Ammonium removal from primary and secondary ...

k30

9/8/00

6:27 pm

Page 179

O. Lahav and M. Green Faculty of Agricultural Engineering, Technion – Israel Institute of Technology, Haifa 32000, Israel Abstract A new process for ammonia removal from sewage effluents is presented. The process uses an ion exchange material, zeolite, both as the separator of NH4+ from the wastewater and also as the carrier for a nitrifying biomass. The process is carried out in a single reactor operating in two modes: an adsorption mode in which the zeolite column acts as a typical ion exchanger and a bioregeneration mode in which the bacteria attached to the zeolite oxidizes the NH4+ to NO3–. The separation between carbonaceous removal and NH4+ removal enables the exclusive selection of nitrifiers at high concentrations attached to the zeolite, thus achieving high bioregeneration rates. This paper summarizes three years of research on the process and focuses on the operation with actual secondary and primary effluents. Process operation showed: (1) no bed clogging occurred due to suspended solids accumulation; (2) residual BOD from the adsorption phase resulted in only minimal heterotroph competition and thus, no fall in the rate of bioregeneration; (3) only a small deterioration in exchange efficiency due to zeolite biofilm coverage as compared to the ion exchange efficiency of “virgin” zeolite. Results show that both secondary and primary effluents can be successfully treated by the process. Keywords Ammonium removal; ion exchange; zeolite; nitrification; biological fouling

Introduction

The popular method of nitrogen removal from wastewater using combined carbonaceous removal and nitrification – denitrification processes suffers from two major operational drawbacks: (1) the forced combination between heterotrophic carbon oxidizers and autotrophic nitrifiers induces operation at low F/M values and large volume aeration basins; (2) the limited ability to optimize the nitrification operational environment (heterotrophic competition, presence of toxic substances, organic shock loads) results in relatively low NH4+ removal rates especially in cold temperatures. A process for ammonium removal was developed to overcome these drawbacks by separating the nitrification step from the carbonaceous removal step and concentrating the NH4+ from the wastewater into a small volume reactor in which high nitrifier concentrations and optimal conditions for nitrification (temperature, oxygen, pH, NH4+ concentrations) can easily be maintained. Results of process operation on simulative effluents are given elsewhere (Lahav and Green, 1998, 2000). This paper focuses on the operation with actual secondary and primary effluents.

Water Science and Technology Vol 42 No 1–2 pp 179–185 © IWA Publishing 2000

Ammonium removal from primary and secondary effluents using a bioregenerated ion-exchange process

Process description

Ion exchange mode (separation stage): A column filled with zeolite (chabazite) is used for ammonium ion exchange from secondary or primary effluents. When NH4+ concentration breakthrough occurs (or after a programmed period of time) the system switches to the bioregeneration mode. Bioregeneration mode (nitrification stage): The same column containing the ammonium rich zeolite is used during the bioregeneration mode as a fluidized bed reactor for biological nitrification with the zeolite acting as the carrier for the biofilm. The microorganisms can only oxidize ammonium released into solution, therefore, a cation containing regenerant

179

k30

9/8/00

6:27 pm

Page 180

solution is recirculated through the bed in order to desorb NH4+.

O. Lahav and M. Green 180

I. desorption:

Z – NH4+ + Na+ ↔ Z – Na+ + NH4+

II. nitrification:

NH4+ + 2O2 → NO3– + 2H+ +H2O

The amount of the NH4+ desorbed and its concentration in the recirculated solution is a function of the total cation concentration in the solution and the solution volume. After a short time, the solution reaches an apparent equilibrium concentration of ammonium (reaction I), while simultaneously, the biomass starts to oxidize ammonia (reaction II). The oxidation of the liberated ammonia to the nitrate anion in the second reaction, shifts the equilibrium in the first reaction to the right and desorption continues until the ammonium concentration in the solution drops to negligible values. At this point, the amount of NH4+ remaining in the zeolite to the next adsorption phase is a function of the cation composition and concentration of the recirculated regenerant solution. During the regeneration mode, the reactor operates in an almost batch mode (no outflow and minimal inflow) and pressurized oxygen is supplied for the nitrification process together with bicarbonate to maintain constant pH. The oxidation of the desorbed ammonium to nitrate anions allows for the reuse of the regenerant during many cycles of nitrification. The addition of external cations is limited only to the amount of sodium bicarbonate buffer added. At the end of both adsorption and regeneration modes, backwash is practiced. After the adsorption mode, backwash removes suspended solids thus preventing bed clogging and heterotroph bacteria accumulation in the bed which might compete with the nitrifier population. At the end of the bioregeneration mode, backwash removes the remaining regenerant solution from the bed which may deteriorate ion exchange efficiency at the beginning of the next adsorption phase. This nitrate-rich backwash water is considered a product rather than a pollutant and therefore denitrification is not necessary. A schematic figure of the system is shown in Figure 1.

