Rectangular Storage Tank - Thermal Movements ?

[rohit307]

I am working on designing a Rectangular Water Storage Tank. (AWWA D-100)

I need to consider Stresses and Load combinations that include Thermal Movement due to Temperature change - *Correction - Design requirement Ambient DeltaT (Effects for Solar heat gain to Night time - sky heat Loss) For ambient temperature change from 105F to -10 F.

I am thinking i need to consider the actual effects of Heating and cooling (Conduction, Convection and Radiation Heat gain and loss) to figure out the Delta T needed for Design/ Self-Straining Force check on the Tank (steel shell and bracing)? Thoughts?

I am curious if there are procedures or guidelines for it in any code or Practice for engineering a rectangular tank. I would appreciate any inputs. If Ansys/FEA approach is the best then do share articles or information on how to approach it in the software.

Tank is 55'x9.5' and 13' height closed Rectangular Tank has a good bit of internal bracing for the standard Hydrostatic and (Wind / EQ combinations) . I am trying to figure out how to go about the thermal ?

 

[JStephen]

For a typical cylindrical tank operating at ambient temperature, there is zero consideration given to these effects. Which is one reason you won't find much in AWWA D10 addressing the issue.
For tanks constructed of two different metals (steel tank with aluminum roof, for example), thermal effects are considered, typically by providing adequate flexibility at the interface.
The current API-650 allows tanks to be built partially from carbon steel and partially from stainless steels, and the differences in expansion need to be taken into consideration in that case.
API-650 covers tanks with design temperatures up to 500 degrees F, and for elevated temperature tanks, some effects need to be taken into consideration. One issue is allowing for lateral movement between anchor bolts and anchor bolt chairs and bottom plate. One issue is metal fatigue due to thermal cycling (which is not a consideration under 200 degrees design temperature).
For this tank, I would think the major consideration would be allowing for minor differential movement between the tank and any anchor bolts or attached piping.
For tanks subject to ice formation, there are some additional considerations. Elimination of any internal ladders, braces, or other items that ice can hang up on is one. Typical municipal tanks are prevented from freezing by water circulation. For tanks that are completely stagnant (fire protection water, for example), NFPA 22 has limited information on insulation and heating.

 

[rohit307]

@JStephen Thank you for the response. This tank has a mixer to keep water circulating during low temperatures so don't see Icing being an issue.

What I am bewildered is that zero consideration to these effects is given ? I am coming from another industry so Tank Engineering is very new to me, I was assuming steel being steel the stresses from Alpha*E*Delta T show stresses that seem pretty high. Assuming a 55-ft long shell welded at corners and welded internally at braces, there are stresses developed all across for a Delta T of 50 .

Am I oversimplifying this perhaps but assuming Fully Fixed at both ends between bays or even edges of wall. We are looking at ( 6.5*10^(-6)* 29*10^6*50) = 9425 psi, That is the additional Axial stress you would want to account for on top of your other load combinations .

Am i missing something very conceptual and basic and looking at it erroneously ?

 

[HTURKAK]

Am i missing something very conceptual and basic and looking at it erroneously

I am sorrow to say this .. but ,you are missing something . You can assume the tank wall temperature would be similar with the water content . If the tank can expand freely , temperature stresses will not develop.
I do not have any idea for the roof and bottom plates. If the tank is supported on short legs, the bottom plate will also has the same temperautre with liuid content. The roof plate can experience ambient temperature .
Pls provide some descriptive sketches to get better responds.

 

[rohit307]

@HTURKAK

It is a Closed Rectangular Tank -

Top and Bottom plates are same thickness and material as the Shell -

The edges are all welded . The tank is resting on a less than 1% slope Skid. There are direct - Anchor bolts to the Foundation.

Tank is Storing Water at normal temperature. Assume all internal walls , top and bottom are in contact with liquid.

The tank is welded at all edges so expand freely is not a right representation, Tank has internal bracings to support the hydrostatic load and other loads which are also welded to the interior. Essentially creating vertical and horizontal panels.

I am looking at one panel that is welded at all 4 edges and internal is liquid temperature external is ambient hot air - Assume Delta T of 50F

 

[JStephen]

On the "fully fixed at both ends"- what is it fixed to? Normally a tank bottom would be resting on sand, concrete, or pad material, but not fixed to it, either.
The shell and bottom are attached to each other, but there is no mechanism to make them at substantially different temperatures.
If you are assuming the outside of the shell plate is 50 degrees hotter than the inside, I suspect that will give a much larger heat flux than would actually be generated.

 

[rohit307]

The edges are welded so the shell is welded to top and bottom , hence the full fixity. I was assuming the same that with a 0.5" thick shell the heat flux will be much higher and i wouldn't need to take as crazy a Delta T consideration. The outside temp is higher (or lower with the top and bottom max as listed) inside contents temp and the inner wall temp is causing the Delta T.

It is being installed on a grillage that has 1% slope. So bottom of the tank will have outside as ambient temp and inside as the contents temp.

The fixity was not to the foundation. If anything, we do have (8) chaired Anchor rods - Don't see much force as Wind/EQ load requirement is low.
@JStephen - I see you mentioned about the issue of allowing for movement for the AR's - Any reference on how to go about it ? Just alpha* E* Delta T figure out elongation and allow for that or the additional stresses it imposes on the tank wall and anchor rod itself ?
May be do a slotted hole for it ?

 

[JStephen]

The heat transfer to both inside and outside surfaces will be due primarily to convection. In figuring convection heat transfer, the driving factor is difference between fluid temperature out away from the surface, and the surface temperature. IE, you do not assume surface temperatures are equal to fluid temperatures here. So you maybe able to reduce the thermal difference considerably.

The implication in API-650 is that simply slotting bolt holes on the anchor chair and allowing adequate clearance between tank chime and bolt is all that is required for thermal design of anchors. See Section 5.12.11.