Substrate removal in long sewer lines

e

WOl. Sci. Tech. Vol. 31. No.7, pp. 213-218, 1995.

Pergamon

Copynght@ 1995 lAwQ

0273-1223(95)00338-X

Printed in Great Britain. All rights reserved. 0273-1223195 $9'50 + 0-00

SUBSTRATE REMOVAL IN LONG SEWER LINES Adem Ozer* and Ersin Kasirga** * Dokuz EylUl University. Engineering Faculty,Environmental Engineering Department, Bomova-IZMIR ** Engineering-Science,lnc. 100 West Walnut Street, Pasadena, CA 91124, USA

ABSTRACT Wastewaters generated from large dwelling areas are collected and disposed of by means of long sewer uunk lines. The wastewater treatment potential of sewers is not a novel phenomenon. however. it has not been thoroughly investigated in the past. In this study. possible treatment efficiencies are estimated in the presence of sufficient oxygen. Suspended and attached growth kinetics are studied to explore the significance of aeration in sewer uunks. Although there are several eXlstmg models describing consumption mechanisms of soluble substrates and reaction rates. a new method is investigated in this paper for shortening the parameters used in biological rate equations. Soluble substrate consumptions of suspended growth microorganisms are taken into consideration by means of an experimental approach and they are estimated by using respiration rates. Relationships between respiration rate, substrate utilization rate and biodegradable organic matter concentrations are observed during the experimental program. An empirical relationship which provides the utilization rate in correspondence with diminishing substrate concentrations along a sewer line is developed based upon the expertmental results. Expected treatment effiCIencies are determined considering utilizatton rates of attached and suspended growth microorganisms.

KEYWORDS Biofllm; suspended growth; substrate removal; treatment in sewer trunk lines. INTRODUCTION Long sewer lines have a potential for treatment which is usually attributed to the activities of attached and suspended growth microorganisms. Existing models describing this phenomenon are investigated as well as searching for better ones. A heterogeneous system is used to develop a model which takes into account the substrate flux rate through the biofilm surface. as well as diffusion and biochemical reactions. The model does not consider the liquid phase and is based upon the assumptions that substrate transfer is characterized by molecular diffusion and substrate utilization converges to a Monod equation. The equation of continuity which is written for the substrate changing between the liquid and the solid phases in biofilm. is used for developing the model. A second order non-linear differential equation. which is derived from the equation of continuity. has such boundary conditions that substrate concentration at the liquid phase-biofilm interface would be the interfacial substrate concentration and the slope of the substrate profile is zero at the solid phase-biofilm interface. It was attempted to obtain general solutions by making variables dimensionless in the differential equation so that the substrate profile of the biofilm could be found when the equation was solved. Atkinson and How (1974) showed how to determine the parameters for the biological rate equation by utilizing experimental results performed on an inclined plate biofilm reactor. Sansarci (1980) developed a 213 JWST 31-7-P

214

A. CZER and E. KASIRGA

method to detennine kinetic parameters by means of flux-concentration observation pairs. In this study. these p~eters are calculated for the biofilm. Observed and calculated flux rates are shown as Nob and Nca respecttvely. The biofilm substrate utilization rate is independent of the thickness when the biofilm grows larger than a certain thickness. Atkinson and Howell (1975) defmed this value as the sufficient biofilm thickness. The flux rate for the sufficient biofilm thickness can be found depending upon interfacial concentration and model parameters. This flux rate of Ns is defined as the flux rate of a biofilm with infinite thickness and the biofilm thickness which consumes a certain percentage of this flux is called Mideal by Atkinson and Ali (1978). Commencing from the Modulus of Thiele. flux consumptions for the sufficient biofilm thickness are detennined. Levenspiel (1972) defined that should the Modulus of Thiele be less than I. diffusional resistance can be omitted or should it be more than 1. a strong diffusional resistance is mentioned for a reaction. Harremoes (1978) investigated concentration changes in biofilin based upon the assumption of zero-order reaction rate. Nevertheless. Czer (1982) pointed out that concentration at the biofilm-solid phase interface would be zero for a biofilm thickness corresponding to the Modulus of Thiele of 1. where flux rate is given as N I. Although the Modulus of Thiele approach for the biological rate equation is slightly different. the concentration at the biofilm-solid phase interface is expected to be zero. In order to search for the validity of this. flux rates of N 1 were calculated by using the biological rate equation. These rates were then compared to the flux rates of Nideal obtained for the biofilm of ideal thickness and to Ns values. This comparison has shown that NI fluxes define the sufficient biofilm thickness more accurately than N ideal values do. since N 1 fluxes demonstrate much better fitness with Ns fluxes even for small interfacial concentrations. Nideal values. however. showed greater deviations than N 1 values for small interfacial concentrations.

