A twelve-year-old commercial array is not old, exactly. But it is running modules that were mid-tier when installed, inverters approaching end of service life, and a design that assumed a very different cost structure.
Meanwhile it holds something increasingly valuable: an existing grid connection, an established site, and in some markets a legacy tariff arrangement.
Inverter replacement only. The most common and the least disruptive. Inverters are the shortest-lived major component – typically 10-15 years – and modern replacements are more efficient, better at partial load, and offer grid support functions that older units lack. Modules stay, racking stays, cabling mostly stays.
Partial repower. Replace failed or badly degraded module strings while retaining the majority. Sounds efficient and frequently is not, because mixing module generations in one array creates mismatch. Only sensible if the replacements can be electrically segregated onto their own MPP trackers.
Full repower. New modules, new inverters, reusing the racking and cabling where they remain sound and where the new modules fit the existing rails. Effectively a new plant on an existing site, with a substantial saving on the parts that do not need replacing.
Module efficiency has improved considerably. An array designed around 260 W modules can often be replaced with modules well above 400 W in the same physical footprint. That is a large capacity increase on the same roof, the same racking positions, and the same connection.
Whether you can use the extra capacity depends on your connection agreement and your consumption. If you are export-limited, additional DC capacity still helps by filling more of the day at the cap. If you are consumption-limited, it may not.
The comparison is between continuing as-is and the repower. Continuing means accepting declining output, rising failure rates and inverters that will need replacing regardless. Repowering means capital now against materially higher output for a fresh asset life.
Rough guide to when it usually works:
| Situation | Typically |
|---|---|
| Inverters at end of life, modules performing acceptably | Inverter replacement only |
| Modules degraded well beyond warranty curve | Investigate cause first, then full repower |
| Site is export-limited and array undersized for the cap | Full repower attractive |
| Roof needs replacing within five years | Coordinate both, or defer |
| Legacy tariff at risk from modification | Usually leave alone |
The single best opportunity to repower is when the roof covering needs replacing anyway. The array has to come off regardless, and the marginal cost of putting new modules back instead of old ones is far lower than a standalone repower.
If your roof has ten years left and your inverters have three, plan the sequence now. Replacing inverters at year three and then removing everything at year ten for roof works is a waste that a little forward planning avoids.
Modules removed in a repower are usually still functional at perhaps 85% of original output. Secondary markets exist, and reuse is preferable to recycling on both environmental and economic grounds. Recyclers can handle the remainder – in the EU, producer responsibility obligations under WEEE apply, so there is an established route. Factor the residual value or disposal cost into the case; it moves the number more than people expect.