Electric Grid of Japan before the Great Depression: A Comparison with Germany

Abstract

Japan is one of the most successfully electrified countries in the world, but it does not have a unified national electric power grid: the country is divided into two zones with different standards for alternating current (50 and 60 Hz). To explore the root of Japan's split grid, this essay adopts technology historian Thomas Hughes’ socio-technical system frameworks and compares the history of national grid formation in Germany and Japan in the 1920s. In both countries, foresighted engineers and entrepreneurs noticed the necessity of a unified national grid—something we see from the case study of Matsunaga Yasuzaemon and Arthur Koepchen. However, government policy in Japan regarding frequency unification was inconsistent, whereas the German provincial governments set up clear visions for grid construction. Moreover, Japanese ideology of the day preferred market competition, whereas in Germany electric utilities coordinated with each other. For Germany, the 1920s heralded the beginning of a European grid, whereas Japan's bifurcated frequency standard hindered the formation of a unified national grid. And Japan's split grid can even result in the electric power system's fragility in the face of natural disasters such as the Great East Japan Earthquake of March 2011.

Keywords:

• Electrification
• technological style
• socio-technical system
• history of technology
• electric power industry

1 Introduction

The history of Japan's electrification encompasses both the miraculous and the puzzling. William Edward Ayrton (1847–1908), the world's first professor of electrical engineering, taught in Japan from 1873 to 1878 (Takahashi 1990; Uchida 1991). Thanks to his dedication, the educational level of Japanese electric engineering has been estimated to be among the highest in the world ever since (Okochi and Uchida 1980: 135–136). Japan's earliest electric utility, Tokyo Electric Light (forerunner of today's Tokyo Electric Power, or Tepco), was set up in 1883, about the same time as other advanced countries such as Britain, Germany, and the USA. In 1889, engineer Tanabe Sakuro 田辺朔郎 (1861–1944), still in his twenties, designed Asia's first hydroelectric power station, in Kyoto (Tanabe 1894). Before the First World War, Japan pioneered high voltage alternating current (hereafter HVAC) transmission and constructed a system using 115 kV, the highest voltage in the world except for the USA at the time (Hausman, Hertner, and Wilkins 2008: 20–21). The proportion of households using electric power in Japan had already reached 89 percent by 1935, making Japan one of the most electrified countries on earth (Kurihara 1964: 181). These achievements were miraculous for a latecomer to industrialization. There is little wonder that technology historian Thomas Hughes (1983) mentioned in the preface to Networks of Power that he would have wished to study Japan had he known the Japanese language.

But if Hughes had examined Japan from his comparative history perspective, it is quite likely that he would have been puzzled by some facts. Firstly, Japan does not have a unified frequency (or “cycle”, hereafter used interchangeably with frequency) standard for alternating current. Secondly, Japan does not have a unified national grid. These two facts are related to each other. A unified grid is based on a common standard for frequency. If one wants a unified grid without a unified cycle, then frequency converters have to be built to interconnect smaller systems running on different alternating current standards. However, frequency converters are very expensive (Yokoyama 2011).

Considering Japan's electrification achievements, the lack of cycle standards and the absence of a unified national grid would not appear to be problematic. After all, the quality of Japan's electricity, measured by the average length of blackout, is considered to be among the highest in the world (Suzuki 2020: 206). However, its bifurcated cycles and consequently non-unified national grid become controversial when compared with other countries and when natural disasters occur. As early as 1945, the United States Strategic Bombing Survey pointed out that Japan was an exception among industrialized countries in that it did not have a unified cycle standard for electric power (USSBS 1947: 16). Half a century later, researchers recognized that there was no country except Japan where the electric power grid had two cycle zones (50 and 60 Hz) (Kadoi 1991).

Moreover, the absence of a unified national grid worsened eastern Japan's electricity shortage after the Great East Japan Earthquake of March 2011. With the capacity of the frequency converters between eastern Japan (50 Hz) and western Japan (60 Hz) being relatively small, Tokyo's electric power system fell into a 10-day rolling blackout even though western Japan did have surplus capacity (Torayashiki and Maruya 2016). It is true that, before the earthquake, Japan had already constructed frequency converters to interconnect the 50 and 60 Hz regions in preparation for a future natural catastrophe (Akama 2006). Still, the earthquake suddenly revealed the fragility of Japan’s non-unified electric power system, despite Japanese technology generally being considered world leading. Understandably, Mitsubishi Electric Power Products commented that it was shameful that Tokyo had to fall into blackout even though the western Japanese grid could supply a lot (Fairley 2011).

So, why did Japan not achieve a unified national grid? This is the question, which motivated this essay. Scholars usually hold that the bifurcated cycle standards in Japan date back to the different origins of the electrical machinery used in western and eastern Japan (Fairley 2011; Kadoi 1991; Kikkawa 2004; Suzuki 2020; Yamaguchi et al. 2018). In 1895, Tokyo Electric Light (TEL) installed the 50-cycle a.c. (alternating current) generators manufactured by the German firm AEG. Two years later, Osaka Electric Light (OEL) installed the 60-cycle a.c. generators manufactured by the American firm General Electric. The influence of TEL and OEL, the two biggest utilities in eastern and western Japan respectively, between them settled the battle over electrical systems in Japan and laid the foundation for the bifurcated cycle standards. This view, then, understands the different cycle standards as a result of path dependency from the 1890s. However, this explanation is static; it does not analyze the response of Japan's socio-technical system to the split-cycle situation in the years after the 1890s. In reality, Japanese engineers, entrepreneurs, and politicians did try to build a unified national grid with a standardized cycle, especially during the 1920s, as this essay will explain.

I adopt a “socio-technical system” approach to examine the early history of national grid building. This framework (also called the “large technological system” in some scholarly works, see König 2009: 90–93) was developed by technology historian Thomas Hughes (1983: 5–14) to refer to the interplay between a technology’s physical component and its nonphysical, social components. The physical component might be anything from electric power systems, infrastructure, or mass tourism, to something such as organ transplantation (König 2009: 90). The nonphysical components usually include things such as inventors, engineers, entrepreneurs, business organizations, politicians, regulation, finance, ideology, etc., as Hughes sets out in Networks of Power. The physical and nonphysical components interact with each other, leading technology to present itself in varying “technological styles” in different regions and countries around the world.

Hughes’ framework has had a huge impact on students in the electric power industry. Bakke’s (2017) book The Grid criticized myopic utility companies, uncomprehending regulation, and simplistic advocates of renewable energy. For Bakke, the grid is “a machine, an infrastructure, a cultural artifact, a set of business practices, and an ecology” (Bakke 2017: 10). This is exactly how the socio-technical system approach looks at electricity. The same holds for Cohn's (2017) book, also entitled The Grid. For Cohn, the story of the American grid is more than just a story of technology but instead a complicated story of cooperation between private enterprise, public control, and government regulation. As we will see below, this is also true for the grids in Germany and Japan.