EXPERIMENTAL STUDY In order to determine the kinetic parameters of attached and suspended growth microorganisms. extensive experimental studies were performed. Wastewater was aerated in a 10 illlindrical container with a conical were collected at bottom which was operated as l!Jw,tcJU~!lctor. Total and filtered wastewater samples intervals of 1 to 2 hours, were analyzed for total and filtered chemical oxygen demand (COD) values. As soon as reductions in the COD values were observed. respiration rates were determined at 10 to 15 minutes time intervals. An equation to determine the substrate utilization rate of suspended growth microorganisms was derived by evaluating respiration rates with total and filtered COD samples. Parameters for the biofilm, however, were calculated by means of a batch biofilm reactor which consisted of a sewer pipe with a diameter and length of 9 cm and 297 cm respectively. Wastewater kept at a constant temperature was recycled by means of a pump. The slope of the channel was 0.0 II and velocity of the wastewater was about 0.5 mls. Under these conditions. a half-full flow was realized with a biofllm surface area of 4.200 cm 2. The entire system was kept in the dark in order to simulate the prevailing conditions in a real sewer. After formation of the biofllm on the inner surface. experimental studies were started. Wastewaters used in the reactor were changed daily and biofllm growth was observed. After three weeks. biofilm thickness reached a stable stage. By analyzing the wastewater samples taken from the reactor during the experimental period, respiration rates, and total and filtered COD values were recorded in time.

whlcl1

Experimental studies carried out for the samples taken from the aeration tank demonstrated that substrate utilization of suspended growth microorganisms can be determined with relation to respiration rate measurements. Therefore, biofilm utilization rates were found by subtracting the suspended growth microorganism utilizations determined by respiration rate measurements from total substrate utilization rates. Since the biofilm surface area was known. utilizations were obtained as flux rates.

215

Substrate removal in long sewer lines

EVALUATION OF THE RESULTS Aeration tank experiments and determination of the respiration rates Variations of the respiration rate curves obtained in this study have shown similarity with the ones developed by Pomeroy and Parkhurst (1972). These curves fIrst increase and after reaching a peak a steady decrease takes place. Perhaps it is to be expected that curves which represent long flow durations would reach the peak faster than the curves of fresh wastewaters.

In order to calculate substrate utilization rates of suspended growth microorganisms. relatively simple equations are used by considering the experimental results. Respiration rate is mostly affected by substrate concentration and mass of microorganisms as well as yield coeffIcient and other physical chemical effects. A simple relationship is developed rather than a complex one which consists of several kinetic parameters. Relationships between respiration rate. unfiltered COD (microorganism mass) and substrate concentration were investigated separately. In order to fInd out the nonbiodegradable part of the wastewater. a method which was proposed by Shredder (1977) was used. In compliance with this method. wastewater was mixed with activated sludge at a certain ratio and aerated in a batch reactor. Decrease of the filtered COD value was observed and it was found that this value became constant after a certain period. This constant value represents the nonbiodegradable portion of the domestic wastewater. In this study. nonbiodegradable filtered COD values ranked between 76 - 82 mgll. Substrate concentration introduced as S was obtained by subtracting the mean constant COD value of 80 mgll from the filtered COD value measured during the experiments. A linear relationship between the observed respiration rates and the biodegradable portion of filtered COD values was investigated and a correlation coeffIcient of 0.813 was obtained. This equation is given below.

