1st concrete tank

huangyhg

1st concrete tank
greetings all,
i am designing my first concrete tank and had a few questions.  the tank is (3) 36' sq. chambers in a line, sharing walls.  it is a total of 23' tall (plus the slab depth), is burried 20', and does not have a top. the water table is 11' below the top of the tank.
i have completed the slab design using risa 3d and compression only springs, by the way jae's 9-18-2000 suggestion to use upper and lower bound soil constants to envelope the solution has already helped me. (i don't have soil info...yet). fyi the slab required had a d=23".  which seems thick to me but is helping me with the bouyancy problem anyway.
now i am ready to start the wall design/detailing. here are my thoughts, i'd appreciate any comments:
1. the walls could be designed as a cantilever (conservitive), by using pca's "rectangular concrete tanks" publication moment/shear coefficents, or by fe analysis. i'm leaning to designing as a cantiliver. how have you performed this design? how ever performed, i will consider the different loadings, backfilled/not backfilled, empty/full chamber.
2. waterstops will be required at the joints.  i am concerned about their placement in relation to the joint reinforcement. if the reinf. has say 2" cover i am not going to be able to fit the ws between the bar and the inside of the tank.  on the other hand if i put the ws more toward the center of the wall the fluid will be able to reach the unprotected reinf. it will even get worse at the divider walls (on either side of the middle cell) where they meet the perimeter wall and the slab. any experience with standard construction/design would be greatly appreciated.
3. lastly, thanks for you patience, i intend to design the corner reinforcement to restrain the wall's end reaction, is that correct? i will probably end up having hooks for each mat of horizontal wall reinforcement.
thanks in advance
jt
preng,
why to design the walls as cantilevers?    the pca's publication has tables within the parameters of your tank.    a  b/a = 36/12  for interior water load, and b/a = 36/20 for exterior soil/groundwater load could be used.        for many years before the advent of the digital computers concrete tanks were designed using that pca's publication.
just a question, why have waterstop at the interior divider walls?  i can understand the exterior walls, so that the tank doesn't leak to the soil, but what is the problem if it leaks tank to tank?  the installed cost of pvc waterstop runs around $2.00/lf maybe that is cheep insurance?
dlew,
thanks for your comments.  i think that my post was misleading in that the groundwater table level is 11' below the top of the tank, the interior fluid level will normally be 20' deep but i am designing for mistakingly filling the tank(s) to the top. all that said your point about pca's pub. is understood.  the ratios b/a will have to be interpolated between two tables, is that correct? i suppose my reluctance to use the pca pub is that i am unfamiliar with it.
jheitd,
thats a thought, although the tanks will be leak tested prior to use and i am sure that each chamber will need to be leak proof.  i will look into it further though.  have you designed a tank (sewage aeration) and not found the need for ws on the interior divider walls?
thanks
jt