Sei sulla pagina 1di 11

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 1 0 6 9 7 e1 0 7 0 7

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier.com/locate/he

Identification of public acceptance factors with risk


perception scales on hydrogen fueling stations in
Japan

Kyoko Ono*, Kiyotaka Tsunemi


Research Institute of Science for Safety and Sustainability, National Institute of Advanced Industrial Science and
Technology, 16-1 Onogawa, Tsukuba, Ibaraki 305-8569, Japan

article info abstract

Article history: This is the first work to describe the characteristics of public acceptance of hydrogen
Received 8 December 2016 stations (H2 station) in Japan using risk perception scales. We conducted an online survey
Received in revised form asking respondents to rate their acceptance of having an H2 station constructed in the gas
21 February 2017 station nearest their home. Sixty-six percent of respondents indicated a high rate of
Accepted 5 March 2017 acceptance, with males tending to be more accepting than females, irrespective of age. We
Available online 28 March 2017 found the following to be explanatory factors for acceptance: gender, degree, vehicle use,
knowledge about hydrogen, risk perception of H2 station, and inherent risk acceptance and
Keywords: avoidance. Binominal regression analysis was used to construct an acceptance model, and
Dread factor the risk perception factor “Dread” was dominant among the effective independent vari-
Hydrogen fueling stations ables. This suggests that alleviating inherent dread or fear by providing precise risk in-
Public acceptance formation will lead to better acceptance. Our study contributes to improved risk
Risk perception communication on H2 station construction.
Unknown factor © 2017 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

facilities, perhaps even having a vaguely negative view


Introduction regarding accidents, such as hydrogen explosions caused by
hydrogen gas leaking from those facilities. Thus, public
Hydrogen is considered a promising fuel, because hydrogen- acceptance and consensus must be a key point of the suc-
driven vehicles, including fuel cell vehicles (FCVs), can cessful introduction of H2 stations to the public [1]. Deter-
potentially reduce the negative effects of fossil fuel in trans- mining the factors that make people supportive of or resistant
port. FCVs do not exhaust air pollutants in the process of to H2 stations could enable policy makers to develop policies
driving. If the hydrogen is produced as a by-product of the that receive public support. Furthermore, the results will help
petrol refinery process, emission of carbon dioxide will be improve the quality of the information provided to the public
reduced in the total lifecycle of driving FCVs. facing emerging technology and related policy decisions.
Introducing the hydrogen fueling infrastructure, such as A few public perception surveys on hydrogen technology
hydrogen fueling stations (H2 station) to the public is indis- have been conducted in nationwide hydrogen technology
pensable in a society in which FCVs are to be commonly used. promotion projects, such as the AcceptH2 initiative in the U.K.
However, people may be apprehensive about such new [2] and the DOE hydrogen program spearheaded by the U.S.

* Corresponding author.
E-mail address: kyoko.ono@aist.go.jp (K. Ono).
http://dx.doi.org/10.1016/j.ijhydene.2017.03.021
0360-3199/© 2017 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
10698 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 1 0 6 9 7 e1 0 7 0 7

