Master of Science Thesis
KTH School of Industrial Engineering and Management Energy Technology, EGI
TRITA-ITM-EX 2019:722
Division of Heat and Power Technology SE-100 44 STOCKHOLM
Optimisation of charging strategies and energy storage operation for a
solar driven charging station
Jindan Gong
-i-
Master of Science Thesis EGI 2019 TRITA-ITM-EX 2019:722
Optimisation of charging strategies and energy storage operation for a solar
driven charging station
Jindan Gong
Approved Examiner
Björn Laumert
Supervisor
Monika Topel
Commissioner Umeå Energi AB
Contact person Jörgen Carlsson
-ii-
Abstract
The Swedish energy sector is undergoing transformational changes. Along with a rapid growth of renewables and a shift towards electromobility, the transformation is expected to bring challenges to the power system in terms of grid instability and capacity deficiency. Integrating distributed renewable electricity production into the electric vehicle (EV) charging infrastructure is a promising solution to overcome those challenges.
The feasibility of implementing such a charging infrastructure system in northern Sweden is however uncertain, as the solar resources are scarce in the long winter period. This study aims to maximise the value of a solar powered EV charging station, placed in a workplace environment in Umeå.
An integrated system model of the charging station is developed, comprising separate models of a solar PV system, a battery energy storage system (BESS), the workplace EV fleet and the building Växthuset, onto which the charging station will be installed. Three scenarios are developed to study the charging station’s system performance under different EV charging strategies and BESS dispatch strategies. Two additional scenarios are developed to study the potential grid services that the charging station can provide in the winter period. A techno-economic assessment is performed on each scenario’s simulation results, to measure their effect on the charging station’s value. It involves analysing the charging station’s profitability and how well the BESS is utilised by the end of a ten-year project period. The charging station’s grid impact is further assessed by its self-consumption of solar power, peak power demand and the grid energy exchange.
The assessed charging station values indicate that the overall grid impact was reduced with dynamic EV charging strategies and that the BESS capacity utilisation was strongly influenced by its dispatch strategy.
The charging station further implied a net capital loss under the explored scenarios, even while the dynamic charging strategies brought by a slightly increased economic value. Moreover, the studied winter scenarios showed a great potential for the charging station to provide ancillary services to the local distribution grid while maintaining an efficient BESS capacity utilisation. The winter period’s peak power demand was significantly reduced by optimising the BESS operation to shift peaks in the building’s load profile, and peaks caused by the additional EV charging demand and the EV heaters, to off-peak hours. On this basis, future research is recommended for improved simulations of the charging station operation and to study additional value-added features that the solar driven charging station can bring.
Key words: Electric vehicle (EV), charging station, decentralized energy resources, linear programming, charging strategies, energy storage, energy system modelling
-iii-
Sammanfattning
Sveriges energisystem genomgår en omfattande omställning. Förändringar i form av en ökad andel förnybar elproduktion och elektrifieringen av transportsektorn förväntas medföra stora utmaningar för elsystemets nätstabilitet och överföringskapacitet. Att integrera in distribuerad, förnybar elproduktion som en del av laddinfrastrukturen för elfordon ställer sig som en lovande lösning för att möta de väntande utmaningarna.
Möjligheterna att tillämpa en sådan lösning i norra Sverige är däremot mindre självklara, då solresurserna är knappa under vintertid. Det här examensarbetet syftar till att maximera nyttan av en soldriven laddstation för elbilar, placerad på ett arbetsplatsområde i Umeå.
En integrerad energisystemmodell av laddstationen har skapats, bestående av systemmodeller av solpaneler, ett batterienergilager, arbetsplatsens elbilsflotta samt byggnaden Växthuset, som laddstationen ska anslutas till. Tre scenarier har utformats för att undersöka hur laddstationens prestanda förändras beroende på olika laddstrategier för elbilarna och batterienergilagrets styrning. Ytterligare två scenarier har utvecklats för att utforska möjliga nättjänster som laddstationen kan bistå med under vintertid. Laddstationens värde har vidare bedömts utifrån systemets prestanda i de olika scenarierna. Bedömningen grundar sig på laddstationens lönsamhet och hur välutnyttjat batterienergilagret är efter en kalkylperiod på 10 år, samt på specifika påverkansfaktorer på elnätet. Faktorerna omfattar konsumtionen av egenproducerad el, toppeffektuttaget och nätöverföringarna orsakade av laddstationen.
Från värderingen av laddstationen framgår det att de dynamiska laddstrategierna ledde till en, överlag, minskad påverkan på elnätet samt att styrningen av batterienergilagret hade stor inverkan på dess utnyttjandegrad. Laddstationens nettonuvärde förblev negativt i de tre scenarierna, även om de dynamiska laddstrategierna, ökade dess ekonomiska värde till en viss del. Vidare tyder simuleringen av vinterscenarierna på att det finns en stor potential för laddstationen att erbjuda tjänster för lokalnätet och samtidigt nyttiggöra sig av batterienergilagret. Växthusets toppeffektuttag reducerades märkbart genom att optimera batteristyrningen till att flytta effekttoppar orsakade av Växthusets ellastkurva eller elbilarnas laddning och uppvärmning, till de timmar där lasten var lägre. Med detta i bakgrund föreslås vidare studier som fokuserar på den integrerade energisystemmodellen för att förbättra simuleringarna, samt att undersöka möjligheterna till att erbjuda fler nättjänster, som ökar laddstationens mervärde.
Nyckelord: Elbilar, laddinfrastruktur, distribuerad elproduktion, linjärprogrammering, laddstrategier, energilager, förnybara energisystem
-iv-
Foreword
This thesis report constitutes a final degree project of 30 ECTS, completing a master’s degree in Sustainable Energy Engineering and a five-year degree programme in Energy and Environment. The thesis was conducted at Umeå Energi Elnät AB and at the department of Energy Technology, KTH.
First and foremost, I would like to express my gratitude to my supervisor at Umeå Energi, Jörgen Carlsson, for his invaluable guidance and help throughout the thesis process, and for giving me the incredible opportunity to take part of such an interesting research project in my own hometown. An extended gratitude is directed towards Patrik Nordenstam at Akademiska Hus, for his generous help in clarifying several areas of the project, and Lena Ahlgren from Umeå Energi, for taking her time to give helpful advice on solar PV systems.
From the unit of Heat and Power Technology, KTH, I would like to gratefully acknowledge my examiner Prof. Björn Laumert, for realising this thesis project, and my academic supervisor Dr. Monika Topel Capriles, for her guidance, constant support and for inspiring my interest of energy system modelling.
