Evsel atıksuların havasız biyolojik arıtımı üzerine bir araştırma

Abstract

Bu çalışmada, havasız çamur yataklı reaktörlerde evsel aüksularm antılabilirliği araştırılmıştır. Evsel atıksu olarak İTÜ Kampüsündeki bir kanalizasyon bacasından alman atıksu kullanılmıştır. Birinci bölümde, tezin konusu ve kapsamı açıklanarak önemine değinilmiş ve tez kapsamında araştırılan konular belirtilmiştir. İkinci bölümde havasız arıtma prosesi ile ilgili genel bilgiler verilmiş ve bu sistemlerde işletmeye alma ile proses kontrolünde dikkat edilecek hususlar özetlenmiştir. Tezin üçüncü bölümünde havasız arıtma sistemleri tanıtılmış ve yaygın olarak kullanılan havasız sistemlerin birbirlerine göre üstünlükleri tartışılmıştır. Bunlara ilaveten havasız çamur yataklı reaktörlerin boyutlandınlmasmda dikkat edilecek hususlar verilerek, havalı ve havasız sistemlerin maliyet mukayesesi yapılmıştır. Dördüncü bölümde havasız arıtma sistemlerini modellemede kullanılan kinetik modellere ait bilgiler verilmiştir. Tezin beşinci bölümünde, evsel atıksu ve eşdeğeri sentetik aüksularm havasız arıtımı konusunda yapılan çalışmalar verilmiş ve çalışmalardan elde edilen sonuçlar değerlendirilmiştir. Tezin altıncı bölümünde; deneysel çalışma planı, çalışmada kullanılan reaktör ve atıksuyun özellikleri ile reaktörün hidrolik modelini belirlemede faydalanılan modellerle ilgili teorik bilgiler verilmiştir. Yedinci bölümde deneysel çalışmadan elde edilen sonuçlar tartışılmış ve reaktörün hidrolik özelliklerini belirlemek üzere LiCİ ile yapılan deneyler ve elde edilen sonuçların Geliştirilmiş Gama Dağılım Modeline uygulanması verilmiştir. Hidrolik bekleme süresi 6-24 saat arasında değiştirilmek suretiyle üç yıllık bir sürede boyunca evsel atiksularm havasız çamur yataklı reaktörde antılabilirliği ile ilgili deneysel araştırma yapılmıştır. Reaktör 330 ile 830 arasında beslenmemiş diğer zamanlarda ise toplam debi yarımşar saatlik sürelerde kesikli olarak verilmiştir. Bu tür besleme düzeni özellikle küçük yerleşim birimlerinden gelebilecek debileri temsil etmek üzere seçilmiştir. Çalışma süresinde ortalama olarak % 67 - 78'lik KOİ giderimlerine ulaşılmıştır. Çalışma verileri İkinci Mertebe Kinetiğine ve Çok Değişkenli Lineer Korelasyon Modeline uygulanmış ve sonuçların bu modellere oldukça yüksek bir korelasyonla uyum sağladığı görülmüştür. Bu çalışmalara ilaveten, arıtılmış evsel atıksuda N ve P gidermek amacıyla yapılan kimyasal arıtma deneyleri tablolar ve şekiller halinde verilerek, elde edilen sonuçların değerlendirilmesi yapılmıştır. Sekizinci bölümde ise deneysel çalışmadan elde edilen sonuçlar özetlenmiş ve konuyla ilgili gelecekte yapılacak çalışmalarda dikkat edilmesi gereken hususlar belirtilmiştir.

Successful application of anaerobic systems on treatment of medium and high strength industrial wastewaters by modern reactors put the agenda on the anaerobic treatment of domestic wastewaters. The aim of this investigation is to observe the treatment potential of domestic wastewaters in the anaerobic upflow sludge blanket reactors(UASBR), determination of hydraulic and kinetic factors affecting the performance of the system and to develop a compact treatment system for the regions of Turkey with a mild and warm climate by using the experimental data from the study. The experimental studies have been continued on an anaerobic upflow sludge blanket reactor about three years in laboratory conditions. LITERATURE REVIEW The anaerobic treatability studies of domestic wastewaters can be separated into three groups with respect to the characteristics of the wastewaters. These are: - Purely domestic wastewaters - Municipal wastewaters including industrial effluents - Low strength synthetic wastewaters While COD removals of 70 to 90 % are observed for anaerobic treatability studies of domestic and low strength synthetic wastewaters, lower COD removal efficiencies may be observed for domestic wastewaters containing industrial effluents due to possible inhibitory substances. Anaerobic treatment of diluted wastewaters has created some problems due to low growth yields of anaerobic microorganisms. The growth yields of methane bacteria controlling the last phase of the anaerobic treatment process are very low and then- doubling time is about 3 to 5 days. The minimum doubling time is occurred only when the organic carbon concentrations are high and micro and macro nutrients sufficient. As a results, the active biomass wash-out rate must be lower than the synthesis rate to maintain the stability of the reactor. That's why, the wash out of biomass from the reactor is minimized and very short hydraulic retention time are not applied. -XV- While scientific and technological information level on anaerobic treatment were increased, the anaerobic treatment process used for high strength organic wastewaters, now has also been used for diluted wastewaters. Anaerobic treatment can be considered as a compact, simple and economic process and it can be easily applied for the effluents from restaurants, hotels and holiday resorts. Even in the septic tanks operated at appropriate conditions, considerable levels of COD removals can be achieved. Using of lower volumetric loadings in anaerobic treatment of partially soluble waste waters are more appropriate. In fact, complex wastewaters containing high amount of particulate organics can also be treated in granular upflow sludge blanket reactor. In the UASBR having flocculent sludge, influent suspended solids cause