UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Ammonia recovery from municipal wastewater through a struvite formation-thermal decomposition cycle Wilson, Connor Walter 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2013_fall_wilson_connor.pdf [ 2.41MB ]
Metadata
JSON: 24-1.0103342.json
JSON-LD: 24-1.0103342-ld.json
RDF/XML (Pretty): 24-1.0103342-rdf.xml
RDF/JSON: 24-1.0103342-rdf.json
Turtle: 24-1.0103342-turtle.txt
N-Triples: 24-1.0103342-rdf-ntriples.txt
Original Record: 24-1.0103342-source.json
Full Text
24-1.0103342-fulltext.txt
Citation
24-1.0103342.ris

Full Text

AMMONIA RECOVERY FROM MUNICIPAL WASTEWATER THROUGHA STRUVITE FORMATION-THERMAL DECOMPOSITION CYCLEbyCONNOR WALTER WILSONB.A.Sc., The University of British Columbia, 2010A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THEREQUIREMENTS FOR THE DEGREE OFMASTER OF APPLIED SCIENCEinThe FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES(Civil Engineering)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)September 2013? Connor Walter Wilson, 2013ABSTRACTiiABSTRACTBench-scale batch experiments were performed to assess the potential for ammonia removaland recovery from municipal post-digestion waste streams via struvite (MgNH4PO4?6H2O)crystallization using thermally decomposed struvite as a source of magnesium andorthophosphate. To simulate this process, newberyite (MgHPO4?3H2O), a synthesized surrogatefor thermally decomposed struvite, was added to various ammonia solutions including syntheticstruvite crystallizer effluent and synthetic dewatering centrate. The main objective of this studywas to develop the concept of the proposed technology through evaluation of the effects ofchemical and physical factors on the rates and mechanisms of ammonia removal,orthophosphate solubilization, and overall newberyite-to-struvite conversion efficiency.A model was developed using PHREEQC software to simulate each batch experiment and topredict the solid and liquid phase compositions that would result from these systems attainingchemical equilibrium. Experimental and model-predicted results were employed for thedelineation of near optimal conditions for efficient transformation of newberyite into struvite.Ammonia removal efficiencies as high as 87% were achieved while maintaining orthophosphateresiduals as low as 10 mg/L PO4-P. Measurements of liquid phase compositions at reactiontimes approaching equilibrium compared well with that predicted by the model. Resultssuggested an optimum within a region of 1 to 3 hour reaction times, pH between 7 and 8,temperature between 10? and 25? C, and at a newberyite dose that provides a suspensionMg:N:P molar ratio of 1:1:1.Although the results of the present study illustrate the potential of this technology, it isrecommended that further research be performed employing the newberyite-containing materialproduced by the pilot-scale struvite thermal decomposition reactor located at the University ofBritish Columbia.PREFACEiiiPREFACEThis thesis is original, unpublished, independent work by the author, C. Wilson.TABLE OF CONTENTSivTABLE OF CONTENTSABSTRACT................................................................................................................................. iiPREFACE.................................................................................................................................. iiiTABLE OF CONTENTS............................................................................................................. ivLIST OF TABLES..................................................................................................................... viiiLIST OF FIGURES .................................................................................................................... ixLIST OF ABBREVIATIONS AND SYMBOLS ............................................................................. xiACKNOWLEDGEMENTS ......................................................................................................... xiiCHAPTER 1: INTRODUCTION.................................................................................................. 1CHAPTER 2: BACKGROUND ................................................................................................... 52.1 Motivation for ammonia removal and recovery........................................................................... 52.1.1 Conventional ammonia synthesis .........................................................................................62.1.2 Demand for nitrogenous fertilizers.......................................................................................72.1.3 Ammonia as a fuel.................................................................................................................82.1.4 Economic benefits of recovery of ammonia at municipal WWTPs.....................................102.2 Existing ammonia recovery options for post-digestion side streams.........................................12CHAPTER 3: RESEARCH OBJECTIVES .................................................................................15CHAPTER 4: LITERATURE REVIEW.......................................................................................164.1 Chemistry of magnesium and phosphate compounds ...............................................................164.1.1 Aqueous equilibria affecting speciation of magnesium, ammonium, orthophosphate, andcarbonate ............................................................................................................................................174.1.2 Solubility products ..............................................................................................................184.2 Struvite and newberyite morphology.........................................................................................184.3 Factors affecting struvite formation ...........................................................................................204.3.1 Supersaturation ..................................................................................................................204.3.2 Mg:P and N:P molar ratio....................................................................................................214.3.3 Hydrodynamics ...................................................................................................................224.4 UBC struvite crystallization process............................................................................................23TABLE OF CONTENTSv4.4.1 Pilot-scale UBC struvite crystallizer design .........................................................................234.4.2 Pilot-scale UBC struvite crystallizer operation....................................................................264.5 Struvite decomposition products as ammonia removal agents .................................................264.5.1 Struvite decomposition in solution ? ?Wet process?..........................................................284.5.2 Ammonia removal following ?wet process? .......................................................................314.5.3 Thermal decomposition of struvite in air ? ?Dry process? .................................................334.5.4 Ammonia removal following ?dry process? ........................................................................354.6 Potential ammonia gas capture techniques ...............................................................................374.7 Conclusions for development of present study..........................................................................38CHAPTER 5: MATERIALS AND METHODS.............................................................................405.1 Description of performed batch tests.........................................................................................405.2 Ammonia solutions ? Synthetic wastewaters.............................................................................415.3 Synthetic newberyite ..................................................................................................................425.3.1 Synthesis .............................................................................................................................425.3.2 Analysis ...............................................................................................................................425.4 Materials and equipment ...........................................................................................................435.4.1 Batch test method and apparatus ......................................................................................435.4.2 pH monitoring .....................................................................................................................445.4.3 Conductivity monitoring .....................................................................................................445.4.4 Ammonium monitoring.......................................................................................................445.4.5 PHREEQC-2 chemical equilibrium model............................................................................455.5 Sample collection and preservation ...........................................................................................455.6 Analytical methods .....................................................................................................................465.6.1 Magnesium .........................................................................................................................465.6.2 Ammonia and orthophosphate...........................................................................................475.6.3 Carbonate alkalinity ............................................................................................................475.6.4 XRD identification of solid phases.......................................................................................475.6.5 Crystal morphology.............................................................................................................475.7 Statistics ......................................................................................................................................485.8 Terminology ................................................................................................................................485.8.1 Molar ratio ..........................................................................................................................48TABLE OF CONTENTSvi5.8.2 Solubility product (Ksp-N and Ksp-S)........................................................................................495.8.3 Supersaturation ratio ..........................................................................................................495.8.4 Removal efficiency ..............................................................................................................50CHAPTER 6: RESULTS AND DISCUSSION ............................................................................516.1 Fundamentals of newberyite dissolution-struvite formation mechanism in the presence ofammonium..............................................................................................................................................516.2 Transformation of newberyite into struvite in ammonia solution:   Phase 1 ? SuspensionMg:N:P molar ratio of 1:1.1:1 .................................................................................................................516.2.1 pH effect on rate and efficiency of ammonia removal .......................................................526.2.2 pH effect on rate and extent of orthophosphate solubilization .........................................556.2.3 Temperature effect on rate and efficiency of ammonia removal ......................................586.2.4 Temperature effect on rate and extent of orthophosphate solubilization ........................606.2.5 XRD analysis of solid phase mixtures ..................................................................................626.2.6 Chemical composition of solid phase mixtures ..................................................................636.2.7 Comparison of experimental results and model predicions...............................................656.2.8 Initial rates