Balancing Crop Productivity and Energy Generation: A Hybrid Framework for Custom-Designed Agrivoltaic Systems with Dynamic Shadow Coverage Index

1. Introduction

The ever-increasing battle between food production and renewable energy for land allocation is a defining battle of 21st century. By 2050 the global population is to increase to over 9.7 billion people, thereby requiring a 60% increase in agricultural production along with the switch to low-carbon energy to alleviate climatic change [2,3]. This battle between the production of energy and food may be addressed through the use of Agrivoltaic systems (AVS), the co-location of photovoltaic panels (PV) and crop cultivation on the same agricultural land [4,5]. While enhancing efficiency of Land Use in dual usage of solar resources, AVS generally produce Land Equivalent Ratios (LER) over 1.5 [6,7].

For India, it remains particularly important due to its position as the world’s most populous country and third largest contributor to GHG emission, coupled with its target of 500 GW of renewable energy sources by 2030[8] and its need to secure the food security of 1.4 billion people [9]. However, most of the current studies into the performance of AVS have been carried out in temperate Europe and look at shade tolerant crops such as lettuce, spinach and berries [10,11]. The majority of the world, specifically the subtropical regions such as the Indo Gangetic plains with their extreme solar radiation, summer temperatures (>40 C) and monsoons remain significantly understudied, especially with relation to major staple crops.

The primary issue in agrivoltaics is distributing sunlight between crops and photovoltaic panels to ensure sufficient production of both crops and electricity. So far, the field of agrivoltaics has not thoroughly assessed design strategies, mainly because the best way to divide sunlight varies with location and the type of crops grown [12,13]. While the variety of agrivoltaic pilot projects mirrors the diversity in agriculture, transferring insights across different environments is difficult without a clear understanding of how system components interact and how local climate conditions influence these interactions [14]. Existing models fail to account for essential factors affecting plant function, thus missing important aspects of how agrivoltaic systems work together. The use of dimensionless metrics like LER, GCR, and WUE enhances the application of agrivoltaics in practical settings [6,15,16].

Direct contribution of integrated solar agriculture with UN SDG7 Affordable and Clean energy by providing distributed renewable electricity source [17,18]. Also, in providing transition to sustainable, safe and reliable and efficient energy sector since conflict on usage for the land use reduced by AVS and a source of dual income (food and energy) the Net Zero Transition is accessible for the food producers and users [19,20]. Due to the Net Zero objectives of numerous countries, we anticipate the expansion of land use innovative solutions, such as the one proposed with the AVS system will be a key element in achieve net zero emissions without impacting food security [21,22].

Despite the extensive number of articles about AVSs there is still much to learn, in particular most AVS systems prioritize power density (kW/ha) at the expense of crop physiological requirements. Present conventional designs generally focus on achieving high Ground Coverage Ratios (GCR >0.5) and low panel mounting heights (<3m) to minimize the land cost while maximizing energy production. While this is a reasonable assumption from an energetic perspective, there has been no recognition of the drastic long term shade stress being imposed on the underlying crop by these conventional PV systems leading to reduced yields and nutritional content and altered morpho-physiology [20]. This is exacerbated in systems targeting important shade sensitive staple crops such as potato (Solanum tuberosum L.), a crop requiring a large input of photosynthetically active radiation (PAR) in specific phases i.e., during tuber initiation and bulking stages (van der Werf et al., 2020). Furthermore, conventional methods of measuring performance lack utility for comparing AVS performance; the widespread use of the static, GCR cannot fully take in to account dynamic, space-time variability of shadows induced on crops by seasonal solar position and panel arrangement, and there is need for a time dependent, Shadow Coverage Index (SCI) that can take these aspects in to account. Filling these gaps will be key to leveraging the full benefits of agrivoltaics to enable the sustainable, secure, and efficient use of land that contributes to not just SDG 7, but the broader climate mitigation goals as well.