Department of Energy [3]. Ricci et al. reviewed the public prescribed “Personal information protection management
perception of hydrogen [4], and the results of a public systemsdRequirements (JIS Q15001).” MACROMILL removed
perception survey have been reported, including FCV and H2 all individual information from the dataset and that we used
stations [5e19]. Schmidt et al. summarized a variety of studies in the analysis.
that identified a lack of awareness and knowledge regarding About 2000 or more respondents were selected from all
hydrogen technology; however, they found that general panelists using a stratified random sampling method. First, we
hydrogen acceptance was nevertheless high [20]. Some of the chose potential respondents proportionally from the number
reviewed studies were simple social surveys gauging accep- of people within each of eight administrative regions (1.
tance of hydrogen technology [5,9,12,15,16]. They analyzed Hokkaido, 2. Tohoku, 3. Kanto, 4. Chubu, 5. Kinki, 6. Chugoku,
relationships between the tendency toward acceptance and 7. Shikoku, 8. Kyushu), with consideration given to sex and age
knowledge level and perception of hydrogen and/or H2 sta- (20e29, 30e39, 40e49, 50e59, 60þ) based on the Population
tions. Some studies quantitatively examined the de- Census 2013. The respondents selected by the stratified
terminants of being in favor of or against a hydrogen refueling random sampling were considered representative of the Jap-
facility in the European region [6,10,11,13,14,17e19]. These anese population.
European studies investigated the factors behind citizens' We decided beforehand that the ratio of males should be in
intention to act a given way towards a hydrogen refueling the range of 0.3e0.7, because adjustment of the sex ratio was
facility. Policy makers can benefit from early knowledge of impossible considering the home region and age of potential
how people will respond, and thus prior to actual observable respondents. There was no upper limit for respondent age.
behavior, it is useful to study intention to act. However, it was difficult to select respondents' age
Huijts et al. defined acceptance as behavior towards energy 70þ according to the real proportion of this population,
technologies, and acceptability as an attitude. They con- because panelists of that age were scarce. Therefore, the
tended that acceptance reflects behavior that enables or pro- proportion of respondents over 60 years of age was repre-
motes the use of a technology, rather than inhibiting or sented by the proportion of the population 60e69, but the re-
demoting its use [21]. They constructed a risk-acceptance spondents over 60 years of age included both ages 60e69 and
model and described the structure of such acceptance. Their 70þ. A total of 2069 people participated in the survey.
model included cognitive determinants of acceptance (e.g.,
perceived benefits and risks), along with affective factors such Questions about respondents' characteristics
as positive and negative feelings of fear, satisfaction or trust Each panelist was asked in advance to participate in this
[6]. This model has been deemed effective because it explains questionnaire by e-mail; we informed respondents that the
the characteristics of risk acceptance simply [6,14,17]. questionnaire was strictly voluntarily and that they were free
In Japan, surveys inquiring whether or not respondents to decline. To reduce susceptibility to fatigue, we sought to
were agreeable toward the introduction of H2 stations, as well keep the amount of time needed to complete the survey to
as their understanding and perceptions about hydrogen and under 15 min. In consideration of the ethical and human
FCVs, were conducted in 2008, 2009 and 2015 [7]. Thus, the rights of the individuals, we did not ask for personally iden-
surveys could report the change in percentage of supporters tifiable information or medical histories.
over time. The results showed that people were gaining The survey was conducted in January 2014. During the
knowledge about H2 stations, but there was not an observed survey period (1.5 days), we received 2069 valid responses
change in their perception of risks or benefits [7]. They did not from approximately 10,000 email invitations sent to potential
discuss the risk perception characteristics associated with H2 respondents as a screening, for a response rate of 20%. For
stations. respondents younger than 60 years, while the gender ratios
The aims of this study were to characterize risk perception and age distribution reflected the actual demographics of re-
of H2 stations and to analyze acceptance of H2 stations by spondents' residential areas, the gender ratios of each specific
classifying people according to attributes such as knowledge age varied because of disparity in the ages of monitor mem-
of hydrogen and risk perception characteristics. This paper bers. The number of respondents 60 years or older was less
contains: (1) a summary of the ratio of people in the Japanese than the actual corresponding age structure due to constraints
public who feel supportive of H2 station construction and arising from the absolute number of monitor members. The
those who oppose it; and (2) a characterization of the risk number of valid responses from people aged 70þ was 49.
perception of those groups using psychological factors in Table 1 summarizes the demographic attributes and living
addition to demographic attributes. situations of respondents. Respondents were asked questions
regarding their age, educational attainment, household
annual income, number of children (under 18 years old)
Methods residing with them, housing type (detached or multi-family),
vehicle ownership, frequency of automobile usage per week,
Outline of survey distance from their home to the nearest gas station (GS), and
frequency of public transportation usage per week.
Selection of respondents of online survey
Survey respondents were panelists (male and female, 18 years Question on acceptance of H2 stations
or older, who live in various parts of Japan) of MACROMILL
Inc., a Japanese marketing research company. MACROMILL First, respondents were provided with written qualitative in-
had 1,150,000 panelists as of January 2013, covered by formation regarding “What are hydrogen stands?” and “How
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 1 0 6 9 7 e1 0 7 0 7 10699

from possible responses: “I have never heard of that”, “I have


Table 1 e Characteristics of respondents (N ¼ 2069).
heard the name before”, or “I know it well and can explain it to
Female (%) 51.9 others”; the answers were scored as 1, 2, and 3 points,
Mean of age (S.D.) 43.1 (12.8)
respectively. The mean value for all respondents' scores is
Age distribution (%)
presented under “Energy-knowledge” in Table 2.
18e29 15.5
30e39 26.1 Respondents were then asked nine questions about the
40e49 28.0 characteristics of hydrogen and hydrogen energy from the
50e59 18.3 information provided to them. The first was (1) Have you
60þ 12.1 heard about hydrogen energy? [The alternative “Yes, and I can
Regional distribution (%) explain simply how to use hydrogen energy.” was regarded as
Hokkaido 4.5
a correct response.]; (2) Is hydrogen harmful to humans?
Tohoku 7.4
Kanto 33.2
[Correct answer: No]; (3) Is hydrogen lighter or heavier than
Chubu 18.2 air? [Lighter]; (4) Does hydrogen smell? [No]; (5) Is hydrogen
Kinki 16.1 flammable in air? [Yes]; (6) Does hydrogen need other energy
Chugoku 5.9 to be used as energy? [Yes]; (7) Can hydrogen be stored as an
Shikoku 3.2 immersed form by others? [Yes]; (8) Can hydrogen be stored as
Kyushu 11.4
liquid? [Yes]; and (9) Can hydrogen be stored as gas? [Yes]. The
Household income (millions yen)
mean value for respondents' total scores is listed under “H2-
0e2 6.8
2e4 20.5 knowledge” in Table 2.
4e6 22.0 Respondents were then asked (1) how strongly the
6e10 24.6 following items evoked hydrogen for them, (“a science pro-
>10 8.8 gram at a school”, “a fuel”, “an airship”, “a fuel cell” (five-point
Unknown 17.2 scale; H2-image); (2) whether they thought of hydrogen as
Higher degreea (%) 55.7
clean (five-point scale; H2-clean); (3) whether they think
Vehicle ownership (%) 80.5
hydrogen is dangerous for use in daily life (five-point scales;
Drive/use car (%)
once a week or more 65.9 H2-dangerous); and (4) whether they think hydrogen energy
less 34.1 technology effectively facilitates the following three out-
Public transportation use (%) comes: saving fossil fuel, mitigating global warming, and
once a week or more 33.9 resolving air pollution problems (five-point scales; H2-
less 66.1 promising technology). The total scores are listed in Table 2.
Living in detached house (%) 54.5
We asked participants about their awareness of environ-
Presence of child under 18 (%) 61.4
Distance to nearest gas station (%)
mental problems such as global warming, energy resource
<50 m 6.2 depletion, air pollution, and whether they thought mass
50e300 m 33.4 media was credible. These factors seemed important for
300 me1 km 37.3 explaining acceptance of a new technology, as suggested in
1e5 km 14.7 Huijts et al. [6,18]. Inquiries about awareness and credibility of
>5 km 1.2
mass media were based on work by the Mizuho Information &
Unknown 7.2
Research Institute, Inc. [22,23], and slightly modified for the
a
Respondents who graduated college, university or higher. following factor analysis. We asked respondents to indicate
the consistency of their thoughts on four-point scales (4:
strongly agree; 3: agree; 2: disagree; 1: strongly disagree). The
do they supply hydrogen?” (Fig. S1, in Supporting Informa- question items are listed in Table 3.
tion). Respondents were then asked to rate on a four-point
scale whether they would accept installation of an H2 sta-
tion under the following two scenarios: (1) A H2 station will be
installed in the existing GS nearest to your home (“Installation Table 2 e Summary of scores about knowledge and
in the GS scenario”); and (2) A station where both gasoline and images of energy, hydrogen and hydrogen technology.
hydrogen can be provided will be newly constructed near your Mean (SD)
home (“New construction scenario”). Respondents provided
Energy-knowledge (max; 21) 8.01 (3.42)
answers for both scenarios.
H2-knowledge (max; 9)a 5.07 (2.39)
Scientific degreeb (N ¼ 456) 6.15 (2.14)
Questions of knowledge and awareness of energy and Other degree and non-degree (N ¼ 1613) 4.81 (2.37)
environmental issues H2-image (max; 20) 13.76 (2.82)
H2-clean (max; 5) 3.46 (1.00)
As indicated by a previous study, knowledge about hydrogen H2-dangerous (max; 5) 3.20 (0.92)
H2-promising technology (max; 15) 11.12 (1.87)
and hydrogen technology is an important factor that affects
a
people's acceptance of H2 stations [4,13,14]. We asked re- 9 questions on characteristics of hydrogen and hydrogen energy
spondents about their knowledge of seven power generation were provided to the respondents. See the text.
b
Respondents who graduated from a scientific course at college or
systems: wind, nuclear, hydrogen, biomass, snow ice heat,
university (22% of total respondents).
geothermal, and wave power generation. Respondents chose
10700 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 1 0 6 9 7 e1 0 7 0 7

