TE-LP-003—Laboratory practice for the selection, calibration and use of direct reading thermometers

Version: 7

Record of changes
Version Date Description
2 Added "controlled copy" statement in the document header of the original release (2006-06-06).
3 Reworked procedure in sections 1.2 and 2.1, tables 3, 2.5, 2.8, 2.10, 3.2, 3.4 and 3.6.
4 2009-07-08 Added new ambient and in service temperature requirement under 2.2 and 3.3 (changed from (22 ± 5) °C to (23 ± 5) °C).
5 2015-10-28
  • Revised tolerance structure, new calibration tolerance and specifications for stability and effect due to environmental conditions.
  • Carried out a final review, combining content from the current tolerances and field calibration requirements, made editorial corrections and fixed equations.
  • Added option to include other sources of uncertainty in uncertainty calculation (needed for MC primary and secondary thermometers).
  • Changed recommended range for liquid heated product applications up to 120 °C instead of 125 °C.
6 2016-10-18 Corrected the temperature symbols and updated the format
7 2017-07-18 Removed requirement to include uncertainty sources on certificates in section 3.3.8.

Table of contents


1 Introduction

1.1 Purpose

The purpose of this document is to provide recommended practice for the selection, calibration and use of direct reading digital thermometers that will be used for calibration and inspection work conducted by Measurement Canada (MC) staff or accredited or registered organizations. The document also serves as a framework and guideline for the technical evaluation of organizations providing thermometer calibration data to MC.

1.2 Scope

This recommended practice applies to direct reading electronic digital thermometers used for calibration and inspection work under the Weights and Measures Act and Regulations and the Electricity and Gas Inspection Act and Regulations.

1.3 Applicable documents

1.4 Abbreviations and symbols

MC: Measurement Canada

2 Specifications and requirements

2.1 Accuracy and measurement uncertainty requirements

The criteria for thermometer certification are specified in the Weights and Measures Regulations as an accuracy tolerance of ± 0.25 °C for temperatures up to 50 °C and ± 0.6 °C for temperatures above 50 °C.

To ensure the above tolerances are met, the following uncertainty components are applied:

  1. Maximum calibration uncertainty
    The maximum allowed calibration expanded uncertainty (k = 2), for the purpose of thermometer certification, is 0.15 °C. The thermometer calibration measurement uncertainty is determined at the time of calibration by the laboratory performing the calibration of the thermometer. The calculation of the calibration measurement uncertainty is described in section 3.3.
  2. Thermometer model performance specifications
    Thermometer model performance specifications are detailed in section 2.3. This section addresses the sources of measurement uncertainty that are not included in the calibration uncertainty estimate. These include the thermometer stability and susceptibility to ambient conditions. Only those thermometer models meeting these criteria are eligible for certification.

2.2 Operating conditions and resolution

The thermometer's operating conditions class will be rated as per Table 1.

Table 1: Classification of digital thermometer operating conditions
Class Application
Laboratory conditions 23 °C ± 5 °C (indoors)
Field conditions −20 °C to 40 °C (outdoors)

The recommended resolution and minimum measurement range of thermometers is dependent on the intended application (see Table 2).

Table 2: Recommended ranges and resolutions for thermometers
Application Minimum range and min. resolution
General field inspection −30 °C to 50 °C × 0.01 °C
Liquid heated product inspection −30 °C to 120 °C × 0.01 °C
Calibration of thermometers used in a gas prover for temperature correction 0 °C to 35 °C × 0.01 °C
Temperature correction in gas measuring apparatus 0 °C to 35 °C × 0.01 °C
Bell prover air temperature(s) 0 °C to 35 °C × 0.01 °C

2.3 Performance specifications

2.3.1 Thermometer stability

The thermometer must be designed to be stable to within ± 0.1 °C for the duration of the verification period.

