Life-cycle ecological footprint assessment of grid-connected rooftop solar PV system

ABSTRACT

Solar photovoltaic (PV) power generation system is generally considered to be land-intensive in view of the diffuse nature of solar energy. However, a comprehensive assessment in this regard would involve bio-productive land for all types of resource consumption during the lifespan of the system. This study presents such an assessment for a grid-connected rooftop solar photovoltaic (RSPV) system located in a tropical climate. The life cycle ecological footprint (EF_T) methodology has been used to derive the results based on material & energy consumption, water & manpower requirements along with waste disposal. This methodology is illustrated through an existing 1 MW_p RSPV system installed at an educational institution in India. The total EF_T is evaluated as 22.69 global hectare (gha) per year, out of which the environmental impact of solar PV panels is 76% of the total. The EF_T per kW_P capacity is 0.454 gha/kW_P and the EF_T per unit solar panel collector area is 0.074 gha/m². The EF_T of the RSPV system per unit electricity generation is 1.59 × 10⁻⁵ gha/kWh. However, considering only emissions from the grid-electricity in India, an approximate reduction potential of ecological footprint through RSPV-based electricity is estimated as 95%.

KEYWORDS:

• Life cycle assessment
• bio-productive land
• ecological footprint reduction potential
• solar PV system

1. Introduction

India being a tropical country receives adequate solar radiation for nearly 300 days annually amounting to about 2300–3200 sunshine hours/year. Almost all the regions receive 4–7 kWh of solar radiation per square metre depending upon the location. The country’s solar installed capacity reached 28.18 GW as of 31 March 2019 (CEA, 2019). In the year 2015, the target was raised to 100 GW of solar capacity (including 40 GW from rooftop solar) by 2022, targeting an investment of US$ 100 billion (Ahmad, Samuel, and Amar 2016; MNRE, 2018).

The Life Cycle Ecological Footprint-based studies for power generating systems were not found in literature. However, several studies have been reported in the area of life cycle energy and carbon emissions from rooftop solar photovoltaic (RSPV) systems. Among the renewable energy systems, photovoltaics have minimal environmental impacts (i.e. 3.33% of coal-fired power generation) (Yu et al. 2017). The environmental impacts of the 52.7 kW_p rooftop PV system were reported by Eskew et al. (2018). The system produced climatic change impact of 0.079 kg CO_2-eq/kWh and was estimated to avoid 1 × 10⁶ kgCO_2-eq over its lifetime. Ludina et al. (2018) reviewed and summarised impact assessment methods of life cycle assessment (LCA) namely, cumulative energy demand (CED), energy payback time (EPBT), and greenhouse gas (GHG) emission rate and found that multi-crystalline silicon PV technology has energy consumption of 1034–5150 MJ/m², the EPBT of 0.8–4.17 years, and greenhouse gas emissions rate of 12.1–569.0 gCO_2eq/kWh. Peng, Lun, and Yang (2013) examined the EPBT and GHG emissions from mono-crystalline silicon (mono-Si), multi-crystalline silicon (multi-Si), amorphous silicon (a-Si), CdTe thin film and CIS thin film. They found that for mono-Si PV systems, the EPBT varied from 1.7 to 2.7 years and GHG emission rate was in the order of 29–45 gCO_2eq/kWh. For multi-Si PV system, the EPBT and GHG emission rates were 1.5–2.6 years and 23–44 gCO_2eq/kWh, respectively; and for thin-film PV systems (a-Si, CdTe and CIS), the EPBT and GHG emission rates were within the range of 0.75–3.5 years and 10.5–50 gCO_2eq/kWh, respectively. Fthenakis and Alsema (2006) studied the EPBT and GHG emissions for rooftop mono-Si and multi-Si PV modules and concluded that under the solar irradiation of 1700 kWh/m²/year, the EPBT for rooftop mono-Si and multi-Si were 2.7 and 1.7 years, respectively; and GHG emissions for mono-Si and multi-Si were 45 and 37 gCO_2eq/kWh, respectively. Kato et al. (2001) conducted life cycle assessment (LCA) on CdS/CdTe PV modules of residential rooftops to estimate and compare primary energy requirements, EPBT, and CO₂ emissions with multi-Si and a-Si at different annual production rates. The study showed that the EPBT of CdS/CdTe PV modules were 1.7 year at 10 MW/year, and 1.1 year at 100 MW/year, which are lower than the EPBT of a-Si and multi-Si. The life cycle GHG emissions of CdS/CdTe PV modules ranged from 8.9 to 14.0 gCO_2eq/kWh for annual production from 10 to 100 MW/year. Koo et al. (2017) reported that the electricity generation of a-Si and CIGS base smart PV system blind for windows was about 81.3 kWh/m² and 145.8 kWh/m², respectively. Ordóñez et al. (2010) reported that in Andalusia (Spain), the estimated PV solar energy capacity for residential buildings (if installed on all of them) was 9.73 GW/year. With the total roof surface area of 265.52 km², this would satisfy 78.79% of energy needs of the city.

