Link¨oping University — Department of Management and Engineering Master s thesis, 30 credits — M.Sc. Energy - Environment - Management Spring 2021 — LIU-IEI-TEK-A–21/04042-SE
Implementation of Battery Energy
Storage Systems in Residential Buildings
- A case study of a multifamily building in southern Sweden, exploring
profitability, self-sufficiency, and environmental performance.
Agnes Berg & Emelie Detert
Examiner: Patrik Rohdin / Link¨oping University
Company Supervisors: Iulia Minda, Adam Engstr¨om / E.ON Sweden
Supervisor: Stefan Blomqvist / Link¨oping University
June 12, 2021
Link¨oping University
SE-581 83 Link¨oping, Sweden
+46 013 28 10 00, www.liu.se
Abstract
Energy storage is of increasing interest as an enabler of incorporating renewable intermittent power in the power systems globally. There are several technologies for energy storage, and this thesis focuses on battery energy storage systems (BESS). Previous research has shown that it is difficult to install BESS with a payback time within the battery lifetime, making it a challenge to realise profitable investments. The complexity of developing an optimal control of the battery is also documented in research as an-other challenge. Optimal sizing of the BESS could be a solution to the challenge of reaching profitabil-ity. The thesis is identifying and analysing some important technical and energy-related parameters affecting the performance of BESS installations. Identification and analysis of parameters affecting the performance will help build insight into the optimization of BESS and help enable the development of more efficient sizing and operation. By developing an algorithm simulating the BESS when con-trolled using two different strategies, this thesis additionally contributes to the research by displaying the complexity of battery control, which is realised by the energy management system (EMS). Thereby the thesis is adding to the research base for the future development of smarter and more optimal EMS. The main research methodologies used in the thesis was a literature study and a case study. The re-sults suggested that the energy management strategy used in the battery control was gravely affecting the performance in terms of economic profitability, self-sufficiency and environmental impact. It was also implied that it is difficult to develop an efficient battery control to reach the full potential of the storage system. The main conclusions in this paper are that the most important parameters to con-sider when implementing a battery storage in a residential multifamily building are battery technology, battery capacity, building load, renewable energy generation, energy management strategy as well as the electricity prices and investment cost. The energy management strategy most favourable for the case building studied was found to be a combination of optimizing the self-sufficiency and performing peak shaving. It would also be preferable to further develop the battery control to also take electricity prices and balance services into consideration. For this, AI and machine learning could be integrated in the control of the system. According to the case study results, the lithium ion battery technology had better potential for reaching economic profitability while the nickel metal hydride technology showed better potential in terms of environmental performance. The choice of battery technology and energy management strategy should however be adjusted to the customer specific demands and prerequisites.
Preface
With this master thesis we complete our Master of Science in Energy — Environment – Management at Link¨oping University. The purpose has been to contribute to the pursuit of enabling improved BESS installations and thereby support the energy transition towards a fossil free power system in Sweden. We would like to take the opportunity to thank the Local Balancing Team and Julia Brunfeldt at E.ON. A special thanks to our supervisors Adam Engstr¨om at E.ON Energy Distribution and Iulia Minda at E.ON Customer Solutions for being so supportive and for giving us a warm welcome into the team, inviting us to discussions and meetings. For this we are very grateful. We would also like to express our gratitude towards our supervisor Stefan Blomqvist, our examiner Patrik Rohdin and our opponents at Link¨oping University for great support and interesting discussions. Additionally, we would like to address our appreciation towards everybody else helping us realizing this thesis and our family and friends. Thank you for always setting time aside to support us in ups and downs.
It has been a pleasure to realise this project together and during the journey we have learned a lot from each other. We now look forward to bring the insights we have gained with us as we begin our professional lives.
Agnes Berg & Emelie Detert Link¨oping, June 12, 2021
Contents
1 Introduction 1 1.1 Background . . . 1 1.2 Aim . . . 2 1.2.1 Research questions . . . 2 1.3 Delimitations . . . 21.4 Outline of the Report . . . 4
2 Theory 5 2.1 The Swedish Energy and Power System . . . 5
2.2 The Swedish Electricity Market . . . 8
2.2.1 Electricity Price Development . . . 8
2.2.2 End-user electricity price . . . 9
2.3 Economic and administrative Instruments . . . 10
2.4 Solar Power - Photovoltaics . . . 12
2.5 Energy Storage Systems (ESS) . . . 13
2.6 Battery technologies . . . 13
2.6.1 Important Characteristics . . . 14
2.6.2 Lithium-Ion Batteries . . . 15
2.6.3 Nickel-Metal Hydride Batteries (Ni-MH) . . . 15
2.6.4 Sodium Sulfur Batteries . . . 16
2.7 Battery Energy Storage System (BESS) . . . 16
2.7.1 Battery Energy Storage System Applications . . . 17
2.7.2 Components of the Battery Energy Storage System . . . 18
2.8 Sizing of Battery Energy Storage Systems in literature . . . 20
2.9 Environmental aspects of the Power System . . . 22
3 Case Study Description 25 4 Methodology 27 4.1 Pre-study . . . 27 4.1.1 Literature Study . . . 28 4.1.2 Interviews . . . 29 4.1.3 Parameter Analysis . . . 30 4.2 Case Study . . . 30
4.2.1 Planning of Case study . . . 31
4.2.2 Data collection . . . 31
4.2.3 Development of MATLAB Algorithm . . . 31
4.2.4 Simulations of Battery Energy Storage System Implementation . . . 33
4.2.5 Calculations . . . 35
4.2.6 Data Interpretation . . . 36
4.3 Development of Battery Design Tool . . . 37
4.4 Methodology Critique . . . 37 5 Results 39 5.1 Pre-study: Interview . . . 39 5.2 Pre-study: Analysis . . . 41 5.2.1 Technical parameters . . . 41 5.2.2 Economic parameters . . . 43
5.2.3 Energy related parameters . . . 43
5.3 Visualisation of Case building . . . 44
5.4 Initial simulation: Implementation of battery storage . . . 50
5.4.1 Self-sufficiency strategy . . . 51
5.4.2 Peak shaving strategy . . . 54
5.5 Parameter Variations . . . 57
5.5.1 Power Target . . . 57
5.5.2 C-rate . . . 57
5.5.3 No night charge . . . 59
5.6.1 Photovoltaic Production - decrease and increase . . . 61
5.6.2 Load increase . . . 63
5.6.3 Electricity price increase . . . 65
5.6.4 Investment cost reduction . . . 66
5.7 Design Tool . . . 67
6 Analysis of Results 69 6.1 Case building . . . 69
6.2 Summary of impact on battery storage performance . . . 69
6.3 Self load rate . . . 70
6.4 Degree of self-sufficiency . . . 71
6.5 Degree of peaks shaved . . . 72
6.6 Carbon dioxide savings . . . 72
6.7 Profitability: Payback time and Net present value . . . 73
6.8 Battery selection . . . 74
6.9 Comparison of tool and case . . . 74
7 Discussion 76 8 Conclusion 80 9 Future studies 81 10 Appendix 91 10.1 Interview Protocol . . . 91
10.2 Electricity prices, E.ON 2019, SE 4 . . . 92
10.3 Case Study Results . . . 92
List of Figures
1 The Energy System in Sweden divided into three steps, Supply, Transformation &
Trans-mission, and Consumption from The Swedish Energy Agency [23]. . . 5
