马萨诸塞州加德纳市是该州第一个使用聚砜（PS）超滤膜的社区，超滤膜的容量为100000 MWCO（~0.1µm） 地表水源上的物质。然而，在新膜系统安装后不久，该系统就被关闭了 无法通过制造商的额定生产能力，且 秋季周转期的问题。尸体解剖 膜表面有一层厚厚的二氧化硅污染层。各种酸、碱和 对消毒剂方案进行了测试，同时提高了反洗能力，但没有成功。这一点值得商榷 这个问题是膜本身的问题，还是水源的作用。像 因此，决定租用一个全尺寸的微滤膜单元，采用不同的 膜材料（聚偏二氟乙烯或聚偏二氟乙烯），但孔径相似（~0.1 µm）并并排测试两个系统。 经测定，PVDF膜在室温下的性能优于PS膜 水晶湖水源水的全面中试。通常PS系统会丢失 随着时间的推移，生产变得更加混乱。然而，PVDF膜， 展示了一个弹性系统，该系统具有有限的不可逆污垢，通常 更高的比通量。实施并实施了统一修正污垢指数（UMFI） 提供了对系统的相对性能的洞察，因为运营变化 可作为操作员在确定 膜操作的变化如何影响膜性能。 总的来说，从初步试验结果中得出以下观察结果: 在原水中加入明矾凝固剂对两种膜都至关重要 获得更高的通量率；和 PS膜需要大约10平方英尺的膜才能生产 在运行过程中，当膜被污染时，需要1加仑以上的水 大于20平方英尺的膜面积。当明矾加入时，PVDF 膜需要不到5平方英尺的膜来生产1加仑的水 水PVDF膜表现出与PS膜相似的趋势 明矾被关掉的时候。包括3个参考文献、表格和图表。

The City of Gardner, Massachusetts, was the first community in the state to use ultrafiltration membranes with a 100,000 MWCO (~ 0.1 µm) using a polysulfone (PS) material on a surface water source. However, shortly after the new membrane system was installed, the system was unable to pass the rated production capacity by the manufacture, with a particular problem during the fall turnover period. An autopsy of the membrane surface suggested a thick silica fouling layer. Various acid, base, and disinfectant schemes were tested along with increased backwashing capabilities, with no success. This left in question whether the issue was inherent with the membranes or rather, a function of the source water. As such, it was decided to rent a full scale microfiltration membrane unit with a different membrane material (Polyvinylidene di-fluoride or PVDF), but similar pore size (~ 0.1 µm) and test the two systems side-by-side. It was determined that the PVDF membranes performed better than the PS membranes at the full scale pilot test for the Crystal Lake source water. Generally the PS system lost production and became more fouled over time. The PVDF membranes, however, demonstrated a resilient system that had limited irreversible fouling and generally a higher specific flux. The Unified Modified Fouling Index (UMFI) was implemented and provided insight into the relative performance of the system, as operational changes were made and could be used as an index that the operator could implement in determining how a change in membrane operation affects membrane performance. In general, the following observations were made from the pilot testing results: dosing the raw water with an alum coagulant was essential for both membranes to obtain the higher flux rates; and, the PS membrane required approximately 10 square feet of membrane to produce 1 gallon of water during a run, and as the membranes fouled, required upwards of greater than 20 square feet of membrane area. When alum was dosed, the PVDF membranes required less than 5 square feet of membrane to produce 1 gallon of water. The PVDF membranes exhibited a similar trend as the PS membranes when the alum was turned off. Includes 3 references, tables, figures.