Customer's response to dynamic pricing in utility energy Tariff quality and reliability with the time of use: An Empirical case study of household electricity customers in Indonesia

Abstract

Time of Use (ToU) is one of the types of dynamic pricing strategies at which a pricing scheme employs a variable rate rather than a fixed one causing electricity tariffs over a particular consumption period to rise once in high consumer demand at peak times and to fall once the demand is low at off-peak times. Owing to the important factor of perceiving customers' response to the successful implementation of the ToU tariff scheme, this study examines the household electricity customers' response in Indonesia - especially residents of Java Island and Bali Island - to future implementation of this dynamic pricing strategy in utility energy tariff quality and reliability. Data on the behavior of the PLN customers need to be thoroughly gathered and then analyzed quantitatively for well-informed decisions on a set of variable rates based on peak/off-peak times of the day and the season to establish in a board of directors' room. Hence, survey work is conducted, and a questionnaire is carefully designed and developed in this study to help collect and analyze the data on consumer behavior. The results show that customers’ response exhibits a promising outcome. At the total number of 451 respondents, 63 % of the survey participants expect to opt into the ToU scheme and the reason for this is their sensitivity to a bargain price of the electricity rate. However, 37 % of the respondents find ToU burdensome to implement and very demanding, as a result they opt out of the ToU scheme. From the results of the survey, it can also be inferred household energy-related activities that customers are quite willing to shift are washing and ironing clothes where a shift in these activities would lead to potential savings in the electricity bills of 26 % off the regular price. From the use of the quantitative method in this study, the total potential daily load reduction of energy consumption with the ToU tariff scheme could reach 3,23 kW for each customer. The findings conclude that customers have presented an encouraging response to future implementation of ToU and express their thoughts ToU seems to bring potential benefits of saving electricity bills and looks as if it is fairly easy for customers to adapt to shifting their energy usage from peak times to off-peak times.

1. Introduction

Successful network modernization consists not only of deploying new infrastructure and technologies but also of leveraging technology to promote service and active customer engagement. In the case of power companies, the use of this technology can be implemented through the use of smart meter technology which is also designed to support the Home Energy Management System (HEMS). Unlike conventional meters in general, smart meters can record customer energy consumption in real-time at half-hour intervals or even more often [1]. This certainly makes it easier for customers to find out how much energy has been consumed at each time interval per day and its use has the potential to have a good economic impact [1,2].

Different pricing strategies often referred to as dynamic pricing are part of the business case for the application of smart meter technology [1]. With dynamic pricing, different electricity rates will be charged to customers based on time, BPP (biaya pokok penyediaan listrik or cost of electricity supply), and network conditions. Customers will be given several offers related to the level of variability in electricity prices and are encouraged to use energy more wisely and efficiently in responding to their electricity tariff choices [3]. One of the dynamic pricing strategies commonly adopted in determining electricity rates is the Time of Use (ToU) scheme [4]. In ASEAN, the only country that adopted the ToU scheme is Malaysia. Malaysia adopted this scheme for commercial and industrial electricity customers. Residential customers currently still apply fixed pricing for the electricity that has been used for every kWh [5]. The implementation of the ToU itself is marked by the determination of electricity unit prices according to different consumption periods, where electricity tariffs during peak loads will be higher than electricity rates at off-peak load times and vice versa [6]. Several studies have stated that this electricity tariff has the potential to stimulate customers to shift their electricity consumption from peak load periods to off-peak load periods so that it can have an impact on reducing costs and increasing system efficiency [1,7].

In Indonesia, the electricity supply utility company, i.e. Perusahaan Listrik Negara commonly known as PLN, which has a statutory obligation to provide generation as well as transmission and distribution of electricity nationwide throughout the Indonesian archipelago, has the potential to adopt the ToU price scheme in line with the increasing demand for electrical energy, especially in the household group. To adopt the ToU price scheme, customers have to install the smart meter in their homes. Currently, as of June 2023, more than 103.615 customers have installed smart meters, and this will be increased to reach 1.217.256 customers by the end of the year [8]. A smart meter can save electricity use as the customers can monitor their electricity spent in real-time. This can be one of strategies to embody the PLN's Green Plan that related to the energy efficiency [9]. Nevertheless, Indonesia itself is still implementing a flat pricing scheme to be offered to its customers [10].

According to the statistic report of PLN in 2022, Java-Bali has the most household customers than any other island customer in Indonesia, with 48,352,329 combined customers or 61.73 % of all PLN's household customers in Indonesia [11], so Java-Bali customers have the biggest potential to be implemented this ToU price. Even so, the implementation of the ToU scheme in the household electricity category has never been realized and is a new thing for customers. Although theoretically and practically in several countries the implementation of the ToU scheme can be said to be effective and provide benefits for both customers and electricity providers [1], in fact it is still unclear whether in Indonesia the implementation of the ToU scheme can get a positive response and be accepted by customers. Customers may give a negative response because it is not easy to adjust and understand the effectiveness of the ToU tariff design which tends to be more complex than pricing with a flat rate scheme. In fact, consumer behavior and response are important factors in the successful implementation of dynamic pricing and maximize the company's long-term profit [12]. Seeing this, customer responses to the implementation of the ToU tariff scheme need to be studied and analyzed. Many studies have been conducted to look at customer responses to the implementation of several dynamic pricing schemes, but there are still few that only focus on ToU rates. In addition, research related to customer response in Indonesia, especially the Java-Bali region, has not been found. Therefore, to fill this research gap, an analysis of how the Java-Bali customers respond to the planned implementation of the ToU tariff scheme needs to be analyzed further (see Fig. 1).

Fig. 1.

Typical Dispatch Generator on Java-Bali System (Source: PLN LITBANG).

2. Problem definition

The existing condition of the PLN load for household groups in Java-Bali shows a significant difference on weekdays (Monday-Friday) and weekends (Saturday-Sunday). Based on expert judgment analytics, weekdays three peak load periods, that is in off-peak (1–7 a.m.), mid-peak (12 a.m., 8 a.m.,10–11 p.m.), and on-peak period (9 a.m.-9 p.m.), whereas weekend has two peak load periods, that is in the off-peak period (12 a.m.-5 p.m.) and on-peak period (6–9 p.m.), as shown in Fig. 1, Fig. 2. This fluctuating system workload condition will greatly affect the electricity supply costs (BPP) that need to be issued by PLN. Planning the implementation of the Time of Use (ToU) tariff scheme can be done as an alternative in solving this problem.

