Induction motors and power factor

[Guineafowl]

Hi all,

I have a 1 HP capacitor start motor (old, c.1950) that runs an air compressor in my little workshop. When it's running, the lights in the shop and the whole house dim and brighten rhythmically, at about 5 Hz. This, of course, is related to the compression cycles of the compressor, but how is such a low power appliance having such an effect on the line voltage?

I have researched power factors and realize that the old motor is likely to have a low one, but I have two questions:

1. Is the brightness of the house lights oscillating between normal and low, or becoming brighter than normal then dimmer, oscillating about the normal?

2. Would inserting a capacitor in series with the run windings help, or is this too simplistic?

Our house supply is UK standard, i.e. 230V AC 50Hz.

 

[Integrand]

Oscillation of the lights between normal and low, and becoming brighter than normal then dimmer, are both possible for different reasons. The latter case can be caused by wiring faults/degradation which constitute a fire hazard. You should have your service company check the relevant circuits as soon as possible. After that potential cause is resolved, others can be considered if needed.

 

[Guineafowl]

Thanks, Integrand. I'll have to check, but I think the compressor did the same to the previous owner's lights. Certainly worth checking the wiring though. What sort of fault might be there? The workshop is powered by a spur from the downstairs ring.

The lights oscillating about the normal makes sense to me - the motor is drawing excess energy as reactive power (dim) then returning it to the supply (bright). According to my clamp meter, the current drawn is about 5 A (unloaded) which would equate to 1150 W, far more than the rated 1 HP, or 750 W.

 

[anorlunda]

I think you have two independent questions. The motor behavior is one. Inadequate or faulty wiring is another. We are taught to always suspect faulty wiring when house lights dim repeatedly.

But your identification of 5 hz oscillations sounds unlikely. Do you mean 5 second period, 1/5 hz? Fluctuations at 5 hz approach stroboscopic frequencies. Although people's abilities vary greatly, I think very few people would be able to identify a specific number like 5 hz for flickers that fast.

 

[jim hardy]

The workshop is powered by a spur from the downstairs ring.

Here in the US where 200 amp service is typical a 1hp motor won't dim the lights noticeably.
Your pulsations are as you said from the increased current drawn by the motor on every compression stroke. So they're from real current not reactive.
Compressors are an unconstant load. The motor is developing considerably more than 1hp on the upstroke and considerably less on the downstroke.
That a 1hp motor can modulate voltage that much does seem curious though.

What is the amp rating of service to your house?

Perception of flicker is complex because Incandescent lamps aren't linear and neither is the eye.
A friend complained of lights dimming whenever his airconditioner compressor started. We measured and found his house voltage was too high by almost ten percent.
We also checked all the connections in his panel, found no measurable voltage drop and nothing running warm..
Calling the power company to correct the high voltage, which they did by changing a transformer tap , made the effect no longer noticeable.
And his incandescent bulbs last a lot longer now.

old jim

 

[Guineafowl]

I think you have two independent questions. The motor behavior is one. Inadequate or faulty wiring is another. We are taught to always suspect faulty wiring when house lights dim repeatedly.

But your identification of 5 hz oscillations sounds unlikely. Do you mean 5 second period, 1/5 hz? Fluctuations at 5 hz approach stroboscopic frequencies. Although people's abilities vary greatly, I think very few people would be able to identify a specific number like 5 hz for flickers that fast.

I've checked again - I can count about 40 dim-bright cycles in ten seconds, so 4 Hz is more accurate. They exactly match the noise of the compressor, and only occur with it on.

Here in the US where 200 amp service is typical a 1hp motor won't dim the lights noticeably.
Your pulsations are as you said from the increased current drawn by the motor on every compression stroke. So they're from real current not reactive.
Compressors are an unconstant load. The motor is developing considerably more than 1hp on the upstroke and considerably less on the downstroke.
That a 1hp motor can modulate voltage that much does seem curious though.

What is the amp rating of service to your house?

Perception of flicker is complex because Incandescent lamps aren't linear and neither is the eye.
A friend complained of lights dimming whenever his airconditioner compressor started. We measured and found his house voltage was too high by almost ten percent.
We also checked all the connections in his panel, found no measurable voltage drop and nothing running warm..
Calling the power company to correct the high voltage, which they did by changing a transformer tap , made the effect no longer noticeable.
And his incandescent bulbs last a lot longer now.

old jim

The service head cutout fuse is 100A, but this is with 230V. Also, both incandescents and CFLs are affected - we hardly have any incandescents now, as our lovely EU has pretty much banned them.

I suppose my ordinary meters won't measure my supply voltage accurately enough to rule out the situation above.

