effects of storage on phosphorus recovery from urine

4 downloads 11939 Views 155KB Size Report
(UBC) in Vancouver, Canada, crystallizes small pellets of garden-ready ... storage of urine in the context of struvite recovery, i.e. how do different ..... These data.
Environmental Technology, Vol. 29. pp 807-816

© Taylor & Francis, 2008

EFFECTS OF STORAGE ON PHOSPHORUS RECOVERY FROM URINE E. TILLEY1*, J. ATWATER2 AND D. MAVINIC2

1

2

Eawag: Swiss Federal Institute of Aquatic Science and Technology, Überlandstrasse 133, 8600 Dubendörf, Switzerland Department of Civil Engineering, University of British Columbia, 6250 Applied Science Lane Vancouver, BC, V6T 1Z, Canada

(Received 1 November 2007; Accepted 27 May 2008 )

Taylor and Francis Ltd

10.1080/09593330801987145

ABSTRACT In laboratory experiments using synthetic urine the effect of temperature, faecal contamination, dilution and headspace on urine to be used as a feedstock for struvite recovery were examined. The effects of adding different quantities of magnesium on the amount of phosphorus that could be removed from solution was also examined. An average of 62% of phosphorus could be removed in the form of struvite when magnesium was added to the urine solution after ureolysis had forced the precipitation of calcium and magnesium minerals. Dilution and the presence of faecal urease were found to affect the rate of ureolysis but not the purity of the struvite recovered. These results indicate that, by simply storing urine until it achieves a pH of 8 or greater, struvite can be recovered from source-separated urine with only a magnesium addition.

Keywords: Urine, struvite, MAP, phosphorus, sanitation

[7, 8, 9]. Although ubiquitous, the enzyme urease is concentrated in faecal matter and catalyzes the hydrolysis of urea to ammonia and carbon dioxide by the simplified reaction:

INTRODUCTION

Phosphorus, one of the three essential nutrients for plant growth, is a finite resource that is mined from natural deposits – deposits that are quickly disappearing. It has been projected that half of the current economically recoverable phosphate resources will be exhausted in 60–70 years [1]. To address this coming shortage, innovative technologies have been developed to recover phosphorus from human sewage – the same sewage that is often responsible for the eutrophication of water bodies [2]. Those technologies require that the pH be raised with a caustic and a source of magnesium be added so that phosphorus can be precipitated in the form of struvite. Struvite (MgNH4PO4· 6H2O) is a bioavailable phosphate fertilizer that has been recovered from wastewater and sold in Japan since the 1990s [3]. Technology developed at the University of British Columbia (UBC) in Vancouver, Canada, crystallizes small pellets of garden-ready struvite [4]. However, in both cases, struvite is being harvested from dilute waste streams and requires the addition of both caustic and magnesium. Urine, which accounts for only 1% of the volume of wastewater [5], actually contains over 70% of the N and over 90% of the P excreted by humans, making it an ideal feedstock for struvite recovery [6]. Struvite crystals form naturally in urine (kidney stones) when urine becomes supersaturated with ammonia as a result of urea hydrolysis by bacterial urease

CO( NH 2 )2 + H 2O → 2 NH 3 + CO2

(i)

In urine, the ammonia that is generated causes the pH to increase to approximately 9 [10]. The possibility of recovering struvite from urine was proposed as early as 1996 [11], but only now is the technology catching up. Toilets that allow for the collection of urine separately from faecal matter are gaining acceptance, as is the idea that urine itself could be used a fertilizer [5, 12, 13]. Some work has addressed the influence that different storage conditions (e.g. temperature, faecal contamination) have on the dynamics of urine, but none has examined the storage of urine in the context of struvite recovery, i.e. how do different storage conditions affect the quality of struvite that can be recovered [10, 14, 15]. As well, some researchers have recovered struvite from fresh urine but, for most separating and collection systems, fresh urine would be difficult, if not impossible, to obtain [16, 17]. The goal of this work was to better understand the potential of urine for use as a feedstock for struvite recovery. An important element of this work was the attention paid to calcium. Calcium is known to be one of the main impurities in struvite [18], and its presence can inhibit the formation of magnesium phosphates [19]. Different storage conditions

