DC Optimizer vs Microinverter vs String Retrofit For Shaded Roof Decision Matrix
Why this matters
A partially shaded roof is the single most common reason a residential array underperforms its model. With a plain string inverter, the weakest module in a series string drags the whole string down because the current path is shared; one shaded module clamps the operating point of every panel wired with it. Module-level power electronics (MLPE), either DC optimizers or microinverters, break that coupling. Choosing among optimizers, microinverters, and a straight string retrofit changes harvest, cost, fault behavior, and code compliance under NEC 690.12 rapid shutdown. Picking wrong means either leaving energy on the roof or spending on hardware the shade profile never needed.
This matrix is for the tech standing on a roof deciding what to quote, or diagnosing why an existing string system underproduces in afternoon shade.
The options
String inverter (no MLPE). All modules in a string wired in series to a central inverter. Maximum power point tracking (MPPT) happens at the string level, so the string finds one operating voltage and current for all modules together. Cheapest hardware, fewest failure points, highest single-point-of-failure exposure. On an unshaded, single-plane roof it is often the best value. NEC 690.12 still requires a rapid shutdown solution; a string system needs a separate rapid-shutdown device per module or a compliant initiator at the array boundary.
DC optimizers. A small DC-DC converter bolts to each module and feeds a string inverter designed for it. Each optimizer runs its own MPPT, buck/boost converting so the string current stays uniform even when one module is shaded. The string inverter still does the DC-AC conversion. Optimizers satisfy 690.12 module-level shutdown natively. Conduction losses through the optimizer are small but real (a few tenths of a percent) on unshaded modules.
Microinverters. A full DC-AC inverter per module (or per two modules). Each module produces AC independently; modules are paralleled on an AC trunk. Complete electrical independence, no high-voltage DC on the roof, and 690.12 compliance is inherent. Highest per-watt hardware cost and the most roof-mounted active electronics to fail, but each failure costs only one module's output.
When A/B/C wins
String wins when the roof is a single plane with negligible shade, the budget is tight, and the owner accepts that a future shade source (new tree, neighbor's addition) would force a retrofit. It also wins on large ground mounts where shading is controlled by row spacing, not obstructions.
Optimizers win on moderate, predictable shade (a vent stack, a single chimney clipping a corner of the array for part of the day) and on long strings where you want one inverter location and centralized monitoring with module-level data. They also win when the design needs high DC voltage for long conductor runs; the string architecture keeps current low and voltage high, reducing wire size.
Microinverters win on heavy or chaotic shade (multiple dormers, trees on more than one azimuth), on complex multi-plane roofs where you would otherwise need many short strings, and where the owner wants no high-voltage DC on the structure for safety or fire-service reasons. They also simplify expansion: add modules without rebalancing strings.
A quick rule for shade severity: if any module is shaded for more than two to three hours on more than one face of the roof, MLPE pays back. If shade hits one module on one plane briefly, optimizers usually suffice. If there is effectively no shade, do not pay for MLPE to solve a problem you do not have.
String and optimizer systems carry hazardous DC voltage (often 300 to 600 V) on the roof until rapid shutdown is triggered. Never assume a module is de-energized in daylight. Verify rapid shutdown function and measure conductor-to-ground voltage before touching DC wiring. Arc-fault and ground-fault protection per NEC 690.11 and 690.41 must be confirmed operational on any retrofit.
Field decision flow
- Map the shade. Walk the roof and note every obstruction's azimuth and the hours it casts on the array. A shade study or on-site solar pathfinder reading beats guessing.
- Count the planes. One plane, no shade points to string. Two or more planes, or any tree on multiple azimuths, points to microinverters.
- Score the shade severity. Brief, single-module, single-plane shade points to optimizers; persistent multi-plane shade points to microinverters.
- Check the conductor run. Long DC home runs favor the high-voltage string or optimizer architecture for smaller wire; very short runs neutralize that advantage.
- Confirm 690.12 path. String needs added rapid-shutdown devices; optimizers and microinverters satisfy it natively. Factor the added device count into the string option's true cost.
- Confirm AC vs DC service constraints. Microinverters add AC current on the trunk and may bump conductor or breaker sizing on the AC side; verify the service can absorb it.
- For a retrofit diagnosing existing underproduction: if a string system loses afternoon production matching a known shade source, the fix is MLPE (optimizers if the inverter can be reused with a compatible model, microinverters if a full re-architecture is warranted), not inverter replacement alone.
When two options tie on harvest, choose the one with fewer roof-mounted active components for long-term service simplicity, unless the owner explicitly values the redundancy microinverters provide.
References
- NEC (NFPA 70) Article 690.12, Rapid Shutdown of PV Systems on Buildings
- NEC (NFPA 70) Article 690.11, Arc-Fault Circuit Protection (Direct Current)
- UL 1741, Inverters, Converters, Controllers and Interconnection System Equipment for Use With Distributed Energy Resources
- IEC 62446-1, Photovoltaic (PV) systems - Requirements for testing, documentation and maintenance