Two modules with the same specification and a 40% price difference get compared on the unit price, and the cheaper one usually wins. Over the life of the installation the unit price is the smallest line in the total. The rest is cells, gas, labour and the cost of the week the instrument stopped working.
The four lines that make up the cost
Purchase price. Paid once. For a fixed detector it is typically 15% to 25% of the five-year total.
Replacement cells. Paid on a schedule set by the chemistry, not by the supplier.
Calibration. Gas, regulator, labour and the paperwork. Paid every time, for as long as the instrument exists.
Unplanned failure. A detector that fails without warning does not cost a cell. It costs a shutdown, an investigation, and in a regulated plant, a report.
Cell lifetime is chemistry, not quality
| Sensing principle | Typical life in clean service | What shortens it |
|---|---|---|
| Electrochemical, H2S / CO | 2 to 3 years | High humidity, continuous low-level exposure, high temperature |
| Electrochemical, O2 | 1 to 2 years | Consumable by design, ages from the day it is first exposed to air |
| Catalytic bead, %LEL | 3 to 5 years | Poisons: silicone, lead, sulphur compounds, halogenated solvents |
| NDIR, CO2 / CH4 | 5 to 10 years | Source ageing, contamination of the optical path |
| PID lamp | 1 year or about 5000 hours | Continuous operation, dirty windows |
Every one of these numbers assumes clean service. A catalytic bead in a foundry, a paint booth or a silicone sealant line can be dead in months. The specification sheet quotes the clean-service figure because that is the only figure that can be quoted.
This is why the question to ask a supplier is not “how long does the cell last” but “under what conditions does that number halve”.
The arithmetic that decides it
Take a fixed detector with one H2S channel, installed where a calibration costs a fixed amount in gas and technician time.
- Cell A costs 1.0 unit, lasts 2 years. Over five years, replacements are needed at year 2 and year 4: 2.0 units of cell cost.
- Cell B costs 1.6 units, lasts 4 years. Over five years: 1.6 units.
Cell B is 60% more expensive to buy and 20% cheaper to own. Add the labour for two extra interventions, both of which require a permit, a ladder and a second person in most plants, and the gap widens. Add the risk that the year-4 replacement falls in the week of a customer audit and the comparison is finished.
The same arithmetic runs on calibration interval. A channel that drifts 1% of reading per month needs a shorter interval than one that drifts 0.2%, and every interval step costs the same gas and the same hour of a technician.
The line that is not on the invoice
Unplanned failure is where the real variance sits, and it is the reason two suppliers quoting the same cell price can produce very different outcomes.
A cell that dies gradually is noticed. Sensitivity falls until the instrument fails its own calibration check, and the replacement gets scheduled. A cell that fails suddenly is a different matter: cracked electrolyte, a poisoned bead, a flooded membrane. The instrument continues to display a number, and the number is zero, and zero looks like clean air.
Two features separate these cases, and neither shows up in a cell price comparison:
Diagnostics that run continuously. A cell whose response is monitored against an internal reference can flag a sensitivity drop before it reaches the alarm point.
A calibration record that is read. A plant that reviews calibration history catches a channel trending down over three intervals. A plant that files the certificates does not.
How to write the comparison
Put five columns in the spreadsheet and fill them for the whole five-year term, not the first year:
- Purchase price per point
- Cell price and expected interval, in the stated service conditions
- Calibration interval and cost per visit, multiplied by visits
- Expected instrument life, since a five-year model and a ten-year model are not comparable at the same price
- A failure provision, which is an estimate and should be labelled as one
Then divide by the number of measurement points and years. That figure is the one that compares, and it will frequently reverse the ranking produced by the purchase price.
Cell lifetimes, calibration intervals and prices vary by sensor, by application and by site conditions. Treat the figures above as a structure for comparison rather than a specification; confirm intervals against the datasheet and your own calibration history.