Gas Generator Fighting Each Other...

Hey guys,

Hope you all are doing great!

A tech observed a strange phenomenon yesterday. He said that he had two 100KW,415V,50Hz gas generators with identical ratings, parameters and components. Both managed by Deepsea 8610 controllers. When he tried to sync them both, they synced perfectly on no load. But as soon as he applied a load, even as small as 5kW(through a load bank), they both started to behave oddly i.e. their currents will shoot as high as 250A and it seemed like one generator is taking the other as load and then the other takes the first as load and this loop is cyclic. Then they both trip.

Anyone with similar experience? Any ideas?

Regards,
Mutahir
 
Hi!
Well I am not bad.. Hope you are doing well...
Can you tell us what kind of load bank is.. Resistive capacitive.. Inductive...
Is there kind of PMS i/load sharing Installed in Deep-sea or external module.. Also try to have a read on ctcross current compensation..
We need a sld or similar drawing.. To support and also control logic diagram of the deepsea module...

Cheers
James
 
This behaviour is excatly what would be expected of two generators operating in parallel with each other (synchronized together) when both governors are set to operate in Isochronous speed control mode.

When two generators are syncrhonized together to supply a small set of load(s), it is customary for one generator's governor to be set to operate in Isochronous speed control mode and the other generator's governor to be set to operate in Droop speed control mode. In this way, the two generators will "share" the load acting as one generator. When two generators synchronized together and both governors are set to operate in Isochronous speed control mode, they will each try to control the frequency of the load, and the load will usually be seen to switch back and forth between the two generators. In this mode, there are two chiefs and no Indians--and the two chiefs each want to be the master controller and cannot share the load in order to maintain a stable frequency. When one generator is set to be the "master" (the Isoch unit will be the master) and the other unit is set to Droop, the Droop will behave nicely and only operate with the load the operator has chosen for it, and the Isoch unit will increase or decrease its load--up to its limits--to maintain load AND frequency of the system.

If you require more help, you will need to tell us what governor mode each of the two generator-sets is set to be operating in when the two are synchronized together.
 
Mrasool,

AND, you're going to have to tell us if there's any kind of load sharing scheme active when the two generator-sets are operating in parallel (synchronized).
 
MRasool,

I did some World Wide Web sleuthing early this morning and see that Deep Sea 8610 controllers are synchronizing and load sharing controllers.

So, it would seem that ControlsGuy25's advice is the best at this point--there is something amiss with the load sharing "scheme." (I put "scheme" in parentheses because it would appear that there isn't a load sharing scheme if both gen-set governors were in Isoch when synched and loaded....)
 
MRasool,

I did some World Wide Web sleuthing early this morning and see that Deep Sea 8610 controllers are synchronizing and load sharing controllers.

So, it would seem that ControlsGuy25's advice is the best at this point--there is something amiss with the load sharing "scheme." (I put "scheme" in parentheses because it would appear that there isn't a load sharing scheme if both gen-set governors were in Isoch when synched and loaded....)
CSA,

Thanks for these comments I appreciate to learn & participate on this forum with people around the world...

Mrasool,

Can you give a quick update on our comments...

Stay blessed
 
Generator battles can be exciting. I worked on a diesel-electric anchor-handling offshore vessel for 5 years or so. It had 4 diesel gen-sets, numbered 1 through 4. It was built in the early 80s with analog electronic controls. We could run any combination of any number of generators at a time, synchronized together. Each generator had its own controller that sent analog signals to the other generators. The lowest numbered generator was supposed to be master, but if the frequency adjust knob on a higher-numbered generator was set higher than the master, they would fight. In the terms used above, the lowest-numbered generator would be isochronous while the others would be in droop mode.

When synchronizing in another generator with a higher number, we had to remember to turn its knob down "a little" after closing the breaker. If we later shut down the lowest-numbered generator, we had to then remember to adjust the knob of the new master to keep us at 60Hz and make sure the other generators online were all turned down lower. After changing masters, we would stand by in the control room to make sure it stayed stable.

The fight wouldn't always start right away and would vary in intensity as the load changed, but it was enough to see the lights flickering if you didn't catch it. It could get pretty intense in the engine room with the lights flickering, the governor linkages surging back and forth, and the turbochargers changing pitch. You immediately knew that *something* wasn't right.
 
Hello Gentlemen! Thank you for all of the wonderful inputs you give.

Update time!

So we explored a lot of things during our troubleshooting and in the end reached to below solution:

The CAT Gensets have Voltage and Frequency Knobs on the engine controller. These knobs are sometimes used by FM to make minor adjustments after they work on the engines.

Recently, FM carried out a Top-end Overhaul of both the engines. During their work, they changed the Diaphragm vales(on the gas line) and throttle valve(on the engine). When they did that, they tweaked the V and F knobs to adjust the Voltage and Freq on the CAT local panel.

During my checkup of the Deepsea 8610 Modules(responsible for load sharing and KW/KVAR control), I noticed that each controller is tunned to control 'V' and 'F' of the genset. The tunning of Deepsea is essentially controlling the CAT CDVR(Digital Voltage Regulator) and Woodward Governor(both are mounted on the generator) through the 8610(mounted in the switchgear room). This tuning requires that the AVR Analog and the Governor Analog be as close to '0' as possible with the nominal Frequency and Voltage. So we adjusted the local V and Freq to control our AVR and Governor Centre and Range in the software and brought it closed to '0'. The critical thing I noticed here was that if I change the local V and Freq from the engine manually, I will have to retune. If I don't the deepsea control will not be effective.

We did this on both Gensets.

