Double wall arrangement is foreseen in API 620 and 625 for low pressure storage tanks for refrigerated and liquefied products.
If the question is for API 650 ( i assume ), API 650 suggests the use of dikes etc. for containment.

You may look EN 1993-4-2: 2007 ..Double wall arrangement is foreseen if the primary tank contains toxic or dangerous fluids and may look EN 14015 for containment requirements.

I understand that when there is a space constraint for having a conventional bund secondary containment, a double shell is seen used. My question is whether this is qualified by API 650.

Because that's not a double wall tank, it's simply an extreme form of a tank bund.

so long as the bund wall can take the entire contents of the tank without overflowing, you can make it from whatever you want.

What you refer to is very odd and therefore not addressed in a design code intended for more normal installations.

Usually you find other factors such as fire precautions and separation distances mean you don't need such a tank or that it wouldn't be permitted.

I've never seen anything like that in API 650.

I would avoid a common bottom plate with the tank to separate tank from bund.

Also the zone in between becomes a zone 0 or Zone 1 and becomes horrendous to get any access into it, never mind filling with rubbish, sand, animals etc etc.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

Quote (svi


..I understand that when there is a space constraint for having a conventional bund secondary containment, a double shell is seen used. My question is whether this is qualified by API 650.)

API 650 does not specify double shell . If there is a space problem for containment, ( if the liquid content is petroleum, petroleum product , etc ), you should consider dike wall ( may be RC wall with height suitable for the space constraints..)

I have seen RC dike wall with 7.0 m but not double shell for API 650 tanks..IMO, it is not matter to design the double shell but the increased safety and fire risks could be the main concern..
I will suggest you, consider posting this thread at Storage Tank Engineering forum to get better responds..

Double shelled tanks are most suitable for diameters up to about 10 - 12 meters on a continuous concrete foundation

I certainly hope that we are not having this insane third-world conversation "because the client said so" ....

I also certainly hope that this is not the type of client who, when he learns of the massive double shell tank Capital costs, will simply say "Never mind ... Do it the old way"

It is important to remember that a double walled tank this large increases various types of Capital costs dramatically. This due to much higher costs for foundations, paint and coatings as well as leak detection systems.

If this huge tank is subject to foundation settlement (as many large tanks are)the double shelled tank design may put the inner shell at risk.

MJCronin
Sr. Process Engineer

API 650 does not include guidance on any type of containment as that is out of the scope of the standard. I find double wall tanks to be fairly common, particularly where space is a major constraint. The outer wall typically is a form of containment, like a dike or bund wall just taller. I also run into double wall tanks that have product in both spaces, and once made two floating roofs for such a tank configuration. API 650 does not care why you build a cylindrical tank but it gives guidance for the designer and operator of such tanks. As mentioned above and in the linked question there are many issues to consider when using a tank inside another serving as a steel dike including piping flexibility, differential settlement between the two tanks, safe access, drainage, monitoring, etc. I've worked on tanks between 10 feet to 150 feet in diameter with the steel containment tank usually but not always 10 feet larger in diameter. In all cases that I know of the the steel containment tank was designed using API 650 since it serves the same purpose and has the same loads.

As ever IFRs gives the best answer backed up by years of experience. Take note.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

https://res.cloudinary.com/engineering-com/image/upload/v1633569575/tips/Google_Earth_v9oeam.pdf

These are outside Philadelphia, PA.

The inner tanks are 150 foot diameter. The outer tanks have large windgirders on the inside of the tank shell.

Thank you all for the responses.

IFRs: Good picture and post. Curious to know about the floating roof for the annulus.

This is the same that I wanted to say that these configurations are in use around the world. Is it not time for API 650 to acknowledge to include. Agree that secondary containment is not scope but here the containment is another tank. There should be clarity in the STD to say that such configurations are acceptable with the corresponding provisions and considerations.

LittleInch: Incidentally there were 3 tanks at an Island tank farm that had this configuration. The tanks are now due for refurbishment. This calls for a major re-investment. Many of the tanks are to be re-built. The current effort is to re-locate some of them so as to create the safety intertank spacing required by the current NFPA. Possibly to meet the insurance requirement. There were shortfalls and not meeting the safety spacing requirement. At the outset, It was the intent to do away with the double-walled ones. However in the process of layout, based on the number of tanks / operational storage needs and the limited plot space availability on the island, it has again driven the selection to the double-walled design.
Yes, the space between the shells is a confined space that needs inspection access. It cannot be filled up with anything. The nozzles from the inner wall penetrate the outer wall with a bellow sealing. This includes the foam and fire water

MjCronin: My common sense says not to. I intend to bring some sanity into the selection. Nevertheless, it needs something as the tank STD to negate it. I do not know how the original double-walled tanks (30 year+) came into being and what were the design considerations. The operational driver is the experience with the existing double-walled tanks over this time. Double-walled tanks do not seem to be uncommon from Many a tank design person says that they go by the premise as if they are designing 2 tanks and therefore justified by the requirements in the API 650. I am driven to think this isn't so and want to know why there has never been a request for an interpretation. I did raise one now. But would not know when or what response will that be.

