Post Tenebras Lux
←︎ Projects

Industrial Abandonware

We live in a world constrained by energy. A lack of ever larger and incessant flows of energy has always been, and always will be, the most important physical barrier to human flourishing. If energy was drastically cheaper than it is now, as may happen in certain locales over the coming decades, and freely convertible in form, the ultimate commodity cost of a great bulk of polymers, metals, fuels and chemical ingredients would fall as well, decreasing the cost of housing, food, most manufacturing and more.

Historically, the failure to adopt energy-expensive (on a relative basis) but more efficient processes has prevented the development of sustained economic and technological progress: In Greece, the aelopile steam engine. In Rome, watermills. In Eliabethan England, the failure to adopt rolled Coal balls and the general transition from the repricing of wood. In the late 19th century, the failure to adopt early forms of solar energy. The Fischer-Tropsch process and the ongoing reclamation by Terraform Industries. These all touch on the same phenomenon. Throughout history, the relative cost of an energy process has repeatedly halted humanity. As we go forward to a future with cheaper energy in many areas, our ability to see forward in time and see which branches of the tech tree may become, once again, available to us after being lost. This work will be part economic analysis, touching on Jevon's paradox, part Energetic analysis, digging into Government data as well as alternative sources to reconstruct timelines of a variety of industrial processes that may fit the thesis.

What

I'm writing an essay arguing that industrial efficiency is multidimensional, and that energy cost can, and often has, forced us to leave behind a process that wins on many or all other axes — yield, purity, labor, speed, quality. My initial research and calculations split the candidates into two categories. Family A is the thesis as originally stated: processes that have been and remain abandoned because of their own energy cost, displaced by an energy-cheaper route (lead case: calcium-carbide acetylene chemistry). However, I have increasingly come to view with interest Family B: processes that were already energy-light but got abandoned anyway — for speed, scale, or convenience on the back of cheap fossil energy — and which "reclaim themselves" as energy and carbon reprice (sail cargo, lime mortar, natural dyes). This project looks at both Families, and will compare them to the historical cases, hoping to find industrial processes we will soon be able to conduct economically.

Methodology

For each candidate I separate the kill mechanism — thermodynamic (process inherently energy-hungry) vs. economic (a cheap fossil substitute appeared) vs. confounded (regulated out, e.g. Watermills in Rome likely succumbing to the relative price of slave labour, mercury-cell chlor-alkali killed by Minamata poisoning, not energy cost). The x-axis is the relative price of a specific energy carrier, not energy per tonne. Numbers get re-anchored from secondary/vendor sources to neutral primaries (DOE bandwidth studies, EU BREF, Ullmann's, neutral LCA). Some data contridicts itself, so I'll be using 10y moving averages from data sources as feasible. The important calculation will be a sensitivity analysis: for each process, at what relative energy price would it win again, and at what externality cost? Control cases where the low-energy route simply won outright (solar salt, natural trona vs. Solvay).

Possibilities

I am motivated to make the thesis falsifiable per example. If the real reason we abandoned a process was fuel/feedstock substitution or regulation rather than an efficiency reversal, that example fails. I intend to note this for all relevant examples, though if it becomes many this will be listed with less detail. The Family A claim is most at risk of being a just-so story (each case is one substitution event, easy to narrate, hard to generalize); the Family B claim is the one that sensitivity analysis will be performed on, looking at the reclaim argument, so I intend to publish both such that root causes and differences in the nature of processes may elide, with possible strategies for guessing at which industrial processes are more or less likely to be reclaimed, as well as how confident we will be in our predictions.

A parallel look applying the same methodology to historical data, pre-1980 (using 1970s relative price increases as causal mechanism) will look at the strength of the proposed mechanism.

