Power generation, circulating current problem 1

[skeandhu]

Could anyone tell me the best method for making two entirely different generator sets load share. The main generator is driven by an old Ruston TA gas turbine with a mechanical governor set at 5% speed droop, the standby generator is a Caterpillar diesel set bought as a prime stand alone generator. When the two are paralleled the circulating currents are 2000A plus! Obviously a new governor will have to be purchased for the Caterpillar so it will be able to load share. Is it possible to accomplish this by connecting the two AVR'S together using VAR load sharing relays and motorised potentiometers? Any suggestion would be gatefully accepted.

 

[bigamp]

It sounds like the circulating current may well be reactive. I assume you are operating the two generators in parallel with each other but isolated from the grid, is this correct? If you are to operate like this, it is essential for both generator's AVR's to have provision for operating in parallel, some form of voltage droop, cross current compensation etc. is required to ensure you get equal reactive load sharing.

Possibly you will need to replace AVR's. If you do then Basler make an excellent range of products and are worth looking at. I have used their digital microprocessor based DECS in the past and found it to be excellent (and reasonably priced). Having AVR's with under and overexcitation limiting would also help.

IMO, you are correct to be looking at the governors. Does the cat governor have droop? It will need to have if it is to share kW load. Check out the Woodward website for more information.

Regards

 

[wbd]

Verify that the generator winding pitches are the same.

 

[Shortstub]

Are they directly connected to the same bus, or is there an intervening transformer.

 

[MShaat]

hi,
skeandhu

The problem of circulating current is reffered to the incorrect reactive power sharing as bigamp mentioned; this means there are incorrect/inaccurate settings of voltage magnetude and/or voltage droop.
You have to note that whatever the method of reactive load sharing, the need for parallel droop kit connected to AVR is indespensible to get a stable reactive power sharing.

wbd
Could you please explain the link between the winding pitch and circulating current: are there certain constrains in paralleling full pitch to 5/6 or 2/3 pitch alternators.
Thanks in advance

Regards

 

[skeandhu]

Thanks for all the replies, but i think i should explain further, The Caterpillar is for Isoch operation only, as you suggested a governor will have to be purchased. This is a small Production platform in the gulf run by Aramco. Inside the breaker cubicle for the Ruston there is a cross current compensation set up, Rheostat, current transformer which is hooked up to the AVR but obviously is not connected to the Caterpillar. To let you understand the Engineers here are not very good and why they bought a stand alone generator is beyond me. While looking on the net i tried a Company called Selco (we use their Reverse Power Relays) a VAR load sharing relay T4900 looks promising.

 

[busbar]

Are you able to manually adjust the AVRs to reduce circulating curent? If not, it is possible automatic controls may not be able to do any better.

 

[coingy]

Skeandhu,

In order to share correctly you should have the controls set for droop (as Bigamp states) and the AVR's and Governors set regardless of rating of the sets to the same settings, typically 4%. Having one set in Isoch will cause problems as its flat response will fight for the load or reactive load as there will be no stability with the Ruston sets.
The QCC circuit should be connected between the Rutsons and the Cat in order that sharing of the Mvars takes place automatically.

Hope this helps, Coingy

 

[jbartos]

Suggestion: Before any paralleling takes place, computer modeling might be performed to obtain necessary data for paralleling, protection and grounding.

 

[bitoy]

An individual droop CT's should be allowed to feed back to the AVR's. These CT's are short circuited or bypassed through relay contacts during single or isochronous operation.

 

[Electrifier]

as i have seen on caterpillar generators, DROOP CT is normally connected in parrallel of droop potentio. If CT polarity is reversed, it will cause to pick excessive reactive power at start and draw a high amount of stator current. If the polarity is ok, You can adjust VAR share by droop ptoentio as mentioned above. We are doing this at our several parallel applications of CAT SR series generators by just agjusting the droop potentio. Also check the Polarity of voltage input at AVR as per specification.

 

[codling]

As mentioned above, you need to check the winding pitch on the alternators to ensure that they match. Most likely, considering you are getting 2000A of circulating current, they do not match. I would look at "Power Systems Analysis" by Arthur Bergen pages 125 thru 127 which discusses this problem as it occurs when paralleling transformers.

In practical applications we have had to adjust the AVR controls to see if the problem can be balanced by adjusting excitation level. If not, then the problem is reactive and you will need to add reactance to one of the machines on the grounding leg. The reactor impedence can be calculated using Bergen as a reference.

Hope this helps.

Jim

 

[francisd]

Have you measured the 'circulating' current to check if its 3rd harmonic? Are both generators grounded?