A recent trip to a Latin American country helped remind me of so much that we often take for granted in the United States. Their government requires all electric utilities to monitor the quality of the supply at the distribution substations, primarily as a flicker study, though also momentary and temporary sags and interruptions are to be counted each month.
Driving through the countryside at night would make one wonder why one should undertake such an ambitious program, since many residents have only candles for light, and there isn’t much that the utilities can do about the flickering. But after reviewing the data from one substation, it is clear why the utilities are concerned about the cities’ problems. The perceptibility short term (Pst) was averaging 0.7 (as shown in the figure above), with numerous excursions over 1.0, which is generally considered the threshold where most people will observe the light flicker from voltage fluctuations.
Voltage fluctuations that result in the light flicker follow the same rules as other power quality phenomena: Ohm’s and Kirchoff’s Laws. The human element, the eye-brain response to the changes in light intensity, is the difference in the equations. Two significant factors here are that the human eye doesn’t generally respond to variations above 20 Hz or so. That is how we can perceive movies as being continuous, even though they are made up of a series of still images being flashed on the screen at 24 times per second. On the other end of the spectrum, the eye-brain perception is most acute at around 9 Hz, where it takes just a quarter of a percent change in voltage for most people to observe the lights flickering.
Flicker monitoring has been around for most of the last century, with many people still relying on the “old GE curve” to determine the level of such. There are no shortages of sources, with arc furnaces, arc welders, resistance welding ma-chines, lamp dimmers, rolling mills, large electric motors with variable loads, heat-pumps or air conditioning, medical im-aging machines (X-ray, MRI, CAT Scan), large capacity copy machines, even appliances such as refrigerators, being poten-tial sources. Of these, arc furnaces and welders are perhaps the most notorious in this list.
Yet flicker doesn’t seem to get the attention in the United States as in most of the rest of the world. Maybe it is because this type of industry isn’t often located in residential and commercial areas, or because the grid is “stiffer,” or because people don’t realize it as the source of problems. There are theories out there that some of the “sick building” syndrome can be attrib-uted to flicker, making some people slightly nauseous. Maybe variations in fiber thickness, causing textiles and rugs to have a slight color change, are attributed to a mechanical problem with the equipment.
It is difficult, if not impossible, to measure the voltage fluctuations using your digital voltmeter. The voltage changes and the frequency of those changes need to be compared against the eye-brain response curves. This mathematical process takes consid-erable amount of processing power in power quality analyzers. If the source is a large horsepower motor starting periodically, then you probably could use the old GE flicker curves. Then again, the source of the problem was already known, so what to do about it is all that is left to figure out. But in most cases, a power quality monitor that complies with either IEEE 1459 or IEC 61000-4-15 standards for measuring Pst is necessary. If the monitor you use doesn’t have such capability, most instrument rental com-panies carry such, which takes the guess work out of determining if there is a problem and the location of its source.
Since the voltage fluctuations follow the same rules as sags and swells, monitoring the current at the same time as the volt-age will give you a clue as to where it is coming from. If the current goes up significantly when the Pst value goes up, the source of the problem is most likely downstream from the monitor. If the current stays steady or goes down when the Pst goes up, then look upstream.
Since the Pst value is computed over a 10-minute interval, it may take a while to get good data. For the impatient, some in-struments output a value called Pinst (the instantaneous perceptibility). This number doesn’t follow the rule of 1.0 being the limit of perception, but it jumps dramatically when the voltage is fluctuating.
While flicker doesn’t get the attention blackouts get, it is another parameter to be aware of when doing a periodic audit of the facilities power system. And it may be an indication of a larger problem lurking in the dark. EC
BINGHAM, a contributing editor for power quality, can be reached at 732.287.3680.