# Electrical System Harmonics

I'm a mechanical engineer so I don't know much about electric circuits and power generation to begin with. Can somebody explain what a harmonic is in the power grid and what is the reason why it is bad? I understand it comes from non-linear electric loads but what does that mean?

• It would help if you took a step back and provided a bit more context to your question. Start with why you need to know this. A harmonic is simply a frequency at an integer multiple of the fundamental frequency of a wave. It has no inherent badness. Commented Jun 22, 2017 at 17:20
• odd harmonic currents come from iron magnetic hysteresis in motors and transformers. many smaller harmonics come from peak voltage charged diode bridge cap current spikes. In 3 phase Y transformers the harmonics tend to be in phase and add to the expected balanced low Neutral current and may exceed the primary 3 phase wire current if excessive thus overheated conductors in transformers and also eddy current losses at higher frequencies. in magnetics. Commented Jun 22, 2017 at 17:30

## 3 Answers

A particular reason why the electricity companies think harmonics are bad is that they have to supply them (which means a generally slightly thicker cable on average) AND they can't usually bill the user for them. There are exceptions of course (for higher energy users) and they are encouraged (by the cost of their bill) to keep harmonics low and power factor as close to unity as possible.

A harmonic is a term that nearly always applies to a non-linear load distorting the normally sineusoidal load current. Basically it's not a higher power consumption but it does mean the infrastructure has to be able to cope with the basic billable currents and the generally non-billable harmonics: -

• What's the layman's definition for a non-linear load? And power companies need those thicker wires because of the spikes in the total current drawn that are higher than expected? Also how do harmonics affect the power grid as a whole? Does it induce harmonic current back on the system? Commented Jun 22, 2017 at 17:34
• I don't know of a definition that a layman can understand other than to equate it to the progressive acoustic distortion from a hifi when the volume is increased to high. You are warping your question into a much bigger question. Do some research on the terms I've used and raise a new question if necessary but don't move the goalposts (or evolve the question) because people like me don't get enthralled by that. Commented Jun 22, 2017 at 17:42
• During high demand periods, higher currents mean thicker wire but, the harmonics don't get smaller i.e. the cable still has to be a tad thicker and because they are higher frequency, those harmonic currents occupy the skin of conductors and produce larger-than-normal-current heating effects. Commented Jun 22, 2017 at 18:09
• Thanks it makes a lot more sense now. In addition to the heating damage that can result, do harmonics do anything else detrimental to a system though? Commented Jun 22, 2017 at 18:53
• They can cause icreased induction to other conductors (because induction is proportional to frequency) and, if the harmonics range up high enough in frequency they cause EMI and can interfere with radios and possibly, if taken to extremes might cause problems with other equipment such as watt meters. I think you should research this one because I'm out of ideas. Commented Jun 22, 2017 at 18:58

The following may help understand non-linear loads and how they generate harmonics.

simulate this circuit – Schematic created using CircuitLab

Figure 1. A simple bridge rectifier, smoothing capacitor and resistive load.

The circuit of Figure 1 should be easy enough to understand. When the top AC input is positive current flows through D2, R1 and D3 back to the other AC line. C1 charges up and maintains voltage on the load. For a mechanical engineer this would be analogous to having two pulsed hydraulic supplies, four non-return valves (D1 - 4), a pressure reservoir (C1) and the load (R1).

Form the above it should be seen that current (flow) on the first pulse will follow the supply pressure but that on subsequent pulses no current will flow until the reservoir pressure is exceeded by the supply pressure. The result is that the required current flows in pulses - see the black curves in Figure 2.

Figure 2. Rectified AC (red), capacitor voltage (blue) and rectifier current (black). The AC current (to the left of the rectifier on Figure 1) will look similar to the black curve except that each half cycle will alternate above and below the line.

It should also be intuitive that since the rectifier provides current in pulses that the peak current may be many times the average current.

Since any periodic wave can be formed by the addition of a sine wave of the fundamental frequency and its harmonics (in varying amounts) the distorted current demand creates harmonics in the current supply.

Figure 3. This fabulous illustration of the Fourier Transform by Lucas V. Barbosa on Wikipedia's Fourier transform page shows the transformation of a periodic waveform from the time domain to the frequency domain. The frequency plot shows the relative strength of the harmonics with clarity that could not be obtained from staring at the time plot.

If you understand how roller bearing create pulses from irregular surfaces , then you can imagine current gets bumps and harmonics from irregular impedance. It causes more heat and wear. In magnetics once the force reaches saturation like a spring bottoming out, the force rises sharply causing damage from excess current.

The nonlinear behavior of all magnetic cores is just like backlash on a belt drive but perhaps smoother. The hysteresis and stiffness or steep slope of some vs others that are smoother in motors of all kinds affects the amount of harmonics generated. Also SMPS supplies need to be active power factor corrected now if for large AC-DC supplies in IEC standards. 50% THD reduces PF by 10% which adds to the stored energy in the grid and increases conduction losses.

Harmonics are also generated by line frequency current pulses such as old linear diode bridge cap linear supplies that are anything but linear on the front end. THe current only charges ~ 10% or <30% depending on load and rating and then discharges to the regulators the rest of the time between each half cycle. These pulses create more damaging even harmonics which add in phase in the transformer neutral wire, while odd harmonics tend to cancel when in a differential mode or delta 3 phase system. (perfect 50% square waves have no even harmonics)

,Thus the harmonics are like bumps in a roller bearing surface impedance that add excess heat, stress forces, lower efficiency , insulation capabilities which all results in lower lifetime and higher cost of ownership.

• But where do these "bumps" come from? Is it just irregular demands for current like adding 10 computers to a circuit all of a sudden and then removing them soon after? Commented Jun 22, 2017 at 18:54
• They come from devices that don't draw power over the entire waveform. Cheap switch-mode power supplies are a good example, in that they only draw power from the sine wave when it's peak is above a certain value. Commented Jun 22, 2017 at 20:22
• An unregulated supply with 10% Vpp ripple will have pulse current about 10x the average oad for 10% of the time then none for 90% of the time. This is the current bump that affects THD and effective lower %PF as the center of the peak shifts earlier away from peak sine when load increases, PF drops slowly but THD is high. THe diode bridge is like a comparator if the input rectified sine > output DC it charges the cap via diode drop limited only by ESR of cap, diode and transformer loss . Otherwise off ( no current) Commented Jun 23, 2017 at 12:55