【作者】 邓欢欢；

【导师】 赵建夫； 李建华；

【作者基本信息】 同济大学 ， 环境科学， 2007， 博士

【摘要】 本文通过人工湿地对农村富营养化水体中污染物去除性能和规律进行的小试和中试试验，对湿地植物的生长特性及其吸收氮磷的特点、湿地基质微生物数量分别与碳源代谢等进行了较为系统和深入的研究，通过试验结果可以得到如下结论：4种类型湿地(香蒲水平潜流、风车草水平潜流、香蒲垂直潜流、风车草垂直潜流)对富营养化水体均具有较好的净化能力。在水力停留时间为3d，进水COD、总氮、总磷分别为45～110mg／l、1．5～5．2mg／l和0．22～1．54mg／l的条件下，出水COD能达到地表水Ⅳ类水标准(GB3838-2002)，总氮除8月为Ⅳ类外其余均能达到Ⅲ类水标准，总磷除8、9月为Ⅲ～Ⅳ类外其余均能达到Ⅱ类水标准。在水华暴发的7、8、9月间湿地对浮游植物的去除效果良好，在进水叶绿素a浓度高达103．5mg／m~3～252．4mg／m~3时，湿地的去除率可达82．5％～99．2％之间。对污染物面积负荷与湿地面积去除率的研究表明，在处理这类低浓度富营养化水体时，4种类型的湿地对COD、总氮、总磷的面积去除率与污染物面积负荷在全年期间均呈良好的线性关系，显示了在处理低浓度水时良好的处理与适应能力。总体来说人工湿地能够有效地发挥净化农村富营养化水体的作用，出水能够达到较好的水质水平。在4种类型湿地去除效果中比较，在试验初期和第二年初春(3月)，两种植物类型湿地中，无论是垂直潜流还是水平潜流，污染物各个指标香蒲湿地均略好于风车草，但在随试验进行到第一年10月至1月间，第二年4月后，风车草又略好于香蒲，显示植物生长发育周期不同对湿地处理效果影响较大。在两种流态湿地中比较，COD是水平潜流一直略好于垂直流；而总氮和氨氮在香蒲湿地中垂直流在秋冬季节要略好于水平潜流，其它时间相当，而风车草湿地中两种流态没有明显差别；总磷在香蒲湿地中垂直流在冬春要略差于水平潜流，风车草湿地两种流态差别不大。总体来看植物对去除效果的影响要大于流态的影响，但这种影响在处理这种低浓度富营养化水体上并不很明显。在水平潜流湿地中，从前向后植物高度和分生植株数量都呈递减趋势，而垂直潜流湿地中的植物株(丛)间差异不明显。收割后的植物地上生物量显示风车草地上部分株(丛)均生物量要远大于香蒲，不同单元间约为香蒲的1．51～3．01倍，水平潜流湿地前后植物地上部分生物量差异显著。对不同湿地植物NP元素进行测定，结果显示：风车草植物N元素含量远高于香蒲；不同单元植物地上部分N元素差异显著，在水平潜流湿地中从前向后N元素含量呈梯次递减，而P元素差异较小，垂直潜流湿地株(丛)间差异不显著且NP含量较水平潜流高。人工湿地处理低浓度富营养化水体时，植物更容易受到N元素缺乏胁迫影响，而P影响较小。比较人工湿地不同单元间植物地上部分吸收NP量，两种植物的水平潜流湿地中从前向后4个单元间吸收量差异均极为显著，从前向后显著递减，这种差异性与生物量或植物NP浓度相比而言，与生物量相关性更强，显示植物生长受到污染物浓度沿程降低(特别是N元素)的显著影响，而垂直潜流湿地中植物吸收量与水平潜流湿地最大的第一单元相当。研究植物地上部分吸收NP量对湿地总去除量的贡献可知：植物吸收的作用很大，反映了在处理低浓度富营养化水体时，植物吸收发挥了重要的作用，通过收割植物地上部分能将植物吸收的NP部分迁移出人工湿地，延长湿地使用年限。使用三维荧光扫描技术对溶解性有机物在湿地中的转化研究结果表明进水的三维荧光光谱可表征出4类溶解性有机物，在湿地净化过程中，这4类有机物表现出不同的趋势，表征受污染状况的2个常见类蛋白峰(类酪氨酸和类色氨酸)均出现降低的趋势，显示进水中部分溶解性有机物的被利用和削减，而2个类富里酸峰(可见光类富里酸和紫外类富里酸)则出现增加，证实了有机物在湿地中并不只是简单的被吸收利用，而是存在分解转化，这种转化更明显的表现在向不易被生物利用的腐殖化方向发展。在湿地垂直方向上，上层孔隙水中类蛋白峰较强，而下层孔隙水中类富里酸峰较强；在水平潜流湿地水平方向上前部类蛋白峰较强，而后部类富里酸峰较强，较易被利用的类蛋白峰强变化与微生物数量和碳源代谢活性大小趋势相符。在两种流态湿地中，细菌总数都是沿程逐渐降低，水平潜流湿地上层基质细菌总数要比中下层高约1个数量级，但细菌总数与各段污染物去除率之间没有明显相关性。Biolog测试得到的平均色度变化(AWCD)显示，就微生物群落对能源碳的代谢能力而言：在两种流态湿地样点中比较，不同季节下，微生物群落代谢能力有差异，在秋季，水平潜流湿地前部下层低于其它三点，微生物活性而垂直潜流上层较低于下层；在春季，较低的这两点都出现活性恢复。对Biolog测试数据进行不同类别碳源类型的代谢分析：各采样点对聚合物、糖类、羧酸和氨基酸4类碳源利用较好，但对胺类和酚类利用差异较大。对Biolog测试96h数据标准化后使用主成分分析，结果表明，无论是秋季还是春季，两种流态人工湿地床体不同部位样点均能被分开，显示样点间各自具有不同的碳源利用特性，各自对碳源具有不同的利用特性。根据Biolog测试的数据计算不同样点的碳源利用功能多样性指数显示：在不同季节不同流态的湿地中，各采样基质点丰度、均匀性、多样性指数均有差别，但丰度指数相差较小，一般为上层>下层；均匀性指数则显示，在秋季水平潜流湿地中，前部下层低于其它三点，在垂直潜流湿地中，上层较低于下层，该点显示了较强的碳源利用偏好，在春季，较低的这两点都出现显著升高；多样性指数变化趋势与丰度指数类似。对污染物在湿地床体空间沿程变化的研究表明污染物间变化规律各有差异。COD在水平潜流中无论暖季寒季均呈现沿程梯形递减规律，而垂直方向上从上向下是逐渐增加；而垂直潜流则没有表现出明显的沿程降低趋势，在寒季会出现1／2段点高于床体其它各点和出水的情况，但两种流态前1／4段去除均占主要贡献；总氮空间变化规律与COD类似，但垂直潜流在寒季从沿程1／2段开始变化很小；总磷与COD、总氮规律也相近，但垂直流在暖季从沿程1／2段开始基本没有变化，就COD、总氮、总磷而言，前1／4段均占去除的主导作用，而在风车草湿地中这一点表现的更加明显。在水平潜流中污染物浓度沿程削减与植物的生长特性(株高和分生植株数)及吸收NP量有较好的一致性，在水平方向上沿程变化与微生物数量也较为一致，但在湿地下层，与秋季微生物群落碳源代谢活性变化并不一致。在垂直潜流湿地中，污染物浓度沿程变化与微生物数量略为一致，与微生物群落碳源代谢活性春季趋势一致，但与秋季趋势并不一致。氨氮的空间变化规律与其它3个指标不一致，在水平潜流湿地中，暖寒季均表现为沿程1／4段增加较多，然后逐渐降低；在垂直潜流中暖季是在1／4段升高，然后在1／2段降低，然后又沿程升高一些，寒季则为在3／4段前沿程升高，后1／4段又降低的趋势，这与流态上的差别使得湿地中有机氮分解的规律存在差异有关。研究表明湿地对氨氮的去除并不理想，出水氨氮浓度高于进水的情况在一年中大部分时间内存在，由总氮和氨氮沿程变化可知，很可能进水中部分有机氮被截留后在厌氧条件下并没有得到全部去除，而是分解释放，有部分释放的氨氮没有被硝化—反硝化作用去除，而是随水流出。而通过溶解性有机物的研究可知，湿地沿程表征受污染程度的类蛋白荧光强度的削减可能正标示着类蛋白类有机物分解释放出氨氮。但全年湿地出水氨氮浓度月均值均小于0．57mg／l，基本达到Ⅱ类水标准，因此氨氮处理不理想并不影响出水水质。在小试试验结果基础上开展了应用高速泳动床—人工湿地复合系统工程实际处理富营养化水体，调水改善农村天然水体水质工程研究，结果表明高速泳动床能降低氨氮浓度、提高湿地进水溶解氧含量，强化后续湿地工艺去除能力。人工湿地对高速泳动床出水的去除效果较好，COD出水能达到地表水Ⅲ～Ⅳ类水标准，总氮出水能达到地表水Ⅳ类水标准，总磷出水能达到地表水Ⅱ类水标准，叶绿素a去除率由初期的75％左右提升到90％以上。湿地对氨氮的去除效果不理想，但出水均<1．0mg／l，能达到地表水Ⅲ类水标准。复合垂直潜流湿地在较高的负荷情况下(停留时间约少1／3)，仍能达到与水平潜流湿地相当的去除效果，对总氮和氨氮去除率还略高于水平潜流湿地，显示了较强去除能力。总体来看，在300t／d的水量负荷下，高速泳动床—人工湿地联合工艺能显著改善补充进中心湖的调水水质，随着系统的进一步成熟(包括湿地植物的生长、基质微生物群落的发育成熟)，整套系统的处理能力有望提高，能为增强中心湖水体交换，改善水质提供有力的支撑。更多还原