807

and magnesium dosing regimes were evaluated as ways to optimally prepare urine solutions such that a maximum amount of pure struvite could be recovered. To do this, two specific studies were conducted to more closely examine different facets of urine preparation and storage. There were two parts to this study: the goal of Part A was to verify the effects of temperature, contamination, dilution and headspace on stored urine and, in Part B, the effect of variable magnesium additions on urine that had undergone ureolysis was determined.

conditions. To examine the effect of temperature, headspace, dilution and faecal contamination, a 32-solution matrix was used. Temperature affects biological activity (i.e. urease production) and the rate of ureolysis [20]. The amount of headspace influences the partial pressure of ammonia and may alter the amount that volatilizes [10]. Dilution is known to affect the final composition of stored urine [21] and highly concentrated solutions are sometimes slower to undergo reactions due to the shielding effect of ions. Urine is easily contaminated by faecal matter during collection. In this research, wastewater served as a proxy for faecal matter [22]. Two levels of temperature (4°C and 23°C), two levels of nominal dilution (no dilution and dilution by 50% with distilled water), two types of headspace systems (open jar and closed jar) and four levels of faecal contamination (0%, 5%, 10% and 25% wastewater (WW) by volume) were used. The matrix of the 32 urine solutions is shown in Table 2. Solutions #1–16 were stored at 4°C and solutions #17–32 were stored at 23°C. Both sets of 16 were prepared in the exact same manner. Each solution was prepared in a one litre Nalgene, wide-mouth jar with a screw-top lid. The jars were soaked in bleach and rinsed with distilled water to remove any biological contaminants. Half the solutions were used at full-strength and half were diluted by 50% with distilled water. Solutions that were “closed” were stored with lids on, except during sampling, and “open” solutions were left open for the duration of the experiment. Primary effluent (or “wastewater”) from the UBC wastewater treatment pilot plant was used as a source of faecal contamination. Different amounts (0, 50, 100 or 250 ml) of the initial volume (1 l) were replaced with wastewater such that each solution had the same final volume. Every attempt was made to keep the wastewater-free solutions sterile (i.e. 0% WW), but since

MATERIALS AND METHODS

Synthetic urine is commonly used in urological research when experimental constraints do not permit the use of real urine. Synthetic urine was used herein since the collection of urine from human subjects is both arduous and timeconsuming, as well as being subject to ethical and safety review processes. All experiments were conducted using synthetic urine as per the method given by Griffith et al. [8]. Table 1 lists the quantities and constituents used for synthetic urine. Considering the large volume required, and the fact that these investigations were preliminary, synthetic urine was deemed appropriate. Once the processes and parameters of interest had been identified using the synthetic matrix, further research using real urine could be conducted to refine and validate the results. Part A The goal of Part A was to determine the behaviour and composition changes that occur under different storage Table 1.

Composition of synthetic urine (reproduced from Griffith et al. [8]).

Species CaCl2·H2O MgCl2·6H2O NaCl Na2SO4 Na3C6H8O7·2H2O (sodium citrate dihydrate) Na2C2O4 (sodium oxalate) KH2PO4 (potassium phosphate monobasic) KCl NH4Cl CO(NH2)2 (urea) C4H7N3O (creatinine)

Concentration (g l−1)

Ca: 4.3 Mg: 3.2 SO4: 16 citrate: 2.3

0.020 2.8

oxalate: 0.149 PO4: 20.5

NH4: 19

Total Na = 118 mEq Total K = 42 mEq PH = 5.8

808

Matrix of solutions in Part A.