Research notes — candidate ledger

Working notes, not draft prose. Each candidate carries a kill/unlock mechanism and a falsifiability flag. Two orthogonal axes are useful here:

  • A vs. B (as above): killed by its own energy cost (A) vs. energy-light but killed by cheap fossil energy and reclaimable as energy reprices (B).
  • Direction of the repricing that reclaims it: ↓ unlock — a known-but-too-energy-costly process switched on when an energy carrier got cheap (canonical: electricity → aluminium); vs. ↑ reclaim — a low-energy process reclaimed as bulk energy reprices upward (the Family B pattern).

A third timing worth flagging separately from A/B: latent / never-durably-adopted — demonstrated and known, but shelved for want of an enabling input, sometimes oscillating in and out of use rather than being abandoned once. The coal balls are the type specimen.

By era

Rome / Greece.

  • Aeolipile (Hero, 1st c. AD) — ⚠️ weak; sub-1% efficiency, near-zero torque, no high-pressure metallurgy for ~1,600 yrs. Reclaimed by demand + machining in the coal era, not by energy repricing. Use to dismiss, not to lean on.
  • Watermills / Barbegal (2nd c. AD) — real under-deployment, but the "cheap slave labour suppressed it" story is confounded: 1990s re-dating puts Barbegal in the slave-rich early 2nd c., against the labour-shortage prediction. Keep only as a flagged confound (already noted in Methodology).

Tudor / Elizabethan / early Stuart England — strongest historical cluster. Driver = the wood-fuel crisis (charcoal, the iron industry's carrier, repriced sharply upward via deforestation). ↑ pattern throughout.

  • Coke/coal iron smelting — Dud Dudley, patent 1621; technical success, commercial failure (sulfur, scale, monopoly politics); dormant ~80 yrs until Abraham Darby (1709) once charcoal scarcity bit. Clean "known-but-not-deployable → reclaimed as incumbent carrier repriced up + technique matured."
  • Coal-firing of glass, brewing, brick, salt, lime (reverberatory furnace, c. 1610–1640) — enabling re-engineering (separating fuel from charge) let known trades switch carriers once wood got dear. Whole-economy fuel transition.
  • Coal balls / cole-balles — see dedicated note below.
  • Lee's stocking frame (1589) — ❌ known-but-suppressed for labour/political reasons (Elizabeth I refused the patent). A confound; do not count as energy.

Renaissance. Thin for the thesis — its "ahead of its time" artifacts (Leonardo) are paper designs without enabling inputs, not shelved working processes. The real action here is the start of the charcoal→coal pivot above. The gap is itself informative: abandonware needs a working process to abandon.

Modern — the clean ↓ unlock cases.

  • Aluminium — ✅✅ canonical. Deville's chemical route made it a precious metal (~$17/lb, 1859, ≈ silver; Napoleon III's banquet service; ~200 t total world output 1856–1889). Hall–Héroult electrolysis (1886) + cheap hydro (Niagara, 1895) → ~$0.30/lb within a decade. The metal didn't change; a cheap carrier arrived. Sited literally at a dam.
  • Electrochemistry writ large (chlor-alkali, electrolytic H₂) — water electrolysis demonstrated 1800; the whole corpus sat economically dormant ~90 yrs until cheap electricity. The single biggest "depricing reclaimed a known process" event is this, of which aluminium is the dramatic member.
  • Birkeland–Eyde arc nitrogen (1903) — best control/falsifier: viable only on ultra-cheap Norwegian hydro (~60–70 kWh/kg-N), then displaced by Haber–Bosch (~8 MWh/t NH₃), a route with a lower energy floor. Lesson: cheap energy can unlock a wasteful process, but a lower-floor rival still wins. Live flip-back candidate (plasma NOx) as renewable electricity cheapens — good sensitivity-crossover test.
  • Mouchot solar engine (1878) — ↑ mirror: working solar concentrator, gold medal 1878, abandoned when the Cobden–Chevalier treaty + rail/shipping made coal cheap. A known process killed by a rival carrier de-pricing; reclaims itself as fossil reprices up.
  • Fischer–Tropsch (1925) — pure price-toggle; deployed only under embargo/autarky (Nazi Germany, apartheid Sasol), dormant when oil is cheap. (Terraform Industries = the current ↓ reclaim bet as electricity cheapens — already in the intro.)
  • Supporting one-liners: EVs (1900 dominance → cheap oil → battery/carbon reclaim); wood-gas/producer-gas vehicles (WWII, gone when petrol returned); heat pumps (Kelvin 1852, deployment gated by electricity economics).

