ASSESSING THE POTENTIAL OF HYBRID PV ...

32nd European Photovoltaic Solar Energy Conference and Exhibition

ASSESSING THE POTENTIAL OF HYBRID PV–BATTERY SYSTEMS TO COVER HVAC LOADS UNDER SOUTHERN EUROPEAN CLIMATE CONDITIONS J. Solanoa, L. Olivieria, E. Caamaño-Martína, G. Almeida Dávia a Instituto de Energía Solar, Universidad Politécnica de Madrid Av. Complutense 30, 28040 Madrid, Spain. Tel.: +34 91 5495700x4323; fax: +34 91 5446341; e-mail: [email protected]

ABSTRACT: This paper presents theoretical and experimental works that are being carried out in a grid-connected residential building prototype available at the Solar Energy Institute. The house is provided with five rooftop photovoltaic (PV) systems distributed on different planes, a battery energy storage system (BESS) and heating, ventilation and air-conditioning system (HVAC) based on two air-to-air direct expansion reversible heat pumps. Thermal loads, HVAC consumption, PV generation and storage system are simulated using different dynamic models and validated with actual data coming from the monitoring experimental campaign, in order to determine the energy and economic suitability of using hybrid PV-battery systems in combination with reversible heat pumps and optimized control strategies to cover HVAC loads in residential buildings under southern European climate conditions. Results corresponding to the Madrid case-base show that for the building without BESS the selfsufficiency rate is about 20%, however the implementation of storage systems could duplicate this rate, depending on the BESS capacity and the PV generator nominal power. Keywords: Hybrid, PV System, Battery Storage and Control, Building Integration, HVAC.

1

experimental evidence of the possibilities of performing an Active Demand Side Management of domestic electricity consumption was demonstrated worldwide. In particular, by using local forecasts of the expected PV generation, the so-called “deferrable” loads, for example washing machine, dryer and dishwasher, which constitute about 25% of daily load, these loads are displaced in time to make the best direct use of PV electricity by using Artificial Intelligence techniques. Together with a small local electrical storage system (7 kWh), an annual selfsufficiency of 90% can be achieved [10-12].

INTRODUCTION

HVAC systems represent between 40% and 60% of energy demand for buildings in Europe [1] and are also of increasing importance in buildings worldwide [2,3]. Despite the importance of thermal systems in the energy balance of buildings, a review of researches in the field of control allow concluding that specific strategies for combining local PV generation with conditioning equipment powered by electricity (especially reversible heat pumps) have not been proposed yet. The interest of this approach has been also shown by the IEA-Solar Heating and Cooling Programme, particularly within its Task 53 [4], whose main objective is to analyze the use of solar driven systems for cooling and heating, including PV driven systems, from both technical and economic points of view. In this work HVAC loads are analyzed under different climatic conditions. With this aim, Madrid base case has been selected in order to cover the main characteristics of the southern European climate. The selection has been made taking into account not only the Köppen climate classification [5] but also the climatic zones defined in the Spanish Technical Building Code (CTE) [6] in order to have a higher resolution about the influence of the Spanish climate peculiarities. The main objective of this work is to assess the benefits of using hybrid PV-battery systems in combination with reversible heat pumps and optimized control strategies to cover HVAC loads in residential buildings under southern European climate conditions. This work is being developed in Spain, particularly at the premises of the Solar Energy Institute (IES-UPM) where a prototype of a Zero-Energy Residential Building, called “Magic Box”, exists since 2006 (Fig. 1), which was the first European entry in the “Solar Decathlon” international competition and has been recognized internationally as outstanding urban-scale PV system integration project [7,8]. In the prototype several research initiatives have been carried out, amongst which the project “Residential electricity demand side management with PV technology” [9] is to be mentioned. Within this project, the first theoretical and

Figure 1: Upper and frontal (south) views of “Magic Box” prototype In a previous work by the authors [13], it was shown that annual billing for HVAC demand in commercial buildings can be reduced up to 50% only with the use of PV without storage systems, mainly because thermal loads match reasonably well with PV energy production. In contrast, in the residential sector, most of the HVAC consumption occurs time-shifted with respect to the PV generation, therefore the addition of a BESS is required in order to decouple generation and consumption, increasing the self-consumption rate. According to this approach, the aim of this work is to propose optimized sizing strategies for BESS and PV systems, in order to cover HVAC loads in an energy and cost efficient way. 2

METHODOLOGY

2.1 Simulations To analyze the behavior of the global system under different BESS control strategies, the following parameters are being simulated:

2399


a)

b)

c)

d)

HVAC thermal loads and corresponding electricity consumption of the air-to-air direct expansion heat pumps, on the basis of annual dynamic simulations performed using DesignBuilder and EnergyPlus tools. PV energy produced by “Magic Box” rooftop mounted PV systems, using data of incident irradiance on the inclined surfaces of the PV arrays and air temperature measured with a local meteorological station. Performance of the storage system, based on models of batteries developed within this research. Energy and cost savings that the PV-storageHVAC system will produce in comparison with a conventional system, in which the heat pumps are powered by grid electricity.

2.2 Experimental measurements In order to validate the theoretical simulations, the following variables are monitored: outside temperature, outdoor relative humidity, indoor temperature (six sensors), indoor relative humidity, indoor concentration of CO2, electricity consumption of the heat pumps and electricity generated by the PV system. All these variables are being registered every minute. 2.3 HVAC loads estimation In order to estimate the annual thermal loads of the building and the associated electricity demand of the HVAC system, the following approach has been used. Firstly, some of the building characteristics such as location, geometry, materials, internal gains and setup temperatures were defined in DesignBuilder. Next, the model was exported to EnegyPlus in order to accurately define HVAC systems. In particular, to define the ductless air-to-air direct expansion split systems the following EnergyPlus objects were used: - AirLoopHVAC:UnitaryHeatPump:AirToAir:MultSpeed - Coil:Cooling:DX:Multispeed - Coil:Heating:DX:Multispeed - Fan:OnOff This configuration is the same used in the tool BEopt developed by NREL [15] to define Mini-Split Heat Pumps, i.e. the same kind of equipments installed in the experimental building. Next, the performance characteristics were tuned according to the manufacturers specification and the annual simulations were performed with a time-step of 10 minutes. In the next step, the output of the simulation was compared with the experimentally measured data in order to have a feedback about the reliability and accuracy of the model. In particular, for some days, simulated and measured electricity consumptions of the HVAC system were compared, showing a good agreement (Fig. 4). The comparison was not performed on a annual basis because the experimental campaign was still under development when the simulations were carried out.

Figure 2: Simulation diagram of the system.

1.6 1.4

Measured Simulated

1.2 1.0 0.8 0.6 0.4 0.2 0.0 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00

HVAC electricity demand [kW]

In Fig. 2 a diagram of the input and output variables of the whole process simulation is shown, and Fig. 3 show the electrical scheme of the system.

Time [h]

Figure 4: Simulated and measured electricity demand for HVAC from February 12 to 21. Figure 3: Electrical scheme of the entire system. 2.4 Photovoltaic System The PV generation system of “Magic Box” consists of six independent monocrystalline silicon PV arrays with 7 kWp of total nominal power. This system was designed in order to exploit the different tilts of the sun along the year. To achieve this goal, the arrays are distributed in different south-oriented surfaces. Each PV array has an associated string-type inverter, therefore, the PV AC power is supplied to the AC bus. In Table I, the main features for each array are shown: tilt angle, power

Simulations have been made with different durations (minute-based, hourly, daily, monthly, yearly, etc.) and with different configurations of PV and battery sizes. BESS can be simulated by varying storage capacity, state of charge (SoC), efficiency and general behavior according to different algorithms. These control strategies were implemented in MATLAB® - SIMULINK and they have been based on the smart grid simulator GridSim [14].

2400


nominal of PV generator (PG), maximum power inverter (Pinv), and the annual expected PV energy production (EPV).

W

Table I: Characteristics of PV arrays Arrays Inclination PG(kW) Pinv(kW) EPV(kWh)

1 12º 1.45 1.50 1745

2 12º 1.44 1.50 1683

3 25º 1.34 1.50 1774

4 25º 1.34 1.50 1774

5 38º 0.80 1.00 1026

6 90º 0.61 0.60 572

SoC (%)

2500

100

2000

90

1500

80

1000

70

500

60

0

50

-500

40

-1000

Due to the fact that the HVAC consumption is just a part of a residential load, specific PV arrays have been selected (arrays 3 & 4), so that annual generation will be comparable to the annual HVAC load (3548 kWh vs. 3230 kWh).

30

Ppv Pload Pbat Pgrid SoC

-1500 -2000

20 10 0

0

2

4

6

8

10

12

14

16

18

20

22

24

hour

Figure 5: Daily power flow of strategy 1. PPV: experimental PV power; Pload: experimental HVAC load; Pbat: simulated battery power; Pgrid: simulated grid power; SoC: Battery state of charge.

