Gas Turbine SRV Leak Observation

Hello Guys,

I wanted to share a recent experience of mine on our Hitachi H-25 Gas Turbine Unit
(with Mark-VI Control System) and wanted your recommendation / expert guidance in
devising the way forward. We had our unit startup scheduled prior to which I checked
the system, verified the calibration / stroking of IGV, GCV and SRV and found no
abnormality. However, when we initiated startup and lined up Natural Gas on the
upstream of SRV, machine sent a trip "Gas Fuel Intervalve Pressure High Trouble". This is
a pre-ignition trip which is actuated when intervalve P2 gas pressure is greater than 0.30
MPa (60 sec delay) prior to GT firing. Upon checking, we found that P2 pressure was
0.41MPa and this is why Mark-VI was sending the trip command.
Following checks were carried out to find the exact root cause of the issue:

1. Operation of Bleed SOV G20VG was verified and was found to be okay. However,
this SOV was preventively replaced with a new one
2. Both Interstage Pressure Transmitters G96FG-1 and 2 were verified and calibrated
and were found to be healthy
3. Gas presence in intervalve chamber downstream of SRV was verified physically
through Pressure Transmitters vent points and SOV vents and gas presence was
observed
4. SRV position was physically verified and also through LVDT feedback and it was
ensured that SRV was closed physically as well

Based on above checks, it was quite apparent that the issue was due to slight leaking of
SRV valve. Valve operation was OK but it was not providing tight shut off at zero
position. We were able to start the machine by raising the setpoint to 0.5MPa. Machine
started normally and is operating fine till now but I have some basic queries which I'd
like to share here:

1. During GT operation, SRV position remains relatively constant to ensure constant

pressure at GCV upstream so will this slight leaking in SRV affect turbine
operation on maximum load?
2. We did some internal hazard analysis and didn't think that raising the trip
setpoint to 0.5MPa could have adverse effect on turbine mechanically. This trip is
just to ensure that gas does not enter turbine manifold in offline condition and
 we've already catered that by first checking that GCV is not leaking and is holding
pressure at upstream and second by ensuring an interlock which depressurizes
gas circuit whenever GT is stopped or tripped. I'd like to know if there's
something we've missed or if this modification can be harmful?
3. It is quite evident that SRV needs mechanical overhauling but OEM (Hitachi) does
not recommend on-site overhauling and advises to send existing GCV-SRV set to
their factory and in the meanwhile, purchase a new set from them as a 'roll-in
roll-out' concept. This is, of course, a very expensive solution. Any
recommendation or way out regarding this will also be appreciated.

 
Like Reply

Makster,

1. During GT operation, SRV position remains relatively constant to ensure constant

pressure at GCV upstream so will this slight leaking in SRV affect turbine
operation on maximum load?

That's 100% not correct. During GT operation, as the GCV opens/closes to change load
the intervalve pressure will change--and the SRV responds appropriately to keep the P2
pressure stable across the entire load range (100% TNH to Base Load and back to 100%
TNH). Also, leaking of the SRV will not affect operation at maximum load, only when the
SRV should be closed.

1. We did some internal hazard analysis and didn't think that raising the trip
setpoint to 0.5MPa could have adverse effect on turbine mechanically. This trip is
just to ensure that gas does not enter turbine manifold in offline condition and
we've already catered that by first checking that GCV is not leaking and is holding
pressure at upstream and second by ensuring an interlock which depressurizes
gas circuit whenever GT is stopped or tripped. I'd like to know if there's
something we've missed or if this modification can be harmful.

The only thing which may be affected is there will be a "hard light-off" because of the
increased pressure and fuel trapped between the SRV and GCV that is released into the
fuel piping, nozzles and combustors when the GCV is opened. The normal sequence of
events during firing when it's time to admit fuel and ignite it is for the ignitors to be
energized first, then the GCV opens (releasing any trapped fuel between it and the SRV)
and then the SRV opens to bring the P2 pressure up to reference. The intervalve vent
solenoid, 20VG-1 closes when the unit starts turning, and that's when--if the SRV is
leaking--fuel and fuel pressure will begin to build up in the intervalve cavity. And, it will
continue to build pressure all during the purging sequence and up until the GCV
opens.... If the SRV is leaking when the unit is at rest and the pressure in the intervalve
cavity is building up with 20VG-1 open, then it's a pretty significant leak. Closing 20VG-1
with a leaking SRV is going to allow pressure and fuel to build up even higher (with no
vent to atmosphere through 20VG-1). This will, as was written above, most likely result
in a "bang" of some magnitude when the fuel is ignited.

