Process Instrumentation and Measurement
Process Instrumentation and Measurement tests selection, installation, calibration and interpretation of field instruments. The exam favours practical loop questions: a transmitter is only correct if its range, accuracy, wiring and process connection suit the application.
Core rules and concepts
- Distinguish accuracy from precision. Error is indicated minus true value; systematic errors include zero and span errors, while random errors affect repeatability. Hysteresis, dead band and drift are separate specifications.
- For a Pt100 RTD, R0 = 100 Ω at 0 °C and α ≈ 0.00385/°C, so sensitivity near 0 °C is about 0.385 Ω/°C. Three- or four-wire connections reduce lead resistance error. Thermocouples generate a Seebeck emf and need cold-junction compensation and correct extension-grade wire.
- A 4-20 mA loop uses a live zero: 4 mA is the lower range value, not zero, and 0 mA indicates an open circuit or power failure. Check loop burden: the supply must exceed the transmitter minimum voltage plus the voltage drop across barriers, indicators and receiver resistance.
- HART superimposes a digital FSK signal on the 4-20 mA current without disturbing the process signal; it supports configuration and diagnostics but is not a high-speed control network.
- Flow measurement follows the technology: orifice plates produce ΔP proportional to Q², so square-root extraction is needed; magnetic flowmeters need conductive liquid; Coriolis meters measure mass directly; vortex meters have a low-flow cutoff.
- DP level measurement must correct for density; radar depends on dielectric properties and installation; capacitance level is affected by build-up.
Worked example
A Pt100 reads 80 °C. Using R = 100(1 + 0.00385 × 80), R = 130.8 Ω. A transmitter scaled 0-200 °C produces 4 + 16 × (80/200) = 10.4 mA. If the true temperature is 82 °C, the indicated error is -2 °C, or -1% of span. State whether that is within the stated tolerance before accepting the instrument.
Traps
Using 0.385 Ω per °C for a 100 °C change instead of 38.5 Ω; ignoring lead resistance; fitting the wrong thermocouple type; treating 4 mA as zero; and ignoring impulse-line or density effects.