Finally, from the bottom of my heart, I am immensely grateful to each and every person who made these last five years… words cannot…
-v-
Nomenclature Abbreviations
AC Alternating current
BESS Battery energy storage system BEV Battery electric vehicle CAPEX Capital expenditures
DC Direct current
DoD Depth-of-discharge
EPA Environmental Protection Agency EV Electric vehicle
HEV Hybrid electric vehicle ICE Internal combustion engine IRR Internal rate of return
IVA Royal Swedish Academy of Engineering and Sciences KPI Key performance indicators
LCOS Levelized cost of storage
MPP Maximum power point
NPV Net present value
OC Optimised charging
OPEX Operational expenditures PHEV Plug-in hybrid electric vehicle
PV Photovoltaic
RTC Real-time controlled charging
SMHI Swedish Meteorological and Hydrological Institute SOC State-of-charge
UC Uncontrolled charging
V2G Vehicle-to-Grid
VAT Value-added tax
Characters
Solar PV equipment
𝑁𝑃𝑉,𝑎𝑟𝑟𝑎𝑦 Number of PV modules in the array [ - ]
𝑁𝑃𝑉,𝑠𝑡𝑟𝑖𝑛𝑔 Number of PV modules in each string [ - ]
𝑁𝑠𝑡𝑟𝑖𝑛𝑔 Number of strings in the array [ - ]
-vi-
𝑃𝑐𝑎𝑝,𝑃𝑉𝑎𝑟𝑟𝑎𝑦 Power capacity of the PV array [W]
𝑃𝑃𝑉𝑟𝑎𝑡𝑒𝑑 Nominal power of the PV module [Wp]
Sun-earth geometry
∆𝑡𝐷𝑆𝑇 Daylight saving time [h]
∆𝑡𝐸𝑂𝑇 Equation of time [min]
𝑡𝑐𝑙𝑘 Local clock time [h]
𝑡𝑠𝑜𝑙𝑎𝑟 Solar time [h]
𝛹𝑙𝑜𝑛𝑔 Local longitude [°W]
𝛹𝑠𝑡𝑑 Time zone meridian [°W]
𝛾𝐴 Solar azimuth angle [rad]
𝜃𝑧 Solar zenith angle [rad]
𝑛 Gregorian calendar day of the year [ - ]
𝛿 Declination angle [rad]
𝜑 Local latitude [rad]
𝜔 Hour angle [rad]
Solar radiation
𝐺𝐵 Beam irradiance [W/m2]
𝐺𝐷 Diffuse irradiance [W/m2]
𝐺𝑅 Ground reflected irradiance [W/m2]
𝐺𝑇 Total irradiance [W/m2]
𝛽𝑐 Surface tilt angle [rad]
𝛾𝑐 Surface azimuth angle [rad]
𝜃𝑐 Solar incidence angle [rad]
𝜌 Ground albedo value [ - ]
Electricity yield
𝐸𝑔𝑎𝑝 Band-gap [eV]
𝐺𝑁𝑂𝐶𝑇 Total irradiance under NOCT conditions [W/m2]
𝐺𝑆𝑇𝐶 Total irradiance under STC [W/m2]
𝐼𝑆𝐶 Short circuit current [A]
𝐼𝑚𝑝 Current at maximum power point [A]
-vii-
𝐼𝑜 Saturation current [A]
𝐼𝑝ℎ Photo current [A]
𝑁𝑠 Number of PV cells in series in each module [ - ]
𝑃𝐴𝐶,𝑎𝑟𝑟𝑎𝑦 AC power output of the array and inverter [W]
𝑃𝐷𝐶,𝑎𝑟𝑟𝑎𝑦 DC power output of the PV array [W]
𝑅𝑠 Series resistance [Ω]
𝑇𝑆𝑇𝐶 PV cell temperature under STC [°C]
𝑇𝑎𝑚𝑏 Ambient temperature [°C]
𝑇𝑐𝑒𝑙𝑙 PV cell temperature [°C]
𝑉𝑂𝐶 Open circuit voltage [V]
𝑉𝑚𝑝 Voltage at maximum power point [V]
𝑛1 Ideality factor [ - ]
𝛾𝐼𝑆𝐶,𝑆𝑇𝐶 Temperature coefficient of short circuit current [A/°C]
𝜂𝑖𝑛𝑣 European efficiency of the inverter [%]
𝐼 Output current [A]
𝑁𝑂𝐶𝑇 Nominal operating cell temperature [°C]
𝑆𝑇𝐶 Standard test conditions [ - ]
𝑉 Output voltage [V]
𝑘 Boltzmann’s constant [J/K]
𝑞 Charge of an electron [C]
𝛼 Thermal voltage [V]
BESS model
𝐵𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦_𝑟𝑒𝑡𝑎𝑖𝑛𝑒𝑑 BESS retained capacity [%]
𝐸𝐵𝐸𝑆𝑆 BESS level of charge [kWh]
𝐸𝐵𝐸𝑆𝑆𝑚𝑎𝑥 Maximum BESS level of charge [kWh]
𝐸𝐵𝐸𝑆𝑆𝑚𝑖𝑛 Minimum BESS level of charge [kWh]
𝐸𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦 BESS nominal energy capacity [kWh]
𝐿𝑐𝑦𝑐𝑙𝑒 Total BESS cycle life [cycles]
𝐿𝑡𝑜𝑡𝑎𝑙 Total BESS charge life [kWh]
𝐿𝑢𝑠𝑒𝑑 Used BESS charge life [kWh]
𝑃𝐵𝐸𝑆𝑆 BESS power flow [kW]
𝑃𝐸𝑉 Aggregated EV load [kW]
-viii-
𝑃𝑃𝑉 Solar PV power production [kW]
𝑃𝑠𝑒𝑙𝑓_𝑑𝑖𝑠𝑐ℎ𝑎𝑟𝑔𝑒 BESS self-discharge [kW]
𝑡𝑠𝑡𝑜𝑟𝑎𝑔𝑒 BESS storage time [h]
𝛬 BESS life depletion [%]
Scenarios
𝐶𝐵𝐸𝑆𝑆(𝛬) BESS cost function [SEK]
𝐶𝑒𝑙 Electricity spot market price [SEK]
𝐶𝑝𝑒𝑎𝑘 Grid power fee [SEK/kW]
𝐶𝑡𝑟𝑎𝑛𝑠𝑚 Grid transmission fee [SEK/kWh]
𝐸𝐸𝑉,𝑑𝑎𝑦 Daily EV charging demand [kWh]
𝐸𝐸𝑉 EV battery level of charge [kWh]
𝐸𝐸𝑉𝑚𝑎𝑥 Maximum EV battery level of charge [kWh]
𝑃𝐵𝐸𝑆𝑆 BESS power flow [kW]
𝑃𝐵𝐸𝑆𝑆𝑑𝑖𝑠 BESS discharge [kW]
𝑃𝐸𝑉 EV system fleet load [kW]
𝑃𝐸𝑉_ℎ𝑒𝑎𝑡𝑒𝑟 EV heater electricity demand [kW]
𝑃𝐸𝑉𝑚𝑖𝑛 Minimum EV charging demand [kW]
𝑃𝑃𝑉 Solar PV power production [kW]
𝑃𝑐ℎ𝑎𝑟𝑔𝑒 EV charging rate [kW]
𝑃𝑐ℎ𝑎𝑟𝑔𝑒𝑚𝑎𝑥 Maximum charging rate [kW]
𝑃𝑔𝑟𝑖𝑑 Net grid load [kW]
𝑃𝑣ä𝑥𝑡ℎ𝑢𝑠𝑒𝑡 Växthuset electricity demand [kW]
𝑆𝑂𝐶𝐵𝐸𝑆𝑆 BESS state-of-charge [%]
Techno-economic assessment
𝐵𝐸𝑆𝑆𝑢𝑡𝑖𝑙𝑖𝑠𝑎𝑡𝑖𝑜𝑛 BESS capacity utilisation factor [ - ]
𝐶0 Total investment cost [SEK]
𝐶𝐵𝐸𝑆𝑆𝑐ℎ𝑎𝑟𝑔𝑒 BESS charging cost [SEK]
𝐸𝐵𝐸𝑆𝑆𝑑𝑖𝑠 BESS electricity discharge [kWh]
𝐿𝐸 BESS economic life [years]
𝐿𝑇 BESS technical life [years]
-ix-
𝐿𝑡𝑜𝑡𝑎𝑙 BESS charge life [kWh]
𝑅𝑡 Yearly net cash flow [SEK]
𝑟 Discount rate [%]
𝛬 Yearly BESS capacity depletion [%]
-x-
Table of Contents
Abstract ... ii
Sammanfattning ... iii
Foreword ... iv
Nomenclature... v
Abbreviations ... v
Characters ... v
List of Figures ... xiii
List of Tables ...xiv
1 Introduction ... 1
1.1 Background ... 1
1.2 Problem statement ... 2
1.3 Aims and Objectives ... 2
1.4 Methodology ... 2
1.5 Delimitations ... 3
1.6 Previous studies ... 4
1.6.1 Contributions to research ... 5
1.7 Thesis outline ... 5
2 Theoretical background ... 6
2.1 The Swedish power system ... 6
2.1.1 Production ... 6
2.1.2 The power trading market ... 7
2.1.3 The power grid ... 7
2.1.4 Demand response ... 8
2.1.5 End-user customers ... 8
2.2 Solar PV ... 9
2.2.1 Solar PV cell characteristics ... 9
2.2.2 Solar PV system configuration...11
2.2.3 Single diode equivalent circuit ...12
2.2.4 System orientation ...12
2.3 Battery energy storage ...13
2.3.1 Battery terminology ...14
2.3.2 Batteries ...14
2.3.3 Battery aging ...15
2.3.4 NiMH battery characteristics ...15
2.4 EVs ...16
2.4.1 EV types ...16
-xi-
2.4.2 EV heating systems...16
2.4.3 EV charging infrastructure ...17
2.5 Study location - Umeå ...18
2.5.1 Campus Umeå ...19
2.6 RUGGEDISED U6 project description ...19
3 Model development ...21
3.1 Software programmes...21
3.1.1 Python ...21
3.1.2 Pandapower ...21
3.1.3 PVlib python ...21
3.1.4 Gurobi Optimizer ...22
