To what extent can battery storage capacity contribute to matching photovoltaic energy supply with energy demand to achieve peak shaving?

To what extent can battery storage capacity

contribute to matching photovoltaic energy supply

with energy demand to achieve peak shaving?

Pre-Master Thesis

Abstract:

In this simulation study, the demand threshold and the battery capacity size are varied to determine the effects on the battery capacity requirements to achieve peak shaving using PV-power in small cities and villages. The model simulates and provides results in the form of the state of charge of the battery and the failure rate of experiments over a period of one year. The results suggest a battery capacity size between 120.000 kWh and 200.000 kWh combined with a demand threshold of 40.000 kWh (92,9% of average hourly demand), will peak shave 10.9% of the yearly energy demand using PV-power.

Keywords: Peak Shaving; Battery Energy Storage; Photovoltaic Energy; Simulation; State of

Charge; Failure Rate

Supervisor: J. E. Fokkema

Table of Contents

List of tables & figures ...4

List of tables & figures

Table 1. Component list ...12

Table 2. Experiments list ...13

Figure 1. Peak shaving explained ...5

Figure 2. Conceptual model of the simulation ...11

Figure 3. Base case scenario with battery capacity size 10.000 kWh and demand threshold 40.000 kWh 14 .. Figure 4. Base case scenario, number of times SOC within the range (15% - 85%) ...15

Figure 5. Failure rate (%) with varying battery capacity size and varying demand thresholds ...16

Figure 6. Number of hours SOC within the range of 15% - 85% ...17

Figure 7. Failure rate (%) Double PV-power supply ...17

1. Introduction

The urgency to limit climate change is great. Under the Climate Act (2019), the Netherlands is obligated to reduce its greenhouse gas emissions by 95% in 2050 compared to 1990. According to the Ministry of Economic Affairs and Climate, flexibility to cope with the loss of supply or demand of energy is already present in the Netherlands due to the availability of demand-side response, storage and controllable power. Future growth and development of wind and solar energy will further strengthen this flexibility.

Electricity demand and supply vary greatly throughout the day (Chen et al., 2009), in the morning and evening the demand is greater than during the day. However, the solar supply is highest during the day, resulting in a mismatch between demand and supply. The demand for energy has peaks and valleys, especially the peak in demand forms a big challenge for renewable energy sources. Solutions to deal with the peak demand are becoming increasingly important. Growth in peak load raises the chances of a power failure and may lead to a sudden daily increase in the cost of energy. This situation causes an overload on the current energy supply network and poses a challenge to balance supply and demand to meet peak loads for grid operators(Brown & Cortes-lobos, 2010).

Peak shaving is one of the solutions to counter peak demand. Uddin et al. (2018) describe peak shaving as a process in which the peak demand curve is flattened by reducing the peaks in curve through it into times of lower demand. Uddin et al. describe different strategies for peak shaving such as Demand Side Management, integration of Energy Storage Systems and Integration of Electrical Vehicle to the grid. The strategy this paper focuses on is the integration of Energy Storage Systems. Integration of Energy Storage Systems is a method which can be achieved by using Battery Energy Stored Systems (BESS) to store excess power at times with low demand to be used at times of high demand to flatten the curve. Therefore, peak shaving allows for a reduction in electricity costs for consumers but also creates a more manageable request curve for the provider as described by Johnson et al. (2011).

In recent literature, it has been shown that peak shaving using BESS can be economically viable. Economic benefits are achieved several stakeholders such as the house owner with solar panels, the local grid operator and the distribution network operator. The main economic benefits are achieved by charging the BESS on low tariff times and discharging the BESS on high tariff times.

The main problem while using BESS to achieve peak shaving is the sizing of the battery storage and determining the right time to charge the battery. The latter issue will not be considered in this paper. The requirements for battery storage will be looked at in-depth. The goal of this paper is to determine the effects of varying demand thresholds on the battery capacity requirements to achieve peak shaving using solely PV-power in small cities and villages. The battery requirements are determined based on two variables, the capability to supply the demand above the demand threshold and the capability to remain within the desired range of the State Of Charge (SOC) (15% and 85%) (Linden, David & Reddy, 2011). The SOC is calculated as a ratio of the average battery energy to total battery capacity per hour and is used to extend the cycle life of the battery. When the battery requirements are known, this paper can serve as an example for further implementation of similar practices in small cities or villages.

