The building model, used in the design study, is 10 m width, 40 m length and 40 m height (=10 stories) [21] with 40 households [22] . According to the statistics data collected in Japan [22] , the average floor space of dwelling of Japan is about 100 m² per a household. The height of one floor is assumed as 4 m. Assuming that four households stay per floor, the floor space is 400 m². This study proposed the building model utilizing the accelerated wind between buildings for a power generation by wind turbine and assessed the power generation performance of wind turbine of 50 kW class whose height was 39 m [21] . Since this study considers the application of this building model for combining with PV system in the future, the building size of 10 m width, 40 m length and 40 m height is set at the initial condition. The BIPV system has considered on the roof/top of the building, i.e., with 400 m² areas, and a FC using electrolytic H₂ produced when there is the surplus power from BIPV [23] .

The power generated by PV system is calculated by using the following equation [24] :

E P V = H × K × P / 1 (1)

where E_PV is hourly electric power of PV system (kW∙h), H is hourly amount of solar radiation (kW∙h/m²), K is power generation loss factor (−), P is system capacity of PV (kW), 1 is solar radiation under standard state (AM1.5, hourly solar radiation: 1 kW∙h/m², module temperature: 25 degree Celsius) (kW/m²). The instantaneous solar radiation data by 10 sec of the reference [19] are integrated as hourly data for calculating the hourly electric power of PV system.

In this study, the high-performance PV P250α Plus produced by Panasonic has been considered. This module has conversion efficiency and maximum power rating per module is 19.5% and 250 W [25] . The size of PV module is 1580 mm × 812 mm × 35 mm. P has been calculated by installing this PV module on the roof of the building model, which is 75 kW_p (=300 solar modules). To calculate K, the performance value of state-of-the-art commercial devise is used. K is calculated by using the following equation [24] :

K = K p × K m × K i (2)

where K_p is power conversion efficiency of power conditioner (−), K_m is correction factor decided by module temperature (−), K_i is power generation loss by interconnecting and dirty of module surface (−). In this study, K_p and K_i are set at 0.96 and 0.95, respectively. K_p is assumed by referring to the performance of commercial power conditioning device VBPC259B3 manufactured by Panasonic [26] . K_m is calculated by the following equation [24] :

K m = 1 − ( T m − T s ) C 100 (3)

where T_m is PV module temperature (degree Celsius), T_s is temperature under standard test condition (= 25 degree Celsius) (degree Celsius), C is temperature correction factor which is 0.35 [27] (%/degree Celsius). The temperature characteristics of PV module which is adopted for this study is referred. T_m is calculated by using the following equation [24] [28] :

T m = T a + ( 46 0.41 U m 0.8 + 2 ) H − 2 (4)

where T_a is ambient air temperature (degree Celsius), U_m is wind velocity over module of PV (m/s). In this equation, the convection heat transfer by wind around the PV module is considered.

The meteorological data, such as solar radiation, the ambient air temperature, and wind velocity of the cities involved in the study have been taken from the data base of the project “PV300” for the period from August, 2013 to July, 2014 [19] . Table 1 has listed a sample of PV300 data, which shows the data of air temperature and amount of horizontal solar radiation at 10 sec intervals on 1st August, 2013 for Nagoya in Japan. According to the previous study [24] , the optimum tilt angle of solar panel installed on a roof/top of the building is 0 degree, resulting that the horizontal solar radiation has been used for calculating the power generated by the PV system in this study.

In this study, it has been assumed that the surplus power generated by the PV system over the electricity demand of households [20] living in the considered building, would be used for electrolytic H₂ production (i.e. long term energy storage in the form of electrolytic H₂). To optimize the size of BIPV system,

number of households has been varied by 40, 20, 16 and 12 which correspond to 10, 5, 4 and 3 stories of the building, respectively. In this study, the building which has 4 households per floor is assumed. Then, the multiple number of 4 is based for assessment. To keep a building structure, the lowest limit of stories is set at 3 which is the standard low height building, indicating 12 households. The Type-S electrolyzer manufactured by IHT [29] [30] whose H₂ production rate, power consumption and electrolysis efficiency are 760 N∙m³/h, 4.45 kW∙h/N∙m³ and 79.5%, have been used in this design study. The amount of electrolytic H₂ could be produced by the surplus power generated from the PV system is calculated by the following equation:

V H 2 = E s / P e (5)

where V_H2 is amount of electrolytic H₂ produced (N∙m³), E_s is surplus power generated by PV system (kW∙h), P_e is power consumption (kW∙h/N∙m³). In this study, it is assumed that the electrolyzer can be operated following the power generation characteristics of PV system every time and the produced H₂ can be stored as well as used instantaneously [24] .

