As a filter runs, the floc it captures gradually chokes the passages between the grains of media, and it takes more and more pressure to push water through. That growing resistance is head loss, and watching it is one of the main ways a plant knows when a filter needs cleaning. This guide explains what head loss across a filter bed is, how it rises through a filter run, how SCADA level and pressure instruments compute it, and how terminal head loss serves as one of the triggers, alongside turbidity breakthrough and run time, that initiates a backwash.
Filter Head Loss in one line: Filter head loss is the loss of hydraulic head, essentially the pressure drop, across a filter bed as water passes through it, and it rises steadily during a filter run as captured floc clogs the media and resists flow. It is monitored by measuring the water levels or pressures above and below the bed and taking the difference, which SCADA can compute and trend continuously. When head loss reaches a preset terminal value, it becomes one of the triggers, along with rising filtered-water turbidity and elapsed run time, that ends the run and initiates a backwash.
Head loss is the hydraulic price of pushing water through the media. Even a perfectly clean bed presents some resistance, giving a small clean-bed head loss, because the water has to thread through the narrow, winding gaps between grains. What makes head loss useful as an operating signal is that it does not stay constant: as the filter runs and captures floc, those gaps progressively fill and narrow, the resistance grows, and the head loss climbs. In a gravity filter this shows up as the water having to stand higher above the bed to drive the same flow through the increasingly clogged media.
The rise in head loss over a run is, in effect, a running tally of how much the bed has clogged. Early in the run, with a clean bed, head loss is low and increases slowly. As solids accumulate the head loss climbs faster, and toward the end of the run it can rise steeply as the remaining passages become tight. The shape of that curve tells an experienced operator a great deal about how the filter is loading, and an abnormally fast rise can flag a filter that is being overloaded or a media problem, while a curve that never rises much can hint at short-circuiting or a bed that is not capturing solids as it should.
It is worth distinguishing head loss from filtered-water turbidity, since both are watched but they measure different things. Head loss tells you how clogged the bed is and therefore how much longer it can usefully run before it must be cleaned, while turbidity tells you whether the water coming out is still clean. A filter can have plenty of head-loss margin left yet begin to let turbidity through, or reach its head-loss limit while still producing clear water; the two together give the full picture of the filter's state.
Head loss is fundamentally a difference between two points, so measuring it means measuring the hydraulic head above the bed and below it and subtracting. In a gravity filter the head above the bed is essentially the water level over the media, which a level instrument reads, and the head after the bed is set by the underdrain and outlet arrangement; the difference between the upstream level and the downstream reference is the head loss across the bed. On pressure filters the same idea is applied with pressure transmitters upstream and downstream of the bed, and the differential between them is the head loss.
Bringing these signals into SCADA lets the head loss be computed and trended continuously rather than read spot by spot. A platform such as Merobix can take the level or pressure readings, calculate the differential across each filter, and plot it through the run so operators watch the head loss climb in real time. Because the computation is just a subtraction of measured values, the accuracy of the head-loss figure depends on the level or pressure instruments behind it, which is why those instruments are kept calibrated and their readings sanity-checked; a drifting level sensor will quietly bias the head-loss reading.
Trending head loss across the whole bank of filters is where the monitoring pays off. Seeing every filter's head loss together shows which filter is closest to needing a backwash, helps stagger the backwashes so the plant is not caught with too many filters offline at once, and reveals a filter whose head-loss curve is behaving unusually. Alarms on head loss give notice as a filter approaches its terminal value, so the backwash can be planned rather than forced, and an alarm on an abnormally fast rise can catch overloading before it turns into a turbidity breakthrough.
Every filter is assigned a terminal head loss, the maximum head loss at which the run is ended and the filter is taken out for backwashing. It is set from the available driving head and the point beyond which pushing more flow through the ever-tighter bed is no longer sound, since an over-clogged bed can begin to release captured floc or draw negative pressures within it. Reaching terminal head loss is a clean, physical indication that the bed has held about as much solids as it usefully can, and it is one of the standard reasons a run is stopped.
In practice terminal head loss is one of three triggers watched together, and a run ends on whichever arrives first. The others are filtered-water turbidity approaching a breakthrough threshold, which says the water is starting to suffer regardless of how clogged the bed is, and elapsed run time reaching a set maximum, which caps how long a filter runs even if head loss and turbidity are still fine. Head loss tends to be the trigger when the water is heavily loaded with solids; turbidity breakthrough tends to lead when floc is fine or poorly formed; and the time cap catches lightly loaded filters that would otherwise run indefinitely.
SCADA is what turns these triggers into a smooth operation. The control system watches head loss, turbidity, and run time on every filter continuously and initiates the backwash sequence when any one crosses its limit, so the decision is consistent and immediate rather than dependent on someone noticing. Trending head loss also lets operators anticipate the trigger and coordinate backwashes across the bank in advance. Used this way, terminal head loss stops being just a number in a design manual and becomes a live, monitored condition that keeps each filter cleaned at the right moment to protect finished-water quality.
Head loss is the difference in hydraulic head across the filter bed, so it is measured by reading the head above the bed and below it and subtracting. In a gravity filter the head above the bed is the water level over the media, read by a level instrument, and the difference from the downstream reference is the head loss; on pressure filters, pressure transmitters upstream and downstream give the differential. SCADA takes these readings, computes the differential for each filter, and trends it continuously through the run.
Terminal head loss is the maximum head loss at which a filter run is ended and the filter is taken out for backwashing. It is set from the available driving head and the point beyond which pushing flow through the increasingly clogged bed is no longer sound, since an over-clogged bed can release captured floc or develop negative pressures. Reaching it indicates the bed has held about as much solids as it usefully can, and it is one of the standard triggers that ends a run.
Alongside terminal head loss, a filter run is ended by filtered-water turbidity approaching a breakthrough threshold, which shows the water is starting to suffer regardless of how clogged the bed is, and by elapsed run time reaching a set maximum, which caps how long a filter runs even if head loss and turbidity are still acceptable. A run ends on whichever of the three arrives first, and SCADA watches all of them continuously so the backwash is initiated at the right moment.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.