This paper describes the construction of a bespoke, agricultural-based agrivoltaic system designed to optimize crop production at the expense of power density. Situated on the research farm of the Amity University, Noida (India) the proposed system uses an elevated mounting height to optimize ventilation and reduce heat stress along with a wide inter-row spacing allowing more lateral light, pole-based structures to minimize soil coverage and thus intended to optimize Land Equivalent Ratio (LER) by allowing crop shading indices less than that required. It utilizes the hybrid approach of field experimentation coupled with PVsyst-based energy modeling simulations for better optimization and looks at the system from three different scenario; (i) spatial/temporal and land usage/shading effects, (ii) energy yields and (iii) crop yield/quality/physiology, to determine the perfect balance between crop yield and energy production.

The specific objectives of this study targeted at providing a methodical assessment of the custom-made, agriculture-focused agrivoltaic system using a complete hybrid framework: First, design and characterize a custom agriculture-oriented agrivoltaic system that is specific for subtropical conditions, specifying all of the design parameters of structural characteristics (i.e., height of mounts, space between rows, configuration of pole arrays), geometrical variables (i.e., tilt angle of modules, orientation direction, array configurations), and electrical information (i.e., panel capability, configuration and capacity of system inverter and system configuration). Second, develop a new, dynamic Shadow Coverage Index (SCI) to quantify the shading situation where Solar position simulations derived from the software PVsyst coupled with phenological events of potatoes throughout the whole growing season are integrated to represent spatiotemporal variation of shading. Third, examine and investigate crop characteristics during a parallel outdoor experimental study on potatoes in the specific design and in monocropped open field, specifically potato growth components (height, area of the leaves and count of the leaves), tuber characteristics (unit tuber count, unit tuber weight, unit dry mass of tuber), and total final yield in the above mentioned two planting configurations. Fourth, investigate and evaluate system performance, via determining Land Equivalent Ratio (LER) which highlights the interrelationships between the power generation and cultivation aspects so as to determine the land-equivalent efficiency of the current specific design configuration. Lastly, develop and propose a reproducible framework that can provide guidance for determining appropriate design configurations for a balance between crop yield and energy supply based on evidence derived from observations from a subtropical region agrivoltaic plant system, such that comparable planting arrangements are applied to new facilities.

2. Materials and Methods

2.1. Study Area and Climatic Conditions

The field experiment was conducted in the rabi (winter) season, at research farm of Amity Institute of Organic Agriculture, Amity University, Noida, Uttar Pradesh, India (28o 32′29.44” N; 77o 19′ 56.73” E, 200m amsl). This site is characterized by a subtropical semi-arid climatic zone, dominated by hot summers (April-June, mean temp 38-42C), cool winters (December-February, mean temp 5-15oC) and a prominent monsoon period (July September). The crop was grown during winter season to avoid extreme temperatures due to heat stress and as part of normal farming activities in the study region. Global Horizontal Irradiance (GHI) at the location is 1,618.8 kWh/m/year, with maximum 764 hr sun light per year. Table 1 and Table 2 provide a detailed description of the study area’s location and climate and weather conditions, respectively.

2.2. Custom-Designed Agrivoltaic System

2.2.1. Structural Design

A pole-mounted, fixed-tilt agrivoltaic structure of land area 629 m² was constructed on the site. The key design parameters chosen to prioritize crop growth are as follows (Table 3):

An east-west orientation of the panels with north-south rows arrangement was adopted, which help in balanced shading distribution to the underlying crops. Structure configuration was based on a 5x3 matrix, with a total of 32 panels mounted on 16 poles (two panels per pole arranged in vertical orientation), as shown in Figure 1. A hot-dip galvanized steel structure (UL 2703 certified) was utilized to ensure corrosion resistance of the mounting system. The detailed specification of SPV system including inverter and structure has been mentioned in Appendix A as A1.

2.2.2. PV-Array Design

Monocrystalline modules with a capacity of 325 Wp are mounted on the structure at an optimal tilt and azimuth of 24° and 180°, respectively, at a height of 4m for easier operation with underlying agricultural activity. The row-to-row spacing was selected to be 8 m. The DC power generated by the array was directed to a DC combiner box, converted into AC, and fed to a 10-kW grid-connected inverter. The output power from the grid-connected inverter was connected to the AC distribution board and synchronized with the utility grid. Data of system’s performance was logged using an integrated data acquisition system [23].

Table 4. SPV System Configuration.

