Reactive power compensation for a Control Power Transformer (CPT) 3

[bacon4life]

What are typical reactive excitation losses in a 480V/120V single phase 500 VA Control Power Transformer for a motor control center? Would 60 VAr be reasonable?

Can I put a 120V capacitor in the control panel to provide power factor correction? The MCC already has power factor correction downstream of the motor contactor, but when the motor does not run the power factor appears to be very poor.

 

[waross]

Measure the primary current with no load on the transformer. Multiply by 480 Volts or 277 volts depending on the primary connection. That will be your VA.
Measure the resistance of the primary.
Use I²R to calculate W.
Root [(VA)²) + (W²)] = VARs
Then calculate the VAR hours per month.
Now calculate the monthly saving on your power factor penalties that you will gain by correcting the power factor of a 500 VA transformer.
Now calculate the payback period for the capital cost of the capacitor and the cost of installation.
Add the cost of your time and look at the grand total, and the payback period.
Now destroy all of your calculations before the financial officer sees the payback period.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[bacon4life]

I started with the financial analysis first. Eliminating over $1000/month per site in power factor adjusted demand charges would certainly cover my time to look into this a bit more. Power factor charges come in all sorts of varieties, and this particular rate structure has extraordinarily large costs for loads with very low average power because the billing demand is the combination of peak demand and the average power factor:

Billing demand = peak demand / average power factor.

If this was a simple $/kvarh charge, Bill's comments would be spot on.

 

[davidbeach]

Perhaps a unity power factor rectifier and convert the controls to DC? If the rate structure is power factor, without regard for leading or lagging, you want perfection. Make that rectifier into a battery charger and you gain UPS functionality but might need to add undervoltage trips here and there since loss of motor power will no longer equal loss of control power.

What's the application, alternate station service for a plant outside your service territory?

 

[waross]

single phase 500 VA

What are typical reactive excitation losses in a 480V/120V single phase 500 VA Control Power Transformer for a motor control center? Would 60 VAr be reasonable?

That's 60VAR x 24Hr x 30Days / 1000 = 43.2 KVARHr.
Now look at your bill. Subtract 43.2 KVARh from the monthly total KVARHr and recompute the charges.
How much will be your monthly saving if you correct the PF for a 500 VA transformer in a plant paying over $1000 per month in penalties?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[waross]

Oh the answer to your other question:

Can I put a 120V capacitor in the control panel to provide power factor correction?

Yes. That will work.
I had a customer who had a PF penalty of about $4 per month. One estimate was about $1000, the other estimate was about $2000.
I added a small single phase capacitor hidden inside a panel. The capacitor was about the size of a run capacitor for a 1/2 HP motor. That added enough VARHrs per month to remove the penalty even though the service was three phase.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[davidbeach]

If it's the type of situation I'm picturing, he has an alternate station service from Podunk PUD for a power plant somewhere. Much of the time the load is the CPT and its really lower power factor. A very small number for the denominator. Occasionally they need to use the alternate station service source and run up a high peak demand, a large number for the numerator, but for not enough time to seriously affect the average PF. It might even be cheaper to stay on the alternate source enough of the time to push the average PF up to a reasonable value without changing the peak demand. That might bring the demand charges down enough to more than make up for the increased energy charges.

 

[krisys]

My friend bacon4life,

The term reactive power compensation is for the higher capacity systems. As a general practice, the 500VA control supply transformers will not require such compensation.

Are you talking some special case? Thumb rules/charts are available for the reactive power compensation of standard Motor Control Center or motors at low voltage.

All others are normally highly engineered solutions!

 

[jghrist]

this particular rate structure has extraordinarily large costs for loads with very low average power because the billing demand is the combination of peak demand and the average power factor:

Billing demand = peak demand / average power factor.

The reactive power losses on a 500 VA CPT will not significantly affect the average power factor of the plant unless the plant load is trivial, which would make the peak demand trivial.

 

[waross]

