Grid Tariff Design for Efficient Utilisation of the Distributor Grid: A qualitative study with actors on the Swedish electricity market

(1)

IN THE FIELD OF TECHNOLOGY DEGREE PROJECT

INDUSTRIAL ENGINEERING AND MANAGEMENT AND THE MAIN FIELD OF STUDY

MECHANICAL ENGINEERING, SECOND CYCLE, 30 CREDITS STOCKHOLM SWEDEN 2020,

Grid tariff design for efficient utilisation of the distribution grid

A qualitative study with actors on the Swedish electricity market

MATILDA HAIKOLA MALIN SÖDERBERG

KTH ROYAL INSTITUTE OF TECHNOLOGY

(2)

Master of Science Thesis

KTH School of Industrial Engineering and Management

Grid tariff design for efficient utilisation of the distribution grid

A qualitative study with actors on the Swedish electricity market

Matilda Haikola

Malin Söderberg

(3)

Master of Science Thesis TRITA-ITM-EX 2020:342

Grid tariff design for efficient utilisation of the distribution grid

A qualitative study with actors on the Swedish electricity market

Matilda Haikola Malin Söderberg

Approved

2020-06-04

Examiner

Per Lundqvist

Supervisor

Per Lundqvist

Commissioner

ÅF Pöyry AB

Contact person

Pablo Giaconi

(4)

Abstract

The Swedish electricity system is transitioning due to the establishment of climate policy goals and trends related to technology and demographics. The transition has resulted in an increased demand for electricity. The increased demand for electricity in combination with lack of forecasts, planning and coordination between actors in the electricity sector has led to the occurrence of grid congestion. Extending the network is time-consuming and requires substantial investments. Instead, an alternative is to utilise the available grid capacity more efficiently by implementing flexibility solutions. Flexibility can be achieved by implementing incentives such as grid tariffs. This solution has recently gained much attention in Sweden, but it is not apparent how grid tariffs should be designed.

The purpose of this thesis is to investigate how distribution grid tariffs could be designed to incentivise different actors to contribute to flexibility in a way that results in an efficient use of the electrical grid. A qualitative study was performed, collecting empirical data through semi- structured interviews with actors in the Swedish electricity market. The aim is that the results from this thesis will act as a basis for DSOs planning to design grid tariffs with the purpose to utilise the grid more efficiently.

The findings present a ToU capacity charge with off-peak periods that are free of charge as the preferable main price signal in the tariff to achieve efficient utilisation of the grid. It is further argued that other structural elements can complement the ToU capacity charge. A small fixed charge could be added in order to contribute to the cost reflectiveness of the grid tariff. A small energy charge could be incorporated in order to provide consumer with incentives to be flexible below the current metered maximum power and strengthen the signal from the ToU capacity charge. A small energy charge can avert difficulties related to providing incentives below the current metered maximum, as it still can provide some incentives to be flexible, or strengthen the signal from the ToU capacity charge. Further, the energy charge can ensure sustainability if customers respond well to a ToU capacity charge and to compensate solar PV customers.

Furthermore, recommendations to further enable the grid tariffs potential to provide price signals include shifting the focus of the revenue cap from CapEx to OpEx and exploring the hampering signals of the energy tax as well as contradicting price signals from the wholesale electricity price.

Keywords: Grid tariffs; Demand response; Time-of-use; Real-time-pricing; Critical-peak-pricing;

Distribution grid; Distributed energy resources; Flexibility markets.

(5)

Sammanfattning

Det svenska elsystemet genomgår en förändring till följd av införandet av klimatmål och trender relaterade till teknik och demografi. Denna förändring har resulterat i ett ökat effektbehov. Ett ökat effektbehov i kombination med bristande prognostisering, planering och samordning mellan aktörer inom elsektorn har lett till uppkomsten av kapacitetsbrist. Att bygga ut elnätet är tar tid och kräver större investeringar. Ett alternativ är att istället utnyttja det befintliga elnätet mer effektivt genom att implementera flexibilitetslösningar. Flexibilitet kan uppnås genom att införa incitament i form av elnätstariffer. Denna lösning har nyligen fått mycket uppmärksamhet i Sverige, men det är inte klart inte hur dessa elnätstariffer ska utformas.

Syftet med detta arbete är att undersöka hur distributionsnätets tariffer kan utformas för att stimulera olika aktörer att bidra med flexibilitet på ett sätt som resulterar i en effektiv användning av det befintliga elnätet. En kvalitativ studie genomfördes där empiriska data samlades in genom semistrukturerade intervjuer med aktörer på den svenska elmarknaden. Syftet är att resultaten från detta arbete ska fungera som ett underlag för nätägare som planerar att utforma

elnätstariffer med syftet att utnyttja nätet mer effektivt.

Resultaten visar att en ToU-effektavgift med gratis off-peak perioder bör vara den huvudsakliga prissignalen i en elnätstariff som ämnar att utnyttja det befintliga nätet mer effektivt. Det visar även att andra strukturella element kan komplettera ToU-effektavgiften. En mindre fast avgift kan adderas i syfte att göra elnätstariffen mer kostnadsriktig. En mindre energiavgift kan införas för att ge kunder incitament att vara flexibla även under den nuvarande uppmätta maximala effekten och stärka signalen från ToU-effektavgiften. Vidare kan energiavgiften säkerställa tillräckliga intäkter för nätägaren om kunderna svarar bra på en ToU-effektavgift och för att kompensera kunder med solceller. Ytterligare rekommendationer för att möjliggöra prissignaler genom elnätstariffer inkluderar att skifta fokus på intäktsramen från CapEx till OpEx och utforska de hämmande prissignalerna från energiskatten och de motstridiga prissignalerna från elhandelspriset.

Nyckelord: Elnätstariffer; Efterfrågeflexibilitet; Distributionsnät; Flexibilitetsmarknader;

(6)

List of tables

Table 1. Outline of the thesis. ... 5

Table 2. The main regulatory principles for grid tariffs. ... 12

Table 3. Grid tariffs theoretical impact on demand response ... 17

Table 4. Decision-making principles and behavioural biases associated with cost-reflective pricing. ... 25

Table 5. Interviewees in the orientational interviews. ... 30

Table 6. Interviewees in the focus interviews. ... 31

Table 7. Phases of template analysis ... 32

Table 8. Techniques to ensure the trustworthiness of qualitative research. ... 34

Table 9. An overview of the interviewed DSOs and their grid tariff structures according to the conceptual framework... 45

Table 10. Evaluation of the energy charge. ... 68

Table 11. Evaluation of the capacity component. ... 72

(10)

List of figures

Figure 1. Tariff design methodology reworked by the authors. ... 8

Figure 2. Fixed pricing for the energy or capacity charge ... 9

Figure 3. ToU pricing for the energy or the capacity charge ... 9

Figure 4. CPP for the energy charge. ... 10

Figure 5. RTP for the energy charge. ... 10

Figure 6. Dimensions of the grid tariff structure with inspiration from. ... 11

Figure 7. Load shifting ... 15

Figure 8. Peak shaving ... 15

Figure 9. Valley filling. ... 15

Figure 10. Strategic conservation. ... 15

Figure 11. Change in electricity use during off-peak hours and peak hours in response to ToU capacity charges. ... 19

