How much gas one unit displaces. TEPCO's Kashiwazaki-Kariwa Unit 6 has 1,356 MW of capacity, restarted on 21 January 2026 and entered commercial operation on 16 April - the first TEPCO unit back in service since the Fukushima accident. IEEFA's April 2026 briefing puts the displacement at roughly 1.1 million tonnes of LNG per year; other sources, working from annual generation and a 62 billion cubic feet basis, put it nearer 1.3 million tonnes. Framed against chokepoint risk the number is clearer: Japan moves about 4 million tonnes of LNG a year through the Strait of Hormuz, roughly 6% of its total LNG imports, and official estimates say Unit 6 alone offsets about 30% of that. For a country whose power sector leans heavily on imported gas, this converts transit risk directly into domestic baseload.

But displacement is gradual, and it has side effects. As of March 2026, 15 of Japan's 36 reactors were operating, about 37 GW in total; if every operable unit returned, Japanese LNG demand could fall by a further 12.5 million tonnes a year. The constraint is pacing: regulatory review is slow and many units remain in the queue. The grid binds too - IEEFA records the first renewable curtailment in TEPCO's service area after KK6 returned (1 March), with curtailment peaking at 3,290 MW on 29 March when KK6 ran at full output, more than seven times the level on 21 March when the unit was offline. New inflexible baseload is not a one-for-one decarbonisation gain, and grid flexibility will slow the substitution. The long-run supply account points the same way: Japanese LNG imports have fallen from about 11 Bcf/d in 2018 to about 9 Bcf/d in 2025.

Short-term prices are moving the opposite way. JOGMEC's weekly record shows the December-delivery JKM in the high-USD 31s/MMBtu on 23 October, up from the high-USD 29s a week earlier (16 October); market quotes on 30 October put JKM at USD 33.36/MMBtu, TTF at USD 32.02 and Henry Hub at USD 2.90, a JKM-Henry Hub spread of USD 30.46. Setting spot against term highlights the split: Japan's average landed LNG price in August 2026 was USD 14.16/MMBtu (JPY 117,875/tonne), with US cargoes at USD 12.89, ASEAN USD 14.91, Middle East USD 14.89 and Russia USD 14.83; August imports were 4.99 million tonnes, down 6.9% year on year; and the same month's JCC was USD 101.66/bbl. In other words, roughly a factor of 2.4 separates the average landed price from the spot marker - these are no longer the same market. Inventories are tight as well: METI reported on 21 October that power-generation LNG stocks stood at 2.54 million tonnes as of 18 October, down 0.10 million tonnes on the week and down from 2.64 million tonnes two weeks earlier.

What it means for SNSUC. First, do not plan for Japan as a growth market: nuclear restarts, efficiency gains and renewable expansion mean Japanese LNG demand is structurally subtracting. The reachable opportunities sit in seasonal peaks and in windows when the inventory buffer is thin, not in volume expansion. Second, a 2.4x gap between average landed cost and spot means portfolio management is worth far more than single-cargo trading - term negotiations cannot use spot as a reference, and spot trades cannot be underwritten on term-contract margins; the two books need separate accounting and separate risk control. Third, on the LPG side: when gas prices are high, fuel substitution in industrial and residential segments shifts toward LPG, but LPG is still priced off CP and freight, so quotes should separate benchmark price from freight and insurance rather than letting gas volatility leak into LPG contracts. Fourth, with only two to three weeks of power-generation LNG inventory, Japanese buyers pay more for certainty of delivery window than they save on price: for a supplier, specifying laycan, vessel type and berthing window in detail beats conceding on unit price. The above is a market observation based on public data and does not constitute investment advice.