3.2 Integrated system model development ...22
3.2.1 Charging station system model ...22
3.2.2 PV system model ...23
3.2.3 BESS model ...30
3.2.4 System load ...35
3.2.5 Integration ...38
3.3 Scenarios ...39
3.3.1 Linear programming ...39
3.3.2 Charging strategy scenarios ...40
3.3.3 Winter scenarios ...42
4 Techno-economic assessment ...44
4.1 Charging station economy ...44
4.2 Key performance indicators ...45
4.2.1 Scenario assessment ...46
5 Results ...48
5.1 Charging strategy scenarios ...48
5.1.1 Self-consumption of solar PV power ...48
5.1.2 Net grid load ...50
5.1.3 Charging strategy scenario KPIs ...52
5.2 Winter scenarios ...52
5.2.1 Net grid load ...53
5.2.2 Winter scenario KPIs ...55
6 Discussion ...56
6.1 Integrated system model ...56
6.1.1 PV system model validation ...56
6.1.2 Integrated system model validation ...57
-xii-
6.1.3 Integrated system model weaknesses ...58
6.2 Charging strategy scenarios ...58
6.2.1 Uncontrolled charging ...58
6.2.2 Real-time controlled charging ...59
6.2.3 Optimised charging ...59
6.2.4 Charging strategy scenario comparison ...59
6.2.5 Charging strategy scenario limitations ...60
6.3 Winter scenarios ...60
6.3.1 Reference winter scenario ...60
6.3.2 Optimised winter scenario ...61
6.3.3 Winter scenario comparison ...61
6.3.4 Winter scenario limitations ...62
6.4 Charging station in context...62
7 Conclusions ...63
7.1 Future work ...63
8 Bibliography ...64
-xiii-
List of Figures
Figure 1. Overview of the analysis approach. ... 3
Figure 2. The Swedish power production by energy source 2017 [18]. ... 6
Figure 3. Nordic merit order curve [19]... 7
Figure 4. The four bidding areas of Sweden [21]. ... 8
Figure 5. Solar PV cell working principle [29]. ...10
Figure 6. Solar PV cell I-V curve, power-voltage curve and MPP [30]. ...10
Figure 7. Solar PV cell performance under: (a) varying solar irradiance; (b) varying cell temperature [31]. .11 Figure 8. Structure of a solar PV array [31]. ...11
Figure 9. String inverter configuration [33]. ...12
Figure 10. Single diode equivalent circuit model [35]. ...12
Figure 11. Representation of: (a) Solar positions angles; (b) declination and hour angles for point P [38]. 13 Figure 12. Solar incidence angle, surface azimuth angle and surface tilt angle [38]. ...13
Figure 13. (a) Geographical location of Umeå [66]; (b) total irradiance in Sweden [64]. ...18
Figure 14. Total irradiance by season in Umeå [65]. ...19
Figure 15. Overview of the integrated system model. ...22
Figure 16. Schematic design of the charging station system model. ...23
Figure 17. Four-parameter mathematical model of a single diode equivalent circuit [76]...28
Figure 18. Flowchart of the BESS dispatch algorithm. Based on [16]. ...31
Figure 19. BESS charge retention at +20°C based on manufacturer data [84]. ...32
Figure 20. Flowchart of the self-discharge algorithm. ...33
Figure 21. BESS charge life capacity degradation based on manufacturer data [84]. ...34
Figure 22. Probability distribution of arrival and departure times of the EV system fleet model. ...36
Figure 23. Outline of the separate system model integration...38
Figure 24. Schematic flowchart of the RTC smart charging algorithm. ...41
Figure 25. The self-consumption of PV power with an UC strategy. ...48
Figure 26. The self-consumption of PV power with an RTC strategy. ...49
Figure 27. The self-consumption of PV power with an OC strategy. ...49
Figure 28. The UC net grid load, in relation to the electricity spot market price. ...50
Figure 29. The RTC net grid load, in relation to the electricity spot market price. ...51
Figure 30. The OC net grid load, in relation to the electricity spot market price. ...51
Figure 31. The charging station system’s net grid load resulted from the two winter scenarios. ...53
Figure 32. The net load on the external grid under the reference winter scenario. ...54
Figure 33. The optimised BESS operation’s effect on the net grid load. ...54
Figure 34. Validation of the PV system model’s power production. ...56
Figure 35. Comparison of total solar irradiance in 2015 and 2018 [104]. ...57
-xiv-
List of Tables
Table 1. A brief comparison of established secondary batteries [43, 44, 45]. ...15
Table 2. EV charging equipment overview [61, 62]...18
Table 3. Technical description of the RUGGEDISED U6 project [8]. ...20
Table 4. Technical data on the IBC Solar PV modules [77]. ...24
Table 5. Coefficients of the equation of time [37]. ...25
Table 6. Summary of the BESS model technical parameters. ...30
Table 7. Summary of the EV system fleet model technical data. ...35
Table 8. The modelled EV system fleet size over the year of 2018. ...37
Table 9. The SOCs at arrival and the daily EV charging demands. ...37
Table 10. Bidirectional AC-DC power conversion efficiencies. ...39
Table 11. Economic parameters of the charging station equipment. ...44
Table 12. Umeå Energi’s network tariff scheme for low voltage network customers [26]. ...45
Table 13. Compilation of the charging strategy scenarios’ KPIs. ...46
Table 14. Compilation of the winter scenarios’ KPIs. ...47
Table 15. Compilation of the charging strategy scenarios’ KPIs. ...52
Table 16. Compilation of the winter scenarios’ KPIs. ...55
Table 17. General validation of the integrated system model. ...57
-1-
1 Introduction
The introduction chapter is written to establish the research topic and put it into context to recognise its current importance. The focus of this thesis is narrowed down by defining specific aims and objectives, along with stating its scope and boundaries. The chapter continues with contrasting this thesis with existing research and is lastly closed by presenting an overview of the report structure.
