Are you asking for the probability of the type TWO error, when the mean is 50.5?

That doesn't make much sense either since 50.5 is not in the region defined by your alternative hypothesis.

The probability of the type one error is alpha, which you say is .025.