Pacca, Sivaraman, and Keoleian (2007) compared poly-Si (13% efficiency) with a-Si (6.3% efficiency) module-based rooftop solar PV systems (33 kWh system capacity) that were installed in the University of Michigan. The EPBT of poly-Si and a-Si are 2.4 and 7.4 year, respectively. For the CO₂ emissions, it comes in the range of 34.3 and 72.4 g of CO₂/kWh, respectively. The life cycle assessment of multi-Si module-based solar PV systems with the irradiation of 1427 kW h/m²/yr installed in Japan showed that the EPBT and GHG emission rates were 2.4 years and 20 g CO_2-eq./kWh, respectively (Kota et al. 1998). Stoppato (2008) presented the EPBT and emissions results of an LCA of poly-Si module panels in 27 countries. The EPBT ranged from 3.368 years in Spain to 6.522 years in the United Kingdom, and CO₂ emissions ranged from 0.002 kgCO₂/kWh in Norway to 0.841 kgCO₂/kWh in Australia.

Osma-Pinto and Ordóñez-Plata (2019) studied the measuring factors influencing the performance of rooftop PV panels in warm tropical climates. They found that, on average, a green roof-top solar PV system can generate between 1.0 ± 0.4% and 1.3 ± 0.4% more energy per day than a PV system installed on a concrete roof at an average height of 50 cm above the roof surface. In addition, air velocities exceeding 1 m/s may enhance the daily average energy generation by 1 ± 0.4% and 3 ± 0.4%.

In India’s context, Prakash and Bansal (1995) evaluated the energy consumption in mono-Si solar PV module production in India, and examined its implications for large-scale introduction of solar PV plants in the country. They evaluated the energy payback period as approximately 4 years, which was comparable to energy payback periods of similar modules produced internationally. Nawaz and Tiwari (2006) examined a 1.2 kW_p (capacity) PV system for mud house at IIT Delhi and reported that the EPBT is in the range of 7–26 years and the net CO₂ saved is 0.24–0.77 kg/kWh, depending on solar panel’s efficiency and climatic condition of India.

Table 1 summarises the results from previous studies based on rooftop solar PV system with poly-Si-based module. This summary indicates that for different solar irradiations, system capacities, and module efficiencies, the EPBT of the rooftop solar PV (poly-Si module) for most of the cases is in the range of 0.8–7.4 years. The GHG emissions produced during the life of the rooftop solar PV is in the range of 12–840 gCO_2-eq/kW h.

Table 1. Literature review summary

S.No.	Authors	System Specification	Irradiation (kWh/m^2/yr)	EPBT (yr)	Emission	Remark
1	Eskew et al. 2018	52.7 kWP, 16.2% efficiency	-	2.5	0.079 kgCO2-eq/kWh	Roof-top; poly-Si module,
2	Ludina et al. 2018		1200–1700	0.8–4.17	12.1–569.0 gCO2eq/kWh	Rooftop, poly-Si module, EE 1034–5150 MJ/m^2
3	Peng, Lun, and Yang 2013	11.6–14% efficiency	1117–1700	1.5–2.6	23–44 g CO2-eq./kW h	2699–5150 MJ/m^2
4	Wild-Scholten 2009	13.2% efficiency	1700	1.75	29 g CO2-eq/kW h	Rooftop, frameless, poly-Si module
5	Stoppato 2008	Range of efficiency of most module are about 16%	890–1754	3.368–6.064	0.058 − 0.840 kgCO2/kWh	Rooftop, poly-Si module, Annual production (kWh/kWp) 754–1460
6	Pacca, Sivaraman, and Keoleian 2007	33 kWP, 13% efficiency	4–6.5 kWh/m^2/day	2.7–7.4	34.3 g CO2-eq/kW h	Rooftop, poly-Si module
7	Jungbluth, Dones, and Frischknecht 2007	13.2% efficiency	1117	2.9	60 g CO2-eq/kW h	Rooftop, Frame, poly-Si module
8	Nawaz and Tiwari 2006	1.2 kWP, 8–11% efficiency	800–1200 W/m^2	13.26–18.14	38.64 kg CO2/yr/m^2	Rooftop, poly-Si module, EE 1380 kWh/m^2
9	Battisti and Corrado 2005	10.7% efficiency	1530	3.3	NA	Rooftop poly-Si module
10	Ito et al. 2003	12.8% efficiency	1675	1.7	12 g CO2-eq/kW h	Rooftop, poly-Si module

Ecological Footprint: The Ecological Footprint (EF) concept was developed in the mid-nineties by Mathis Wackernagel and William Rees (Wackernagel and Rees 1996). It can be used to examine various measures such as the feasibility of resource use, distribution of the world’s natural resources, waste assimilation, and overall sustainability of a system. The indicator incorporates all inputs (resources) and converts them into a single parameter called ‘global hectare (gha)’. One global hectare (gha) is equivalent to 1 ha of bio-productive land with world average productivity (Bastianoni et al. 2006). The expression for evaluating EF is as follows:

Ecological Footprint (EF) = ∑ ($\frac{C_i}{Y_i}$).e_i,

where C_i is annual consumption of ith product (ton/year), Y_i is yield factor (annual productivity) of ith product (ton/ha), e_i is equivalence factor of different bio-productive land types. The equivalence factor (e_i) of different land types (cropland, pastureland, forestland, marine land, etc.) is provided in Table 2.