This essay adopts an international comparative approach, comparing Japan with Germany, given that the puzzle of the Japanese national grid cannot be fully grasped unless in a comparative context. Firstly, as mentioned above, the absence of a cycle standard and unified grid does not seem to have been problematic for Japan's spectacular achievements in electrification. We might indeed have forgotten about the situation had not a natural disaster brought it into the spotlight. Secondly, as a historian of technology with a research specialty in electric power, the author owes much intellectual debt to Thomas Hughes and feels it advisable to carry on his grand-scale international comparison method (Hughes 1983). Thirdly, Germany's history of electrification shares similarities with Japan's. Both countries started electrification in the 1880s; both had adopted high-voltage alternating current transmission by the 1910s; and both were among the most electrified countries in the world by the 1930s, being more electrified than even the USA and Britain. By 1935, 85 percent of German households had been electrified, whereas the percentage for the USA was 68 and for Britain 44 (Kurihara 1964: 181). The fact that Germany and Japan were electrified at about the same pace and at about the same time makes it easier to compare them than to compare Japan with latecomers such as China. Moreover, Germany and Japan are of a relatively small size compared with countries such as the USA and China. The German national grid looks smaller than an American regional one (for example see Cohn 2017: 3, 32, 43) if one superimposes their maps. The United States is essentially composed of three systems, one for the West, one for the East, and one for Texas, though all three run at 60 Hz and are interconnected. But Japan's land area is 55.7 percent that of Texas alone, while Germany’s, at 52.7 percent, is similar. The same problem arises if we overlay grid maps of Japan or Germany onto China. For these reasons, the better comparison is between the two smaller countries of Japan and Germany.

The First World War accelerated the development of high-voltage alternating current technology and deepened the socio-technical system's awareness of electric power interconnection. By the 1920s, the earliest national grid plans had appeared in the developed world, and the global engineering community regularly met at the World Power Conference to share information (Russ 2020). American foreign capital financed the electric power industry across the world (Hausman, Hertner, and Wilkins 2008). The 1920s proved to be a golden period for national grid construction that was only interrupted by the Great Depression, and for these reasons, this historical survey focuses on that decade and will elaborate on what happened after the 1920s in the concluding section.

The structure of this essay is organized as follows. Beginning with an overview of electric power's socio-technical system in Germany and Japan, it then discusses the four most important factors of the socio-technical system: political regulation, technical planning, entrepreneurial activity (through the cases of Matsunaga Yasuzaemon and Arthur Koepchen), and market behavior between free competition and cartel. In the conclusion, it will also discuss how the experience of the 1920s still has an impact on future interconnectivity in Northeast Asia.

2 An Overview of the Socio-technical System of Germany and Japan’s Grids

We begin with an overview of German and Japanese electrification in the 1920s, focusing on their macroeconomic status, business contexts, political backgrounds, technical standards, system design, and international finance—all parts of the nonphysical components of the “socio-technical system”.

The 1920s was a period of crisis for both Germany and Japan, but the economies of both countries did nevertheless steadily grow until the crash of 1929 and the Great Depression. As visualized in Miyazaki Hayao’s animation The Wind Rises風立ちぬ, Japan experienced the Great Kanto Earthquake of 1923, the credit crisis of 1927, and deflation, yet Japan's real per capita GDP increased by 23 percent from 1920 to 1929 (Fukao, Nakamura, and Nakabayashi 2017: 282). Germany suffered war reparations and astronomical inflation, yet real GDP increased by 18 percent from 1924 to 1929 (Spoerer and Streb 2013: 49). Against the background of macroeconomic growth, electricity generation in Japan increased by 27 percent per year from 1920 to 1929, and in Germany by 22 percent in the same decade (Kurihara 1964, Appendix; Stier 1999: 15). The highest voltage in Japan rose from 110 kV to 154 kV, and that in Germany from 100 kV to 200 kV (Teishinshō 1941: 222–223). The 200 kV systems in Germany grew by 1,512 km between 1920 and 1930, while the 154 kV systems in Japan grew by 2,819 km (Frank 1929: 964; Shibusawa 1938: 20).

The major technical characteristics of the five electric utility companies with the largest generation capacity are listed in Table 1 and Table 2. These tables follow the “technological style” paradigm of Hughes (1983: 409). Electric utilities in Germany and Japan have similarities in origin and concentration, but differences in ownership. Generally, these utilities can be divided into two categories, the first being regional systems developed from urban systems, such as Tokyo Electric Light, Tōhō Electric Light, RWE (Rheinisch-Westfälisches Elektrizitätswerk), and VEW (Vereinigte Elektrizitätswerke Westfalen), and the second category being utilities that were designed as regional grids, such as Ujikawa Electric Power, Nippon Electric Power, Elektrowerke, and ASW; most of them were born in the 1910s and 1920s with the advance of high-voltage alternating current (HVAC) technology. Naturally, to exploit economies of scale, most of the utilities concentrated power generation in one or two power plants and transmitted them under high voltage to load centers. When it came to ownership, Japanese utilities were mainly privately owned, or, to use Cohn’s (2017) terminology, investor-owned, whereas German utilities were mainly government-owned.

Table 1 Size and Concentration of the five largest Japanese Systems

Firm (foundation year)	Tokyo Electric Light (1883)	Nippon Electric Power (1925)	Daido Electric Power (1919)	Toho Electric Power (1887)	Ujikawa Electric Power (1906)
System Capacity (year)	581MW 1933	223MW 1933	313MW 1933	119MW 1933	111MW 1932
Biggest Power Stations (% of System Capacity)	Inawashiro(96 MW)(17%) Nakatsugawa(57MW)(10%) 1933	Amagasaki(140MW)(63%)Yanagikawara(50MW)(23%)1933	Osaka(96MW)(31%)Shogawa(47MW)(15%)1933	Nagoya(52MW)(43%)Kamiaso(23MW)(19%)1933	Fukuzaki(40MW)(36%)Uji(32MW)(29%)1932
Transmission Systems(km/year) above 100kV	1387(km/1933)	764(km/1933)	989(km/1933)	198(km/1933)	N/A
Ownership	private	private	private	private	private

Source: Compiled from Teishinshō denkikyoku (1934).

Table 2 Size and Concentration of the five largest German Systems

Firms(foundation year)	RWE (1898)	VEW (1923)	Preussenelektra (1923)	Elektrowerke (1915)	ASW(1923)
System Capacity (year)	1,060MW /1931	245MW /1928	149MW/1928	735MW /1931	317MW/1928
Biggest Power Stations (% of System Capacity)	Goldenbergwerk(500MW)(47%)	Gemeinschaftswerk Hattingen(100MW)(41%)	Kraftwerk Ahlem(49MW)(33%)	Kraftwerk Golpa-Zshornewitz(440MW)(60%)	Kraftwerk Bohlen(167MW)(53%)
	Pumpspeicherwerk Herdecke(140MW)(13%)1931	Gersteinwerk(98MW)(40%) 1928	Kraftwerk Borken(43MW)(29%) 1928	Kraftwerk Tattendorf(160MW)(22%) 1931	Kraftwerk Hirschfelde(105MW)(33%) 1928
Transmission Systems(km/year) above (100 kV)	4139(km/1933)	506(km/1933)	400(km/1931)	2438(km/1931)	578(km/1928)
Ownership	mixed	government	government	government	government

Source: Compiled from Unterausschuß III (1930); Das spezial-Archiv der deutschen Wirtschaft (1933).

Abbreviations: RWE = Rheinisch-Westfälisches Elektrizitätswerk, VEW = Vereignite Elektrizitätswerke Westfalen, Preussenelektra = Preussische Elektrizitäts- Aktiengesellschaft, Elektrowerke = Elektrowerke AG, ASW = AG Sächsische Werke.

