Module-level power electronics change how a loss propagates, not whether it happens. A shaded module produces less energy whether or not it has its own converter attached. What the converter prevents is the shaded module dragging the rest of its series string down with it. That is a genuine and sometimes large benefit on a complicated roof, and it is a much smaller one on a clean, uniform, unshaded plane.
The problem these devices were invented to solve
Modules wired in series share one current path, so the least capable member sets the ceiling for the whole string. Put a chimney shadow across one module for three hours a morning and the other eleven modules in that string spend those hours operating below their capability. Bypass diodes inside the module limit how bad this gets by shunting current around a badly shaded cell group, but a conducting bypass diode means that portion of the module contributes nothing at all. Module-level power electronics break the shared-current constraint by giving each module its own conversion stage. The shaded module delivers what it can, its neighbors deliver what they can, and the two are no longer chained together. On an array with dormers, vent stacks, a neighbor’s tree and three roof planes, that architectural change can recover a meaningful share of annual production.
What the marketing usually leaves out
Recovering mismatch loss is not the same as recovering shade loss. If a module is in full shadow, no electronics anywhere will make it produce, and the energy is simply gone. Independent studies and design tools model this correctly, but sales conversations often blur the distinction and imply that per-module conversion turns a shaded roof into an unshaded one. It does not. The right way to evaluate the benefit is to have the array modeled twice, once with a string architecture and once with module-level conversion, using an actual shade study of your roof. On a heavily obstructed site the difference will be obvious and worth paying for. On an open south-facing plane the modeled difference is often small enough that the extra hardware is being bought for other reasons.
What else the architecture buys you
Per-module conversion brings a set of benefits that have nothing to do with shade. It provides module-level monitoring, which turns troubleshooting from a guessing game into reading a screen and makes a failing module visible long before it shows up in a monthly total. It usually satisfies the module-level shutdown function required for rapid shutdown, since each device can stop sending voltage down the line. It removes the constraint that every module in a string must share an orientation, which makes awkward roofs designable at all. And it decouples array expansion from string sizing, so adding modules later is often simpler. For many residential projects those practical advantages, not the shade arithmetic, are the real justification.
The costs that arrive later
Every module-level device is an electronic assembly mounted on a roof, under glass, through thousands of thermal cycles, in weather. That is a demanding environment for power semiconductors and electrolytic components. Failures are not common in good products, but when one happens the repair requires roof access, module removal and often scaffolding, and the labor frequently exceeds the part cost. A string inverter, by contrast, sits at eye level in a garage or on a wall where replacement is straightforward. There are also compatibility constraints: optimizers generally work only with the manufacturer’s own inverter, and a discontinued product line can leave an owner with no in-kind replacement. Ask what the warranty covers on labor, how long the manufacturer commits to parts availability, and what happens to the system if one device fails.
How to decide for a specific roof
Start with a real shade study rather than an impression. Design software and site survey tools can quantify how much of each module’s annual irradiance is obstructed and when. If the array is uniform and the obstructions are minor, model both architectures and compare. If the roof has multiple planes, significant obstructions, or a shading pattern that moves across the array through the day, the module-level option is usually the stronger design regardless of the arithmetic, because it also makes the system monitorable and expandable. Whichever way you go, the module-level shutdown requirement in your jurisdiction has to be satisfied by some means, and that is a code question for the designer and the authority having jurisdiction, not a marketing question.
What to check before you act
- Ask for a shade study and a modeled comparison of both architectures.
- Confirm whether the warranty pays for roof labor on a failed device.
- Check that optimizers and inverter come from a compatible product family.
- Verify the design satisfies module-level shutdown as adopted locally.
Common questions
Will optimizers fix a tree that shades half my roof?
No. They stop the shaded modules from limiting the unshaded ones, but the shaded modules still make little power. Trimming or relocating the array does more than any electronics can.
Can I add optimizers to an existing string array later?
Sometimes, but the inverter usually has to be compatible or replaced, and the string design has to be reworked. Treat it as a redesign by a qualified installer, not an accessory.
The practical takeaway
Buy module-level conversion for what it actually delivers: containment of mismatch, per-module visibility, design freedom on awkward roofs and a shutdown path. Do not buy it expecting shaded modules to produce like sunny ones. Get the roof modeled both ways and let the shade study decide.
References and further reading
- NREL: photovoltaics research including system architecture and performance
- NREL PVWatts: models expected production for a given array configuration
- NFPA: background on the National Electrical Code, including photovoltaic shutdown requirements
Numerical scenarios are illustrative unless identified otherwise. Follow the exact product instructions; component ratings and local installation requirements can differ.