Component	Specification
PV Module Rating	Mono-crystalline, 325 W
Number of Modules	32
PV Array Capacity	≈10 kW
Inverter Sizing	10 kW grid-connected inverter
Inverter Efficiency	97–98%
Mounting Structure	Fixed tilt
Monitoring	Data logger
GM Structure	HDG (Hot-dip Galvanized)

Table 4. SPV System Configuration.

Component	Specification
PV Module Rating	Mono-crystalline, 325 W
Number of Modules	32
PV Array Capacity	≈10 kW
Inverter Sizing	10 kW grid-connected inverter
Inverter Efficiency	97–98%
Mounting Structure	Fixed tilt
Monitoring	Data logger
GM Structure	HDG (Hot-dip Galvanized)

Table 5. SPV System Sizing for 10kW PV System.

Parameter	value
Module rated power (Wp)	325
Number of modules	32
Total capacity (kW)	10
Module per string 1	16
Number of strings	2
Array Configuration	16 x 2

Table 5. SPV System Sizing for 10kW PV System.

Parameter	value
Module rated power (Wp)	325
Number of modules	32
Total capacity (kW)	10
Module per string 1	16
Number of strings	2
Array Configuration	16 x 2

2.3. Simulation Framework

The custom-designed agrivoltaic system model evaluates metrics such as the Ground Coverage Ratio (GCR), Shadow Coverage Index (SCI), Land Equivalent Ratio (LER), crop yield, and energy output within a hybrid agrivoltaic framework. This integrated approach facilitates field experiments and simulations, employing a multi-objective decision-making strategy to optimize and identify trade-offs between energy production and agricultural outputs, also described in Figure 2. Energy yield simulations were conducted using PVsyst, employing typical meteorological data sourced from the synthetic Typical Meteorological Year (TMY) Meteonorm 8.1 database. Shading analysis was performed using hourly time-step simulations with the SketchUp software tool. However, in Figure 3 three-dimensional design of custom-designed AVS model shown by using SketchUp 3D software tool, Figure 4., module spatial arrangement layout simulated in PVsyst Ver.7.4, respectively. Further system energy flow schematic diagram of 10 kW Solar PV System is illustrated in Figure 5.

2.3.1. Ground Coverage Ratio (GCR)

GCR is an important designing parameter which determines proportion of PV array area to be covered over the total land area available. High GCR will yield higher energy density, but with lower light available for crops.

$$\mathit{G}\mathit{C}\mathit{R}={\displaystyle \frac{{\mathit{A}}_{\mathit{P}\mathit{V}}}{{\mathit{A}}_{\mathit{L}\mathit{a}\mathit{n}\mathit{d}}}}$$

(1)

where as,

A_PV = Total projected area of PV modules on the ground (m²)

A_Land = Total land area allocated for the PV system (m²)

Calculation:

Module dimensions: 1.96 m × 0.99 m = 1.9404 m² per module

Tilt angle = 24°

Pitch = 8 m

Total land area = 629 m²

Total useable land area = 600 m²

Projected width per module:

W = 1.94 × cos (24°) = 1.77” m”

Total module area:

A_PV = 1.77 × 32 = 56.64 m²

GCR = 56.64/600 = 0.0944 ≈ 9.44%

2.3.2. Shadow Coverage Index (SCI)—Dynamic Modelling

Shadow coverage Index (SCI), a non-dimensional parameter in range 0-1, determines the ratio of a surface area to be covered by shadow (at a given time or over a period). SCI addresses the limitations of GCR where time variability in solar trajectory and array arrangement is taken into consideration for determining shadow profile across the crop.

SCI Classification:

SCI < 0.3 → Low shading, crops receive sufficient sunlight

SCI 0.3–0.6 → Moderate shading, suitable for shade-tolerant crops

SCI > 0.6 → High shading, may significantly reduce crop productivity

The shadow area depends on solar geometry, panel dimensions, and tilt angle:

$$A_{shadow}=L_{shadow}\times W_{panel}$$

(2)

where,

L_shadow = Shadow length (m)

W_panel = Width of the PV panel (m)

The shadow length is expressed using solar elevation angle:

$$L_{shadow}={\displaystyle \frac{H_{panel}}{\mathrm{t}\mathrm{a}\mathrm{n}\left({\theta }_s\right)}}$$

