The need for re-powering a plant may occur even before 25 years, if you get to the point where the technology continues to come down in cost – so that 10 years in, you can actually see the economic benefit of the cost of replacing the current modules with new modules that are have a higher power output.
In this way, you can potentially see an endless re-powering approach, which is very effective because you’re using the same infrastructure, and you already have sunk development costs.
If re-powering isn’t an option, the most obvious option is decommissioning – in that case we recycle all the panels at the end of their useful life, and turn about 90 per cent of the material into new panels.
Can you elaborate on the grid constraints that these larger projects encounter?
Generally, you find better land availability where there’s not as useful transmission availability. As a result, a lot of the plants that have been constructed in North America have been massive. I don’t think you’re going to see a lot of these going forward – largely because it is difficult to find that optimum balance between land availability and transmission availability – but more importantly you’re really not leveraging the modularity of photovoltaic (PV) solar as much as you could.
Solar PV essentially realises all economic economies of scales at about 20 megawatts (MW), so if it doesn’t cost you any more on a unit basis to build 20 MW as opposed to 500 MW, then what you should be doing is leveraging more strategic places on the grid where you don’t have those grid constraints, where it’s easy to find a site, where it’s closer to distribution load, or it’s not as much of a pain to connect, or it doesn’t have the same line loss impact.
We see the sweet spot in the 10–50 MW range predominantly, and then in some scenarios at 50–100 MW.