reason for difference in power.prop.test and epiR::epi.sscompb results

Question

I was trying to rework the examples in Epidemiology Study Design and Data Analysis by Mark Woodward using R. Example 8.12 deals with calculation of sample size to test for a population proportion of 0.3 against the one sided alternative that it is greater than 0.32.

I tried two commands

power.prop.test(n = NULL, p1=0.3, p2=0.32, power =0.9, alternative = "one.sided")
epiR::epi.sscompb(n = NA, treat = 0.3, control = 0.32, power = 0.9, sided.test = 1 )

The results (sample size) given by the two functions are very different.

Two-sample comparison of proportions power calculation 

              n = 9157.159
             p1 = 0.3
             p2 = 0.32
      sig.level = 0.05
          power = 0.9
    alternative = one.sided

NOTE: n is number in *each* group

$n.total
[1] 4568
$n.treat
[1] 2284
$n.control
[1] 2284
$power
[1] 0.9
$lambda
[1] 0.937500 1.066667

The result by epi.sscompb agrees with the value in the book.

Can someone tell me if there is an error in the code? If there is no error in the code, why are the results different?

Thanks, ajith