Thanks to you both.

 

[Salvador]

Well EU indeed has ruled out the old bulbs , sadly as they are great at specific areas.
As for your problem , do you have an analog voltmeter? you could plug it in one of your wall sockets and watch as the compressor runs how much the arrow swings , then you could get an approximate knowledge of how big the voltage swing is , or you can use a digital multimeter but you need one that can show the difference in voltage fast enough , just with analog it is easier for the eye.

Anyway I doubt it's the power factor , seems more likely the voltage is sagging under load.
Can you try out something different , like take some other high power tool or instrument and test , preferably something that has a linear load aka resistance , like a water heater , electrical oven , a toaster I dunno.Just switch it on and off slowly and watch what happens then , do the lights dim then maybe use a voltmeter also to see how big is the difference then.
Well that's what I would do to see further into the problem.

 

[Guineafowl]

I have an analogue voltmeter - my 'poor man's oscilloscope'.

Normal voltage - 255 V
With compressor - 250 V and oscillates minutely +- 2v at most
With 10 kW shower - 235 V
With 3 kW kettle - 248 V

?

 

[Tom.G]

Your voltage readings indicate a problem somewhere between your fuse box and the power company's generator. This is based on the assumption that the voltage readings are taken with the loads being on different branch circuits. The voltage drops are about linear with the current being drawn, indicating a problem common to all of them.

I would guess similar probabilities of service head/cutout fuse box or the power company. The closer you are to the seacoast, the more likely it's the power company. The junctions in the power lines tend to corrode in the salt air.

 

[jim hardy]

Normal voltage - 255 V
With compressor - 250 V and oscillates minutely +- 2v at most
With 10 kW shower - 235 V
With 3 kW kettle - 248 V

Where did you measure that ?

Let's look at that voltage drop..
10kw at nominal 240 volts is about 41.67 amps
and that drops your voltage from 255 to 235, 20 volts
R = E/I . 20 volts /41.67 = 0.48ohms , and 41.67 amps through 0.48 ohms is 833 watts. . That's a LOT of heat, about as much as a toaster.
Something should be getting very warm when you run that 10kw shower.
Open your breaker box and look for something discolored. Feel the front of the panel and the breakers for a hot spot.
If you don't find one , call the power company, tell them you have excessive "flicker" and ask for a check of your service.
Could be a loose connection anywhere. It's in their interest to not let it wreck something. They can look for it with an infrared camera.

With compressor - 250 V and oscillates minutely +- 2v at most

That gives you an idea just how sensitive the eye is to brightness of incandescent lamps.
Doubtless the needle on your meter is damped by air and would swing a little further if it could.
Run compressor and shower together - is the fluctuating brightness less noticeable at that lower voltage?

System voltage varies according to how much power is being pushed past your house to other customers further away.. So check incoming voltage at a different time of day, too. Show those readings to the guys who come to check your service.

old jim

 

[Salvador]

Surely I think almost anywhere the max voltage fluctuates depending on which time of day you measure , for evenings I have 235 to 240 then at mornings or active day hours when all the factories run their max i have about 225/230.

Anyway your measured 255 is a bit too high I think, 260 is the maximum voltage on many single phase appliances that have some sort of semiconductor inner stuff, surely a kettle and a lamp will run happily on various different voltages, but a tv and a pc psu may not.
Another thing i find interesting is that only a 2 volt swing can be seen by the eyes when looking at a incandescent lamp.As for your calculations Jim, it can be one of either two possibilities, if there's a problem in his house wiring then those 800+ watts of heat can cause trouble over such short wire distance but if the voltage sags all the way to the nearest transformer for whatever reason then no noticeable extra heat would be felt or seen.
Here in the countryside were old wires are still used many would get an even bigger drop in voltage for a 10kw load , no wires would ever get warm simply because they are steel wires running on poles and they are very long , so with outside conditions like wind blowing and low temperatures they don't give a damn so to speak they just drop the voltage and that's it.

But hmm , a 10KW load is actually quite a big one.I'm not even sure how much f a voltage drop I would get in my house.Since I don't have such a load not even combining all of my stuff.But for em it's different because i live in an apartment block and each of the phases in the incoming cable here feed atleast a dozen flats so surely the combined power is quite large.Ok Guineafowl what more you can do , if I were you I would now try one more thing , can you locate whether the voltage drop happens in your house or in the incoming wires , the first trick that comes to mind is this , can you take out the fuses and disconnect your house from the supply or simply turn off all your electrical appliances , then take a high power load like your 10kW shower and connect it right at the fuse box were the cables run into your house , in other words as close as possible to were the wires come in from the street main cables.then when you do this measure the voltage drops again and then you would see , if the voltage drops are the same , then surely your house wiring is not to blame but the voltage drop then happens along the line outside your house and so then you must call your power company and complain.P.S. I'm not sure why my writing went from black to blue but there's nothing I can do so hope you like the change. :D

 

[Guineafowl]

Jim Hardy - I measured the compressor voltage drop at a workshop socket, and the other two in the kitchen. My meter is only a simple Draper one, so the absolute values might be out, but I feel the relative values are relevant.

