Calibration of Temperatures Sensors.
A vital but often misunderstood procedure within the food
Presented by N.E.M Business Solutions
The Pt100 platinum resistance thermometer is by far the most common sensor in the food industry, and yet its inner workings and how to maintain it are often a mystery to technical staff.
Platinum resistance thermometers (PRTs) offer excellent accuracy over a wide temperature range (from -200 to 850 C). Standard Sensors are are available from many manufacturers with various accuracy specifications and numerous packaging options to suit most applications. Unlike thermocouples, it is not necessary to use special cables to connect to the sensor.
The principle of operation is to measure the resistance of a platinum element. The most common type (PT100) has a resistance of 100 ohms at 0 C and 138.4 ohms at 100 C. There are also PT1000 sensors that have a resistance of 25 ohms and 1000 ohms respectively at 0 C.
Sensor manufacturers offer a wide range of sensors that comply with BS1904 class B (DIN 43760): these sensors offer an accuracy of ±0.3 C at 0 C. For increased accuracy, BS1904 class A (±0.15 C) or tenth-DIN sensors (±0.03 C).
Related standards are IEC751 and JISC1604-1989. IEC751 also defines the colour coding for PRT sensor cables:
Often food manufacturers use an
"outside" contractor to carry out calibrations and
have no idea how these calibrations are carried out or what is actually calibrated.
Within the food industry temperature control is
The following procedure is
kind permission of Geest Prepared Foods
and the Author David McLucas, Engineering Manager.
This document represents the type of information that should be standard throughout the industry.
Temperature Equipment Calibration Procedure
This paper describes in detail the methods used on the Geest Sutton Bridge site to calibrate the temperature controller and the associated three wire Pt100 Resistance Thermometer Detector, herein referred to as RTD's. This method is used for all RTD's on site and is commonly referred to as simply the "temperature probe calibration".
This process has been documented to increase the knowledge base of the Sutton Bridge personnel not actively involved in the calibration process. This will facilitate a better understanding of the process and increase the awareness of the detail of what is involved in the calibration procedure. It will also demonstrate that the current procedure checks and confirms the accuracy and reliability of the RTD's, their associated controllers and the interconnecting wiring.
Note; the term temperature controller refers to the equipment that the RTD connects to, i.e. the PLC RTD input card (analogue or digital) and the specific method of displaying the temperature. On the Geest Sutton Bridge site the temperature readings are displayed on SCADA screens that are located throughout the facility.
The calibration procedure is a fully documented and certified process where all the necessary information associated with each calibration is recorded on the Calibration Certificate. A copy of the Calibration Certificate is retained by the Site Engineering Department and by the calibration sub contractor. This form also has a space to specify any corrective actions that may be identified as necessary during the calibration procedure. The calibration process also acts as an independent planned maintenance check that supplements the routine monthly in-house calibrations.
The sub contractor, (following their own in house ISO 9001:2000 procedures that are certified by Lloyds Register), undertakes the annual site calibration process for the temperature equipment using master equipment which is independently calibrated by external sources accredited to UKAS.
A single temperature measuring system comprises of: -
The RTD itself, · the interconnecting wiring, · the PLC RTD input card (located in the control cabinet), · the site computer network connection and · the output temperature display shown on the SCADA screen, i.e. the controller.
2. Summary Explanation of Values referred to in Text
Value A. Selected resistance value, in ohms, injected into system. Value B. Conversion of value A to a temperature value using the calibration chart. Value C. SCADA screen temperature value.
Value D. Measured resistance reading from the RTD sensor, in ohms. Value E. Conversion of value D to a temperature value using the calibration chart. Value F. Calibrated thermocouple measuring the same heat source as the RTD being calibrated.
3. Pictures/Diagrams of a Three Wire Pt100 RTD
PHOTOGRAPH OF HEAD WIRING IN A TYPICAL THREE WIRE Pt100 RTD (showing the two red and one white wires)
4. Three Wire Pt100 RTD's
A three wire Pt100 RTD is a Resistance Thermometer Detector (RTD) which has three wires (there are two and four wire options available) that are connected to a Platinum (Pt) resistance element. At a temperature of 0oC the resistance of the element is 100 Ohms.
The resistance element in such a device is a piece of almost pure Platinum that has been subjected to a controlled manufacturing process. This is then insulated and enclosed in a protective sheath. The three wires connected to the resistance element are also enclosed in the protective sheath but are extended beyond the end of the sheath for easy connection to the field wiring.
It is not the purpose of this paper to explain the method of operation of RTD's or the advantages in using them. Sufficient to state that such devices are widely used for temperature measurement. Their suitability is based on the fact that the resistance that an electrical conductor (e.g. the Platinum element) exhibits to the flow of an electric current is related to its temperature.
