The filter is not producing a stable fouling signal because its flow is changing. A level controller moves the control valve as intermittent inlet flow changes the 2 m³ tank level; the pumps and valve then move the filter operating point. Since filter differential pressure rises with flow as well as cartridge loading, raw differential pressure cannot indicate cartridge condition unless readings are compared at the same flow.
Moving the valve downstream does not remove this dependency. Measure flow, validate both pressure signals, check pump operation, and then compare or normalize differential pressure at a defined reference flow. Look at the trend first. Tuning does not fix wiring.
What reading should define the problem?
Trend tank level, valve position, line flow, both filter pressures, calculated filter differential pressure, and the running state of each pump on one time base. The installation reports a valve pressure drop of 170–220 kPa near 20% opening and expects filter differential pressure to rise toward 30 kPa over about 12 hours. Those values must not be interpreted without the simultaneous flow.
| Signal | Source | Wrong-value symptom |
|---|---|---|
| Tank level | Level measurement on the 2 m³ pump tank | Noise or pulsing drives unnecessary valve motion |
| Valve position | Level-controller output or valve feedback | Position changes without matching hydraulic response indicate stiction, saturation, or feedback error |
| Line flow | Flow measurement in the common discharge line | Missing or unstable flow data prevents valid comparison of filter loading |
| Upstream pressure | Pressure transmitter before the cartridge filter | Spikes, offsets, or trapped gas corrupt the high-side value |
| Downstream pressure | Pressure transmitter after the cartridge filter | Timing mismatch or offset creates false differential-pressure movement |
| Filter differential pressure | Upstream pressure minus downstream pressure | Rises and falls with flow even when cartridge resistance is unchanged |
| Pump state | Run feedback for each parallel pump | A pump transition appears as a filter-pressure event |
If differential pressure follows flow, continue with flow compensation. If it changes while flow is steady, check pressure measurement first. If flow, valve position, or discharge pressure changes when a pump starts or stops, investigate pump interaction before modifying the control loop.
Does filter pressure move with flow?
At a fixed cartridge condition, differential pressure is a function of flow and liquid properties. Using the stated square-law relationship, doubling flow produces four times the differential pressure, while halving flow produces one quarter. The working normalization is:
ΔP_ref = ΔP_measured × (Q_ref / Q_measured)²
Here, Q_ref is a selected reference flow and Q_measured is the simultaneous measured flow. Use a stable, routinely attainable reference flow that produces a useful transmitter reading. Reject calculations when measured flow is zero or too low for reliable division.
The square-law model assumes an incompressible liquid with sufficiently stable density and no flashing through the valve. Cartridge behavior can depart from a pure square law, particularly as viscosity or flow regime changes. Validate the exponent by collecting differential pressure at several stable flows with cartridges in a known condition. If normalized pressure still changes systematically with flow, derive an empirical filter curve from those points instead of forcing the square-law model.
Are both pressure signals trustworthy?
If differential pressure moves while line flow is steady, hold process adjustments and test the measurement chain. Two independent pressure transmitters can produce a noisy difference when each measures a much larger static pressure than the 30 kPa filter differential of interest. Range, accuracy, zero offset, damping, and scan timing all affect the subtraction.
- Hold flow as steady as the process permits and record both transmitter values at the same sampling time.
- Check each transmitter zero and calibration against a suitable reference.
- Inspect process taps and sensing connections for blockage, leakage, trapped gas, liquid pockets, or elevation-induced head differences.
- Confirm that the two signals use compatible engineering units and scaling.
- Compare the calculated difference with a direct differential-pressure measurement or calibrated test instrument across the filter.
If both absolute pressures rise and fall together but their difference remains stable after synchronized acquisition, the common pressure movement is not filter fouling. If only one signal jumps, repair that channel before changing valve position or controller tuning.
Will moving the control valve after the filter fix it?
