stress.