HRSG Water Hammer at Start-up
HRSG Water Hammer at Start-up
(OP)
I've been looking at an issue here at my facility for the last 4 months or so. I'll start with some background, we have a GE LM6000 turbine with a EIT HRSG behind it. Back in June of this year, a loud intermittent banging was observed by one of our personnel. I began investigating that and found that cause after a bout 2 weeks (leak by on a blowdown line going to a flash tank would cause occasionally cause it under the right conditions). We resolved that issue; however, while I was investigating that issue, I found a much more frequent and severe water hammer happening every time we start the unit (we had 200+ starts in the last year).
When we startup, the feedwater line is pressurized against closed control valves (Class 5 or 6 shutoff). Exhaust gas begins to flow and heat the HRSG. During this time, the LP FW temp begins to drop (it typically loses about 120+ degrees). The temp probe is downstream of the closed LP FW control valve just before the water enters the drum. The water upstream of the control valve typically cools off while the unit is off, since the line comes from the economizer to the CV outside the unit, though it is insulated. To me, it seemed that leak by of the control was the most likely issue. We were introducing cold water from the line and flashing it across the CV. Additionally, the volume of water in the FW line that is external to the unit is ~50 gallons. I did some other testing where we pressurized the line and used the reading from the flow meter and it seemed to line up that once the cool water was evacuated from the line, the hammering stopped.
However, we did another test were we did not pressurize the FW line. We started the turbine and began heating the HRSG. During this test, the same temperature drop and water hammer were experienced. I have attached trends from both scenarios (top one is with pressurizing FW before firing the turbine) where you can see a slight increase in pressure in the FW, but nowhere near the normal operating pressure. I cannot figure out why this would occur. I have been in communication with the HRSG OEM and they do not seem to have any idea what may be causing it. I have sent them full data from all points on the HRSG and FW systems and they haven't come to any conclusions.
We also had the unit open for annual inspection last month and found the FW header inside the LP drum was completely loose from it's supports. I borescoped the pipe and didn't see anything that would concern me. We reattached the header to the supports and just started the unit last Friday with the same results as before (large drop at LP FW temp probe and water hammer). I also spoke with some of our senior maintenance people and they stated this has happened since the unit was new (about 7 years old), but it was never investigated. Since you can only hear it when you are halfway up the HRSG stair tower, nobody noticed or cared enough to look for a root cause.
I don't want to start throwing new components or doing mods to the unit unnecessarily. I still believe CV leak by is most likely cause. However, to confuse matters even more, we closed all the manual isolation valves at the CV station and pressurized the line. The same temperature drop downstream of the CV was observed, but no hammering occurred. I attached piping arrangement drawings from the economizer outlet to the LP drum inlet for review. All of that said, anyone have experience with anything like this?


When we startup, the feedwater line is pressurized against closed control valves (Class 5 or 6 shutoff). Exhaust gas begins to flow and heat the HRSG. During this time, the LP FW temp begins to drop (it typically loses about 120+ degrees). The temp probe is downstream of the closed LP FW control valve just before the water enters the drum. The water upstream of the control valve typically cools off while the unit is off, since the line comes from the economizer to the CV outside the unit, though it is insulated. To me, it seemed that leak by of the control was the most likely issue. We were introducing cold water from the line and flashing it across the CV. Additionally, the volume of water in the FW line that is external to the unit is ~50 gallons. I did some other testing where we pressurized the line and used the reading from the flow meter and it seemed to line up that once the cool water was evacuated from the line, the hammering stopped.
However, we did another test were we did not pressurize the FW line. We started the turbine and began heating the HRSG. During this test, the same temperature drop and water hammer were experienced. I have attached trends from both scenarios (top one is with pressurizing FW before firing the turbine) where you can see a slight increase in pressure in the FW, but nowhere near the normal operating pressure. I cannot figure out why this would occur. I have been in communication with the HRSG OEM and they do not seem to have any idea what may be causing it. I have sent them full data from all points on the HRSG and FW systems and they haven't come to any conclusions.
We also had the unit open for annual inspection last month and found the FW header inside the LP drum was completely loose from it's supports. I borescoped the pipe and didn't see anything that would concern me. We reattached the header to the supports and just started the unit last Friday with the same results as before (large drop at LP FW temp probe and water hammer). I also spoke with some of our senior maintenance people and they stated this has happened since the unit was new (about 7 years old), but it was never investigated. Since you can only hear it when you are halfway up the HRSG stair tower, nobody noticed or cared enough to look for a root cause.
