Power factor correction experiment

[T.M.Haja Sahib]

The instability problem exists exactly because the voltage is reduced and the motor is operating on the wrong side of the speed torque curve. Reducing the voltage further will not fix that.

The problem is solved because of the change in the shape of the torque-speed characteristic of the motor at low voltage.

Look at the motor characteristics I posted in #171. Anything below about 0.97 pu is in that unstable region. No amount of control can change that. It is what it is.

It is because the operating point is on the torque-speed characteristic corresponding to rated voltage or somewhat less.If it is on the torque-speed characteristic corresponding to very low voltage,it is longer a problem.It is witnessed by practical applications.

If it was, you would be getting people employing the very much simpler (W3C) soft-start power circuit rather than the double conversion arrangement of a variable frequency inverter for speed control. But they are not. Don't you wonder why?

Each type of controller has merits and demerits.One type of controller suitable for a particular application may not be suitable for another application.I made use of phase control drive of induction motor for application of proving my point that speed of induction motor can be varied greatly with reduction in voltage.

 

[gar]

111124-2359 EST

Back to my original post.

I am using a TED 1000 system for whole house monitoring. The commercial power factor correction capacitor for home use arrived yesterday.

Ran a quick GR 1650-A bridge measurement. There is a single capacitor in the box plus some transient limiting devices. The capacitor is placed across the 240 supply.

The capacitance is 70 ufd, and d = 0.008 @ 1 kHz. At 60 Hz the reactance is 38 ohms, and 240 current is 6.33 A.

On the TED system it seems to reduce total power by about 140 W at 1830 W or 7%, and earlier today at a higher power level about 4%. I haven't done any more direct bench testing. Is the effect real for an unknown reason, a TED problem, or something else. I will rerun my direct motor test, and the capacitor alone with the Kill-A-Watt. Then the same tests with the TED System in a standalone mode.

So at the moment a mystery.

.

 

[Besoeker]

The problem is solved because of the change in the shape of the torque-speed characteristic of the motor at low voltage.

It changes the operating point. Torque is approximately proportional to the square of the voltage. Reduce the voltage and you very soon end up on the wring side of the speed torque curve. A point to note is the slip at which you get to break down torque doesn't change much with voltage. My post #171 clearly indicates that to be around 0.97 pu speed at both 80% voltage and 100% voltage. The one you posted in #193 shows the breakdown torque is about the same at 50% voltage and 100% voltage. Incidentally, I think those curves (#193) are not representative of a real world motor, or at least none that I have encountered. They would give you a 4-pole motor running at 1350 rpm. Typically it would more likely be around 1485 rpm for a good modern machine.

I made use of phase control drive of induction motor for application of proving my point that speed of induction motor can be varied greatly with reduction in voltage.

I don't why you continue to insist that the speed of induction motor can be varied greatly with reduction in voltage.
It can't. And I suspect you know that.
Look back at YOUR OWN example in post #151.
The voltage is reduced to 1/sqrt(3) of rated voltage. The large voltage reduction gives a speed change of <1%

 

[mivey]

Hi mivey,
You appreciated Besoeker.But you did not tell even a single word of appreciation to me for my efforts, as self-assumed manager of this thread,in moving the thread in a specific direction.I feel let down.
Now coming back to your remark
''I'm not sure what else you could add,''
I want to say Besoeker as an expert has concealed some important information so as not to concede his defeat.I hereby expose that secret information to you.
The torque-speed characteristic of the induction motor displayed in the diagram of post#170 is based on what is called open loop control of induction motor.In that method,motor speed control is feasible within a very narrow range,say,1.2%.But there is another method called phase controlled induction motor drive with feedback control by which the induction motor can be run at any speed below its synchronous speed by reducing the supply voltage alone.
A block diagram of phase control drive is shown below.
http://www.st.com/internet/com/end_application/375.jsp

OK. I appreciate the info you posted, but I'm still waiting for the post that ties it all together into a valid argument. The final thrust that will silence Besoeker's arguments, so to speak. Unfortunately, it appears at the moment that Besoeker is the only one shooting the silver bullets.

You are definitely putting stuff out there, but I am not quite sure what. Trying to follow you is like herding cats.

