Background

Most energy consumed globally is generated from conventional resources. According to the International Energy Agency (IEA), 81% of the total primary energy consumption globally in 2017 was derived from fossil fuels [1, 2]. However, there are many drawbacks to using fossil fuels, since the generation of emissions from the combustion of these fuels has adverse effects on climate change, global warming [3], and humanity [4]. Moreover, increasing energy consumption will ultimately lead to exhausting conventional resources in the future [5]. According to Shafiee and Topal [6], in approximately 35 years from 2005, natural oil supplies will be completely depleted, while natural gas and coal stocks will be exhausted in 37 years to a 107 years’ time, respectively. As a result, coal will potentially be the only fossil fuel resource still existing after 2042, and it, too, will be depleted in 2112 based on current and future energy demands [6]. Therefore, research relating to these drawbacks and limitations, afforded to energy derived from fossil fuels, has attracted the attention and interest of many scientists to investigate renewable energy resources in many countries worldwide [7].

As the overall situation relating to the generation and consumption of electricity in Vietnam is unstable, and based on current projections, it is expected that power cuts will occur regularly from 2020, especially during the hot season [8]. Coal, gas and hydropower resources account for over 90% of the electricity generated in Vietnam, with 35% of total power production based on hydropower [9,10,11]. However, hydropower dams give rise to certain problems for the environment and society. For instance, in areas where dams are constructed, local communities must be aware [12,13,14] that the local ecosystem is threatened, and the environment harmed. In contrast, dams also bring several advantages, such as the generation of power, land irrigation and control of flooding [15,16,17]. In Vietnam, more than 200,000 people have been relocated due to the construction of hydropower dams, which involved the destruction of 44,557 homes and the loss of 133,930 hectares of land [18]. Moreover, the volume of river discharge and the capacity for power generation may decline due to deforestation in constructing hydropower dams [19].

Wind power, biomass energy and solar power are all sustainable sources of energy, the use of which is supported by the Vietnamese government through feed-in tariffs (FITs), whereby consumers can sell the electricity they generate to the electricity grid. Solar power particularly, is seen as a suitable source in generating electricity in the Central Highlands and the southern area of Vietnam [20]. The Central Highlands have a high potential for the generation of solar energy since it has an annual ‘sunshine’ duration of between 2000 and 2600 h [21]. After evaluating the solar radiation at 171 sites in Vietnam, Polo et al. [20] showed that the Central Highlands and South Vietnam, with the highest potential concentration of solar energy, had the capacity to generate an average of about 5.6 kWh/m2 of electricity per day. Daklak Province has an available rooftop space of 22.4 km2 for the installation of the system [22]. The Vietnamese government is also seeking to promote the installation of rooftop solar electricity generation systems at the second-highest FIT price for electricity that is generated directly from the burning of biomass [23,24,25,26].

Smart grid rooftop solar electricity systems offer further benefits compared to off-grid and hybrid solar systems given they are cheaper to instal and do not require a battery, which needs to be replaced every 2 to 5 years. Smart grid systems are also appropriate for installation in the Central Highlands since more than 95% of houses and other dwellings in this region are currently connected to the electricity grid [27].

There are five provinces in the Central Highlands of Vietnam, namely, Kontum, Gialai, Daklak, Daknong, and Lamdong, where Daklak is optimally located in the middle of the Central Highlands and regarded as the metropolis of the area [28]. This province also has the highest population in the region, with 1,919,147 people and 435,688 households, and the total duration of sunshine in 2018 was 2431.30 h [29]. However, even though rooftop solar electricity systems are available in the province, as at the 30th May 2019, only 18 from the 435,688 households had installed such systems, and almost all are smart grid systems [30]. It was therefore interesting to investigate why the number of people who had installed smart grid rooftop solar systems was so low and what are, the factors affecting the intention of people to instal such systems, and what percentage were willing to pay for installing instal the system. Accordingly, this study focused on these three issues.

Literature review

Definition of a smart grid

A smart grid (SG) is defined as a smart electricity/power grid, a future grid, a mobile grid, or an internal grid that is an evolution of the twentieth-century electrical grid. Conventional power grids are typically used to distribute power from a few central generators to a vast number of users. In contrast, a smart grid uses two-way electricity and information to create an integrated and distributed advanced electricity supply network [31].

