|
Biofilm, Volume 5, Paper 1 (BF00001) 2000 Analyzing Flow Characteristics and Influence of Biological Growth on Dispersion in Aerated Submerged Fixed-Film Reactors (ASFFR) Ramin Nabizadeh*, Ali Reza Mesdaghinia, Simin Nasseri, Amir Hossain Mahvi and Mahmood Shariat Tehran University of Medical Sciences, School of Public Health and Institute of Public Health Research, Department of Environmental Health Engineering, P.O.BOX: 6446-14155 Tehran, Islamic Republic of IRAN * corresponding author email: r_nabizadeh@yahoo.com Received:March 21st 2000 Code Number: BF00001 ABSTRACT The Aerated Fixed Activated Sludge System is one of the biological systems used to treat different types of wastewater containing organic substances. This paper presents the results of the preliminary phase of a study to determine the performance pattern of such a system treating petrochemical wastewater on a pilot scale. As the hydraulic characteristics of such systems are important factors affecting the behaviour, design and interpretation of data from the reactor, the hydraulic regime of the system was determined using a pulse input of Rhodamine B into the system. Various air supply and water flows were used as independent variables during the tracer study. Experimental data in 16 different conditions were collected and dispersion number (d/ul) in each condition was determined using a computer program which was developed in this research. Furthermore, the statistical patterns were used to evaluate the fit of data in each condition with the ideal completely-mixed model. The results showed that, in a wide range of air supply and water flows, the completely mixed condition is achievable in this system; biological growth of film at different hydraulic loading rates did not change the pattern of mixing significantly. The content of this article is so organized that the complete methodology and computative approaches for performing flow pattern characterization in water and wastewater treatment systems are easily available. Key Words : fixed activated sludge, aerated submerged fixed-film reactors, hydraulic regime, dispersion number, mixing INTRODUCTION Aerobic biological wastewater treatment processes are extensively used for removing organic substances. Aerated Fixed Activated Sludge, also known as Aerated Submerged Fixed-Film Reactor (ASFFR) in recent publications, has been used for more than 60 years, and now is an alternative for treating various kinds of industrial wastewaters including petrochemical industries. Stability, high efficiency in COD removal, long cell residence time, independence of the system to the performance of secondary settling tank, short-term and simple start up and ease of operation due to omitting a sludge recirculation line, are the special specifications of this system which make wastewater treatment simple. As uniformity and effective contact between substrate and biomass in such a system play an important role in its performance and behaviour, it is necessary to determine the hydraulic characteristics in practice. It should be noted that the completely mixed condition is preferred in order to take the best advantage of the total area of packed media in practice . Furthermore, many factors such as water flow, air supply, shape and morphological characteristics of packed media and filter porosity also affect the flow pattern in the system. Therefore, before starting up such biological reactors, tracer studies should be performed to determine the status of the hydraulic regime. Tracer studies should be based on theoretical principles and also definite procedures in order to make the results of different tests comparable. After collecting