Solution # (23°C)

Nominal Dilution

Open/ Closed

WW (ml)

1 2 3 4

17 18 19 20

0 0 0 0

open open open open

0 50 100 250

5 6 7 8

21 22 23 24

0 0 0 0

closed closed closed closed

0 50 100 250

9 10 11 12

25 26 27 28

1/2 1/2 1/2 1/2

open open open open

0 50 100 250

13 14 15 16

29 30 31 32

1/2 1/2 1/2 1/2

closed closed closed closed

0 50 100 250

Solution # (4°C)

Concentration (mmol l−1)

0.65 0.651 4.6 2.3 0.65

1.6 1 25 1.1

Table 2.

urease is ubiquitous, especially in an Environmental Lab., and especially for those solutions that were stored open, some contamination was inevitable. Unless stated otherwise, the following methods were used. The pH of all solutions was measured with a pH probe calibrated with standard solutions (Fisherbrand) and the probe was rinsed with distilled water between samples. For chemical analyses, approximately 5 ml of sample were withdrawn with a syringe and filtered into sample vials with a 0.45 mm Millipore filter. Samples were preserved to a pH of 2 with 5% H2SO4 and stored at 4°C until they could be analyzed. Ammonia and phosphate were measured with a Lachat QuikChem 8000 flow injection instrument. Calcium, magnesium, aluminium, iron and potassium analyses were made using a Varian Inc. SpectrAA220 Fast Sequential Atomic Absorption Spectrophotometer. Samples were prepared with the appropriate matrix modifiers as per the instrument manual; atomic absorption was used to measure all of the elements except for potassium, which was measured with atomic emission. A Bruker D8 Advance powder X-ray diffractometer, equipped with copper radiation and a graphite monochromator, was used to identify the crystal structure. Crystal constituents were measured by dissolving 100 mg of struvite crystals in 50 ml of 0.5% nitric acid and analyzing the solution using atomic absorption, as described above.

The 18 solutions were stored at room temperature with lids on. Solutions #1–9 were identical to solutions #10–18 except for the amount of magnesium added. From Part A, it was known that a precipitate formed when the solution pH passed 8. When the precipitate formed, the solution was filtered using a Whatman 934-AH (1.5 mm) glass filter. The precipitate was saved and the filtrate was resampled for all parameters. The remaining filtrate (500 ml) was dosed with a concentrated magnesium solution to achieve either a 1:1 PO4:Mg ratio (solutions #1–9) or 1:2 (solutions #10–18). Magnesium additions were based on the assumption that the PO4-P concentration in full-strength urine was 650 mg l−1 and that dilutions were perfectly linear. After magnesium was added, the mixture was shaken vigorously for one minute and allowed to react for one hour, at which point all the solutions were resampled for all parameters. The precipitate that formed was filtered out and the remaining filtrate sampled for a final time. Since the precipitates were of interest, the wastewater was filtered with a FisherBrand P8 large pore filter to remove large particulate matter in an attempt to minimize particulate matter that attached to precipitates. Each solution (#1–18) was monitored daily for changes in pH and sampled for ammonia, phosphate, magnesium and calcium. RESULTS AND DISCUSSION

Part B Part A The aim of the second experiment was to determine the effect of magnesium on urine solutions that had been hydrolyzed, and to see if struvite could be formed in urine after ureolysis had set in. Urine that had been hydrolyzed was defined as having an elevated pH (> 8) and having surrendered nutrients and metals in the form of insoluble minerals (spontaneous precipitation). Eighteen solutions were examined in this part of the study. Two concentrations of magnesium at three levels of contamination were investigated (0, 100 and 250 ml wastewater) and an extra level of dilution was investigated (full strength, half strength, quarter strength). Table 3 shows the solutions used in Part B. Table 3.

Matrix of solutions in Part B.

