The provision of energy, in particular the availability of electrical energy, is critical for societies worldwide to function. The demand for energy globally is ever increasing and thus every effort should be made to make the supply of energy greener and more sustainable. Smart grid technology has an essential role to play in this respect, for example, by enabling more efficient grid control and by better integrating renewable energy sources. The high share of the injected power by renewable energy require alternative control mechanisms for the power grid. Smart microgrids are a promising concept for the structuring of the power grid of the future. Grid simulation allows to explore the opportunities and drawbacks of microgrids as well as examining the hypothetical grid topology and infrastructure. Simulation allows potentially dangerous scenarios to be tested without putting costly hardware at risk. Although simulation of transmission networks, particularly AC power flow analysis, has been widely investigated, the simulation of distribution networks and microgrids remains at an early stage. The first part of this dissertation contains a general model for smart grids, with respect to the flow networks of electrical energy, information, and payment. The flow network model serves as a reference model for the subsequent evaluation of current open-source software for power system simulation. The results for four simulators, used to simulate the IEEE 14-bus test case, will be presented. It will become apparent that established software for power grid simulation lacks the capacity to simulate renewable energy sources, as well as the residential demand for individual households. The second part of this body of work is concerned with RAPSim (Renewable Alternative Power System Simulation), one recently developed software tool for smart microgrids, which supports models for renewable energy generation. RAPSim has an easy-to-use graphical user interface and offers opportunities for extensions, but also for user-specific modifications, which will also be described in this work. RAPSim's capabilities are demonstrated using two test cases, including the IEEE 14-bus test case. Finally, areas for further development are outlined and the case is made for RAPSim's suitability for smart grid simulation. Load disaggregation or non-intrusive load monitoring are efficient ways to identify which devices are operating in a given set of devices. Techniques of this kind can allow small-scale users, such as individual households, to play an active role in the smart grid. Data for both real-world and simulated consumption profiles are used to assess the performance of load disaggregation, where certain power consumption cases present particular difficulties. Recently, there have been attempts to quantify the difficulties presented by different consumption profiles, without the use of load disaggregation. In the third part of the dissertation, such a measure for consumption profiles is proposed. The measure, named as the proficiency of power values, is based upon entropy values and is motivated by information theory. The relevance of proficiency will be highlighted by comparing eighteen real-world device sets, which have been measured on different measurement campaigns of household power consumption.