Paralleling 1

[Guardiano]

Hi,

I have on site three cat generators, 500 kVA capacity and generating at 400 V 50 Hz.Two of the sets are identical,both engine and generators,with VR6 HMI.The third one has got a VR3 HMI.The sets generate on a common bus and the 400 V is fed to a 5500 V station through a 400/5500 V transformer.When paralleling, each set behaves independently and someone is required permanently to adjust the voltage on the generators.The support is very poor where I am and I would like to make the right decision.I have been requested to purchase additional cards to effect the laod sharing. I would really appreciate comments from forum members.

Guardiano

 

[Skogsgurra]

Stand by for Bill (waross). He knows all about it.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[itsmoked]

Or Catserve.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

Catserveng has me on Cat specific problems. Let's wait for him to log in.

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

 

[rbulsara]

Yes, waiting for catserveng is worthwhile.

For all I know you may just have a connection or setting issue. However, VR3 and VR6 are now archaic. I know you are somewhere in no man's land but it may be time to make a thorough upgrade to the system. Provide consistent and modern AVR,LSM and governors for all units. Even that method is becoming obsolete and all microprocessor based controls are becoming more prevalent.

The "lack of service" is only going to get worse with time as modern parts are electronic and even require laptop to program! Parts are no longer field repairable. This is not an issue where you can get parts easily but would be an issue for you.

Rafiq Bulsara

http://www.srengineersct.com

 

[Guardiano]

Thanks all for the replies. I hope catserveng would pop in very soon.
I have also noted something which is worth commenting I believe. The magnetising current on both sets at noload is around 275 A which I think is quite high for a 500 kVA set.

Guardiano

 

[ScottyUK]

High magnetising current? Sounds like your machine is in voltage control. With a synchronous machine you should be able to float VAR-neutral on the system, but if the machine is fighting the grid trying to regulate voltage then you will see some large reactive current flows.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[rasevskii]

Guardiano-

If you mean magnetizing current as the stator current at no-load then there are indeed vars flowing between the units, also known as circulating current. That means the cross-current compensation is not working or is reversed in one or more units.

I cannot comment on the VR3 or 6, wait for catserveng to answer. But rafiq's remarks on the prevalence of oversophisticated controls unsupported in remote locations is absolutely true. It depends on whether this is a new problem that just happened at your site, or has this been the same from the beginning. Where are you located?

The cross-current compenasation is a CT in the center phase, that is phase B if your phases are marked A-B-C.
Does this exist and is it wired up?

regards, rasevskii

 

[catserveng]

Sounds like you have a system with the units operating in both frequency and voltage droop. VR3 AVR's and VR6 AVR's usually don't play well together in a cross current compensation circuit. Are the VR6's a retrofit or an original AVR?

Let's back up a bit, I need some more details. What engine models and what kind of governing? Newer engines will have their governing integral with the ECM (Engine Control Module), older engines will have either Woodward or CAT hydramechanical governors or Woodward electronic governors. Does the kW load share work as desired or do you have issues with it?

On the VAR sharing, do all of the units have droop CT's installed? On the VR6 equipped machines you have the option to use the original CAT nominal 1 amp CT (looks like a bar type CT instead of a donut or window) or a 5 amp nominal CT, which type do you have? On the VR3 equipped machine if it has a droop CT it will be the old style CAT CT, the VR3 does not work correctly with a "normal" type CT.

As noted above the VR3 is an obsolete AVR and frankly I change them out whenever I get the chance. The VR6 is still a production item and is actually a Basler AVC63-12 regulator, the CAT version is only available in 240 VAC sensing.

If you are satisfied with the operation of the kW load share and only wish to address the VAR share and voltage control, I think your best solution would be as follows,

Replace the VR3 AVR with a VR6 (or Basler AVC 63-12).
Install 5 amp nominal CT's on all the units, based on the info provided the rated current is about 720 amps, use 800:5 CT's installed on the B phase.
Set all AVR voltage droop pots the same, I would start at about 30% ot the potentiometer, It's likely you don't have a resistive/reactive load bank so this will be a good starting point. Also make sure you set the stability pots the same, I usually start at 50%.
Use the Basler instructions and wire the droop CT's in a cross current compensation circuit. (diagram on page 16 of this document available from Basler

http://www.basler.com/downloads/VR_parallel.pdf

please take the time to read the whole thing, I think you'll find it very helpful, also try this,

http://www.basler.com/downloads/9337200991D.pdf

This is the Basler AVC63-12 manual, it also has a cross current description and schematic and I find it better than the CAT manual for the VR6).

I have done a large number of smaller systems, two to five generator sets paralleling using a Woodward kW load share line system like the 2301A and the LSM (Load Share Module)use and the AVR's connected in cross current.

A note on cross current, it works best when the units are the same size or well matched proportionally, and the same type AVR with similar droop and stability settings. When installed and setup correctly I find the usually VAR share to less than 5% and voltage droop is very low, usually no more than 2%. Systems trying to use CAT 1 amp CT's usually don't work worth a darn, some people get them to work for awhile, but I don't find them reliable. Also some folks feel you can do a cross current system with no voltage droop. You can make it work but I find them not very stable. Also if you go to a cross current system, keep people from messing with the voltage adjust pots. On the last system I did (and several before) the pots on the front of the panel are dummies. If you install and adjust the system correctly, the voltage will stay pretty close, and your VAR share will stay good. Biggest mistake is someone tries to maintian the system using panel meters that are not accurate. Set everything up with one good quality DMM, record the settings and let it do it's job, I have systems I have setup over 10 years ago that have stayed working well.

If you find that you desire a better form of control, and/or want to improve you kW share as well, there are several options. I lean towards controls made by Woodward, but there are others as well. More details about your system and what you would like to achieve can provide you with better recommendations.

Hope that helps,

Mike L.

 

[Guardiano]

Thanks Mike for your expert comments.I regret for my late reply but there was no internet connection the last three days.
Both VR6 are original ones.The two sets with the VR6 AVR have the CAT3412TT engine and the generators are of the SR4 type. The third set,with the VR3 AVR, has the CAT3412B engine and SR4B generator.They all have the woodward governors mounted on them.
Under no load that is with only one set feeding the 1600 kVA 400 V/5500 V transformer, the no load current is around 90 A.When the second set is paralleled, the no load current on both sets rose to around 200 A.What could be the cause of the high circulating current ?
Besides could you post me the manual for the VR6 and the VR3
AVR?I haven't got any documents on them.

Thanks.

Guardiano

 

[rbulsara]

That is just the reactive current being exchanged between the two gens. The voltages are not matched (when running individually) and need to be set correctly on both of them to match when in parallel. This appears to be just the settings issue, excitation, voltage and speed should match.

Run them both individually and make sure that all parameters are as close as possible to each other.

Rafiq Bulsara

http://www.srengineersct.com

 

[rbulsara]

And then parallel and adjust voltage/speed further as needed..of course.

Rafiq Bulsara

http://www.srengineersct.com

 

[nawao]

Your droop setting is wrong. Try to adjust each set to 4% droop. To do so turn the droop pot fully counterclockwise, load individually each set and adjust droop pot to get 4% voltage droop at full load. You can make it proportionally, if you load the set at 50% adjust the pot at 2% voltage droop. Once you have adjusted the three sets this way try to parallel and tell us how is the behaviour