12 replies to this topic

[NATU]

Dears,

I need your help for understanding the Closed Drain System. Below is the matter

As per Engineering Standards Closed Drain System Drum shall be installed either at grade or below grade. However, in one of our unit the elevation of CDS Drum (Z99-D-0018) is 5000 mm above ground. So my question is,

is it ok to have elevation of 5000 mm for CDS Drum?

Because normal draining procedure is, Before being drained to CDS Drum, equipment is isolated & Depressurized.

If this is the procedure then we may not have sufficient pressure in vessels to be drained to CDS Drum which is at higher elevation.

If you install CDD at elevated position like 5 m then you need to raise elevation of upstream process equipment and lines as well which will be very costly.

Also in this case we may required to place CDS header on pipe rack. Please correct me if I am wrong.

If you have any similar experience in this subject, I should be grateful for your kind sharing.

Thanks,
Natu

Attached Files

•  CDS Drum 1.pdf   198.6KB   644 downloads

[ankur2061]

NATU,

As long as there is a net positive elevation difference between the drain header (6"-P-X010-1CC9P) and the inlet nozzle, there should not be any problem in draining to the CDS drum due to gravity. When I say net positive elevation difference it means that drain header is not a riser to the drain vessel from a elevation lower than the inlet nozzle or in other words the elevation difference between the drain header and the inlet nozzle connection is postive. You need to refer the isometric for the drain header to find out that it is continuously sloping down or run down towards the inlet nozzle of the CDS drum. Yes, I would definitely prefer to have a CDS drum below grade because it more or less precludes my checking of each and every drain sub-header and the main header to see that I have a net positive elevation difference available between the CDS drum inlet connection and the drain header(s). In fact, I would prefer to have a CDS drum not only below grade level but also buried for advantages of not having to provide a PSV sized for fire case for the CDS drum and safety related issues of toxic or flammable fluids accumulationg in a below-grade pit where the CDS drum might be located. But many times layout constraints do not permit what is the most desirable configuration and solution needs to be found with some improvisation keeping in mind that operational and maintenance safety is not compromised.

Engineering standards are guidelines to help decision-making but are not gospel truth and cannot substitute common sense or engineering judgement of any engineer where ingenuity is required.

Hope this helps.

Regards,
Ankur.

[TS1979]

NATU

Liquid can only drain from high energy point (pressure + elevation) to low energy point. As long as the energy (pressure + elevation) of the upstream of the closed drain vessel is larger than the energy (pressure + elevation) at the liquid surface of the drum. You are OK.

Normally, the vessel at raised elevation is to provide necessary NPSH for the downstream pump.

[Dacs]

Is that an existing plant? Because we're talking about 8m of static head!

Because if so, the only viable procedure for you to drain your vessels completely is using inert gas (such as N2) to pressurize the vessel again (after isolation and depressurization) to provide drainage, and this in itself poses some safety issues.

Of course, if all the vessel elevation is higher than your drain drum, then there won't be any issues at all, although I'm wondering how'd you drain your pump casing (assuming it goes to the same drain drum as well).

For the elevation of the drain drum itself, the only reason that I can think of why it's elevated is because of NPSH. Drain headers are usually sloped (going to the drum) without pockets and usually are placed in trenches (since most drain drums are below grade).

I can't help but notice that this is a Saudi Aramco plant. Strange because the projects I've involved with them require drain drums at below grade (there's even a standard for that, I forgot the exact name though).

[fallah]

NATU,

Without having the 3D model of the plant in hand it is hard to judge about the rightness of the made decision on installing the CDS drum in an elevation above ground. Of course in your case, due to using horizontal centrifugal pumps for drum emptying, seems you had to install that drum in elevated position regardless of if there are net positive elevation differences between incoming lines and the drum.
In buried CDS drum vertical submersible pumps to be used to transfer the liquid inventory.

Fallah

Edited by fallah, 16 October 2012 - 02:53 AM.

[NATU]

Dear All,

Thank you so much for response & showing me the way forward. This is the Closed Drain System of PENEX unit (Isomerization Unit). Please kindly note BTL Elevation of almost all the vessels is less than 5m, which is less than the elevation of Closed Drain Drum. Also all the pump casing drains are goes to the same drain drum. So as Mr. Dacs said, only viable procedure is “By using Nitrogen gas to pressurize the vessel again & drain it”.

Dear Ankur Sir,

As you told “Many times layout constraints do not permit what is the most desirable configuration”. As this is the new plant I will figure out if there is any such constraint for this unit.

But one more thing I could not understand is, Logic behind providing PSV protection for this Closed Drain Drum?

As this CDD is open communication (floating) with Flare KO Drum. There is no isolation (automatic) in between line of CDD & Flare KO Drum. So we can take the credit of this line for relieving during over pressurization of CDD during Fire or Possible Gas Blow By scenario. Only thing is, I require to size this line for the controlling case of CDD. So with this I can eliminate the requirement of PSV for CDD. Please let me know your view on this.

Dear Mr. Dacs Sir,

As you told only viable procedure is “By using Nitrogen gas to pressurize the vessel again & drain it”. But it seems to be expensive. Now I am figuring out the safety issues of this procedure,

Could you please shade some light on this matter? What are safety issues of draining using pressurized Nitrogen gas?