Figure 12. Reduction of mean maximum demand by households in Sala Heby two years after the implementation of a ToU capacity based tariff. ... 20

Figure 13. The arc of price responsiveness. ... 21

Figure 14. Illustration of the research process. ... 28

Figure 15. The actors in the electricity system. ... 37

Figure 16. Total price for electricity in Sweden 2020. ... 42

Figure 17. The load profile for 30 January 2019 ... 43

Figure 18. The load profile for 21 June 2019 ... 43

Figure 19. The load profile for 2019 ... 44

(11)

Acronyms

BRP - Balance responsible party CPP - Critical peak pricing

DER - Distributed energy resources DG - Distributed generation

DSF - Demand side flexibility DSO - Distribution system operator

Ei - Energimarknadsinspektionen (Swedish Energy Market Inspectorate) PTR - Peak time rebate

PV - Photovoltaic RTP - Real time pricing

Svk - Affärsverket Svenska kraftnät ToU - Time-of-use

TSO - Transmission system operator

(12)

Acknowledgements

This study was conducted as a Master Thesis project on behalf of ÅF Pöyry AB (AFRY) and the School of Energy Technology at KTH Royal Institute of Technology. The thesis concludes the studies to obtain a Master of Science in Engineering, degree program in Industrial Engineering and Management with a specialisation in Sustainable Power Generation at KTH Royal Institute of Technology.

We would like to extend our gratitude to everyone who has in any way contributed to this study.

To begin with, we want to thank our supervisor Per Lundqvist at KTH for his guidance, feedback and the inspiration he has given us during this process but also previous studies at KTH. Furthermore, we would like to express our great appreciation to our colleagues at AFRY for their advice and support. Special thanks to Simon Siostedt, Anne Grevener and Pablo Giaconi at AFRY for their continuous feedback. We wish to express our gratitude to the interviewees whose expertise and knowledge has created and formed the results of this study. It has been truly inspiring to talk to experts within the field, and this degree project would not have been possible without all of you.

Finally, we would both like to thank our families and friends for their unconditional support in whatever challenges we take on, no matter where they might lead.

Matilda Haikola and Malin Söderberg Stockholm, June 2020

(13)

1. Introduction

This chapter provides an introduction to the thesis. A short background to the subject is presented, followed by the problem statement and the aim of the thesis and the related research questions this thesis will try to answer. The last sections present the delimitations and an outline of the thesis.

1.1 Background

The Swedish electricity system is currently in a transition as a result of several factors such as climate policy goals and technological and demographic trends (Swedish Energy Agency, 2019a).

The Swedish state has set the target to have 100% renewable electricity by 2040 (IRENA, 2020), which entails a significant increase in the share of variable energy sources (Swedish Energy Agency, 2018). The increase in the share of variable energy sources can potentially cause balance issues in the electrical grid and threaten the reliable operation of the system (Huang and Wu, 2019). Furthermore, the demand for electricity is expected to increase due to trends such as electrification of the transport sector, the construction of new computer centres and emerging electricity-intensive industries. The increased demand for electricity will result in a higher power requirement, which puts pressure on the capacity in the electric grid (Sahlén et al., 2019). The capacity of the grid refers to the physical properties of the electrical grid that limit the power that can be delivered and how much electricity the grid can transport (Wiesner et al., 2019). When the power requirement exceeds the available capacity in the grid, congestion problems occur.

In Sweden regions such as southwest Skåne, Stockholm, Mälardalen, Uppsala and Gothenburg are experiencing problems with congestion or risk experiencing problems with congestion in the near future as a result of congestion in the transmission grid (Wiesner et al., 2019). The

combination of population growth and urbanisation has resulted in the increased electricity and power demand mainly being located in metropolitan regions in Sweden (Sahlén et al., 2019).

Wiesner et al. (2019) conclude that congestion problems in Sweden are not only a result of increased electricity demand but also a lack of forecasts, planning and coordination between actors in the electricity sector. Short term consequences from congestion problems include limited opportunities to establish or expand businesses and infrastructure. Long term consequences are related to the implementation of climate actions (Wiesner et al., 2019).

Managing congestion requires either network extensions or the implementation of incentives to decrease use (Neuteleers et al., 2017). Investments aimed at extending the grid capacity will be required to handle the increased power requirement. However, these projects usually take 10-15 years to complete, and the high growth rate of urban areas put pressure on the lead time for new capacity. Investments in the grid also entail a higher cost for the final customer, and the number of hours where the grid capacity is reached are relatively few (Sahlén et al., 2019). Hence, there is a need for alternative solutions for utilising the available grid capacity more efficiently. By altering load patterns, further investment in available grid capacity can be reduced or deferred in the longer run, which lowers overall costs in the system to the benefit of customers

(EURELECTRIC, 2016).

(14)

Flexibility is thus seen as an essential solution to handle the increasing share of intermittent energy and manage the local capacity deficiency. The demand side has been identified as a segment with the potential to provide flexibility (Sahlén et al., 2019). Flexibility can be achieved by implementing incentives for consumers to utilise the grid more efficiently (Neuteleers et al., 2017). The electric grid constitutes a natural monopoly, meaning it is not socio-economically viable to build competing electrical grids (Ek and Hallgren, 2012). As network industries are natural monopolies where costs are covered through grid tariffs (Bergaentzlé et al., 2019), one way to incentivise customers to utilise the grid and the available energy more efficiently is thus through price signals via grid tariffs. A grid tariff is a fee paid by customers to distribution network companies for the delivery of electricity. By designing it properly, it is possible to lead the system towards a more efficient operation by modifying the consumption and generation pattern of different actors (Rodríguez Ortega et al., 2008).

Using grid tariffs as a means to utilise the available capacity in the grid more efficiently has recently gained much attention in Sweden. At the beginning of 2019, the Swedish Electricity Act (1997:857) was revised to make it possible for electricity distribution companies to conduct a pilot project on a small customer segment, testing grid tariffs that result in more efficient utilisation of the grid (Kaplin, 2019). Changes in the Swedish Electricity Act (1997:857) made in 2018 means Ei is issuing new regulations on how grid tariffs shall be designed in order to promote efficient use of the grid (Carlsson, 2020). In January of 2020, it was reported that the project had changed focus to investigate further the permission of locational signals in the grid tariff (Tvingsjö, 2020a). In May of 2020, it was declared that a suggested revision of the Swedish Electricity Act (1997:857) regarding locational signals had been sent in by Ei (Tvingsjö et al., 2020).