1.1 Background
In light of the 2017 climate goals follow-up, by the Swedish Environmental Protection Agency (EPA), Sweden is foreseen to come short of the 2030 milestone targets towards reaching net zero emissions by 2045. The emission gap in 2030 is expected reach 2-4 million tonne CO2-equivalents in the non-emission trading sector, taking both working and planned measures into consideration. Looking at the stricter 2030 target for the domestic transport sector, the gap is expected to amount to 1-3 million tonne CO2-equivalents [1]. For Sweden to close the gaps and fall in line with the nation’s commitment to the Paris Agreement of limiting global warming to 1.5°C, requires urgent, extensive and unparalleled actions in all sectors of society [2].
The transition to a sustainable energy system is highlighted as a central focus area for reaching the 2045 climate goal of net zero emissions, in which domestic transport is considered a key sector [1]. Both the Swedish EPA and the Royal Swedish Academy of Engineering and Sciences (IVA) believe in the electrification of vehicles as a prominent pathway for decarbonising the transport sector [1, 3]. In their Electricity Crossroads project, the IVA estimates the electricity use in the transport sector to reach 10-16 TWh beyond 2030 [3]. Without the use of incentives, the future’s electric vehicle (EV) fleet could cause a surge in demand during critical hours that eventually would require reinforcements of the electric grid to avert overload. The EVs may however also bring about possibilities to balance the grid by acting as distributed energy storage devices or controllable loads [4]. On the supply side, the 2045 climate goal involves a conversion to a 100% renewable electricity generation mix [1]. The IVA accentuates the prospects of distributed technologies to realise a fully renewable electricity generation, emphasising on small scale installations of solar PV connected directly to the local distribution grid [4].
A coupling between EVs and solar photovoltaics (PV) is promising in theory. Combining the two technologies can be mutually beneficial by allowing a higher penetration of both EVs and solar PV without compromising the grid capacity, power quality or a reliable electricity supply. The EVs’ ability to serve as flexible demands can balance the intermittent electricity supply of renewables. Decentralised solar PV can in turn charge EVs on-site, reducing the net load on the electric grid [5]. Testing and implementing such EV charging infrastructure solutions is a central action area in RUGGEDISED, an EU smart city project to foster the development of smart, resilient cities across Europe [6]. The city of Umeå has been selected as one of three lighthouse cities in RUGGEDISED, to act as a testbed for smart city solutions. Umeå Energi is a local energy company involved that takes on smart EV charging infrastructure solutions through project RUGGEDISED U6. Within the main campus of Umeå University, an EV charging station will be installed on an existing building. A system of on-site solar PV and a local battery energy storage will be used to supply power for daytime EV charging and to balance the building’s electricity loads [7, 8]. As this thesis is written, a small test-facility is under establishment. Decisions on placement of the PV modules and charging station are made and system components are under procurement [8].
-2-
1.2 Problem statement
Although the concept of coupling solar PV with EVs is promising to address coming strains on the electric grid, the practicality of implementing such a system in Umeå is more uncertain. Djurdjevic [9] has identified 5 climatic regions in Europe with varying potential for solar power production, of which northern Sweden belongs to the region with the lowest ranking. Moreover, Nilsson [10] has studied the possibilities of a sports centre in Umeå to achieve electricity self-sufficiency with a roof-top solar PV system and various storage technologies. In the outcome of the study, the analysed energy system was found to have a low feasibility of self-sufficiency, mainly due to insufficient solar resources in the winter months.
The two studies imply that a solar driven EV charging station in Umeå might not be a feasible project and that it could benefit being further assessed prior to installation. Using models to simulate the charging station operation allows for exploring different operation strategies to mitigate the uncertainties and gain insights on its feasibility.
1.3 Aims and Objectives
The overall aim of this thesis is to examine the feasibility of introducing renewable decentralised energy technologies and local battery energy storage to the EV charging infrastructure in a Nordic climate. The specific aim is to assess smart charging schemes and energy storage dispatch strategies to maximise the value of a workplace solar charging station for EVs on Umeå University campus. The aim is achieved by reaching the following objectives:
• Develop an integrated energy system model of the charging station
• Establish and implement scenarios involving dynamic charging schemes and optimised dispatch strategies for the battery energy storage system (BESS)
• Analyse the effects on the BESS capacity utilisation, the charging station’s impact on the power grid and its profitability
• Analyse the potential for the charging station to provide grid services in a winter case
1.4 Methodology
The overall approach to fulfil the aims and objectives of this thesis, begins with performing a literature review of existing works. The literature review is conducted on solar PV powered parking lots for EVs, with the intention of grasping the state-of-the-art research. It further reviews reports and articles on the Swedish power system and the charging station system technologies to establish a theoretical background for the conduct of study. Moreover, representatives from Umeå Energi and Akademiska Hus are interviewed to set the premises for modelling their joint research project and define key performance indicators (KPIs) for the assessment.
The following step of the research approach involves the charging station model development. The charging station is modelled as an integrated system model by joining separate energy system models of the charging station components. It is developed in Python and pandapower, based on technical descriptions from the interviews and mathematical models found in the literature. Scenarios are defined concurrently, as they constitute an integral part of the model. The scenarios are developed based on strategies for EV charging, BESS dispatch and heating system operation found in previous studies and projects. They are implemented into the model by the software programmes pandapower and Gurobi Optimizer.
-3-
In the last step, the scenarios are analysed by performing a techno-economic assessment of each scenario’s simulation results. The assessment measures the charging station’s grid impact, profitability and BESS utilisation, by means of the established KPIs. The potential grid services are analysed separately for the winter case. An overview of the scenario analysis is illustrated in Figure 1. The achieved results are lastly discussed to reason how they fulfil the thesis’s aims and objectives and to identify the limitations of the research.
Figure 1. Overview of the analysis approach.
1.5 Delimitations
The scope of this thesis is limited to study an energy system consisting of the electricity consumption of building Växthuset, the solar PV system, the BESS and the electricity demand of a modelled workplace EV fleet and their heaters. The analysis concerns fully electrified, battery-powered electric cars. The charging of plug-in hybrid electric vehicles (PHEV), electric bicycles and other electric modes of transport is thereby omitted. It further considers smart charging concepts, where the EVs are simulated as loads. Vehicle-to- Grid (V2G) integration of the EVs is not encompassed in the analysis. The charging station system is examined through an energy system model, disregarding the analysis of power quality, voltages and other electrical parameters of a power flow.
