Paralleling generator troubleshooting

[Mztr45]

Have two 3306 caterpillar 175kw generators with 2301 speed and load share controls and spm a synchronizers for each. When one of the generators is online and holding the load and second generator is running at rated producing power but not on the bus I have an issue when attempting to parallel. The instant I move the switch into parallel mode the online generator immediately ramps down in speed about 20 rpm and the oncoming will ramp up almost the same. This is confirmed watching frequency meter and rpm gauges.

These generators were paralleling and load sharing just fine a few days ago. Trying to nail down possible culprit where the problem is coming from. I'm thinking one of the synchronizers is on its way out and sending a garbage signal to the online generator even though it's not the oncoming generator. The same switch position is used for paralleling whether it's #1 to #2 or #2 to #1.

 

[waross]

You probably never noticed it before.
If two generators are running at the same speed, but out of sync, you may have to wait a long time for them to drift into synchronization.
Nest practice when synchronizing is for the incoming set to be running slightly faster.
That ensures that it picks up a little load and avoids possible nuisance reverse power trips.
20 RPM plus 20 RPM, equals 40 RPM difference or 2.2%. At 60 Hz, that is 1.32 Hz.
You should come into synchronization in less than one second.
It may be that something has happened to change both speed changes, they may be better at 10 RPM or 5 RPM, but the fact that both are the same RPM suggests than the issue may be elsewhere.
I would first look for an issue with the sync check relay.
A suggestion; If the voltages are not matched, or if the sync check thinks that the voltages are not matched, it may be preventing the sets from sync'ing. Then you would see the RPM difference that came and went so fast on a normal sync that you never noticed it.


--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[waross]

Further to the above.
I installed a standby set with a very fast transfer switch.
The transfer was so fast and the internal clearances so little that an internal grid to generator flash-over and/or motor damage were possible if the transfer was not done when the generator and grid were in sync.
The transfer switch had a sync check relay, but no speed adjustment.
The sync check relay waited until the phases drifted into sync before allowing a transfer.
This set was operating in droop, so the frequency varied from 60 to 61.8 Hz.
When the set was fully loaded and running at 60 Hz, it sometimes took several minutes to drift into synchronization.
The supplier had screwed up the order and the set was undersized by a factor of 208/240. (I wonder how that ratio happened?)
Hence the set often ran at 100% load.
The solution? Tweak the Hz range up to 60.5 Hz to 62.3 Hz.


--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[Mztr45]

Thanks for the reply. I'm fairly in tune with how this plant runs and it's definitely something new occurring, a .5ish hz change is pretty noticeable on a meter with a span of 58 to 62.

In my past experience with other plants, like you described, put switch into parallel, adjust oncoming generator to slightly faster than the online generator, confirmed by sync meter showing slowly clockwise and then close breaker at approximately noon position, with a sync check relay preventing breaker close if otherwise. I've never seen an online generator change speed during this process outside of load changes and they react back to adjusting for 450v @ 60hz.

For what it's worth, the issue started out where I was trying to bring on the second generator and both engines were going even further apart. I shut everything down including the engines, got the one generator going again put it onto the dead bus and then got the other gen going and everything performed as normal, online gen held steady with the load, put switch into parallel, oncoming generator began adjusting speed and I closed the breaker. They were load sharing as nicely as can be expected Over the course of about 8 hours, the load slowly shifted over to the first generator and was almost at the threshold of reverse power. Singled back up and my online gen was now only producing 420v, readjusted that back to 450v and now where I am. Can get the breaker to close briefly and then just opens shortly after from reverse power.

I'm already aware of some issues on the one generator, which I'm fairly certain is a fuel delivery issue at higher load demand, but it only has problems when it's online by itself, up to now it's still been playing nice as long as it's being paralleled to the other. The typical load is only about 100kw, so it's not like they have to work hard sharing that, having them paralleled is more for redundancy in case one goes offline than being able to handle normal load.

