Total degrees of freedom in a two-replicate $2^3$ design are not simply twice the treatment degrees of freedom

Under an unconfounded $2^3$ factorial experiment with factors, A, B, C in two replicates arranged in a randomized block design which of the following statements is not correct?