Table 2. Equivalence factor (GFN, 2016)

Bio-productive land	Equivalence factor ei(gha/ha)
CO2 absorption land ($e_{C{O}_{2}land}$) Forest land ($e_{forestland}$) Crop land ($e_{cropland}$) Pasture land ($e_{pastureland}$) Sea productive/marine land ($e_{marineland}$)	1.28 1.28 2.52 0.43 0.35

This study focuses on the life cycle ecological footprint assessment of an RSPV system located in tropical climatic zone of India. This study is significant, as a huge enhancement in installation and commissioning of RSPV systems in India is envisaged in the near future. This study will help in the direction of overall estimation of natural resource demand for India’s renewable energy transition. Such a study may also be helpful in exploring the potential reduction feasible in the EF of the power sector of the country.

2. Methodology

The method used in this study to compute the life cycle ecological footprint of an RSPV system is explained below:

2.1. Life cycle ecological footprint of a roof-top solar PV system (EF_T)

RSPV systems are responsible for natural resource consumption for manufacture &transportation, manpower requirement, waste generation, etc., during their life span. In order to evaluate EF_T, the direct and indirect resource consumption related to RSPV systems are converted into their equivalent productive land that is needed to produce or absorb their impacts in the form of land categories (such as CO₂ land, forestland, cropland, etc.). The various stages of the RSPV system’s life cycle are shown in Figure 1. The principle of assessment of the EF_T is depicted in Figure 2. The estimation of EF_T is given by Equation (1)(1) ${\text{EF}}_{\text{T}}{\text{= EF}}_{\text{m}}{\text{+ EF}}_{\text{e}}{\text{+ EF}}_{\text{t}}{\text{+ EF}}_{\text{w}}{\text{+ EF}}_{\text{mp}}{\text{+ EF}}_{\text{we}}\text{,}$(1) :

(1) ${\text{EF}}_{\text{T}}{\text{= EF}}_{\text{m}}{\text{+ EF}}_{\text{e}}{\text{+ EF}}_{\text{t}}{\text{+ EF}}_{\text{w}}{\text{+ EF}}_{\text{mp}}{\text{+ EF}}_{\text{we}}\text{,}$(1)

where, EF_m, EF_e, EF_t, EF_w, EF_mp and EF_w represent the life cycle ecological footprint of RSPV systems’ material consumption, direct energy consumption, transportation, water consumption, manpower and waste generation of the RSPV systems, respectively. All the above parameters cumulatively assess the EF_T of RSPV system.

Figure 1. System boundary of the life cycle stages of RSPV system

* represents Heavy Duty Vehicle (HDV).

Figure 2. The life cycle ecological footprint of RSPV system

2.1.1. Ecological footprint of materials consumption (EF_m)

The EF_m may be determined by considering the embodied energy of total material consumption during the lifecycle of an RSPV system. The EF_m of an RSPV system is calculated by using Equation (2)(2) ${\text{EF}}_m=\{{\sum }^{}(M_i.E{E}_i).{\lambda }_m(\frac{1-Aoc}{Af})\}.e_{C{O}_{2}land},$(2) :

(2) ${\text{EF}}_m=\{{\sum }^{}(M_i.E{E}_i).{\lambda }_m(\frac{1-Aoc}{Af})\}.e_{C{O}_{2}land},$(2)

where $M_i$ is the material consumed during the lifecycle of an RSPV system and $E{E}_i$ is the embodied energy associated with that material; ${\lambda }_m$ represents corresponding emission factor (tCO₂/MJ); A_oc represents percentage of CO₂ absorption in oceans i.e. 0.3 (SIO, 2017), and A_f is CO₂ absorption factor of forests i.e. 2.68 tCO₂/ha (Mancini et al. 2016). The term $e_{C{O}_{2}land}$ represents the equivalence factor of CO₂ absorption land (Table 2).