2 The Power System in Sweden from Svenska Kraftn¨at [28]. The system is divided into electricity producers, power grids, electricity consumer, electricity market and trading companies. . . 6
3 The share of renewable electricity in relation to total electricity generation 1990-2018, in per cent from The Swedish Energy Agency [23]. . . 7
4 Average weekly price (system price) on Nord Pool Spot 1996-2016 from the Royal Swedish Academy of Engineering Sciences [45]. . . 8
5 Spot price development 1996 to end of august 2020, with expected future prices, from Sweden Energy [47]. . . 9
6 Daily variation of global solar radiation in Lund 1989-2014. The solid line show average values for each hour of the day for January, April, July and October. The dashed line show the lower and upper limits for the 95th percentile. [73] . . . 12
7 Schematic figure of a battery energy storage system. The electricity generated by the PV modules are transferred through the PV inverter, to the building (load), the battery or fed into the grid. The battery can be charged by the PV power or by the grid and discharged to supply the building load. The power flows are measured by meters, and the information is send to the Energy Management System (EMS) and the cloud-platform. 17 8 Schematic picture of an AC coupled BESS connected to PV-modules, inspired by B.D. Olaszi et. al. [103]. In AC coupled systems the battery is connected to the PV system (PV-modules and inverter) via a charge regulator and a battery inverter on the AC side of the PV inverter. Hence, two inverters are needed which result in more conversion losses than in DC-coupled systems. . . 18
9 Schematic picture of a DC coupled BESS connected to PV-modules, inspired by J. Weniger et. al. [102]. The DC coupled battery energy storage system is connected to the DC-link of the PV inverter. In this way, only one inverter is needed and conversion losses are lower than in AC-coupled systems. . . 18
10 Self-sufficiency and peak shaving strategy applied on a net load curve. The battery is charged when the net load is below zero for both control strategies. Instead the bat-tery is discharged when the net load is above zero for self-sufficiency (light grey box) and above the target for peak shaving algorithm (bronze coloured box). . . 20
11 A simplified overview of the process of the master thesis. First, the aim and the purpose of the thesis were set. Then the pre-study covering a literature study and interviews fol-lowed. Important parameters regarding BESS installations were identified in the last step of the pre-study. The case study consisted of a data collection phase, development of MATLAB algorithms, simulations and lastly calculations and data interpretation of the simulations results. Finally, a battery design tool was developed and conclusions and discussion could be finalised. . . 27
12 The Self-sufficiency algorithm controls the hour, the net load, and the battery level and takes decision to charge or discharge the battery based on this. . . 32
13 The Peak Shaving algorithm controls the hour, the net load and the battery level and makes decisions to charge, discharge or stay in idle mode (no charge or discharge) based on this. . . 32
14 Battery system with important parameters for calculations in MATLAB on the system. . 33
15 Overview of all simulations performed in the case study, divided into initial simulations, parameter variations and sensitivity analysis. n is referred to the number of simulations performed. . . 34
16 Interview answers from theme 1 - experiences of the participants. . . 39
17 Interview answers from theme 2 - details about sizing. . . 40
18 Interview answers from theme 3 - dispatch strategies. . . 40
19 Interview answers from theme 4 - Results & Outcomes. . . 41
20 Energy profile in 2019 showing the total electricity demand, i.e. load (in blue) for the case building and the total electricity production from its photovoltaic panels (in red). . . 44
21 Net load profile in 2019, showing the difference between load and production. When be-low zero there is potential for storing energy. . . 45
22 Duration diagram for the overproduction in 2019. The overproduction is sorted from largest to smallest value of produced solar electricity during the entire year. It can be
seen that out of the 8760 hours of the year, overproduction occurred for below 1900 hours. 45 23 Energy profile in summer, week 28-31 in 2019, showing the total electricity demand, i.e
the load (in blue), for the case building and the total electricity production from its pho-tovoltaic panels (in red). . . 46 24 Energy profile in winter, week 2-5 in 2019, showing the total electricity demand, i.e. load
(in blue) for the case building and the total electricity production from its photovoltaic
panels (in red). . . 46 25 Net load for the week with the day with maximum electricity production in 2019, 5th
July. Showing the difference between total load and total production. . . 47 26 Energy profile for the week with the day with maximum electricity production in 2019,
5th July. Showing the total electricity demand for the case building and the total
elec-tricity production from its photovoltaic panels. . . 47 27 Energy profile in the day with maximum electricity supply in 2019, 4th December.
Show-ing the total electricity demand for the case buildShow-ing (blue) and the total electricity pro-duction (red). It can be noted that there is a peak in the morning and three more peaks of increasing power throughout the day and in the evening. . . 48 28 Energy profile in the day with maximum electricity supply in 2019, 4th December.
Show-ing the total electricity demand for the case buildShow-ing (blue) and the total electricity pro-duction (red). It can be noted that there is a peak in the morning and three more peaks of increasing power throughout the day and in the evening. . . 49 29 Cost function for battery and inverter installation per installed kWh of Li-ion and
Ni-MH battery. y1 (light blue) is the function for the NI-Ni-MH battery, y2 (dark blue) the
function for Li-ion battery. . . 50 30 Self load rate, s, (in blue) and degree of self sufficiency, d, (in red), depending on battery
capacity. Same results for both battery technologies when using self-sufficiency energy
management strategy. . . 51 31 Carbon dioxide savings as a function of the battery capacity, increasing with increasing
capacity. Same results for both battery technologies when using self-sufficiency energy
management strategy. . . 51 32 Payback time (PBT) as a function of the battery capacity in blue for Li-ion battery and
red for Ni-MH battery. . . 52 33 Net present value (NPV) as a function of the battery capacity in blue for Li-ion battery
and red for Ni-MH battery. . . 52 34 State of Charge (SoC) in the battery when using self-sufficiency energy management
strategy and the 10 kWh batteries. The SoC is displayed over the summer week with the highest PV production during 2019. The energy profile for the same period is shown in Figure 25 and it can be concluded that the load is not so high that the battery gets
immediately discharged, resulting in the flat tops seen in this figure. . . 53 35 State of Charge (SoC) in the battery when using self-sufficiency energy management
strategy and the 10 kWh batteries. The SoC is displayed over the winter week with the highest load during 2019. The energy profile for the same period is shown in Figure 26 and it can be concluded that the load is so high that the battery gets immediately
dis-charged after being dis-charged at night, resulting in the peak formed tops seen in this figure. 53 36 Self load rate, s, and degree of self sufficiency, d, depending on battery capacity. Higher
values are naturally seen for s than for d. . . 54 37 Carbon dioxide savings when implementing Li-ion battery (blue line) and Ni-MH (red
line) using the peak shaving energy management strategy. . . 54 38 Payback time for BESS installation with Li-ion battery (blue) and Ni-MH battery (red).
