Section 500.215  Sample Testing of Meters

a)         A utility may, at its option and upon giving notice to the Commission, adopt scientific sample procedures for new and in service meters.  Such procedures shall be supervised by an individual trained in statistical sampling techniques.  During the first five years, or as directed by the Commission, following the adoption of such procedures, the utility shall file the results of the sample testing technique as part of the Annual Report – Form 21, on page 541.  At least once a year a chi-square test must be made to verify the randomness of the sample.

b)         Meter lots for new meters must be established consisting of meters of a single type and size, manufactured under the same conditions, and at essentially the same time.  All sample testing procedures must be in accordance with Inspection Level II of Military Standard 105-D of the Department of Defense (also see Supply and Logistics Handbook H-105).

c)         Meter lots for in service meters must be established consisting of meters of a similar type, size and year of installation.  In the ninth and subsequent years that meters are in service, they shall be tested in accordance with Inspection Level II of MIL-STD 105-D of the Department of Defense (see also Supply and Logistics Handbook H-105).

d)         In order to comply with the two percent slow or one percent fast accuracy limits of Section 500.190, the plan adopted for new meters must provide an acceptable quality level not to exceed 1.0% in order to assure a process average of at least 99%.  New meters must be 100% tested by the manufacturer. The test records must accompany the meters and will be considered as the initial test.

e)         The plan adopted for meters in service must provide an acceptable quality level of 6.5% in order to assure a process average of at least 93.5%.  A meter will be defined as deviant if the average of its check-rate and capacity-rate accuracy tests is not between three percent and three percent slow.