2.1.2. Ecological footprint of direct energy consumption (EF_e)

The EF_e depends on the amount of direct energy consumed during the life cycle of RSPV system. The estimation of EF_e is given by Equation (3)(3) $E{F}_e=({\sum }^{\text{E\_l}}.{\alpha }_e+{\sum }^{}{C}_{fuel}.{\alpha }_{fuel}).(\frac{1-A_{oc}}{A_f}).e_{C{O}_{2}land},$(3) :

(3) $E{F}_e=({\sum }^{\text{E\_l}}.{\alpha }_e+{\sum }^{}{C}_{fuel}.{\alpha }_{fuel}).(\frac{1-A_{oc}}{A_f}).e_{C{O}_{2}land},$(3)

where $E_l$and $C_{fuel}$ are the amount of electricity and fossil fuel used, respectively; ${\alpha }_e$(i.e. 0.82 tCO₂/MWh; Ministry of Power Central Electricity Authority, Government of India (MPCEA) 2018) and ${\alpha }_{fuel}$ (i.e. 3.17 kgCO₂/kg of diesel; EEA, 2013) are the emission factor for electricity and emission factor for fossil fuel, respectively.

2.1.3. Ecological footprint of transportation (EF_t)

The EF_t depends on three parameters: first, RSPV system materials’ transportation from the production site to the onsite RSPV project; second, manpower transportation from their houses to the onsite RSPV project; and third, waste disposal from the project site to landfill area/recycling plant. The assumptions for the estimation of the EF_t are as follows:

1. RSPV system materials and wastes are transported through Heavy Duty Vehicles (HDV).

2. Labourers use diesel-fuelled buses to reach the site of the RSPV project from their houses.

3. The average distance travelled by labourers is 5–10 km and wastes are transported to a distance of 10–15 km.

Estimation of the EF_t of material, manpower and wastes is done by Equation (4)(4) $E{F}_t=\{({\sum }^{}\frac{X_{mi}.T_{mi}}{C_{HDV}}+{\sum }^{}\frac{Y_{wj}.T_{wj}}{C_{bus}}).F_{HDV}+\sum \frac{Z_k.T_k}{C_{bus}}.F_{bus}\}.{\alpha }_{fuel}.(\frac{1-{\text{A}}_{oc}}{Af})e_{C{O}_{2}land},$(4) :

(4) $E{F}_t=\{({\sum }^{}\frac{X_{mi}.T_{mi}}{C_{HDV}}+{\sum }^{}\frac{Y_{wj}.T_{wj}}{C_{bus}}).F_{HDV}+\sum \frac{Z_k.T_k}{C_{bus}}.F_{bus}\}.{\alpha }_{fuel}.(\frac{1-{\text{A}}_{oc}}{Af})e_{C{O}_{2}land},$(4)

where $X_{mi}$and $T_{mi}$ are consumption and the average distance of transportation of ith material during RSPV system life, respectively. $Y_{wj}$ (i.e. 10 km) and $T_{wj}$ (i.e. 12 km) are waste generation and the average distance travelled by the jth material, respectively. $Z_k$ and $T_k$ (i.e. 7 km) are number of labourers and average distance travelled by them, respectively. $C_{HDV}$ (i.e. 3.5 tonnes) and $C_{bus}$ (i.e. 50 passengers) represent the capacity of HDV and Bus, respectively. ${\alpha }_{fuel}$ is emission factor of fossil fuel (i.e. 3.17 kgCO₂/kg of diesel; EEA, 2013), F_HDV (i.e. 0.222 kg of fuel/km) is average fuel efficiency of HDV and F_bus (i.e. 0.222 kg of fuel/km) is average fuel efficiency of the bus (Baidya and Borken-Kleefeld 2009).

2.1.4. Ecological footprint of labour/manpower (EF_mp)

The manpower requirement for an RSPV system is categorised into three categories: (a) skilled labour, (b) semi-skilled labour and (3) unskilled labour. The total manpower requirements during an RSPV lifespan are estimated in terms of full-time equivalent (FTE). FTE is the ratio of the total numbers of paid hours during a period (part time, full time, contracted) by the number of working hours in that period. An employment opportunity of one FTE is a job that exists for one person to work on for the duration of 1 year. The calculation assumes 260 working days per calendar year (CEEW and NRDC, 2014, 2015). The total number of manpower requirements during the lifecycle of an RSPV system is estimated according to a report by the Skill Council of Green Jobs (SCGJ, 2016); which is depicted in Table 3.

Table 3. Employment estimate for RSPV system

Employee (Type)	Employee requirement (Labour days/MW)	Charges per person (Rs/day^a)
Skilled	3.03^b x 260 = 788	1500
Semi- skilled	18.06^b x 260 = 4696	1000
Unskilled	2.5^b x 260 = 650	350

^aOne working day is equivalent to 8 hours of work; ^bFTE estimation according to SCGJ report (SCGJ, 2016).

