NF薄膜程序可同時去除病毒、溶解性有機物及硬度，並保留適當礦物質，達適飲性水質。然而運用難題在於薄膜阻塞，所引起濾程之縮短。故本研究以桌上型薄膜模組進行過濾程序，了解阻塞現象，包括薄膜表面結構、pH、掃流速度、溶質物種及其交互作用等因素，以期改善處理效能。結果顯示粗糙度大的薄膜表面，易於凹處沉積微粒，不利清洗後的清水通量恢復。
高pH時，雜質與薄膜的表面電位均較負，故有較大靜電斥力，可明顯提高去除率。僅含背景鹽之溶液的清水通量在等電位點有其峰值。含鈣有機物的清水通量，以錯合鈣為主導，高pH高阻塞。高掃流速度可減低阻塞程度，亦可降低離子濃度極化效應，進而提高鈣離子及導電度去除率，但對溶解性有機碳的去除率影響不大。
單一雜質試驗中，比較同濃度之3種溶解性有機物 (DOM)、4種Particles及Ca2+，結果顯示NF薄膜在相同操作條件下，除水楊酸以外，均有相似的清水通量下降趨勢，不因物種因素而有太大差異。且當進流濃度控制在某範圍時，濾餅形成後，即使薄膜不清洗，清水通量仍相當穩定。
混合溶質試驗中，於Particles、DOM、Ca2+之間，做任兩物種混合液之比較，結果顯示因DOM與Ca2+會產生錯合，而在膜面上形成密緻吸附膠層，使得NF薄膜之阻塞最嚴重。故對NF膜的前處理，去除Ca2+與DOM其中任一種或防其化學鍵結，比去除Particles更重要。
有機物在含鈣的情況下，藻酸的阻塞程度高於腐植酸，推測其主因為錯合鈣量較高。此結果意味著對優養化的水源，進行NF薄膜處理，是比一般天然有機物水源不利，需有適當的前處理單元。再者於比較鈣和有機物何者居多的阻塞程度，吾人得到去除有機物尤重於去除鈣的觀念。故相對於Particles或Ca2+，DOM對於NF膜的阻塞佔更重要角色。
比較含鈣有機物的阻塞沉積量與鈣錯合量對阻塞程度的相關性，發現無論進流水是含腐植酸或藻酸，均以鈣錯合量對阻塞程度的相關係數較高，故以鈣錯合量作為NF膜的有機物阻塞指數，應用於實際天然水源，可改善以往僅用LSI及SDI為薄膜阻塞指數的預測準確性，而為添加酸劑及抑垢劑提供最佳劑量的參考。

Nnanofiltration membrane can remove greater part of natural organic matter (NOM), precursor of disinfection by-products, and multivalent ions. Its rejection of monovalent ions is much lower than that of multivalent ions, and therefore retains some of the mineral substances in the permeate. However, the major problem of applications is the fouling of membrane. In this research, Rapid Bench Scale Membrane Test (RBSMT) is used to study the factors affecting the fouling of the NF membrane. Those factors studied include membrane surface morphology, cross-flow velocity, and water quality of the feed water, such as pH, type and concentration of the solute and suspended solids.
The results show the rougher the membrane surface is, the easier the fouling in valley, and unfavorable in flux recovery rate by cleaning. Increasing cross-flow velocity could reduce concentration polarization and fouling rate, and increase rejection rate for calcium and conductivity. However, its affect on dissolved organic carbons (DOC) rejection was minor.
And impurities rejection increased with increasing pH. This probably is due to the higher electrostatic repulsive force between impurities and membrane surface, as both usually gain more negative charge at higher pH. For feed water consisted of pure water and background electrolyte (0.003 M NaClO4), the flux has highest value at isoelectric point of the membrane. For feed water contained calcium and organics, flux was mainly affected by complexed calcium.
For single species impatrity testing, comparing the same concentration of salicylic acid, alginic acid, humic acid, 0.1 μm SiO2, 0.01 μm SiO2, kaolin, bentonite and Ca2+, results show that NF membrane have the same flux decline under the same operation condition, regardless of the kinds of solutes, except for salicylic acid. The higher the pH value, the higher the flux decline rate. For feed water contained organics without calcium, flux was mainly affected by their solubility in water. And when impurity concentration was in certain range, after cake formation and initial flux decline, the flux value would remain at certain value for long time, even without cleaning.
For dual impurities testing, it is noticed that systems containing dissolved organic matter (DOM) and calcium have the highest fouling rate. This is probably due to the formation of tenacious gel layer, made from the adsorption of DOM-Ca complex membrane surface. Therefore, in the preatment of NF, the removal of either DOM or Ca is more important than removing particles.
When DOM was mixed with Ca, the fouling of alginic acid was more serious than humic acid, probably because the quantity of complexation formation of the former was larger than that of the latter. This also indicates that when the source water of a NF system is entrophic, the pretreatment to prevent excessive fouling is especially important. Comparing the fouling phenomena with feed water under various concentration ratio of Ca and DOM, we notice that the role palyed by DOM is more significant than that of Ca or particles.
Finally, comparing the correlation between surface deposition quantity and flux decline rate versus that between calcium complexation qantity and flux decline rate, it was found the correlation between the latter was stonger, no matter the DOM involved was humic acid or alginic acid. Therefore, calcium complexation quantity may be used as fouling index in the field.

