|
" A Scenario Approach for Operational Planning with Deep Renewables in Power Systems "
Modarresi, Mohammad Sadegh
Xie, Le
| Document Type
|
:
|
Latin Dissertation
|
| Language of Document
|
:
|
English
|
| Record Number
|
:
|
1053264
|
| Doc. No
|
:
|
TL52381
|
| Main Entry
|
:
|
Modarresi, Mohammad Sadegh
|
| Title & Author
|
:
|
A Scenario Approach for Operational Planning with Deep Renewables in Power Systems\ Modarresi, Mohammad SadeghXie, Le
|
| College
|
:
|
Texas A&M University
|
| Date
|
:
|
2019
|
| Degree
|
:
|
Ph.D.
|
| student score
|
:
|
2019
|
| Note
|
:
|
100 p.
|
| Abstract
|
:
|
This work is both enabled by and motivated by the development of new resources and technologies into the power system market operation practice. On one hand, penetration level of uncertain generation resources is constantly increasing and on the other hand, retirement of some of the conventional energy resources like coal power plants makes market operations an attractive topic for both theoretical and state-of-the-art research. In addition, as generation uncertainty increases, it impacts the true cost of energy and causes it to be volatile and on average higher. This work targets flexibility enhancement to the grid to potentially eliminate the impact of uncertainty. Two different viewpoints in two different markets for electricity is targeted. This dissertation looks at the real-time market generation adequacy from the Independent System Operator’s point of view, and the day-ahead scheduling of energy and reserve procurement from the market participant’s point of view. At the real time scale, the emphasis is on developing fast and reliable optimization techniques in solving look-ahead security constrained economic dispatch. The idea is when forecast accuracy gets sharper closer to the real-time and slower power plants retiring in recent years, market participants will spend more and more attention to the real-time market in comparison to the day ahead operation in terms of the energy market. To address it, a data-driven model with rigorous bounds on the risk is proposed. In particular, we formulate the Look-Ahead Security Constrained Economic Dispatch (LAED) problem using the scenario approach techniques. This approach takes historical sample data as input and guarantees a tunable probability of violating the constraints according to the input data size. Scalability of the approach to real power systems was tested on a 2000 bus synthetic grid. The performance of the solution was compared against state-of-the-art deterministic approach as well as a robust approach. Although the real-time market is primarily for energy trading, the day-ahead market is the market for ancillary service trading. In this dissertation, at the day-ahead scale, the focus is on providing ancillary service to the grid by controlling the consumption of millions of privately owned ii pool pumps in the US, while benefiting from energy arbitrage. A conceptual framework, a capacity assessment method, and an operational planning formulation to aggregate flexible loads such as inground swimming pool pumps for a reliable provision of spinning reserve is introduced. Enabled by the Internet of Things (IoT) technologies, many household loads offer tremendous opportunities for aggregated demand response at wholesale level markets. The spinning reserve market is one that fits well in the context of swimming pool pumps in many regions of the U.S. and around the world (e.g. Texas, California, Florida). This work offers rigorous treatment of the collective reliability of many pool pumps as firm generation capacity. Based on the reliability assessment, optimal scheduling of pool pumps is formulated and solved using the deterministic approach and the scenario approach. The case study is performed using empirical data from Electric Reliability Council of Texas (ERCOT). Cost-benefit analysis based on a city suggests the potential business viability of the proposed framework.
|
| Descriptor
|
:
|
Electrical engineering
|
| Added Entry
|
:
|
Xie, Le
|
| Added Entry
|
:
|
Texas Aamp;M University
|
| |