Let me know your ideas.

Many thanks for your time.

Regards,
Rajiv Natkar

[Dacs]

One obvious safety issue that I see is when your N2 pressure is higher than the vessel design pressure.

To be honest, I don't think this is widely used because the drain system is usually designed with self draining (gravity flow) in mind.

Truth be told, you may not be needing this because maybe all of your vessels are above 8m.

[ankur2061]

Dear Ankur Sir,

As you told “Many times layout constraints do not permit what is the most desirable configuration”. As this is the new plant I will figure out if there is any such constraint for this unit.

But one more thing I could not understand is, Logic behind providing PSV protection for this Closed Drain Drum?

As this CDD is open communication (floating) with Flare KO Drum. There is no isolation (automatic) in between line of CDD & Flare KO Drum. So we can take the credit of this line for relieving during over pressurization of CDD during Fire or Possible Gas Blow By scenario. Only thing is, I require to size this line for the controlling case of CDD. So with this I can eliminate the requirement of PSV for CDD. Please let me know your view on this.

Rajiv,

Your P&ID shows the vent connected from the CDS drum to the flare header and also shows PSVs connected to the flare header which I had not noticed earlier.

In case if you have an open circuit to the flare header from the CDS drum you may avoid providing PSVs for the CDS drum.

The advantage of a buried drum is that the fire case is not applicable at all and thus a PSV for fire case doesn't apply. PSV may still be applicable for a gas blow-by case after due analysis of the gas blow-by scenario.

The argument that has been given for not providing a PSV to a buried CDS drum is that since it is buried no fire-case is applicable and even if a gas blow-by case is applicable and the vent from the CDS drum is a open circuit to the flare system, a PSV is not required. Most advocates of a PSV on a open circuit closed drain drum have only one argument that since the vessel is being designed as per ASME Section VIII, Div. 1 it requires a PSV as per code. This also is not true, since the latest version of ASME Section VIII, Div. 1 allows exceptions for not providing PSVs on vessels designed as per the code.

One interesting case that I had argued for not providing a HP steam separator designed as per ASME Section VIII, Div. 1 and receiving wet steam from a HRSG was that the separator was fully rated which means that the design pressure of the upstream equipment and the separator was the same and the upstream equipment was protected by dual PSVs.

Looking at your P&ID where the CDS drum vent is connected to the flare with a LO valve (open circuit) and is probably floating with the flare, you probably don't even need a PSV if your vent is properly sized for the maximum vapor flow resulting either out of an external fire or from a gas-blow by.

Another example of a vessel designed as per ASME Section VIII. Div. 1 but not provided with a PSV is the Flare KOD which is connected to the flare with an open circuit and floats with the flare system

Hope this helps.

Regards,
Ankur.

[fallah]

Ankur,

In fact, a CDS drum operates same as a KOD with a free connection to flare network, so as you mentioned a CDS drum doesn't need to PSV against overpressure. Of course, the connecting line should be sized such that can cover all overpressuring scenarios (gas blowby, sudden vaporization,...) without passing MAWP of the drum.

Fallah

[RoyenG]

Hi NATU,

if possible, why don't you trace the strory/philosophy behind choosing this high level of KOD instead of satisfying pump NPSH only?
and see the sources that connected to the Closed drain header, may be connections only from PSVs and BDVs, while liquids will be connected to the other drum.

and agree with fallah, PSV does not required in this CDS.

RoyenG

[RoyenG]

Hi

I do not know your procedure to purge equipments in the project,
But from note 1 of your P&ID says that valve to the flare header will be closed during T&I/steam purging.
This will be the source of vessel to become overpressurized, and PSVs installed.

Royen

[NATU]

Dear Royen,

For tracing this I have already asked client to provide draining philosophy. I am waiting for that document.

Dear All,

Thanks a lot for sharing your valuable inputs. Now I have better understanding of this system. I had taken this point for discussion with client. I will let you know the final conclusion.

Thanks again.

-Rajiv Natkar

[Bhupat]

Dear all,

I know its too late to reply on the topic, but still let me share something might be of use.

 

Two type of closed drain system are as followed:

1. Above ground

2. Below ground

 

Ideal design for closed drain system is below ground and which is widely accepted in most of the refinery world-wide. But again some client prefer to have system above ground based on their operational experience.

 

Above ground closed drain system are possible in process unit, where the equipment are located at elevation and mostly pressurized system. So, pressurized gas e.g. N2 is used to drain the associated system e.g. pumps/vessel etc. This is not the case in tank farm, where you have mostly low pressure system API 650 / API 620 tank. Neither N2 can be used to pressurized the system nor equipment elevation raise is affordable solution and therefore no option left to choose the gravity system.

 

Pressurized system has many disadvantage compare to gravity system.The main disadvantages are

- Large N2 Consumption

- Liquid may trap, if not purged properly.

 

Below ground drain drum has disadvantage of confined space entry in case drum is located in vault design and buried drum has difficulty in maintenance. Therefore some time of least preference of some client.

 

I have difference in opinion on "To satisfy Pump NPSHa" Increase in elevation of drum is not right choice. One can not buy this argument.