moored behind the breakwater. The last functional performance criteria would 

 thus be that xf. of the waves transmitted by the breakwater during a 50-year 

 storm would be in excess of 1 m. The return period convention is in keeping 

 with traditional practice, though the phrase "with 2 percent probability per 

 year" would be a more accurate description of the storm of interest. Esti- 

 mated probability per year might be a more appropriate increment in terms of 

 providing even steps of cost between alternatives, but either convention will 

 serve. The value of y.% should relate to some consideration of the actual 

 number of waves of H* or greater necessary to cause a measurable effect. A 

 storm whose peak conditions lasted 3 hr with T^ = 10 sec would include 

 roughly 1,080 waves. A small value of x^ is appropriate, on the order of 

 1 percent, which for the example condition would include 10 or 11 waves. 

 These waves would not likely occur in sequence, but a few of them might. 



91. The shorter return periods of 20 or <10 years might be too risky 

 for a small boat harbor where relatively fragile vessels and mooring facili- 

 ties are planned immediately on the lee of the breakwater. These criteria are 

 reasonable, however, when losses due to cargo handling inefficiencies or ves- 

 sel transit time are all that is at stake. The 50-year storm is, on the other 

 hand, a very conservative criterion for wave transmission. At least four 

 functional performance criteria should be addressed to assure identification 

 of an optimum design. 



92. The second column of Table 7 includes choices for structural in- 

 tegrity criteria in terms of the damage to the armor layer, as might be esti- 

 mated by Equation 14. The 7»D(Hjj) value chosen should be consistent with 

 the incipient damage level, as measured in model experiments pertinent to \ 

 the breakwater design at hand. H^ is the wave height applied in analyti- 

 cal stability relations. Return periods of 30 years or less for the storm 

 represented by H^j will plainly involve substantial expected damage and 

 therefore should be investigated only for minor breakwaters where repairs can 

 be easily accomplished or postponed without significant adverse consequences. 

 Long return periods greater than 50 years are important to address, however, 

 since rubble-mound breakwaters require such a tremendous commitment of equip- 

 ment and materials to repair. The risk of affordable quarrystone being un- 

 available 30 or 40 years in the future might be great, even though it may be 

 readily available at present. Repair of breakwaters in remote areas involves 

 high mobilization and demobilization costs, even for small repair efforts. 



61 



