11 replies to this topic

[herrani]

Hi all

There is an ongoing discussion at my refinery, regarding best practice for venting/blanketing of underground tanks, and I would very much appreciate your opinion.

We have a number of closed drain system containing hydrocarbons, which are drained into underground storage tanks. The drained hydrocarbons are not expected to be very light, but it depends on which system the drain is coming from. At present we have the following scenarios:

1. Underground tank with a free air vent (equipped with flame arrestor, and to a safe location)
2. Underground tank with a closed vent, connected to the main flare header

The older systems are installed following the first arrangement, but they require some type of area classification around the vent, and it is expected that the tank vapour area has an explosive atmosphere.

The newer systems are installed following the second arrangement, but experience serious problems during flare relief situations, due to backpressure from the flare system, and lack of liquid seal or other methods to prevent backflow of flare gas into the main flare header.

It is my opinion that underground tanks should remain with an atmospheric vent, or be connected to a gas blanketing system other than the flare system in cases where this is unavoidable. Unfortunately, I have not found any standard covering this situation.

I would be very interested in hearing the experience of others from the forum on this issue.

Edited by herrani, 08 July 2010 - 07:08 AM.

[proinwv]

It seems to me that you, in fact, are an expert based upon your experience.

A gas blanketing system, with O2 monitoring and pipe away to a disposal, seems to be a distinct possibility in your case. Of course, this will be more expensive than what is presently being done.

[fallah]

The newer systems are installed following the second arrangement, but experience serious problems during flare relief situations, due to backpressure from the flare system, and lack of liquid seal or other methods to prevent backflow of flare gas into the main flare header.

Back flow from where to where?I think you mean from flare header to slop drum.Please confirm,otherwise clarify your statement.

The slop drum would always be in pressure equilibrium with relevant flare header,because a line including a CSO valve connects the drum to flare header and on the other hand there are check valves downstream of all drain lines routing toward closed drain system.

Would you please more explain in detail about serious problems during flare relief situations.

Edited by fallah, 08 July 2010 - 12:29 PM.

[herrani]

Hi fallah

Yes you are right, the valve connecting the vent of the underground tank to the flare header is locked open, and when there is a flare relief the pressure in the main flare header increases, leading gas into the underground tank.

The high pressure in the underground tank is high enough to lower the level below the inlet point and flare gas propagates to the closed drain design piping (design pressure of underground tank = design pressure of flare tip). The drain points connecting with the process have double block and bleed valves, but we have experienced gas leaks from the closed drain system directly to the atmosphere due to untight flanges and similar stuff.

I also feel that it is not a good idea to use a safety system for blanketing purposes, but have not been able to find any design practice or standard that covers venting and blanketing of underground tanks.

Edited by herrani, 09 July 2010 - 05:38 AM.

[fallah]

Hi Herrani,

You fail to answer my question and clarification request in my previous post.
Untight flange could be changed to tight flange and it isn't a general fault of closed drain system.Piping class of drain lines up to slop drum are equal or higher than the piping class of the flare network which they are connected to.
You didn't mention if check valve is considered in downstream of all drain lines routing toward your closed drain system.

Anyway,eventhough specifying destination of the vents on top of drums of closed drain system among different flare networks (HP,MP,LP,LLP) should be performed accurately,in the systems i have ever seen there had been no such problems you described in this thread.

If any point remained please submit it for better clarification.

[herrani]

Hi fallah

Thank you for your reply. I hope this will provide sufficient clarification.

The closed drain system consists of a number of drain pipes from each piece of equipment, which are connected to a main drain header. All the piping is underground. Each drain point is equipped with double block and bleed valves, which are normally closed, and no check valves. The draining will usually take place under pressure.

There are no valves installed in the main header (manual nor check).

The closed drain header enters the underground drain collection tank, close to the bottom of the vessel.

During commissioning of the system, we experienced high pressures in the flare header due to PSV relief from another section of the refinery, and this resulted in the activation of gas detection alarms in the process area where this closed drain system is located. We are also worried about possible consequences of a leak in the underground piping system.

What I wonder is:

1. Why can't we use a vent to the atmosphere? I can see that this was the original design practice in older places of the refinery (1980s) but that newer systems are not designed this way
2. Is it a good idea to use the flare header (safety system) for blanketing purposes? Somehow, this does not sound correct. Are you saying that you have seen/designed this kind of installations in the past?
3. Are there any applicable standards/engineering design practices that cover this discussion?

