Shunt Capacitor Placement

[johnnyball]

Does anyone have a good hand calculation method or methods for locating shunt capacitor banks on a three-phase 4 wire multigrounded distribution feeder, strictly from the standpoint of minimizing line losses?

 

[pwrtran]

IEEE 1306-1992 provides more detailed information on shunt capacitor application. A good practice is to place the capacitor closer to the large load so that it can provide reactive currentlocally not from the feeder source, thus, reduce the line losses. Another aspect you also need to consider is the switching open points. Capacitors should always stay with the large loads when doing the switching.

 

[RRaghunath]

scopidia,
I think you mean IEEE 1036!

Johnnyball,
If you are meaning compensating the VARs consumed by OHL, this may not be significant at the voltage level you have (3-phase, 4-wire LV system) and you can focus only on load end VAR compensation.

 

[johnnyball]

Thanks for the info. guys, but what i'm getting at is a way to minimize energy losses due to the reactive load. I'm looking for a proven method to optimally locate fixed and var switched capacitors without the luxury of having some of the newer modeling software that is available these days. I've heard of a method that places a value of capacitor at a point where the line kvars are one-half the value of the capacitor kvars. I understand that one needs a knowledge of the load cycle as well when combining fixed and switched capacitors. Can anyone elaborate on this, or are there any other methods for this? I'm interested strictly in line loss reduction and have no concern for the voltage aspect of the capacitors( this is already taken care of).

 

[magoo2]

The one-half rule means you locate the capacitor bank where the cumulative load is equal to or greater than one-half the bank rating. In other words, with a 600 kvar bank, you look for the location where the load meets or exceeds 300 kvar.

Because the load distribution on a given circuit is generally non-uniform, it is likely that you won't find a point that exactly meets the 300 kvar load. This is why you include the greater than as well as equal condition. It can be shown that you can just look at the reactive load and ignore the reak (kW) portion of the load in this effort.

Consider the peak load condition first and suppose you want to add 2 - 600 kvar banks. Pick the location where the kvar >= 300. Then you recalculate the reactive profile since every point in front of the cap bank will be reduced by 600 kvar. Repeat the process to find the second location.

This works fine for a peak load condition, but it generally will put the banks too far out on the circuit and the losses will be higher for any lower load condition.

And this doesn't address your concern about load cycle and switched versus fixed banks.

To satisfy the load cycle over the year, you really should be applying the capacitors for the average reactive load situation. If you define a reactive load factor (RLF) = average reactive load/peak reactive load, then you modify the one-half rule to say that you put the capacitor bank where the average reactive load equals or exceeds one-half the bank rating. This works best in terms of energy savings over the year.

This is frequently converted to the rule that says you apply the capacitor bank where the peak reactive load/RLF >= one-half the bank rating.

This rule works for both fixed and switched capacitor banks.

 

[magoo2]

I didn't mean to say this:

This is frequently converted to the rule that says you apply the capacitor bank where the peak reactive load/RLF >= one-half the bank rating.

What I should have said is:
apply the capacitor bank where the peak reactive load >= one-half the bank rating/RLF.

So if RLF =1, you have the rule for minimizing losses under the peak load situation. For the energy situation, where RLF <1, this rule results in locating the same size cap banks closer to the source end of the circuit.

 

[pwrtran]

johnnyball

You can determine the cap bank locations without going through excessive calculation.

1) improve PF - put the bank at the station
2) voltage support - close to the end of the line, or approximate 2/3 of the length from the station
3) reduce line losses/relief some feeder capacity - close to the the large Var customers

I would like to first find out where the load center(s) and open points are, then place cap banks accordingly. You don't want to see your cap banks switched away from the large var customers to an alternate feeder that has its own banks or doesn't need extra vars.

 

[johnnyball]

Interesting facts guys. This was the information that I was after.

Have any of you kept up with the loss reductions possible using capacitors? Our company, a rural municipal, was having line losses anywhere from 7.5 to 8.5 %. We reworked some of the existing cap installations, and our losses dropped to approx. 5%. This continued for a year or more until our losses gradually crept back up to 7%, where it remains steady. We researched virtually everything, but could not find any one cause for the change in loss. The system was basically unchanged for the time period involved. The loss amounts to considerable money. On our system it is about $100 000 for every percent of loss.

 

[johnnyball]

I found a really good paper on this topic on the EPRI website. Look under project ID # 1010655.

 

[KVX10]

Just another idea...
Do you have a SCADA system with a load management option? If so, you can use the already installed SCADA system using load management switches that control oil switches to open or close in the capacitors.

 

[jghrist]

For uniformly distributed loads, locate capacitors 2/3 of the distance from the substation to minimize losses.

For uniformly decreasing distributed loads, locate capacitors 1/2 of the distance from the substation.

If the load is concentrated, locate the capacitors near the load.

For loads that are none of the above, use engineering judgement to place the capacitors with the above rules of thumb in mind. I have found, from varying capacitor locations in a load flow program, that the difference in losses for large variations from the ideal location are small.

 

[magoo2]

For uniformly distributed loads, locate a 2/3 capacitor 2/3 of the distance from the substation to minimize losses. This rule only applies for a 2/3 capacitor and is frequently misquoted. This is also a special case of the 1/2 rule.

With a uniformly distributed load, the load at any point along the feeder equals 1 - x, where x = 0 corresponds to the substation and x = 1 corresponds to the end of the circuit. Try to apply a 2/3 capacitor. 1/2 Rule says you appy it where the load equals 1/3. At x =2/3, the load equals 1/3.

Could consider smaller banks like 1/3. 1/2 Rule says to apply it where the load equals 1/6. Works out to be x = 5/5 for this size.

My point is mentioning this is that the 2/3 rule is a very restrictive case that applies for uniformly distributed loads. The 1/3 Rule is more general and more accurate.

From a high level, you should apply capacitors where you have significant loads and this accomplishes this quite well.

 

[jghrist]

For sizing capacitor banks, I would determine the minimum var load of the circuit and install fixed banks to correct for this. Loss reduction is proportional to the average var load.

My experience is that switched capacitors cannot be justified by loss reduction because of the cost of the switches and controls and the fact that they are not on line all of the time. Switched capacitors can be justified if needed:

1. To meet power factor requirements at peak load to avoid penalties.
2. To reduce line amps and avoid replacing overloaded conductors or equipment.
3. To improve voltage.
4. Possibly to reduce peak losses if the demand cost of power is high.