Electrical energy is primarily generated in thermal power plants. The boilers used in a thermal power plant is an important part of energy production. Predicting the important factors which leads to energy loss or wastage of the fossil fuel, coal will be majorly important because we could reduce the wastage before boilers are used. Since it will be the analysis of relationship between physical properties of the raw materials, e.g., coal, water, air , a one time analysis’s results won’t change because physical properties of a matter stays constant. A boiler requires coal, heat and air for input and excludes hot flue gas which could become a reliable source of heat, suitable for purposes of drying away moisture from industrial equipments ; the boiler also leaves unburnt carbon as waste, there are no further applications for it yet. Managing boilers would mean less or no wastage of coal. This analysis report is the detailed analysis of Boilers’ Indirect method of calculating efficiency. It provides detailed analysis about unburnt carbon along with all the factors where heat is lost. The applied exploratory analysis on thermal energy production to ensure that coal is burns properly without any wastage, leaving less or no unburnt carbon.
Introduction
Research Objective and Novelty
The study focuses on improving boiler efficiency in thermal power plants, which are major contributors to global electricity. Since these plants heavily depend on coal combustion, optimizing this process is vital for energy conservation. The primary research goal is to analyze unburnt carbon in ash—an indicator of poor combustion—and reduce it toward 0% using Indirect Efficiency Method recommended by the Energy Conservation Department of India.
B. Methodology Overview
Data Collection
Real-time sensor data from the 2023–2024 fiscal year stored in Excel sheets.
Parameters sourced from 25 boiler operation variables.
Tools Used
Python with pandas, numpy, matplotlib, seaborn
Exploratory Data Analysis (EDA) methods: line plots, covariance heatmaps, boxplots
Efficiency Analysis
Focus on boiler heat losses via the Indirect Method
Boiler Efficiency Formula: Efficiency (%) = 100 - Sum of Heat Losses
Special attention to heat loss due to unburnt carbon in ash (bottom, ESP, cyclone, APH)
Data Cleaning & Preparation
Handled missing values using imputation.
Standardized column names and data types.
Outliers removed via boxplots for normalization.
Feature Engineering
Grouped ashes by location to find key contributors to efficiency loss.
Identified Excess Air as a major factor influencing unburnt carbon.
Avoided PCA due to potential data loss.
C. Boiler Components Involved
The system analysis considered components like:
Steam drum, furnace, superheater, air preheater, heat exchanger
Circulator pumps, condenser, safety valve, alternator
These components directly influence combustion quality and heat transfer efficiency.
Ensured data consistency and unit standardization.
Recognized the impact of excess air on unburnt carbon and combustion inefficiency.
Identified that bottom ash is the most representative ash for analysis.
H. Common Mistakes in EDA Noted
Ignoring missing data or data distributions.
Failing to investigate variable correlations.
Inadequate visualization or overreliance on statistics.
Confirmation bias and poor documentation.
Conclusion
Tons of coal is burnt everyday in roughly each state in in each country with people unaware of how to reduce the fossil wastage. This research has bought various new insights towards the topic of reduction of “unburnt ash carbon” which is a proper waste product. Sharing this research, development in thermal plant giants have started , focusing on the outcomes of this research, and have found massive improvement with their unburnt coal wastage, and directly increasing the efficiency.
References
[1] https://beeindia.gov.in/sites/default/files/4Ch1.pdf
[2] Applying VLSI EDA to energy distribution system design, https://ieeexplore.ieee.org/abstract/document/6742872/references#references
[3] Design and testing of the Organic Rankine Cycle, https://www.sciencedirect.com/science/article/abs/pii/S0360544200000633
[4] Data science with Python and Dask
[5] Hands-On Exploratory Data Analysis with Python: Perform EDA techniques to understand, summarize, and investigate your data, https://books.google.com/books?hl=en&lr=&id=QcHZDwAAQBAJ&oi=fnd&pg=PP1&dq=EDA+boxplot&ots=tQRDRh0ndh&sig=H9qD09muWHzXn5bdPulK4T9jU9o