Info

Number of days

Figure 5.19 Application, during the first 6 days, of waste water containing 30 mg/l NH4+ - N, with a reactor of previously unused clinoptilolite. In the first three days, a breakthrough of nitrate was observed.

The removal efficiency of the simultaneous nitrification and denitrification (SND) and the effluent concentration of ammonium-N and nitrate-N were measured throughout this run. Table 5.14 summarizes the results of RUN 1 where an organic carbon source was applied to the system in stoichiometrically correct amounts. Table 5.15 shows some examples of mass balance for the simultaneous nitrification and denitrification. The amount of SND is equal to the amount of ammonium-N which in the reactor is totally converted via nitrate-N to nitrogen gas (N2).

Figure 5.20 shows the relation between the loading of ammonium-N in kg N/m3 voidvolume* day versus the simultaneous nitrification and denitrification (SND) reaction rate in kg N/m3 * day. The SND reaction rate as bed volume, is calculated in kg N/m3 * day, as:

SND(kgNlm3bed volume * day) =[N,J-\NJ*-^-*8/30 (5.26)

where: HRT is the hydraulic Retention Time, in hours. Ninfl = NH/-Ninfl+N02 -Nin(1 + N03 - N lnfl Neff = NH4+ - Nef( + N02- - Ne(f + N03 - N e„

The factor 8/30 is the conversion factor between void volume and bed volume

The maximum amount of SND obtained during RUN 1 was 13.5 kg N/m3 void volume * day (= 3.6 kg N/m3 * bed volume * day) and the efficiency of SND is up to 99 % with a loading of up to 14 kg N/m3 void volume * day under the following conditions: temperature 20 °C, DO 2-3 mg/l, pH 7.7-7.8 and stoichiometric addition of organic compound as ethanol to obtain denitrification.

Because Fig. 5.20 yields a linear relationship between loading and SND the theoretical maximum amount of SND is not found. The linear equation (5.27) can be obtained from Fig. 5.20 by linear regression, r2 for the linear regression is 0.99.

Data obtained from several sampling ports along the reactor presented in Fig. 5.17 yield the following equation estimating the simultaneous nitrification and denitrification, at a specific height "z" along the reactor in kg N / m3 bed volume * day, knowing the SND at height "i".

SND kg N / m3 voidvolume ' day

Loading kg N / m3 voidvolume * day

Figure 5.20 Relation between the loading of ammonium-N in kg N/m3 bedvolume* day versus the simultaneous nitrification and denitrification (SND).

For a different substrate concentration k may be inserted in equation (5.28).

For ammonium concentration in the influent below 100 mg/l NH4" - N ; k=0.080 (s=18.1%). Between 100 and 500 mg/l NH4" - N; k=0.065; (s=29.2%) and between 500 and 1000 mg/l NH4" - N; k=0.034 (s=21.9%). s is the standard deviation.

Table 5.15 Mass balance of SND using clinoptilolite as media.

Input

unit mg/l

Total N

30,2

100,4

500,7 1000,4

Output

Removed by SND

Total

30,2

100,4

500,7 1000,4

* The removal by SND is found as the difference between influent and effluent waste water samples.

The relationship between the amount of SND removal in mg/l and the hydraulic retention time (HRT) is shown in Fig. 5.21, for the different feed concentrations shown in Table 5.15. From Fig. 5.21 it can be seen that the SND removal does not decrease by reduction of HRT with the ratio of HRT used during RUN 1. A greater flow through the reactor increases the daily removal capacity of the clinoptilolite medium, and the efficiency remains the same. This may be explained by considering that the surface that the bacterial population can occupy on the media is so large, that the maximum utility of the surface is not reached during the experiments. Therefore the use of a lower HRT and thus higher daily loading, yields a larger biomass, and therefore a higher capacity. Perhaps a higher flow can mechanically wash out the dead and older bacteria from the reactor and thereby also provide new surface for fast development of a fresh, new biofilm.

Table 5.16 Efficiency of SND during RUN 1, using clinoptilolite as media.