Ownership has been closely related to electrification's political background. In short, Germany and Japan differed greatly in how they thought about electricity utilities: Germany preferred regulation, whereas Japan preferred free competition. In Japan, the Ministry of Communications 逓信省, which regulated public utilities ranging from railways, telegraphy, and shipping through to the postal system, centralized electric utility regulation from 1894 onward. The Ministry adopted a policy toward electricity that encouraged competition. In the legislative process leading to the Electric Utility Law of 1911, public opinion was that electricity should be a free market (Tsūshōsangyōshō 1979: 28–31, 102–105), and this was exactly how Japan's electricity market looked in the 1910s and 1920s. It was not until 1931 that the Ministry of Communications began to regulate electricity tariffs. Things were different in Germany: from the 1880s, electric utilities were under municipal government regulation, which developed into government ownership, as represented by the case of Berlin (Dame 2011: 220–239; also see Hughes 1983, chapter VII).

Both Germany and Japan set technical standards for voltage and frequency, but Japan witnessed little progress in frequency standardization. The Ministry of Communications standardized high-voltage transmission to 110 kV in 1914 and upgraded it to 154 kV during the First World War. According to engineer Shibusawa Motoji, voltage standardization was very successful, but not so for frequency unification (Shibusawa 1929). In 1922, in the Osaka-Nagoya region 87 percent of power was at 60 Hz, while 83 percent of Tokyo was at 50 Hz and 73 percent of northeast Japan was at 60 Hz (TōhōDenryokuChōsabu 1923a). The split cycle situation did not improve during the 1920s, as visualized in a report submitted to the World Power Conference's Tokyo sectional meeting in 1929 by Hitachi engineer Tanaka Bunichiro (Figure 1).

Figure 1 The frequency of Japan’s installed capacity (1927).

Source: Tanaka (1929).

The two countries also reveal great differences in system design. Germany in 1920 was covered by several separate regional grids (Figure 2). At that time, the largest electricity systems had already grown from municipal systems to regional ones. The appearance of regional grids can also be observed on a map of the Japanese system in 1924 (Figure 3), where grids centered on Tokyo, Nagoya, and Osaka can be easily identified. In neither country was the skeleton of a national grid yet in existence. But this no longer held true for the German map in 1929 (Figure 4), where the end of the 1920s had laid the basis for a national grid. Japan had managed great improvements in regional grid building, but the national system as a whole was still split. Two high-voltage systems in the shape of an inverted “V” appeared to link eastern and western Japan, but this was an illusion as the “junctions” were dams with two sets of generators: 50 Hz for Tokyo, and 60 Hz for Osaka and Nagoya (Daidōdenryoku 1942: 189–190). Besides, the grid did not extend beyond the island of Honshu (Figure 5).

Figure 2 The German grid (1920).

Source: Frank (1929).

Figure 3 The Japanese Grid (1924).

Source: Anonymous (1924).

Figure 4 The German Grid (1929).

Source: Frank (1929).

Figure 5 The Japanese Grid (1930).

Source: Shibusawa (1933).

As a final observation, in the 1920s both Germany and Japan's electric power industries relied heavily on international finance. Starting in 1923, the largest Japanese electricity firms issued huge amounts of corporate bonds in New York and London. In New York, the sum of the Japanese electricity bonds issued in 1923–1929 amounted to 39.6 percent of the collective Japanese government and corporate bonds floated thereafter the First World War. Electricity bonds stood at 77.5 percent of all Japanese corporate bonds floated in the USA (Kurihara 1964: 181). Electricity bonds were the only Japanese corporate bond floated in New York in the years 1926, 1927, and 1929 (Bytheway 2014: 126–129). Before 1931, foreign bonds made up about 30 percent of the five largest Japanese electricity firms’ capital formation (Table 1). The total amount of German government and corporate bonds floated in New York from 1924 to 1930 amounted to US$1,464 million (Kuczynski 1933). Electricity bonds made up 20.4 percent of collective German bonds floated in New York. RWE was the largest borrower among German electricity firms. The American bonds for RWE's subsidiaries Schluchseewerk and Rheinkraftwerk-Albbruck-Dogern AG significantly contributed to Germany's national grid formation, as we will see below.

To summarize this overview, the socio-technical systems of Germany and Japan in the 1920s shared similarities in macroeconomic growth, concentration, and international finance, but also presented huge differences in ownership, regulation, ideology, technical standards, and system design. Looking again at Figure 2 and Figure 3, we can see that, in the early 1920s, the grids of both countries were ready at the starting line, with similar potential for growth.

3 Planning a National Grid

Every grand technological project begins as an idea. This is also true for national grids, and we shall begin by looking at the Japanese case. In 1923, Matsunaga Yasuzaemon 松永安左エ門 (1875–1971), head of the electric utility Tōhō Electric Power, translated American engineer William S. Murray’s report, A Super Power System for the Region between Boston and Washington, into Japanese, and proposed a Japanese version of Super Power (TōhōDenryokuChōsabu 1923b). Murray had suggested a concentration of power generation into several large-scale, efficient thermal and hydropower plants, which would be interconnected by grids of between 110 kV and 220 kV (Hughes 1983: 296–297). In Matsunaga’s adaptation to Japan, thermal plants would be concentrated in Osaka, while water resources in the central mountain ranges of Honshu would be fully exploited. Large power plants in Tokyo and Osaka would be knit together by 220 kV trunk lines into a unified Japanese grid. On the supply side, this national grid interconnected the major hydropower sites in central Japan, while on the demand side it interconnected the three major load centers of Tokyo, Nagoya, and Osaka. Overall, the Matsunaga plan had as its key ideas economy of scale, energy mix, and interconnection (Figure 6). But in retrospect, the plan also had its drawbacks. Its system design focused on the central part of Honshu, with the other three main islands (Shikoku, Kyushu, Hokkaido) being completely ignored; for all of Japan to be interconnected, one had to wait until after the Second World War.

Figure 6 The Matsunaga plan (1923).

Source: Fukuda (1925).

Matsunaga argued that, if Japan was to reach a national grid, the most urgent task was to standardize frequency. He published a book entitled Concerning the Unification of Cycles 周波数統一について in 1923 to make more people understand the necessity of that point (TōhōDenryokuChōsabu 1923a). Technologically, it was very difficult for different cycle zones to be directly interconnected at that time. The existing technology for interconnection was to build a generator at the border between cycle zones, using an alternating current of one frequency to move the turbine of a generator, which could produce a current of a different frequency. This double investment in plants was uneconomical.

The Japanese engineering community was similarly aware of the rationality behind a large-scale grid. In 1921, Shibusawa Motoji 渋沢元治(1876–1975) of Tokyo Imperial University published a proposal on the construction of a national grid in DenkigakkaiZasshi 電気学会雑誌 (IEE Japan Journal), the leading professional journal of electrical technology (Shibusawa 1921). Shibusawa did not provide a concrete grid plan, but by raising the examples of America, Britain, and Germany, he stressed the importance of interconnecting Japan’s hydro- and thermal power plants and of increasing the scale of power production and the voltage of transmission. Shibusawa also listed cycle unification as the primary task for national grid building. The Matsunaga plan, in fact, also represented the view of the engineering community.

