Supporting Information

S_note1: Assessment of public sewage treatment plants in Japan

Fig. S1. Trends of population connected to sewer treatment plants (STPs) in Japan.

Fig. S2. Per-capita sewer pipeline and Pstp% among economically developed countries

in 2008.

Per capita sewer pipeline length (m/person)

1

12345

6

7

8

9

10

11

12

13

14

15

16

17

1819

Figure S1 presents assessment conditions for public sewage treatment plants in Japan. Japan

Sewage Works Association (JSWA, 1984–2012), 76% of the total population is currently

accessing public sewage systems, but the figure was only 34% in 1984. Although that

percentage has increased, Japan still has a very low rate of per-capita sewer pipeline length

for access to public sewage systems among economically developed countries (JSWA, 2008;

EUREAU, 2008), as shown as in Figure S2.

References

(1.) Japan Sewage Works Association (JSWA): Statistical Year Books for Sewer

Management, JSWA (in Japanese) 1984–2012.

(2) EUREAU (European Federation of National Associations of Water and Wastewater

Services) Statistics: Overview on Water and Wastewater in Europe 2008, Country

Profiles and European Statistics, Brussels 2009.

S_note2: Relation among some service indicators in 2010

Fig. S3. Correlation between wastewater treated amount and and pipeline length.

2

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

Figure S3 shows the relation between sewered population density (here sewer-serviced

population density means sewer-serviced population who live in inhabitable land areas) and

per-capita sewer pipeline length. The sewer-serviced population in Tokyo is very large.

Consequently, the per-capita sewer pipeline length is short. The distributed sewer network is

well accessed and used by residents. However, Aichi, Fukuoka, Nara, Shizuoka, and Nagasaki

prefectures have similar per-capita sewer pipelines but different sewer-serviced population

density. Figure S4 presents the relation between the amount of treated wastewater and

installed sewer pipeline length among prefectures. Osaka and Hyogo have similar lengths of

sewer pipes. However, they differ in their amounts of treated wastewater. A similar relation is

found among Kyoto, Shizuoka, Hiroshima, Gifu, Nigata, Miyagi, and Ibaraki prefectures.

Figure S5 shows correlation between populations connected to sewer treatment plants. There,

Fig. S4. Correlation between sewer-serviced population density and per-capita pipeline length.

Fig. S5. Correlation between population connected to sewer treatment plant and sewer

pipeline line length.

3

40

41

42

44

45

46

47

48

49

50

51

52

53

54

55

56

57

it is interesting to see similar sewer-serviced population but different installed sewer pipeline

lengths in some prefectures. Underlying reasons for that phenomenon of some prefectures are

not explained simply by those correlations. Therefore, we endorsed a decomposition analysis

when finding the underlying reasons for SMUE related with material stocks of sewer pipe and

their direct service indicators.

S_note3: Complete Decomposition Analysis

Decomposition analysis is a powerful tool enabling the identification of the underlying forces

affecting a studied phenomenon. This study examines the driving forces explaining the

differences observed among SMUEs of Japanese prefectures. For this purpose, as shown in

equation (1), the stocked material use efficiency of prefecture i, denoted as SMUEi, is

decomposed into five explanatory ratios.

SMUEi=Ri ×C i× Di × N i . × PS i=SPi

SC i×

SC i

Pi×

Pi

Ai×

A i

Li×

Li

MSiEq. 1

In that equation, the following variables are used.

SPi stands for the service provided by water treatment plants in prefecture i, i.e. the

volume (m3) of wastewater treated by the prefecture.

SCi signifies the wastewater treatment service capacity in prefecture i.

Pi denotes the population with access to water treatment services in prefecture i, i.e.

the number of persons connected to the wastewater treatment plants.

Ai represents the area serviced by sewers in prefecture i.

Li denotes the length of sewers in prefecture i.

4

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

MSi represents the material amount (tons) embodied in sewers.

The first ratio Ri is the capacity usage of installed treatment facilities. Infrastructure is

dependent on the installed capacity rather than on the actual provided service. High-capacity

use is expected to influence the material stock efficiency positively. Ci corresponds to the per-

capita installed capacity. The next ratio, Di, is the population density of the serviced area.

Densely populated areas such as prefecture-level administrative areas with large cities (e.g.,

Tokyo, Osaka, Fukuoka) are expected to have a high EMS because the sewers are insufficient

to serve the dispersed population. The following ratio (Ai/Li) is the area covered by sewers

divided by the sewer length. The ratio indicates the sewer network configuration: a high ratio

corresponds to a network with many branches whereas a low ratio refers to a linear network.

