Loading...

Data-driven Nexus Analysis and Optimization of a Complex Thermo-gasdynamic Energy System and Implementation on an Old National Thermal Power plant in Operational Conditions

Momeni Masuleh, Ghadir | 2022

105 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 55701 (45)
  4. University: Sharif University of Technology
  5. Department: Aerospace Engineering
  6. Advisor(s): Mazaheri, Karim
  7. Abstract:
  8. The performance of the power plant decreases during its lifetime and deviates from its design and initial operation conditions; Maintenance issues, variety of operational patterns, market limitations and financial goals have been caused this situation. Knowing the appropriate actions and finding the optimal operation conditions of the power plant can support the system to restore its initial operational performance and bring it closer to its design condition. In this research, with historical data helps of a steam thermal power plant in Kermanshah, unsupervised machine learning techniques have been used to identify operational patterns, which lead to the identification of optimal operating conditions based on nexus analysis. The historical data used in this research belongs to a 320 MW unit from Bistoon power plant in Kermanshah which includes 75 operational indicators and 8760 samples. The results have been compared with the values of the design conditions obtained by the power plant design documents and the Thermoflow model. The research methodology includes: using the z-score method to normalize the data, the correlation coefficient method to remove redundancy, the principal component analysis method to reduce the dimensions of the big data, the k-means++ clustering technique to process the data, the methods of Elbow, average Silhouette coefficient and the importance index to evaluate the clustering and to find the optimal number of clusters. The purpose of this research is to present strategic optimal operation plans for a thermal power plant and determining the range of stable changes of all operational indicators in order to achieve the objective functions of maximizing the gross production power and cycle efficiency and minimizing the consumption of makeup water and fuel flow. The results of this research identify and present five strategic optimal exploitation programs to achieve the target functions of high gross production power (4.7% improvement), low fuel consumption in special conditions (18.35% improvement) and normal conditions (4.03% improvement), high thermodynamic efficiency (0.52% improvement) and low makeup water consumption (48.65% improvement) that the average optimal values for the target functions are 1.04, 13.22, 1.09, 2.87 and 29.13% far from the values of design conditions respectively
  9. Keywords:
  10. Optimization ; Big Data ; Data Mining ; Machine Learning ; Clustering ; Thermal Power Plants ; Operating Condition ; Water and Energy Nexus

 Digital Object List

 Bookmark

No TOC