Loading...
Nexus of water and energy: resilience through decentralized operation
Joshi, Govind
Joshi, Govind
Citations
Altmetric:
Advisor
Editor
Date
Date Issued
2022
Date Submitted
Collections
Research Projects
Organizational Units
Journal Issue
Embargo Expires
Abstract
Natural disaster events are increasing around the globe in frequency, duration, and severity, thanks to the changing climate. Such incidents, e.g., severe weather events such as hurricanes, tornados, and flooding events, or geological phenomena such as tsunamis and earthquakes, can potentially cause damages to critical infrastructure including power and water networks. Not surprisingly, socially vulnerable groups are disproportionately affected by outages in power and water. The focus of this research is on proposing strategies for reliable and resilient operation of power and water networks, while considering their interdependencies. The ultimate goal here is to enhance the resilience of communities and neighborhoods affected by natural disasters by enabling them to continue to operate their energy and water networks in a decentralized fashion, upon need. Considering the nexus of water and energy offers an opportunity to improve the overall performance of both systems and to provide added reliability and resilience in the wake of large-scale disturbances. In this thesis, a framework is proposed for optimal design and optimal operation of community energy and water systems by utilizing, as much as possible, local energy and water resources. This is achieved in multiple steps as outlined below.
First, an optimal control strategy is presented for a decentralized municipal wastewater treatment plant (WWTP) in order to reduce its overall energy footprint, which will make it more suitable for standalone applications. The system studied uses a sequencing-batch membrane bioreactor (SBMBR). Using the proposed control strategy, the electromechanical components in the SBMBR plant can be controlled in a way that it minimizes the energy consumption of the entire system, while achieving the same treatment load. To model the energy demand of the WWTP, a mixed-integer nonlinear programming (MINLP) model is developed for the Mines Park wastewater treatment facility located in Golden, CO. This formulation focuses on the operation schedules and speeds of air blowers, which affect the concentration of dissolved oxygen (DO), and the air scour blower (ASB) speed and permeation rate, which impact the transmembrane pressure (TMP) of the membranes. The MINLP model is solved to identify an optimal aeration and permeation control, which would minimize the energy footprint of the SBMBR system.
Next, it is mathematically demonstrated how a self-contained community can schedule available energy and water resources optimally so that both demands are met with minimal external support. It is assumed that distributed energy resources (DER) such as PV, wind, and a battery system, along with demand-responsive loads are employed to meet energy demands, while a WWTP built in the community is used to treat and recycle wastewater for potable reuse. The problem is modeled as a multi-objective MINLP optimization model. Simulation results demonstrate that communities can successfully coordinate the locally available water and energy resources to meet both water and energy demands during a disturbance, for instance when the primary source of water or energy is not available. However, due to the uncertainty in both renewable energy generation and electric/water consumption, ensuring a sustainable operation is a challenging task. To address this, a stochastic optimization approach is adopted that allows for modeling the uncertainties in generation and demand. An energy storage system in conjunction with local water storage and wastewater treatment offer the needed flexibility to counteract the variations in wind, solar, and demand. A two-stage stochastic programming model is hence formulated and solved to determine an optimal operation strategy for the combined system.
While the preceding work focuses on operation, a follow-up chapter addresses the problem from a design perspective. A multi-objective mixed integer nonlinear programming (MMINLP) model has been presented to determine the optimal sizes of energy resources such as wind turbines, PV system, and the battery system, as well as the optimal size of the community storage tank (ST) to be able to supply the demand. The electrical power demand of residential customers and power demand of the water distribution system need to be supplied by the microgrid system, whereas, the ST will store clean water either purchased from external sources or obtained from water treatment facilities to supply the water needs of the community.
The proposed methodologies mentioned above can be implemented for rural communities where the centralized municipal water system, centralized wastewater facilities, and/or regional electric grid are not available due to geographical or cost limitations. This idea can also be applicable for small-scale communities whose water and energy networks are temporarily out-of-service as a result of natural or manmade disturbances. Should this happen, the main power system and/or the water network need to be broken into multiple subsystems for standalone and decentralized operation. This is the focus of the last section of this thesis where a solution is proposed to divide a power distribution grid and water distribution network into multiple microgrids and micronets to allow for a localized supply of energy and water. In addition to allowing for continuous supply of power and water, such an approach will also help expedite power restoration efforts as the smaller portions of the network would be self-reliant until the service to the main power grid is restored. A graph theoretic approach is proposed, which is able to weigh nodes and edges based on their individual criticality levels as well as their overall impact on the integrity of the system.
Associated Publications
Rights
Copyright of the original work is retained by the author.