Similarly, the imbalanced distribution of energy sources and load centers was also observable in Germany, though one big difference was that, in Germany, the influence of electrical machinery manufacturers AEG and Siemens had unified the cycle standard to 50 Hz as early as the 1900s (Hughes 1983: 128–129; Nixon 1999: 35–37), meaning that Germany did not face a frequency standardization problem. Germany had rich natural resource endowments in hard coal, brown coal, and hydropower. But the hard coal mines were located mainly in the Ruhr and Saarland, the brown coal mines were located in central Germany and the Ruhr, while hydropower resources were concentrated in southern Germany. Hard coal (anthracite) has a higher thermal efficiency and is suitable for being transported to thermal power plants to cover peak load, whereas brown coal (lignite), with lower thermal efficiency, is more suitable for mine-mouth power plants covering base load. On the demand side, Japanese power consumption was concentrated in the three major cities (Tokyo, Osaka, Nagoya), while the industrial geography of Germany led to more scattered centers : Berlin, the Ruhr, Munich, Frankfurt, Mannheim, etc.

In 1916, Georg Klingenberg (1870–1925), head of AEG’s power plant division, proposed the construction of a Germany-wide grid plan under the leadership of the imperial government of the Second Reich (Gilson 1994: 106–112). This might have been the first concrete national grid plan for Germany, but the Klingenberg plan ignored the hydroelectric resources of southern Germany and was not yet a truly national grid. In 1920, Richard Tröger, a former AEG engineer who worked closely with Georg Klingenberg, published a new national grid plan in Elektrotechnische Zeitschrift, the leading academic journal of electrical engineering. In essence, Tröger proposed the construction of high-voltage transmission lines to interconnect major load centers with hard and brown coal regions and hydro-based southern Germany. The key elements of scale, economic mix, and interconnectivity were all to be observed in his plan.

Arthur Koepchen (1878–1954), a board member and chief engineer of RWE, brought out Germany’s first plan for a 200 kV grid in 1923 (Horstmann 2003: 19) (Figure 7). Alongside other lower-voltage systems planned, the projected 200 kV system would extend from the hard coal-fired and lignite mine-mouth power plants in the Ruhr, across Baden and Württemberg, towards Bavaria, knitting together western Germany’s coal-based systems and southern Germany’s water-based systems. This was an entrepreneurial plan, not only because of its 800 km length, at that time unheard of in Europe, but also because of its 200 kV voltage, a new technology whose first commercialization was in May 1923 in California. The highest voltage standard otherwise at that time was 100 kV. Although the higher voltage could theoretically reduce transmission losses, whether it could be successful in Germany or not was still a matter of controversy.

Figure 7 The Koepchen Plan (1923).

Source: Horstmann (2003).

There is a long way to go from idea to practice. In the early 1920s, engineers in both Germany and Japan had noticed the potential of nationally interconnected systems in exploiting economies of scale and economic mix. However, the realization of their ambitious plans would require the cooption of the whole socio-technical system. We shall first examine, then, the political factors behind national grid formation.

4 Grid Politics: Market or Government?

The Japanese government's policy toward the national grid was inconsistent. As early as 1908, the Ministry of Communications had recommended that the newly launched Ujikawa Electric Power adopt 50 Hz generators instead of 60 Hz, but Ujikawa had rejected the idea (Teishinshō 1941: 218). Ujikawa Electric Power had been created by a group of businessmen from the Kansai region, aiming at developing what was then Japan's largest hydropower plant on the Uji River to supply the Osaka-Kyoto load center (Hayashi 1942). Ujikawa began supplying power in 1913. In that year, 60 percent of installed capacity in Osaka was at 60 Hz and only 12 percent at 50 Hz; among the other Japanese regions (Hokkaido, Tōhōku, Tokyo, Osaka, Chugoku, Shikoku, and Kyushu) only Tokyo was dominated by 50 Hz. In retrospect, it seems that the Ministry of Communications had a basic plan to unify Osaka's frequency with Tokyo by using Japan's largest hydropower station as a benchmark. For Ujikawa, the decision to use 60 Hz was quite rational: 50 Hz had little market outside Tokyo. Nevertheless, Ujikawa's experience indicated that elites in the Japanese government had already noticed frequency standardization as early as the late 1900s.

In the 1910s, the Japanese government showed inconsistent policies towards cycle standards. In 1914, the Ministry of Communications appointed a committee to investigate means of unifying frequency and voltage. The committee suggested that, firstly, 50 Hz should be the standard frequency, and, secondly, 110 kV for transmission and 100 Vat the consumer’s end should be the standard voltages. Voltage standardization was almost completely achieved by the end of the 1920s (Shibusawa 1929: 1155–1156). During WW1, the government even adopted 154 kV instead of 110 kV as transmission voltage to keep up with technological progress and electricity demand.

Frequency unification, however, went awry, the reasons being cost and war. At that time, the biggest load center, Tokyo, was dominated by 50 Hz, but Japan as a whole still had more 60 Hz generators than 50 Hz. Although the government indeed anticipated the huge expenditure of converting 60 Hz systems to 50 Hz, wartime electricity demand moved priority from standardization to promoting power supply. In 1919, the Japanese government finally dropped frequency unification, switching to the principle that both 50 and 60 Hz were acceptable: a 50 Hz system in east Japan, and a 60 Hz system in west Japan (Teishonsho 1941: 220–221).

In the 1920s, the government still kept an eye on frequency unification but again stopped in the face of cost. After the Great Tokyo Earthquake of September 1923, Tokyo fell into a blackout. Nagoya and Osaka had available capacity but could not send electricity to Tokyo because of, among others things, the frequency difference. (Ironically, 90 years later a similar thing happened after the Great East Japan Earthquake in March 2011.) In 1924, the government again gathered a team of specialists to discuss frequency unification. They advised the government on several solutions: build frequency converters, adopt power generators that produce electricity at both 50 and 60 Hz, or unify the frequency to 50 Hz, 60 Hz, or even 55 Hz. Among these solutions, the estimated cost for unifying to 50 Hz amounted to 37 million yen, and to 60 Hz 24 million yen (Asaoka 2012: 148–149). The government could not afford the cost and henceforth promulgated no policy before WW2.

Whereas less electrified countries such as China (Tan 2021: 48–53) and Britain (Hughes 1983: 358) made huge efforts in the 1920s to unify cycle standards as a milestone toward electrification, Japan had already been highly electrified without unified frequency. In residential electrification, the six biggest cities in Japan had reached average electricity coverage of more than 80 percent by 1917. The Japanese industrial electrification rate surged from 61 to 89 percent between 1920 and 1930 (Kosakura 1973: 105). Indeed, in Japan, the socio-technical system (government, investor-owned utilities, investors, and engineers) gave priority to electrification itself, and treated electrification and frequency unification separately, with the former being reached without the latter. The government left the initiative of frequency standardization to investor-owned utilities—in other words, to the market.

To return to the international comparison, Japan's regulation was centralized in the Ministry of Communications, but for a federal country such as Germany, the discussion of politics needs to be divided into two levels: the central government and the local governments.