I mean, we have come from a device on a residential service panel to VFDs for a residence?

 

[mivey]

111124-2359 EST

...On the TED system it seems to reduce total power by about 140 W at 1830 W or 7%, and earlier today at a higher power level about 4%.

Ooooh, the smoking gun! Next thing you know, you will be quoted in their sales literature.

 

[gar]

111126-2302 EST

mivey:

Unfortunately at the moment experiment is disputing theory. So why?

The PF capacitor is not changing the voltage waveform or the voltage very much. My nighttime residual load of around 1 kW is my most consistent load. There is some day to day temperature variation, but I don't believe enough to change the load much.

At the present $100 cost it would be useful if some others tried an actual experiment. Today I recorded the spinning disk readings at 0945 and 2330. The difference was 32 kWh. The TED system appears to be under reading with the capacitor present. In the morning I will get better evidence on this. Since I have started the whole house test I have not wanted to pull the capacitor from the main panel.

There has to be an explainable reason and I tend to suspect TED, and possibly a result of the use of current transformers. The Kill-A-Watt I suspect uses a resistor for a current shunt.

I hope those manufacturers don't read this and as you say use it for a testimonial. I like to discuss some things as I go along just to get comments. These are very crude experiments to just get a feel, but also to provide explanations for others that might run the same experiment.

I have seen an Internet comment by someone with a TED system that claims that they have seen a power (energy) saving with a PF capacitor in a residential application based on the TED system.

This I believe was on the TED Forum. I got banned from the TED Forum, probably because of probing questions, and I do not like their use of power line communication. They are very touchy about why PLC is the best route. I do not believe their patent claims read on the actual circuit they are selling. This would relate to how they describe the input signals. I do not believe you can put a patent number on a product that is not read on by that patent.

.

 

[T.M.Haja Sahib]

It changes the operating point. Torque is approximately proportional to the square of the voltage. Reduce the voltage and you very soon end up on the wring side of the speed torque curve. A point to note is the slip at which you get to break down torque doesn't change much with voltage. My post #171 clearly indicates that to be around 0.97 pu speed at both 80% voltage and 100% voltage. The one you posted in #193 shows the breakdown torque is about the same at 50% voltage and 100% voltage. Incidentally, I think those curves (#193) are not representative of a real world motor, or at least none that I have encountered. They would give you a 4-pole motor running at 1350 rpm. Typically it would more likely be around 1485 rpm for a good modern machine.



I don't why you continue to insist that the speed of induction motor can be varied greatly with reduction in voltage.
It can't. And I suspect you know that.
Look back at YOUR OWN example in post #151.
The voltage is reduced to 1/sqrt(3) of rated voltage. The large voltage reduction gives a speed change of <1%

Not only you but the three professors from Indian Institute of Technology,Chennai (see post 174) had it wrong presumably:You all assumed that motor operation is possible only in the conventional stable region of the motor characteristic.This is because the torque-speed of constant torque load is such that in the conventional unstable region of motor characteristic,it is either below or above the motor characteristic, as the voltage is reduced so that the motor accelerates to the operating point on the stable region of the motor characteristic or it stops altogether respectively.But in the case of fan type loads,the torque-speed characteristic of the load may be such that (i.e it satisfies the stability condition:its increase (decrease) in torque with speed is greater (less ) than that by motor) the operating point lies on the conventional unstable region of the motor characteristic i.e the conventional unstable region is actually stable operating region in this case.

 

[T.M.Haja Sahib]

111126-2302 EST

Unfortunately at the moment experiment is disputing theory. So why?

The PF capacitor is not changing the voltage waveform or the voltage very much.

On the contrary,it seems voltage change brought on by the capacitor is the only way to prove or disprove the energy saving due to connection of capacitor in this case.

 

[Besoeker]

Not only you but the three professors from Indian Institute of Technology,Chennai (see post 174) had it wrong presumably:

Those three are wrong, I'm wrong, your posts #151 and #193 are wrong. But you are right? Is that what you mean because that's how it comes across, intentionally or otherwise.

You all assumed that motor operation is possible only in the conventional stable region of the motor characteristic.This is because the torque-speed of constant torque load ....

