Frame-III (MS-3002) gas turbine high seismic vibration at Brg#4 casing (LP shaft)

We have GE twin shaft gas turbine (MS3002) with second stage variable nozzles and equipped with Speedtronic MK-II control system.
GT under went HGPI (hot gas path inspection) on Nov, 2018 and right after aforesaid maintenance, its LP shaft radial vibration and Brg#4 side casing seismic increased from 1 mm/sec to 8 mm/sec. Due to above issue we have to compromise load and consequently unit runs at lesser LP speed. Could anyone please suggest what has go wrong and cause of rise in LP shaft seismic vibration as during HGPI, maintenance done on HP rotor, combustion inspection and fuel nozzle and NOT LP SHAFT and FIELD INSTRUMENTATION.
Further, I have shared operating parameters on various dates before and after HGPI maintenance for review and confirm parameters are appropriate or otherwise.

Meanwhile, we have checked undermentioned.

1. GCV, SRV and NCV stroke checking and found appropriate.
2. HP speed settings check and found as per Control specs.
3. PCD TX range, calibration found as per recommended settings.
4. Second stage Variable Nozzle operation / stroke check which is smooth (-5 Deg to 15 deg)
5. P2 pressure calibration checked and found appropriate.
6. LP and HP speed pickups checking.
7. Temperature and nozzle control settings verified.
8. Dry-run test for unit startup sequence complete simulation, which is normal.
Looking forward affirmative response as usual and we would welcome in case of any query.




A favour in return, and I'll give a jab at this. A MKII normally have more than one vibration amplifier cards, if my memory is correct.
So, each transducer has it own amp;
But the amplified signals goes to another card, for level detection and latch and latching.

If the input transducers are connected together, [single card or at output of each amp] and are out of phase, you can end up with the problem described.

My memory of the diagramatic setup is not clear, but I've observed this problem before.

Before doing anything else, I would isolate each transducer first by disconnecting the wire at the JB.

Hmmmmm....there is open circuit detection, but there is a resistor jumper somewhere on the caRd for that.
Thanks for your reply.
Inadvertently missed to mention earlier regarding seismic transducer and its loop.
Please note that we already checked loop and simulate transducer loop from site to control room with Vibration Shaker and verified indication, alarms limits, which is appropriate.

To follow up with you last response.
You said you checked the loop with a shaker and used the oscillator on SVDC

But that only simulates one transducer at a time...and that's maybe the reason why no constructive interfarance was seen.

There could also be another problem...related to the cards and there age.
If you would like to take this discussion off line further, you could send me an email vcegce@gmail.