The EF_mp is associated with the metabolic rate of labour/manpower for the different type of activities (Martínez-Rocamora, Solís-Guzmán, and Marrero 2017). However, the impact of manpower’s mobility has already been considered in Section 2.1.3. It is assumed that nearly 60% of the daily metabolic calories (i.e. 2400 kcal/day; National Institute of Nutrition (NIN) 2018) are burned by manpower during work activities. The EF_mp is determined by Equation (5)(5) $\mathrm{E}{\mathrm{F}}_{mp}=\frac{d_w}{365}.f_d\left\{{\sum }^{\left(\frac{C_{f_i}}{Y_{f_i}}\right)}.e_i+{\sum }^{\left(C_{fue{l}_j}.{\lambda }_j\right)}.\frac{\left(1-S_{oc}\right)}{A_f}.e_{C{O}_{2}land}\right\}$(5) .

(5) $\mathrm{E}{\mathrm{F}}_{mp}=\frac{d_w}{365}.f_d\left\{{\sum }^{\left(\frac{C_{f_i}}{Y_{f_i}}\right)}.e_i+{\sum }^{\left(C_{fue{l}_j}.{\lambda }_j\right)}.\frac{\left(1-S_{oc}\right)}{A_f}.e_{C{O}_{2}land}\right\}$(5)

where d_w is total number of labour-days worked during the life of RSPV system; f_d is food intake fraction of an Indian adult diet assigned to work activity (i.e. 60%), C_fi is food consumption category i (kg/person-year); Y_fi is food production yield of ith food item (kg/ha-yr). C_fuelj is fuel consumption category j (i.e. LPG, kerosene, etc.) during cooking (kg/person-year), λ_j is emission factor of fuel category j. The consumption of different food items in Indian households and their EF are depicted in Table 1A (Annexure).

2.1.5. Ecological footprint of water use (EF_w)

The EF_w depends on water use during the lifecycle of an RSPV system. Groundwater is generally used in RSPV system construction in India, therefore CO₂ absorption land is needed to compensate for the electrical energy for uplifting water from underground to overhead tanks. The EF_w of water use is determined by Equation (6)(6) $E{F}_w=C_w.\{E_w.{\alpha }_e.\frac{(1-A_{oc})}{A_f}\}{e}_{C{O}_{2}land},$(6) :

(6) $E{F}_w=C_w.\{E_w.{\alpha }_e.\frac{(1-A_{oc})}{A_f}\}{e}_{C{O}_{2}land},$(6)

where C_w is total water use during all phases of RSPV system (m³); E_w is electricity consumption to extract groundwater (kWh/m³); ${\alpha }_e$ is emission factor of electricity used for groundwater extraction.

2.1.6. Ecological footprint of wastes (EF_we)

RSPV systems’ wastes generally result in landfill disposal after transportation of the waste materials. Its transportation is included in Section 2.1.2. The EF_we is determined by Equation (7):

(7) ${\text{EF}}_{\text{we}}={\sum }^{}\frac{(1-\text{r}){\text{X}}_{{\text{W}}_{\text{i}}}}{{\text{M}}_{{\text{w}}_{\text{i}}}}.{\text{e}}_{\text{landfill}}\text{,}$(7)

where X_wi represents the total volume of waste generated during RSPV system life (m³) and r represents the recycled fraction of the PV system (if any). M_wi represents the dump capacity of waste disposal (m³/ha); and $e_{landfill}$ represents the equivalence factor of the type of bio-productive land for the wastes to be disposed (0.48 gha/ha pasture land; GFN, 2016).

2.2. Details of RSPV and climatic condition of the location

The RSPV system used in this study is of 1 MWp capacity; installed and commissioned by Renew Solar Energy Pvt. Ltd, India. The System is installed at Motilal Nehru National Institute of Technology, Allahabad, Prayagraj, UP, India. The photographic and satellite images of the RSPV system are depicted in Figure 3(a,b) respectively.

Figure 3. (a) Photographic image of RSPV system, (b) Satellite image of RSPV system

The city of Prayagraj is located in a tropical climatic zone, having geometrical coordinates of 25º27ʹ N and 81º51ʹ E. The ambient air temperature reaches up to 50ºC in summer time during peak hours, while during winter time, it drops up to 2ºC during night time. Average annual precipitation has been recorded as 978.5 mm with average rainy days of 47.9 during the year 2018. Also, the mean monthly sunshine hours are recorded as 2773.7 hours with average daily solar radiation of 4.93 kWh/m²/day. The relative humidity varies between 29% and 81% during the year.

3. Results and discussion

3.1. Ecological footprint of the RSPV system (EF_T)

The EF_T of the RSPV system as evaluated from Equation (1)(1) ${\text{EF}}_{\text{T}}{\text{= EF}}_{\text{m}}{\text{+ EF}}_{\text{e}}{\text{+ EF}}_{\text{t}}{\text{+ EF}}_{\text{w}}{\text{+ EF}}_{\text{mp}}{\text{+ EF}}_{\text{we}}\text{,}$(1) is 453.84 gha (20 years of life), while the annual average EF_T is about 22.69 gha/yr. The EF_T per kW_P capacity is 0.454 gha/kW_P and the EF_T per unit solar panel collector area is 0.074 gha/m². A break-up of the EF_T in its various components is depicted in Figure 4. Solar PV panels contribute as much as 76% to EF_T, followed by Balance of System (BOS) that contributes 15%. The detailed break-up of the results is as follows:

Figure 4. Break-up of the EF_T distribution

3.1.1. Ecological footprint of materials consumption (EF_m)

The embodied energy of solar PV modules and Balance of System (BOS) of the RSPV system are given in Table 4.

