Experimental Research And Model Study On Solids Circulation Rate In A Double Fluidized Bed

Double fluidized bed gasification technology can achieve heat, electricity, gas multiple co-generation, Near-Zero Emission of CO₂ and high carbon conversion, and also can realize clean, highly effective use of biomass energy. The key to keep double fluidized bed running steadily for the use of gasifying biomass is to make sure the control of solids circulation rate is appropriate. The solids circulation rate reflects the running state of double fluidized bed to some extent, the higher the solids circulation rate, the fewer temperature difference between the gasification and combustion zones, and the gasification zone can keep high temperature to provide calorie for gasification. Moreover, higher solid flux between the two fluidized beds conveys more char from the gasification to the combustion zone, which reduces the required amount of additional fuel. At present, it is difficult to monitor the solids circulation rate online, so a model for predicting the solids circulation rate is important for the stability of operation and detailed engineering purposes in the double fluidized bed, and it is important for double fluidized bed design.This paper built a double fluidized bed cold model and made a systematic test on the cold bench. The experimental result indicates the effect of several factors concerning gas velocity in the bubbling fluidized bed and riser, static bed height and particle size on the solids circulation rate; Based on the pressure balance and material volume conservation of double fluidized bed, a fluid dynamics model was established, through the comparison of predicting and experimental value, this work shows that the predicting error is obvious when the solids circulation rate was less then 2kgm^-2s^-1, while almost all predicted value's relative error from the tested is within±20% when the solids circulation rate is more than 2kgm^-2s^-1, and the forecasting value curve match with the tested good; With the help of the software of Matlab, this paper established BP neural network with momentum added which gives an efficient simulation in the solids circulation rate, When the number of hidden layer and nodes were respectively set to 1 and 15, the training time is shorter and the stability of the net work is good, and the whole-performance of neural network is the best, the results show that the relative error of testing samples is also within±20% and the Mean Diversion Extent is no more than 8%. It proves that the two models have a better forecasting ability.