 

[waross]

I would check all parameters independently and compare with reported values.
You may have a loose or corroded connection somewhere.
Wait and see if Catserveng posts in.
It may be a failing synchronizer, but more likely a compromised signal into the synchronizer circuit.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[TugboatEng]

I have had issues with speed control on Caterpillar engines recently. Actually, persistently.

Is the speed signal transmitted through a relay or switch? I have recently learned of the concept of a "wetting current". Most control relays have a minimum switching current of 35-50mA. This is above the common 4-20mA process signals and well above the range of voltage signals, Caterpillar uses a lot of PWM which is a voltage signal.

There are relays with gold plated contacts that have minimum current ratings of 1mA. However, this is still well above the currents seen with voltage based signals.

My solution has been to use signal converters to convert from voltage to 0-20mA, place the relay with gold plated contacts in the 0-20mA circuit, and use a second converter to convert back to the voltage signal. As long as your signal moves above 5% you will meet the necessary minimum current.

Caterpillar does not use the correct relays in their equipment.

 

[waross]

I have used sealed reed relays for low power signals.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[waross]

Hi Tug, I believe that there is also a voltage threshold as well as a current threshold.
Many years ago when solid state logic was first applied to industrial control schemes in place of relay logic we encountered similar problems.
With currents below your 35-50 ma limit we found that a voltage of about 100 Volts or more would also puncture the surface film.
The early solid state logic implementations often used 100 Volts of more AC pr DC for field inputs from push buttons or limit switches.
The panels had a row of input modules that accepted the 100+ Volts from the field devices and converted the signal to a logic level.
Of course this was digital or on-off inputs not analogue inputs.
But there is more than one way to deal with surface film, sometimes.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[Mztr45]

Hey TugboatEng. As far as I'm aware the speed signal isn't sent through a switch or relay. It's just a single mag pickup that first terminates in the local control on the overspeed box and then splits off to the local rpm gauge, a remote rpm gauge and the 2301. Not a great design choice in my opinion, would rather see a dedicated pickup just for the 2301 like all the other 3508 and 3406 systems I've worked with. But those were also 20 year old boats and not a 40 year old like this one.

Right now my plan after hitting the pier in a day or so where i can go dark safely is to just swap the two synchronizers and see if that changes anything. If not, going to groom the 2301s for both, the one has some gain adjustment needed to clear up some idle speed instability anyways and the other takes a long time to go from idle to rated. Who knows maybe I'll find something loose that I didn't catch the first time I gave everything the tug test. I'm reluctant to call the synchronizer bad just because I've never seen a 2301 or synchronizer fail without some sort of obvious event involved. But it's the only thing I can think of that would be getting introduced into the speed control upon turning it to parallel mode.

I did fire up the other gen again and set the online gen a bit faster to 1810rpm/60.3hzish, matched speed with the other and flipped the parallel switch on and the online gen shot up to around 1860rpm/ pegged my meter at 62hz and I immediately put it back.

 

[TugboatEng]

When I say speed signals I am talking about the signal from the synchronizing controller to the governor/ECM of the engine. There are often relays in-between. Cat does put relays in their control cabinets. Cat uses AgNi contacts in their relays and these are not correct for the application. The older the boat gets the more likely it is for the problem to occur. We experience issues with speed control within 6 months to 10 years. Our Cat engines are propulsion engines and the problem manifests as zero to partial throttle response.

Waross, I had looked into reed switches but I want sure if they would work in the marine environment with its associated vibration. My crash course was here recently on the forum. I really appreciate everyone's input.

 

[Mztr45]

Gotcha. That is something I can verify, whether my voltage coming from the 2301 is still the same when it reaches the plug to the governor actuator when it's switched into parallel. It wouldn't make sense to me to ever have this sent through anything else but who knows, i haven't found any drawings for the current set up only the original. I know with the 3508s with an ecm it's a little different. Have had to redo deustch connectors for the pwm signal getting sent back on those and loose wires on the j1 plug at the ecm itself.