(3)

where,

H_panel = Effective height of the PV panel above ground (m)

Θ_s = Solar elevation angle (degrees)

Thus,

$$A_{shadow}={\displaystyle \frac{H_{panel}}{\mathrm{t}\mathrm{a}\mathrm{n}\left({\theta }_s\right)}}\times W_{panel}$$

(4)

For multiple PV rows:

$$\mathit{S}\mathit{C}\mathit{I}={\displaystyle \frac{\mathit{N}\times {\mathit{A}}_{\mathit{s}\mathit{h}\mathit{a}\mathit{d}\mathit{o}\mathit{w}}}{{\mathit{A}}_{\mathit{t}\mathit{o}\mathit{t}\mathit{a}\mathit{l}}}}$$

(5)

where, N = number of PV rows.

SCI calculation using 3-D shading model of PVsyst (Version 7.4) was done considering solar geometry (azimuth, altitude of the Sun), panel dimension, tilt, mounting height, row to row gap. Time-step simulations for 10 minutes interval were run for the cropping season (December to March) and overlapped with potato phenological stages (emergence, vegetative stage, tuber initiation, tuber bulking, and maturity) in order to identified sensitive periods of shading in potato cultivation.

2.3.3. Energy Yield Simulation (PVsyst)

Energy generation for PV array was simulated using PVsyst V7.4 (PVsyst SA, Switzerland). Input data parameters for simulation were:

Hourly meteo-data for the study location from a synthetic Typical Meteorological Year (TMY) Meteonorm 8.1 database.

Module’s manufacturer specifications (temperature coefficient, efficiency) and other system properties.

System losses considered were 3% as soiling loss, thermal losses according to the NOCT of the module and 2% as inverter loss.

Shading losses computed from the SCI model.

Core Equations:

PV Output Power:

$$P=A\cdot G\cdot \eta$$

(6)

where,

A = Panel area (m²)

G = Irradiance (W/m²)

η = Efficiency (%)

Performance Ratio:

$$\mathit{P}\mathit{R}={\displaystyle \frac{{\mathit{Y}}_{\mathit{f}}}{{\mathit{Y}}_{\mathit{r}}}}$$

(7)

where,

Y_f = Final yield (actual energy output)

Y_r = Reference yield (theoretical maximum)

The simulation production included monthly as well as seasonal AC energy (kWh), specific yield (kWh/kWp), and performance ratio (PR).

2.3.4. Land Equivalent Ratio (LER)

LER is used to determine relative efficiency of agrivoltaic system when agricultural and solar energy yield of the combination system are measured in relative to independent solar and agricultural systems. LER >1 for the combination system indicate AVS uses land more efficiently than the individual monoculture and PV farms. The working calculation has shown in Appendix A, A4: LER Calculation—Extended

$$LER=\left[{\displaystyle \frac{Y_{AVS}}{Y_{Control}}}+{\displaystyle \frac{E_{AVS}}{E_{PV-only}}}\right]\times \left(1-LL\right)$$

(8)

where,

Y_AVS = Crop yield under agrivoltaic system

Y_Control = Crop yield under conventional monocropping

E_AVS = Energy yield from agrivoltaic system

E_(PV-only) = Energy yield from PV-only system (no crop)

LL = Land loss factor (LL = 0 for overhead systems)

Interpretation:

LER = 1: Agrivoltaic system performs equivalently to separate systems

LER > 1: Agrivoltaic system is more land-efficient

LER < 1: Separate systems are more land-efficient

2.4. Crop Cultivation and Experimental Design

2.4.1. Experimental Details

The experiment and image sample collected on strip plot (shown in Figure 6) design with two treatments replicated three times; plot dimensions per treatment were 6m. Meanwhile, weather condition data has been shown in Appendix A, A2: Detailed Weather Data (November 2024–March 2025).

The treatments considered were as under:

• T1: PV array mounted crop planting (AVS treatment), and
• T2: PV array-less, conventional monocrop planting (Control).

