Techno-Economic and Environmental Assessment of Rooftop Photovoltaic Systems in Baghdad Using SketchUp-PVsyst Simulation

Iraq has substantial solar resources and a strong rationale for distributed photovoltaic (PV) deployment [1], [2]. Baghdad also experiences high ambient temperatures that reduce PV module output [3], while dust deposition can impose additional optical losses [4]. In dense urban rooftops, parapets, service rooms, neighboring structures, and row-to-row geometry can create near-shading that is difficult to represent with simple two-dimensional sizing methods [5], [6]. These factors make realistic rooftop geometry, thermal behavior, soiling, and shading important design considerations rather than secondary corrections.

Previous work has addressed complementary parts of this problem. Al-Ghezi et al. [2], [3] quantified Baghdad solar-resource and temperature effects, while Musa and Shneishil [6] used 3D shading and PVsyst for rooftop layout optimization in Baghdad. Ebhota and Tabakov [7], Moustafa et al. [8], and Kumar et al. [9] demonstrated the value of simulation tools for orientation, operating-temperature, and system-design studies. However, most studies examine a limited set of design variables or emphasize technical output without a consistent economic and environmental comparison at roof scale.

Module technology creates an additional design dimension. Zhang et al. [10] and Ali et al. [11] summarize the device-level and thermal-performance potential of advanced heterojunction (HJT) architectures, while Camilo and Santos [12] demonstrate the importance of reporting technical and financial indicators together. A gap therefore remains for Baghdad rooftops: few studies jointly assess realistic 3D near-shading, module technology, mounting strategy, layout density, and their combined techno-economic and environmental consequences.

Accordingly, the present study addresses the following design question: which rooftop PV configuration gives the most practical balance between energy yield, financial performance, and realistic constructability for an urban building in Baghdad? To answer this question, a 3D rooftop model of the College of Education building at Mustansiriyah University was developed in SketchUp/Skelion and imported into PVsyst for near-shading analysis. Seventeen variants were then compared by combining monocrystalline and HJT modules, fixed/seasonal/two-axis mounting strategies, portrait/landscape layouts, row-spacing alternatives, and several loss combinations.

The main contribution of the paper is the integration of three normally separate perspectives in one framework: (i) realistic 3D rooftop and near-shading modelling for a Baghdad case study; (ii) side-by-side technical comparison of module technology, mounting strategy, orientation, and spacing; and (iii) translation of these differences into economic and environmental indicators. This framework supports a practical design recommendation for future rooftop PV deployment in Baghdad and comparable hot urban environments.

The study site is the College of Education building at Mustansiriyah University in Baghdad, Iraq (latitude 33.3$^\circ$ N, longitude 44.4$^\circ$ E). The available rooftop consists of two similar blocks with a combined gross roof area of about 968 m$^2$. The PVsyst simulations used Meteonorm 8.0 meteorological data for Baghdad. The annual global horizontal irradiation (GHI) used in the model is about 1,986.4 kWh/m$^2$/year, equivalent to an annual daily average of approximately 5.44 kWh/m$^2$/day. This high solar resource is consistent with local Baghdad studies [2]. At the same time, hot summer conditions and dust exposure make the site suitable for examining the trade-off between high irradiance, thermal losses, soiling, and practical rooftop limitations.

The workflow began with the construction of a detailed roof model in SketchUp. The digital model included the principal rooftop obstructions and nearby geometries needed to reproduce the installation conditions. The Skelion plug-in was used to place PV rows and configure their spacing and tilt, after which the scene was exported to PVsyst. The same geometry was then imported into the PVsyst Near Shadings environment so that geometric and beam-shading effects could be included consistently in the annual simulations.

