| Author: | Dai, Mingkun |
| Title: | Supply-based cooling distribution management of air-conditioning systems for demand limiting and building-grid interaction |
| Advisors: | Wang, Shengwei (BEEE) |
| Degree: | Ph.D. |
| Year: | 2025 |
| Subject: | Air conditioning Buildings -- Energy conservation Hong Kong Polytechnic University -- Dissertations |
| Department: | Department of Building Environment and Energy Engineering |
| Pages: | 1 volume (various pagings) : color illustrations |
| Language: | English |
| Abstract: | Energy and environmental issues are critical concerns that have attracted great attention in recent decades. The building energy consumption plays a significant role in the broader context of these issues, given the increasing demand for commercial spaces and the need for sustainable development alongside urbanization and industrialization. Air-conditioning systems account for a substantial portion of a building's energy usage, making their efficient control vital for overall energy performance. Proper control strategies have the potential to unlock significant energy savings and provide energy flexibility services to power grids. Conventional process control utilized in building central air-conditioning systems can be viewed as demand-based feedback control. In this situation, the control of cooling distribution in the air-conditioning system depends on the individual cooling demands of each air-conditioned space. However, demand-based control fails to effectively manage the cooling distribution when cooling supply is limited. The limitations of conventional demand-based control become evident in the following scenarios: First, air-conditioning systems in commercial buildings are usually switched on in advance to precool the indoor spaces to create an acceptable working environment by office hours. However, the central cooling systems often cannot provide enough cooling supply capacity due to high demand during the morning start period, especially in hot seasons. In this situation, imbalanced cooling distribution often results in significant differences of cooling-down speed among different building zones, requiring an extension of precooling time and leading to considerable energy waste. Moreover, air-conditioning systems have great potential to provide energy flexibility services to the power grids of high-renewable penetration. Direct load control, by switching off some operating chillers, is the simplest and most effective means for air-conditioning systems in buildings to respond to urgent power reduction requests from power grids. However, the implementation this approach in today's buildings, which widely adopt demand-based feedback controls, could lead to serious issues, including disordered cooling distribution and additionally energy consumption. Therefore, this PhD study aims to theoretically and practically develop smart cooling distribution management strategies of air-conditioning systems, focusing on demand limiting and building grid-interaction. To begin with, the concept of supply-based control is proposed as an effective approach for cooling distribution when the cooling supply is limited. To implement this approach in today's buildings, a reconfigurable supply-based feedback control is proposed. This system integrates supply-based feedback control for demand limiting control under limited cooling supply and demand-based feedback control during normal operation with sufficient cooling supply. In particular, this strategy can be conveniently deployed in today's conventional digital controllers. The proposed strategy incorporates a control loop reconfiguration scheme and a setpoint reset scheme, facilitating effective demand limiting control and enabling smooth transitions between the two control modes. The control loop reconfiguration scheme reconnects the controlled variable and resets the control parameters when switching from one mode to another while determining the proper timing for this transition. The commonly used PID control function is adopted. The setpoint reset scheme establishes the setpoint of the feedback loop in demand limiting mode. The proposed control strategy is implemented in a commonly used digital controller to conduct hardware-in-the-loop control tests on an air-conditioning system involving six air handling units (AHUs). Test results show that the reconfigurable control achieves commendable control performance. Proper chilled water distribution enables even thermal comfort control among building zones during demand response and rebound periods. Temperature deviation among building zones is maintained below 0.2 K most of the time. Power demand reductions of 11.6% and 27% are achieved during demand response and rebound periods, respectively, when using the proposed reconfigurable control compared to conventional controls. Advanced control strategies for specific enhancement are proposed in the following sections. Specifically, to address the problems and challenges during the morning start period, an iterative learning control strategy for building air-conditioning systems under limited cooling supply is introduced. This simple control strategy can determine the AHU water valve openings during the morning start period to achieve uniform cooling among building zones by updating the valve opening control values of individual AHUs using iterative learning control. A reinforcement learning approach is developed for setting the control parameters by adopting a classical reinforcement learning method, namely Q-learning. The proposed control strategy is model-free and does not require extra sensors or additional experimental work for thermodynamic characteristic parameter identification. Validation tests are conducted, and results show that the proposed control strategy could reduce the daily precooling time by up to 12.1% during typical days in Hong Kong by achieving uniform cooling. Daily energy consumption could be reduced between 5.1% and 17.8% by shortening the morning start period, corresponding to a weekly electrical energy savings of between 1,376 kWh and 2,916kWh in the test building. To address the problems and challenges of building grid-interaction, an event-driven control strategy is proposed to effectively unlock building energy flexibility for fast demand response using air-conditioning systems. The proposed control strategy determines the optimal AHU water valve openings based on real-time indoor environment data from different air-conditioned zones for even distribution of the limited cooling supply after shutting down part of the operating chillers during the demand response period. A cooling distribution control scheme is proposed and used in the control strategy for even cooling distribution. An event-driven scheme is introduced into the cooling distribution control for the first time to minimize adjustments of the valve openings. This event-driven scheme could avoid frequent adjustments of the AHU valves, reducing unnecessary wear and tear during the control process. Validation tests demonstrate that the limited cooling supply can be distributed properly and that the same indoor air temperature profiles can be achieved eventually among the indoor spaces. The power demand of the chiller plant is reduced by 170 kW (5%) with the proposed event-driven control while maintaining the same comfort levels as existing time-driven control. The average accumulated valve travel distance is also reduced by 54.6%, significantly decreasing the wear and tear of the AHU valves. To improve control generality and scalability, a distributed cooperative control strategy for air-conditioning systems based on the multi-agent system is proposed to facilitate fast demand response. The control architecture is deployed on field control stations of corresponding terminals (i.e., valves and dampers) based on environmental variable measurements of individual air-conditioned spaces. The multi-agent system comprises agents that serve as local controllers for their respective terminals, working collaboratively to achieve even cooling distribution. Each agent performs on-site control using information collected from its own terminal and its neighbors through a distributed architecture. Validation tests demonstrate that the proposed control approach can efficiently manage uneven temperature increases in different zones of the building. During the demand response event, a significant reduction of 2,562 kWh of electricity is achieved, accounting for 19.7% of the electricity consumption compared to the conventional control. Finally, to further facilitate the integration of the reconfigurable feedback control in existing building automation systems, strategies implementing reconfigurable feedback control are proposed for supply-based cooling management throughout the entire daily cycle of the building daily, including demand limiting, morning start and soft stop. The implementation involves a detailed control strategy along with corresponding hardware placement. Hardware-in-the-loop control tests are conducted to validate the deployment plan. Test results indicate that the reconfigurable supply-based feedback control method can be conveniently deployed in today's practical building automation systems. Significant energy savings are obtained during the morning start period (i.e., 9.1 %) and soft stop period (i.e., 13.3 %). Besides, power limiting can be further reduced by an additional 30.8 % during the demand limiting period. In conclusion, the development and implementation of the control strategies and real application plans in this PhD study can theoretically and practically provide guidance for demand limiting and building-grid interaction utilizing air-conditioning systems. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/13591