2.3.2 Variation due to ambient conditions

The thermometer measurement variations due to ambient conditions, for the rated temperature range, must be within ± 0.1 °C. This performance characteristic may be derived from:

  1. measurement and tests by MC
  2. manufacturer's data and compliance statement
  3. second party measurement and tests

2.3.3 Probe design and construction

The temperature sensor must be designed and constructed to operate without leakage or deformation over the expected range of operating pressures, temperatures and environmental conditions, including the immersion depth of the sensor in liquid when required.

Where thermometers will be used to determine flowing temperature stability such as in the testing of automatic temperature compensation (ATC), the time constant of the thermometer in flowing liquid must be less than 15 seconds. The recommended probe diameter is 3 mm (or 1/8 inch) and the recommended lead wire length is approximately 6 feet.

2.3.4 Resistance to thermal shock

Thermometer probes intended for field applications must be able to withstand thermal shocks of 50 °C without loss of accuracy.

2.3.5 Field thermometer packaging requirements

Thermometers intended for field use must be equipped with a carrying case which is water resistant, dustproof and dimensionally stable over the range of environmental conditions expected during the service life of the device.

2.3.6 Construction

The construction must be mechanically and electrically sound. The materials, finish, etc., must be of a quality to provide assurance of long life and sustained accuracy.

2.3.7 Battery power

When power is supplied by a battery, the thermometer must be equipped with a low power indication.

2.3.8 Markings

The following information must be indelibly marked on, or be capable of being displayed by, the instrument:

  1. manufacturer's name
  2. model number
  3. serial number

In the case of thermometers with multiple sensors, each sensor must be marked by a unique identifier.

2.3.9 Provisions for sealing

The thermometer design must prevent inadvertent adjustment of the thermometer.

3 Calibration

Calibration includes the following operations:

  1. visual inspection of the thermometer
  2. adjustment and calibration
  3. obtaining As Found and As Left results
  4. calculating the calibration measurement uncertainty
  5. producing the calibration certificate

3.1 General requirements for the reference thermometer and calibration equipment

3.1.1 Reference thermometer

Traceability

When used for the calibration and certification of thermometers as a local standard, the calibrator's reference thermometer must be traceable to a reference standard listed in the Weights and Measures Act and Regulations.

When used for the calibration and certification of thermometers as measuring apparatus under the Electricity and Gas Inspection Act and Regulations, the calibrator's reference thermometer must be traceable to a national reference standard maintained by National Research Council of Canada.

Measuring range

The reference thermometer must not be used outside its certified measuring range.

Monitoring

The reference thermometer must be monitored for stability with an ice bath, water triple point cell or other suitable methods with similar accuracy and stability. Monitoring must be conducted on a monthly basis as a minimum.

3.1.2 Calibration bath

A calibration bath of suitable operating range with demonstrated stability and homogeneity of ± 0.02 °C or better throughout this range must be used.

3.2 Calibration conditions and method

The documented calibration process will ensure the following.

Confirm that the thermometer model meets the specification outlined in section 2.

The thermometer is inspected and the inspection results as well as the seal condition are recorded.

Before beginning the calibration procedure, the thermometer under calibration and the measurement standard is preconditioned at ambient conditions for at least 12 hours. Ambient temperatures in the calibration laboratory are maintained within 23 °C ± 5 °C.

The thermometer sensors are sufficiently immersed in the temperature bath to address immersion errors. If this cannot be achieved, immersion error is considered in the estimate of the uncertainty.

Temperature of the temperature bath is allowed to reach the set point and stabilized before readings are taken.

As Found results, at a minimum of two temperature points, are obtained prior to adjustments. It is recommended that the ice point be included.

The thermometer is adjusted as per the manufacturer's instructions or other suitable method which will minimize the thermometer measurement error.

In the case of thermometers that cannot display measured resistance and are adjusted based on measurement errors, a suitable deviation function is selected for the thermometer. The deviation function is the equation used to model the expected thermometer corrections over the calibration range. Often a linear correction deviation function will be suitable when the thermometer error can be characterized by a slope and offset. Second and third order equations can also be used for the deviation function where the thermometer error is not expected to be linear.

