Commercial vs Residential Solar

Commercial and residential solar design diverge at the electrical service. A house is a single-phase 120/240 V load sitting on a framed, sloped roof. A commercial building is three-phase, usually 120/208 V or 277/480 V, on a low-slope deck, billed partly on peak demand and reviewed against stricter fire-access and labelling rules. The plan set that comes out the other end is a different document.

Key takeaways

• Service type drives everything downstream. Residential is single-phase 120/240 V. Commercial is three-phase, commonly 120/208 V wye on small buildings and 277/480 V wye on larger ones.
• Interconnection stops being a busbar calculation on a main panel and becomes a question of feeder taps, supply-side connections and switchgear under NEC 2023 Article 705.
• Structural approach splits by roof type: penetrating attachments into rafters on residential slopes, ballasted or mechanically attached racking on commercial low-slope decks.
• The 2021 International Fire Code gives non-residential roofs their own dimensioned pathway rules, and it renumbered the solar section from 1204 to 1205.
• Arc-flash labelling under NEC 2023 110.16(B) applies to non-dwelling equipment at 1,000 A and above. It never applies to a house.
• Demand charges change the design objective from annual kWh offset to shaving a coincident kW peak, which changes array size, orientation and storage.

Most installers treat the jump from residential to commercial as a scaling exercise: same design, more modules. It is not. The two run on different services, different structural logic, different fire-code sections, different labelling thresholds and a different economic objective, and if you quote commercial work off a residential process the gaps surface in plan review. Here is what our commercial solar PV design team treats differently.

How does the electrical service differ on a commercial building?

Residential service in the United States is single-phase, three-wire, 120/240 V. That single fact sets the inverter list, the conductor sizes, the busbar rating and the backfeed breaker math for a typical residential solar design. Commercial service is three-phase, and there are two configurations you will meet constantly.

120/208 V wye is the one most designers underestimate. Small commercial buildings, strip retail, restaurants, clinics, small offices and older multifamily are frequently 120/208 V three-phase four-wire, not 277/480 V. It gives you 120 V line-to-neutral for receptacles and lighting and 208 V line-to-line for equipment. 277/480 V wye shows up on larger facilities, warehouses, manufacturing and anything with substantial motor or HVAC load, because higher voltage means lower current for the same power and therefore smaller conductors.

Utilities publish which services they will provide and at what size. Puget Sound Energy’s non-residential electric service handbook is a good example: it lists 120/208 V four-wire and 277/480 V four-wire as the standard three-phase options, and limits single-phase 120/240 V service to loads up to a maximum demand of 100 kW. Verify the actual service before you draw anything, because the design consequences are not cosmetic.

Getting the voltage wrong ripples through the whole set. Inverter selection is voltage-specific, and many three-phase string inverters produce a lower rated AC output at 208 V than at 480 V, so a 208 V site needs more inverters for the same array. Conductor sizing, voltage-drop calculations, conduit fill and the need for a step-up transformer all move with it. On the DC side, string sizing follows the inverter’s MPPT window rather than the service voltage, but the inverter you can use does not, so the two decisions are coupled. Pin the service first.

How does interconnection change under NEC 2023 Article 705 at commercial scale?

On a house, interconnection is usually one calculation. You backfeed a breaker into the main panel and demonstrate compliance with the busbar allowance in NEC 2023 705.12(B)(3)(2), the 120 percent rule: the sum of 125 percent of the inverter output circuit current and the rating of the overcurrent device protecting the busbar cannot exceed 120 percent of the busbar rating, with the breakers at opposite ends. Keep those two percentages straight. The 125 percent figure belongs to continuous inverter output current; the 120 percent figure belongs to the busbar.

At commercial scale that calculation frequently fails, and the design has to move. NEC 2023 restructured Article 705 into a cleaner set of options. Section 705.12 now splits into feeders and feeder taps at 705.12(A) and busbars at 705.12(B), which matters because commercial buildings often have the room to land a source on a feeder or a tap rather than fighting a distribution panel busbar. Section 705.11 covers source connections to a service and was rewritten in the 2023 cycle with dedicated subsections for conductors, connections, service disconnecting means, bonding and grounding, and overcurrent protection. Section 705.13 addresses energy management systems, which can be used to limit current at a connection point instead of upsizing gear.

Then there is the utility. A residential interconnection application is usually a form. A commercial one can involve a screening study, protective relaying, a dedicated utility disconnect with specific accessibility and labelling, revenue-grade metering and a witness test. California projects add a tariff layer on top; our explainer on Rule 21 for solar covers who it binds and what it puts on the drawing. Whichever cycle your jurisdiction has adopted, name it on the plan set, because jurisdictions are spread across several NEC editions and a reviewer will check the one they enforce. Our guide to which NEC articles govern a plan set maps 690, 705, 706 and 710 against the sheets they land on.

How do structural and racking decisions differ by roof type?