Flow

Retention

Loading

Temp

DO

PH

nh;

-N

NO,

-N

SND

Effici

SND

L/h

time

kg N/m3

°C

mg/l

Hit

Eft

Effici

mf

Eft

kg N/m1

ency

Kg N/m'

(HRT)

void-

mg/l

mg/l

ency

mg/l

mg/l

void-

%

support

volume

%

volume

media

'day

'day

'day

1.2

6,66

0,10B

20

2,0-3,0

7,7-7,8

30,0

1,49

95,0

0,10

1,30

0,098

90,7

0,026

2.0

4,04

0,178

20

2,0-3,-0

7,7-7,8

30,0

0,22

99,3

0,10

2,12

0,165

92,7

0,044

3,6

2,22

0,324

20

2,0-3,0

7,7-7,8

30,0

0,51

98,3

0,20

0,22

0,319

98,5

0,085

0,9

8,88

0,270

20

2,0-3,0

7,7-7,8

100,0

1,44

98,6

0,40

4,63

0,255

94,4

0,068

1,5

5,33

0,450

20

2,0-3,0

7,7-7,8

100,0

1,61

99,3

0,10

3,95

0,425

70,4

0,110

2,0

4,10

0,585

20

2,0-3,0

7,7-7,8

100,0

0,70

98,6

0,70

6,25

0,412

92,5

0,182

2,5

3,25

0,738

20

2,0-3,0

7,7-7,8

100,0

1,38

98,0

0,10

1,65

0,683

96,9

0,412

5,3

1,50

1,600

20

2,0-3,0

7,7-7,8

100,0

1,96

98.0

0,40

0,49

1.550

99.9

0,598

1,5

5,33

2,250

20

2,0-3,0

7,7-7,8

500,0

0,67

99,9

0,46

0,49

2,248

99,9

0,598

2,6

3,03

3,960

20

2,0-3,0

7,7-7,8

500,0

0,71

99,9

0,42

0,30

3,956

99,9

1,052

4,7

1,70

7,059

20

2,0-3,0

7,7-7,8

500,0

0,72

99,9

0,70

0,23

7,055

99,9

1,877

0,6

10,03

2,392

20

2,0-3,0

7,7-7,8

1000,0

2,23

99,8

0,12

1,97

2,383

99,6

0,634

2,4

3,38

7,100

20

2,0-3,0

7.7-7,8

1000,0

5,51

99,4

0,26

0,66

7,059

99,4

1,878

4,7

1,70

14,118

20

2,0-3,0

7,7-7,8

1000,0

42,3

95,8

0,35

0,70

13,516

95,7

Figure 5.21 HRT vs. SND at different influent concentrations during RUN 1. RUN 2.

The results from RUN 2 were obtained with waste water containing ammo-nium-N, nitrate-N and an organic compound source in form of ethanol, applied to the clinoptilolite reactor. The clinoptilolite medium does not bind nitrate as it does with ammonium.

The aim of this run was to observe if an addition of both ammonium-N and nitrate-N would be converted simultaneously to nitrogen gas.

Table 5.17 shows the efficiency of SND during RUN 2. The reactor was thus able to denitrify both the added amount of nitrate and the amount produced during the nitrification of the added ammonium.

The efficiency of Run 2 was comparable with Run 1, for the influent concentrations applied.

The conclusion is that it is possible to perform denitrification of nitrate added in excess of the nitrate produced by conversion of ammonium by nitrification.

Figure 5.21 HRT vs. SND at different influent concentrations during RUN 1. RUN 2.

The results from RUN 2 were obtained with waste water containing ammo-nium-N, nitrate-N and an organic compound source in form of ethanol, applied to the clinoptilolite reactor. The clinoptilolite medium does not bind nitrate as it does with ammonium.

The aim of this run was to observe if an addition of both ammonium-N and nitrate-N would be converted simultaneously to nitrogen gas.

Table 5.17 shows the efficiency of SND during RUN 2. The reactor was thus able to denitrify both the added amount of nitrate and the amount produced during the nitrification of the added ammonium.

