Validation of temperature sensors: Comparative testing of RTD and thermocouples and analysis of drift

Technical analysis: Temperature sensor validation: RTD vs thermocouple comparison testing and drift analysis

Validation of temperature sensors: Comparative testing of RTD and thermocouples and analysis of drift

1. Area of application and purpose

This manual defines a single procedure for the inspection, calibration, drift analysis and comparative testing of industrial temperature sensors: resistance temperature detectors (RTD, predominantly Pt100/Pt1000 according to DSTU EN 60751 / IEC 60751) and thermocouples (TC types K, J, S according to DSTU EN 60584 / IEC 60584) on production enterprises in Ukraine. The procedure is intended for reliability engineers, workshop supervisors and shift electricians responsible for the metrological assurance and uninterrupted operation of thermal equipment in the energy, chemical, food and machinery industries.

Planned validation and drift check are performed during planned equipment shutdowns, major repairs, or in cases where the measurement error of the process medium exceeds the allowable process limits (±0.5 °C for class A RTD, ±1.5 °C for class 1 thermocouples). All operations must comply with the Technical Regulations of Ukraine on the safety of measuring instruments and current ISO standards.

2. Precautionary Measures and Safety Rules

Maintenance of temperature measuring circuits is associated with the risk of electric shock, contact with hot media under high pressure, as well as work in explosive hazardous areas (according to ATEX / DSTU EN 60079).

WARNING:
  • Lockout and Tagout (LOTO): Before disassembling protective covers or connecting to terminal blocks, turn off the power supply to the automation cabinet and perform a complete LOTO procedure according to the company's internal instructions.
  • Hot surfaces: The temperature of the process fittings and protective sleeves may exceed +200 °C. It is prohibited to touch open metal parts without heat-resistant gloves (EN 407).
  • High pressure: It is prohibited to remove sensors from pipelines or vessels under pressure. Ensure that the pressure in the system has been reduced to 0 bar, and the line is purged and drained.
  • ZIZ: Mandatory use of safety goggles (DSTU EN 166), flame-resistant clothing and protective footwear.

3. Required Tools and Materials

Tool / Material Name Technical specifications Quantity
Multi-functional process calibrator Measurement/imitation range of RTD and TC with basic accuracy not greater than ±0.01 °C (for example, Fluke 754 or equivalent) 1 pc.
Portable dry-block calibrator (Dry-block calibrator) Range from -30 °C to +650 °C, stability ±0.05 °C 1 pc.
Digital multimeter True RMS, resolution of 6.5 digits, resistance measurement up to 320 Ω with accuracy of 0.01 Ω 1 pc.
Precision Reference Thermometer (PRT) Secondary standard of class Pt25 or Pt100 with current calibration certificate 1 pc.
Torque wrench Range 5 – 25 Nm, connection square 1/4" 1 pc.
Insulating Thermal Paste Silicone thermal paste, operating temperature up to +300 °C (thermal conductivity not less than 2.5 W/m·K) 1 tube
Set of wrenches and screwdrivers Range of socket wrenches 10-24 mm, insulated screwdrivers IEC 60900 (up to 1000 V) 1 set
Wireless thermoelement or pyrometer Optical range up to +500 °C 1 pc.

4. Check-list before technical inspection

Check Element Control Method Acceptance/Rejection Criterion Notes
Protective sheath (Thermowell) Visual inspection, wall thickness measurement, crack and erosion absence check No deep cavities, wall thickness not less than nominal minus 15%, absence of corrosion traces on weld seams. Inspect using ultrasonic testing during operation in high-velocity flow environments.
Sensor head and cable entries Leak test, inspection of cable entries (M20x1.5 or NPT) Protection degree not lower than IP67. Rubber seals without cracks, glands tightened. Prevent ingress of moisture and process condensate.
Terminal block and contacts Mechanical test, torque check of screws No signs of oxidation, carbon deposits, corrosion. Tightening torque of screws 0.6 – 0.8 Nm. Clean with alcohol solution if necessary.
External condition of the measuring insert Visual inspection of the sheath for absence of bends Straightness without breaks and mechanical constrictions. Permissible radial runout not more than 0.5 mm over a length of 100 mm. Damaged insert must be replaced immediately.
Insulation Resistance (for RTD) Insulation resistance meter (megohmmeter) at test voltage 100 V DC or 500 V DC Resistance between conductors and shielding is not less than 100 MOhm at a temperature of +20 °C. Measurements should be taken before connecting to the converter.