1.2 Problem statement

Grid tariffs are recognised as a tool for increasing efficiency in the electrical system. Currently, it is not apparent how grid tariffs should be designed to provide an optimal price signal for efficient use of available resources (THEMA, 2019). A particular tariff design can provide an efficient price signal in theory but may experience issues in practice. Limiting factors such as the knowledge of the customer (THEMA, 2019) and other conflicting principles such as cost reflectiveness and full cost recovery for the distribution system operator (DSO) need to be considered during the design process (Rodríguez Ortega et al., 2008).

New actors, markets and technologies are emerging that may affect the electricity market

conditions and hence how to design the grid tariff in order to facilitate or incentivise efficient use of the distribution grid. Our traditional power system, where the power is “trickling down” from the highest voltage level to the lowest voltage level as large power plants are delivering electricity to end consumers, is changing due to the emergence, and increasing implementation, of

distributed energy resources (DER) (Perez-Arriaga, 2016). DER can be defined as resources connected to the distribution grid, i.e. the lower voltage level, such as solar photovoltaics (PV), small wind farms, energy storage, demand side management and electric vehicles. DER will likely

(15)

impact the design, operation, organisation and regulation of the distribution network (Perez- Arriaga, 2016).

Furthermore, new local flexibility markets for more efficient use of the electrical grid are being implemented in Sweden in Uppland, Gotland, Skåne and Västernorrland. The implementation of these markets is a part of the EU financed project CoordiNet, that aims to demonstrate how DSOs and TSOs (transmission system operators) shall act in a coordinated manner and use the same pool of resources to procure grid services more reliably and efficiently (Energiforsk, 2019).

As industry customers are sensitive to price changes and are actively trading their flexibility (Alvehag et al., 2016), it is interesting to focus on smaller customers in this context. For example, residential customers and other property owners that might experience issues in contributing with their flexibility. Actors such as energy service companies and aggregators become essential to enable such customers to actively participate with their flexibility (Sahlén et al., 2018).

Questions arise how grid tariffs should be positioned concerning the increasing implementation of DER and the emergence of flexibility markets.

A large share of previous studies in the area has been focused on assessing the effect on

consumption and economic performance of a selected tariff design (Bartusch et al., 2018, 2014, 2011; Koliou et al., 2015; Svalstedt and Löf, 2017). However, few studies give an overview of different grid tariff structures regarding customer response and their practical implications for grid operators. The impact of energy sources such as solar PVs has been studied by some, see for example; Cambini and Soroush (2019), Angela Picciariello et al. (2015), A. Picciariello et al.

(2015) and Simshauser (2016). However, there is a shortage of studies providing a holistic perspective on the design of grid tariffs in these changing market conditions. As expressed by A.

Picciariello et al. (2015), the design of the tariff structure remains an open question.

1.3 Aim and research questions

With the given background and problem formulation, the overarching purpose of this thesis is to investigate how distribution grid tariffs could be designed to incentivise different actors to contribute to flexibility in a way that results in an efficient use of the electrical grid. The aim is that the results from this thesis will act as a basis for DSOs planning to design grid tariffs with the purpose to utilise the grid more efficiently.

In order to fulfil the purpose of this thesis, the following research questions will be answered:

Main RQ: How can grid tariffs be designed in order to create incentives for actors in the electricity market to utilise the grid more efficiently?

RQ1: How do the actors on the electricity market interpret the efficient use of the grid?

RQ2: How can grid tariffs be designed in order to alter customer behaviour towards utilising the grid more efficiently?

RQ3: How can grid tariffs be designed in order to ensure a sustainable operation of grid operators’

businesses?

RQ4: How can grid tariffs be designed with consideration to other solutions for utilising the grid more efficiently, including flexibility markets and distributed energy resources?

(16)

1.4 Delimitations

The focus of this thesis is to evaluate different grid tariff structures. Other steps in the design process, such as calculating the revenue requirement and computing the final rates for the components in the tariff structure are not considered in this thesis. The recommendations in the thesis regarding grid tariff design are mainly concerned with distribution tariffs provided by DSOs in the local grids. Hence, feed-in tariffs and connection charges will not be considered in this thesis. This thesis is limited to the geographical area of Sweden since laws and regulations regarding the electricity sector are determined nationally. The current regulations of DSOs and the design of grid tariffs act as a basis for the thesis but are considered flexible as they might change in the future. For example, the revenue cap is, to some extent considered, but only in terms of its existence and not the factors affecting the size of it. Furthermore, different conditions apply for different types of customers. Therefore, only smaller customers with the maximum fuse size of 63 A are considered in this thesis. These customers include smaller businesses, offices and residential facilities.

1.5 Contribution

This study examines how grid tariffs should be designed in order to achieve efficient utilisation of the distribution grid. The study contributes to the discussion regarding the design of grid tariffs. It provides a simple overview of the tariff design options based on a conceptual framework and practical implications of these. Further, this thesis contributes to previous theoretical and empirical findings by including the perspective of several actors involved in the Swedish electricity system and hence provide a more holistic picture of the subject.

(17)

1.6 Outline

In table 1, the different chapters and the outline of the thesis is presented.

Table 1. Outline of the thesis.

Chapter Description

1 Introduction In the introduction, a background to the problem is provided. The problem statement and the related research questions are

presented.

2 Literature and theory

Previous literature on the subject as well as established theories that will be used to analyse the result is presented. Focus is put on grid tariff structures, the effects of grid tariff on customer behaviour and principles related to the design of tariffs and decision-making criteria.

3 Method This chapter describes the research design, research approach, data collection and analysis used to carry out the thesis. Methods to ensure research quality and fulfilment of ethical aspects are described.

4 Overview of the Swedish electricity system

A brief overview of the Swedish electricity system is presented to provide a background about the conditions in Sweden regarding the electrical grid, the market design and involved market actors.

Readers who are well acquainted with the market structure in Sweden can skip this chapter.

5 Results and analysis

The answers from the semi-structured interviews are presented based on themes and the conceptual framework presented in the literature and theory chapter.

6 Discussion The empirics are discussed using the regulatory principles for tariff design, the decision-making criteria, and previous findings derived from the literature and theory chapter.

7 Conclusions Answers the research questions presented in the introduction chapter based on empirics and literature.

(18)

This page is intentionally left blank.

(19)

2 Literature and theory

The literature and theory chapter will act as a basis for analysing the results in this thesis and is divided into five parts. In the first part, different grid tariff structures are presented, followed by a presentation of the fundamental principles of tariff design. The third part goes deeper into the concepts of demand-side flexibility and demand response, and their relation with grid tariffs. In the fourth chapter, literature regarding grid tariffs in the presence of distributed energy resources is presented—finally, the focus shifts to decision-making principles derived from the field of behavioural economics.