In Vietnam, SG development was adopted by the Government in 2012 through Decision 1670/QD-TTg of October 2012, which encourages investments in renewable energy and SGs. SGs have two-way power, where customers can sell the surplus power from their renewable energy supply to the grid and buy electricity from the grid [32].

SG rooftop solar electricity systems

SG PVs can be mounted on the rooftops of buildings, offices or other commercial and industrial buildings. In comparison to other forms of solar energy, an SG rooftop PV offers a two-way exchange, where electricity can be directly targeted either from a power utility to consumers or in the opposite direction when the consumers’ systems generate more power than they consume. As such, an SG rooftop PV can help the PV system to run smoothly and in a stable manner in the following ways [33]:

  • If the customer’s need is > than the capacity generated by the system, the inverter will withdraw the power required from the grid;

  • If the customer’s need is = to the capacity generated by the system, the power generated from the rooftop solar electricity system will be used to meet the customer’s requirements;

  • If the customer’s need is < than the capacity generated by the system, the surplus electricity will be sent to the grid.

The process of determining the relationship between the generation of electricity and the need of consumers’ and controlling or regulating the flow of power occurs continuously and automatically, without the need for user intervention.

In Vietnam, the surplus energy acquired from domestic solar power systems and the power available from the grid has been determined by two-way meters from 1st July 2019, based on Decision No. 02/2019/QD-TTg. Nevertheless, consumers can sell power to EVN at an FIT price of USD 9.35/1kWh independent of the power they purchase from the grid [26].

Smart grid rooftop solar systems are highly suitable for the Central Highlands as 95.17% of households in this region are currently connected to the electricity grid [27]. The benefits of SG rooftop PVs are lower installation costs and the fact that consumers will not need to replace the batteries every 2 to 5 years. In addition, unused electricity from the system can be sold to EVN at a reasonable price.

Methodology

Study area

As mentioned earlier, the Central Highlands of Vietnam have a high potential for solar energy, and almost all households in the area are currently connected to the electricity grid. Therefore, it is a suitable region for the installation of smart grid rooftop solar electricity systems. Daklak Province comprises a city, named Buon Ma Thuot, a town, and 13 districts. This study was conducted in Eatam Ward in Buon Ma Thuot City (see Fig. 1). This ward is close to the centre of Buon Ma Thuot and has 1,154 households [29]. The economy of Eatam is reasonably well developed, having a large university; the Tay Nguyen University.

Fig. 1
figure 1

(source: https://daklak.gov.vn/web/english)

The study area

Full size image

Sample size and data collection

The target population for this research comprised 1154 households in Eatam Ward [29]. The sample was calculated based on the following formula [34]:

$$n = \frac{N}{{1 + Ne^{2} }},$$

where n is the size of the sample, N is the size of the population, and e is the sampling error \((\pm 5\mathrm{\%}\)). Thus, the sample was determined as:

$$n = \frac{{1,154}}{{1 + 1,154x0.05^{2} }} = 297.03.$$

Furthermore, the sample size adopted was 300 households.

A draft questionnaire was developed based on a review of previous studies, with the selected items appropriately adjusted to the Vietnamese situation. The questionnaire (refer to Appendix 1) commenced with an introduction about smart grid rooftop solar energy systems, after which, the items that constituted the main part of the questionnaire were addressed to the respondents. A pre-survey was conducted before the main study to ensure that all the items in the questionnaire were capable of being understood by the respondents in the main survey. After the pre-survey, the main survey was conducted in the research area between May and July 2019 with the 300 households purposively selected by convenience sampling. The survey investigated the intention of the participating households to instal rooftop solar systems in the future. Only those households that had not installed smart grid rooftop solar electricity systems participated in the survey. The sample was selected so that only those households whose dwellings had a rooftop deemed appropriate for the installation of a rooftop solar power system were included. The questionnaire consisted of two parts. Part 1 contained questions relating to the factors that affected their intention to instal a rooftop solar power system, while Part 2 collected information relating to the demographics of the households.

In Vietnam, unmarried people often live with their parents and are unable to make their own personal decisions. They also do not have their own house, so the installation of solar systems is not possible. Therefore, only households that comprised a married couple and their relatives were included in the survey. In some families, the husband answered the survey items, while in others, the items were answered by the wife. The items were all addressed to the participants in the local language, Vietnamese, by the researchers, who recorded the respondents’ answers.