1. It is quite evident that SRV needs mechanical overhauling but OEM (Hitachi) does
not recommend on-site overhauling and advises to send existing GCV-SRV set to
their factory and in the meanwhile, purchase a new set from them as a 'roll-in
roll-out' concept. This is, of course, a very expensive solution. Any
recommendation or way out regarding this will also be appreciated.

LOTS of sites around the world perform their own SRV/GCV maintenance. There are kits
for it. I'm sure the OEM would LOVE to sell a complete SRG/GCV assembly, but, if you
have a competent mechanical department or have a local shop that you have
confidence in--it's not a difficult task. The most dangerous part is opening the "lid" of
the assembly without launching the springs and hurting someone (and, then, of course,
closing the lid and compressing the springs when reassembling). But, it's been done in
the field hundreds of times, maybe even more--this combined SRV/GCV assembly has
been used around the world for decades.
 
Like Reply

 #3

Hi Makster,
All of the SRV's I've worked with have a elastomer seat. If it does not seat properly,
chances of further deterioration is also high. And soon after a shut down, you will be
able to replace the seat within couple of hours while turbine in turning gear). In my
opinion, it would be worthwhile to replace the seat rather dealing with specially over
speed protection. Sice GCVs are class 4 leakage class, it is important to have a tight SRV.
Pls ignor if your SRV does not contain a elastomer seat.
regards,
 
Like Reply
Makster,

I wanted to also add that 20VG-1 vents through what's usually referred to as a "goose-
neck vent." (It vents through some type of atmospheric vent, regardless of what it's
called; just follow the tubing and you will find the atmospheric vent. 20VG-1 is a
normally-open solenoid-operated valve, and it is de-energized when the unit is at zero
speed. 20VG-1 is usually a pretty small solenoid valve, with either 3/8-inch or 1/2-inch
tubing connections. So, it doesn't pass much fuel when it's opened.... It's just there
mainly to relieve the pressure between the SRV and GCV so that during starting/firing
there isn't fuel/pressure trapped between them which can cause a "hard" light-off (a
bang of some magnitude; I have seen the sides of HRSGs bulge because of SRV leakage
when firing!!! it's pretty scary!!!).

SO, if pressure is building up between the SRV and GCV when the unit is at zero speed
that means the SRV is leaking so much that 20VG-1 can't alleviate the fuel/pressure fast
enough to keep the pressure to near zero. Which means the leak is NOT good.

As was written above, usually, when the unit starts turning (breaks away from zero
speed) during starting, 20VG-1 is energized to close it--and it stays closed during
starting and accelerating and loaded operation. SO that means that is the SRV is leaking,
and leaking "badly" that pressure/fuel will build up between the SRV and GCV during
purging and prior to firing....

My suggestion is to check the atmospheric vent to try to determine how much fuel is
being passed by 20VG-1 when the unit is at zero speed. You could put your hand over
the opening of the atmospheric vent to see how much pressure builds up as you hold
your hand over the opening. You should also try forcing 20VG-1 to a logic "1" (to close
20VG-1) while the unit is at zero speed and monitoring the pressure between the SRV
and GCV. If it increases quickly you know the leak past the SRV is pretty serious. You
should ALSO be watching the P2 pressure during the next start--because if it's
increasing above even 1 barg that can lead to high exhaust temperatures at light-off
and the "hard" light-off being referred to. Hard light-offs are not good for combustion
hardware (liners; transition pieces; nozzles) or exhaust components (diffusers; transition
ducts) or HRSGs (if the unit exhausts into a "boiler").

This is pretty serious stuff. I know of a site that was having such hard light-offs that the
induced flow through the axial compressor was causing the inlet filter house implosion
doors to be sucked open!!! It was a simple-cycle unit with a short exhaust stack so the
"bang" wasn't as noticeable as with an HRSG but damage was found on some of the last
stage blades (rotating and stationary) of the axial compressor. (This had been going on
for years.... Good thing it was a peaker unit and didn't run very much. But, it was started
once every month and people were aware of and watched for the opening of the
implosion doors. (The springs on the implosion doors are VERY strong--and it takes a lot
of pressure differential to cause them to open, and then when they close--they SLAM
shut with a lot of noise). Hard to believe it was ignored for so long.... Frame 5s are
BEASTS and can take a lot of abuse but after so long there is ultimately going to be
damage. The rotor had to be removed and the bill was more than USD 1,000,000.000 for
repairs to the unit after all was said and done.

Hope this helps!