Published online 3 April 2006
Published in J Environ Qual 35:742-748 (2006)
DOI: 10.2134/jeq2005.0179
© 2006 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
TECHNICAL REPORTS
Behavior of an Aerated Submerged Fixed-Film Reactor (ASFFR) under Simultaneous Organic and Ammonium Loading
R. Nabizadeh* and
A. Mesdaghinia
Department of Environmental Health Engineering, School of Public Health, Tehran University of Medical Sciences, P.O. Box 14155-6446, Tehran, Iran
* Corresponding author (rnabizadeh{at}sina.tums.ac.ir)
Received for publication May 10, 2005.
 |
ABSTRACT
|
|---|
The performance of an aerated submerged fixed-film reactor (ASFFR) under simultaneous organic and ammonium loading and its effect on nitrification was studied. Organic loadings varied in the range of 1.93 to 5.29 g chemical oxygen demand (COD) m–2 d–1 and NH4–N loadings were in the range of 116 to 318 mg NH4–N m–2 d–1. Increments of loading rates were obtained both by increasing the flow rate and increasing the influent substrate in individual pilot runs. Results showed that with organic loading rates up to 3.97 g COD m–2 d–1, complete nitrification was achievable. Although high organic loading such as 5.29 g COD m–2 d–1 could cause nitrification to stop, shifting to lower organic loadings made nitrification start and set rapidly to its previous steady-state concentrations. Comparison of results showed that in the ASFFR, nitrification would be severely affected by an organic loading rate of 5.29 g COD m–2 d–1 by increasing either the flow or the influent substrate. It should be noted that the average value of dissolved oxygen was 3.4 mg L–1 with an air supply of 15 L min–1, and there was no indication of oxygen limitation. The results of this study show the flexibility of ASFFRs under changing organic loads. Furthermore, for achieving complete nitrification and optimum application of these reactors for protecting receiving water from the environmental hazards of ammonium, the maximum organic loading that would present complete nitrification should be considered.
Abbreviations: ASFFR, aerated submerged fixed-film reactor • BOD, biological oxygen demand • COD, chemical oxygen demand • HRT, hydraulic retention time
|
INTRODUCTION
|
|---|
ATTACHED GROWTH SYSTEMS have proven to be efficient alternatives for treatment of carbonaceous and nitrogenous pollutants. Trickling filter was used as the first attached growth system, and its performance in reducing carbonaceous substrate and nitrification was well documented by Metcalf & Eddy (1991).
The simplicity, flexibility, and high retention time of fixed activated sludge also has been reported by Mesdaghinia (1986). The practical kinetic approach for interpreting a newer modification of the fixed-film system, known as the aerated submerged fixed-film reactor (ASFFR), was described by Hamoda (1989). Park et al. (1996) conducted an experimental study and showed the high performance of ASFFRs in treating petrochemical wastewaters with high organic loading rates. The satisfactory application of an ASFFR in removing high ethylene glycol loads was reported by Nabizadeh et al. (2000). High solids retention time of the ASFFR provides more appropriate conditions for the growth and activity of nitrifying bacteria, so the system was used for simultaneous carbon removal and nitrification. It has been shown that the ASFFR process is able to handle continuous severe organic loading increasing from 5 to 120 g m–2 d–1 biological oxygen demand (BOD) with a slight decrease in organic removal efficiency from 97.9 to 88.5% for BOD and 73.6 to 67.8% for COD (Hamoda and Al-Sharekh, 1999). Hamoda (1989) showed that, along with such performance in C removal, nitrification only decreased at higher rates. Nitrification in biofilm under variable loads and low temperature (7°C) was studied by Canler et al. (2003).
Sen et al. (1994) used biofilm modules in an activated sludge system to enhance nitrogen removal. It was concluded that under optimum conditions, nitrification was enhanced by 43% with the addition of fixed media in the aeration zone of an activated sludge basin, which was operating near the minimum mean cell residence time for nitrification in a continuous flow system without media. Gilmore et al. (1999) studied the two-stage biological aerated filter system to optimize ammonia reduction in the second nitrifying filter media. It was found that overall nitrification efficiency for an average temperature of 12.4 ± 0.1°C was >90% when NH4–N loading to the second stage was 0.6 kg m–3 d–1 or less.