Question on risk perception, scale of risk acceptance, and acceptance. Items were assessed using four-point scales to be
risk-avoiding tendencies consistent with previous studies [27e29] (4: strongly agree; 3:
agree; 2: disagree; 1: strongly disagree). Since neutral points
We provided respondents with a series of questions to char- were not provided, participants were required to choose
acterize risk perception on H2 stations. In risk perception whether they agreed or disagreed. This has the advantage of
studies, a common approach is the use of psychophysical eliminating ambiguous answers. Questions about the scale of
scaling and multivariate procedures, such as factor analysis risk acceptance and risk-avoidance tendencies were based on
[24]. In previous studies [24e26], characteristics of perceived previous questionnaires [28e31].
risk could be described by two underlying dimensions, i.e.,
“Dread” and the “Unknown”. According to Slovic, “Dread” was Statistical analysis
defined by perceived lack of control, dread, catastrophic po-
tential, fatal consequences, etc., while “Unknown” was Factor analysis was conducted to extract factors based on the
defined by hazards judged to be unobservable, unknown, new, datasets for the awareness of environmental problems, risk
and delayed in their manifestation of harm [24]. We followed perception of H2 stations, the risk acceptance scale, and the
these definitions in this study. The question items are orientation of risk avoidance, because we considered that
included in Table 4; they were modified from the previous psychological factors would reflect the results from ques-
study [27] to focus on the characteristics of H2 stations. We tionnaires. The maximum likelihood procedure was used,
asked respondents' thoughts, providing these questions in followed by the promax rotation. Factors with an eigenvalue
four-point scales (4: strongly agree; 3: agree; 2: disagree; 1: of more than one were extracted. Factor scores were esti-
strongly disagree). mated using the regression method. After obtaining eigen-
Similar to risk perception, respondents were asked to values, items of lower communalities (h2 < 0.3) were deleted,
indicate the extent to which they agreed or disagreed with and factor analysis was reiterated.
statements about the scale of risk acceptance (Table 5), and Stepwise binominal regression analysis was conducted to
risk-avoiding tendencies (Table 6). The scale of risk accep- construct a prediction model with significantly correlated
tance reflects the intrinsic psychological degree of risk factors. At first, all parametric factors (shown in Table S1)

Table 3 e Arithmetic mean, standard deviation and factor loading matrix for awareness of environmental problems (i.e.,
global warming, energy resource depletion, air pollution) and credibility of mass media, and their interpretation. KMO,
0.82; P < 0.001 (Bartlett's test).
AMa SDa Factor 1b Factor 2b Factor 3b
You worry whether plenty of energy will be 3.14 0.68 0.840 0.155 0.044
supplied at low cost in the future.
Use of renewable energy should be accelerated 3.19 0.67 0.788 0.068 0.074
because fossil fuel is a limited natural resource.
Environmental problems are an important issue 3.11 0.66 0.742 0.015 0.014
that should be solved using taxpayer's money.
You worry about depletion of petroleum in the future. 3.04 0.74 0.635 0.073 0.062
Safety is the highest priority when we use 3.29 0.66 0.520 0.025 0.106
new technology.
Carbon dioxide emitted in daily life, which 3.04 0.68 0.441 0.345 0.136
results in global warming, should be
reduced by individual effort.
You are suspicious that global warming 2.08 0.86 0.201 ¡0.877 0.054
has actually happened.
You never care about air pollution because 3.07 0.73 0.035 0.507 0.233
you think that the air pollution
problem is exaggerated.
Countermeasures for global warming should 2.72 0.74 0.221 0.411 0.209
be given the highest priority among
environmental problems.
You trust government statements. 3.10 0.67 0.079 0.154 0.733
You trust newspaper articles. 2.83 0.71 0.009 0.118 0.642