Solution # (2:1 Mg:PO4)

Nominal Dilution

Open/ Closed

WW (ml)

1 2 3

10 11 12

0 0 0

closed closed closed

0 100 250

4 5 6

13 14 15

1/2 1/2 1/2

closed closed closed

0 100 250

7 8 9

16 17 18

1/4 1/4 1/4

closed closed closed

0 100 250

Solution # (1:1 Mg:PO4)

Temperature Solutions that were stored at 23°C all reached a maximum pH of about 9.3, whereas the solutions stored at 4°C, did not uniformly reach a maximum pH; the values were more stratified. Some 4°C solutions reached higher peak values (i.e. 9.5) than the corresponding solution that was stored at 23°C. This discrepancy can be explained by the higher pK values for ammonia at lower temperatures, as shown in Table 4 [23]. Thus, as the temperature decreases, the pH at which equilibrium between ammonia and ammonium is attained increases. At 23°C, the pK is approximately 9.3, which corresponds to the pH observed at equilibrium, whereas the solutions stored at 4°C should have reached a maximum pH near 9.9. However, since the temperature and headspace were not

Table 4. pK 10.081 9.903 9.730 9.564 9.401 9.246

809

Dissociation constants for ammonia [23]. Temperature (°C) 0 5 10 15 20 25

held constant during sampling, the system did not reach the theoretical upper limit. Ammonia is a product of ureolysis and is an indicator of urease action, which was clearly affected by temperature. Shown in Figure 1 is the effect that temperature had on the ammonia generation in the urine mixtures over the course of 30 days. Every solution stored at room temperature (23°C) approached or exceeded 5000 mg NH3-N l−1, whereas none of the solutions stored at 4°C reached 5000 mg l−1. The stratified values seen in Figure 1B are due to the effects of dilution with distilled water and wastewater. Every line plotted below 7500 mg l−1 represents a dilute solution, and the lines above this point represent solutions that are full-strength urine. Changes in ammonia in solutions stored at A) 4 °C and B) 23 °C.

A)

-1

Ammonia (mg NH3-N l )

12500

Sample 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

10000

7500

5000

2500

0 0

5

10

15

20

25

30

Day

B) 12500

Sample 17 18 19 20 21 22 23 24 25 26 27 29 30 31 32

10000

3

-1

Ammonia (mg NH -N l )

Figure 1.

Stratification within these groupings is due to the replacement of a volume of urine (0–25%) with wastewater. The relation between pH and ammonia is further demonstrated in Figure 2. When pH is plotted against ammonia concentration for all 32 solutions for all sample points, a backwards “L” shape is formed. The plot highlights several interesting features: ammonia levels did not significantly increase (above 2000 mg l−1) until after a pH of 9 was reached; pH never exceeded 9.5 regardless of ammonia concentration, and the gap between pH 7 and 9 indicates that the rise in pH between these two points was rapid – so much so that, in most cases, no data could be captured with the sampling strategy. Thus, the potential for ammonia generation in urine that reaches a pH of 9 or higher is great.

7500

5000

2500

0 0

5

10

15

20

25

30

Day Figure 1.

Changes in ammonia in solutions stored at A) 4°C and B) 23°C.

810

-1 Ammonia (mg NH3-Nl )

12500

10000

7500

5000

2500

0 5

5.5

6

6.5

7

7.5

8

8.5

9

9.5

10

pH Figure 2.

Changes in pH vs. ammonia concentration.