RES

=0.0077 + 0.00064 S

(1)

where RES (mgll min) and S (mgll) represent the respiration rate and the biodegradable part of the filtered COD respectively. After investigating the linear relationship between mass of microorganism and respiration rate. a correlation coeffIcient of 0.76 was found. It was also imperative to know the change of microorganism concentrations along the pipe in order to use this equation. Therefore. it was decided that Equation I. which shows the relationship between the substrate concentration and the respiration rate. would be used to predict the substrate utilization of the microorganisms. Experiments with the bjofilm reactor and determinatjon of the kinetic parameters Kinetic parameters of Atkinson and How's (1974) biofllm model are k •• k2 and k 3. The least-squares method was applied to fInd out these parameter values. The curve was the N 1 flux equation and the data are flux• concentration pairs obtained from the experiments. Observed and biodegradable portions of the COD values were used in the calculations. The biodegradable flItered COD values provided a better fIt compared with the observed flItered COD values. Kinetic parameters of k l=0.043 s-2. k2 85 cm- I and k3 = 21.76 cm 3/mg were found. Although a suffIcient biofllm thickness of I mm was calculated. the biofilm thickness always ranged between 2 and 3 mm in the reactor. Figure I demonstrates the relationship between biological degradable COD versus observed and calculated flux rates from the N I flux equation.

=

The required channel lengths for a certain concentration decrease are calculated by applying the kinetic parameters to Equation 2.

A. OZER and E. KASIRGA

216

(1 + 2 Bo )111 - 1 (1 + 2 Bo )111 + 1 (1 + 2Bo)II1-(1 +2BJlI2 + I n - - - - - - - - = - Z

(2)

(1 + 2 Bi )111 - 1

(1 + 2 Bi )111 + 1

where Bi and BO show the influent and the effluent dimensionless average liquid phase substrate concentrations respectively while Z is the dimensionless length of the reactor. These dimensionless values can be put in dimension by using the model parameters. The diffusion coefficient of the biofilm was chosen as 0.5 cm 2/d in the calculations.

O.S

Nob, Nca(N 1 )

Q5

10- (mg/cm 2s)

• •

•

4

0.4

0.3 0.2 01

• • 0.02

• 004

0.06

008

•

• •

01

0.12

0.14

•

0.16

018

S(mg/cm 3 )

Figure 1. Biological degradable COD vs observed and calculated flux rates.

DEGREE OF TREATMENT ALONG THE PIPE Change of the substrate concentration by the activities of biofilm and suspended growth microorganisms along the canal can be predicted using Equations I and 2 together, when the dissolved oxygen concentration in wastewater is increased close to the saturation value. Equation 2 provides the required length of reactor (pipe length) to obtain a desired effluent concentration for a given initial concentration. Utilized substrate amount per unit length of the canal could be found assuming that the biofilm reaction rate is constant for a small range of concentrations such as 170 - 168 mg/l COD. Utilization of suspended growth microorganisms could be found by the use of Equation I for the same range of concentration per unit length of pipe. Therefore, the term of length should be used instead of time in the reaction rate and this can be done easily by considering the average flow velocity. Total utilized substrate per unit length of the pipe was obtained at a certain concentration range as a summation of the biofilm and the suspended growth microorganism utilizations. Required pipe length for a small decrease of concentration such as 170 to 168mg/l COD was found with this total reaction rate. The same procedure was also repeated for concentration ranges of 168-166. 166-164. and 164-162 mg/l COD in order to find out the required lengths. By the use of this approach, length can be found for a desired effluent concentration. Nevertheless. increasing the concentration step from 2 to 20 mg/l COD did not affect the results, which enabled one to do the calculations without the use of a computer. In Equation 2, which was used to estimate the biofilm utilizations. concentrations and lengths are dimensionless. These values were converted into diml!nsional