2.1 Grid tariff structures

Due to the shared nature of the grid, the cost of providing service to one user depends on others being provided with the same service, and also upon how the other users use the system

(Sakhrani and Parsons, 2010). The distribution tariff, also called distribution use of system charge, is meant to cover the recurrent capital and operating costs for network expansion, operation and maintenance (A. Picciariello et al., 2015). It is estimated that residual (or fixed) costs account for about 70-90 % of the grid companies costs (Ei, 2020b). The costs related to the distribution network can be divided into the following groups (Ek and Hallgren, 2012; Lydén et al., 2011):

● Customer administration costs (measurement and billing)

● Network costs (investment, return on capital, maintenance and operation)

● Distribution costs (grid losses)

● Overlying grid costs (variable and fixed)

There exist several theories for how these costs should be allocated, and these have developed over time. The first theory-backed cost allocation methodology was the accounting approach, where the main objective was to recover all the cost items in the company accounts. In the more recent literature, it is possible to differentiate between two major approaches for allocating costs and hence designing grid tariffs; a cost-causality based approach and an approach based on economic theory and marginal costs. Even though the latter aims to achieve a better economic signal, it has some shortcomings, hence the former is the most usual solution applied (Reneses and Rodríguez Ortega, 2014).

In general, the design of a distribution grid tariff can be separated into two parts. The first part is setting the total allowed revenue, or revenue cap, for the distribution grid operator (Rodríguez Ortega et al., 2008). A revenue cap is required due to the monopolistic character of the electrical grid, where grid companies otherwise could take advantage of their position and charge higher prices (Sakhrani and Parsons, 2010). The calculation of the allowed revenue has received much attention in the literature, while the tariff structure design can be seen as an open question (A.

Picciariello et al., 2015). In the Swedish context, the allowed revenue is set by The Swedish Energy Market Inspectorate (Ei) which is referred to as the revenue cap.

The second part of the distribution tariff design is the allocation of this allowed revenue among all the users of the network (Rodríguez Ortega et al., 2008). It involves the actual design of the tariff structure, as the tariff structure determines how the costs are distributed among the users

(20)

(Nijhuis et al., 2017). This second part can be conducted based on three steps, as illustrated by figure 1. In the first step, a tariff structure is defined, followed by an allocation of the allowed revenue to each component in the structure, and finally, the final rates are computed. The process is not merely sequential as illustrated by the picture as the steps need to be performed simultaneously. Even if the tariff structure is the starting point, the final structure will be dependent on how costs are allocated, which is the second step (Rodríguez Ortega et al., 2008).

Figure 1. Tariff design methodology reworked by the authors (Rodríguez Ortega et al., 2008).

The allocation of the revenue requirement and calculation of final rates will not be considered in this thesis. Hence the different options for the grid tariff structure will be the focus of the remaining part of this chapter.

2.1.1 Structural elements

Usually, the grid tariff can consist of three different structural elements: a fixed charge (€/period), an energy charge (€/kWh/period) and a capacity charge (€/kWpeak/period) (A. Picciariello et al., 2015; Rodríguez Ortega et al., 2008; THEMA, 2019). The fixed charge is an invariant fee independent of the customers’ consumption (A. Picciariello et al., 2015) and should represent costs such as billing and metering of the customer (THEMA, 2019). The energy charge depends on consumption and is a cost per electricity consumed. The capacity charge is a cost based on the maximum power used during a specific time range (Kirkerud et al., 2016). There are several alternatives available for the design of this component. One option is to let the customer decide in advance the level of standby power they require. This arrangement can be realised through fuse tariffs based on the fuse size of the customer, or subscription tariffs based on an agreed upon level of power the customer subscribes on (THEMA, 2019). The other option is to base the charge on the customer’s metered maximum power consumption during a specific period (A. Picciariello et al., 2015). The number of points of measurements can vary depending on the DSO, and the charge can be time differentiated.

The final tariff structure can be composed of one or all of these above mentioned components.

A tariff, including both an energy and a capacity charge, can be referred to as a two-part tariff (EURELECTRIC, 2013). A two-part tariff can also refer to a fixed component together with either an energy charge or a capacity charge. A tariff, including all three structural elements, can be called a three-part tariff (Simshauser, 2016). From now on, when referring to a two-part tariff, the definition provided by EURELECTRIC (2013) is used.

(21)

2.1.2 Time differentiation

According to Perez-Arriaga (2016), one of the major design issues for charges and prices in the power system in the future will be deciding the level of granularity in time and location, i.e. how far one should go with time and geographical differentiation. Regarding time differentiation, it is possible to distinguish between two grid tariffs; static and dynamic grid tariff. Static tariffs are determined in advance (IRENA, 2019a). There are disagreements regarding the concept of dynamic grid tariffs where some claim that tariffs based on fixed time periods are dynamic. In contrast, others claim that dynamic tariffs are determined in real-time based on actual system conditions. In general, dynamic tariffs can be defined as a tariff that varies over time based on demand (Ek and Hallgren, 2012).

Grid tariffs can be more or less dynamic (THEMA, 2019). The scale ranges from static flat-rate pricing with a charge for a fixed amount of energy or a predefined capacity to dynamic real-time pricing (RTP) where rates vary each hour. The rates used in RTP depend on the actual system conditions and can reflect either the wholesale electricity price (Bergaentzlé et al., 2019; Kirkerud et al., 2016) or the situation in the grid (THEMA, 2019), see figure 5. In between these two extremes, there are fixed pricing, time-of-use (ToU) pricing and event-driven pricing. Fixed pricing means there is a fixed price per kWh or per kW (EURELECTRIC, 2013; THEMA, 2019), as illustrated by figure 2. ToU pricing refers to predeclared tariffs varying during different time intervals, where tariffs are high during peak periods and low during off-peak periods. These intervals can consist of a number of hours during the day, or rates can vary by day, week, month or season (Kirkerud et al., 2016; Koliou et al., 2015), see figure 3. Event-driven pricing includes critical peak pricing (CPP), variable peak pricing (VPP) and peak time rebate (PTR). In CPP the prices are

abnormally high during event days consisting of critical peak periods (40-150 hours per year) and discounted during the rest of that specified day during non-critical periods (Koliou et al., 2015).

The peak pricing remains the same for all time intervals. CPP is illustrated in figure 4. VPP is similar to CPP but differs in the sense that peak prices vary between different time intervals.

Peak time rebate is another form of event-driven pricing, where rebates are provided for consumption below a certain predetermined level during peak hours (Dutta and Mitra, 2017).

Figure 2. Fixed pricing for the energy or capacity charge Figure 3. ToU pricing for the energy or the capacity charge, inspired by (Yan et al., 2018).

(22)

Figure 4. CPP for the energy charge inspired by (Yan et

al., 2018). Figure 5. RTP for the energy charge inspired by (Yan et al., 2018).

Not all structural elements allow for all levels of time differentiation. The fixed charge is, by definition, limited to flat-rate pricing. Unlike the energy charge, the options for varying the capacity charge based on the metered maximum power outtake are in the literature often limited to flat-rate pricing, fixed pricing and ToU-pricing (THEMA, 2019). Some authors further mention variable pricing as an alternative for the capacity charge. In variable pricing, there are different prices for each defined capacity level (EURELECTRIC, 2013; A. Picciariello et al., 2015). In this study, variable peak pricing is assumed to be included in flat-rate pricing.

