Grounding and DC-side errors survive to the field. Here's why.

Most solar projects that fail do not fail in the energy model. They fail somewhere between an approved drawing and the field, in the details that were never fully reconciled: grounding, DC-side protection, conductor sizing, and the assumptions buried in a stamped set.

These are not exotic failure modes. They are boring ones. That is what makes them dangerous — they pass permitting, they pass a modeling review, and they pass a casual design check because nothing about them looks wrong at a glance.

Why these errors get through

A solar design travels through a lot of hands. The energy model validates production. The structural and civil drawings validate the layout and foundations. The electrical design is often assembled from prior projects and adjusted for the new site, and that is exactly where a copied assumption becomes a site-specific error.

Grounding is a good example. An electrode system that performed on one site may be inadequate on a site with different soil resistivity, different local utility requirements, and different fault-current contributions from an interconnected system that did not exist when the template was built. On paper, it still looks complete.

DC-side issues follow the same pattern. String and source-circuit conductor sizing, overcurrent protection coordination, and disconnect placement are frequently carried over from a reference design without re-checking them against the actual module, inverter, and site configuration.

What a useful review checks

A review that is meant to catch field-level errors is not a documentation audit. It reconciles the design against the specific site and the specific equipment. In practice that means:

• Soil resistivity and fault-current assumptions against the actual grounding design
• Source-circuit and feeder ampacity against real temperature, conduit fill, and termination ratings
• Overcurrent protection coordination across the DC and AC sides
• Disconnect and labeling compliance against the adopted code cycle and the utility's requirements
• As-procured equipment against what the design assumed

None of this requires guesswork. It requires reading the design the way a field crew will build it, and asking where the assumptions came from.

The economics of timing

The cost of these findings is almost entirely a function of when they are found. A grounding correction at 30% design is a mark-up and a resubmittal. The same correction at 90% design is a change order. The same correction after mobilization is a field rebuild.

I have seen this play out directly. On a 120 MW utility-scale build, design review identified a grounding error that would have required re-trenching across 40 acres. Caught before mobilization, the fix was a drawing change. The project avoided more than $200K in rework and kept its COD date.

The most expensive design review is the one you skip. The second most expensive is the one you do late.

The lesson is not that every project is hiding a grounding error. It is that the window where findings are cheap is short, and it closes earlier than most schedules assume. Review belongs at design milestones, before procurement locks in equipment and before a crew is standing on the site.

That is the job of independent owner's engineering: not to second-guess for its own sake, but to reconcile the design against the ground it will actually be built on — while the answer is still just a revision.