Neither my #170 nor your #193 related to constant torque loads. Your post #151also mentions a pump and the context suggests that it is probably centrifugal, not constant torque.
Figure 27 from your post #27 has a centrifugal load curve. Actually, as far as I can recall, NO constant torque application has been presented here.

is such that in the conventional unstable region of motor characteristic,it is either below or above the motor characteristic, as the voltage is reduced so that the motor accelerates to the operating point on the stable region of the motor characteristic or it stops altogether respectively.But in the case of fan type loads,the torque-speed characteristic of the load may be such that (i.e it satisfies the stability condition:its increase (decrease) in torque with speed is greater (less ) than that by motor) the operating point lies on the conventional unstable region of the motor characteristic i.e the conventional unstable region is actually stable operating region in this case.

Think you have this vice versa.
An operating point will be reached where load torque and motor torque are equal. This is normally in the low slip region. If there is increased load, the motor slows down a little to a point where the motor torque matches the new load torque and all is fine and dandy and stable as it should be - and usually is.
There can also be anomalous situations where motor torque matches load torque on the "wrong" side of the speed torque curve and it just sits there at that speed. Bad things can happen and reduced voltage is not a viable control method to avoid this.

Look back the motor characteristics in my post #170.
First, note that the load is centrifugal, not constant torque.

All values in pu.
Look at the data for 0.6 speed. Typical fan and pump variable speed don't generally require a speed range wider than 0.6 to 1.0.
At 0.6 speed and rated voltage the motor torque is 0.84 and the load torque is about 0.25.
Under these operating conditions, motor current is about 6.24pu.
So, by reducing the motor voltage we could reduce the torque it develops by 0.25/0.84.
With me so far?

Motor torque is approximately proportional to the square of the voltage and current roughly proportional to it.
So, 0.55 pu voltage to provide the torque. And about 3.4pu current.
That is 3.4 times rated current.
Not a viable option.

 

[Besoeker]

On the contrary,it seems voltage change brought on by the capacitor is the only way to prove or disprove the energy saving due to connection of capacitor in this case.

A kWh meter is the way to do it.
The sales pitch for these derives is the money that you will/may/might/might not save on your bill. That, for most of us, is based on kWh.

 

[gar]

111127-1343 EST

Last nights experiments were with outside temperature in the 55 F range. Power company read 9.25 kWh from 2330 to 0810, and TED read 7.8 kWh. There is some estimate added in the TED value because its total resets at midnight. Capacitor still remains on the main panel with its TED system.

I have a second TED 1000 system setup for standalone measurements. It should be somewhat comparable with the one monitoring the main panel. I have 3 Kill-A-Watt EZ units and 2 standard units.

Bench tests today with TED 1000 and Kill-A-Watt EZ 110926, the same one as post #1. Used 90 ufd of post #1.

At 0 load both TED and EZ read 0 watts.


With a 800 W resistance heater, very small fan, and with and without 90 ufd capacitor:
Heater only TED reads 800 W, and EZ reads 789 W, 120.4 V.
Heat & cap TED reads 770 W, and EZ reads 795 W, 120.4 V.

90 ufd capacitor only as load:
TED reads 30 W, and EZ reads 6.2 W, 4.24 A, 527 VA, 0.01 PF, 124.8 V.

Post #1 motor (Montgomery Ward, 1/3 HP 115 V motor, 60 Hz, and 5.9 A), no mechanical load, and no capacitor:
TED reads 180 W, EZ reads 141 W, at 116.9 V.

Add 90 ufd capacitor in parallel with motor:
TED reads 150 W, EZ reads 140 W, at 117.0 V.

Clearly the phase angle of the load current has an adverse effect on the reading of the TED 1000 system.

Thus, I have to tell my readers that use of the TED 1000 system is not valid for evaluation of the effect of a power factor capacitor at the main panel. Further the 1000 has serious problems with good accuracy on inductive loads. But this does not mean the TED 1000 is not useful to help one reduce their electrical power consumption. An inaccurate but relative measurement may be useful, but not when evaluating the effect of a power factor correction capacitor at the main panel.

It should be noted that TED is using a Cirrus CS 5461 A power conversion chip. This is apparently one that was designed and has been used in commercial kWh meters.

So what are the commercial electronic kWh meters on the side of your house doing to prevent errors from non-zero phase shifts of the current being measured?