Each transducer associated with the thermometer is calibrated at a minimum of six (6) evenly distributed test points across its measuring range or, as a minimum, the range given in Table 2. Where a nonlinear deviation function or calibration equation is used to model the thermometer response, a minimum of three (3) test points per equation variable is used.

The coefficients of the deviation function or calibration equation, for each sensor, are determined by the least squares fit.

In the case of thermometers that are to be certified for use with corrections, the deviation function is used to generate the correction table for the sensor.

If the thermometer is adjusted, an "As Left" ice point test is conducted.

3.3 Calibration measurement uncertainty calculations

Uncertainty calculations must be based on the recommendations of the Guide to the Expression of Uncertainty in Measurement (GUM). Uncertainty values will be reported at 95 % confidence interval or with a value of k = 2.

The measurement uncertainty of the thermometer calibration results must include the sources of uncertainty described in sections 3.3.1 to 3.3.5.

3.3.1 Uncertainty of the measurements by the reference thermometer

The sources of uncertainty for the reference thermometer will include the uncertainty due to:

  1. the reference thermometer calibration (ureference)
  2. stability of the reference thermometer (ustability)
  3. any other sources that will contribute to the uncertainty of the temperature measurement by the reference thermometer such as:
    • immersion error of the reference thermometer (uimmersion)
    • hysteresis of the reference thermometer (uhysteresis)
    • self-heating of the reference thermometer (uself heating)

3.3.2 Uncertainty due to the stability and uniformity of the temperature bath

This uncertainty component is obtained by measurement of the temperature bath's stability and uniformity or from the manufacturer's specifications.

3.3.3 Uncertainty due to the linearity and repeatability of the thermometer under calibration

This uncertainty component is estimated by calculating the standard deviation of residuals for the determined deviation function or the calibration equation.

3.3.4 Uncertainty due to thermometer resolution

The standard uncertainty due to thermometer resolution is:

Equation 1 (the long description is located below the image)

Description

The standard uncertainty due to thermometer resolution is equal to the minimum graduation size divided by the square root of 12.

3.3.5 Thermometer bias error

This section applies to thermometers that have a resolution ≤ 0.01 °C and are to be used without a correction table. The maximum bias error, Errorbias, is taken as the maximum measured error over the measuring range. The maximum bias error must be less than ±0.05 °C.

3.3.6 Combined calibration measurement uncertainty

The combined calibration measurement uncertainty of the thermometer under calibration is calculated as follows:

Equation 2 (the long description is located below the image)

Description

The combined calibration measurement uncertainty is equal to the square root of the sum of squares of the uncertainties of the reference thermometer, the stability and uniformity of the temperature bath, the linearity and repeatability of the thermometer and thermometer resolution.

Where:

3.3.7 Expanded calibration measurement uncertainty

Uncertainty values will be reported with an expansion factor of k = 2, which correspond to a 95 % confidence interval as follows:

Equation 3 (the long description is located below the image)

Description

The expanded uncertainty is equal to two times the combined uncertainty.

In the case of thermometers where the reading will be used without the application of a correction, the thermometer expanded measurement uncertainty will be estimated as follows:

Equation 4 (the long description is located below the image)

Description

The expanded calibration measurement uncertainty is equal to two times the combined uncertainty plus the thermometer bias error.

3.3.8 Expanded thermometer measurement uncertainty

The estimated thermometer measurement uncertainty may be reported in the calibration certificate instead of the calibration measurement uncertainty when the relevant sources of measurement uncertainty estimates are known. This approach is recommended for MC calibration laboratory primary and secondary thermometers.

The thermometer measurement uncertainty is obtained by the root sum square of the calibration measurement uncertainty; see section 3.3.6, and the uncertainty contributions of the relevant uncertainty sources such as the thermometer stability and susceptibility to environmental conditions.