It has been assumed that the H₂ produced by the electolyzer would be used to generate power through a polymer electrolyte fuel cell (PEFC) system [24] . H₂ is converted into electricity by FC following the below equation:

H 2 + 1 / 2 O 2 = H 2 O + η f Q (6)

where η_f is power generation efficiency of latest PEFC stationary system based on lower heating value (= 0.39) [31] , Q is lower heating value of H₂ (=242) (kJ/mol). It is assumed that the energy loss for operating pump to preserve and provide gases is ignored [18] . Figure 1 illustrates the schematic drawing of the proposed integrated BIPV + FC system.

In this study, a monthly self-sufficiency rate of the proposed combination system consisting of the PV and FC has been investigated for Nagoya (Latitude: 35.10˚N, Longitude: 136.54˚E), Toyota (Latitude: 35.4˚N, Longitude: 137.9˚E), Tajimi (Latitude: 35.19˚N, Longitude: 137.7˚E), Takayama (Latitude: 36.8˚N, Longitude: 137.15˚E), Ogaki (Latitude: 35.21˚N, Longitude: 136.36˚E), Hamamatsu (Latitude: 34.42˚N, Longitude: 137.43˚E) and Shizuoka (Latitude: 34.54˚N, Longitude: 138.18˚E). They are the main cities in Tokai region located in the center of Japan. The self-sufficiency rate is defined as the power supplied (from the combined PV and FC system) to the electricity demand of the households living in the building. The hourly time change in the self-sufficiency rate in the day, when the daily mean amount of horizontal solar radiation per month has been obtained and estimated.

This study has investigated the power production from the PV system using the meteorological data base of PV300 [19] including air temperature, solar intensity and wind speed at 10 sec intervals. Since the electricity demand of households [20] is available as hourly data, the power production from the PV system is

summarized as hourly data by integrating the instantaneous data available at 10 sec intervals.

As an example, Table 2 has listed the hourly power from PV system (installed in Nagoya for a year) which shows the seasonal characteristic of the PV system. From this table, it can be seen the hourly power of PV system increases from the morning up to the noon and decreases from the noon up to the evening as expected. In addition, it is obvious that the hourly power of PV system is higher in spring and summer, while it is lower in autumn and winter. The data shown in the Table 2 has been summed as the typical daily data for the location in the month which then is multiplied by the total days for each month as the monthly data. The monthly power production from the PV system is shown in Figure 2 to investigate the seasonal and local characteristics. It has been observed that the monthly power of the PV system is higher in spring and summer, while it is lower in autumn and winter irrespective of the cities. According to the comparison among seven cities, the monthly power of the PV system from November to January in Takayama is smaller than that in the other cities. In addition, the monthly power of the PV system for Hamamatsu is high through the year. The reasons will be discussed later.

Table 3 has listed the hourly electricity demand of 40 households for a year. In this study, the variation of electricity demand profile for different cities is not considered like the previous study [18] . Namely, the same electricity demand profile has been used for all seven cities in this study. When considering the electricity demand profile of other number of households, the data in Table 3 have been recalculated. It is found that the hourly electricity demand increases from the morning and keeps almost a constant value during a day. After that, the hourly electricity demand increases rapidly from the evening and keeps almost a

constant value up to the midnight. In addition, it is obvious that the electricity demand is higher in summer and winter, especially at the night in summer. However, it is lower in spring and autumn.

Figure 3 has shown the monthly self-sufficiency rate of the FC system in the 40 house case. The self-sufficiency rate is calculated by dividing the power produced by the FC system by the electricity demand, which is not covered by the PV system. Here, the power produced by the FC system has been calculated by using the surplus power of PV system and the Equations (5) and (6). If the self-sufficiency rate of the FC system is over 100%, the proposed integrated system can cover the electricity demand of households living in the building. In this

study, it has been assumed that the surplus power of PV system is used to electrolyze water and the H₂ produced is stored for fueling the FC to make up the shortage of power during the required time.