Table 4. Embodied energy of rooftop solar PV system

Component Embodied energy(MJ/m^2)
Solar PV module poly crystalline(multi-Si) 2699–3120 (Jungbluth, Stucki, and Frischknecht 2009; Wild-Scholten 2009; Alsema and Wild-Scholten 2007) BOS Inverter (per kWP capacity) 503 (Pacca, Sivaraman, and Keoleian 2007) Frame mild steel 32.24 (MJP/kg) (Praseeda, Reddy, and Mani 2015) Installation 448 MJ/kW (Sumper et al. 2011)

The EF_m of RSPV system as calculated by Equation (2)(2) ${\text{EF}}_m=\{{\sum }^{}(M_i.E{E}_i).{\lambda }_m(\frac{1-Aoc}{Af})\}.e_{C{O}_{2}land},$(2) is 412.46 gha (i.e. 91% of the total EF_T). The detailed specifications of the RSPV system along with their environmental impact are provided in Table 5. The EF of solar PV panels is 83.5% (i.e. 344.25 gha) of the EF_m of RSPV system.

Table 5. Specifications of the different parts of Installed solar PV system

Component	Specifications		EF(gha)
Multi- Si panel	Rated power Module efficiency Open circuit voltage – Voc (Volts) Short circuit current – Isc (Amps) Dimensions: length x width x height Operating temperature range No of Panels PV String size Total Panel area Embodied energy (EE)	320Wp 16.41 % 45.7 9.06 1970 mm x 990 mm x 35 mm - 40 to 85°C 3125 Nos. 20 series; 8 parallel 6094.68 m^2 2699–3120 MJ/m^2 panel area (Jungbluth, Stucki, and Frischknecht 2009)	344.25
Balance of System (BOS)
Frame	Mild steel – 52.6 ton EE – 32.24 MJ/kg (Praseeda, Reddy, and Mani 2015)		52.62
Inverter	Model Nominal input Voltage PV input operating Voltage range Maximum operating PV input current Efficiency EE	KSG-50 K 620 V 250–950 V 108 ADC 98% 503 MJ/kWp (Pacca, Sivaraman, and Keoleian 2007)	9.76
Cables	PV panels to inverter – 900 m. Inverter to metering system – 300 m. Grounding cables – 120 m Material of all cables is copper.		1.04
Concrete	Cement – 450 packets (50 kg each) Aggregate – 47.8 m^3 Sand – 23.9 m^3		4.79

3.1.2. Ecological footprint of direct energy consumption (EF_e)

The environmental impact of direct energy of RSPV system mainly occurs in installation and commissioning phase of system. Due to lack of data about the direct energy consumption in RSPV system, the assumption of energy consumption in installation of rooftop system is taken as 448MJ/kW_p (Sumper et al. 2011). For one MW_p RSPV system, the direct energy consumption has been estimated as 4.48 x 10⁵MJ. The EF_eas calculated by using Equation (3)(3) $E{F}_e=({\sum }^{\text{E\_l}}.{\alpha }_e+{\sum }^{}{C}_{fuel}.{\alpha }_{fuel}).(\frac{1-A_{oc}}{A_f}).e_{C{O}_{2}land},$(3) is 8.69 gha (i.e. 1.9% of the total EF_T).

3.1.3. Ecological footprint of transportation (EF_t)

Location of installation site of RSPV is an important factor in deciding the EF of transportation. In this study, the solar PV modules, steel frames, inverter and cables were transported from Delhi city to Prayagraj city by heavy duty vehicles (distance: approx. 700 kilometres). The EF of transportation as calculated by Equation (4)(4) $E{F}_t=\{({\sum }^{}\frac{X_{mi}.T_{mi}}{C_{HDV}}+{\sum }^{}\frac{Y_{wj}.T_{wj}}{C_{bus}}).F_{HDV}+\sum \frac{Z_k.T_k}{C_{bus}}.F_{bus}\}.{\alpha }_{fuel}.(\frac{1-{\text{A}}_{oc}}{Af})e_{C{O}_{2}land},$(4) is 5.82 gha (i.e. 1.1% of the total EF_T).

3.1.4. Ecological footprint of manpower (EF_mp)

The total manpower requirement is 6134 labour-days. The EF_mp as estimated by Equation (5)(5) $\mathrm{E}{\mathrm{F}}_{mp}=\frac{d_w}{365}.f_d\left\{{\sum }^{\left(\frac{C_{f_i}}{Y_{f_i}}\right)}.e_i+{\sum }^{\left(C_{fue{l}_j}.{\lambda }_j\right)}.\frac{\left(1-S_{oc}\right)}{A_f}.e_{C{O}_{2}land}\right\}$(5) is 5.11 gha (i.e. 1.1% of the total EF_T).

