Shikoku Electric Power - Company History

Shikoku Electric Power did not begin with a founder. After the war, Japan’s entire generation and transmission network sat inside a single wartime entity, Nippon Hassoden, and the occupation authorities wanted supply responsibility pinned to identifiable regions. When the enabling bill failed to pass the Diet, the reorganization was imposed in 1950 by Potsdam ordinance instead, splitting the country into nine private regional utilities. On 1 May 1951 the Shikoku company was constituted in Takamatsu with capital of $1.1M (¥400m), taking over the local assets of Nippon Hassoden and the wartime distributor Shikoku Haiden, under a first president drawn from Nippon Hassoden.

The shape of the company was decided by arithmetic. A single prefecture on Shikoku was too small a market to carry the cost of a power station, so all four prefectures were placed under one operator, vertically integrated from generation through transmission, distribution and retail, with a monopoly inside its service area. Three years later, in May 1954, the shares were listed on the Tokyo Stock Exchange, giving the new utility a capital-market channel to fund what came next.

What came next was construction. Shikoku ran mostly on hydro, which meant supply anxiety every dry season while industrial output recovered and households electrified. The answer was oil-fired thermal capacity: the Anan station in Tokushima started up in July 1963 and Saijo in Ehime in November 1965, shifting the region from a hydro-dependent mix to a thermal-led one. For its first fifteen years the company’s strategy was, in effect, the state’s electricity-development plan translated into an annual construction budget.

Demand on Shikoku grew at roughly 10% a year through the petrochemical, paper and shipbuilding boom of the early 1970s, and in July 1971 the Sakaide station in Kagawa completed a three-thermal system with Anan and Saijo. Sakaide No.1 was a 195 MW unit and Japan’s first combined-cycle plant, pairing gas and steam turbines for better fuel efficiency and faster starts. Between the years ended March 1965 and March 1975, electricity sold rose about two and a half times. Alongside the plants the company began building its own supply chain — instrumentation, construction and industrial-services subsidiaries founded from 1961 through 1970 — a habit of doing things in-house that would later matter more than the plants did.

Then the fuel arithmetic broke. Thermal generation ran almost entirely on heavy oil, and when crude quadrupled in the year after October 1973 the company posted a profit decline for the year ended March 1974. Diversifying away from oil became the industry’s answer, and for Shikoku that meant nuclear power — for which a site had already been offered. In 1969, after the company abandoned a candidate site at Tsushima on geological grounds, the assembly of Ikata, a depopulated town at the tip of the Sadamisaki peninsula, voted to invite a plant in, seeking development money and property-tax revenue. Licensing followed in November 1972 and construction began in 1973.

In August 1973 thirty-five local residents sued in Matsuyama District Court to overturn the reactor licence — the first administrative suit in Japan against a nuclear siting permit, and a case that ran until the Supreme Court dismissed the final appeal in October 1992. Construction went ahead regardless: Ikata No.1 (566 MW) started in September 1977, No.2 in September 1982 and No.3 (890 MW) in December 1994, giving 2,022 MW across three reactors and lifting nuclear to about 40% of output. One siting decision, taken in a town that asked for it, would set the amplitude of the company’s earnings for the next forty years.

With three reactors running on low fuel cost as baseload and thermal following the peaks, the late 1990s were the steadiest years the company ever had. Deregulation arrived in stages rather than at once: a bidding system for wholesale supply in 1995, under which Shikoku signed its first two independent-power contracts; retail choice for extra-high-voltage customers in 2000; extension to high-voltage users in 2004–05; and tariffs loosened from approval to notification. The nine utilities kept their monopoly over households while learning, slowly, to buy and sell power in a market.

The more consequential move of these years was made long before the market opened. In July 1984 the company spun its information-systems department into a separate firm in Takamatsu — not a carrier, merely an in-house data processor. It took a Type I telecommunications licence only five years later, in June 1989, and renamed itself STNet; the vessel had been built first and the business poured in afterwards. STNet sold corporate lines, data centres and internet service to the utility’s own customer base, launched a free consumer internet service in 1996, was made a wholly owned subsidiary in 2002, and absorbed two regional cable operators in 2006 and 2007. It became the most profitable non-electricity business in the group.

The other extension went abroad. Technical-cooperation projects began in 2003, and in July 2008 the company took a 5% stake in the Ras Laffan C power and desalination project in Qatar — 2.73 GW, fully operational in April 2011 — its first overseas independent power project. The point was not scale but the competitive-bidding experience a regulated monopoly could not otherwise acquire; Oman, Chile, the United States, the UAE, Indonesia, Taiwan and Myanmar followed. Domestically the last big thermal project, Tachibana-wan No.1 (700 MW), started in June 2000, and by 2005 management was targeting a tenth of group profit from non-electricity businesses — a target the 90%-plus electricity share of consolidated sales kept out of reach.

The March 2011 disaster removed the company’s foundation. New safety rules and a slow restart review kept all three Ikata reactors offline from 2011 through 2015, and the 40% of output they had carried was replaced by thermal generation at enormous fuel cost. After a profitable year ended March 2012, the company recorded net losses for the three years that followed, and in February 2013 it applied to raise regulated household tariffs by an average of 10.94% — presented as a last resort, and paired with a cost-cutting plan averaging ¥28.1bn a year over three years, built on a roughly 16% cut in staff pay, deferred maintenance and reduced hiring.

Restart, when it came, was not stability. Ikata No.3 came back in August 2016 but was halted twice more by injunctions from the Hiroshima High Court, in December 2017 and January 2020. Units 1 and 2 were retired in 2016 and 2018 rather than take on safety investment that could not be recovered, leaving a single 890 MW reactor where 2,022 MW had stood and nuclear’s share down from about 40% to the low teens. Meanwhile full retail liberalization on 1 April 2016 ended the territorial monopoly outright, and new entrants took roughly 10% of electricity sold inside the region. In April 2020 the transmission and distribution business was legally separated into Shikoku Electric Power Transmission & Distribution, leaving the parent with generation and retail and the group in a five-segment structure.

The last shock was fuel again. LNG and coal prices after the invasion of Ukraine drove the deepest loss of the whole period in the year ended March 2023, wiping out the earnings recovery that the restarted Ikata No.3 had delivered; the fuel-cost adjustment cap was removed for liberalized contracts in July 2022 and regulated tariffs were raised by an average of 28.74% from June 2023. Recovery was abrupt: net profit of ¥60.5bn for the year ended March 2024 and a record $450.8M (¥68bn) for the year ended March 2025. The plan announced in September 2025 sorts the group into three layers — electricity and telecoms as core, international power as the expansion field, decarbonized supply and energy solutions as the frontier — and aims to double international profit to ¥8bn by fiscal 2030. Seventy-five years after being created by decree to sell electricity in four prefectures, the company is trying to earn somewhere else as well.