Data analysis

The research model was designed according to previous research. Customers mostly purchase electricity based on the payment of a monthly bill rendered in arrears for the actual amount of electricity consumed during the preceding period. However, installing a rooftop solar power system involves a substantial, long-term financial commitment, and there are uncertainties associated with the payback period and future utility costs as well as the likelihood of improvements occurring in rooftop solar technologies. Ek found in 2005 that individuals who show a positive attitude towards wind power pay more attention to the unfavourable consequences of environmental problems to humans and the ecosystem resulting from conventional electricity generation, and are probably, therefore, more focused on collective or altruistic values [35]. Attitudes are composed of views and opinions toward objects that vary in specificities [36]. Therefore, people may have beliefs and may care or be concerned about specific problems such as the effects of fossil fuels on climate change and global warming. Consequently, this may lead them to adopt certain behaviours, as posited in the theory of planned behaviour [37]. According to Roberts, consumers will opt for environmentally friendly products if they have good knowledge of such products or the contribution of the products to the health of the environment or of other individuals [38]. Faiers and Neame meanwhile, suggested that consumers may be uncertain and confused if they do not have sufficient knowledge about products [39]. Previous empirical research has concluded that environmental concerns and values, and the innovativeness of consumers will have an impact on people’s intention to instal solar electricity systems [40,41,42]. Government incentives have also been found to have a strong influence on such intentions [42].

Based on the theory of planned behaviour (TPB), the intention of households to instal smart grid rooftop solar electricity systems would be influenced by their awareness of rooftop solar electricity generation and other environmental concerns, the innovativeness and attitudes of the households, and the incentives being offered by the government to instal such systems, as illustrated in Fig. 2.

Fig. 2
figure 2

Research model [40,41,42]

Full size image

A probit binary model was used in this research to analyse the factors that affect the intention of households to instal smart grid rooftop solar systems, as illustrated in Eq. 1 below:

$$\user2{Y}{\text{ }} = {\text{ }}\beta x + \varepsilon ,$$
(1)

where Y is the dependent variable: intention to instal smart grid rooftop solar system, β is the coefficient valuation, and x, independent variables: awareness about rooftop solar electricity, environmental concerns, innovativeness, attitudes, government incentives. The independent variables were measured using items employed in earlier studies (see Table 1) [41,42,43].

Table 1 Items used to measure independent variables
Full size table

In this research, the dependent variable for each household (Yi) was assessed as a binary variable as follows:

Yi = 0: the household does not intend to instal a rooftop solar electricity system;

Yi = 1: the household intends to instal a rooftop solar electricity system; where Yi is the intention to instal of family i

Since Y can only have values of 0 and 1, it cannot be explained by a linear regression model. However, a probit model can show the probability of Y falling between 0 and 1 based on a cumulative normal distribution given any Z-score, with (Z) € [0, 1].

Thus, it can be stated that:

$$\begin{gathered} Y{\text{ }} = \phi ~\left( {{\text{X}}\beta ^{\prime} + {\text{ }}\varepsilon ^{\prime}} \right), \hfill \\ \phi ^{{ - 1}} \left( Y \right){\text{ }} = {\text{ }}X\beta ^{\prime} + {\text{ }}\varepsilon ^{\prime}, \hfill \\ \end{gathered}$$
$${\mathbf{Y}}* = {\text{ }}{\mathbf{X\beta ^{\prime}}} + {\text{ }}{\mathbf{\varepsilon ^{\prime}}}.$$
(2)

Here, ε’ ~ N (0,σ2), where Y* ranges from −∞ to ∞. Y* is also called a latent variable.

Whether or not a family intends to instal a rooftop solar electric system can be expressed as the relationship between the linear regression model (Eq. 1) and the latent variable (Eq. 2) as follows:

$$\begin{gathered} Y\, = \,0{\text{ if}}Y*\, < \,0{\text{ }}({\text{the household is not likely to instal a solar electricity system}}), \hfill \\ Y\, = \,{\text{1 if}}Y*\, > \, = \,0{\text{ }}\left( {{\text{the household is likely to instal a solar electricity system}}} \right). \hfill \\ \end{gathered}$$
(3)

Thus, a probit model was used to estimate the probability of people installing a solar electricity system.