Fig. 2.

Java-Bali System Daily Load Profile for Household Groups.

(Source: PLN LITBANG)

The contribution of this study in particular are.

• 1.Knowing customer responses, especially for household customers in Java-Bali to the ToU tariff scheme;
• 2.Knowing the potential use of electricity by the customer, so that it can help balance and/or reduce the electrical load by shifting some activities that allow it to be shifted to another period with a lower load. This is in line with the main function of HEMS, i.e. optimizing energy use in households with smart meters through energy savings and reduced consumption;
• 3.Knowing the prediction of the network load curve after the implementation of the ToU tariff scheme.

There are also limitations imposed in carrying out this study, including.

• 1.Customer responses to the ToU tariff scheme will only be analyzed for household customers;
• 2.The secondary data used is data on customer load and the cost of supplying electricity (BPP) in 2019;
• 3.The secondary data to be used is fully provided by PT PLN (Persero) Center for Research and Development of Electricity;
• 4.The survey will only be conducted online with the sample of the study i.e. household electricity customers in Java-Bali.

3. Methodology

3.1. Subject of research

The subjects of this study are Indonesian people who use household electricity and live on the islands of Java and Bali. Specifically, the customer in question is a customer who subscribes to household electricity with a subscription power of 1300 VA to 6600 VA.

3.2. Research steps

In general, several steps need to be carried out in this research, these steps are visualized through a flow chart as shown in Fig. 3., and each step can be explained as follows.

Fig. 3.

Research Step Flowchart.

3.2.1. Study of Literature

At the literature study step, a search for several references related to the topic of this study was carried out to get an overview related to research instruments that can be used in finding out consumer responses and also data processing methods.

3.2.2. Secondary Data Collection

In this study, secondary data were obtained directly from PLN Indonesia. Some of the secondary data needed in this study are consumer load data and the marginal cost incurred by PLN in meeting the electricity consumption of its customers. The data provided is data from 2019, especially for the Java-Bali area.

3.2.3. Research Instrument Design

In order to find out how the customer responds to the application of the Time of Use (ToU) tariff scheme on household electricity, a research instrument in the form of a questionnaire is used. The questionnaire in this study is the result of adaptation and development of research [1]. There are two variables that are part of the research questionnaire, namely Demographics and Customer Response (CR).

In the customer response variable, four main questionnaire items are used to determine customer ratings of the ToU tariff scheme that has been described in the questionnaire. Three of them are statements that are measured using a five-scale Likert scale, where 1 represents strongly disagree to 5 represents strongly agree with the given statement. The next item is a statement regarding whether the customer is willing or not willing to join this tariff scheme. If the customer is willing to join, they will find out more about the reasons for their willingness, the possible activities to be moved, and also the amount of electricity bill savings they want through three open-ended questions and multiple choices. Meanwhile, if the customer is not willing, they will also be asked the reasons underlying their choice. The four items used in the questionnaire are listed in Table 1.

Table 1.

Adaptation andDevelopment ofQuestionnaireItems forCustomerResponseVariables.

Questionnaire Items	Items Code
I think the application of the Time of Use tariff scheme is a good idea	CR1
With the Time of Use tariff scheme, I can save more on my electricity bill	CR2
For me, learning to adapt to the Time of Use tariff scheme is an easy thing	CR3
If the Time of Use tariff scheme is offered to me now, then I am willing to join this tariff scheme	CR4

3.2.4. Selection of Respondent Criteria

This research is still limited to household electricity customers in Java-Bali. Using a research instrument in the form of a questionnaire, several criteria were determined for respondents who can participate in this study, including (1) living in Java/Bali, (2) living in a household, (3) subscribing to electricity for a household group of at least 1300 VA, and (4) knowing the daily use of electricity at home. The selection criteria for the minimum subscription power of a customer, which is 1300 VA, is carried out by considering the possibility that customers can shift their electricity use at certain times.

3.2.5. Pilot Study of Research Instrument

The pilot study is the preparatory stage before the actual survey process is carried out. This stage is carried out with the aim of testing the research instrument in this case a questionnaire that will be used later. The pilot study steps are visualized through a flow chart as shown in Fig. 4. In this study, a pilot study was conducted on 35 respondents who had met the criteria for research subjects. The data that has been obtained from the pilot study process will then be used to test whether the questionnaire items used in the questionnaire are valid and reliable. Questionnaire items that will be tested for validity and reliability are items that use a Likert scale and the Yes/No option in measuring respondents' views, namely items CR1-CR4.

Fig. 4.

Pilot Study Step Flowchart.

Testing the validity of the questionnaire items is done by using the data on each item to find out how well the item can measure a concept correctly [13]. The most common methods used in testing the validity of questionnaire items are the Pearson correlation and Spearman methods. Basically, to test the validity of the questionnaire items used, these two methods will both compare the correlation value of the item to the total (r count) with the value of r in the distribution table. If the value of the item to total is greater than the value of the r table, then the questionnaire item can be said to be valid [14]. Pearson's method itself belongs to the parametric statistical method, while Spearman's belongs to the non-parametric statistical method. The parametric statistical method was chosen when the data used were normal. Conversely, if the data is not normal then non-parametric statistical methods tend to be used. Therefore, to be able to determine which validity testing method will be used, it is necessary to first test the normality of the data. The normality test itself is used to find out how the distribution of the data is held. Data with a normal distribution can be seen from the increasing number of data whose values are close to the average value of the overall data and the decreasing number of data that is far from the average value. Normality testing can be done in several ways, such as skewness and kurtosis values, Kolmogorov-Smirnov test, and Shapiro-Wilk test. Shapiro-Wilk is the most robust test for all types of distributions and sample sizes. In this study, the Kolmogorov-Smirnov and Shapiro-Wilk tests will be conducted to test the normality of the data. Conclusions related to the normality of the data can be drawn by looking at the significance values generated in the two tests. If the resulting significance value < 0.05, it can be said that the data is not normally distributed.

Conversely, questionnaire item reliability testing is also needed to find out how the stability and consistency of an item measure a concept. Where in this study, reliability is measured by testing how the consistency of the answers from the respondents themselves. Reliability can be tested using the coefficient value of Cronbach's Alpha (α). According to Ref. [13], if Cronbach's Alpha value is less than 0.6 then the reliability of the questionnaire item is classified as poor, if the value is in the range of 0.7 then it is acceptable, and if the value is more than 0.8 then it is classified as good.

