API MPMS 13.2
Manual of Petroleum Measurement Standards Chapter 13-Statistical Aspects of Measuring and Sampling Section 2-Methods of Evaluating Meter Proving Data
|Publication Date:||1 November 1994|
Chapter 13.2 will address procedures for evaluating any meter's performance where meter proving factors are devel-oped in accordance with Chapter 12.2. The data in examples used in this chapter are intended to be typical of custody transfer operations of low-vapor-pressure fluids using dis-placement or turbine meters in accordance with Chapters 4, 5, and 6 of the Manual of Petroleum Measurement Standards. However, the procedures in Chapter 13.2 can be used for non-custody transfer metering applications and for custody trans-fer metering of high-vapor-pressure and gaseous fluids where meter proving data are available.
Procedures and examples are given for analyzing the ran-dom uncertainties associated with meter proving data (see note). Procedures and examples are also given for evaluat-ing and controlling trends in meter factors with control charts and control logs to ensure that meter factor variations exhibit a random nature that results in the propagation of a lower average uncertainty in measurements with time and throughput.
Since no single document can cover all of the statistical procedures and applications being practiced, procedures other than those appearing herein may be used. Alternate sta-tistical procedures are not expected to duplicate the compu-tational values provided by procedures in this chapter. However, alternative procedures should achieve the same purpose intended by the procedures given herein. When al-ternative statistical computational procedures are to be used and affect metered quantities, parties directly involved with the metering operations should be notified prior to their implementation.
Some of the procedures in Chapter 13.2 are suitable for hand calculations and graphs employed by field personnel. These procedures are discussed and illustrated in 13.2.5 of this chapter. However, the statistical evaluations and control charts are generally too rigorous for manual field computa-tions, and computers should be employed. The statistical procedures in this document may serve as a guide for devel-oping software for computer applications.