217

Substrale removal in long sewer lines

forms by using the values of kinetic parameters obtained during the experimental studies. In order to calculate the utilization of suspended growth microorganisms. 0.5 mls flowing velocity and traveled length per minute in kilometres were written in Equation I to obtain the utilization in mgII km. Results obtained by this means are shown in Table I. Table I. Required pipe lengths in km to obtain given treatment efficiencies for various diameters Diameter Treatment efficiencies and remaining concentrations as mgll COD 12% 24% 35% 47% 59% 71% 82% 94% (cm) 90 70 50 30 10 130 110 150 20 40 60 80 100 140 180 200

0.8 1.4 1.9 2.3 2.6 3.0 3.4 3.5

1.7 3.0 4.0 4.8 5.5 6.4 7.1 7.4

2.8 4.8 6.4 7.7 8.7 10.3 11.4 11.9

3.9 6.9 9.2 11.0 12.5 14.7 16.4 17.1

5.3 9.3 12.4 15.0 17.0 20.1 22.4 23.3

7.1 12.4 16.6 20.0 22.7 26.9 30.0 31.3

9.5 16.8 22.4 27.0 30.7 36.4 40.7 42.4

14.3 25.0 33.3 39.9 45.3 53.5 59.6 62.1

In Table I. concentration changes in a sewer trunk with no connections is shown for different diameters. In this Table. it is assumed that the initial concentration of the wastewater and respiration rate were 170mgll COD and 7 mglI.h respectively. In this study. however, 366 mgll COD concentration for the biodegradable part of the filtered COD values and respiration rates of 31.8 mglI.h are found. This demonstrates that the higher the substrate concentrations. the higher the reaction and respiration rates which eventually result in the adequacy of shorter lengths to obtain equal concentration decreases. For example, required length to provide a decrease from 360 to 350 mgII COD would be shorter than the length to provide a decrease from 170 to 160 mg/l COD. Although the initial conditions where respiration rates show an increase are omitted in the study, the yield of suspended growth microorganisms could reach significant levels. However, the assumptions on which Table I is based neglect the factors which may shorten the pipe lengths due to the complex rapid reactions at the beginning of the experiments. These could also be related to some other parameters besides the fact that it was aimed to be on the safe side when the calculations were undertaken. It is also expected that sewerage connections would increase the wastewater concentration and increase respiration rates. Pomeroy and Parkhurst (1972) obtained such conclusions. In order to use sewers as a treatment alternative, provisions of aerobic conditions and certain flow time or pipe length are prerequisites. Sometimes, flow durations might be so long in the pipes in large cities that maintaining long flow durations are not favored from the hydraulic point of view but this may bring about some degree of treatment Natural oxygenation of wastewater is usually provided by surface aeration, fall in manholes and turbulent effects created on the connections. Potential energy between the head and the tail of the pipe is consumed during the aeration. The available energy is constant for a pipe and additional oxygen intrusion may be needed in case of a lack of energy. Oxygen concentrations maintained in wastewater have an economic significance and substrate utilization rates of suspended growth microorganisms are independent from oxygen concentrations down to very low values. However, this cannot be generalized for biofilm. Examples show that liquid phase oxygen concentrations do not affect biofilm substrate utilization rates (Harremoes, 1978; Atkinson and Williams. 1971). Nevertheless. it is also expected that the liquid phase oxygen concentration would affect the active biofilm thickness or at least would become the limiting factor. Due to this reason, experiments performed in this study should be reproduced for low oxygen concentrations as well. CONCLUSION

In this study, the wastewater treatment potential of long sewer trunks under natural and artificial aeration conditions was investigated. Substrate utilization rates of attached and suspended growth microorganisms

