5 replies to this topic

[ZAS]

Dear All,

I hope the expert would be able to help me here.

See attached schematics and ISO. The nature of our operation require frequent start and stop of a High Pressure (HP) pump. The line is maintained full of LNG at -153 deg C and 8 barg. At start up the, HP pump, pressure rises to 98barg in 10seconds.

In the attached schematic, the yellow highlighted is the 2" TRV line which vigorously vibrate due to surge from HP pump start-up. The Isometric of the TRV line is also shown in the schematic attached, which has about 5 elbows.

To avoid this vibration, we have installed a valve on the TRV line. When starting up the pump, we isolate the TRV. Once the discharge valve is opened, then an operator can bring back the TRV line in service.

We think this is not a good solution as we keep isolating a TRV on pump start-up and we have introduce human factor in the operation.

I will appreciate if you can elaborate on the physics inside the pipeand suggestion on how we can re-engineer this to eliminate the surge.

Kind Regards

ZAS

Attached Files

•  Drawing Tool.xls   1.14MB   39 downloads

[kkala]

A non expert trivial view, hoping of more effective help.
1. Pump head is about 11 bar, so its normal discharge pressure could be about 11+8 = 19 barg (plus static head from LNG tank). After start up it develops 98 Barg in 10 s, so there must be water hammer in the discharge line (8" to vaporiser & 6" of kickback flow).
2. TRV set pressure is 148.4 barg, so we would expect no flow in the 2" line even during pump start up. Yet vibrations in this line indicate flow at rather excessive rate (neglecting resonance). Vibrations may be intensified by the five elbows, assumed to be upstream TRV.
So either TRV set pressure is wrong, or developed discharge pressure is not 98 Barg but 148.4 barg or higher. The latter could expose piping in pump start up to pressures higher than design pressure.
You may check Thermal Relieve Valve (TRV) first (calibration, leakage), then measure discharge pressure at start up and see how to reduce water hammer (surge) at the discharge line, to either evaporator or to LNG tank (kickback is assumed to return there). The pressure increase is rather high, probably the reason can be traced. Water hammer analysis can quantify the results, but usually you have to locate the "suspect" first. Following merely concerns some thinking on it.
3.1 Pump shut off can be about 8+1.25x11~ 22 barg (plus static head from LNG tank); probably a start up with discharge valve closed, then gradually opened, could alleviate surge. Is this action realized by the "FC" valve? If so, the valve had better be at manual mode in start up, slowly opened. But why a valve at discharge is CSO?
3.2 If pipe is really full of liquid, chances of water hammer are reduced, especially at start up. You may want to verify whether whole pipeline is full before start up.
3.3 I assume that the shematic is simplified; there must be additional valve at discharge to justify the TRV. Check valves at discharge (if there are not) could alleviate water hammer, reducing back flow. This concerns mainly pump stop; but back flow may empty parts of the pipeline, when pump is idle.
3.4 If surge on discharge line is settled, most probably vibrations on the 2" line will disappear at start up. They may appear when the TRV gets in function, which will last few seconds to my understanding (as thermal relieve valves do).
3.5 It is agreed that present situation (isolating PRV in start up and bypassing it) is not recommended, even though TRV contribution to pressure relief would be insignificant (pump flow = 456 m3/h).

Edited by kkala, 15 June 2011 - 06:53 AM.

[ankur2061]

ZAS,

The service is LNG which is a clean fluid. You can think of having an expansion bellow in the inlet line to the TRV. This definitely should help in dampening the vibrations due to surge during start-up instead of closing the inlet valve. A vibration analysis will be required for the entire piping configuration to select the right kind of bellow for vibration dampening.

Regards,
Ankur.

[cnaren]

You probably have some vapour formation when the pump is stopped. If the TRV line goes vertical to a higher elevation, that probably explains the vibration - At start-up, the vapour try to rise into the "dead" zone, while the liquid tries to come down, giving you the hammering effect.

You probably hit the right solution by providing a bypass to TRV, which acts as a vapour vent. To remove the human element, make this valve open automatically when the pump is switched on, and close using a timer after a few seconds.

Dear All,

I hope the expert would be able to help me here.

See attached schematics and ISO. The nature of our operation require frequent start and stop of a High Pressure (HP) pump. The line is maintained full of LNG at -153 deg C and 8 barg. At start up the, HP pump, pressure rises to 98barg in 10seconds.

In the attached schematic, the yellow highlighted is the 2" TRV line which vigorously vibrate due to surge from HP pump start-up. The Isometric of the TRV line is also shown in the schematic attached, which has about 5 elbows.

To avoid this vibration, we have installed a valve on the TRV line. When starting up the pump, we isolate the TRV. Once the discharge valve is opened, then an operator can bring back the TRV line in service.

We think this is not a good solution as we keep isolating a TRV on pump start-up and we have introduce human factor in the operation.

I will appreciate if you can elaborate on the physics inside the pipeand suggestion on how we can re-engineer this to eliminate the surge.

Kind Regards

ZAS

[kkala]

You probably hit the right solution by providing a bypass to TRV, which acts as a vapour vent. To remove the human element, make this valve open automatically when the pump is switched on, and close using a timer after a few seconds.

My original interpretation of "we isolate the TRV" is that the valve downstream TRV (in circle on the sketch) is closed and the valve bypassing TRV is also closed. cnaren has given an interesting interpretation in case that valve bypassing TRV is opened during start up. Probably this is what it happens with the valves.

Edited by kkala, 16 June 2011 - 12:43 AM.

[Profe]

Hi ZAS

What is the specification used to install this TRV and why?

What is the purpose of this relief pressure?
Is to protect downstream equipment from overpressure based on centrifugal pump shutoff pressure?
Does the NGL is bottled? Which is the destination of the kickback line?
Perhaps what is needed is a PRV, and not requires a TRV?

For your concerns:

A thermal relief valve is used when a liquid is blocked (it est: a line or pipe between two block valves) in a pipe, and the pipe is exposed to sunlight. Although liquid does not expand much, since it fully fills the pipe, the forces developed can seriously stress the pipe. There is no difference between a safety relief valve for any other purpose and a thermal relief valve. However, it is usually small in size, "D" orifice. For a screwed valve this would be 3/4 x 1 inch.
Sizing a thermal relief valve is generally neglected since the smallest orifice size available is generally considerably (10 to 100 times) greater than the required capacity. However, if sizing is required, review API 520 and API 521, sec. 3.14.

From API 521:

3.14.2 SIZING AND SET PRESSURE
The capacity requirement is not easy to determine. Since every application will be relieving liquid, the required capacity of the relieving device will be small; specifying an oversized device is, therefore, reasonable, A ¾-inch x 1-inch nominal pipe size (NPS ¾ x NPS 1) relief valve is commonly used. If there is reason to believe that this size is not adequate, the procedure in 3.14.3 can be applied. If the liquid being relieved is expected to flash or form solids while it passes through the relieving device, the procedure in 3.20.1 is recommended.
Proper selection of the set pressure for these relieving devices should include a study of the design rating of all items included in the blocked-in system. The thermal relief pressure setting should never be above the maximum pressure permitted by the weakest component in the system being protected. However, the pressure-relieving device should be set high enough to open only under hydraulic-expansion conditions. When thermal relief valves discharge into a closed system, the effects of back pressure should be considered.

Rule of thumb: The simple selection, just get one that will relieve before the design pressure is exceeded. Not like a safety relief valve that is the subject of a considerable study.

I think that a little bit of knowledge will make us the best
Good luck.

Fausto

Edited by Profe, 17 June 2011 - 10:37 AM.