Harmful effect on power systems from using big VFDs or tons of computers at home

[ICanHas]

An old small industrial VFD or new bare-minimum built CHINA MADE VFD eBay devoid of all but the essential parts can have a power factor of 0.5 to 0.6 with a total harmonic distortion of 100-150%.

The power company provides medium voltage power to a distribution transformer in your subdivision to provide the low voltage power. A subdivision with 10 houses with average power usage level will have something like 34.45kW of demand. This is computed by 7.03kW demand x 10 homes x 0.49 coincidence factor) Such a subdivision maybe fed from a 50kVA transformer. The short circuit rating on this type of transformer is about 4,000A wth a 5.2%Z rating.

When some DIYer installs a surplus industrial/eBay China made 3 hp VFD without harmonics filtration out of ignorance or selfishness, it wreaks havoc on the local power system and cause this to happen to the entire subdivision:



http://www2.schneider-electric.com/m..._solutions.pdf

The losses in the power company transformer also increases substantially which means the power taken from primary side is greater than it is with a 3kW non-harmonic load. The additional loss contributes to increased "carbon footprint" and energy consumption which is anti-eco.

You can expect an input power of 3kW after motor and drive losses at full output. The highest 15 minute average is used for demand in the power industry and since you can anticipate it can operate more than 15 minutes at a time at full load, 3kW will be used for demand.

VFDs and computer power supplies contain power electronics that produce e EXTREME amounts of harmonic currents. The problem is that power system for subdivision of 10 homes is not designed to see 3kW of sustained harmonic load. CFLs do not escalate to this level of problems unless you switch on 230 13W CFLs all at once and leave them on for well over 15 minutes in a subdivision of 10 homes. Yeah, that ain't gonna happen in real life.

A large, straight rectifier power supplies with straight rectifier input is common in internet general purpose CHINA MADE VFDs.

Why is it a problem when someone use power polluting loads that is not even meant to be used at home like a 3hp VFD or run ten computers to do "coin mining" mining and do dumb stuff like "omg I'll save 40W between mah ten computers if I bypass the PFC".

A 0.5PF 3hp VFD will have an input characteristics of
240V
input amps: 25A RMS, 6kVA total
PF 0.5
60Hz components: 12.5A, 3kW
THD of 100% fundamental: 12.5A, 3kVA
3rd harmonic is probably in the order of 2.4kVA
PSC used in air conditioners produce no harmonics and power factor is close to 1.0.

That one customer is drawing 3kVA of harmonic demand on the subdivision transformer at the point of power coupling. 3kVA/34.45kVA =8.7% TOTAL DEMAND DISTORTION (TDD)
IEEE 519-1992 recommendation is not more than 5%.

Computers, ECM, X13 and such from all the other homes will add up to about 2kVA for a combined total of 5kVA or 14.5% TDD leading to situation that breaks the camel's back.

100 CFLs turned on at the same time across 30 houses would not be an issue, because, the demand is small RELATIVE to the total demand and transformer size that would be used to feed 30 houses. The effect of you alone 3kVA of harmonic current depends on the size of the subdivision transformer on which you're connected.

The attached PDF tells you the effect of when proportion of harmonics relative to total demand and transformer size is excessive. This is an example of large portion of electronic battery charger being used at the same time by one customer. High proportion of harmonic producing load reduces efficiency of power company transformers and reduce efficiency of induction motors used by other customers by distorting the voltage waveform. Both of these have negative environmental impact.

This is why harmonic loads often cause a problem with generators. The harmonic producing dirty load is HUGE relative to the capacity of your home generator which is much smaller than the neighborhood distribution transformer. When you have to "oversize" generator to work with harmonic loads, you get a first hand experience of concept of wasted capacity.

Until recently, harmonic loads relative to total demand was negligible in residential power systems. Increased use of CFLs and harmonic producing embedded systems type appliances is changing this.

A 3 hp VFD in an industrial facility or an apartment complex with a 400kW demand fed from a 1,000kVA transformer is not a problem, because it's small in proportion to the local power system.

[jeff5may]

And your dissertation says what? Inverter-drive mini-splits are evil? Edison was better than Tesla? Time machines are cool? I don't get it.

[ICanHas]

Originally Posted by jeff5may

And your dissertation says what? Inverter-drive mini-splits are evil? Edison was better than Tesla? Time machines are cool? I don't get it.

OEMs recognize the power quality issues in these large harmonic producing loads and consumer oriented products like production spec mini-splits do have harmonic mitigations.

[jeff5may]

So could I rob power from the grid by leeching off the harmonics and rejecting the fundamental frequency? If I could catch that reflected power, I don't think the power utility would mind. I don't think they would come looking for me.

[ICanHas]

Originally Posted by jeff5may

So could I rob power from the grid by leeching off the harmonics and rejecting the fundamental frequency?

That's exactly how devices cause problems on the grid.

If you were going to apply brakes on your bicycle you can:

apply smooth and steady braking throughout the rotation.... "clean load"
applying 1 x brake pad force over the entire 75 inch circumference of the rim

or you could apply "dirty load" much harder in very short bursts and only when the wheels are in 12 and 6 o c'lock positions.
10 times the force applied to 3.75 inch section at a time, twice per rotation.