From the PVsyst database, two 680 W module families were selected: an n-type monocrystalline module (RCM-680-8DBNM) and an HJT module (RCM-680-8DBHM), whose simplified schematic is shown in Figure 1. Both modules have the same rated peak power, making the comparison more meaningful because differences in annual output can be attributed primarily to thermal behavior and system configuration rather than nominal module power. The selected inverter family was an SMA 30 kW string inverter, scaled appropriately for the varying array capacities. The main characteristics of the selected modules are compared in Table 1. The complete simulation workflow is summarized in Figure 2, while Figure 3, Figure 4, and Figure 5 illustrate the rooftop geometry and corresponding simulation stages.

The configurations were arranged so that the impact of module technology, mounting strategy, orientation, row spacing, and cumulative losses could be isolated. For the layout-only variants, the landscape cases accommodate 192 modules at 0 m spacing, 154 modules at 0.5 m spacing, and 116 modules at 1.0 m spacing, whereas the portrait cases accommodate 162, 147, and 120 modules, respectively. This difference in installed direct current (DC) capacity is important because wider row spacing reduces self-shading but also decreases the number of modules that can fit on the rooftop.

The row-spacing check was based on simple shadow geometry for a tilted PV row [13]. Let $L$ be the sloped module dimension in the north-south row section, $\beta$ the module tilt, and $\alpha_\mathrm{{min}}$ the critical solar elevation angle. The minimum clear horizontal distance required for the shadow from the upper edge of one row to reach the lower edge of the next row, together with the corresponding minimum horizontal pitch, are given by:

Here, $D_\mathrm{{min}}$ is the shadow-avoidance clear distance between adjacent rows and $P_{\mathrm{row,min}}$ is the horizontal pitch between corresponding lower edges of successive rows. For the fixed landscape-layout geometric check, $L$ = 1.303 m, $\beta$ = 30$^\circ$, and $\alpha_\mathrm{{min}}$ $\approx$ 33.3$^\circ$ (approximately the solar-noon elevation at the winter solstice for latitude 33.3$^\circ$ N). These values give $D_\mathrm{{min}}$ $\approx$ 0.99 m and $P_{\mathrm{row,min}}$ $\approx$ 2.12 m. The simulated clear-spacing scenarios of 0, 0.5, and 1.0 m were implemented directly as design alternatives in the rooftop layout; they were not derived from the incorrect expression in the previous version. Thus, the 0 and 0.5 m cases intentionally accept greater self-shading in exchange for higher installed capacity, whereas the 1.0 m case is close to the geometric shadow-avoidance distance at the selected critical condition. The definitions of all 17 simulated PV system variants are summarized in Table 2. The seasonal HJT case is denoted as HLSN consistently throughout the text and figures.

To move the simulations away from idealized output and closer to practical operation, several additional losses were introduced stepwise. The base soiling loss was set to 3%. Module aging was represented using the PVsyst degradation setting corresponding to year 10 of operation. System unavailability was set to 2%, equivalent to approximately 7.3 days/year, and auxiliary electricity consumption was included in the fully loaded cases. These terms were accumulated progressively in the loss-build-up variants so that their combined impact could be observed. Standard PVsyst thermal, electrical, mismatch, wiring, and inverter losses remained active according to the selected components and system model. The geometric definition of inter-row shading variables used in this analysis is illustrated in Figure 6.

Figure 7 and Figure 8 show two of the mounting options implemented in PVsyst: seasonal tilt and two-axis tracking.

The first four variants keep the installed DC capacity constant at 192 modules (130.56 kWp) so that module technology and mounting strategy can be compared without the confounding effect of array size. They use the same nominal dense-layout basis, and the additional near-shading/soiling/aging/unavailability/auxiliary scenario terms used in the later loss-build-up cases are not applied to this controlled comparison.

