Aiming at practical implementation of exchanging electricity between a distributed smart grid and a smart house, this paper proposes a new optimization technique of power management which takes battery degradation into account. Although the battery degradation is a very important issue of the power management, no conventional approach has considered it. First, we define the battery degradation model by making the relationship between the degradation and the battery status, i.e., SOC, charging speed, and temperature clear. Next, a simulator to calculate the total expense of the smart house is proposed. By using the simulator in every case of design space, an optimization flow for expense minimization is executed. Furthermore, it explores the optimization points by changing the battery and PV sizes, and has found the solution which expense is 50% less than the worst solution. Moreover, exploration of the optimum solution for future is shown, under assumption that the battery cost would be decreased.