Seismic design of swimming pool 2

[patswfc]

I am reviewing the design of a circular pool, 25m diameter 2m deep.
The pool will be in the ground and is a standalone structure.
The site is in France with lower seismic requirements.

A colleague has suggested that the pool doesnt need to be designed for seismic as it
is in the ground. Does anyone know if this is correct and can point to a reference
to back this up?
Underground structures, such as tunnel can experience damage during seismic events,
therefore I would think that the pool would also need some sort of check for this.
P.S. I have only a very basic understanding of seismic design.

Thanks

 

[EDub24]

I don’t know of any reference that says this. It could just be engineering judgment. It doesn’t hurt to run the numbers and check. I’m not sure what are the requirements are for pools but ACI 350.3-06 gives seismic design parameters for liquid-containing structures. I think the intent was for water treatment plants or structures like that but it’s probably safe to assume the criteria can be extended to include pools. Tunnels see damage because they are so long and are affected more by ground movement which is not the case with pools or other smaller underground/below-grade structures. You will see some additional dynamic soil loading on the walls in addition to the hydrodynamic loading on the trailing wall although if you are in an area of low seismic ghen the hydrostatic loading might still be able to counteract all of that. Anyways, run the numbers and verify seismic doesn’t control and then from now on you can use your judgement and say it’s OK by inspection.

 

[racookpe1978]

South coastal areas of France are near the Italy (active volcano), Sicily, Alpine, Perrenies mountain-building regions between Spain and France, so seismic activity is possible. (Heck, I've been wakened up by an earthquake in the Appalachian foothills in west GA, USA, so earthquakes do happen in unexpected places.)

For a open-topped recreational pool, cracking of the concrete (loss of water, flooding nearby ground and basements) would be the hazard, economic loss is limited to draining and re-repair of the cracks. With low damage probable, over-building for a low-probability earthquake would be the problem. Compare this to, for example, an open-topped cooling water pool for reactor wastes. Great damage near-certain if the pool cracks and is not refilled!

Seismic analysis (sideways ?) earth movement and choosing a realistic acceleration to evaluate is relatively inexpensive compared to everything else. But doing something about the design to avoid the cracking could be very, very expensive: More rebar and mesh, thicker concrete, different shape to eliminate a stress-riser at a sharp corner, etc.

 

[hokie66]

Not sure about France, but the main intent of seismic provisions is to protect life and limb. Damage to man made structures is not assured, but safe egress is the aim. An inground swimming pool is probably not a high priority structure in that regard.

 

[JoshPlumSE]

Swimming pools are basically just retaining walls. The worst load conditions are generally when they're empty and not filled.

Seismic is not generally an issue for a few reasons.
1) Seismic isn't really an issue for retaining wall either. At least there isn't a history of retaining wall failures creating life safety issues. At least not for that height.
2) There isn't really a life safety issue if the pool walls were to get damaged. At least not for in ground pools.
3) The curved nature of the pool walls and floor actually really help the pool strength. That's almost always ignored in design, but it really will help it out a good deal.

I did some ACI 350 sloshing analysis on an above ground swimming pool once. But that's way overkill a below ground pool.

 

[patswfc]

Thanks everybody for your replies.

EDub24, I agree that tunnels would be a lot more susceptable to damage due to their length compared to a pool.

racookpe1978 and hokie66, avoiding collapse is the main criteria. I would hope that as soon as an earthquake was felt, anyone in the pool would quickly make their way out of it. My concern is that the wall could collapse inwards potentially traping, or worse, someone in the pool. The fact that it is a circular tank should help reduce this risk in my mind.
The second criteria of damage limitation obviously isnt as critial as: a) this doesnt involve a risk to life and b) the client has confirmed that they are OK with whatever damage may occur during a seismic event! That said, I would want our design to follow good practice and try to limit damage as much as is practical.

Joshplum,
We have a paper relating to retaining walls, where ka(e) is calculated based on the Mononobe-Okabe equation. We would generally design retaining walls based on ka (or Ka(e) in this case) for stabilty checks and Ko for strcutural checks. When we calculate ka(e) for this project, it is less than the ko value that we have used for the structural design of the tank. As this structure isnt technically a retaining wall, I am wondering if it is pobbible for the walls to experience Ko(e) or even Kp(e) (if there are such values) in the event of an earthquake, which would likely be higher than Ko that we have used in the design. I cant seem to find any info on ko(e)/kp(e).
I have just found Mononobe-Obake equation for Kp(e) in an annex of eurocode 8-5, so will get the design checked for this and amend if more onerous than the standard ULS/SLS design requirements.

Thanks all for your help.

 

[MotorCity]

Your colleague and others above are correct, the pool does not need to be designed for seismic forces.

Seismic forces are based on ground acceleration. Structures located below grade move with the soil during a seismic event, so they are both moving at the same velocity. In other words, no change in velocity means no acceleration relative to each other, means no net force relative to each other.

The opposite is true for above grade structures. As the ground moves, the structure will sway back and forth in an attempt to "catch up" with the ground movement, thus resulting in different velocities and resulting in acceleration.

I seem to remember article(s) in Structure Magazine that offer insight on where to take the base level of a structure for the seismic design of a full basement, partial basement, walkout basement, etc.
The in-ground pool would be similar to a basement structure.

Also as others have said above, an in ground pool represents a fairly low threat to life during a seismic event.

 

[BridgeSmith]

"My concern is that the wall could collapse inwards potentially traping, or worse, someone in the pool."

Assuming that's even possible with a circular pool, which it doesn't seem like it would be, the water would have to displace and flow around the wall. That would be relatively slow-motion collapse. That said, would it really be that difficult or expensive to embed some welded wire fabric in the wall?

 

[glass99]

We used to use the Mononobe-Okabe soil pressures for designing bridge abutments and retaining walls for seismic forces.

A key question is whether the pool is assumed to be empty or not. If its full, the seismic soil pressures are likely less than the hydostatic pressure of the water. A thin wall swimming pool is stable because passive soil pressures will be much more than hydostatic pressure, and active pressure will be much less. Also: is the soil wet?

 

[BridgeSmith]

"A key question is whether the pool is assumed to be empty or not."

If the pool was empty, presumably no one would be in it, so the life-safety hazard is non-existent. The potential for damage would also be minimal, since as MotorCity pointed out, an in-ground pool will just move with the ground.

 

[glass99]

Yes, generally speaking pools are not designed to be empty, but sometimes they get drained in the winter, and I could see a client in a seismic zone not wanting their pool walls to collapse right before the season starts. For example, if I was Greek and hosting an Olympic event.

 

[patswfc]

Thanks all for your additional comments.

Motorcity, I see where you are coming from about not having to design for seismic as the whole structure is in the ground. However would this not only be the case if the pool tank and ground movements are in perfect harmony. As i have stated before, I dont have a great knowledge on seismic design and am maybe overthinking this, but i dont think that this harmony of movement between the ground and the tank is a safe assumption. I have found the attached paper which discusses the use of dynamic earth pressures for underground structures that are relatively small (i.e. not like a tunnel, which would be more susceptible to damage).
In the end, we have checked the structure for Kp(e) earth pressures based on the Mononobe-Okabe equation in the eurocode. This assumes that a seismic wave first hits the tank moves in to the ground on the other side, before it has moved!! This is probably way over the top (not really reasonable assumption), however proposed design didnt need amended for this condition.

Glass99, The pool is likely to be empty for 3/4 the year (autumn to spring). This large pool is for someones holiday home. They apparently will only be there over the summer!