The conditional probability of customer i, being likely to instal a solar electricity system (i.e. Yi = 1) given xi, is expressed as:

$$P\left( {{\text{Yi}} = 1} \right) = P\left( {Yi = 1|xi} \right) = P\left( {\beta 'xi + \varepsilon _{{1i}} \ge 0} \right){\text{|}}xi{\text{)}} = P\left( {\frac{{\varepsilon _{i} }}{\sigma } \ge \frac{{ - \beta 'xi}}{\sigma }} \right) = \emptyset \left( {\frac{{ - \beta 'xi}}{{\sigma _{1} }}} \right),$$
(4)

where \(\emptyset {\text{~is}}\) the cumulative standard normal distribution function.

The conditional chance of Y = 0 (i.e. prefers not to instal a solar electricity system) is given by:

$$P\left( {Yi = 0|xi} \right) = 1 - \emptyset \left( {\frac{{ - \beta 'xi}}{{\sigma _{1} }}} \right).$$
(5)

Therefore, the probability of the observed household is:

$$\mathop \prod \limits_{{I = }}^{{N_{1} }} \left[ {\emptyset \left( {\frac{{ - \beta 'xi}}{{\sigma _{1} }}} \right)} \right].\mathop \prod \limits_{{N_{1} - 1}}^{N} \left[ {1 - \emptyset \left( {\frac{{ - \beta 'xi}}{{\sigma _{1} }}} \right)} \right],$$
(6)

where N1: household wants to instal a rooftop solar system, and N-N1: household does not want to instal a rooftop solar system.

In this model, a higher value of β’i indicates that the household is more likely to instal a solar electricity system. The value of β’i indicates that a one-unit change in Xi leads to a β’i change in the Z-score of probability Y. The SPSS v.20 software program was used to calculate the probabilities and to conduct other statistical analyses.

Determination of reliability

The reliability of the items in the research instrument through which the data were measured was determined using Cronbach’s alpha, and the corrected item–total correlation. Cronbach's alpha is a coefficient of coherence that shows the close connection of items to a group. The purpose of Cronbach's alpha is to determine if the observed variables measure the same for a concept that is being measured. The contribution of the variables is reflected by the corrected item–total correlation, whereby it is allowed to eliminate inappropriate variables in the research model. The reliability is considered acceptable if the Cronbach’s alpha is ≥ 0.7, and the corrected item–total correlation is > 0.3 [44].

Exploratory factor analysis

An exploratory factor analysis (EFA) was used to determine the factors that affect the intention of households to instal rooftop solar electricity generation systems. The conditions of the analysis were as follows:

  • Factor loading > 0.5 and Kaiser–Meyer–Olkin (KMO) 0.5 ≤ KMO ≤ 1;

  • Bartlett test (Sig. < 0.05) and the average variance extracted > 50%; and

  • Variables with KMO < 0.5 were removed from the model [44].

Multicollinearity diagnostics

Multicollinearity is the phenomenon of independent variables strongly correlating with each other. A regression model with multicollinearity will render the results of quantitative analysis not meaningful and incorrect. Variance inflation factors (VIFs), which measure collinearity and tolerance, were applied in this research to determine the multicollinearity phenomenon. Menard [45] stated that a tolerance of below 0.20 is a major concern, while a tolerance of just under approximately 0.10 probably implies a severe collinearity problem. Because a VIF is the opposite of the tolerance, a tolerance of 0.20 relates to the rule of 5 and a tolerance of 0.10 to the rule of 10 [46]. Hair et al. [47] proposed that a VIF of less than 10 is representative of irrelevant collinearity [48]. Kennedy [49] claimed that a VIF > 10 indicates harmful collinearity for standardized results [50].

Results

Profile of respondents

The study was conducted with a sample of 300 households in the Daklak province, of which 51% were males, and 100% were married. The electricity consumption for the majority of the people was at levels 3 and 4, with percentages of 39.7% and 29%, respectively. The detailed information of the respondents is given in Table 2.

Table 2 Profile of respondents
Full size table

Factor analysis

Reliability statistics

Before analysing the factors that affected the intention to instal a smart grid rooftop solar electricity system, the Cronbach’s alpha was calculated to test the reliability of each of the items in the survey.

The reliability was considered acceptable if the Cronbach’s alpha was ≥ 0.7, and the corrected item–total correlation was > 0.3 [44] As can be observed from Table 3, the reliability of all the items in the survey was acceptable except for the items, Awareness1 and Awareness3. Removing those two items from the model and repeating the process produced the revised Cronbach’s alpha for factor 1, as shown in Table 4.

