In addition, for each of these parameters there are different scales of 

 satisfactory performance. For example, manned installations require a high 

 degree of confidence in their stability and, therefore, can tolerate very little 

 motion; whereas, unmanned and relatively insensitive seafloor installations 

 are often capable of tolerating larger settlements without impairment of their 

 mission. In the extreme case, an installation involving numerous identical 

 structures (each of which is unmanned and duplicates to some degree the 

 mission of the others) may be capable of tolerating (for the sake of economy) 

 some failures. In a situation such as this, the scale of performance behavior 

 may be such that fewer than two failures (in the overall installation involving 

 a large number of individual implants) may be considered satisfactory. 



The scale of performance may also be influenced by factors such as 

 soil or sediment province, physical environment (such as water depth, current 

 velocity, and biologic activity), and design life of the installation. In spite of 

 the wishes or needs of the owner or operator of an installation, such factors 

 may force a shift of performance scales. For example, performance satisfac- 

 tory at 6,000 feet may be unacceptable at 60 feet. Such a shift is, in essence, 

 attributable to the state-of-the-art of certain technologies which limit or restrict 

 performance. 



To ascertain the scale of performance and the parameters affecting 

 performance, the behavior of the in-situ foundation must be monitored. The 

 monitoring of foundation performance serves six purposes: (1 ) it initially 

 focuses objective thought on the type of performance which is required, on 

 the level of performance which is satisfactory, and on the parameters which 

 should be considered for satisfactory performance; (2) it keeps the operators 

 informed of the condition of the installation so that remedial steps can be 

 taken if they become necessary; (3) it evaluates the success of the foundation 

 design procedure and the assumptions made therein; (4) it points out behavior 

 parameters which may not have been considered at the time of the design; 



(5) it begins to give a statistical view of foundation behavior and failure; and 



(6) it forms a library of past experience or case histories, which can be used 

 in future analyses and comparisons. 



Such monitoring of foundations on land has been common throughout 

 the ages. Earliest design techniques were based strictly on observations and 

 experience (the empirical approach). More recently, the need for performance 

 monitoring, as a means of improving foundation design capabilities, has been 

 pointed out in prominent technical literature (Casagrande, 1965; Feld, 1965). 



As discussed earlier, a variable and dissimilar number of behavior 

 parameters collectively (and often mutually exclusive) contribute to an instal- 

 lation's degree of satisfactory behavior. The parameters which are most 

 commonly important, and thus worth monitoring, are (in probable decreasing 



