Electricity is distributed throughout the electrical power network in 3-phase voltage. This power reaches households as a single-phase voltage, generally 115vac or 240vac. This is achieved by allocating households with either phases A, B, or C of the final 3-phase power distributed to the street through a low voltage transformer. A present problem confronting the electrical power industry is identification of which particular phase a household is connected to. This information is often not tracked and the mechanisms for identifying phase require either manual intervention or costly signal injection technologies. Phase information is important as it is a foundation for the larger problem of balancing phase loads. Unbalanced phases lead to significant energy losses and sharply reduced asset lifetimes. In this paper we propose a new approach to compute household phase. Our techniques are novel as they are purely based upon a time series of electrical power measurements taken at the household and at the distributing transformer. Our methods involve the use of integer programming and solutions can be retrieved using branch and bound search algorithms implemented by MIP solvers such as CPLEX. Furthermore, as the number of measurements increase, continuous relaxations of integer programs may also be used to retrieve household phase efficiently. Simulation results using a combination of synthetic and real smart meter datasets demonstrate the performance of our techniques and the number of measurements needed to uniquely identify household phase.