HTURKAK: I took the cue from an earlier thread in this group. besides cross-posting is to be avoided.

Quote (svi

...Incidentally there were 3 tanks at an Island tank farm that had this configuration. The tanks are now due for refurbishment. This calls for a major re-investment. Many of the tanks are to be re-built. The current effort is to re-locate some of them so as to create the safety intertank spacing required by the current NFPA. Possibly to meet the insurance requirement. There were shortfalls and not meeting the safety spacing requirement. At the outset, It was the intent to do away with the double-walled ones. However in the process of layout, based on the number of tanks / operational storage needs and the limited plot space availability on the island, it has again driven the selection to the double-walled design.)

Dear svi,

The subject tanks having dia 80 m and the ht should be in the range of 18.0 m . The roof should be EFR ..The picture posted by MR. IFR has diameter 45.0 m with fixed roofs and the radial space between exterior dike wall should be in the range of 12.0 m.. and apparently these are not double wall tanks..I have the opinion that API 650 does not address double wall tanks ..

If the available plot space is limited , you may consider to increase the height ( of course if soil conditions allows ). In past, i have designed 22m ( h) tanks and saw 25.0 m ( h ) at northern Europe. This option will allow more space between the tanks and more space..

I will suggest you ;
- Consider the option of increasing the ht.
- Consider dike walls ( not double wall )
- In case of double wall , provide fixed roof for the annular space..

These are my thoughts based on 40+ years experience.. and remember, this is WORLD WILD WEB ..

P.S. Have ever seen an when oil storage tank in fire ?

Good Luck ..


-

Perhaps we should stop referring to the original poster's situation as a double wall tank. It is really just a tank inside a tank. Same for the picture I posted. The outer tank is mostly empty during its life but when it is used it is just a tank. Since both are tanks serving the same purpose of holding liquid they can both be designed using API 650. The design of fire protection, piping and piping flexibility, leak prevention and detection, common or separate bottom plate, foundation settlement, confined space, open or covered annulus, etc are not in the scope of API 650.

svi - the outer floating roof was similar in construction to the nromal inner one, but it had seals on both inner and outer edges. Not hard to do or build.

The outer tank is a dike wall, I don't think NFPA limits the spacing between dike walls but I don't remember what they say about this situation regarding tank to tank spacing if the dike is tall enought to contain the proper volume.

@IFRs
I've never worked in tanks so bare with me here. For my own curiosity

I looked at the linked photo and got a chilling image of an operator being in-between the tanks when a failure occurs. Specifically someone attempting mitigation or the like during the initial phases of failure. IE; tank's leaking! Lets drain it! Then a burst.

Is there a method of retreat for the operator in this case? Or is there an SOP not to go between the tanks? Or is the simplest answer that this a non-issue that an early coffee fed morning has conjured up in my head.

If the inner tank is leaking, no one in their right mind would go into that space. Standard operating procedures should prohibit such a dangerous act. Same as if the dike was a larger earthen wall. In this case the outer tank would be serving its function and contain the leak while the inner tank was emptied.

IFRs: I coined the word double wall, since there wasn’t any. Double wall, if the inner one leaks the second wall becomes the tank shell.

The annular floating roof, that must be for liquid filling on both sides. Will that be double wall?
will you have a shot. This one must be from a drone?

IFRs: “I don't think NFPA limits the spacing between dike walls but I don't remember what they say about this situation regarding tank to tank spacing”

NFPA provides the minimum inter-tank spacing of the regular tank. Wall to to wall spacing so as to minimize radiation effect to an adjacent tank. With the additional wall that acts as a shield. If both tanks are using an outer wall for secondary containment, The inter-tank distance is between the inner wall of one tank to the outer wall of the adjacent tank. The adjacent outer wall being without content may need consideration. Is this not a consideration in the case of tank in the picture?

When the annulus space is used for storage. Or during testing of the outer shell. The external pressure loading onto the outside of the inner wall needs consideration.

These and other considerations need to be addressed by API, NFPA entities.