The story of steam

Steam has stayed the same because it was productized early and productized well. Stark dates the modern boiler to the Babcock and Wilcox combustion boiler patent of 1867, and the significant thing about it was not the combustion but the factory: it replaced systems built brick by brick on site with a mass-manufactured unit, and industry then built itself around that unit. An engineer from 1867 would recognize today’s boiler on sight. The improvements since have been in efficiency and in criteria pollutants, sulfur oxides, nitrogen oxides and particulates, mostly driven by regulation, but the form factor and the business model are the same. Most of what industry needs is saturated steam, meaning steam sitting in equilibrium with liquid water, which is where the very high heat transfer of the phase change is available, and that is delivered generally at 225 degrees C and below. Superheated steam, which is essentially a pure gas, is used higher up and often as a chemical reactant.

The economics are simple and unforgiving. Fuel can be 70% to 90% of the operating cost of a boiler, capital is a small share of the lifetime cost, and boilers run 20 years and often far longer, so the cost of steam in North America is effectively the price of natural gas plus a little. That is why electrification has barely happened despite being available: resistive and electrode boilers are off the shelf today and amount to a one-for-one exchange of the gas price for the electricity price, which in the United States means a two to three times increase in operating cost. Biomass and renewable natural gas carry similar penalties for the same reason. Note what this does to the usual intuition about a sleepy incumbent technology. The boiler is not unchanged because nobody tried; it is unchanged because the thing it burns is extraordinarily cheap and piped to every industrial site in the country.

That sets up the reversal, which is the most useful part of the episode. If the problem is that electricity costs more than gas, the fix is to get more heat out per unit of electricity, which means a heat pump, and the way to raise a heat pump’s coefficient of performance is to start from a warmer source than ambient air. Every facility has warm waste streams doing nothing. Hence thirty years of industrial heat pump projects built around waste heat capture, and hence, Stark argues, thirty years of a market that never scaled. His objection has two parts. The first is capital and repeatability: capture is the expensive step, lower-temperature heat requires larger heat exchangers, and waste heat is mismatched in time, temperature and location relative to steam demand, so every installation is a bespoke engineering project. A manufacturer with twenty plants cannot copy the solution from one site to the next, which forecloses any volume manufacturing benefit. The second part is thermodynamic: waste heat is waste because it is low grade, and as temperature falls the usable fraction of that energy falls with it, so you do more work to extract less. What has actually scaled in energy technology, he argues, is the modular and repeatable, which is the boiler’s own lesson turned against the waste heat approach.

So AtmosZero takes the opposite trade deliberately, and Stark is candid that it is a trade rather than a free win. Going air source means a bigger temperature lift, from ambient to above 100 degrees C, and theoretically a lower coefficient of performance and more capital in the heat pump itself, handled by stacking multiple stages and managing the resulting complexity. What is bought with that efficiency penalty is a standardized, mass-manufacturable product. The remaining question is whether the resulting efficiency is enough to bridge the gap between electricity and gas prices, and here Stark splits the market rather than claiming one answer. Where steam is needed below about 200 degrees C and the facility is a price taker at the end of the wire, a high enough coefficient of performance lets it buy ordinary industrial tariff electricity without having to play the electricity market at all, and that is where he thinks heat pumps win. Where the facility is very large, has time-of-day pricing, a power purchase agreement or behind-the-meter renewables, or needs higher temperatures, he points to thermal storage instead, charging cheaply and discharging heat on demand. Europe is a different calculation again, because gas there is now priced off imported liquefied natural gas rather than Russian pipeline supply, and because energy security has become a reason to electrify independent of the price spread.