After the tuning, we made a sync attempt(at 50KW load total). The gensets did not sync successfully. But the response had improved considerably. We decided that we should put some more load(we decided that we will keep atleast 50KW on both Gensets and then make a sync attempt. Both are rated at 108KWs. This time around, when the load was 50KWs each, the synced perfectly. We loaded both to 100KW and brought the load down two to check if a problem happens, but they took load and shed it normally when in sync.

We made multiple sync attempts then and all were successful if the load on both the gensets was 50KW or above. Below this rating, the sync was unsuccessful leading to same high circulating currents.

We instructed OPS to keep the load at 50KW or above on individual gensets below attempting their sync.

Now, we are almost half way there. We want them to sync easily below 50KW as well.

Any ideas?

I was thinking of tweaking the PID settings of CAT CDVR and seeing if the response of Voltage adjustment improves.

Regards,
Mutahir
 

Attachments

Hello Gentlemen! Thank you for all of the wonderful inputs you give.

Update time!

So we explored a lot of things during our troubleshooting and in the end reached to below solution:

The CAT Gensets have Voltage and Frequency Knobs on the engine controller. These knobs are sometimes used by FM to make minor adjustments after they work on the engines.

Recently, FM carried out a Top-end Overhaul of both the engines. During their work, they changed the Diaphragm vales(on the gas line) and throttle valve(on the engine). When they did that, they tweaked the V and F knobs to adjust the Voltage and Freq on the CAT local panel.

During my checkup of the Deepsea 8610 Modules(responsible for load sharing and KW/KVAR control), I noticed that each controller is tunned to control 'V' and 'F' of the genset. The tunning of Deepsea is essentially controlling the CAT CDVR(Digital Voltage Regulator) and Woodward Governor(both are mounted on the generator) through the 8610(mounted in the switchgear room). This tuning requires that the AVR Analog and the Governor Analog be as close to '0' as possible with the nominal Frequency and Voltage. So we adjusted the local V and Freq to control our AVR and Governor Centre and Range in the software and brought it closed to '0'. The critical thing I noticed here was that if I change the local V and Freq from the engine manually, I will have to retune. If I don't the deepsea control will not be effective.

We did this on both Gensets.

After the tuning, we made a sync attempt(at 50KW load total). The gensets did not sync successfully. But the response had improved considerably. We decided that we should put some more load(we decided that we will keep atleast 50KW on both Gensets and then make a sync attempt. Both are rated at 108KWs. This time around, when the load was 50KWs each, the synced perfectly. We loaded both to 100KW and brought the load down two to check if a problem happens, but they took load and shed it normally when in sync.

We made multiple sync attempts then and all were successful if the load on both the gensets was 50KW or above. Below this rating, the sync was unsuccessful leading to same high circulating currents.

We instructed OPS to keep the load at 50KW or above on individual gensets below attempting their sync.

Now, we are almost half way there. We want them to sync easily below 50KW as well.

Any ideas?

I was thinking of tweaking the PID settings of CAT CDVR and seeing if the response of Voltage adjustment improves.

Regards,
Mutahir
I told you that, the load scheme was not implemented /tuned well.....!
 
Guys,

Can anyone guide me on the PID settings of AVR?

My original settings in both CAT Digital Voltage Regulator were:
P-200
I- 500
D- 30
Loop Gain- 50

I changed to:
P-200
I- 750
D- 30
Loop Gain- 60

Both the GGS are at 1% Droop Mode.

After this change, the GGs starting syncing at 40KW and above. Before they were syncing only at 50KW and above.
But they is very little Voltage dip that I can feel in the lights of the switchgear after this change. How can I smoothen the Voltage output further by these gains?

Regards,
Mutahir
 
Guys,

Can anyone guide me on the PID settings of AVR?

My original settings in both CAT Digital Voltage Regulator were:
P-200
I- 500
D- 30
Loop Gain- 50

I changed to:
P-200
I- 750
D- 30
Loop Gain- 60

Both the GGS are at 1% Droop Mode.

After this change, the GGs starting syncing at 40KW and above. Before they were syncing only at 50KW and above.
But they is very little Voltage dip that I can feel in the lights of the switchgear after this change. How can I smoothen the Voltage output further by these gains?

Regards,
Mutahir
Hello.

These kinds of operations /tuning should be fixed during commissioning phase.. Is that the case here...

Did you have heard about droop voltage compensation..

What is actually grid/utility configuration..

I told you before that load scheme was not tuned well it looks like still that voltage droop/dips remains..

Can you make a test and record trends datas... So you can share here and we can provide you with a better solution..

James.
 
You may also look at KVAR fluctuations if occuring ...as we not there ...and do not have overall view of the plant yet....

PID settings tuning require some tests to be performed ...so follow OEM instructions and come back here for any clarifications...

I worked about 7 Years as Gas /Steam Turbine Generator and I know well about those PID tuning and Grid system stability...

Any time!
James
 
I see on the last secreenshsot that GOVERNOR & AVR got same values for SW1/SW2 is that correct ...
Yes, it was like this before. But we tuned the Governor and AVR settings after that.

  1. At first when this problem started the GGs weren't syncing at all at any value of KW.
  2. After tune-up of deep sea we were able to sync them at 50K and above.
  3. After further tune-up of PID settings now we are able to have GGs sync at 40KW and above.

Now we are at a stage where we need to tune further to achieve sync even below 40KWs. I will do some further tuning and share results later.
 
Yes, it was like this before. But we tuned the Governor and AVR settings after that.

  1. At first when this problem started the GGs weren't syncing at all at any value of KW.
  2. After tune-up of deep sea we were able to sync them at 50K and above.
  3. After further tune-up of PID settings now we are able to have GGs sync at 40KW and above.

Now we are at a stage where we need to tune further to achieve sync even below 40KWs. I will do some further tuning and share results later.
Do you mean that sw1&sw2 values are correct..
 
Top