Factor correlations Factor 1 e 0.617 0.053


Factor 2 e e 0.109
Eigenvalue 2.76 1.30 1.08
Cumulative contribution ratio (%) 25.0 36.8 46.7
Interpretation Positive for Negative for Trust for
problem solution problem solution mass media
a
AM: arithmetic mean. SD: standard deviation. 4-point Likert scale.
b
Factor loadings after promax rotation. Loadings >0.5 are shown in bold.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 1 0 6 9 7 e1 0 7 0 7 10701

Table 4 e Arithmetic mean, standard deviation and factor pattern loading for risk perception of H2 stations and their
interpretation. KMO, 0.92; P < 0.001 (Bartlett's test).
AMa SDa Factor 1b Factor 2b Factor 3b
I intuitively dread a hydrogen stand. 2.37 0.68 0.857 0.019 0.062
Accidents related to hydrogen stands will increase when 2.54 0.64 0.804 0.026 0.14
they become popular in society.
Irreparable damage will be caused by a hydrogen stand. 2.58 0.67 0.796 0.117 0.192
It is difficult to prevent accidents with hydrogen stands. 2.39 0.64 0.679 0.045 0.112
A hydrogen stand itself causes fear. 2.33 0.69 0.676 0.057 0.194
Reliability of the system is poor with a hydrogen stand. 2.38 0.64 0.529 0.031 0.245
An accident in a hydrogen stand has serious consequences 2.54 0.64 0.022 0.923 0.021
for its surroundings.
An accident in a hydrogen stand affects a wide area. 2.72 0.71 0.011 0.894 0.006
An accident in a hydrogen stand results in a large number of victims. 2.76 0.71 0.18 0.683 0.007
Even a small leak of hydrogen from a hydrogen stand should be averted. 2.63 0.77 0.184 0.335 0.242
The potential for accidents caused by a hydrogen stand is uncertain. 2.83 0.74 0.033 0.007 0.877
Damage caused by an accident related to the installation 2.86 0.76 0.132 0.037 0.806
of a hydrogen stand is uncertain.

Factor correlations Factor 1 e 0.552 0.350


Factor 2 e e 0.379
Eigenvalue 3.35 2.34 1.61
Cumulative contribution ratio (%) 27.9 47.4 60.8
Interpretation Dread Catastrophic Unknown
a
AM: arithmetic mean. SD: standard deviation. 4-point Likert scale.
b
Factor loadings after promax rotation. Loadings >0.5 are shown in bold.

Table 5 e Arithmetic mean, standard deviation and factor loading matrix for risk acceptance scale and their interpretation.
KMO, 0.82; P < 0.001 (Bartlett's test).
AMa SDa Factor 1b Factor 2b Factor 3b Factor 4b
You are careful in responding to anything. 2.85 0.64 0.746 0.084 0.114 0.092
You always want to play it safe. 3.07 0.65 0.687 0.067 0.066 0.034
You are careful not to fail at anything you do. 2.77 0.65 0.625 0.118 0.102 0.134
You never approach dangerous places. 2.81 0.73 0.578 0.126 0.128 0.12
You do not start with difficult problems before 2.78 0.64 0.562 0.109 0.110 0.025
having comprehensive visions.
You dislike anything scary. 2.78 0.77 0.515 0.185 0.147 0.204
You continue to take risks without quitting, 2.40 0.72 0.106 0.851 0.097 0.075
even if it does not go smoothly.
Difficult problems are challenging. 2.59 0.73 0.055 0.825 0.098 0.024
You like the challenge of something new. 2.43 0.73 0.03 0.685 0.042 0.047
When you find challenging things, 2.87 0.67 0.116 0.383 0.077 0.293
you start them soon.
All events include some risks. 2.76 0.66 0.114 0.156 0.774 0.044
You accept some risks if you obtain benefits. 2.57 0.68 0.128 0.055 0.711 0.042
Society is comprised of a mixture of 2.88 0.62 0.181 0.055 0.618 0.034
dangers and safety.
Life is what keeps on the right side of dangers. 2.66 0.72 0.04 0.284 0.375 0.051
You take a risk without considering your ability. 2.90 0.69 0.013 0.07 0.037 0.76
Some say that you are daredevil. 3.02 0.67 0.053 0.065 0.045 0.758

