Produced Water tank blanketing 2

[enguest]

Gents,

Now I am working on the design of Produced Water tank blanketing by fuel gas. Tank to be connected to the LP flare. Flare system operation pressure is 0.1barg. I would like to supply the fuel gas with pressure below 0.1 barg to avoid sweeping and continious lose of fuel gas to the flare.
Is it right and could you please advice any industrial standard which is apprpriate for this case?

Thanks.

 

[HeinrichK]

I suspect you're talking about an atmospheric storage tank. Consider the following:
-Tanks can't handle well negative pressure. Filling it with fuel gas makes it hard to use a straight-forward vacuum breaker (FG + air = awkward).
-Hanging an atmospheric tank to a LP flare leads to a very unwanted back-flow scenario. (or overpressure scenario during relief)

So if the produced water is a clean service, why not use nitrogen blanketing to atmosphere?

Regarding the standard, I guess API 2000 will help you on the storage tank design...

 

[zdas04]

I've worked with thousands of produced water tanks over the years. Two tanks were blanketed with fuel gas, three were blanketed with nitrogen, the rest were open to atmosphere. The stated reason for the blaketing on the 5 tanks was corrosion mitigation, but corrosion in the head space on produced water tanks is exceedingly rare.

If you just must do it, it is your gas that you are using and you need to do your sums very carefully. If the pipe to the flare is too small then you could easily rupture the tank during a rapid inflow scenario. If the fuel-gas supply pipe or regulator is too small you can collapse the tank on a pump-out. API tanks have an amazingly small allowable pressure range for the head space and it takes less than 2 psig positive pressure to launch a tank lid (a 15 ft diameter tank with 2 psid is exposed to 51,000 lbf) while -1 psig will crush it like a beer can.

Good luck with this.

David

 

[mfelzien]

enguest,

In API 2000 you will find what you need. There is a simple calculation to handle the fire-case and off breathing required (This calculation will need to be issued with the process datasheet). Something about seismic requirements also. I Also remember something about NACE and the importance of the control valve for the off gas line.

Hope this helps a bit.

Yours,

Mike F.

 

[enguest]

Gents,

Thank you very much for all your help.

HeinrichK,
I would prefer to use nitrogen as well but tank is quite big, so client doesn't want to put nitrogen in.
Back flow scenario is possible and now I am trying to investigate the consequences of this back flow.
I could find only one problem with the backflow - sucking the air from the atmosphere, but this can be not an issue because flare is big. If you have any other points on this could you please write about them?
Unfortunately i not working on the design of this tank, i am working only on the installation of blanketing system and would like to get standards about blanketing to base my blanketing system desing on these standards.

zdas04,

These are good points, I checked pipes sizes before but will check them again.

Design pressure of this tank is only + 0.4 BAR(G) TO 0.0025 BAR(G) VACUUM, so i am restricted to play out of this range.
Setting the pressure below the LP flare operating pressure I am sure that sometimes the backflow will occur (see above) since the maximum back pressure from the flare is 0.3 barg.

mfelzien

I will definitely check the NACE. Fire case is not appropriate, but the rest of possible cases i took into accout (thermal - API 2000 Table 2B, pumping and fuel gas line control valve failure).

 

[HeinrichK]

The problem wth backflow from the flare is that you don't know what's in it. condensable HCs? H2S?, NH3? Other things that can pollute the water or corrode the tank?

I don't think there's a general accepted standard on blanketing, it is just good engineering practice. (with API 2000 giving the required in/out flows)

@Zdas4: are the fuel gas blanketed tanks open to the flare or is there a control valve between tank and flare? And if there is a valve how do you protect against vacuum?

 

[zdas04]

The very day that I took over responsiblilty for those two tanks I turned off the fuel gas and opened the thief hatch until I could get a proper Anardo valve installed. Back then onshore gas wells rarely had a flare, so the backpressure valve on the exhaust just went to atmosphere. The supply regulator must have been big enough to prevent vacuum since it had been on the tank for 4 years before I removed it.

David

 

[HeinrichK]

Simply relying on a pressure regulator wouldn't make me feel comfortable. I've seen several pictures of tanks looking like crushed cans...

 

[JimCasey]

Let me propose a simple system:
>Tank
>Fuel gas source assumed to have adequate capacity to supply the needs of the tank
>Blanketing regulator, also sized for the needs of the tank
>COnservation vent with pipeaway to flare, and flame arrestor on the vacuum inlet. Seems like I remember Protectoseal makes one.
Calculate the demands of the tank per API2000. The maximum inflow occurs on a hot afternoon when a rain-shower suddenly chills the tank and the internal atmosphere contracts. Size the regulator and vacuum valve appropriately. The vacuum relief valve should NEVER have to work, but it protects the tank if you lose fuel gas or the blanketing regulator. You may need an emergency vent for fire case. These are basically just weighted manhole covers that pop open if the pressure(volume) exceed the capabilities of the vent to flare.

So in a normal case, the atmosphere displaced by pump-in flows thru the conservation vent to the flare. Backflow is prevented 'cause the conservation vent is unidirectional.
Replacement of tank atmosphere on pump-out is handled by the blanketing regulator.
If the blanketing gas system fails, the vacuum relief valve allows air in to protect the tank. Flame arrestor prevents ignition source from entering.
If there is a fire and the PW boils off, the emergency vent opens to protect the tank. Steam will displace any air in the tank and internal deflagration will not happen.

 

[zdas04]

This is a produced water tank for gods sake. It is unlikely to burn (and if it does, the fire will fail the thin walls and dump a tank load of water on the fire). I don't know why you would blanket it. I don't know why you would spend this much creative energy on it. It is just a water tank.

David