#780 1939 · Don A. Sillers and Alexander Clarke (Peerless Manufacturing) · Natural gas utilities / industrial safety
Adding a smell to gas wasn't the hard part — keeping the smell proportional as the flow changed was, and that's what actually got patented
the problem
Odorized gas still escaped detection whenever pipeline flow rate changed
background
On March 18, 1937, a leak of odorless natural gas accumulated undetected inside the New London, Texas, consolidated school until a spark in the industrial-arts shop triggered an explosion that killed 294 students and staff — still the deadliest school disaster in US history. Investigators traced the cause directly to the gas's complete lack of smell, and the Texas legislature responded almost immediately, mandating that all natural gas sold for domestic or industrial use be artificially odorized so a leak could be detected by smell before it could explode.
The obvious response — add a foul-smelling chemical to the gas stream — was already being tried, using devices that simply let a pungent liquid evaporate into the passing gas. That approach worked passably at a steady, constant flow, but real pipelines never run at constant flow: demand rises and falls through the day, and pressure fluctuates with it. Evaporation-based odorizers dosed gas at whatever rate the liquid happened to evaporate, disconnected from how much gas was actually flowing past at that moment — meaning the exact conditions that were also most likely to accompany a real leak, surges and pressure swings, were also the conditions where the odorizer was least reliable.
what everyone would do
The direct response to an odorless-gas disaster was already underway before Sillers and Clarke's patent: add a foul-smelling chemical to the gas stream so a leak could be smelled before it exploded, using devices that simply let a pungent liquid evaporate into the passing gas. That fix worked at a steady flow, but real pipelines never run at constant flow — demand and pressure shift throughout the day — and an evaporation-based odorizer had no way to track how much gas was actually passing at any given moment, so its dosing drifted exactly when conditions were least steady.
what they saw
Sillers and Clarke saw that the actual engineering problem wasn't "how do we make gas smell bad," it was "how do we guarantee the smell stays proportional to the gas no matter how the flow changes" — and evaporation could never answer that second question, because it responds to temperature and surface area, not to gas volume. Building the odorizer's dosing pump around an actual flow meter meant the two quantities that had to track each other, gas volume and odorant volume, were mechanically coupled to the same physical measurement, instead of one being measured and the other merely hoped to keep pace.
the move
Just months after the New London disaster, Peerless Manufacturing founder Don Sillers and Alexander Clarke patented a fundamentally different approach: instead of letting odorant evaporate passively, they built the dosing mechanism around an actual gas meter, using the metered flow itself to mechanically drive a dipper-type pump that injected odorant in direct proportion to the gas passing through — so the amount of odorant added rose and fell in lockstep with the volume of gas, not with an unrelated evaporation rate.
why it works
Because the dipper-type pump was mechanically driven by the same metered gas flow it was dosing, the volume of odorant injected rose and fell in direct proportion to the volume of gas passing through the line, holding steady across flow-rate and pressure swings of up to 50 percent — precisely the fluctuating conditions that had made evaporation-based odorizers marginally ineffective. That reliability is what let a single installed unit run for decades, in one case 25 years without maintenance, and let natural gas build a safety reputation that odorization by evaporation alone had never earned.
the payoff
Because dosing was mechanically tied to metered flow rather than to evaporation, the Peerless Type M odorizer stayed proportional across a wide range of flow rates and pressure changes — up to 50 percent variation — the exact conditions that had defeated earlier evaporation-based designs. One installed unit, Serial No. 2105, ran continuously from 1942 to 1992, including a stretch of 25 years without maintenance, and by the end of World War II natural gas had earned a reputation as a fundamentally safe fuel largely because of odorizers built on this principle.
where it breaks
The mechanism only holds its guarantee within the flow and pressure range the meter and pump were built to handle — outside that range, or if the meter itself fails or is bypassed, the coupling that makes dosing proportional breaks down and the system silently reverts to the same unreliability evaporation-based odorizers had. It also only solves detection, not prevention: an odorized leak still requires someone nearby to notice the smell and act on it, which is why odorization has always been paired with, not a substitute for, pipeline inspection and leak-detection programs.
what came after
The American Society of Mechanical Engineers designated the Type M odorizer a National Historic Mechanical Engineering Landmark in 1992, and metered, flow-proportional odorization — rather than passive evaporation — remains the underlying engineering principle behind natural gas safety odorization systems used worldwide today.
references
- [1]Meter-type Gas Odorizer (Engineering Landmark #163)American Society of Mechanical Engineers, 1992asme.org
- [2]Peerless Manufacturing Type "M" OdorizerBullock Texas State History Museum, 2017thestoryoftexas.com