will probably require fewer tests than the first conceptual investigation, 

 but with a larger scale and more detailed model the cost per test will be 

 higher. 



4. Armor Stability 



a. Economic Considerations - Preliminary studies of the history of 

 conventional breakwaters along the Pacific coast made it clear that it was 

 not economically feasible to design the seaward revetment to be completely 

 stable against all possible storms. All breakwaters in deep water with a 

 severe exposure are expected to and generally have suffered occasional 

 damage. The extreme consequences of complete failure of the island's 

 revetments made this problem much more critical than for breakwater design. 

 A rational approach was therefore developed which consisted of six steps : 

 [1) prediction of the frequency of occurrence of large storm waves at the 

 site; (2) correlation of predicted storms with laboratory tests of revet- 

 ment sections; (3) estimates of cost and damage for various trial designs; 

 (4) evaluation of damage to the various designs; (5) economic analyses of 

 various designs; and (6) selection of final revetment sections. 



Table 4 siommarizes the calculations for estimating average annual 

 repair cost for a trial design of the island's seaward face revetment. 

 Columns 1 and 2 were developed by oceanographic study of the island site. 

 For column 3, the maximum wave of a storm was assumed to be 1.9 times the 

 significant wave of the storm. Column 4 correlates the storm waves with 

 laboratory waves. 



Based on observations of the model tests, it was assumed that the maxi- 

 mum wave of a wave train best describes the destructive ability of the wave 

 train. The maximum wave of the laboratory wave tests was about 1.16 times 

 the laboratory designated wave height for the test, a figure derived from 

 a study of a typical wave record of the wave channel tests. Based on these 

 assumptions a wave height = 1.9/1.16 = 1.64 times the significant wave of a 

 predicted storm was used in the modified Iribarren formula to determine the 

 armor for a no-damage design. Columns 5 and 6 are based on damage estimates 

 obtained in the laboratory wave tests. Considerable refinement of such dam- 

 age estimates was later published. (Hudson, 1959.) The estimated repair 

 cost in column 7 is based on a unit price for armor materials replaced or 

 recovered, and a lump sum cost for mobilization and demobilization. Column 

 8 is column 7 divided by column 2 and gives the incremental part of the 

 average annual repair cost contributed by each class of storms. 



Figure 7 is a plot illustrating the economic analysis of several trial 

 designs. Curve a. is the average annual repair cost for various designs 

 computed as illustrated in Table 3, and curve b is the present worth of 

 the average annual repair cost for a 25-year period at an interest rate of 

 6 percent. Curve a is the estimated construction cost of each of the 

 various trial designs. The capitalized cost, curve d, is the sum of 

 curves b and a and its low point represents the most economical design. 



25 



