PFC Cap Bank Repair vs Replace Economics 1

[jktwn]

We currently have 10 200KVAR capacitor banks scattered around the plant on the 480V side that have failed in the past 25+ years and were not repaired. At the plant service entrance there is a 2250KVAR bank on the 12.5KV side that has been maintained and is working.

Our power factor runs at 0.8 lagging and we need to increase to 0.9 minimum to avoid rate penalties beginning late next year.

Not knowing the failure mode of the existing units, the question is whether it is more economical to purchase additional KVARs on the 12.5KV line or repair the old units in the plant and maybe buy a smaller PFC bank to go on the 12.5 side.

Calculations indicate we would need an additional 2500KVAR of PFCs to achieve 0.9 PF.

Any suggestions would be appreciated.

 

[wareagle]

Normally you install caps at the load to reduce the demand and also to reduce line loss and improve voltage drop. That said if these are not a major problem I would rather install the caps at the service. If you are being billed a "Demand Charge" let me also recommend that you don't stop at the 0.90 pf but add a few more and go to 0.99 or 1.0. You can do the math and see what the rate of return is. I think you will find it is worth the extra expense. What I have seen is the return is 1 or 2 years and gravy after.

 

[digitrex]

I believe high voltage capacitors are more expensive. If you need to add HV switchgear for additional HV capacitors, the cost will be even higher.

I had some bad experiences with LV capacitors with voltage rating equals to the service voltage and many of them failed. I changed to higher voltage rating (125%) and they are OK for almost 10 years so far. When you select higher voltage capacitors, you only need to re-calculate the equivalent kVAR at the service voltage.

As wareagle said, it is normally recommended to install the capacitors as near to the load as possible.

If the capacitors had been 25+ years old, I suppose your PF controller/regulator should be very old as well. Why not replace the whole capacitor banks instead of repair?

 

[Skogsgurra]

I think that you also will have to consider the risk of resonance between the load side capacitors and the reactance in the transformer.

This was not a big problem 10+ years ago. But as use of VSDs has increased lately, it is more and more of a concern. The reason why your load side capacitors are non-functional may be because of resonance, which increases voltage and current in the capacitor until an internal or external fuse blows. The resonant frequencies are often in the 250 - 780 Hz range (5th - 13th harmonic in 50 - 60 Hz mains frequency).

I have had several such cases where the installation had to be changed so that all compensation was made on the medium voltage (3 - 11 kV) side. Even if you do not have much VFDs today, you will surely have those within a few years and then your low voltage compensation will be at risk. Other loads that cause harmonics and risk for resonance are controlled infrared heating where the phase control injects lots of harmonics.

There are sites where you can read about this.

http://www.ecmweb.com/powerquality/electric_power_factor_correction_3/

is one and such companies as Siemens and ABB have lots of application information.

Gunnar Englund

http://www.gke.org

 

[keewee]

Wxashcra was the 2250 kVAr correction bank on the 12.5 kV supply installed because the 480 V correction wasn't working ?

Now in New Zealand the power companies are scaling back their power factor correction efforts and requiring the user to correct their own power.

Wareagle if you have a system of acheiving a 1.0 power factor then you better get a patent real quick ! :]

Wxashcra I think Skogsgurra is right and you should investigate why the 480 V capacitor banks have failed before you replace them. Is there any resonance filters ?

 

[jktwn]

Keewee - The 2250 bank was installed back in the late 80s because our local utility was threatening to impose pf penalties, however, they never got around to enforcing them. We installed the PFC bank at that time. The LV banks were installed in the late 70s and early 80s as the plant was being built.

We are planning on commissioning a power quality study to check for harmonics on the LV side. Quite a few VFC drives have been installed recently and there is some concern that they may be impacting PQ in the plant. The resonance issue has me a little worried - I have heard horror stories of major power distribution system damage caused by some well-intentioned EE's misapplication of PFC.

The reason for even looking into the issue of repairing the existing 480V PFC capacitors is that if we can get them back into operation economically that should get us a good way down the road to 0.9 pf and no penalties.

I agree with Skogsgurra that we need to understand the underlying cause of the cap failures before simply fixing and turning them back on. We have an abundance of older style DC SCR drives up to 1000hp, lots of ac motors, and quite a few VFDs.

Digitrex do you remember the failure mode of the LV caps that were rated at line voltage?

thanks,

 

[digitrex]

wxashcra,
Our old capacitors failed quite long time ago. We didn't investigate into the details. I only remember each capacitor unit had several small capacitors connected in parallel. Some of them failed (burnt) and it didn't cause any tripping until we found the failure after measuring unbalance capacitances around the delta connection. May be those capacitor banks were poor quality.

 

[dpc]

I agree with skogsgurra that placing capacitors close to the motors, as we have done in the past, is rapidly becoming just that, a thing of the past.