Increasing with increasing capacity, reaching a maximum for 70-80 kWh batteries. . . 55 39 Net present value for BESS installation with Li-ion battery (blue) and Ni-MH battery
(red). Decreasing with increasing capacity, looking as if reaching a minimum for Ni-MH. . 55 40 State of Charge (SoC) in the 10 kWh Li-ion battery when using peak shaving energy
management strategy. The SoC is displayed over the summer week with the highest PV production during 2019. The energy profile for the same period is shown in Figure 25 and it can be concluded that the load is not so high that the battery gets immediately discharged, resulting in the flat tops seen in this figure. However, the PV production around hour 4480 is clearly not enough to reduce the load peak occurring shortly af-ter, hence the battery is not discharged during the coming, lower load peak the next day since the power target was set above that level. . . 56
41 State of Charge (SoC) in the 10 kWh Li-ion battery when using peak shaving energy management strategy. The SoC is displayed over the winter week with the highest load during 2019. The energy profile for the same period is shown in Figure 26 and it can be concluded that the battery is not able to reduce the load peak occurring the fifth of December at about hour 8120 and that the power target must have been adjusted so to be higher than the coming load peaks over the week which result in the battery not
discharging and therefore holding the maximum State-of-Charge throughout the week. . . 56 42 Self-sufficiency rate (s) for varying C-rate with the two battery technologies. . . 58 43 Degree of self-sufficiency (d) for varying C-rate with the two battery technologies. . . 58 44 Payback time for varying C-rate with Li-ion battery (blue line) and Ni-MH battery (red
line) of 10 kWh. . . 59 45 Net present value for varying C-rate with Li-ion battery (blue line) and Ni-MH battery
(red line) of 10 kWh. . . 59 46 Yearly carbon dioxide savings for varying C-rate using peak shaving strategy. . . 59 47 Self load rate, s, and degree of self sufficiency, d, depending on battery capacity for
self-sufficiency strategy when the battery is not charged from the grid. . . 60 48 Self load rate, s, and degree of self sufficiency, d, depending on battery capacity for peak
shaving strategy when the battery is not charged from the grid. . . 60 49 Carbon dioxide emission savings when using self-sufficiency strategy and not charging
the battery from the grid. . . 60 50 Carbon dioxide emission savings when using peak shaving strategy and not charging the
battery from the grid. . . 60 51 Net present value (NPV) when using self-sufficiency strategy and not charging battery
from grid during the night. . . 61 52 Net present value (NPV) when using peak shaving strategy and not charging battery
from grid during the night. . . 61 53 Self load rate, s, as function of photovoltaic production for different Li-ion battery
ca-pacities. . . 61 54 Self load rate, s, as function of photovoltaic production for different Ni-MH battery
ca-pacities. . . 61 55 Degree of self-sufficiency, d, as function of photovoltaic production for different Li-ion
battery capacities. . . 62 56 Degree of self-sufficiency, d, as function of photovoltaic production for different Ni-MH
battery capacities. . . 62 57 Carbon dioxide savings as function of photovoltaic production for different Li-ion battery
capacities. . . 62 58 Carbon dioxide savings as function of photovoltaic production for different Ni-MH
bat-tery capacities. . . 62 59 Net present value as function of photovoltaic production for different Li-ion battery
ca-pacities. . . 63 60 Net present value as function of photovoltaic production for different Ni-MH battery
ca-pacities. . . 63 61 Self load rate, s, as function of the load for different Li-ion battery capacities. Ss stands
for self-sufficiency, and ps for peak shaving. . . 63 62 Self load rate, s, as function of the load for different Ni-MH battery capacities. Ss stands
for self-sufficiency, and ps for peak shaving. . . 63 63 Degree of self-sufficiency, d, as function of the load for different Li-ion battery capacities.
Ss stands for self-sufficiency, and ps for peak shaving. . . 64 64 Degree of self-sufficiency, d, as function of the load for different Ni-MH battery capacities. 64 65 Carbon dioxide savings as function of the load for different Li-ion battery capacities. . . . 64 66 Carbon dioxide savings as function of the load for different Ni-MH battery capacities. . . 64 67 Net present value (NPV) as function of the load for different Li-ion battery capacities. . . 65 68 Net present value (NPV) as function of the load for different Ni-MH battery capacities. . 65 69 Payback time (PBT) depending on battery capacity for a 20 % electricity spot price
in-crease. . . 65 70 Net present value (NPV) depending on battery capacity for a 20 % electricity spot price
increase. . . 65 71 Payback time (PBT) depending on battery capacity for a 30 % decrease in investment
cost. . . 66 72 Payback time (PBT) depending on battery capacity for a 50 % decrease in investment
73 Net present value (NPV) depending on battery capacity for a 30 % decrease in
invest-ment cost. . . 67 74 Net present value (NPV) depending on battery capacity for a 50 % decrease in
invest-ment cost . . . 67 75 Tool interface with case building data input marked in neon yellow color. . . 68 76 Tool interface with available battery-inverter packages showing investment cost and
yearly savings needed for payback within battery life. . . 68 77 Electricity cost savings as a function of battery capacity, regardless of battery technology. 92 78 Electricity cost savings with Li-ion battery (blue line) and Ni-MH battery (red line). . . . 92 79 PBT when using self-sufficiency strategy and not charging battery from grid. . . 94 80 PBT when using peak shaving strategy and not charging battery from grid. . . 94 81 Self load rate, s, and degree of self sufficiency, d, depending on battery capacity for a 20
% electricity spot price increase. . . 94 82 CO2savings depending on battery capacity for a 20 % electricity spot price increase. . . . 94
83 s and d depending on battery capacity for a 30 % decrease in investment cost. . . 95 84 s and d depending on battery capacity for a 50 % decrease in investment cost. . . 95 85 CO2savings depending on battery capacity for a 30 % decrease in investment cost . . . . 95
86 CO2savings depending on battery capacity for a 50 % decrease in investment cost. . . 95
87 Electricity cost savings when using self-sufficiency strategy and not charging battery
from grid. . . 95 88 Electricity cost savings with Li-ion battery (blue line) and Ni-MH battery (red line)
when using peak shaving strategy and not charging battery from grid. . . 95 89 Maximum peak in supplied power from the grid with Li-ion battery and a target of 8
kWh, as used in the initial case simulations. . . 96 90 Maximum peak in supplied power from the grid with Ni-MH battery and a target of 8
kWh, as used in the initial case simulations. . . 96 91 Maximum peak in supplied power from the grid with Li-ion battery and a target of 0 kWh 96 92 Maximum peak in supplied power from the grid with Li-ion battery and a target of 16
kWh . . . 96 93 Maximum peak in supplied power from the grid with Ni-MH battery and a target of 0
kWh . . . 96 94 Maximum peak in supplied power from the grid with Ni-MH battery and a target of 16
kWh. . . 96 95 Maximum peak in supplied power from the grid without battery. . . 97
List of Tables
1 Comparison of the different battery technologies covered in the theory. * No self-discharge
mechanism of the active materials [92]. . . 16
2 Case building specifications. . . 25
3 Technical specification of the and electrical system in the case study. . . 25
4 Electricity supply contracts costs and grid contract costs. 25 % VAT is added on all costs. 25 5 Battery packages available at E.ON customer solutions, i.e. battery size (capacity) with battery power and corresponding inverter power and price per installed kWh of storage. . 26
6 Search strategies used in the literature study. In the selection process, relevant articles were chosen based on scientific value, credibility and actuality after reading the title and abstract. . . 29
7 Interview questions. . . 30
8 Inputs that are typed into the Excel battery design tool. . . 37
9 Outputs that are given from the Excel battery design tool. . . 37
10 The technical, economical and energy related parameters identified by the writers of the master thesis when analysing the theory from Chapter 2.8 and the interview results from Chapter 5.1. The parameters identified as being of extra interest and thus further stud-ied in the case study simulations are marked in yellow. . . 41
11 Variable electricity and network (grid) cost for the case building during 2019 estimated from data of supplied electricity and assumed electricity contract and grid contract. . . . 49
12 Case building parameters in 2019, later used to compare with battery implementation. s is referred as the self load rate, d as the degree of self-sufficiency. . . 50
13 Batteries simulated in the case study with capacity, name used in the simulations, and total investment cost. . . 51
14 Reduction of power peaks in percent with the implementation of battery with a self-sufficiency strategy compared to without battery. The simulation results with self-self-sufficiency strategy were identical for Li-ion and Ni-MH battery technologies. . . 52
