Reactive droop compensation issues with three 250 KW Cats 2

[iamjohnmack]

thread238-172850: power factor not stable
Hello. I am in need of some advice on an issue we are having with two of three Caterpillar 250Kw units that are paralleled to supply a stand alone floating maintenance barge. The condition we are experiencing is that units #one and #two together run well and share the load equally and power factors adjust well. When you parallel either #one or #two with #three individually the power factors do not react well and the loads are not shared equally. We have surmised that it is cross current based on the bus amperage and individual unit amperage gages comparisons. And the fact that the power factor meters lag and lead respectively.
As I read through some of the threads, which were very helpful, I have some questions.
The gensets are SR4 CATS with upfitted VR6 regulators. We have had the Cat techs in to work on the system. They tuned in the 2301a controls and got the units to share pretty well. But, told us based on the age of the equipment that the power factors are as good as it gets.

Finally,
Question one: Can anyone tell me if the burdeon resistor is internal to the VR6?

Question two: How can you tell if the polarity of the CTs are correct?

I suspect that the polarity of the CT on unit #three is not correct but am leary of reversing it without some more information.

Sorry this got so long. Wanted everyone to have as much info as possible

 

[catserveng]

1. If you refering to the burden resistor shown in some cross current connection diagrams, then no, that resistor is used to trim the currents when using different type regualtors.
2. Easiest way is to set the droop pot to about mid position and add load to a stand alone unit that has some reactive power content and watch the voltage droop as the load is added, if the CT is reversed the voltage will increase instead of drop. The CT being reversed is not likely based on your description since usually if it is when paralleled one unit will hog all the reactive load and the other will go into loss of excitation, especially on these size units.

Go to Basler's website, get the manual for an AVC63-12 (the Basler version of the VR6), also in the downloads area next to the manual you'll see a link for an application note for VAR/PF in Parallel, get that as well, an excellent reference on how to setup a cross current circuit.

When the units had the VR6's installed, did they replace the droop CT's or leave the old ones? The original CAT bar typr CT installed in most older SR4's could be a little inaccurate, and not provide a good enough signal to make VAR sharing work correctly. When I upgrade I also install new droop CT's with 5 amp secondaries and sized for full load current.

Do the units have some voltage droop? If I don't have a resistvie/reactive load bank to set a system up I start with the droop pots set all the same, usually start at mid scale.

A properly installed cross current or droop VAR share system on three like sized units with the same AVR's and droop circuits should be able to proportionally VAR share within 2%. Differing AVR's and droop circuits will make that worse but have been able to get them within 10% in most cases.

Hope that helps, Mike L.

 

[iamjohnmack]

CatServ thanks for the info. I will research baslers site for the info you suggested.
When the VR6's were installed we did not replace the CT's. We used what was on the machine. I recently had the CAT parts guy research based on arrangement number what size the CT's were and he informs me that the CT's are the one amp variety. At one point the CAT Techs replaced one of the CT's (#2unit). I assumed that he replaced like for like.
The units are set for some droop. Not sure, but I want to say around 3 or 4%. I will look when I return to work this evening.
The upgrade of CT's seems like an extremly good idea. I will make that happen.
We do have access to resistive load banks. And I think we will address this as soon as we are able to shut down for maintenance and I have the replacement CT's in hand. Any other input/suggestions would be greatly appreicated.

 

[catserveng]

A good source for CT's is Flex-Core,

http://www.flex-core.com,

I usually use something like their 194 series, size the CT for full load amps, you didn't say what your nominal voltage is so you'll need to figure or read it from the nameplate if you can. Put ALL B phase conductors thru it so you get a good signal, do the same on all three units.

If you don't have a load bank, use your own load, just come up with some larger loads that equal at least 50% rated power. Apply it to each unit and assure the voltage droops are as close as possible, about 3% voltage droop will work fine for most cross current systems.

Also make sure the Volts/Hz (Basler calls it UF Knee)setting is the same on all units, I usually set it for 59.5 Hz on a 60 Hz system, otherwise if you hit it with a big load they won't respond the same and share like they should.

Link to AVC63-12 manual,

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

Link to Parallel Paper

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

Good Luck! Mike L.

 

[iamjohnmack]

CatServ-Great Stuff. I will sift through all of this and order the CT's. The Nomimal is 480. I am assuming that I will size these for VA and look for 5 amp secondaries. I will make sure that the UF Knee is set. Right now I think we are at 58.5 from what I remember. I will check my notes.
Thanks for all of the great advice. I will let you know how it goes and if any of this effects a positive change. I am not satisfied with the approach that this is as good as it gets in terms of power factor adjustment. This system worked pretty well until we had to change the VR3s to the VR6s.

 

[e5fornow]

Just from past experience the VR6 does not do cross current very well. If the load demands can take it I would remove the cross current lines and run it in droop. You should be able to use the internal droop pots to balance out the reactive load. If you are required to run in cross current I would make sure the regulator is wired correctly. Terminal 5 for 1A secondary and terminal 5A for 5A. Also make sure the external burden resistor has been removed and any external pots are gone. If you follow the Basler cross current diagram the .1 Ohm resistor is a starting point I have had to go as high as .3 Ohm, also I am not sure if it was in the manual but the resistor need to be rated for 15Watts also you want as little resistance in the cross current lines so make sure you remove the door bell transformers which were often used with the VR3s. Good Luck.

