ANSI B109.1:2000 pdf free download – DIAPHRAGM-TYPE GAS DISPLACEMENT METERS (Under 500 Cubit Feet Per Hour Capacity)
43.1 Objectives.
The primary purpose of in-service performance testing is to provide service-life information on which the user may base a meter utilization program. nie testing and maintenance procedures, meter design and the level of accuracy specified must be such that a realistic balance exists between the benefits realized from high accuracy levels and the cost of achieving these levels. Any program established should be reviewed periodically with a view toward improvement in light of the current state of the art.
43.2 Test Programs.
Statistical sampling and variable interval plans provide for differences in meters and operating conditions and encourage improvements in meter design and meter maintenance programs. Statistical sampling programs use statistical analysis techniques to monitor the accuracy of meters in service. Variable interval plans also rely on meter test data for the same purpose, but use different methods.
4.3.2.1 Statistical Sampling.
A statistical sampling program for testing gas meters should conform to the general provisions set forth in this section and shall be based on accepted principles of statistical sampling. Meters shall be divided into homogeneous groups such as year set, manufacturer, case type, diaphragm material, etc. However, any group of meters may be combined with another group with similar operating characteristics to provide an adequate sample size for analysis and control. Meters in any given group may be further subdivided according to location. age or other factors that may be disclosed by test data to have an effect on the performance of the meters. Subsequently, groupings may be modified or combined as justified by the performance records. A sample shall be taken each year from each group. The sample taken shall be of sufficient size to demonstrate with reasonable assurance the condition of the group from which the sample is drawn. Meters removed from service on a routine basis may meet the sample requirements.
4.3.2.2 Variable Interval.
A variable interval plan for testing gas meters should conform to the general provisions set forth in this Section. As stated above, meters shall be divided into homogeneous groups such as year set, manufacturer, case type, diaphragm material, etc. However, any group of meters may be combined with another group with similar operating characteristics to provide an adequate sample size for analysis and control. Meters in any given group may be further subdivided according to location, age or other factors that may be disclosed by test data to have an effect on the performance of the meters. Subsequently, groupings may be modified or combined as justified by the performance records. The number of meters to be tested from each group during the current year shall be a percentage of the total original group and be dependent upon the results of the in-service performance tests made during the preceding year, or years, up to a maximum of three. The number or percentage of meters to be tested in each group is a function of the percentage of meters found outside the acceptable percent accuracy limits of 98% and 102% for check rate. The relationship used to determine the number of meters to be tested shall be designed to meet the objectives set forth in section 4.3.1 and shall provide for increasing the number of meters tested with increases in the percentage of meters outside the acceptable limits. An example of a variable interval plan is one in which the formulas may provide for a removal rate of 4.5% when the percentage of meters outside the acceptable percent registration limits equals 20% and for a maximum removal rate of 25% when the percentage of meters outside the acceptable limits equals or exceeds 40%. The minimum number of meters to be removed from each group should be 200 meters or 10% of the total group, whichever is the lesser.