Why curtailment is becoming a feature rather than a bug

I was looking at some recent capacity factor data this morning and it reminded me of that classic engineering trap where you spend four hours automating a five-minute task. We see a similar psychological hurdle in grid debates regarding solar overbuild.

For decades, the energy sector operated under a scarcity mindset. Fuel was the expensive variable, so efficiency was the god we worshipped. Every electron had to be accounted for, and wasted generation was seen as a failure of system design. But the rapid drop in solar LCOE has fundamentally broken that logic.

When the marginal cost of generation approaches zero, the economics of wasted energy flip. It is becoming increasingly clear that it is cheaper to overbuild solar capacity by 20% or 30%, and accept significant curtailment during peak production hours, than it is to try to perfectly match generation to load with expensive seasonal storage or firming assets right now.

The efficiency trap here is the obsession with utilizing 100% of potential output. We still see arguments treating curtailment as a system failure or a sign of grid incompetence. In reality, if you aren't curtailing a significant percentage of your VRE production, you probably haven't built enough of it.

We are moving from a fuel-intensive model to a capital-intensive model where the asset is cheap enough to sit idle or be wasted part of the time. It feels counterintuitive to anyone raised on the old thermal baseload orthodoxy, but the math on overbuilding plus moderate storage beats the perfect efficiency model almost every time now.

I’m curious if anyone has seen recent modeling on the inflection point for this. At what LCOE price point does the just build more panels strategy hit diminishing returns against long-duration storage costs?