Residential structural design is a point-load problem on a sloped, framed roof: identify rafter or truss size and spacing, land an attachment in structure, flash the penetration, and show that added dead load plus wind uplift and snow stay within member capacity under the adopted IBC 2021 or IRC 2021 provisions. The obstacles are physical, and a thorough site survey usually settles them before an engineer opens the file.

Commercial low-slope roofs are a different problem entirely, because you are now designing against a membrane and a deck rather than against rafters. The deck may be steel, concrete or wood; the membrane may be TPO, EPDM, PVC or built-up; and the roof almost certainly carries a warranty that a careless penetration voids. Three approaches dominate.

• Ballasted. The array is held down by weight rather than fasteners, so the membrane is not breached and the roofing warranty is generally preserved. The trade is load: ballast blocks are heavy and the layout is not uniform, because uplift is highest at corners and edges. Unirac’s RM10 EVO installation guide is representative: the system is limited to roof pitches under 2:12, individual blocks run roughly 26 to 38 lb depending on supplier, and the manual is explicit that module loading and ballast counts must be calculated for each individual project rather than assumed.
• Mechanically attached. Used where the roof cannot accept the added ballast weight, where design wind speeds are high, or where the roof lacks the parapet and friction assumptions ballasted layouts rely on. Each attachment becomes a detailed, flashed penetration coordinated with the roofing contractor.
• Hybrid. Ballast across the field with mechanical attachments at corners and perimeter zones, which is often the cheapest way to satisfy uplift without loading the whole deck.

Racking manufacturers certify to UL 2703 for mechanical load and system fire classification, but the governing wind and seismic demands come from ASCE 7 and the SEAOC PV1 and PV2 guidance the manufacturers reference. That is why a commercial ballast layout is a stamped, project-specific drawing rather than a catalogue detail. Carports and ground-mounted arrays add a foundation design on top; those run through our ground mount plan sets workflow.

What does the 2021 fire code require for rooftop access and pathways?

Start with a numbering trap that costs people revision cycles. In the 2018 International Fire Code, solar photovoltaic power systems lived in Section 1204. In the 2021 IFC, Chapter 12, Section 1204 is portable generators and solar photovoltaic power systems moved to Section 1205, with access and pathways at 1205.2 and the non-residential requirements at 1205.3. If your title block or your general notes still cite 1204.3 under a 2021 adoption, you are pointing a reviewer at the wrong section.

For Group R-3 dwellings, the code works in ridge setbacks and pathways. Arrays occupying 33 percent or less of the plan-view roof area require a clear setback of not less than 18 inches on both sides of a horizontal ridge; above 33 percent coverage that setback increases to 36 inches. Pathways from the lowest roof edge to the ridge are required on separate roof planes, and panels may not be placed below emergency escape and rescue openings. Several exceptions apply, including sprinklered dwellings, so read the adopted local text.

Buildings other than Group R-3 get a separate, more dimensional regime at 1205.3:

• Perimeter pathways (1205.3.1). A minimum 6-foot-wide clear perimeter around the edges of the roof, reduced to a minimum 4-foot-wide clear perimeter where either axis of the building is 250 feet or less.
• Interior pathways (1205.3.2). Pathways at intervals not greater than 150 feet throughout the length and width of the roof, plus a pathway not less than 4 feet wide in a straight line to roof standpipes or ventilation hatches, and a pathway not less than 4 feet wide around roof access hatches.
• Smoke ventilation (1205.3.3). A pathway not less than 4 feet wide bordering all sides where non-gravity-operated smoke and heat vents occur, and a pathway not less than 4 feet wide on not fewer than one side where gravity-operated dropout smoke and heat vents occur. Between array sections the code accepts either a pathway not less than 8 feet wide, or a pathway not less than 4 feet wide bordering 4-foot by 8-foot venting cutouts every 20 feet on alternating sides of the pathway.

These setbacks are not trivial on a large roof. A 6-foot perimeter plus interior and ventilation pathways can remove a meaningful share of the usable area, so model the pathways before you commit to a system size, not after. Local amendments compound it: Washington’s adoption of the 2021 IFC, for instance, adds a requirement that multiple arrays be separated by a 3-foot-wide clear access pathway. That kind of overlay is covered in our piece on permitting requirements that catch installers out, and it is a common reason a set comes back marked up. See also why solar permits get rejected.

When do arc-flash labels and other commercial-only markings apply?

Labelling is one of the cleanest dividing lines between the two worlds, because the trigger is written as an occupancy test. NEC 2023 110.16(A) requires an arc-flash hazard warning on electrical equipment in other than dwelling units where that equipment is likely to require examination, adjustment, servicing or maintenance while energized. NEC 2023 110.16(B) goes further: in other than dwelling units, a permanent label is required on service equipment and, new in this cycle, on feeder-supplied equipment rated 1,000 A or more. The prior threshold was 1,200 A and covered services only, so a project designed off an older habit will under-label.

