热水循环加热厌氧反应器稳态数值模拟分析

Steady-state Numerical Analysis of Anaerobic Reactor by Hot Water Circulating Heating

摘要:

对于厌氧发酵系统，温度分布是否均匀是影响产气效率的关键因素。通过在10L厌氧反应器中分别对水和TS为12.4％（质量分数）的发酵料液进行加热的热平衡试验，运用Fluent软件对厌氧污泥传热特性进行仿真模拟，通过传感器监测温度变化，得出水的温度分布与高TS料液有很大差异，而且在没有搅拌装置的情况下进行高TS料液的传热仿真，不能将其物性参数默认为水。在对水进行传热仿真时发现，对于低TS料液进行仿真要考虑料液间的对流换热，因此需调用Boussinesq假设，调用该假设后，模拟与试验最大误差仅为4.2%。对换热管壁面厚度设置作出分析并得出：在Fluent中设置虚拟壁面与在Gambit中直接绘制壁面模型具有相同效果，而前者可以简化模型并减少计算时间。此外，在进行高TS仿真时发现最适合这种热水循环加热系统的湍流模型为可实现k—ε模型（Realizable k—ε）。

Abstract:

Temperature distribution has a great influence on the efficiency of gas production for anaerobic fermentation system. Fluent is one of the most advanced software for fluid simulation. It is suitable to apply it in simulation analysis of the anaerobic reactor without stirring. Two kinds of raw materials which were water and digested slurry with TS of 12.4% were applied in two 10L anaerobic reactors to conduct anaerobic fermentation experiments. Totally 16 temperature sensors were installed at each anaerobic reactor in order to detect temperature continuously. The temperature figures indicated that two types of materials were definitely different. Therefore, applying the Fluent heat transfer simulation to digested slurry anaerobic fermentation system cannot only utilize water as physical parameter. However, the error was only 4.2% between experiment with raw material of digested slurry with TS of 12.4% and Fluent simulation, when water and Boussinesq hypothesis were simultaneously defined as physical parameters. It was also explored that setting up a virtual wall in Fluent can dramatically simplify the model calculation time, and it had the same effect with defining real wall model in Gambit. In addition, it was found that realizable k—ε model was the most suited to the anaerobic fermentation system of digested slurry with high TS. The aims of the research were to explore suitable Fluent setting and optimization method for anaerobic fermentation and provide guidance for further work. Optimizing model and improving efficiency in unsteady-state numerical analysis will be the main body of the further work. 