Most published high-concentration modules specify ±4 % to ±6 % of reading above 1 %vol. At a 22 %vol trip point, that is a band of ±0.88 to ±1.32 %vol — wide enough that a reading sitting exactly on the trip point tells you very little about which side of it you are on.
This page gives you the arithmetic and a calculator that takes your own threshold. Our own module is in the comparison, and it does not clear the bar either.
First, the safety direction is reversed
If you come from mine environment monitoring, the intuition you have is the wrong one here.
In environment monitoring, lower is safer. The danger is concentration rising toward the lower explosive limit, and thresholds sit around 1 %vol.
In a drainage pipeline, higher is safer. Methane in air is explosive between about 5 and 15 %vol. Drainage gas is deliberately kept above that range. The danger is concentration falling into it.
That reversal is why the critical zone is not 0–5 %vol. It is the band just above the explosive range, where a falling concentration is approaching a state you must not enter.
Two published sources put numbers on this band.
The German Gasabsauge-Richtlinien, in the mining regulations of the Arnsberg district of North Rhine-Westphalia, requires a fixed recording measuring device on the collection main and near the compressor. Below 22 % methane the compressor must shut down automatically. Below 30 % a warning is issued. Below 5 % oxygen, special measures apply.
The New Zealand Royal Commission report on Pike River records the same order of magnitude from a different direction: drainage pipeline gas is not diluted to less than 30 % methane in air, safely above the explosive range.
Which jurisdiction these numbers belong to
These are a German regulation and a New Zealand inquiry finding. Equivalent thresholds for China, Poland and Russia are not established in public sources, and the Chinese figures commonly quoted — 1.0 % alarm, 1.5 % power cut — belong to environment monitoring, not to the drainage line. Use your own project’s threshold. The calculator below takes it as an input for that reason.
What the accuracy figure actually buys you
An accuracy of ±5 % of reading at a displayed value of 22.0 %vol means the true value lies somewhere between 20.90 and 23.10 %vol.
Two different costs sit on the two sides of that band.
A trip that should have happened and did not is a safety exposure: the compressor keeps running while the true concentration has already fallen below the set point.
A trip that should not have happened is lost production: the plant stops on a reading that was low by a margin the specification permits.
The width of the band determines how often you are exposed to each. It is not a specification detail. It is the resolution at which your interlock can actually make a decision.
The arithmetic, worked
The calculation is one line. Error band = displayed reading × accuracy percentage. At 22 %vol with ±5 % of reading: 22 × 0.05 = 1.10 %vol.
At a 22 %vol trip point:
| Accuracy specification | Error band | True value can lie between | Wider than ±0.5 %vol? |
|---|---|---|---|
| ±3.5 % of reading (ours) | ±0.77 %vol | 21.23 – 22.77 | Yes |
| ±4 % of reading | ±0.88 %vol | 21.12 – 22.88 | Yes |
| ±5 % of reading | ±1.10 %vol | 20.90 – 23.10 | Yes |
| ±6 % of reading | ±1.32 %vol | 20.68 – 23.32 | Yes |
At a 30 %vol warning point the same four specifications give ±1.05, ±1.20, ±1.50 and ±1.80 %vol.
If you want the band to stay inside ±0.5 %vol of the trip point, the required accuracy is 0.5 ÷ 22 = ±2.27 % of reading at a 22 % threshold, and 0.5 ÷ 30 = ±1.67 % of reading at 30 %.
Where that ±0.5 %vol comes from
That criterion is ours, not a regulation. No standard we have found specifies how wide an error band an interlock may tolerate. The real allowance comes from your own interlock design and your own tolerance for false trips. We use ±0.5 %vol because it is a round number that keeps the band clear of the set point — and the calculator takes your figure instead if your design says otherwise.
Run it on your own numbers
The band is 2.2× wider than your acceptable band.
Nothing you type here is sent anywhere that identifies you, and this page sets no cookies. The calculation runs in your browser.
Where our own module lands
The Pipeline Insertion TDLAS Methane Measurement Module is specified at ±3.5 % of reading above 1.00 %vol, and ±0.05 %vol below 1.00 %vol. Figures apply at 25 °C, 101.3 kPa absolute, methane in nitrogen balance, after warm-up.
At a 22 %vol trip point that is a band of ±0.77 %vol. It is wider than ±0.5 %vol, the same as the others. Against the best published incumbent figure of ±4 %, it is an improvement of about 12 %. That is not a difference an engineer changes supplier for, and we do not present it as one.
Why we did not write ±2.5 %
The arithmetic above shows ±2.27 % is what would actually clear a 22 % threshold at a ±0.5 %vol tolerance, so ±2.5 % would have read well and would have been close to a real target. We publish ±3.5 % because that is the figure the platform holds today. Tightening a specification after measurement is a safe direction. Publishing the tighter number first and retreating from it is not.
So accuracy is not where we ask you to choose us. It needs to be no worse than what you have, and ±3.5 % of reading meets that. The differences that are real sit elsewhere:
- Pressure envelope. 20 to 200 kPa absolute, compensation range stated separately from the mechanical rating. Published incumbent modules commonly start at 30 kPa, and a Russian type-approved drainage parameter station is registered for 53.3 to 114.7 kPa absolute. A drainage line under deep suction sits below where those modules are specified.
- In-pipe flow-through cell. Some modules described as pipeline type are diffusion cells with an optional clamp fitting. In a line with flow and suction, those two behave differently.
- We tell you whether the band clears your tolerance. We have not found another supplier publishing this calculation. It costs us nothing and it is the reason this page exists.
- Documentation for your certification route. The inputs your type testing needs, prepared against the route your jurisdiction actually uses.