By using tools in the form of IBM SPSS Statistic 26 software and also Microsoft Excel 2016, it was found that the significance value for the data of the four items tested both with the Kolmogorov-Smirnov and Shapiro-Wilk tests was <0.05 so it can be said that the data in the pilot study this is not normally distributed. Therefore, in the validity test, a test belonging to the non-parametric statistical method, namely Spearman, will be used. The results of validity testing using pilot study data with the Spearman method assisted by IBM SPSS Statistic 26 software, the item-to-total value for all questionnaire items >0,334 (r table for N = 35 at a significant level of 5 %), indicates that all items used in the questionnaire can produce valid data and it can also be concluded that the questionnaire items can measure the concept of Customer Response (CR) appropriately. In reliability testing, the coefficient value of Cronbach's Alpha for all questionnaire items is > 0,6, which indicates that the data for each item is reliable, so it can be concluded that all items in the questionnaire have good stability and consistency to measure a concept. Referring to the results of the validity and reliability of the questionnaire items, it can be said that the questionnaire is feasible to use.

3.2.6. Determination of The Number of Samples

In this research survey, the survey population is all household electricity customers with a minimum subscription power of 1300 VA who are domiciled on the islands of Java/Bali. With a total population of 9.061.291 customers who subscribe to at least 1300 VA in Java-Bali, the minimum number of survey samples can be determined by calculating using the Slovin formula as below (margin of error = 0.05).

$$SampleSize=\frac{9061291}{1+\left(9061291\times {0,05}^2\right)}=399customers$$

From the calculations that have been carried out using the Slovin formula, the minimum number of samples needed in this survey is 399 samples.

3.2.7. Data Collection

The survey to collect customer responses to the Time of Use (ToU) tariff scheme in this study was conducted using a questionnaire that was disseminated indirectly or online via Google form. Customers who can provide responses or participate as survey respondents are those who meet the following four criteria, (1) domiciled on the island of Java/Bali, (2) live in a household, (3) subscribe to electricity for household groups of at least 1300 VA, and (4) know the daily use of electricity at home. Customers who do not meet these criteria will be eliminated and not taken into account in the survey. From the distribution of the questionnaires that have been carried out, the number of samples obtained is 513. However, as many as 62 samples cannot be used in this study because they do not meet the criteria of the respondents that have been determined, so they need to be eliminated. Therefore, the number of samples that will be analyzed for responses regarding the ToU tariff scheme is 451, which by calculation can be said to have met the minimum sample size requirements in this study.

3.2.8. Data Preparation

All the data that has been obtained will be analyzed later to be able to answer the objectives of this study. However, before the data is analyzed, as a first step, it is necessary to do some data testing to find out how the quality is, namely through normality, validity, and reliability testing which has the same principle in testing the normality, validity, and reliability of the questionnaire items. which has been done before. The data from the distribution of the questionnaires to be tested for quality is the data used to measure customer views, namely the data on the questionnaire items CR1 to CR4.

Using the same steps as testing the validity and reliability of questionnaire items using pilot study data, at this stage, all the data that has been obtained will first be tested for normality. By looking at the significance values generated in the Kolmogorov-Smirnov test and the Shapiro-Wilk test, it can be concluded that the overall data obtained are not normally distributed (significance = 0,000). This can be due to the research data in the form of a Likert scale which is classified as ordinal data. Therefore, furthermore, testing the validity of the data must be carried out using a non-parametric method, namely Spearman. From the results of validity testing with the Spearman method, the item-to-total value of all data for all items is > 0,093 (r table for N = 451 at a significant level of 5 %). While in reliability testing, the coefficient value of Cronbach's Alpha for all items using the entire questionnaire data is > 0,6. Therefore, it can be said that all the data on the questionnaire items are valid (the data obtained are in accordance with the data that occurred in the subjects studied) and reliable (the data obtained are consistent). In other words, the overall data obtained is suitable for use and analysis.

3.2.9. Data Processing and Analysis

At this stage, processing and analysis of the overall data that has been obtained from the questionnaire distribution stage will be carried out. Processing and analysis are carried out so that it can answer the research objectives that have been designed previously.

The responses of all customers from the questionnaire will be processed corresponding to their statement item. The statement number 1 to 3 will be calculated by multiplying the number of responses on each response scale by the weight of each response scale to get the total score. The response strongly disagrees weights 1, disagrees has a weight of 2, doubt has a weight of 3, agrees has a weight of 4, and strongly agrees has a weight of 5 [15].

The total score of each item will then be transformed into a percentage (index %) through the following equation [15].

$$\mathbf{index}\%=\frac{\mathbf{total}\mathbf{score}}{\mathbf{maximum}\mathbf{total}\mathbf{score}}\times \mathbf{100}$$ (1)

After transforming the value, the % index value will later be used to make a final decision regarding the responses/responses of all respondents based on the % index interval classification which is formed based on the number of Likert scales or responses used. The % index interval is 100 % divided by 5 (the number of Likert scale used) which is 20 % [13]. Classification of index % interval more clearly can be seen in Table 2.

Table 2.

Index %Classification.

Index % Interval	Classification
0 % - 19,99 %	Strongly Disagree
20 % - 39,99 %	Disagree
40 % - 59,99 %	Hesitant
60 % - 79,99 %	Agree
80 %–100 %	Strongly Agree

3.2.10. Conclusion

In conclusion, it will be known how consumers respond to the ToU tariff scheme, activities that consumers think it is possible to shift, as well as forecasts for potential load reductions due to the implementation of the ToU scheme.

4. Results and discussions

4.1. Respondents Profile

Of the 451 respondents who participated in this study survey, the majority of respondents were male (59 %), aged 31–41 years (47 %), and domiciled in Java (91 %). If classified, most of the respondents' areas of residence are in urban areas (82 %). The respondent's employment status is dominated by private employees (34 %), with the latest education at Bachelor's degree (46 %). The respondents also having an income in Indonesian Rupiah (IDR), between IDR 5.000.001 and IDR 10.000.000 per month (29 %). The number of family members who live together in one house ranges from 4 to 6 people (61 %) and the type of house occupied by the majority is private/own house (86 %). The majority of respondents subscribe to electricity for households with a PLN subscription power of 1300 VA (50 %) and the average monthly electricity bill is between IDR 250.001 and IDR 500.000 (45 %).