15 Results from battery storage implementation with increasing capacity using the self-sufficiency strategy. s is referred as the self load rate, d as the degree of self-self-sufficiency. Further, PBT is the payback time and NPV the net present value. . . 54
16 Reduction of power peaks in percent with the implementation of a battery with a peak-shaving strategy compared to the case without battery for both technologies. The bold marked values are the ones with best reduced peaks. . . 55
17 Results from battery storage implementation with the peak-shaving strategy. S is re-ferring to the self load rate and d to the degree of self-sufficiency. Further, PBT is the payback time and NPV the net present value. . . 57
18 Reduction of power peaks in percent with the implementation of battery with a peak-shaving strategy for different target powers, compared to the case without battery. . . 57
19 Results of change in power target (0 kWh and 16 kWh) for the peak-shaving strategy. The change (%) is referred to as the change compared to peak-shaving with 8 kWh tar-get. S is the self load rate, d is the degree of self-sufficiency, PBT is the payback time and NPV the net present value. . . 57
20 Reduction of power peaks in percent with the implementation of battery with a peak-shaving strategy with different C-rates, compared to the case without battery. . . 58
21 Results of battery storage implementation with the peak-shaving strategy for the small-est batteries (10 kWh) and varied C-rates. The change (%) is referred to as the change compared to the initial cases with C-rate 0.5 and 0.3. S is the self load rate, d the degree of self-sufficiency, PBT is the payback time and NPV the net present value. . . 59
22 Payback time (PBT) and net present value (NPV) with a 20 % increase in electricity prices. Ss stands for self-sufficiency and ps for peak shaving. Further, PBT is the pay-back time and NPV the net present value. . . 66
23 Payback time (PBT) and net present value (NPV) with a 30 % decrease in investment cost. Ss stands for self-sufficiency and ps for peak shaving. . . 67
24 Payback time (PBT) and net present value (NPV) with a 50 % decrease in investment cost. Ss stands for self-sufficiency and ps for peak shaving. . . 67
25 Summary on how parameters affecting the different indicators for self-sufficiency algorithm. Parameters improving the indicators are marked with green (from light green -low improvement, to dark green - high improvement). Instead, parameters that worsen the indicators are marked with yellow (low impact), orange (moderate impact), and red (high impact). . . 70 26 Summary on how parameters affecting the different indicators for the peak-shaving
al-gorithm. Parameters improving the indicators are marked with green (from light green - low improvement, to dark green - high improvement). Instead, parameters that worsen the indicators are marked with yellow (low impact), orange (moderate impact), and red (high impact). . . 70 27 The proposed battery-inverter packages for the case building if a battery energy storage
system was to be installed. The most economic profitable batteries for each technology was chosen since the self-sufficiency and carbon dioxide savings were still improved and the payback time and net present value for larger batteries were considered completely indefensible. The range in the results is the result from different energy management strategies. The peak shaving gave better results for payback time (PBT) and net present value (NPV) while the self-sufficiency gave better results for self-load rate (s), degree of self-sufficiency (d) and carbon dioxide savings. . . 74 28 Electricity prices, E.ON. Variable for SE4 2019. . . 92 29 Results of battery storage implementation with the self-sufficiency strategy with a
in-crease and dein-crease of production.The change (%) is referred as the change compared to the initial case. For the max production cases, the pay-back time were negative, and
were thus discarded. . . 93 30 Results of battery storage implementation with the peak-shaving strategy with a
in-crease and dein-crease of production. The change (%) is referred as the change compared to the initial case. For the max production cases, the pay-back times were negative, and were thus discarded. . . 93 31 Results of battery storage implementation with the self-sufficiency strategy with a 5
per-cent increase of the load. The change (%) is referred as the change compared to the ini-tial case. . . 93 32 Results of battery storage implementation with the peak-shaving strategy with a 5
per-cent increase of load. The change (%) is referred as the change compared to the initial
case. . . 93 33 Results of battery storage implementation with the self-sufficiency strategy with a 10
percent increase of load. . . 94 34 Results of battery storage implementation with the peak-shaving strategy with a 10
per-cent increase of load. The change (%) is referred as the change compared to the initial
List of Nomenclature
CIF: Total cash inflow [SEK] Cinv: Investment cost [SEK] C: Charge [Q]d: Degree of self-sufficiency [%] DOD: Depth of Discharge [%]
EBC: Electricity charged into the battery [kWh] EBC: Electricity discharged from the battery [kWh] ELoad: Electricity consumed in the building [kWh] Enetload: Electricity net load [kWh]
EP V: Electricity produced from PV [kWh] n: Lifetime [Years]
N P V: Net present value [SEK] P BT: Payback-time [Years]
P V AF: Present value annuity factor [-] r: Interest rate [%]
s: Self load rate [%] SOC: State-of-Charge [%] t: Time [s]
1 Introduction
In the following chapter, the background and the motivation to the thesis is presented. This is followed by the purpose, delimitations and the outline of the report.
1.1 Background
The global demand for energy and energy related services to satisfy human social development, eco-nomic development, welfare and health is increasing [1] and energy-related carbon dioxide emissions are significant, representing two-thirds of all greenhouse gases [2]. This development raises concerns regarding the possibility to reach crucial sustainability goals, such as keeping the temperature rise be-low 2 °C, in accordance with the Paris Agreement [3]. Renewable energy sources are associated with opportunities for mitigating climate change [1], and is regarded as one of the most effective and effi-cient solutions to address the issues of oil depletion, carbon dioxide emissions and increasing energy demand [4]. Measures are taken to mitigate climate change on national and international level, by po-litical and business leaders as well as by people rising in massive demonstrations [5] to put pressure on world leaders. National and international goals are aiming to move towards a fossil free society with increased usage of renewable energy sources. The sustainable development goals (SDGs), adopted in the 2030 Agenda for Sustainable Development in 2015 by all United Nations Member States are a set of 17 goals calling for action by all countries [6]. The EU strategy is in line with the aim to become cli-mate neutral by 2050 [7] and the European 2030 Clicli-mate Target Plan is the Commission’s proposal to cut greenhouse gas emissions by at least 55 percent by 2030. This increased ambition to cut greenhouse gas emissions reflects in the renewable energy directive with the binding renewable energy target of at least 32 percent in 2030 [7]. The directive includes new provisions for enabling self-consumption of re-newable energy [7] and the Swedish government has set the ambitious climate goal to achieve zero net greenhouse gas emission by 2045 [8]. The strategy used to reach the climate goal is heavily dependent on general economic instruments in combination with targeted efforts to support development and in-troduction of new technology [9].
The negative environmental impact on the globe is forcing countries to replace fossil fuels with re-newable energy sources [10]. Technological innovation, particularly in the field of rere-newable energy, is viewed as an enabler of the energy transition [2]. Of all electricity worldwide, 25 percent was produced from renewable sources in 2017 [2] and in 2019 it reached almost 27 percent which is the highest level ever recorded. However, this is not enough to meet the sustainable development scenario [11] and the world is not on course to reach the SDGs related to energy, based on existing and announced policies [3]. Additionally, the carbon dioxide emissions are not decreasing fast enough to reach the Paris Agree-ment [3]. Nonetheless, renewable energy sources are increasing significantly and had the highest rate of generation growth in 2019 compared to other energy sources. The share of renewable power gener-ation in Sweden increased from 11 percent to 17 percent in 2020, corresponding to an overall increase in renewable production even though the total power generation decreased during 2020 [12]. There has also been a rapid increase in photovoltaic installations globally, due to national subsidies and decreased prices [13]. More specifically, building-applied photovoltaic systems is an important market segment, offering the opportunity for self-consumption of electricity generation [13]. Roof top mounted photo-voltaic systems have been made economically interesting in Sweden in the recent years due to a rapid decrease in photovoltaic module costs. Especially, multifamily houses with large roof top area make large-scale photovoltaic systems with lower installation costs possible and are particularly interesting for potential prosumers, i.e., consumers that can also be producers [14].