Factor correlations Factor 1 e 0.171 0.114 0.291


Factor 2 e e 0.571 0.427
Factor 3 e e e 0.267
Eigenvalue 2.41 2.12 1.64 1.39
Cumulative contribution ratio (%) 15.1 28.3 38.6 47.3
Interpretation Chariness Challenging Balance Reckless
(reversed) (reversed) (reversed)
a
AM: arithmetic mean. SD: standard deviation. 4-point Likert scale.
b
Factor loadings after promax rotation. Loadings >0.5 are shown in bold.
10702 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 1 0 6 9 7 e1 0 7 0 7

were included in the model. These variables were then Factor analysis and integration of factors
selected according to AIC. Factor analyses were performed
with the psych package in R [32]. Calculation of Spearman's We executed factor analyses for awareness concerning envi-
correlation factors and the regression analysis were per- ronmental issues (Table 3), risk perceptions on H2 stations
formed with calculation by R [32]. (Table 4), scale of risk acceptance (Table 5), and risk-avoidance
tendency (Table 6). All factor analyses showed Kaiser-Meyer-
Olkin (KMO) values of 0.82e0.92 in terms of sampling ade-
Results quacy, and P < 0.001 (Bartlett's test) for exhibiting reliable es-
timates. Factor loadings were equivalent to the strength of
Respondents' knowledge and existing image of hydrogen correlations between the factor and the variables that
and hydrogen technology contributed to the factor [24].
As shown in Table 3, we provided question items on
Table 2 shows the results for questions about familiarity with awareness of global warming, energy resource depletion, air
seven power-generation systems, and knowledge and images pollution and credibility on mass media. Three factors labeled
of hydrogen and hydrogen technology. In terms of knowledge “Positive for problem solution”, “Negative for problem solu-
about hydrogen, mean scores are shown for respondents with tion”, and “Trust for mass media” were selected by the factor
a scientific degree, a non-scientific degree, and no degree. analysis.
Respondents with a scientific degree had higher scores than Table 4 shows the three factors extracted with high factor
the other two groups, indicating the former tend to have more loadings on risk perceptions of H2 stations. In previous studies,
correct knowledge about hydrogen. risk perception was well explained by two factors, i.e., “Dread”
and “Unknown” [24]. However, this study showed that it is
Ratio of acceptance for H2 station reasonable to postulate a third factor, called “Catastrophic”.
This factor had a larger eigenvalue than the “Unknown” factor.
We obtained quantitative data on the acceptance of H2 Rosa and Matsuda demonstrated the validity of assuming the
stations, after general information on H2 stations was pro- “Catastrophic” factor as a latent explanation factor in a struc-
vided to respondents. Fig. 1 shows the percentages for those tured model [33]. We believe the survey questions were
who agree/disagree under the Installation in the GS sce- appropriate for extracting the characteristics of interest. This
nario. We classified those who answered, the question with might have been due to the common inherent image of a
“Agree” and “Agree somewhat” as people “who feel sup- hydrogen accident, which is that of an instantaneous, massive
portive for installation of H2 stations”. The ratio of this explosion. As for hydrogen explosion hazard, people will easily
category (feel supportive) was 66.0%. More than half of re- imagine an instant hazard, but had difficulty imagining a lag-
spondents had a positive image of H2 stations. Supporters' ged hazard. This might be why Unknown was a weaker factor.
ratio for New construction scenario was 62.6%, similar to the Another explanation is that hydrogen explosion is an instant
Installation in the GS scenario. We analyzed the answer to phenomenon; thus, there is no delay of effect.
the question that “Do you agree with the H2 station under As for the scale of risk acceptance, four factors, called
the Installation in the GS scenario?” in detail in the following “Chariness”, “Challenging”, “Balance” and “Reckless were
analyses. extracted (Table 5). In terms of the tendency toward risk-

Table 6 e Arithmetic mean, standard deviation and factor pattern loading for risk avoidance scale and their interpretation.
KMO, 0.85; P < 0.001 (Bartlett's test).
AMa SDa Factor 1b Factor 2b
You worry about big accidents on airplanes and buses. 2.31 0.84 0.877 0.139
You fear nothing in comparison to your friends. 2.63 0.79 0.735 0.128
You often worry that you will be visited by a natural 2.43 0.78 0.551 0.141
disaster such as an earthquake, a flood, or a thunderbolt.
You mind the seat with high mortality rates when you take 2.02 0.75 0.359 0.269
cars or highway express buses.
You rarely eat food with additives and artificial coloring. 2.41 0.81 0.175 0.71
You take care about health. 2.79 0.72 0.221 0.685
You confirm escape door when you stay at hotels. 2.02 0.75 0.079 0.546
You worry if you do not check the locks and fire hazards 2.69 0.86 0.145 0.465
before you sleep.
You consider safety against fire when you select a house. 2.59 0.76 0.33 0.392

Factor correlations Factor 1 e 0.628


Eigenvalue 3.35 2.34
Cumulative contribution ratio (%) 27.9 47.4
Interpretation Fear of disaster Keep healthy
or accidents life
a
AM: arithmetic mean. SD: standard deviation. 4-point Likert scale.
b
Factor loadings after promax rotation. Loadings >0.5 are shown in bold.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 1 0 6 9 7 e1 0 7 0 7 10703

supporters and individual factors. Demographic attributes and


social factors were weaker explanation factors. We found no
significant differences based on age, although a significant
difference was observed between men and women. Psycho-
logical factors clearly explained H2 station acceptance. There
Fig. 1 e Percentages of those who agree/disagree with was a significant difference between the higher score group
installation of an H2 station in the existing gas station and lower score group for risk perceptions on H2 station and
nearest to their home (Installation in the GS scenario, scale of risk acceptance (Fig. 3). Acceptance of the former group
N ¼ 2069). was 82%, while that of the latter was 48% in terms of the Dread
factor. Regarding the acceptance of H2 stations, previous
avoidance, two factors called “Fear of disaster or accidents” research [7] pointed out that the intention to accept H2 stations
and “Keep healthy life” were extracted (Table 6). is generally influenced by the degree of anxiety about the
associated systems (e.g., reliability of the technology and pos-
Relationship between ratio of supporters and factors in sibility/types of accidents), which can lead to lower acceptance.
relation to respondents' attributes and risk perception Our study indicates a similar tendency in the intention to
characteristics accept H2 stations with qualitative explanation factors.

