Kinetics and thermodynamics of struvite crystallization as it applies to phosphate recovery from
municipal wastewater for agricultural fertilizer production
Kurt N. Ohlinger1, Thomas M. Young2, Edward D. Schroeder2
1
California State University, Office of Water Programs
6000 J Street, Sacramento, CA 95819-6025. ohlinger@ecs.csus.edu
2
University of California, Davis, Dept. of Civil & Environmental Engineering
One Shields Avenue, Davis, CA 95616
Abstract
In response to struvite accumulation problems at the Sacramento Regional Wastewater Treatment Plant, a
685,000 m3/day facility in California, studies were undertaken to understand the fundamentals of struvite
formation and to develop control mechanisms based on that understanding. The study was conducted in three
phases. Mechanisms affecting struvite solubility were studied in the first phase and an improved model for
predicting struvite saturation conditions was developed. Kinetics of struvite formation was studied in the second
phase to evaluate the cause(s) of preferential accumulation in systems with relatively uniform precipitation
potential. Phase three consisted of utilizing a pilot-scale fluidized bed reactor to precipitate struvite from SSB
supernatant at the SRWTP, incorporating knowledge gained from the solubility and kinetics experiments to
optimize the reactor design.
Introduction
Accumulation of struvite (MgNH4PO46H2O) on pipe walls and equipment surfaces of anaerobic digestion and
post-digestion processes is a problem that plagues the wastewater treatment industry. Struvite is well known for
plugging pipes and fouling pumps, aerators, screens, and other equipment. Remediation is often impractical
and, when possible, is costly in terms of labor, materials, and system downtime (Ohlinger et al, 1998). At the
Sacramento Regional Wastewater Treatment Plant (SRWTP), a 685,000 m3/d (181 mgd) facility in Northern
California, struvite accumulation has rendered extensive piping systems inoperable and has created a
continuous, and costly, maintenance requirement for removing struvite from equipment operating in postdigestion sludge storage basins (SSB). All attempts to remove struvite accumulation from SSB supernatant
system pipe walls have failed and the 5.6 km (3.5 mi.) piping system must be replaced. System replacement,
however, without discovery and implementation of controls to prevent further struvite formation will set the
stage for further struvite accumulation. Similar problems have been reported at treatment facilities elsewhere
(Borgerding, 1972; Mohajit et al., 1989; Horenstein et al., 1990; Mamais et al., 1994).
Approach and Results Summary
In the phase one study, struvite equilibrium experiments were conducted in a controlled environment.
Equilibrium was approached from two directions, formation in supersaturated solution and dissolution of
struvite solid in undersaturated solution. Struvite synthesized in the laboratory and struvite collected from the
field were used in the dissolution experiments. A new value for the struvite solubility constant was derived
from the equilibrium experiments. The derived solubility constant was applied to the conditional solubility
product model for predicting the struvite saturation condition presented in water chemistry texts by Stumm and
Morgan (1970) and Snoeyink and Jenkins (1980). Using the derived solubility constant and including the
effects of ionic strength on ion activity and all complexation species, the model accurately predicted struvite
precipitation potential at the SRWTP. Using the same model with the commonly accepted struvite solubility
constant failed to predict struvite precipitation in the SRWTP system where precipitation was occurring. Both
curves are plotted on Figure 1 with field data from SRWTP SSB supernatant. Figure 1 illustrates
supersaturation conditions accurately predicted by the model using the experimentally derived solubility
constant and undersaturation predicted by the model using the commonly accepted solubility constant.
In phase two, a study of the kinetics of struvite formation, the causes of preferential struvite accumulation in
systems with relatively uniform precipitation potential was investigated. Experiments were designed to study
the rate controlling mechanisms for the two primary components of precipitation, nucleation and growth. A lab
scale investigation of nucleation and growth during the induction period revealed that the induction period was
dominated by the nucleation process. The results further showed nucleation to be a reaction controlled process
dependent primarily on struvite supersaturation level that could not, therefore, cause preferential struvite
accumulation. Figure 2 illustrates the relation of struvite induction time to the supersaturation level. Figure 3
illustrates the relative effects of changing mixing energy and supersaturation on struvite induction time and,
together with Figure 2, supports the conclusion that the induction period is primarily controlled by the reaction
limited nucleation process (Ohlinger et al, 1999).