The technical assessment follows standard PV performance definitions commonly used in PV performance studies [14]. Four indicators were used: annual alternating current (AC) energy production, reference yield, final yield, and performance ratio.

where, $E_{\mathrm{AC}}$ is the annual grid-connected AC electrical energy (kWh), $P_{\mathrm{AC},t}$ is the AC output power (kW) at annual simulation time step $t$, $\Delta$$t$ is the time-step duration in hours, and N is the total number of time steps in one year. $H_{\mathrm{POA}}$ is the annual plane-of-array irradiation (kWh/m$^2$), $G_{\mathrm{STC}}$ is the standard-test-condition irradiance (1 kW/m$^2$), and $P_0$ is the installed DC capacity (kWp). $Y_r$ represents the available solar resource, $Y_f$ is the useful annual AC energy per unit of installed capacity, and PR indicates how effectively the system converts the available plane-of-array irradiation into useful AC energy after modeled losses.

The financial evaluation used Net Present Value (NPV), Internal Rate of Return (IRR), Simple Payback Period (PBP), a discounted NPV-based Return on Investment (ROI), and Levelized Cost of Energy (LCOE) [12]. Each configuration used its reported initial investment and the common base-case assumptions in Table 3. Public vendor quotations for the selected module and inverter families were used only as price benchmarks; the comparative calculations use the total investment values reported for each variant.

The initial investment $C_0$ includes PV modules, inverter units, mounting structure, DC and AC wiring, protection and connection components, installation and commissioning, and other balance-of-system items connecting the rooftop system to the grid. For the two-axis tracking case, the tracking structure, drive/control system, and related installation cost were also included. Land purchase, taxes, financing charges, battery storage, roof strengthening, and major component replacement were outside the cost boundary. The same boundary was applied consistently across configurations.

Here, $C_0$ is the initial investment, $a$ is the discount rate (0.02 in the base case), $n$ is the 25-year analysis period, $C_{\mathrm{O\&M},j}$ is the operation and maintenance (O&M) cost in year $j$, $E_j$ is the annual energy used in the cash-flow model, and $C_j$ is the annual net cash flow. Under the stated base case, $C_j = E_j V_e - C_{\mathrm{O\&M},j}$, where $V_e = 0.063\ \mathrm{USD/kWh}$ and $C_{\mathrm{O\&M},j} = 0.01\,C_0$ per year. The simple payback period is undiscounted and uses the constant annual net cash flow. The ROI reported here is explicitly $NPV/C_0 \times 100\%$, so it is a discounted NPV-based ratio rather than an undiscounted lifetime ROI. IRR is the discount rate $r$ that makes Eq. (6) equal to zero. No electricity-price escalation or additional year-by-year degradation was applied in the financial model; where the PVsyst aging setting is included, its effect is already embedded in the corresponding annual-energy result.

Eq. (9) was used to calculate the avoided CO$_2$ values reported in Table 5. $EF_{grid}$ is the adopted base-case grid emission factor of 0.594 kgCO$_2$/kWh, and the factor 1/1000 converts kilograms to tonnes. The lifetime calculation uses the annual energy value for each variant over the 25-year study period. No additional year-by-year degradation correction is added to the CO$_2$ summation; cases that include the PVsyst aging-loss setting already contain that penalty in their annual-energy result.

Before comparing all variants, two complementary simulation outputs were used to interpret the results. Figure 9 shows the sun-path and shading analysis for the fully loaded HJT landscape case with 1.0 m spacing (HL1SSAUA), illustrating annual solar access for the modeled roof. Figure 10 shows the PVsyst loss diagram for the dense 0 m all-loss HJT landscape case (HLSSAUA); its grid injection of 181,982 kWh matches Table 4. The sun-path plot indicates that many surrounding obstructions act mainly at low solar angles, while the annual loss diagram quantifies the combined near-shading effect for the modeled geometry.

Figure 10 shows that near-shading irradiance loss is approximately 18.1%, larger than the approximately 7.3% temperature loss in this dense all-loss case. The near-shading term is an aggregate geometric loss from row-to-row self-shading together with the modeled rooftop and surrounding obstructions; the exported loss diagram does not separate these contributions numerically. Additional penalties include the 3% soiling assumption, electrical and inverter losses, aging, auxiliaries, and unavailability. Consequently, the results do not support treating surrounding-building shading alone as the dominant loss. Instead, dense row geometry and high operating temperature are both important in Baghdad, and the relative importance changes with layout spacing and loss scenario.

Table 4 summarizes the technical results of the 17 simulated configurations. Installed DC capacity is calculated directly from the module count using 0.680 kWp per module, and the specific production values are reported relative to those exact capacities. Figure 11, Figure 12, Figure 13, and Figure 14 visualize the monthly behavior of the principal groups of variants.

The two-axis tracking HJT case (HLT) produced the highest annual energy output, 272,308 kWh/yr, corresponding to 2,086 kWh/kWp/yr. Tracking increases plane-of-array irradiation throughout the day and year. Its performance ratio (73.63%) is nevertheless lower than those of the fixed and seasonal HJT cases because the reference yield also increases while thermal and system losses remain. The HLSN gives the highest PR (85.22%), while the fixed landscape HJT case (HLF) is very close at 85.19%. This is practically important because a simple fixed structure can provide nearly the same conversion-quality indicator without requiring seasonal mechanical adjustment.

The HJT advantage over the selected monocrystalline module is also visible in the controlled comparison. At the same installed capacity of 130.56 kWp, annual energy increases from 239,314 kWh/yr for MLF to 244,305 kWh/yr for HLF, while PR rises from 83.44% to 85.19%. This difference is consistent with the lower Pmax temperature coefficient of the HJT module in Table 1 (-0.24%/$^\circ$C versus -0.30%/$^\circ$C). Under Baghdad's high summer temperatures, the lower temperature sensitivity helps the HJT module retain higher specific production and PR.

The influence of row distance must be considered together with installed capacity. For the landscape shading-only cases, specific yields rise from 1,544 kWh/kWp/yr at 0 m spacing (HLS) to 1,699 kWh/kWp/yr at 0.5 m spacing (HL05S) and 1,754 kWh/kWp/yr at 1.0 m spacing (HL1S), showing that additional spacing mitigates mutual shading. At the same time, annual energy falls from 201,576 to 177,922 and then to 138,335 kWh/yr because fewer modules fit on the roof. Thus, wider spacing improves output per installed kWp but reduces total electricity from the available roof area. The 0.5 m case may offer a compromise when maintenance access and shading reduction are priorities, whereas denser packing can be preferable when maximizing total annual rooftop energy is the primary objective.

Figure 13 shows the sequence of losses and why practical non-idealities should be included in the simulation. Starting from HLS (201,576 kWh/yr), adding soiling (HLSS), aging (HLSSA), unavailability (HLSSAU), and auxiliary consumption (HLSSAUA) reduces annual energy to 181,982 kWh/yr, a cumulative decrease of approximately 9.72%. This reduction is the cumulative effect of several losses rather than a single penalty. The result reinforces the importance of cleaning, reliable operation, and appropriate thermal/electrical design for rooftop PV systems in Baghdad.

The economic results are summarized in Table 5. They show that the configuration with the highest energy output is not automatically the most attractive financial option. This distinction is essential for practical design decisions.

The two-axis HJT case illustrates the distinction between maximum energy output and economic attractiveness. HLT requires the highest initial investment (79,720 USD) and delivers the maximum annual energy and avoided operational emissions, but its simple payback extends to about 4.9 years and its discounted LCOE is approximately 0.0179 USD/kWh, higher than that of the simpler fixed and seasonal HJT cases. Conversely, some smaller spaced configurations show favorable normalized financial indicators because of their lower initial investment, although they deliver less total annual energy from the available roof. Financial indicators should therefore be interpreted together with total energy contribution, roof utilization, and constructability.

HLF remains a practical roof-scale option. It produces 244,305 kWh/yr with a discounted LCOE of approximately 0.0116 USD/kWh, a simple payback period of about 3.1 years, and an NPV of approximately 245,366 USD under the stated assumptions. HLSN produces more energy and gives a slightly better modeled financial result, but periodic tilt adjustments may be less practical under local dusty conditions. Some portrait or widely spaced cases can also appear attractive in relative financial terms because of their smaller investments, but they use less installed capacity and therefore provide lower total annual energy and avoided operational CO$_2$ from the roof.

Environmental performance was expressed as operational grid CO$_2$ avoided over the 25-year study period. Using the adopted 0.594 kgCO$_2$/kWh factor and the annual energies in Table 4, avoided emissions range from approximately 1,840.6 to 4,043.8 tCO$_2$. Because this operational metric is directly proportional to generated energy, HLT has the largest avoided-emission value in the present calculation. Possible life-cycle effects associated with extra tracking materials, maintenance, or component replacement are not quantified here and should therefore not be used to reverse that ranking without a separate life-cycle assessment.

Table 6 positions the present study relative to representative Scopus-indexed literature. Earlier work demonstrates the value of PVsyst design analysis, urban near-shading models, HJT technology reviews, yield benchmarking, and financial evaluation, while the present study combines these elements in a single Baghdad rooftop framework.

The findings indicate several potential contributions of rooftop PV to Baghdad's urban energy transition. Electricity generated close to the point of use may reduce daytime grid imports, and replacing a portion of grid electricity can reduce operational CO$_2$ emissions. However, this study does not include a building load profile, time-of-use grid analysis, or backup-generator dispatch model; therefore, peak-demand reduction and backup-generation displacement should be regarded as potential benefits rather than demonstrated outcomes. The results directly show that environmental performance depends on practical rooftop choices such as geometry, module spacing, shading, maintainability, and reliability.

From a practicality perspective, the fixed landscape HJT system is well suited to the case-study roof. It avoids the moving parts and control requirements of tracking, maintains a high PR, and is simpler to operate and maintain in dusty conditions. These characteristics can help the system remain close to expected performance over its service life. The environmental relevance of rooftop PV therefore depends not only on simulated yield but also on selecting a design that is maintainable, durable, and economically transparent in a real urban setting.

This work provides a techno-economic and environmental assessment of rooftop PV systems for a building in Baghdad using a SketchUp/Skelion rooftop model and PVsyst near-shading analysis. Seventeen configurations were compared under Baghdad's climatic conditions, and the simulations showed that the selected HJT module outperforms the comparable monocrystalline module because of its lower power temperature coefficient. Annual energy was maximized by the two-axis HJT system, while the seasonal HJT system achieved the highest PR. Considering energy production, cost, structural simplicity, maintenance requirements, and operational environmental benefit together, the fixed landscape HJT system remains a practical candidate for the case-study roof.

The row-spacing analysis reveals an important rooftop trade-off. Increasing clear spacing reduces mutual shading and increases specific production, but it also reduces the number of modules that can be installed and therefore lowers total rooftop energy. The preferred spacing consequently depends on whether the design objective is maximum total roof energy, maximum specific yield, or greater maintenance access. For this roof, the dense fixed landscape HJT layout provides high total energy and a strong techno-economic result, while the 0.5 m option becomes useful when access and shading reduction receive greater priority.

The avoided-emission calculation also demonstrates the operational environmental significance of rooftop PV for the Baghdad case study. Under the stated assumptions, the simulated systems avoid approximately 1,840.6--4,043.8 tCO$_2$ over 25 years relative to grid electricity. Future work should compare the simulations with measured operating data and evaluate sensitivity to electricity value, grid emission factor, cleaning frequency, and degradation rate. Future work should also investigate storage and load-matching options. Overall, realistic shading modelling is essential for rooftop PV design, and fixed landscape HJT systems are viable candidates for reducing operational emissions in Baghdad.