According to Figure 3, the monthly power of the FC system is higher in spring and summer, while it is lower in autumn and winter irrespective of cities. It can be considered that the power generation characteristics of FC system follows the power generation profile of the PV system. In addition, it has been seen from Figure 3 that the monthly self-sufficiency rate of the FC system is below 100% irrespective of the cities. Therefore, the proposed BIPV + FC system is not able to cover completely the electricity demand of 40 households living in the building.

The households’ number is changed to investigate the proper households’ number for same installment area of solar array on the roof/top of building. Figures 4-6 have shown the monthly self-sufficiency rate of FC system assumed to be installed in Nagoya, Toyota, Tajimi, Takayama, Ogaki, Hamamatsu and Shizuoka for 20, 16 and 12 households’ cases, where the stories of buildings are 5, 4 and 3, respectively. Figures 7-9 show the monthly mean temperature, mean wind velocity and sunshine hours [32] whose date and time followed the data base of the project “PV300” [19] , respectively.

It can be seen from Figures 4-6, the self-sufficiency rate of FC system is higher with decreasing households’ number. This is because that the surplus power of PV system increases with decreasing households’ number, i.e., decreasing electricity demand. In addition, the self-sufficiency rate of FC system can be over 100% in spring and summer while that is below in autumn and winter. This trend matches with the power generation characteristic of PV system. Though the electricity demand increases in summer, the power of PV system seems to be sufficient, resulting in the high self-sufficiency rate of FC system in summer. Comparing seven cities, the self-sufficiency rate of FC system from November to January in Takayama is lower than that in the other cities. Since Takayama has a lot of snow in winter which can be explained by short sunshine hours shown in Figure 9, resulting that the power generated from the PV system is smaller compared to the other cities. Though the monthly mean temperature in winter is lower in Takayama as shown in Figure 7, it is not effective to increase the power generation from the PV system. Therefore, the power generated from FC system decreases due to less H₂ generated. The self-sufficiency rate of FC system in Hamamatsu is high through the year, and it is over 100% even in winter for the 12 households’ case. Hamamatsu has long sunshine hours through the year

according to Figure 9. Furthermore, Hamamatsu has the highest monthly mean wind velocity through the year among seven cities. Strong wind prevents the temperature of the PV modules installed from increasing, thus the power generated by the PV system increases, according to Equations (1), (2), (3) and (4).

Figure 10 shows the annual self-sufficiency rate of FC system assumed to be installed in Nagoya, Toyota, Tajimi, Takayama, Ogaki, Hamamatsu and Shizuoka for 40, 20, 16 and 12 households’ cases. Though the monthly self-sufficiency rate varies, the surplus H₂ obtained in the month whose monthly self-sufficiency rate is over 100% can be stored and provided for the month whose monthly self-sufficiency rate is below 100%. Therefore, the annual self-sufficiency rate of FC system is a good indicator to determine the optimum households’ number which can cover the electricity demand through the year.

According to Figure 10, the annual self-sufficiency rates of FC system for 16 and 12 households’ cases are over 100% irrespective of considered cities. Since it is better to provide the power to many households, this study has decided that 16 households’ (=4 stories) case is the optimum for the proposed BIPV + FC system. In addition, it is observed that the self-sufficiency rate of FC system for 20 households is over 100% in Hamamatsu. Therefore, Hamamatsu is the most appropriate city for installing the proposed BIPV + FC system among the seven cities studied.

The correlation between the households’ number and self-sufficiency rate of FC system per solar PV installation area for two cities in Aichi Prefecture, three cities in Gifu Prefecture and two cities in Shizuoka prefecture has been shown in Figures 11-13, respectively. In these figures, the regression formula of the approximate curves and R² error are also shown.

Formula show the correlation between the households’ number and self-sufficiency rate of FC system per solar PV installation area can be expressed in the form of y = ax^b well. It might be thought that the self-sufficiency rate of

FC system per solar PV installation area will increase for the households’ number which is smaller than 12. However, this study assumed that the lowest limit of stories were set 3 which was the standard low height building. Therefore, this study hasn’t considered the case for households’ number which is smaller than 12.