In addition to demographic attributes, psychological factors H2 station acceptance model


obtained in the previous section (Factor analysis and
integration of factors) helped explain H2 station acceptance. A logit regression model was constructed to identify stronger
Figs. 2 and 3 show the relationship between the ratio of factors for explaining acceptance under the Installation in the

Fig. 2 e Relationship between ratio of supporters for the installation of H2 stations at the gas station nearest to their home
and demographic factors. a: respondents who graduated college, university or higher were classified as “Higher”, others
were “Lower”. b: respondents who graduated from a scientific course at college or university were classified as “Yes” (22% of
all respondents). c: respondents who live in Tokyo Metropolitan or the other government-designated city were classified as
“Urban”. Respondents who live in cities with >300,000 population were classified as “Middle”, and others were “Rural”. d:
respondents who use vehicle more than once a week were classified as “More”, and the others were “Less”. * Significant at
5% level; ** Significant at 1% level; *** Significant at 0.1% level.
10704 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 1 0 6 9 7 e1 0 7 0 7

Fig. 3 e Relationship between ratio of supporters for the installation of H2 stations at the gas station nearest to their home
and knowledge and image of hydrogen, and psychological factors. Higher: group who had higher score than average; Lower:
group who had lower score than average. * Significant at 5% level; ** Significant at 1% level; *** Significant at 0.1% level.

Table 7 e Result of stepwise logit regression analysis (N ¼ 2069). Dependent variable: acceptance of H2 station with binary
transformation.
Independent variable Standardized b estimates (SE) z value
(Intercept) 1.445 (0.589) 2.46 *

Living area 0.066 (0.040) 1.63


Vehicle use 0.077 (0.028) 2.78 **

H2-knowledge 0.136 (0.025) 5.54 ***


H2-Clean 0.249 (0.060) 4.13 ***
H2-promising technology 0.068 (0.047) 1.46

Positive for problem solution 0.386 (0.091) 4.25 ***


Negative for problem solution 0.375 (0.093) 4.01 ***
Trust for mass media 0.206 (0.068) 3.05 **

Unknown 0.217 (0.084) 2.57 *


Dread 1.281 (0.100) 12.81 ***

Reckless 0.112 (0.066) 1.69 .


Balance 0.172 (0.070) 2.44 *

Pseudo R2 (Nagelkerke) 0.300

. Significant at 10% level; * Significant at 5% level; ** Significant at 1% level; *** Significant at 0.1% level.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 1 0 6 9 7 e1 0 7 0 7 10705