[1] Lyonnaise, D. E., Water treatment membrane process, AWWA Research Foundation (1996).
[2] Pontalier, P. Y., Mechanisms for the selective rejection of solutes in nanofiltration membranes, Separation and Purification Tech. v.12, 2 (1997) 175–181
[3] Peeters, J.M.M., Retention measurements of nanofiltration membranes with electrolyte solutions, J. of Membr. Sci. 145 (1998) 199-209
[4] Braghetta, A., Francis, A. D., Nanofiltration of natural organic matter : pH and ionic strength effects, J. of Environ. Engi. (1997).
[5] IRC manual for bench- and pilot-scale treatment studies, EPA 814/B-96-003 , April (1996).
[6] Noble, R.D., Membrane separations technology principles and applications (1999).
[7] FilmTec membranes technical manual, April (1995).
[8] Taylor, J. S. and Mulford L. A., Membrane protocol to meet the ICR, presented at the AWWA Membrane Technology Conference, August 13-16, Reno, Nev. (1995).
[9] Laîne’, J. M., and Anselme, C., Impact varies pretreatment on nanofiltration Performance in treating water, presented at the AWWA Membrane Technology Conference, August 13-16, Reno, Nev (1995).
[10] Yiantsios, S.G., The effect of colloid stability on membrane fouling, Desalination, Vol. 118, 3 (1998) 143-152
[11] Chellam, S., Jacangelo, J.G., Bonacquisti, T. P., and Schauer, B. A., Effect of pretreatment on surface water nanofiltration, J. AWWA, 89 (1997)10-77.
[12] Leenheer, J.A, Characterization and digenesis of strong-acid carboxyl groups in humic substances, Applied Geochemistry 18 (2003) 471-482
[13] Croue, J. P., Characterization of natural organic matter in drinking water, AWWA Research Foundation (2000).
[14] Schnitzer, M., Characterization of organic matter associated soil clay, Soil Sci. 145 (1988) 448-454.

[15] Plummer, J. D.; Edzwald, J. K., Effect of ozone on disinfection by-product formation of algae, Water Sci. and Tech. 37 (1998) 49-55.
[16] Nilson, J. A. and DiGiano F. A., Influence of NOM compostion on nanofiltration, J. AWWA (1996) 53-66
[17] Cheng, F. L., Effects of humic substance characteristics on UF performance, Wat. Res. (2000) 1097-1106.
[18] Schäfer, A. I., Fane, A. G., and Waite, T.D., Nanofiltration of nature organic matter: removal, fouling, and the influence of multivalent ions, Desalination, (1998) 118.
[19] Jucker, C., Adsorption of aquatic humic substances on hydrophobic ultrafiltration membranes, J. of Membr. Sci. 97 (1994) 37-52.
[20] Vrouwenvelder, H. S., Biofouling of membranes for drinking water production, Desalination, 118 (1998) 157-166.
[21] Souriraian, S., Characteristics of porous cellulose acetate membranes for the separation of some organic substances in aqueous soultions, I&EC Prduct Research and Development, 4 (1996) 201-206.
[22] Childress A. E., Menachem Elimelech, Relating nanofiltration membrane performance to membrane charge (electrokinetic) characteristics, Environ. Sci. Technol. 34 (2000) 3710-3716.
[23] Cheryan, M., Ultrafiltration and microfiltration handbook, Technomic (1998) 46.
[24] Schippers, J. C., and Verdouw, J., The modified fouling index, a method of fouling characteristics of water, Desalination 32 (1980) 137.
[25] Sung, L. K. and Taylor, J. S., Using film theory and ion coupling to model mass transfer in diffusion controlled membrane system, In proceedings of the 1995 membrane technology conference, Reno, Nevada. (1995)
[26] Belfort, G., Fluid mechanic of membrane modules, J. of Membr. Sci., 40 (1989) 123-147.
[27] Gregor, J. E., Interactions of calcium and aluminum ions with alginate, Water Res. 30 (1996) 1319–1324.
[28] Maxime, P., Effect of aging on UF membranes by astreaming potential (SP) method, J. of Membr. Sci., 154 (1999) 213 – 220.
[29] Wei Y., Humic acid fouling during microfiltration, J. of Membr. Sci., 157 (1999) 1– 12.
[30] Osmonics, SEPA CF membrane cell instruction manual (1995).

[31] Vrijenhoek, E. M., Influence of membrane surface properties on initial rate of colloidal fouling of reverse osmosis and nanofiltration membranes, J. of Membr. Sci. 188 (2001) 115-128.
[32] Bowen, W. R.; Tedora. D., Atomic microscopy studies of nanofiltration membranes: surface morphology, pore size distribution and adhesion, Desalination, 129 (2000) 163 – 172.
[33] Seungkwan, H., Chemical and physical aspects of natural organic matter (NOM) fouling of nanofiltration membranes, J. of Membr. Sci., 132 (1997) 159-181
[34] Amy, G., NOM rejection by, and fouling of NF and UF membranes, AWWA Research Foundation (2001).
[35] 葉宣顯等人，本省自來水水源中溶解性有機物成分之分析及現有淨水程序對其去除效率之評估，台灣省自來水股份有限公司研究報告 (1998)。
[36] 楊子岳，桌上型快速薄膜試驗評估NF薄膜程序在表面水處理上之運用，國立成功大學碩士論文 (1999)。
[37] 高山鎮，薄膜阻塞控制之研究，國立成功大學碩士論文，(2000)。
[38] 林俊德，極微濾薄膜在自來水淨水工程上之應用，國立交通大學碩士論文 (2000)。
[39] 葉宣顯等人，澄清湖高級淨水處理模型廠試驗研究 (第二年)，台灣省自來水股份有限公司研究報告 (2001)。