I think you, Tom.G and Salvador are all saying the same thing - that the line voltage is dropping across the supply's 'internal resistance'. Nothing is hot at the consumer unit (checked while wife was in shower) so I imagine the drop is coming from the overhead cables. The shower is run off a dedicated 40 A MCB and 10mm2 cable (no more than 10m long).

As Salvador correctly states, my (very rural) supply comes in on very old bare steel cables - there are six poles solely to get the supply to my house, and this is tapped off a very long steel line that is a long way from the substation. There is a pole-mounted transformer on the fifth pole from me - I guess from P=I2R that most of the power loss and voltage drop occur between this and my house, given the lower efficiency of lower voltage lines. Would anyone know what sort of stepdown is occurring at this transformer?

Seems like this discussion of power factors might have unearthed a problem with my supply!

 

[Salvador]

i don't know about UK , but were I live and also back in the former USSR we had a simple line system, now to exclude the higher lines from this equation, the lower line would look like this , the rural countryside lines are all 20kV lines, then when such a line reaches a local small village or town it goes into a transformer and outcomes 3 phase 0.4kV or 380v line with 3 wires and a neutral. so the last pole in rural areas from the transformer carries 4 wires , 3 phases and a neutral.

Some larger homes or farms have one small step down transformer themselves so the 20kV line is stepped down right at their doorstep and the incoming 380v 3 phase goes just a few meters in length.

I don't know the transformer you have there but do you have a three phase wiring coming into your house or single phase?

 

[Guineafowl]

i don't know about UK , but were I live and also back in the former USSR we had a simple line system, now to exclude the higher lines from this equation, the lower line would look like this , the rural countryside lines are all 20kV lines, then when such a line reaches a local small village or town it goes into a transformer and outcomes 3 phase 0.4kV or 380v line with 3 wires and a neutral. so the last pole in rural areas from the transformer carries 4 wires , 3 phases and a neutral.

Some larger homes or farms have one small step down transformer themselves so the 20kV line is stepped down right at their doorstep and the incoming 380v 3 phase goes just a few meters in length.

I don't know the transformer you have there but do you have a three phase wiring coming into your house or single phase?

Single. The pole transformer nearest me has only two wires in and out.

 

[Salvador]

two wires out would mean single phase supply to your house , but two wires in is a bit strange to me since single phase is usually only 230 volts not higher.

Anyways so your house get supplied with single phase over rather old steel/aluminum wires is that correct?
In that case I wouldn't be surprised if you get 20 volt drop when connecting a 10kW load which is a rather big one.

 

[Guineafowl]

two wires out would mean single phase supply to your house , but two wires in is a bit strange to me since single phase is usually only 230 volts not higher.

Anyways so your house get supplied with single phase over rather old steel/aluminum wires is that correct?
In that case I wouldn't be surprised if you get 20 volt drop when connecting a 10kW load which is a rather big one.

Correct.

Does anyone here work for a power company in the UK? My house was derelict when I bought it, and the electricity had been disconnected. Before reconnecting, the supplier asked what type of supply I wanted, so I just asked for a standard household one, not really knowing any different. On the paperwork that came after the connection, I seem to remember seeing a reference to a 6 kW supply. I thought no more about it until now (this was 5 years ago).

If I were to complain to the supplier, do you think they'd object to the fact that I've had a 10 kW shower installed? It's been there for a good 4 years with no problems, so I'm tempted to let sleeping dogs lie.

 

[jim hardy]

...but if the voltage sags all the way to the nearest transformer for whatever reason then no noticeable extra heat would be felt or seen.

agreed. If that wire is short it will be visibly warmer on an IR camera. I was thinking of a loose terminal which would really show up .
Glad to hear you checked for hotspots inside, though.

There is a pole-mounted transformer on the fifth pole from me - I guess from P=I2R that most of the power loss and voltage drop occur between this and my house, given the lower efficiency of lower voltage lines. Would anyone know what sort of stepdown is occurring at this transformer?