Nitrate is a stable end product with low toxicity and does not harm aquatic life in the concentrations typically present, but ammonia is highly toxic at low concentrations. In many industries, such as the petrochemical industry, fish farming, fertilizer manufacturing, and many others, high amounts of ammonia and organic matter are present in the wastewater. Fluctuation of organic matter concentration in the wastewater of these industries may cause an irreversible shock to the conventional wastewater treatment systems (Nemerow, 1991). Therefore, an efficient COD and nitrification removal system should be considered to protect the receiving water from the potential hazards of ammonia.
Heterotrophic bacteria grow very efficiently, doubling in population about every 8 h, but nitrifiers are much less efficient, and typically require 24 h to double in population (Golz, 2005). Biological treatment processes are classified as either suspended growth or fixed film. In suspended-growth processes, the waste is added to a large aerated tank where it is converted to cellular material by suspended bacteria. These suspended bacteria must be removed by a solids separation device before the recirculating water can be returned to the culture tank. To give the autotrophic bacteria time to reproduce, a portion of the separated solids must be continually recycled to the aeration tank, and the amount of solids that are recycled must be continually monitored. Golz (1999) noted that the suspended growth process is not frequently used in aquaculture systems due to the intensive operational requirements. It was also concluded that bacterial attachment provides sufficient reproduction time for nitrifiers in fixed film systems, and this feature makes biofilm nitrification an appropriate treatment method for recirculating aquaculture systems (Golz, 2005). It should be considered that in attached growth systems, the operators are not able to do much to re-stabilize the biofilm if it is removed from the filter media due to shock of organic loads. Therefore, for the optimum environmental application of these systems, their flexibility to the variations of organic loadings should be studied. In this study, the organic and ammonia loadings were increased simultaneously to evaluate flexibility and performance of a single-stage ASFFR reactor in nitrification under such conditions, because, in many industries, the ammonia concentrations in the waste stream would increase along with the organic content of wastewater.
|
MATERIALS AND METHODS
|
|---|
Specifications of the ASFFR Pilot
The specifications of the ASFFR (Fig. 1) and its supporting media are presented in Tables 1 and 2, respectively. The ASFFR pilot was made up of Plexiglass and was filled with 29 600 pieces of PVC cylinder. A perforated plate was installed at the top of the packed media to keep them submerged in the reactor. The sampling ports were used to take the samples from different depths of the reactor. Although all these sampling ports were not used in this study, they facilitated the tracer study, which was performed to determine the flow regime in various hydraulic loadings in the preliminary stage of reactor tests. The tracer study was performed to determine the reactor's flow regime with the mentioned packing media. It has been previously indicated that a well-mixed condition would be provided in the reactor (Nabizadeh et al., 2000). Air and synthetic wastewater were injected at the bottom of the reactor to maintain an upflow current of liquid. An air compressor was used to supply the air requirement of the reactor. An air pipe was equipped with a filter to remove dust and other impurities. The air-flow rate was determined by a flow meter and was also adjusted using an adjusting value to keep the dissolved oxygen at the appropriate level. The flow of incoming wastewater was adjusted by a dosing pump that was calibrated according to the desired flow rate. Avoiding the flow variation during the study, an equalization tank was used in the suction line of the pump. The equalization tank was connected to the wastewater storage tank through a floater. The floater could open and close the valve and maintain a constant static head in the suction line. Finally, a collection unit was installed at the top of the packed media to discharge effluent from the ASFFR.