2.1.3 Geographical differentiation

Geographical differentiation can be divided into nodal pricing, zonal pricing and uniform pricing.

Nodal pricing acknowledges that location is an important aspect of electricity which should be reflected in its price (Holmberg and Lazarczyk, 2015). In nodal pricing, marginal losses and grid congestions are reflected in the prices in each node. Hence, in nodes with a power surplus, the price will be lower than in nodes with power deficit (THEMA, 2019). Zonal pricing aggregates nodes into zones with uniform pricing considering only transmission congestion between the zones (Lin and Magnago, 2017; IRENA, 2019). Grid congestion inside a zone is managed by redispatch resulting in situations in which the prices on constrained and unconstrained sides of the zone have to diverge. In uniform pricing, transmission constraints are neglected, and transmission is separated from energy trading. As a result, the uniform pricing will often fail to give adequate signals for efficient use of transmission capacities (Lin and Magnago, 2017).

(23)

2.1.4 Composition of the grid tariff

The following dimensions of the grid tariff structure have been identified as options during the design of the tariff; structural elements, time differentiation and geographical differentiation as illustrated by figure 6. Building on the literature presented above and the authors own reflections, the available options for time differentiation of each component are represented in figure 6.

These identified dimensions will act as a conceptual framework in the study.

Figure 6. Dimensions of the grid tariff structure with inspiration from (Eid et al., 2016; EURELECTRIC, 2013).

2.2 Regulatory principles for distribution tariff design

There exists a consensus in the literature regarding a number of regulatory principles that grid tariffs should observe. Despite that these principles provide guidelines on how grid tariffs should be designed, there is still room for interpretation (Reneses and Rodríguez Ortega, 2014).

Rodríguez Ortega et al. (2008) argue that some principles are boundary setting, while others provide clues about how things are to be done. Which principles that need to be considered and the relative importance of these can be argued to differ. In table 2, some generally accepted principles are presented.

(24)

Table 2. The main regulatory principles for grid tariffs (EURELECTRIC, 2013; Rodríguez Ortega et al., 2008;

Sakhrani and Parsons, 2010; Strbac and Mutale, 2005).

Principle Description

Cost-reflectiveness The tariffs should reflect the costs incurred by each customer connected to the grid.

Economic efficiency

Price signals should be given to customers and DSOs through the grid tariffs that incentives a behaviour that maximises the social welfare, both in the short term by efficient operation and the long term by following the path of least cost development (investments). Hence network services should be provided at the lowest possible cost, and overinvestment should be discouraged.

Non-discriminatory (equity)

This principle, also called equity, is dependent on how discrimination is defined. In general, rates can be regarded as non-discriminatory when customers are charged the same amount for using the same service or good, independent of customer characteristics or indented use of the electricity. Others say that there should be no discrimination between customers within the same customer group (e.g. rural and urban customers).

Sustainability This principle regards the fact that grid tariffs need to provide a complete cost recovery of all allowed network costs and a reasonable return on capital for the DSOs. It is thus important to consider during the calculation of the revenue requirement.

Additivity The sum of all components should add up to the total revenue requirement.

Stability The tariff design should be stable in the short term while gradually changing in the long term in a way that can be forecasted by interested parties, in order to minimise uncertainty regarding investments and regulation.

Simplicity Network tariffs should be designed in a way that makes them easy to understand and implement.

Transparency The tariff design methodology should be transparent and available to all users of the grid to provide predictability.

Consistency The tariff design should be consistent with the regulatory framework in each country.

(25)

Among these principles, the cost-reflective principle has gained a lot of attention in the literature and is, by some, considered to be one of the major objectives in grid tariff design (Belmans and Jargstorf, 2015; Nijhuis et al., 2017; Ruester et al., 2014). Sakhrani and Parsons (2010) suggest two more principles; universal access and allocative efficiency.

It is important to note that these principles may be conflicting and the fulfilment of all of these principles or objectives might become complex (A. Picciariello et al., 2015; Rodríguez Ortega et al., 2008). For example, the non-discriminatory principle might lead to the network tariff being evened out for different customers even if they incur different costs, and hence creating a

conflict with the cost-reflectiveness principle. Furthermore, simplicity and cost-reflectiveness can be difficult to achieve at the same time; the same goes for economic efficiency and simplicity.

For that reason, it might be necessary to prioritise some of the principles over others during the rate design (A. Picciariello et al., 2015). THEMA (2019) emphasises that economic efficiency is difficult to achieve if the customers cannot respond to the price signals due to limited

understanding of the price signals or how to act based on them. According to EURELECTRIC (2013), the challenge in regulation is to establish a trade-off between these competing objectives.

Belmans and Jargstorf (2015) further elaborates on this and mentions additional policy objectives that need to align with the tariff design, and these are:

● Affordable retail prices for electricity

● Reduction of energy consumption

● Increase in renewable generation

● Reduction in CO2 emissions

These additional objectives make the tariff design more complex by adding trade-offs, for example, with the cost-reflectiveness principle. Distributed generation can add further trade-offs as it can lead to increased cross-subsidisation between users. Under a net-metering or net-billing scheme, users with distributed generation can lower their grid costs, despite not contributing to the costs they cause to the system. This leads to other users having to cover for the lack of revenues for the grid operator. The cost-reflective tariff could address this, but then there is a trade-off as reduced fees is a part of the business case for distributed generation, and this might contradict the policy objective to increase renewable generation. Another trade-off can be seen for the reduction of overall consumption when a capacity component is introduced to make the tariff more cost-reflective. This component entails that the price of electricity is reduced and so is the incentive to reduce consumption (Belmans and Jargstorf, 2015).

2.3 Demand side flexibility and demand response

One emerging trend in the literature regarding grid tariffs is their role in attempts to utilise demand-side flexibility and their involvement in demand response programs. DSOs play a crucial part in stimulating demand response due to the fact that they hold the necessary technical

infrastructure (Bartusch and Alvehag, 2014). There is a wide variety of definitions of the concept demand-side flexibility (DSF) and demand response. The Swedish Energy Market Inspectorate defines DSF in the following way (Alvehag et al., 2016):

(26)

“Demand side flexibility is a voluntary change in the demand for electricity from the grid during shorter or longer periods, caused by of some type of incentive.”

Another definition of DSF is provided by CEER (2014), who defines it as:

“Demand-side flexibility can be defined as the capacity to change electricity usage by end-use customers (including residential) from their normal or current consumption patterns in response to market signals, such as time-variable electricity prices or incentive payments, or in response to acceptance of the consumer’s bid, alone or through aggregation, to sell demand reduction/increase at a price in organised electricity markets or for internal portfolio optimisation...”

The definition provided by Ei considers the fact that the electricity consumption does not have to be equal to the demanded electricity from the grid as customers can have storage or

production facilities (Alvehag et al., 2016), which is not considered in CEER’s definition.