The efficiency of Run 2 was comparable with Run 1, for the influent concentrations applied.

The conclusion is that it is possible to perform denitrification of nitrate added in excess of the nitrate produced by conversion of ammonium by nitrification.

Table 5.17 Efficiency of RUN 2.

Flow

Retention

Loading

nh4*-

n

n03'-

N

SND

time

kg N / m3

Inf

eff

Inf

eff

kg N/m3

(hrt)

matrix per day

mg/1

mg/1

mg/1

mg/1

matrix per day

1,2

6,67

0,059

30,0

0,28

30,8

7,59

0,051

1,2

6,67

0,043

30,0

0,26

14,1

0,10

0,042

3,0

2,67

0,501

30,0

0,88

177,9

105,3

0,244

1,2

6,67

0,501

30,0

1,41

99,0

30,5

0,377

0,9

8,89

0,150

30,0

1,29

70,1

0,10

0,095

1,2

6,67

0,055

30,0

0,12

27,3

0,20

0,055

Temp;

2Ö1 C, dü; 2-3 mg/1

in bulk solution,

pH = 7.7

-7.8. Nitrite-N

was not detected in the samples.

RUN 3

During RUN 3, ammonium-N was added to the clinoptilolite reactor without any continuous additon of a carbon source for denitrification. This should make it possible to recover the nitrate or nitrite produced during the nitrification process. The aim of this run was to be able to determine the efficiency of the nitrification process alone. For RUN 1 and RUN 2 the efficiencies of nitrification and denitrification are difficult to separate.

The clinoptilolite applied was fresh, to ensure that no organic compounds were left from previous experiments which would lead to uncertainty about to obtained results.

Nitrosomonas, can however (as the only bacteria in the biofilm) use C02 from the atmosphere to synthetize biomass (La Cour Jansen and Henze 1990).

Three times, during the 36 days of the test period, a shock-load of organic compound, in the form of ethanol, was added. On each occasion it resulted in a sudden development of the SND process.

Figure 5.22 shows the amount of ammonium-N nitrified, the amounts of produced nitrite-N and nitrate-N. Only between 25 and 30 percent of the loaded ammonium-N is nitrified. This low yield of nitrification is presumably due to the following two factors.

1) At low pH it is difficult to obtain sufficient biomass concentration to convert the amount of applied ammonium-N.

2) Because of the high nitrite-N and nitrate-N concentrations, the nitrification process can be inhibited by its own products.

At day 13 and 22 respectively 50 mg/l and 100 mg/l of COD (organic compound) were added to the waste water. At days 33, 34 and 35 1000 mg/l of COD were added.

Figure 5.22 indicates that the concentration of nitrite-N produced during the test period, was subject to great fluctuation. At the two first COD additions the concentration of nitrite-N increased and, therefore, at least some of the added COD were used to produce Nitrosomonas biomass.

The amount of produced nitrate-N increased when the nitrite concentration had reached its peak-value. This is natural because Nitrobacter (N02" conversion to N03") is not developed until nitrite has been produced. Nitrite, however, both inhibits the nitrification and acts as a substrate for nitrobacter. The second step of the nitrification

At day 13 and 22 respectively 50 mg/l and 100 mg/l of COD (organic compound) were added to the waste water. At days 33, 34 and 35 1000 mg/l of COD were added.

Figure 5.22 indicates that the concentration of nitrite-N produced during the test period, was subject to great fluctuation. At the two first COD additions the concentration of nitrite-N increased and, therefore, at least some of the added COD were used to produce Nitrosemonas biomass.

The amount of produced nitrate-N increased when the nitrite concentration had reached its peak-value. This is natural because Nitrobacter (N02 conversion to N03) is not developed until nitrite has been produced. Nitrite, however, both inhibits the nitrification and acts as a substrate for nitrobacter. The second step of the nitrification process (see Chapter 3) is therefore difficult to initiate.