When integrating renewable energy into the power grid, a major issue is its intermittent and stochas-tic characterisstochas-tics [4]. For photovoltaics aimed to supply residential buildings, there is also a distinctive timing issue with a mismatch of photovoltaic generation and load in the building ??. The photovoltaics generate electricity in the middle of the day, whereas, the highest loads occur in the morning and in the evening. Renewable energy production is moreover, strongly dependent on local weather and cli-mate conditions. The power system in Sweden is changing from being centralized and stable towards a system with the integration of smaller and geographically dispersed generation, needing to be more flexible [15]. Supplying momentarily demand in kilowatts can be a challenge in a system like this. Ar-eas where the transmission grid is undersized and can not transmit enough power are particularly chal-lenged. The southern distribution area (SE4) in Sweden is such an area, making it dependent on dis-tribution from other parts of the country and from neighbouring countries [15]. An increased wish for electricity self-sufficiency by self produced photovoltaic power been noticed [16] and is particularly mo-tivated in areas with undersized grids. Peak power plants are used whenever the load in the power
sys-tem is high and are used as backup generation when the load is low to secure electricity supply. This is however, associated with high costs and high emissions of CO2eq [17].
Energy storage can stabilise fluctuations in demand and supply, helping to balance the electricity grids and thus play a key role in future power systems with more renewable power generation [18]. Battery energy storage systems (BESS) have been broadly accepted as one of the potential solutions to the challenges imposed by the intermittent and fluctuating features of solar and wind energy [4]. Studies have also shown that storage can increase the benefit of photovoltaic production, even in small scale [19]. There are several EU initiatives on batteries, underlining the importance that battery technol-ogy will have for electrification, which is viewed as one of the main tools to reach decarbonisation [18]. The battery energy storage system market is further expected to grow with a compound annual growth rate of 33 percent from 2019 to 2024 with promising opportunities in the residential, non-residential and utility industries. Major market drivers are the increasing demand for grid-connected solutions, the high demand for lithium-ion technology in the renewable energy industry and declining prices of lithium-ion batteries [20]. As the number of business and private power prosumers with installed pho-tovoltaic production increase, also on the Swedish market, so does the interest for storage applications. In addition, the congestion problem experienced, especially in Sk˚ane, increases the demand for self-sufficiency [15]. A development towards higher self-self-sufficiency in both industries and residential clusters producing and storing their own power could help reduce the congestion problems and support the fu-ture energy supply safety. The use of batteries is related to a intense use of energy and materials in the extraction and production phase of the life cycle. This puts pressure on an efficient use of materials and batteries [21]. The economic concern of battery systems, nevertheless, remains the major barrier to be overcome before battery energy storage systems can be a mainstream storage solution in the energy sector [22] [4]. Optimizing the size and using the right operation of the battery energy storage systems is therefore a vital issue in order to balance the trade-off between using battery energy storage systems to improve renewable energy system performance and to achieve profitable investment [4].
1.2 Aim
The master thesis aims to investigate important aspects of battery energy storage system sizing in res-idential buildings. A case study of a resres-idential multifamily building was performed as the major part of the master thesis in order to fulfill this aim. By identifying and analysing the parameters that are affecting the performance of the battery energy storage system, the aim was also to help build insight into the optimization of the systems. Contribution to the development of more efficient sizing and op-eration of battery energy storage systems will thereby be given by this thesis.
1.2.1 Research questions
The following research questions were set up to be answered in the study to fulfil the aim of the thesis. 1. Which are the most important parameters to consider for optimal performance when
implement-ing a battery energy storage systems in a residential multifamily buildimplement-ing?
2. Which Energy management strategy is most favourable in terms of economic profitability, self-sufficiency and environmental performance for battery energy storage system implementation in a multi-family building like the one studied in the case?
3. Which battery technology is most favourable in terms of economic profitability, self-sufficiency and environmental performance for battery energy storage system implementation in a multi-family building like the one studied in the case?
1.3 Delimitations
The performance evaluated in the thesis was restricted to economomic profitability (payback-time and net present value), self-sufficiency (self-load rate, and degree of self-sufficiency) and environmental im-pact (carbon dioxide savings). Moreover, the report was limited to investigate residential multifamily buildings. More specific, one residential multifamily building with photovoltaic installations. This cus-tomer segment was seen as relevant since photovoltaic installations are significant in multifamily build-ings and that cost saving possibilities when adding a battery storage are promising. The case data was taken from 2019 since the electricity consumption in 2020 was considered abnormal due to the COVID-19 pandemic and data from previous years was not available. For the profitability the system boundary was set to include only the purchase price of battery and cost for electricity supply without taking end-of-life costs or incomes from fed-in electricity into consideration. In the environmental calculations, the
system boundary was set to the carbon dioxide emissions from production of electricity, excluding the emissions from raw material extraction, production, transportation and end of life of the battery and the photovoltaic system. The batteries included in the case and the tool, were set to two different tech-nologies in different sizes; lithium-ion batteries (Li-ion) and nickel-metal hydride batteries (Ni-MH). The lithium-ion batteries were of two different types, depending on the size, offered by E.ON. No other battery or energy storage technologies were used since the chosen ones currently have the highest po-tential for residential buildings. Moreover, the battery was sized and evaluated without taking the dy-namic aging into account, due to the complicity of the modelling. Additionally, the installed maximum power of the photovoltaic panels was assumed as fixed and the energy management strategy of the bat-teries limited to self-sufficiency and peak shaving. The algorithms for self-sufficiency and peak shaving were lastly, not optimized, due to a very complex structure and were thus out of scope for the thesis.
1.4 Outline of the Report
1.0 IntroductionIn the introduction chapter, the background to the thesis is presented, in order for the reader to un-derstand the importance of the studied field. The chapter also includes the research questions that are the basis for the thesis.
2.0 Theory
The theory chapter serves to present a background to the theoretical context for the study. This in-cludes, for example, an introduction to the power system, the Swedish electricity market, economic and administrative instruments, photovoltaics, and energy storage systems.
3.0 Case Object Description
In this chapter, the case object and the batteries used in the study are described. 4.0 Methodology
The methodology used for the thesis is described and motivated in the chapter. This includes the re-search method, the literature study, the data collection, the case study and tool development. More-over, the credibility of the research is discussed.
5.0 Results
The results are presented in this chapter. Firstly, the outcomes of the pre-study is presented. This consists of the interview results and parameter analysis. After this, the visualisation of the case build-ing energy profile is displayed. It is followed by the case study simulation results of implementbuild-ing bat-tery storage. The results from simulations of parameter variation and sensitivity analysis is then given. Lastly, the results of the development of a battery design tool is presented.
6.0 Analysis of Results
In this chapter, the results from the simulations in the pre-study and case study are analyzed and com-pared with literature findings. The analyse is concentrated on how the battery energy storage system performance was affected by different parameters such as battery capacity and building load.
