9 replies to this topic

[Chem01]

For Ammonia storage, new double wall tank installation is on its way. Tank has ~ 5000 MeT capacity, inside tank and annulus do not communicate with each other. Design pressure of inner tank is higher and that of outer tank.
Liquid NH3 from inner tank can be transferred to other users, this liq. line has a shutoff valve which is installed inside the annulus. The reason of being inside annulus is if the upstream flangs of the valve has some leakage it should be contained within the tank and not outside. But this result in another problem of maintenance of the valve for which special safety precautions will be required in future when entering inside the annulus space for maintenance.
I've not seen such installation at other plants, Is there any altrernate to avoid installation of this valve?
Regards,

[Art Montemayor]

Chem01:

We need more details and a drawing of the tank and its dimensions - especially the annulus space. Otherwise, there is little anyone on the Forum can do except ask the same questions and/or speculate. I believe everyone knows that a double-wall tank is sometimes used in low temperature service in order to invoke a vacuum annulus or an annulus filled with insulation and inerted with nitrogen - very similar to a cryogenic cold box. Which one is yours?

If you have already shipped the tank to site, then you (or your company) have approved the final fabrication drawings and all other related specifications. That means you approved of the flanged(?) block being located in the tight annular space. It is probably too late to change the mechanical design. I don't know why you would approve something first and then question it later. Please explain this and also reveal your possible alternatives and/or problems with the fabricated design.

Await your detailed reply.

[Chem01]

@Art, thanks for your reply.
I'll be in office tomorrow and ll provide dwngs and the data.
Interestingly Tank designer has also provided a reference from a book to support installation of valves inside annulus, i'll also share that.

Note that the annulus is to contain any leakage from the inner tank, there will not be any material insulation inside annlus. It will remain under N2 pressure.

The tank is under construction, its engineering study was completed by offsite and utility deptt. and
now the package has been handed over to Ammonia engineers. All drawings are already approved but
we can change design to some extent with a change order.

Regards,

[astro]

Immediate thoughts without more detail for alternatives to installation of a block valve in the annular space are:

1. Install a block valve external to the tank;
2. Top entry for liquid outflow using submersible pumps (as typically seen in LNG service).

I've regularly seen application of option 1 above for ammonia service. Option 2 (which is not an option that I've seen applied in practice) has capital cost and operating implications that result, presumably, in the application of option 1 that accompanies the acceptance and management of the risk associated with the penetration of the outer tank.

It's a matter for assessing the Owner's appetite for risk while considering the relevant regulatory obligations, sound engineering practice and how the risk issues balance, viz.

a ) the risk of failure to operate the annular space block valve when required;
b ) the risk of a loss of containment incident upstream of the first external block valve.

[Chem01]

Art,
I've attached few files for further clarification.
I believe there is no standard that pushes for installation of valve inside annulus, i guess designer is forcing so that he can make us a reference for other installations.
Tank height = 19.8 m, frm bottom to top
Height of cylinderical portion = 15.5 m
Dia of inner tank = 26.9 m
Width of annulus = 0.8 m

Astro,
Option-2 seems feasible but i've not seen this for Amm storage tank.

Thats right, installing a valve outside like other plants poses risk when leakage at valve upstream line / flange occurs wheras installation inside annulus gives rise to risk factor as maintenance has to enter inside annulus for valve routine checks etc.

Attached Files

•  Amm_storage_tank_PID.pdf   422.21KB   170 downloads
•  Guide_to_Storage_Tanks.pdf   193.51KB   148 downloads
•  Amm_storage_tank_PID.pdf   422.21KB   131 downloads
•  Amm_tank_elevation.pdf   434.79KB   103 downloads

[astro]

The guide that you uploaded that deals with the block valves in the annular space refers to a UK site where a full containment design was not possible due to "the lack of suitable in-tank pumps" [1st paragraph 21.3.2].

I would suspect that the drive for such a high integrity design may have been due to concerns for societal risk, i.e. fatalities outside the site boundary. What are your project's risk drivers? Maybe you can relax higher containment integrity based on your local circumstances (e.g. remote location from the nearest neighbouring community)?

From a consequence analysis, what improvement does the additional hardware offer and is it worth the costs (installation and maintenance) over the life of the tank?

