During the 1996 NEC cycle, a task group composed of interested parties was created to recommend to the National Electrical Code Committee the direction its standards should take to improve the safeguarding of persons and property from conditions that can be introduced by nonlinear loads. This group was designated the NEC Correlating Committee Ad Hoc Subcommittee on Nonlinear Loads. The scope of this subcommittee was as follows:
(1) To study the effects of electrical loads producing substantial current distortion upon electrical system distribution components including, but not limited to
a. Distribution transformers, current transformers, and others
b. Switchboards and panelboards
c. Phase and neutral feeder conductors
d. Phase and neutral branch-circuit conductors
e. Proximate data and communications conductors
(2) To study harmful effects, if any, to the system components from overheating resulting from these load characteristics
(3) To make recommendations for methods to minimize the harmful effects of nonlinear loads considering all means, including compensating methods at load sources
(4) To prepare proposals, if necessary, to amend the 1996 National Electrical Code, where amelioration to fire safety may be achieved
The subcommittee reviewed technical literature and electrical theory on the fundamental nature of harmonic distortion, as well as the requirements in and proposals for the 1993 NEC regarding nonlinear loads. The subcommittee concluded that, while nonlinear loads can cause undesirable operational effects, including additional heating, no significant threat to persons and property has been adequately substantiated.
The subcommittee agreed with the existing Code text regarding nonlinear loads. However, the subcommittee submitted many proposals for the 1996 NEC, including a definition of nonlinear load, revised text reflecting that definition, fine print notes calling attention to the effects of nonlinear loads, and proposals permitting the paralleling of neutral conductors in existing installations under engineering supervision.
As part of the subcommittee's final report, nine proposals for changes to the 1993 NEC were submitted. All were accepted without modification as changes to the 1996 NEC. Also included in this report and now pertinent to the 2002 NEC 310.15(B)(4)(c) is the following discussion.
SHOULD NEUTRAL CONDUCTORS BE OVERSIZED?
There is concern that, because the theoretical maximum neutral current is 1.73 times the balanced phase conductor current, a potential exists for neutral conductor overheating in 3-phase, 4-wire, wye-connected power systems. The subcommittee acknowledged this theoretical basis, although a review of documented information could not identify fires attributed to the use of nonlinear loads.
The subcommittee reviewed all available data regarding measurements of circuits that contain nonlinear loads. The data were obtained from consultants, equipment manufacturers, and testing laboratories, and included hundreds of feeder and branch circuits involving 3-phase, 4-wire, wye-connected systems with nonlinear loads. The data revealed that many circuits had neutral conductor current greater than the phase conductor current, and approximately 5 percent of all circuits reported had neutral conductor current exceeding 125 percent of the highest phase conductor current. One documented survey with data collected in 1988 from 146 three-phase computer power system sites determined that 3.4 percent of the sites had neutral current in excess of the rated system full-load current.
According to 384-16(C) of the 1993 NEC (for the 2002 NEC, refer to 408.16), the total continuous load on any overcurrent device located in a panelboard should not exceed 80 percent of its rating (the exception being assemblies listed for continuous operation at 100 percent of its rating). Because the neutral conductor is usually not connected to an overcurrent device, derating for continuous operation is not necessary. Therefore, neutral conductor ampacity is usually 125 percent of the maximum continuous current allowed by the overcurrent device.
Also important for gathering electrically measured data from existing installations is the following.
Measurement of Nonsinusoidal Voltages and Currents
The measurement of nonsinusoidal voltages and currents may require instruments different from the conventional meters used to measure sinusoidal waveforms. Many voltage and current meters respond only to the peak value of a waveform, and indicate a value that is equivalent to the rms value of a sinusoidal waveform. For a sinusoidal waveform the rms value will be 70.7 percent of the peak value. Meters of this type are known as ?average responding meters? and will only give a true indication if the waveform being measured is sinusoidal. Both analog and digital meters may be average responding instruments. Voltages and currents that are nonsinusoidal, such as those with harmonic frequencies, cannot be accurately measured using an average responding meter. Only a meter that measures ?true rms? can be used to correctly measure the rms value of a nonsinusoidal waveform.
