**How do you define the biological limit**

I think it very much depends on how you define a biological limit. Fertility depends on genetic quality, which varies from one individual to another. For example, in your calculation you took lifespan as a constant but it obviously is not a constant. Fecundity is reduced by all these small mutations segregating in the human populations that reduce the fecundity of any single human. There exist no human that is free from all deleterious alleles. In the below answer, I will assume that by biological limit you mean the mean fitness of a female that would live in the same environment than we do today but would not carry any single of the deleterious locus that are segregating in the population.

**Mutation load Theory**

The decrease in the mean fitness of an idealized population where no deleterious mutations would exist compared to the realized mean fitness of a population is called the mutation load. The mutation (in an infinite population) is a function of the genome-wide mutation rate only. Let the genome-wide mutation rate in human be $U=2.2$, the mutation load $L$ in the human population is equal to $L=1-e^{-2.2}=0.89$. The estimates taken from [this video][1]. You may want to have a look to the video in order to have a better understanding of why $L=1-e^{-U}$.

**So what does mutation load theory predicts?**

The above calculation means that segregating deleterious allele reduce the fitness of the average woman by 89%! If we assume that an average woman can have 10 babies (this number is just a random guess) if she wants to, then the maximum biological limit of a "perfect woman" (in a world where there is no more competition than the one in which we are living) is $\frac{10}{1-0.89}≈91$  babies!

According to [wikipedia][2], the world record is 69.


  [1]: https://www.youtube.com/watch?v=jctvV5xrWw4&spfreload=10
  [2]: http://en.wikipedia.org/wiki/List_of_people_with_the_most_children