Environmental Technology, Vol. 22. pp 1263-1272
© Selper Ltd, 2001
RECOVERY OF AMMONIA AS STRUVITE FROM
ANAEROBIC DIGESTER EFFLUENTS
I. ÇELEN1* AND M. TÜRKER1,2
1Department
of Environmental Engineering, Gebze Institute of Technology, Gebze, Kocaeli Turkey
2Pak-Food Industries, PO Box 149, 41001, Izmit, Turkey
(Received 6 February 2001; Accepted 20 March 2001)
ABSTRACT
The effects of environmental conditions on ammonia removal as struvite (Magnesium ammonium phosphate, MAP) were
studied in a laboratory scale batch reactor. MAP precipitation was carried out by adding phosphoric acid and magnesium
source either as MgCl2 or MgO. The effect of temperature, pH, Mg:N:P ratios were studied. Temperature did not
significantly affect ammonia removal between 25-40 0C and over 90% removal was obtained. The effect of pH, however, was
significant and highest removal was reached at pH 8.5-9.0. The various stoichiometric ratios of ammonium to Mg and P
have been tested and slight excess of Mg and P found to be beneficial for higher recovery of ammonia as struvite. However
further increase in Mg and P ratios did not result in further ammonia removal which is also costly for the practical
application of the process. When MgO was used as M source, the ammonia recovery was 60-70% whereas the use of MgCl2
has increased this figure up to 95%. In addition a two step purification process was developed to recover MAP crystals from
impurities of the anaerobic digester. Firstly, precipitates were dissolved in acid and impurities were removed by
centrifugation. The clarified supernatant was re-precipitated by adjusting its pH with caustic. It was shown that in the
two steps process white MAP crystals could be obtained with over 85% recovery to be used for another applications.
The economical analysis of the process has shown that ammonia in the digester effluents can be recovered at the cost of
$7.5-8.0 kg-1 NH4+-N. The rate of reaction is very fast and is completed almost in minutes. This simplifies the process design
resulting in a smaller reaction vessel.
Keywords:
Struvite, magnesium ammonium phosphate, ammonia recovery, fertilizer, precipitation.
INTRODUCTION
Nitrogen compounds are present in some industrial as
well as in domestic wastewaters in significant quantities 1.
Due to human impact on the nitrogen cycle, the reactive
nitrogen compounds are accumulating on earth causing
eutrophication in receving waters and deterioration of water
quality [2]. Significant costs are associated with extra
treatment required to reduce discharge concentrations. There
are a number of physicochemical and biological techniques
available for the treatment of nitrogen-containing waste
streams.
The
techniques
such
as
biological
nitrification/denitrification and breakpoint chlorination
reduce nitrogen compounds to dinitrogen gas. However,
alternative technologies exist to convert ammonia into a
reusable and saleable useful product. One such technology is
the recovery of ammonia as struvite which has a potential as a
fertilizer.
Struvite or magnesium ammonium phosphate (MAP)
precipitate in the presence of Mg2+ (M), NH4+ (N) and
PO43- (P) according to following reaction when the
thermodynamic solubility product, Ks, is exceeded:
Mg2  NH 4  PO43  6H 2 O  MgNH 4
PO4 .6H 2 O
(i)

Ks  Mg
2
NH PO 

4
3
4
(ii)
However, it has been shown that in crystalization
experiments, the precipitation of struvite reduces the pH
which suggests that HPO42- would precipitate in the reaction
rather than PO43- according to following reaction:
Mg 2   NH 4  HPO42  6H 2 O  MgNH 4 PO 4 .
6H 2 O  H 
(iii)
In the literature, several papers have addressed the
recovery of ammonia or phosphate as struvite from industrial
and domestic wastewaters [3-7]. The precipitation of struvite
may present some problems in wastewater treatment plants
causing deposits in pipe walls [8, 9]. However, struvite has a
1263
potential use as a fertilizer. It has been shown to be a highly
effective source of nitrogen, magnesium and phosphorus for
plants and can be used as a slow release fertilizer at high
application rates without damaging plant roots [10, 11].
The time course of the reaction between Mg2+ (M), NH4+
(N) and PO43- (P) to form struvite has been studied in order to
establish required equilibrium time. The time course of
Table 1. The approximate composition of wastewater.
MATERIALS AND METHODS
Struvite precipitation experiments were carried out in a batch
reactor with a volume of 200 ml mixed with a magnetic stirrer.
The temperature of the reactor is controlled at the desired
value by thermostatic controller. The pH measurements are
made with pH meter E588 Metrohm Herisau and pH is
adjusted either with HCl or NaOH solutions. The chemicals
used were commercial grade and their contents were
determined before the experiments. All analyses were carried
out according to Standart Methods [12].
