Central Limit Theorem Problem

I believe I am struggling with a CLT problem, but regardless, what do I do with this question?

Using historical data, the success of seed germination rate for a certain variety of lettuce is 31%. A nursery has recently sowed 10,000 of these seeds, and then shortly after received orders for lettuce plants that can only be filled from successful germinations from this sowing.

1) If the order is for 3,000 lettuce plants, what is the probability that they can fill all of these orders?
2) Redo Part (a) for an order of 3,300 lettuce plants.
3) Failing to deliver on orders is something that cannot be 100% avoided in this line of work. However,
one can decide on a maximum acceptable level of risk of this failure occurring. If this particular
nursery is willing to accept a risk of such failure – ie, of not having enough germinated plants from a
sowing to meet all orders – of up to but not exceeding 5%, then what is the largest total amount of
plants it should take orders for, for each sowing of 10,000 of these seeds?

CLT looks about like the following: If the actual average (mean) success rate is u out of the population of all sowings then, if we sample this average [sow a bunch of 10K seeds and measure the average success rate], the sample average will be normally distributed with a mean of u and standard deviation of \(\displaystyle \frac{\sigma}{\sqrt{n}}\) where n is the sample size [the 10K]. If the variance of the underlying distribution is unknown, the sample variance can be used as an estimate.

We should expect the true average is close enough to 31% that we can use the 0.31 * 10000 = 3100 for the mean. However, as given, we have no idea of what the standard deviation is for success of sowing nor what the actual underlying distribution is of the success rate. At least, this is the way I read the question.

So, the best I can come up with for answers is
(1) Probably pretty good.
(2) mmm, maybe not so good.
(3) Well we could assume a normal distribution and variance but our answer would depend on that, so I really don't know.