Paralleling Substation Transformers 1

[rockman7892]

I’m looking at an application where a local utility plans to add a second 20MVA transformer in parallel with an existing one in an existing substation. New transformer will be tapped from the existing incoming 138kv line with a new breaker and isolation switches added on high side. Secondary side will feed into an existing 13.8kv switcher lineup through an existing spare feeder breaker.

The secondary switchgear will have two transformers in parallel feeding the common bus without a bus tie and (4) load feeders connected to this common bus.

Each transformer will have its respective transformer diff protection with the mains and feeders in switcher having 50/51 protection.

Are there any issues with this application and is this something that is commonly done? Some questions that come to mind..

1) Increases short circuit level on secondary switchgear. Should not be an issue with (2) 20MVA units at 9.67%Z each proving aprox 20kA to 13.8kv gear.

2) Circulating currents....LTC’s on both units will need to have a master slave controller in order to avoid circulating currents?

3) Are there any potential issues with having a common bus with no bus tie? I’m used to switcher that has a bus tie and have not seen an application with two transformers feeding a common bus.

4) Other than the respective differential protection on each transformer is there any additional protection schemes that need to be considered for this application?

 

[Mbrooke]

1) Typically 30/40/50MVA units feed 13.8kv systems in the POCO world. I can't see two 20MVA units being much of a problem. Just make sure fault current on any down stream equipment is within the device ratings.

2)Yes, you will need an LTC paralleling control. There are some really good PDFs which I will upload.

3) None off the top of my head- but someone else may chime in.

4) You could add reverse power elements but from my vantage point I don't see the need as you described the setup.

 

[Mbrooke]

In terms of those parellel LTC, some good PDFs on the subject:

https://files.engineering.com/getfi...e=LTC_Transformers_and_Associated_Control.pdf

https://files.engineering.com/getfi...2&file=Paralleling_of_LTC_Transformers_#1.pdf

https://files.engineering.com/getfi...c&file=Paralleling_of_LTC_Transformers_#2.pdf

https://files.engineering.com/getfi...e&file=Paralleling_of_LTC_Transformers_#3.pdf

https://files.engineering.com/getfi...543c17&file=Time_Delay_of_LTC_VR_Controls.pdf

https://files.engineering.com/getfi...d8c38a60fc&file=Voltage_Reduction_Systems.pdf

 

[Mbrooke]

Let me know if the links work- there should be 6 PDFs

 

[crshears]

Ideal to ensure second 20 MVA trafo has identical characteristics so as to facilitate identical loadings and minimize circulating currents; the latter can be problematic and downright hinky if trafos with different tap counts are paralleled, or if they have different angular displacements per tap. At one location I recall, a pair of trafos ran on the same tap from tap 1 to 9, with a one-tap spread once the second trafo got to tap 10, with a two-tap spread once it got to 19, and a three-tap beyond tap 28; specific primary headboard selections had been chosen to achieve this, as any other options made for huge circulating currents...better not to go there at all if you can avoid it.

Because of wildly varying system requirements, more often than they cared for the operators of these trafos had to take them out of automatic tap control, re-balance their VAR flows by resetting the tap spread, and go back to auto; I personally have not seen any instances where automatic tapchanger controls have been programmed to mimic what these operators had to do manually. If you know of an instance where this has been implemented successfully, I wouldn't mind knowing where and how.

Indeed the pair of quadrature boosters / phase shifting transformers in the Ontario x New York tie lines at Cornwall-Massena for years had both different tap counts AND different phase angle displacements per tap, which made it impossible at times to get anywhere close to balancing the comparative loadings of the two lines. One end-of-life unit of the pair is presently being replaced; I have no knowledge of whether the replacement unit's characteristics are being chosen to match those of the remaining unit...

