Ultrafiltration for Municipal Wastewater Reclamation and Potable Reuse

1. Introduction
Ultrafiltration (UF) is an important component of treatment schemes for water reuse due to its effective pretreatment for downstream processes like reverse osmosis. UF also plays a significant role in direct and indirect potable reuse because it can physically remove pathogens. However, UF fouling is a problem in many reuse process trains and is detrimental to long-term filtration performance. This research focuses on UF fouling mechanisms and its mitigation by coagulation pretreatments using iron salts. Earlier work largely considered conventional coagulation and electrocoagulation using aluminum salts, so this will be one of the first to implement electrocoagulation using iron salts. The motivation to evaluate electrocoagulation with iron salts is that it can induce Fenton reactions generating strong oxidants like ⋅OH that are capable of attenuating viruses thereby gaining virus removal credits and possibly improving other performance.
2. Materials and methods This research was conducted at Texas A&M University in cooperation with Orange County Water District to investigate the effects of coagulation on ultrafiltration.
1) Two secondary effluents. One was from the City of Bryan, TX wastewater treatment plant (WWTP), and the other was from the Texas A&M University WWTP. Both employ conventional activated sludge, and samples were collected after ultraviolet (UV) disinfection.
2) Filtration system setup and procedure. The hollow fiber UF membranes (Memcor©: S10N) were sealed in the home-made module. A constant filtration flux of 60 L/(m2⋅h) was employed for a 22-minute duration. Backwash cycles of 80 L/(m2⋅h) for 15 seconds were followed, as commonly practiced in municipal applications. Each experiment typically consists of 9 cycles of such filtration and backwash. Data were acquired using a written program in LabVIEW software to continuously monitor the pressure and automatically control the process.
3) Coagulation pretreatment. Coagulation and flocculation jar tests were performed with iron chloride added for conventional coagulation and with iron plate electrolysis for electrocoagulation.
4) Blocking law mathematical model. The exponent in constant flux blocking law formula is modelled to indicate sizes of colloids deposited in membrane pores or on the membrane wall.

3. Results Counter-intuitively, the secondary effluent with lower turbidity and TSS (shown in attached image 1) caused much higher fouling (shown in attached image 2). After comparing water quality parameters of two effluents, one hypothesis was proposed: UF fouling is driven by the feed water particle size distribution relative to the membrane pore size, especially when the new membrane is used. After measuring particle size distribution of two feedwaters (shown in attached image 3) and validating the size measurement by mathematical modelling, the hypothesis was proved, and the following membrane fouling mechanism was revealed: UF fouling is determined by nano-particle size distribution in the feed water and its comparison to the membrane pore size.
Nano-colloids were smaller than the nominal membrane pore size before pretreatment, allowing their deposition in the porous matrix of the membrane and making them difficult to be removed by backwashing. Coagulation pretreated flocs are much larger than membrane pores and deposited on the external membrane surface as a loose cake layer, thus easily removed by backwashing and alleviating both irreversible and reversible fouling. As a result, UF productivity was largely determined by colloid size. Electrocoagulation using iron performs the same as conventional coagulation with iron in terms of UF fouling mitigation, even though it can achieve virus inactivation. To exclude the synergistic effect of organics and harness ions and isolate the colloidal size effect, the permeate of Texas A&M WWTP effluent ultrafiltration was ultra-filtered as a feedwater because it contained above 90% of harness ions and total organic concentration in the effluent. It showed much less fouling than the effluent, so organic fouling effect mechanism was excluded in this study.

4. Significance to the industry Based on our findings, monitoring the UF feed water particle size distribution (especially in the nano-colloid range2 100 nm) is recommended during wastewater reclamation to optimize reuse process train performance and enhance water recovery. Conducting bench or pilot scale testing on sampled plant effluent will help inform the design of a more efficient UF system and further benefit the whole reuse process train.