to decrease in the methanogenic activity of the biomass. That's why, there is no guarantee for the granulation process in the treatment of wastewaters with high amount of suspended solids. The following conclusions can be given considering the anaerobic treatability results from the related literature so far:. In the anaerobic treatment of domestic wastewaters, COD removals of 30 to 85 % are possible depending on the share of industrial effluents in the municipal wastewaters.. Domestic wastewaters can be treated anaerobically at high organic and hydraulic loading rates. Hydraulic retention times as low as one hour are applicable.. Excess biological sludge produced in the anaerobic treatment of domestic wastewaters is very low compared to the aerobic treatment processes.. Anaerobic treatment systems, particularly UASBR' s, are very feasible and attractive alternatives for the treatment of domestic wastewaters.. UASBR's are simple, compact and low cost reactors, so they are easily applicable.. The removal of nutrients in the effluent from the anaerobic treatment systems may be required especially in the case of discharging enclosed water. MATERIAL AND METHODS Anaerobic treatability studies were performed in a lab-scale UASBR with a net volume of 10.35 1. The reactor was fed intermittently during the course of the study. Characteristics of the feed to the reactor were given in Table 1. The same reactor had previously been used in another anaerobic treatability study carried out by synthetic low strength wastewater and the sludge in the reactor was used as the seed for this study. Such an opportunity reduced to the period of start-up. The reactor was not fed between 330 to 800 am. Total daily feed was pumped to the reactor in the rest of the day within 30 minutes intervals for each hour. Such a feeding regime was applied to simulate the flowrate of the domestic wastewaters from the small communities. Operating conditions in this study were given in Table 2. -XVI- Table 1. Characteristics of Domestic Wastewater Table 2. Operation Conditions of Experimental Studies EXPERIMENTAL STUDIES IN THE MODEL REACTOR The domestic wastewater from a sewer system at ITU campus was used as the feed. The subjects following were investigated in this study: - Effects of intermittent feeding on the reactor performance - Determination of hydraulic flow characteristics of the reactor by the tracer experiments. - Determination of relations between the hydraulic parameters and the treatment process responses. - Investigation of reactor biomass characteristics - Effect of temperature on the process efficiency - Determination of kinetic parameters for the treatment process. - Evaluation of final effluent and determination of additional treatment requirement. RESULTS AND DISCUSSION The study was continued about 3 years and carried out in three different stages. Characteristics of the feed were different in each stage as indicated in Table 3. The reactor was not feed between the days 63 to 273 and 378 to 654 from the start-up due to a heavy water shortage in Istanbul and summer holiday in the University. The reactor was operated at 7 different steady state conditions during the study and the related results were summarized in Table 4. Evaluation of the experimental study results were presented below. -XVII- Organic Trading Rate and COD Removal Organic loading rates were varied in the range of 0.2 to 1 kg COD/m3-day during the study. Influent and effluent COD's and related COD removals for each steady state operating conditions were presented in Table 5. Steady state COD removals varied in the range of 67 to 87 % in this study. Effluent COD's were generally 50 mg/1 at steady state operating conditions. However, effluent COD's have increased as the influent COD's raised and this resulted in relatively low COD removals. Effluent suspended solid concentrations were generally less than 50 mg/1. Temperature. pH and Alkalinity The two most important indicators of the stability of anaerobic reactors are pH and alkalinity. These parameters were continuously measured during the study. The physico-chemical stability of the reactor was continuously monitored by measuring pH and alkalinity in the reactor. The reactor pH and alkalinity ranges were between 6.4 and 9.7, 100 to 750 mg/1, respectively in the whole study period. Concentration of ammonia were increased when pH raised to about 9 for a particular time period of experimental studies, and this resulted in a substantial reduction of reactor performance. The reactor temperatures were in the range of 30+5 for 5 steady state operating condition while 25+5 °C for the remaining two steady state. Table 3. Characteristics of Domestic Wastewater as Different Periods of Along the Study ( ) Temperature belong to every steady state -xvin- Table 4. Steady State Operating Parameters Reactor Biomass Concentrations Biomass concentration measurements carried out on samples with drawn from different heights in the UASBR revealed that the distribution of the biomass along the height of the reactor was not uniform. The amount of biomass was gradually decreased as the volumetric or hydraulic loading rate increased. However, the performance of the reactor was not affected significantly even for average biomass concentrations as low as 6000 mg/1 and the stability of the treatment process could be maintained. Hydraulic Modelling bv Tracer Studies Tracer experiments have been performed with LiCl in order