Startup Phase of the ASFFR Pilot
After packing the pilot with supporting media, the hydraulic retention time (HRT) was adjusted to 11 h and, at this stage, the concentration of influent COD was 500 mg L–1. Ethylene glycol was used as a carbonaceous source and ammonium chloride (NH4Cl) as a nitrogen source. The ammonium ion is largely predominant at neutral or slightly basic pH: the ratio of the ammonium to the ammonia concentration is equal to 100:1 at a pH of 7.4 and a temperature of 20°C. Therefore, the NH3 portion in equilibrium was very low, and it could not vary significantly during the short storage period. The concentration of influent NH4–N was 30 mg L–1. To avoid sharp pH changes in the reactor caused by nitrification, alkalinity of the reactor content was adjusted to 340 mg L–1 as CaCO3 using sodium bicarbonate. For preparing synthetic wastewater, tap water was used and a sufficient amount of phosphorus was added according to the BOD to N to P ratio of 100:5:1 to provide the required phosphorus for microbial growth. Ammonium was determined according to Method 4500-NH3 B (American Public Health Association, 1998). Nitrate concentration was analyzed based on spectrophotometric analysis with sodium salicylate (Fresenius et al., 1988, p. 229–230). Before filling the reactor, 20 L of adapted sludge from an ASFFR reactor, capable of complete nitrification, was added to provide sufficient adapted microbial population for organic and ammonia degradation in the startup period. The adapted sludge was collected by withdrawing and washing the biofilm on the media of an ASFFR reactor, which was working with satisfactory organic removal and complete nitrification. Air supply was adjusted to 5 L min–1 to provide required soluble oxygen in the bulk liquid. The pilot was operated 14 d without recordings of effluent COD, ammonium, and nitrate, but the biofilm growth on the supporting media was observed. After 2 wk when the biological growth was established on supporting media, effluent COD and nitrate were determined. As shown in Fig. 2, effluent COD reached the steady-state condition after 18 d and its concentration was 23.5 mg L–1. Although after 18 d of the pilot run the ASFFR showed a steady-state condition according to the effluent COD concentration, nitrification was not initiated since the nitrate effluent concentration was 0 mg L–1. The effluent nitrate concentration remained unchanged up to 30 d of the pilot run. To enhance the nitrification, inflow was cut off for 10 d and changed from pilot to batch. Furthermore, 10 L of sludge from a secondary sedimentation tank of a municipal wastewater treatment unit was added to the reactor. After 10 d (Day 40), the pilot was fed for 3 h d–1. After 43 d of the pilot run, continuous flow again was stabilized and effluent nitrate tended to be steady. The concentration of COD and nitrate were monitored up to 60 d after the pilot run, and Fig. 2 shows that from Days 50 to 60 the effluent concentrations of COD and nitrate were steady and complete nitrification occurred in the ASFFR pilot. The ammonia concentration was also nondetectable in the last 10 d.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 2. Effluent chemical oxygen demand (COD) and nitrate concentrations at the startup period: (A) system turned to batch, (B) intermittent flow started for 3 h d–1, and (C) continuous flow established.
|
|
Loading Strategy for Evaluating the Flexibility of the ASFFR
In this research, seven pilot runs have been studied to review nitrification in the ASFFR both with increasing the organic loading through increasing the concentration and also through increasing the flow rate. It should be noted that loading rates are presented according to the mass of substrate per total surface area of packed media per day. The influent concentration of NH4–N increased proportionally to the influent COD concentration. Table 3 presents the loading specification of each of seven pilot runs. After startup of the pilot with an 11-h detention time, flow rate increased to change the detention time to 8 h, and after that influent concentration was increased until complete disappearance of nitrification. At this stage, the pilot inflow turned to the first loading rate (8 h, influent COD = 500 mg L–1, and NH4–N = 30 mg L–1). After the reestablishment of a steady-state condition, the loading rates were increased by increasing the flow rate with constant influent COD and ammonia concentration. For the first run (HRT = 11 h, influent COD = 500 mg L–1, and NH4–N = 30 mg L–1), air supply adjusted to 5 L min–1. For other pilot runs, air supply increased proportionally to the loading rates to keep oxygen concentration above 3 mg L–1, since the oxygen concentration proved to be a limiting factor both in organic removal and nitrification. Dissolved oxygen, influent COD, effluent NH4–N, and effluent nitrate were detected daily during the steady-state condition. Parameters such as pH, alkalinity, and VSS were also analyzed routinely, but they are not considered basic variables that will be analyzed in this article.