4.2. Customer response to time of use(ToU) Tariff scheme

In the research questionnaire, an explanation is given to consumers regarding the provisions of the implementation of the Time of Use (ToU) tariff scheme that has been planned by PLN, starting from the classification of off-peak and on-peak periods and the amount of bills that may arise using this scheme, as shown in Fig. 5, Fig. 6 irespectively.

Fig. 5.

Time Period Classification Graph for Weekday and Weekend.

Fig. 6.

Comparison of Electricity Bills That Can Appear with FR and ToU Tariff Schemes.

Classification results about the time period that obtained using the K-Means Clustering method are validated using Expert Judgement Analysis. The result can be concluded that the load in the off-peak period and mid-peak for the weekend condition tend to make no difference. In order to avoid bias, time period

classifications, that is off-peak and on-peak. The mid-peak period will be classified as an off-peak period. While the weekday period is valid and does not need to be changed. The classification period for the weekday and weekend after Expert Judgement Analysis is shown in Table 3.

Table 3.

Classification Period for Weekday And Weekend Condition Based on Expert Judgement Analysis

Classification	Condition
	Weekday	Weekend
Off-Peak	1 - 7 a,m. WIB	12 a.m.–5 p.m.; 10 p.m.–11 p.m. WIB
Mid-Peak	12 a.m.; 8 a.m.; 10 p.m.–11 p.m. WIB	–
On-Peak	9 a.m.–9 p.m. WIB	6 p.m.–9 p.m. WIB

Fig. 6 shows the graph of comparison of electricity bill that can appear with flat rate and ToU tariff schemes. The graph can be changed subject to the amount of saving or an increase of the bill can be done depending to the number of the shifted electricity consumption. The implementation of the ToU scheme tariff will appear the 3 possibilities related to the monthly electricity bill.

These 3 possibilities will appear is.

• 1.The bill will be the same as the flat rate tariff, if the consumers do not shift the electricity consumption. (Quo status)
• 2.The bill will be increased if the consumers shift the electricity consumption partially from the off-peak period to the on-peak period.
• 3.The bill will decrease if the consumers shift their electricity consumption partially from the on-peak period to the off-peak period.

Referring to the explanation of the provisions for the implementation of the ToU scheme, the responses or responses of all customers who became respondents in this study to each statement item were shown in Fig. 7, Fig. 8, Fig. 9.

Fig. 7.

Customer Response to Statement Item 1 Regarding the ToU Tariff Schemes Is Good Idea.

Fig. 8.

Customer Response to Statement Item 2 Regarding the Savings after Using ToU Tariff Schemes.

Fig. 9.

Customer Response to Statement Item 3 Regarding the Ease of Accommodation of the ToU Tariff Scheme.

The results of calculating the total score for each statement item on each response scale can be seen in Table 4.

Table 4.

TotalScore forEachStatementItem.

Response Scale	Item CR1		Item CR2		Item CR3
	Count	Total Score	Count	Total Score	Count	Total Score
Strongly Agree (Weight: 5)	115	575	120	600	98	490
Agree (Weight: 4)	158	632	137	548	122	488
Hesitant (Weight: 3)	96	288	107	321	134	402
Disagree (Weight: 2)	41	82	52	104	58	116
Strongly Disagree (Weight: 1)	41	41	35	35	39	39
Total score for each item	1618		1608		1535

Where the maximum total score that may be obtained in this study is 2255 (highest weight 5 multiplied by the number of respondents 451). By using Equation (1)., the results of the transformation of the total score into the index value % are as follow s Table 5.

Table 5.

Index Value % forEach Statement Item.

Statement Item Code	Index% Value
CR1	72 %
CR2	71 %
CR3	68 %

Based on the % index value in the table above, it can be seen that for statement item 1, an index % value of 72 % is generated, which when referring to T able III, this figure falls into the agreed category. Therefore, it can be concluded that all customers who participated in the study (451 customers) agreed that the planned implementation of the Time of Use (ToU) electricity tariff scheme, especially for the household group, was a good idea. While in the item 2 statement, the index value % is 71 %. So it can also be concluded that all customers agree that the implementation of the ToU electricity tariff scheme can help them to save on their expenses or electricity bills. In statement item 3, the index value % generated is 68 %. Based on the % index value, it can also be concluded that all customers agree that the process of adjusting to the ToU electricity tariff scheme is an easy thing. However, if you look at it, the index % value generated for the item of ease of adjustment is still lower than the other two items. This is quite reasonable, because the ToU electricity tariff scheme among households is indeed a new thing, so customers still have little doubts about the ease of adjusting later.

In addition to the three statement items that have been processed and analyzed, there is one more item that is also used to find out how the customer responds or responds to the ToU electricity tariff scheme. The item is a question item regarding the customer's willingness to join the ToU electricity tariff scheme if this scheme is really to be implemented by PLN. In the distributed questionnaire, two alternative answers can be chosen by the customer in answering this question item, namely Yes or No. From the results of the questionnaire distribution, 63 % of respondents responded that they were willing to join the ToU electricity tariff scheme, while the remaining 37 % responded that they were not willing to join the tariff scheme. The customer's decision regarding their willingness to join the ToU tariff is based on several reasons as shown in Fig. 10.

Fig. 10.

Some of the Customer's Reasons Regarding Willingness to Join the ToU Tariff.

Description:

A = The existence of customer sensitivity to prices which can lead to bill savings.

B = The ToU scheme is more fair, transparent, and makes it easier for customers and PLN.

C = Wants to contribute to PLN's plan to implement go-green.

D = Want to add insight and try the electricity tariff scheme which is still relatively new.