【Abstract】 The study was based on constructed wetlands （CWs） in small-scale and pilot-scale reactors treating the rural eutrophic water. The treatment performance and mechanisms of CWs, growth character of plants, metabolic properties and functional diversity of the microbial communities in CWs were analyzed.Four types of CWs, Typha angustifolia/gravel bed horizontal subsurface flow constructed wetland（HFCW）, Cyperus alternifolius/gravel bed HFCW, Typha angustifolia/gravel bed vertical subsurface flow constructed wetland（VFCW）, Cyperus alternifolius/gravel bed VFCW, had good treatment performance. Under the condition that COD, TN and TP concentration of the influent were 45～110mg/l, 1.5～5.2mg/l and 0.22～1.54mg/l respectively on hydraulic rention time （HRT） of 3d, COD concentration of the effluent could gain the level of class IV of surface water quality according to GB3838-2002, TN concentration could meet the standard of classIII except class IV in August, TP concentration could stabilized at the level of class II except class III and IV in August and September. In the period of algal bloom（July, August and September）, the removal rate of Chla in CWs was 82.5%～99.2% under the condition that the concentration of Chla was about 103.5mg/m³～252.4mg/m³. The area loading rates of COD, TN, TP were found linearly correlated with their removal rates. Compared to four types of CWs, removal rates of Cyperus alternifolius CWs were higher than that of Typha angustifolia CWs from October to January of the next year, but lower from May to September in both HFCWs and VFCWs. Between two flow patterns Typha angustifolia CWs, COD removal rate of HFCWs was higher than that of VFCWs in all times, but TN and NH₄⁺-N removal rates were lower than that of VFCWs in autumn and winter, TP removal rates was higher than HFCWs in winter and spring. The contamination removal rates of contamination in two flow patterns Cyperus alternifolius CWs had no remarkable difference.The growth characters of plants were different among four types of CWs. Along the beds in HFCWs, the plant height and tillered plant number both sloped down. But in VFCWs, there were no remarkable difference. The aboveground biomass of Cyperus alternifolius was much more than that of Typha angustifolia. The concentration of nitrogen in Cyperus alternifolius was much higher than that in Typha angustifolia, but the concentration of phosphorus had no much variance. Among the difference cells in CWs, the variance of nitrogen concentration was more remarkable than phosphorus. The concentration of nitrogen distribution could be summarized as follow: hf-I > vf > hf- II > hf-III> hf-IV. It could be explained that the plant in CWs treating the rural eutrophic water was easier to be menaced by nitrogen not by phosphorus. The total nitrogen and phosphorus absorbed by plant aboveground biomass in different cells decreased along the beds in HFCWs. This variance was positively correlated with the aboveground biomass, while showed no apparent relationship