Do the math.
500 VA is barely enough power to make a decent cup of coffee.
If you are seeing $1000 a month in penalties it makes sense to correct the power factor. The payback on power factor correction is generally less than one year.
Instead of percentage PF, you should be looking at KVARHrs per month. You may use the PF meter to help determine the KVAR of a particular load, but the savings are in reducing the KVARHrs per month. If you do that the PF% will take care of itself.
There are two general approaches to PF correction.
1> The science of PF correction.
A power factor controller is installed to switch capacitors in and out as needed to keep the plant PF close to unity.
2> The art of PF correction.
Before the advent of electronic KWHr meters it was common to base PF penalties on the monthly average PF.
One meter measured the monthly KWHr consumption and possibly the KVA demand.
A second meter measured the monthly KVARHr consumption.
Root (KWHr² + KVARHr²) = KVAHr
Monthly KWHr / monthly KVAHr = Average %PF.
If the PF was below 90% the penalty would be 1% per % PF. eg: A PF of 89% would pay 11% penalty. A PF of 45% would pay a 55% penalty. There were other more expensive calculations used by some utilities.
The first step in the art of PF correction was to select a target average PF. It may be 95%.
The KWHrs would be used to calculate the allowable KVARHrs at 95% PF.
Actual KVARHrs minus calculated KVARHrs = KVARHrs of correction needed.
Step two; Bulk correction. Too much leading power factor will drive the voltage up and burn out the night lights when the plant is idle.
The next step was to estimate the KVAR required to correct the PF of all the transformers in the plant. It was often safe to double the correction on the transformers. If the transformers required 8 KVAR of correction it may be safe to connect 15 KVAR of correction, 24/7.
Then the largest motors that ran the most hours in the month would be corrected.
At each step, the monthly added KVARHrs were checked against the needed amount of correction.
Often the large motors would have double correction. The codes often prohibited directly connecting more KVAR than needed to correct the motor to unity, if the capacitor current passed through the O/L relay.
However the KVARs may be increased by bypassing the O/L relay. Sometimes by connecting ahead of the O/L relay at the motor starter, sometimes using a contactor switched with the motor to control the capacitor bank.
In a sawmill, most motors ran two shifts for 5 days a week. 10 KVAR x 8 Hr x 2 shifts x 5 days per week x 4 weeks x 1.08 correction to 30 days = 3456 KVARHrs per month.
The dry kiln fans generally run 25/7.
10KVAR X 24Hrs x 30Days = 7200 KVARHrs per month.
10 KVAR connected to the dry kiln fans is more than twice as effective as 10 KVAR connected to the head saw.
During times when the plant was idle, the PF may be leading. The meter may be running backwards and subtracting KVARHrs. Some utilities installed rachets on the KVARHr meter to prevent metering leading PF.
All part of the art of PF correction.
Don't be misled by high or low percentages until you calculate, percentage of how much.
Anecdote.
A trucker was overloaded by about 10,000 lbs at the government scales. He not only was fined hundreds af dollars, he had to send for a second truck to lighten his load.
The next morning he told his wife;
"I have to lighten my load. If you pack one sandwich instead of two sandwiches in my lunch today that will be a 50% reduction in weight."
The only difference the second day was that he was hungry as he waited for the second truck to arrive.
Can you change your 500 VA transformer for a 250 VA transformer and cut your percentage in half?



Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[waross]

Remember this is not a POWER transformer.
This is a 500 VA CONTROL transformer.
This will fit in a lunch box.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[bacon4life]

Yes, I know this is not a typical situation, which is why I am asking for any potential drawbacks for installing a capacitor on a controls transformer. Harmonic resonance and ferroresonance can be issues on utility scale capacitor banks, but I have not worked with low voltage capacitor banks before.

The application example is a pump that runs only during extreme weather events. If no extreme weather events occur, the pump is tested once per month for about 30 minutes. For a typical month:
Peak demand = 50 kW @ 0.95 pf
Monthly energy usage from load= 25 kWh + 15 kVArh
Monthly reactive energy usage from dual CPTs = 86 kVArh = 0.060 VAr * 24*30*2
Monthly average power factor = 0.27
Monthly demand adjustment= 50 kW *0.95/0.27 = 170 kVA
Monthly savings = $1088 = (170 kVA-50 kVA)* $9/kVA

In a month where the pump runs for many hours, the power factor correction capacitors attached to the terminals of motor improve average power factor to 0.95 so the demand charge is based on just 50 kVA.

@David- Yes, an alternative approach would be to consume more real power (i.e. test for longer duration or install strip heaters ) to improve power factor. I had been a bit worried about explaining this pf issue to the operations folks before posting this thread. Based on the responses so far from very knowledgeable EEs who understand pf correction, I am now terrified of trying to explain this solution the operations folks.

@Bill- Is control power transformer the wrong term? I assumed MCCs typically used something along the lines of the [URL unfurl="true"]http://ab.rockwellautomation.com/Power-Supplies/Control-Power-Transformer#selection[/url]

 

[wroggent]

How is the pump turned on? Perhaps the CPT could be disconnected when the pump is not in use?

