W=Fx doesn't seem right as nothing is moving any distance. While the spring was extending the rocket was putting work in to it equal to 1/2 Kx^2, but I'm not interested in that. I want to know the power required to maintain the equilibrium.