2.4.2. Soil Characteristics

The soil at the experimental location is a deep loamy soil with well-developed alluvium, moderate in clay and low in organic carbon content with a pH of approximately 8.4. For Control T2 the soil parameters are pH (1:5 suspension): 8.48 (Alkaline), Total Nitrogen (TKN): 0.037% by Mass (Low), Available P: 17.3 mg/Kg (Moderate), Available K: 178 mg/Kg (Moderate). Whereas the same for T1 are observed as pH (1:5 suspension): 8.39 (Slightly Alkaline), Total Nitrogen (TKN): 0.042% by Mass (Low), Available P: 18.5 mg/Kg (Moderate), Available K: 187 mg/Kg (Moderate).

2.4.3. Agronomic Practices

The agronomic practice during the experiment is as given below, in Table 6:

2.5. Data Collection and Statistical Analysis

The data collected during the process has shared as Appendix A, A3: Raw Crop Data.

2.5.1. Crop Parameters Recorded

The following parameters were recorded for both treatments:

• Leaf area (cm²)—measured using leaf area meter
• Plant height (cm)—measured from ground to apical meristem
• Number of leaves per plant
• Number of tubers per stand
• Unit weight of potato (g/tuber)
• Dry matter weight (g)—oven-dried at 70 °C to constant weight
• Final yield per plot (kg)

2.5.2. Statistical Analysis

As the installation is under study under a pilot scale, crop parameters were evaluated under descriptive, rather than confirmatory statistical framework. The values shown are the mean ± standard deviation (SD), and % changes are provided as indicative measure in pilot research study. This framework is ideal to showcase a design concept and simulation validated technique, rather than performing crop physiological characterization.

3. Results

3.1. Energy Yield Assessment

The PVsyst simulation results for the 10.4 kWp custom-designed AVS are presented below, in Table 7.

The key performance indicators are:

• • Annual Energy Yield: 1,976 kWh/year (normalized: 1,055 kWh/kWp/year)
• • Specific Yield: 2.98 kWh/kWp/day
• • Performance Ratio (PR): 79.1%
• • System Capacity Utilization Factor: ~12.1%

With a performance ratio of 79.1%, a fairly high efficiency is found with allowances for the various losses like soiling (3%), thermal losses and inverter losses (2%) and shading losses. While the energy produced under low GCR (0.09) was comparatively less when compared to conventional PV farm, agrivoltaic cultivation resulted in a better crop yield under this particular conditions.

3.2. Shading Dynamics—Shadow Coverage Index (SCI)

3.2.1. GCR–SCI Relationship

It has been observed that GCR is non-linearly related with SCI. At custom-built GCR of 0.09, SCI was less than 0.2 throughout cropping seasons-clearly within the ‘low shading’. Thus, minimal shading overlap occurred with good light penetration which is considered ‘crop-friendly’ condition with agrivoltaic system.

3.2.2. Seasonal SCI Variation

The SCI exhibited distinct diurnal and seasonal patterns from November to March, indicated in Figure 7:

• Morning ramp (8–10 AM): SCI rises sharply with rapid shading onset after sunrise
• Midday plateau (10 AM–2 PM): SCI stabilizes at approximately 0.09–0.12, indicating relatively uniform shading
• Afternoon divergence (2–5 PM):

  ○

  January–February: Peak SCI (~0.14–0.15) — strongest late-day shading

  ○

  November: Secondary peak (~0.13), then declines

  ○

  December: Continuous decline, weakest shading overall

  ○

  March: Gradual rise, moderate shading

The key finding herein is that during the potato tuber initiation (30-50 DAS) and tuber bulking phases (50-90 DAS)-which are the most critical stages of potato production-the SCI remained below 0.2 confirming the custom design has successfully avoided deep-shading effect at these yield determining stages.

3.3. Crop Performance

3.3.1. Vegetative Growth Parameters

Growth and yield parameters were measured for each scenario, T1 & T2, are represented in following Table 8 and Figure 8.

Leaf Area: Over the observed period leaf area increased in both treatments. At the early stage (30 DAS), there was a marginal increase in leaf area in AVS over the control due to favorable micro-climate with diffused light and low stress while at mid stage (30-60 DAS), steady increase with close trend was observed between the two treatments. At late stage (60 DAS), control had greater leaf area when compared to AVS treatment.

