High-rise cold storage pile foundation

[p.firtinidis]

I'm quoting a high-rise cold storage (2.800m2 X 30m height) founded on a friction-pile raft foundation, where the raft is also the final floor. Raft thickness is 1,00m due to the racking's anchorage demands. The freezer temperature is -22 Celsius (-8 F) and the water table is about 2m below floor level. Client's geotechnical study calls for a pile mesh of about 5X5m and Φ1200 reinforced piles.
The clients' plans include floor insulation and a glycol-heated sub-floor as is usual but my concern is the piles present a thermal bridge that long-term will cause the ground around them to freeze. Since summer air temperature reaches 40oC (100 F), I also believe that seasonal freezing-unfreezing cycles will cause significant soil degradation (especially around the perimeter piles), so in addition to frost heaving I'm concerned for the system's seismic performance.
Any experience or thoughts on how to calculate/mitigate such issues?
Thanks,

 

[oldestguy]

A difficult problem for sure. One possible mitigation of heave potential would be to first replace all soil to a given depth (that's also a question) with non frost susceptible soil. Granular soil with less than 5% passing the number 200 sieve is non frost heaving type. Under this plan there would be a weight of non-frost susceptible soil around the upper pile lengths adding to load on the piles sufficient resistance to any uplift by deeper soil freezing. I know this is a way to minimize soil heaving by moisture-expansive soil types, but for freeing, not sure. Knowing the ground water table may help also, since water gives off heat when freezing, limiting the depth of the frozen front.

Is it at all possible to use a floating thick slab, perhaps with some soil improvement or replacement and eliminate the piling?

 

[p.firtinidis]

Thanks for the reply. Indeed the freezing depth is among the first that need to be quantified, I think a coupled THM FE model would be necessary to find the long-term 0 oC isothermal. Since the piles are large in comparison to their mesh a large model is needed (cant assume axis symmetry and simplify the model to one pile, so I would guess 1/4 of the raft would have to be modeled). I really liked your latent heat idea, perhaps freezing depth is limited by the water table and if indeed the problematic range is limited perhaps replacing the in-situ silty sand with a coarser material would help. Of course reliable seasonal GW table measurements/assumptions would have to be taken. To your last point, I'm afraid plastic and differential settlements make a shallow foundation solution untenable.

 

[BigH]

Check out references on ice rinks and cold storage buildings. A few that I am aware of:

http://web.mit.edu/parmstr/Public/NRCan/CanBldgDigests/cbd061_e.html

https://www.plastifab.com/technical-library/design-manual/0406_Ice Rinks.pdf

Throwing this in:

https://www.ngm2016.com/uploads/2/1...es_in_thawing_a_frozen_ground_sinnathamby.pdf