21S

A. OZER and E. KASIRGA

were studied separately and their contributions to the treatment were considered. Applying the developed method, required sewer lengths which flow half full and have a velocity of 0.5 m/sec were calculated for different diameters to obtain desired treatment efficiencies. In order to consider the biofilm efficiency, several models are available in the literature and the one which is given by Atkinson and Daoud (1968) was used. This model provides substrate utilization as flux and consists of three parameters. A new method which utilized results of batch experiments was developed to determine the values of these parameters. The relationship which allows one to calculate the flux independently from the biofilm thickness is derived for the biofilm with sufficient thickness. This relationship appears to be in agreement with the other relationships given in the literature. It produces better results than the similar ones for low substrate concentrations. Biofilm model parameters ofkl 0.043 s-I, k2 = 85 em-I and k3 21.76 cm 3/mg were found respectively.

=

=

It was shown that the substrate utilization rates of suspended growth microorganisms could be estimated using the respiration rates. Experiments have shown that the substrate utilization rate converges at the respiration rate for sewage after a certain time period. It was found that biofilm is dominant for small diameters while suspended growth microorganisms play a more important role in the treatment mechanisms of larger sewers. Based on this result, it is thought that the oxygen concentration of the wastewater should be high enough to provide an adequate amount of oxygen to the biofilm in sewer trunks with small diameters. However, since the substrate utilization rates are independent of oxygen concentration, maintaining low oxygen concentrations is expected to result in certain savings for large sewer trunks. Further studies of the relationship between the biofilm substrate utilization rates and the oxygen concentrations should be performed and detailed research is also needed to explore the best technology for oxygen supply into sewers. REFERENCES Atkinson, B., Daoud, 1. S. (1968). The Analogy Between Microbiological Reactions and Heterogenous Catalysis. Trans. o/the Inst. o/Chern. Eng., v.46. pp. T19-T24. Atkinson, B., Williams, D. A. (1971). The Performance Characteristics of a Trickling Filter with Hold-up of Microbial Mass Controlled by Periodical Washing. Trans. o/Inst. o/Chem. Eng .. v.49, pp.2 I5-224. Atkinson, B., How, S. Y.(1974). The Overall Rate of Substrate Uptake by Microbial Films. Part 11- Effect of Concentration and Thickness with Mixed Microbial Films. Trans. of the Ins. a/Chern. Eng.,v.52. Atkinson, B., Howell, J. A. (1975). Slime Hold-up, Influent BOD and Mass Transfer in Trickling Filters. ASCE, Journal of the Environmental Engineering Division, v.lOl, n.EE4 (August 1975),pp.5SS-60S. Atkinson, B., Ali, A. R. (1978). The Effectiveness of Biomass Hold-up and Packing Surface in Trickling Filters. Water Research, v.12, pp.l47-156. Harremoes, P. (1978). Biofllm Kinetics. In: Water Pollution Microbiology, v.2, Wiley and Sons, Inc. New York, pp.71-I09. Levenspiel, O. (1972). Chemical Reactions Engineering. New York, John Wiley and Sons, Inc. Ozer, A. () 982). The PUrification Potential 0/ Sewers Under Aerobic Connilitions. Izmir. Ege University, Civil Eng. Fac•• Dept of Env. Eng., Doct. Thesis. Pomeroy, R. D., Patkhurst, J. D. (1972). Self Purification in Sewers. Advances in Proc. 6th Int. COnference. Jerusalem Water Pollution Research, Oxford, Pergamon Press. Per. Press. 6th International Water Pollution Research Symposium- June 18-23. 1972. Sansatci. H. (l980).lnvestigation a/Substrate Removal MechaniSms in Biofilms. lTV. Civil Eng. Fac. Doctoral ThesiS. Schroeder. ED (1977). Water and Wastewater Treatment. Tokyo. McGraw-Hili. Kogakusha Ltd.