Various wiring configurations are possible with RTD's however a three-wire RTD is a standard arrangement used widely in this type of application. They have a relatively simple wiring arrangement and provide accurate readings, even with long wiring runs between the various parts of the system. In addition a three-wire lead will resolve any problems caused by any effect of the temperature range on the wiring itself.
Various classes of Pt100 RTD's are available. The classes distinguish the accuracy of the RTD. Class B RTD's are the type used on the Sutton Bridge site.
At 0oC they are accurate to within + 0.3oC · At 100oC they are accurate to within + 0.8oC
Other classifications are available with an improved tolerance, e.g. a class 1/10 instrument is accurate to + 0.03oC at 0oC and + 0.12oC at 100oC.
A similar range of classifications is available with Pt1000 probes. These are the same as the Pt100 range except for the fact that at 0oC their resistance is 1,000 ohms. These are used in applications where a high resistive load is required.
5. Temperature Controller (SCADA) Calibration
As can be seen in the two illustrations included in Section 3 there are three wires connecting the controller (or temperature display) to the RTD. The first step of the calibration process is to disconnect the three controller wires from the RTD. The two controller wires that were connected to the two red wires in the RTD head are twisted together to form a single cable connection. This joining can be done at the probe head or at the resistance source. (Note; the third wire was connected to the white wire in the RTD head).
A calibrated resistance source, typically a Fluke meter, is then connected between the two twisted wires (red wire connection in the RTD head) and the remaining single wire from the controller (white wire connection in the RTD head). See diagram below.
A specific resistance value is selected (value A) on the Fluke meter and is injected into the controller circuit. Any resistance value selected (a parameter expressed in ohms) has a known equivalent temperature (value B) that can be obtained by reading the appropriate value from the calibration chart. This action provides an accurate "reference" temperature (value B).
NOTE: - The calibration chart referred to above is the resistance/temperature relationship for Platinum resistance thermometer detectors. This is the published data set of the ITS-90 scale, i.e. the International Temperature Scale of 1990, which was adopted on the 1st January 1990. This international standard defines the temperature/resistance values from minus 200oC up to 850oC for Pt100 RTD's.
This "reference" temperature (value B) is then compared with the temperature reading that is checked and recorded at the same time and is shown on the SCADA screen (value C). The SCADA screen reading must be within the allowable tolerance range, +/-1 degree of the reading obtained from the calibration chart, for the calibration process to continue.
If the reading is within tolerance then this process is repeated two more times to give three readings across the working range of the instrument. For CIP temperature instruments the resistances selected are normally 100.00, 119.40 and 138.50 Ohms. The temperature equivalent values are 0, 50 and 100oC respectively.
A more appropriate range is used for chillers and freezers, typically: -
Normal calibration temperature values for a freezer operating at minus 18oC using a 3 point calibration would be minus 25oC, minus 10oC and 5oC. and · Normal calibration temperature values for a chiller operating at 5oC using a 3-point calibration would be minus 10oC, 0oC and 10oC.
If any reading, or check, is outside the allowable tolerance at any stage of the process the calibration procedure is terminated.
If the calibration process is terminated then it may be possible to adjust the controller software to bring the reading back within tolerance. This process is equivalent to altering and adjusting a potentiometer on a non-PLC controlled instrument. If there is any adjustment of the controller then the entire calibration process is restarted.
In the unusual event of this type of adjustment not resolving the error a new calibrated controller input card would be purchased, installed and re-calibrated in situ. The in situ calibration ensures that the controller is fit for purpose after installation and has not been damaged in any way after the manufacturer's test and certification procedure was completed.
When completed successfully the above calibration process checks and confirms the accuracy of the controller. It also confirms that the wiring between the RTD and the controller, including any PLC cards, is problem free.
The controller readings obtained during the current calibration process are then compared with the previous calibration readings to check and confirm the year on year performance of the controller. This direct comparison can be undertaken because it is possible to inject the same, precise, known, accurate resistance values into the controller during the calibration process each year. Since the controller displays the injected resistance value as a temperature value then the controller temperatures displayed during the calibration process should be the same, within the allowable tolerance, each year.
Thus it can be understood how this check provides confirmation of the ongoing reliability of the controller.
6. RTD Calibration
The RTD probe itself is calibrated immediately after the controller calibration has been successfully completed. At this point in time the RTD and the controller are still electrically disconnected. A calibrated resistance meter (again, typically a Fluke meter) is connected across the twisted red wires and the white wire in the RTD head, i.e. a similar wiring arrangement to that described and shown in the previous section. In this instance though, the meter value, which is recorded, (value D) is a measured resistance value, expressed in ohms. This value is created by the temperature that the disconnected RTD element is experiencing at that precise moment.