After each
experiment, the supernatants and the precipitates were
analysed for Mg, N and P to check the consistency of
experimental results and all balances closed within acceptable
limits.
RESULTS AND DISCUSSIONS
The Composition of Wastewater
The wastewater used in this work to recover ammonia
is the effluent of an anaerobic digester treating molases-based
industrial wastewater, the chemical composition of which
is given in Table 1. The effluent contains approximately
1400 mg l-1 ammonia and negligible amounts of phosphate
and magnesium. Therefore they are added in stoichiometric
quantities in struvite precipitation studies. Calcium was also
present in negligible quantities. However, potasium
concentration was not negligible at around 2150 mg l-1.
The Effect of Time
Parameter
NH4+
Mg2+
PO43Ca2+
K+
COD
pH
Concentration in
water (mg l-1)
1400
21.4
24
21.2
2150
3240
7.9
struvite precipitation has been followed by analysing
ammonia concentration in the supernatant when both MgCl 2
and MgO were used as Mg source, as shown in Figure 1.
The ammonia concentration reached its equilibrium
value immediately after the mixing with phosphate and
magnesium source with either MgCl2 or MgO and remained
constant during the rest of the test period. Therefore, forty
minutes reaction time has been accepted as a safe time for the
reaction to proceed to equilibrium.
Effect of pH
Ammonia nitrogen is distributed between NH4+ and NH3 as a
function of pH according to the following equilibrium
reaction:


NH3  H2O  NH4 OH
where the equilibrium constant, Ka, is defined as [13]:
1264
(iv)
Ka 
Figure 1. Time course of struvite precipitation.
Table 2. Ammonium lost to air after 40 minutes at different
10
(v)
 5.7 *10
pH.
NH3H3O 
NH 

4
The high pH values favor NH3 which is in equilibrium
with air according to Henry’s Law and facilitates air stripping
of ammonia. The effluent containing NH4+ (1354 mgl-1) at 37
0C was kept at pH 7.9, 8.5 and 9.0 for 40 minutes. The results
of the experiment are shown in Table 2 At pH 9.0 17.9% of
ammonia is lost to air whereas at the same condition no loss of
ammonia is determined at pH 7.9. The effect of pH on
ammonia recovery as struvite was studied at various
stoichiometric ratios of ammonia to Mg and P. The results are
presented in Figure 2 when MgCl2 was used as magnesium
source. As the pH of the medium increased, the percent of
ammonia removed as struvite increased as a function of
stoichiometric ratios of Mg:P relative to ammonia.
Reasonably high removal ratios were obtained when Mg:N:P
ratio was 1.2:1:1.2 even at pH as low as 6.0.
pH
NH4+
(mgl-1)
Ammonia
lost to air (%)
7.9
8.5
9.0
1354
1205
1112
11
17.9
Effect of Mg:N:P Ratios
According to reaction (i) magnesium, ammonium and
phosphate are required in equimolar quantities to form MAP.
However, the experimentally obtained ratios may differ for
optimum (or better) ammonia removal as struvite due to the
presence of some other species present in the effluent that
may form by-products. The results obtained with various
Mg:P ratios relative to ammonia are presented in Figure 3.
1265
Figure 2. The effect of pH on ammonia removal at different Mg:N:P ratios.
Figure 3. The effect of relative molar ratio of Mg:P to N on ammonia removal.
While a slight excess of Mg and P resulted in better removal of
Effect of Magnesium Source
ammonia, the ratios above 1.2 did not yield higher recovery of
ammonia as struvite. Similarly, N and P were kept at
In addition to MgCl2, MgO has been tested as an alternative
equimolar concentrations and Mg concentrations were
Mg source. MgO is mentioned in the literature as a potential
increased to improve ammonia recovery. This resulted in
Mg source mainly due to its cheaper price and its basic
improved recovery of ammonia as a function of pH as shown
character in pH adjustment [14].
in Figure 4. Similarly, this time Mg and N were kept at the
When MgO was used as Mg source, N:P ratio was kept
same molar ratio and the relative ratio of P was increased.
at 1:1.2 and Mg ratio was changed from 1.2 to 1.6. As the M
However this experimental protocol did not yield better
ratio increased, relative removal ratio of ammonia increased
ammonia recovery at or closer to optimum pH 9.0 as shown in
from 60% to 70%. Both Mg sources were also compared
Figure 5.
under identical conditions. In these experiments Mg:N:P
ratios were
1266
Figure 4.
The effect of relative molar ratio of Mg to N:P on ammonia removal.
