S-S-02—Measurement uncertainty and meter conformity evaluation specifications


Category: Statistical methods
Specification: S-S-02 (rev. 1)
Issue date: 2009-10-28
Effective date: 2009-10-28
Supersedes: S-S-02


Table of contents


1.0 Scope

1.1 This specification establishes requirements for the determination and use of measurement uncertainty for measurement standards and its use in the evaluation of conformity of meters for pattern approval, verification, reverification, or other inspection purposes, including auditing, monitoring, and general enforcement of the legislation.

1.2 This specification additionally provides guidance on the establishment of specification limits for meter performance characteristics.

2.0 Authority

This specification is issued under the authority of sections 12, 18, and 19 of the Electricity and Gas Inspection Regulations.

3.0 Normative references

3.1 Guide to the expression of uncertainty in measurement (GUM):1993 (corrected and reprinted 1995), BIPM/IEC/IFCC/ISO/IUPAC/IUPAP/OIML.

3.2 ISO 10576-1:2003, Statistical methods – Guidelines for the evaluation of conformity with specified requirements – Part 1: General principles.

4.0 Measurement uncertainty requirements

4.1 General

4.1.1 Measurement uncertainty shall be determined in accordance with GUM (reference 3.1).

4.1.2 The determination of measurement uncertainty shall take in consideration such factors as:

  1. influence quantities
  2. test duration
  3. equipment test and display resolution
  4. equipment usage procedures
  5. the relationship of test conditions to those at the operational location

4.1.3 Measurement uncertainty shall be stated in the form of an uncertainty interval.

4.1.4 Annex A below specifies the requirements for rounding observations and functions of observations.

4.2 Certification of measurement standards

4.2.1 All measurement standards intended to be used for the conformity evaluation of meters shall be calibrated by a laboratory formally recognized as being capable of providing dissemination of the measurement unit of interest through an unbroken chain to the relevant nationally-recognized reference standards and be accompanied by statements regarding the measurement uncertainty of each calibration result.

4.2.2 The determination of the measurement uncertainty referred to in 4.2.1 above shall be fully documented, complete with an uncertainty budget analysis identifying Type A and Type B sources and details of all technically-relevant factors considered in its determination.

4.2.3 The certificate for the measurement standard shall include the standard uncertainty and coverage factor used, in addition to the uncertainty interval, for each calibration result.

4.2.4 Full compliance with 4.2.1 and 4.2.2 above is a prerequisite to the certification of a measurement standard.

5.0 Meter conformity requirements

5.1 Specification of requirements

5.1.1 The requirements of section 4 of 10576-1 (reference 3.2) shall apply, where the entity is a meter and the quantifiable characteristics include both quantitative (performance) and qualitative (non-performance) characteristics.

5.1.2 The meter's performance characteristics shall be defined over the range(s) of the meter's design and intended use.

5.1.3 For the purposes of 4.1.3 of 10576-1 (reference 3.2), the procedure shall be documented and authorized by Measurement Canada prior to use.

5.1.4 The establishment of specification limits for meter performance shall take into consideration:

  1. established specifications for meters performing the same function
  2. meter manufacturing process capability
  3. minimum standards for measurement system capability
  4. the effects of differences between the meter's test conditions and usage conditions
  5. the effects of individual meter performance on overall measuring system performance

5.2 Meter conformity evaluation

5.2.1 The requirements of sections 5, 6, and 7 of 10576-1 (reference 3.2) shall apply.

5.2.2 The uncertainty interval of the measurement result used for determining conformity or marginal conformity to the requirements shall be based on a coverage of at least 95 %. The measurement result shall be corrected for all known test equipment or measuring standard biases prior to determining conformity or marginal conformity.

5.2.3 For meter verification and reverification purposes, the target for meter performance conformity is the midpoint of the interval associated with the specification limits, with performance results becoming less acceptable as they depart from target; excessive performance results near specification limits shall be cause for corrective action.

5.2.4 The conformity of a meter's performance quality characteristics shall be determined over the range(s) of meter's design and intended use.

5.2.5 Where the performance test points specified for a meter do not include the endpoints of the range(s) of meter's design and intended use, the conformity evaluation procedure shall address this additional uncertainty so that conformity is assured over the range(s).

5.2.6 Meter performance is considered conforming when the test value is equal to or within the test limits. Alternatively, where the end points of the expanded uncertainty interval associated with a measurement result is not within or equal to the specified limits, the meter is considered to be nonconforming or marginally conforming as the case may be. The applicable meter's verification specification shall specify the use of a one-stage or two-stage conformity test.

Note: Compressed specification limits separate conforming units from marginally conforming units. The test limit is the limit established when the specification limit is adjusted for the associated measurement uncertainty. The limits are defined in the meter's verification specification.

5.2.7 For meter verification and reverification purposes, it is sufficient for the purposes of section 7 of 10576-1 (reference 3.2) to report conformity and nonconformity in alternative forms provided that they are clearly defined in the relevant meter test procedure and convey the same meaning as the statements in reference 3.2.

5.2.8 Quality records necessary to support conformance to this specification shall reflect all details described in this document.

6.0 Revision

The purpose of revision 1 is to:

Annex—A Rounding and significant figures

A.1 Scope

A.1.1 This annex specifies the requirements for rounding observations and functions of observations which are used in the calculation of statistical quantities and in statistical decision-making procedures.

A.1.2 The requirements of this section are applicable for specifying the number of significant digits and for correcting situations where observations and their functions are reported to more figures than are required or justified for the purpose in view.

A.2 Rules for rounding

A.2.1 To round a number to n significant figures, the following rules shall be followed:

  1. when the figure immediately to the right of the n th figure is less than 5, the n th figure shall be kept unchanged;
  2. when the figure immediately to the right of the n th figure is greater than 5, the n th figure shall be increased by 1;
  3. when the figure immediately to the right of the n th figure is equal to 5, and
    1. there are no figures or only zeroes following this 5, the n th figure shall be increased by 1 if odd or kept unchanged if even, or
    2. there are any figures other than zero following this 5, the n th figure shall be increased by 1, whether odd or even.

A.2.2 A number shall always be rounded off in one stage (i.e. one step) to the number of figures that are to be recorded.

A.2.3 For purpose of meter conformity decisions clause A.2.1c) above is modified: when the figure immediately to the right of the n th figure is equal to 5, the n th figure shall be increased by 1.

A.3 Observations and calculations

A.3.1 All observations shall be rounded to the appropriate number of significant figures prior to performing any further processing of them.

A.3.2 For purpose of meter conformity decisions measurement results shall be rounded to the least discrimination as found in either the test value e, or the uncertainty value.

A.3.3 For purpose of meter conformity decisions the meter's verification specification shall state the minimum required measurement discrimination.

A.3.4 All intermediate and final calculations involving observations shall be performed to produce results containing at least six significant figures.

A.4 Presentation of results

A.4.1 Observations shall be presented using the number of significant figures indicated by clause A.3.1 or A.3.2 above and any decision shall be based accordingly on these values.

Date modified: