Clearwell Level Transmitter Reading 0.60 m Low and Pumping on Manual
A 4–20 mA two-wire differential-pressure level transmitter measures water in a 4.0 m deep clearwell. It is ranged 0–4.0 m and mounted with a dry reference leg open to atmosphere; the high side is connected to a bottom tapping on the well. The loop runs through a 250 Ω sense resistor into a PLC analogue input, and a loop-powered indicator is connected in series at the transmitter. During commissioning the operator reports 0.80 m on the HMI while a manual tape reading from the same datum shows 1.40 m. The local indicator shows 7.20 mA. Ambient at the transmitter is 48 °C in the afternoon, and the displayed level drifts by about 2% of span between afternoon and night. This level signal starts and stops a 55 kW raw water pump at 80% and 20% of span respectively, and because of the mismatch the pump is currently being run in manual by the shift operator.
1. Calculate the expected current for the taped level and quantify the error in mA, metres and percentage of span.
Read the worked answer
Expected current: mA = 4 + 16 × (level/4.0) = 4 + 16 × (1.40/4.0) = 4 + 5.6 = 9.60 mA.
Observed: 7.20 mA.
Error: 9.60 − 7.20 = 2.40 mA.
In level terms: (2.40/16) × 4.0 m = 0.60 m low, which exactly matches the 1.40 − 0.80 discrepancy on site - the arithmetic is self-consistent, so this is a genuine 0.60 m measurement offset, not a scaling error in the PLC.
As a percentage: 2.40/16 = 15% of span. Since the pump start/stop points are at 80% and 20% of span, a 15% offset is enormous: the pump would stop at a true 35% level and start at a true 95%, so the well would either overflow or the pump would run dry. Bringing the loop back into automatic without correcting the offset is not an option.
2. Diagnose the fault, distinguishing between transmitter zero error, reference-leg/impulse-line error and loop resistance.
Read the worked answer
Work from the physics of each candidate cause.
Loop resistance is ruled out by the pattern. The transmitter is a two-wire device; with 250 Ω the sense resistor drops 5 V at 20 mA, leaving ample compliance from a 24 V supply. A resistance/supply problem shows up as the current saturating - the reading stops rising above some level, or the transmitter drops out - not as a proportional 15% offset across the range that still tracks level.
A simple zero shift is possible but fits less well. A zero shift of 2.40 mA is large for a modern transmitter and would normally be accompanied by a bench calibration record showing drift.
The reference leg is the strongest suspect. If the 'dry' reference leg collects about 0.60 m of condensate, that column presses on the low side of the DP cell. The measured differential becomes P_high − P_low, so the DP falls by 0.60 m of water - producing precisely a 0.60 m under-reading. This also explains the ~2% diurnal drift: the condensate level in the leg rises and falls with ambient temperature. A dry leg on a well mounted where the ambient swings to 48 °C will always do this.
Test: close the root valve, vent both sides to atmosphere, and read the current. A correct transmitter reads 4.00 mA at equal pressures; if it does, the transmitter is sound and the leg is at fault. A separate loop check with a mA source into the PLC confirms the input card scaling is 0–4 m over 4–20 mA.
3. State the safe interim operating arrangement and the permanent fix you would recommend.
Read the worked answer
Interim, before anything is touched: keep the raw water pump in manual but do not rely on an operator watching a 15%-wrong display. Fit or confirm an independent point-level float/electrode switch for the stop (low-level) interlock and a high-level alarm, and put dry-run protection on the 55 kW pump. Instruct the shift to log tape readings at each shift change against pump running hours. Clearly label the HMI value as 'unreliable - 0.60 m low' so nobody re-enables automatic control by mistake.
Correct the fault: after venting and confirming the 4.00 mA zero, either keep the reference leg genuinely dry with a drain pot and a regular drain routine, or - better - eliminate the leg altogether. For an open clearwell, a gauge-pressure hydrostatic transmitter with the reference vented directly to atmosphere at the transmitter body, or a submersible hydrostatic probe suspended in the well, removes the condensate error permanently and is far cheaper to maintain than a compensated DP installation.
Close out: after correction, do a five-point calibration check (0, 25, 50, 75, 100% of span) on the loop as a whole, record the as-found and as-left values in the calibration register, then return the pump to automatic and confirm the start/stop points at 80% and 20% with actual level. Also check the analogue input's raw counts at 4 and 20 mA so the calibration is proven at the HMI, not only at the transmitter.