This chapter describes the assumptions based on the case study, followed by the methods employed, consisting of three main activities: data acquisition, RSPP design, and design result analysis. The analyses in each activity are done using codes written in MATLAB.
Case study
The Ontowiryo Mosque is located at 109.96° Longitude, −7.85° Latitude, in Wonosari village, Ngombol subdistrict, at the south side of Purworejo district, Central Java. The configuration of the roof, azimuth, tilt, and area, is presented in Fig. 1 and Table 1. The mosque is connected to the grid with a maximum electricity capacity of 5,500 VA. As a community center, installing RSPP on the mosque further enhances its purpose, such as providing street lighting to its surroundings, given the lack of such service in the neighborhood of the case study. This purpose is of interest for further studies as the RSPP analyzed in this paper is limited to supplying the energy use in the mosque, described in the following section.
Figure 1
(a) Indoesia Map showing the location of the Ontowiryo Mosque, modoified from FreeVectorFlags.com with CorelDRAW 201841,42, (b) Google maps image showing the mosque that is facing south-west. The mosque roof has two sides, west (W) and east (E). Image is modified with CorelDRAW 201842, (c) The South Façade adapted from the mosque’s Detailed Engineering Design43.
Table 1 Ontowiryo Mosque roof configuration.
Data acquisition
The data acquisition process prepares the required data for the RSPP design process, including site condition data and hourly electricity load profile data. The site condition data are obtained from Solcast, which provides a comprehensive set of hourly site condition data44. This data includes air temperature, cloud opacity, Diffuse Horizontal Irradiance (DHI), Direct Normal Irradiance (DNI), Global Horizontal Irradiance (GHI), sun azimuth angle, sun altitude angle, wind speed, relative humidity, and precipitable water. Site condition data are available for the chosen location for 12 years (2008–2019) or 4,383 days, containing 105,192 hourly data.
Figure 2a shows the average daily GHI across 12 years of data with a relatively constant value with an average of 5.4 kW/m2 per day. The slight fluctuation is due to the seasonal difference between years, where one year may have a longer dry/wet season and vice versa. The dry season in Indonesia spans in the middle of the year, while the rest is the wet season. Figure 2b presents the variation of irradiances throughout the year. GHI is relatively constant, while higher DHI values are observed at the end and beginning of the year. This condition is due to the wet season with more frequent rain and cloudy skies. Due to the relatively constant irradiance value across years, the site condition data used in the design and simulation can be represented by the data from one of the available years, with its average daily GHI being closest to the average. Thus, 2008, with an average daily GHI at 5.41 W/m2, is used for the design and simulation. Since 2008 is a leap year, only data from 365 days is used to simulate a year of operation.
Figure 2
(a) average daily GHI at each year between 2008–2019. (b) Average daily irradiance (DHI, DNI, and GHI) for each month in 2008.
The hourly load profile data is ideally generated from historical data by taking a sample of electricity consumption in one day. This method is unavailable for the considered case study as the mosque is currently under construction. Therefore, the load profile of the mosque is estimated by calculating the lighting and electronic equipment that is possibly used for specific activities and at certain times, presented in Tables 2 and 3. The list of lighting used in each room is obtained from the mosque’s Detailed Engineering Design. Electrical appliances used for the mosque operation include a sound system, desktop, router, refrigerator, vacuum cleaner, and water pump.
Table 2 Lightings number, power, location, and operational period.
Table 3 Electrical appliances power and operational period.
The final load profile of the mosque is presented in Fig. 3. The electricity use of the mosque is 16.89 kWh/day and 6,166.49 kWh/year. To conclude, this research work with 8,760 hourly data simulates a full year operation of RSPP.
Figure 3
The hourly load demand profile of Ontowiryo Mosque.
