The correct answer is 3
Replications are several experimental trials with the same definitions of factors (levels). Replications are subject to the same sources of variability, independently of each other. You can replicate combinations of factor levels, groups of combinations of factor levels, or complete experiments.
For example, if you have three factors with two levels each and test all combinations of factor levels (full factorial design), a replication of the entire experiment would have 8 runs (23). You can choose to run the experiment once or with multiple replications.
The design of an experiment includes a step to determine the number of replications that must be performed.