Figure 5.
The effect of relative molar ratio of P to Mg:N on ammonia removal.
maintained at 1.2:1:1.2 respectively and the results are
In addition, the caustic consumptions were compared
presented in Figure 7. The ammonia recovery as struvite
for both Mg sources for pH adjustment and the results are
reached over 95% when MgCl2 was used whereas when MgO
shown in Table 3. Less caustic was consumed with MgO for
was used rather lower values between 48% to 58% were
pH adjustment indicating the more basic character of MgO in
obtained, similar to the results presented in Figure 6.
comparison to MgCl2.
1267
Figure 6. The effect of molar ratio of MgO to fixed N:P at 1:1.2 on ammonia removal.
Figure 7. The comparison of MgCl2 and MgO on ammonia removal at Mg:N:P ratios of 1.2:1:1.2.
Table 3.
The effect of Mg source on caustic consumption for pH adjustment.
NaOH consumptions (30%) (ml)
Mg source
MgCl2.6H2O
MgO
pH 8.0
pH 8.5
pH 9.0
5.5
2.25
5.75
2.35
6.0
3.5
Effect of Temperature
The temperature of effluent may affect the equilibrium
composition of struvite yielding different ammonia
recoveries. There are a number of conradicting experimental
results reported in the literature regarding the effect of
temperature [14]. Therefore the influence of temperature
on ammonia removal as struvite has been studied between 2540 0C at Mg:N:P ratio 1.2:1:1.2. It was found that temperature
has a negligible influence on ammonia precipitation as
struvite between the temperature range studied as shown in
Figure 8. Almost over 95% ammonia recovery as struvite was
obtained over the range of temperatures studied.
1268
Purification and Recovery of Struvite
In the struvite precipitation experiments, struvite
precipitated together with impurities present in the effluent.
If the purified struvite is required for a particular application,
this can be achieved with a two step purification process.
Here we have studied clarification and mass balance of
precipitated struvite obtained from coloured effluent
containing impurities. The flow diagram and results of mass
balance are shown in Figure 9 and the photograph of the steps
is shown in plate-1.
Figure 8. The effect of temperature on ammonia removal.
Figure 9. Purification and mass balances of struvite recovery in the two step process.
1269
Plate 1.
The photograph of the two steps purification and recovery of struvite.
The original effluent was coloured and white struvite
crystals were precipitated together with impurities in the first
step. After discarding the supernatant and dissolving the
precipitate in the required volume of acid, the impurities were
removed from the suspension by centrifugation. Finally,
clarified and dissolved struvite can quantitatively be
recovered by adjusting its pH by addition of caustic solution.
The mass balance of this purification process has shown that
85.8% of the ammonia present in the effluent could be
recovered as white struvite crystals after the two steps
clarification procedure. In fact the supernatant of the final
step containing equilibrium concentrations of Mg, N and P
can be recycled back to the first step to increase the efficiency
of the recovery process.
Economic Analysis of the Process
The cost of ammonia recovery as struvite was studied
based on the experimental results presented here, in order to
assess the economical viability of the process in comparison to
existing nitrogen removal technologies. In this preliminary
assessment, investment and utility costs such as electricity
and water etc. are not taken into account and only the cost of
chemicals, Mg source, P source and caustic have been
considered in the calculations as exemplified in figure 10. The
commercial value of struvite was not considered, however,
the market price of struvite will determine the applicability of
this process in practice. The market prices of the chemicals
used in the calculations are given in Table 4. The results of
the economical analysis for the experimental results presented
here are given in Table 5 and Table 6 where the contribution
of each chemical to overall cost of struvite precipitated is
calculated. The cost of ammonia removal is the function of
removal ratio and cost of the chemicals added per ammonia
present in the effluent. The relative cost of MgCl2 is highest
compared to those of H3PO4 and caustic when it is used as Mg
source, amounting to approximately 55-65% of the overall
cost. However the cost of the process per ammonia fixed as
struvite depends on the removal ratio as well as the
stoichiometric ratio of chemicals used. When MgO is used as
Mg source, main cost factor is the cost of H3PO4 since the price
of Mg from MgO is relatively cheaper than that of MgCl 2.
However, the cost of the process per ammonia removed as
struvite does not change much since the recovery of ammonia
is relatively low in comparison to the MgCl2 used. In order to
reduce the overall cost of the process a cheaper caustic source
such as Ca(OH)2 can be used for pH adjustment. However,
there is a risk of calcium phosphate precipitation with much
higher pKs value than struvite which eventually reduces the
availability of phosphate for struvite thus increasing the need
for phosphate.
