MonitoringUpdated 6 November 2026

What happens to the trend when a filter restarts?

A restart leaves a time signature. Comparing the turbidity trace with the filter log helps operators work out what the PTV 1000 was measuring as the filter came back online.

By Fowzie Smith-Tahir LinkedIn

Two technicians walking beside a reservoir

A filter has just been washed. Its turbidity trace rises, settles and then lifts again. The operator coming onto shift sees two peaks, but only one entry in the log, marked “backwash complete”. The trace gives a precise time for each reading at the analyser. Working out what was happening at the filter at either moment takes a little more information.

Give each change in the trace a place in the restart

A restart usually spans several changes in operating state. Depending on the plant, there may be a rest period, a gradual increase in flow or a period of filtering to waste before water is directed onward. Recording those transitions separately lets a reviewer ask whether an excursion occurred at first flow, during a valve change or after the filter had been running steadily.

Start with the physical source of the sample. A PTV 1000 drawing from an individual outlet can show how that filter responds. On a combined header, the reading represents the mixture reaching the sample tap and can be dominated by other filters already online. If the question is whether one bed is taking longer to settle than it used to, an individual outlet gives the comparison more meaning. See our guide to filter outlet monitoring for the implications of each sampling point.

Put the process and analyser on the same timeline

Water takes time to move from the sample tap through the sample line and into the measuring chamber. A flow change or interruption can alter the delay. During commissioning, agree how the site will establish a representative travel time and record the conditions under which it was observed. Use that observation to interpret the timeline and keep the original instrument timestamps in the record.

Imagine a rise recorded just after a valve opens. If the sample line still held water from the preceding stage, the rise could represent an earlier event. Looking at the sample flow and the plant event log together makes that possibility visible. Confirm that the analyser and historian clocks agree and keep any sample-flow or instrument status beside the trace. The installation guidance explains why the route and drainage matter to the measurement.

Compare restarts under similar conditions

Bring up a previous restart from the same filter and put the traces on a common time scale. Look at the time it takes to return to the filter’s usual running profile as well as the height of the first peak. A narrow peak during a repeatable transition tells a different story from a reading that stays elevated once normal flow and sampling have resumed. Keep the filter’s operating conditions with each record so the comparison makes sense across shifts.

When the shape changes, check what else changed on that run. Air entering the sample path, a disturbance to sample flow or recent work on the measuring chamber can all change what the analyser sees. If a reference sample is taken, record where and when it was collected, which method was used and whether it represented the same water. The US EPA turbidity guidance manual discusses filter ripening and restart practices in section 5.7. Use your own plant procedure for decisions about returning a filter to service.

Leave a record the next shift can use

A useful handover might say that the first rise appeared while the filter was still discharging to waste, that the sample flow was steady and that the trace returned to its familiar profile before the later valve change. If the timing is uncertain, say which event or sample condition still needs checking. This makes the record actionable when another operator sees the same pattern on the next wash.

Once a team records restarts consistently, a longer settling time or an excursion during normal flow is easier to pick out. Keep the trace, filter identifier and event times together. That gives the next reviewer enough context to decide where to investigate, from the filter and its valves to the sample line and the instrument itself.

Let’s talk about your application.

Tell us which outlet you want to monitor and what you see when the filter returns to service.

Advice for your application. Discuss your sample point, expected range and the PTV 1000 optical options.
Practical service support. Discuss commissioning, calibration, maintenance or an instrument issue with our team.
Clear next steps. Share the site conditions and what you need to resolve, so PMA can direct your enquiry to the right team.
An enquiry, not a mailing list. Your details are used to respond to you. You won’t be added to a marketing database.