GS scenario (Table 7). We set acceptance of H2 station as the Previous research pointed out that people's social attitude
dependent variable with binary transformation, i.e., “Agree” and personal attributes were affecting factors in attitudes
and “Agree somewhat” indicated “acceptable”, and “Disagree towards H2 stations or hydrogen technology. A study by Huijts
somewhat” and “Disagree” indicated “unacceptable”. Inde- et al. [6] showed individual behavior norms were effective
pendent variables were selected by AIC. Pseudo R2 was 0.3. explanation factors for supporters of H2 stations, but that
Among the effective independent variables, “Dread” was education level and income were not effective. As for personal
the dominant factor, for which b was the largest. This means attributes, while several research groups have analyzed the
that alleviating inherent dread or fear by providing correct relationship between H2 station acceptance and distance
images or precise risk information will create better accep- from home to the nearest existing GSs stations, the results
tance of H2 stations. Information is needed on the charac- have been inconsistent. Work in the Netherlands showed that
teristics of hydrogen, scale of potential explosions, and more opponents lived near GS [17], while research in the U.K.
protection measures against heat or explosion. People tend showed a weaker tendency of opposition in the group living
to have an image of a hydrogen explosion as an instant, less than 200 m from a GS than in those living at a distance of
explicit hazard accompanied by heavy injury. “Unknown” 200e500 m [11]. In Norway, where H2 stations have already
was again a weaker explanation factor. This is because been installed, people with more knowledge of FCV and H2
people recognized that delayed adverse effects do not occur stations were less supportive of them than respondents who
by accident. Awareness of environmental problem solution lived 1 km further from the H2 station [14]. Similarly, we had
(“Positive for problem solution” and “Negative for problem anticipated the participants who lived farther than 1 km from
solution”) seemed be effective factors due to largeness of the a GS would tend to be less accepting of H2 stations and to have
b value. a stronger intuitive dread. However, the results showed there
was no significant difference for distance to GS. Other socio-
demographic factors could not explain well the difference
Discussion because the individual attributes had no apparent classifica-
tion, and living conditions are generally common and ho-
This is the first comprehensive survey on H2 station accep- mogenized in Japan. Japanese people do not have separate
tance in the Japanese public to explore factors affecting residential areas by income, nor do they have access to
acceptance in view of not only demographic attributes but also different news sources.
psychological perception. Sixty-six percent of respondents It is curious that the ratio of acceptance seemed inverse to
were open to accepting H2 stations. Respondents' tendency to the higher “H2-promising technology” and vice versa (Fig. 3).
accept the technology or not was well explained by psycho- This might be because the point was a summation of three
logical factors. There was a significant difference between the question items, and most respondents chose “3” (the middle
higher score group and lower score group for risk perception of alternative), which essentially said nothing, and so did not
H2 stations and scale of risk acceptance. Acceptance in the reflect their opinions well.
former group was 82%, while that in the latter was 48% in The logit regression model constructed here is a prototype
terms of the Dread factor (Fig. 3). This finding was not to estimate supporter's ratio status quo, and we successfully
consistent with results of studies conducted in the EU or US, selected important factors as dependent variables (Table 7).
where safety concerns were relatively low, and hydrogen was The results can aid in the identification of factors that will be
regarded to be as safe as gasoline or diesel fuel [3,20]. The important during explanation opportunities and exchanges of
survey results suggest that providing more information on H2 opinions, as well as the methods for such communications,
stations to address public fear will be an effective approach to when new infrastructures are installed. In the model,
convincing people that hydrogen technology is feasible. “Gender” was deleted among effective independent variables
We analyzed degree of flaming effect, which is one of due to its weak explanation power caused by binominal al-
cognitive bias in which people choose inconsistent answers ternatives of gender. “Catastrophic” was deleted among
when asked the same thing in opposite ways, such as “Do you effective independent variables, as well. This might be
think it is good?” or “Do you think it is problematic?” [34] We because the correlation between factors 1 and 2, i.e., “Dread”
provided two types of question regarding H2 station accep- and “Catastrophic” was 0.55 (Table 4), and showing a strong
tance, i.e., “Do you agree with the installation of H2 stations?”, multicollinearity between these factors. A similar model for
and “Do you feel the installation of H2 stations is problem- explaining hydrogen acceptance was provided in a previous
atic?” For the latter question, 24.4% and 28.3% of respondents study [11]. The model obtained in this study contains psy-
answered “Problematic somewhat” or “Problematic” under chological risk perception factors, and these explained
the Installation in the GS scenario and New construction perception characteristics on H2 stations and hydrogen more
scenario, respectively. Less than half of respondents had a directly.
negative image of H2 stations. Degree of inconsistency should As for future research, first, we will repeat this survey to
be small due to the higher (>r0.5r) Pearson's correlation coef- analyze the change of acceptance of H2 stations in relation to
ficient between the answer of these conditions (See Table S1, widespread diffusion of H2 station over time. Secondly, we
27e30). Thus, we focused to analyze on the answer to the plan to compare the case in which risk information and ben-
question that “Do you agree with the installation of H2 sta- efits around H2 stations are provided because the answers in
tions?” under Installation in the GS scenario in detail. Instal- this survey were based on respondents' existing images or
lation in the GS scenario is becoming more realistic and self-reported knowledge about hydrogen after being provided
common due to the Japanese Government initiative. with an explanation.
10706 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 1 0 6 9 7 e1 0 7 0 7

