With Japan largely now having normalised its previously high storage levels, LNG import levels should start to revert to making small incremental gains. As we move through summer, the large incremental drops in power sector gas demand, which have been driven by both mild weather and the impact of last year’s nuclear restarts, should begin to ease. Assuming seasonally normal weather this summer, some restocking would mean that total LNG imports in summer 2019 should rise by about 0.9 Mt, leading stocks to finish the season around 4.5 Mt, up by 0.1 Mt y/y.
Japanese LNG imports fell by 0.64 Mt (8%) y/y to 7.3 Mt in March. The decline concluded a hefty 2.3 Mt y/y fall (9%) in Q1 19 LNG imports, which, combined with weak demand, is likely to have largely eroded most of the y/y storage surplus that Japan has carried since late summer 2018. Like March, Japanese HDDs for April are forecast to be higher y/y, although still comfortably below the seasonal norm. But this month’s demand uptick, combined with efforts to tighten a very loose supply-demand balance in Q1 19, will help draw down stocks to a more comfortable 4.05 Mt by end-April (broadly flat y/y), from November 2018’s record high of 5.2 Mt, according to our balances.
Incremental available nuclear capacity will begin to ease as summer progresses and therefore so will the size of the y/y drop in power sector gas demand, especially when compared against winter 2018’s significant y/y loss. Final data is not yet available, but November 2018 to February 2019 is when the most significant historical y/y loss in gas burn in the power sector likely occurred—estimated to total 2.2 Mt y/y or 10% according to our balances—as that is the period in which cumulative nuclear capacity peaked, at about 4.1 GW. In contrast, we expect to see a much smaller 0.25 Mt y/y (1%) drop in gas demand in the sector across April-September.
There could be a small y/y uptick in gas into power in late Q3 19, when nuclear maintenance constraints are scheduled to be highest in the summer season—and if unusually hot weather supports power sector cooling demand. In September, Japanese nuclear capacity is scheduled to be 1.2 GW lower y/y, the only time in the season it will be lower, according to the current maintenance schedule. That September maintenance will be partially offset by nuclear capacity that has returned over the past 13 months, but will ultimately leave aggregate nuclear capacity slightly lower y/y that month.
However, we are more likely to see very small incremental losses in summer 2019 power sector gas demand if a return to seasonally normal weather dampens aggregate power consumption.
Across the rest of Q2 19—when available nuclear capacity is scheduled to be considerably higher y/y because of a lighter maintenance schedule—we can expect the largest y/y drops in power sector gas demand of the season.
Aggressive withdrawals in Q1 19 have now left more room for summer LNG imports. We have revised our forecast for LNG import growth in summer 2019 to be 0.88 Mt y/y (+2%), up slightly from our last forecast of 0.6 Mt y/y, leaving and end-of-summer carryout of around 4.5 Mt, up by 0.1 Mt y/y. This forecast assumes seasonally normal weather.
Where Japanese stocks ultimately peak around the start of the heating season will depend much on early winter demand. Milder-than-average weather in October-November 2018 led the country to make an unusually strong early winter injection, pushing stocks to peak at a record high of 5.2 Mt by end-November 2018, when the heating season generally begins in earnest. A stockbuild more in line with seasonally normal weather would leave peak winter stocks at about 4.9 Mt by end-November 2019, 0.3 Mt lower y/y. A heavier October-November stockbuild similar to that seen over that period in 2018 would push stocks to peak at 5.3 Mt, a touch above last year’s record high.
That would raise the risks for a very bearish summer 2020 balance for Japan. With just one addition to the nuclear fleet over winter 2018 (the 0.9 GW Ikata 3 unit, in late October), there should be very little incremental gas demand lost because of no additional nuclear power capacity y/y set to come online over winter 2019, so it will be up to winter weather and the res-com sector to soak up supply. A seasonally normal winter would mean Japan would end winter 2019-20 with stocks at a record high for the time of year (roughly 4.2 Mt), but a mild winter would leave stocks even higher, meaning very little room for imports in summer 2020.
|Fig 1: Japanese nuclear generation, TWh||Fig 2: Japanese gas into power, Mt, y/y|
|Source: METI, HJKS, Energy Aspects||Source: METI, Energy Aspects|