That sounds like a LONG LONG run for the low side. I had a house with 200 ft run and the electric company was worried about "Flicker" so set me to a high tap. It caused lights to dim noticeably when airconditioner cycled (often in Florida) and short bulb life. After a year i got them to set me to a lower tap, bulb life improved and electric motors ran quieter. Dimming was less noticeable at lower voltage.

Stepdown ratio ? High side is probably between 4 and 24 kv
http://electrical-engineering-portal.com/primary-distribution-voltage-levels
the higher the voltage the taller the insulators.

If I were to complain to the supplier, do you think they'd object to the fact that I've had a 10 kW shower installed? It's been there for a good 4 years with no problems, so I'm tempted to let sleeping dogs lie.

Inquire. My utility will move your low side service wire underground just to avoid trouble calls from trees. If yours will do that you have it made.

What you really need though is the transformer closer to the house.
Here's a voltage drop calculator , estimate your length and try it
http://www.rapidtables.com/calc/wire/voltage-drop-calculator.htm

Trivia for the day - bulb life is inversely proportional to voltage^¹³. . 5% overvoltage shortens life by ~half .

 

[Salvador]

well maybe when they asked you about what kind of load will you run at your house they were thinking what power transformer to use for your house.
When you said a standard load they just installed a lower power transformer , probably, so it's normal that it sags under a high load.
I think you just need to calculate the extra expenses for redoing your connection vs the needs you have and then see whether you still need it/want it or no.

 

[Guineafowl]

Bulb life inv proportional to V^13? Wow. I assume (and hope) this doesn't apply to CFLs and LEDs.

Salvador - I think the old transformer was still there, just deactivated. Two years ago I had flickering lights/brownouts and they put in a new one, presumably the same again but newer.

I'll make some discreet enquiries - thanks to all for your help, by the way. Very interesting.

 

[anorlunda]

Trivia for the day - bulb life is inversely proportional to voltage^13

Wow. That's a bit of trivia I never heard before. It makes we want to relate a story I heard from a former boss. Sorry if this is a bit off-topic.

He was on duty and bored one night in the late 1940s operating the power grid in an Amazonian region of Brazil. He got curious about how load changes with voltage. So he started increasing the voltage in small steps, recording the power at each step with the intention of plotting a curve. When he got up to around 30% overvoltage, he was startled by an abrupt drop in load. Then he realized what he had done. He burned out all the light bulbs in the region. That was the end of his operator career..

 

[jim hardy]

Bulb life inv proportional to V^13? Wow. I assume (and hope) this doesn't apply to CFLs and LEDs.

Some places you'll see 12th power
point is, it's dramatic.
http://www.allegromicro.com/~/media/Files/Technical-Documents/an295012-A-Primer-On-Driving-Incandescent-Lamps.ashx

$\frac{250}{248}$^^3.5 = 1.028, 2.8% brightness change for 0.8% voltage change

 

[Tom.G]

HIJACK WARNING!

Guineafowl said: ↑
Bulb life inv proportional to V^13? Wow. I assume (and hope) this doesn't apply to CFLs and LEDs.
That's why some people would replace incandescents in difficult-to-reach places with 130V rather than standard 120V lamps; they last about 3 times as long.

Nop, doesn't apply to CFLs and LEDs. CFLs and LEDs are supplied with a constant-current source. The apartment building I'm in is fed with 3-phase power and we lost a phase about a year ago. I was on the failed phase. The voltage at the wall outlet measured 88V, versus the 120V nominal. The CFLs weren't affected at all, the one incandescent lamp was very dim, and the refrigerator kept running -- for a while. When the 'fridge cycled off it would not start again, only sit there and hum. To avoid burning up the motor, I shut off the breaker... and bought 10 pounds of ice. The flourescent lights in the elevator were flickering but the building manager wasn't concerned (knowledgeable?) enough to use the stairs! The older style flourescents in the elevator used a high impedance stepup transformer called a ballast. These may or may not ride out such a severe voltage drop. In this instance it was borderline.

 

[anorlunda]

The apartment building I'm in is fed with 3-phase power and we lost a phase about a year ago. I was on the failed phase. The voltage at the wall outlet measured 88V, versus the 120V nominal.

What you described can be very dangerous. I just searched and failed to find a story from pre-Internet days.

As I remember it: A single phase fault at night caused low voltage in a remote Inuit village in Canada.. The voltage was high enough to let the furnace oil pumps run in the furnaces, but not high enough for the igniters to light. The oil pooled all night long. In the morning, the fault was repaired, normal voltage restored, the igniters sparked and more than 300 simultaneous fires started in that poor village. Needless to say, that incident resulted in drastic revisions to the logic in oil furnace safety codes.

I regret not finding an authoritative source for that story. Perhaps one of our Canadian PF members can find it.