Few things spring to mind.
I get about 80 Tonnes (78 long) of buoyancy, bearing in mind that seawater is a bit denser than fresh. It would pay know the dimensions of your cylinders to more than one significant figure - especially the diameter. Your final buoyancy calcs will want to take account of the shape of the cylinders (which won't be cylindrical!)
The reason you need to know the buoyancy quite accurately is that by the time you add the structure and the payload, the amount of buoyancy you have in reserve is going to be much, much less than you started off with: This works a bit like the Micawber principle.
You will need to think about how much you can tolerate losing a few tanks (either through flooding, or through detachment)
With buoyancy coming from tanks that are evenly distributed and each almost completely immersed, you are going to need to look closely at stability. Unlike a traditional hull, this kind of arrangement doen't automatically add buoyancy in the right place when you load it unevenly. Look at what happens if something makes it roll, if you don't load up evenly, while you're moving things aboard or if you lose a few adjacent tanks.
If you attach each tank individually to a platform above water level, all the structural strength and stiffness is going to have to come from the platform - which is probably something you're otherwise keen to keep quite thin. Don't design for an evenly distributed load, because you won't get one. Even if you can be sure the raft is never going to be exposed to a big swell (having your tanks mostly submerged actually works a bit to your advantage in a swell), you are going to have to think about what happens if you put the structure aground on a falling tide (imagine the whole structure supported only at opposite corners - or perched on a pinnacle somewhere under the middle).
Similarly, in the event of a collision or grounding, the attachment of the first tank in the firing line (probably one of the ones at the periphery) is going to find itself the weak link between getting on for a hundred tonnes of moving metal and something much bigger and is going to need to be designed to withstand that.
Mooring and tug attachment both generate quite beefy point loads which will need to be spread through the structure.
I think this project is unconventional enough to need the detailed involvement of a professional naval architect. This isn't me touting for business (I don't have the right skills and live in the wrong part of the world anyway); Remember that something this size is going to turn out to be environmentally hostile, not to mention personally inconvenient, if it ends up unevenly distributed along forty miles of foreshore.
A.