Figure 1 Schematic description of the process

k30

9/8/00

6:27 pm

Page 181

Table 1 Primary and secondary effluent data Parameter

Units

NH4+ concentration (Average)

mg/l

Average BOD

mg/l

Primary effluent

38.6

38.6

(26 samples, SD=7.4)

(50.8 samples, SD=3.8)

18.5

297.9

(3 samples, SD=3.7)


102.3

718.9

(21 samples, SD=57)

(10 samples, SD=251.4) 54–58

(Ca2+)

mg/l

54–58

(Na+)

mg/l

143–154

143–154

(K+)

mg/l

18–26

18–26

(Mg2+)

mg/l

30–35

30–35

Total bacteria concentration

no./ml

105.5

107.5

O. Lahav and M. Green

Average COD

mg/l

Secondary effluent

Table 2 Operational data for the primary and secondary effluents experiments Operational parameter

Primary effluents

Secondary effluents

Units

Adsorption phase length

3

3

hr

Average NH4+ load

1.92

1.74

g NH4+/hr

Regeneration phase length

3.5

3.0

hr

Minimal O2 conc. during regeneration

3

3

mg/l O2

8

8

litre

15

15

litre

Regenerant leaving the system each cycle Total regenerant volume

Materials and method Physical description of the experimental system

The system consisted of 9 litre (90 mm diameter) reactor filled with 2000 g of chabazite zeolite (liquid volume of 7 litres). The zeolite used was a natural Herschelik-Sodium Chabazite (CABSORB-ZS500H) – Chabazite – distributed by GSA Resources Inc., Arizona, USA, with an ion-exchange capacity of 2.5 meq/g. Aeration was provided by pure oxygen pressurized column. Three pumps were used for feeding and recirculation. A constant pH was maintained using a pH controller connected to buffer solution pump. Measurement of oxygen uptake rate (O.U.R.)

Nitrifiers O.U.R.: 1/2 gram of zeolite covered with biofilm was mixed in a 50 mg/l NH4+–N and 100 mg/l K+ solution . The pH was held constant at 7.5 and phosphate was added as needed. When the solution was aerated to saturation, the air supply was stopped and oxygen concentrations were measured every two minutes. Heterotrophs O.U.R.: a similar procedure was used except that the solution contained 50 mg/l glucose and nutrients (P, K, N) as needed. N-Allyl Thiourea (1 ml per 100 gram solution) was added to inhibit nitrifiers. Heterotrophs concentration in the biofilm

Biofilm was removed from the zeolite by VORTEX (5 times, 30 seconds each time). The bacteria was incubated on R2A agar plate for 24 hours at 37ºC. Effluents characteristics

The secondary and primary effluents characteristics are summarized in Table 1. A complete cycle was sampled every two days as follows: hourly input and output ammonium concentrations during the adsorption cycle, NH4+ and NO2- concentrations profile every 1/2 hour during bioregeneration.