2.5 Battery Energy Storage System Before installation and experimental analysis of the BEES under different control strategies, a theoretical analysis of the dynamic behavior under the expected HVAC load conditions has been done by means of simulations. This assessment is extremely important in any profitability analysis of BEES [16,17], as well as for providing insight on sizing methods for HVAC applications. In this sense, it is worth mentioning that specific powering of HVAC applications have been much less studied compared with more traditional PV applications (e.g. lighting, telecom., signaling, etc.). According to [18], the optimal storage size for residential buildings using PV systems was about 4.5 kWh in 2013 and it will be increased significantly to 7.0 kWh in 2021 under different scenarios of electricity costs. In this work the value of 7 kWh for the case study will be used, and variations between 0 to 10 kWh for further analysis. The house is simulated with a lead-acid battery stationary bank (it is the most used storage element in PV system [19]) coupled to the AC bus by means of a bidirectional inverter. The battery bank is divided in 4 batteries, each battery has a capacity of 145 Ah (C10) and a voltage of 12 V. Therefore, the total battery bank voltage is 48V with a capacity around 7 kWh. The battery inverter does not only implement the current conversion, but also allows for controlling the power flows in the house with high-level software battery controller that can develop different strategies [20].

As shown in Fig. 5, the battery supplies the HVAC demand until it reaches the minimum state of charge allowed (in this case SoCmin = 20%), then the power required is imported from the grid. PV generation is used first to supply the local demand, the surplus of PV charges the battery to 100%, and the excess is exported to the grid. 2.6.2 Strategy 2: Reduction of grid power (Watts balance) In this strategy, as in strategy 1, the PV power supplies demand in self-consumption mode, excess of PV charges the battery, and if the battery is full, PV surplus is exported to the grid. The difference with strategy 1 is that the battery is discharged only to supply power peaks. HVAC demand is supplied firstly by PV and then by the grid, but only up to a limited established power. If demand exceeds this limit, the battery is discharged to supply higher powers. Under any circumstance there is power exchange between the grid and battery. Fig. 6 shows operation of strategy 2, wherein it can be seen that the battery reduces power peaks from the grid, but in terms of energy savings the outcome is minimal.

2.6 Electrical Storage Control strategies The aim of implementing control strategies in the BEES is to maximize the use of local PV production in terms of both energy (Strategy 1) and power (Strategy 2). 2.6.1 Strategy 1: Maximize the use of PV Energy (kWh balance) In this strategy, the PV power supplies demand in self-consumption mode, excess of PV charges the battery, and if the battery is full, PV surplus is exported to the grid. Likewise, HVAC demand is supplied firstly by PV, if it demands more power, it is obtained from the battery, and finally it is imported from the grid. Under any circumstance there is power exchange between the grid and battery. Fig. 5 shows the operation of this strategy for a specific day of the year (March 3rd, 2016), where both PV generation and electricity consumption of HVAC are real measured data, whereas battery power, grid power and SoC are simulated data.

Figure 6: Daily power flow of strategy 2. PPV: experimental PV power; Pload: experimental HVAC load; Pbat: simulated battery power; Pgrid: simulated grid power; SoC: Battery state of charge.

2401


The potential benefit of this control strategy can be explained taking into account the impact that the contracted peak power has on the electricity bill, which strongly depends on the regulatory framework. In Spain for instance the power term has a heavy impact on the electricity bill, so a BESS control strategy focused on reducing power peaks supplied by the grid is worth of study. 3

is outside the period of PV generation. However, in the summer months, most of HVAC demand is shifted within the period of PV generation, hence self-consumption is increased and battery SoC is kept in high levels (average of daily values: 88%, being the corresponding average in winter period: 76%), as shown in Fig. 7b. The control strategies used in batteries, allow to reduce consumption of both energy and power from the grid. Strategy 1 is useful in electricity markets whose value of energy (kWh) is high, while strategy 2 is useful in electrical markets as the Spanish one, where the contracted power (kW) has a high cost in billing (38 €/kW/year in 2015, 212% greater than in 2013). For case study, in strategy 1 (Fig. 5), the contracted power (according to the different levels of permitted power) is 1.725 kW, however, using strategy 2 in Fig. 6, the contracted power is reduced to the smallest reasonable value: 0.805 kW (47% less). The combination of both strategies, as shown in Fig. 7a, enables to limit the power from the grid in the months of peak demand (winter) and to reduce grid electricity imports in summer months.

