Abstract:This study establishes detailed neural network modeling procedures for the combustion optimization of a power generation unit, equipped with a Selective Catalytic Reduction (SCR) system. In particular, artificial intelligence neural networks were applied due to its capability in treating large pools of data of great nonlinearity. The objective of this study was to reduce nitrogen oxide (NOx) emissions in the stack gas, lower the cost of emissions compliance and minimize the net unit heat rate at this power generation unit at full-load conditions. These procedures include building a database with real coal power plant operating data, modeling the database and using the genetic algorithm to optimize the operating conditions. The operation is optimized with respect to the steam temperature, selective catalyst reactor, and separated over-fire air conditions. Verifications of the results demonstrated that by carrying out the reported study, the pollutant concentrations of the boiler are significantly decreased and the catalyst layer is effectively preserved, while the operation can be maintained stable near the full-load condition. It was found that the average ammonia (NH3) flow to the SCR under optimized conditions was 229 kg/h, while the average NH3 flow under non-optimal operational conditions was 260 kg/h, resulting in a reduction in NH3 consumption of approximately 12 % from the non-optimized operational condition. In the meantime, the SCR inlet (boiler exit) NOx was reduced from 523 mg/Nm3 (non-optimized operational condition) to 444 mg/Nm3 (optimized operational condition), which represent about 15 % reduction. Based on the experience with SCR from manufactures and operators, the SCR catalyst life expectancy has a direct relationship with the SCR inlet NOx concentration and NH3 injection rate. The lower the NOx concentration at the SCR inlet and the less the NH3 injection rate, the longer of the SCR catalyst life will be. Thus, it is expected that the SCR catalyst layer replacement frequency could be reduced by 10 to 20 % over the normal catalyst replacement cycle, while meeting the same environmental NOx emissions limit. Additionally, a Delta Heat Rate was also calculated using best available data from the optimized operational conditions, including boiler load, boiler excess oxygen (O2), fly ash unburned carbon and attemperation flow. Results were compared with the average Delta Heat Rate calculated from baseline tests. It was found that the average Delta Heat Rate from baseline tests are −60 kJ/kWh, while the delta heat rate from the optimized operation condition hours is −118 kJ/kWh. This represents about 0.57 % improvement on unit heat rate from the baseline average.

Comparative study of machine learning methods integrated with different optimisation algorithms for prediction of thermal performance and emissions in a pellet stove

Combustion Optimization Study of Pulverized Coal Boiler Based on Proximal Policy Optimization Algorithm

Comparison of Random Forest and Neural Network in Modeling the Performance and Emissions of a Natural Gas Spark Ignition Engine

STUDY ON PREDICTION MODEL FOR TEMPERATURE OF FLUE GAS EXHAUSTED FROM UTILITY BOILERS BASED ON PROPERTIES OF FUEL COAL

Comparative analysis and prediction study for effluent gas emissions in a coal-fired thermal power plant using artificial intelligence and statistical tools

An Artificial Neural Network Model to Predict Efficiency and Emissions of a Gasoline Engine

Machine learning assisted prediction of exhaust gas temperature of a heavy-duty natural gas spark ignition engine

Building energy consumption prediction and optimization using different neural network-assisted models; comparison of different networks and optimization algorithms

Modeling and Coordinative Optimization of Nox Emission and Efficiency of Utility Boilers with Neural Network

Comparative investigation of the usability of different machine learning algorithms in the analysis of battery thermal performances of electric vehicles

A Numerical and Experimental Performance Assessments of an Improved Natural Draft Biomass Cook Stove

Employing machine learning algorithm for properties of wood ceramics prediction: A case study of ammonia nitrogen adsorption capacity, apparent porosity, surface hardness and burn-off for wood ceramics

Combustion optimization of a coal-fired power plant boiler using artificial intelligence neural networks

Predictive modeling of biomass gasification with machine learning-based regression methods

Prediction of thermo-hydraulic properties of flow in an innovative plate heat exchanger using machine learning algorithms

Environmental Pollution Prediction of NOx by Process Analysis and Predictive Modelling in Natural Gas Turbine Power Plants

Optimization of Circulating Fluidized Bed Boiler Combustion Key Control Parameters Based on Machine Learning

Tabular Machine Learning Methods for Predicting Gas Turbine Emissions

A Hybrid Neural Network‐Based Improved PSO Algorithm for Gas Turbine Emissions Prediction

Prediction of CO‐NOx emissions from a natural gas power plant using proper machine learning models