data of tracer study in a reactor, an analytical approach is required for interpretation. Here the principles of two different methods that are used in this research are briefly discussed. Method 1 Determination of Dispersion Number (d) One of the popular approaches for the determination of hydraulic regime in a reactor, is calculating dispersion number (d). Dispersion number in an ASFFR may be defined as below. (1) d = Dispersion number of reactor (dimensionless) D= Upflow Dispersion Coefficient () u= Mean Flow Velocity () l=Height of the Reactor (L) In practice, Dispersion Number can be calculated through a tracer study. In this approach an appropriate tracer such as Rhodamine-B is injected either in a pulse or continuously. After injection, concentration of tracer in the outlet of the system is measured periodically. Based on concentration data at different times, the normalized variance of data distribution may be calculated as below: (1) (2) (3) Where : ti= elapsed time ci=tracer concentration at ti =mean residence time ( time to the centroid of the distribution) = variance = normalized variance (dimensionless) The variance and Dispersion Number for a reactor are calculated from the effluent concentration-time data from a pulse input. After the dye injection at the inlet of the reactor, output samples are collected at time intervals from 0.5 min to several minutes. The shape of the residence time distribution is determined by graphing the concentration versus time. The dispersion number (D/ul) can be computed using Eq. 3 through a trial and error procedure. In this reasearh a computer program has been developed for determining dispersion number and calculating the parameters required for graphing time distribution curves. This program is a FOXPRO routine and source code is presented in Table 1. Each dispersion number then should be compared with typical ones according to (Viessman, 1990) in order to determine the hydraulic regime of the reactor. Table1- Computer Programme for Computing Dispersion Number Using Tracer Test Data. Set Talk Off Set Echo Off set dele on CLEAR ALL CLEAR STORE SPACE(8) TO NAME @2,2 SAY "ENTER THE DATA FILE:" GET NAME READ IF !used("&NAME") Select 0 use &NAME ELSE Select &NAME ENDIF
dbfname=dbf() Clear STORE 0 TO Ndata,C0,TR @ 7,5 SAY " Tracer Concentration:" GET C0 Picture "999999.99" @ 8,5 SAY " Theoretical Detention Time:" GET TR Picture "9999999.99" READ Count all to NUM GO TOP FOR i=1 To NUM Replace CAL_T_TR WITH T/TR ; CAL_C_C0 WITH C/C0 ; THEO_C_C0 WITH EXP(-T/TR) ; TC WITH T*C ; T2C WITH T^2*C SKIP ENDFOR BROWS SUM C To Sigma_C SUM TC TO Sigma_TC SUM T2C TO Sigma_T2C TBAR= Sigma_TC/Sigma_C VARIANCE= (Sigma_T2C/Sigma_C)-(TBAR^2) Norm_var= Variance/(TBAR^2)
D_UL=TRY(Norm_var) Set print On Set Printer to DULOUT.TXT Count all to NUM GO TOP ? '" ORDER OF VARIABLES FROM LEFT TO RIGHT:"' ?'"TIME,CONCENTRATION,CALCULATED T/TR,CALCULATED C/C0,THEORITICAL C/C0,T/TBAR,TC,T2C"' ? ? FOR i=1 To NUM ? ALLTRIM(STR(T,12,4))+" "+ALLTRIM(STR(C,12,2))+" "+ALLTRIM(STR(CAL_T_TR,12,6))+" "+ ; ALLTRIM(STR(CAL_C_C0,12,6))+" "+ALLTRIM(STR(THEO_C_C0,12,6))+" "+ ; ALLTRIM(STR(T/TBAR,9,3))+" "+ALLTRIM(STR(TC,12,2))+" "+ ; ALLTRIM(STR(T2C,12,2)) SKIP ENDFOR ? ? dbfname ? SPACE(5)+'"'+"Initial Concentration:"+STR(C0,14,2) ? SPACE(5)+'"'+"Theoritical Detention Time :"+STR(TR,14,2) ? SPACE(5)+'"'+"Sigma C:"+STR(Sigma_C,14,2) ? SPACE(5)+'"'+"Sigma T*C:"+STR(Sigma_TC,14,2) ? SPACE(5)+'"'+"Sigma T^2*C:"+STR(Sigma_T2C,14,2) ? SPACE(5)+'"'+"MEAN RESIDENCE TIME:" + STR(TBAR,7,2) +" (time)"+'"' ? SPACE(5)+'"'+"VARIANCE : " + STR(VARIANCE,7,2) + " (time^2)"+'"' ? SPACE(5)+'"'+"NORMALIZED VARIANCE:" + STR(Norm_var,8,4) + " Dimentiomless"+'"' ? SPACE(5)+'"'+"CASE:" + STR(CASE,8,4)+ " Dimentiomless"+'"' ? SPACE(5)+'"'+"DISPESION NUMBER (D/UL) : " + STR(D_UL,20,3) + " Dimentiomless"+'"' Set print OFF Set Printer to CLEAR ALL CLEAR MODI FILE DULOUT.TXT FUNCTION TRY PARAMETERS CONTROL PUBLIC CASE JUMP=0.00001 DO While .T. ? CASE," ",CONTROL ," ",JUMP WAIT CASE=(2*JUMP)-(2*JUMP^2)*(1-EXP(-1/JUMP)) IF (CASE<=0) _OUTPUT=OPTIMIZ(JUMP,CONTROL) Return(_OUTPUT)
* JUMP=JUMP/10 * EXIT ENDIF IF (CASE<=CONTROL) JUMP=JUMP*10 ELSE EXIT ENDIF ENDDO INITIAL=JUMP/10 FINAL=JUMP GAM=INITIAL/100000 FOR I=INITIAL TO FINAL STEP GAM CASE=(2*I)-(2*I^2)*(1-EXP(-1/I)) *? CASE," ",CONTROL ,INITIAL ,FINAL *WAIT IF ABS( (CASE-CONTROL) )<=0.0001 Return(I) ENDIF ENDFOR FUNCTION OPTIMIZ PARAMETER JUMP,CONTROL INITIAL=JUMP/100 FINAL=INITIAL*10 GAM=INITIAL/100000 Do While .T. FOR I=INITIAL TO FINAL STEP GAM CASE=(2*I)-(2*I^2)*(1-EXP(-1/I)) IF ABS( (CASE-CONTROL) )<=0.0001 Return(I) ENDIF ENDFOR INITIAL=FINAL FINAL=INITIAL*10 GAM=INITIAL/100000 LOOP ENDDO
Method 2 : Fit of collected data in a Tracer study with Completely Mixed Model Theoretically the effluent concentration of tracer with a steady-state flow and completely mixed regime in the reactor, can be formulated as below: 4) Where : effluent tracer concentration of reactor initial tracer concentration at the beginning of the test. T = elapsed time Tr =Theoretical Hydraulic detention time Converting the Eq. 4 to a linear equation is simply performed as below: 5) 6) In complete mixed reactors, graphing versus yields a line and in ideal conditions all data points are exactly located on the line. In this condition, of linear regression of versus is equal to 1. Any deviation of hydraulic regime from completely mixed condition affects and decreases the quantity of this parameter to less than 1. Thus fitness of data with complete mixed model and determination of of linear regression in experimental data can be useful to determine whether a complete mixed hydraulic regime is achievable in the reactor. Relatively less data points are required, which is one of the advantageous features of this method. MATERIALS AND METHODS Hydraulic characteristics in an ASFFR were studied in the preliminary phase of comprehensive research on the behaviour of this system in removing various organic loading rates. Fig. 1 illustrates the pilot plant which was used in this research. Fig.1- Scheme of Aerated Submerged Fixed-Film Reactor in the Study Studies were conducted in a 1.8 m tall, 25 cm diameter Plexy-Glass Cylinder, packed with 11000 tabular PVC media. The length and diameter of each PVC medium were 1.2 cm and 0.6 cm, respectively. The useful volume of the reactor was 66.18 L with packed ratio of 15.68% . Upflow current was provided using an adjustable dosing pump with a range of 4-35 L/hr. The air supply of reactor was provided using a compressor with tank capacity of 150 L and 6 atm pressure,. The compressor was connected to the pilot plant with an oil-tarp and adjustable valves and also a flowmeter(0-40 L/min) . The pilot plant was equipped with sampling ports every 30 cm. The concentration of Rhodamine-B was determined according to spectrophotometry at 555 nm. A 50 mg/L reference solution was prepared and then diluted to concentrations of 20-50-80-100-200-400-600-800-1000 ug/L. Based on the absorbance of each prepared standard at 555nm and linear regression of data the following equation was set. (7) Independent variables in this phase were water flows and air supply. Water flow was changed from 4.616 to 14.153, 23.356 and 30.763 L/hr respectively and each flow was studied without air supply to simulate plug-flow condition and also with air supply of 10, 20 and 