RESULTS AND DISCUSSION

Due to large number of input variables (17 according to Fig. 1), a specific case has been chosen to show the overall behavior of PV-battery hybrid system: • Location: Madrid • PV generator nominal power: 2.68 kWp • Orientation & Tilt angle: south, 25º. • Annual PV energy generation: 3548 kWh • BESS capacity: 7 kWh, Nominal Voltage: 48V • Minimum SoC: 20%; Maximum SoC: 100% • Annual HVAC electricity consumption: 3230 kWh • Control strategies: strategy 2 in winter (November to March), strategy 1 for rest of the year.

3.2 Self-sufficiency & self-consumption The amount of PV energy locally consumed is called absolute self-consumption when it is instantaneously supplying the electrical demand consumption. However, this value (expressed as a percentage) can be relative to the total generation (self-consumption), or to the total consumption (self-sufficiency) [21]. Necessary energy to supply the HVAC demand comes from two sources: PV (directly or indirectly, through the battery) and the grid. The percentage of demand supplied from PV is the self-sufficiency of the system. Likewise, the energy produced by the PV system is directed toward three destinations: HVAC, battery and the grid. Of the total PV generation, the percentage of energy directed toward both battery and HVAC is the self-consumption of the system.

The main results are shown below: annual power flows, self-sufficiency and self-consumption, HVAC demand duration curve, billing and economic savings. 3.1 Annual power flows In Fig. 7 the simulation of both power exchange with the grid and power exchange with the battery are shown with a minute-based resolution over a year. In winter months, the battery is discharged to supply power values greater than 805 W (strategy 2), and for the rest of the year (from day 90 to 304), the grid does not exceed the established maximum power, so battery uses strategy 1 to reduce energy consumption imported from the grid, as shown in Fig. 7. The annual simulation of the entire system identifies in which periods of time (seasonal, monthly, hourly, etc.) the HVAC demand is higher. Fig. 5 and Fig. 6 show data measured on a winter day wherein electrical consumption

Figure 7: a) Annual power interaction in “Magic Box”. PPV: PV power; Pload: HVAC load; Pbat: battery power; Pgrid: grid power. b) SoC: battery state of charge.

2402


In Fig. 8 it shows the percentages of self-sufficiency and self-consumption for the case study.

3.4 Billing and economic saving According to Spanish regulated prices currently in effect [22], specifically the 2.0 DHS tariff period (residential tariff with hourly prices), the value in euros that would be paid to the utility has been calculated over a year for the three scenarios outlined above. The invoice is divided into three values: power, energy, and taxes, as shown in Fig. 10.

Figure 8: Annual energy accumulated. Origin of energy consumed by HVAC and self-sufficiency (left); final use of PV generation and self-consumption (right). Using a PV system to supply directly the HVAC demand (specific case presented) only reaches 20%, i.e. 80% of energy comes from the grid (Fig. 9). When batteries are added, the energy from grid can be reduced to 64%, as observed in Fig. 8a and Fig. 9. In Fig. 8b the final use of electricity locally generated by the PV system is shown, wherein the 18% covers instantly the HVAC demand, 15% is used to charge the batteries, and also was observed that exists 67% of PV energy that is not used directly and therefore it is exported to the grid. This energy can be used to supply other different types of demand with different load profiles that will significantly increase the selfconsumption.

Figure 10: Annual electricity bill under three scenarios: original demand (left); after using PV system (center); and, after using hybrid PV - battery system (right). In the specific case study, Fig. 10 shows that the use of the PV system only reduces 17% the initial billing, and adding batteries reaches 42% of economic saving. It should be noted that annual simulations have been made with each strategy individually, but with the combination of both strategies the highest results in economic savings are obtained (5 percentage points more than strategy 1, and 10 percentage points more than strategy 2).

3.3 HVAC demand duration curve The grid load duration curves shown in Fig. 9, which represent the grid electricity demand to supply the HVAC load, compare three scenarios: the first one (blue) is without using PV nor battery; the second curve (red) shows the scenario when using only PV (no BEES), and the third curve (yellow) shows the scenario when using PV and BEES.