However, CEER’s definition acknowledges that DSF can be activated through a wide range of signals, from dynamic grid tariffs to contractual-based direct load control that may be automated (CEER, 2014). In that respect, the customer does not have to be active in all cases, as implied by Ei (Alvehag et al., 2016). This report will be based on a definition of DSF which is a

combination of the two:

“Demand-side flexibility can be defined as the capacity to change the demand for electricity from the grid by end-use customers (including residential) from their normal or current demand patterns in response to market signals or some type of incentive.”

Furthermore, CEER’s definition highlights that DSF refers to the capacity to change consumption and not the change itself, which is called demand response (EURELECTRIC, 2013; SETIS, 2014). According to EUCELECTRIC (2013), demand response can be defined as:

“… changes in electric usage by end-use customers from their normal consumption patterns in response to changes in the electricity price or incentive payments designed to alter timing, level of instantaneous demand or total electricity consumption.”

Demand response can be signalled via two different mechanisms; price-based and incentive- based mechanisms (Alvehag et al., 2016; Bartusch and Alvehag, 2014; Koliou et al., 2015; SETIS, 2014; Shen et al., 2019). Price-based demand response also called implicit demand-side flexibility, involves variable prices reflective of grid conditions or active hourly market (Koliou et al., 2015), where customers are motivated to shift their consumption to lower their total electricity cost.

This can be achieved using time-varying electricity prices and grid tariffs (Alvehag et al., 2016).

On the contrary, incentive-based demand response, also called explicit demand-side flexibility, compensate customers for increasing or decreasing their consumption upon request with an ex- ante contract (Koliou et al., 2015). The request can, for example, be based on conditions in the grid, such as shortage of capacity (Bartusch and Alvehag, 2014). Common arrangements for utilising incentive-based demand response are direct load control, demand bidding, emergency demand response, capacity markets, curtailable services and ancillary services markets (Goldman et al., 2010; Koliou et al., 2015).

(27)

Demand response can be carried out in different ways. Customers can either actively respond to price signals or use automated solutions (SETIS, 2014). For example, when using incentive-based demand response, incentive payments can be given to customers for allowing appliances to be remotely controlled by the DSO, i.e. direct load control, or the customers can upon agreement control the appliances themselves to reduce or increase consumption during a period (Song, 2019). In price-based demand response, customers can react to price changes by actively

adapting their consumption or changes in consumption can be automatically adapted (European Smart Grids Task Force, 2019) for example by using smart appliances. Solutions in which the load is automatically altered based on price signals have been suggested to be preferable, as it can lower the burden on customers and the response to the price signals can be predicted to a greater extent (Schreiber et al., 2015).

Activating demand-side flexibility can result in different changes in demand patterns, i.e.

different types of demand response, and these are illustrated in figure 7-10: Load shifting, Peak shaving, Valley filling and Strategic conservation (Alvehag et al., 2016; Koliou et al., 2015).

Figure 7. Load shifting, inspired by (Alvehag et al., 2016; Koliou et al., 2015)

Figure 8. Peak shaving, inspired by (Alvehag et al., 2016; Koliou et al., 2015)

Figure 9. Valley filling, inspired by (Alvehag et al.,

2016; Koliou et al., 2015). Figure 10. Strategic conservation, inspired by (Alvehag et al., 2016; Koliou et al., 2015).

(28)

Load shifting means that the customers displace parts of the load from peak-hours to off-peak hours (Koliou et al., 2015), or from high price hours to low price hours. This can relate to electricity use that cannot be avoided but can be used at another point in time, such as charging an electric car or electricity for space heating. During peak shaving, the customer lowers its electricity use during peak hours, without compensating for this another point in time. Valley filling occurs when customers increase their electricity use during off-peak hours (Alvehag et al., 2017). During strategic conservation, the customer lowers its overall electricity use (Koliou et al., 2015).

2.3.1 Price-based demand response via grid tariffs

A proper design of the grid tariff should, according to the literature promote efficient short term usage of the grid, while also ensuring optimal long term development (A. Picciariello et al., 2015).

Efficient usage of the grid can be interpreted as consumers shifting their load from peak hours to off-peak hours, which may result in network reinforcements and investments being delayed (Rodríguez Ortega et al., 2008). According to EURELECTRIC (2013) load shifting is one part in ensuring an energy-efficient use of the infrastructure, but they also add peak shaving and valley filling, which all can be achieved with demand response. Sahlén et al. (2019) also acknowledge peak shaving as a part of efficient utilisation of the grid.

Grid tariffs have the potential of utilising demand-side flexibility by activating price-based demand response. There are studies which have shown that grid tariffs might have more potential to provide price signals compared to electricity prices according to Alvehag et al.

(2016). The different grid tariff structures previously presented will have different effects on the load patterns of the customer and hence lead to different demand responses. A summary of the theoretical demand response of some grid tariffs is provided in table 3. Related trade-offs with the previously presented regulatory principles can also be found in table 3.

(29)

Table 3. Grid tariffs theoretical impact on demand response adapted from (EURELECTRIC, 2013; Koliou et al., 2015).

Grid tariff Impact on load Regulatory trade-off

Fixed energy charge Strategic conservation - Cost reflectiveness - Sustainability

- Economic efficiency + Simplicity

ToU energy charge Load shifting, Peak shaving + Cost reflectiveness - Sustainability

+ Economic efficiency - Simplicity

Fixed/ToU capacity charge

Load shifting, Peak shaving, Valley filling

+ Cost reflectiveness + Sustainability

+ Economic efficiency + Simplicity

Two-part tariff:

- Fixed capacity charge

- Fixed/ToU energy charge

Peak shaving, Load shifting + Cost reflectiveness + Sustainability

+ Economic efficiency - Simplicity

EURELECTRIC (2013) indicate that all the grid tariff options, except the fixed energy charge, have a good representation of the induced cost. As peak demand is one of the main drivers for costs, the tariff options incentivising reduction of peak consumption during peak hours, have a higher potential for reducing network costs. This applies to all options except the fixed energy charge, which simply incentivises an overall consumption reduction, i.e. strategic conservation (EURELECTRIC, 2013). The ToU energy charge, as well as the two-part tariff, is claimed to have higher complexity and measurement requirements compared to the other options (EURELECTRIC, 2013).

The tariff designs including a capacity component (the capacity based tariff or the two-part tariff), is stated to better guarantee revenue adequacy for the DSO, i.e. achieve the sustainability principle. With large energy-based tariffs, the charge per kWh might need to increase to offset this loss of consumption and customers might be incentivised to introduce energy efficiency measures as a way to lower overall consumption (EURELECTRIC, 2013). The need for

sequential increases is especially valid when regulated revenues are held constant. A tariff where the energy charge is dominating is more suitable for an environment where there is a high energy demand growth, which is not the case today (Simshauser, 2016). Koliou et al. (2015) mean that several studies recommend that future tariffs move away from charges solely based on energy and incorporate capacity charges to more accurately reflect the impact customers have on network costs. More recent literature agrees that alternative solutions to energy-based tariffs

(30)

should be found, due to the problems with cost recovery and their negative impact on demand side flexibility (Bergaentzlé et al., 2019).

