Does dry gas PSV always need drainage on discharge pipe ? 1

[AutoDirt]

For a project im working on, I have a dry gas line that has a PSV on it.

The end of the discharge is a gooseneck so that rain water cannot freely enter

When I read API 520 and 521, it mentions draining on discharge pipe is needed if it’s not free draining and if liquid accumulation can occur

This is a dry gas and rainwater cannot freely get in

However, should I also consider if moist air can condense in the discharge pipe this forming a head of water ?

Thanks !

 

[The Obturator]

We're going to need more info to advise on this. Maybe a sketch/photo? What is the valve type and size (inlet x outlet), set pressure, how is this 'gooseneck' attached and what material is it? (This goose could fly off very easily during valve opening).

What API is basically cautioning you on, is accumulated liquid that could possibly build up and over the seat height (roughly the centre of the outlet of the valve), to act as a head of pressure. ie., a back pressure that could affect the opening (set) point.

Depending on your valve design, internal contours/depths around the nozzle etc., it would be wise to have an open vent at the lowest point. That is if there is indeed a possibility of a liquid build up as described above.

 

[LittleInch]

A small weep hole is usually recommended to avoid any liquid build up which can also shoot out at high speed if the valve lifts.

A 180 degree bend sounds not a good idea for a PSV as velocities could easily bend this or snap it off. OK for vents and very low velocity gases.

 

[AutoDirt]

Hi everyone

Thank you all for the responses

We are at the beginning stages of this design so don’t have a lot of info now but basically it’s on PSV at a hydrogen unit at a refinery.

CGA-G-5.5 gives a common design for the vent outlet to look like the one above in the photo

I’m having a hard time picturing how rain can get into this

Material: 316SS

Set pressure: 400#

Hydrogen goes through a dryer upstream which has a target dew point of -90F with an interlock of -40F

We don’t have the valve info finalized yet

Thank you all for the input

 

[georgeverghese]

And what would happen if this dryer was completely deactivated / poisoned, and the dewpoint analyser was not working - what would dewpoint be then leaving this PSV?

 

[AutoDirt]

Roughly +20F

That may seem high but ambient temperatures even in the area I’m at don’t get that low

 

[LittleInch]

You can get dew/ condensation and sideways rain.

But there is a big difference for me between a "vent" which implies low velocities below 2 to 3 m/ sec and a PSV release which could easily be much higher velocity.

 

[The Obturator]

Although you state that this type of vent pipe is a typical design from CGA-G-5.5, it is not of a typical outlet pipe for Pressure-relief Valves meeting API-520,-521,-526, ASME VIII, XIII etc.

Note I do not have CGA-G-5.5 to comment more on.

 

[georgeverghese]

You may most likely have to purge this PSV exit line with N2 to prevent flashback / detonation for this highly flammable H2. Purging this line with N2 is much preferred to installing a flame arrestor / detonation arrestor which is often prone to blockage. N2 flowrate should enable a line velocity at the exit some what higher than flame velocity for H2, which is 2.8m/sec in order to prevent H2 flashback. N2 flow should be continuously monitored at DCS and a control room alarm should be enabled for low / no flow of N2. be enabled Fit an screw on insect screen at each exit. Ask a process safety engineer if there are other special requirements for this exit pipe in H2 vent service.

Is the H2 recycle gas compressor ( upstream of dryer) oil lubricated on the service gas side ? If so, drill a weep hole at the low point of the PSV tail pipe.

At some added minimal risk, if you dont have N2 to purge, and there are no other vessels in the exit path of the PSV, choose a 300lb rated PSV and 300lb piping to withstand detonation in case of accidental ignition.

 

[shvet]

For dangerous clean gases (N2, H2, CH4, steam) we practice to connect a weep hole to a small-bore drain and route that downward to a safe location that takes into account the exposed dispersion areas. There are reasons to increase metallurgy of the drain to prevent rust blockage.
As an option we locate a PSV and the weephole in an remote zone that is blocked by an administrative mean. Usually this is an isolated area chained and tagged with an alram sign that a zone behind is unsafe. I know that some users use wire mesh for fencing such kind of a exposed area.

Note that your exhaust has not a toroidal ring which is the industry-recognized practice preventing H2 self-ignition.
Note that a weephole exposed area can be surprisingly extensive taking into account the PSV's built-up pressure and unfavorable wether conditions.

Can you comment - if this vent to be installed in a refinery then what's wrong with APIs&ASMEs? Why CGA?

 

[LittleInch]

You may most likely have to purge this PSV exit line with N2 to prevent flashback / detonation for this highly flammable H2. Purging this line with N2 is much preferred to installing a flame arrestor / detonation arrestor which is often prone to blockage. N2 flowrate should enable a line velocity at the exit some what higher than flame velocity for H2, which is 2.8m/sec in order to prevent H2 flashback. N2 flow should be continuously monitored at DCS and a control room alarm should be enabled for low / no flow of N2. be enabled Fit an screw on insect screen at each exit. Ask a process safety engineer if there are other special requirements for this exit pipe in H2 vent service.

