projected d and F were similar to those tested. There was, however, some 

 question whether the effects of beach erosion should be calculated based on a 

 Q versus F' plot with all the measured data (30 to 100 percent DWHAWB) or 

 based only on the 100 percent DWHAWB measured data. This question was re- 

 solved by choosing the more conservative (100 percent DWHAWB) data plot (Fig- 

 ure 11) to calculate the effects of beach erosion for the design storm event 

 where erosion potential is greatest. The method for calculating changes in 

 overtopping rates for the maximum H that can exist at each swl is ex- 

 plained, and the respective beach erosion elevations are tabulated in Appen- 

 dix B. The percent difference and the percent decrease in overtopping rates 

 between the data trend for the Phase II seawall at the +1.0 NGVD beach erosion 

 elevation and the projected trend for the same wall at a +3.4 NGVD beach ero- 

 sion elevation for hurricane conditions at the three swl's tested are given in 

 Table 3. These numbers are based on Figure 11 where only those data collected 

 at 100 percent DWHAWB were used. The percent difference values in Table 3 

 were used to predict the change in overtopping for the surge hydro- 

 graphs (Lillycrop, Pope, and Abel, in preparation). 



Table 3 



Overtopping Comparisons Using Hurricane Conditions at the 



+1.0- and +3.4-ft NGVD Beach Elevations 



swl, ft 

 +9.5 

 +8.0 

 +7.0 



Percent Difference* 



46 



22 



9 



Percent Decrease** 

 54 

 78 

 91 



* Percent difference in Q for data at +l-ft NGVD beach elevation and pro- 

 jected values at +3.4-ft NGVD beach elevation. 

 ** Percent decrease in Q for beach elevation at +3.4-ft NGVD versus +1.0-ft 

 NGVD. Percents are based on Q values calculated in Table B3 (Appendix B) 

 using only the 100 percent DWHAWB data points. 



Wave Setup and Seiche in the Wave Tank 



29. In almost all wave tank tests, there can exist local wave setup and 

 seiche. Both of these phenomena can occur in the prototype, but their ampli- 

 tude and overall effects may not be the same as in model tests; thus, it is 

 important that they be identified and, to the best extent possible, accounted 



26 