Of the total 286 customers who are willing to join the ToU tariff scheme, 70 % of them stated that price sensitivity was the main reason for their decision. Where electricity tariffs tend to be cheaper in the off-peak period, customers perceive that this scheme is more attractive and profitable for customers, although electricity rates also tend to be more expensive in the on-peak period, but they are more sensitive to focus on higher tariffs. inexpensive. Therefore, customers who have the flexibility to set the time of electricity usage will try to move activities to off-peak periods and reduce electricity usage in on-peak periods to save their electricity bills. As many as 18 % of customers think that the ToU scheme is more fair, transparent, and easier for them and PLN because the impact of implementing the ToU scheme can also be seen directly. On the customer side, it will be easier for them because the bills paid are according to their usage, customers can find out their electricity usage in detail, and customers can also predict the electricity bills they have to pay. In addition, for customers, the ToU scheme which provides cheaper electricity tariffs in the off-peak period is a form of incentive given by PLN to customers for their contribution to the implementation of go-green, which is PLN's goal for the plan of implementing the ToU scheme. Meanwhile, on the PLN side, it will be easier for them because the electricity burden borne by them can be more evenly distributed between off-peak and on-peak periods from the implementation of the ToU scheme. Furthermore, as many as 17 % of customers also reasoned that they wanted to contribute to PLN's plan in an effort to implement go-green or save electricity use by being more efficient in using electricity on a daily basis. Another reason was given by 6 % of other customers, who they were willing to join the ToU scheme because they wanted to increase their knowledge and try the ToU electricity tariff scheme which is still relatively new.

In spite of that, several reasons underlying the decision of customers who are not willing to join the ToU electricity tariff scheme are shown in Fig. 11.

Fig. 11.

Some of the Customer's Reasons for Unwillingness to Join ToU Tariffs.

Description:

A = The implementation of the ToU scheme seems burdensome and manages customers.

B = There are still doubts regarding the impact, benefits, and technicalities of implementing the ToU scheme and it still takes time to adjust.

Of the total 165 customers who are not willing to join the ToU tariff scheme, 69 % of them feel that the implementation of the ToU scheme seems burdensome and manages customers. This is because customers find it difficult to move their daily activities just like that. This happens because in some cases in the household, activities that use electricity are carried out by household assistants and housewives. This makes them feel that it is not possible or less flexible to remember off-peak times and postpone activities to be carried out during the off-peak period even though the activities that need to be done are also quite a lot and sometimes uncertain because it suits their needs. Moreover, the off-peak period tends to occur at a time that is quite early in the morning and late at night, so that according to customers the implementation of this scheme will disrupt their time to rest. In addition, the existence of a pandemic condition has also made some people spend a lot of time at home because of the need to work from home or online school. According to customers, their electricity usage has also become less and requires more effort to provide education to other family members so that customers can really feel the impact and benefits of this scheme. With all the effort they need to spend, they feel unfair because they also still have to pay bills. Some customers also feel that the current tariff is already cheap and still object if there is an increase in tariffs, even if only during the on-peak period.

Furthermore, as many as 43 % of customers still have doubts about the impact, benefits, and technicalities of implementing the ToU scheme and need time to adjust to the scheme. Customers are still not sure whether this scheme can really have an impact or be effective in saving electricity and feel that the implementation of this scheme makes electricity use complex. Besides, in several cases of customers, their electricity usage between off-peak and on-peak periods tends to be almost the same and according to them the implementation of the ToU scheme will give the same results as the flat rate scheme or in the worst case their bills may tend to increase due to the usage cycle. electricity that does not change even increases due to daily needs. According to customers, the ToU scheme could cause PLN to lose demand in the on-peak period and instead accumulate it in the off-peak period. These doubts occur because customers have not been able to see the impact and benefits of implementing the ToU scheme directly, so it is also necessary to test this scheme first before it is actually set as an electricity tariff scheme in Indonesia. During the trial period, PLN can also give customers a discount on electricity bills so that customers can adjust to the ToU scheme to the fullest. Customers also still feel that they do not fully understand the technical implementation of the ToU scheme. Because of this, they are afraid that the electricity tariffs for the off-peak and on-peak periods will have a very wide range and it is possible for the off-peak and on-peak periods to fluctuate because PLN sees trends in customer electricity usage. Therefore, customers still expect further socialization provided by PLN to customers. The existence of these doubts makes customers prefer if the electricity tariff applied is the same (flat rate). For customers, the ToU tariff scheme is considered more suitable for business and industrial customers. If later this scheme is implemented, then PLN needs to prepare carefully related reports related to customer electricity usage in detail and tools to monitor electricity use in real-time so that the benefits of implementing the scheme can be felt.

In the context of efficiency and effectiveness in the use of electricity bills, there are also several suggestions given by customers to be able to achieve this. According to customers, PLN can look for more efficient production and operational centers so that the tariffs offered are still easy and cheap and start to socialize with customers related to the use of solar cells. Regarding the difference in electricity rates offered, PLN can also offer different tariffs based on the purpose of using electricity (for productivity or not) or based on the subscription power of its customers.

4.3. Potential transfer of electricity usage by customers

The implementation of the Time of Use (ToU) electricity tariff scheme will indirectly stimulate customers to move some of the activities in their households that are often carried out during the on-peak period to the off-peak period. In this study, it is also intended to find out what activities according to customers are possible to be shifted to the off-peak period. In the questionnaire, several activities in the household have been provided that customers can choose if they feel that these activities are possible to be shifted. The results of customer responses are shown in Fig. 12.

Fig. 12.

Customer Response to Activities Most Likely to Be Shifted.

Based on the results of the customer responses above, it can be seen that the two activities that customers think are most likely to be shifted if the ToU electricity tariff scheme is implemented are washing clothes with a washing machine where there are a total of 209 customers (73 %) who choose this activity and also ironing clothes with a total of 209 customers (73 %) customers who chose as many as 173 customers (60 %). This is quite reasonable because in household activities these two activities are usually carried out only once a day and the schedule or time period for these activities is on average always the same every day so that it will be easier for customers to move these activities to another time period. whose load is lighter.

However, two activities are least chosen by customers because they feel that these activities are difficult to move. The two activities are turning on the electric stove (12 % of customers) and also turning on the water heater (19 % of customers). This can happen because the activity of turning on the stove in one day can be done many times and at an uncertain time every day according to need. The same thing also happens to the activity of turning on the water heater, so customers find it more difficult to move the activity, especially during the off-peak period.

