1. Define your site

Draw a boundary on the map, upload a KML or GeoJSON file from Google Earth, or enter an address. The tool auto-detects elevation, timezone, and weather data for your location.

2. Weather & terrain

OptiSolar fetches a Typical Meteorological Year (TMY) from PVGIS, which combines 10+ years of satellite-measured solar irradiance, temperature, and wind data into a representative annual profile. For terrain, USGS elevation data is sampled across your site to model how slopes affect row placement, shading, and grading cost.

3. Layout generation

For each candidate design, the optimizer places tracker rows inside your lot boundary using a polygon-slicing algorithm. It accounts for setbacks, exclusion zones, access roads, and row-length constraints. Each row gets a drive motor, torque tube, and structural posts costed into CAPEX.

4. Energy simulation

Energy production is modeled using pvlib-python, the industry-standard PV performance library. The simulation includes:

• Single-axis tracking with backtracking to eliminate inter-row shading on flat ground
• Bifacial irradiance via the infinite-sheds model, capturing rear-side gains from ground albedo
• Terrain-aware shading using a vectorized 2D cross-section model that accounts for elevation differences between rows
• Loss modeling including soiling, wiring, mismatch, availability, thermal derating, and inverter efficiency
• Inverter clipping from DC/AC oversizing

5. Economics & LCOE

For each design, a full CAPEX breakdown is computed: modules, inverters, balance of system, tracker hardware, land, grading, and interconnection. LCOE is calculated using a 30-year discounted cash flow that includes the federal Investment Tax Credit (ITC), 5-year MACRS depreciation, O&M, property tax, insurance, and annual degradation.

6. Optimization

The optimizer uses differential evolution, a population-based search algorithm, to explore combinations of:

• Ground coverage ratio (GCR)
• Tracker azimuth (axis orientation)
• Maximum rotation angle
• Axis height
• DC/AC ratio
• Ground albedo
• Grading strategy (none, terrace, optimized slope)
• Racking type (single-axis tracker or fixed-tilt)

Each combination is screened against bankability gates (shading < 2%, performance ratio > 80%, clipping < 5%) before being evaluated. Designs that fail are rejected. The search runs in two phases: a fast analytical scan across the full design space, then a high-fidelity bifacial simulation polish on the winning design.

7. Baseline comparison

Every result is compared against an un-optimized baseline: a standard tracker layout at GCR 0.33, azimuth 180° (true north-south), and DC/AC ratio 1.25. This represents a reasonable but generic design that a developer might pick without optimization. The savings figures show the LCOE improvement from optimizing to your specific site.

Limitations

• The optimizer assumes a single module type (550 W bifacial, 2-in-portrait) and a single inverter size (3.3 MW central). Real projects may mix equipment.
• Weather data is TMY (typical year), not an actual year. Real production will vary ±10% year-to-year.
• Grading costs are estimated from elevation samples, not a full geotechnical survey.
• Interconnection costs are parameterized but do not model specific utility requirements or queue positions.
• The 5,000-acre lot limit prevents excessive compute on this hosted instance.

Want custom assumptions?

Equipment specifications, loss parameters, and bankability thresholds are fixed in this hosted tool. If your project needs different module types, custom loss budgets, or site-specific interconnection modeling, book a call with Jairus to discuss a dedicated deployment.