RSPP design and analysis
The solar irradiance received by the rooftop area arrives in three forms: direct irradiance (({G}_{M}^{dir})), diffused irradiance (({G}_{M}^{dif})), and irradiance reflected by the ground (({G}_{M}^{ground}))45. The solar irradiance data obtained in the previous activity (DNI, DHI, and GHI) are used to calculate each value of these forms. Equation (1) defines the direct irradiance where (gamma) is the angle of incidence (AOI), the angle between the PV module surface normal, and the incident direction of the sunlight. Given the sun’s position that constantly changes, AOI is a function of panel azimuth angle ({A}_{M}), panel height angle ({a}_{M}), sun azimuth angle ({A}_{S}) and the sun elevation angle ({a}_{S}). Equation (2) define the (mathrm{cos}gamma).
$${G}_{M}^{dir}=DNI cdot mathit{cos}gamma$$
(1)
$$mathit{cos}gamma =mathit{cos}{a}_{M}mathit{cos}{a}_{S}mathit{cos}left({A}_{M}-{A}_{S}right)+mathit{sin}{a}_{M}mathit{sin}{a}_{S}$$
(2)
Equation (3) defines the diffused irradiance, a function of the Sky View Factor (SVF). SVF is the fraction of the sky from which the module can receive diffused irradiance expressed in Eq. (4). Equation (5) defines the irradiance reflected by the ground where (alpha) is the albedo of the ground which determine the reflectivity coefficient of the ground. The total irradiance received by the module (({G}_{M})) is obtained by summing the three components of irradiance to determine the irradiance received by the rooftop and, hence, the estimated yield of solar energy on the mosque roof.
$${G}_{M}^{dif}=DHI cdot SVF$$
(3)
$$SVF=frac{1}{2}left(1+mathit{cos}{theta }_{M}right)$$
(4)
$${G}_{M}^{ground}=GHI cdot alpha cdot (1-SVF)$$
(5)
$${G}_{M}={G}_{M}^{dir}+{G}_{M}^{dif}+{G}_{M}^{ground}$$
(6)
This paper studied a grid-connected RSPP modeled in Fig. 4. The system is modeled in the hourly time step for one year, containing 8760 data steps, and repeated for 25 years (2021–2045) of the RSPP assumed lifetime46. RSPP generation in one hour is calculated as a product of the area of the module, module number (({n}_{M})), irradiance received by the module (({G}_{M})), inverter efficiency (({eta }_{inv})), module degradation rate (({eta }_{deg})), and module efficiency (({eta }_{M})), presented in Eq. (7). Module efficiency is a function of the module’s irradiance level and temperature (({T}_{M})) as presented in Eq. (8) until 1245.
Figure 4
$${E}_{RSPP}(t)={eta }_{M}left({T}_{M},{G}_{M}right) cdot {eta }_{inv} cdot {eta }_{deg} cdot {area}_{M} cdot {n}_{M} cdot {G}_{M}$$
(7)
$${eta }_{M}left({T}_{M},{G}_{M}right)={eta }_{M}left(25^circ{rm C} , {G}_{M}right) cdot left[1+kappa left({T}_{M}-{T}_{STC}right)right]$$
(8)
$${eta }_{M}left(25^circ{rm C} , {G}_{M}right)=frac{FF cdot {V}_{oc}left(25^circ{rm C} , {G}_{M}right) cdot {I}_{sc}left(25^circ{rm C} , {G}_{M}right)}{{G}_{M} cdot {area}_{M}}$$
(9)
$${V}_{oc}left(25^circ{rm C} , {G}_{M}right)={V}_{oc}left(STCright)+frac{n{k}_{B}{T}_{STC}}{q}mathit{ln}left(frac{{G}_{M}}{{G}_{STC}}right)$$
(10)
$${I}_{sc}left(25^circ{rm C} , {G}_{M}right)={I}_{sc}left(STCright)left(frac{{G}_{M}}{{G}_{STC}}right)$$
(11)
