Eight Times Too Fast

The Gas Management System is the most ordinary thing Stanley Meyer ever built, and it is the best.

It exists in two documents that say the same thing: Hydrogen Gas Fuel & Management System For An Internal Combustion Engine, filed on 2 November 1990 and published 14 May 1992, and its granted American twin, US 5,293,857, issued 15 March 1994. That American grant is the last patent Meyer ever received. There is also a bench tape in this archive in which two Meyer brothers walk a camera along the real rack, card by card, with an oscilloscope running.

What it is

An engine management computer for a hydrogen car. Sense the driver and the engine; decide; work four outputs — the cell driver, an air gate, an exhaust gate, and four injectors clocked off a light gate under the distributor cap. Sequential injection, closed-loop resonance tracking, and a full alarm and shutdown tier, in 1990, in a garage.

The argument

Hydrogen burns about eight times faster than petrol. A piston engine cannot use a flame that fast. So dilute the hydrogen with gases that will not burn until the flame slows to petrol's speed.

Every one of those sentences is correct, and the industry reached the same conclusion about twenty years later.

Three things worth checking

His burn-rate table is right. Every row of Table I sits inside the measured laminar burning velocity for that fuel. His hydrogen-to-petrol ratio of eight is really seven point six.

"Exceedingly narrow" sounds wrong — until he defines it. Hydrogen has the widest flammability range of any fuel in common use: 4 to 75 per cent in air against petrol's 1.4 to 7.6. But two paragraphs above the table he defines his own term:

As used herein, the "combustion envelope" refers to the range within which combustion of a fuel gas is possible, given a predetermined quantity of combustible fuel and its ratio to the combustion media

That is a tuning window, not a flammability window — how far the mixture can drift from the point you set before the engine misbehaves. On the rich side, for hydrogen, it genuinely is narrow: rich hydrogen pre-ignites and backfires, which is the symptom he names. And Figure 1 is a plot of exactly that: burn rate against engine speed, hydrogen as a spike, petrol as a broad arch.

He had two knobs, and the patent says so. Running lean gets hydrogen's flame down to petrol's speed by lambda 1.84 on its own. The usual reading is that he refused that road to protect a "sacred" 2:1. His own claims say otherwise:

claim 5: mixing a hydrogen containing fuel gas with at least one of ambient air and exhaust gas

The 2:1 is the ratio inside the fuel gas, fixed by the water — "the fuel gas mixture produced by the fuel cell intrinsically includes the optimum 2:1 ratio". Add any air to that and the whole charge is lean. Figure 3 draws the air path; the bench tape names both gates. The engine as drawn is a lean-burn engine with two diluents: the air gate leans it, the exhaust gate dilutes without leaning.

And read the dilution paragraph to the end. It is usually quoted half way:

excess oxygen insures complete combustion of the hydrogen component; to the extent that oxygen is in excess, it is a non-combustible dilutent. Too much of an oxygen excess may result in the production of undesirable NOx exhaust gases; however, this is not a significant problem and may be resolved … by the introduction of more exhaust gas rather than air as a dilutent.

That is not an apology. It is the lean-burn/NOx trade-off, stated correctly, with EGR chosen as the tool — which is what a calibration engineer does today.

The air processor is a corona ozoniser

The usual dismissal prices total ionisation: 14.9 eV a molecule against 2.96 eV returned by burning one hydrogen molecule, and ions recombining in ~5 µs against a 17 ms trip down the manifold. Both are true and neither is the claim — nobody ionises all the air. At ~10 ppm the energy cost is negligible, and what survives 17 ms is not ions but ozone and excited oxygen, which live minutes. Ozone-seeded lean flames burn faster and hold a lower lean limit; the field is called plasma-assisted combustion. Measurable: an ozone sensor at the manifold and the lean-limit lambda, processor on and off.

What the cell is asked for — as a ledger, not a verdict

a small car at cruise wants	~660 L/min of mixed gas
Faraday's ceiling	10.45 mL/min per ampere
so, at 20 kW	~63,800 A
at a more realistic 12 kW	~38,000 A
Meyer's own claimed yield	115 mL/min/A — 11× Faraday
which would still need	~5,800 A

Meyer's whole claim is that the process is not Faraday electrolysis, so a Faraday calculation cannot convict him — it is the premise he denies. It sets the target. And even his own best number asks for a hundred alternators.

The measurement that settles it is not amps. It is gas energy out against watts into the cell — because "amp leakage" is a claim about what leaks, not a statement of what the supply delivers. In 33 minutes of bench tape, nobody points the camera at a meter.

See Gas yield and Faraday efficiency, Coefficient of performance, Phase-locked loop: lock, hold and damping and Pulse train timing.

What he got right anyway

Flame arrestors written into a patent claim — and named as such in the text ("The quenching tube prevents burn-back or flash-back of hydrogen in its delivery tube"), and drawn nine years earlier in WFC 435 as a "Spark Arresting Gas Line", 1981. (Do not size one from it: his upper bound, 635 µm, is hydrogen-in-air's quenching distance; the certified safe gap is 290 µm, and the 2:1 gas straight off the cell quenches nearer 200.) Exhaust recirculation, argued for on exactly the grounds a combustion engineer would use today — lower peak temperature, less nitrogen oxide. Per-event gas metering rather than per-second flow. A sweep-and-lock resonance tracker with a lock lamp on every one of eleven channels. Related: He Chose the Losses, Amp Leakage.