CR

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

 

[RRaghunath]

rockman7892,
I see two concerns with the proposed arrangement:
1) What happens when you have an uncleared outgoing feeder fault?
2) What happens if there is fault in one of the transformers?
In both the above cited cases, you will lose both the transformers. Is this OK?
If the load is within the capacity of one transformer, why not keep one transformer only connected to the switchboard and let the other transformer automatically get connected in case the subject transformer trips!
This is a popular scheme employed when the switchboard is designed with a single bus, without a tie breaker.
The automatic changeover always confirms that the transformer trip is not due to bus fault or uncleared downstream feeder fault.

 

[Palletjack]

Common bus...
What are your thoughts regarding maintenance?
If you need to do XF#2 testing, pick a low load time and transfer all loads to XF#1 and test XF#2, then Reverse to finish testing or maintenance.

 

[protoslash]

When you add a paralleled tx, you form a spot network.

A big protection consideration is backfeeding into primary fault. Your new station will be fed from two 138kV lines in parallel, if one of the 138kV faults, this 138KV source will trip. However the other healthy 138kV line will feed your substation bus and backfeed into the faulted 138kV line.

 

[stevenal]

I've never done the master-slave (client-server?) style, but you would need to deal with tie breaker open and transformer out of service situations, including protective trips. I have used the circulating current and delta var methods of paralleling, and they deal with all these circumstances nicely. A fair amount of equipment and wiring is involved. Lots of information available at [link Beckwithelectric.com]Beckwithelectric.com[/url].

 

[Palletjack]

I was just reading over the OP again..
One would think a station with that capacity you wouldn’t use an OLTC XF but rather feeder regulation.
That way the paralleling problems with an LTC are out of the way..
Different places, different ways...

 

[jghrist]

Is this and industrial or utility application? A lot of utility distribution equipment has a 10 kA fault limit. Like fuse cutouts and loadbreak elbows.

 

[davidbeach]

There's any number of configurations. I know of examples where up to four transformers run on a common LTC control scheme to keep them in step. One three transformer group staggers the step changes so that the feeders see three steps with 1/3 the magnitude as there would be with a single step by all three transformers together.

 

[bacon4life]

Like jghrist said, it may over-duty many pieces of downstream equipment.
Be sure your regulation and protection functions all work no matter whether one or two banks is in service.
Double check the rating of the spare feeder bay, as it is common to have a 2000A rated main and only 800A feeders.

 

[Palletjack]

Maybe I’m missing something here. I realize paralleling station transformers is done in different parts of the country, but I still can’t figure out why.

Put a bus tie switch in, break it into two stations. Lose only half your load rather than all of it should something happen in your 87 zone.
Maintenance is easier to do, since you can tie the bus together when needed and kill one XF for testing or maintenance. A shoulder month should be able to be fed from one.

Paralleling the XFs increases your fault amps, and stresses your breakers and OCRs.

So what are the real advantages of paralleling XFs in a station?[pre][/pre]

 

[waross]

You are missing the purpose of the tie breaker.
If one transformer is out of order, the tie breaker may be closed to pick up the load even though the remaining transformer may be temporarily overloaded.

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

 

[stevenal]

Both situations below assume one transformer can carry the entire load short term.

Independent transformers with open bus tie: In the case of a transformer fault, half the load is lost until bus tie can be closed.

Parallel transformers with closed bus tie: In the case of a transformer fault no load is lost and the remaining transformer continues to regulate.

 

[crshears]

If the downstream equipment was all purchased and installed in anticipation of the eventual twinning of the supply trafos, running with two in parallel should not be insurmountable.

If on the other hand the fault current infeed capability in the new configuration will far exceed the interrupting ratings of the existing equipment but you still wish to operate in parallel on the low side, there will be major capital replacement costs in your future...and if the owner won't spring for the cash you'll have no choice but to run with a normally open tie breaker and either some manner of ATS or, if the application doesn't warrant that, a manual transfer scheme.

CR

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

 

[ElectricRob]

Replying to Palletjack,

Splitting the bus and installing a tie breaker is probably the best option, but that adds to the cost of the project.
In 25 years, I've never submitted a budget that didn't get cut (nor have I ever seen a CFO take the blame for the resulting equipment failures).