Capacitor Banks

[Fitzdrew516]

Hey guys -

I got my degree in Power Systems Engineering so I have done plenty of power factor calculations back in the day. I don't remember all the steps exactly, but I still have all my books and notes so I feel pretty confident I can figure out all the calculations again.

Almost all of the work I do is on the industrial side of things so I see (from utility bills) some low(ish) power factors fairly often. So my question is - How are these capacitor banks installed in a real world application? If I do the calculations and determine I need a "x number" farad capacitor to correct the PF then what exactly needs to be specified and how are these connected from a real world standpoint?

FYI, to keep it simple I would go about it by correcting it at the switchgear instead of at the individual motors. So I assume I would just connect a capacitor bank to the switchgear in some way, but I guess I'm just unsure exactly how it's done or what typically gets specified. Anyway - I'm rambling so if anyone with some experience in this area could make it a little bit clearer I'd appreciate it. It's an option I'd like to be able to confidently offer my clients.

Thanks,
-Drew

 

[templdl]

In my opinion a pf is at given point in time. By selecting a given capacitor value for the pf my concern would be the risk of ending up with a leading pf at times. So either your calculations are taking this into consideration or that you have the capability to switch the capacitor bank in and out automatically to prevent a leading pf. In the switching process there is a risk of a voltage spike which masy not be desirable. I have found this to be quite troublesome with nuisance tripping the older technology breaker peak sending solid state trip units that are sensitive to current spick as a result of the voltage see pick. So watch for those little gremlins that can pop up.

 

[Fitzdrew516]

In my opinion a pf is at given point in time. By selecting a given capacitor value for the pf my concern would be the risk of ending up with a leading pf at times. So either your calculations are taking this into consideration or that you have the capability to switch the capacitor bank in and out automatically to prevent a leading pf. In the switching process there is a risk of a voltage spike which masy not be desirable. I have found this to be quite troublesome with nuisance tripping the older technology breaker peak sending solid state trip units that are sensitive to current spick as a result of the voltage see pick. So watch for those little gremlins that can pop up.

Thanks for the info! I typically gather a year of utility bills so that gives me a good idea of the average PF for the facility in question. Then my idea is that when doing the calculation my aim wouldn't be for unity, but rather a PF of maybe .93 or something like that to allow some fudge factor to eliminate the possibility of a leading PF.
If I didn't want to this I believe Square D makes a pretty decent system that automatically adjust per measured power factor/inductance. Not 100% sure about that though.

 

[GoldDigger]

The other approach to take if you have a few large lightly loaded motors is to place the capacitors after the motor contactor so that they will only be online when the motors are running.
If the low PF is largely the result of uncompensated non linear loads (distortion PF) then capacitors aline may not be a good solution.

 

[Fitzdrew516]

The other approach to take if you have a few large lightly loaded motors is to place the capacitors after the motor contactor so that they will only be online when the motors are running.
If the low PF is largely the result of uncompensated non linear loads (distortion PF) then capacitors aline may not be a good solution.

For sure! Thanks for the input, GoldDigger. Usually the facilities I work on have a lot of smaller motors, not just a few culprits; that's why I think I would stick towards correcting it at the MSB.

 

[Iron_Ben]

The other approach to take if you have a few large lightly loaded motors is to place the capacitors after the motor contactor so that they will only be online when the motors are running.
If the low PF is largely the result of uncompensated non linear loads (distortion PF) then capacitors aline may not be a good solution.

I think that Golddigger's point about attacking the large, lightly loaded motors is a good one since they are so contributory to low power factor.

I want to note that pf penalty varies greatly from utility to utility. At the POCO where I worked, the penalty kicked in when pf got below 0.9. And the lower it was, the more the penalty. BUT, our penalty was laughably small. I never once saw a case where the payback from installing capacitors made economic sense. We served some rice mills and sugar mills that would go as low as 0.65. The penalty was only about 1.5% of the total bill. Not a good deal for us, but the terms of the penalty were set not by us but by the Public Service Commission so there was effectively nothing we could do but grin and bear it. Other utilities I know of (different states) were allowed to change more.

 

[Phil Corso]

FitzDrew...

If you need the simple calc, it is:

If the original apparent load, kVAo, the original power factor, PFo, and the required power factor, PFr, are known, then the formula for capacitive reactive power, neglecting capacitor loss, is:

kVAc = K x P, where, K = tan [ Acos ( PFo ) ] - tan [ Acos ( PFr ) ], and, P = kVAo x PFo, in kW.

Regards, Phil Corso

 

[Phil Corso]

FitzDrew... correction:

If the original apparent load, kVAo, the original power factor, PFo, and the required power factor, PFr, are known, then the formula for capacitive reactive power, kVAc, neglecting capacitor loss, is:

kVAc = K x P, where, K = tan [ Acos ( PFo ) ] - tan [ Acos ( PFr ) ], and, P = kVAo x PFo, in kW.

Phil

 

[Ingenieur]

Hey guys -

I got my degree in Power Systems Engineering so I have done plenty of power factor calculations back in the day. I don't remember all the steps exactly, but I still have all my books and notes so I feel pretty confident I can figure out all the calculations again.

Almost all of the work I do is on the industrial side of things so I see (from utility bills) some low(ish) power factors fairly often. So my question is - How are these capacitor banks installed in a real world application? If I do the calculations and determine I need a "x number" farad capacitor to correct the PF then what exactly needs to be specified and how are these connected from a real world standpoint?

