Understanding historical reservoir operation rules, dynamics, and deficiencies - opportunities for improvement

Chen, Yanan

Permalink

https://hdl.handle.net/2142/127493 Copy

Description

Title

Understanding historical reservoir operation rules, dynamics, and deficiencies - opportunities for improvement

Author(s)

Chen, Yanan

Issue Date

2024-12-05

Director of Research (if dissertation) or Advisor (if thesis)

Cai, Ximing

Doctoral Committee Chair(s)

Cai, Ximing

Committee Member(s)

• Stillwell, Ashlynn
• Zhang, Zhenxing
• Zhu, Tingju

Department of Study

Civil & Environmental Eng

Discipline

Civil Engineering

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• data-driven reservoir simulation model
• reservoir operation rules
• reservoir operational changes
• forecast informed reservoir operation

Abstract

Reservoirs play a vital role in surface water management, serving various purposes such as flood control, irrigation, water supply, and hydroelectricity. However, reservoir operation has been complicated by rapidly changing environments, including climate, water demand, and operation police changes. This dissertation aims to enhance understanding of historical reservoir operation rules, dynamics, and deficiencies across the contiguous United States (CONUS) and explore ways to improve operation performance. The research is organized into three interconnected parts. The first part develops a generic data-driven reservoir operation model (GDROM) using machine learning (ML) to extract real-world reservoir operation rules. Tested on 467 reservoirs of varying capacities and functions across the CONUS, the GDROM demonstrates comparable accuracy in release simulation to other ML models, with improved interpretability due to its transparent structure. The second part identifies reservoir storage and operational changes between 1990 and 2019 for 256 reservoirs across the CONUS and analyzes the underlying causes. The GDROM model is used as a tool to identify operational changes, including the cases of both operational effectiveness and deficiencies through analyzing the relations of storage and operational changes under environmental changes. To address the operational deficiencies detected in the second part, the third part proposes a new forecast-informed reservoir operation (FIRO) decision support framework that integrates analytically derived hedging policies with empirically based flood control rules. Within this framework, the GDROM is integrated to capture real-world operation specifications that reflect operators’ actual response to various hydrologic conditions. The FIRO framework is tested on Folsom Lake in California, which demonstrates significant improvement in water conservation benefit in the post flood season without increasing flood risks during the flood season. Leveraging historical operation records and the developed date-driven reservoir operation model, this dissertation provides insights into real-world operational rules, dynamics, and existing deficiencies within the CONUS, offering guidance for operation design in response to future changes. Moreover, this research establishes a new FIRO decision support framework, which introduces flexibility to address operational deficiencies and improve reservoir performance under varying hydrological conditions.

Graduation Semester

2024-12

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/127493

Copyright and License Information

Copyright 2024 Yanan Chen

Owning Collections