Answer:
The probability that the service desk will have at least 100 customers with returns or exchanges on a randomly selected day is P=0.78.
Explanation:
With the weekly average we can estimate the daily average for customers, assuming 7 days a week:

We can model this situation with a Poisson distribution, with parameter λ=108. But because the number of events is large, we use the normal aproximation:

Then we can calculate the z value for x=100:

Now we calculate the probability of x>100 as:

The probability that the service desk will have at least 100 customers with returns or exchanges on a randomly selected day is P=0.78.