I don't want to start throwing new components or doing mods to the unit unnecessarily. I still believe CV leak by is most likely cause. However, to confuse matters even more, we closed all the manual isolation valves at the CV station and pressurized the line. The same temperature drop downstream of the CV was observed, but no hammering occurred. I attached piping arrangement drawings from the economizer outlet to the LP drum inlet for review. All of that said, anyone have experience with anything like this?







RE: HRSG Water Hammer at Start-up
RE: HRSG Water Hammer at Start-up
I'm amazed you didn't read my whole post. From paragraph 3 in my original post:
RE: HRSG Water Hammer at Start-up
RE: HRSG Water Hammer at Start-up
Did you confirm the low spot in the piping downstream of LCV-1325 was drained each time?
What is the connection E59, which seems to tie-in downstream of LCV-1325? Is that a chemical injection line?
RE: HRSG Water Hammer at Start-up
I agree, but they have no reason to. I told them I just want to fix the issue at our cost.
The hammering is in the FW line at the LP drum.
No, did not confirm each time. I believe it is typically not drained since we start and stop so often.
Chem feed.
RE: HRSG Water Hammer at Start-up
RE: HRSG Water Hammer at Start-up
I am going to confirm with operations if they drain that line. Most times, it would refill because of level maintained in the drum (above 50) that would flow back into that line. One other item to note, operations previously stated that they will fill the drum just by pressurizing the FW line and letting leak by fill it (apparently they have been doing this for years). No heat is added during this time. I asked them to check right after they started and right before they stop to see if the hammering is there and sure enough it was.
RE: HRSG Water Hammer at Start-up
I believe the two circumstances, while similar, are different occurrences.
RE: HRSG Water Hammer at Start-up
http://65.57.255.42/ubbthreads/ubbthreads.php?ubb=...
You need to check the P&ID drawings with the source of vibration, and go through the options that can produce water in the steam lines/tubes during the start-up. It is too difficult to guess. Check if there is drain pocket in the discharge side of the headers and the drain valves functioning.
RE: HRSG Water Hammer at Start-up
No need to check with P&IDs. I am 100% certain the source is in the feedwater line between the control valve and the LP drum inlet (a total of ~ 25 feet of pipe).
RE: HRSG Water Hammer at Start-up
RE: HRSG Water Hammer at Start-up
I believe that the temperature drop we're seeing is the result of TT-1130 being "pre-heated" by the drum prior to operations pushing "cold" feedwater to bring up the level - TT-1130 is located directly adjacent to the drum connection. After the feedwater pressure drops back to zero, the TT reading begins to recover (albeit slowly) to its previous reading until feedwater is again re-introduced, where we see an almost identical drop again. Given how close the temperature transmitter is to the drum, and how much heat it is absorbing prior to flow being established, I take back my previous recommendation about draining that segment prior to start-up...better for the temperature transmitter that it remain flooded, most likely.
RE: HRSG Water Hammer at Start-up
I agree and that has been my thought from the start. But that one test where we didn't pressurize the line and still had the same symptoms is what is making me second guess it.
I have considered heat tracing the external section of line to warm the feedwater so there is not as much temp difference between the upstream and downstream side of the CV at initial start. Otherwise we could purge the line through drains and dump the 50 gallons of cool water.
RE: HRSG Water Hammer at Start-up
RE: HRSG Water Hammer at Start-up
That's very interesting - I assume the control valve was closed during this test as well. Do you suppose that the noise is being generated by the contraction of the hot pipe near the drum as the "cold" feedwater enters it? It's not immediately clear to me how it's supported from the drawings your originally posted, but I suppose that's a possibility albeit for a short run like that I would find it surprising.
Looking at your recent test where the hammering occurred, it appears the LP FW pipe saw a quick drop in temperature of ~140 °F. For 4" SA105 pipe, it's contracting about 2" per 100' of pipe.
RE: HRSG Water Hammer at Start-up
I did another test yesterday after my last post and again this morning as things cooled off. For the first two tests, the FW temp probe recorded drops of ~ 50 F and 90 F. No hammer occurred during either of these. The test this morning had a drop of 130 F, but I terminated it due to the amount of hammering. These drops were what I expected based on the HP econ outlet temp (which is approximately equal to the "cold" water temp in the section of external line). As soon as we depressurized the line, the hammering stopped (due to no flow is most apparent reason). The hammering started right around 100 deltaT from the starting point.
RE: HRSG Water Hammer at Start-up
RE: HRSG Water Hammer at Start-up
RE: HRSG Water Hammer at Start-up
Of course, the operations department could add it to their rounds to check the delta T and drain the line manually - manpower intensive and, as often as you appear to be starting and stopping, likely not the most effective approach.
RE: HRSG Water Hammer at Start-up
As far as ops checking this on rounds, they barely do anything on rounds, so if I put this on the list, I know it would never get done.