An online analyser is most useful when an unusual value leads to a considered action. Without an agreed response, an alarm can create two unhelpful extremes: it is ignored because it has caused noise before, or it is treated as definitive proof before the sample, method and process have been checked.
The better approach is to agree an investigation plan before commissioning. That plan defines what the measurement is for, which trend conditions deserve attention, who checks the sample route, when a process check or confirmatory sample is needed and how the outcome is recorded.
This is topical in the UK water sector. The Drinking Water Inspectorate’s 2025 report asks companies to use monitoring data to identify deteriorating trends and manage risks proactively. The Environment Agency’s current business plan includes increased monitoring and inspections at wastewater treatment works. These priorities support better operational monitoring discipline; they do not change a site’s permit, statutory sampling method or laboratory obligations.
Define the purpose before setting an alarm
An alarm threshold is not automatically a regulatory limit, a process setpoint or a 3S specification. It is a local operational decision. Start by writing the purpose of the measurement: trending a source change, supporting a treatment-stage check, protecting downstream equipment, guiding an investigation or another defined task.
Then distinguish three kinds of alert that may need different responses:
- a result moving away from the expected operating trend;
- a sudden step change that needs a process or sample-system check; and
- an instrument or sample-system condition that could make the result unreliable.
The response should reflect the defined purpose. An operational trend may prompt a check of plant status and an additional sample; it should not be described as a compliance failure unless an applicable approved programme explicitly says so.
Build a first-response sequence
The first response should be simple enough to use under pressure. A good starting sequence is:
- Check whether the trend is persistent, rather than one isolated cycle.
- Check instrument status, alarms, reagent level and programmed routine history.
- Check sample flow, take-off condition, reservoir, filtration and drain route.
- Review process changes, dosing, flows, source changes and recent maintenance.
- Take a confirmatory sample or use the site’s agreed comparison route where appropriate.
- Record the event, checks performed, result and decision.
The exact sequence must be adapted to the site and parameter. It should make clear who owns each action, the expected response time and the escalation route if the value remains unexpected.
Separate analyser routines from sample-system reality
The supplied 3S-CL brochure lists automatic calibration, validation and cleaning as standard features, with intervals that can be selected. It also states that reagent level is monitored, and that calibration data, alarms and analysis results are stored in the integrated data logger with USB storage. These capabilities can support a defined operating routine.
They do not prove that the water at the take-off is representative, that an external filter is clear or that a laboratory result is comparable. The older supplied user manual provides useful operational detail but is dated 2017; use the current product brochure as the authority for current product claims, and check the installed configuration and current service documentation before creating site instructions.
The automatic calibration, validation and cleaning guide explains the role of those routines. Use it alongside, rather than instead of, a sample-system and process investigation plan.
Decide what to record
Record enough information that a later reviewer can understand what happened. At minimum, include the time, displayed value and units, relevant trend period, analyser status, routine checks, sample-system observations, process conditions, any confirmatory-sample details and the action taken.
The 3S-CL brochure says results and alarm data can be stored and exported from the data logger. Whether and how that data is incorporated into a site record remains a site decision. It may help to establish a short, repeatable event form rather than rely on free-text notes.
Use laboratory results carefully
A laboratory result can be a valuable part of an investigation, but the comparison must be designed. Check sample time, sample location, preparation, analytical method, units and reporting basis. A disagreement does not automatically prove that either result is wrong.
Read the online water analyser versus laboratory results article before setting a comparison procedure. It provides the complementary question: are the two results actually measuring comparable water in a comparable way?
Make escalation proportionate
Not every alert needs the same response. A site may define an initial operational check, a supervisor review, a process specialist review and a regulated-response route. Build these with the relevant site owner, permit holder and laboratory arrangements. Do not publish universal alarm values or generic compliance statements because they depend on the parameter, process, site controls and applicable regulatory framework.
This is particularly important for wastewater: the Environment Agency’s guidance explains that permits can specify monitoring and reporting requirements, and total nitrogen is a defined aggregate measure that is not automatically supplied by one nitrate or nitrite trend. Keep operational alarms and regulated results clearly labelled.
A commissioning checklist for alarm readiness
Before live use, confirm that the team has:
- an agreed measurement objective and named owner;
- a documented normal range or trend expectation, reviewed for the site;
- clear alarms for analyser/sample status as well as measured value;
- a first-response checklist and out-of-hours route;
- defined inspection points for sample, filtration, drainage and reagents;
- a planned laboratory or manual comparison method where appropriate;
- a record format for investigations and corrective actions; and
- current PMA support details and access arrangements.
This does not require complicated software. It requires agreement that turns data into an appropriate operational response.
Contact PMA about commissioning and support
If a 3S application is being planned or recommissioned, discuss the measurement objective, sample conditions and required operating routines with PMA. Start with the 3S online colorimetric analyser, the automatic-routines guide and 3S Support. Contact PMA to arrange an application discussion.
Frequently asked questions
Is an online analyser alarm a compliance alarm?
Not automatically. The alarm should be labelled according to its site-defined operational purpose. Compliance monitoring is governed by the applicable permit, approved methods and reporting arrangements.
Do automatic calibration and cleaning remove the need for checks?
No. They support analyser operation. Teams must still check the sample route, process context and comparison evidence when investigating an unusual result.
What should happen after a one-off unexpected result?
Use the site’s agreed first-response plan: inspect the trend and status, check the sample system and process conditions, then take any appropriate confirmatory action. Avoid concluding from a single number alone.