 

[Skogsgurra]

What? Are you out of your mind? Compensate a 500 VA transformer? Forget it.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[davidbeach]

Come on guys, read the problem statement. The uncompensated CPT is not causing any electrical issues but, due to a perverse rate structure, it is a significant cost impact while not compensated.

Running a 400W heater, at 480V, looks like it would push your average PF to over 95%. What would the extra 288kWhr a month cost? At $0.10 per that would be $28.00 per month to save $1000 per month. That's an investment that most accountants would go for. Then in a few years somebody can try to figure out how to save the $28.00 per month being wasted by the heater and discover the $1000 bills again.

 

[waross]

Control Power Transformer is a valid term. I used the term Control Transformer to differentiate from the term Power Transformer.
Your power factor penalties are the most Draconian that I have ever seen.
At even $1.00 per kilo Watt hour you will be paying $1088 in penalties on $25 consumption.
By all means add a small capacitor to the transformer.
Just be sure that the control transformer is what is what is causing the penalties.
Given your usage profile I suspect that you are confusing demand charges with PF penalties.
Your calculation of demand charges is confusing.
This sort of thing didn't happen when the electro-mechanical meters measured demand directly.
Can you scan and post a copy of the power bill? Black out the names and account number.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[bacon4life]

@Waross-The utility billing mechanism for PF penalty is to adjust the demand charge, so it unclear what the proper term should be. The charges are strictly due to the billing mechanism of combining a peak demand register with an average power factor register; it does not matter whether the meter is mechanical or electronic. The attached image shows a before and after comparison for adding a 100 VAr compensation to one of the sites. There is also a demand ratchet clause, which makes the billing even more complicated. For this particular site the savings would have been about $300 per month.

@david- I don't want to encourage wasteful consumption or leave a trap for the future. That being said, the ideal heater sizing is highly variable both across dozens of installations and for any specific installation by time of year.

@Skogsgurra-Yes, I was out of my mind when I looked at the billing history for these sites. So far, compensating the CPT is least bad option I have come up with.

 

[jghrist]

The attached image shows a before and after comparison for adding a 100 VAr compensation to one of the sites.

It seems that your major problem is a high demand rate and a horrible load factor. For the months where there is load, the load factor is very low. For instance, in the 6th row, you have 57 kW of demand, but only 59 kWh of energy. This is an average load factor of 59/(57*24*30) = 0.084. With such a low load factor, a small fixed kvar compensation is large compared to the very low real energy. In this case, you could install a small capacitor anywhere (it doesn't have to be associate with a CPT) to compensate for the reactive load when load is very low.

 

[waross]

Wow! Talk about creative bookkeeping.
Customer;
"How much is 2 plus 2?"
Creative accountant;
"How much do you want it to be?"
Thank you for the very complete information.
Demand charges are peak charges and are reasonable charges. Peak demand is measured in KVA. The power factor is already included in the peak KVA.
It is not valid arithmetic to compute maximum demand from average values.
If the peak kW is used then the PF at the time of the peak must be used to get a valid value for peak KVA.

Explanation of demand charges:
Some months you only run a pump for 30 minutes and it takes 50 kW and 60 KVA. There are 720 hours in a month.
The utility must provide capacity to supply your load whenever you want it.
The energy charge is less than $1.00. That can not pay for the infrastructure. The utility must have the equipment available to supply 60 KVA whether you use it or not. Hence the demand charge based on maximum KVA demand.
Once you use a demand of 60 KVA, the utility wants to be paid for the lines and transformers that will supply the 60 KVA even if you don't use it. It is there in reserve whenever you need it.
Some demand ratchets reduce 10% each month. Some persist for a number of months. I have seen cases where the demand charges carried on for a year after a company went out of business and the power service was disconnected.

Step one: Add 100 VAR compensation to each site.
Step two: Have a lawyer and a math professor look at the tariffs. You may have been grossly over-billed. Even though the tariffs may allow the calculation the public utilities commission may not.
Demand charges are a justified charge, not a penalty.
PF charges are a penalty and may be onerous but not to the extent of $648 with consumption charges of $2.36
That's a penalty of over 27,000%
It looks like $6000 of overcharges for 19 months service.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Skogsgurra]

OK bacon, I understand now. I have never seen such a terrible load profile before. Combined with such harsh billing. Looks like there is some crop processing going on and that the plant is shut off most of the time. But you still need some auxiliary power to "bootstrap" the plant when needed. And probably also for alarms and supervision.
Solar comes to mind. A standard RV solar panel, a battery and a likewise standard sine inverter will not cost you much and will allow communication both ways, if needed, as well as "pressing the button" when you want to run the plant. But I do agree that a capacitor is simpler. Sorry for questioning your mental health.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.