Study On Accuracy Evaluation And Application Of Online Coal Consumption Monitoring System

Under the pressure of energy shortage and environmental pollution,energy-saving generation dispatching is a significant strategic initiative to guide thesustainable development of power industry because it can minimize energyconsumption and pollutant emissions. As the important technical support andreference basis of energy-saving generation dispatching, online coal consumptionmonitoring system has great significance on its scientific and fair implementation.For the problems and controversies in the practical application of the system, thestudy focusing on the turbine heat rate and boiler efficiency was studied toguarantee the accuracy of the system.Online calculation of turbine heat rate involves a large number of temperature,pressure and flow measuring point data. In order to ensure the accuracy of theparameters involved in the calculation, a large number of historical operatingparameters were processed using data mining methods to get the reference valuesfor important parameters and the corresponding tolerance range were given. Thework can enhance the parameter sensitivity analysis and multiple data validationtechniques in the system. Comparative tests show that the calculation program in thesystem is precise; regularity of the calculation model is correct and the accuracy ofthe results is high.Accuracy of the boiler efficiency calculated online is mainly affected by coalcomposition, exhaust gas temperature and oxygen, carbon content of fly ash andslag. For the complexity of coal, online calculation program of boiler efficiencyunder complex coal conditions was proposed. Through the depth analysis of theinherent relationships between elemental compositions of coal，a coal databaseapplicable to the program was established with clustering algorithms and othermathematical methods, achieving online access to coal in the process of boilerefficiency calculation online. Practical results show that: the trend of coal calorificselected from virtual coal database is highly consistent with that tested by powerplants, with the biggest difference being about700kJ/kg.So it can meet the needs ofonline calculation in engineering.For the fluctuations and sampling error of exhaust gas temperature and oxygen,a large number of historical operating parameters were processed using data miningmethods to get the reference values and the corresponding tolerance range weregiven; For the problems caused by anthropogenic inputs offline of carbon content of fly ash and slag, the upper and lower limits was formulated with the distributioncharacteristics of the data after statistical studies of historical data for each unit.Experiments show that the introduction of the uncertainty of these four parametersmake the upper and lower deviation of boiler efficiency is about1%. Uncertaintyanalysis of boiler efficiency improved credibility of the results calculated.Through these studies, the accuracy of online coal consumption monitoringsystem is well guaranteed, which is of great significance for the smoothimplementation of energy-saving generation dispatching. Meanwhile, the accuracyevaluation method has some reference and guidance for the similar systems in otherregions and even the online monitoring system in other areas.