30 L/min. Hence, 16 different state have been evaluated. In each run, after pulse input of tracer and 1 minute rapid mixing of the content of reactor, initial concentration was determined. Then samples were collected from the effluent of the system in time periods of 20-60 min (based on flow rate) and concentration of Rhodamin-B determined. Sampling and determination of tracer concentration were continued up to the the minimum detectable concentration. For calculating the Dispersion Number, a program was developed in FOXPRO. This program can compute Dispersion Number and parameters which are used to graph time distribution curves. Required data of the program are : tracer concentrations (C), related elapsed time(ti), initial concentration of tracer in reactor(0), and hydraulic detention time (Tr), The following steps presented here describes how to use the program. Step1 : A file with following fields structure shoud be made in FOXPRO Field Field Name Type Width Dec 1 T Numeric 12 4 2 C Numeric 12 2 3 CAL_T_TR Numeric 12 6 4 CAL_C_C0 Numeric 12 6 5 TC Numeric 12 2 6 T2C Numeric 12 2 If a file with this structure exists previously, its structure can be copied to the desired file as follow: COPY STRU TO File name Step 2: tracer concentrations (ci) and related elapsed time (ti) are entered to the file. Step 3 : Execution of the routine DISPERSION as below: DO DISPERSION At this step, the following statement appears. In response, the name of data file should be addressed. ENTER THE DATA FILE : Then two question statements appears as follows which should be answered correctly. INITIAL TRACER CONCENTRATION: THEORETICAL DETENTION TIME: After these steps the computation performs and output will appears. For example output file of data for water-flow of 30.763 L/hr and air supply of 10 L/min is illustrated in Table 2 . Table 2- Example of Output File of Designed Program (for water-flow of 30.763 L/hr and air supply of 10 L/min
After performing the tracer study, biological startup of pilot was performed using glucose as carbon source. Nitrogen and phosphorous were added proportional to influent BOD using ammonium chloride and ammonium phosphate . Influent BOD:N:P was adjusted according to ratio of 100:5:1. The reactor was inoculated with microbial mass consisting of the secondary sludge of a sanitary wastewater treatment plant. After reaching steady state condition, the carbonaceous substrate was replaced with ethylene glycol . In this stage 4 hydraulic loading levels were studied. Influent COD was adjusted to 500 mg/L and was not changed during the study . In this way organic loading was increased by increasing of hydraulic loading of wastewater to the reactor. Pilot plant effluent was monitored during each run and after reaching steady state condition samples were collected from each sampling ports simultaneously and concentration of soluble COD, alkalinity, phosphorous, and nitrate determined to show the effect of biological growth on the uniformity of the system. RESULTS AND DISCUSSION According to the described methodology, dispersion numbers were determined for 16 different states in the tracer study. Data in Table 3 shows that, apart from the state when no air was supplied to the system, in all other conditions dispersion numbers were more than 0.2. This means that well mixed condition were maintained. Variation of water flows and air supplies in the experienced range in this study did not show any significant change in dispersion number. When there was no air supply to the reactor, dispersion number variation was significant and reactor showed Plug-Flow behaviour. Table 3- Dispersion Number in Different Condition During Tracer Test.