|
RESULTS AND DISCUSSION
|
|---|
In the first pilot run with HRT = 11 h, influent COD = 500 mg L–1, and NH4–N = 30 mg L–1, after 18 d the pilot reached a steady-state condition based on COD concentration. But complete nitrification was not achieved until 50 d of operation. As it was the first run, this time was required for autotrophic bacteria to grow and attach to the supporting media. Average COD removal in the steady-state period was calculated as 95.6%. The average value of dissolved oxygen was 5.9 mg L–1 with an air supply of 5 L min–1. As shown in Fig. 2, effluent COD and nitrate reached steady state in 50 d of operation. As ammonia concentration was not detectable, it can be concluded that complete nitrification occurred. Therefore, with the organic loading of 1.93 g COD m–2 d–1, complete nitrification was achievable.
In the second run, flow rate was increased to adjust the detention time to 8 h. Influent COD and NH4–N were 500 and 30 mg L–1, respectively, and remained constant. As shown in Fig. 3, a new steady-state condition was established within 3 d. The average value of dissolved oxygen was 5.1 mg L–1 with an air supply of 8 L min–1. Average COD removal in the steady-state period was calculated as 94.2%. After establishment of a steady-state condition in this run, effluent ammonia was not detectable and complete nitrification was achieved. According to the results of this run, complete nitrification can be achieved under the organic loading rate of 2.65 g COD m–2 d–1 and NH4–N loading rate of 159 mg NH4–N m–2 d–1 in the pilot-scale ASFFR.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 3. Effluent chemical oxygen demand (COD) and nitrate concentrations for Run 2 (hydraulic retention time [HRT] = 8 h, influent COD = 500 mg L–1, NH4–N = 30 mg L–1).
|
|
In the third run, flow rate was not changed or adjusted to maintain HRT of 8 h, but to increase the loading rate of COD and ammonium, influent COD and NH4–N were increased to 750 and 45 mg L–1, respectively. Changes of effluent COD and nitrate due to this rapid loading change are presented in Fig. 4. As presented in Fig. 4, after the application of this loading rate, the changes in effluent COD and nitrate tended to reach a new steady-state condition within 3 d, which shows the flexibility of the ASFFR in sustaining sock loads. The average value of dissolved oxygen was 4.5 mg L–1 with air supply of 10 L min–1. Average COD removal in the steady-state period was calculated as 94.3%. After establishment of a steady-state condition in this run, effluent ammonium was detectable and its average concentration was about 4 mg L–1 as NH4–N. Because the ammonium concentration was not considerable and it was about 9% of influent ammonia, it can be concluded that under the organic loading rate of 3.97 g COD m–2 d–1 and NH4–N loading rate of 238 mg NH4–N m–2 d–1, sufficient nitrification still occurs.
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 4. Effluent chemical oxygen demand (COD), nitrate, and ammonium concentrations for Run 3 (hydraulic retention time [HRT] = 8 h, influent COD = 750 mg L–1, NH4–N = 45 mg L–1).
|
|
In the fourth run, HRT remained the same as the third run to 8 h. Organic and ammonium loadings were increased by increasing influent COD and ammonium. Influent COD and NH4–N were increased to 1000 and 60 mg L–1, respectively. As presented in Fig. 5, at first influent COD began to increase, but after about 5 d a new steady-state condition prevailed. The COD removal efficiency of 93.8% achieved was sufficient considering such high organic loading. The effect of organic loading on nitrification was considerable (Fig. 5). After 10 d of starting this load, nitrate concentration decreased to zero and nitrification completely ceased. Along with the decrease of nitrate, the concentration on ammonium increased and reached its highest level in 11 d after applying 40 mg L–1 as NH4–N. The average value of dissolved oxygen was 3.5 mg L–1 with an air supply of 15 L min–1; there was no indication of oxygen limitation. Therefore it can be concluded that under the organic loading rate of 5.29 g COD m–2 d–1 and an NH4–N loading rate of 318 mg NH4–N m–2 d–1 nitrification would cease. Although nitrification seems to have disappeared in this load, organic removal is still sufficient and stable.