To determine what the effects of varying demand thresholds on the battery capacity requirements to achieve peak shaving in small cities or villages using photovoltaic power are, the following research question is put forward; ‘To what extent can battery storage size contribute to matching photovoltaic energy supply with energy demand to achieve peak shaving?

literature as it researches the effects of combining a varying demand threshold with varying battery capacity sizes on battery capacity requirements to achieve peak shaving in small cities and villages.

This paper is organized as follows. In the following chapter, the theoretical background of this paper is addressed. The existing literature on battery storage to achieve peak shaving is described and analysed.

Thereafter, the methodology will provide an overview of how the simulation is set-up and which data will be used and why. Also, the conceptual model which forms the basis of the simulation is shown. In the results chapter, the different scenarios combined with different battery sizes are analysed and will form the basis for the last two chapters, the discussion and the conclusion. The discussion chapter will reflect on the choices and assumptions made in this paper. Lastly, the conclusion will provide an overview of the most important remarks drawn out of this paper and will discuss possible future research.

2. Literature review

In this paper economic benefits of battery storage to achieve peak shaving are not the main focus nor the application of solar panels in residential neighbourhoods. The main focus is on the battery storage size and the available surplus of energy from photovoltaic power to achieve peak shaving in small cities and villages. Peak shaving achieved by using battery storage with PV-power is researched extensively (Leadbetter & Swan, 2012; Pimm et al., 2018; Zheng et al., 2015). However, this paper contributes to this existing literature in the fact that it combines varying demand thresholds with PV-power and how this affects battery requirements to achieve peak shaving. When battery energy storage systems are used, as described in this paper, it could provide substantial, increasing economic benefits due to a decrease in prices for storage systems (Nykvist, B.; Nilsson, 2015).

another advantage which is the flexibility possible in the placement of a battery in comparison to other ESS or peak shaving strategies.

Zheng et al. (2015) developed a controlling mechanism for BESS by setting a demand limit (hereafter referred to as the demand threshold). They use a constant demand limit and three varying demand limits depending on the season. Whenever there is demand above the set limit, the control unit discharges to meet the demand above the demand limit. When energy demand is less than the demand limit, the storage system will absorb power (provided that the SOC is within its margins).

The research in this paper employs a similar mechanism, however, it only uses a constant demand threshold and the focus in this paper is on battery requirements to achieve peak shaving. In the paper of Zheng et al. (2015) the main focus lies on the economic benefits of using this method to achieve peak shaving.

Leadbetter and Swan (2012) conducted a simulation of battery energy storage systems for residential electricity peak shaving in Canada. They simulated with a range of battery types and sizes and inverter-size specific to a variety of residential houses. Similar to Zheng et al. (2015) they employed a demand threshold above which the battery discharged to match the demand above the threshold. The battery is charged during a nightly five-hour period using grid energy. This research employs a similar mechanism, however, instead of using grid energy to charge the battery it uses solely PV-supplied energy to charge the battery. The simulation only uses grid energy when there is insufficient PV-energy available to fulfil demand. Additionally, this paper focuses on the implications the mechanism has for smaller cities and villages and combines with a demand threshold to determine battery requirements for peak shaving using PV-power.

Li et al. (2018) as it implements a demand threshold above which the battery stored energy is used to match the demand. In the paper of Li et al. (2018) PV-power is used whenever there is demand, to achieve self-sufficiency, which is not the goal in this paper.

Pimm et al. (2018) employ a similar method as researched in this paper. They set out to answer what the maximum possible achievable peak shaving is, using battery storage in residential areas using PV power. However, they simulate for only one week in the summer and one week in the winter in one year for 100 houses. Additionally, the do not consider possible degradation of the battery, and thus a desired State of Charge is not present. They comment on this as possibilities for future work. Their model shows that small-scale electricity storage (e.g. 2 kWh and upwards) has the potential to reduce peak demand at a low voltage substation in the UK by over 50%.

The research in this paper does include a desired SOC (between 15% and 85%) in its simulation, it also conducts a simulation for the entire year and not only for two weeks in a year. This presents a full picture of the possibilities of peak shaving using battery storage with PV-power during a year.

