A new analysis model for the comparison of control systems for smart grids has been created. The model has been demonstrated by performing a comparative analysis on simulations of two independently developed control systems: IntelliGator and Triana. Both control systems operate in a hierarchical structure, but have a different architecture. The IntelliGator system is based on real-time control using a Walrasian market system. The Triana system uses global planning through Iterative Dynamic Programming and decentralised real-time control using Model Predictive Control. A microgrid case study (Flex Street) has been developed that incorporates renewable and controllable technologies with various penetration rates in a Dutch residential sector. Simulations of the control systems have been performed on Flex Street, given a peak-shaving objective. They have been assessed on relative peak reduction, energy efficiency and savings and abated emissions due to a reduction in transport losses. Real-time control using a market-system without forecasting and planning has been able to achieve a high share of the potential benefits. It has accomplished a relative peak reduction of 32 to 67%, depending on the amount of controllability offered in each scenario. Global planning in combination with decentralised control has been able to achieve a relative peak reduction of 32 to 71%, while efficiently reserving storage capacity. Finally, results from the simulations have been used to demonstrate some implications of smart grid implementation. Results show that the annual cost reduction on the consumer side is relatively small under the current tariff structure in the Netherlands. Savings occur mostly on the side of the grid operator, due to the cutback on grid capacity. A reduction in the required grid capacity of 19 to 54% can be achieved. The analysis model provides a flexible tool for comparative analysis of smart grid control systems, and promises to help developers, policy makers and consumers in their decision-making.