One possibility for storing the oil produced by an offshore well is to make use of a moored submerged buoyant tank. Tests were conducted with a 1:49 scale model of such a tank, with neutral buoyancy mooring lines, in the 6 ft. deep by 8 ft. wide by 200 ft. long Wave-Towing Tank, University of California, Berkeley, under contract to the California Research Corporation (presently the Chevron Research Corporation).
The "prototype" was a circular cylinder approximately 190 ft. long by 22½ ft. in diameter with hemispherical heads, with an empty weight of 874 kips, and with a 60½ ft. longitudinal radius of gyration. Two general series of tests were conducted:
Vertical mooring lines only, and
vertical, breast, bow, and stern mooring lines.
The "prototype" water depth was 245 ft. and the top of the submerged tank was 185 ft. above the bottom Measurements were made of mooring line forces and tank motions with the model in head, quartering, and beam seas.
For the case of vertical mooring lines only, it was found that the dynamic force in the mooring lines due to wave action could be predicted rather well for beam seas by use of the "wave force on a pile" theory. Furthermore, it was found that the limits of dynamic forces could be predicted for the model in head seas using the same theory. The high forces measured in the vertical mooring lines due to a shifting of the buoyancy load from one line to another, when they were not exactly parallel, leads to the conclusion that extreme care will be necessary during installation. The greatest motion of the tank occurred in head seas in waves of 21 sec., prototype (the same value as the natural period of surge of the moored tank), with the surging motion of the tank being as much as 6 times the wave height and 4 times the total horizontal motion of the surface water particles.
When breast, bow, and stern mooring lines were added to the system, the motion of the submerged buoyant tank became negligible. It was found, however, that at some point in the transference of the buoyant force from the vertical to the lateral mooring lines (the percentage load transference depending upon the wave height and period to which the system was subjected) the vertical mooring lines would just slacken at one point in the cycle. This resulted in an impact type of loading upon take-up with high forces and considerable vibration. The magnitude of the peak force under these conditions could not be estimated due to the response characteristics of the force recording system.
A plexiglass tube was used for the tank and aluminum tubes were used for the mooring lines. Availability of plexiglass tube sizes (one being 5.5 in.O.D., 1/8 in. wall thickness) fixed the scale at 1:49 as the closest possible to the desired scale. This geometric scale fixed the time scale as 1:7.0 and the force scale at 1:117,650.*