With the increased use of adjustable frequency drives, these local capacitors are being subjected to levels of harmonic currents that they were never intended to handle. It is common to go into an old facility and find a lot of caps with blown fuses or bulging cans.

Most new industrial installations are using MV capacitor banks with automatic switching and appropriate harmonic filtering.

 

[wareagle]

Keewee
"Wareagle if you have a system of acheiving a 1.0 power factor then you better get a patent real quick.

I think you know how. We don't have all the information but add caps until you get the PF you want. Done it a number of times and saved a lot of money. I know 12 kv caps come in 300, 600, 900 and 1200 kvar banks.

wxashcra
What is your demand at 0.80 pf?

 

[keewee]

Sorry Wareagle I was having a pedantic laugh there. A 1.0 P.F. is theoretically impossible and we call it "unity" ( is it called unity in The States ? ). You're absolutely right about adding caps until the desired PF is achieved. What PF is required by your PoCo's. Most of ours require 0.95.

Wxashcra, why was the 12.5kV correction installed if there was already correction on the 415V side ? Was the correction on the 415V side not enough ? The reason I ask is in light of DPCs comments on the move away from localised correction capacitors. Is this because smaller caps at the motors are more vulnerable than larger caps at the supply switchgear ? So is it better to correct PF at the supply ?

 

[jktwn]

Wareagle, Overall demand is just above 20MVA and we operate 24/7. Our demand is close to flat which makes our utility company happy.

Keewee, I wasn't involved in the decision, but back then we had a very capable retired power company consultant in the plant and his recommendations were acted on without much questioning.

Current management likes the idea of simply adding another PFC bank on top of the existing one on the 12.5kv side since it would seem to be a simple approach to solving our low pf problem.

It's sounding like the concensus is to go with PFC on the 12.5kv at the incoming swithgear. The only problem for us with that approach is we will have to add another switch which means a plant shutdown for a day.

I really appreciate the feedback - there's a lot of great experience and knowledge out there.

 

[wareagle]

wxashcra
Using your information of 20000 kva and 0.8 pf your net kvar = 12000 kvar. Using a $5.00/kva charge your demand charge is about $100,000 per month. Improving the pf to 0.90 with an addition of 4250 kvar the net demand is 17778 kva and demand charge of $88890 per month. If you choose to go further and improve the pf to 0.99 by adding 9725 kvar, the demand result is 16161 kva and a charge of $80805. So improving the pf from 0.90 to 0.99
reduces your charge from $88890 to $80850 or a saving of
about $8000 per month or $96000 per year. That a lot of money to pocket for years to come.

Keewee
The utility in my area has only one rate for customers this size. The demand charge is about $5.00 per kva and the customer is not required to correct to any specific
pf. They do so to reduce the monthly demand charge.

 

[dpc]

I recommend that you do some harmonic measurements or analysis of the existing system and specify appropiate harmonic filters on the new cap bank. The cap bank suppliers can help you with this.

For keewee:

"A 1.0 P.F. is theoretically impossible and we call it "unity" ( is it called unity in The States ? ). You're absolutely right about adding caps until the desired PF is achieved. What PF is required by your PoCo's. Most of ours require 0.95."

Achieving a 1.0 pf is not theoretically impossible, just difficult due to variable loads and varying pf as load changes. Power factor requirements are being tightened here in the US. Bonneville Power Administration (power wholesaler) penalizes for pf below 0.97 and retail utilities are passing this requirement along to end users.

Putting the caps at medium-voltage level provides better isolation from the harmonic distortion caused by the 480V adjustable freq drives. And since the cap bank is lumped in one spot, it makes it possible to apply harmonic filtering more economically.

 

[LionelHutz]

I'll just point out that wareagle has calculated, you require 4250kVAR and fixing your 200kVAR caps gets you 2000kVAR. So, even if you fix the caps you'll still 2250kVAR short of achieving 0.9pf. And, as already mentioned I'd go even larger than this so that you'll still reach 0.9pf when you add another load to your system.

It's not a real great idea to mix capacitors in a system like this without doing some studies or measurements to ensure a problem won't occur. It's actually not a good idea to install any capacitors without studying the possible harmonic effects.

A 1.0 P.F. is theoretically impossible and we call it "unity"

I don't understand this. It's impossible in practice to correct to exactly 1.0pf but it's not theoretically impossible. You can overcorrect from a negative power factor to a positive power factor so it's not like reaching a 1.0pf is impossible either.

I just talked to a customer today where the 1800kVAR bank we supplied for motor power factor correction is trapping some harmonics from a large cycloconverter. The overcurrent relay is picking up for a second every so often as the cycloconverter operates. They didn't want to look at their power system to determine if harmonics would be an issue. I stuck enough protection on the bank so it should be ok. Good thing I'm using 5610VAC caps on a 4160VAC system.