No—not by itself. At the same flow and cartridge condition, placing the control valve before or after the filter does not remove the filter’s flow-dependent differential pressure. The valve changes total system resistance, and the pump operating point determines the resulting flow.
A downstream valve raises pressure upstream of the valve, including pressure through the filter. That backpressure changes absolute pressures but does not directly add to the pressure difference across a clean, fixed-resistance filter. It can indirectly change differential pressure by changing flow. Valve relocation also changes the minimum pressure in the circuit, so review pump suction conditions, valve flashing or cavitation risk, filter housing pressure limits, and the valve’s controllable pressure-drop range before modifying piping.
If relocation produces steadier flow because it changes the hydraulic response, the raw trend may look better, but the measurement still requires equal-flow comparison. Treat relocation as a hydraulic design decision, not a filter-monitoring correction.
Are the parallel pumps changing the operating point?
Different impeller sizes do not automatically make two parallel pumps fight, but they can produce unequal head-flow curves and poor load sharing. Both pumps operate at the same common discharge head. The higher-head pump may carry most of the flow, while the lower-head pump contributes little, operates near shutoff, or experiences reverse-flow risk when stopped if isolation devices do not prevent it.
Run a controlled comparison at steady tank conditions: operate pump one alone, pump two alone, and both together. Record common flow, suction pressure, discharge pressure, valve position, and filter differential pressure for each state. Compare the readings with each pump’s manufacturer curve for its installed impeller.
If starting the second pump causes a repeatable step in flow and filter differential pressure, the pressure event is hydraulic rather than cartridge loading. Check pump rotation, impeller identification, speed, discharge isolation, and check-valve operation. Select a compatible operating arrangement before retuning level control.
How should the resolving procedure be verified?
- Repair any pressure-channel calibration, scaling, tap, or timing problem found during the steady-flow test.
- Add a reliable common-line flow measurement if one is not already available.
- Select
Q_reffrom a stable normal operating region and record the clean-cartridge baseline at that flow. - Calculate
ΔP_reffrom synchronized flow and pressure data. Apply a low-flow validity limit rather than allowing the calculation to divide by a value near zero. - Trend raw flow, raw differential pressure, normalized differential pressure, valve position, tank level, and pump states.
- Test each pump state and several stable flow points. Confirm that raw differential pressure moves with flow while normalized differential pressure remains near the clean baseline.
- Use normalized differential pressure for the cartridge-change indication only after repeated clean-to-loaded cycles show that it tracks loading independently of normal flow variation.
A successful correction separates three effects: inlet pulsing appears primarily in tank level and controller action; pump or valve changes appear in flow and absolute pressures; cartridge loading produces a sustained rise in equal-flow or normalized filter differential pressure. Do not adopt the expected 30 kPa over 12 hours as an automatic trip rule until the same criterion is verified for the installed cartridges, liquid, and operating flow.
Frequently Asked Questions
Why does filter differential pressure decrease with a dirty cartridge?
The line flow may have decreased because the level controller closed the valve or the pump operating point changed. Compare differential pressure at the same flow or calculate ΔP_ref before judging cartridge condition.
Why does the differential pressure spike when the second pump starts?
Starting the second parallel pump changes the common flow and discharge operating point. Record pump state and flow with the pressure trend; a repeatable simultaneous step identifies a hydraulic event rather than sudden filter loading.
Why does moving the control valve downstream not stabilize filter differential pressure?
The filter still sees changing flow, and its pressure loss changes with that flow. Valve location changes the absolute pressure profile, but equal-flow comparison is still required.
When should troubleshooting stop and official support be contacted?
Stop adjusting when calibrated instruments disagree, a pump operates outside its manufacturer curve, reverse flow is suspected, or relocation could violate pump, valve, filter-housing, or process pressure limits. Contact the pump, valve, filter, or instrumentation manufacturer through its official support channel with synchronized trends, calibration records, pump curves, cartridge data, and the proposed piping arrangement.