7.0 Discussion
In this chapter a discussion about the findings of the pre-study and case study is developed. Addi-tionally, other interesting topics regarding battery energy storage systems and battery sizing are dis-cussed. Some further discussion about the comparison between the battery energy storage system siz-ing through simulations, as in the case study, and simple sizsiz-ing, as with the tool, is also presented. 8.0 Conclusion
The chapter answers the research questions and present the final conclusions from the study. 9.0 Future studies
2 Theory
This chapter gives some theoretical background about the Swedish power system and electricity market with economic and administrative instruments regulating it. Some theory on photovoltaic power genera-tion and energy storage follows before some theory on battery technologies. Some definigenera-tion and theory of battery energy storage systems (BESS) is then presented, after which a summary of BESS in litera-ture is given. At the end of the chapter, environmental impact of the power system and of batteries is presented.
2.1 The Swedish Energy and Power System
The Swedish Energy system consists of three main steps; supply, transformation & transmission, and lastly consumption of energy, as in Figure 1. The system needs to always be in balance, namely the supply must be equal to the energy use and losses. Since mid-1980, the total energy supplied has been more or less the same at around 550-600 TWh per year [23]. However, the supply from fossil energy sources has been cut in half from the 1970s. The final energy usage in 2018 was 373 TWh, with a to-tal supply of 552 TWh. The Swedish energy system uses different energy carriers, e.g., electricity, bio fuels, petroleum products, district heating, coal, and natural gas. Electricity has the largest share of final energy use per energy carrier [24]. The total energy generation in 2020 decreased with 5.6 TWh, compared to the previous year, which can be linked to the COVID-19 pandemic leading to lowered op-eration in industries and less traveling [25].
Figure 1: The Energy System in Sweden divided into three steps, Supply, Transformation & Transmission, and Consumption from The Swedish Energy Agency [23].
The power system is the energy carrying part of the energy system. It is complex due to electric power being consumed at the same moment as it is produced. In order for the system to be in balance there has to be enough electricity and if it is not consumed where it is produced, it also has to be transferred to the user [26]. As in Figure 2, the system consists of electricity producers, the grid with three differ-ent levels, the electricity consumers and a market with the electricity market and electricity trading companies [23]. Power grid systems are mainly divided into the two different categories transmission grids and distribution grids. In Sweden the distribution grid is divided into the regional grid and the local grid [27]. The transmission grid, which operates on a voltage level of 220 kV - 400 kV, transports large amounts of power a large distance and is administered by Svenska Kraftn¨at, the Swedish Trans-mission System Operator (TSO). The national grid is also interconnected to the neighbouring coun-tries through DC transmission connectors. Moreover, the regional grid (20 kV - 130 kV) links the trans-mission grid with the local grid, and further connects power intensive industries and power generation plants. The regional grid is mainly operated by the power grid companies E.ON Energy Networks, Vat-tenfall Power Distribution and Ellevio. Lastly, the local grid (0.4 kV - 20 kV) is used to transfer power from the regional grid to the end customers, e.g., residential facilities. Power generated in small-scale installations is also fed into the local grid [27].
Figure 2: The Power System in Sweden from Svenska Kraftn¨at [28]. The system is divided into electricity producers, power grids, electricity consumer, electricity market and trading companies.
In an synchronous electricity system, there must be a constant balance between production and con-sumption of electricity for the system to work. The voltage frequency should always be 50 Hz and is regulated both automatically and manually by the increase and decrease of production or decrease of consumption [29]. Since the EU Commission regulation 2017/2195 (EB), establishing a guideline on electricity balancing, entered into force in December 2017, several efforts have been initiated on Swedish, Nordic and EU level in order to meet the new requirements [26]. There are several types of frequency reserves used to keep the balance on the grid. Frequency Containment Reserve (FCR), his-torically referred to as primary frequency regulation, are reserves with active effect used to damp the changes in frequency. There are two types of FCR. FCR-N is used for normal operation with frequency between 40.9 and 50.01 Hz and FCR-D is used for disturbed operation activated when the frequency is below 49.9 Hz. Frequency Restoration Reserves (FRR) are reserves for active effect with the purpose of restoring the frequency to the nominal level. FRR is divided into aFRR with automatic activation (earlier called secondary frequency regulation) and mFRR with manual activation (earlier tertiary fre-quency regulation) [26]. Variation in weather, from one day to another but also between seasons, af-fects both production and consumption and is a challenge for keeping the balance [30]. Additionally, large-scale introduction of renewable energy sources in the power systems makes the power generation more intermittent and increase the need for frequency control [31] [22].
The use of electricity per capita in Sweden is high compared to other countries in Europe [32], mostly due to cold winters, the fact that electricity is used for heating, and energy-intense industry. Histori-cally, nuclear and hydro-power have dominated the Swedish electricity generation. [24]. However their share of generation has decreased from 96 percent in 1990 to 80 percent in 2018, mostly due to an ex-pansion in renewable electricity generation dominated by wind power and shut down nuclear plants [24]. In 2020, two nuclear reactor was shut down, which led to a decrease in nuclear power with 27 per-cent, compared to 2019 [25]. Instead the hydro power increased with 10.2 percent and the wind power also reached a production record in 2020, with an increase of 7.7 TWh (38.6 percent) compared to 2019 [33]. The installed renewable generation increases, see Figure 3, and the solar power follows this trend. In 2018, 0.2 percent of the Swedish electricity generation was made up of solar power. However, from 2018 to 2019, the number of installed grid-connected solar PV systems was increased by almost 70 per cent to 698 MW [23].
Figure 3: The share of renewable electricity in relation to total electricity generation 1990-2018, in per cent from The Swedish Energy Agency [23].
Since the beginning of the 21st century, the electricity use has been slightly decreasing in Sweden, de-spite a strong increase in the population [24]. The sector with the highest consumption of electricity is the housing and service sector, followed by the industrial sector. Within the housing and service sector, 90 percent of the energy used belongs to households and premises. About half of the energy is used for heating and hot water and the second half is used as electricity. For single-family houses, electricity is the main heating source for heating purposes, while district heating is the dominating source in multi-family houses and offices [24]. The second biggest user of electricity in Sweden is the industrial sector where 35 percent of the energy used is as electricity. Since the 1970s, the electricity consumption in the industry has increased by over 10 TWh, partly due to the big shift from oil during the oil crisis and the electricity intense production of pulp. However, due to energy efficiency initiatives in the past years, the consumption has decreased.
Moreover, is the power system in Sweden developing fast. Renewable generation plants are being in-stalled in a faster pace than ever before at the same time as nuclear plants and co-generation plants are shut down and decommissioned, which puts pressure on the power system [26]. Consequences are less rotational energy (inertia) and voltage regulation in the system, and additionally, a shortage of power supply in bigger cities during the winter. In the transportation sector a huge transformation is expected, especially within road traffic. The amount of electrical vehicles (EVs) has increased four times from 2016 to 2019 [24] and from 2019 to 2020, the registration of new private cars driven on electricity increased with 77.9 percent [34]. As the amount of electrical vehicles and new electricity intense industries are ramping up in Sweden, the electricity consumption is expected to increase from about 140 TWh to 165 TWh or more by 2040 [26]. This would lead to a decrease in electricity over-production from 20 TWh to 10 TWh in 2040. Along with this, the demand for power is increasing and within a short future, distribution system operators can have power shortages of hundreds of megawatt in their grids, especially in the regions Stockholm, Uppsala, Malm¨o and M¨alardalen county [26]. This needs to be solved by electricity produced further away and a higher use of the transmission grid. A series of large investments in the grid is therefore planned by the transmission system operator Sven-ska Kraftn¨at. Distributor system operators are also making efforts to meet the challenges. This is done by safeguarding that production units deliver during high consumption periods, introducing flexibility in consumption, and limiting consumption during high consumption periods. One example is the ini-tiative CoordiNet in which Svenska Kraftn¨at together with E.ON Energy distribution, Vattenfall has developed a platform for local flexibility [26]. Another is the PARITY project, which E.ON Energy Solutions is a part of. The project aims to investigate local flexibility markets with storage based on smart contracts using blockchain [35]. Another flexibility solutions involving batteries are vehicle-to-grid (V2G) solutions, based on the fact that electric cars can store and dispatch electricity [36]. They can then be used in peak-shaving purposes or so called valley filling, charging at night when the de-mand is low. In the future vehicle-to-grid can be used in combination with PV and battery systems with lower costs for smart buildings [37].