View larger version (33K):
[in this window]
[in a new window]
|
Fig. 5. Effluent chemical oxygen demand (COD), nitrate, and ammonium concentrations for Run 4 (hydraulic retention time [HRT] = 8 h, influent COD = 1000 mg L–1, NH4–N = 60 mg L–1).
|
|
In the fifth run, loading turned to the same characteristics as Run 2 (HRT = 8 h, influent COD = 500 mg L–1, NH4–N = 30 mg L–1). This step of the research was performed to test the flexibility of the ASFFR. The main objective in this run was to see whether the system could turn back to its previous performance after nitrification ceased. As shown in Fig. 6, after changing the loading rate to the conditions of Run 2, the system rapidly responded and nitrification began to increase while ammonium tended to decrease. After 8 d, a steady-state condition was set up and nitrification got back to its normal complete state. If the autotrophic biofilm was removed and washed out through the reactor, the ASFFR could not show such a performance rapidly, since these bacteria grow very slowly and it would take a long time to establish an active nitrifying biomass again. Therefore, it can be concluded that high organic loading would suppress the nitrification process, but the capacity of regeneration still exits and when loading conditions shift rapidly to the normal state, there would be a chance to restart the nitrification. Figure 7 presents the changes of influent COD, effluent nitrate, and effluent NH4–N through Runs 2 to 5.
View larger version (33K):
[in this window]
[in a new window]
|
Fig. 6. Effluent chemical oxygen demand (COD), nitrate, and ammonium concentrations for Run 5 (hydraulic retention time [HRT] = 8 h, influent COD = 500 mg L–1, NH4–N = 30 mg L–1).
|
|
In the sixth run, flow rate was increased to adjust the detention time to 6 h. Influent COD and NH4–N were 500 and 30 mg L–1, respectively, and remained constant. This step of research was performed to review the effect of increasing the loading rate due to flow increment on the process of nitrification. As shown in Fig. 8, a steady-state condition in view of effluent COD was set after 14 d of operation, and complete nitrification was also achieved after 30 d of operation. It showed that the average value of dissolved oxygen was 4.2 mg L–1 with air supply of 12 L min–1. Average COD removal in the steady-state period was calculated as 92.8%. Complete nitrification could be achieved under the organic loading rate of 3.53 g COD m–2 d–1 and NH4–N loading rate of 212 mg NH4–N m–2 d–1.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 8. Effluent chemical oxygen demand (COD), nitrate, and ammonium concentrations for Run 6 (hydraulic retention time [HRT] = 6 h, influent COD = 500 mg L–1, NH4–N = 30 mg L–1).
|
|
In the seventh run, flow rate was increased to adjust the detention time to 4 h. Influent COD and NH4–N were 500 and 30 mg L–1, respectively, and remained unchanged. As shown in Fig. 9, a steady-state condition of COD was set after 14 d of operation, but complete nitrification could not be achieved during 40 d of operation and the ASFFR failed to show a sufficient performance of nitrification in this run. Therefore, it should be noted that under the organic loading rate of 5.29 g COD m–2 d–1 and NH4–N loading rate of 318 mg NH4–N m–2 d–1 nitrification would be considerably decreased. Although nitrification seems to have disappeared in this load, organic removal is still sufficient and stable. The loading rate of Run 7 was equal to the loading rate of Run 4. Comparison of results shows that in the ASFFR, nitrification would be severely affected in this organic loading rate either by increasing the flow or by increasing the influent substrate. The average value of dissolved oxygen was 3.4 mg L–1 with an air supply of 15 L min–1. Average COD removal in the steady-state period was calculated as 92%.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 9. Effluent chemical oxygen demand (COD), nitrate, and ammonium concentrations for Run 7 (hydraulic retention time [HRT] = 4 h, influent COD = 500 mg L–1, NH4–N = 30 mg L–1).