with the concentration of nitrogen and phosphorus. According to the contributing of nitrogen and phosphorus about removal method analysis, it was found that plant uptake played very important role.In this study, researcher selected effective determination methods like dissolved organic carbon （DOC）, UV-Visible absorption and Three-Dimensional Excitation Emission Matrix （3DEEM） analysis, to conduct a research on distribution and transport of dissolved organic matter （DOM） by CWs. There were four types of fluorescence peak in the influent spectrum of 3DEEMs, which were attributed by four kinds of DOM: visible fulvic acid-like, UV-fulvic acid-like, tryptophan-like and tyrosine-like. The positions of four types of fluorescence peak in the difference sampling interstitial water and effluent were similar. The protein-like fluorescence peaks （tryptophan-like peak and tyrosine-like peak） in interstitial water and effluent samples of CWs were observed weaker than that in influent samples. The fulvic acid-like fluorescence peaks （visible fulvic acid-like peak and UV-fulvic acid-like peak） in interstitial water and effluent samples of CWs were stronger than that in influent samples. It showed that DOM was not only utilized by the microorganisms and plants in CWs, but also transformed to other forms. In this study, DOM in influent was likely to transformed to fulvic acid-like matter which was a sort of stodgy food to microorganisms. The microbial experiment results revealed that the abundance of bacteria in four types of CWs substrates in top layer was larger than in middle and bottom layer. In HFCWs, the abundance of bacteria was higher in front end than after end. There were no apparent relationship between the abundance of microorganism in substrate and the removal efficiency of COD, TN and TP.The metabolic properties and functional diversity of the microbial community in four types of CWs were analyzed by the sole-carbon-source utilization profiles using ECO Biolog microplates. There were four sampling points in HFCWs substrate bed（ upper forepart layer, lower forepart layer, upper back-end layer and lower back-end layer） and two sampling points in VFCWs substrate bed（upper layer and lower layer）. The results showed that there was apparent difference among the microbial communities in different substrates, as well as in different seasons. In autumn, the variation of average well color development explained that the metabolic activity of lower forepart layer substrate microorganisms was remarkable lower than that of other three substrate microorganisms in HFCWs, as well as the activity of upper layer was lower than that of lower layer in VFCWs. But in the next spring, the activity of lower forepart layer in HFCWs and upper layer in VFCWs increased a lot. Among the 31 carbon sources, polymers, carbohydrates, carboxylic acids and amino acids had greater rates of color development for all substrates. But the metabolic capability of amines and phenolic compounds was much different. The results of principal component analysis showed the different structures and metabolic properties of different substrate microorganisms. The results of functional diversity indices showed that there was greater variance in different substrate microorganisms. In HFCWs