1、采用好氧方式对有机物的去除研究表明，MBR对有机物的去除效果非常好，平均去除率为96.2%，出水平均COD为16.7mg/L，远低于国家循环冷却水和景观用水标准，可以回用。MBR去除有机物的关键力量仍是生物反应器，膜分离则直接和间接地起到了强化作用。

2、硝化研究表明，在膜生物反应器中，由于污泥泥龄长，而且溶解氧充足，有利于硝化菌生长，因此对NH3-N的去除效果非常好，多数情况下，出水NH3-N低于1mg/L，去除率大于90%。

3、反硝化研究表明，由于MBR污泥浓度高，内部存在厌氧微环境，起到了在好氧条件下同步硝化反硝化的作用。试验条件下，NH3-N平均去除率为99.1%，TN平均去除率为56.1%。

4、MBR工艺的污泥特性表现为：污泥产量低，沉降性能差，但以膜分离实现泥水分离可以保证系统的正常运行。MBR工艺由于膜的截留作用，活性污泥会在反应器内逐渐积累而升高，但这种升高并不是无限制的。研究表明，MBR污泥浓度出现了一个峰值，随后进入相对稳定状态；与传统活性污泥CAS工艺相比，MBR的污泥颗粒平均粒径小，这对提高传质效率很有益处。试验中，Dp,MBR(50%)=127.12µｍ，Dp,CAS(50%)=227.24µｍ；在一定的容积负荷下，VSS/SS比活性随着运行时间的延长表现出下降的趋势。

5、 活性炭－膜生物反应器组合工艺研究表明，投加粉末活性炭后MBR对COD的去除率较之未投加时有所提高（由投加前的97%提高到99%），尤为重要的是，投加PAC后，膜内外的COD差值得到了很大程度的降低（由投加前的48mg/L降至投加后的14mg/L）；投加PAC对NH3-N的去除效果影响不大。

6、本次试验对凹凸棒－膜生物反应器组合工艺研究并没有取得理想的效果，这除了试验周期较短以及试验进水浓度较低等原因外，经分析，最主要的是凹凸棒对COD、活性污泥等的吸附性能不及PAC的吸附性能理想。看来，这种组合工艺还需要进一步的研究，尤其是凹凸棒本身的性能、特点以及应用于污水处理的预处理方式需要进一步研究。

7、MBR膜污染机理和膜污染防治研究表明，水力反冲洗能够有效地恢复膜通量，但周期不能太长。试验运行60天后，对膜组件进行了水力反冲洗，效果虽好，但维持时间过短； MBR运行半年后，对膜组件进行了化学浸泡清洗。选择的清洗剂有HCl、NaOH和NaClO。结果表明，在膜污染较严重时，选择化学清洗还是非常有效的。经过化学清洗，本试验的膜通量恢复了70%。

8、试验后期，尝试了向MBR反应器中投加活性炭来改善膜污染的程度与速度。发现，随着PAC投加量的增大，水头损失的增长速度明显减慢。未投加PAC时，水头损失增长速度为16mm/d，而PAC浓度为2g/L时，增长速度减至1.7mm/d，取得了非常理想的效果。这主要是由于：一方面，PAC的强烈吸附作用改变了活性污泥絮体的结构，使其变得更大，因而在膜表面形成的泥饼层比较疏松；另一方面，PAC的存在减少了胞外聚合物EPS的含量，因此活性污泥的粘度更小，膜通量就会随之增大。

(1) The research of organism removal adopted aerobic mode shows that MBR is very effective to remove organisms – the average removal rate is 96.2% and the average effluent chemical oxygen demand (COD) is 16.7mg/L, far lower than the national standard of circulatory cooling water and water used in sights. So, it is reclaimable and reusable. Removing organisms is mainly dependent on the bioreactor, while membrane separation intensifies the action directly or indirectly.

(2) The research of nitrification shows that the removal effect of NH3-N is satisfactory, because the long sludge retention time (SRT) and sufficient oxygen dissolved in MBR is favorable for the growth of nitrifying bacteria. In most cases, the effluent NH3-N is lower than 1mg/L and the removal rate is higher than 90%.

(3) The research of denitrification shows that because of the high sludge concentration, an anaerobic micro-circumstance exists in the interior. So, MBR has the function of simultaneous nitrification and denitrification. Under the experiment condition, the average removal rate of NH3-N and TN are 99.1% and 56.1% respectively.

(4) The speciality of MBR technics presents low output of sludge and low capability of sedimentation. But it can guarantee the system to run normally by using membrane separation process to realize the separation of sludge from water. Using MBR technics, activated sludge will pile up and rise gradually in the reactor because of the holding-up function of membrane, but the rise is not unlimited. According to the researches, a peek value of the sludge concentration appears first in MBR, and then followed a relatively stable state. Compared with the conventional activated sludge (CAS) process, the average diameter of sludge particle (Dp) in MBR is smaller, which is beneficial to improve the substance transfer efficiency. In the experiment, Dp MBR (50%) = 127.12µm, Dp CAS (50%) = 227.24µm. Under the specific definite cubage load, the ratio of volatile suspended solids to suspended solids (VSS/SS) indicates a downtrend along with an increase of runtime.

(5) The research of the BAC-MBR combined technology shows that the removal rate of COD in MBR increases after adding PAC - from 97% to 99%. It is especially important that after adding PAC the difference value of COD between internal and external membrane decreases observably - from 48mg/L to 14mg/L. However, adding PAC seems ineffective to remove NH3-N.

(6) The research of the Attapugite-MBR combined technology in the experiment has not led to satisfactory effects. According to analysis, besides the limited period of experiment and the low concentration of the influent, the uppermost reason is that the sorption capability of attapugite to COD, activated sludge, etc. is lower than that of PAC. It seems that this kind of combined technology still need further study, especially in the capability and speciality of the attapugite itself, as well as the pretreatment mode used in wastewater treatment.

(7) The research of membrane contamination mechanism and protection of MBR shows that waterpower counter-rinse can restore the membrane flux effectively, but the period cannot be too long. After 60 days’ operation in the experiment, although the effect of waterpower counter-rinse to the membrane groupware is satisfactory, the maintainable duration is too limited. After six months’ operation of MBR, we chose HCl, NaOH and NaClO as detergents to rinse membrane groupware by using chemical soaking and rinsing method. The results show that chemical rinse is very effective in serious membrane contamination. By chemical rinse, 70% of membrane flux was resumed in this experiment.

(8) In the last stage of the experiment, we tried to add activated carbon into MBR in order to reduce the membrane contamination degree and velocity. Along with the increase of PAC quantity, the velocity of water-head-loss increase slows down evidently. When the concentration of PAC reaches 2g/L, the velocity of water-head-loss increase reduces to 1.7mm/d, compared with a 16mm/d before adding PAC. So, the effect is perfect. It is mainly because, on one hand, the strong sorption of PAC changes the structure of sludge flocculus, making it larger, and, as a result, forms a loosened cakelike sludge layer on the surface of the membrane; on the other hand, the existence of PAC reduces the content of EPS, so the viscidity of activated sludge is lower and the membrane flux increases correspondingly.