The State of Charge is considered in this paper to ensure a long cycle life of the battery. Xu et al. (2018) describe battery ageing as ‘each cycle causes independent stress, and the loss of battery life is the result of the accumulation of all cycles.’ The State of Charge either increase or decrease the ‘independent stress’ for each cycle. Xu et al. (2018) performed multiple simulations with different SOC levels and concluded obvious differences between high voltage SOC and lower voltage SOC in the level of degradation of the battery.

They explain that this is possibly due to over-voltage caused by a high SOC during the cycling. This is in line with the findings of Linden, David & Reddy (2011) who concluded that it is best to keep the SOC within the range of 15% and 85% every hour. This research paper will use this range for the SOC to ensure a longer battery cycle life.

2012). Additionally, Resch, Bohne, Kvamsdal and Lohne (2016) described that denser cities (which most large cities are) consume less energy per capita, however, this is not known about smaller cities, which implies smaller cities are less energy efficient. This could present new insights from the results compared to existing literature.

3. Methodology

3.1 Problem description

This paper aims to understand how peak shaving strategies affect battery capacity requirements while using photovoltaic power. To be more specific, the main problem addressed in this paper is concerned with the effects of the combination of varying a demand threshold and varying battery sizes are on battery capacity requirements are to achieve peak shaving. The demand threshold is the limit above which energy demand is considered a peak. When energy demand is above this threshold, the battery will supply the energy instead of the energy grid.

We consider parameters and variables within this simulation. The parameters used in this simulation are the Photovoltaic power supply Pt (kWh) per hour and the energy demand Dt (kWh) of Groningen per hour. Also, the demand threshold T (kWh) and the battery storage size S (kWh) are used and varied within this simulation. The independent variables in this simulation are the Failure rate F, in the amount of kWh demanded but not able to supply by photovoltaic energy, and the State of Charge SOC, in hours of which it was within the range of 15% and 85%. The SOC ranges from 15% to 85% to ensure longer battery life (Linden & Reddy, 1995).

3.2 Conceptual model

and the energy demand will be bought from the energy grid. When it is not possible to store photovoltaic energy, because the battery is full, it will either be sold to the grid or disposed of.

A different route is taken if the first question in the conceptual model was answered with ‘no’. The first follow-up question is still whether or not there is demand above the demand threshold. Depending on the answer to the aforementioned question, the next step should be to check if there is energy stored from previous cycles in the battery. If there is sufficient energy stored in the battery this energy will be used to match the demand above the threshold. However, even when there is an insufficient amount of energy stored in the battery, demand has to be met. Resulting in energy being bought from the grid which is the regular practice for a household without a battery energy storage system. Though, this will be noted as a failure in the simulation.

Table 1. Component list

The data for the energy demand of the year 2020 in the Netherlands used in this work is provided by NEDU, The data for hourly solar power supply is provided by PVGIS of the year 2016, the data was first delivered in 15-minute intervals but these are transformed to hourly data to make the simulation more manageable. The data PVGIS provides is energy generated in energetic values, in which efficiency losses are already dealt with. So, there is no need to correct this data for solar panel efficiency losses. The data of PVGIS is based on 230,000 solar panels in Groningen (RTV Noord, 2018) and by using the formula, area in m2 _*

efficiency (18%) / 100 (PVGIS, 2020), to calculate peak power, needed to form the data, it resulted in peak power of 68,310 kWp.

Component Detail Include/exclude Comment

Photovoltaic power supply

Symbol P In kWh

Include Changes hourly

Energy Demand Symbol D

In kWh

Include Changes hourly

Battery Storage Size Symbol S In kWh

Include Changes per

simulation Demand Threshold Symbol T

In kWh

Include Changes per

simulation State Of Charge Symbol SOC

In hours of which it was within the range of 15% and 85%

Include Outcome of

simulation

Failure rate Symbol F

Measured in kWh battery was not able to supply

Include Outcome of

simulation

Energy Saved Symbol ES

In kWh

Include Outcome of

simulation

Cost saved Exclude No cost modelled

Battery types Exclude Variation in battery

types not modelled Degradation of

battery live

Exclude Degradation not

modelled

3.3 Experimental setup

By changing the demand threshold and battery capacity size, different combinations and situations are simulated. By changing the demand threshold, the timing when the stored battery energy is used is changed. This impacts the needed amount of stored energy to match the demand. If the demand threshold is lower, the needed amount of stored energy is higher. When the demand threshold is higher, the energy demand will longer be supplied by the energy grid. Differentiating in battery size allows for respectively more or less stored photovoltaic energy resulting in different capabilities to match the peak demand. When this is combined with the (varying) demand threshold, the best and worst combinations will be shown in the simulation. Allowing to conclude which battery storage size with demand threshold is most optimal for peak shaving. The reason for using only four parameters is to keep the simulation simple yet effective and able to fully answer the research question.