Table 3 Reliability statistics
Full size table
Table 4 Cronbach’s alpha of the first factor after removal of items, Awareness1 and Awareness3
Full size table

After removing the two items, all the variables were considered to be suitable for the model. An EFA was, therefore, conducted with all the variables without Awareness1 and Awareness3.

Exploratory factor analysis

In this study, the KMO was found to be equal to 0.785, with the significance of the Bartlett test being equal to 0.000, and the sum of the variance extracted being 76.791%. All the factor loadings were higher than 0.5, as illustrated in the rotated component matrix illustrated in Table 6.

Multicollinearity diagnostics

The results would be incorrect and would not be meaningful if there was multicollinearity. Therefore, the study had to be diagnosed for multicollinearity before any analysis could be performed. As shown in Table 5, all the VIFs were well under 5, and the tolerance was greater than 0.2. Therefore, there was no multicollinearity in the model. Thus, all the variables were appropriate for analysis.

Table 5 Multicollinearity diagnostics
Full size table

Probit binary model

After testing the reliability of the items, conducting the EFA, and removing the results of the items, Awareness1 and Awareness3, from the data, the probit binary model was used to explore the intention of the sample of 300 households to instal a smart grid rooftop solar system. As noted above, the intention to instal was a binary dependent variable with a value of 0 indicating that the people who had no intention of installing a solar electricity system, and a value of 1 indicating a positive intention. As shown in Table 6, the independent variables used in the model were labelled as Household_awareness, Environmental_concern, Household_innovativeness, Attitude, and Government_incentive, which corresponded with the five factors derived from the following items:

Awareness = mean (Awareness2, Awareness4, Awareness5).

Environmental_concern = mean (Environmental_concern1, Environmental_concern2, Environmental_concern3).

\(Household\)_innovativeness = mean (Customer_innovativeness1, Customer_innovativeness2, Customer_innovativeness3).

\(Household\)_attitude = mean (Attitude1, Attitude2, Attitude3).

Government_incentive = mean (Government_incentive1, Government_incentive2, Government_incentive3, Government_incentive4).

Table 6 Rotated component matrix
Full size table

The model that was used to predict the dependent variable was as follows:

$$\eqalign{ Intention~to~ & install* \cr & = ~{{\beta '}_0} + {{\beta '}_1}Household\_Awereness + {{\beta '}_2}Environmental\_Concern \cr & + {{\beta '}_3}Household\_Innovativeness + {{\beta '}_4}Household\_attitude \cr & + {{\beta '}_5}Government\_Incentiv{e_e} + \varepsilon ' \cr}$$

From the total of 300 households, 99, representing 33% of the surveyed sample, indicated that they intended to instal smart grid rooftop solar systems on their houses (refer to Table 7). However, most households (67%) had no intention of installing such a system. The reasons that were given by those who did not intend to instal rooftop solar systems varied; some said that they did not have enough money to instal such a system. In contrast, some did not know about rooftop systems and had doubts about their efficiency, and wanted to know what guarantee existed in that such systems would operate as they were intended to.

Table 7 Categorical variable information
Full size table

From the results in Table 8, the Sig. of Customer_innovativeness was 0.74, which is higher than 0.05, and thus, the effect of Customer_innovativeness on the intention to instal was not significant. The remaining four Sig. values were all less than 0.05, and therefore, those four factors had a significant effect on the dependent variable. Most of the coefficient β values were positive, with the exception of Intercept β’o, which means that the effects were all in the same direction, i.e. the effects were all positive. The β’ of the awareness variable was the highest, indicating that this factor had the greatest effect on the intention to instal a solar electricity system. Government incentives were also found to be an important factor in the people’s decision to instal smart grid rooftop solar electricity systems in their houses.

Table 8 Parameter estimates
Full size table

In fact, most of the people who answered the questionnaire had limited knowledge concerning smart grid rooftop solar electrical systems, and many thought that such systems were only intended to generate hot water via solar power. Therefore, the researcher needed to provide an introduction concerning smart grid solar systems before administering the main survey.

Table 9 shows the probabilities of households deciding to instal smart grid rooftop solar systems in their houses. As can be seen, only 2% of the sample recorded a probability from 0.401 to 0.500, and none of the probability measurements exceeded 0.500. The majority of the probability measurements of the intention to instal solar electricity systems were between 0 and 0.010. The detailed probabilities of the intention to instal smart grid rooftop solar electricity systems for each household appear in Appendix 2.

