Scaling a photoreactor is not a matter of multiplying dimensions by a factor, because the physics refuses to scale uniformly. Light attenuates exponentially with path, transfer scales with geometry, and heat scales with volume while cooling scales with area. Three threads must be re-engineered, not enlarged.
Bigger vessels get deeper, and only the illuminated shell reacts, apparent rates per volume fall even with more total light. Countermeasures: keep paths thin by growing area instead of depth, plate and flat designs; or move light inside, immersion lamps and fiber delivery. The validated photon environment of the lab scale must be reproduced, not approximated.
Mixing times grow with scale and stirrer power grows steeply. Re-verify suspension and heat distribution with tracers and multi-point sampling, lab stirring does not translate, it must be redesigned. In flow systems, scale by numbering-up to keep the transport environment identical.
Absorbed light becomes heat proportional to volume, while surface cooling grows only with area. Thermal management that was passive in the lab, a fan, ambient loss, becomes active at scale, chillers, cooling channels, temperature mapping under real irradiation.
At each scale, compare kinetics, selectivity and space-time yield against the smaller benchmark. Hold your acceptance criteria before the first scaled run, and treat any gap as a design signal, not a nuisance.
Perfectlight's scale chain runs from lab photoreactors to pilot and demonstration systems; see engineering solutions.