Plant Height: Plant height continuously increase in both treatments with 94 DAS for AVS crops and until 108 DAS for control crops. Similar trend, AVS shows smaller stature (10.8% height decrease at 90 DAS) as compared to control at all stages of crop growth due to shading effect which is directly associated with low irradiation.

No. of Leaves: There were always significantly more leaves at the control treatment than AVS. However, significant increases were first observed after ~ 59DAS and the effect grew significantly at a late stage (62-65 DAS) of vegetative growth where AVS results in lower number of leaves at the end of the vegetative stage.

3.3.2. Yield Attributes

The yield in each case after crop harvesting is summarized in following Table 9 and Figure 9:

Number of tubers/stands: Throughout potato growth-higher number of tubers was observed at the control treatment compared to AVS treatment. At the early stages the difference between control and AVS was higher (4-5 tubers in 59-66 DAS) but as it grown further it significantly decrease and reached to approximately 1 tuber in 72-100 DAS.

Unit Weight of the tubers: Increase in unit weight of tubers was observed in both the treatments but AVS shows higher unit weight over control throughout all the stages of the crop growth, and significantly more gain at approximately between 60-75 DAS where it levelled off while control shows slow steady increase. Finally, unit weight was significantly increased at AVS treatment compared to control (30.0 gm vs 47.5 gm), the reason can be attributed towards favourable micro-climate which improved the overall moisture and heat condition resulting in better tuberization under AVS compared to control.

Dry matter: AVS treatment shows significantly more dry matter then control throughout all the growing periods, a very sharp increase in dry matter from ~0.2 g to 0.88 g was observed at approximately 70 DAS while slow increase from ~0.15 g to 0.20 g occur in control which clearly means that biomass production capacity increased under agrivoltaic system due to improved microclimate

Final Yield: Potatoes were harvested with the observed AVS yielded of 18.3 kg/plot and that of Control is 15.0 kg/plot, thus 22.0% of higher yield was recorded for custom-designed AVS.

3.4. Land Equivalent Ratio (LER)

The Land Equivalent Ratio calculated for both T1 & T2 conditions are summarized in Table 10.

LER in the case of for T1 has been observed to be 1.65, which clearly depicts that the land is being utilized efficiently up to 65% higher compared to individual monoculture and PV land usage and this is much higher as compared to the expected value of LER>1.0 for any agrivoltaic system. The following Table 11 illustrates the trade-off dynamics of the study.

4. Discussion

4.1. Physiological Response of Potato to Dynamic Shading

Potato plants under the custom AVS demonstrated a classical shade-avoidance response: marginally longer stems (+16.8% at early growth stage) but decreased leaf area and number as compared to the control. Such morphological adjustments correspond to the potato response to reduce PAR, which increases biomass allocation for stem elongation and tuber filling in reverse way, as identified [24].

Surprisingly, the 58.3% increase in unit tuber weight obtained under AVS indicates that the moderate shading (SCI < 0.2) during tuber bulking perhaps provided additional favorable factors such as (i) attenuated heat stress which resulted in more efficient partitioning of assimilates toward tuber production and (ii) enhanced soil moisture retention by reducing evaporation rate, and (iii) favorable microclimatic condition for tuber growth. The massive 340% increase in dry matter content further consolidates the above interpretation and shows increased efficiencies in biomass accumulation under the agrivoltaic microclimate.

The total yield improvement by 22.0% (18.3 ± 2.1 kg versus 15.0 ± 1.8 kg per plot) is broadly consistent with recent meta-analyses that the potato yields ranged from -20% to +11% with different agrivoltaic system designs and climates. The custom-design with intentionally low GCR and elevated mounting height obtained at the upper limit of this range indicating the possibility for agriculture-centric designs to improve productivity in the tropics.

4.2. Trade-Off Between Energy Generation and Crop Yield

The trade-off results demonstrate a non-linear relation between the GCR, SCI and the system outputs, and the LER of 1.65 achieved at GCR=0.09 can be considered significantly positive as the combined food-energy output exceeded that of the individual systems by 65%. This value falls within the upper range of LER values (1.2-1.8) that have been reported in a global review of agrivoltaics.