Taxonomy

| Pattern | Kill/block | Unlock/reclaim | Cases | |---|---|---|---| | ↓ latent until carrier cheapened | too energy-costly | a new cheap carrier (electricity) | aluminium, electrochemistry, Birkeland–Eyde | | ↑ reclaimed as fossil reprices up | cheap fossil undercut it | carbon/energy repricing | Mouchot, sail cargo, coal balls, EVs | | price-toggled (no progress, just switched) | cheap incumbent fuel | incumbent gets dear / cut off | Fischer–Tropsch, producer-gas | | confound — not energy | labour / politics / regulation | — drop from thesis | Lee's stocking frame, (Barbegal's contested cause) |

The coal-ball case (Anton Howes)

Coal-dust + dampened loam/clay, hand-rolled into balls and dried; brought from Liège by Nicolas Romero, promoted by Hugh Plat (published 1603), re-claimed by Richard Gosling (1628), demonstrated again later by the Royal Society and Rumford. Never durably adopted; resurged during fuel crises (e.g. the 1593–1603 Spanish war) and faded whenever coal was cheap and abundant.

Verdict: Family B (↑ reclaim), not a ↓ unlock — do not file next to aluminium. An energy-light process (it reclaims waste fines + cheap loam) squeezed out when bulk fuel is cheap and reappearing whenever fuel reprices up.

Why it earns its place — it answers the memo's own worry that Family A is a just-so story of single substitution events: coal balls are N > 1. The same process crosses the viability threshold and uncrosses it repeatedly across four centuries (1600s → Rumford → Victorian crises → WWII fuel agglomerates → modern biomass pellets). History runs the sensitivity experiment for us; the crossover is observable, not counterfactual.

Two caveats (both flow from the Methodology):

  1. The toggle variable is really the fuel-price-to-labour-price ratio + convenience, not energy price alone — Plat bundled them with veteran make-work, i.e. a labour subsidy hiding inside the "fuel saving." This reads as a confound but is actually the point: it shows the killing axis can be labour, not thermodynamics or fuel substitution, which is exactly the multidimensional-efficiency claim. It forces the x-axis to be more than one carrier's price.
  2. Scope: the Elizabethan version is domestic/poor-relief fuel economy, not heavy industry. Use it as the historical seed of an industrial lineage (coal/peat briquetting → WWII agglomerates → biomass pelletizing), which is industrial and is fuel-scarcity-toggled, rather than overclaiming that hand-rolled 1603 balls were an industrial process.

Sources

  • Aeolipile / Hero — en.wikipedia.org/wiki/Aeolipile
  • Barbegal watermills — Science Advances, PMC6124920
  • Dud Dudley / Tudor-Stuart fuel crisis — britannica.com/biography/Dud-Dudley; englishhistoryauthors.blogspot.com (Tudor and Stuart Energy Crisis: Coal)
  • Aluminium history — en.wikipedia.org/wiki/History_of_aluminium; ACS Hall Process landmark; Science History Institute
  • Birkeland–Eyde — en.wikipedia.org/wiki/Birkeland–Eyde_process; plasma-NOx techno-economics, PMC8133363
  • Mouchot — en.wikipedia.org/wiki/Augustin_Mouchot
  • Fischer–Tropsch — en.wikipedia.org/wiki/Fischer–Tropsch_process; CRS report RL34133
  • Coal balls — Anton Howes, Age of Invention: "All Fired Up" and "How Coal Really Won" (ageofinvention.xyz)