Continuing, the BESS capacity utilisation is evaluated by its energy throughput. The charging station’s impact on the grid is measured by its self-consumption of solar power, peak power demand and the amount of electricity transmissions it requires. The profitability is determined using the net present value (NPV) and internal rate of return (IRR) as metrics. They are in turn defined by cash-flows of the charging station capital expenditures (CAPEX) and operational expenditures (OPEX) as costs, and revenues from the EV charging service that the charging station provides. The analysed grid services consist of reducing the energy system’s peak power demand through load shifting.
-4-
1.6 Previous studies
Several recent articles have studied the integration of distributed renewable energy technologies into the EV charging infrastructure [5, 11, 12, 13, 14, 15, 16]. Nunes, Figueiredo, Brito [5] have performed an extensive literature review to issue the state-of-the-art of solar parking lots for EVs and identify related opportunities and challenges. The topical research mainly addresses smart charging concepts, V2G systems and their potential for grid power balancing. Findings of the research show that solar parking lots provide an opportunity of promoting the adoption of EVs and displacement of CO2-emissions. The challenges lie in the field testing and profitability of such energy systems. The literature review continues to highlight technical and legal frameworks for solar parking lots and exploring new mobility concepts, such as automatization of vehicles, as prospects of future work.
The research of [11, 12] cover quantitative studies of using decentralised renewable energy to charge EVs.
Denholm, Kuss and Margolis [11] have simulated large-scale deployments of solar PV and PHEV systems to identify the potential co-benefits of integrating the two technologies into the Texas utility grid. The outcome of the study shows that such an energy system advances petroleum displacement and leaves potential for reducing the size of EV batteries. The work by van der Kam and van Sark [12] have analysed the deployment of EVs, under numerous smart charging schemes, to facilitate a microgrid’s self- consumption of solar power. The main findings indicate that EVs and smart charging schemes can make significant contributions to a well-balanced supply and demand. The system is however not deemed favourable when employed on a larger scale.
In [13, 14], the studies have applied solar driven charging stations in office building environments. Su et al.
[13] have aimed to bring forward an EV owner-friendly charging strategy by making use of solar PV in an office site in China. The study shows that solar parking lots in office sites allow for a higher penetration of EVs without affecting the electric grid. It is further concluded that the benefits of a higher adoption of EVs and PV can be reaped without adapting the EVs owner’s charging behaviour. Tulpule et al. [14] have performed a comprehensive study on two PV powered workplace charging stations in the Unites States with varying solar resources and financial structures. The study consists of an assessment of the charging stations’ economic performance and environmental impact under different charging schemes, followed by a parametric study to find the optimal PV system size. The main findings demonstrate that the charging stations can be profitable for the parking garage owners within the lifetime of the PV system. Optimising the charging scheme has less impact on the charging station economics but results in larger displacement of CO2-emissions as the PV power supply is better utilised for charging the EVs.
The work of [15, 16] continues to examine solar powered workplace charging stations in case studies of European cities. The work of van Roy et al. [15] concerns an existing office building microgrid in Belgium.
It has investigated how different EV charging strategies affect the office building’s consumption of self- generated electricity. The outcome of the study shows that employing straightforward dynamic charging strategies can lead to significant reductions of the system’s grid impact and that smart charging strategies result in a higher self-sufficiency. Chandra Mouli, Bauer and Zeman [16] have performed a study on a solar charging station with a local battery storage, applied to an office building in the Netherlands. The authors conclude that the Dutch solar resources are insufficient during the winter season, hindering the charging station from being self-reliant. As the charging station is meant for workplace charging, the EVs’ ability to balance the solar power supply is limited to weekdays. The main findings further indicate that the best option to complement the Dutch solar resources is a smaller storage size, to balance diurnal solar variations.
-5- 1.6.1 Contributions to research
The literature focuses on smart charging schemes to reduce the grid impact of solar driven charging stations.
The explored smart charging schemes often rely on forecasts of the EV charging demand, solar power production amongst other parameters, neglecting systems where the future knowledge is limited. The most relevant case studies examine the charging stations in locations with mild winters and moderate seasonal variations of length of daytime, leaving a lack of understanding about the performance of solar powered charging stations in areas with seasonally scarce solar resources. Moreover, integrating local storage to support the smart charging strategies in reducing the charging station grid impact is not extensively discussed in the literature.
This thesis contributes to the existing pool of research by examining the profitability and feasibility of a solar powered charging station with local energy storage in a Nordic climate. It is based on local meteorological data, travel behaviour statistics and financial structures. The charging station’s deployment in the subarctic climate of Umeå [17] is further accentuated through the in-depth scenarios of the winter months, involving additional electricity loads from EV heaters. The thesis builds on the work of [15] by pursuing the study of simple smart charging strategies in contrast to smart charging strategies developed through linear programming. Moreover, it extends the work of [16] by including a local battery energy storage in the analysis and exploring different dispatch strategies for the battery, and their effect on the charging station system performance and battery wear.
1.7 Thesis outline
This thesis is structured into seven chapters. The thesis opening of chapter 1 sets the context of the research topic, together with stating the aims and objectives and correlating them to the existing literature. The chapter further describes the research approach for addressing the aims and objectives. Chapter 2 constitutes the theoretical background, presenting theories and concepts relevant to the thesis work. The theoretical background encompasses an overview of the Swedish power system, the investigated technologies and the area of study. Chapter 3 provides a comprehensive description of how the integrated system model is developed and how the explored scenarios are defined. The methods behind the techno- economic assessment are further explained in chapter 4. In chapter 5, the obtained results from implementing the scenarios into the integrated system model and performing the techno-economic assessment are demonstrated. Chapter 6 continues with analysing and discussing the results and thesis limitations. The conclusions of the thesis, together with recommendations for future work, are lastly summarised in chapter 7.
-6-
2 Theoretical background
The theoretical background is written to present relevant theories and concepts behind the analysis and the development of the energy system model. The chapter summarises the fundamentals of the Swedish power system to understand how electricity is priced for the end-users. It further introduces solar PV, battery and EV technologies to understand the basis of the charging station system model. The area of study is lastly presented together with a brief outline of the RUGGEDISED U6 project description.
2.1 The Swedish power system
An overview of the Swedish power system is presented, from production to consumption, in order to understand what impact the charging station might imply on the grid and the potential for adding value to the charging station.
2.1.1 Production
The Swedish power system consists of an almost entirely fossil-free electricity production, mainly relying on hydropower and nuclear power, illustrated in Figure 2. The total electricity production in Sweden amounted to 161 TWh in 2017, of which around 58% originated from renewable energy sources [18].
Figure 2. The Swedish power production by energy source 2017 [18].
The short-term dispatch strategy of the electricity generation technologies can be simplified to follow the merit order system, based on the marginal generation costs and the available capacity. The technologies are ranked in ascending order, to bring the lowest marginal cost technologies online first, and bring higher marginal cost technologies online as the demand increases, illustrated in Figure 3. In reality, the dispatch is regulated by the long-term planning as well, where generation flexibility and limitations of the grid play a great role [19].
-7-
Figure 3. Nordic merit order curve [19].
