Ed Porter

NEM: battery capture rates recovered to 53.5% in Q3 2026, up 4.6 points after two quarters of decline. Queensland led at 65.8%, helped by stabler prices and a longer-duration fleet. Duration is increasingly the differentiator: Limondale, an 8-hour asset, captured 91.7%.

🎧 Transmission: SSE and Statkraft have done a battery revenue swap: SSE locked in a fixed price for its battery income, Statkraft took on the variability. Both end up with less risk, and it needs far less paperwork than a wind PPA. This episode covers how the structure works and who's next.

MISO: forecast buildout to 2031 reaches 9.8 GW of BESS and 23 GW of gas. The split is regional. MISO North takes 93% of the batteries, anchored by Illinois and Michigan procurement targets, while MISO South builds gas for data center load across Entergy territory.

CAISO: Cluster 16 drew 88 GW of applications, down 75% from Cluster 15's 347 GW, and projects are smaller too, averaging 380 MW against 640 MW last cycle. Batteries still feature in 85% of capacity, but merchant revenues at $2.40/kW-month, 64% below 2023, have thinned the field.

Great Britain: from AR4 on, a CfD pays nothing when the reference price is negative. But that price blends two day-ahead auctions that disagree near zero. One Scottish wind farm switched off in 15 hours where the blend stayed positive, missing about £45k. 30 GW follows this rule by 2032.

Netherlands: TenneT's grid charges took half the revenue of a 2-hour BESS in the year to September 2026, €140k of €279k/MW/yr. The bill scales with contracted import, so a 1-hour battery earns most at just 30-40% of rated power. For longer durations, cutting imports barely helps.

Netherlands: a 2-hour BESS with a full grid connection earned €305k/MW/yr over the first nine months of 2026. The jump from 1 to 2 hours adds the most, €92k/MW/yr, because it unlocks upward aFRR capacity. Below that, there isn't enough stored energy to back the obligation.

NEM: the proponent implementation stage, from connection agreement to construction, has nearly tripled in two years, from a median of 6.5 months in June 2024 to 20.3 months in June 2026. No deadlines govern this stage, and revising a project to add a battery can reset the clock.

Great Britain: connections reform has pushed 6.5 GW of the BESS pipeline back by at least six months. The pipeline now holds 20.7 GW across 217 projects. Slower buildout under the same market conditions means higher revenues for the batteries that do connect on schedule.

ERCOT: queued gas requests reached 79 GW in Q3 2026, eight times the January 2023 level. But almost none has synchronized. Gas added just 10 MW this quarter, against 2.13 GW of solar. The queue reflects intent, not delivery, and the gap between the two is widening.

NEM: in Modo's Central forecast, an eight-hour battery going live in 2030 returns 11.1% across the mainland and 12.1% in NSW, the best of any duration. A 12-hour build earns 17% more revenue but costs 37% more, so the returns fall past eight hours.

NEM: batteries did their job. From July to September, the median midday price rose from $3/MWh in 2025 to $28/MWh in 2026 as the fleet soaked up solar. Two-hour spreads in NSW and Queensland fell to about a third of the year before. Each new build now needs a different case.

Italy: Terna issues a grid offer in about two months, but the permit is where projects wait. At MASE, median review time rose from 12 months in 2024-25 to 19 in 2026 as requests piled up. Start to finish, a standalone battery now takes about five years to reach operation.

ERCOT: Modo now tracks 92 GW of data center load in advanced planning as of October 21, requesting grid access from Dallas-Fort Worth out to West Texas and the Panhandle. These projects are heavily phased, so what matters for generators is the ramp schedule, not the headline number.

ERCOT: Modo's central scenario has TB2 spreads dipping to $2.8/kW-month in 2027, below 2026's $3.7, as battery additions outpace evening demand growth. They recover to a $10.4/kW-month peak in 2032, when the margin between firm supply and data center load is tightest.

PJM: solar made up 37% of applied capacity in the last transition cycle. In Cycle 1, that share fell to 6%. Gas now leads at 81 GW, with storage close behind on project count. The driver is capacity accreditation: gas and storage carry higher ELCC ratings, worth more with prices at the cap.

🎧 Transmission: Germany: 50Hertz went from a handful of battery connection requests a year to more than 100, and roughly 180 projects, over 80 GW, applied this year. That's near Germany's entire peak demand, at just one of four TSOs. First come, first served is gone, replaced by maturity scoring.

ERCOT: Q3 2026 added a record 2.2 GW of battery capacity, taking the fleet to 18.7 GW / 33.4 GWh. But most of that reflects decisions made two to three years ago. 1.3 GW of it signed interconnection agreements when revenues averaged $193/kW-year, against $24 today. The 2027 pipeline points lower.

NYISO: the 2026 cluster is the first to include natural gas, 2.9 GW of it, mostly three repowerings on Long Island and in Brooklyn. If they lift net capacity at the evening peak, they'd narrow the TB4 spread premium that makes Zones J and K the best battery sites in the state.

Netherlands: TenneT's grid charges took half the revenue of a 2-hour Dutch BESS in the year to September 2026, €140k of €279k/MW/yr. The bill scales with contracted import capacity, so a 1-hour battery earns most at just 30-40% of rated power. A 2-hour one is better off near full.

NEM: Q3 2026 added 830 MW / 1,410 MWh of BESS, the third-largest quarter on record despite running 58% below Q2. All four new projects were 2-hour systems, so fleet duration slipped to 2.2 hours. But that reverses next quarter: 4-hour builds make up 59% of what's under construction.

MISO: 68 GW of BESS sits in the interconnection pipeline, but only 5 GW holds a signed agreement. Developers have withdrawn 52% of everything requested since 2015. The headline interest is real, but the attrition rate tells you how little of it will actually reach the grid.

CAISO: BESS revenues fell to $2.25/kW-month in September, down 12% year on year. The striking part is that 13% of the month's revenue came from a single price spike on 9 September, when demand ran 2.6 GW above forecast during an extreme heat warning. Strip that day out and the month looks thin.

Netherlands: a 2-hour BESS with a full grid connection earned €305k/MW/yr over the first nine months of 2026. The jump from 1 to 2 hours adds the most value, €92k/MW/yr, because it unlocks upward aFRR capacity. A 1-hour battery can't store enough to guarantee the obligation.