Search

Electric water heaters (EWHs) are usually equipped with inbuilt thermostats to measure water temperature near the installed positions. Most of the existing EWHs have only two thermostats and often do not have water flow sensor installed. Estimation of state of heat energy (SOHE) inside the hot water tank, which is crucial for efficient control and operation of the EWH, under such imperfect system conditions is very challenging. This paper, therefore, designs a transactive EWH agent (TEWHA) to better estimate SOHE in presence of measurement errors and imperfect system knowledge. The TEWHA, built-in Python, approximates the EWH physics using limited measurements obtained from the ground truth GridLAB-D model. The performance of the TEWHA is validated against ground truth system for three scenarios: a) ideal condition where TEWHA has complete knowledge of the system, b) imperfect condition where TEWHA has incomplete system information, and c) noised condition where TEWHA measurements contain random errors. Finally, an optimization problem is formulated and solved to demonstrate how TEWHA could use the estimated and measured parameters for a transactive control of EWH.

This paper presents an Artificial Neural Network (ANN)-based building-level hourly electrical load forecasting method that takes into account HVAC set points as one of the input parameters, in addition to the historical building load and outdoor weather data. The data presented in this paper deal with cooling load only. ANN is used to train and test the dataset, and the ANN-based load forecasting model provides the predicted load for each hour of the day. Three training algorithms are explored, including Levenberg-Marquardt, Scaled Conjugate gradient back-propagation and Bayesian Regularization (BR). Findings indicate that the BR-based neural network offers the best performance in terms of forecasting accuracy. In addition, a case study using a commercial building in Chicago, Illinois is presented where performances of the developed ANN-based models are compared. The forecasting error is around 5% or less for hour-ahead load forecasting, and around 8% or less for 12-hour ahead load forecasting.

Quantifying energy savings and peak demand reduction potentials in a building due to adjustment of HVAC set point during a demand response (DR) period is critical for active DR program participation and optimal demand-side resource allocations. This has been a topic of great challenge given the complex nature of physical characteristics and HVAC system thermal dynamics in commercial buildings. Based on the building model developed in eQUEST validated against measured power consumption data from a smart meter, this paper investigates how energy savings and peak demand reduction potentials of a building are impacted by HVAC set point changes during a DR period. Simulation results reveal that the days with high outdoor temperature are expected to achieve a predictable percentage of daily energy savings and peak load reductions when HVAC set points are raised in summer. However, the inevitable demand restrike, i.e., increase in building peak demand after a DR event, is a concern.

With the increasing deployment of smart meters and Internet-of-Things technologies, buildings are becoming smarter, and capable of participating in demand response (DR) programs. Heating, Ventilation and Air Conditioning (HVAC) units, as one of the major loads in buildings and due to their cyclic operating characteristics, are excellent candidates for peak load management. A prerequisite to HVAC control for peak load management is to know their power consumption. While many studies assume that HVAC units consume fixed rated power during their operation, field measurements show that this value is not fixed but varies with outdoor temperature. Performing DR using the fixed power consumption model may result in unexpected load reduction, i.e., the reduction may be lower or higher than expected. Although power consumption measurements of individual HVAC units can provide necessary data, it is prohibitively expensive to install a power meter for each HVAC unit. To reduce hardware investments, this paper proposes an algorithm to derive power-temperature models of individual HVAC units from a single power meter that measures the power consumption of all HVAC units. Research findings indicate that the power consumption of individual HVAC units can be precisely modeled and disaggregated from a single power meter data.