2.1.2 The power trading market
Nord Pool is the power trading market of the Nordic and Baltic countries, where hourly spot market prices for electricity are decided by day-ahead trading. Power producers publish the available capacities for each hour of the coming day. The buyers then submit their bidding prices in a closed auction. After the bidding window is closed, one single electricity price for each hour is set by market forces at the intersection between the supply and demand curves, exemplified by the dots in Figure 3. The spot market price is essentially determined by the cost of producing the last required kWh to meet the hourly demand. As the auctioned available capacities follow the merit order curve, the hourly electricity system market prices can be considered as a reflection of the electricity demand [20].
2.1.3 The power grid
The Swedish power grid is divided into three systems that maintain different voltage levels and involves different actors. The transmission grid is operated by Svenska Kraftnät, a state-owned authority. It is a high voltage system network for transmitting electric power over long distances, from large-scale producers to the regional distribution network. The transmission grid operates at voltage levels of 220 kV and 400 kV.
The regional distribution grid maintains voltage levels of 20 kV – 130 kV for distributing electric power from the transmission grid and medium-scale producers, to the local distribution grid and large-end users.
The local distribution grid is a low voltage network, operating at voltage levels of 0.4 kV – 20 kV. The purpose of the local distribution grid is to distribute electric power from the regional network and small- scale producers, to low voltage end users. The low voltage end-users lastly consume AC power at voltage levels of 400/230 V [21].
There is a spatial mismatch between the production and consumption of electric power in Sweden, putting a strain on the national transmission grid. The national power grid is consequently divided into four bidding areas, portrayed in Figure 4. Areas SE1 and SE2 cover the northern Sweden, faced with overproduction, while SE3 and SE4 cover the southern Sweden, faced with a power shortage. Transmission limits between the areas and electricity price area differentials are enforced to promote electricity production in the southern Sweden, and by that, stimulate the supply and demand balance within each area to avoid congestions in the transmission grid [20].
-8-
Figure 4. The four bidding areas of Sweden [21].
2.1.4 Demand response
Demand response is an electricity demand management strategy for addressing grid instability issues caused by a fluctuating demand. Actions are taken on the costumer side to change the volume and time-of- occurrence of the consumed electricity, in response to the power market prices. The three main services that demand response strategies can provide to the power system are: peak shaving, valley filling and load shifting. Peak shaving involves reducing the demand congestion by curtailing the electricity consumption during peak hours. As a result, the total electricity consumption is reduced. Valley filling involves increasing the demand during off-peak hours for a smoother demand profile. The overall electricity demand is thereby increased with valley filling. Load shifting involves changing the time-of-use of deferrable appliances to shift the electricity demand from peak hours to off-peak hours, resulting in both reduced peaks and a smoother demand profile. It does not change the total amount of consumed electricity. Apart from improved grid stability, demand response strategies may further imply less emissions, by avoiding fossil fuelled electricity generation to meet demand peaks, and holding off grid reinforcement measures [22].
2.1.5 End-user customers
Demand response and distributed generation are considered measures for deferring network reinforcements [23]. In Sweden, end-user participation in demand response and distributed generation is encouraged through pricing schemes affecting the end-user electric utility bill, and subsidies for solar PV systems and energy storage technologies for households [24].
2.1.5.1 Demand response participation
Through the electric utility bill, end-user customers are obliged to pay for their consumed electricity and the access to, and operation and maintenance of, the power grid. The electricity supply costs are the costs of used electricity and is charged by the electricity suppliers. There are two types of electricity plans for end- user customers: fixed-rate electricity plans, where the electricity price per kWh is constant; and variable-rate electricity plans, where the electricity price fluctuates, following the hourly spot market prices [25].
-9-
The network operation costs are the costs for accessing the power grid and its operation and maintenance.
It is charged by the network operators. Umeå Energi offers two types of network contracts: a fuse subscription contract for households, consisting of a fee for the transmitted electricity, miscellaneous fees and a subscription fee based on the main fuse rating; and a power subscription contract for companies and businesses, consisting of a transmission fee, miscellaneous fees and a subscription fee based on the yearly peak power demand, regardless of the main fuse rating [26]. Moreover, there are additional taxes and fees, accounting for about 30-40% of the total costs in the electric utility bill. This thesis focuses on the electricity supply costs and network operation costs as they are deemed as appropriate indicators for gained charging station value and provided grid services [27].
2.1.5.2 Distributed generation
In Sweden, the adoption of distributed generation systems is stimulated through state-enforced fiscal incentives for solar PV and energy storage systems. There are currently three main incentives for distributed generations systems, concerning the installation and operation of solar PV systems and household energy storage:
• A solar PV installation program, providing a direct capital subsidy covering 20% of the investment costs. The solar PV installation program expires in 2021 [24].
• A green electricity certificate system, where producers of renewable electricity receive certificates for each produced MWh. The certificates are sold, providing an additional a source of income, on top of the electricity sales [24].
• An energy storage installation program, providing a direct capital subsidy covering up to 60% of the investment costs for household installations. The energy storage installation program will be in effect until 2020 [24].
2.2 Solar PV
The fundamental concepts of the solar PV system operation, configuration and orientation are introduced to provide the relevant background information on solar PV systems for the charging station modelling approach.
2.2.1 Solar PV cell characteristics
Solar PV is a renewable energy technology that converts solar energy into electric energy through the photovoltaic effect. The technology is built on systems of solar PV cells, where each cell is a separate unit for energy conversion. The solar PV cells consist of p-type and n-type semiconductor materials that are joined to form an electric field across the cell. When the cell is exposed to solar radiation, electrons are freed in the illuminated areas. The free electrons are subject to the built-in electric field, resulting in a flow of electrons, a photo-induced current [28]. The energy conversion process is illustrated in Figure 5.
-10-
Figure 5. Solar PV cell working principle [29].
The operation and performance of a solar PV cell can be represented by an I-V characteristic curve. It is a graph of the photo-induced current as a function of an external applied voltage and the solar radiation. The curve is bordered by the short-circuit current (𝐼𝑆𝐶), the maximum current induced when the cell is short circuited, and the open-circuit voltage (𝑉𝑂𝐶), the maximum voltage across the solar PV cell occurring when the cell circuit is open. As the power output of the solar PV cell is the product of the cell voltage and cell current, an equivalent power-voltage curve can be derived from the I-V characteristic curve. From each curve, an optimal cell operating point can be determined, where current-voltage pair generates the maximum power output. The operating point is denoted as the maximum power point (MPP) and the underlying current and voltage as the maximum power current ( 𝐼𝑚𝑝) and the maximum power voltage (𝑉𝑚𝑝), respectively [28]. An indicative I-V characteristic curve, power-voltage curve and their unique MPP are demonstrated in Figure 6.
Figure 6. Solar PV cell I-V curve, power-voltage curve and MPP [30].
The performance of the solar PV cell is highly affected by the solar irradiance and the cell temperature.
Their influence on the cell performance can be visualised by studying the I-V characteristic curves under the varying irradiance and temperature conditions. The main effect of changing the solar irradiation at a fixed cell temperature involves a linear increase of the 𝐼𝑆𝐶 and a higher cell efficiency when the solar irradiance increases, illustrated in Figure 7a. When changing the cell temperature at a fixed level of solar irradiance, the main effect involves a linear decrease of the 𝑉𝑂𝐶 and a lower cell efficiency when the cell temperature increases, illustrated in Figure 7b [31].
-11-
Figure 7. Solar PV cell performance under: (a) varying solar irradiance; (b) varying cell temperature [31].