Is the H2 recycle gas compressor ( upstream of dryer) oil lubricated on the service gas side ? If so, drill a weep hole at the low point of the PSV tail pipe.

At some added minimal risk, if you dont have N2 to purge, and there are no other vessels in the exit path of the PSV, choose a 300lb rated PSV and 300lb piping to withstand detonation in case of accidental ignition.

Good point. When we're looking at vents and relief lines for hydrogen service (which has suddenly become a major source of work for pipeline studies) we now assume it will catch fire and become a flare due to the very low ignition energy of Hydrogen. You don't want fire pointing downwards.

 

[shvet]

You don't want fire pointing downwards.

(1) a very expensive, (2) complicated, (3) occupied a lot of space, (4) requiring a personnel attention, and (5) not a robust mean creating (6) a dangerous O2 depleting zone around a the vent exhaust and PSV drain both - it should be added.

 

[LittleInch]

For dangerous clean gases (N2, H2, CH4, steam) we practice to connect a weep hole to a small-bore drain and route that downward to a safe location that takes into account the exposed dispersion areas. There are reasons to increase metallurgy of the drain to prevent rust blockage.
As an option we locate a PSV and the weephole in an remote zone that is blocked by an administrative mean. Usually this is an isolated area chained and tagged with an alram sign that a zone behind is unsafe. I know that some users use wire mesh for fencing such kind of a exposed area.

Note that your exhaust has not a toroidal ring which is the industry-recognized practice preventing H2 self-ignition.
Note that a weephole exposed area can be surprisingly extensive taking into account the PSV's built-up pressure and unfavorable wether conditions.

Can you comment - if this vent to be installed in a refinery then what's wrong with APIs&ASMEs? Why CGA?

Have you got details, diagram, links for that "toroidal ring" please?

 

[georgeverghese]

This document may be useful as a checklist

https://www.eiga.eu/uploads/documents/DOC211.pdf

This std also recommends a min design pressure of 40barg (class 300 ) for the vent pipe to withstand detonation, and this would include the PSV also. Note it doesnt suggest the use of this toroidal ring on the vent exit.

 

[shvet]

@AutoDirt
some pointers can be found at the EIGA's guide

https://www.eiga.eu/uploads/documents/DOC211.pdf

The snuffing steam design can be found in ExxonMobil in the next post. Note the pictures in the next post - ExxonMobil guides the exhaust upward with no a rain cap.

API 520-2-2020

11.2 Safe Practice for Installation of Drain Piping
Design, operation, and maintenance of drain piping that is part of the discharge system warrant the same level of care that is applied to the rest of the system. The drain piping installation shall not adversely affect the relief device performance. Flammable, toxic, or corrosive fluids shall be routed to a safe location. Procedures or controls shall be sufficiently robust to prevent accumulation of liquids that could prevent the relief device from operating properly. Drain piping may require purging or heat tracing to maintain its functionality.

BP's std. GP 44-80

11.5. Purge gas for flare and atmospheric vent systems
Purge gas systems installed on flare and atmospheric vent systems shall comply with the following general requirements:
a. A continuous purge system shall be used for closed flare and atmospheric vent systems to prevent air ingress and a potential explosion hazard.
It is not necessary to purge individual atmospheric relief devices.
b. In special cases, atmospheric vent systems may use a robust design that is capable of withstanding an internal explosion (detonation). Use of this approach shall be subject to entity EA approval in consultation with fire and blast experts.
It is possible to build a vent system that is sufficiently strong enough to withstand an internal explosion inside the vent system. This is not done in flare systems due to their size and complexity in most processes. However, explosion rated vent systems are occasionally used to eliminate purge gas. The explosion rated design option is best suited for small well pads or NUI offshore platforms, where relatively few items of equipment and a simple process is connected to the vent system.
Internal vent header pressures created by explosion overpressure can be as high as 5 500 kPag to 6 900 kPag (800 psig to 1 000 psig) in the vent system and shall be considered in the design for explosion rated vent systems. This overpressure was calculated based on an 8 in vent pipe approximately 6 m (20 ft) long.