The economical analysis based on the experimental
results presented in this work falls into the same range carried
out by other researchers. Siegrist et al [15] considered
electricity and maintenance in addition to chemical cost of the
process and came up with $9.10-11.38 kg-1 NH4+-N. Andrade
and Schuiling [14] claimed that the cost is between $4.55-9.92
kg-1 NH4+-N at high ammonium concentrations. Siegrist [16]
has calculated the cost of the process $9.72 kg-1 NH4+-N.
1270
Figure 10.
The simplified diagram for the calculation of cost of struvite from the experimental results.
Table 4.
Table 5.
Market prices of the chemicals used in the experiments.
Chemical
Price ($ kg-1)
H3PO4 (75%)
MgCl2.6H2O
MgO (85%)
NaOH (100%)
NH4+
0.40
0.31
0.44
0.12
0.23
Economic analysis of the process when MgCl2.6H2O is used as Mg source.
Mg:N:P Molar Ratio and pH
NH4 Removal (%)
Cost
($ kg-1 NH4+-N)
Cost
($ kg-1 MAP)
Cost of the Chemicals (%)
H3PO4 : MgCl2 : NaOH
1:1:1 and pH 8.0
1:1:1 and pH 8.5
1:1:1 and pH 9.0
1.2:1:1 and pH 8.0
1.2:1:1 and pH 8.5
1.2:1:1 and pH 9.0
1.4:1:1 and pH 8.0
1.4:1:1 and pH 8.5
1.4:1:1 and pH 9.0
1.2:1:1.2 and pH 7.5
1.2:1:1.2 and pH 8.0
1.2:1:1.2 and pH 8.5
1.2:1:1.2 and pH 9.0
1:1:1.2 and pH 7.5
1:1:1.2 and pH 8.0
1:1:1.2 and pH 8.5
1:1:1.2 and pH 9.0
78.7
83.4
90.2
88
86
94.5
88.5
87.4
96.9
93.2
94.3
97
95.36
81.6
82
82.6
83.8
7.72
7.48
6.98
7.89
8.13
7.45
8.8
8.79
7.99
7.96
7.95
7.7
7.9
8.3
8.31
8.02
7.99
0.62
0.66
0.47
0.55
0.61
0.49
0.6
0.62
0.48
0.52
0.51
0.5
0.5
0.65
0.71
0.6
0.63
35.5 : 56.7 : 7.8
35.3 : 56.4 : 8.3
34.8 : 55.3 : 9.9
31.7 : 60.5 : 7.8
31.4 : 60.2 : 8,4
31.2 : 59.4 : 9.4
28.3 : 62.9 : 8.8
28.5 : 64 : 7.5
28.2 : 63 : 8.8
35.2 : 56 : 8.8
34.8 : 55.4 : 9.8
35 : 55.6 : 9.4
34.6 : 55.2 : 10.2
35.4 : 56.3 : 8.3
35.2 : 56.1 : 8.7
35.4 : 56.3 : 8.3
35.4 : 56.3 : 8.3
1271
Table 6.
Economic analysis of the process when MgO is used as Mg source.
Mg:N:P Molar Ratio and pH
NH4 Removal (%)
Cost
($ kg-1 NH4+-N)
Cost
($ kg-1 MAP)
Cost of the Chemicals (%)
H3PO4 : MgO : NaOH
1.2:1:1.2 and pH 8.0
1.2:1:1.2 and pH 8.5
1.2:1:1.2 and pH 9.0
1.4:1:1.2 and pH 8.0
1.4:1:1.2 and pH 8.5
1.4:1:1.2 and pH 9.0
1.6:1:1.2 and pH 9.0
48.9
54.4
57.8
53.2
64.9
66.9
68.8
8.07
7.26
7.14
8.14
6.73
6.59
6.4
0.84
0.79
0.75
63.2 : 28.8 : 8
63.2 : 28.8 : 8
60.5 : 27.5 : 12
60.9 : 31.3 : 7.8
60.4 : 31.1 : 8.5
59.8 : 30.7 : 9.5
59.6 : 30.6 : 9.8
In all these works, including ours, the commercial value of
struvite is not taken into account. Webb et al [17] has
calculated the cost of chemicals as $14.9kg-1 NH4+-N. When
the market price of struvite being $8.75 kg-1 NH4+-N is
deduced from the cost, the actual cost comes down to
$6.15 kg-1 NH4+-N. As a result, local availability and the price
of chemicals and use of struvite determine whether this
process can potentially compete with existing nitrogen
removal and recovery technologies currently available on the
market.
2.
CONCLUSIONS
7.
The following conclusions can be drawn from the presented
work:
8.