EURELECTRIC (2013) states that more capacity based tariffs such as two-part tariffs including both an energy charge and a capacity charge, or other tariffs that penalise consumption during peak hours can encourage more efficient use of the network capacity. However, the metered maximum power capacity charge has received criticism for making the customer reduce its load even if it is not beneficiary for the system or needed (Koliou et al., 2015; THEMA, 2019).

THEMA (2014) points out that the introduction of capacity based tariffs has the potential to increase demand response and reduce costs.

Time differentiations such as event-driven pricing and RTP can induce demand response. Event- driven pricing of energy charges, such as CPP, are argued to incentivise customers to either reduce their peak consumption (peak shaving) or shift their load from peak to off-peak periods (load shifting) to protect system reliability. RTP, on the other hand, motivates customers to adjust their electricity consumption according to the demand and supply elastics (Yan et al., 2018).

The above discussion focuses on the variable tariff options and omits the impact of the fixed components in the tariff. The fixed parts of the tariff are argued by Alvehag et al. (2016) to hamper the effects of the variable tariff components when providing price signals to customers.

In the case of fuse tariffs and subscription tariffs (flat capacity charges), one can argue that there is a long term effect on demand-side flexibility as customers can be incentivised to reduce the fuse size or the amount of subscribed power (Alvehag et al., 2016). On the contrary, THEMA (2019) declares that a flat capacity charge (based on fuse size or subscription) creates no incentives for the customer to reduce the power consumption below the set maximum power level.

2.3.1.1 Customers' response to implemented grid tariffs

There have been several studies conducted in Sweden, investigating residential customers' response to different capacity based grid tariffs. Two of these studied the effects of a ToU capacity charge together with a fixed charge based on fuse size, implemented in Sala-Heby and Sollentuna. The two DSOs defined peak hours as weekdays between 7 am and 7 pm and off- peak hours were free of charge. Results show that consumers are able to shift consumption from peak hours to off peak-hours or simply reduce consumption during peak hours and reduce their mean maximum demand without enabling technology (Bartusch et al., 2014, 2011). See figures 11-12.

(31)

Figure 11. Change in electricity use during off-peak hours and peak hours in response to ToU capacity charges obtained from (Bartusch et al., 2014, 2011).

As illustrated by figure 11, households in the Sala-Heby area managed to shift their electricity use from peak hours to off-peak hours by 2.5 % during the summer months two years after

implementation. No obvious changes were observed during the winter months (Bartusch et al., 2011). Responses to the capacity based tariff were also observed among single-family households in Sollentuna who reduced their electricity use during peak hours by 2.3% and 1.2% during summer and winter months, respectively, compared to a reference group as illustrated by figure 11 (Bartusch et al., 2014).

In a follow-up study in Sala-Heby, it was found that there are long term effects of introducing a capacity-based ToU tariff as well for residential customers living in single-family homes

(Bartusch and Alvehag, 2014). This can further be confirmed by the study conducted in

Sollentuna in 2014 as the capacity based tariff was implemented during the late 1990s (Bartusch et al., 2014). In Sandviken Energi Elnät, a different ToU capacity charge was introduced for the customers with fuse sizes between 35-63A. The capacity-based tariff varies each month and is based on the average maximum power outtake during each month the two previous years. A noticeable overall reduction in consumption was observed after the introduction of the capacity- based tariff. However, it was not possible to say if there was a shift from peak hours to off-peak hours (Bartusch et al., 2018).

(32)

Figure 12. Reduction of mean maximum demand by households in Sala Heby two years after the implementation of a ToU capacity based tariff (Bartusch et al., 2011).

As illustrated in figure 12; it was further found that households in Sala-Heby responded to the capacity based tariff by reducing the mean maximum demand for capacity by 11.7 % and 9.5% in peak periods of the summer and the winter seasons, respectively, as illustrated by figure 12, two years after implementation (Bartusch et al., 2011).

No effect in terms of demand response was seen among households in condominium apartments in Sollentuna (Bartusch et al., 2014), which align with the findings in Sala-Heby where the demand response was higher for single-family houses and rental apartments compared to households in condominium apartments (Bartusch and Alvehag, 2014). Difficulties to

understand the difference between energy and power among customers has been observed both in Sala-Heby as well as in Sandviken when introducing a capacity-based tariff (Bartusch et al., 2018, 2011). It is claimed that this might result in a reluctance to adapt to changes (Bartusch et al., 2011).

With enabling technology, a slightly higher shift was observed during a pilot project in Gotland.

A ToU grid tariff (not specified whether it is an energy or capacity charge) was implemented together with two other price signals, Nord Pool Spot prices and a wind component. In the studied population, including both people with and without automatic control, 55% managed to steer away more than 5% of their total consumption during peak hours three years after

implementation. The population with automatic control as a whole lowered their consumption with an average of 4% compared to prior to the test (Svalstedt and Löf, 2017). All the studies mentioned above have a common limitation, namely small samples that may not be

representative of larger populations.

Customers' response to energy-based tariffs has not been studied to a large extent in Sweden;

however, there are multiple international studies on the subject. In a meta-analysis, Faruqui et al.

(2017) studied different time-varying energy charges, including ToU, CPP, VPP and PTR, with the majority being ToU rate designs. More than 60 pilots were included in the database, and they included both grid tariffs and electricity retail prices. The authors concluded that customers respond to price signals from time-varying energy charges by reducing their peak load and that

(33)

the peak-to-off-peak price ratio in relation to the % peak reduction creates an arc-like shape, see figure 13.

Figure 13. The arc of price responsiveness (Faruqui et al., 2017).

A peak-to-off-peak price ratio of 2:1 entails a 5% reduction in peak usage, while a price ratio of 4:1 results in a 10% reduction of peak usage, on average. In the presence of enabling technology, the magnitude of demand response is even stronger, which is illustrated in figure 13. Enabling technology includes devices that provide customers with the ability to actively respond to price signals, such as in-home displays and smart thermostats (Faruqui et al., 2017).

In a Danish study, it was concluded that RTP based on the spot price does not have an impact on households' electricity consumption as the interviewed households “found the real-time pricing scheme too complicated and time-consuming to follow” (Friis and Haunstrup

Christensen, 2016). In another study conducted in Sala-Heby, it was found that when households were faced with a ToU capacity charge and an RTP energy charge they seemed to give

precedence to the ToU tariff over the RTP, due to the economic impact being perceived as larger by adapting to the ToU tariff. Furthermore, households tended to translate the RTP signal into a ToU version. The authors argue that the ToU tariffs might be more effective in terms of reducing peaks than RTP schemes (Öhrlund et al., 2019).