The final ratio (Li/MSi) is the inverse of the material intensity of pipelines, usually expressed

in the literature in units of kilograms per meter. A high ratio therefore corresponds to sewers

having small diameter.

Decomposition analysis is based on the consideration that SMUE in each prefecture results

from changes in the national level material stock efficiency: SMUENat. Equation (3) shows

that these changes are distinguished for each ratio with the superscript “eff” referring to the

“effect.” Each effect is subsequently calculated based on a Taylor decomposition analysis. An

example for the capacity ratio (Rieff) is shown in the box equations.

SMUENat .=RNat .×CNat . × DNat . × N Nat . × PSNat

SMUEi=SMUENat+ Rieff +C i

eff +Dieff +N i

eff+ PSieff Eq.2

Therein, the following variations and definitions are used.

5

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

Rieff =Ri

eff , I+Rieff , II+Ri

eff , III +Rieff , IV+R i

eff ,V Rieff ,I=CNat . DNat .N Nat . PSNat × Δ Ri

Rieff ,II=1

2 ( DNat . N Nat . PSNat ΔC i Δ Ri+CNat .N Nat . PSNat Δ Di Δ Ri+RNat . CNat . DNat . PSNat Δ N i Δ Ri+RNat . CNat . DNat . NNat . Δ PSi Δ Ri )

Rieff ,III=1

3 ( N Nat . PSNat Δ Di ΔCi Δ Ri+DNat . PSNat Δ N i ΔC i ΔR i+ DNat . NNat . Δ PSi ΔC i Δ Ri+CNat . N Nat . Δ PSi Δ Di Δ Ri+CNat . PSNat . Δ N i Δ Di Δ Ri+CNat . DNat . Δ PSi Δ N i Δ Ri )

Rieff ,III=1

4 (CNat . Δ PSi Δ N i Δ Di Δ Ri+DNat . Δ PS i Δ N i ΔC i Δ Ri+NNat . Δ PS i Δ Di ΔC i Δ Ri+PSNat . Δ N i Δ Di ΔC i Δ R i )

Rieff ,V =1

5Δ PSi Δ N i Δ Di ΔCi Δ Ri

S_note4: Material stocks in each prefecture in 1984 and 2012

6

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

Figure S6 presents material stocks in respective prefectures in 1984 and 2012. It was readily

observed that material stocks increased in all prefecture-level administrative areas. For

example, within 28 years, there material stocks in some prefectures, such as Osaka, more than

doubled.

S_note5: 47 prefecture-level administrative area locations on the Japan map

Japan Map

Fig.S6. Material stocks of sewers in respective prefectures in 1984 and 2012.

7

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

Ref.: http://ww2m.biglobe.ne.jp/ZenTech/English/Japan/Map/Japan_Prefectures_Map.htm

S_note6: Tokaido and Sanyo Shinkansen lines on the Japan map

Ref.:

https://en.wikipedia.org/wiki/T%C5%8Dkaid%C5%8D_Shinkansen

Fig. S9. Sanyo Shinkansen line on the Japan map.

Ref.: https://en.wikipedia.org/wiki/Sany%C5%8D_Shinkansen

Fig. S7 47 prefecture-level administrative area locations on the Japan map.

Fig. S8. Tokaido Shinkansen line on the Japan map.

8

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

S_note 7: Diameter share of respective prefecture stocks

Figure S10 presents the diameter shares of respective prefecture stocks ranked from the least

to most efficient prefecture. As a first explanatory factor of the high diversity of material

stock efficiency we encountered, as shown in Figure S10, indeed, the material intensities of

larger pipeline with diameter of larger than 2000 mm category are, respectively, 151 and 19

times more important than diameter of less than 600 mm and 600–2000 mm pipeline

categories. Although it appears that Hiroshima has the largest share of large pipelines (72%),

and although Akita’s share of pipelines larger than 600 mm reaches as high as 88%, highly

material efficient prefecture-level administrative areas such as Tokyo and Kyoto have shares

of large pipelines also reaching as high as 61% and 50%, respectively. Pipeline size alone is

insufficient to explain the high diversity of observed material stock efficiency.

9

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

S_note 8: Some testing for observing simple correlations among some parameters in

2010

Fig. S10. Diameter share in prefectures (2010).

10

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

From those figures above, we were able to observe some correlations. However, these were

insufficient to explain the diversity of encountered prefecture behaviors. We therefore

conducted correlation analysis using decomposition analysis.

Fig. S10. Some testing for observing simple correlations among some parameters.

11

186

187

188

189

190

191

192

193

194

195

196