Efficiency of Power Factor Correction Capacitors 1

[Dobber1978]

I have gotten involved in an energy study and the question about adding power factor correction to try and reduce line current came up. The client's design criteria currently states 0.9PF and we are looking at increasing this to 0.93PF.

As this is a energy study the question of how efficient the power factor correction equipment is has come up. I figure it is going to have some losses but I can't seem to find what these losses will be. I have sent a request to a power factor correction vendor to see what they say but would appreciate input from non-vendors as well.

My thoughts for losses
- copper losses, likely low if designed properly
- contactor losses, again should be low. Contactors are needed as this will be a switched set up providing about 2MVAR of correction with all stages online.
- internal resistance of the capacitor losses, no idea what this would be for large capacitors.

So is my thinking correct? Anything to add?

Thanks
Dob

 

[Skogsgurra]

Tan(delta) is always given for PF capacitors. And that is the internal losses for fundamental frequency.

Then you probably have some (or a lot of) harmonics which find their way through the capacitors. Easier if the harmonic number is high. And that produces heat in the ESR, which is very tightly coupled to the tan(deta).

In extreme cases, the harmonics will increase the skin effect in all parts and that increases losses.

Your local code and its comments probably contain a lot of relevant information.

Rules of thumb may work, but if you are dealing with the difference between PF .92 and .93, they may be insufficient.

Gunnar Englund

http://www.gke.org

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

 

[crshears]

Pertinent point: there may also be internal discharge resistors specified/provided to drain off any residual trapped charge when caps are switched o/s, the idea being to avoid overvoltages approaching 2.0 per unit across contacts if the cap is switched into service again before sufficient time has elapsed to dissipate such charges in the absence of draining resistors, the downside of course being that when the cap is in service, these resistors comprise a source of constant energy loss.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[waross]

Further to crshears comment here is an excerpt from the Canadian Electrical Code:
26-222 Drainage of stored charge of capacitors
(1) Capacitors shall be provided with a means of draining the stored charge.
(2) The draining means shall be such that the residual voltage will be reduced to 50 V or less after the capacitor
is disconnected from the source of supply
(a) within 1 min in the case of capacitors rated at 750 V or less; and
(b) within 5 min in the case of capacitors rated at more than 750 V.
(3) The discharge circuit shall be
(a) permanently connected to the terminals of the capacitor bank; or
(b) provided with automatic means of connecting it on removal of voltage from the line.
There are a few other exceptions which will not apply to switched capacitors.
Capacitor approvals are generally accepted by both the CEC and by the NEC so I expect you may find a similar rule in the NEC.
Do the math and allow a generous safety factor and you will have a reasonable estimate of the value of the internal discharge resistors.
Generally correction capacitor banks are quite warm to the touch but not painful. (Some electricians will drag their hand across a capacitor bank whenever they pass by. No or little heat indicates an open fuse.)
I would estimate the heating losses in 5 KVAR to 20 KVAR PF correction caps to be a few 10s of watts, not in the kiloWatt range.
When capacitors are used to eliminate PF penalties the payback is quite short.
We will be very interested in hearing what the payback interval is when caps are used to reduce line current I²R losses.


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

 

[waross]

I would set the power factor controller to correct to 100%
Most of the time the PF will be running at 100%.
Only when all equipment is running and the PF drops below 93% will you gain any benefit from extra caps.
That is when the controller is calling for maximum correction and there is not enough capacity available for full correction.
You may find that the actual saving is not worth the added cost of extra capacitors.
Rather than add another switched stage, I would add extra capacity to the first few stages. You may not need to add or change any contactors.

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

 

[HamburgerHelper]

Waross's discharge rates for cap bank resistors line up with what I see at utilities as well.

I would just add that where you put a capacitor bank inside of a plant can make a difference on payback. ETAP had a module to help with sizing and locating a capacitor bank and doing all the payback calculations for you. I haven't seen anyone use something like this. In plants, space is often too much of an issue to not just put it where it fits. Ideally, though you would like to put the capacitor bank next to whatever needs vars.

https://etap.com/product/optimal-capacitor-placement

Utility penalties for a poor power factor at the point of interconnect are often a driving force for installing capacitor banks. Losses will change roughly proportional to the square of your power factor. (0.9/0.93)^2 = 9.3% change in losses.


------------------------------------------------------------------------------------------
If you can't explain it to a six year old, you don't understand it yourself.

 

[bacon4life]

Since thin film capacitors have pretty low internal losses, the discharge resistors and the fusing makes up a significant portion of total losses on substation banks. Not sure if the fuses are as big a deal on low voltage banks.

How did your client come to choose 0.93 power factor? It sort of seems strange to introduce the hassle of installing a multistage switched bank to only go from 0.90 power factor to 0.93 power factor.

 

[LionelHutz]

To have any chance of success at lowering line losses enough to justify the cost, you'd want the capacitors as clos to the loads as possible, not at the plant distribution. You must have a very long feeder or customer owned transformers between the meter and the location you're planning to put this bank.

Still, I'm doubtful you could find enough savings to justify the cost. Capacitors save enough money to justify the expense by reducing utility power factor charges, not by reducing losses.

 

[waross]

Let me put this another way.
Plant PF generally varies with loading.
That's why we use PF controllers rather than bulk correction.
SET YOUR PF CONTROLLER TO 100%
That will give you the maximum saving most of the time without spending any money.
Monitor your sustem and determine how often and for how long the PF drops below 93%
I doubt that the potential saving will pay for your engineering time to date.

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

 

[veritas]

Considering the question of the OP, my gut feel says that the savings in reduced supply chain (network) losses due to the capbank will outweigh the losses of the bank itself. This of course presupposes that the bank has been optimally sized and located in the network as close as possible to the VAR load (usually at the MCC for LV industrial loads). Since the capbank draws leading current from the source and this counters the lagging current the source needs to supply to mainly motors, the overall effect is reduced current from the source which means reduced losses in the supply chain and of course the positive spin off of increased network capacity and voltage support.

 

[LionelHutz]

Oh, there could easily be some energy savings, but the energy savings alone won't give a reasonable payback payback on a capacitor bank.

I'm doubtful the energy savings would even justify adding extra capacitor bank steps to get from the 90% no penalty power factor penalty level to a 93% or 100% "extra energy saving" power factor level. But, possibly the 93% is being investigated because it's just energizing all the steps of the already proposed or installed bank. It does make sense to maximize the savings of the installed equipment.

 

[waross]

I am still having trouble understanding why a power factor controller is set at 93% rather than 100%
In the extremely unlikely case that the load and power factor are so stable that the controller is always maxed out at 93%, bulk correction would have been a lot less expensive that using a PF controller.


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

 

[LionelHutz]

It's rather trivial to size a bank to only be capable of 93% when the system is at full load, which can save a lot of money on capacitors compared to over sizing the bank to allow a 1.0 power factor at full load.

 

[waross]

Back to your original problem.
Look at you I²R losses in your feeders from the revenue meter to the PF correction connection.
Calculate how much the current will drop when you improve the PF to 93%. Calculate the savings per hour on the reduced I²R losses.
Estimate the hours per month that the PF would have actually be down to 90%.
Calculate the total savings per month.
Please share your findings with us.
I think that the actual dollar savings will be very small, and it is not worth looking further for the internal losses of the capacitors.
This plant may have a different PF profile than the plants that I have been familiar with and I am willing to be corrected.


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

 

[Dobber1978]

Thanks for all of the input.

If the client does want further study of this option I will post what the findings are in terms of potential power savings. At this point we have many opportunities and they will likely only select a few to study in depth.

Thanks again.