Headspace In a closed system, ammonia and ammonium will stay in equilibrium, but ammonia is volatile and, in an open system, it will volatilize; this results in an overall loss of ammoniacal nitrogen from the system. To determine what effect the loss of ammonia had on the dissolved ammonia concentrations, solutions were either stored with a lid (closed) or no lid (open); however, it should be noted that lids were removed approximately once every three days, for about 10 minutes, for sampling. To calculate losses, the difference in measured ammoniacal nitrogen concentrations for each solution pair was calculated. Solutions stored at room temperature reached steady-state ammonia levels and so the average of the last three measurements was used to calculate the difference between initial and final levels. No comparison can be made for solutions stored at 4°C, because the temperature of the headspace increased during measurements thus creating a vacuum and pulling ammonia out of solution. The average percentage loss is presented in Table 5. To calculate the percent loss, the maximum value was taken as the average of the last three values. The percent difference between the maximum and initial values was calculated for both solutions in the solution pair (since each had a different maximum value), and the two percentages were averaged. These data indicate that as a percentage, dilute urine is more prone to nitrogen loss. There does not appear to be a linear correlation with the amount of wastewater in solution although, generally, there is a greater loss with an increased amount of wastewater. Figure 2.

Table 5.

Changes in pH vs. ammonia concentration.

% WW 0 5 10 25

Percentage loss of ammoniacal nitrogen. Full strength

Dilute (50%)

8* 13 11 16

9 19 14 21

* based on the average of 2, rather than 3, highest values Dilution The effect of dilution is best demonstrated by examining the pH changes of the solution; ammonia changes are not appropriate since ammonia generation is a function of initial urea concentration, which is obviously related to dilution. Compared with full-strength solutions, dilute solutions reacted, and reached a steady state pH, more quickly. Solutions kept at 4°C were generally non-reactive and are not useful for comparison. Most full-strength solutions began to increase in pH later than dilute solutions and reached a pH value of at least 9 later than the corresponding dilute solution. The difference was more pronounced in the solutions that had not been dosed with wastewater. Since the dilute solutions reacted more quickly, it appears that less contamination was needed to induce ureolysis. Since the solutions were not mixed, diffusion was the only means by which urease could come in contact with urea in order to hydrolyze it. Full-strength solutions are quite saline and the concentration of solutes would

811

The precipitate was collected but, because it was a mixture of settled wastewater, organic residue and crystals, the exact composition could not be determined. Although the minerals formed were not identified, the amount of calcium, magnesium and phosphorus that formed these minerals was determined. Figure 4 compares the percentages of calcium, magnesium and phosphorus removed due to spontaneous precipitation. Solutions that did not precipitate spontaneously, i.e. the sterile solutions that were not affected by ureolysis, are not shown. For the sake of comparison, Figure 4 shows both solution sets (i.e. solutions #1–9 and solutions #10–18); the two sets are identical. (In this and the following figures, undosed solutions that were used as controls, and unreacted solutions are not displayed). An average of 89% of magnesium, 83% of calcium and 31% of phosphorus were removed due to spontaneous precipitation. Consequently, about 70% of phosphorus remained in solution and was available to be harvested in the form of struvite. The solution that remained (after spontaneous precipitation) was ideal for struvite recovery since it contained only about 15% of the original calcium (which can interfere with optimal crystal formation). Work done with calcium and magnesium phosphates indicates that, unless the ratio of magnesium to calcium is greater than 1, struvite will not preferentially form [19]. The mixed-mineral precipitate was subsequently filtered out, and the remaining urine solution was dosed with magnesium in either a 1:1 or a 2:1 Mg:PO4 ratio (the magnesium dose took into account dilution and was adjusted accordingly). When magnesium was added to the urine solutions, a white precipitate quickly formed and settled out. Figure 5 shows the phosphorus balance, i.e. how much phosphorus was lost to spontaneous precipitation, bound in struvite or left in solution. Since some of the precipitated struvite adhered to the vessel walls, it was difficult to obtain a very accurate mass balance. The mass allocated to struvite formation was assumed to be the difference between the concentration in solution before and after magnesium addition. It can be seen from Figure 5 that the majority of phosphorus was removed from solution in the form of struvite, while the majority of calcium was removed via spontaneous precipitation. Figure 6 indicates that a small percentage (up to 8%) of calcium appeared to form into struvite; however, this may not be entirely true since subsequent data showed that very little calcium contaminated the struvite. These data simply indicate that measurements of calcium in solution, before and after struvite formation, were subject to a margin of error. The second set of nine solutions (solutions #10–18) was dosed with twice as much magnesium as the first set (solutions #1–9); however, there was no significant difference between the sets. The removal of phosphorus and calcium varied primarily with dilution. The fifth solution pair (#5 and #14) seems to have had the best removals of phosphorus to struvite; this pair was comprised of 50% urine and 10% wastewater. The crystals that formed as a result of the magnesium addition were analyzed with X-ray crystallography and were

