## P-delta effects on Elevated Tank Per AWWA

## P-delta effects on Elevated Tank Per AWWA

(OP)

I'm asking a question I've came across in the past, but haven't had to deal with until recently.

Per AWWA D100-21 13.3.3.2 it says to include P-delta effects for a lateral drift equal to three times the elastic deflection. ASCE 7 has something similar. Correct me if I am wrong here, but they are basically giving me a Large p-delta value that they want me to design to, but I don't need to use a lateral load to generate that drift and have my structure resist that load. In other words, my structure only needs to resist calculated lateral seismic loads prior to the factor, and also the additional moment created that results from the factored drift.

The reason I ask is because this is in principle entirely different than how AISC 360 handles P-delta effects. P-delta effects in AISC 360 uses initial imperfections that are generated used notional load which are derived from gravity loads. AISC also makes adjustments to the stiffness, but I'm not going to get into that here. These notional loads are effectively added to lateral loads and the p-delta analysis is ran iteratively to determine loading and calculated drift. How does everyone reconcile these different approaches? And if you are using software for the column and bracing design, How do you incorporate the AWWA requirements into the model without using larger than life notional loads? Additionally, Does AWWA not require analyzing p-delta effects when seismic analysis isn't necessary?

Per AWWA D100-21 13.3.3.2 it says to include P-delta effects for a lateral drift equal to three times the elastic deflection. ASCE 7 has something similar. Correct me if I am wrong here, but they are basically giving me a Large p-delta value that they want me to design to, but I don't need to use a lateral load to generate that drift and have my structure resist that load. In other words, my structure only needs to resist calculated lateral seismic loads prior to the factor, and also the additional moment created that results from the factored drift.

The reason I ask is because this is in principle entirely different than how AISC 360 handles P-delta effects. P-delta effects in AISC 360 uses initial imperfections that are generated used notional load which are derived from gravity loads. AISC also makes adjustments to the stiffness, but I'm not going to get into that here. These notional loads are effectively added to lateral loads and the p-delta analysis is ran iteratively to determine loading and calculated drift. How does everyone reconcile these different approaches? And if you are using software for the column and bracing design, How do you incorporate the AWWA requirements into the model without using larger than life notional loads? Additionally, Does AWWA not require analyzing p-delta effects when seismic analysis isn't necessary?

## RE: P-delta effects on Elevated Tank Per AWWA

See 13.3.3.4 for an increase factor that is also evaluated.

Note that D100 is written on an allowable stress basis, and seismic load is calculated on that basis as well.

See 4.6.2 of AWWA D100-21 for a P-delta requirement unrelated to seismic.

## RE: P-delta effects on Elevated Tank Per AWWA

You basically described exactly how I figured it says me to do it. AISC method includes provisions using allowable stress so it shouldn't be a problem.

I think the reason I am so confused by because of my comparison to this method of p-delta effect to AISC. They aren't even close to the same thing. AISC uses a fictional load to create the deflection. So software that has been built around using AISC doesn't have any way of doing what AWWA requires unless you are adding a fictional load to create a deflection of 3 times the drift. So is everyone else doing these checks by hand, or is there software that is able to do this? I suppose I could reduce the stiffness to 1/3, but that seems like way overkill. I'm kinda leaning to just doing it the way AISC says to do it and call it a day.

Section 4.6.2 seems to apply to pedestal tanks, not to column supported. The column supported section in Section 4 doesn't seem to address p-delta effects.

## RE: P-delta effects on Elevated Tank Per AWWA

AWWA D100 has its own allowable stress provisions that are different from, and more conservative than, AISC. It allows AISC allowable stresses for seismic design of leg tanks, but I doubt the overall design is normally done via software tailored to AISC.

AWWA D100 doesn't require P-delta analysis for some cases, but whether it ought to be included is a different issue. Where I run into this is in the design of small-diameter foundations with estimated differential settlement- so there's a potential for a distinct tilt, which is not well addressed by the codes.