On the other hand as shown in Table 4, fit of data with the complete mixed model indicated that in all test condition, except when no air supply exists, more than 90 percent of data fit with the completely mixed model was obtained. Table 4- results of Tracer test Data Fitness with Complete Mixed Model according to
Figs. 2-5 illustrate the variation of (Ln C/C0) versus (T/Tr) in each of the 16 test condition. Even when there was no air supply in the system, for lower water flows (longer detention times), dispersion occurred in reactor due to diffusion. In other cases the role of air supply in mixing was less relative to cases with shorter detention time. Fig.2- Ln(C/C0) Vs. T/Tr for Wastewater Flow of 4.416 L/hr Fig.3- Ln(C/C0) Vs. T/Tr for Wastewater Flow of 14.153 L/hr Fig.4- Ln(C/C0) Vs. T/Tr for Wastewater Flow of 23.153 L/hr Fig.5- Ln(C/C0) Vs. T/Tr for Wastewater Flow of 30.763 L/hr As mentioned, pilot plant effluent was monitored during each experiment and samples collected from each sampling port simultaneously in steady state conditions to show the effect of biological growth on mixing. Concentrations of soluble COD, alkalinity, phosphorous, and nitrate, determined to show the effect of biological growth on the uniformity of the system, are shown in Table 5 for different loadings. As the variation of parameters at each sampling port is not significant for different loadings, it is concluded that in the range of loadings used in this study there was not significant variation of concentration of the measured parameters along the reactor. Therefore biological growth of film on supporting media did not change the mixing pattern significantly. Table 5- Concentration of COD,NO3-, ALK, P at different Ports after reaching Steady State Condition
* Test Conditions: 1- 4.616 L/hr (detention Time=11 hrs) 2- 14.153 L/hr (detention Time=8 hrs) 3- 23.153 L/hr (detention Time=6 hrs) 4- 30.763 L/hr (detention Time=4 hrs) ND= Not detectable CONCLUSIONS In this research flow characteristics of an Aerated Submerged Fixed-Film Reactor before start-up and also after reaching steady state condition with different detention time was studied. Tracer tests were performed and data analyzed to determine dispersion number in each condition. Samples from different ports along the reactor were analyzed to characterize the effect of biological growth on mixing. The results showed that in the experienced range of flow and air supply, the well-mixed condition is available. Furthermore analyzed samples from different ports showed that in different applied hydraulic loadings, there was no significant lack of uniformity of concentration in the system. So biological growth did not pose any critical lack of uniformity of concentration during the test. REFERENCES Arceivala Soli, J, (1988), Wastewater Treatment for Pollution Control, McGraw-Hill ASTM, ( 1997), Standard Test Method for Open-Channel Measurement of Time Travel Using Dye Tracers, Vol 11.01 Fang, H.H.P, Yeong C.L.Y., (1993), Biological Wastewater Treatment In Reactors With Fibrous Packing, Journal of Environmental Engineering, Vol 119, No.5, 946-957 Hamoda, M. F., Abd-El-Bary, M. F.,(1987), Operating Characteristics of the Aerated Submerged Fixed-Film Bioreactors, Wat. Sci. Tech., Vol 21,939-947 Kato, D, Sekikawa Y, (1972), Fixed Activated Sludge Process for Industrial waste Treatment, Journal Water Pollution Control Federation, 44, 401-413 Louis, R.J., Randall, C.W., (1995), "Utilization of a Sponge media Integrated Fixed-Film Activated Sludge Process for treatment of a high strength, high ammonia industrial wastewater", WEF 68th Annual Conf. & Exposition, Miami Mesdaghinia, A., (1986), " Fixed Activated Sludge Makes Sewage Treatment Simple", Wat. Sci. Tech., Vol 18/7-8, PP 193-198 Park,T.J., Lee, K.H., (1996)," Petrochemical Wastewater Treatment With Aerated Submerged Fixed-Film Reactor Under High Organic Loading Rate ",Wat. Sci. Tech., Vol 34, No. 10, PP 9-16 Rusten, B., (1984), " Wastewater Treatment With Aerated Submerged Biological Filters", WPCF, Vol. 56, No.5,PP 424-431 Smith & Loveless, Inc, (1983), " Check FAST First for Dependable, Economical Treatment",USA, Technical Report,Kansas Smith & Loveless, Inc, (1988), "Fixed Activated Sludge Treatment System ", Technical Report,USA,Kansas Viessman, W Jr., Hammer, Mark J., (1990), " Water Supply and Pollution Control ",PP 287-291, Harper & Row
Copyright remains with the author. Published by Bioline Publications
The following images related to this document are available:Photo images[bf00001e.jpg] [bf00001c.jpg] [bf00001d.jpg] [bf00001b.jpg]Line drawing images[bf00001a.gif] |
|