Single-junction or tandem? - The next platform decision for solar PV manufacturing.

For fifty years the solar industry has driven module prices down along a remarkably stable learning curve, built on a single platform of wafer-based crystalline silicon. With mass-production cells now within a few percentage points of the practical efficiency limit for single-junction silicon, the industry faces a choice between continuing to remove cost from the existing platform and moving to perovskite-on-silicon tandem cells. Remaining on silicon is estimated to reduce the learning rate from around 24% to 10-15% per doubling, leaving module prices 15-35% above the historical trend by the late 2030s, while a maturing, increasingly replacement-driven global market will slow the learning curve further in calendar time. Tandem cells could restore a steeper curve, but only at a very large transition cost, with unresolved risks in stability, manufacturability and end-of-life toxicity, and with a supply chain whose R&D capacity has been hollowed out by the exit of Western equipment and materials vendors and whose Chinese successors face severe overcapacity and mounting losses. Since solar is already the cheapest source of new-build generation in most markets, the gains from tandem appear small relative to its cost and risk, and the more prudent course for the Chinese industry would be to remain on single-junction silicon, leaving others to bear the expense of proving tandem before adopting it.