181

k30

9/8/00

6:27 pm

Page 182

Results and discussion


The system was operated continuously for 38 days (4 cycles a day) with secondary effluents and then 21 days with primary effluents. The operation data for the experiments with primary and secondary effluents are presented in Table 2. Operating conditions were similar for both experiments except for the adsorption period. The size of the system and all other technical parameters were similar. During the experiments the system operated with a regenerant solution in the steady state concentration of 60–65 meq/l, resulting from regenerant discharge of 8 litres and the amount of ammonium adsorbed in the adsorption phase (an average of 5.76 g NH4+–N for the operation with primary effluents and 5.22 g NH4+–N for the operation with secondary effluents). The operational parameters selected (Table 2) were based on results obtained when treating simulated effluents at a wide range of operational conditions. The length of the bioregeneration phase that was required to oxidize the total amount of ammonium adsorbed in the adsorption phase was 3.5 hours for operation with primary effluents and three hours for operation with secondary effluents. Figures 2– 5 show the results of NH4+ concentration profiles during the bioregeneration phase and in the effluent of the adsorption phase. The results are similar to those obtained with simulated effluents comprised only of a NH4+ solution enriched with typical cation concentrations and no organic matter (Lahav and Green, 1998). The similarity in the nitrification efficiency during bioregeneration indicates a negligible effect of heterotrophic competition as a consequence of the possible accumulation of organic solids during the adsorption phase. Regarding the adsorption results, lower ammonium concentrations in the effluent were observed when primary effluents were treated, as compared to those observed with secondary effluents. This was probably the result of the longer retention time (20% more) that was applied during these experiments. The rational for the longer adsorption retention time was that in order to maintain a constant ammonium load the flow rate was reduced by 20% to compensate for the higher ammonium concentration in the primary effluents (50 mg/l instead of 40 mg/l). During the bioregeneration cycle the maximal nitrite concentration was approximately 10 mg/l, a value much lower than those obtained in the experiments with simulative effluents. Probably the reason for these lower nitrite concentrations was the higher concentration of micronutrients in the real effluents as compared to simulative effluents.

Figure 2 NH4+ concentration profile during bioregeneration with secondary effluent 182

Figure 3 NH4+ concentration profile in the adsorption phase effluent (experiment with secondary effluent)

k30

9/8/00

6:27 pm

Page 183

Figure 5 NH4+ concentration in adsorption phase effluent (experiment with primary effluent)


Figure 4 NH4+ concentration profile during bioregeneration with primary effluent

Table 3 Suspended solids concentration in the influent and effluent in the adsorption phase Effluent type

Secondary Primary

Mean TSS in inflow

Mean TSS in outflow

TSS removal efficiency

Mg/l

mg/l

%

36.1

5

84


(7 samples, SD=1.3)

181

46



75

Bed clogging due to suspended solids accumulation

The original mean diameter of the zeolite particles was 1.5 mm. The mean particle size changed during operation as a result of two phenomena: (1) abrasion during the start-up phase, where the particles were not yet covered by biofilm; (2) establishment of biofilm coverage on the particles (width: 30–40 micron). The above two phenomena resulted in bed stratification: smaller particles with a relatively high biofilm covering and low specific weight were found in the upper part of the reactor and larger particles with a relatively thin biofilm layer at the bottom. The distribution of the particles in the reactor was as follows: about 30% of the volume was occupied by 1.5 mm particles, about 60% was occupied by 1 mm particles, and 10% were particles less than 1 mm. At the top 2–3 centimetres of the reactor, very small zeolite particles covered in a thick layer of biomass were observed. During the adsorption phase these particles formed an impervious layer which did not allow for high downflow rates, and emphasized the necessity for the upflow technique that was implemented in the process. The bed functioned as a fairly good filter for suspended solids in the adsorption phase. This could be problematic for two reasons: head losses may develop as a result of suspended solids accumulation and more important, organic suspended solids might be trapped in the reactor and contribute organic material for heterotrophic development in the bioregeneration phase. Thus, effective backwash was essential for normal operation of the process. It was found that a backwash of two bed volumes with bed expansion of 30% is sufficient to remove all soluble BOD and most (over 95%) of the suspended matter. It should be mentioned that the reactor physical design was not optimized and with proper design it might be possible to reduce the backwash volume needed. Suspended solids concentration in the influent and the effluent in the adsorption phase is presented in Table 3.

183

k30

9/8/00

6:27 pm

Page 184

Table 4 Heterotrophic and nitrifier concentrations and oxygen uptake rate (O.U.R.) results in a reactor fed with actual effluents

Time


End of simulative stage

Heterotrophic

Total biomass

O.U.R.

O.U.R.