The guide also gives the impression that stress corrosion cracking resulting in the need to physically enter the tank is almost a foregone conclusion [last paragraph 21.3.2]. Counter to that view is that there is a valid approach that avoiding entry into an ammonia tank is a positive for its integrity because a thermal cycle plus the ingress of oxygen (and the need to remove it) is avoided. There is a personnel safety benefit, as well, of designing for minimal confined space entry.

The care taken to commission the tank and purge O2 and then N2 from the tank is a major determinant of the tank's long term integrity. That's a step in bringing the plant online that you don't want to rush.

This paper discusses a leak before break approach in a 30,000t double walled ammonia tank:
International Journal of Pressure Vessels and Piping
Volume 77, Issue 13, November 2000, Pages 783-789
Advanced safety features of an ammonia tank
P. A. McGowan
Offering a different approach to addressing a failure.

I've been involved with designing an installation for acoustic emission testing. Here are some references off the web:
Acoustic Emission brochure
AE Inspection
AE for NH3 Tanks
EFMA Ammonia Tank Storage Inspection
EFMA Ammonia Tank Inspection Guide

In the case of my experience, the design and construct contractor's order was changed so that during the tank hydrotesting the AE equipment was in place. Not only did this set a data base line for integrity assessment, the AE testing picked up a minor fabrication defect that was rectified before putting the tank into service. When avoidance of stress raisers is so critical, AE has a potentially highly valuable role to play. In the case of this tank, there were no annular space valves and the liquid outlet penetrated both walls at the base of the tank.

Given the NDT tools at the Owner's disposal, they were delivered a means of monitoring tank integrity throughout its service life confident that the risk of a problem would be minimal.

[Chem01]

Astro,
Thanks for your detailed reply.
Tank designer has clarified to us the design basis is full containemnt.
This tank is very close to process plant, within 100 meters of plant area. In other plants i've seen storage area is located at some greater distance from critical plant areas.
There is no dyke around the tank, which also requires full containemnt.
Outside the plant boundry there is a village about 400 m away, keeping in view above all full containment seems a better option.

This is our input that i've given, i know maintenance ppl are against it and they'll counter part. Anyways I can foresee our decision of instalaltion within annulus wil be accepted.

Regarding accoustic emission testing i'll try to pursue mangement to go for it. In my previous experience a double wall NH3 tank was de-comissioned once in ~ 17 years. For AE do we need to purchase the system right now? or we can wait for few years.
Thanks

[Art Montemayor]

Chemo1:

I now have a clearer picture of what your scope and basic data are. Thank you for the information and follow up. Your input is vital for our members’ comments and their values added.

After looking and studying the documents and reading your scope, I have several comments:

1) I think it is commendable to scope out a 100% secondary containment system on atmospheric liquid ammonia storage – particularly with a close proximity to a village. This concept is going to be expensive and I concur with your concern to make all the possible study and investigations NOW prior to commissioning the final product in order to realize a complete, working successful project. I believe it is of vital importance for the project engineers to ensure that the thought, effort, and expense being made in the interest of public safety are completely justifiable and successful. It is not often that capitalistic efforts are concentrated on public well-being; this is an opportunity to make that effort bear fruit.

2) However, I have reservations about the practicality of the details being incorporated in the detailed design. I find scope contradiction in justifying a 100% secondary containment system in the interest of public safety, while designing for a questionable, hazardous, confined space entry into the cold (-33 oC), narrow (0.8 m), and inerted (nitrogen) annulus. I can foresee a lot of hazards and dangers in maintaining a set of block valves (one of them pneumatically actuated) within such close quarters. My extensive plant field experience prevents me from accepting what is described in the magazine article and the P&IDs furnished. I have some deep and serious concerns and, hopefully, these can be resolved with further details which are missing from the information furnished.

3) The first serious concern is the maintenance of the operability of the single source ability to drain the tank of liquid: the block valves in the annulus. The indication is that instrument air will be used to activate the main block valve. I challenge the use of instrument air, unless it can be shown that the moisture in the IA will be so low that it will not pose an operating problem due to water ice build-up in the -33 oC annulus. Unless you can use cryogenic-derived nitrogen gas, I doubt you can guarantee that no water ice will plug the IA lines in the annulus.