3.1.5. Ecological footprint of water use (EF_w)

Water is mainly used during installation and operation phase of an RSPV system. To calculate water use during the entire life of RSPV system, 3.5–5 litres of water per m² of solar PV surface (EIA, 2016) have been considered. The EF_w of RSPV system as calculated by Equation (6)(6) $E{F}_w=C_w.\{E_w.{\alpha }_e.\frac{(1-A_{oc})}{A_f}\}{e}_{C{O}_{2}land},$(6) is 0.094 gha (i.e. 0.02% of the total EF_T).

3.1.6. Ecological footprint of waste generation (EF_we)

Recycling/reuse of waste materials may reduce ecological impact but it is not evaluated in this study (r = 0). However, the EF of landfill for the waste generated is evaluated. The EF_we of the RSPV system as calculated by using Equation (7) is 0.0029 gha.

3.2. Potential of EF reduction through RSPV system

The total monthly electricity supply by national grid and RSPV system in the MNNIT campus (for the year of 2018) is depicted in Figure 5. The RSPV system contributed only 10–13.6% of the total monthly electricity consumption of the MNNIT campus, hence all the electricity produced by the RSPV system is consumed within the campus and not exported to the grid. It indicates that the inter-annual variation of electricity consumption in the building does not influence our results, as no fraction of the electricity produced from RSPV is exported to the grid.

The life cycle electricity generation has been calculated with RETScreen software and the simulation results of monthly electricity generation are shown in Figure 6. For the year of 2018, the actual and simulated annual electricity generation are 1523 MWh and 1544.5 MWh, respectively. The maximum positive variation in any month is recorded as 5.68% in the month of June and maximum negative variation is recorded as 4.65% in the month of September. Annual variation in actual and simulation electricity generation is recorded as 1.39%. The estimation of total electricity generation from the RSPV system is about 30,460 MWh, during the lifespan of 20 years. The total life cycle emission mitigation by the RSPV system is about 1288.5 tCO₂. If the time-varying CO_2eq emissions from the grid in future are considered, it will influence only the mitigation potential of the RSPV system. However, it does not influence the total EF_T of the RSPV system.

Figure 5. Electricity consumption in the MNNIT campus

Thus, the life cycle ecological footprint of electricity generation by the RSPV system is only 1.49 × 10⁻⁵ gha/kWh. Considering emission factor of grid electricity in India as 0.82 tCO₂/MWh (Ministry of Power Central Electricity Authority, Government of India (MPCEA) 2018), the estimated ecological footprint of grid electricity generation is 2.72 × 10⁻⁴ gha/kWh (ignoring factors other than emissions). Hence, the EF reduction potential by the RSPV system is about 95%. It may be pointed out here that although the physical land requirements for conventional (e.g. coal based) plants may be lower than that of RSPV systems; however, the total bio-productive land requirement may be very high, particularly in the form of CO₂ absorption land. An accurate assessment of EF reduction potential of RSPV system may be possible when detailed life cycle ecological footprint studies are available for conventional power generating systems.

Figure 6. Monthly electricity generation by the RSPV system

3.3. Economic analysis

The life cycle cost (LCC) is an economic analysis tool that is used in the selection of alternatives. It compares initial investment options and identifies the least cost alternatives for selected lifespan. It consists of initial cost and yearly cost of operation and maintenance. The LCC may be estimated by Equation (7) (Kranti 2010):

(7) $LCC=IC+US(d,N)AC.$(7)

where IC represents the initial cost, AC represents annual cost and USPW is universe-series present worth factor. The USPW is calculated by Equation (8):

(8) $USPW=\frac{1-{(1+d)}^{-N}}{d},$(8)

where d represents discount rate and N represents the number of years for which calculation is performed.

The total initial cost of the RSPV system is 69.15 million rupees. The USPW (d = 5%, N = 20 years) as estimated by Equation (8) is 12.46. The total annual cost is Rs. 15,632. The LCC of the RSPV system as calculated by Equation (7) is 69.34 million rupees.

The payback period of the RSPV system is about 5.2 years (electricity rate at 8.75 Rs/kWh (UPERC, 2019)), Intern Rate of Return (IRR) is about 18.6%, and Net Present Value (NPV) is about 197.11 million rupees. The cash flow diagram (20 years of lifespan) of the RSPV system is depicted in Figure 7. The maintenance cost and salvage value of the RSPV system is about 10,000 Rs/year and 10% of the capital cost of the RSPV system are considered, respectively.

