Location of ambulances is an important decision in the planning of emergency medical services, which enables responding to patients within short response time, thus, reducing morbidities and saving lives. We consider an ambulance location problem that takes into account the variation in demand across the hours of day and also accounts for server level ambulance availability. The proposed problem considers a combination of performances measures including coverage and survival function for different patient types. A mixed integer programming (MIP) model is presented to model the problem. A solution approach based on the genetic algorithm that uses a mixed chromosome representation is developed. The proposed approach is applied on a realistic dataset generated based on an urban location in India. The effectiveness of the proposed solution approach was validated through comparison with results from commercial optimization software (CPLEX).