Peak Load Habits for Sale? Soft Load Control and Consumer Preferences on the Electricity Market.

• Author: Broberg, Thomas

1. INTRODUCTION

   In Europe, and elsewhere, electricity markets are changing and the transformation is characterized by three key factors: (i) deregulation of electricity markets, (ii) new technologies with respect to generation, distribution and use, and (iii) substantial changes in the production mix as a result of energy and climate policy as well as changes in relative production costs for different technologies. These factors in combination with a rigid demand side characterized by daily and seasonal consumption patterns, and consumers that are not exposed to the time of use marginal generation cost, have raised concerns about security of supply. Because of this concern, there is an ongoing discussion of whether energy-only markets, which are the most common market design, should be complemented with a capacity mechanism to ensure enough generation in peak periods (Joskow, 2008a, 2008b; Eurelectric, 2015; Newbery, 2016). Related to this is also the discussion of demand management and demand flexibility, which can be seen as part of such a capacity mechanism (Strbac, 2008; Broberg and Persson, 2016).

   The Swedish electricity generation structure with about 85% hydro- and nuclear power contributes to a relatively flexible and robust power system with modest climate impacts. Nevertheless, the current Swedish interest in demand flexibility is driven by future challenges mainly related to further integration of European electricity markets and the Swedish target of 100% renewable electricity production by 2040. The Swedish aim to increase the share of renewable electricity inevitably leads to a large share of intermittent sources such as solar and wind, while the market integration may lead to less control of domestically produced electricity.

   In line with these observations, the main objective of this study is to estimate Swedish household's willingness to accept load restrictions for electricity use during peak hours. Two types of load restrictions, or control, are considered: "soft control," which refers to a temporary restriction in the maximum possible load (in watt) for high-power appliances and installations; and "hard control," which refers to a complete loss of power for 30 minutes during peak time. We use a stated preference approach asking electricity consumers to choose between hypothetical electricity (delivery) contracts with different attributes concerning maximum load during a specific time of the day. Each hypothetical contract is appended with a monetary compensation, enabling an estimate of the monetary compensation required for load restrictions. The resulting monetary compensation for the "soft" load control can be interpreted as the value of potential lost load (VoPLL), whereas the monetary compensation for complete loss of load corresponds to the value of lost load (VoLL).

   We contribute to the previous literature and the policy discussion in two ways. First, the results of the analysis elicit consumer preferences for demand flexibility, and hence the potential for demand side management (DSM). Second, the analysis gives estimates of VoPLL and VoLL, which is of paramount importance if/when explicit capacity mechanisms are considered. Any capacity mechanism should be designed such that the optimal level of supply security is reached cost-effectively, for which information about the value of lost load is needed (Ovaere et.al. 2019). In addition to these two contributions, we explore households' power consumption for home appliances in the peak hours. A better knowledge of household habits and consumption patterns is important not only for determining the potential for demand response, but also for determining the costs in terms of utility losses associated with curtailment actions. Importantly, this analysis is based on respondents' reported consumption patterns.

   The stated preference approach is commonly used in situations where market values do not exist (Johnston et al., 2017). Several applications in the previous literature relate to assessment of the utility loss following a power outage, i.e. estimation of VoLL (e.g. Doane et al., 1988; Beenstock et.al., 1998; Layton and Moeltner, 2004; Carlsson and Martinsson, 2007; Carlsson et al., 2011; and Reichl et al., 2013). Recently, stated preference approach has also been applied to study demand flexibility (Parsons et al., 2014; Broberg and Persson, 2016; Daniel et al., 2018; and Richter and Pollitt, 2018). Despite its documented weaknesses, (1) the stated preference approach is used in these contexts to overcome difficulties in estimating consumer surplus from electricity consumption using price and quantity data. (2)

   The reminder of the paper is structured as follows. Section 2 provides an extended background and motivation underlying the research questions in focus, as well as a review of the related literature. Section 3 includes a conceptual framework with explanations of VoLL and VoPLL. Section 4 provides descriptions of the methodological approach and data used in the empirical analyses. The empirical results are presented in Section 5. Section 6, finally, is devoted to a discussion and concluding comments.