2.2.2 Solar PV system configuration
Single solar PV cells have low operating voltages, producing insufficient power for practical applications.
The cells are, moreover, fragile and prone to mechanical damage. Solar PV cells are therefore grouped and encapsulated into solar PV modules, that achieve higher power outputs and provide the cells with better protection from harsh environmental conditions. Typical solar PV modules consist of 36 or 72 cells connected in series to raise the operating voltage. Modules are further connected in series, into strings, to raise the operating voltage even more. The strings are then connected in parallel to attain higher current outputs, completing a solar PV array with acceptable voltage and power levels for grid integration [31]. The solar PV array structure is demonstrated in Figure 8.
Figure 8. Structure of a solar PV array [31].
Apart from solar PV array, the solar PV system comprise MPP trackers, to regulate the operating voltage to give the maximum power output, and inverters to convert the DC power output into an AC power output.
The solar PV system can be arranged into various configurations to fit the area of application. The string inverter configuration is a typical arrangement for roof-top PV installations. By connecting each string to its own inverter and MPP tracker, the operating voltage can be regulated for each individual string, without compromising the performance of the other strings [32]. A schematic of the string inverter configuration is presented in Figure 9.
-12-
Figure 9. String inverter configuration [33].
2.2.3 Single diode equivalent circuit
An effective method for defining I-V characteristic curves is by modelling each cell in a solar PV array as a single diode equivalent circuit. The equivalent circuit can be visualised through its circuit diagram in Figure 10. It includes a current source, to represent the photo-induced current; a single diode; a shunt resistance to model the current leakage effect at the cell edges; and a series resistance, to model losses within the cells [34].
Figure 10. Single diode equivalent circuit model [35].
2.2.4 System orientation
The amount of solar radiation reaching the surface of a solar PV array depends on the environmental characteristics: the time and the sun’s position in the sky relative to the system’s surface. As the power output is proportional to the solar irradiance, it is necessary to determine the sun-earth geometry to estimate the performance of the solar PV array and decide on its orientation [36].
The sun-earth geometry describes the sun’s position in the sky. For the sun-earth geometry calculations, the local clock time must be converted into solar time, a time measure based on the movement of the sun across the sky. The conversion involves three correction factors: the equation of time, a longitude correction factor and a daylight savings time correction factor [37]. From the point of view of an observer on earth, the sun’s position can be determined by the solar position angles: the solar altitude angle (𝜃𝑠) and the solar azimuth angle (𝛾𝑠), illustrated in Figure 11a. They are in turn defined by the solar zenith angle (𝜃𝑧), and the declination (𝛿) and hour angles (𝜔), illustrated in Figure 11a and Figure 11b, respectively [36].
-13-
Figure 11. Representation of: (a) Solar positions angles; (b) declination and hour angles for point P [38].
The solar incidence angle (𝜃) is used to describe the angular relationship between the normal of a solar PV system’s surface and the incident sunrays. It is defined by the solar position angles together with the surface azimuth angle (𝛾𝑐) and the surface tilt angle (𝛽𝑐), that assert the surface orientation and slope, respectively [36]. The angles are portrayed in Figure 12. An orientation rule of thumb for fixed tilted surfaces is to set the surface tilt angle equal to the local latitude and the surface azimuth angle equal to zero, facing the surface due south, to maximise the amount of received solar radiation over an entire year [39].
Figure 12. Solar incidence angle, surface azimuth angle and surface tilt angle [38].
2.3 Battery energy storage
An overview of battery energy storage is presented to provide a background for the BESS model development. It introduces necessary battery terminology, to understand battery characteristics and operation, and goes through established battery technologies and battery aging in general. The main characteristics of the NiMH battery are, lastly, described, as it is the selected technology for the RUGGEDISED U6 project.
-14- 2.3.1 Battery terminology
Battery pack – A structure of battery modules grouped together to supply power at a higher voltage level.
The battery modules are in turn assemblies of battery cells [40].
Capacity – The theoretical maximum amount of energy that the battery can store or deliver. It is commonly measured in Ah, as the product of a discharge current, corresponding to a specific discharge rate, and the number of hours that the battery can maintain the specific discharge rate. An energy capacity in Wh can be derived from the Ah by multiplying it with the battery’s nominal voltage [41].
Specific energy – A battery characteristic describing the battery’s energy capacity per unit mass [41].
C-rate – A measure of the battery’s discharge current, normalised against its rated capacity. A C-rate of 1 refers to the discharge current required to deplete a fully charged battery in 1 h [41].
State-of-charge (SOC) – The instantaneous amount of energy stored in the battery, as a fraction of the rated energy capacity [41].
Depth-of-discharge (DoD) – The amount of energy withdrawn from the battery as a fraction of the rated energy capacity. The DoD and SOC together constitute the battery charging window, as 𝑆𝑂𝐶 = 1 − 𝐷𝑜𝐷 [41].
Cycling – Commonly refers to a complete charge of a battery, followed by a complete discharge, at a specific DoD. In this thesis, a charging cycle and discharging cycle refers to a session of continuous charging and discharging, regardless of the effective DoD, following the work of [42].
Cycle life – The battery life expressed as the number of cycles that the battery can undergo before its capacity degrades below a certain level. It is estimated by cycling the battery continuously with a standard discharge cycle of a specific DoD [42].
Self-discharge – The temporary capacity loss caused by internal chemical reactions within the battery cell, where the amount of stored energy is reduced without connecting the battery to an external circuit. The rate of self-discharge varies with time [41].
Energy efficiency – A measure of the amount of energy that can be extracted from the battery as a fraction of the amount of energy required to charge the battery [41].
2.3.2 Batteries
Battery energy storage refers to the technologies used for storing electricity by absorbing electrical energy, transforming it into chemical energy and releasing it as DC power at a later point in time. Battery energy storage technologies provide useful tools for improving the power system stability. They are further anticipated to enhance the integration of renewable electricity generation technologies, by supporting the
-15-
grid with system flexibility to balance the intermittent power supply. Moreover, the technologies are expected to play a crucial role in electrifying the transport sector [43].
Batteries can be divided into two types: disposable primary batteries and rechargeable secondary batteries.
Secondary batteries, in turn, encompass a variety of battery storage technologies developed for different areas of application [44]. The main characteristics of the established secondary battery types are presented Table 1.
Table 1. A brief comparison of established secondary batteries [43, 44, 45].
Secondary batteries – main characteristics
Lead-Acid NiCd NiMH Li-ion NaS Reference
Specific energy
[Wh/kg] 30-45 30-45 40-80 60-200 100-250 [45]
Typical
discharge time Hours Hours Hours Hours Hours [45]
Efficiency [%] 75-90 60-70 65-75 85-98 70-85 [45]
Cycle life 1,000 – 5,000
1,000 – 3,000
1,000 – 3,000
1,000 – 20,000
5,000-
10,000 [43, 44]
Toxicity Very high Very high Low Low High [46]
2.3.3 Battery aging
The performance of batteries deteriorates over time due to an internal aging process within the battery cells.
Battery aging manifests as decreased storage and power capacities, and higher self-discharge rates. Although battery aging is inevitable, the process is highly influenced by the charging and discharging operation.
Understanding the underlying mechanisms of the aging process can advance an effective battery management to inhibit power and capacity fade [41].
