Optimal Energy Management for SmartGrids Considering Thermal Load and Dynamic Pricing

More active participation of the demand side and efficient integration of distributed energy resources (DERs) such as electric vehicles (trVs), energy storage (ES), and renewable energy sources (RESs) into the existing power systems are important design objectives of the future smart grid. In general, effective demand side management (DSM) would benefit both system operators (e.g., peak demand reduction) and electricity customers (e.g., cost saving). For building and home energy scheduling design, heating, ventilation, and air-conditioning (HVAC) systems play a very important role since HVAC power consumption is very significant and the HVAC load can be scheduled flexibly while still maintaining user comfort requirements. This thesis focuses on energy scheduling design for two different application scenarios where HVAC and various DERs are considered to optimize the benefits electric users. The first part of the thesis studies the joint scheduling optimization of EVs and HVACs, which aims to minimize the total electricity cost considering user comfort requirements. The proposed design exploits EVs as a dynamic storage facility where the energy stored in each EV can be used to charge other EVs (EV2EV) or to supply to HVAC systems (EV2HVAC) during the high-priced periods. Various system and design parameters such as user temperature comfort preference, household occupancy and EV travel patterns as well as detailed modeling of building thermal dynamics are captured in the proposed model. Under our design, optimal power consumption profiles of HVACs and optimal charging/discharging profiles of EVs can be obtained by solving a simple linear programming (LP) problem. Numerical studies show that the HVAC systems tend to consume more energy during off-peak hours to precool (preheat) buildings in summer (winter) and consume less energy during the on-peak periods while still maintaining the indoor temperature within the predefined comfort range thanks to building thermal inertia. Furthermore, ollr design enables a subset of EVs to be discharged to supply electricity to the HVAC systems and other EVs during on-peak hours. This is confirmed to result in significant cost saving, allow more flexibility in setting the tradeoff between cost and user comfort, and reduce energy demand during on-peak hours. The second part of the thesis investigates an optimal power scheduling and bidding problem for a community-scale microgrid (MG) under the day-ahead (DA) pricing. The considered MG consists of RESs (e.g., wind turbines, solar panels), conventional generating units (e.g., fuel cells, microturbines), a number of buildings with their associated loads, and an optional battery storage facility. The proposed optimization framework aims to balance between maximizing the expected beneflt of the MG and minimizing the MG operation cost considering user thermal comfort requirements and other system constraints. The underlying problem is formulated as a two-stage stochastic program where first-stage decisions include commitment statues of all conventional units and hourly bid quantities that the MG aggregator submits to the DA market while the second-stage decisions comprise porû/er dispatch decisions, actual power exchange between the MG and the main grid, battery charging/discharging decisions, and amount of involuntary load curtailment and renewable energy curtailment. The thermal dynamic characteristics of buildings is exploited to compensate for the variability of renewable energy generation. Numerical results show that integrating flexible HVAC load scheduling into energy management framework of the MG aggregator can indeed increase significantly the MG profit and reduce the amount of renewable energy curtailment, which also helps mitigate the high energy imbalance charge caused by bid deviation.