$$FF=frac{{I}_{mpp} cdot {V}_{mpp}}{{I}_{sc} cdot {V}_{oc}}$$
(12)
(eta left(25^circ{rm C} , {G}_{M}right)), ({V}_{oc}left(25^circ{rm C} , {G}_{M}right)), and ({I}_{sc}left(25^circ{rm C} , {G}_{M}right)) are the module efficiency, open-circuit voltage, and short circuit current as a function of irradiance level. Data obtained from the module’s datasheet are temperature coefficient ((kappa)), open-circuit voltage (Voc), short circuit current (Isc), area of the module, and module’s degradation rate. The values for Boltzmann constant (kb), elementary charge (q), and quality factor (n) are obtained from47. FF is the Fill Factor, the ratio between power generated at maximum power point and the product of Voc with Isc. GSTC and TSTC are the standard conditions for irradiance and temperature when the PV modules were tested, at 1000 Wm−2 and 25 °C. The module’s temperature (({T}_{M})) is determined through the fluid-dynamic model which considers heat sources from sun irradiance, convective and radiative heat exchange from the front and rear side of the module. Equations (13–25) present the calculation for ({T}_{M}),
$${T}_{M}(t)=frac{{alpha }_{M}{G}_{M}+{h}_{c}{T}_{a}+{h}_{r,sky}{T}_{sky}+{h}_{r,gr}{T}_{gr}}{{h}_{c}+{h}_{r,sky}+{h}_{r,gr}}$$
(13)
$${h}_{c}={h}_{c}^{T}+{h}_{c}^{B}$$
(14)
$${h}_{c}^{T}=sqrt[3]{{h}_{forced}^{3}+{h}_{free}^{3}}$$
(15)
$${h}_{forced}= {w}^{0.8}$$
(16)
$${h}_{free}= frac{0.21 cdot k{(Gr cdot Pr)}^{0.32}}{{D}_{h}}$$
(17)
$$Gr=frac{gleft({T}_{M}-{T}_{a}right){D}_{h}^{3}}{{v}^{2}}$$
(18)
$${D}_{h}=frac{2LW}{L+W}$$
(19)
$${h}_{c}^{B}=R cdot {h}_{c}^{T}$$
(20)
$$R=frac{{alpha }_{M}G-{h}_{c}^{T}left({T}_{INOCT}-{T}_{a}right)-{epsilon }_{top}sigma left({T}_{INOCT}^{4}-{T}_{sky}^{4}right)}{{h}_{c}^{T}left({T}_{INOCT}-{T}_{a}right)+{epsilon }_{top}sigma left({T}_{INOCT}^{4}-{T}_{sky}^{4}right)}$$
(21)
$${alpha }_{M}=left(1-Rright)left(1-{eta }_{M}left(STCright)right)$$
(22)
$${T}_{INOCT}= {T}_{NOCT}+18$$
(23)
$${h}_{r,gr}={epsilon }_{back}sigma left({T}_{M}^{2}+{T}_{gr}^{2}right)left({T}_{M}+{T}_{gr}right)$$
(24)
$${h}_{r,sky}={epsilon }_{top}sigma left({T}_{M}^{2}+{T}_{sky}^{2}right)left({T}_{M}+{T}_{sky}right)$$
(25)
where ({alpha }_{M}) is the absorptivity of the module, ({T}_{a}) is the ambient temperature, ({h}_{c}) is the convective heat transfer coefficient of the module, ({h}_{r,sky}) and ({h}_{r,gr}) are the radiative heat transfer to the sky and the ground, or roof in case of RSPP, and ({T}_{sky}) and ({T}_{gr}) are the sky and ground temperature. The process of finding the module’s temperature is an iterative process as the ({h}_{r,sky}) and ({h}_{r,gr}) are functions of ({T}_{M}). (Gr) is Grashof number, (Pr) is Prandtl number which equals to 0.71 for air, (sigma) Stefan-Boltzmann constant, ({D}_{h}) is the hydraulic diameter of a rectangle with width (W) and length (L), (k) is the air thermal conductivity and (v) is the air viscosity. ({T}_{INOCT}) is the installed nominal operating cell temperature which for direct mount equals to Eq. (23). ({epsilon }_{back}) and ({epsilon }_{top}) are the emissivity of the back and front surface of the module, which equals to 0.89 and 0.84 respectively.
