# Not Frequently a Problem: How Load Affects Frequency

Three sine waves 120 degrees apart from each other. Image Credit: Richard Bingham

One of the most fundamental parameters in power quality is often taken for granted. The late-19th century “Battle of the Currents” between Tesla and Westinghouse (who favored alternating current) against Edison (who favored direct current) was decided in favor of AC. This meant the electric utility voltage distribution grid would be the “alternating” shape of the sine wave, like a roller coaster going up and down forever.

The naturally occurring wave shape appears in many aspects of science, from the speed of an unimpeded pendulum swing to the voltage produced from a rotating coil in a magnetic field.

The figure above shows a set of three-phase voltage waveforms. The waveforms each go from 0 volts (V) up to the positive peak, back to 0V, down to the negative peak and back to 0V in one cycle, highlighted in yellow for the red waveform. In this example, that cycle lasts 16.66 milliseconds (msec), which translates to 60 hertz (Hz) in frequency, the common frequency used in North America. Europe and some other parts of the world use 50 Hz, and some shipboard and aircraft systems use 400 Hz or 800 Hz. We rely on this periodicity to keep a lot of things constant (from keeping time with old electromechanical clocks to large-scale mechanical processes), and basically the same number across the entire grid. Usually, it is almost constant over time, with the average value (middle purple lines) being just +/- 0.01 Hz from 60 Hz, and the min/max excursions being less than +/- 0.05 Hz.

The three waveforms shown in the figure represent a typical three-phase voltage system. Just like the pistons in a car engine that fire in a sequence to keep the engine turning smoothly with nearly constant power, the three waveforms cross the zero axis in a sequence of 5.55 msec apart from each other, represented by the three purple boxes labeled 1, 2 and 3. Since one complete cycle in a circle is 360 degrees, each of these is 120 degrees apart from each other. This also is another PQ parameter that we often take for granted. In a perfect world, the magnitude, frequency and phase relationship of the three voltage waveforms would remain the same. While we are aware of the effects of the voltage magnitude changing with sags, swells and interruptions, what makes the frequency or phase angles change, and what are the consequences?

The grid is made up of many large-scale voltage generators interconnected by transmission wires. With the growth of renewable energy sources, this paradigm has changed, and the distributed energy resources (DERs) are much larger in number connected through the distribution system. The voltage they produce must be at the same frequency at the same phase relationship with high precision, or the power flows don’t happen the way the electric utilities (and the consumers) want them to. The worst-case scenario would be if there were two voltage waveforms were 180 degrees out of phase of each other. As one would go up, the other would go down. Connect them together, and there would be no usable voltage (and likely a large fireball).

When a large demand is suddenly placed on a rotating electromechanical generator, the generator’s rotational speed slows down. This causes the voltage frequency produced to reduce. Conversely, suddenly removing a large load can cause the generator to speed up until the control systems can adjust. If the change is large enough, there may not be a complete recovery. Adding this voltage source into the grid is going to cause instability issues.

While many PQ monitors measure frequency, they often use only the 10 second average value as the IEC standards require. Many don’t record the phase relationship between the voltages and currents.

In the past 10 years, utilities have deployed a dedicated instrument, called a phasor measurement unit (PMU) to get a more accurate measurement across their territory. With microsecond-accurate time stamps, this phase angle and frequency data can be compared across the entire grid.

Phase angle, as measured by the phasors, is the first indication of instability. Changes in frequency tend to occur closer to the onset of a problem. By the time the voltage changes, it’s often already too late to correct the error. More likely, it’s going to cause localized issues, and if more substantial action isn’t taken quickly, it’s going to get bad for a lot of people, such as with the large, cascading blackouts.

What will be the impact of 25–50 percent of the power provided to the grid coming from distributed energy resources, most of which aren’t controlled by the utilities? Solar and wind power use conversions from DC to 60 Hz AC or varying hertz AC to DC to 60 Hz AC, respectively. While they don’t have the same characteristics as the electromechanical generators, they present a new situation that the old mathematical models for operational predictions of stability don’t handle well. So far, the answer is … we really don’t know yet.