1.
3.
4.
5.
6.
The rate of struvite precipitation is very fast and is
completed in minutes.
The optimum pH for ammonia recovery is 8.5-9.0 over
the range studied here. However, the higher the pH, the
higher the loss of ammonia to air due to stripping since
high pH favors NH3. Struvite precipitation has also been
observed as low as pH 6.0
For optimum recovery of ammonia, slight excess of Mg
and P are required.
MgCl2 is a better Mg source than MgO. However less
caustic is used for pH adjustment when MgO is used.
The reaction temperature did not have an influence
on the equilibrium concentration of ammonia between
25-40 oC.
In a two step purification process struvite can
quantitatively be purified from impurities present in the
effluent.
The cost of ammonia recovery is around $7.5-8.0 kg-1
NH4+-N as a function of removal ratio and amount of
chemicals added. This process can be viable when the
value of struvite is taken into account.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
Wiesmann, U., Biological nitrogen removal from wastewater, Adv. Biochem. Eng. Biotechnol., 51, 113-154 (1994).
Galloway, J.N., Schlesinger, W.H., Levy, H., Michaels, A. and Schnoor, J.L., Nitrogen fixation: antropogenic enhancementenvironmental response, Global Biochem. Cyc., 9, 235-252 (1995).
Battistoni, P., Fava, G., Pavan, P., Musacco, A. and Cecchi, F., Phosphate removal in anaerobic liquors by struvite
crystallization without addition of chemicals: preliminary results, Water Res., 31, 2925-2929 (1997).
Battistoni, P., Pavan, P., Prisciandaro, M. and Cecchi, F., Struvite crystallization: a feasible and reliable way to fix
phosphorus in anaerobic supernatants, Water Res., 34, 3033-3041 (2000).
Tünay, O., Kabdasli I., Orhon, D. and Kolçak, S., Ammonia removal by magnesium ammonium phosphate precipitation in
industrial wastewaters, Water Sci. Technol., 36, 225-228 (1997).
Shin, H.S. and Lee, S.M., Removal of nutrients in wastewater by using magnesium salts, Environ. Technol., 19, 283-290
(1997).
Schulze-Rettmer, R. The simultaneous chemical precipitation of ammonium and phosphate in the form of magnesium
ammonium phosphate, Water Sci. Technol., 23, 659-667 (1991).
Snoeying, V.L. and Jenkins, D., Water Chemistry, John Wiley and Sons Inc., New York, (1980).
Loewenthal, R.E., Kornmuller, U.R.C. and Van Heerden E.P., Struvite precipitation in anaerobic treatment systems, Water
Sci. Technol., 30, 107-116 (1994).
1272
10. Gaterell M.R., Gay R., Wilson R., Gochin R.J. and Lester J.N. An economic and environmental evaluation of the
opportunities for substituting phosphorus recovered from wastewater treatment works in existing UK fertiliser markets,
Environ. Technol., 21, 1067-1084, (2000).
11. Priestley, A.J., Cooney, E., Booker, N.A. and Fraser, I.H., Nutrients in wastewaters-ecological problem or commercial
opportunity. In: proceedings of the 17th federal convention of the australian water and wastewater association, Melbourne,
1, 340-346 (1997).
12. APHA, Standard Methods for the Examination of Water and Wastewater, 17th ed. Amer. Public Health Assoc., Washington, D.C.
(1989).
13. Stumm, W. and Morgan, J.J., Aquatic Chemistry, Wiley-Interscience, New York (1996).
14. Andrade, A and Schuiling, O., Literature Search, Digest and Assessment Concerning Struvite Recovery for Recycling from
Waste Waters, Final Report for CEEP (CEFIC)., Geochem Research BV / Utrecht University Holland (1999).
15. Siegrist, H., Gajcy, D., Sulzer, S., Roeleveld, P., Oschwald, R., Frischknecht, H., Pfund, D., Morgeli, B. and Hungerbuhler, E.,
nitrogen elimination from digester supernatant with magnesium-ammonium phosphate precipitation, In: Chemical Water
and Wastewater Treatment II, (Eds.) Klute, R. and Hahn, H., Springer Verlag, Berlin, 457-464 (1994).
16. Siegrist, H., Nitrogen removal from digester supernatant-comparison of chemical and biological methods, Water Sci.
Technol., 34, 399-406 (1996).
17. Webb K.M., Bhargava S.K., Priestley A.J., Booker N.A. and Cooney E., Struvite (MgNH 4PO4 .6H2O) precipitation: potential
for nutrient removal and re-use from wastewaters. Chemistry in Australia, 62, 42-44 (1995).
1273