The Free Option to Congest the Grid: Japan's Battery-Storage Interconnection Queue and the 2026 Entry-Cost Reforms

Do entry fees clean up interconnection queues? No jurisdiction had repriced queue entry while under observation, so the question has been argued through models. The most prominent model, a structural analysis of PJM by Johnston, Liu, and Yang, first answered yes.

Its revised 2026 version finds that fees barely move completed capacity, because the projects that congest the queue most are also the most valuable. Japan has now run the missing experiment, twice in one winter, and this paper pre-registers its evaluation. Cumulative battery-storage connection-study applications in Japan reached 113 GW by March 2025, on the order of national peak demand, while grid-connected utility-scale storage stood at 0.23 GW.

From January 2026, new applications must document land-use rights. From April 2026, gridscale batteries face a doubled security deposit, forfeited on voluntary withdrawal. We assemble the first research panel from the grid coordinator's disclosures, 2018-2025, and document the run-up in application counts.

Battery study applications grew 16-fold in two years and reached 86% of study applications by late 2025. They spiked in December 2025, the last month before the first reform, a timing consistent with deadline-driven filing that the phased rules made individually rational. Congestion pools in the Tokyo demand center, not in the renewable-rich north.

The battery series has an explosive pre-trend, so we specify an interrupted time-series design with counterfactual bracketing, not difference-indifferences. Outcomes, specifications, and detection thresholds (a 42-77% collapse) are fixed here, before the first post-reform data release. We also state in advance which outcomes would remain inconclusive.