Should Water Load of Water supply tank taken in Mass source for seismic analysis

[Ahmad Uzair Arshad]

Dear Engineers,

I am designing an elevated water supply tank (90 feet above ground) of 200,000 gallons, Should I take weight of the water in the seismic weight of the structure for calculation of base shear.
Most Codes define fluid loads separately, meaning they are neither placed in live nor in dead loads.

Now should I take the water load in seismic weight of the building?

Thank you.

 

[bones206]

Yes, definitely. In the US we would follow ASCE 7 Chapter 15.7 as well as the applicable AWWA design standard for characterization of seismic design loads.

Here is an excerpt:

 

[HTURKAK]

Some portion of the water content ( impulsive mass ) will act together with supporting structure ..The convective mass portion will act different..You would calculate the impulsive force and convective force separately and combine them.

I will suggest you to look ;
- ACI-350.3 Seismic Design of Liquid-Containing tanks
- AWWA D 103

And the following doc. which you may download for free from

http://www.sefindia.org

 

[JStephen]

For a typical elevated tank (steel leg-type or single-pedestal water tower as commonly used in the US), the tank contents are all considered in the "impulsive" category. If you have a building-supported tank or some other odd structure, I suppose that might vary. Part of the logic there may be the difficulty in determining sloshing characteristics for of odd-shaped tanks). For ground-supported tanks, in most cases, consideration of the convective motion will lead to a reduction in the total forces rather than an increase.
And for those same type of elevated tanks, typically 90% of the seismic forces are due to the water mass, rather than the tank itself, so it is essential to include the water weight in the seismic loading.

 

[Ahmad Uzair Arshad]

Thank you for your response,
I have a question that may sound weird, In my opinion, water will act as a damper because of inertia, which will reduce the effect of forces on the supporting structure.
However tank walls will endure more force due to impulsive load, I have designed the walls for dynamic pressure calculated using "ACI-350.3 Seismic Design of Liquid-Containing tanks".
Now there are 3 scenarios that I have modeled in my structure.
1. Dynamic water pressures calculated using ACI-350 acting along with the seismic case, that is dynamic force in X-direction is applied in combination with seismic force in the x-direction.
2. Dynamic water pressure is applied in the opposite direction of seismic force, keeping in mind that water will act as an inertial force(reactive force) that will damp the displacement amplitude.
3. All dynamic water pressure (all directions) is applied simultaneously, canceling each other, along with seismic forces.

In all of the above scenarios, water weight was considered in the calculation of base shear.
Following are the results for each scenario.
1. extreme load effects, 12 inc deflections are noted at 90 feet above the ground, which doesn't look reasonable.
2. quite low load effects, 2-inch deflections at the same height(90').
3. Moderate Load effects, 5-inch deflections at the same height(90').


The third one is similar to as described by bones206.
But I am still a little inclined to take water load as a damping agent, am I missing something important?

Thank you

 

[JStephen]

By "dynamic water pressure", are you meaning pressures associated with sloshing only? (The convective component as described in some of the posts above)? Or do you mean all pressures other than static? (IE, impulsive + convective as described above)?
In your first post, you mentioned "take the water load in seismic weight of the building". If this is a concrete structure or is incorporated into a building of some kind, you could have a structure considerably more rigid than a common water tower.
And if you have substantial seismic forces, you can easily wind up with unreasonable forces or deflections, and have to adjust the type of tank, type of supports, etc., to get a reasonable design.

 

[Ahmad Uzair Arshad]

dynamic pressure meaning total pressure water is exerting in the dynamic state (convective + impulsive) combined using SRSS as defined in ACI 350.
the structure is a 100 ft high concrete water supply tank. with horizontal bracings at 15' intervals. bottom two-story may be used as office floors.
Actually, I am quite confident that the walls of the tank will be designed for this pressure but what about the whole structure.
This water is not attached to the structure, hence it will not act with the structure. on the contrary under seismic forces i will act like an inertial damper.
just like a heavy pendulum.
This feels rational.
or Am I missing something?

 

[bones206]

This graphic may help. Note how the impulsive component is idealized as a mass rigidly connected to the tank wall. The convective component is idealized as a spring that can oscillate.

Section 15.7.10.2 above is instructing us to ignore the convective component and consider the entire water volume to be rigidly connected (i.e. 100% impulsive), unless certain conditions are met that permit you to consider convective component effects.

 

[HTURKAK]

- The tank wall(s) would be designed for the pressure combination = static+impulsive + convective .. this is true..
- The whole structure would be designed for the comb. with seismic loading . The seismic loads would be calculated for impulsive and convective portion and the effects would be calculated with SRSS combination.
- I suspect you have confusion with tuned mass dampers ( TMD )
- I also suggest to look worked examples at the doc. which i released at previous respond.

If you post the details of elevated tank ( square, circular, stage properties, seismicity etc.) you may get more useful responds..

 

[Ahmad Uzair Arshad]

Thank you.
I get it now, the water load will be considered in base shear calculation & tanks walls will be design for dynamic pressures of the water.