The algorithm of the model compares the load and generated energy from RSPP every hour. When the load exceeds the RSPP generation, generated energy will be consumed while importing energy from the grid to supply the deficit. Vice versa, when the RSPP generation is higher than the load, the excess of the generated energy will be exported to the grid. Equations (26 and 27) present the import and export mechanism of the model. According to the regulation, the imports and exports are accumulated each month to determine the monthly electricity bill. PLN compensates 65% of the exported energy to offset the imported energy. The offset amount is limited that the reduced imported energy can not be lower than the monthly minimum load, equal to 40 h usage of the maximum electricity capacity, which results in 220 kWh monthly minimum load or 2640 kWh annual minimum load (({E}_{min load})). If this limit is reached, the export excess will be deposited in the grid and offset up to the following three months’ imports before being reset to zero. Due to the limitation of the model, the offset balance reset mechanism is not modeled, and the model accumulated the imports and exports annually. Hence, the model may not account for the lost export deposits due to the reset mechanism. The net load, the reduced load due to the utilization of the RSPP, can be obtained by offsetting the imports with the compensated exports, expressed in Eq. (28). This value will be used in calculating the annual revenue of the RSPP, as further explained in the financial model in Fig. 5.
Figure 5
The RSPP gains its revenue from the annual electricity bill saving, calculated by subtracting the initial load with the final load after the utilization of RSPP multiplied with the electricity tariff (({C}_{el})). In each year, the financial model compares the net load with the minimum load set by the regulation for 25 years of the RSPP assumed life span. The net load is a year function due to the module’s degradation rate, while the minimum load is a constant. If the net load is higher than the minimum load, the bill equals the net load. Vice versa, when the net load is below the minimum load, the bill equals the net load, and the excess is deposited in the grid. In practice, this excess can offset future imports when the RSPP generation is low. However, this model only uses one-year irradiance data and repeats them for the assumed life span of the RSPP. Hence, RSPP generation output is relatively constant, which RSPP with a relatively large capacity may result in unutilized deposited exports. The annual revenue of the RSPP ((R(Y))), which accounts for the total cash inflow each year, is obtained from subtracting OPEX from the annual electricity bill saving. Darghouth et al. estimated that the CAPEX and OPEX for SPP in Indonesia is 1365.76 US$ and 24.38 US$/kW per year48. The annual revenues and the CAPEX of the RSPP are then used to calculate the system’s Net Present Value (NPV). Based on a market study in 2017, a typical proportion of RSPP CAPEX in Indonesia is consisting of 47% PV modules cost, 13% inverter cost and 40% other complementary components as well as installation cost49.The cost for PV modules and inverter in this paper are taken from a renewable energy technology distributor website since the estimate from Dargouth et al. lacks technical specification details needed for the simulation50. Just as other renewable energy technologies, RSPP is capital intensive. Therefore, the PV module with the lowest price/kW is employed to keep the CAPEX down and this study employs Suntech STP375S 375 W. Each module was modeled in the simulation at various RSPP capacities with the maximum electricity capacity as the upper limit or at 5.5 kWp. This study also employs 9 inverter types from Growatt of different sizes by matching the power output requirement of each simulation.The CAPEX includes PV modules (({C}_{PV})), inverter (({C}_{inv})) and other components and installation (({C}_{other})) as presented in Eq. (31). Table 4 presents the price/kW of the PV module, inverter and other costs used in the calculation. The technical specifications of the PV modules and inverters can be found as Supplementary Table S1 and S2 online.
Table 4 CAPEX components price per capacity.
The MEMR Ministerial Regulation regulates the electricity tariff in Indonesia. For social sector consumers with 5500 VA electricity capacity, the tariff has been 6.27 US¢/kWh for the last eight years. This tariff is subsidized as it only accounts for 65% of the BPP on average. The statistics report of PLN presents The BPP values from 2011–202,048,49, while the PLN’s Electricity Supply Business Plan (RUPTL) presents the projection of BPP values from 2021 to 203050. The BPP values beyond 2030 are estimated using a second-order polynomial regression function generated from the known values of BPP from 2015–2030. The BPP drop in 2015 was due to the increase in power system efficiency51. The BPP spike in 2025 is due to the significant increase in renewable energy adoption to achieve the Paris Agreement in that particular year. It is estimated that the BPP will rise at a slower rate due to the coal PP phase-out starting in 2030 and the decreasing cost of renewable energy technologies52. Thus, the projected electricity tariff from 2022 onwards is set at 65% of the BPP presented in Fig. 6. The currency exchange rate used in this paper is 1 USD = 14,351 IDR and 1 GBP = 1.33 USD.