 

[iamjohnmack]

E5fornow....Thanks for the response. The terminology here is a bit difficult to keep up with. In the past week I have had more of an education on parallel droop compensation or reactive droop compensation, crosscurrent compensation or reactive differential compensation than I ever wanted to . I guess what I hear you saying is if we are running crosscurrent compensation get rid of it with the VR6. We are running reactive droop compensation. When the VR6s were installed we removed the external droop pot and wiring that was associated with the VR3. Now where I think this went wrong is that we did not replace the CTs and are using the 1 amp CTs that were existing. I dont feel that this signal is adequate for the VR6. Everything I am reading is saying 5Amp 5 Amp 5Amp... I am not even sure that the 1 amp Ct's are hooked up right. There is no way to know which lead is which. I am going to take it through the troubleshooting to try and verify that when we get back to the boatyard next week. There, I can run one unit standalone and add some of the larger inductive loads and get a meter on it to see how it is reacting. I have ordered new solid core 5 Amp CTs and will install them when recieved. I have never really had to troubleshoot this circuit in the past. It has always been pretty solid. It has been interesting bringing myself up to speed. This site is a lifesaver!! It is not very reassuring that you are saying that the VR6s do not like to play nice in terms of reactive load. I hope this is NOT just as good as it gets.....Thanks Again for the post. I'll let you guys know how it turns out. Good Luck with E6.

 

[catserveng]

I guess my experience is different than those of e5fornow, I've done quite a few systems with 2,3 and 4 same size generators setup the same and had very good results. I've VERY poor experience in mixed systems, The more closely matched the tail ends and control system, the better it works.

I do go behind a lot of people working on systems that don't work right for a number of reasons, mainly poor installation and wiring practises, too little droop, and improper or incomplete intial setup.

e5fornow makes a good point about the isolation transformers, some systems got them installed, some didn't. Burden resistor sizing has a lot to do with CT match and where your loads are most of the time, and adjustable resistor is usually the best, and get one In most retrofits I clean up I find they were done half assed by people who really didn't know how to do it in the first place, Frankly, I find if they don't know about and use the Basler article, they shouldn't be doing it, at least in my opinion. Your comment about not being sure if everything is hooked up right is pretty typical. There needs to be a good overall drawings, the wiring needs to be correctly marked, the droop's need to be correctly set, manual voltage adjust pots should be locked with stem locks, and to do it right you really either need a proper test load and develop a good test plan using your loads you can do repeatedly.

A lot of systems can be marginally done and work ok. A lot of systems are really operating in droop because they are poorly done and the operators really don't know any differently.

I've also done a lot of systems in droop, an they also work just fine as long as the droops are set the same and no one messes with the voltage adjustments. The actual system droop will be slightly less than the single unit droop. Not sure why, maybe one of the other folks here can explain that, but in most cases that is what I've seen.

I rarely do droop or cross current control systems on new installations any more, active VAR/PF control using a Woodward, Basler or PLC based type control is really where most systems I deal with are going, especially mixed systems.

Hope that helps, good luck! Mike L.

 

[iamjohnmack]

CatServ....I'll be the first to admit that I have zero experience with this in terms of putting a meter and screwdrivers on the reactive load sharing portion of these. In the end, unless you are in Pennsylvania, I will most likely have Caterpillar come in on this and handle the settings. After we install the new 5 Amp Ct,s and verify all of the wiring. As I said before, this has been an education. This web site and guys like you have been a true relief. To at least understand the circuits and be able to speak and listen intelligently about it is a benefit. At least the units I have are identical. And based on your advice in previous posts I truly believe that this can be alot better. We returned to the boatyard yesterday, and I will be able to check things out. The CT,s are on their way. I will keep everyone posted as to how this turned out. It may benefit someone in the future. Thanks again!

 

[iamjohnmack]

CATSERV.....After the CT's finally arrived and we installed them. After many hours of studying the Basler material, I decided to give adjusting this a shot. As suggested, I approached each unit individually and added about 50% heavy motor load. Made adjustments. The only problem I had with this was that the # two unit reacted as if the CT output was installed reversed. I verified and reverified that the CT was installed correctly (B phase through H1 side of CT and X1 to #6 on VR6). It is identical to the other machines, however the DRP adjustment pot would react exactly opposite of the other two. CCW would make voltage fall. I was not sure what to do with this so I reversed the CT output and it the reacted as the other two machines. ??????. The end result was that units #1 and #3 are really close and respond well to load changes. However, when you attempt to parallel one with two or three with two it does not do well. The power factors mirror each other but the voltages are the same. I am reading the voltages at the sensing terminals on the VR6s and at the primarys of the xfmers in the switchgear. As I am typing this it sounds impossible based on what I know(which is not much) but these are the conditions. Urrrrrrgh!!! It should be noted that the metering on the switchgear is about 25 years old. I am ordering new Digital/analog meters from Cromptom/Tyco Electronics. Any suggestions from anyone would be greatly apprieciated.

 

[catserveng]

When you setup for the single unit testings, is the old cross current wiring completely disconnected? My most likely guess is that you still have something in the cross current loop active.

Other possiblity, I've only seen two in 30 years, is a mismarked CT. You are talking about the droop adjustment pot on the VR6, right? The old VR3 droop pots are gone, correct?

Do you have an as-built drawing of your system?

If you'd like to get a hold of me direct go to

http://www.pactectric.com,

maybe I can walk you tru some thigns more directly.

Mike L.