of newberyite dissolution and struvite formation...........................................686.3 Transformation of newberyite into struvite in ammonia solution:   Phase 2 ? SuspensionMg:N:P molar ratio of 1:1.4:1 .................................................................................................................696.3.1 Newberyite dose effect on rate and efficiency of ammonia removal ................................706.3.2 Newberyite dose effect on rate and extent of orthophosphate solubilization ..................736.4 Transformation of newberyite into struvite in synthetic crystallizer effluent ...........................766.4.1 Ammonia removal...............................................................................................................786.4.2 Orthophosphate residual ....................................................................................................796.4.3 Chemical composition of solid phase mixtures ..................................................................816.4.4 Newberyite and struvite supersaturation...........................................................................826.4.5 Caustic consumption...........................................................................................................856.4.6 Crystal morphology.............................................................................................................866.5 Transformation of newberyite into struvite in synthetic centrate .............................................896.5.1 Ammonia removal...............................................................................................................916.5.2 Online ammonium residual monitoring..............................................................................926.5.3 Orthophosphate residual ....................................................................................................946.5.4 Chemical composition of solid phase mixtures ..................................................................95TABLE OF CONTENTSvii6.5.5 Newberyite and struvite supersaturation...........................................................................966.5.6 Caustic consumption...........................................................................................................986.5.7 Crystal morphology.............................................................................................................98CHAPTER 7: CONCLUSIONS................................................................................................102CHAPTER 8: RECOMMENDATIONS .....................................................................................104REFERENCES .......................................................................................................................105APPENDIX A: INSTRUMENT OPERATIONAL SETTINGS.....................................................115APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS.............................................116APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES ....................................130APPENDIX D: MODELLING RESULTS ..................................................................................151LIST OF TABLESviiiLIST OF TABLESTable 1 ? Modelled savings for two WWTPs implementing ammonia recovery from dewateringcentrate (ThermoEnergy Corporation, 2007).............................................................................11Table 2 ? 2013 prices for various ammonium-derived fertilizers (USDA ERS, 2013).................11Table 3 ? Experimentally determined pKsp at 25? C for various magnesium and phosphatecompounds ...............................................................................................................................18Table 4 ? Range of operation and performance for UBC struvite crystallizer pilot studies .........26Table 5 ? Summary of struvite decomposition studies using thermal-alkali treatment ...............31Table 6 ? Summary of ammonia removal studies using struvite decomposed under thermal-alkali conditions.........................................................................................................................32Table 7 ? Summary of experimental parameters for duplicate batch tests.................................40Table 8 ? Annacis Island WWTP centrate compared to studied solutions .................................42Table 9 ? Synthetic newberyite chemical composition and concerned experiments ..................43Table 10 ? Suspension characteristics at t = 0 h for Mg:N:P molar ratio 1:1.1:1 newberyite dosebatch tests ................................................................................................................................52Table 11 ? Suspension characteristics at t = 0 h for Mg:N:P molar ratio 1:1.4:1 newberyite dosebatch tests ................................................................................................................................70Table 12 ? Suspension characteristics at t = 0 h for Mg:N:P molar ratio 1:1:1 newberyite dose insynthetic crystallizer effluent batch tests ...................................................................................78Table 13 ? Suspension characteristics at t = 0 h for Mg:N:P molar ratio 1:1.05:1 newberyitedose in synthetic centrate batch tests .......................................................................................91LIST OF FIGURESixLIST OF FIGURESFigure 1 ? Historic trends in world population, fertilizer consumption, and meat production(Erisman et al., 2008)................................................................................................................. 8Figure 2 ? Star-shaped dendritic (a), X-shaped twinned (b), coffin-shaped (c), and rod-like (d)struvite crystals (Abbona et al., 1985) .......................................................................................19Figure 3 ? Pseudo-octagonal newberyite crystals amongst dissolving tabular struvite crystal (a);rhombohedral newberyite crystals (b and c) (Boistelle et al., 1983; Kontrec et al., 2005; Babi?-Ivan?i? et al., 2002)...................................................................................................................20Figure 4 ? General schematic of UBC struvite crystallization process (top) and crystallizerinjector (bottom) (Fattah, 2004).................................................................................................25Figure 5 ? Conceptual schematic of NH4 removal through reuse of Mg and PO4 ......................28Figure 6 ? Schematic diagram of reusing MAP residues for ammonia removal by acid dipping asproposed by Zhang et al. (2004) ...............................................................................................29Figure 7 ? NH4-N removed for various pH conditions at (a) 10?, (b) 25?, and (c) 35? C .............54Figure 8 ? Residual PO4-P for various pH conditions at (a) 10?, (b) 25?, and (c) 35? C .............57Figure 9 ? NH4-N removed for various temperatures for (a) no pH control, (b) pH 7, (c) pH 8,and (d) pH 9..............................................................................................................................59Figure 10 ? Residual PO4-P for various temperatures for (a) no pH control, (b) pH 7, (c) pH 8,and (d) pH 9..............................................................................................................................61Figure 11 ? N:P molar ratio of solid phase mixtures sampled at 1 h (No pH = no pH control) ....64Figure 12 ? N:P molar ratio of solid phase mixtures sampled at 3 and 12 h (No pH = no pHcontrol)......................................................................................................................................64Figure 13 ? Comparison of real and model-predicted NH4-N for 25? C......................................66Figure 14 ? Comparison of real and model-predicted PO4-P for 25? C......................................66Figure 15 ? Comparison of real and model-predicted N:P molar ratios of solid phase mixtures 67Figure 16 ? pH vs time for no pH control at 10? C .....................................................................69Figure 17 ? NH4-N removed for various pH conditions at (a) 10?, (b) 25?, and (c) 35? C ...........72Figure 18 ? NH4-N removed at 25? C for Mg:N:P molar ratio of (a) 1:1.1:1 and (b) 1:1.4:1........73Figure 19 ? Residual PO4-P for various pH conditions at (a) 10?, (b) 25?, and (c) 35? C ...........75Figure 20 ? PO4-P residual at 25? C for Mg:N:P molar ratio of (a) 1:1.1:1 and (b) 1:1.4:1 .........76Figure 21 ? Reactor Configuration 1: Ammonia recovery from primary crystallizer effluent .......77Figure 22 ? Comparison of real and model-predicted NH4-N.....................................................79LIST OF FIGURESxFigure 23 ? Comparison of real and model-predicted PO4-P.....................................................81Figure 24 ? N:P molar ratio of solid phase sampled at 1 and 4 h...............................................82Figure 25 ? Newberyite and struvite supersaturation ratio at pH 7 ............................................83Figure 26 ? Newberyite and struvite supersaturation ratio at pH 8 ............................................84Figure 27 ? Theoretical struvite supersaturation ratio immediately after mixing of liquid reagents.................................................................................................................................................85Figure 28 ? Comparison of real and model-predicted caustic consumption...............................86Figure 29 ? x40 magnified newberyite (a); 1 hour samples from pH 7-10? C (b), pH 7-25? C (c),and pH 8-25? C (d)....................................................................................................................87Figure 30 ? x10 magnified 4 hour samples from pH 7-10? C (a), pH 7-25? C (b), pH 8-10? C (c),and pH 8-25? C (d)....................................................................................................................88Figure 31 ? x40 magnified 4 hour samples from pH 7-10? C (a), pH 7-25? C (b), pH 8-10? C (c),and pH 8-25? C (d)....................................................................................................................89Figure 32 ? Reactor Configuration 2: Ammonia recovery from dewatering centrate ..................90Figure 33 ? Comparison of real and model-predicted NH4-N.....................................................92Figure 34 ? Comparison of probe and sample measured NH4-N...............................................93Figure 35 ? Comparison of real and model-predicted PO4-P.....................................................95Figure 36 ? N:P molar ratio of solid phase sampled at 1 and 4 h...............................................96Figure 37 ? Newberyite and struvite supersaturation ratio at pH 7 ............................................97Figure 38 ? Newberyite and struvite supersaturation ratio at pH 8 ............................................97Figure 39 ? Comparison of real and model-predicted caustic consumption...............................98Figure 40 ? x40 magnified synthetic newberyite batches (a and b); 1 hour samples from pH 7-10? C (c), and pH 7-25? C (d)....................................................................................................99Figure 41 ? x10 magnified 4 hour samples from pH 7-10? C (a), pH 7-25? C (b), pH 8-10? C (c),and pH 8-25? C (d)..................................................................................................................100Figure 42 ? x40 magnified 4 hour samples from pH 7-10? C (a), pH 7-25? C (b), pH 8-10? C (c),and pH 8-25? C (d)..................................................................................................................101LIST OF ABBREVIATIONS AND SYMBOLSxiLIST OF ABBREVIATIONS AND SYMBOLSARP Ammonia Recovery ProcessCRT crystal retention timeEUR EuroFBR fluidized bed reactorGBP Great British PoundHRT hydraulic retention timeKsp-N newberyite solubility productKsp-S struvite solubility productMLD million litres per dayn number of samplesPS-eq conditional solubility product of struvite at equilibriumPS-reactor in-reactor conditional solubility of struviteQTG quasi-isothermal thermogravimetrys standard deviationSN supersaturation ratio with respect to newberyiteSS supersaturation ratio with respect to Struvitet Student?s t-valueUBC University of British ColumbiaUSD United States DollarUSDA ERS United States Department of Agriculture Economic Research ServiceUSEPA United States Environmental Protection AgencyUSGS United States Geological SurveyWERF Water Environment Research FoundationWWTP wastewater treatment plantXRD X-ray diffractionACKNOWLEDGEMENTSxiiACKNOWLEDGEMENTS? My partner, Aubrie, for encouraging me to pursue an impactful and fulfilling career andfor her support throughout my degree? My supervisor, Dr. Donald Mavinic, for his insight, enthusiasm, and for bringing thenutrient recovery group to life? My mentors, Dr. Sergey Lobanov and Frederic Koch, for their unparalleled knowledgeand experience? Paula Parkinson and Timothy Ma for kindly providing their analytical and experimentalexpertise? Natural Sciences and Engineering Research Council of Canada for their generousfunding? My loving family? My friends for cheering me onCHAPTER 1: INTRODUCTION1CHAPTER 1: INTRODUCTIONNitrogen is essential to life and is among the nutrients consumed in the largest quantities byorganisms. Wastewater nitrogen is derived primarily from organic matter originating fromhuman, animal and food processing wastes. Organic nitrogen is decomposed by bacteria torelease ammonia. Aqueous ammonia takes the form of both ammonium (NH4+) and un-ionizedammonia gas (NH3). In typical environmental conditions the majority of ammonia exists asammonium. However, un-ionized ammonia is known to be more toxic to organisms living inreceiving bodies (Environment Canada, 2001; Randall and Tsui, 2002). Ammonia discharge isreduced through wastewater treatment to prevent these toxic effects as well as to avoideutrophication of downstream aquatic or marine environments. To achieve lower nitrogendischarge goals, wastewater treatment plants (WWTPs) employ biological nitrogen removalprocesses, such as nitrification, which is commonly followed by denitrification to ultimatelyconvert ammonia to atmospheric nitrogen.Aerobic or anaerobic digestion of the biosolids produced during wastewater treatment results inthe release of a significant fraction of the nutrients that were previously removed.  Ammonia iscontained in liquid biosolids streams at levels which generally far exceed that of the raw influentwastewater. Therefore, digester supernatant, as well as the filtrate or centrate generated duringdigested sludge dewatering, are returned upstream for further treatment. These streams carryhigh ammonium and orthophosphate concentrations and, in the presence of magnesium, theymay be supersaturated with respect to magnesium ammonium phosphate (MgNH4PO4?6H2O),also known as struvite. Struvite is a sparingly soluble salt that commonly forms in systemswhich convey post-digestion streams. Deposits may appear as scale on the walls of pipes orwithin equipment. This has the potential to damage pumps and dewatering equipment, as wellas significantly reduce the diameter of pipes resulting in a loss of hydraulic capacity and loweroperational efficiency. Therefore, chemical removal of orthophophate or routine system cleaningwith acid is often required, consequently increasing process complexity and maintenance costs.To avoid the operational issues surrounding unwanted formation, controlled crystallization ofstruvite may be employed to reduce its supersaturation in wastewater. One technology that hassuccessfully been used for this purpose is the fluidized bed reactor (FBR) designed at Universityof British Columbia, also known as the UBC struvite crystallizer (Dastur, 2001; Adnan, 2002;CHAPTER 1: INTRODUCTION2Britton, 2002; Huang, 2003; Fattah, 2004). The UBC struvite crystallization process provides thechemical and hydrodynamic conditions favourable for the incorporation of struvite crystals intoagglomerates and eventually pellets. After sufficient reaction time, struvite pellets are harvestedand may be sold as a high-purity, slow-release fertilizer containing equimolar parts magnesium,ammonium and orthophosphate. This process was made commercially available by OstaraNutrient Recovery Technologies Inc., who has commissioned six municipal struvite recoveryfacilities to date (Ostara Nutrient Recovery Technologies Inc., 2007; Britton et al., 2009; OstaraNutrient Recovery Technologies Inc., 2011).Municipal wastewater is generally much higher in nitrogen relative to phosphorus. Therefore,with the addition of sufficient magnesium, the amount of struvite that can be formed from thesestreams is limited by its orthophosphate concentration. Pilot and full-scale studies have proventhat the UBC struvite crystallization process is capable of consistently providing orthophosphateremovals from post-digestion streams between 80% and 99%. However, this process onlyremoves, on average, between 5% and 10% of the nitrogen; this means that the crystallizereffluent is still rich in ammonia (Britton, 2002; Huang, 2003; Fattah, 2004; Ostara NutrientRecovery Technologies Inc., 2007, 2011; Britton et al., 2009). Following struvite recovery, thecrystallizer effluent is returned to upstream processes where ammonia is restabilized. If externalmagnesium and orthophosphate was added to this stream, a significant portion of the residualammonia could theoretically be recovered as struvite, while considerably increasing plantammonia removal capacity and reducing aeration costs.Recent research has focussed on the recovery of phosphorus from wastewater to offset theworld?s reliance on localized and limited phosphate rock reserves for global food security (LeCorre et al., 2009; WERF, 2010; Liu et al., 2012). Since ammonia is synthesized commerciallyfrom atmospheric nitrogen, the main constituent in air, through the Haber-Bosch process, it isimpractical to convert high quality sources of orthophosphate to struvite in order to recover it.Nevertheless, the ammonia production industry relies heavily on natural gas as a non-renewable precursor for hydrogen (Smil, 2001; Erisman et al., 2008). It is a frightening conceptthat the availability of an ingredient so widely used in a fertilizer that is crucial in feeding theworld is dependent on a fossil fuel that has, in the past, experienced market volatility (Mohr andEvans, 2011; Maggio and Cacciola, 2012). Further, the processes of synthesizing ammonia andremoving it from wastewater are both energy and resource intensive. If an inexpensive sourceCHAPTER 1: INTRODUCTION3of external magnesium and orthophosphate were available, the potential is there to recoverammonia from nutrient-rich sidestreams, rather than biologically convert it back to atmosphericnitrogen or lock it up in waste solids.A possible source of magnesium and orthophosphate could be derived from recovered struviteitself. However, this requires the removal of the ammonium that is bound within struvite. Severalstudies have shown that struvite can be used as a precursor to produce other magnesiumphosphate materials under various experimental conditions (Abdelrazig and Sharp, 1988;Sarkar, 1991; Sugiyama et al., 2005; Bhuiyan et al., 2008; Kurtulus and Tas, 2011, Novotny,2011). Newberyite (MgHPO4?3H2O) is an ideal material as it contains no ammonium, butprevious studies suggest that significant chemical addition would be required to produce it froma struvite suspension (Boistelle et al., 1983; Zhang et al., 2004; Babi?-Ivan?i? et al., 2006).Using a proprietary technology recently developed at UBC, pelletized struvite can be convertedto a nearly pure source of crystalline newberyite through relatively low temperature thermaldecomposition in air.Newberyite is expected to fully dissolve in ammonia-rich wastewater and recrystallize as struvitewith control of temperature and/or pH. This presents an opportunity to efficiently recyclemagnesium and orthophosphate while removing and recovering ammonia. In essence, asignificant portion of the struvite formed through these reactions could be thermallydecomposed to produce more newberyite. Further, there is the potential to recover the ammoniagas evolved during this process which represents commercial value as a high-purity source ofreactive nitrogen. Eqns. 1, 2, and 3 illustrate the theory of magnesium and orthophosphaterecycling through the systematic reactions of newberyite dissolution, struvite recrystallization,and struvite thermal decomposition.MgHPO4?3H2O(s) + H+? Mg2+ + H2PO4- + 3H2O (1)Mg2+ + NH4+ + HPO42- + 6H2O? MgNH4PO4?6H2O(s) + H+ (2)MgNH4PO4?6H2O(s) + Heat? MgHPO4?3H2O(s) + NH3(g) + 3H2O (3)CHAPTER 1: INTRODUCTION4For the described technology to be justified economically, it is essential that the outlinedprocess offers high ammonia recovery efficiency and a short reaction time. Further, it ishypothesized that newberyite will dissolve in wastewater and that nearly 100% of theorthophosphate released can be utilized in the formation of struvite. To explore thesehypotheses, this research employed a systematic approach. For a variety of experimentalconditions, the reaction rates were observed for the conversion of newberyite to struvite in thepresence of ammonium. This work employed batch tests to examine several combinations ofsynthetic newberyite with an ammonia solution, synthetic crystallizer effluent, and syntheticdewatering centrate. The extent of reactions with respect to time was evaluated based on liquidand solid phases compositions, as well as solid morphologies. This data was compared to theoutputs of a chemical equilibrium model to verify its usefulness in future development of theproposed ammonia removal and recovery technology.CHAPTER 2: BACKGROUND5CHAPTER 2: BACKGROUND2.1 Motivation for ammonia removal and recoveryNitrogen is one of the primary nutrients essential to life and ammonia is taken up by terrestrialorganisms either directly or indirectly to satisfy nutritional needs. As a part of the nitrogen cycle,ammonia can be produced from atmospheric nitrogen (N2) by select organisms or it can beconverted to atmospheric nitrogen by combined nitrification and denitrification. Ammonia issynthesized at a massive scale by the fertilizer industry. Approximately 100 Mt of reactivenitrogen is synthesized annually worldwide using the Haber-Bosch process. To put thisanthropogenic effect in perspective, about 150 to 200 Mt nitrogen is fixed naturally per year onearth, mostly by symbiotic diazatrophs (Smil, 2001). Not all of the ammonia applied to land isutilized by crops and losses can be attributed to runoff and leaching. It has been estimated thatonly 17% of the 100 Mt of reactive nitrogen synthesized for global agriculture in 2005 wereactually consumed as food by humans (Erisman et al., 2008; Aiking 2011).Humans excrete a significant fraction of the nutrients contained in the food they ingest.Alongside agricultural sources, these nutrients find their way back into the environment asmunicipal wastewater effluents and organic matter deposited to landfills. Anthropogenic loadingof nutrients is the main cause for eutrophication of receiving water bodies. Eutrophication oflakes and coastal estuaries results in enhanced productivity and the formation of algal bloomsthat are detrimental to local ecosystems. This is extremely difficult to remedy as nutrients arecycled between organic matter deposited in sediment and the water column. In severe cases,this boost in organic carbon content can result in increased heterotrophic activity in thesediments contributing to oxygen-deficient zones. These are also referred to as ?dead zones?,as their formation will decimate local aerobic organism populations. Pollution from thesesources also increases un-ionized ammonia levels in aquatic and marine environments, which isknown to be toxic to many organisms (Environment Canada, 2001; Randall and Tsui, 2002). Forthese reasons, nitrogen discharge is heavily regulated in many countries and biological nutrientremoval processes were developed for municipal wastewater treatment. These processesessentially lock up the nitrogen originating from chemical fertilizers in solids for disposal and/ordestroy it through conversion to atmospheric nitrogen.CHAPTER 2: BACKGROUND6The current paradigm of synthesis and subsequent loss, disposal, and destruction of ammoniais wasteful. Ammonia fertilizer manufacturing and municipal wastewater treatment are bothresource and energy intensive tasks. Municipal wastewater should be viewed as an ammoniaresource and nutrient-rich waste streams should be exploited through the recovery of ammoniain forms that could be employed by agriculture or other industries.2.1.1 Conventional ammonia synthesisThe atmosphere contains 78% nitrogen and, unlike phosphorus, it is not considered a limitedresource on earth. Using the Haber-Bosch process, ammonia can be synthesized fromatmospheric nitrogen gas as needed. Eqn. 5 provides a general explanation of this process.N2(g) + 3H2(g)? 2NH3(g) ?H = -92 kJ/mol (5)At a temperature and pressure of approximately 400 to 450? C and 10 to 30 MPa respectively,hydrogen and nitrogen gas are combined to form ammonia (Smil, 2001). This reaction isgenerally performed in the presence of an iron catalyst. The Haber-Bosch process of today isnot particularly efficient and allows for a nitrogen-to-ammonia molar conversion of about 15%per pass and, therefore, unreacted gases are recycled further to achieve overall conversions ofabout 98% (Smil, 2001). Approximately 85% of the ammonia worldwide is produced fromhydrogen gas generated through steam reforming of light hydrocarbons of which 80% ofprocesses utilize natural gas (Smil, 2001; Wood and Cowie, 2004). Ammonia synthesis isresponsible for about 5% of the world?s natural gas consumption (Maxwell, 2005). Further,providing the conditions of high temperature and pressure are energetically expensive. It isestimated that global ammonia production accounts for 1.3% of the world?s fossil fuel-derivedenergy use, contributing considerable greenhouse gas emissions (Smil, 2001; Erisman et al.,2008). However, there is debate concerning the role that ammonia production plays in globalwarming, since enrichment of aquatic and marine environments by synthetic ammonia is acause for eutrophication and, therefore, enhanced carbon dioxide sequestration. Althoughreactive nitrogen is a renewable resource, the cost and availability of ammonia-based fertilizersdepend heavily on that of fossil fuels and energy.CHAPTER 2: BACKGROUND72.1.2 Demand for nitrogenous fertilizersThe cultivation of food crops results in the transfer of nutrients from the soil to plant matter.When these crops are harvested, generally only a portion of the nutrients from these plantresidues are reintroduced in the soil. The nitrogen cycle is broken with respect to conventionalagriculture, since the majority of the nutritional elements that went into growing crops are lost tothe environment, disposed of in landfills, or emitted to the atmosphere. Synthetic fertilizers arebelieved to provide 60% to 80% of the nitrogen requirements for cultivation of high staple crops(Smil, 2001). Without these inputs, the world population would be significantly lower than it istoday. As a result of the Haber-Bosch process for ammonia synthesis, the number of peopleone hectare of arable land can support has increased from approximately 1.9 to 4.3 between1908 and 2008. Today, it is estimated that 80% of the ammonia produced is used formanufacturing nitrogenous fertilizers (Erisman et al., 2008). Further, assuming the globaladoption of extremely basic, vegetarian diets, preindustrial agriculture could provide for onlyabout 40% of the today?s world population (Smil, 2001).Synthetic fertilizers have allowed for massive growth of the livestock industries resulting in aworldwide shift towards more meat-based and dairy-based diets. These changes to human dietsfurther increase demands as meat and dairy production indirectly require more fertilizer thancereal and vegetable crops. During the mid-1990?s, about one third of the nitrogen used to growcrops was fed to domestic animals (Smil, 2001). This is of concern because livestock proteinconversion efficiencies for chicken, pork, and beef are approximately 20%, 10%, and 5%respectively, and little more than half of the nitrogen contained in animal manures is usedglobally for cultivation of crops (Smil, 2001; Aiking 2011). Figure 2 illustrates historic trends withrespect to world population, fertilizer application, and meat production. It is predicted that theworld population will increase by about 2.3 billion in the next 40 years and feeding the world willnot be possible without global shifts towards more efficient transfer of synthesized ammonia tofood, preservation of land fertility, lower overall meat and dairy consumption, and moresustainable production of fertilizers (Aiking, 2011).CHAPTER 2: BACKGROUND8Figure 1 ? Historic trends in world population, fertilizer consumption, and meatproduction (Erisman et al., 2008)2.1.3 Ammonia as a fuelIt is inevitable that humans will be forced to turn away from non-renewable fossil fuels to moresustainable energy sources with fewer environmental impacts. Although hydrogen is commonlymentioned as a synthesizable alternative, it has a far lower octane level when compared toother fuels burned by vehicles and its adoption suggests significant challenges with regard tosafe storage and distribution (Yin et al., 2004; Christensen et al., 2006; Lan et al., 2012).Ammonia can provide more energy per unit volume than hydrogen (Zamfirescu and Dincer,2009). Hydrogen is also three times more expensive than ammonia, with respect to the volumeof stored energy (Zamfirescu and Dincer, 2008). Further, ammonia is safer than hydrogen dueto its narrow flammability limits and it is generally considered non-explosive. It also possesses acharacteristic odour which alarms those nearby of its presence. Since ammonia is alreadywidely used in industry, distribution methods are well established and it is stored for combustionin a similar fashion to propane, making it attractive for vehicular operations (Zamfirescu andCHAPTER 2: BACKGROUND9Dincer, 2008; Zamfirescu and Dincer, 2009). Some advanced ammonia internal combustionengines are designed to use compression ratios several times higher than that of theconventional. Alternatively, conventional internal combustion engines can run on a mixture of80% ammonia and 20% gasoline with minor modifications (Zamfirescu and Dincer, 2009). Up to60% of the energy used in a turbocharged diesel engine can be supplied by ammonia with aconversion efficiency of close to 100% (Reiter and Kong, 2011). As compared to gasoline, liquidpetroleum gas, compressed natural gas, and methanol, ammonia has the lowest cost perenergetic unit when used in the 100 km driving range (Zamfirescu and Dincer, 2008; Zamfirescuand Dincer, 2009). Ammonia is also an excellent refrigerant. It is estimated that engineperformance may be improved by 10% by using onboard ammonia for engine cooling and airconditioning (Zamfirescu and Dincer, 2009).A major advantage ammonia has over fossil fuels is that its combustion produces no carbondioxide or sulfur oxides. Nevertheless, the burning of ammonia in internal combustion enginescould potentially result in the release of some nitrogen oxides. However, this emission may beminimized through the optimization of air-to-fuel ratios (Zamfirescu and Dincer, 2009). Althoughcombustion of ammonia proposes a concern based on its toxicity, ammonia pollution may bemitigated using well established reversible adsorption techniques (Elm?e et al., 2006). Currenttrends in research suggest a future advancement of hydrogen fuel cell technology but, again,one of the major obstacles for the shift towards hydrogen-based energy is the need for safetechniques of hydrogen storage and transport. These issues could be avoided by usingammonia and other storage materials as indirect sources of hydrogen. In place of hydrogen,ammonia may be transported in compressed cylinders or as decomposable materials such asMg(NH3)6Cl2 or (NH4)2CO3. This allows for on-site production of hydrogen for fuel cells throughthermal cracking or catalytic methods (Yin et al., 2004; Christensen et al., 2006; Elm?e et al.,2006; Lan et al., 2012) Alternatively, ammonia may be used directly in alkaline, alkalinemembrane, and solid oxide fuel cells (Zamfirescu and Dincer, 2008; Hejze et al., 2008;Zamfirescu and Dincer, 2009; Lan et al., 2012). Recovered sources of ammonia from wasteproducts represent ideal candidates for hydrogen production, compared to widely used and non-renewable hydrogen precursors such as natural gas.CHAPTER 2: BACKGROUND102.1.4 Economic benefits of recovery of ammonia at municipal WWTPsThe main benefit of ammonia recovery from post-digestion streams is the nitrogen removalaspect. At municipal WWTPs, it is not uncommon for return streams to contribute 15% to 20%of the nitrogen loading to secondary operations; yet, these inputs make up only 1% of theinfluent flow (Fux and Siegrist, 2004). To satisfy strict effluent quality regulations, sidestreamprocesses for nitrogen removal are an attractive alternative to costly expansion of mainstreambiological processes. For perspective, a New Hampshire study (2010) examined 18 secondaryWWTPs processing from 0.5 to 25 MLD, with an average effluent total nitrogen of about 18mg/L. To upgrade these plants to meet 8 and 3 mg/L total nitrogen limits, capital costs wereestimated at between 45M to 60M and 58M to 67M USD respectively (Kessler, 2010). Theoperational costs for these plants are approximated at 13 to 15 USD/kg total nitrogen removedfor a limit of 8 mg/L and 70 to 86 USD/kg for a limit of 3 mg/L. This cost may vary significantlyfrom larger BNR plants (35 to 490 MLD). Case studies from nine BNR plants suggestsoperational costs between 0.30 and 2.20 USD/kg of total nitrogen removed, with the majorityattributed to aeration costs (USEPA, 2008). Decentralized nitrogen removal processes are anoption for treatment of post-digestion streams; however, these generally involve high externalcarbon consumption and/or high aeration costs. With the assumption of an 85% ammoniaremoval from a 150 m3/d digester supernatant stream at 1000 mg/L NH4-N, operational costs fora partial nitritation-anammox system have been estimated at 2.50 EUR/kg total nitrogenremoved, while a nitrification-denitrification system would cost between 3.05 to 4.10 EUR/kg(Fux and Siegrist, 2004).Physical-chemical techniques for sidestream ammonia removal have the potential toconsiderably reduce the operating costs of central biological operations, while reducing nitrousoxide emissions. A preliminary cost analysis by Evans and Thompson (2009) comparedtechnologies for 90% recovery of the ammonia in digester supernatant from a moderately sizedsecondary WWTP. This comparison suggested that capital expenditure divided over 20 years,plus operating costs for steam stripping-condensation, vacuum distillation-acid scrubbing, or airstripping-acid scrubbing systems would amount to an expense of between 1 and 2 GBP/kg ofNH4-N recovered. Another study by ThermoEnergy Corporation (2007) modelled the savingsthat could be obtained for two 500 MLD WWTP scenarios by recovering approximately 90% ofthe ammonia contained in dewatering centrate using their patented technology (see SectionCHAPTER 2: BACKGROUND112.2). The first scenario was characterized by digestion of sludge from combined carbon andnutrient removal (single sludge plant) while the other handled sludges from separate carbon andnutrient removal operations (two sludge plant). Table 1 outlines the results of this modellingstudy. Considerable savings in aeration energy, sludge disposal, and chemical costs could begained by removing the ammonia resolubilized during digestion (ThermoEnergy Corporation,2007).Table 1 ? Modelled savings for two WWTPs implementing ammonia recovery fromdewatering centrate (ThermoEnergy Corporation, 2007)Savings Single Sludge Plant Two Sludge PlantMethanol Reduction 38% 19%Alkalinity Reduction 10% 13%Sludge Reduction 6% 3%Aeration Energy Reduction 10% 13%The other major incentive of ammonia recovery is the potential for internal revenue generation.Table 2 lists 2013 values for various ammonium fertilizers as provided by the US Department ofAgriculture Economic Research Center (2013). Recently, the value of recovered reactivenitrogen has been estimated between 260 to 770 USD per tonne (WERF, 2010; Orentlicher,2012). If only 15% of the influent total nitrogen was recovered from a WWTP treating 1.5 tonneN/d, internal revenues as high as 63,000 USD per year could be generated.Table 2 ? 2013 prices for various ammonium-derived fertilizers (USDA ERS, 2013)Fertilizer Price (USD/tonne)Anhydrous ammonia 932Urea 44%-46% nitrogenAmmonium nitrateAmmonium sulfate651598574Nitrogen solutions (%30) 451CHAPTER 2: BACKGROUND122.2 Existing ammonia recovery options for post-digestion side streamsSeveral techniques exist for the recovery of ammonia from concentrated wastewater such asmunicipal biosolids streams. One of the simplest methods is through the addition of magnesium,orthophosphate, and caustic to remove ammonia via the formation of struvite. Struvite crystalsare easily separated from the wastewater and can be employed as a fertilizer (Le Corre et al.,2009; WERF, 2010; Liu et al., 2013). After moderate capital expenditures of around 200,000EUR for a crystallizer, this type of operation has been estimated to cost as much as 6 EUR/kgnitrogen recovered (Cilona et al., 2009). Although struvite has been valued at between 180 and300 EUR/tonne, recent research is focussed on the recovery of phosphorus from waste, ratherthan the conversion of high quality orthophosphate to struvite (Le Corre et al., 2009; WERF,2010; Liu et al., 2012). Some techniques for ammonia recovery that stand out include ammoniaair stripping-scrubbing, absorption using membrane contactors and adsorption by ion exchange.Similar to the technology proposed in this study, each requires their own combination ofresource and energy inputs.Air stripping is a well established method of wastewater ammonia removal. Air is contacted withwastewater to allow for the transfer of ammonia from the liquid to gas phase. However, at thenear neutral pH of wastewater, ammonium dominates and, therefore, significant caustic additionis required to raise the pH high enough to convert the majority of ammonium to un-ionizedammonia. Additionally, ammonia stripping systems may employ various combinations ofaeration, steam application, vacuum induction, wastewater conveyance and heating whichcontribute significant operational and maintenance costs (Elston and Karmarkar, 2003; Evansand Thompson, 2009; Orentlicher, 2012; Ulbricht et al., 2013). These operating costs have beenapproximated to be between 1 GBP/kg to 6 EUR/kg NH4-N removed or 13.50 EUR/h alongsiderelatively high capital costs of 230,000 to 300,000 EUR for infrastructure (Evans and Thompson,2009; Cilona et al., 2009). A similar process is used to recover the ammonia. The ammonia off-gas is contacted with an acid solution (acid scrubbing) and the result of absorption is anammonium salt solution which can be marketed to various industries (Elston and Karmarkar,2003; Cilona et al., 2009). Ammonia stripping is well suited to ammonia-rich wastewater such ashuman urine. For instance, a bench scale study by Ba?ak?ilardan-Kabakci et al. (2007)demonstrated 97% volatilization of urine ammonia, of which nearly 100% was recovered in apacked bed, sulfuric acid scrubber.CHAPTER 2: BACKGROUND13A recently developed technology utilizes gas permeable membranes to transfer ammonia fromwastewater to an acid absorption phase. Essentially, the wastewater is in contact with bundlesof hollow fibre membranes and the ammonia is driven across the membrane into a flowing acidsolution by the concentration gradient between the two liquid phases. This gradient remainsstrong as the ammonia reacts with the acid to form an ammonium salt solution that can berecovered. Cilona and colleagues (2009) achieved greater than 85% ammonia recovery frompower plant condensed flue gas using membrane gas transfer. Compared to other ammoniarecovery processes, they claimed that membrane contactors would contribute relatively lowcapital costs of around 150,000 EUR while operation (including chemical, pumping, andmembrane replacement costs) is estimated at 1.21 EUR/kg nitrogen recovered or 0.17 EUR/m3for the 200 mg/L NH4 wastewater processed at 25 m3/h. In another application with industrialwastewater, up to 95% recovery of ammonia has been reported using a similar membranecontactor (Ulbricht et al., 2013). Similarly to ammonia stripping, this technique requiresammonia to be present as un-ionized ammonia and, therefore, requires high caustic dosing.Further, operation and maintenance costs are involved in absorbent pumping, and membranereplacement (Cilona et al., 2009; Ulbricht et al., 2013).Ion exchange techniques have been used frequently to remove ammonia from wastewater.Adsorbents for these processes include zeolites which are natural mineral materials andsynthetic ion exchange resins. One promising adsorbent is clinoptilolite, which can be placed inpacked bed columns and adsorb wastewater ammonium (Hedstr?m, 2006; Beler-Baykal et al.,2011; Allar and Beler-Baykal, 2013). Using this technology, ammonia removals as high 97%have been achieved in treating human urine and greater than 86% ammonia recovery has beenobtained through regeneration of exhausted clinoptilolite (Beler-Baykal et al., 2011; Allar andBeler-Baykal, 2013). Interestingly, a recent study showed that clinoptilolite can also be used toremove and recover up to 99% of the orthophosphate in human urine (Allar and Beler-Baykal,2013). Additionally, clinoptilolite is a common soil conditioner and exhausted material can beused directly as a fertilizer.A more complex ion exchange method for the recovery of ammonia from wastewater is theAmmonia Recovery Process (ARP) patented by ThermoEnergy Corporation (Fassbender, 2001;ThermoEnergy Corporation, 2007). The first stage of ARP involves the adsorption of wastewaterammonium on selective ion exchange resins. These resin columns are then regenerated usingCHAPTER 2: BACKGROUND14a zinc and sulfuric acid solution. After regeneration, a solution containing recovered ammonia aswell as zinc and sulfate remains. With the addition of more acid to this solution, the zinc doublesalt, ammonium zinc sulfate hexahydrate ((NH4)2SO4ZnSO4?6H2O), can be crystallized. Thedouble salt crystals are harvested and heated to produce ammonia and sulfur trioxide (SO3)gases leaving behind solid zinc sulfate. The off-gas from this process is absorbed in sulfuricacid and this mixture can be concentrated by evaporation to produce marketable solidammonium sulfate. In a pilot study at the Oakwood Beach WWTP in New York, the ARP provedsuccessful in recovering nearly 100% of the ammonia from municipal dewatering centrate. Thistechnology consists of multiple processes some of which require heavy chemical addition orheating contributing to operating costs of around 2.64 USD/m3 for a 650 mg/L NH4 wastewaterprocessed at a rate of 8 m3/d. However, the complexity of this technology?s design lends to itsvery high capital cost of as much as 44M USD (Fassbender, 2001).CHAPTER 3: RESEARCH OBJECTIVES15CHAPTER 3: RESEARCH OBJECTIVESThe purpose of this research is to demonstrate the potential for wastewater ammonia removalby struvite crystallization using synthetic newberyite as a surrogate for thermally decomposedstruvite. It also works to explore the possibility that these materials can by employed to producestruvite suitable as a feedstock for the UBC-developed thermal decomposition process.The objectives of this study are defined as follows:1. To develop the physical-chemical concept for the ammonia removal stage of theproposed ammonia recovery technology specifically ammonium uptake during theconversion of newberyite to struvite2. To verify optimal conditions of the process as defined by chemical equilibrium modellingand experimental results3. To delineate the rates and mechanisms of the process4. To provide recommendations regarding process operation, including optimal physical-chemical parameters and cost-effectivenessCHAPTER 4: LITERATURE REVIEW16CHAPTER 4: LITERATURE REVIEW4.1 Chemistry of magnesium and phosphate compoundsIn most cases, struvite can form between pH 6 to 9 in nutrient-rich wastewater if sufficientmagnesium is present. Struvite crystallization is promoted also by decreases in temperaturewhich occur during the conveyance of post-digestion streams. In water at 25? C, struvitesolubility decreases with the increase of pH up until its minimum solubility around pH 10.3 afterwhich struvite becomes more soluble (Ohlinger et al., 1998; Ohlinger et al; 1999; Bhuiyan et al.,2007). However, struvite is not the only magnesium compound that could exist in conditionsinherent of most wastewaters. Other crystalline solid phases such as newberyite, bobbierite(Mg3(PO4)2?8H2O), or magnesite (MgCO3) could form and perhaps limit the magnesium andorthophosphate available for struvite formation (Taylor et al., 1963; Boistelle and Abbona, 1983;Kontrec et al., 2005; Babi?-Ivan?i? et al., 2006; K?nigsberger and K?nigsberger, 2006; Shand,2006). Eqns. 6 to 9 demonstrate the formation of these magnesium compounds.Struvite: Mg2+ + NH4+ + HPO42- + 6H2O? MgNH4PO4?6H2O(s) + H+ (6)Newberyite: Mg2+ + H2PO4- + 3H2O? MgHPO4?3H2O(s) + H+ (7)Bobbierite: 3Mg2+ + 2HPO42- + 8H2O? Mg3(PO4)2?8H2O(s) + 2H+ (8)Magnesite: Mg2+ + CO32-? MgCO3 (9)Newberyite can exist even in the presence of high ammonium if pH drops to between 5 and 6(Boistelle and Abbona, 1983; Kontrec et al., 2005; Babi?-Ivan?i? et al., 2006). In contrast,bobbierite and, in the presence of carbonate, magnesite may form in significant quantities abovea pH of 8 in magnesium enriched nutrient solutions (Taylor et al., 1963; K?nigsberger andK?nigsberger, 2006; Shand, 2006). Further, newberyite or bobbierite crystallization could occuralongside struvite due to increases in solution temperature. This is because struvite solubilityincreases as solution temperature is enhanced above 25? C, while newberyite and bobbieriteCHAPTER 4: LITERATURE REVIEW17solubility decreases slightly (Boistelle and Abbona, 1983; Kontrec et al., 2005; Babi?-Ivan?i? etal., 2006; K?nigsberger and K?nigsberger, 2006).4.1.1 Aqueous equilibria affecting speciation of magnesium, ammonium,orthophosphate, and carbonateThere are many simultaneous reactions occurring in wastewater which affect the speciationand, therefore, the activity of aqueous magnesium, ammonium, orthophosphate, and carbonate.Eqns. 10 to 32 lists some of the potential aqueous equilibria that are indirectly involved in theformation of magnesium and phosphate compounds such as struvite, newberyite, bobbierite,and magnesite (USGS, 2013).H+ + OH- H2O (10)4Mg2+ + 4H2O Mg4(OH)44+ + 4H+ (11)Mg2+ + H2O MgOH+ + H+ (12)Mg2+ + 2HPO42- Mg(HPO4)22- (13)Mg2+ + PO43- MgPO4- (14)Mg2+ + H2PO4- MgH2PO4+ (15)Mg2+ + HPO42- MgHPO40 (16)Mg2+ + 2HPO42- + 2H+ Mg(H2PO4)20 (17)2Mg2+ + 2HPO42- Mg2(HPO4)20 (18)H3PO4 H2PO4- + H+ (19)H2PO4- HPO42- + H+ (20)HPO42- PO43- + H+ (21)NH3 + H+ NH4+ (22)Mg2+ + NH3 MgNH32+ (23)Mg2+ + 2NH3 Mg(NH3)22+ (24)Mg2+ + 3NH3 Mg(NH3)32+ (25)NH4+ + HPO42- NH4HPO4- (26)CO2 + H2O H2CO3 (27)H2CO3 HCO3- + H+ (28)HCO3- CO32- + H+ (29)Mg2+ + HCO3- Mg(HCO3)+ (30)CHAPTER 4: LITERATURE REVIEW182Mg2+ + HCO3- Mg2CO32+ + H+ (31)Mg2+ + 2HCO3- Mg(HCO3)20 (32)4.1.2 Solubility productsThe solubility products (Ksp) of magnesium salts vary with temperature according to the negativeenthalpy change of their formation reaction but are generally measured at 25? C in water. Thefundamentals of Ksp and how it is related to solid phase saturation are explained in more detailin Section 4.3. Table 3 lists some experimentally determined Ksp values for the solid phases ofinterest to this study.Table 3 ? Experimentally determined pKsp at 25? C for various magnesium and phosphatecompoundsSolid Phase pKsp = -Log10Ksp ReferenceStruvite13.1513.2613.3613.3613.6813.47Taylor et al., 1963aOhlinger et al., 1998Babi?-Ivan?i? et al., 2002Bhuiyan et al., 2007Koutsoukos et al., 2007Lobanov et al., 2013Newberyite5.825.785.88Taylor et al., 1963bVerbeeck et al., 1984Lobanov et al., 2013Bobbierite 25.2025.47Taylor et al., 1963bLobanov et al., 2013Magnesite 7.527.80Pokrovsky et al., 1999B?n?zeth et al., 20114.2 Struvite and newberyite morphologyStruvite and newberyite crystallize with a variety of habit and morphology depending greatly onfactors such as supersaturation, pH, temperature, elemental molar ratios and crystal age.Struvite takes orthorhombic morphologies and single, twinned, and dendritic crystal shapeshave been reported. Single crystals may be rod-like, prismatic platelets, coffin-shaped orCHAPTER 4: LITERATURE REVIEW19needle-like while dendrites may be X-shaped twins, multi-branched, or star-shaped (Abbona etal., 1985; Babi?-Ivan?i? et al., 2002; Kontrec et al., 2005). Generally, smaller, more elongatedand dendritic crystals are believed to form at higher supersaturation. At low supersaturationlarger, rod-like or tabular types are observed (Abbona et al., 1985). Newberyite crystals aretrigonal and generally take rhombohedral or pseudo-octagonal morphologies (Boistelle et al.,1983; Babi?-Ivan?i? et al., 2002; Kontrec et al., 2005). Figures 2 and 3 provide some examplesof reported crystal morphologies for struvite and newberyite respectively.Figure 2 ? Star-shaped dendritic (a), X-shaped twinned (b), coffin-shaped (c), and rod-like(d) struvite crystals (Abbona et al., 1985)a bc dCHAPTER 4: LITERATURE REVIEW20Figure 3 ? Pseudo-octagonal newberyite crystals amongst dissolving tabular struvitecrystal (a); rhombohedral newberyite crystals (b and c) (Boistelle et al., 1983; Kontrec etal., 2005; Babi?-Ivan?i? et al., 2002)4.3 Factors affecting struvite formation4.3.1 SupersaturationA solution?s degree of supersaturation with respect to struvite is the primary factor in itscrystallization. Struvite supersaturation ratio (SS) is directly dependant on magnesium,ammonium, and orthophosphate activity. However, solution pH, temperature, and ionic strengthalso play a role in supersaturation and, therefore, struvite formation. Eqn. 33 demonstrates SSas a relationship between constituent activities and solubility product (S?hnel and Garside,1992).= { } (33)The most common way of controlling supersaturation ratio in a struvite crystallization process isthrough pH adjustment. Although nutrient-rich wastewaters such as post-digestion streams maybe initially basic, the formation of struvite results in the release of hydrogen ions and, therefore,chemical addition is generally required to maintain basic conditions and maximizeorthophosphate removal through nucleation and growth of struvite crystals. It is widely acceptedthat an in-reactor pH of 8 to 9 will allow for effective struvite crystallization (Andrade anda b cCHAPTER 4: LITERATURE REVIEW21Shuiling, 2001; Le Corre et al., 2009; Huang et al., 2011). This may be accomplished throughthe addition of a basic solution of sodium hydroxide, magnesium oxide (MgO), or brucite(Mg(OH)2). Raising the pH affects supersaturation ratio indirectly by increasing theorthophosphate activity through a shift in the orthophosphate equilibria. For instance, additionalhydroxide neutralizes hydrogen ions resulting in a shift from dihydrogen phosphate to hydrogenphosphate, which is utilized in the formation of struvite as demonstrated by Eqns. 34 and 35.H2PO4- HPO42- + H+ (34)Mg2+ + NH4+ + HPO42- + 6H2O? MgNH4PO4?6H2O(s) + H+ (35)Solution temperature is also indirectly involved in supersaturation. As temperature increases,struvite solubility product increases according to the negative enthalpy change of the struviteformation reaction. An increase in struvite solubility product, Ksp-S, results in a reduction ofstruvite supersaturation ratio, SS. In general, struvite recovery potential is higher forwastewaters with lower temperature due to its decreased solubility.The ionic strength of the crystallizer feed affects the activity of struvite?s constituents,magnesium, ammonium, and orthophosphate. Higher ionic strength results in more interactionbetween solution ions and, therefore, lower effective availability or activity of these constituents.This effect is quantified using activity coefficients denoted as ? which are a function of solutionionic strength, ion valence, and ion effective diameter. Activity is defined as the product of thespecific activity coefficient for an element and its molar concentration. An increase in ionicstrength results in a decrease in activity coefficient and, therefore, a reduction of constituentactivity. Hence, a less saline wastewater has higher constituent activities and is moresupersaturated with respect to struvite.4.3.2 Mg:P and N:P molar ratioStruvite is composed of equimolar parts magnesium, ammonium, and orthophosphate.Therefore, struvite crystallization is limited to the lowest molar concentration between theseconstituents regardless of solution supersaturation. Unless the local water contains significantCHAPTER 4: LITERATURE REVIEW22hardness, wastewater is generally magnesium limited and supplemental magnesium must beadded to effectively remove orthophosphate as struvite. In a struvite recovery process, this isaccomplished with the addition of a concentrated magnesium feed stock containing dissolvedmagnesium chloride (MgCl2 or MgCl2?6H2O), magnesium oxide, or brucite. Studies havedetermined that struvite crystallization is enhanced in solutions with magnesium in excess oforthophosphate. To maximize orthophosphate removal efficiency, a Mg:P molar ratio of 1.05:1to 1.3:1 has been recommended (Jaffer et al., 2002). Further, Huang et al. (2003) found thatpellet size, hardness and density were increased above this Mg:P range in a pilot study using aUBC struvite crystallizer.Municipal wastewater streams generally carry an ammonia molar concentration many timeshigher than that of orthophosphate. This is advantageous with regard to controlled struvitecrystallization, as increased N:P molar ratio has been shown to enhance orthophosphateremoval (M?nch and Barr, 2001). Fattah (2004) observed a positive correlation between N:Pmolar ratio and orthophosphate removals during their struvite recovery pilot study. However,there are a lack of studies which systematically examine the effect of N:P molar ratio on struviteformation.4.3.3 HydrodynamicsThe hydrodynamic conditions and turbulence induced within a reactor are believed to play a rolein struvite crystallization. Ohlinger and colleagues (1999) demonstrated that an increase inmixing rate results in more rapid accumulation of struvite. Further, it is posited that crystalnucleation rates may also be increased in cases of enhanced turbulence. However, aggressivemixing may also result in crystal breakage and attrition (Franke and Mersmann, 1995). Withfluidized bed struvite reactors such as the UBC struvite crystallizer, turbulence is dependent onupflow velocity as well as struvite load, particle size distribution and reactor scale. Theperformance of these technologies relies on a lower bound upflow velocity that allows for thefluidization of struvite particles and adequate exposure of particle surfaces to the feed solution.Further, an upflow velocity too high will result in the loss of fine struvite crystals, in turn, affectingstruvite agglomeration rates. Although the UBC struvite crystallizer is capable of producingrelatively large struvite pellets, the effect of turbulence on struvite pelletization is not wellunderstood due to the complex hydrodynamics inherent to its design. Nevertheless,CHAPTER 4: LITERATURE REVIEW23examination of struvite grown in this type of reactor suggests that higher upflow velocity resultsin a harvest of larger pellets with higher density and hardness (Huang, 2003; Fattah, 2004).4.4 UBC struvite crystallization processIn 1999, UBC acquired funding from British Columbia Hydro to commence research onphosphorus recovery from various waste streams, with the intention of producing a fertilizersuitable for the enrichment of oligotrophic streams. Bench-scale studies led to the developmentof the UBC struvite crystallizer (Dastur, 2001; Adnan, 2002). Many projects have beencompleted in this area, but the pilot studies utilizing real waste streams from Penticton, LuluIsland and Annacis Island WWTPs were integral in the progression from batch tests tocommercialization of struvite recovery in less than a decade (Britton, 2002; Huang, 2003;Fattah, 2004).4.4.1 Pilot-scale UBC struvite crystallizer designThe pilot-scale crystallizers used at these WWTPs followed the general characteristicsillustrated in Figure 4. They are composed of a FBR, an injector, an external clarifier, andstorage for chemical additives and feed. The success of this technology originates from thedesign of the injector and the FBR. Feed, recycle feed, supplemental magnesium, and causticare mixed at the injector. The high turbulence and supersaturation ratio resulting from influentmixing allows for the rapid nucleation of struvite. These nuclei grow into crystals as the fluidflows upward through the FBR. The FBR is made up of four zones of varying diameter. Frombottom to top, this includes the harvest zone or wasting zone, the active zone, the fines orreaction zone, and the seed hopper. With the largest diameter, the seed hopper is essentially aclarifier, which retains the small crystals (seeds) long enough for them grow larger. As zonediameter decreases, turbulence increases due to higher upflow velocities with the highest beingat the injector. The active and harvest zones are characterized by high turbulence and,therefore, high particle collision frequency. Struvite fines entering these zones agglomerate intolarger particles along with nuclei. Over a period of hours to days, these agglomerates grow intoround, hard pellets formed by layers of struvite smoothed by attrition and abrasion. The pelletseventually remain in the harvest zone where they can be recovered by draining a portion of thecrystallizer. Feed and recycle bypasses allow for continuous operation during harvesting. TheCHAPTER 4: LITERATURE REVIEW24lengths of each zone have been varied from 45.7 to 275 mm to provide various hydraulicretention times (HRT). However, HRT and, therefore, orthophosphate and magnesium residualsin the crystallizer effluent are controlled primarily by recycling. The external clarifier facilitatesremoval of struvite fines lost to the effluent, while a portion of the supernatant is returned to thecrystallizer.CHAPTER 4: LITERATURE REVIEW25Figure 4 ? General schematic of UBC struvite crystallization process (top) andcrystallizer injector (bottom) (Fattah, 2004)CHAPTER 4: LITERATURE REVIEW264.4.2 Pilot-scale UBC struvite crystallizer operationThe operation of a struvite crystallizer is largely waste specific. The principle operationalparameters of interest to this process are the in-reactor supersaturation ratio, in-reactor Mg:Pmolar ratio, feed temperature, total reactor flow, recycle ratio, HRT and crystal retention time(CRT). Parameters used to evaluate struvite recovery performance are effluent PO4-P, percentremoval of orthophosphate and ammonia-nitrogen and pellet diameter. Table 4 summarizes therange of operation and performance for important struvite recovery pilot studies.Table 4 ? Range of operation and performance for UBC struvite crystallizer pilot studiesOperational ParameterPilot-StudyBritton, 2002Penticton digestersupernatantHuang, 2003Lulu & Annacisdigester supernatantFattah, 2004Lulu dewateringcentrateaSupersaturation ratio (SS) 1.1~2.2 1.1~1.9 1.0~1.9bMg:P molar ratio 1.0~16.8 2.0~21.7 1.1~30Temperature (? C) 16~25 10~20 15~29Total reactor flow (L/min) 2.4~10.2 3.1~4.8 8.3~23.1cRecycle ratio 3.0~23 4.0~10.3 6~12HRT (mins) Not reported 3.6~9.4 4.0~9.5dCRT (days) 12~47 8~20 Not reportedPerformanceEffluent PO4-P (mg/L) 3.9~43.6 3~13.5 2~54% PO4 Removal 0~91 88~98 24~100% NH4 Removal 0~26 1~22 5~10Pellet diameter (mm) 0.5~2.1 1.5~3.5 1.4~3.6a. Cube root of the ratio of conditional solubility product of the solution leaving the injector to that ofequilibrium (PS-reactor/PS-eq)1/3b. Mg:P molar ratio of the solution leaving the injectorc. Recycle flow divided by the influent flow to the reactord. The volume of crystal bed divided by the volumetric rate of crystal harvest4.5 Struvite decomposition products as ammonia removal agentsAn ideal material for the stabilization of aqueous ammonium would contain both magnesium andorthophosphate. Struvite can be decomposed to various products either in air or in solution. Themajority of the ammonium contained in struvite may be eliminated through release to solution orvolatilization to the atmosphere. However, if some ammonium remains in the solid phase, itsCHAPTER 4: LITERATURE REVIEW27capacity to remove wastewater ammonium via dissolution and struvite reformation is limited.Decomposed struvite containing minimal residual ammonium has a greater aqueous ammoniaremoval potential. In other words, the required dose of this material would be smaller than onethat contains more residual ammonium as demonstrated by Eqns. 36 and 37.2Mg(NH4)0.5H0.5PO4?XH2O + NH4+ + (6-2X)H2O? 2MgNH4PO4?6H2O + H+ (36)MgHPO4?XH2O + NH4+ + (6-X)H2O? 2MgNH4PO4?6H2O + H+ (37)A struvite decomposition product is suitable for removal of ammonia from wastewater if it:1. Contains ammonium at an N:P molar ratio of less than one (ie. ammonia removalefficiency of material increases as N:P approaches zero)2. Is less thermodynamically stable than struvite and will dissolve readily to releasemagnesium and orthophosphate in the wastewaterAmmonia removal agents can be produced by either wet or dry processes of struvitedecomposition. Several studies have proven that these residues can be used to removewastewater ammonia via struvite recrystallization and that the struvite formed can be repeatedlydecomposed. As illustrated by Figure 5, ammonia could theoretically be removed continuouslyfrom wastewater by reusing magnesium and orthophosphate as cycled between struvite andstruvite decomposition products. This section reviews the current state of knowledge on struvitethermal decomposition products and their effectiveness as ammonia removal agents.CHAPTER 4: LITERATURE REVIEW28Figure 5 ? Conceptual schematic of NH4 removal through reuse of Mg and PO44.5.1 Struvite decomposition in solution ? ?Wet process?By adjusting solution conditions such as pH and temperature, struvite instability can be induced.For instance, struvite may convert partially or fully to another magnesium phosphate bychanging one or both of these parameters. In the interest of ammonia recovery, thisdecomposition should result in the release of struvite ammonium to solution while preservingmagnesium and orthophosphate in a solid phase and this suspension can be settled anddecanted. The ammonia in the supernatant can be recovered while the solid phase can bereused as a source of magnesium and orthophosphate for removal of more ammonia fromwastewater. In this case, the transformation from struvite to a magnesium phosphate phase,which is low in or devoid of ammonium occurs in excess water and is, therefore, referred to asthe ?wet process?.StruviteCrystallizationWet or DryStruviteDecompositionStruviteEffluentRecoveryHeat + H+/OH-orHot airWastewaterNH4+ + PO43-NH4+orNH3(g)NewberyiteMg2+ + PO43-Mg2+ + OH-CHAPTER 4: LITERATURE REVIEW29Struvite can be replaced by newberyite through simply adding acid to a suspension. Work byBoistelle et al. (1983) demonstrated that the drop in pH induced by struvite formation caneventually provide the conditions for complete conversion of struvite to newberyite if the initialsolution pH is low enough. As pH decreases, struvite supersaturation ratio eventually reachesunity (SS = 1) and struvite will begin to dissolve as long as newberyite forms simultaneously.Struvite will completely dissolve if the solution is sufficiently supersaturated with respect tonewberyite (SN > 1). Complete replacement of struvite by newberyite was shown to occur at apH between 4 and 5.5, with the process favoured at higher temperature. Babi?-Ivan?i? andcolleagues (2006) expanded on this topic revealing the reaction kinetics of the struvite-newberyite conversion with adequate mixing. Using various combinations of initial pH and initialSN:SS ratios, it was confirmed that struvite is fully converted to newberyite in a pH range of 4 to6. At a temperature of 25? C, this process could be completed in 30 minutes, given an initialsolution of low pH and high SN:SS ratio.Figure 6 ? Schematic diagram of reusing MAP residues for ammonia removal by aciddipping as proposed by Zhang et al. (2004)A chemical reuse and ammonium recovery application was proposed by Zhang et al. (2004)employing a struvite-newberyite system as shown in Figure 6. The two-stage process involvesammonia removal from wastewater by struvite crystallization followed by acidification of theseparated struvite to produce a suspension containing newberyite. By adjusting the pH andtemperature of a struvite suspension to 5 and 60? C respectively, a mixture of newberyite andCHAPTER 4: LITERATURE REVIEW30struvite was produced in 90 minutes. The ammonium-rich acid supernatant is recovered as afertilizer and the newberyite formed is reused to remove additional wastewater ammonia.Although the extent of the struvite-to-newberyite conversion was not reported, this solid phasewas reused successfully as an ammonia removal agent.Struvite can be decomposed to other magnesium phosphates by thermal-alkali treatment of thesuspension. T?rker and Celen (2007) demonstrated this with struvite formed in anaerobicallypretreated industrial wastewater containing molasses. The ammonia was removed via struviteformation by adding magnesium and orthophosphate in excess. With the addition of caustic tothe struvite suspension at a OH:NH4 (initial NH4+ concentration in wastewater) molar ratio of 1:1,ammonia is released to the liquid phase. At this high pH, the ammonia may be volatilized byheating. 81% and 100% removal of ammonia from the alkali suspension was achieved at 110?C and by distillation respectively in only minutes. They suggested that the evolved ammonia gascould be recovered in boric acid. Although it was noted that it could contain Mg3(PO4)2 and/ormagnesium pyrophosphate, the dominant material in the resultant solid phase was notidentified. Nonetheless, these residues proved to be a suitable source of magnesium andorthophosphate for high removal of ammonia when added to another sample of wastewater.Researchers from the State Key Laboratory of Environmental Aquatic Chemistry in Beijing,China, have adopted a similar approach for ammonia recovery from various wastewaters byreusing struvite decomposition residues. Their studies verify and expand upon the resultsobtained by T?rker and Celen (2007). Each of these studies followed a similar conceptualprocess flow involving two main stages. In the first stage, ammonia was removed from thewastewater by producing struvite. The second stage involved the decomposition of collectedstruvite by thermal-alkali treatment. The decomposed struvite produced in each study was laterused successfully to remove more ammonia from the same wastewater. The supernatantproduced in the second stage was considered a recoverable source of ammonia. The type ofwastewater, the OH:NH4 molar ratio of the alkali struvite suspension, the heating temperatureand time, the percent release of ammonia during struvite decomposition and the suggestedcomposition of decomposed struvite are summarized for each of these studies in Table 5.CHAPTER 4: LITERATURE REVIEW31Table 5 ? Summary of struvite decomposition studies using thermal-alkali treatmentWastewaterOH:NH4MolarRatioHeatingTemp. (?C)HeatingTime(hours)% Releaseof NH4-NSolidPhases;PossibleImpuritiesReferenceMolassesindustry 1:1 ?110 <3a81-100Not identified;Mg3(PO4)2,Mg4P2O7T?rker andCelen, 2007Landfillleachate 1:1 90 2 >96%bAmorphousMgNaPO4;Ca, K, AlHe et al., 2007Saponification 1:1.1 100 3 Not reportedNot identified;Mg3(PO4)2,Mg4P2O7Huang et al.,2009Coking 2:1 110 3 90bAmorphousMgNaPO4;Mg3(PO4)2,Mg4P2O7Zhang et al.,2009Piggery 1:1 110 3 Not reportedbAmorphousMgNaPO4;Ca, K,Huang et al.,2011a. Percent removal of solubilized NH3 from liquid phase by volatailization; Percent elimination of NH4+from struvite was not reportedb. XRD identified amorphous phase; MgNaPO4 was suggested as dominant material by authors4.5.2 Ammonia removal following ?wet process?Ammonia can be removed from wastewaters using residues that were produced bydecomposing struvite in excess water. One of these residues is newberyite. By dissolvingstruvite in acid and heating the resulting solution, newberyite can be formed. This process canbe reversed by collecting newberyite and adding it along with caustic to an ammonia solution.The newberyite dissolves and the released magnesium and orthophosphate are utilized torecrystallize struvite. Zhang and colleagues (2004) proposed that this reversible process couldbe exploited for continuous ammonia removal and recovery using the two-stage applicationshown in Figure 6. Newberyite was added to an ammonia solution to provide a Mg:N:P molarratio of 1:0.5:1 and pH was maintained at 8.5 for 4 hours to produce struvite. The mixture ofstruvite and newberyite was collected and the struvite portion was converted back to newberyitein the acidification stage as described in Section 4.5.2. The newberyite material from this stagewas added once again to the same solution. This procedure was repeated five more times. Inthe struvite formation stage, of the first five cycles, greater than 98% ammonia removal wasCHAPTER 4: LITERATURE REVIEW32achieved. However, between 2% and 5% of the magnesium and orthophosphate was lost fromthe solid phases to the supernatants of both the ammonia removal and acidification stages. Itwas believed that enough was lost by the fifth cycle to reduce the ammonia removal efficiencyof the sixth stage to 88%.Another decomposition product that could be used to remove ammonia from wastewater is theamorphous phase produced from struvite under alkaline conditions at elevated temperatures.Table 5 reviews the conditions of various thermal-alkali treatments of struvite to form thismaterial. Magnesium and orthophosphate can be cycled through alternating ammonia removaland thermal-alkali stages. The type of wastewater, the pH and reaction time of ammoniaremoval/struvite formation stage, the percent removal of ammonia from the wastewater for thefirst and last cycles and the residual orthophosphate concentration in the supernatant from theammonia removal stage are summarized for each of these studies in Table 6.Table 6 ? Summary of ammonia removal studies using struvite decomposed underthermal-alkali conditionsWastewater pH ReactionTime (h)%Removalof NH4-NCycle No.PO4Residual(mg/L)ReferenceMolassesindustry 8.5 Not reported92771st5th Not reportedT?rker andCelen, 2007Landfillleachate 9.0 2.096841st6th 2-10 He et al., 2007Saponification 9.0 0.5 99a991st6th <1Huang et al.,2009Coking 9.5 1.5 85701st5th Not reportedZhang et al.,2009Piggery b9.4-8.5 1.0 80651st5th <5-70Huang et al.,2011a. Initially Mg and PO4 are in excess and separate stages were employed for dissolution ofdecomposition products and residual PO4 recoveryb. Required no adjustment of pH; pH varied as struvite formed; Initially Mg was in excess asmagnesiteCHAPTER 4: LITERATURE REVIEW334.5.3 Thermal decomposition of struvite in air ? ?Dry process?Under certain conditions, struvite may become thermodynamically unstable in air, even atambient temperature. Depending on its morphology, struvite is believed to transform completelyinto newberyite in open systems. However, this phenomenon occurs slowly over a period ofseveral months (Cohen, 1966; Whitaker, 1967; Ribbe, 1969). In the early 1900?s, research ondry thermal decomposition of struvite was performed to improve a common technique used fordetermination of magnesium or orthophosphate concentration, which involved the formation ofstruvite and ignition to magnesium pyrophosphate. Errors in the results of this technique wereattributed to the fact that the conditions at which orthophosphate transforms to pyrophosphatewere not agreed upon universally. Kiehl and Hardt (1933) completed a study which investigatedthis transition by determining the dissociation pressures of various magnesium phosphates. Inopen atmosphere at temperatures between 40? and 60? C , it was suggested that struvite losesfive moles of its water of crystallization forming dittmarite. Based on the composition of thegaseous phase, struvite loses both ammonia and water above 60? C. In heating dittmarite,pyrophosphate was detected in solid phases whenever ammonia evolved. This study declared250? C as a suitable temperature to quickly and completely decompose struvite to magnesiumpyrophosphate.The first study that presented evidence of an intermediate phase occurring during heatingbetween dittmarite and magnesium pyrophosphate was completed by Paulik and Paulik (1975a,1975b). This research used a thermogravimetric method which was novel at the time toinvestigate thermal decomposition of struvite under quasi-isothermal conditions. The methodreferred to as quasi-isothermal thermogravimetry (QTG) employs automation to maintain aconstant heating temperature when the mass loss rate passes a preset threshold. QTG allowsfor more accurate determination of the temperature at which phase transformations occur, aswell as demonstrating the kinetics of these reactions. Paulik and Paulik verified thatapproximately five moles of water of crystallization is lost from struvite, with simultaneouselimination of ammonium at temperatures above 90? C. However, the ammonia evolvedamounts