Although the system energy ratio (EAVS/EPV-only) was relatively low at 0.43, this occurred due to the deliberate sacrifice in power density. However, the system design is economically reasonable because:

• The 22.0% additional crop yield has sufficiently offset the 57.0% decrease in energy generation in the system total value,
• The custom design enables year-round agricultural activities beneath the panels, improving overall land-use efficiency, and
• During subtropic areas hot weather conditions are a limiting factor for the tuber productivity, moderate shading helps to retain quality of and market value of the product and increases total value of the crops due to increased unit yield.

This study hence challenged the conventional idea which emphasizes power maximization as the optimal configuration of AVS and demonstrates the necessity of considering crop limitations to optimize system design.

4.3. Justification of Custom-Design over Conventional AVS

A typical conventional AVS with high GCR (e.g., >0.5, mounting height <3m) would have: (i) remained an SCI > 0.4 during tuber bulking, which would have resulted in a loss of ~40% to 60% of potato yield, (ii) obtained far below a 1.5 LER. To avoid these problems, the following design modifications for the custom AVS were imposed: (1) elevated mounting height (4m) to maintain high air circulation rates and alleviate heat stress, and allow the passage of field machines; (2) widely spaced rows (8 m) so as to allow side-light penetrates, hence maintaining an SCI < 0.2 during potato tuber formation; (3) low GCR (0.09) to minimize deep shading over agricultural floor and yet provide adequate energy generation; and (4) pole-foundation design that minimize the soil footprint and preserve the soil structure. All the improvements are consistent with the current system design principles for temperate and tropical areas in which heat stress is the main limiting factor affecting the yield.

4.4. Implications for Subtropical Agrivoltaics and Future Directions

The study proves the feasibility of the agriculture-first approach to boost the land-use efficiency in the subtropics, and the hybrid experimental-simulation framework for the customized AVS design allows for application and evaluation of AVS design with specific crop requirements. There are:

• For Researchers: Researcher will benefit from the introduced SCI metric to investigate the impact of AVS design on the crop physiology.
• For Policymakers: Policymakers are presented with solid evidence that AVS enables highly efficient land-use that meets the PM-KUSUM objectives of deploying renewable energy for agriculture in India.
• For Farmers: Farmers can be attracted to invest in dual-input agriculture due to double income stream of food and energy without further compromises in agricultural yield.

For future work the researchers suggest to: Replicate study over a multi-year and multi-season crop production in order to confirm the identified results; Evaluate system performances with multiple staple crops (e.g., Wheat, maize and pulses) grown under similar designs; Incorporate economic analysis to compute net present value (NPV) and payback period (PBP) of integrated food energy systems; Monitor, under actual conditions, the system dynamics such as local microclimatic conditions and soil health changes along with crop growth and energy output; and Construct comprehensive techno-economic model with consideration of location-specific climate risks.

5. Conclusions

The present study successfully proved the potential of the custom-designed agriculture-focus agrivoltaics system for greater Land Use Efficiency (LER = 1.65) in addition to boosted potato yields (22%) in subtropical Indian conditions. Compromising on optimum power density through reduced GCR (0.09), larger height (4m), and extended plant spacing (8 m) helped sustain the Shadow Coverage Index (SCI) under 0.2 during the tuber bulging phase and instead boosted, rather than hindered, crop yields.

This newly developed hybrid experiment–simulation frame provides an easily replicable, site specific-centric protocol for the optimal AVS configuration. Furthermore, inclusion of the SCI parameter dynamically within the PVSyst has provided an objective framework to connect the PV system parameters to crop physiology’s thresholds, which has long been missing within standard AVS designs.

The results of this research not only contradict the standard approach in which the value of GCR signifies good performance and land-use but also proves that optimized system design is not dependent solely on land–use or power but rather based on crops requirements which can ultimately result in an integrated AVS design far more efficient to crop production and resource-utilization in subtropical environment.

Considering the ongoing conditions of heat stress, food insecurity and shortages of land in subtropical regions, an integrated AVS in an agriculture-centric frame might serve as a good way for an eco-friendly and sustainable food-energy nexus, directly contributing to SDG 7 (Affordable and Clean Energy) and supporting an equitable transition to a Net Zero future.