be such that diffusion may be inhibited. In dilute solutions, diffusion would occur more readily and, thus, a small amount of urease would more easily diffuse throughout the solution. Faecal contamination Urease is ubiquitous in nature; so, although measures were taken to reduce cross contamination, it was difficult to prevent foreign urease from contaminating “sterile” solutions. Despite the fact that contamination did occur, the solutions were not contaminated for about the first two weeks of storage. The onset of urea conversion is related to the amount of wastewater, and therefore urease, present in solution; solutions with 25% WW react almost immediately, while solutions with only 5% WW are delayed several days before complete conversion. Within a given level of contamination, e.g. 5%, the data are stratified such that diluted solutions react before full-strength solutions (with the same degree of contamination) but after solutions with higher degrees of contamination. There was little difference between any of the solutions with 25% wastewater; urea hydrolysis is nearly instantaneous and no clear pattern can be discerned. Part B The results of Part A showed that phosphate minerals naturally precipitate out of solution as a result of ureolysis, herein referred to as ‘spontaneous precipitation’. Spontaneous precipitation rids the solution of calcium, producing a urine matrix that has all of the conditions necessary for high-purity struvite formation: high pH (9.3), high ammonia concentrations (> 1000 mg l−1) and low calcium content. Other work with real urine collection systems indicates that calcite, hydroxyapatite and struvite are the minerals that precipitate in urine collection systems [20]. However, since they co-precipitate, it is not possible to separate the struvite from the other minerals. The results of Part B show that, by adding magnesium to the urine solution remaining after spontaneous precipitation, pure struvite can indeed be harvested. pH The urease induced a pH increase for each of the nine solution pairs; as well, a pH drop was incurred when magnesium was added to the high pH solution is (shown in Figure 3). Except for the magnesium dosage, the matched solution pairs were subjected to the exact same conditions and behaved almost identically (recall that solutions #1–9 were dosed in a 1:1 Mg:PO4 ratio and solutions #10–18 were dosed in a 2:1 ratio). The pH drop shown in Figure 3 indicates the point at which magnesium was added and, chemically, the release of protons occurred as struvite is formed [24]. Figure 3.

Effect of magnesium addition on pH.

Phosphorus, calcium and magnesium The spontaneous precipitation of solids from solution consistently happened when the urine mixture, regardless of composition, reached a pH greater than or equal to 8.

812

Figure 4.

Percent loss of phosphorus, calcium and magnesium to spontaneous precipitation.

Figure 5.

Allocation of phosphorus.

Figure 6.

Allocation of calcium.

9.5 Sample #

9

pH

8.5

1

10

2

11

8

3

12

7.5

4

13

5

14

6

15

7

16

8

17

9

18

7 6.5 6 5.5 5 0

24

48

72

96

120

144

168

Hours Figure 3.

Effect of magnesium addition on pH.

Phosphorus

Calcium

Magnesium

100 90 80

Percent (%)

70 60 50 40 30 20 10 0 2

3

5

6

8

9

11

12

14

15

17

18

Solution # Figure 4.

Percent loss of phosphorus, calcium and magnesium to spontaneous precipitation.

813

Spontaneous precip

Struvite formation

In solution

100% 90% 80% 70%

Percent

60% 50% 40% 30% 20% 10% 0% 2

3

5

6

8

9

11

12

14

15

17

18

14

15

17

19

Solution Figure 5.