FYI, to keep it simple I would go about it by correcting it at the switchgear instead of at the individual motors. So I assume I would just connect a capacitor bank to the switchgear in some way, but I guess I'm just unsure exactly how it's done or what typically gets specified. Anyway - I'm rambling so if anyone with some experience in this area could make it a little bit clearer I'd appreciate it. It's an option I'd like to be able to confidently offer my clients.

Thanks,
-Drew

the caps are usually rated in kvar
do a power triangle for the existing
do one for where you'ld like to be (S and Q change, P is constant)
the difference in Q is cap size in var

as others said
get good data, i, v and p
make sure motors are sized appropriately, lightly loaded is a common mistake
is ther a cost benefit? Do an analysis
operational benefit, ie, lower current
you may want to stage/sequence them with controls based on i, var, pf, etc
also over v limits
there are standard packages google s & s power factor control

location is an engineering decision
local to loads/central location both have pro's and cons
I lean towards central on a main switchgear cb
but I've used both

a good primer
http://www.eaton.com/ecm/groups/public/@pub/@electrical/documents/content/sa02607001e.pdf

 

[Fitzdrew516]

FitzDrew... correction:

If the original apparent load, kVAo, the original power factor, PFo, and the required power factor, PFr, are known, then the formula for capacitive reactive power, kVAc, neglecting capacitor loss, is:

kVAc = K x P, where, K = tan [ Acos ( PFo ) ] - tan [ Acos ( PFr ) ], and, P = kVAo x PFo, in kW.

Phil

Thanks Phil!

 

[Fitzdrew516]

the caps are usually rated in kvar
do a power triangle for the existing
do one for where you'ld like to be (S and Q change, P is constant)
the difference in Q is cap size in var

as others said
get good data, i, v and p
make sure motors are sized appropriately, lightly loaded is a common mistake
is ther a cost benefit? Do an analysis
operational benefit, ie, lower current
you may want to stage/sequence them with controls based on i, var, pf, etc
also over v limits
there are standard packages google s & s power factor control

location is an engineering decision
local to loads/central location both have pro's and cons
I lean towards central on a main switchgear cb
but I've used both

a good primer
http://www.eaton.com/ecm/groups/public/@pub/@electrical/documents/content/sa02607001e.pdf

Dang. That document has a lot of good info in it. I'm printing it out now for a closer read. I definitely would do a cost benefit/ROI analysis on this before presenting it to the client; I nerd out over that stuff.
I have a good understanding of the calculation aspect of the design based on stable and consistent metering data, but more of what I'm interested in is how it is usually specified and shown on drawings if I were to design it, and how exactly would a capacitor bank get hooked up to the MSB (assuming that's the route I would choose)? How would this kind of set up look on a single line?
I googled "s & s power factor control", but the results weren't super clear; I'm unsure if the results were returning the info you intended me to see. Provide a link if you're referring to something specific and I'll check it for sure.

Thanks,
-Drew

 

[Fitzdrew516]

I think that Golddigger's point about attacking the large, lightly loaded motors is a good one since they are so contributory to low power factor.

I want to note that pf penalty varies greatly from utility to utility. At the POCO where I worked, the penalty kicked in when pf got below 0.9. And the lower it was, the more the penalty. BUT, our penalty was laughably small. I never once saw a case where the payback from installing capacitors made economic sense. We served some rice mills and sugar mills that would go as low as 0.65. The penalty was only about 1.5% of the total bill. Not a good deal for us, but the terms of the penalty were set not by us but by the Public Service Commission so there was effectively nothing we could do but grin and bear it. Other utilities I know of (different states) were allowed to change more.

I've noticed it greatly varies between areas and POCOs. Sometimes I see no penalty and other times I've seen it add a good 25% to the bill ( I think that job was a California job - go figure). However, what you can always depend on is a very cryptic utility bill because they don't want the average customer to know how bad they're getting screwed over .

 

[Ingenieur]

Dang. That document has a lot of good info in it. I'm printing it out now for a closer read. I definitely would do a cost benefit/ROI analysis on this before presenting it to the client; I nerd out over that stuff.
I have a good understanding of the calculation aspect of the design based on stable and consistent metering data, but more of what I'm interested in is how it is usually specified and shown on drawings if I were to design it, and how exactly would a capacitor bank get hooked up to the MSB (assuming that's the route I would choose)? How would this kind of set up look on a single line?
I googled "s & s power factor control", but the results weren't super clear; I'm unsure if the results were returning the info you intended me to see. Provide a link if you're referring to something specific and I'll check it for sure.

Thanks,
-Drew

sorry s & c power factor or intellicap

 

[vicdog]

One connection option is to attach to a spare breaker in the switchgear. Properly sized of course. If automatic bank, controller CT would need to reference total current drawn by the switchgear. Sometimes you can scab the power meter CT. Banks are usually referenced in kVAR. harmonics is a big consideration so a power quality recording session should be done.

 

[topgone]

One connection option is to attach to a spare breaker in the switchgear. Properly sized of course. If automatic bank, controller CT would need to reference total current drawn by the switchgear. Sometimes you can scab the power meter CT. Banks are usually referenced in kVAR. harmonics is a big consideration so a power quality recording session should be done.

To add, the kVAR steps should properly be chosen else the automatic cap bank hunts and contactors will be very noisily cutting in and out.

 

[Sahib]

The sequence 1,2,4,8,16,32..... is interesting in that any number less than its sum can be represented as a sum of its numbers For example take first three members 1,2,4 from the sequence. Its sum equal to 7. Any number up to 7 can be represented using 1,2,4.By choosing Kva of capacitors in the sequence, hunting may be minimised.

 

[Ingenieur]

A 2 stage 1/3 and 2/3 arrangement is common
low load 1/3
medium load 2/3
full load 1/3 + 2/3 = full rating