I would use a binomial for this question, definitely. As you have a fixed number of trials, and a fixed probability of success/failure
Let X = the number of defective light bulbs in a sample of 50,
than X~B(50, 0.02)
than just use the formula, as you cant use the tables.