Most members of IEC, IEEE and CSA working groups assume that small devices cause negligible harmonic distortion on the power system. Without any measurement, simulation or technical study substantiating this assumption there is a perception that any low-power equipment below 75 W is producing a negligible distortion level. Since field measurements have shown that significant harmonic levels exist in the voltage supply at all times, it is important to find the origin of this distortion.
Triplen harmonics are the most significant harmonic levels recorded over the past five years, as observed in CEATI project 5142 “Harmonic Distortion Level Evolution at Residential PCCs in North America”. These measured distortions come almost exclusively from single-phase equipment, since three-phase equipment generates very small levels of triplen harmonics. An indication that single-phase devices remain continuously connected is a substantial level of the triplen harmonic recorded at all times, even during the night when most single-phase equipment is expected to be powered off.
Let us assume that each device produces 50 mA of third harmonic. Since the third harmonic current of small devices is generally nearly equal to the fundamental, individually this equipment consumes only 6 W (50mA × 120V). If each house connects 10 of these devices, then the harmonic3 (h3) becomes 0.5A per house. On average, MV power feeders supply about 1500 residential customers. Consequently, each MV feeder must absorb at least an equivalent of 750 A of the third harmonic current, based on 120 V. At 14.4kV, this current appears to be negligible with only 6.25A; however, this assumption is incorrect.
According to an internal survey recently conducted by a utility in Canada, the 95th percentile of its MV feeder impedance is 46.8Ω at 60 Hz, with the X/R ratio of 10.49 in average. A rough computation can provide a general figure of the impact of this harmonic current on the power system. Using the internal survey data for the MV feeder impedance, and neglecting the feeder conductor skin effect for the resistive portion, the impedance for h3 becomes approximately equal to 130Ω at the 95th percentile. In addition, the impedances found for networks operated by other utilities in Canada significantly exceeded this value. Nevertheless, let us assume that the current is injected in the middle of the feeder with the impedance of 65Ω at 180 Hz. This current produces a voltage drop of 341 V on a 14.4 kV base, which already represents 2.4% of the total voltage magnitude and is equal to 47% of the IEC-defined compatibility level. Although this is only a very rough estimate, it indicates the importance of such small devices, which can easily maintain the distortion level on the power system over a certain limit. This indication raises the question regarding the impact of small devices permanently connected to the power system.
This project aims to better estimate the impact of small power devices connected in each household as part of the typical residential load. For this purpose, it uses the harmonic levels measured at several locations within the US and Canada in the context of CEATI project 5142. The work performed also includes the current and voltage measurement of more than 100 devices. The values obtained from these measurements allowed for the development of three harmonic models for mixed devices operating in standby mode, as well as one harmonic model for a mix of 10 LED lightbulbs of various supply circuits. The measured harmonic levels were also used to adapt the harmonic compensation factors kp (see IEC61000-1-4) for Canadian and US networks.
CYMDIST software was used to simulate the actual networks supplied by substations operated by five individual Canadian utilities. Applying the harmonic models to these simulations and considering the compensation factor kp, the average and 95th percentile of the harmonic voltage levels produced by small devices permanently connected in each household was calculated. The simulations also considered the LEDs connected with these devices. In addition, the report comments on the design of the power supply of several small devices and LEDs in use today. Finally, the results of field measurements of the network impedances up to 96 kHz at 120V and at240 V are presented.
Compliance assessment, Harmonic, 2-9 kHz components, Harmonic angle, Power quality, PQ measurement, PQ survey methodology