نمذجة وأداء التحكم بالبيانات السيئة لمرجل ذي اسطوانة في محطات الطاقة الحرارية

Modeling and Performance of Bad Data Control of Drum Type Boiler in Thermal Power Stations

Abstract

Thermal power stations are the main source of electrical energy by converting the chemical energy in fuels through many processes. The boiler is one of the main parts of power stations and its environment is harsh and critical. So that the response of the station according to the load demands depends upon the response of the boiler. Due to the harsh environment of the boiler, bad data may be happened due to different causes. The presence of Bad data in the control system of power station is one of the most important challenges because it affects the continuity of operation of power stations in an appropriate way. Therefore, bad data should be processed to ensure the operation of power stations to guarantee the uninterrupted production of energy. This requires an estimator that has the ability of detecting, identifying the presence of Bad data and replace the bad data by an estimated data.
The estimator is built according to a mathematical model applied to each section of the boiler, as well as following the weighted least squares approach (WLS), as a statistical technique, to estimate an alternative value to any bad measured reading. The mathematical relations based on basic principle of heat transfer and fluid flow.
The estimator depends on the J-index value, four tests are applied to identify bad data present in the measurements taken from the metering system. In this study, a static state estimator is applied to Al-Dura power station in Baghdad. According to the different tests performed on the estimator, it is found that the present estimator can treat 10 bad data in the measurement system simultaneously. The minimum error ratio that can be detected as bad data by the estimator was 4% and the redundancy of the estimator related to the boiler of Al-Dura thermal power station was 2.125.
The control systems of thermal power stations require a suitable mathematical relation to describe the dynamic behavior of the boiler because the response of the power station to the variations in load demands depends on the response of the plant boiler. The dynamic response of some of the variables of the boiler is tested by using abrupt changes for the pressure, fuel flow-rate, heat flux inside the furnace, mass flow rate of steam, mass flow-rate of water. The required time (settling time) to reach the pressure and mass flowrate of steam was about 4 minutes to reach the targets (steady state) for both increasing or decreasing these variables which depend on increasing or decreasing load demand. For the heat flux, the settling time was about 4 minutes for the increasing case and about 5 minutes for decreasing case to reach the maximum or minimum targets.