|
|
|
CONCLUSIONS
|
|---|
It was shown that under organic loading rates up to 3.97 g COD m–2 d–1, complete nitrification was achievable with NH4–N loading rates up to 238 mg NH4–N m–2 d–1. This research showed that in the ASFFR, nitrification would be severely affected at an organic loading rate of 5.29 g COD m–2 d–1 by increasing either the flow or the influent substrate. It should be noted that the average value of dissolved oxygen was 3.4 mg L–1 with an air supply of 15 L min–1 and that there was no indication of oxygen limitation. Furthermore it was shown that increased organic loading rate had inhibited nitrification, but nitrification efficiency could be rapidly recovered once the organic loading rate decreased to the previous step. Since the autotrophic bacteria that perform nitrification grow slowly, it can be concluded that ceasing the nitrification process in the ASFFR due to high organic loads does not necessarily mean that autotrophic biofilm is destroyed or washed out. A fast and flexible restart of nitrification with reduction of the organic loading rates of the ASFFR confirms this idea. It should be noted that the biodegradable organic loading rate must not exceed 4 g COD m–2 d–1 to remove COD and ammonium nitrogen simultaneously in the ASFFR.
The results of this study showed that the ASFFR could be used for combined COD removal and nitrification below an organic loading of 4 g COD m–2 d–1, which provides the designers with guidance in setting the design parameter value to optimize the performance of the biological treatment unit. The flexibility of the ASFFR under changing organic loads presents it as a reliable technology for environmental applications where wastewater contains organic matter and ammonium nitrogen simultaneously, and the requirements of organic matter and ammonium should be met to protect aquatic life from potential hazards of the pollutants.
|
REFERENCES
|
|---|
- American Public Health Association. 1998. Standard methods for the examination of water and wastewater. 20th ed. APHA/AWWA/WEF, Washington, DC.
- Canler, J.P., J.M. Perret, F. Lengrand, and A. Iwaema. 2003. Nitrification in biofilters under variable load and low temperature. Water Sci. Technol. 47(11):129–136.
- Fresenius, W., K.E. Quentin, and W. Schneider. 1988. Water analysis. Springer-Verlag, New York.
- Gilmore, K.R., K.J. Husovitz, T. Holts, and N.G. Love. 1999. Influence of organic and ammonia loading on nitrifier activity and nitrification performance for a two-stage biological aerated filter system. Water Sci. Technol. 39(7):227–234.
- Golz, W.J. 2005. Biological treatment in recirculating aquaculture systems. Available at http://mysite.verizon.net/res6m3ph/sitebuildercontent/ sitebuilderfiles/wjg_tchr.pdf (verified 27 Jan. 2006). Article posted by the author.
- Hamoda, M.F. 1989. Kinetic analysis of aerated submerged fixed-film (ASFF) bioreactors. Water Res. 23(9):1147–1154.[CrossRef]
- Hamoda, M.F., and H.A. Al-Sharekh. 1999. Sugar wastewater treatment with aerated fixed-film systems. Water Sci. Technol. 40(1): 313–321.
- Mesdaghinia, A. 1986. Fixed activated sludge makes sewage treatment simple. Water Sci. Technol. 18(7–8):193–198.
- Metcalf & Eddy. 1991. Wastewater engineering: Treatment, disposal, and reuse. 3rd ed. McGraw-Hill, New York.
- Nabizadeh, R., A.R. Mesdaghinia, S. Nasseri, A.H. Mahvi, and M. Shariat. 2000. Analyzing flow characteristics and influence of biological growth on dispersion in aerated submerged fixed-film reactors (ASFFR). Available at http://www.bioline.org.br/request?bf00001 (verified 27 Jan. 2006). Bioline Int., Toronto.
- Nemerow, N.L. 1991. Industrial and hazardous waste treatment. Van Nostrand Reinhold, New York.
- Park, T.J., H.K. Lee, S.D. Kim, and W.C. Kim. 1996. Petrochemical wastewater treatment with aerated submerged fixed-film reactors (ASFFR) under high organic loading. Water Sci. Technol. 34(10):9–16.
- Sen, D., P. Mitta, and C.W. Randall. 1994. Performance of fixed film media integrated in activated sludge reactors to enhance nitrogen removal. Water Sci. Technol. 30(11):13–24.
TOP
ABSTRACT
INTRODUCTION