The parameters, demand threshold and battery size, for the base case scenario are based on the parameter data of the energy demand per hour and the solar supply per hour (from respectively NEDU and PVGIS). The average energy demand per hour is taken (43.298 kW) and the average solar supply per hour is taken (7876 kW). The parameter demand threshold and battery size for the base case are the rounded-out numbers of the average supply and demand. Making the demand threshold 40.000 kWh and the battery size 10.000 kWh. All other parameters remain unchanged.

In the following simulation, different combinations of battery storage sizes and demand thresholds will be simulated, to determine the effects on the SOC and the failure rates. These used values for the different battery sizes and demand thresholds are displayed in table 2.

Table 2. Experiments list

Experiments Battery storage size Demand threshold

Base Case 10.000 kWh 40.000 kWh

Varying battery storage size and demand threshold

10.000 / 50.000 / 100.000 / 120.000 / 150.000 / 200.000 kWh

40.000 / 50.000 / 60.000 / 70.000 / 80.000 kWh Double PV-energy supply 10.000 / 50.000 / 100.000 /

120.000 / 150.000 / 200.000 kWh

40.000/ 50.000 / 60.000 / 70.000 / 80.000 kWh

4. Results

The results of the experiments will be discussed in the following section. 4.1 Different scenarios

The base case scenario was determined using the average of the energy demand and the average of the photovoltaic supply. Resulting in a demand threshold of 40.000 kWh and a battery capacity size of 10.000 kWh.

In figure 3, the base case scenario over the entire year is displayed. The amount of kWh is displayed on the y-axis and the months are displayed on the x-axis. The figure shows an overview of the average monthly photovoltaic supply, the demand over the threshold of 40.000 kWh, the State of Charge at the beginning of each period (hour) and the average number of failures.

It is visible that in the winter months (January & February & March & November and December) the average solar supply is lower than the demand above the threshold, in the simulation every hour there is not enough PV-power available, either directly or stored, is counted as a failure which is shown in figure 3 and the SOC will be 0%. When this occurs, the remaining energy will be supplied by the energy grid. When the winter months end, this changes. The solar supply is capable of supplying the requested demand above the threshold.

16

Base case scenario

kW h -14000 -6500 1000 8500 16000 Months

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Average of Photovoltaic supply (kWh) Average of Over Demand Treshold T (kWh) Average of SOC -start of the period Average of Failure (kWh)

For the base case scenario with a battery capacity size S of 10.000 kWh and a demand threshold T of 40.000 kWh, the SOC is in total 562 (6,4%) hours within the desired range of 15% and 85% of the total 8785 hours. This is caused by the fact that the battery is too small to store enough energy to supply the demand. The Failure rate F is far beyond desired, which is on average 80,4%. This is caused mainly by the winter months in which the solar supply was not sufficient for the energy demand. A failure rate of 80,4% means 54.697.489 kWh in a year which is not able to be supplied, meaning this has to be supplied by the energy grid. Which defeats the purpose of peak shaving, because the peaks still exist.

In the simulation, experiments were also performed to test the response of the SOC and the Failure rate, by varying the battery capacity size in combination with varying demand thresholds. In figure 5, the failure rate per combination is shown and in figure 6, the SOC per combination. In total, this present 30 combinations of battery sizes and demand thresholds.

17

Base case scenario

Number of times SOC within range

X hours S O C i n ra nge 0 18 35 53 70 Months

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Total

Failure rate (%)

Figure 5 displays the failure rate. The failure rate presents the percentage of the energy of the total demanded energy surpassing the demand threshold which could not be supplied by photovoltaic energy. The different demand thresholds are displayed on the x-axis, the failure rate percentage is displayed on the y-axis. The different battery capacity sizes are displayed within the graph using different colours. It is visible that the battery sizes of 10.000 and 50.000 kWh peak out the highest for every demand threshold, which means these battery sizes are not suitable to achieve peak shaving based on the failure rate. The larger battery sizes of 120.000 kWh, 150.000 kWh and 200.000 kWh have a relatively low failure rate, especially as the demand threshold increases. At a demand threshold of 80.000 kWh, the three highest battery capacity sizes have a failure rate of 0%, which means that there was never a moment that the batteries were unable to supply the demand. Based on the failure rate, the best combinations to achieve maximal peak shaving capabilities would be a demand threshold of 80.000 kWh with a battery size larger than 120.000 kWh, as the failure rate is 0%.

Figure 6 displays the number of hours the State of Charge SOC was within the range of 15% - 85% per combination. The different demand thresholds are displayed on the x-axis, the number of hours the SOC was within range is displayed on the y-axis. It becomes clear that the demand threshold of 40.000 kWh has the largest number of hours within the range for almost all battery sizes except for a battery size of 10.000 kWh. The main reason for these results is the fact that with the larger battery sizes, the SOC is larger than 85%.

SOC within is range

X hours S O C w it hi n ra nge 0 1000 2000 3000 4000 Demand threshold (kWh) 40000 50000 60000 70000 80000 10.000 50.000 100.000 120.000 150.000 200.000

These results mean that the battery cycle life is lengthened when a demand threshold of 40.000 kWh is used with a large battery size of 150.000 kWh. This, however, contradicts the conclusion drawn from the failure rate, so this implies a trade-off needs to be made whether to choose for a longer battery cycle life (with SOC within range of 15%-85%) or an increased capability of peak shaving (and thus a low failure rate) before these results become applicable in practice.

The double PV-energy supply scenario repeats the experiments as shown above but with the photovoltaic supply doubled. The results of these experiments are shown in figure 7 & 8.

19

Failure rate (%) Double PV power supply

F ai lure ra te (%) 0% 23% 45% 68% 90% Demand threshold 40000 50000 60000 70000 80000 10.000 50.000 100.000 120.000 150.000 200.000

Figure 7 shows a similar graph as for the failure rate when the PV supply was not doubled. However, as the demand threshold increases the failure rate also decreases. The battery size of 10.000 kWh is incapable of achieving peak shaving as the failure rate lies on average above 70%. The larger the battery size and the higher the demand threshold, the lower the failure rate becomes. Similar to the experiments without doubled PV-supply, the best combinations based on the failure rate, to achieve maximal peak shaving capabilities would be a demand threshold of 80.000 kWh with a battery size larger than 120.000 kWh, as the failure rate is 0%.

Figure 8 displays the number of hours the State of Charge SOC was within the range

of 15% - 85% per combination. The different demand thresholds are displayed on the x-axis, the number of hours the SOC was within range is displayed on the y-axis. As visible in figure 8, the demand threshold of 40.000 kWh has the largest number of hours within the range of the SOC. The largest battery size of 200.00 kWh with a demand threshold of 40.000 kWh shows the best result. However, this contradicts the results based on the failure rate, which presented a high demand threshold of 80.000 kWh and a large battery size of 120.000 kWh or larger as the best solution (up to 200.000 kWh). So, even with double PV-power supply the trade-off still needs to be made whether a longer battery cycle life or an increased capability for peak shaving is desired before these results become applicable in practice.

SOC within range - Double PV power supply

X hours S O C w it hi n ra nge 0 1000 2000 3000 4000 Demand threshold (kWh) 40000 50000 60000 70000 80000 10.000 50.000 100.000 120.000 150.000 200.000

4.2 Discussion

The main problem this paper addresses is the peak in energy demand which puts a strain on the energy grid. By including renewable energy, in this case, photovoltaic power, and storing this to use for when peaks appear in demand, peak shaving is achieved. To research this problem the following research question is constructed: ‘‘To what extent can battery energy storage contribute to matching photovoltaic energy supply with energy demand to achieve peak shaving.’

To answer this question different scenarios were simulated, to determine how varying the demand threshold in combination with varying the battery size affects battery requirements to achieve peak shaving. The results per scenario indicated a different conclusion for each variable per scenario. Based on the failure rate it is clear that a demand threshold of 80.000 kWh, which is 185% of the average hourly energy demand, in combination with a battery size of 120.000 kWh or larger (up to 200.000 kWh), is best to achieve maximum peak shaving capabilities.

However, based on the State of Charge, the combination of a demand threshold of 40.000 kWh, which is 92,2% of the average hourly energy demand, with a battery size of 120.000 kWh till 200.000 kWh is best to achieve the longest battery cycle life. The failure rate for these combinations is between 29,4% and 45,4%.

These results are contradicting and form a trade-off, more peak shaving capability and short battery cycle life, or less peak shaving capability and longer battery cycle life. To decide on what trade-off is best suitable for achieving maximum peak shaving capabilities, the energy saved by peak shaving is compared. For the combination recommended based on the failure rate, demand threshold of 80.000 kWh and a battery size of 120.000 kWh till 200.000 kWh, the yearly amount saved by peak shaving is 0.3% of the yearly total energy demand. As opposed to the combination based on the SOC, demand threshold of 40.000 kWh and a battery size of 120.000 kWh till 200.000 kWh, the yearly amount saved by peak shaving is 10.9%.

energy demand? When the answers to these questions become clear, this simulation can serve as an example for applying peak shaving strategies in small cities and villages.

Within this simulation, the demand threshold remains at a constant level throughout the year. The demand threshold is varied per experiment but not within the year. Zheng et al. (2015) performed a similar study but with the focus on the economic benefits of peak shaving with multiple storage systems. In their study, they varied the demand threshold throughout the year on three different levels. The demand threshold is dependent on the season, the three thresholds are based on the winter, summer and spring/fall. This paper opted to keep the demand threshold constant, therefore the demand did not surpass the threshold (mainly in the summer) while there was sufficient photovoltaic supply. This means the photovoltaic power was used to charge the battery, but any remaining energy would be sold back to the energy grid. This would increase the practicality and realism of the model.

When the results of this paper are compared to the results of existing literature, the saving for overall peak shaving is relatively higher. In Martins et al. (2018) a percentage of 6% was of the maximum peak load, in Li et al. (2018) they concluded that a reduction of 1.1% on the peak load can be achieved. In this paper, 10,9% of the yearly energy demand is saved by peak shaving, which is relatively high compared to the other papers. However, if the varying demand limit as performed in Zheng et al. (2015) is included this percentage will be more representative for practice.

5. Conclusion

The research in this paper answers the main research question: ‘To what extent can battery energy storage contribute to matching photovoltaic energy supply with energy demand to achieve peak shaving.’ This paper contributes to the existing literature in the fact that it combines varying demand thresholds with PV-power and how this affects battery requirements to achieve peak shaving. When battery energy storage systems are used, as described in this paper, it could provide substantial, increasing economic benefits due to a decrease in prices for storage systems.

when put into practice. Mainly because wind energy is less season dependent and time-dependent which would increase the amount of energy available for peak shaving. Secondly, the assumption is made that the storage in the form of a battery takes the form of single larger devices or multiple smaller devices which are coordinated in their operation.

By specifying if the battery storage takes the form of a single larger device or multiple smaller devices, the conversion losses when energy is transported through cables could be measured. When one single larger device is used, there will be fewer conversion losses in comparison to when using multiple smaller devices. Thirdly, there is no difference in the type of battery. However, different kinds of batteries have different advantages and disadvantages which are not taken into account in this study. Which could impact possible outcomes in terms of efficiencies and state of charge intervals. Lastly, the economic benefits generated by achieving peak shaving are not calculated for the simulations which could have shown the feasibility of this solution more clearly.

Possible future research could incorporate other renewable energy sources in the mix to achieve peak shaving to increase the completeness of the simulation. Another possible aspect that could be looked at in the future is the optimal amount of batteries and location placement in combination with the energy grid and the renewable energy generation. In practice, this simulation could be used to simulate what the consequences would be of different battery storage sizes and demand thresholds. By doing so, costly real-world tests could be avoided. Lastly, this simulation could be repeated with a varying demand limit throughout the year, as performed by Zheng et al. (2015). Which would increase the practicality and realism of the model.

