AJOPSIS technical note, 2026-10-29 — written by our own engineering team.
Mercury reaches a cement kiln from two directions: the raw meal, which carries trace mercury that is released at calcining temperature, and the fuels, including alternative fuels and waste, which can add considerably more. The kiln system then does something that makes cement harder to monitor than a power plant: the raw mill and the preheater recirculate mercury internally, so the stack does not see a steady fraction of what enters. The raw mill running or stopped changes the picture within minutes, and mercury can be released in episodes rather than as a smooth concentration.
Mercury leaves a kiln as elemental mercury, as oxidised compounds such as mercuric chloride, and bound to dust. Elemental mercury has convenient optical properties and can be measured down to very low levels, which is why Hg⁰ analysers are common. But the emission limit refers to total mercury, and a monitor that sees only Hg⁰ reports low whenever oxidised mercury is present — which is precisely the situation where chlorine is available in the process, and in cement plants the chloride input from raw materials and fuels varies. Total mercury is defined as Hg⁰ plus everything that can be converted to Hg⁰.
The arrangement used by OPSIS keeps the optical analyser where it is and adds a sample path. Hg⁰ is measured across the duct by UV-DOAS, in the same way as other gases. For total mercury, flue gas is drawn through a heated probe and heated line to a converter held at about 200 °C, where bound mercury compounds are released as elemental mercury; the gas then passes through a measuring cell in front of the same analyser. An ejector pump moves the sample, and flow, temperature and pressure are monitored, with oxygen measurement as an option. The converter works without added chemicals, and because the gas never enters the analyser itself, the optical part of the system stays clean.
There are no moving parts in the sample path beyond the ejector, and the converter is reported to run twelve months continuously without measurable loss of efficiency — although efficiency should still be verified, because a converter that degrades silently produces a low total mercury reading with no other symptom. The OPSIS total mercury system has been type-tested to EN 15267 by TÜV, certificate 0000035012_01/20 (2012), as documented by the manufacturer.
For a cement plant that already runs a DOAS analyser for NOx, SO₂, CO, HCl and HF, the step to total mercury is small: a software change where mercury is not yet enabled, plus a sampling system with its measuring cell and multiplexer. No second analyser is needed, and the same instrument continues to report the other components. That matters on a kiln where space on the platform, power and shelter are already tight.
Three things, in order. Converter efficiency, checked at least annually and after any process upset that could poison the catalyst. The heated line, because a cold spot condenses mercury out of the sample and the loss is invisible in the reading. And the raw mill cycle, since mercury behaviour changes with it — a monitoring system that is stable across that cycle is worth more than one with a better detection limit on a test bench.
This note draws on OPSIS product documentation for the total mercury system and on the type-test certificate it cites (EN 15267, TÜV certificate 0000035012_01/20, 2012), together with our own experience of commissioning and servicing mercury systems. It is a technical note, not a substitute for the manufacturer's documentation or for the applicable emission standard. For the Chinese version see 水泥厂汞排放监测.
AJOPSIS technical note, written by our own engineering team, 2026-10-29. Chinese reader? See the Chinese version.
Send the gases, ranges, duct or path dimensions and required certification — we will respond with a technical proposal.