Ieee std 1159-1995, ieee recommended Practice for Monitoring Electric Power Quality

Annex A
(informative) 
Calibration and self testing
A.1 Introduction
Electrical measurements and the ancillary Þeld of meter calibration are two aspects of the same industry. As
new advances are made in measurement instrument technology, new calibration technology is demanded to
keep these instruments at peak performance and maintain their traceability to national standards.
Calibration requirements should be based upon monitoring objectives and the nature of loads, not on the spe-
ciÞc instrument used. We should ask ourselves, ÒWhat are the consequences if the measurements do not
meet speciÞcation?Ó If we are measuring voltage to ANSI C84.1-1989 [B1] speciÞcations, the tolerance is
±5%, steady state. There can be occasional excursions outside this boundary. What are the accuracy and cal-
ibration requirements for compliance with ANSI C84.1-1989? The document does not describe any. It does
say that the measured values must be in rms values but does not indicate whether the values can be deter-
mined by means of peak recording volt-meter technology that uses an algorithm to convert to rms. This
method is not accurate with distorted waveshapes. ANSI C84.1-1989 deÞnes steady-state as sustained volt-
age levels and not momentary voltage excursions. What if, however, we use an instrument that takes periodic
snapshots and then calculates an average? Is that method accurate? How can one say that calibration has
meaning to that instrument in an absolute sense?
Other factors to be considered include the assumptions of the instrument maker. If we are to measure true
rms values accurately under all distorted conditions, then there are only two options to considerÑwe can
either digitally sample multiple points on the wave of a whole cycle and calculate the true rms value, or we
can measure the heat generated in a resistor. Can one sample a token number of cycles and average them?
Possibly, except that voids in the data may result.
The accuracy of measuring the magnitude of the ac voltage both steady-state and transient disturbances can-
not be separated from the process of how the instrument records the measurement. Peak detecting, averaging
and special algorithms only work where there is limited waveform distortion.

---

IEEE
MONITORING ELECTRIC POWER QUALITY
Std 1159-1995
61