Hydrodynamic - ACI 350.3 Housner vs USACE EM Westergaard

[ATSE]

To my structural friends experienced with seismic / hydrodynamic loading on concrete walls using both Army Corp approach and ACI 350.3 approach.

Comparing the USACE / Westergaard ("Army Corp") equations for seismic hydrodynamic loading on walls vs ACI 350.3, the shear and moment results are substantially different.
See figures below.
For a moderately seismic site with 30' deep water and 250' long tank (or stilling basin or check structure bay...), the seismic component for ACI 350.3 is about 2 to 3 times larger than USACE / Westergaard.

Has anyone else run into this challenge? Inviting comments regarding any reconciliation for the large discrepancy in resultant forces.

ACI 350.3:

US ACE EM-1110-2-2100 / 584:

 

[ATSE]

Still looking for comment.

 

[JStephen]

I live in the round-tank world, so can't help much on the specifics.
For larger tanks, the sloshing period becomes very long and the convective contribution becomes fairly small, check if that's happening in your case.
Make sure the calculated force on the wall is the force on each wall, not the combined effects of forces on near and far walls.
If either procedure is considerably out of date, that could be the issue.
The current round-tank equations show impulsive pressure as a quadratic, not linear as shown in the first figure.
And the codified methods wind up with multiple factors multiplied, make sure you don't wind up comparing a factored/reduced/ultimate/allowable load to something else.

 

[GC_Hopi]

I doubt you can reconcile this. You would need to look into the derivation of the equations and usually the equations are simply derived differently. It's just different strokes for different folks. I have checked the seismic loading for tanks with ASCE 7 and ACI 350 and they give different results because the codes use different coefficients in the equations.

Edit: One recommendation is to graph the equation so you can get a feel for the difference. Sometimes these equations give similar results for a specific range then diverge