Edited by herrani, 12 July 2010 - 03:52 AM.

[fallah]

Hi Herrani,

The document included the "process closed drain design" for a plant is titled "Drainage Philosophy".

Short answer to your questions:

1-For environmental and safety issues

2-I think connecting slop drum's vent to flare network doesn't mean: "Using Flare Header for Blanketing Purpose",it means the vapor of process fluid should be routed to a proper sink,i.e. flare network.

3-Each project considering the nature of relevant process and containing fluids would have its own Design Basis may differs from another project.

Now some additional concept:

Usually two types of drainage would be considered:

-Operational Drainage
-Maintenance Drainage

To me,a check valve is a mandatory device should be installed to prevent back flow due to high pressure in slop drum.

Hope above helps you out.

[herrani]

Hi fallah

Just to make sure that I understand your comments:

1-For environmental and safety issues

The area classification drawings of the underground tanks vented to the atmosphere, and not to the flare show zone 2 only. This means that emissions are minimal. Plus, if we provide flame arrestors in the vent line the risk of explosion will be much lower. Of course, if we were dealing with a very volatile product I would agree with you, but most of the slops systems have heavy hydrocarbons.

2-I think connecting slop drum's vent to flare network doesn't mean: "Using Flare Header for Blanketing Purpose",it means the vapor of process fluid should be routed to a proper sink,i.e. flare network.

There is a pump installed in the vessel to empty it, which is activated automatically with a LH switch. When the tank is emptied, flare gas is used to "refill" the tank vapour space as a blanketing gas instead of air. We believe this is the main reason for choosing this concept, since actual hydrocarbon evaporation is expected to be low.

3-Each project considering the nature of relevant process and containing fluids would have its own Design Basis may differs from another project.

Now some additional concept:

Usually two types of drainage would be considered:

-Operational Drainage
-Maintenance Drainage

Of course, we have our own specifications for closed drain systems, and underground piping, but none of these specifications give any recommendations on how to vent/blanket undergroound tanks. I have also found no recommendations in API, nor Shell or similar design guidelines.

These systems are a combination of both operational and maintenance drainage, but the expected flowrates in the main drain header are expected to be small on average. If we empty some of the big vessels for maintenance purposes, the drained volumes will increase, but this will only happen from time to time.

To me,a check valve is a mandatory device should be installed to prevent back flow due to high pressure in slop drum.

If we install valves in the underground piping system, we will need some kind of valve pit to ensure maintenance. We do not have this arrangement in any part of the refinery. All buried piping is free from valves or similar equipment that could need periodic maintenance. In my head, the only option would be to provide some kind of liquid seal, based on the operating liquid level of the underground tank, but that has not been taken into account during the design phase.

Which brings me back to my original question: if we have low evaporation of the tank contents, why is it a bad idea to use a vent to the atmosphere? It sure is a simpler solution!

Edited by herrani, 13 July 2010 - 03:36 AM.

[fallah]

Hi Herrani,

I meant the check valve to be installed where the drain line is still aboveground and before it is going to be underground.Therefore no need to be worry about valve pit,....

In gas refinery there are high evaporation of the tank contents and your suggestion to use a vent to the atmosphere would be almost impossible.Your suggestion may be applicable for oil refinery but i am not well aware of it.

[herrani]

Hi again fallah

Thank you very much for the discussion.

If we are to install a check valve, it should be just upstream of the underground tank, to avoid the drain system to be filled with flare gas. I agree that this should be possible.

I have seen this type of installation in systems collecting the drains of compressor filters in natural gas export stations as well, and that is why I am in doubt of what is the best solution for this problem.

I am looking forward to hearing more about this from other forum members!

[karzan]

Hi Dears
In your discussion, one question was created in my mind,
check valve on outlet gas line from closed drain is needed or not?
I think that if we have blanket gas for closed drain so we have to
consider check valve, but in case of flare balanced closed drain I am not
sure about it.

[fallah]

Hi Dears
In your discussion, one question was created in my mind,
check valve on outlet gas line from closed drain is needed or not?
I think that if we have blanket gas for closed drain so we have to
consider check valve, but in case of flare balanced closed drain I am not
sure about it.

Flare balanced closed drain,needs to be equipped with a check valve at the inlet line to the relevant drum,as well.