The same calibration chart referred to previously is again used to determine the actual temperature (value E) that the RTD is experiencing; i.e. value D, converted from ohms to an equivalent temperature value expressed in degrees Centigrade.
A separate independent calibrated thermometer, normally with a digital readout, is then used to measure the temperature that the RTD being calibrated is measuring. The use of the separate independent calibrated temperature thermometer provides an accurate "reference" temperature (value F) for subsequent comparison with the RTD reading.
This "reference" temperature can be obtained in a number of ways. For example, two typical methods used are; locating the calibrated thermometer adjacent to the RTD; or by placing it into the outlet of an open sampling port. Whichever method is used it is important that the two temperature instruments are subjected to the same heat load simultaneously and that the readings are obtained simultaneously.
The temperature (value E) obtained by measuring the resistance value of the RTD and converting it to degrees Centigrade (value D to E) is then compared with the "reference" temperature read from the separate independent calibrated thermometer (value F).
If there is a difference in the values outside the allowable tolerance range, +/-1 degree, then the RTD is discarded and replaced with a new calibrated RTD. RTD's are always discarded if they fail their calibration check because they have no adjustment capability.
Whenever a new RTD is fitted it is immediately re-calibrated in situ. The in situ calibration ensures that the RTD is fit for purpose after installation and has not been damaged in any way after the manufacturer's test and certification procedure was completed.
In applications that have an easily accessible and a relatively safe medium to measure, e.g. an open vessel heating water, then the RTD will be calibrated across the normal operating temperature range. However in other applications where the medium may be hazardous or where the temperature is maintained at a relatively constant temperature, e.g. a CIP caustic rinse recovery tank, then the RTD is calibrated at only one temperature as this reflects the normal operating condition of the system.
If the RTD reading is within the allowable tolerance range then the RTD calibration is complete.
This procedure checks the accuracy of the RTD. The fact that the RTD has remained within tolerance year on year confirms the reliability of the RTD.
There is no year on year calibration comparison of the RTD readings (as there is with the controller readings) because the temperature reading being measured can vary at the time of checking. For example the actual ambient temperature of an empty tank or the temperature of water in a recirculating system subject to steam heating via a heat exchanger cannot be guaranteed to be the same each year unless rigorous control measures are introduced with the sole purpose of achieving this aim.
7. Stability of RTD's
Platinum (Pt) RTD's are highly stable measuring devices. During manufacture detectors are heat treated to homogenize the crystal structure and remove any oxides. Consequently, typical drift values for a Pt100 sensor is 0.05°C per annum. If the temperature range is confined to 25 - 150°C, as on the Sutton Bridge site, the drift can be as low as 0.005°C a year.
Experience has shown that RTD's rarely drift out of tolerance. They fail completely and suddenly by going open circuit or by short-circuiting. An open circuit failure is confirmed by a reading at the upper end of the instrument range or by an error message on the controller. A short circuit failure is confirmed by a reading at the lower end of the instrument range or by an error message on the controller.
There are no moving parts in a Pt 100 RTD. It is basically a piece of platinum (which is the resistance element) which has the wires connected to it, all encased in an insulated metal sheath. This simplicity of design and the control of the manufacturing process are reasons for the reliability of the probe.
The method of calibration described above means that it is not always necessary or usual to remove the RTD from its fitting/location for calibration. This is advantageous because it minimises the risk of damage to the RTD during the calibration process.
In situ calibration also eliminates any potential errors associated with variations of the insertion length of the probe stem between its operational and calibration locations (if it had to be removed for calibration). RTD's are "stem sensitive" instruments, unlike thermocouples which are tip sensitive. In order to obtain accurate readings with an RTD at least 25mm of the stem should be exposed to the heat source being measured. Variations in the insertion length can result in the probe reading being affected because of slight differences in the heating effect of the environment around the probe.
Other types of RTD sensors are available, other classifications and different resistance materials can be used. However the Sutton Bridge site has standardised on Pt 100 RTD's, this is a typical selection choice for the type of application and the accuracy required.
Since the site start up in January 1999 only one controller has been identified as faulty and only two RTD's have failed (one of which had a wire twisted off of the terminal as a result of removal and replacement). All items were immediately replaced when the fault was discovered.
This low failure rate is further confirmation of the reliability of the temperature controllers and RTD's.
There are 26 separate RTD's on site and each one is calibrated annually by an external subcontractor. This is in addition to the routine monthly in-house calibration checks completed.
Geest Sutton Bridge
The following table shows the resistance in Ohms at 2°C Intervals:
|Deg C||PT100||Deg C||PT100||Deg C||PT100||Deg C||PT100|
N.E.M Business Solutions Tel / Fax : 01823 680119 Mobile 07768 981196