The life cycle power generation is evaluated as 30,460 MWh, the levelized cost of electricity (LCOE) of the RSPV system is estimated as Rs 2.27/kWh. However for off-grid and hybrid systems, the LCOE of poly-crystalline module-based PV system is about 0.32–0.37 $/kWh (Jahid et al. 2018, 2019). This estimated value of LCOE is close to the LCOE of solar PV (i.e. 42–62 $/MWh) of India reported by Deshmukh et al. (2019).

Figure 7. Cash flow diagram of the RSPV system

4. Conclusions

Renewable energy systems are considered as clean sources of energy through land-intensive. However, it consumes some fossil energy and other resources as well as produces emissions and wastes during its lifespan. Therefore, it imprints some environmental impact on the planet. In this study, the life cycle EF of RSPV system is estimated as 453.84 gha. The EF_T per kW_P capacity is 0.454 gha/kW_P and the EF_T per unit solar panel collector area is 0.074 gha/m². The EF contribution of solar PV panels during their lifespan is 344.25 gha (i.e. approx. 76% of the total EF_T), while the EF of BOS is 68.22 gha (i.e. approx. 15% of EF_T). The estimation of lifecycle electricity generation from the RSPV system is about 30,460 MWh.

The LCC of RSPV system is calculated as Rs. 69.3million for the lifespan of 20 years; the solar PV panels and BOS contribute 49% (34 million Rs) and 43% (30 million Rs) of the total LCC of the RSPV system, respectively. The LCOE of the RSPV system is estimated as Rs 2.27/kWh.

The EF of electricity generation by the RSPV system is only 1.49 x 10⁻⁵gha/kWh. The EF of electricity generation by the RSPV system is about 95% lower than that of the grid electricity. The results obtained from such a study may be helpful in proposing and evaluating strategies for reduction in EF of power generating systems. The life cycle ecological footprint assessment of building integrated RSPV system should also be performed to facilitate sustainable building design.

Further studies need to be carried out to assess the life cycle ecological footprint of other renewable energy systems (e.g. ground-mounted solar PV, wind, biomass, small-hydro, etc.), so as to facilitate a comparative assessment of different power generating systems as well as their EF reduction potential.

Abbreviation

EF: ecological footprint

EF_m: life cycle ecological footprint of material consumption

EF_e: life cycle ecological footprint of energy

EF_mp: life cycle ecological footprint of manpower

EF_T: life cycle ecological footprint

EF_t: life cycle ecological footprint of transportation

EF_w: life cycle ecological footprint of water use

EF_we: life cycle ecological footprint of waste disposal

e_i: equivalence factor

gha: global hectare

LCA: life cycle cost

RSPV: Roof-top solar photovoltaic

BOS: Balance of System

LCOE: Levelized cost of electricity

GHG: greenhouse gas

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Ajay Biswas

Ajay Biswas is a postgraduate student of Department of Mechanical Engineering at the Motilal Nehru National Institute of Technology, Allahabad (UP), India. His academic interest lie in the domain of renewable energy systems.

Dilawar Husain

Dilawar Husain is currently pursuing doctoral programme in the Department of Mechanical Engineering at the Motilal Nehru National Institute of Technology, Allahabad (UP), India. His academic interests include life cycle analysis, renewable system technologies and sustainability.

Ravi Prakash

Dr Ravi Prakash is currently working as a Professor in the Department of Mechanical Engineering at the Motilal Nehru National Institute of Technology, Allahabad (UP), India. His academic interests include energy management, life cycle energy analysis, and sustainability. He is a Fellow of the Institution of Engineers (India) and a member of the International Solar Energy Society.

Annexure

Table 1A. Main raw materials for food preparation per capita in India

Monthly Total annual CO2 emission Yield Production EF consumption consumption factor (ton/ha) (gha) (NSSO, 2014)

Cereals 9.28 kg 111.36kg 2.39 (OGD, 2017) 0.117 Pulses 0.90 kg 10.8 kg 0.69 (OGD, 2017) 0.039 Vegetable 8.4 kg 100.8 kg 1.61 (OGD, 2017) 0.157 Beef 0.06 kg 0.72 kg 32 (Chambers, Simmons, and Wackernagel 2004) 0.011 Mutton 0.08 kg 0.96 kg 72 (Chambers, Simmons, and Wackernagel 2004) 0.006 Milk 5.4 litre 64.8 litre 458 (Chambers, Simmons, and Wackernagel 2004) 0.062 Fish 0.252 kg 3.024 kg 0.035 (Chambers, Simmons, and Wackernagel 2004) 0.030 Fruits 0.654 kg 7.848 kg 2330 (IHD, 2014) 1.58 × 10^−6 Edible oil 0.85 kg 10.2 kg 0.38 0.068 Wood 4.3 kg 51.6 kg 1.5–1.6 (kgCO2/kg) 73 m^3/ha (FSI, 2015) 0.028 LPG 1.9 kg 22.8 kg 3.31 (kgCO2/kg) 0.025 Kerosene 0.40 litre 4.8 litre 2.58 (kgCO2/litre) 0.004 Total annual EF/person 0.549