The expanded thermometer measurement uncertainty values will be reported with an expansion factor of k equal to 2, which corresponds to a 95% confidence interval.

3.4 Calibration pass/fail

The thermometer must not be certified if the expanded (k = 2) calibration measurement uncertainty is greater than 0.15 °C.

The thermometer must not be certified if the thermometer indication error, at one or more calibration points, exceeds ± 0.75 °C.

3.5 Certification period

Unless otherwise authorized, thermometers must be calibrated at least once every two years.

3.6 Sealing

Thermometers that have physical access to any means of adjustment must be protected by a seal to indicate any adjustments that can affect the thermometer calibration.

The values of the programmable parameters that can affect measurement accuracy of the thermometers must be included on the calibration certificate.

3.7 Calibration status indicator

The thermometer must be labelled to indicate the status of calibration by the application of a verification mark or tag indicating:

  1. the expiry date of the calibration certificate, and
  2. unique identifier of the laboratory which performed the calibration.

3.8 Certificate of calibration

The following information must be contained in the certificate of calibration for the instrument:

  1. name of the instrument's manufacturer
  2. name and address of the owner of the instrument
  3. a description and unambiguous identification of the instrument
  4. date of the calibration
  5. range of the calibration
  6. estimate of calibration measurement uncertainty
  7. traceability of the thermometer standard
  8. correction table (as applicable)
  9. name and address of the laboratory which performed the calibration
  10. any special limitations for the use of the instrument
  11. the expiry date of the verification certificate
  12. title: calibration certificate
  13. certificate number
  14. page identifier
  15. identification of the calibration method used
  16. the signatory's name and signature
  17. the value of programmable parameters that can affect measurement accuracy of the thermometers, such as Callendar-Van Dusen coefficients or other
  18. when the thermometer is received with a broken seal, a note to this effect

3.9 Request for the designation of a local standard by accredited laboratories

A request for the designation of a local standard by an accredited laboratory must be made with the form provided in the Appendix.

4 Application and use for inspection duty

The thermometer must not be used outside its calibrated temperature range.

The thermometer must not be used outside the rated ambient temperature for which it is certified.

The owner is responsible to ensure the thermometer is suitably rated for the hazardous location classification as required.

Unless otherwise specified in this document, the thermometer must be used as per the manufacturer's recommendations.

The thermometer must be monitored by performing ice-point verification on a monthly basis. The ice point verification compares the thermometer ice point reading to that of the "As Left" ice point value recorded on the certificate of calibration. If this differs by more than ± 0.1 °C, the thermometer cannot be used for inspection purposes and must be re-certified.

Temperature measurement errors due to non-ideal location of the probe and dynamic effects will be taken into consideration when reviewing the suitability of a thermometer for a measurement application.

Appendix: Designation of local standard request

Thermometer model Space to insert thermometer model

Thermometer serial number Space to insert thermometer serial number

Conditions of seals when received Space to insert conditions of seals when received

Calibration report/certificate #Space to insert calibration report/certificate

This is to request that the above thermometer be designated as a local standard. The thermometer has been calibrated and meets the following criteria:

  1. The calibration is traceable to a reference standard listed in the Weights and Measures Act and Regulations.
  2. The suitability of the thermometer model is verified by Measurement Canada's thermometry laboratory or the thermometer has been previously calibrated by Measurement Canada.
  3. The calibration measurement uncertainty is less than or equal to 0.15 °C.
  4. Accesses to the thermometer adjustments have been sealed and the values of programmable parameters that can affect measurement accuracy of the thermometer are recorded on the calibration certificate.
  5. The "As Left" ice point is recorded on the calibration certificate.
  6. The "As Left" thermometer error before application of correction is within ± 0.75 °C.

Signed by Space to insert signature

Accredited laboratory

Date modified: