For most of the last century, electricity was a small share of how the world actually ran. The grid handled the easy stuff — lights, motors, appliances, the household. The hard stuff was burned. Steel was made by burning coal. Cement was made by burning gas. Chemicals were made by cracking petroleum. Ships, planes, and trucks burned distillates. Industrial heat, which is most of industrial energy by volume, came almost entirely from combustion.

About 87% of industrial energy is still fossil. I think this is the most underappreciated number in the energy conversation. People look at the percentage of electricity that's renewable and feel like the transition is well underway, but electricity is only about a fifth of total energy. The other four-fifths is mostly molecules being burned, and the renewable share of that is approximately zero.

The interesting thing about that 87% is that it's not a problem to solve. It's a description of the surface area of the next economy. Everywhere you see "fossil," there's a transition that hasn't happened yet, and in most cases the transition runs through electrochemistry.

The pattern is roughly the same in every sector. Whatever the legacy process burned, the new process uses electricity to do the same chemistry in a more controlled way. Hydrogen used to come from cracking methane; now you can make it from water and electricity. Ammonia used to come from combining methane-derived hydrogen with nitrogen at high pressure; now you can make it from electrolytic hydrogen plus catalytic nitrogen reduction. Industrial heat used to come from burning gas; now you can deliver it electrically, sometimes more efficiently. Steel used to require coke; now there are direct reduction routes that use hydrogen as the reducing agent.

The thing all of these have in common is that they happen inside anelectrochemical cell, and every electrochemical cell has a membrane at the heart of it. So when we talk about industrial electrification, what we're really talking about is a several-decade buildout of electrochemical hardware —fuel cells, electrolyzers, flow batteries, electrolysis reactors, electrodialysis stacks — most of which barely exist at industrial scale today but all of which use the same basic substrate.This is the part that I think investors and policy makers tend to miss. They evaluate hydrogen as a market. They evaluate flow batteries as a market. They evaluate carbon capture as a market. Each market on its own looks uncertain, because each one depends on subsidies or mandates or commodity prices that are hard to predict. Looked at separately, the case for any one of these technologies is shaky.

Looked at together, the case is much stronger. Because the same substrate and key components run through all of them, an improvement to the membrane is an improvement to all of them simultaneously. The same argument can be made for better manufacturing processes for bipolar plates, porous metals, electrochemical catalysts, and stack assembly instrumentation. You don't have to be right about which application wins. You only have to be right that at least one of them does, and that whichever one does will use the same chemistry as the others.

This is the kind of position that's hard to take a deliberate bet on, because most investors are organized around picking applications. Almost nobody is organized around picking the substrate. But the substrate is where the leverage is. Anyone who had bet on planar silicon wafers in the 1980s was, in effect, betting on every chip company that would exist for the next forty years, without having to pick any one of them.

The 87% number is going to come down over the next several decades. The pace will depend on a lot of things — capital costs, electricity prices, policy, customer adoption. But the direction is clear, and the shape of the buildout is clearer than I think most people realize. Whatever combination of trucks, ships, datacenters, chemicals, and grid systems ends up driving the transition, they all converge on the same hardware like proton exchange membranes.

That's the part that makes this interesting as a place to build a company rather than just an interesting trend to watch.

‍