Spontaneous precip

Allocation of phosphorus.

Struvite formation

In solution

100% 90% 80%

Percent

70% 60% 50% 40% 30% 20% 10% 0% 2

3

5

6

8

9

11

Solution # Figure 6.

Allocation of calcium.

814

12

Table 6. Solution # 2 3 5 6 8 11 12 15 16 17 18

Calcium content in struvite crystals. % Calcium (by mass)

Crystal Purity (%)

0.04 0.03 0.05 0.12 0.05 N.D* N.D* N.D* 0.04 N.D* N.D*

99.96 99.97 99.95 99.88 99.95 >99.98 >99.98 >99.98 99.96 >99.98 >99.98

The spontaneous precipitation that occurs naturally removes most of the calcium from solution leaving a urine matrix that is still high in phosphate and ammonia and suitable for magnesium dosing and struvite formation. For dilutions greater than 1/4, there was essentially no difference in the quality of struvite that is recovered; all solutions had a significant amount of phosphate remaining and very low concentrations of other interfering metals. Recovered struvite was over 99.5% calcium free and different magnesium dosage ratios (1:1 or 2:1 Mg:PO4) had no noticeable effect on the quality of struvite. In addition to the above, the following specific observations were made: ●

* ND= Non-Detect value below 0.25 mg l−1 ●

verified to be struvite. The struvite crystals were then analyzed for calcium to determine the purity. As shown in Table 6, every crystal was at least 99.95% pure and there was no significant difference between the crystals from different solutions.

● ●

CONCLUSIONS

Current struvite technology utilizes anaerobic sludge digester supernatant as a feedstock to which a caustic and a magnesium source are sometimes added to increase the pH and supersaturation ratio to induce precipitation. This technique is effective although chemical additions are costly and the concentration of phosphorus in the feedstock can be low. This research investigated the possibility of using phosphorus-rich urine as a feedstock for struvite and found that pure struvite could, in fact, be recovered with minimal chemical additions. It was found that the natural changes in urine that result from ureolysis can be used to optimally prepare stored urine for struvite recovery. Urine that has undergone ureolysis has a high pH (9.3), elevated ammonia levels and low levels of calcium (that can interfere with struvite formation). By using ureoloysis to raise the pH, rather than a chemical addition, significant savings at the full scale could be realized.



Low temperatures significantly slow the rate of ureolysis and consequently the rate of ammonia generation and pH decrease. Nitrogen is lost through the volatilization of ammonia when urine is stored in the open although, compared to the total amount of ammonia present, the loss is small. Dilute urine is affected by ureolysis and reaches a maximum pH more quickly than full-strength urine. The onset of ureolysis is related to the amount of faecal matter that contaminates the solution; the more urease that contaminates the urine, the faster it will react and reach a maximum pH of approximately 9.3. Phosphate minerals will precipitate from stored urine that undergoes ureolysis; an average of 31% of phosphorus, 83% of calcium and 89% of magnesium will be lost to the spontaneous precipitation of minerals from solution.

By storing synthetic urine that is diluted up to four times, allowing it to hydrolyze, isolating it from precipitated minerals, and dosing the decanted matrix, pure struvite can be recovered from the prepared solution. ACKNOWLEDGEMENTS

The authors would like to thank the Natural Sciences and Engineering Research Council of Canada and the CIHR/ Bridge Program for their generous financial support.

REFERENCES

1. 2. 3. 4.

Steen, I., Phosphorus availability in the 21st century: Management of a non-renewable resource. Phosphorus and Potassium, 217, 25–31 (1998). Randall, C. W., Potential societal and economic impacts of wastewater nutrient removal and recycling. Water Sci. Technol., 48, 11–17 (2003). Ueno, Y. and Fujii, M., Three years experience of operating and selling recovered struvite from full-scale plant. Environ. Technol., 22, 1373–1381 (2001). Britton, A., Koch, F., Mavinic, D., Adnan, A., Oldham, W. and Udala, B., Pilot-scale struvite recovery from anaerobic digester supernatant at an enhanced biological phosphorus removal wastewater treatment plant. J. Environ. Eng. Sci., 4, 265–277 (2005).

