Occupant-Centric Modeling and Control for Low-Carbon and Resilient Communities

Wang, Jing

Graduate Thesis Or Dissertation

Occupant-Centric Modeling and Control for Low-Carbon and Resilient Communities Público Deposited

Analytics

Baixar PDF

Citations

Citeable URL: https://scholar.colorado.edu/concern/graduate_thesis_or_dissertations/0p096813c

Abstract

Global climate change and resulting frequent extreme weather events have highlighted the significance of energy sustainability and resilience. Communities, which refer to a group of buildings located geographically together, are important units for energy generation and consumption. Hence, the research of community energy sustainability and resilience has drawn much attention during the past decades. However, there remain many challenges surrounding community energy modeling and control to achieve the low-carbon and resilient goals.

First, few tools are readily available for community-scale dynamic modeling and control-based studies. To address this gap, a community emulator was developed, which was designed to be hierarchical, scalable, and suitable for various applications. Data-driven stochastic building occupancy prediction was integrated into the emulator using logistic regression methods. Based on this work, we publicly released a library for net-zero energy community modeling using the object-oriented equation-based modeling language Modelica.

Second, building load control informed by real-time carbon emission signals is underdeveloped as utility price-driven control has so far been dominant. To better facilitate community energy sustainability through decarbonization, we proposed four rule-based carbon emission responsive building control algorithms to reduce the annual carbon emissions through thermostatically controllable loads. The impact of carbon net-metering, as well as the evolvement of the future energy generation mix, is analyzed on top of both momentary and predictive rules. Based on the simulation results, the average annual household carbon emissions are decreased by 6.0% to 20.5% compared to the baseline. The average annual energy consumption is increased by less than 6.7% due to more clean hours over the year. The annual energy cost change lies between -4.1% and 3.4% on top of the baseline.

Third, the enhancement of community resilience in an islanded mode through optimal operation strategies is often faced with computational challenges given the large number of controllable loads. To tackle this, we proposed a two-layer model predictive control-based resource allocation and load scheduling framework for community resilience enhancement. Within this framework, the community operator layer optimally allocates the available PV generation to each building, while the building agent layer optimally schedules controllable loads to minimize the unserved load ratio while maintaining thermal comfort. We found that the allocation process is mostly constrained by the building load flexibility. More specifically, buildings with less load flexibility tend to be allocated more PV generation than other buildings. Further, we identified the competitive relationship between the objectives of minimizing unserved load ratio and maximize comfort. Therefore, it is necessary for the building agent to have multi-objective optimization.

Finally, to account for the uncertainties of occupant behavior and its impact on resilient community load scheduling, we developed a preference-aware scheduler for resilient communities. Stochastic occupant thermostat-changing behavior models were introduced into the deterministic load scheduling framework as a source of uncertainty. KRIs such as the unserved load ratio, the required battery size, and the unmet thermal preference hours were adopted to quantify the impacts. Uncertainties from occupants’ thermal preferences and their impact on load scheduling are then studied and addressed through chance constraints. Generally, the proposed controller performs better in terms of the unmet thermal preference hours and the battery sizes compared to the deterministic controller.

Creator