By moving some activities in the household to an off-peak period, customers will be able to experience savings, especially on their electricity bills. In this study, information was also explored regarding the amount of bill savings desired by customers from the implementation of the ToU electricity tariff scheme. Of the 63 % of customers (286 customers out of a total of 451 customers) who are willing to join this tariff scheme, the amount of savings they want on monthly electricity bills from moving their activities due to the implementation of the ToU scheme looks quite varied. If calculated as a whole, the average nominal amount of electricity bill savings they want per month is IDR142.731,-. It is calculated from the amount of monthly savings that can be done based on the number of customer activities that want to be shifted using the ToU Tariff Scheme. If it is calculated based on the size of their current electricity bill, the desired percentage reduction in bills is also quite diverse, from the smallest is 3 % to the largest is 100 % of the current total bill. In other words, customers want them to be free from electricity costs. On average, the percentage reduction in electricity bills is 26 % from their current desired bill. On the other hand, if you look at the ToU tariff designs that have been formed previously, the maximum percentage reduction that can be accommodated by PLN is only 11,5 %. This is can be explained as the maximum of billing reduction compared to BPP that be gathered from PLN. It can be seen that customers are quite hopeful that the strategy of setting electricity tariffs can reduce their electricity bills as effectively and efficiently as possible.

4.4. Potential for network load reduction due to time of use (ToU) Tariff scheme

The process of predicting the daily network load curve for the implementation of the Time of Use (ToU) tariff scheme will refer to information about the activities that customers are willing to shift, which is listed in sub-chapter C. From the activities that have been selected by the customer, later I will try to calculate the total daily load first using the Household Energy Calculator provided online by the Directorate General of New, Renewable Energy and Energy Conservation (http://kalkulator.ebtke.esdm.go.id). The calculation of the total daily load can be briefly explained by multiplying the average usage time per day and the average equipment power. Before calculating the daily load for each activity or use of equipment, it is necessary to first specify the specifications of each equipment. The specifications for each equipment will refer to research [11], where the specifications used in calculating the load are the average specifications of the equipment owned by the population in Indonesia. The specifications for each equipment that will be used as material for calculating the load are listed in Table 6 below.

Table 6.

Specifications ofEach Household Appliances.

Equipment Name	Average Usage Time/day	Average Equipment Power	Equipment Type
Washing machine	2,1 h	<250 Watt	–
Iron	1,8 h	300-350 Watt	Regular (plate)
TV	6,7 h	69 Watt	Tube TV
Air Conditioning	8 h	554 Watt	Type: Standard ½ PK Temperature: 20 °C
Water Pump	2,8 h	417 Watt	–
Rice Cooker	6,4 h	305 Watt	–
Dispenser	24 h	232 Watt	–
Fan	6,4 h	53 Watt	Type: Standing Fan
Blender	1,5 h	215 Watt	–
Water Heater	1,9 h	680 Watt	–
Electric Stove	8,3 h	800 Watt	–

By inputting the specifications for each of the above equipment in the Household Energy Calculator, we get the total energy consumption of each activity per month (in units of kWh). From the total consumption per month obtained, it will then be transformed into units of days and converted into load (in units of kW). The results of the total energy consumption of each activity and its load are listed in Table 7 below.

Table 7.

Total Energy Consumption of Each Activity andIts Load.

Activity	Energy Consumption/month (kWh)	Energy Consumption/day (kWh)	Load/day (kW)
Washing clothes with a washing machine	15,75	0,53	0,25
Ironing clothes	16,20	0,54	0,26
Turn on the TV	13,87	0,46	0,22
Turn on the air conditioner	144,00	4,80	2,29
Turn on the water pump	35,03	1,17	0,56
Cook & warm rice with a rice cooker	36,29	1,21	0,58
Turn on the dispenser	167,04	5,57	2,65
Turn on the fan	10,18	0,34	0,16
Turn on the blender	9,68	0,32	0,15
Turn on the water heater	38,76	1,29	0,62
Turn on the electric stove	199,20	6,64	3,16

Based on the results of the load calculation for each activity that can be moved to the off-peak period, then we will try to find the total load value for each activity by multiplying it by the number of responses (the number of customers who chose each activity in question). The results of the complete calculation of the total load for each activity can be seen in Table 8.

Table 8.

Calculation Results of Total Activity Load Willing to be Moved.

Activity	Number of Customers (person)	Load/day (kW)	Total Load/day (kW)
Washing clothes with a washing machine	209	0,25	52,25
Ironing clothes	173	0,26	44,49
Turn on the TV	150	0,22	33,02
Turn on the air conditioner	137	2,29	313,14
Turn on the water pump	119	0,56	66,17
Cook & warm rice with a rice cooker	101	0,58	58,18
Turn on the dispenser	74	2,65	196,21
Turn on the fan	72	0,16	11,63
Turn on the blender	58	0,15	8,91
Turn on the water heater	54	0,62	33,22
Turn on the electric stove	34	3,16	107,50
Total			924,73

From the calculation results above, it can be seen that the total load of all activities that have the potential to be moved is 924,73 kW. This total load is the total load of all activities and customers. Therefore, we will try to find the average total load value for each customer in all activities. The total load value of 924,73 kW will be divided by the number of samples in the study who are willing to join the ToU tariff scheme, which is 286 customers. Based on the calculations, the average total load for each customer who has the potential to move to the off-peak period is 3,23 kW. Furthermore, the total load of each customer can be projected into the total load of the entire population, in this case, the households in Java-Bali. By using the percentage of customers in the study who are willing to join the ToU scheme, which is 63 %, the number of customers in the population who are willing to join the ToU scheme will be sought first. By multiplying the percentage of 63 % by the number of the research population, which is 9.061.291 customers, it is found that the number of customers in the population who are willing to join the ToU scheme is 5.746.185 customers. The total load of the entire population that has the potential to move is obtained by multiplying the total load of each customer by the number of customers in the population who are willing to join the ToU scheme. So that the result is that the daily load of all customers in the population that have the potential to move to the off-peak period is 18.579.237,52 kW or 18.579,24 MW. A comparison of a customer's daily load between the existing condition and the condition after moving several activities to the off-peak period is listed in Table 9 and Fig. 13. Where the existing daily load of a customer is obtained from the average daily load of the entire household class electricity customer. domiciled in Java-Bali and the daily load data for each customer is obtained from secondary data provided by PLN.

Table 9.

Comparison ofLoads perCustomer inExistingConditions andAfterLoadShifts.

Period	Load
	Existing	After moving some activities
Off-peak	3437 kW	6667 kW
On-peak	4684 kW	1454 kW

Fig. 13.

Comparison Graph of Existing Load with Load After a Movement Activity.