In 1919, the Weimar Republic's National Assembly promulgated the Socialization Law to create a national electric power system under the federal government's monopoly of high-voltage transmission (Hughes 1983: 314–315; Stier 1999: 397–400). Ironically, the Weimar government did not begin to make a national grid plan until 1926, when it authorized engineer Oskar von Miller to design a scheme for an all-Germany grid. The Miller plan, published in 1930, proposed a network of high-voltage rings of 200 kV to 380 kV that would pool and distribute electricity throughout the country (Von Miller 1930). However, the Weimar government did not take up leadership in national grid building. The grid-building initiative was left to the regional electricity firms, most of them owned by local governments. This was a major difference compared with the attitude of the Japanese government, which left the arena to investor-owned utilities and market competition.

Indeed, since the 1910s the governments of most of Weimar Germany’s constituent states had taken an active stance in creating regional utilities and building regional grids. Firms like Preussische Elektrizitäts AG (“Preussenelektra” in Table 2) and AG Sächsiche Werke (“ASW” in Table 2), owned by their respective German states, expanded within a decade to become among the largest German utilities. As most states operated or controlled regional electric power systems of their own, they opposed nationalization. Prussia, for example, criticized the nationalization plan because Prussia believed that electric utilities could, under voluntary participation and without coercion from above, build a national grid as well (Hobrecker 1935: 97). This tension between central and state governments was the essence of German national-grid politics. To sum it up, in Germany the path toward a national grid was a political issue, whereas in Japan the initiative was left to the market.

5 Matsunaga Yasuzaemon and the “Electric Power Battle”

Now we return to Matsunaga Yasuzaemon, Japan's pioneer of the national grid, to see how he tried to realize his ambitious plan. His story will be set in comparison with that of his German colleague, Arthur Koepchen. Matsunaga, a liberal-minded graduate of Keio University under the educator Fukuzawa Yukichi 福沢諭吉(1834–1901), believed that the operation and ownership of electric utilities should be in private hands, although he agreed that necessary government guidance was welcome (TōhōDenryokuChōsabu 1923b). After Matsunaga learned that the national grid was not a policy goal for the government, however, he turned to private business for its realization.

Matsunaga's life as a businessman with a critical stance toward government mirrored Fukuzawa Yukichi’s philosophy: individualism, liberalism, and independence in the face of authority. Indeed, Fukuzawa left his footprint on the history of Japan's electrification (Kikkawa 2020). Two of Japan's largest electric utilities (Table 1) were created by his students: Tōhō Electric Power by Matsunaga Yasuzaemon, and Daidō Electric Power by Fukuzawa Momosuke 福沢桃介(1868–1938), who married Fukuzawa Yukichi's daughter. In a metaphor coined by a contemporary observer, Momosuke resembled Fukuzawa Yukichi's interior, whereas Yasuzaemon his exterior. Meanwhile, Kobayashi Ichizo 小林一三(1873–1957), also a Keio University graduate and founder of Osaka’s Hankyu Electric Railway, contributed tremendously to the electrification of the Kansai area. To some extent, Japan's electric power industry mirrored Fukuzawa Yukishi’s ideological preference for free competition.

From 1923 to 1924, Matsunaga negotiated with directors of other big investor-owned electric utilities over setting up a joint company to build a trunk line from the northeast Japanese coast towards the Osaka-Kyoto region. Unfortunately, most of his colleagues were engaged in competing for market share, leaving Mastunagaon his own (Matsunaga 1970: 391–392; TōhōDenryokushiHensaniinkai 1962: 130). The Great Kanto Earthquake of September 1923 destroyed a large part of Tokyo’s public infrastructure but also created new opportunities for electric utilities. Market competition became so fierce that engineer Shibusawa Motoji noted that Tokyo's almost completely ruined electric power system was reconstructed and recovered in no more than three months (Shibusawa 1938: 21).

For Matsunaga, if the other utilities could not be convinced by persuasion, then why not convince them by market competition? Matsunaga launched business campaigns aiming at creating a unified grid by competing with other electric utilities. His concrete plan was a so-called “electric power battle” against Tokyo Electric Light (TEL), Japan's oldest and largest utility.

From 1924 to 1927, Tokyo Electric Power (TEP, not to be confused with TEL or today's Tepco 東京電力), a subsidiary of Matsunaga's firm Tōhō Electric Power, constructed a 154 kV ring network around Tokyo, and two other 154 kV systems from the mountainous regions to the city (TōhōDenryokushiHensaniinkai 1962: 187–205). Matsunaga had two objectives. Firstly, if TEL could be defeated in this competition, then the system from Nagoya to Tokyo could be knitted into one unified grid. Secondly, TEP's energy mix was an experiment. While the common practice was to cover both base and peak load by hydropower, TEP tried an equal balance of coal and hydro (Matsunaga 1970: 406). Although some engineers at that time had already suggested such a novel energy mix, Matsunaga was the first to venture to commercialize it (Kurihara 1964: 124–127). TEP's supply to Tokyo began in February 1927 with aggressive marketing strategies. It adopted a predatory price—37.5 percent lower than TEL. TEP successfully acquired 27 percent of TEL’s former customers within a few months. However, this competition meant parallel distribution lines and huge costs to be borne by TEP, while the lower price meant that the capital invested in the lines had a lower return. Investors would not prove to be happy. This led, unfortunately, to Matsunaga’s undoing. In December 1927, Matsunaga sold TEP's controlling stock to its rival, TEL, under pressure from financial capital including Mitsui Bank as well as J.P. Morgan (Hausman, Hertner, and Wilkins 2008: 174–175). A unified grid from Nagoya to Tokyo was not achieved.

The “electric power battle” was an unforgettable lesson for Matsunaga. It convinced him that a national grid could not be reached through free market competition alone. In 1928, Matsunaga published a new plan in which he recommended a regional monopoly of power supply (Kikkawa 1995: 355). It did not mention the creation of a national grid anymore but indicated the formation of unified regional grids as the priority. In other words, Matsunaga began to suggest that the regional grid was more urgent than the national grid and, further, more important than cycle unification. If the 1923 plan had been idealistic, then the 1928 plan was realistic. Japan's pioneer of a unified national grid and unified cycles abandoned these principles.

6 Arthur Koepchen and Südleitung

In Germany, Arthur Koepchen and RWE also faced many challenges in the realization of the grid plan. The Koepchen plan implied an inter-firm cooperation with utilities in southern Germany. Whether this could be achieved or not was unknown. An interconnection between the southern regions and the coal-based regions in central and western Germany was still unthinkable because of the great energy loss that would be incurred by long-distance transmission. Besides, there was the threat of nationalization. Koepchen and RWE were against the nationalization plan. When Arthur Koepchengot word of the Socialization Law, he was skeptical of it, and in 1919 he argued against nationalization before the National Assembly. In 1920, Koepchen (1920: 481–485) published a chapter in which he pointed out that the biggest problem of nationalization lay in the bureaucratic top-down structure. RWE stated in its annual report that the national grid should be reached voluntarily without coercion from the government (RWE Geschäftsbericht 1919/20). The question was how Koepchen and RWE could prove the workability of a self-regulatory, decentralized alternative.

RWE undertook the construction of the planned 200 kV system in discrete steps. After testing a short trial line in the Ruhr, RWE was still not fully assured of it, until Koepchen visited California by himself in 1924 (Boll 1969: 45). Organizationally, RWE's system expansion towards the south was opened through the acquisition of a controlling interest in the Elektrizitäts-AG vormalsLahmeyer& Co. (EAG) in 1923 (Asriel 1932: 36). A subsidiary of AEG, EAG had know-how in system construction and controlled several utilities in southern Germany. These included the Main-Kraftwerke AG, and the Grosskraftwerk Württemberg AG (Growag). The former was then negotiating with Bayernwerk, the Bavarian state-owned regional grid, over the prospect of interconnection, whereas Growag, in which the state of Württemberg held a minority stake, later paved RWE's way for cooperation with Württemberg.