2. BACKGROUND AND PREVIOUS RESEARCH

   DSM in Sweden has targeted large industrial electricity consumers at moments of imminent power shortages. These moments have typically occurred on days with high power consumption due to exogenous factors, sometimes combined with problems in the power grid or in large-scale nuclear power production. The balancing of intermittent power production, however, requires more adaptable resources that can be activated at short notice during any times of the year. In general, large industrial plants are ill-suited to provide such continuous (dynamic) demand response due to their relatively high start/stop costs. For that reason, interest has shifted towards the household sector (Torriti et al, 2010). The household sector in general, and detached and terrace houses in particular, may have a large potential in this context.

   The basic idea is that DSM programs can be used to create timely load shifting/saving among households. Contracts can be designed so that households are financially compensated if they reduce their power demand at moments when the stability of the power system is threatened. Such contracts may be designed in different ways, but ultimately part of the load is controlled remotely by an external actor (Babar et al., 2014). (3) In the contractual context, a central role is given to aggregators that mediate energy services between suppliers, grid owners and end users. The role of the aggregator is to consolidate the fragmented supply of household power services and package it into products that can be sold on the spot market or the regulating markets.

   At the household level, demand response can work through automatic response and/or through behavioral changes. Activities related to automatic response can be referred to as efficiency activities, and those related to behavioral changes to curtailment activities (see, e.g., Gardner and Stern, 2008). Examples of the former are electricity and appliances for heating, the refrigerator and the freezer, which to a large extent are regulated automatically. Examples of the latter are high-power appliances like the kitchen stove and the coffee machine, and low-power appliances like lighting and computers. Since many single- and two-dwelling buildings in Sweden are heated by electricity, automatic response of heating systems has a significant potential to help balance fluctuations in the power system (EI, 2016). For demand response through curtailment activities the story may be different because it requires a behavioral change.

   Previous research reveals that people demand substantial economic compensation to engage in DSM programs. For example, Broberg and Persson (2016) finds, in a choice-experiment study, that people very much dislike restrictions on the use of household appliances during the evening peak hours. In a related context, there is extensive research related to estimating the value of lost load (VoLL). In general, results from the VoLL literature confirm the findings in Broberg and Persson (2016) that people and firms assign a relatively high value to have access to electricity. In a review, Van der Welle and van der Zwaan (2007) find that the average value for developed countries is in the range 4-40 $/kWh, but also that the value differs substantially between sectors and countries. In a more recent review, Schroder and Kuckshinrich (2015) conclude that VoLL varies substantially within, as well as between, end-user groups, countries and estimation methods. Their reported values range between a few [euro]/kWh to more than 250 [euro]/kWh for non-household end-users and between a few [euro]/kWh and 45 [euro]/kWh for households. Overall, the review reveals that the VoLL is very situation- and time specific, implying that when, where and for how long a blackout occurs are important determinants. The review also reveals that the methodology used to estimate VoLL may explain differences in results. For household end-users, it seems like studies based on the stated preference approach tend to result in lower estimates than indirect approaches, e.g. using a household production function to measure VoLL in terms of the lost value of leisure time (for an example see de Nooij et al., 2007).

   The VoLL for Swedish households has been estimated several times using the stated preference approach. Carlsson and Martinsson (2007) and Carlsson et al. (2011) use an open-ended contingent valuation question and estimate that Swedish households on average are willing to pay about [euro] 0.5-22 to avoid an unplanned blackout at 6 pm wintertime with the duration of 1, 4, 8 or 24 hours. (4) These fairly low WTP estimates are to a large extent driven by the large share of respondents stating zero willingness to pay for avoiding a blackout. (5) In a...