Figure 6
BPP and electricity tariff for case study.
$${E}_{import}left(tright)=left{begin{array}{l}{E}_{load}left(tright)-{E}_{RSPP}left(tright), {E}_{load}left(tright)>{E}_{RSPP}(t)\ 0, {E}_{load}left(tright)le {E}_{RSPP}(t)end{array}right.$$
(26)
$${E}_{export}left(tright)=left{begin{array}{l}{E}_{RSPP}left(tright)-{E}_{load}left(tright), {E}_{RSPP}left(tright)>{E}_{load}(t)\ 0, {E}_{RSPP}left(tright)le {E}_{load}(t)end{array}right.$$
(27)
$${E}_{net load}(Y)=sum_{t=1}^{8760}({E}_{import}left(tright)-0.65 cdot {E}_{export}(t))$$
(28)
$$R(Y)=left{begin{array}{c}{E}_{net load}left(Yright) cdot {C}_{el}left(Yright)-OPEX, {E}_{net load}left(Yright)>{E}_{min load}\ {E}_{min load}left(Yright) cdot {C}_{el}left(Yright)-OPEX, {E}_{net load}left(Yright)le {E}_{min load}end{array}right.$$
(29)
$$OPEX=24.38left({I}_{mpp} cdot {V}_{mpp} cdot {n}_{M}right)/1000$$
(30)
$$CAPEX={C}_{PV}+{C}_{inv}+{C}_{mount}+{C}_{work}$$
(31)
The financial model economic analysis employs the Net Present Value method, which estimates the project outcome economic value: positive or negative. The NPV integrates the initial investment and the expected incomes and costs that occur during the operation of the RSPP into a series of cash flows adapted to the time value of money and risk. Equation (32) presents the calculation for NPV for the 25 years of the RSPP operational life span53. The discount rate ((r)) is determined from the weighted average capital cost (WACC). IRENA assumes a real WACC of 7.5% in OECD countries and China and 10% elsewhere globally for all types of technology54. The IEA assumes a WACC of 8% in developed countries and 7% in developing countries55. Steffen estimated that the WACC for PV system development projects in and outside the OECD countries was 5.4% and 7.4%56. This paper employs a WACC level of 7% for the discount rate.
$$NPV= -CAPEX+sum_{Y=0}^{25}frac{R(Y)}{{left(1+rright)}^{Y}}$$
(32)
RSPP design scenarios
As mentioned in the introduction, the current ecosystem may not be optimal for implementing RSPP on a mosque. Hence, several design scenarios are employed in the operational and financial model to explore the financial feasibility options.
-
The business as usual (BAU) scenario models the RSPPs with two objectives: to have the highest NPV and to meet the load that is not covered with the minimum load limit.
-
The second scenario, the carbon pricing (CP) scenario, models the RSPPs with additional income from the monetized carbon emission reduction. Climate Transparency estimated that approximately 761 g of carbon dioxide are emitted into the air when generating 1 kWh of electricity in Indonesia57. The Environmental Protection Agency (EPA) uses three sets of social cost estimates of carbon emissions (SCC) with different discount levels, as can be seen in Table 5 SCC58 is the discounted monetary value of the future climate change damages due to additional metric tons of carbon dioxide (CO2) emissions59,60. The money collected from the carbon pricing can also fund the energy transition and subsidize renewable energy implementation. This paper uses the SCC at 5% discount factor, closest to the discount factor used in the financial model.
-
The third scenario is the elimination of the minimum load limit (MLL). The minimum load limit prevents the RSPP from entirely supplying the load. This scenario entails removing the minimum imports requirement. Such support policy was temporarily implemented in August 2020 through a MEMR Ministerial Decree to deal with the impact of COVID-1961.
-
The fourth scenario is the rework of the Net-Metering Scheme (NMS). In this scenario, the NMS is reworked to compensate 100% of the exported energy. The GOI has planned this adjustment by revising Ministerial Regulation No. 49/2018 on RSPP in 202162.
-
The fifth scenario entails the enforcement of a non-subsidized electricity tariff (NST). The higher electricity tariff may drive customers to switch their means to meet their energy needs while enabling RSPP to gain more income. In this scenario, the BPP values will be used as the electricity tariff.
Table 5 SCC in US$/ton of carbon emission at different discount factors.
Table 6 Irradiance potential on Ontowiryo mosque rooftop.