From the table and graphs listed above, it can be seen that the implementation of the ToU tariff scheme will indirectly stimulate customers to move their daily activities which are usually carried out during the on-peak period to the off-peak period. From the activities that customers think have the potential to be moved, the total daily load value of each customer which also has the potential to move to the off-peak period is 3,23 kW. The existence of these activity shifts makes the load in the on-peak period tend to decrease and thus will have an impact on reducing customer electricity bills and plant operational costs that must be incurred by PLN.

5. Conclusions

Based on the results and discussions that have been described, the conclusions that can be drawn are the customers’ response exhibits a promising outcome. 63 % of the survey participants expect to opt into the ToU scheme and the reason for this is their sensitivity to a bargain price of the electricity rate. While the rest did not want to join because of the burdensome of shifting their daily activity during the on-peak period. Therefore, the implementation of the ToU scheme should be gradually using the pilot project first before overall implementation.

From the survey results, it can also be inferred that household energy-related activities that customers are quite willing to shift are washing and ironing clothes where a shift in these activities would lead to potential savings in the monthly electricity bills of 26 % off the regular price or equal to Rp142.731. The customers have a positive outlook towards the ToU tariff scheme as an effective and efficient approach to reducing their electricity bills. Nevertheless, it is recommended to further study about consumer responses to ToU tariff with the information about the possible range of saving that can be fulfilled by PLN as an electricity supply utility company.

Using the quantitative method in this study, it can be calculated that the total potential daily load reduction of energy consumption with the ToU tariff scheme could reach 3,23 kW for each customer. If it is calculated for the entire population, the total daily load that has the potential to move to the off-peak period is 18.579,24 MW.

The findings conclude that customers have presented an encouraging response to future implementation of ToU and express their thoughts ToU seems to bring potential benefits of saving electricity bills and looks as if it is fairly easy for customers to adapt to shifting their energy usage from peak times to off-peak times.

Contributing authors consent

All authors have confirmed their substantial contributions to this work and are already in firm agreement with the scientific publication of this paper.

Ethical statement

Identifying information of the respondents, such as names, genders, or specific locations, have been anonymized to ensure participant safety and privacy while informed consent was obtained verbally before participation.

Data availability

Supporting research data set associated with the paper is not available for access publicly since a legally binding contract with the research institution has inhibited the authors from doing so due to sensitive information contained in the data set. Nonetheless data will be made available on request.

CRediT authorship contribution statement

Hari Agung Yuniarto: Writing – original draft, Writing – review & editing, Project administration, Conceptualization, Visualization, Validation, Supervision, Software, Resources, Methodology, Investigation, Formal analysis, Data curation. Nur Mayke Eka Normasari: Methodology, Investigation, Formal analysis, Data curation. Irene Clarisa Gunawan: Validation, Resources, Investigation, Formal analysis. Sella Friscilla Silalahi: Writing – review & editing, Validation, Formal analysis. Deendarlianto: Validation, Resources, Methodology, Investigation, Formal analysis. Indra Ardhanayudha Aditya: Project administration, Validation, Resources, Methodology, Investigation, Funding acquisition, Formal analysis. Arionmaro Asi Simaremare: Software, Investigation, Funding acquisition, Formal analysis, Data curation. Fajar Nurrohman Haryadi: Validation, Software, Resources, Investigation, Funding acquisition, Formal analysis, Data curation.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Biographies

Ir. Hari Agung Yuniarto, ST, MSc, PhD, IPU, ASEAN Eng took his PhD degree in Industrial Manufacturing Engineering and Management from The University of Manchester as well as his MSc degree in Advanced Manufacturing Technology and Systems Management from UMIST (University of Manchester Institute of Science and Technology) both are in the UK, whereas his first degree - ST, equivalent to BEng - in Mechanical Engineering was received from UGM (Universitas Gadjah Mada) in Indonesia. He also has been awarded a professional engineer licence from The AFEO Governing Board (ASEAN Federation of Engineering Organisations) - ASEAN Eng - certifying compliance with practicing engineering within ASEAN countries. He pursued a career in 3 world-class industries, i.e. SCHLUMBERGER Oil Field Services, TOYOTA Manufacturing Corp and FREEPORT-McMoRan Inc before he has eventually served as a senior lecturer of Industrial Engineering in Faculty of Engineering UGM since 2002. His expertise and research interests include Reliable System Quality, System Reliability Engineering, Quality Management & Standardisation, Enterprise Risk Management, Maintenance Engineering, Life-cycle Asset Management, Through-life Engineering, Quality Engineering, and Safety Engineering. He conducts close collaboration researches with universities as well as industries around the world and has published considerable amount of papers in the reputable journals and proceedings (ORCID ID: 0000-0002-2157-1334 ; Web of Science ID: AAQ-8569-2020 ; Publons ID: 3705386 ; Scopus ID: 57192998487).

Ir. Nur Mayke Eka Normasari, ST, Meng, PhD, IPM took her PhD degree in Industrial Management from National Taiwan University of Science and Technology in Taiwan. Her master's degree – MEng is in Industrial and System Engineering from Asian Institute of Technology in Thailand, whereas his first degree – ST, equivalent to BEng – in Industrial Engineering was receive from Universitas Gadjah Mada in Indonesia. She is currently a senior lecturer in Industrial Engineering, Universitas Gadjah Mada since 2007. Her expertise and research interest include, but are not limited to, system modeling, routing problem, supply chain network design, healthcare supply chain, green supply chain, optimization, simulation, pricing strategy, revenue management, customer behavior, and metaheuristic algorithm.

Irene Clarisa Gunawan, ST received the B.S degree in industrial engineering from Gadjah Mada University, Yogyakarta, Indonesia. She is currently a research assistant at the Center for Energy Studies, Gadjah Mada University. She has experience in conducting research related to consumer response to the application of dynamic pricing, for example in cinemas. Her research interests include, but are not limited to, customer behavior, dynamic pricing, and revenue management in service companies.

Sella Friscilla Silalahi, ST received a bachelor's degree in Industrial Engineering from Universitas Atma Jaya, Yogyakarta, Indonesia. She is currently studying for a Master Degree in Industrial Engineering at Universitas Gadjah Mada, Yogyakarta, Indonesia. Her research interest include, but are not limited to data mining for healthcare and politics, system modeling, and machine learning with application in pattern recognition.

