By Glen Shomo, P.E.
The term flicker is often used to describe voltage fluctuations, especially when viewing graphs of RMS voltage trends that include fast, cycle-level variations. These fluctuations are typically caused by varying loads such as welders, arc furnaces and motor start currents. The term flicker, however, should be used to describe the visible change in light intensity, or light flicker, caused by voltage fluctuations on the electric power system. While normal line voltage changes of a few percent are not uncommon, a voltage change as small as 0.25% can produce a visible change in light output from incandescent lamps for some observers. When this light fluctuation becomes irritating is a very subjective issue involving human perception. Humans are generally tolerant of large, infrequent light variations, but are less tolerant of smaller variations that occur more often.
Flicker Tools
Utilities evaluate flicker using voltage sag calculations and estimates of the frequency of occurrence, and by using actual field measurements of voltage fluctuations. These results are often compared to the old GE flicker curve, a common tool for evaluating human light flicker perception, to determine the severity of the voltage modulation with respect to the threshold of visibility or irritation plots. Utilities have also used the trusted GE curve in the reverse fashion for setting general rule-of-thumb voltage sag limits. For example, a residential heat pump that is expected to cycle four times per hour should produce voltage sags no greater than about 4% to avoid flicker complaints, according to the threshold of irritation curve in the GE flicker graph.
The GE flicker curve was developed in the early 1920's using fixed frequency square wave modulation of the voltage applied to incandescent lamps. This simple modulation pattern produced abrupt step changes in voltage and represented a single source of light flicker. This is in sharp contrast to real loads connected to electric power systems today. Many situations encountered today are combinations of several flicker sources with varying frequencies, operation patterns and shapes of voltage modulation that are not addressed by the old GE flicker curves.
Recognizing the need for a new method of evaluating light flicker, the IEC 61000-4-15 flickermeter standard was developed in Europe and later adopted in the U.S. as IEEE 1453-2004. The measurement methodology described in the IEC standard more accurately accounts for complex voltage fluctuations encountered in actual practice by including the effects of multiple flicker sources, frequencies and varying voltage modulation waveforms. In addition, the flickermeter approach standardizes flicker monitoring across different manufacturers–IEC 61000-4-15 compliant instruments should all produce the same results for a given flicker excitation.
The real advantage of the flickermeter method is inherent in its ability to accurately model the human flicker perception. This rather complex modeling is accomplished by five signal processing blocks described in the IEC standard, which represent the lamp-eye-brain response to light flicker–the response of a lamp to supply voltage variations, the perception of the human eye and the memory characteristics of the human brain. A simplified block diagram of this system is shown in Figure 1.
The flickermeter system provides several measurement outputs. IFL, or instantaneous flicker level, represents the real time voltage modulation modified by the lamp-eye-brain response. This output can be plotted as a time interval graph and is useful for tracking down sources of voltage fluctuations.
A statistical analysis block completes the human perception system by providing short and long term flicker severity indexes. Short term flicker severity Pst is evaluated over a 10 minute observation period and is used to evaluate disturbances caused by flicker sources with short duty cycles. According to the IEEE 1453 standard, a Pst value of 1.0 represents the system compatibility level, the level below which customer complaints are not likely to occur. Pst is therefore commonly used to evaluate whether the measured voltage fluctuations are severe enough to cause flicker complaints.
Long term flicker Plt is derived from 12 successive Pst values, or two hours, and is more suitable for evaluating the combined effect of several randomly operating loads such as welders or motors over longer periods of time. A Plt value of 0.8 is considered the system compatibility limit according to IEEE 1453.
 |
Figure 1 |
An additional benefit of the flickermeter system is that the IFL, Pst and Plt values are numerical outputs provided at regular time intervals, which can be presented as graphs in power quality analysis software. Time trends of short and long term flicker perception for the duration of the recording session can be viewed and printed to evaluate variations in flicker severity for individual customers.
While the old GE flicker curve has served utilities reliably for many years and is still useful for evaluating simple load fluctuations from single flicker sources, the newer flickermeter standards provide a much improved method for evaluating human perception of light flicker due to complex loads from multiple sources or customers with varying rates of occurrence. When investigating flicker complaints, utility personnel should include the IEEE 1453 flickermeter measurement when selecting initialization settings for power quality recorders. As the newer flickermeter standard achieves broader acceptance and use, an increasing number of electric utilities are incorporating the improved method into standard practices for evaluating light flicker issues.
More Utility Products Current Issue Articles
More Utility Products Archives Issue Articles