 

[wcaseyharman]

Do you have a system that uses an external bias to control the 2301 like a Woodward EGCP or east Gen.? If so you can you measure and see what’s it’s doing as you attempt to parallel? Curious if it’s the 2301 itself or the device that’s providing the bias (if equipped).

 

[Mztr45]

Been very busy but just wanted to follow up and say thanks for all the replies. I ended up putting in a new synchronizer and everything working great, also got my speed control problems fixed, just needed a little TLC on the linkage. Whoever did it before I got here had it set to max out on the actuator before it maxed on the fuel pump, so it would run 1800 just fine with no load, but once there was a load it would be starved for fuel.

The only question I have leftover from this issue is: After going through the wiring for my synchronizers, I noticed a jumper on both of them that didn't match up with anything I saw in the drawings or reading in the manual. Right now there is a jumper going from the GEN sensing side to the BUS sensing side between terminals 5 & 9. Any ideas on why that jumper exists? My initial thought was maybe for dead bus closure, but that's addressed in the manual and doesn't say anything about using a jumper there.

Also, had a lesson in attention to detail. The first synchronizer I got was similar but not exact to what was previously installed and I still couldn't parallel. Took me a couple hours of headscratching until I figured out that the new synchronizer I put in had a 1% voltage matching threshold while the old one didn't have any threshold. Temporary fix was to tweak the pot for the automatic voltage matching (even though there was nothing in place to utilize this function) until I could hear the internal relay start to click and then I knew that the generator was producing voltage very close to the threshold for matching to the other gen, hold the breaker close switch at that point and eventually would match long enough to let it close. Once I had it that close I could just use my voltage adjust knob and listen for that relay while adjusting voltage for further parallel needs untiL i got another synchronizer without the voltage matching.

 

[che12345]

Mztr45 (Electrical)(OP)15 Sep 23 14:09
"...Have two 3306 caterpillar 175kw generators with 2301 speed and load share controls and spm a synchronizers for each. When one of the generators is online and holding the load and second generator is running at rated producing power but not on the bus I have an issue when attempting to parallel. The instant I move the switch into parallel mode the online generator immediately ramps down in speed about 20 rpm and the oncoming will ramp up almost the same. This is confirmed watching frequency meter and rpm gauges...."
It is nice to hear that the system is back to normal operation after replacing the automatic synchronizer. But the root cause is not established.
I have the following opinion which were not mentioned by learned contributors so far.
1. The device you have mentioned correctly is an automatic synchronizer, NOT a syn check relay. The key difference is the automatic synchronizer has a set of contact for voltage raise/lower and another separate set contact for speed fast/slow. These may be either dry contacts or with voltage output. A syn check relay does NOT perform these functions.
2. An automatic synchronizing relay receives the voltage input from the BUS as the reference and compare the voltage of the same phase from the GEN. Voltage raise/lower is connected to the AVR. The speed fast/slow is connected to the governor of the GEN. There is no control on the voltage or frequency of the BUS.
3. Your observation that the BUS and GEN changes in frequency is unlikely caused by the automatic synchronizer.
Che Kuan Yau (Singapore)

 

[Pandit1]

When multiple generators are running in parallel, the load distribution among them is dependent on their speed regulation. To ensure the smooth operation of the generators, it is important to check their speed regulation data and maintain a slightly higher incoming frequency than the running machine. The synchronization of the generators should be done when the synchroscope is rotating clockwise to avoid any interruption in the load-sharing process.

 

[crshears]

Back in my generating days we referred to this as 'pushing into the system' when synchronizing; this philosophy applied to almost every generator, the only exception being any submerged turbine hydraulic units [based on hydrodynamic theory], which were always synched while running slightly slow.

CR

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

 

[wcaseyharman]

What model synchronizer is it? Do you have a manual?

What comes to mind regarding the jumper, although it’s probably not for this case, is PTs were often open delta with B phase grounded. An A phase from each side of the breaker could be brought into each polarity terminal of a synchronizing device and the non polarity terminals jumpered together and grounded (since B phase for both sets of PTs were referenced to ground).