Bavaria, however, regarded RWE's expansion with skepticism, as it threatened the market for their state-owned utility. In 1923, at Bavaria's initiative, representatives of Bavaria, Baden, and Württemberg met in Stuttgart and reached an agreement that consisted of three points (Pohl 1996: 180–182). Firstly, their state-owned utilities would henceforth negotiate with either EAG or RWE jointly. Secondly, Bavaria, Baden, and Württemberg would speed up the southward interconnection between their state-owned utilities and, respectively, the water-based systems in Tirol, Switzerland, and Vorarlberg, to prevent the potential monopoly of RWE from the north. Thirdly, the three states would cooperate in interconnectivity between themselves. In the ensuing years, Bavaria, Baden, and Württemberg followed this agreement throughout RWE's southward expansion.

Nevertheless, such collusion was not intended to block interconnection and cooperation with RWE. In 1924, RWE’s Main-Kraftwerke signed an energy exchange contract with Bavaria's state-owned Bayernwerk. Since RWE could thereby indirectly interconnect its system with Bayernwerk’s, it revised its strategy and changed the direction of the 200 kV system. A comparison between the original plan of 1923 and the 200 kV system established in 1928 reveals that the southern half, instead of extending towards Bavaria, advanced towards Württemberg and Baden. Hence the German name Südleitung, “southern transmission line” (see Figure 4 and Figure 7).

The cooperation with Württemberg played a significant role in the Südleitung. In 1925, RWE joined via EAG in Growag’s (i.e. the Württemberg state’s) project of building a dam in the Vorarlberg Alps (RWE Geschäftsbericht, 1925/6). Vorarlberg was rich in waterfalls but had a limited local market; Württemberg could import cheap hydropower from Vorarlberg; EAG provided know-how in system construction; RWE reached an economic mix with hydropower from the Alps. Although Bavaria warned Württemberg again over the “threat” of RWE, Württemberg granted RWE the concession over the construction of the Südleitung across its territory while, on the other hand, speeding up the construction of its interconnection line with Baden to check RWE’s potential monopoly (Pohl 1998: 71–72).

In addition to cooperation with Württemberg, in 1925 RWE set out on the strategy of interconnection with Baden (RWE Geschäftsbericht, 1925/6). The state-owned Badenwerk, similar to Württemberg's, speeded up the construction of its interconnection southwards with Switzerland, on the one hand, while engaged in cooperating with RWE on the other. Badenwerk connected itself with Switzerland in April 1926 and with RWE in December (Badische Landes-Elektrizitäts-Versorgungs AG Geschäftsbericht 1926/7). The economic mix of energy sources soon took effect. In the summer of 1927, 14 million kWh of energy was transmitted from the Baden-Switzerland power pool to RWE, while, in winter, 16 million kWh from RWE flowed southwards to Baden and Switzerland (Dehne 1928: 1206). The cooperation between RWE, Badenwerk, and the Swiss systems went further in 1928 when they together participated in setting up the Schluchseewerk AG (RWE owning 50 percent, Badenwerk 37.5 percent, and Swiss utilities 12.5 percent) (RWE Geschäftsbericht, 1928/9). The Schluchseewerk pump-storage dam could receive summer excess energy from the Badenwerk and Swiss systems and release it to the RWE power pool when necessary. It was also a cooperation that improved the economic mix of all parties and made more efficient use of natural resources.

By the end of the 1920s, the power pool of the Südleitunghad reached an energy mix from brown coal, hard coal, running water, and natural and pump-storage dams, with otherwise disparate systems interconnected by it. They were electrically one, but organizationally not integrated. Although the construction of the Südleitung caused tension, especially from Bavaria, such tension, in effect, stimulated grid building. The physical interconnection came together with complex inter-firm networks, such as those among the three southern states and their state-owned utilities, and those between RWE and the southern utilities.

7 Japan: The Grid Under Market Competition

Although Matsunaga Yasuzaemon, with foresight and ambition, had pioneered Japan's national grid construction, other Japanese electric utilities were also building high-voltage transmission systems at the same time. The difference was that Matsunaga had in mind an interconnected national system (although in retrospect it only included Honshu and was not truly national), whilst other electric utilities planned isolated generation and transmission ties. An examination of Japan's electric power market as a whole can help us understand why Matsunaga's grid plan failed in the 1920s. The discussion below focuses on the Tokyo area, Japan's largest load center.

Competition was the keyword for Tokyo's electric power market. In 1907, Tokyo Electric Light pioneered Japan's high-voltage transmission in constructing a 55 kV line from Komabashi to Tokyo (Line 1 in Figure 5). The earliest high-voltage system of more than 100 kV running to Tokyo was the Inawashiro Line (115 kV, 225 km) constructed by the Inawashiro Hydropower Company in 1913 (Kikkawa 2004, 56). At that time, the Inawashiro Line ranked as one of the world's most advanced transmission systems. Many other regional electric utilities such as Katsurakawa Electric Power and Keihin Electric Power joined in the competition for market share.

In the 1920s, Tokyo Electric Light took pains to unify the various local systems by merger and acquisition (Tokyodentō 1936: 152–158), and its system expanded rapidly. The lines it acquired included the Inawashiro Line as well as the Koshin Line (Line 8 in Figure 5), constructed by the former Keihin Electric Power. Meanwhile, Tokyo Electric Light's own system building continued. A 216 km-long 154 kV transmission line was constructed in 1924 from the Shinano River to Tokyo (Line 3 in Figure 5). After the acquisition of Inawashiro Hydropower, in 1926 a new system with a higher voltage (154 kV, at that time the highest in Japan) was constructed by Tokyo Electric Light from Inawashiro to Tokyo (Line 2 in Figure 5) to replace the old Inawashiro Line.

The problem for Tokyo Electric Light's regional grid, however, lay in the lack of interconnection. This situation made Tokyo’s hydro-based system unstable, as different river systems had different seasonal fluctuations in water flow. Competing firms separately built transmission lines toward Tokyo in isolation from each other. As already mentioned, a ring network around Tokyo was constructed by Matsunaga.

Competition again explains the Tokyo-directed grids of Nippon Electric Power and Daidō Electric Power. In 1920, Nippon Electric Power (NEP) began to build the Yanagikawara Hydropower Plant on the Kurobe River (Nippon Denryoku 1929: 166–188). As its 50-megawatt capacity appeared to be too much for consumption in Kansai alone, in 1927 NEP decided to build the NEP Tokyo Line (Line 7 in Figure 5) (Nippon Denryoku 1929: 356). By then, competition between Tokyo Electric Light and Tokyo Electric Power, as mentioned before, was approaching an end, and the NEP Tokyo Line initiated a new round of competition.