.

 

[topgone]

KILOWATTHOUR METERS

KILOWATTHOUR METERS

For all the investigations people do, those people at the utility worries more about these aspect of making sure the correct registration is achieved by their meters. It is their ultimate bottom line at stake. Having said that, the metering department calibrates each and every meter before sending those meters out. Being one those who worry about these things (in my other life before), the company reserves the right to reject or accept kWHr meters from all the different meter manufacturers. We sort of cook these meters at the lab for 24 hours and see if the registrations are within tolerable limits (this is after the as-found and as-left test runs; unity PF then 0.5 PF phantom load tests).

 

[T.M.Haja Sahib]

Look back the motor characteristics in my post #170.
First, note that the load is centrifugal, not constant torque.

All values in pu.
Look at the data for 0.6 speed. Typical fan and pump variable speed don't generally require a speed range wider than 0.6 to 1.0.
At 0.6 speed and rated voltage the motor torque is 0.84 and the load torque is about 0.25.
Under these operating conditions, motor current is about 6.24pu.
So, by reducing the motor voltage we could reduce the torque it develops by 0.25/0.84.
With me so far?

No.I won't.You are trying to mislead me.
See the underlined sentence above.The motor torque curve is above the load torque.So the current being taken by the motor i.e 6.24 p.u is not only for load but also for its acceleration purpose till it develops its maximum torque.So 6.24 p.u current value for load should not be used.

 

[Besoeker]

No.I won't.You are trying to mislead me.

Now why would I want to do that since you're doing such a fine job of it on your own?

See the underlined sentence above.The motor torque curve is above the load torque.So the current being taken by the motor i.e 6.24 p.u is not only for load but also for its acceleration purpose till it develops its maximum torque.So 6.24 p.u current value for load should not be used.

You really don't understand the curves.
The 6.24 pu current is what the motor will draw at rated voltage from the supply at 0.6 pu speed regardless of mechanical loading.
That's what the curve shows you. It is what it is. No ifs, no buts.

 

[T.M.Haja Sahib]

The 6.24 pu current is what the motor will draw at rated voltage from the supply at 0.6 pu speed regardless of mechanical loading.
That's what the curve shows you. It is what it is. No ifs, no buts.

That is the problem.The effect of motor acceleration on starting current of an induction motor can not be determined separately from the torque-speed characteristic of motor,I do not think your argument in post#209 is valid.You may try with a real life application,as you seem to be well-equipped:Take a ordinary ceiling fan.Without a voltage regulator,plot the torque-speed characteristic and with voltage regulator,repeat.You will see the fan operating on the 'unstable region' of motor characteristic with low speed.Does the current exceed the motor rated current?A multiple of times?

 

[Besoeker]

That is the problem.The effect of motor acceleration

Forget acceleration. The situation at 0.6pu speed with rated supply voltage applied. is that the motor will draw 6.24 pu current.
That's what the curve for current shows.

 

[T.M.Haja Sahib]

Forget acceleration. The situation at 0.6pu speed with rated supply voltage applied. is that the motor will draw 6.24 pu current.
That's what the curve for current shows.

What about the real life application mentioned there?

 

[Besoeker]

What about the real life application mentioned there?

The curves I gave you are for a real life motor on a centrifugal load.

 

[GeorgeB]

Take a ordinary ceiling fan.Without a voltage regulator,plot the torque-speed characteristic and with voltage regulator,repeat.You will see the fan operating on the 'unstable region' of motor characteristic with low speed.Does the current exceed the motor rated current?A multiple of times?

Shaded pole motors (all single phase AFAIK), have virtually nothing in common with industrial 3 phase squirrel cage induction motors. The certainly can have their speed controlled via voltage (phase control via a triac as one method) but don't confuse them with industrial motors. Their torque is voltage related ... speed is only by the approximately cubic increase in power required with a fan load.

UNLOADED, bare shaft, there is substantially no speed control by voltage change.

 

[Besoeker]

Shaded pole motors (all single phase AFAIK), have virtually nothing in common with industrial 3 phase squirrel cage induction motors.

Think it's called moving the goalposts. Maybe that's just a UK expression.
And thanks, BTW.