to determine the coefficient of dispersion. LiCl was injected instantaneously into the entrance of the reactor so that the average concentration of Li in the reactor was about 10 mg/1. The Li concentrations at the exit of the reactor were measured for a duration of about three times of the hydraulic retention time in each experiment. The results from the experimental study were used to Gamma Distribution and Modified Gamma Models. It was observed that these results were more suitable for modified Gamma Distribution Model. The parameters of the cited model were determined as J3 = 0.9, T = 0.1, p =2. Considering these results, the anaerobic upflow sludge blanket reactor system can be hydraulically thought as two completely mixed and one plug flow reactor in series (Figure 1). When diluted wastewaters, like domestic wastewaters, are treated in anaerobic upflow sludge blanket, generally high hydraulic loadings rate are applied. The findings obtained from Modified Gamma Models show that the UASBR' s can be considered as two completely mixed reactor in series for practical applications. -XK- i o O 1 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Dimensionless time 5.5 6 6.5 B - 0.9 ; T. 0.1 Q « 31 i/day 0.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 Dimensionless time 8 « 0.9 ; T ? 0.1 Q ? 20.7 l/day 2 2.5 3 3.5 4 4.5 5 5.5 6 6.6 7 Dimensionless time Figure 1. The Comparison of the E(0) - 6 Values Obtained from the Experimental Study and Modified Gamma Models Curves. -XX- KINETIC EVALUATION Experimental data from the study were applied to Multiple Variables Linear Correlation and Lineralized Second Order Substrate Removal Kinetics Models. It was observed that both models were perfectly fit to the experimental results. The reaction rate coefficient of the second order COD removal kinetics was determined as k2 = 0.217 day"1 (= 0.009 hour"1) in this study. This value was lower than k2 = 1.665 day"1 (= 0.069 hour1) which was found for the synthetic wastewater with COD's less than 700 mg/1. This result shows that domestic wastewater might be treated with a rate of seven times lower than that of the synthetic wastewater. Coefficient of correlation for measured and predicted values by the models is 0.928 and the effluent COD from the UASBR can be defined the following equation: Smodel ~ So' & e 1.3466 - 0.002 ) The biodegradability of the wastewater can be identified by linearized model coefficient^). The values of model coefficients are increased as the biodegradability of the wastewater is decreased. POST TREATMENT OF THE UASBR EFFLUENTS Since the concentration of nitrogen and phosphorus in the effluent from the UASBR are higher than the discharge standards, an additional post treatment may be required. Chemical purification using Ca(OH)2, NaOH and cement with and without aeration were applied to the effluent from the UASBR for this purpose. Chemical post treatability studies with Ca(OH)2 and NaOH were not successful for ammonia removal. On the other hand, chemical treatment with Ca(OH)2 together with 2 -XXI- hours of aeration gave satisfactory results. The optimum dosage of Ca(OH)2 was determined as 300-350 mg/1 (which corresponds to a pH of 11) in the experiments (Table 6). Average removals of TKN, NH3 and Total P were 65, 74 and 90 % respectively indicating such a treatment is feasible for practical purposes. Total coliform concentration of raw domestic wastewater was about 0.9- 1Ö7 coli/100 ml. This value was decreased to about 0.2-2- 106 in the UASBR effluent and to 0.2- 105 after chemical post treatment with Ca(OH)2 plus 2 hours of aeration. As a result of this findings, it can be concluded that total coliform removals in the anaerobic and chemical post treatment stages are about 80-94% and 75-98% respectively. Table 6. Results of Additional Treatment of Anaerobic Effluent (Chemical Treatment and Aeration) CONCLUSIONS The results at this study can be summarized as follows: - Hydraulic characteristics of the UASBR' s can be defined by the modified Gamma Distribution Model. Tracer study results have shown that, the UASBR used in this study can hydraulically be considered as a completely mixed and one plug flow compartments in series. The volume of the plug flow compartment constitutes about % 10 and the whole system may be considered as 2 completely mixed reactors in series for the practical applications. - Domestic wastewaters can be anaerobically treated in mesophilic UASBR' s with COD removals about 70% without any chemical addition. - The concentration of the steady state biomass in the UASBR' s are balanced about 20 000 mg/1 for domestic effluents and the excess biological sludge production is almost negligible. - The biomass in UASBR' s treating domestic effluents are flocculent in type not in granular structure. -xxn- The reactor effluent suspended solids concentration are very low and rarely exceeds 50 mg/1. COD' s in the UASBR effluents are generally about 50 mg/1 at steady state and rarely reached to 70 mg/1 for transient conditions. Concentrations of ammonia, phosphorus and total coliforms in the UASBR effluent are not satisfied discharge standards to surface waters. Chemical post treatment by Ca(OH)2 with 2 hours of aeration provide significant reductions both in ammonia and phosphorus. Anaerobic treatability study results from the UASBR perfectly fit the second order multiple substrate kinetic and multivariate linear correlation models at steady state operating conditions.