Prof. Dr. Ir. Deendarlianto, ST, MEng received the B.S. degree in Mechanical Engineering from Universitas Gadjah Mada, Yogyakarta, Indonesia, and the Ph.D. degree also in Mechanical Engineering from the University of Tokushima, Japan. Next, He did post-doctoral fellow at the Fluid Dynamics Institute, Helmholtz-Zentrum Dresden-Rossendorf, Germany from 2009 to 2011. He is currently a Professor with the Department of Mechanical and Industrial Engineering, Universitas Gadjah Mada. His research interest includes, but are limited to, interfacial behavior in multiphase flow system and the development of the analytical model on the energy policy options.

Indra Ardhanayudha Aditya, ST, MT received a master's degree in electricity and energy management from the University of Indonesia. He has a bachelor degree in Engineering Physics from the Institut Teknologi Bandung (ITB). He currently works at PT PLN Research Institute as a researcher with research interests ranging from renewable energy, energy economics and energy policy.

Arionmaro Asi Simaremare, ST, MEngSc after finishing his bachelor degree in power engineering from the Institute Teknologi Bandung (ITB) in 2010, He started his career as a power distribution engineer in PT PLN (Persero), a state owned electric company in Indonesia. After more than 5 years in the power distribution sector and learning how power distribution operation and maintenance works, including customer complaint management, He decided to pursue higher education and finished his master of engineering science degree in renewable energy engineering from The University of New South Wales. Since 2018, He started his career at the PLN Research Institute as a researcher with research interests ranging from renewable energy and energy economics to energy policy. ORCID ID: https://orcid.org/0000-0003-3694-6482 Scopus Author ID: 57208863948.

Fajar Nurrohman Haryadi, ST, MT currently works at PLN (an electric state-owned electricity company in Indonesia) and has been researching multidisciplinary analysis, including market research of Rooftop PV customer and Electric Vehicle customer in Indonesia. He has strong interest in Customer Behavior related to Technology Adoption (Electric Vehicle, Solar Rooftop PV).

APPENDIX.

Appendix 1. . Normality Test for Pilot Study Questionnaire

Item Code	Kolmogorov-Smirnov Sig.	Shapiro-Wilk Sig.	Notes
CR1	0,003	0,002	Not normal
CR2	0,000	0,001	Not normal
CR3	0,007	0,002	Not normal
CR4	0,000	0,000	Not normal

Appendix 2. . Validity Test for Pilot Study Questionnaire

Item Code	Correlation Value of Item to Total	R-Table	Notes
CR1	0,973	0,334	Valid
CR2	0,948	0,334	Valid
CR3	0,920	0,334	Valid
CR4	0,837	0,334	Valid

Appendix 3. . Reliability Test for Pilot Study Questionnaire

Item Code	Cronbach's Alpha Coefficient	Notes
CR1	0,769	Reliable
CR2	0,775	Reliable
CR3	0,784	Reliable
CR4	0,852	Reliable

Appendix 4. . Normality Test for Overall Data

Item Code	Kolmogorov-Smirnov Sig.	Shapiro-Wilk Sig.	Notes
CR1	0,000	0,000	Not normal
CR2	0,000	0,000	Not normal
CR3	0,000	0,000	Not normal
CR4	0,000	0,000	Not normal

Appendix 5. . Validity Test for Overall Data

Item Code	Correlation Value of Item to Total	R-Table	Notes
CR1	0,934	0,093	Valid
CR2	0,933	0,093	Valid
CR3	0,892	0,093	Valid
CR4	0,775	0,093	Valid

Appendix 6. . Reliability Test for Overall Data

Item Code	Cronbach's Alpha Coefficient	Notes
CR1	0,811	Reliable
CR2	0,818	Reliable
CR3	0,857	Reliable
CR4	0,922	Reliable

Appendix 7. . Respondent Profile

Respondent Demographic Item	n	Percentage
Gender
Male	265	59 %
Female	186	41 %
Age (year)
20–30	75	17 %
31–41	213	47 %
42–52	99	22 %
53–63	53	12 %
≥64	11	2 %
Region
Java Island	412	91 %
Bali Island	39	9 %
Residential Area
Perkotaan	369	82 %
Pedesaan	82	18 %
Occupation
Private Employee	151	34 %
Civil Servant	95	21 %
BUMN Employee	53	12 %
Entrepreneur	50	11 %
Housewife	49	11 %
Pensiunan	18	4 %
Lecturer	9	2 %
College Student	6	1 %
Others (Labor, Contract Employee, Doctor, Non-civil Employee, Driver online, freelancer, etc)	20	4 %
Education
Elementary School	2	0,4 %
Middle School	2	0,4 %
High School	37	8 %
Diploma	36	8 %
Bachelor	207	46 %
Master	146	32 %
Doctoral	21	5 %
Income per Month
< IDR 2.500.000	44	10 %
IDR2.500.001 – IDR5.000.000	119	26 %
IDR5.000.001 – IDR10.000.000	129	29 %
IDR10.000.001 – IDR15.000.000	74	16 %
IDR15.000.001 – IDR20.000.000	30	7 %
> IDR20.000.000	55	12 %
Number of family members living together (person)
1–3	158	35 %
4–6	275	61 %
7–9	17	4 %
≥10	1	0,002 %
Type of house lived in
Ownership	390	86 %
Rent	58	13 %
Apartment	3	1 %
PLN Subscription power
1300 VA	224	50 %
2200 VA	144	32 %
3500 VA	39	9 %
4400 VA	16	3 %
5500 VA	24	5 %
≥6600 VA	4	1 %
Average of electricity monthly bill
≤ IDR250.000	82	18 %
IDR250.001 – IDR500.000	203	45 %
IDR500.001 – IDR750.000	79	18 %
IDR750.001 – IDR1.000.000	46	10 %
> IDR1.000.000	41	9 %

Appendix 8. . Bill Electricity Savings from Each Customer Subscription Power

PLN Subscription power	Desired average monthly bill saving	Desired percentage of monthly bill saving
1300 VA	IDR113.673,-	29 %
2200 VA	IDR154.273,-	25 %
3500 VA	IDR158.810,-	18 %
4400 VA	IDR278.125,-	23 %
5500 VA	IDR217.105,-	21 %
≥6600 VA	IDR283.333,-	23 %
Overall	IDR142.731,-	26 %

Appendix A. Supplementary data

The following is/are the supplementary data to this article.