Influences on struvite growth rate were also investigated in phase two by monitoring the accumulation rate on
surfaces of coupons placed in SSB supernatant at the SRWTP. Mixing intensity, struvite supersaturation level,
CO2 stripping, surface roughness, and coupon material were investigated to determine the relative influence of
each on struvite growth rate. Mixing intensity was determined to exert the greatest influence on struvite growth,
indicating transport limitations on the surface integration rate. The findings were consistent with field
observations of preferential struvite accumulation in areas of high mixing intensity.
Phase three experiments consisted of operation of a fluidized bed reactor (FBR) designed to rapidly precipitate
struvite from process fluids, thereby reducing struvite precipitation potential and preventing struvite
precipitation in downstream processes. The system exploited the struvite solubility decrease associated with pH
increase and the optimum surface integration rate on struvite crystal medium and in a high mixing energy
environment.
The struvite solubility model developed in phase one has wide applicability as a predictive tool for designers
and operators of wastewater treatment processes. The model can be used to assess struvite precipitation
potential and the effectiveness of proposed remediation measures. Results from phase two provide designers
with knowledge of configurations and materials selection to minimize impacts of struvite accumulation by
eliminating conditions conducive to preferential accumulation. Phase three experiments demonstrated that the
FBR process can provide controlled struvite removal from process streams and demonstrated the first order
kinetics of struvite precipitation in an FBR, illustrated in Figure 4 (Ohlinger et al, 2000). Design guidelines for
scale-up of a struvite precipitation system are presented.
References
Borgerding, J. (1972) Phosphate deposits in digestion systems. J. Water Pollut. Control Fed., 44(5), 813-819.
Horenstein, B.K., Hernandez, G.L., Rasberry, G., Crosse, J. (1990) Successful Dewatering Experience at
Hyperion Wastewater Treatment Plant. Wat. Sci. Technol.¸ 22(12), 183-191.
Mamais, D., Pitt, P.A., Cheng, Y.W., Loiacono, J., Jenkins, D. (1994) Determination of ferric chloride dose to
control struvite precipitation in anaerobic sludge digesters. Wat. Environ. Res., 66(7), 912-918.
Ohlinger, K.N., Young, T.M. and Schroeder, E.D. (2000) “Post Digestion Struvite Precipitation Using a
Fluidized Bed Reactor,” ASCE Journal of Environmental Engineering 126(4), 361-368.
Ohlinger, K.N., Young, T.M. and Schroeder E.D. (1999). “Kinetics Effects on Preferential Struvite
Accumulation in Wastewater,” ASCE Journal of Environmental Engineering 125(8), 730-737.
Ohlinger, K.N., Young, T.M. and Schroeder, E.D. (1998). “Predicting Struvite Formation in Digestion,” Water
Research, 32(12), 3607-3614.
Mohajit, Bhattarai, K.K., Taiganides, E.P., Yap, B.C. (1989) Struvite Deposits in Pipes and Aerators, Biol.
Wastes, 30, 133-147.
Snoeyink, V.L., Jenkins, D. (1980) Water Chemistry, John Wiley & Sons, New York.
Stumm, W., Morgan, J.J., (1970) Aquatic Chemistry, John Wiley & Sons, New York.
3.00
pKso=13.26,
u=0.1
4.00
pKso=12.6,
u=0.1
5.00
SSB
supernatant
pPs
6.00
7.00
8.00
9.00
10.00
4.0
6.0
8.0
10.0
12.0
14.0
pH
Figure 1. Prediction of struvite saturation condition in SRWTP SSB supernatant using commonly accepted
model inputs (dashed line) and experimentally derived model inputs (solid line).
4.0
3.5
log (time, sec.)
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0
10
20
30
[log (Sa)]-2
Figure 2. The effect of struvite supersaturation (Sa) on the induction time required to form detectable struvite
precipitate from supersaturated solution at 22C.
70
60
Induction time (sec)
50
40
Sa=2.1
Sa=2.4
Sa=2.7
30
20
10
0
300
500
700
900
1100
Stir Speed (rpm)
Figure 3. The effect of mixing energy and supersaturation (Sa) on the induction time required to form detectable
struvite precipitate from supersaturated solution at 22C.
1.1
1.0
PS Reduction
0.9
Predicted
full media
50% media
0.8
0.7
0.6
0.5
0
1
2
3
4
5
6
7
Hydraulic Detention Time (hr)
Figure 4. First order reaction kinetics of struvite precipitation in a continuous-feed fluidized bed
reactor, illustrated using the reduction in conditional solubility product (PS)
8