Daidō Electric Power constructed the Daidō Tokyo Line (Line 5 in Figure 5) in 1929. In retrospect, the Daidō Tokyo Line was arguably a foolish idea. Daidō Electric Power had already wholesaled electricity to Tokyo in 1924 through TEL's Koshin Line (Line 8 in Figure 5). But with the zeitgeist of market competition, the cooperation mechanism between the two had broken up by the late 1920s (Kurihara 1964: 152–156). There was an agreement between the two firms regarding demarcation, but TEL reneged on it. In revenge, Daidō Electric Power decided on the construction of the Daidō Tokyo Line. In Figure 5, it is easy to see how the Daidō Tokyo Line runs parallel with TEL’s Koshin Line (Line 8). In reality, the Daidō Tokyo Line didn't come on stream until 1934 (Daidōdenryoku 1942: 173–174); the power transmission facility lay idle for five years, and remains a good example of the double investment and inefficiency of the Japanese electric power system building during the 1920s.

Because of the vicious spiral of competition, most Japanese utilities had become unprofitable by the end of the 1920s. Several proposals to change the status quo were raised by utility managers, political parties, and financial capital. The final compromise that could be reached was a loose cartel. The Electric Power League (電力連盟DenryokuRenmei) was set up in 1932 and consisted of the five largest Japanese utilities that together controlled 60 percent of Japan's electricity (Kikkawa 1995: 181–182, 189). Meanwhile, the government revised the Electric Utility Law to reduce market competition among electric utilities.

Overall, the systems constructed in the 1920s around Tokyo were not efficient and did not lead to a unified national grid. The engineering community recognized this at the time. Shibusawa Motoji noted that, except for the Inawashiro Line (Line 1 in Figure 5) and the Koshin Line (Line 5 in Figure 5), none of the other lines were fully operational (Shibusawa 1932: 266). In other words, they did not carry the optimal amount of electricity that their capacity could allow. They did not achieve an economy of scale, though they had the potential. For investors, they did not earn money, as they should have. As public utilities, they wasted social resources. As engineer Koyama Ryuichi commented in his report to the World Power Conference of 1929:

The present state of the electrical field in Japan shows that conditions for electrical installation and methods of power supply are far from satisfactory. Especially the creation of so many power supplying districts, here and there, without any control, has resulted in serious waste of the capital invested and power produced, eventually hampering the interest of the parties involved … . [T]here are so many lines of different systems crossing each other, incidentally resulting in unnecessary extensions of the numerous lines … [A]ll the lines owned by different suppliers should be combined more efficiently into one system. (Koyama 1929: 1227–1228)

8 Germany: The Grid in Cooperation

From a comparative perspective, “electric power war” was not a term unique to Japan. German scholars also talked about the Elektrokrieg, an example of which is the dispute between RWE and Prussia from 1926 to 1928. The cause of the competition was a conflict over interest in the supply area, especially the regional load center in Frankfurt. As the RWE Südleitung involved piecemeal acquisition of controlling interest in cities, towns, and rural districts along its length, Prussia, which owned several utilities, felt its market share threatened by RWE. Although Frankfurt, based on that city's calculations, refused RWE's electricity supply, Prussia withheld from granting RWE the concession to cross the Main River until Frankfurt had signed a supply contract with Prussia. Moreover, RWE acquired controlling stock of a coal field to the east of Prussia's supply area, while Prussia similarly acquired a coal field to the west of RWE’s, coal being vital for thermal power generation. An armistice came in 1927 when RWE and Prussia reached a demarcation agreement and swapped their stock holdings in these coal fields (Hughes 1983: 424–425; Stier 1999: 299–301).

Similarly to Japan, the end of the “electricity war” in Germany accompanied the formation of a cartel, though the cartels in the two countries had different origins. The primary objective of the German cartels was to coordinate grid building, rather than to prevent resource waste as in the Japanese case.

The first German cartel, AG für Deutsche Elektrizitätswirtschaft (AdE), was set up in 1928 by Preussische Elektrizitäts AG, Elektrowerke AG, and Bayernwerk, with the objective of joint planning and construction of large-scale power plants to increase system efficiency and profitability. The concrete project was the construction of a 200 kV transmission line from Hamburg to Bavaria (see Vereinigte Industrieunternehmungen AG Geschäftsbericht 1929/30; Elektrowerke AG Geschäftsbericht, 1928; Preussische Elektrizitäts AG Geschäftsbericht, 1928). By 1931, AdE stockholders also included RWE, ASW, and VEW (AG für Deutsche Elektrizitätswirtschaft Geschäftsbericht 1931). Unlike Japan's Electric Power League, AdE was purposed as a physical component, with the Hamburg-Bavaria interconnection plan borne in mind as a part of the national grid.

The second German electricity cartel, Westdeutsche Elektrizitätswirtschafts AG(WAG), founded in 1929, might be primarily understood as a natural consequence of the Südleitung. Its stockholders included RWE, Badenwerk, Main-Kraftwerke, and other regional utilities that had participated in the Südleitung (Hughes 1983: 426). Its objective though was, broadly speaking, the same as AdE, the main difference being that WAG already had an existing power pool and the know-how associated with it.

WAG was also an organizational alternative to nationalization. In 1931, RWE officially stated in its annual report that, although Miller's national plan included a scheme for a 1,470 km, 220 kV grid for western Germany, WAG had already operated a transmission system of 1,550 km without coercion from the central government. “Our viewpoint is that the natural course of development should not be commanded by any organization from above,” as RWE stated in its company report (RWE Geschäftsbericht, 1930/31). In short, the German national grid was a consequence of cooperation from below between electric utilities, state governments, and the engineering community.

9 Conclusion

As we have seen, entrepreneurs in both Germany and Japan raised proposals for a national grid at the beginning of the 1920s, but the outcomes were quite different. Germany created the basic version of a national grid and paved the way for further development, whereas Japan laid the basis for a future national grid—but not a unified one with a standardized frequency. Both the German and Japanese systems by the end of the 1920s can be considered as the skeletons of a national grid, but their physical aspects embodied huge differences.

The reason for those differences is complicated and is to be found in the socio-technical system comprised of regulation structures, political structures, business rivalries, market structures, and ideological preferences. In Germany, state-level government-owned firms were critical of nationalization and came up with a strategy to find a decentralized alternative from below. Politically, the federal structure prevented the enforcement of nationalization from above. The cooperation between Germany’s state-level regional electricity firms led to multi-core cooperation structures as well as cartels reaching an economic mix of power sources. In politically centralized Japan, by contrast, the government encouraged the free market and viewed the unification of cycles as less important than electrification itself. The initiative of grid construction was left to the investor-owned electric utilities themselves. However, when it came to market competition, Japanese utilities found it difficult to cooperate in cycle unification and interconnection. Historians of technology believe that different countries have their historical contexts and that there is a variety of paths toward electrification; unlike Germany, Japan in the 1920s lacked a consistent policy, government guidance, and cooperation among electric utilities. Cohn (2017) was quite right: a grid is not just wires, poles, and cables, but a system of cooperation.

What happened after the 1920s? The national grid building in Germany led by RWE paved the road for an international grid in Europe after the Second World War. In 1948, RWE published a book in celebration of the firm's fiftieth anniversary. In it, Koepchen wrote an article that introduced the experience of the Südleitung and depicted the prospect of further interconnection between Germany and Austria by a new 400 kV transmission system (Scholler 1948: 15–16). Hermann Roser of REW wrote another article about Das Europa-Netz, the European network (Scholler 1948: 19–26). In the long run, RWE's grid-building efforts in the 1920s foreshadowed the unified European network as it stands today (see Lagendijk 2008).