CCPS's Pressure Relief Guideline 1997

5.8. Release to Atmosphere
A broad approach is presented for designing stacks (including pressure relief device tail pipes) for the release of emergency relief effluent directly to the atmosphere ...
5.8.2 General Design Considerations
In every stack installation, careful consideration should be given to two potential problems: accumulation of liquid in lines terminating at the stack, and accidental ignition by lightning or static electricity. To avoid these problems, consider the following measures (API 521):
* The piping to the vent stack should be designed to avoid pockets that could collect liquid, and should be sloped to one or more low-point drains. These drains can be installed to function automatically by use of a liquid seal. The height of the seal should provide a head equivalent to at least 1% times the back pressure at maximum relief load to prevent release of vapors through the seal. When significant quantities of liquid may accumulate, a small vapor-liquid disengaging drum may be installed at the base of the stack. Rain caps on the stack sometimes are used to prevent entry and collection of rain water
* Flammable vapors exiting from a vent stack may be ignited by lightning or static electricity. Experience has shown that vapors with a high hydrogen content are particularly susceptible to ignition by static electricity. Such vent stacks should be grounded and equipped with a toroidal ring on the outlet (Parry 1994; API 521). API 521 also recommends installing a remote-operated snuffing steam connection at the top of the vent stack to snuff out fires caused by lightning or other sources. Since snuffing steam may not be available during some emergencies, the design should take into account the thermal radiation levels at a nearby locations where personnel might be exposed, and to possible flash-back and explosion in the vent line.
The discharge flow rate during an emergency can vary widely, and dispersion effectiveness may be diminished as the flow rate declines. Turn-down in the pressure relief system can be minimized by use of pressure relief valves in place of rupture disk devices; however, the potential for on-off operation with relief valves should be recognized. In some instances, blowers have been incorporated into stack systems to aid dispersion. As indicated in the next section, compensation for lower discharge velocities can be made by increasing the stack height.
Tall stacks should be designed to withstand wind loads and associated flow induced vibration.

 

[shvet]

@LittleInch
See below and bolded in CCPS above . Note that ExxonMobil and CCPS do not distinguish hydrogen and hydrogen-hydrocarbons as API does.

API 521-2020

5.8.4.2.3 Release of Hydrogen-rich Streams
This conclusion pertains to hydrocarbon vapor releases. Experience indicates that streams with a high hydrogen content are susceptible to ignition by static electricity as a result of the described mechanism because of electrostatic discharges at the sharp edge of the vent outlet. NASA investigated this phenomenon [123] and found that such electrostatic discharges can be prevented by installing a toroidal ring on the vent stack outlet. This ring inhibits the static discharge at the vent stack exit by removing the sharp-edged geometry of the vent outlet, which is conducive to spark formation.

Toroidal ring prevents gas ignition at vent stack outlet - NASA Technical Reports Server (NTRS)

Torodial ring welded to the vent stack outlet prevents static discharges which ignite combustible gases in a venting system. The ring inhibits the flow of current by removing the cause of turbulence characteristics of a sharply defined vent exit.

ntrs.nasa.gov

ExxonMobil's std. GP 03-02-04

4.3.4. Addition: Ignition of a Relief Stream at the Point of Emission
1) Snuffing steam is required for PR valves that discharge greater than 50 mole % hydrogen or vapors above their autoignition temperature to the atmosphere. The Owner’s Engineer will specify when flammable vapors below their autoignition point require snuffing steam. Snuffing steam shall be installed as follows:
a) At least a NPS 1 in. (25 mm) steam line shall be tied directly to a separate connection on the discharge riser per Figure A–6. The steam isolation valve, normally closed, and a 3/4 in. (20 mm) bleeder, normally open, shall be located adjacent to each other at a safe location (typically grade) accessible to the operator. The discharge of the bleeder shall be directed to an open drain or catch basin. Or,
b) At least a NPS 1 in. (25 mm) line with a block valve and hose connection shall be separately connected to the discharge riser, per Figure A–7. The block valve and hose connection shall be located within 20 ft (6.1 m) of a steam utility station.
2) The snuffing steam line shall be traced and insulated if vapors leaking from the PR valve contain enough water to condense, freeze, and block the steam line.
3) Discharge risers for PR devices in hydrogen service at greater than 50 mole % hydrogen shall be equipped with toroidal rings according to Figure A–8.

Shell's std. 85.45.10.10

5.8.4.2.3 Release of Hydrogen-rich Streams
This conclusion pertains to hydrocarbon vapor releases. Experience indicates that streams with a high hydrogen content are susceptible to ignition by static electricity as a result of the described mechanism because of electrostatic discharges at the sharp edge of the vent outlet. NASA investigated this phenomenon [116] and found that such electrostatic discharges can be prevented by installing a toroidal ring on the vent stack outlet. This ring inhibits the static discharge at the vent stack exit by removing the sharp edged geometry of the vent outlet, which is conducive to spark formation.
1. The end of the discharge pipe shall be cut off squarely.
2. A toroidal ring shall be installed where hydrogen content of the vapor exceeds 20mol% to minimize the risk of ignition by static electricity.
Table 17: Toroidal ring diameter of Hydrogen-rich streams
Pipe diameter-------------------Toroidal ring diameter
Less than DN200 (NPS8)----------13 mm (1/2 in)
DN200 to DN300 (NPS8 to NPS12)--20 mm (3/4 in)
Greater than DN300 (NPS12)------25 mm (1 in)

 

[LittleInch]

@LittleInch
See below and bolded in CCPS above . Note that ExxonMobil and CCPS do not distinguish hydrogen and hydrogen-hydrocarbons as API does.

API 521-2020


ExxonMobil's std. GP 03-02-04


Shell's std. 85.45.10.10

Thank you. Very useful. Not seen that before but will take note.

 

[AutoDirt]

Wow thank you everyone for the wealth of information and sources!

Keep the good work