2.3.1.2 Customers' incentives to provide flexibility

Findings from a study conducted in Sweden show that two critical driving forces for

contributing with flexibility among customers are the will to contribute to the environment and the economic gain (Bartusch et al., 2014, 2011). Other studies also argue that the motives to contribute are not solely financial. The perceived social responsibility to act (Strengers, 2010), as well as doing something for the general good (Murtagh et al., 2014) are mentioned as additional motives. Further, in a later study Bartusch et al. (2018) found that the economic incentives were perceived by customers as somewhat limited. Bartusch et al. (2018) and Sahlén et al. (2019) both point out that the price signal from the grid tariff is hampered by its small size in relation to the total electricity costs, including VAT, energy tax and the electricity retail price. Svk suggests that grid tariffs, electricity retail prices and the energy tax should be coordinated in order to create

(34)

higher incentives for flexibility (Svk, 2015). It is further argued by EURELECTRIC (2013) that one unambiguous price signal must be given to customers in order to have a successful

involvement of customers.

The problem with contradicting price signals from the grid and the electricity trade is mentioned in several studies. Bartusch et al. (2018) evaluated an RTP capacity based tariff that could be implemented in order to reflect the current conditions in the grid more accurately. They analysed the correlation between consumption in Sandviken and the spot prices from Nord Pool and received an average correlation coefficient of 0.69 with significance level 0.01. Nonetheless, due to significant variations, they concluded that there is a risk of providing customers with

ambiguous price signals by introducing a grid tariff with RTP. Öhrlund (2016) compared a ToU capacity charge with spot prices from Nord Pool and concluded that the price signals are ambiguous in 28.7 % of the hours over three years. Alvehag et al. (2016) show similar problems with a three-part tariff. However, there is one study which suggests that the price signal from the grid tariff can become more potent if the customer chooses a variable electricity price, as the problems with grid congestion often coincide with cold weather and high electricity prices (Sahlén et al., 2019). However, this is not confirmed in the study. Bartusch et al. (2018)

emphasise that contradicting price signals is not a primary concern today for grid operators, as customers rarely choose variable electricity prices.

Bergaentzlé et al. (2019) emphasise that the relative share of grid tariffs in the final bill affect how grid tariffs can provide price signals for flexibility and that other charges such as taxes can limit the price signals and flexibility. High energy taxes and governmental charges are mentioned as an obstacle, especially in Sweden for utilising demand-side flexibility via grid tariff (Bartusch et al., 2018). As a grid tariff based on RTP combined with an hourly varying electricity retail price would to a large extent give opposing signals, Bartusch et al. (2018) evaluated a cost-based energy tax as an alternative way to enhance the price signal from the grid tariff. The study concluded that a cost-based energy tax creates a more dynamic price signal to the customer, but the problem with contradicting price signals remains as it is based on the spot price.

Studies indicate that customers have low responsiveness to changes in the electricity price. Based on a meta-analysis by Labandeira et al. (2017) the price elasticities of electricity in the empirical literature is on average −0.126 in the short-term and −0.365 in the long-term, where absolute values between 0 and 1 mean that the demand is inelastic. These estimations do not distinguish between different types of consumers which use electricity for different purposes and which affect the price on demand. In the study, it is explained that it is possible to further distinguish between residential, industrial and commercial consumers. The absolute values of the long-term elasticity are higher as consumers tend to react to price changes by implementing more energy- efficient equipment (Andruszkiewicz et al., 2020). Andruszkiewicz et al. (2020) declare that knowledge regarding price elasticity of electricity demand values is crucial in order to properly design grid tariffs, especially with regards to ToU tariffs and zonal tariffs.

(35)

2.4 Grid tariff design and distributed energy resources

Optimal use of distributed energy resources is stated to contribute to using the existing capacity of the grid more efficiently (Ruester et al., 2014). Schittekatte et al. (2018) further argue that if grid tariffs do not consider the rise of DER, the DSOs cost recovery, and a fair allocation of costs might be at risk. Included in the definition of DER are distributed generation, electric vehicles, energy storage and demand-side management. As the latter has been addressed above, this section will go through literature related to distributed generation, electric vehicles and energy storage.

2.4.1 Distributed generation

Focus in the literature has been on studying the emergence of distributed generation and its effect on grid tariff design. Distributed generation (DG) can be defined as the generation of electricity using different energy sources close to the load demands, in contrast to large power plants that are connected to the transmission grid (Perez-Arriaga, 2016). A. Picciariello et al.

(2015) argue that there is a need for change in tariff design to include DG. The regulatory principles for tariff design might need to change, and shift focus to principles such as sustainability and cost reflectiveness.

At low penetration levels, DG can lower network costs, including grid losses and investment in capacity. Connection costs and energy losses are vital components in the costs of distribution that can be related to the number of DG. Both recent and early studies show that DG can help lower losses in a network by placing capacity near the load. However, it has been found that system losses might increase at high penetrations rates. More recent technical studies

demonstrate that overall system losses might follow a trajectory by which the net benefits are achieved at a certain level of DG and diminish after this level (Cambini and Soroush, 2019). In a case study on DG in Swedish low voltage grids, it was found that grid losses decreased until a penetration level of 1kW per household, after which they started to increase. It was argued that at higher penetration levels, a majority of the generated electricity above 1kW is exported (Widén et al., 2010).

It is stated by Cambini and Soroush (2019) that one major challenge in grid tariff design is to compensate DG prosumers for their contribution to the environment and their economic benefits. They suggest implementing a three-part tariff for DG customers including a fixed charge reflecting the DG connection charges, an energy charge and a variable charge that represent grid losses. This tariff is argued to compensate DG customers for their contribution while they are also paying for their required investments, and result in an economic “degree of prosumption”.

Schittekatte et al. (2018) instead argue that energy charges over-incentivises solar PV-adoption.

With a tariff dominated by energy charges, customers with DG can lower their cost without necessarily lowering their costs inflicted on the grid. Customers with DG will not require less capacity because they still use the grid, especially during peak hours (EURELECTRIC, 2016).

They will consequently shift their induced network costs on customers who do not have DG,

Grid Tariff Design for Efficient Utilisation of the Distributor Grid: A qualitative study with actors on the Swedish electricity market

Grid tariff design for efficient utilisation of the distribution grid

A qualitative study with actors on the Swedish electricity market

MATILDA HAIKOLA MALIN SÖDERBERG

Master of Science Thesis

KTH School of Industrial Engineering and Management

Grid tariff design for efficient utilisation of the distribution grid

A qualitative study with actors on the Swedish electricity market

Matilda Haikola

Malin Söderberg

Abstract

Sammanfattning

Table of contents

List of tables

List of figures

Acronyms

Acknowledgements

1. Introduction

1.1 Background

1.2 Problem statement

1.3 Aim and research questions

1.4 Delimitations

1.5 Contribution

1.6 Outline

2 Literature and theory

2.1 Grid tariff structures

2.1.1 Structural elements

2.1.2 Time differentiation

2.1.3 Geographical differentiation

2.1.4 Composition of the grid tariff

2.2 Regulatory principles for distribution tariff design

2.3 Demand side flexibility and demand response

2.3.1 Price-based demand response via grid tariffs

2.4 Grid tariff design and distributed energy resources

2.4.1 Distributed generation