815

5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

Berndtsson, J. C., Experiences from the implementation of a urine separation system: Goals, planning, reality. Build. Environ., 41, 427–437 (2006). Hanæus, J., Hellström, D. and Johansson, E., A study of a urine separation in an ecological village in northern Sweden. Water Sci. Technol., 35, 153–160 (1997). Griffith, D. P., Bragin, S. and Musher, D. M., Dissolution of struvite urinary stones - experimental studies in vitro. Invest. Urol., 13, 351–353 (1976). Griffith, D. P., Musher, D. M. and Itin, C., Urease – primary cause of infection-induced urinary stones. Invest. Urol., 13, 346–350 (1976). Griffith, D. P., Urease stones. Urol. Res., 7, 215–221 (1979). Hanæus, A., Hellström, D. and Johansson, E., Conversion of urea during storage of human urine. Vatten, 52, 263–270 (1996). Larsen, T. A. and Gujer, W., Separate management of anthropogenic nutrient solutions (human urine). Water Sci. Technol., 34, 87–94 (1996). Lienert, J., Haller, M., Berner, A., Stauffacher, M. and Larsen, T. A., How farmers in Switzerland perceive fertilizers from recycled anthropogenic nutrients (urine). Water Sci. Technol., 48, 47–56 (2003). Pahl-Wostl, C., Schonborn, A., Willi, N., Muncke, J. and Larsen, T. A., Investigating consumer attitudes towards the new technology of urine separation. Water Sci. Technol., 48, 57–65 (2003). Hellström, D., Johannson, E. and Grennberg, K., Storage of human urine: Acidification as a method to inhibit decomposition of urea. Ecol. Eng., 12, 253–269 (1999). Hellström, D. and Kärrman, E., Nitrogen and phosphorus in fresh and stored urine. Environ. Res. Forum, 5, 221–226 (1996). Ban, Z. and Dave, G., Laboratory studies on recovery of N and P from human urine through struvite crystallisation and zeolite adsorption. Environ. Technol., 25, 111–121 (2004). Lind, B. B., Ban, Z. and Byden, S., Nutrient recovery from human urine by struvite crystallization with ammonia adsorption on zeolite and wollastonite. Bioresour. Technol., 73, 169–174 (2000). Huang, H., Mavinic, D., Lo, K. and Koch, F., Production and basic morphology of struvite crystals from a pilot-scale crystallization process. Environ. Technol., 27, 233–245 (2006). Abbona, F., Madsen, H. E. L. and Boistelle, R., The initial phases of calcium and magnesium phosphates precipitated from solutions of high to medium concentrations. J. Cryst. Growth, 74, 581–590 (1986). Moyo, C., Kissel, D. and Cabrera, M., Temperature effects on soil urease activity. Soil Biol. Biochem., 21, 935–938 (1989). Udert, K. M., Larsen, T. A. and Gujer, W., Biologically induced precipitation in urine-collecting systems. Water Sci. Technol.: Water Supply, 3, 71–78 (2003). Höglund, C., Stenstrom, T. A., Jonsson, H. and Sundin, A., Evaluation of faecal contamination and microbial die-off in urine separating sewage systems. Water Sci. Technol., 38, 17–25 (1998). Perrin, D. D., Ionisation Constants of Inorganic Acids and Bases in Aqueous Solution. Pergamon, Oxford, UK (1982). Babic-Ivancic, V., Kontrec, J., Kralj, D. and Brecevic, L., Precipitation diagrams of struvite and dissolution kinetics of different struvite morphologies. Croat. Chem. Acta., 75, 89–106 (2002).

816