4) I cannot see any alternative to 100% welded valves in the annulus. That calls for 100% welding efficiency and total expansion/contraction stresses being relieved by off-sets in the piping. I would never install an expansion joint in the annulus. This means that the walls of both the inner and outer tanks will have to be mechanically reinforced and insured against excessive expansion stresses – as well as the piping. How this can be done in a horizontal, 0.8 meter space is an engineering challenge. You want to avoid vertical loops because they may be detrimental to NPSH and external pumping.

5) I see a latent danger in sending personnel into an extremely confined space (0.8 meters) that has been inerted and is kept at -33 oC. This could be a requirement if the valves in the annulus fail or need attention. Here, the valve type, quality, reliance, and mechanical specifications are very, very, important. No mention is made of how the manual valves are going to be actuated. Obviously, their handles will have to be actuated through the outer tank walls – but the mechanical mechanism is not mentioned nor detailed. This may be a very important detail and a possible weakness in the design. I see no manner of possibly contemplating the possibility of sending in a human into the annulus while the inner tank is under normal operation. I personally would not issue that order or command because I find the concept to be too perilous. That means that the pump-out and drain valves are the under-pinnings of the total design concept. If they don’t work 100% efficiently, 100% of the time, the operation might be a potential failure. This cannot be left up to chance.

6) I admire the concept of 100% secondary containment. I also believe, as Astro implies, that an external set of block valves are much more mechanically practical and can be designed to complement the containment concept. Jacketed piping and valves (cryogenic valves come to mind), cold boxes, are just some practical and proven concepts already in industrial use that might be applied. I believe the issue of tank integrity is different from that of piping and valve integrity, and should be handled differently. Certainly, the application of Acoustic Emission is a necessity to ensure the mechanical integrity of such an expensive and important application.

7) I don't understand the P&ID note of: "Design Press: 10.0061/0.153 (kg/cm^2 g). I zoomed the drawing up and still see what appears as "10.0061" - a value I don't believe is correct.

I hope these comments are of some help. I look forward to having other members share in contributing their thoughts and opinions on this important subject.

[Chem01]

@Art,
Thanx for a detailed reply, you've shown some dark areas where i was not looking,,,like instt. air at this low temp. I can reply to you point wise:
3) Our instt. air dew point is -40 C, but as it'll be at 4 kg/cm2 for supply to the valve etc. water appears to be aproblem. I've emailed vendor to clarify this point.
4) The block valves and control valve are flanged, whereas the pipe is welded at both ends to the inner tank and the outer tank. To clarify about the stress i've asked vendor to throw some light.

5) The manual valves (isolations of the shut off valve) will be kept open all the time. Their closure is required for handover of shutoff valve and in that scenario, manpower entry is required inside annulus. I echo with you, such an entry in annulus requires extra safety and precautions, special suit and breathing appratus will be required. That i've to dig out if this scheme is finalized.

6) I had in my mind to keep all these valves outside the tank but welded (not flanged). This rules out flange leakage possibility but leakage from bonnet or valve body can still occur.

7) P&ID has mistake, the design pressure is little above atmospheric, i've to confirm from vendor.

I'll update after getting vendor reply,

Regards,

[Art Montemayor]

Chem01:

I'll add my final comments on this topic because it is too serious an application to treat lightly.

I can never justify - nor allow - installing flanged valves in the annulus your application calls for. I seriously caution against applying this type of installation because the valves, their flanges, and the piping will all be under stress while in operation. This cannot be avoided due to the cramped and tight quarters and the short piping runs. All this means that a flange gasket is prone to leak under such conditions.

I can never accept the concept of sending in a human into such an environment. I consider it too dangerous and ill-advised from a design point. I make my engineering judgment based on the basic engineering rule and commandment that I took on when I became an engineer: I will never order or command another individual to carry out an action or go into a situation that I would not do myself, personally. I would never enter that annulus while there was ammonia inventory in the tank; therefore, I can not justify sending another human into that same environment. I will never back away from this professional stance because it has served me well all of my career. I strongly urge you to install all valves and controls externally and leave the annulus area empty. I ask this in behalf of those workers that could be forced to enter such an environment due to need or ignorance.