Let X be the number of entries with an error. X has the binomial distribution. n, the number of trials is 2 000 000 and the success probability, the probability that an entry has an error is 0.027.

the mean of the binomial is n * p = 2 000 000 * 0.027 = 54000

the variance of the binomial is n * p * (1-p) = 2 000 000 * 0.027 * 0.973 = 52542

the standard deviation is the square root of the variance = 229.22

——

now if you have to do any calculations to find probabilities it would be nice to use the normal approximation to the binomial.

if you have n * p > 10 and n* (1-p) > 10 then the binomial can be approximated by the normal distribution with mean n*p and variance n*p*(1-p)

there is also a continuity correction that you must use with this. Please let me know if you want more info on this topic.