After the 1920s, the Japanese government remained interested in cycle unification but remained inconsistent in its policy. In 1942, the government decided to unify Japan's frequency standard to 50 Hz, then in 1945 again to 60 Hz. After the Second World War, as the Japanese economy rose from the ruins, policy priority was given to capacity recovery rather than standardization, similar to the situation after the First World War. So the unification policy was again dropped. Nevertheless, in the 1950s Japan successfully unified the frequency on the island of Kyushu, though with exorbitant expenditure of about 110 billion yen at the time (Denkigakkai 2007). In the following years, eastern Japan was unified to 50 Hz, and western Japan to 60 Hz.

The Japanese electric power system continues to grow. High-voltage transmission developed rapidly, and the voltage standard was raised from 154 kV to 275 kV in the 1950s and to 500 kV in the 1970s; today 500 kV remains the standard (Suzuki 2020: 200). The systems of western Honshu, Shikoku, and Kyushu, all at 60 Hz, were connected in 1962 by underwater transmission systems. In 2000, the connection between Shikoku and Kansai (western Honshu) was further strengthened by high-voltage direct current (HVDC) cable (Figure 8). Hokkaido's system had already been connected with Honshu in 1979, also by underwater HVDC links (Takenouchi 1980). Since 1965, frequency converters interconnecting the 50 and 60 Hz regions have been constructed at Shin-Shinano, Sakuma, and Higashi-Shimizu, all located in the border regions between Japan's two frequency zones (Figure 8). These frequency converters switch alternating current from one frequency system to direct current using a technology called “back-to-back HVDC”, and then synthesize the current to the new frequency (Ito 2015: 83–111). They made it possible for the 50 Hz and the 60 Hz regions to be interconnected until the Great East Japan Earthquake revealed that they were relatively limited in capacity and fragile in the face of natural disasters (Fairley 2011). For an aged society with public finance relying on public debt, it is difficult for Japan in the foreseeable future to increase converter capacity at a great scale, not to mention achieving cycle unification.

Figure 8 Japan’s national grid, nine electric utilities, and their interconnection (2016).

Source: Torayashiki, Tetsuya and Hiroaki Maruya (2016).

To return to our comparative history, to some extent we may say that in the 1920s Japan lost what might have been a golden decade for cycle unification and unified grid building. Thereafter, the Japanese electric power system crossed a point of no return: the Japanese national grid will remain non-unified. But despite that, Japan was, as has been noted, already highly electrified by the beginning of the 1930s, and the quality of Japan's electricity system ranks among the best in the world. So the question is whether it is still even necessary for Japan to unify its cycles. After all, Japan is an exceptional case in the history of technology. Of all the countries in the world, perhaps only Japan is successfully electrified without a unified frequency standard, and it is true that, among developed countries, only Japan lacks a unified national grid. Indeed, a natural disaster can reveal the drawbacks of its bifurcated system, but great earthquakes rarely happen. After all, each country has its own “technological style,” to borrow Hughes’ terminology again. A unified national grid does not have to be the teleological end of the historical development of power grids. The answer to the question of whether it is still necessary for Japan to unify cycle standards and build a unified national grid can be in the negative.

A further point for reflection is the relation between technology and politics. Which of the two, technology or policy, has been more influential in the development of electrification in Japan and Germany? Among the complicated components of the socio-technical system, Hughes noticed the importance of politics. In Networks of Power, he examined the evolution of urban electrification in Berlin, Chicago, and London, and divided them into three types of technology--politics relations: coordination between technology and politics (Berlin), dominance of technology (Chicago), and dominance of politics (London). Based on what has been examined in this essay, we would also characterize Germany’s path toward a national grid as coordination between technology and politics; government-owned utilities such as RWE took the reins of leadership, and provincial utilities cooperated, not without tension, but successfully laid the basis for the national grid.

Compared with Germany, we would characterize Japan’s path towards a national grid as the dominance of technology. The government kept an inconsistent policy, leaving cycle unification and grid building to investor-owned utilities. Similarly to Chicago (Hughes 1983, Chapter VIII), Japan’s investor-owned utilities were able to fully develop their potential without much government intervention. This is one reason why Japan could be so rapidly electrified. But the ideological free market preference prevented Japan's investor-owned utilities from coordinating with each other in national grid building. The United States also had an ideological preference for free competition, but American entrepreneurs could mobilize Wall Street’s financial resources to create holding companies for large-scale grid formation. Japan's reliance on international capital for financing electrification prevented the birth of the holding company structure. This may be one reason why Matsunaga Yasuzaemon did not become a Samuel Insull; all Matsunaga could rely on was market competition. This is certainly a case of Japan’s “technological style”, but a detailed discussion is alas beyond the scope of this essay.

Putting Japan's bifurcated system aside, engineers and entrepreneurs today are planning a power grid at the Northeast Asia scale. The idea of Global Power originated in the 1980s in the United States and Russia. In 1991, an international conference called The World Energy System began to be held in Europe, Japan, and North America. The Melentiev Energy Systems Institute of Russia has conducted several studies on Northeast Asia and Eurasian power interconnection. With progress in HVDC technology, and with the spread of Sustainable Development Goals, new “momentum” (Hughes’ terminology once more) has been injected into the Northeast Asia Grid initiative (see for example Voropai et al. 2019; Yamaguchi et al. 2018 and;Voropai, Podkovalnikov, and Chudinova 2021) The Global Energy Interconnection Development and Cooperation Organization (GEIDCO) was launched in 2016 under the cooperation of State Grid (China), Korea Electric Power (South Korea), Rosseti (Russia), and Soft Bank (Japan). Recent articles in GEIDCO's journal Global Energy Interconnection indicate that engineers are discussing using HVDC to connect Sakhalin (50 Hz) with Hokkaido (50 Hz), Busan (60 Hz) with Kyushu (60 Hz), and Busan with western Honshu (60 Hz), thereby incorporating Japan into the Northeast Asia Grid regardless of Japan's bifurcated system (Ichimura and Omatsu 2019; Zissler and Cross 2020) This is to be reached also despite that Korea is in 60 Hz while China and Russia are in 50 Hz, as long as HVDC ties between these national systems can be built. If these transnational ties can be built, Japan's electric power system will become more robust before another natural disaster takes place, presumably the Nankai Trough Earthquake 南海トラフ地震. Whether the Northeast Asia Grid can be realized will depend on the socio-technical system at a global scale. For the moment, the Northeast Asia Grid faces big obstacles in how to finance the transnational HVDCs and how to alleviate the geopolitical risks among China, Russia, Japan, and two Koreas. A century after the early national grid efforts of the 1920s, it remains to be seen whether the Northeast Asia Grid will be realized in our lifetime.

Acknowledgments

I thank the anonymous reviewers and the editorial team for the critical comments. The essay benefited a lot from discussions with Kanji Hanaki, Terumi Ito and Takafumi Uchikawa in 2020-2022. I am also grateful to William Hausman for sharing with me some of his collections, without which this essay could not be written.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Xia Chenxiao

Chenxiao Xia is a historian of technology affiliated with Shanghai Academy of Social Sciences. Previously, he taught at Osaka University and studied at Kyoto University, Heidelberg University, and ETH Zurich.