A change in the biomass concentrations of both nitrosomonas and nitrobacter is therefore observed during the period of nitrate production. If no nitrite is produced, then nitrobacter is not developed due to a lack of the substrate that nitrobacter uses. On the other hand if the nitrite concentration is low, compared to the nitrate concentration, it was observed that both nitrosomonas and nitrobacter occurred in great amounts.

The applied shock-loads of COD seem, therefore, to have three important concequences in this investigation:

1) Maintenance of a fast formation of SND, during about 1 day.

2) Initiation of the development of Nitrosomonas.

3) Offering a carbon source for the synthesis of nitrobacter as soon as nitrite was available as substrate.

On days 34, 35 and 36 of the experiments, higher amounts of COD were added and a more persistent SND was introduced as during RUN 1. Both the amount of nitrite and nitrate therefore declined rapidly because there was a sufficient carbon source for the denitrification process.

Kinetics

A comparison of the nitrogen removal rate for the following submerged filters; the Biocarbone, Biofor and the SND processes, are outlined in Table 5.18 The kinetic rate of the SND process using clinoptilolite as matrix, was about three times higher than for the Biocarbone and Biofor processes, expressed as kg N / (m3 matrix • day).

addition addition COD addition

Figure 5.22 Results obtained during Run 3.

addition addition COD addition

Figure 5.22 Results obtained during Run 3.

Table 5.18 The nitrification rate for the three submerged processes.

Process Maximum Nitrification rate Reference kg N/nf matrix * day

Biocarbone (OTV) 0.74" Rogella el al. (1990)

Biofor (Degremont) 0.75" Paffoni et al. (1990)

SND' 1.7-3.4 Halling-Sorensen and

Hjuler (1992; 1993)

" Only laboratory experiments " Maximum loading 1,00 kg N / m3 matrix per day.

Pilot-plant experiments

In Vantinge on the island of Fyen in Denmark, a pilot-plant has been built following the same concept as presented for experimental RUN 1. The only difference is that the carbon source is not ethanol, but endogeneous carbon from the waste water. Figure 5.23 shows a photo of the pilot-plant. The pilot-plant consists of 80 m3 bedvolume of clinoptilolite distributed in six connected concrete bassins, with upflow waste water and air distribution.

The plant Is used as the tertiary treatment stage for removal of nitrogen from slaughterhouse waste water. As secondary treatment stage, an activated sludge process unit for combined carbon oxidation and nitrification is used. After a secondary clarifier the waste water is pumped into the SND pilot-plant.

The total SND obtained during the first months of pilot-plant experiments were of the order of 0.45 kg N /m3 bedvolume " day and 1.0 kg N /m3 bedvolume * day.

Table 5.19 show the different influent and effluent concentrations found at the pilot-plant.

Table 5.19 Influent and effluent concentration of important parameters at the SND pilot-plant.

Parameter NH/-N N03-N COD

Influent activated sludge treatment

5500

Influent SND tertiary treatment step.

1200

Effluent SND teriary treatment step.

Figure 5.23 Photo of the SND pilot-plant at Vantinge on the island of Fyen.

Figure 5.24 show a cross-section of a clinoptilolite stone. The porosity of the clay stone makes it possible to obtain aerobic and anaerobic conditions simultaneously. On the surface of the clinoptilolite stone oxygen diffuses into the biofilm and is used for the nitrification process. Ammonium is also diffusing towards the biofilm on the clinoptilolite stone. The ion-exchange ability of the stones binds ammonium on the surface (Jorgensen 1976; Haralambous etal. 1992) and nitrifying bacteria converts it to nitrate. The ion-exchange mechanisms may also play an role in the mechanisms, but is not totally clear.

The concentration of nitrate is highest at the upper layer of the biofilm which is most aerobic. Nitrate diffuses to the more anoxic areas in the lower part of the biofilm, where it is denitrified. Because of the concentration gradient a continuous diffusion to the center of the stone will take place.

Figure 5.25 is a micro-scope photo of the clinoptilolite stone covered by an SND bio-film.

Bulk water

0 0

Post a comment