Table 9 Probability of intention to instal a smart grid rooftop solar electricity system
Full size table

It was also indicated that people would be willing to pay (WTP) to instal a rooftop solar electricity system or would encourage other people to do so if there was support from the government. The largest group (39.3%) of respondents suggested that the government should support 30% of the installation cost for smart grid rooftop solar electricity systems, with the second-largest group (27%) supporting the idea that preferential-rate loans should be made available by commercial banks. Other methods of encouraging people to instal smart grid rooftop solar electricity systems that are not detailed in Table 10 included the idea that people should be able to pay for the installation by instalments over a 3 to 5-year period or that the government should support 50% of the installation cost. Some people also suggested that there should be more communication activities to provide information about smart grid rooftop solar electricity.

Table 10 Willingness to pay for system installation cost
Full size table

The amount of money that households in the area would be willing to pay for installing a smart grid rooftop system

In the introduction to the questionnaire, the cost of installing a smart grid rooftop solar system was indicated to be in proportion to the level of electricity used (see Appendix 1). The households that indicated that they intended to instal such a system were made aware of the cost of installing a system with a generating capacity appropriate for their needs, and they indicated their willingness to pay (WTP) that cost. The breakdown of the amount that the households intending to instal a rooftop solar electricity system were willing to pay (WTP) is shown in Table 10.

Discussion

Research limitations

Even though all areas of Vietnam are appropriate for the exploitation of solar energy [20], the current research concentrated on only one area in Vietnam—the Central Highlands, which has the highest potential due to the duration of sunshine it enjoys annually. In addition, only five factors (awareness, environmental concerns, household innovativeness, household attitudes, and government incentives) were considered, with the data being gathered by a series of items measured using a five-point Likert scale, based on which a model was constructed to determine the effects of the independent variables on the intention of households to instal smart grid rooftop solar electricity systems in the research area. Moreover, only those households occupied by married couples and their families were analysed.

Policy implications

The research provides important information for the Vietnamese government, enabling them to enhance their policy of developing alternative energy resources. Based on the results, the awareness of households of smart grid rooftop solar electricity systems and government incentives were the two most influential factors that affect the intention of the people to instal such systems. Based on these findings, the following two policy measures are proposed for adoption by the government as they are likely to encourage the installation of rooftop electricity generation systems.

Firstly, the Government should identify those households that possess a high awareness concerning rooftop solar power and offer incentives directly to them. The results of this research showed that people with a high level of awareness (i.e. scores of 2.5 to 5 in the Awareness section of the questionnaire) had a higher intention to instal. The Government can also conduct some communication programmes to enhance the awareness of people about rooftop solar electricity so that programmes for the development of alternative energy will be more effective. As happens in Vietnam with every government programme, the People’s Committee will disseminate information to the local people through workshops or meetings and hence, should send invitations to households with the potential to instal rooftop solar systems to attend at a particular date and time. At the meeting, experts in smart grid rooftop solar systems should talk about the functionality and usefulness of these systems so that people will clearly understand the benefits. The Chairman of the Committee should then introduce the incentives that are available to the people who have been invited and have attended the meeting.

In terms of government incentives and based on the recommendations made by the respondents in the research area, the following measures are suggested for implementation in the long-term by the Vietnamese government:

  • Require commercial banks to offer preferential loans with lower interest rates than for those applying for other forms of credit for the installation of domestic electricity generation systems based on renewable energy sources.

  • Offer a monetary gift (e.g., VND 5 million) when people instal a rooftop solar system on their house.

  • Subsidize 10–30% or even 50% of the installation cost of domestic solar rooftop electricity systems in the same way as support is currently being offered to companies that instal such systems and in a similar way as the Indian government is offering such support in their PV rooftop programme.

  • Enhance communication activities concerning smart grid rooftop solar electricity.

Conclusion

The study demonstrated that households have so far installed only a very small number of smart grid rooftop solar electricity systems in the Daklak Province, Vietnam. Moreover, few people are aware of the advantages of using renewable energy resources. However, after the information of rooftop solar electricity systems had been provided in this survey to the respondents in the research area, they understood the concept and 33% indicated an intention to instal such a system in their house, with the majority willing to pay (WTP) between USD 1240 and USD 2220. People with a high level of awareness of smart grid rooftop solar power are more likely to pay to instal such a system, but government incentives and household attitudes are also important factors to consider that influence the intention to instal. However, it was found that environmental concerns and the innovativeness of households were less influential factors.