8 replies to this topic

[barbieros]

Dear friends,

 

We are facing some difficulties to specify a pressure relief valve for a tank.

It's a feed tank of a sour water treatment unit. The pressure control is done by injecting fuel gas in case of low pressure or relieving to flare header in case of high pressure.

 

The pressure relief valve is a pilot-operated valve. The suppliers state that the material in this pilot is an elastomer that doesn't resist to some kind of contaminents.

 

The elastomer provided is called EPR and it doesn't work well with hydrocarbons. Since our tank has a fuel gas control system, this will be a problem. We thought to change the elastomer to other kinds, but this other problems raised (high H2S content, low pressure, etc).

 

Our first alternative is to provide a flushing in the inlet of PRV to assure that elastomer won't be in permanent contact with fuel gas. But we have no idea how to calculate the required flow of flushing. The inlet pipe diameter of PRV is 6".

 

Is there any literature that can guide us or does someone have experience in this topic?

 

Best Regards,

 

Marcelo

[fallah]

Marcelo,

 

The reason of lack of possibility to change the elastomer to other kinds isn't so clear. Indeed, the flushing is normally applied in upstream/downstream lines of PSVs to prevent any plugging in those lines and PSVs theirselves to be used in dirty/high viscous services. But appears to prevent any contact between a gas and a material, here an elastomer, is permanently subject to that gas, flushing cannot be an effective operation and excessive efforts should be done to change either the material or PSV type...

[barbieros]

Fallah,

 

I actually don't know all the elastomers offered by suppliers, but as I was informed, two other types were discussed and discarded:

- Viton, which does not resist H2S .

- Kalrez, which is not so good in low pressures such ours (PRV set = 1kgf/cm² g)

 

We already have a vapour flushing located at inlet line of PRV. But we are not sure if the RO installed will give enough flow.

That tank is not operating yet, but PRV is already bought, so there is not so much we can do about replacing it.

 

Our final solution will be changing the pressure control system from fuel gas to nitrogen, but still fuel gas will be an option of injection, so we would like to guarantee enough vapour flushing anyway.

 

Why do you think flushing cannot be effective?

We read in literature that we should provide enough flow to guarantee a velocity of 1m/s. But this velocity seems too high for us (1m/s in a 6" pipe!). The RO we have in this system will give a vapour velocity in 6" pipe of 0,08 m/s.

 

Is there any velocity we should consider to specify a new RO and replace it?

 

Thank you for your informations.

 

Marcelo

[fallah]

Marcelo,

 

Would you please upload a simple sketch of the tank system. It would be more helpful...

[barbieros]

The tank has 2 PVs controlling the pressure (DCS system). One opens to flare, other receives fuel gas.

There are also two PCVs (auto-operated), as a protection in case of failure in control system, one opens to flare, other receives nitrogen.

PRV is designed to protect tank in case of fire.

 

The sets are:

PCV (N2) = 0,1 kgf/cm²

PV (FG) = 0,5 kgf/cm²

PV (Flare) = 0,7 kgf/cm²

PCV (Flare) = 0,8 kgf/cm²

PRV (Flare) = 1,0 kgf/cm²

 

We want to change fuel gas to nitrogen in control system (PV), but this will take time (project, new pipes, valves, etc.).

For now, what we can do is changing set points of controllers, setting PCV (N2) higher than PV (FG). This will keep nitrogen inside of tank but still FG can be sent.

As another protection, we want to calculate vapour flow as flushing. This flushing already exists, but we don't know if it is enough (12 kg/h).

 

Thank you,

 

Marcelo

 

 tank-prv.PNG   4.29KB   2 downloads

[barbieros]

Fallah,

 

Do you have any suggestion for us?

 

Best Regards,

 

Marcelo

[curious_cat]

 

- Kalrez, which is not so good in low pressures such ours (PRV set = 1kgf/cm² g)

 

 

You mean to say Kalrez is OK at high pressures but not at low pressures? Sounded non intuitive. 

 

Is it a gasket seating force issue?

Edited by curious_cat, 29 November 2013 - 05:37 AM.

[fallah]

Marcelo,

 

You can change the fuel gas to nitrogen in control system (PV) provided that: 

 

- The new set pressures will be in PVs and PRVs set pressures ranges as per vendor technical data...

- The flare system be able to tolerate N2 included in relief through PV open to flare in new conditions... 

- Adequate N2 be available to compensate higher required consumption in new conditions...

 

Indeed, flushing vapor rate might be determined by an online analyser temporarily installed such that take a sample feed from PRV ustream while the vapor rate is increasing till show no fuel gas while fuel gas is incoming through the PRV (new condition) as a worst case in fuel gas consumption standpoint... 

 

Hope above help you out...

[barbieros]

curious_cat:

 

Supplier informed that they do not guarantee Kalrez because of sealing problems. I don't have more information about it. So they only offered EPR and Viton.

 

Fallah:

Thanks for the help. We will try to change to N2.

Installing an online analyser would be the ideal solution, but this would never be approved unfortunatelly.