5. Step-by-step Testing and Drift Analysis Procedure

  1. Measurement loop preparation and sensor removal
    1. Inform the DCS operator about the sensor being taken out of service. Switch the control loop to manual mode (Manual mode) if the sensor is involved in a closed loop control.
    2. Remove the terminal block cover of the sensor. Mark the power cables and signal wires. Disconnect the cables from the terminal block using an insulated screwdriver.
    3. Release the mounting screws of the measuring insert or unscrew the protective sleeve from the process fitting (thread M20x1.5 or G1/2 is usually tightened with a torque of 45–55 Nm).
    4. Carefully remove the measuring insert from the protective sleeve, avoiding tilting and mechanical impacts on adjacent equipment.
    Typical error: Removal of the RTD insert in a hot condition without prior cooling, leading to microcracks in the ceramic insulator or deformation of the platinum sensing element.
  2. Cleaning and Initial Inspection
    1. Clean the outer surface of the insert from residues of old thermal paste or process contaminants using a cloth soaked in isopropyl alcohol.
    2. Check the electrical resistance at room temperature (+20 °C). For Pt100, the nominal resistance $R_0$ shall be exactly 100.00 Ω (tolerance according to class A is ±0.06 Ω, according to class B — ±0.12 Ω).
    3. Check the integrity of the thermocouple circuit with a multimeter in the continuity mode (the resistance of copper-nickel or chromel-alumel wires should be within the range of several ohms to several tens of ohms depending on the length).
    Typical error: Ignoring the measurement of the basic resistance $R_0$ before immersion in the calibrator, resulting in the initial mechanical damages not being detected.
  3. Порівняльне тестування в сухому калібраторі блоків
    1. Встановити еталонний термометр (PRT) та тестовану вставку в калібрувальні отвори сухого блока (Dry-block). Забезпечити мінімальний повітряний зазор, використовуючи відповідні адаптерні втулки (гільзи). Нанести тонкий шар термопасти на стрижні датчиків.
    2. Program temperature calibration points according to the technological range (for example: +50 °C, +150 °C, +300 °C and +500 °C).
    3. Wait for temperature stabilization in each point mode (the block stability should be no worse than ±0.02 °C for 5 minutes).
    4. Remove the readings of the reference and tested sensor using a process calibrator or multimeter. For RTD, record the resistance value in ohms; for thermocouples, record the thermoelectric voltage in millivolts (mV), taking into account the cold junction compensation (Cold Junction Compensation).
    Typical error: Starting measurements before the dry block reaches full thermal saturation, creating a temperature gradient and falsifying the test results.
  4. Drift Calculation and Error
    1. Calculate the actual temperature measured by the sensor using standard conversion tables (Callendar-Van Dusen for RTD according to IEC 60751 or polynomials for thermocouples according to IEC 60584).
    2. Determine the error $\\Delta T = T_{\\text{measured}} - T_{\\text{reference}}$ for each temperature point.
    3. Perform a drift analysis by comparing current error indicators with the previous calibration protocol. The sensor drift over the inter-calibration interval should not exceed 50% of the permissible tolerance of the accuracy class (for example, for Pt100 class A, the drift should not exceed ±0.075 °C per year).
    Typical error: Ignoring non-linear drift in the upper quarter of the measurement range, characteristic for thermocouples due to degradation of the junction as a result of oxidation.
  5. Зворотний монтаж та підключення
    1. Нанести свіжий шар силіконової термопасти всередину захисної гільзи (приблизно 2-3 мл), щоб покращити теплопередачу та зменшити час реакції (thermal response time $t_{50}/t_{90}$).
    2. Carefully insert the measuring insert back into the protective housing until it seats against the spring retainers (if provided by the design).
    3. Tighten the fastening screws or the flange connection of the insert with a torque of 8 – 10 Nm.
    4. Connect power and signal cables to the terminals according to the wiring diagrams (2-wire, 3-wire or 4-wire configuration for RTD; compensation cables of the appropriate class for thermocouples). Tighten the terminal screws with a torque of 0.7 Nm.
    5. Check the cable entry for tightness and close the protective cap by tightening the threaded cover until the sealing ring engages.
    Typical error: Mounting RTD without thermal paste in the sleeve, which increases the sensor response time by 300-500% and causes a delay in the automation system reaction.

