DP Flow Transmitter That Walks With the Time of Day
A 6-in. horizontal natural gas line at 250 psig uses an orifice plate for flow measurement. The taps are at the 3 and 9 o'clock positions. The differential-pressure transmitter (0-100 inH2O, 4-20 mA) is mounted on a bracket 3 ft below the taps, and both impulse lines slope continuously downward from the taps to the transmitter. No drain pots and no vent valves were installed. Within a week of commissioning the indicated flow reads about 15% low compared with the custody transfer meter, and the error changes with ambient temperature (40 F at night, 95 F in the afternoon). An operator wants to simply re-zero the transmitter to fix the reading.
1. Diagnose what is happening in the impulse lines and explain why the reading is low and temperature dependent.
Read the worked answer
This is an installation defect, not a transmitter fault.
In gas service the impulse lines must stay free of liquid. With the transmitter 3 ft below the taps and both lines sloping down to it, every low point and the transmitter body itself becomes a liquid trap. Hydrocarbon and water drop out of the gas, collect in the legs, and change the head each side of the cell sees.
A liquid column standing on the low-pressure leg adds head that subtracts from the measured differential, so the transmitter reads low - consistent with the observed 15%.
As ambient temperature swings from 40 F to 95 F, the volume and level of that trapped liquid change, so the bias changes with it. That is why the indicated flow walks with the time of day rather than drifting randomly.
Note the direction of the error depends on which leg accumulates more liquid. The fact that it reads low points at the low-pressure (downstream) leg, but the underlying cause is the same: the lines were not built to drain.
2. Give the correct installation arrangement for gas service and describe what must be done with this installation.
Read the worked answer
Correct gas-service arrangement:
- Mount the transmitter above the orifice taps.
- Slope both impulse lines at least 1 in/ft continuously downward from the transmitter to the taps - no low pockets anywhere.
- Install drain/vent valves at the low points and at the transmitter so liquid can be blown out.
- Keep the side taps at 3 and 9 o'clock as they are.
For this installation: relocate the transmitter above the orifice and re-make the impulse lines, then blow them down and re-verify against the custody meter.
Do not substitute the wet-leg approach - deliberately filling the legs with a known fluid and calibrating in elevation - because that is appropriate for a true liquid-level application where the fill is controlled, not for a gas line where the amount of dropout is unpredictable and changes with conditions.
3. Describe the correct way to check the transmitter zero with a three-valve manifold, and explain why the operator's re-zero is a bad idea.
Read the worked answer
Live zero check with a 3-valve manifold:
- Close the high-pressure block valve.
- Close the low-pressure block valve.
- Open the equalizer so both sides of the cell see the same pressure.
- Open the manifold vent/drain to bleed off trapped liquid or gas, then close it.
- Read the output - it should be 4.00 mA at zero differential.
- Restore in reverse: close the equalizer first, then reopen the low side, then the high side.
Never open the equalizer with both block valves open - that bypasses the orifice and produces a false flow indication.
Why the operator's re-zero is wrong: the error is a head bias, not a zero offset. Trimming the zero to make today's reading match the custody meter would:
- bury a real installation defect that will keep getting worse as the legs fill,
- introduce a new error at every other flow rate, because the bias does not scale with flow,
- possibly leave the lines partially blocked so the signal becomes sluggish or locks up entirely.
Fix the installation, then calibrate.