N.Y. Comp. Codes R. & Regs. tit. 16, § 500.5 – Tests of meters.

(a) Tests to be made on meters may be considered in two classes, viz.: routine tests as hereinbefore required and tests required upon complaint of consumers or for other special purposes.
(b) The volume of water passed must be sufficient to cause at least one revolution of the pointer on the initial dial except for sensitivity test at “minimum test flow” rate when the said rate shall be maintained for at least five minutes during which time the pointer on the initial dial shall show a steady uniform movement.
(c) Tests of meters over two inches in size may be made in place on the premises by means of an accurately calibrated test-meter of suitable size or by some other reliable apparatus approved by the commission.
1. (1) All routine tests of displacement type meters shall be conducted by testing for accuracy of registration at not less than two rates of flow within “normal test flow limits,” one rate to be approximately that at which the percentage of accuracy is a maximum and the other at approximately the high rate of flow, or in the case of large meters tested in place on premises, at as high a rate as practicable.
2. (2) All tests made upon complaint of consumers or for other special purposes shall be conducted by testing for accuracy of registration at not less than three rates, namely, at the two rates stipulated above for routine tests and also at approximately the low rate of flow.
  (e) Current type.
  All routine and complaint tests of current meters shall be made within and as near as practicable to the low and high rates given under “normal test flow limits” and, if required, at intermediate rates.
  (f) Compound and fire-service types.
  The accuracy of compound and fire-service meters should be determined by making a sufficient number of tests at different rates of flow between the high and low rates given under “normal test flow limits” to permit the construction of a representative accuracy curve; the tests should be of such scope that the rates of flow at the beginning and end of the “change-over” and the maximum error in registration can readily be determined.
(d) Displacement type.