3.5 Variation of parameters The variation of both PV and battery systems characteristics modify the self-sufficiency (Fig. 11), selfconsumption (Fig. 12) and therefore the economic savings (Fig. 13). Thus, keeping a constant the HVAC demand, multiple simulations were performed varying the battery size between 0 to 10 kWh (viable sizes for residential buildings), and the PV system size between 0 to 7 kWp (covering all fields of “Magic Box”).

15

5

30

20 35

15

5

30

20

Battery ( kWh )

25

10 25

Figure 9: Duration curve of imported energy from the grid to supply the demand for HVAC under three scenarios: original demand (blue); after using PV system (red); and, after using PV - battery system (yellow).

40

10

A-long-side with the benefit for the society from increasing clean electricity penetration in the grid, for the residential user another benefit arises from the use of PV & BEES, consisting decreasing imported energy from the grid, and therefore decrease billing, as will be seen below.

Figure 11: Self-sufficiency percentage using PV-BESS with different battery size and different PV nominal power. HVAC load is constant.

2403


consumption from the grid. In this paper, two strategies have been proposed: the first helps to reduce energy consumption, and the second helps to reduce power. In the case study considered (“Magic Box”: 2.68 kWp coupled to a 7kWh/48V Lead-Acid battery with bidirectional inverter; annual PV generation comparable to HVAC demand), when using a combination of both strategies, it is possible to reduce both 47% of contracted power and 36% of energy consumption. b) The implementation of local PV and electrical storage systems to power HVAC loads improves the self-sufficiency rate in varying degrees, depending not only on the PV power, but also on the HVAC load and storage control capacity. This behavior implies that the optimal sizing of PV & BEES must be carried out in each case by taking into account both energy and economic aspects. c) It is clear that, at present, the billing savings by itself might not be enough to encourage the use of PVbattery systems, it will also strongly depend on the electricity tariff structure and energy policy in each country, in addition to PV and storage systems costs which will determine whether the investment will be profitable from the financial point of view.

20

30

Figure 12: Self-consumption percentage using PV-BESS with different battery size and different PV nominal power. HVAC load is constant.

5

35

The authors gratefully acknowledge the financial support by the National Secretary of Higher Education, Science, Technology and Innovation of Ecuador (SENESCYT) for a PhD scholarship to the first author. The authors acknowledge support of the “Fundación Iberdrola España” by means of the “2015 Ayudas a la Investigación en Energía y Medio Ambiente” in the “Smart Grids para la eficiencia en redes eléctricas: caso práctico en la ETSIT-UPM” project. This work has been partially financed by the Spanish Ministry of Economy and Competitiveness within the framework of the project “DEMS: Sistema distribuido de gestión de energía en redes eléctricas inteligentes” (TEC2015-66126-R).

Figure 13: Billing saving percentage using PV-BESS with different battery size and different PV nominal power. HVAC load is constant. The economic savings in billing, which is shown in Fig. 13, is the percentage that is reduced compared to the original scenario wherein the same house does not have a PV system nor a BEES. The self-sufficiency (Fig. 11), self-consumption (Fig. 12) and billing savings (Fig. 13) rate increments are not linear with the BESS capacity, suggesting that the optimal size of the PV-battery system is a crucial factor to be established on the basis of different energy and economic approaches [23].

4

ACKNOWLEDGEMENTS

6 [1]

[2]

[3]

CONCLUSIONS

In this paper, theoretical and experimental works that are being carried out in a grid-connected residential building prototype available at the Solar Energy Institute, has been presented. In particular, the following conclusions have been drawn from the theoretical analysis done on the influence of the specific application (HVAC electrical supply) and the BEES control strategy on the fraction of local energy used in terms of energy (self-consumption, self-sufficiency) and economics (electricity grid bill): a) It is important to know the annual distribution of electrical loads, to determine which season or hourly periods with higher power demand justify the existence of some kind of strategy enabling to reduce

[4]

[5]

[6]

2404

REFERENCES D. H. W. Li, L. Yang and J. C. Lam, "Zero energy buildings and sustainable development implications – A review," Energy, vol. 54, pp. 1-10, 6/1, (2013). L. Perez-Lombard, J. Ortiz and I. R. Maestre, "The map of energy flow in HVAC systems," Appl. Energy, vol. 88, pp. 5020-5031, 12, (2011). K. F. Fong, T. T. Chow, C. K. Lee, Z. Lin and L. S. Chan, "Comparative study of different solar cooling systems for buildings in subtropical city," Solar Energy, vol. 84, pp. 227-244, 2, (2010). IEA Solar Heating & Cooling Programme. Task 53: New Generation Solar Cooling & Heating Systems (PV or solar thermally driven systems); Available at: http://task53.iea-shc.org/, (2015). M. Kottek, J. Grieser, C. Beck, B. Rudolf and F. Rubel, "World map of the Köppen-Geiger climate classification updated," Meteorologische Zeitschrift, vol. 15, pp. 259-263, (2006). CTE. Código Técnico de la Edificación. Boletín Oficial del Estado: BOE 12/09/2013; Available at: http://www.codigotecnico.org/images/stories/pdf/a horroEnergia/DBHE.pdf, (2013).


[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15] [16]

[17]

[18]

[19]

[20]

M. Calvo-Fernández, J.E. Vega, M.A. Egido, E. Caamaño-Martín, “Spanish participation in the 2005 Solar Decathon: design and simulation of the Photovoltaic system”. 20th EU Photovoltaic Conference, pp. 1958-1963, Barcelona (2005). PV database. Building integrated and urban photovoltaic solar energy projects and products. Available at: http://www.PVdatabase.org/projects_viewbest.php, (2016). Spanish Ministry of Education and Science, GEDELOS PV. Project (ENE2007-66135). National Research, Development and Technology Innovation Program 2007-2010. M. Castillo-Cagigal, E. Caamaño-Martín, E. Matallanas, D. Masa-Bote, A. Gutiérrez, F. Monasterio-Huelin and J. Jiménez-Leube, "PV selfconsumption optimization with storage and Active DSM for the residential sector," Solar Energy, vol. 85, pp. 2338-2348, 9, (2011). M. Castillo-Cagigal, A. Gutiérrez, F. MonasterioHuelin, E. Caamaño-Martín, D. Masa and J. Jiménez-Leube, "A semi-distributed electric demand-side management system with PV generation for self-consumption enhancement," Energy Conversion and Management, vol. 52, pp. 2659-2666, 7, (2011). D. Masa-Bote, M. Castillo-Cagigal, E. Matallanas, E. Caamaño-Martín, A. Gutiérrez, F. MonasterioHuelín and J. Jiménez-Leube, "Improving photovoltaics grid integration through short time forecasting and self-consumption," Appl. Energy, vol. 125, pp. 103-113, 7/15, (2014). J. Solano, L. Olivieri, E. Caamaño-Martín and M. Egido, "Climatización eficiente mediante bombas de calor y tecnología solar fotovoltaica: Análisis de vialidad en edificios comerciales en España”. XIII Congreso Ibero – Americano de Climatización y Refrigeración, (2015). ROBOLABO. GridSim Simulator manual; Available at: https://github.com/Robolabo/gridSim, (2016). BEoptTM (Building Energy Optimization). Available at: https://beopt.nrel.gov, (2016). G. Mulder, D. Six, B. Claessens, T. Broes, N. Omar and J. Van Mierlo, "The dimensioning of PVbattery systems depending on the incentive and selling price conditions," Appl. Energy, vol. 111, pp. 1126-1135, (2013). World Energy Council. E-storage: Shifting from cost to value. Wind and solar applications; Available at: https://www.worldenergy.org/estorage-shifting-from-cost-to-value-2016/, (2016). J. Hoppmann, J. Volland, T. S. Schmidt and V. H. Hoffmann, "The economic viability of battery storage for residential solar photovoltaic systems – A review and a simulation model," Renewable and Sustainable Energy Reviews, vol. 39, pp. 11011118, 11, (2014). C. Salame, M. Aillerie, G. Khoury, N. Achaibou, M. Haddadi and A. Malek. Terragreen 2012: Clean energy solutions for sustainable environment (CESSE) modeling of lead acid batteries in PV systems. Energy Procedia 18pp. 538-544. (2012) M. Castillo-Cagigal, "A swarm intelligence approach based on coupled oscillators: an application in demand side management with

photovoltaic distributed generation," (2014). [21] R. Luthander, J. Widén, D. Nilsson and J. Palm, "Photovoltaic self-consumption in buildings: A review," Appl. Energy, vol. 142, pp. 80-94, 3/15, (2015). [22] Agencia Estatal Boletín Oficial del Estado. BOEA-2015-13782; Available at: https://www.boe.es/boe/dias/2015/12/18/pdfs/BOEA-2015-13782.pdf, (2015). [23] J. Weniger, T. Tjaden and V. Quaschning, "Sizing of Residential PV Battery Systems," Energy Procedia, vol. 46, pp. 78-87, (2014).

2405