30' dia x 19' tall water tank with roof 2

[ARS97]

The company I work for (design/build construction company) will be bidding the design, fabrication & construction of a 30' diameter x 19' tall water tank (100,000 gallons) for a coal plant. Most likely, since I specialize in structural steel & reinforced concrete design, I will be recommending that we seek a third party designer who specializes in water tank design. However, management is asking that I at least consider handling the engineering internally. I'm not opposed to it, as long as I have a good grasp on some reference material.

The Gaylord Structural Handbook has some references for tank design, and it's essentially based on hoop tension. This is a pretty simple concept, so designing the walls for the hoop tension seems pretty basic. However, there are two other issues that I'm concerned about:

1. This tank has a roof (with a post in the middle I assume), which will impose a compressive load onto the tank walls, so buckling of the wall is a concern. I'm pretty sure that Troitsky's Tubular Steel Structures has some information on that, but I'll look.

2. I would think that seismic loads would be a HUGE consideration for something like this. 100,000 gallons weighs about 833,000 lbs, so even if we're only dealing with a 5% seismic load, that's about 42,000 lbs! How do you account for such a large lateral force within a thin-shelled circular tank?

I'm sure there are various design codes for water tanks of this nature, but it's not really in my wheelhouse and I don't know which organizations handle this topic.

I will still be recommending that we seek a third party designer, but I'd like to at least be able to talk intelligently about the topic. Any help?

 

[JStephen]

A tank like that would commonly be built under either AWWA D100 (municipal water tank standard) or API-650 (oil tank standard also commonly used for generic industrial storage.) Both standards have sections detailing seismic loading, both standards include consideration of buckling due to seismic. While the seismic loading can be large, relative to, say, a building, the available weight resisting overturn is also large, and seismic loading is likely not a huge issue. The tank would likely be anchored anyway for wind. Generally, with a tank of that size, seismic loading would not affect shell design at all unless in very high seismic area.

A tank that diameter could have a rafter-supported roof with center column or a self-supported cone roof. If a flatter roof is specifically required for access or other reasons, go with the rafters, otherwise, make it self-supporting to eliminate some corrosion issues.

Usually, tank owners will issue specifications themselves or through a consulting engineer, and final tank design will be by the tank company. In some cases, the foundation will be by the owner/consultant, in some cases by the tank company.

 

[ARS97]

JStephen - great info......I'll check those out. Thanks!

 

[Gumpmaster]

JStephen makes good points.

I would definitely specify this out. You'll want a steel tank manufacturer fabricating it. Many of them have their own in-house engineers as well. They'll design something that's the most economical for them to fabricate and it will take them a fraction of the time to design it. It would be a big chunk of work to properly design a steel tank without having ever even seen an example calc before.

You'll have to choose between welded & bolted construction, as well as a good coating system. Periodic maintenance of steel tanks is a big part of their lifecycle cost.

 

[ARS97]

I agree - a third party tank manufacturer is the smart route here. I've been telling them over & over again - yeah, I'm sure I can eventually figure it out, but it's going to be an enormous engineering cost for me to fumble my way through it. Plus, I'm not sure our fab shop or field crews are suited for the QC requirements for such an item. (They tend to have their own "specifications", if you know what I mean.) I'd much rather trust someone who designs these tanks frequently. HOWEVER, we're in a very tough time constraint, and unfortunately, we don't have time to request a third-party bid.

What's going to happen is they're going to throw an conservative price at this tank, and if by some miracle we get it, then we'll hopefully seek a third party designer/fabricator. We just need to make sure we're high enough on the bid. So, I need to have a rough idea on the tank dimensions for bidding purposes only.

We've settled on AWWA D100-11. For a 34' diameter x 15' high tank, I think we can get away with 3/16" shell plate (minimum shell thickness for this size tank, per AWWA). As a conservative estimate, they're probably going with 1/4" shell plate and bottom plate in the quote. The shell plate and base will have full penetration butt weld seams. The stuff that we're going to struggle with are the items such as - two manholes on the bottom shell plate, access hatch on top, external ladder, overflow pipe, vent, roof framing (with center post), etc.

The overwhelming cost is going to be the field labor due to the amount of welding. If anyone has any good references on typical roof/wall details or center post details, by all means, point me in the right direction. Thanks........

 

[racookpe1978]

So, make a few off-the-record phone calls.

The problem with a round tank + rolled (dish) bottom + round walls is the complex multi-curved fitting and setup by the fab shop: Like a ship's hull: All the welded joints are easy to draw. Very, very hard to roll up and get accurately fitted and welded.
Reinforcements under (or around) the intersections between legs and the lower floor to prevent buckling or erosion or collapse) of the tank are equally difficult.

 

[JStephen]

API-650 has typical manway details.
The bottom would normally be lapwelded from the top side only.

 

[ARS97]

racookpe1978 - no, the bottom is flat......no curvature to it. Just a typical cylindrical "reservoir".

JStephen - I seen where it's discussed about the bottom construction....either butt weld with a backer, or lap welded. With the lap weld, how would you handle the uneven surface where the wall meets the bottom? You'd almost have to have the walls sit on a common ring-plate, then fill in the rest of the bottom with lapped plates?

 

[JStephen]

That's one of those things that looks like a problem on paper, but just really isn't in the field.
If you butt-weld with a backup bar, you have exactly the same issue.
Normally, there would be 1/2" padding or grout under the tank shell. But if neither is there, it still isn't a problem.

 

[racookpe1978]

Stephen - I seen where it's discussed about the bottom construction....either butt weld with a backer, or lap welded. With the lap weld, how would you handle the uneven surface where the wall meets the bottom? You'd almost have to have the walls sit on a common ring-plate, then fill in the rest of the bottom with lapped plates?

OK, so a flat bottom makes things simpler.

Try this then to do all three: reinforce the bottom joint, make the bottom joint (wall-to-bottom) easier to weld, porvide a backing plate where the columns intersect the tank. Roll a 4x4x3/8 (or similar) angle to the diameter of the tank.
Fillet weld (inside and outside) tank walls to the angle iron, and tank bottom plates to the angle iron. NDE (dye penetrant) inside and outside welds to verify no pinhole leaks.

Add reinforcement plates under the vertical support tubes, with ribs at each junction. Rolling even a large angle iron is simple, and provides a backing plate and doubler plate that avoids the more tricky, more expensive butt joint. A double fillet takes more welding inches, but is a more forgiving weld.

 

[JStephen]

Normal tank procedures for the last 60 years or so is to just set the shell plate on the bottom plate and weld it with a fillet weld inside and out. That construction is required by API-650 and AWWA D100.

The rolled angle may have been used with riveted tanks, I'm not sure how they made that connection.

 

[ARS97]

Welding it with a double fillet makes sense, plus it's mentioned in AWWA D100-11, but how would that work with a lapped surface? What I was saying was to have a circular ring under the walls that allows for a level surface all-around. Then, the interior bottom plates can be lapped over that ring.

 

[Leftwow]

For engineering, the PIP STE code has nice design examples for a ringwall tank.

 

[JStephen]

See the plate break-down detail in API-650.