Power system flexibility has become a global priority as power systems all over the world transform. By using a range of operational, policy- and investment-based interventions, modern systems can be-come more flexible, hence enabling cleaner, more reliable, more resilient, and more affordable energy. All power system assets, including variable renewable energy, can provide flexibility services if sup-ported by proper policy-, market- and regulatory frameworks. Several countries have introduced mar-ket reforms and regulations to activate variable renewable energy flexibility. Examples of such coun-tries are Australia, Ireland, Spain and the United States. Innovative flexibility retrofit investments have
been demonstrated in existing conventional power plants in the United States, e.g., hybridisation with battery energy storage system [38]. The fact that variable renewable energy cannot always be held at an constant level, for a certain time required to bid in the electricity market, causes a challenge for fu-ture power systems [39]. These challenges can be solved for example by introducing optimal scheduling and forecasting of Energy Management Systems for dispatchable hybrid renewable power plants and optimize the output from both wind, solar, and a battery storage systems. In this fashion, hybrid re-newable plants can take part in the interval-ahead electricity market as conventional units, and thus increase the economic viability of variable renewable energy. Furthermore, by integrating artificial in-telligence (AI) in the systems, for purposes such as generation forecasting and demand forecasting, the costs of variable renewable energy can be decreased and an integration be fostered [40]. This will have an positive impact on storage applications. The use of AI can, for example, predict cycle lives more precise, and support optimal system configuration, by finding a perfect size and control of the storage. Batteries can hence enable more intermittent and variable renewable energy to penetrate the market and will therefore play a key role in transforming the energy system to become more sustainable [41].
2.2 The Swedish Electricity Market
The Swedish electricity market is deregulated, meaning customers can choose from whom to buy their electricity. The distribution, is however, conducted in the electricity network monopoly [42]. The pur-pose of the deregulation was to create a market with competition to achieve lower prices [43]. There are approximately 380 suppliers of power to Nordic and Baltic end-users [44]. Trading of electricity oc-curs mostly on the Nordic power exchange Nord Pool owned jointly by Svenska Kraftn¨at and Nordic and Baltic counterpart Transmission System Opertors (TSOs). However, a smaller portion of the elec-tricity is traded directly between elecelec-tricity producer and suppliers [30]. The elecelec-tricity price is priced per geographic bidding area, of which the Nordic system has 15, and is determined by the supply and demand of electricity in the respective bidding area. Sweden has four bidding areas: SE1, SE2, SE3, and SE4, from north to south [30]. The average price from all bidding areas is referred to as the system price. Nord Pool further has a spot market Elspot for trading per hour for delivery the next day and a market for intraday trading Elbas, to facilitate short term balance [30].
2.2.1 Electricity Price Development
Figure 4 from the Royal Swedish Academy of Engineering Sciences [45] shows the system prices in Nord Pool from the deregulation in 1996 until 2016. As highlighted by the Royal Swedish Academy of Engineering Sciences [45] and Swedish Bureau of Energy [46], the electricity spot prices on the Nordic market are particularly dependent on weather conditions such as precipitation, temperature and wind. But also on other underlying assets such as coal prices and the introduction of emission rights and elec-tricity certificates.
Figure 4: Average weekly price (system price) on Nord Pool Spot 1996-2016 from the Royal Swedish Academy of Engineering Sciences [45].
Figure 5: Spot price development 1996 to end of august 2020, with expected future prices, from Sweden En-ergy [47].
There are no clear long-term trends in the system price development and the variations can be signif-icant both from month to month as well as between years, see Figure 5. Lower production of nuclear power and higher coal prices have historically increased the prices and the introduction of a quota obli-gation for renewable energy, alongside the electricity certificate market, pushed up the prices for con-sumers. However, this has also had a lowering effect on electricity prices due to the introduction of new production into the system [45]. In January of 2021, the average electricity prices increased in all elec-tricity areas in Sweden due to exceptionally cold weather, creating a debate about the power shortage in the south of Sweden, and about the decision to shut down nuclear power plants [46]. Several stud-ies of the Swedish electricity market development predict larger and more frequent variations in the electricity price, more differences in price between the northern and southern electricity areas, and in-creasing possibility to affect the electricity costs [46]. The International Energy Agency released a re-port on energy prices in 2020 which gives an outlook on global energy prices and concludes that resi-dential electricity prices vary significantly across countries [48]. Electricity prices for resiresi-dential use are generally higher and vary more across countries than industrial electricity prices. This is explained by the market regulations keeping industrial electricity prices low to promote competitiveness [48]. As re-newable power generation is introduced in power systems, the general effect on spot prices is that they are lowered. This is due to the low operating cost of renewable energy generation, avoiding the need to turn on expensive peak load generators such as gas turbines [49]. In California and Western Aus-tralia, the market response to photovoltaic penetration has been depressed prices during daylight hours and higher prices in morning and evenings [50]. The prices were peaking at sunset and creating a so called ”duck-curve”, with the shape of a duck head in evening peak prices and the tail represented by the minor morning price peaks. Volatile renewable energy storage dominated markets should result in a positive affects on average spot price, which can generate the incentives for investment in long-run investment in both generation and storage capacity [50].
2.2.2 End-user electricity price
The pricing of electricity for end-users is divided into two contracts, one for the actual electricity sup-ply (electricity contract), and one for the grid transmission service (grid contract). Within the elec-tricity contract, all costs from the elecelec-tricity supplier are included. This includes the elecelec-tricity price, the electricity certificate cost, and the value added tax (VAT) [51], which is 25 percent [52]. Addition-ally, a fixed annual cost is usually added from the power supplier. The VAT is added upon the total costs, both for electricity supply and grid contract. A company buying services and products have the right to get the amount of VAT corresponding to the VAT they generated through selling their own products or services back [53], which applies to real estate companies but not to housing cooperatives (bostadsr¨attsf¨oreningar). The most common electricity contracts in Sweden according to the Swedish Energy Markets Inspectorate [54] are:
• Fixed price with different curing periods. • Variable price with/without curing period.
• Assigned price.
With a fixed price, the customer pays the same price per kWh throughout the entire duration of the agreement, often one, two, or three years long [55]. The fixed price includes electricity certificate and other fixed costs, taxes however are mostly not included. 60 days before the agreement expires, the electricity supplier has to inform the customer about the ending of the agreement. A fixed price could make it easier for the customer to calculate the total price of the power consumption. On the other hand, a fixed price is often more expensive in a long-term perspective, since the power supplier is tak-ing a risk regardtak-ing the price development durtak-ing the duration of the agreement [56]. A variable price, on the contrary, makes is more difficult to estimate the total power cost, and is therefore posing a risk to the customer. Historically, variable cost contracts have been less expensive over time[55]. However, the customers have to take into account the risk of unexpected expenses. The variable price is changing once a month and is historically based on the average price of electricity from Nord Pool market, the electricity suppliers markup and VAT [57]. Nowadays, a new pricing strategy has been implemented, named volume weighted electricity price (volym¨agt elpris). It means that the customer is paying for the purchased price of electricity for the electricity supplier [58]. This price is based on the electric-ity suppliers purchased energy volume for their customers user profile per hour, and is then multiplied with the hourly spot price for the specific bidding area. To the price, fixed costs from Svenska Kraftn¨at and Nord Pool is added. Mix agreements consist of one variable part and one fixed part, for which the customer pays one part of the power consumption in a variable price, and one part in a fixed price [54]. In these contracts it is often possible to tie the variable part to a fixed one. Lastly, the assigned prices are the prices that are assigned if the customers do not actively choose their power price contract [54]. The customer thus has to pay an indefinite price, which in most cases is much higher than the other types of contracts.
