-BF 

 Q„ = Ae * = (1.63 x 10" 2 ) exp [(-31.9) (0.089)]= 9.5 x 10" 4 (26) 



Therefore 



Q 0wen = (9 ' 5 x 10" 4 )(TgH g ) = (9.5 x 10" 4 )(8)(32.2)(5) 



=1.2 ft 3 /sec/ft of seawall (27) 



Discussion 



44. Owen's method is the only method based on experiments with irreg- 

 ular waves. Analysis of the experiments has been presented in two papers, 

 Owen (1980, 1982). However, the data have not been published.* Owen (1980, 

 1982) does not discuss possible scale effects in his small-scale (1:25) 

 overtopping tests. Aaen's (1977) experimental work is discussed later in this 

 report. It indicates that scale effects in overtopping may be very large. 



45. The A and B coefficients in Table 1 are average values of five 

 identical runs. The spread of the five runs allows Owen to determine confi- 

 dence intervals for Equation 20. For an estimate of overtopping Q from 

 Equation 20, the 95 percent confidence interval is from Q/3 to 3Q . Owen 

 implies that this spread is entirely due to the irregularity of the waves. 

 However, it can also be assumed that this spread is partly caused by the 

 influence of other variables not explicitly considered in Equation 20. This 

 confidence interval must be considered when using Owen's method. 



Other Methods 



46. Several methods of estimating overtopping have not been dis- 

 cussed. Cross and Sollitt (1970) derive an analytic expression for over- 

 topping volumes caused by monochromatic waves, but they do not attempt to 

 extrapolate to irregular waves. Kikkawa, Shi-Igai, and Kono (1968) treat 

 monochromatic wave overtopping as a form of weir flow. Jensen and Sorensen 

 (1979) present dimensionless equations and curves for estimating overtopping 

 volumes, but do not clearly describe either the dimensionless variables or the 

 empirical coefficients. Kobayashi and Reece (1983) use an assumed joint wave 

 height and period distribution, a monochromatic-wave runup formula, and 



* Personal Communication, Owen 1984. 



25 



