43 FLANGE 0.D. 
TWENTY. 
2°%4-QN-2x 23) two NPT. 
REQD FOR 
+f oo AVERAGE 
Figure 72. Experimental GRAYLOC flanged test fixture, showing details of 
gasket shape (Mikesell and Brown, 1968). 
Dewatering of a cavity sealed as discussed above and diagrammed 
in Figure 57, does not appear to present any insurmountable problems but 
will require very careful planning. One approach might be to drop a partially 
insulated container of liquified air in the excavated space before closing the 
hatch and providing a siphon (Figure 73) so that as the liquid air boiled and 
displaced the water, the air would remain as a large bubble. The system would 
still be at ocean pressure, and a previously installed pump would be energized 
by power from the surface to reduce the high-pressure air to the desired one 
atmosphere. There would always be the danger that combustible gases from 
a subbottom source would provide an explosive atmosphere. Since these 
gases might not easily be detected or might not be released at full ocean 
pressure, one method might be to charge the space with liquid nitrogen 
initially and later provide oxygen just before the entry of personnel. 
Summary 
Methods for sealing and dewatering a subbottom cavity In rock 
appear to be within the state-of-the-art. Further trade-off studies and 
investigation of potentially superior systems are needed. 
94 