at autumn, the significance of richness indices difference among four substrates was smallness. The diversity and evenness indices of lower forepart layer substrate microbial community were greater than that of other three substrate microbial community, as well as the the diversity and evenness indices of upper layer was lower than that of upper layer in VFCWs. At the next spring, the diversity and evenness indices of lower forepart layer substrate microbial community in HFCWs increased much more, as well as the upper layer substrate microbial community in VFCWs. The spacial distribution analysis of COD, TN and TP in the substrate showed that there were different characters at different pollutants.In HFCWs, regardless of seasons and plants, the concentration of COD, TN, TP decreased along the beds and increased from upper layer to lower layer. However in VFCWs, the concentration variance of COD was small and uncertain in all year. The concentration of TN decreased from upper layer to lower layer in warm seasons, but the concentration stabilized from 1/2 section to effluent in cold seasons. The concentration of TP decreased from upper layer to lower layer in cold seasons, but the concentration stabilized from 1/2 section to effluent in warm seasons. The removal rate of COD, TN and TP in the beginning 1/4 section was in the highest flight to the total removal.The special distribution of NH₄⁺-N was not consistent with that of COD, TN and TP. The concentration of NH₄⁺-N increased from influent to 1/4 sections in CWs and then decreased along the beds. But the concentration of effluent was higher than influent in the most time of a year. The removal of NH₄⁺-N was weakness. The result showed that the concentration of NH₄⁺-N in effluent was below 0.57mg/l （it could meet the level of class II of surface water quality according to GB3838-2002）.The pilot-scale treating system carried out the compound system engineering applying BF bed-constructed wetlands based on small-scale test result. Result indicated that BF bed was able to reduce the concentration of ammonia nitrogen, improve water-entering resolving of oxygen contents of CWs influent. It had a good effect to strengthen the follow-up CWs purification. The pilot-scale CWs had better treatment performance to the BF bed effluent. COD concentration of the CWs effluent could meet the level of class III-IVof surface water quality according to GB3838-2002, TN concentration could get the standard of classIV, TP concentration could stabilized at the level of class II. The removal rate of Chla of CWs increased from 75% to 90% in the last half year. The removal effect of NH₄⁺-N was not satisified. But the effluent concentration of NH₄⁺-N was below 1.0mg/l （it could meet the level of class III of surface water quality according to GB3838-2002）. Integrated vertical constructed wetlands （IVCWs） had a comparative purifying effect to HFCWs under higher loading situation （the HRT of IVCWs was 1/3 less than HFCWs）. In a word, the BF bed-CWs purifying system could improve the quality of supplemental water under the loading of 300t/d. With the CWs being mature （the growth of wetlands plants, the maturity of the wetlands substrate microorganism community）, the effect of purifying is hopeful to improve in the future.更多还原