Published: June 2026 Technical Level: Advanced Category: Power Systems Design
Commercial solar photovoltaic systems in the 100 kW to 5 MW range occupy a design space where utility-scale simplicity meets building-integrated complexity: the systems are large enough to require a formal interconnection study and utility approval process, small enough to be roof- or ground-mounted at the customer's facility, and governed simultaneously by NEC Article 690, IEEE 1547-2018, and the applicable utility tariff's net metering or net billing provisions. This paper develops the electrical design basis for commercial PV systems: string sizing for maximum power point tracker compatibility, DC/AC ratio selection and its implications for inverter clipping losses, rapid shutdown compliance under NEC 690.12, IEEE 1547-2018 interconnection requirements including voltage and frequency ride-through and anti-islanding, and the economic analysis framework that translates energy production estimates into financial metrics under time-of-use and demand charge rate structures. The design methodology is illustrated with a 500 kW rooftop system example that demonstrates the interaction between production optimization, interconnection constraints, and tariff economics.
Commercial solar PV deployment has grown from a niche sustainability initiative to a mainstream capital investment decision for commercial building owners. The economic case is driven by three factors that have converged since 2018: module and inverter costs have declined to the point where installed system costs of $1.00 to $1.50 per watt AC are achievable for rooftop systems above 500 kW; the federal Investment Tax Credit provides a 30 percent tax credit on the installed cost under the Inflation Reduction Act; and time-of-use electricity rates in most utility service territories create a price premium for on-peak generation that aligns well with solar PV's midday production peak. Together, these factors produce simple paybacks of 5 to 9 years for well-designed commercial systems in most U.S. regions, with lifetime internal rates of return of 10 to 18 percent.
The engineering design must satisfy two concurrent objectives: maximizing energy production within the physical constraints of the available roof or ground area, and complying with the interconnection requirements that the utility imposes as a condition of operating the system in parallel with the grid. These objectives occasionally conflict — a production-maximizing design may exceed the utility's allowed export capacity and require curtailment, or a string configuration that minimizes shading losses may produce a DC bus voltage that is incompatible with the selected inverter's operating range. The engineer resolves these conflicts by understanding both the production physics and the interconnection regulatory framework.
A PV string consists of modules connected in series, with the string voltage equal to the sum of individual module voltages. The string must be sized such that the maximum string open-circuit voltage at minimum expected ambient temperature does not exceed the inverter's maximum input voltage rating, and the minimum string operating voltage at maximum expected module temperature is above the inverter's minimum MPPT voltage.
The maximum string open-circuit voltage is:
Where: is the number of modules in series.
is the module open-circuit voltage at standard test conditions (25°C, 1000 W/m²).
is the module's open-circuit voltage temperature coefficient in %/°C (typically −0.25 to −0.35%/°C for crystalline silicon).
is the minimum expected ambient temperature at the site in °C.
NEC 690.7 requires that the maximum system voltage (the maximum string open-circuit voltage for series-connected strings) not exceed 600 V for residential systems or 1,000 V (or 1,500 V for listed systems) for commercial systems. The string is sized to remain within this limit at the coldest temperature the site will experience, because module voltage increases as temperature decreases.
For a worked case, consider a commercial rooftop array using modules with V and a temperature coefficient %/°C, feeding a 1000 V-rated inverter at a site with a record minimum temperature of °C. The per-module open-circuit voltage at the coldest condition is:
The maximum number of modules per string is then the inverter limit divided by this cold-temperature module voltage: , which rounds down to 18 modules. An 18-module string reaches V at the record low temperature, safely within the 1000 V limit, while a 19-module string would reach 1033 V and violate NEC 690.7. This is the calculation that most often governs string length in cold climates — the coldest morning of the year, not the operating condition, sets the maximum string size, and designers who size strings at the standard-test-condition voltage routinely produce strings that overvoltage the inverter on the first hard freeze.
The minimum string MPPT voltage at maximum module temperature determines the minimum number of modules per string. As module temperature increases, the maximum power point voltage decreases. If the string MPPT voltage falls below the inverter's minimum MPPT window, the inverter will not operate the array at maximum power or may disconnect entirely. String sizing must verify that the operating MPPT voltage remains within the inverter's window across the full site temperature range.
The DC/AC ratio — the ratio of the array's nameplate DC capacity to the inverter's AC output rating — is a key design parameter that balances production optimization against inverter clipping losses. A DC/AC ratio of 1.0 means the array and inverter are exactly matched; a ratio of 1.25 means the array can theoretically produce 25 percent more DC power than the inverter can export as AC power during peak irradiance conditions. The inverter clips the excess power by operating the array away from its maximum power point, reducing instantaneous production but accepting this loss in exchange for increased production during the shoulder periods of the daily production curve, when irradiance is below the level that saturates the inverter.
For a south-facing rooftop array in the southwestern United States, a DC/AC ratio of 1.20 to 1.30 typically maximizes the annual energy yield while keeping clipping losses below 2 to 3 percent of annual production. At higher ratios, the incremental clipping losses begin to exceed the incremental production gains from the additional modules, reducing the marginal economics of the last modules added. The optimal ratio depends on the site's irradiance distribution: locations with frequent periods of very high irradiance (Arizona, Nevada) have higher optimal ratios than locations with more diffuse, lower-peak irradiance (Pacific Northwest, Great Lakes region).
IEEE 1547-2018 establishes the performance requirements for DER interconnected with the area electric power system at voltages below 100 kV. For commercial PV systems, the most operationally significant requirements are the voltage and frequency ride-through requirements and the anti-islanding requirements.
The voltage ride-through requirements mandate that the PV inverter remain connected and continue to produce power (or modulate reactive power) through most abnormal voltage conditions rather than tripping immediately on undervoltage or overvoltage. The standard defines three categories of abnormal voltage response, and the applicable distribution system operator specifies which category applies to each interconnection. Category III, which applies in most urban and suburban distribution circuits, requires DER to ride through to lower voltages and for longer durations than Category I, reflecting the higher system stability value of DER that remain connected during grid disturbances.
The anti-islanding requirement mandates that the PV system detect the formation of an unintended island — a condition where the inverter continues to energize a portion of the distribution circuit after the utility source has disconnected — and cease energizing the circuit within two seconds. All modern transformerless and transformer-coupled grid-tied inverters include passive anti-islanding functions (frequency deviation detection, rate-of-change of frequency) and active anti-islanding functions (frequency or voltage perturbation injection). IEEE 1547-2018 does not specify the detection method but requires that the function operate reliably for any load-generation imbalance outside the defined non-detection zone.
NEC 690.12 requires that PV systems on buildings include a rapid shutdown function that, when activated, reduces the DC voltage in the PV array wiring within the building's fire boundary to 30 V or less within 30 seconds of initiating the shutdown. This requirement protects firefighters from the hazard of energized conductors inside the building during a fire suppression operation. Rapid shutdown is implemented either through module-level power electronics (microinverters or DC optimizers) that de-energize each module individually when commanded, or through rapid shutdown initiators at the array junction points that disconnect the modules from the downstream wiring when the building-level rapid shutdown initiator is activated.
The financial performance of a commercial PV system is governed by the interaction between the energy production estimate, the applicable utility rate structure, and the financing terms. The annual energy production in kWh is estimated using simulation tools (PVWatts, SAM, or detailed PVsyst models) that combine the site's solar resource data (TMY3 or TMY4 weather files), the array's tilt, orientation, and shading factors, the inverter's efficiency curve, and the system's DC and AC losses.
The economic value of the production depends on the rate structure. Under a flat energy rate, the value is the product of the annual production and the retail energy rate. Under a time-of-use rate, the value is higher because solar production coincides with on-peak periods when the energy rate is highest. Under a net metering tariff with a retail-rate credit for exported energy, the full production value is realized regardless of whether the building's load exceeds the solar output in each interval. Under a net billing tariff with a reduced export credit, the value of production that exceeds the building load is discounted, incentivizing system sizing to minimize export.
The analysis in this paper connects to several companion studies in this library. Readers concerned with the upstream and downstream engineering will find Battery Energy Storage Systems Integrated with Solar PV develops a closely related aspect of the same problem, while Utility-Scale Renewable Integration extends the treatment into an adjacent domain. For the broader methodological context, DER Integration and Voltage Regulation provides complementary depth.
Commercial solar PV systems in the 100 kW to 5 MW range occupy a design space governed simultaneously by NEC Article 690, IEEE 1547-2018, and the applicable utility tariff, and the design basis developed in this paper shows that the string-sizing and DC/AC-ratio decisions are the technical core from which interconnection compliance and economic performance both follow. The central engineering conclusion is that string sizing must satisfy the maximum-power-point-tracker voltage window across the full temperature range — using the record-low ambient temperature to bound the maximum open-circuit voltage per NEC 690.7 — because a string that exceeds the inverter or equipment voltage rating at cold-temperature extremes is a code violation and a reliability hazard regardless of its mid-range performance. The DC/AC ratio then trades inverter clipping loss against capital efficiency, and the rapid-shutdown and IEEE 1547-2018 interconnection requirements impose the remaining constraints. For the engineer, the durable discipline is to anchor the array design to the temperature-corrected voltage limits first, then optimize the economics within that envelope.
[1] IEEE Standard 1547-2018, Standard for Interconnection and Interoperability of Distributed Energy Resources, IEEE, 2018.
[2] NFPA 70, National Electrical Code, Article 690, 2023 edition, NFPA, 2023.
[3] NREL, PVWatts Calculator User Manual, National Renewable Energy Laboratory, 2024.
[4] NREL, System Advisor Model (SAM) User Manual, NREL, 2024.
[5] IEC 61215, Terrestrial Photovoltaic Modules — Design Qualification and Type Approval, IEC, 2021.
[6] IEEE Standard 929-2000, Recommended Practice for Utility Interface of Photovoltaic Systems, IEEE, 2000. (Superseded by 1547-2018 but referenced for legacy system analysis.)
[7] Lawrence Berkeley National Laboratory, Tracking the Sun: Pricing and Design Trends for Distributed Photovoltaic Systems, LBNL, 2024.
[8] Inflation Reduction Act of 2022, Investment Tax Credit provisions for solar energy, 26 U.S.C. §48, 2022.