(O.U.R.)N

concentration

concentration

heterotrophs

nitrifiers

(O.U.R.)H

mg. VSS /

mg. VSS /

mg. O2 /

mg. O2 /

g. zeolite

g. zeolite

g. zeolite/hr

g. zeolite/hr

(–)

3×10–5

16.6 upper end

0.41

4.43

10.8

0.95

4.45

(–)

(–)

(–)

(–)

(–)

(–)

0.76

6.96

9.15

0.85

5.3

6.23

0.79

5.71

7.22

9.4 lower end After 7 days with

0.09

secondary effluents After 14 days with

0.10 0.13

18.8 upper end 9.8 lower end

0.06




4.68

10.2 lower end


21.0 upper end


0.1

primary effluents


After 18 days with primary effluents

0.09

N/A

Potential heterotrophic competition

The main advantage of separating carbonaceous oxidation from nitrification is the ability to achieve high concentrations of nitrifiers with minimal heterotrophic competition. To study the effect of treating actual primary and secondary effluents, containing both carbonaceous matter and heterotrophic bacteria, the following parameters were monitored during process operation: (1) the change in the total biofilm (heterotrophs + nitrifiers) concentration per gram zeolite; (2) the change in the heterotroph concentration in the biofilm (as determined by growth on specific R2A plates with incubation at 37º C); (3) the changes in oxygen uptake rate of the nitrifiers and the heterotrophs; (4) changes in bioregeneration (or nitrification) rate. The results were compared with results obtained from the simulative effluent runs. The relevant results are presented in Table 4. Evaluation of heterotrophic competition

184

The oxygen uptake rate ratio between nitrifiers and heterotrophs was 10.8 when the reactor was fed with simulative effluent, while a ratio of 6–7 was observed when primary or secondary effluents were treated. These results indicate that the fraction of the heterotrophs from the total population is less that 15%. The total heterotroph concentration increased during the operation from 3×10–5 mg VSS/g chabazite with simulative effluent to an average concentration of 0.1 mg VSS/g chabazite after only seven days, and remained at this concentration throughout the whole operation period with both actual primary and secondary effluents. The total biomass concentration did not change significantly throughout the operation. It remained stable at approximately 10 mg VSS/g chabazite in the lower part of the reactor and about 20 mg VSS/g chabazite in the upper part. The total biomass concentration was similar to that obtained in the experiments with simulative effluents. Based on the bacterial concentrations, the heterotrophic fraction from total biomass was very low (only 1%) which contradicts the oxygen uptake results. This discrepancy probably results from the inaccuracies involved in the techniques for measuring the heterotrophs and nitrifying

k30

9/8/00

6:27 pm

Page 185

Summary and conclusions

A new concept for ammonium removal using zeolite and bio-regeneration has been studied for the treatment of actual secondary and primary effluents. Results showed that the process is capable of high rate ammonium removal and stable performance. Nitrification rate in the bioregeneration phase was stable throughout the experimental period at approximately 7 g N/l reactor/d, similar to the rate obtained using simulative effluents containing no organic material. This result indicates no significant heterotrophic competition in the bioregeneration phase. In addition, the heterotrophic bacteria concentration on the zeolite carrier was found to be constant throughout the experimental period, at no more then 15% of total bacteria population based on oxygen uptake rate tests and bacterial count. During operation, no bed clogging or channeling was observed, due to an effective backwash practiced at the end of each adsorption phase. Results derived from both continuous and batch experiments indicated a reduction in the ion exchange rate (a maximal drop of 30% in the time to breakthrough) for the biofilm covered chabazite as compared to that of virgin chabazite (results are given elsewhere – Lahav and Green, 2000).


bacteria (only partial removal of the biomass from the carrier by VORTEX – see Materials and Methods). Results of both oxygen uptake rate and the bacterial counts showed that the increase in the heterotrophs fraction was minimal. Moreover, the overall bioregeneration rate did not change during the operation (approximately 7 g N/litre reactor/day), indicating no deterioration in nitrifying bacterial activity. During the bioregeneration cycle, the COD did not exceed 40 mg/l and was nearly constant throughout the cycle. At these conditions, the NH4+–N/COD ratio at the beginning of the bioregeneration phase was approximately 10:1, a ratio that favors nitrifiers growth over heterotrophic growth.

References Lahav, O. and Green, M. (1998). Ammonium removal using ion exchange and biological regeneration. Wat. Res., 32(7), 2019–2028. Lahav, O. (1998). A physical – chemical – biological process for the removal of nitrogen compounds from secondary effluents. Doctoral Thesis, Technion – Israel Institute of Technology. Lahav, O. and Green, M. (2000). Bioregenerated Ion Exchange Process: The Effect of the Biofilm on the Ion Exchange Capacity and Kinetics. Water SA, 26(1), 51–58.

185