Wisconsin Administrative Code
Public Service Commission
Chapter PSC 113 - Service Rules For Electrical Utilities
Subchapter IX - Electric Meter Testing Programs and Record Requirements
Section PSC 113.0921 - Statistical sample testing plan for in-service, electronic (solid state) meters and electromechanical, self-contained, single phase and 3-wire network meters

(a) All extended range, surge proof designed meters shall be divided into homogeneous groups based on meter design features and age. The groups shall be further divided into lot sizes categorized by manufacturer, type, serial number, group size or load duty cycle with lot sizes containing a minimum of 26 meters and a maximum of 110,000 meters. The number of lots or lot composition and size may be changed at the end of the sample testing year to allow for increasing or decreasing analysis of accuracy testing requirements on any segment of meters in any lot.

(b) Annually, from each of the assembled lots, a coded sample size specified in Table A-2, Inspection Level IV, page 4 of Military Standard 414, (MIL-STD-414) dated 11 June 1957 and a corresponding actual sample size as shown on Table B-3 page 45, (MIL-STD-414), shall be randomly selected for testing and analysis purposes. Each meter in the lot sample shall be provided with a full load and light load test for accuracy at unity power factor, as specified under s. . A separate statistical analysis shall be performed on each lot sample at each of these 2 load ranges. All meters not registering at either full or light load test points shall be removed from the lot sample and replaced with a different randomly selected meter and not more than two lot sample meters operating outside of the accuracy bandwidth of 95% to 105% at either test load point shall be removed and replaced from the lot sample.

(c) The statistical analysis calculations for both the full and light load accuracy results from the sample lot tests shall be made following the example outlined on page 43 of MIL-STD-414 with the upper and lower specification limits, U and L designated at 102% and 98% respectively. The test criterion for acceptance or rejection of each lot shall be by the Standard Deviation Method, Double Specification Limit with an Acceptable Quality Level (AQL) of 1.00 for the full load analysis and 4.00 for the light load analysis (both normal inspection) as shown on Table B-3, page 45 of MIL-STD-414.

(d) A lot shall be deemed acceptable for continued use if the total estimated percent defective (P) is less than the appropriate maximum allowable percent defective (M) as determined from Table B-3, page 45 of MIL-STD-414, following the procedure of par. (c) for both the full load and light load analysis test points at the respective designated Acceptable Quality Levels. All of the meters in the accepted lot may be retained in use without further accuracy adjustments and will be concluded to have the accuracy characteristics specified in s. PSC 113.0811(1) (c). Meters in the sample lot may be adjusted for acceptable accuracy as required or maintained as necessary and returned to the lot.

(e) A lot shall be deemed unacceptable and rejected for continued use if the total estimated percent defective (P) is greater than or equal to the appropriate maximum allowable percent defective (M) as determined from Table B-3, page 45 MIL-STD-414, following the procedure of par. (c) for both the full load and light load analysis test points at the respective designated Acceptable Quality Levels on any 2 annual sample testing analysis years within a five-year period for the lot or any meters in the lot. All meters in a rejected lot shall be provided with an appropriate test within a period of 48 months from the date of completion of the sample analysis and all the meters tested in the rejected lot shall be adjusted to the accuracies specified in s. PSC 113.0811(1) (c). All meters in a rejected lot that will not be returned to service shall be removed within 48 months from the date of completion of the sample analysis. These meters may be retired without a test. Annual statistical sample testing shall be terminated during the period when all of the meters in a rejected lot are being provided with a test and accuracy adjustment.

(f) All meters in any lot may be tested and adjusted for proper accuracy over a 48-month period at the discretion of the utility without a sample analysis determination specifying the lot test is necessary.

(g) Any meter that is included in this plan, which is removed from service for retirement, may be retired without a test.