The label has to be applied in accordance with acceptable industry practice, carry the date it was applied, and meet the durability requirements of 110.21(B). In practice that means the incident-energy or PPE-category information is coming from an arc-flash study, and someone on the project has to own that scope. It is not automatically the PV designer, but it does need to be assigned in writing, and the plan set should show the label schedule and locations so the inspector is not guessing.

Commercial sets carry markings a residential set never needs: equipment identification for multiple sources, directory placards at each service disconnect, utility-required lockable disconnect labelling, and raceway identification through electrical rooms. Bundle them onto a dedicated placard sheet, as our breakdown of what belongs in a permit-ready plan set shows.

How do demand charges change the way you size a commercial array?

Residential customers are billed almost entirely on energy: cents per kWh, sometimes time-of-use. So a residential design optimises annual kWh production against roof area and budget. Commercial customers are usually billed on two meters’ worth of logic at once: energy in kWh, plus a demand charge in dollars per kW based on the highest short-interval draw in the billing period. That second charge is often a large share of the bill, and it does not care how much energy you produced in April.

The numbers are public. Austin Energy’s commercial rate schedule applies demand charges to commercial customers with an average summer peak demand of 10 kW and higher, and bills secondary-voltage customers between 10 kW and 300 kW at $9.83 per kW of billed demand, rising to $12.56 per kW at 300 kW and above. Read that against a facility with a 400 kW peak and the demand line alone is worth roughly $60,000 a year before a single kWh is priced. Tariffs vary enormously by utility, so pull the customer’s actual schedule rather than assuming.

Once demand is in the model, sizing logic changes:

• You need interval data, not a twelve-month bill summary. The question is whether PV output is coincident with the facility’s peak, and that only shows up in 15-minute or hourly data.
• Orientation and tilt stop being purely about annual yield. A west-shifted or lower-tilt layout that gives up some annual kWh can be worth more if it holds output into the late-afternoon peak.
• Storage or a power control system becomes a sizing variable rather than an upsell, because a battery can clip a peak that PV alone cannot reach.
• Oversizing DC to AC is evaluated against clipping losses and demand coincidence together, not against yield alone.

This is why commercial proposals lean on hourly modelling rather than a rule of thumb, and why the production report is doing real financial work on a C&I project instead of decorating a proposal.

Who stamps the drawings, and how deep does the plan set go?

Plenty of residential permits clear on a well-drawn set with no seal at all, or with a structural letter only when an AHJ asks for one. Commercial is the reverse. A structural PE or SE stamp verifying the deck and framing under the array’s dead, live, wind and seismic demands is close to universal, and an electrical PE stamp is common where load calculations, a supply-side connection or utility-grade switchgear are involved. Wet stamps, ground-mount and roof-mount structural certifications and post-install affidavits are all separate deliverables. Our article on when a solar project needs a PE stamp covers the triggers in detail, and the stamp itself is ordered through engineering stamps. Stamping is licence-bound and jurisdiction-bound, and a short list of jurisdictions falls outside what we can seal; the current list is published on that page, so check it before you promise a client a date.

Sheet depth follows. A residential permit set is typically a cover sheet, site plan, roof plan, attachment and structural details, a single-line diagram, placards and datasheets. A commercial set adds:

• A three-line diagram alongside the single line, showing all three phases, CTs, metering and protective devices.
• Conductor, conduit and equipment schedules with voltage-drop calculations carried through.
• A roof plan dimensioned for fire-code pathways, hatches, vents and setbacks rather than a generic layout.
• A ballast or attachment layout keyed to wind zones, with the supporting structural calculations.
• Equipment pad, switchgear and utility metering details, plus grounding and bonding plans.
• A dedicated placard and labelling sheet.

That is the difference between a handful of sheets and a set that can run into the twenties. If you want a pre-flight list before submitting either type, our pre-submittal checklist and the wider list of design considerations before you draw are the two to run.

How does Avila Solar Drafting handle both project types?

We are a drafting and engineering partner for solar installers and EPCs, not an installer. You keep the customer, the sale and the submittal; we produce the documents, designed to pass first-time review against the code cycle the jurisdiction has actually adopted. Standard plan set turnaround is 2-3 business days, and 1-2 business days on Fast Roof, from the point we have complete and accurate project information. Pricing depends on scope and complexity — current figures are on the solar plan sets page, and revisions are free for six months.

Commercial packages cover rooftop, ground-mount and carport systems up to 500 kW as a standard order, with anything larger custom quoted. If you want to see how the commercial side compares with a residential workflow in practice, read commercial solar plan sets: what changes above residential, and if you are weighing whether to build this capability internally, outsourced solar design vs in-house lays out the comparison. Broader scope sits under solar drafting services and residential work under solar plan sets.

Ready to move a commercial project? Order a commercial plan set, or call 971-410-0655 to talk through a system above 500 kW.

Frequently asked questions