Four main battery operation factors can be associated to the aging process: SOC, temperature, charging and discharging rates, and the DoD. Batteries feature optimal operating windows for the SOC and operational temperature. By operating the battery outside of the given ranges, the cell degradation is accelerated.
Moreover, practising inappropriate charging regimes can reduce the battery durability notably. High charging and discharging rates, as well as cycling the battery at a high DoD exacerbates the aging process, resulting in a shorter battery life. How strong the influence of each factor is depends on the material characteristics of the battery cell [41, 42].
2.3.4 NiMH battery characteristics
This thesis focuses on stationary electrochemical storage through an NiMH battery energy storage system, as it is the chosen technology for the RUGGEDISED U6 project [8]. NiMH batteries are Ni-based secondary batteries. They are composed of two solid electrodes, one nickel and one metal hydrogen alloy, that are separated by an aqueous electrolyte, KOH [47].
-16-
The primary area of application of NiMH batteries are in hybrid electric vehicles (HEV), as they feature a moderate specific energy and an advantageous level of tolerance against overcharging and undercharging, compared with other secondary batteries. The NiMH discharge is further characterised by a flat discharge curve, meaning that power is released at a stable voltage level throughout the discharge. The main drawbacks of the technology lie in the relatively high discharge rate, with up to 20% capacity losses after 24 h. Cost- wise, the NiMH batteries prices lie on the same level as the competitive Li-ion battery technology [45].
2.4 EVs
A brief technological background on EVs is presented to explain the development of the EV system load in the charging station model. The background addresses the development of both EV charging loads and additional EV loads during wintertime, from their heating systems.
2.4.1 EV types
The term EV refers to a variety of vehicle types that rely on electric propulsion to different degrees. It is a trending technology that shifts the automotive industry away from conventional internal combustion engines (ICEs) and are expected to transform the future mobility [48]. In 2018, EVs accounted for 3.2% of the Swedish passenger vehicle fleet, compared to 2.4% the year before [49]. According to the projections of the Swedish Transport Administration, the passenger vehicle fleet is further expected to be 20% electrified by 2030 in order to realise the 2045 national climate goal and the 2030 national target of a fossil fuel independent vehicle fleet [50].
There are three types of established EV technologies in the Swedish vehicle fleet. HEVs and PHEVs both rely on a combination of electric motors and ICEs for propulsion, and accounted for the largest portions of the passenger EV fleet in 2018 [49]. In HEVs, electric motors are added to assist the ICE and the battery packs are charged by regenerative braking, a process of converting the vehicle’s kinetic energy into electric energy when the vehicles decelerate. PHEVs are powered by their electric motors, assisted by the ICEs. The PHEV battery packs can further be charged from external power sources [51]. Battery electric vehicles (BEV) are fully electrified EVs that form a critical part of the strategic plan for a fossil fuel independent vehicle fleet [50]. The powertrain of BEVs solely relies on electric motors and battery packs for propulsion.
Having larger batteries than PHEVs, BEVs support longer electric driving ranges and are charged by external power sources and by regenerative braking. Moreover, as BEVs exclude ICEs from the powertrain, the technology offers zero tailpipe emissions and generate less engine noise than conventional vehicles. The emissions associated with the technology are rather traced to the primary energy sources used in the electricity generation and are apparent in the fuel production process, as well as the vehicle production and disposal process [52].
2.4.2 EV heating systems
The operation of EVs in cold climates has significant impact on the vehicle performance, as cold operating temperatures diminish the EV battery performance and life expectancy. Most EVs can optionally be equipped with on-board heating systems for thermal comfort and maintaining the vehicle’s electric driving range in cold weathers [53, 54]. The heating system of the Nissan LEAF BEV features a warming system for the battery pack, to keep the battery temperature within a certain range, and a warming system for the EV cabin, steering wheel and front seats to maintain thermal comfort [53]. The cabin heater system can further be employed for defrosting the windshield.
-17-
The Nissan LEAF battery warming system operates to automatically draw power from an external source, through the charging connection, when the EV battery temperature falls below a pre-defined temperature limit. When the EV is disconnected from its charging point, the system draws power from the EV battery instead [55]. Equipping an EV with a heating system and operating it to pre-heat the EV battery by an external source enhances the battery performance in cold temperatures significantly [54].
2.4.3 EV charging infrastructure
The existing EV charging infrastructure in Sweden mainly consist of plug-in charging points for BEVs and PHEVs. In 2019, the total number of installed public and private charging points reached 7,800 units [56], corresponding to around one charging point per eight plug-in passenger EVs in use. The private charging points, in household and workplace parking areas, provided for 80-90% of the EV charging power supply in 2019 [49]. The established plug-in charging technologies can be categorised into four modes, offering methods of EV charging at different rates [57]:
• Mode 1 and Mode 2 refers to slow or semi-fast charging. The two modes support charging the EVs with AC power from common household sockets [58]. The Swedish National Electrical Safety Board, however, advises against using household sockets for charging EVs, as the sockets are not designed to withstand high loads for long periods of time [59].
• Mode 3 supports the safest mode of operation for slow, semi-fast or fast AC charging through poles and wall boxes dedicated for charging at higher power levels. Mode 3 charging can be accessed through public charging poles in workplaces and households [58]. It offers charging rates up to 22 kW [60].
• Mode 4 refers to fast charging. An inverter is integrated into the charging system to supply the EV with DC power directly, which enables higher charging power [57]. The mode of charging can commonly be accessed along motorways and in dedicated charging stations, offering charging rates above 22 kW [60]. Compared to the other plug-in charging technologies, mode 4 charging infrastructure is more expensive and less efficient due to the higher current levels and transfer losses [58].
2.4.3.1 EU charging equipment standards
In 2014, the European Commission adopted a directive that established an EV charging equipment standard, to overcome interoperability issues hindering the EV charging infrastructure development. The standard aims to warrant the future charging infrastructure interoperability, which in turn promotes cross- border travelling and economies of scale for the charging equipment manufacturers. The directive standardises mode 3 charging with type 2 charging connectors for AC charging and mode 4 charging with CCS charging connectors for DC charging [61]. An overview of the EV charging equipment is illustrated in Table 2.
-18-
Table 2. EV charging equipment overview [61, 62].
Plug-in charging equipment for EVs
Connector type Type 1 Type 2 CCS CHAdeMO
Plug
Charging mode 1, 2, 3 3 3, 4 4
Geographic distribution
North America, Japan
Europe (standard), China
North America,
Europe (standard) Japan
This thesis focuses on BEVs, referred to as EVs in the remaining thesis, and private charging points with the EU charging equipment standards of type 2 charging connections and mode 3 slow and semi-fast charging.
2.5 Study location - Umeå
Umeå is the twelfth largest city of Sweden, with a population of 127,000. Being located on the northern east coast, portrayed in Figure 13a, the city serves as the capital of Västerbotten County [63]. Umeå lies in a subarctic climate zone [17], having seasonal average temperatures of -8.5°C in the winter and +14°C in the summer. The city is exposed to approximately 1,800 solar hours/year [63] and a yearly average global irradiance of 925 kWh/m2 [64], illustrated in Figure 13b. The solar resources are unevenly distributed over the seasons, with around 30 times higher global irradiance in the summer than in the winter [65], illustrated by the variations of the diurnal global irradiances presented in Figure 14.
Figure 13. (a) Geographical location of Umeå [66]; (b) total irradiance in Sweden [64].