This is how dirty power computers draw power.

Even if you do it so that braking effect per rotation (the same amount of power dissipated per rotation) is the same, the latter is a lot more troublesome.

If I could catch that reflected power, I don't think the power utility would mind. I don't think they would come looking for me.

The brakes are not "producing power" but you can see that there will be vibration that is in multiples of rotation of the wheel. Given the same braking load, the same will have a lot more effect on a bicycle(say subdivision transformer) than on a truck (a 1000kVA transformer feeding a factory).

The modern issue with dirty power drawing inverter drives(VFD) and power electronics is the proliferation of dirty load relative to smooth and steady load. Power factor corrected devices avoid these issues, but China factories and OEM are opposed to it because of cost.

[NiHaoMike]

Work out what the power factor of a non PFC rectifier load would be if you gave it a square wave. (Answer: exactly 1)

Granted, that's a theoretical result only as a true square wave (with infinite slew rate) can not exist, nor is it even desirable. So make it a slew rate limited square wave ("trapezoid wave") where the slew rate is chosen to be reasonably fast, but slow enough to avoid EMI. The flat parts are effectively DC, but the transitions allow arcs to quench just like what happens with AC. Think of it as a compromise between DC and AC.

The question now, of course, is where would you find such a waveform? As it turns out, it's actually a very good waveform for a power inverter to output. During the flat parts of the wave, the transistors are simply left on or off as needed, with PWM only used to control the slew rate. That significantly improves the efficiency. It works even better with non PFC rectifier loads. The loads then only draw power on the flat parts of the waveform, requiring the inverter to only supply a small amount of current on the transitions to overcome capacitance.

It's also possible to design a generator to output trapezoidal waves, and in fact, the motor/generators in many hybrid and electric vehicles do just that.

Look into the future where alternative energy is the norm and rectifier loads like PoL (Point of Load) VFDs are everywhere.

[jeff5may]

No, you don't get what I'm trying to say.

One could build a device that could connect to (someone else's) electric service and "clean up" their dirty power consumption by drawing off power during the times in the AC cycle when there was little or no load. Some of that power could be added back in during times of heavy load to boost the source, making the overall draw seem average. This could possibly fool the electric power meter, maybe even reducing the user's electric bill. The power not added back in could then be siphoned off by me and used at my discretion.

How could this "energy washing" machine be made?

[stevehull]

Yes, it has been done. See Technology Review a couple years back (spin the meter backward).

Steve

[philb]

Get the electric company to install a line drive reactor at the point where they occur.
Several papers are published by the IEEE. Google "vfd harmonics" I couldn't get the correct url posted here without mistakes.

If you'd like your fvd questions answered by power line engeneers, go here:
Eng-Tips Forums

[ICanHas]

Look into the future where alternative energy is the norm and rectifier loads like PoL (Point of Load) VFDs are everywhere.

No, you're looking into the near past. As dirty load concerns grow, regulations will tighten and I expect the expansion of something similar to IEC directives in Europe regulating harmonic producing products that can be sold.

With the advent of real-time monitoring and two way communicating " embedded systems " smart meter instrumentation, it can become economical to implement firmware software based DSP digital for demand billing, harmonics or power factor penalty or auto reporting to sniff out and document source of noxious polluters without the costly task of manual investigation.

Please see the attachment to my first post. Now solution can be as easy as communicating with the customer to work out a plan.

Consider a customer engaged in unreasonable use of service such as running a roomful of bit-coin mining machines with omg so 1337 h4x PFC bypassed power supply or a multi horsepower unfiltered, minimally filtered VFD that was not meant to be used on a residential service and producing disproportionate amounts harmonics.

When power quality issues arise and that customer is unwilling to pay for his proportion of corrective equipment nor is willing to discontinuing the unreasonable use of service, the data gathered can be useful.

The history of making unreasonable levels of power line pollution could be useful in enforcing the terms of tariff "Usage of Service Detrimental to Other Customers" and seek service disconnect or seek judgment for the cost of corrective measures through the legal system.

Originally Posted by philb

Get the electric company to install a line drive reactor at the point where they occur.
Several papers are published by the IEEE. Google "vfd harmonics" I couldn't get the correct url posted here without mistakes.

If you'd like your fvd questions answered by power line engeneers, go here:
Eng-Tips Forums

http://www05.abb.com/global/scot/sco...82003d0201.pdf

Something like that is used to reduce harmonic current flow from getting past the point of common coupling. If the PoCo installs it, its coming out of rate payers. It would be the solution when harmonics at the transformer gets that bad through increased use of rectifier loads, but the idea of something like IEC directive is to prevent that from getting to critical mass in the first place.

See the attachment in this post for an example of harmonics created by one customer affecting another customer. Since the customer was not engaged in unreasonable use and the usage is not avoidable, the power company installed a separate transformer to isolate the harmed customer.

If the cause was from unreasonable use of service like the example I mentioned earlier, it would make sense to bill the offender for the cost of putting the harmed customer on another transformer.