6. References

Brown, M. A., & Cortes-lobos, R. (2010). Transforming industrial energy efficiency. The Bridge, 40, 31–39.

Chen, H., Cong, T. N., Yang, W., Tan, C., Li, Y., & Ding, Y. (2009). Progress in electrical energy storage system: A critical review. Progress in Natural Science, 19, 291–312. https://doi.org/10.1016/j.pnsc.2008.07.014

Christiansen, C., & Murray, B. (2015). Energy storage study. Energy Post, 130, 108–119.

http://www.energypost.eu/energy-storage-megashift-ahead-battery-costs-set- fall-60-2020/%5Cnpapers3://publication/uuid/F1E9D9DD-CDBE-4977-90F9-08D50BD7D171

Hub, E. S. (2020). PVGIS Users manual. EU Science Hub. https://ec.europa.eu/jrc/en/PVGIS/docs/usermanual

IRENA. (2015). Battery Storage for Renewables : Market Status and Technology Outlook. International Renewable Energy Agency, 60, 11-22.

Johnson, M. P., Bar-Noy, A., Liu, O., & Feng, Y. (2011). Energy peak shaving with local storage. Sustainable Computing: Informatics and Systems, 1, 177–188. https://doi.org/ 10.1016/j.suscom.2011.05.001

Leadbetter, J., & Swan, L. (2012). Battery storage system for residential electricity peak demand shaving. Energy and Buildings.

https://doi.org/10.1016/j.enbuild.2012.09.035

Li, Y., Gao, W., & Ruan, Y. (2018). Performance investigation of grid-connected residential PV-battery system focusing on enhancing self-consumption and peak shaving in Kyushu, Japan. Renewable Energy, 127, 514–523.

https://doi.org/10.1016/j.renene.2018.04.074

Linden, David & Reddy, T. B. (2011). Basics Concepts. In Neutrons in Soft Matter (3rd ed.). https://doi.org/10.1002/9780470933886.ch1

Linden, D., & Reddy, T. B. (1995). Handbook of batteries. In Choice Reviews Online, 33. https://doi.org/10.5860/choice.33-2144

World Economic Forum.

Ministerie van Economische Zaken en Klimaat. (2019). Langetermijnstrategie Klimaat. Beleidsnota, 1–20.

Nykvist, B.; Nilsson, M. (2015). Rapidly falling costs of battery packs for electric vehicles. Nature Climate Change, 5, 329–332. https://doi.org/https://doi.org/10.1038/

nclimate2564

Pimm, A. J., Cockerill, T. T., & Taylor, P. G. (2018). The potential for peak shaving on low voltage distribution networks using electricity storage. Journal of Energy Storage, 16, 231–242.

https://doi.org/10.1016/j.est.2018.02.002

Resch, E., Bohne, R. A., Kvamsdal, T., & Lohne, J. (2016). Impact of Urban Density and Building Height on Energy Use in Cities. Energy Procedia, 96, 800–814. https://doi.org/ 10.1016/j.egypro.2016.09.142

RTV, N. (2018). Aantal zonnepanelen in Stad verdubbeld tot 230.000 stuks.

https://www.rtvnoord.nl/nieuws/199842/Aantal-zonnepanelen-in-Stad-verdubbeld-tot-230-000-stuks

Schuster, R. (2020). What is peak Shaving? EDF Renewables.

https://www.edf-re.de/en/peak-shaving-service/what-is-peak-shaving/

Uddin, M., Romlie, M. F., Abdullah, M. F., Abd Halim, S., Abu Bakar, A. H., & Chia Kwang, T. (2018). A review on peak load shaving strategies. Renewable and Sustainable Energy Reviews, 82, 3323–3332.

https://doi.org/10.1016/j.rser.2017.10.056

Xu, B., Oudalov, A., Ulbig, A., Andersson, G., & Kirschen, D. S. (2018). Modeling of lithium-ion battery degradation for cell life assessment. IEEE Transactions on Smart Grid, 9, 1131–1140.

https://doi.org/10.1109/TSG.2016.2578950

Zheng, M., Meinrenken, C. J., & Lackner, K. S. (2015). Smart households: Dispatch strategies and economic analysis of distributed energy storage for residential peak shaving. Applied Energy, 147, 246–257.

https://doi.org/10.1016/j.apenergy.2015.02.039