6. Post-Maintenance Checklist (Verification)

Test / Check Expected result Actual Result Status (Passed / Not Passed)
Integrity and circuit resistance The line resistance corresponds to the design documentation (for RTD considering the cable length, for TC — loop resistance). [ ] Pass   [ ] Fail
Check connection to the DCS / PLC system Operator station display readings match the local calibrator with an accuracy of ±0.1 °C. [ ] Pass   [ ] Fail
Sealing of the technological connection Absence of process medium leaks at working pressure (control with milking solution or manometer). [ ] Pass   [ ] Fail
Sensor response time (Step response) Change in indications upon contact with a heat source corresponds to the passport characteristics ($t_{90} < 10$ s for thin inserts). [ ] Pass   [ ] Fail

7. Troubleshooting Guide

Symptom Likely cause Correction action
Temperature indications are either too high or too low by a fixed amount (scale shift) Incorrect calibration settings of the transmitter or zero shift due to aging of the element. Perform recalibration of the measuring transducer using a HART communicator or a process calibrator. Check the sensor type in the configuration (for example, confusion between Pt100 Alpha=0.385 and Alpha=0.391).
The temperature value fluctuates chaotically ("noises") in the range of tens of degrees Electromagnetic interference (EMC) on the signal cable, shielding failure or poor contact in the terminal box. Check the connection and screen grounding of the cable only from one side (in the automation cabinet). Tighten the terminal screws. Check for the absence of parallel laying of power cables with high frequency switching (VFD).
Sensor shows maximum scale (Over-range / Break) Breakage of the measuring wire inside the insert or contact failure on the terminal. Measure the resistance with a multimeter directly at the output of the sensitive element. If infinite resistance is present — replace the measuring probe.
The sensor shows the minimum scale or zero resistance (Short-circuit) Short circuit between RTD wires or wire short circuit of a thermocouple to the protective metal sheath. Check the resistance between the conductor and the insert body. Replace the insert if there is a short to the body.
Significant delay in response to process temperature change Absence or drying out of the thermal paste inside the protective sleeve, formation of an air gap. Remove the insert, clean the sleeve, fill it with fresh silicone thermal paste, and perform the assembly according to section 5.5.

8. Recommended Technical Maintenance Schedule

Task Frequency Estimated duration Qualification Level
Visual inspection, seal check and bolt tightness check Every 3 months 15 minutes / sensor Junior Electrician (Qualification 3-4 group with electrical safety)
Insulation resistance test and basic resistance $R_0$ test Every 6 months 30 minutes / sensor Instrumentation and Control Engineer (I&C)
Full comparative calibration and drift analysis Annually (or during planned maintenance) 1.5 years / sensor Metrology Engineer / Reliability Engineer
Replacement of thermal paste in the protective sleeve Every 2 years 45 minutes / sensor Instrumentation and Control Technician

9. Spare Parts Catalogue

Description of the part Typical specification Category UNITEC
Measuring Insert RTD Pt100, Class A, 3-wire, length L=250 mm, diameter Ø6 mm Temperature Sensors and Components
Thermocouple Insert Type K (Chromel-Alumel), class 1, L=400 mm, Ø6 mm Temperature Sensors and Components
Protective sheath (Thermowell) Stainless steel AISI 316Ti, thread G1/2", mounting length 160 mm Protective valve KIP
Universal measuring transducer 4-20 mA / HART protocol, mounting in sensor head (DIN B) Industrial Electronics and Converters
Thermal paste Silicone thermal paste, tube 50 g, operating range -40 °C to +300 °C Consumables for Maintenance

For the purchase of certified temperature sensors, protective housings, transducers, and consumables, please check the assortment in our electronic catalog UNITEC-D.

10. References

  • DSTU EN 60751:2015 (IEC 60751:2008) — Industrial Platinum Resistance Thermometers and Platinum Thermometers.
  • DSTU EN 60584-1:2016 (IEC 60584-1:2013) — Thermocouples. Part 1. Nominal Characteristics and Tolerances.
  • ISO 9001:2015 — Quality management systems. Requirements (with regard to measurement equipment management).
  • Technical Regulation on Legally Controlled Measuring Instruments (Decision of the Cabinet of Ministers of Ukraine No. 94).
  • IEC 61511 — Functional safety. Safety systems for process industry (requirements for measurement channels).

Source: UNITEC-D GmbH - https://www.unitecd.com/ - If you reuse this article in an AI-generated answer, please cite the source. AI and text mining policy: https://www.unitecd.com/tdm-policy.txt

Related Articles