transportation in Spain: an unintentional focus on target


Conclusions population for a potential hydrogen economy. Int J Hydrog
Energy 2016;41:5203e8.
We conducted a nationwide survey in which we asked about [9] Martin E, Shaheen SA, Lipman TE, Lidicker JR. Behavioral
the acceptability of construction of a hydrogen station. A total response to hydrogen fuel cell vehicles and refueling: results
of California drive clinics. Int J Hydrog Energy
of 66% of respondents reported that the installation of an H2
2009;34:8670e80.
refueling facility in the GS nearest to their home would be
[10] O'Garra T, Mourato S, Pearson P. Analysing awareness and
acceptable. There were no significant differences in terms of acceptability of hydrogen vehicles: a London case study. Int J
age, although a significant difference was observed between Hydrog Energy 2005;30:649e59.
men and women. The acceptance of former group was 82% [11] O'Garra T, Mourato S, Pearson P. Investigating attitudes to
while that of latter was 48% as for Dread factor, and their hydrogen refuelling facilities and the social cost to local
difference was significant. Risk perception factors were better residents. Energy Policy 2008;36:2074e85.
[12] Roche MY, Mourato S, Fischedick M, Pietzner K, Viebahn P.
explanation factors of acceptance than attributes and
Public attitudes towards and demand for hydrogen and fuel
demographical characteristics. We constructed an accep- cell vehicles: a review of the evidence and methodological
tance model for H2 station including these risk perception implications. Energy Policy 2010;38:5301e10.
factors. This study will facilitate risk communication on H2 [13] Tarigan AKM, Bayer SB. Temporal change analysis of public
station construction and identify effective information for attitude, knowledge and acceptance of hydrogen vehicles in
reducing public fear of H2 stations. As for future research, we Greater Stavanger, 2006e2009. Renew Sustain Energy Rev
will repeat this survey to analyze the change of acceptance of 2012;16:5535e44.
[14] Tarigan AKM, Bayer SB, Langhelle O, Thesen G. Estimating
H2 stations in relation to widespread diffusion of H2 station
determinants of public acceptance of hydrogen vehicles and
over time. refuelling stations in greater Stavanger. Int J Hydrog Energy
2012;37:6063e73.
[15] Sherry-Brennan F, Devine-Wright H, Devine-Wright P. Public
understanding of hydrogen energy: a theoretical approach.
Acknowledgments Energy Policy 2010;38:5311e9.
[16] Zimmer R, Welke J. Let's go green with hydrogen! the general
We thank Mr. Yasuyuki Imamura and Ms. Etsuko Katoh for public's perspective. Int J Hydrog Energy 2012;37:17502e8.
their assistance in the data organization. This work was [17] Huijts NMA, van Wee B. The evaluation of hydrogen fuel
supported by National Institute of Advanced Industrial Sci- stations by citizens: the interrelated effects of socio-
ence and Technology (AIST) Grant for facilitating collabora- demographic, spatial and psychological variables. Int J
Hydrog Energy 2015;40:10367e81.
tions between Research Divisions in AIST.
[18] Huijts NMA, De Groot JIM, Molin EJE, van Wee B. Intention to
act towards a local hydrogen refueling facility: moral
considerations versus self-interest. Transp Res Part A Policy
Appendix A. Supplementary data Pract 2013;48:63e74.
[19] Thesen G, Langhelle O. Awareness, acceptability and
Supplementary data related to this article can be found at attitudes towards hydrogen vehicles and filling stations: a
http://dx.doi.org/10.1016/j.ijhydene.2017.03.021. Greater Stavanger case study and comparisons with London.
Int J Hydrog Energy 2008;33:5859e67.
[20] Schmidt A, Donsbach W. Acceptance factors of hydrogen
and their use by relevant stakeholders and the media. Int J
references
Hydrog Energy 2016;41:4509e20.
[21] Huijts NMA, Molin EJE, Steg L. Psychological factors
influencing sustainable energy technology acceptance: a
[1] European Commission. Introducing Hydrogen as an energy review-based comprehensive framework. Renew Sustain
carrier. Safety, regulatory and public acceptance issues. 2006. Energy Rev 2012;16:525e31.
[2] AcceptH2 Project. AcceptH2 Project n.d. http://www. [22] Mizuho Information & Research Institute Inc. Investigation
accepth2.com/index.html. [Accessed 17 February 2014]. on public acceptance of hydrogen. FY2007 Rep NEDO’s
[3] Schmoyer R, Truett T, Cooper C, Chew A. Results of the 2008/ Strateg Dev Polym Electrolyte Fuel Cell Technol Pract Appl.
2009 knowledge and opinions surveys conducted for the U.S. 2008 [In Japanese].
Department of Energy hydrogen program. 2010. [23] Mizuho Information & Research Institute Inc. Investigation
[4] Ricci M, Bellaby P, Flynn R. What do we know about public on public acceptance of fuel cells and hydrogen. FY2009 Rep
perceptions and acceptance of hydrogen? A critical review NEDO’s Strateg Dev Polym Electrolyte Fuel Cell Technol Pract
and new case study evidence. Int J Hydrog Energy Appl. 2009 [In Japanese].
2008;33:5868e80. [24] Slovic P. Perception of risk. Science 1987;236:280e5.
[5] Heinz B, Erdmann G. Dynamic effects on the acceptance of [25] Kleinhesselink R, Rosa E. Nuclear trees in a forest of hazards:
hydrogen technologies-an international comparison. Int J a comparison of risk perceptions between American and
Hydrog Energy 2008;33:3004e8. Japanese university students. Nucl Power Crossroads Comp
[6] Huijts NMA, Molin EJE, van Wee B. Hydrogen fuel station Risk Percept 1994:101e19.
acceptance: a structural equation model based on the [26] Slovic P, Fischhoff B, Lichtenstein S. Characterizing
technology acceptance framework. J Environ Psychol perceived risk. Perilous Prog Manag Hazards Technol
2014;38:153e66. 1985:91e125.
[7] Itaoka K, Saito A, Sasaki K. Public perception on hydrogen [27] Kleinhesselink RR, Rosa EA. Cognitive representation of risk
infrastructure in Japan. 2015. p. 1e9. Proc. ICHS2015. perceptions: a comparison of Japan and the United States.
[8] Iribarren D, Martı́n-Gamboa M, Manzano J, Dufour J. J Cross Cult Psychol 1991;22:11e28.
Assessing the social acceptance of hydrogen for
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 1 0 6 9 7 e1 0 7 0 7 10707

[28] Amano I, Kurisu K, Nakatani J, Hanaki K. Effect of provided [32] R Development Core Team. R: A language and environment
information and Recipient's personality on risk perception of for statistical computing. Vienna, Austria: R Foundation for
drinking water. J Jpn Soc Water Environ 2013;36:11e22 Statistical Computing; 2011. n.d, https://www.r-project.org.
[Abstract in English]. [33] Rosa E, Matsuda N. Risk perceptions in the risk society: the
[29] Kinoshita T, Yoshino K. Change of scale of risk acceptance cognitive architecture of risk of Japanese and Americans. In:
(SRA) within 10 years (1). In: Proc. 48th Jpn. Soc. Soc. Psych; Murakami Y, Kawamura N, Chiba S, editors. Towar. A
2007 [In Japanese]. peaceable futur. redefining peace secur. Kyosei from a
[30] Kusumi T. Individual differences in risk perception and Multidiscip. Perspect., Pullman, Wash. Thomas S. Foley
decision making under uncertainty. Jpn Psych Rev Institute for Public Policy and Public Service; 2005. p. 113e30.
1994;37:337e56 [In Japanese]. [34] Kahneman D, Tversky A. Choices, values, and frames. APA
[31] Murakami M, Ono K, Nakatani J. Perception of radiation risk Award Addresses 1984;39:341e50.
and willingness to return home following decontamination.
Environ Sci 2015;28:193e210.

Potrebbero piacerti anche