The renewed interest in the deployment of electric vehicles promises enhanced environmental and social compatibility, higher energy efficiency, as well as effective power grid support through the vehicle-to grid energy exchange mode. Focusing on a smaller scale, such as buildings with large parking lots for electric vehicles, the aim of the so-called vehicle-to-building concept is to regulate the power consumption of a building by either throttling the charging rate of electric vehicles or by delivering electricity (discharging) into the building when needed. In this paper, a mathematical model is implemented in MATLAB to peak-shave and valley-fill the power consumption profile of a university building by scheduling the charging/discharging process in an electric vehicle parking lot, using real-world data of power consumption and parking lot occupancy. The simulation of three scenarios with different number of parking spots reveal the feasibility of the proposed approach to effectively flatten the power consumption profile during daytime, which is particularly important for electricity customers that the energy cost depends on the contracted power capacity. In detail, the results indicate that the peak power consumption is reduced by approximately 3% and 20% for the scenarios with the minimum and maximum number of electric vehicle parking spots respectively. (C) 2018 Elsevier Ltd. All rights reserved.