仪表校验

Sophie, PT-1203 came back from the field for its annual calibration. It's a pressure transmitter, range 0-40 bar, accuracy requirement ±0.25% of span. The as-found data doesn't look good. What did you find?

I did a five-point calibration — 0%, 25%, 50%, 75%, 100% — upscale and downscale. The as-found errors are: At 0 bar (0%): -0.05 bar At 10 bar (25%): -0.08 bar At 20 bar (50%): -0.15 bar At 30 bar (75%): -0.22 bar At 40 bar (100%): -0.18 bar The maximum error is -0.22 bar at 75%. The allowed tolerance is ±0.25% × 40 bar = ±0.10 bar. This transmitter is out of tolerance — OOT — by more than double the limit.

(reviewing the data) The error is always negative — the transmitter is consistently reading low. And it's not linear — the error increases from 0% to 75% then drops slightly. That's not a simple zero shift. Sophie, any idea what could cause this pattern?

When I see a negative error that increases with pressure, my first suspect is the sensing diaphragm. It could be coated or corroded — that adds stiffness, so the diaphragm deflects less for a given pressure. Less deflection means lower output signal, which means lower reading. Let me check the process history for this transmitter.

I can tell you — PT-1203 is on the chemical dosing line. It measures the discharge pressure of a sulfuric acid dosing pump. Acid service. If the diaphragm material is standard 316L stainless steel, sulfuric acid will attack it over time.

That's a systematic issue. We need to check all transmitters on acid services — any that don't have Hastelloy or Tantalum diaphragms need to be replaced. We'll issue a preventive NCR. Sophie, for this transmitter, it's beyond repair — tag it as "failed, replace with Hastelloy C-276 diaphragm model." And update the piping material specification so this doesn't happen again.

(at the field junction box, connecting HART communicator) Linda, I'm at FT-2401 — the natural gas flow transmitter. Operations says the DCS reading is fluctuating by ±3% when the flow should be steady. I'm going to do a field calibration check. Let me connect the HART communicator and the pressure calibrator together.

Good. Do an as-found check first — don't adjust anything. I want to see the raw data. Also, check the loop current with your multimeter in series — sometimes the problem is in the loop wiring, not the transmitter.

OK, applying 0% pressure — HART reading 3.998 mA. Spec says 4.000 mA ±0.016 mA. This is borderline. At 50%: 11.998 mA. At 100%: 19.997 mA. The transmitter output on HART looks accurate. But here's the thing — when I measure with the multimeter in series, at 50% I'm getting a fluctuating reading between 11.85 and 12.10 mA. That's a 0.25 mA oscillation. The transmitter's HART output is stable at 11.998 mA — the problem is NOT in the transmitter!

So the transmitter is fine but the loop current is unstable. That points to a loop issue — possibly a loose terminal, a failing I/O card, or a ground loop. Kevin, can you check the DCS analog input card for FT-2401? Look for noise on the channel.

(checking DCS diagnostics remotely) The AI card for FT-2401 is showing a "noise detected" warning — the channel is picking up 50 Hz interference. That's mains frequency noise. Classic ground loop issue — the shield of the instrument cable is probably grounded at both ends instead of just one.

Let me check the shield termination. (opens junction box) There it is — the shield drain wire is connected to the ground terminal here at the field end AND it's also grounded at the marshalling cabinet end. That creates a ground loop — the shield acts like an antenna picking up noise. I'll disconnect the shield at the field end. Per our standard, shield should only be grounded at the control system end.

(disconnects shield, retests) Multimeter reading now — 12.001 mA, steady. The fluctuation is gone. Kevin, how does the DCS channel look now?

Noise warning cleared. Reading stable at 50.0% flow. Problem solved. Good catch, Mike. Another ground loop bites the dust!