Lately we’ve been hearing a lot about gigawatt-scale fabrication shops (fabs) for both PV (NY) and batteries (NV). And they are promoted as being able to provide great economies of scale that will dramatically and magically accelerate the cost reduction learning curve. But the fact of the matter is that it’s not quite so simple in reality. And, just like the revelation that the Wizard of Oz is just a man behind a curtain, the truth may come as a disappointment to many.
Having worked on setting up components of GW fabs (thin-film PV), and from being involved in the equipment sector of both the solar and semiconductor manufacturing business, I can offer a bit of a peek behind the curtain. While there is definite validity to the notion of economies of scale, there are some hard realities that can’t allow for hyper-accelerating the cost reduction learning curve that we have all come to know and love.
Here’s the crux. When you outfit a fab, you are buying the best equipment and recipes available at the time. However, there is a natural learning to any equipment cycle — the first generation is good, but in use, there is a feedback loop that allows for improvements. This is a standard part of any equipment manufacturing company — often called the Continuous Improvement Program (CIP), which intrinsically takes time.
No piece of equipment is even near perfect when it is sold to a production company, but it’s good enough at the time. The equipment vendor works with the production company, takes feedback, allows for changes and improvements, and defines what the next generation of equipment should do differently to have better production. This learning happens on 1 piece of equipment, so having 10 more doesn’t make the learning faster because it’s the same lessons from the same starting point (unless you turn your “fab” into a giant multivariate experiment, which would be very expensive).
What this means is that after one of them has been studied, you now have 11 legacy pieces of equipment that are typically not upgradable and stuck at the same starting line. The 2nd generation of equipment can make the unit (PV or battery) better, cheaper and faster. But for a GW fab you’re carrying all that 1st generation legacy equipment until you can replace them all.
Let’s also look at the supply chain. Conventional wisdom says that buying more of the consumables in bulk is cheaper per unit so GW-size fabs can provide economies of scale. Right? But again, peeking behind the curtain you reveal that there is a legacy issue. For any part of a production process the producer needs to provide the “specs” or specifications of any incoming part, or material. The specs are the best available at the start – but never perfect. The tighter the specs the more expensive the part. As with equipment, there is a learning cycle with specs. So once again the supply chain must be able to change the specs for CIP and if you buy a huge amount for a GW fab, you’re stuck with the legacy specs.
Moving to the other side of the equation, let’s take a peek at the demand side. The most obvious question is: “is there GW demand for the production?” The GW fab makes an assumption that there is demand, but here too there are feedback loops – the sales cycle, the customer’s “specs” (what sells, what works, what doesn’t).
And then of course there the question of price. With legacy equipment and supply chain, the production costs will be higher even as the demand may be only slowly ramping. The GW company is running in the red from costs that are too high costs and income that is too low. The GW fab absolutely must have a sales chain that matures and develops, which also takes some learning cycles.
There’s more to all of this from my experience, but we’re running out of room for this brief commentary, and I can’t rush this learning any faster.
The bottom line is that yes, solar and storage GW fabs can definitely offer economies of scale…but mostly on financial spreadsheets or for very deep-pocketed companies that can withstand the leveraged. In the meantime, the magic is in the CIP cycle. The Wizard of OZ (Giga-Oz) did have his place – but it was for his incremental help to the whole community and his 4 friends for gradual betterment (a heart, a diploma, a medal, and a ride home), not as the magician behind the curtain.
Dr. Skumanich takes an in-depth look at the energy storage industry in his Market Analysis for Energy Storage and Batteries: Forecast to 2030 and to TW levels, which you can purchase here.
Lead image: The Tesla Gigafactory Construction Tour via Flickr/ Steve Jurvetson