反覆過濾程序中濾材過濾特性之研究

Study on the Filter Media Performance during Filtration Cycle

汪美玲(Mei-Ling Wang)

指導教授 ： 童國倫

中原大學/工學院/化學工程研究所/碩士(2004年)

https://doi.org/10.6840/cycu200400049

摘要

編織濾材被廣泛應用於化工相關產業之分離程序，纖維材質和紡紗型式，以及編織方法，對於過濾效能有著深遠的影響，而分離程序中漸增的過濾阻力是工程師常遭遇到的問題。本研究目的是仿效固液分離上的反覆過濾程序，結合統計學的理論模式與實驗，推導一條數學式以表現濾布的過濾特性。
在實驗研究方面，首先建立一套自動化濾材反覆過濾試驗裝置，可用以進行濾材反覆過濾與逆洗效能的實驗分析。研究中，使用恆壓過濾操作方式，選用三種不同構造之濾布，分別是聚丙烯材質的斜紋織濾布與緞紋織濾布，以及聚氨酯不織布，在碳酸鈣泥漿過濾下，探討濾布洗滌方向與操作壓力及泥漿濃度，對反覆過濾特性的影響。理論方面使用純生程序（pure birth process）之機率模式，分析濾材表面單位面積的濾餅變化量，與過濾程序中粒子堵塞濾布孔道的機率；以及洗滌程序中，堵塞的孔道被洗淨的機率，探討三者之間的關聯，並且結合孔道衰退比值，以一條簡化的方程式來表示其相互之間的關係。推衍導得之方程式，能表示反覆過濾與洗滌程序中，濾材阻力的變化與這些參數的關聯，並將堵塞參數 與洗滌參數 以及濾餅沉積量 結合成濾布特性參數 。
實驗研究結果發現，濾布的構造以及孔徑/粒徑相對大小是影響過濾阻力之主因；操作壓力較高與泥漿濃度較高時，擁有最低總過濾阻力者是孔徑最小的聚氨酯不織布，不過操作壓力降低或濃度降低時，聚氨酯不織布的總過濾阻力卻大於聚丙烯的緞紋織濾布。反向洗滌可以把堵塞孔道之粒子推擠出去，所以其效果優於正向。理論結果分析顯示，藉由濾布特性參數 ，可推算單位濾餅變化量使濾布孔道衰退的比值 ，進而由 與 去估算過濾時間 ，表現出濾布特性參數 的數值越小則其濾性越好，估算的過濾時間可以符合實驗值。由 對過濾次數作圖所繪出的曲線，可以判斷 是由堵塞機制或洗滌機制支配，顯示機率理論所作之分析結果可以反映出不同濾布結構與相異操作條件下的反覆過濾特性。

關鍵字

反向洗滌 ； 機率模式 ； 反覆過濾 ； 濾布 ； 堵塞 ； 正向噴洗 ； 濾布特性參數 ； 堵塞參數 ； 洗滌參數

並列摘要

In order to conduct the filtration performance of filter cloths during filtration cycles, an automatic experimental apparatus was constructed. Effects of the woven structure of filter cloths, types of yarn and materials on the washing performance of filter cloth were studied. However, there is not any reliable model to predict the increase of filter media resistance during a filtration process. The other purpose of this study is to develop a stochastic model to predict the filtration properties based upon the obtained filtration data from the testing equipment.
Three types of filter cloths, twill weaved polypropylene fabric, reverse satin weaved polypropylene fabric, and non-woven polyurethane fabric, were examined in this work. Constant pressure filtration (CPF) operation with calcium carbonate suspension was performed under various operating pressure and suspension concentration during each filtration cycle. The increase in the filter cloth resistance during the filtration cycles was examined experimentally with washing at the end of each cycle. Methods of washing, forward spray washing and backward washing, were also compared in this study.
The probability model of the pure birth process is adopted to analyze the variation of number of particles on a unit area of dry cake on the filter fabric, and to describe the probabilities of particle clogging in filtration process and particle cleaning in washing process. Further introducing a parameter of porosity decline ratio, a simplified equation can be derived to interpret the filtration performance of filter fabric. This equation consists of four parameters of clogging parameter η for clean pores, washing parameter γ for clogged pores, quantity of deposited cake on the cloth , and the porosity decline ratio of filter cloth under unit variation of cake mass . Ultimately, a characteristic parameter of filter fabric can be obtained and expressed as a function of η, γ and . This derived equation can be used to describe the variation of filtration resistance during both of filtration and washing processes.
Experimental results depict that structure of filter fabrics and the diameter ratio of particle to fabric pore are two major factors affecting filtration operation during filtration cycles. Under a higher operating pressure and suspension concentration, the non-woven polyurethane gives the lowest total resistance compared to the other two types of filter fabrics. However, under a lower operating pressure and suspension concentration, the non-woven polyurethane shows the highest total resistance compared to the other two types of filter fabrics. Back washing operating mode shows a higher efficiency of filter fabric cleaning than that of forward spray washing mode because it could squeeze the clogged particle out of the fabric easier. Stochastic analysis reveals that the porosity decline ratio of filter cloth under unit variation of cake mass can be predicted based upon the experimentally available characteristic parameter of filter fabric . A lower value of under a specified variation of cake mass indicates a better filtration performance of filter fabric. Once the porosity decline ratio of filter cloth under unit variation of cake mass obtained, the filtration time of total filtration cycles can be estimated based upon the and the experimentally available correlated time coefficient . Validation examination reveals the proposed stochastic model to be a reliable method to predict the filtration performance of filter fabric.

並列關鍵字

clog ； backward washing ； characteristic parameter ； filter cloth ； filtration cycle ； probability model ； forward spray

參考文獻

Backer, S., “The Relationship Between the Structural Geometry of a Textile Fabric and Its Physical Properties-Part IV: Interstice Geometry and Air Permeability,” Text. Res. J., Vol. 21, P.703 (1951).

Chang, D. J., S. H. Chen, S. S. Lin, K. C. Yu, C. Y. Chang, “An Investigation of Reciprocating Filtration System,” J. Member. Sci., Vol.202, P.233 (2002).

Duo, W., N. F. Kirkby, J. P. K. Seville, R. Clift, “Patchy Cleaning of Rigid Gas Filters－I. A Probabilistic Model,” Chem. Eng. Sci., Vol.52, No.1, P.141 (1997).

Doyen, W., B. Baée, F. Lambrechts, R. Leysen, “Methodology for Accelerated Pre-selection of UF Type of Membranes for Large Scale Applications,” Desalination, Vol.117, P.85 (1998).

Defrance, L., M. Y., Jaffrin, “Reversibility of Fouling Formed in Activated Sludge Filtration,” J. Member. Sci., Vol.157, P.73 (1999).