Network contract consists out of two parts; a fixed cost for the subscription fee and one variable cost for the transmission [51].The subscription fee is what the customers have to pay to get access to the grid [51]. For private customers, this fee is depending on the size of the main fuse in the building. The size of the fuse is depending on how much electricity the customer is estimated to take out at their highest peak. For business customers there are two main types of subscription models, where the most commonly used is a fuse subscription similar to the one for private customers [59]. The price for this model is based on the size of the fuse. The second type of contract is power subscriptions, which are used mainly for large consumers such as industries or residential customers. This subscription includes a subscription fee which is independent of the fuse size, instead the customers pay for the maximum power used per month. To be able to hold a power subscription, the facilities have to have an hourly registered electricity meter installed, measuring the highest active power consumed and the energy con-sumed a month. According to the Swedish Electricity Law (1997:857), all businesses with a main fuse bigger than 63 amperes need to have an hourly registered electricity meter installed. Generally speak-ing, what the customer really pays for with the power subscription is the highest consumed hour-value [59]. The transmission fee is the cost for the transmission of the power and is usually measured in ¨ore per kilowatt hour or 15-minutes consumed [51]. This fee is set by the distributor system operator and is meant to cover the operation and maintenance of the grid [60]. Swedish Energy Markets Inspectorate (EI) is controlling so that the transmission fees are according to law.
2.3 Economic and administrative Instruments
There are several economic and administrative instruments that regulate the Swedish power system and encourage fossil free electricity generation. The energy tax on consumed energy by businesses de-pends on the type of organization and where the business is placed [61]. Since 2018, the distributor system operator:s charges the customers for the energy tax, which is then payed to the Swedish Tax Agency. The energy tax in 2021 is 44.5 ¨ore/kWh (VAT included). Moreover, certain tax legislation refers to businesses that are micro-producers of renewable electricity [62]. The business can receive a tax reduction for the electricity that is fed into the grid. To receive the reduction, the production in-stallation has to have the same connection point as the main fuse that connects the business to the grid. Additionally, the fuse must not exceed 100 amperes. Each business can receive reduction for a maximum amount of 30 000 kWh that has been taken from the grid per calendar year, The reduc-tion is 60 ¨ore/kWh, which implies a maximum of 18 000 SEK per calendar year. If the grid contract is shared with others, the maximum of 30 000 kWh still applies. If the business has a photovoltaic in-stallation or several inin-stallations of 255 kW or more the business has to pay taxes and therefore needs to declare how much electricity that has been produced. Furthermore, a company or private customer is not obliged to pay VAT on the sale of electricity if the tax base does not exceed 30 000 SEK for the
year (and for the two years before) [63]. This means that one can sell electricity for 37 500 SEK with-out needing to pay VAT.
The electricity certificate cost was implemented in 2003 in order to stimulate the generation of renew-able energy [64]. When electricity from renewrenew-able sources are generated, the electricity producer re-ceives an electricity certificate for the electricity generated. These certificates are then sold on the mar-ket. The electricity supplier needs to buy these certificates in proportion to how much power is pur-chased, based on a quota. The electricity producers are payed more for the electricity that is coming from renewable sources, which creates incentives for the producer to generate more electricity from re-newable sources. If there is little electricity generated from rere-newable sources, or if the quota increases, the price of the certificates increases. Since 2012, Sweden and Norway have a common market for elec-tricity certificates, aiming to implement 28.4 TWh new renewable elecelec-tricity production until 2020. This goal was already reached in the spring of 2019 [65]. Moreover, the common goal for 2030 within the electricity certificate strategy is to have 46.7 TWh of new renewable electricity production. The re-newable sources that are given electricity certificates are solar energy, geothermic energy, wave energy, wind power, some hydro power, some bio-fuels, and peat in thermal power stations [64]. New installa-tions that were taken into operation after the implementation of the electricity certificate system, have the right to be assigned electricity certificates in 15 years, up until 2035 [65]. During the autumn of 2020, the Swedish government decided to introduce a stopping rule, implying all installations taken into operation after 2021, will not receive the electricity certificate. Owners of photovoltaic installations are also responsible to hold a certain amount of electricity certificates, if they have used more than 60 MWh of their own energy per year, in an installation with more than 50 kW of installed power [66]. The Swedish Electricity Law (1997:857), was adopted in 1997 and includes regulations about electri-cal facilities and trading of electricity [67]. An electrielectri-cal facility in this case is defined as an installation that generates, transfers or consumes electricity. The law commands parties holding a grid concession, to connect electrical facilities and transfer electricity for others on reasonable terms [67]. Moreover, the distribution system operators are not allowed to decline any installation of photovoltaics that already has an existing fuse subscription and which is thus not seen as a new network connection. Although, if the installation results in an upgrade of the fuse, the distribution system operator owns the right to take out a network connection fee [68]. This can, in some cases, lead to high costs and an adjustment of the grid. Photovoltaic owners have the right to receive compensation from the distribution system operator when feeding electricity to the grid, since it lowers the distribution system operators cost for the transmission of electricity [68]. The amount of compensation should be equal to the price of gen-erating electricity locally and therefore vary depending on the electricity price and distribution system operator. The value for the distribution system operator consists of lowered energy losses during trans-mission and reduced fees since electricity does not have to be transferred via other power grids to the consumer. The compensation is usually around some ¨ore per kilowatt hour and is automatically payed by the distribution system operator. Furthermore, micro producers are not obliged to pay for the feed-in to the grid if they fulfill three requirements. These are that they are consumfeed-ing more energy than they feed in to the grid, have a fuse subscription of maximum 63 ampere and generate electricity with a power of maximum 43.5 kW [68]. Instead, the distribution system operator is responsible for measur-ing the electricity transferred from the feed-in on an hourly basis [69]. This usually results in an up-grade of the metering for the micro producer, payed by the distribution system operator. If, however, the micro producer also wants to measure the entire generation, this meter needs to be payed by the micro producers themselves. Moreover, the electricity supplier to the micro producer is obliged to feed in the electricity generated by the facility, if the facility has the right to receive tax reduction for the production of renewable energy and the micro producer has not signed a contract with another electric-ity supplier for receiving the electricelectric-ity. The electricelectric-ity supplier is however not obliged to conclude an agreement and pay for the electricity [69].
In Sweden, tax reduction on Repair, Conversion and Extension exist for private customers, named ROT [70]. This have been used to a great extent in Sweden for renovations and maintenance of, for example, private single-family houses. In the moment of writing, discussions are further held in the Swedish government to increase the so called solar cell support. Another economic instrument imple-mented in 2021, is the tax reduction for green technology [71]. This includes a tax reduction of 15 per-cent on the installation of solar cells and 50 perper-cent on installation of storage of self-produced solar electricity or charging infrastructure for electrical vehicles.
