Simulation and analysis of pressure drop in high temperature twin-candle ceramic filter system

The ceramic candle filter presents an interesting alternative to collect aerosol particles smaller than 2.5[micrometer] (MP2.5) because of its high resistance to high temperature, high pressure, chemicals and corrosion. As cake builds up on the outside surface of the ceramic candle filter, pressure drop, one of the important characteristics of filtration, would be increased rapidly. Meanwhile, the collection efficiency and energy consumption also increased in tandem. So, a pulse jet is injected periodically to shake down the accumulated dust cake, thereby reducing the pressure drop in the system. One technique frequently used to indirectly predict cake formation rate applies computational fluid dynamics (CFD) since it offers low cost, low risk and ease of use. However, nearly all previous CFD investigations focused on single filters. In this research, the cake formation rate on twin ceramic candle filters in a prototype system was modeled and its pressure drops validated with experimental data obtained by Hosokawa Powder Technology Research Institute (HPTRI). A suitable CFD model provided by commercial software code FLUENT was selected and employed to simulate the 3D fluid flow inside the twin-candle unit. The calculated flow field was used to estimate the local flow velocities. Due to the reasonable assumption of 100% collection efficiency and submicrometer aerosol size, the particles would move along the gas streamline with uniform distribution. Thus we could predict the differences in cake formation rate on each part or element of the candle filter from the local face velocity and calculate the cake resistance properties. In addition, the CFD technique was used to investigate the effects of such major parameters as filtration velocity, temperature, dust concentration and inlet feed location, to be varied. It was found that the filtration velocity was the most important parameter affecting the pressure drop of the system.

陶瓷烛式过滤器（ceramic candle filter）是收集2.5微米以下气溶胶颗粒物（MP2.5）的极具潜力的替代方案，因其具备优异的耐高温、耐高压、抗化学腐蚀性能。当滤饼在陶瓷烛式过滤器的外表面逐渐沉积时，作为过滤重要特性之一的压降会快速升高。与此同时，过滤收集效率与能耗也会同步上升。因此，通常会定期喷射脉冲气流以抖落堆积的粉尘滤饼，从而降低系统压降。此前几乎所有的计算流体动力学（computational fluid dynamics, CFD）相关研究均聚焦于单个过滤器，而间接预测滤饼生成速率的常用方法之一便是采用CFD技术，因其成本低廉、风险可控且操作简便。本研究针对原型系统中的双陶瓷烛式过滤器的滤饼生成速率进行建模，并借助细川粉体技术研究所（Hosokawa Powder Technology Research Institute, HPTRI）获取的实验数据对其压降进行验证。我们选用商业软件FLUENT提供的适配CFD模型，对双烛式过滤单元内部的三维流场进行模拟，通过计算得到的流场分布估算局部流速。基于100%收集效率与亚微米级气溶胶粒径的合理假设，颗粒物会沿气流流线均匀分布运动，因此可通过局部面流速预测烛式过滤器各部位的滤饼生成速率差异，并计算滤饼阻力特性。此外，本研究还利用CFD技术探究了过滤风速、温度、粉尘浓度以及进料口位置等关键参数的影响。结果表明，过滤风速是影响系统压降的最核心参数。