to no more than 5% of the struvite ammonium. This suggested that a phase believedto be primarily dittmarite is quite stable between 90? and 230? C. Based on the shape of theQTG curve in this range, it was posited that dittmarite may lose its remaining ammonium, waterof crystallization and constitution by two overlapping reactions each potentially attributingCHAPTER 4: LITERATURE REVIEW34intermediate phases. In the range of 200? to 250? C dittmarite was shown to lose 50% to 80% ofits volatile components. In theory, dittmarite rapidly loses its water of crystallization to formMgNH4PO4 at around 230? C followed by the slow elimination of its remaining ammonium toMgHPO4. Finally, the water of constitution evolves producing magnesium pyrophosphate at500? C. The transformations involved in thermal decomposition of struvite can be representedby Eqns. 38 to 41 (Paulik and Paulik, 1975a; 1975b).MgNH4PO4?6H2O? MgNH4PO4?H2O + 5H2O (38)MgNH4PO4?H2O? MgNH4PO4 + H2O (39)MgNH4PO4? MgHPO4 + NH3(g) (40)2MgHPO4? Mg2P2O7 + H2O (41)In a study by Abdelrazig and Sharp (1988), a similar approach was employed to investigate thethermal decomposition temperatures of struvite and dittmarite. Conventional differentialthermogravimetry confirmed that mass loss from struvite begins to rapidly occur just above 60?C, with a well defined peak occurring at 100? C that was believed to represent high stability ofdittmarite. Similarly to Paulik and Paulik (1975a), QTG results suggested that five moles ofwater of crystallization is simultaneously evolved with a small amount of ammonia during thistransformation. Powder X-ray diffraction (XRD) analyses identified dittmarite and an amorphousphase in a cement containing struvite which was heated at 235? C. When the same materialwas heated at 300? C, it was found to contain only this amorphous phase.In the research following these earlier studies, similar observations were reported. Results fromseveral more experiments suggest that ammonia and water evolve simultaneously duringthermal decomposition of struvite. The temperature at which this begins ranges from 40? to 55?C. The same studies demonstrated rapid decomposition of struvite at temperatures between 85?and 115? C (Sarkar, 1991; Frost et al., 2004; Bhuiyan et al., 2008). Sarkar (1991) and Bhuiyanet al. (2008) also claimed that dittmarite becomes stable in this range. However, Sugiyama andcolleagues (2005) were the only other researchers to detect dittmarite in heated struvite (100? toCHAPTER 4: LITERATURE REVIEW35150? C isotherms) using XRD analysis. Several of these studies also noted the formation of afully X-ray amorphous phase beginning at heating temperatures ranging from 70? to 160? Cextending to as high as 500? C (Sarkar, 1991; Sugiyama et al., 2005; Bhuiyan et al., 2008;Kurtulus and Tas, 2011). This amorphous phase has been suggested to be analogous toMgHPO4 but considering the reviewed literature it might be deduced that this phase couldrealistically be a mixture of multiple phases which might include dittmarite, MgNH4PO4, MgHPO4and magnesium pyrophosphate with composition dependant on heating temperature. Based onthis dependence, the amorphous material would vary in its capacity to remove ammonia fromwastewater.In response to the commercialization of struvite recovery from municipal wastewater, a novelstudy was performed by Novotny (2011) on thermal decomposition of struvite pellets. Thesepellets were approximately 2 mm in diameter and were isothermally heated for 24 hours in atemperature range of 40? to 200? C. He claimed that about 70% of struvite ammonium had beeneliminated at just 80? C along with approximately five moles of water. This contradicts the QTGresults of Paulik and Paulik (1975a) and Abdelrazig and Sharp (1988), that suggested that onlya small amount of ammonia is released below 100? C. Furthermore, 81% and 87% removal ofstruvite ammonium from the pellets were calculated for 160? and 200? C respectively. About70% of the ammonium could be eliminated at heating temperatures between 100? and 200? C in30 to 60 minutes. Based on previous research, the material produced at these temperatureswas believed to be a mixture of struvite and amorphous MgHPO4, but probably contained somemagnesium pyrophosphate; this would explain its poor solubility and, therefore, low ammoniaremoval capacity, as discussed in Section 4.5.4.4.5.4 Ammonia removal following ?dry process?The thermal decomposition of struvite in air is a topic relevant to many fields including analyticalchemistry, mineralogy, thermogravimetry and cement production as is demonstrated in Section4.5.5.  More recently, research has shifted to focus on using these struvite decompositionresidues as a renewable source of magnesium and orthophosphate for wastewater treatmentvia struvite precipitation. Stefanowicz and colleagues (1992) proved that the solid phase productof dry thermal decomposition can be utilized to remove ammonium from wastewater. By heatingstruvite at 150? C for 24 hours, they produced a material which was believed to be mostlyCHAPTER 4: LITERATURE REVIEW36Mg3(PO4)2 but no XRD analysis was implemented. In acknowledgement of preceding research,this decomposition product was more likely a mixture of amorphous magnesium phosphatesand magnesium pyrophosphate. It was demonstrated that the decomposed struvite could beused to treat a concentrated ammonia solution. However, to accomplish this in the set ofreaction times investigated, the residues had to first be dissolved in the solution by acidification.100% removal of ammonia was achieved in 24 hours by maintaining a pH of 9.3 and by addingdecomposed struvite to provide a low N:P molar ratio compared to the stoichiometric ratiorequired for struvite formation. Although the final compositions of solid phases were notreported, it is likely that a significant portion of the magnesium and orthophosphate added to thesolution was resolubilized and lost to the supernatant under these conditions. As the authorsposited, the struvite produced from decomposition residues could potentially be recycled forrepeated ammonia removal. However, a supplementary source of magnesium andorthophosphate would probably be required for consistent process efficiency.Recycling of the struvite produced through dry thermal decomposition has been proven possibleby a group of researchers, most of which are affiliated with the Department of Chemical Scienceand Technology at the University of Tokushima, Japan. A preliminary study by Sugiyama et al.(2005) showed that newberyite would dissolve and reform as struvite in the presence ofammonium, given a solution Mg:N:P molar ratio of 1:1:1 and a constant pH of 8. This resulted ina 77% ammonia removal efficiency after 3 hours. In the following experiment, struvite wasdecomposed in air at 150? C to produce an amorphous phase which contained what wasbelieved to be MgHPO4. When the struvite crystallization process was repeated using thismaterial, a removal of only 41% was achieved. The resulting solid phase was collected anddecomposed once again and the process was repeated a third time, affording a 33% removal ofammonia, perhaps, indicating incomplete elimination of struvite ammonium duringdecomposition. Hence, a portion of the added magnesium and orthophosphate could berecycled repeatedly by combined dry thermal decomposition and struvite formation. The loss ofmagnesium and orthophosphate incurred during this process was not reported.The potential of the dry process of struvite decomposition was later verified by Sugiyama et al.(2007). A layer of MgHPO4?1.2H2O was synthesized on a glass plate using a novel sol-geltechnique. The sheet was immersed in an ammonia solution and, through the formation ofdittmarite, more than 30% of the ammonia was removed in 3 hours. The sheet was then heatedCHAPTER 4: LITERATURE REVIEW37to 150? C for 3 hours to eliminate the ammonium that was taken up and re-immersed in thesame solution. This process was repeated two more times affording ammonia removalefficiencies between 10% and 20%. XRD analysis of the material following decompositionidentified pronounced dittmarite peaks indicating that 150? C is likely too low to eliminate theammonium on the spent sheet. By increasing MgHPO4?1.2H2O availability for ammonia removaland optimizing the decomposition process, this application could be a promising technique forcontinuous wastewater treatment.As described in Section 4.5.5, Novotny (2011) examined the dry thermal decomposition ofcommercially available struvite pellets. Pellets heated at 160? C in air for 24 hours were usedsuccessfully to remove ammonia through dissolution of an amorphous magnesium phosphateand reformation of struvite. When immersed in a 700 mg/L NH4-N solution with a constant pH of8, 40 and 80 g/L doses of decomposition product corresponded to approximately 50% and 93%ammonia removals in 2 hours . These doses provide a very low solution N:P molar ratioresulting in the solubilization of orthophosphate at about 60 and 190 mg/L PO4-P for 40 and 80g/L doses respectively. At a constant pH of 9, residual orthophosphate concentrations werereduced to below 10 mg/L PO4-P due to the higher supersaturation with respect to struvite. Thisresearch demonstrated that decomposed struvite pellets could be used to treat for wastewaterammonia, at the cost of magnesium and orthophosphate release or high caustic addition.4.6 Potential ammonia gas capture techniquesThere are several options for the recovery of ammonia gas released during thermaldecomposition of struvite. Proven methods include condensation and acid scrubbing orabsorbtion. An ammonia-rich distillate or compressed ammonia could be generated from themixture of air, water vapour, and ammonia using cryogenic or temperature-pressure-basedcondensation techniques (T?rker and Celen, 2007; Evans and Thompson, 2009; Orentlicher,2012) This recovered ammonia could also be packaged as a compressed liquid using similarmethods. This type of product would be ideal for fuel combustion or for using ammonia as aprecursor for hydrogen production. Alternatively ammonia-containing gas is contacted withsulfuric, nitric, or phosphoric acid using a scrubber or a membrane contactor to eventuallyproduce an ammonium salt solution (Fassbender, 2001; ThermoEnergy Corporation, 2007;Cilona et al., 2009; Evans and Thompson, 2009; Ulbricht et al., 2013). This solution can be usedCHAPTER 4: LITERATURE REVIEW38directly by industry or can be processed further to produce solid ammonium salts throughevaporation or caustic addition. Alternatively, ammonia could be effectively adsorbed bymagnesium chloride to form magnesium hexamine chloride (Mg(NH3)6Cl2) (Christensen et al.,2006; Elm?e et al., 2006; Zamfirescu and Dincer, 2009; Lan et al., 2012). The stored ammoniacan be easily desorped later through thermal decomposition of the metal amine. However, thismethod requires the removal of water vapour from recovery gas mixtures prior to adsorption, asmagnesium chloride binds water to form hydrates.4.7 Conclusions for development of present studyIn moving forward with the present study, several knowledge gaps exist in literature pertainingto the study of ammonia removal using thermally decomposed struvite. Several studies claimedto have employed a MgHPO4-containing struvite decomposition product to remove ammoniafrom solutions (Zhang et al., 2004; Sugiyama et al., 2005; Sugiyama et al., 2007). However,only the material produced by Sugiyama and colleagues (2007) provided any evidence ofcrystalline MgHPO4 and none of the previous research concerning thermal decomposition ofstruvite named newberyite as a potential product. Although the conversion of struvite tonewberyite has been examined in detail (Boistelle and Abbona, 1983; Kontrec et al., 2005;Babi?-Ivan?i? et al., 2006), no research has yet focussed on investigating the conversion ofnewberyite to struvite in the presence of ammonium. Of the work directed towards ammoniaremoval technologies (Zhang et al., 2004; Sugiyama et al., 2005; Sugiyama et al., 2007; T?rkerand Celen, 2007; He et al., 2007; Huang et al., 2009; Huang et al., 2011), few took a systematicapproach to delineating the multi-parameter effects on rates and mechanisms. In general, thesestudies offered a limited scope of experimental conditions with respect to pH control,decomposed struvite dosing, temperature, and initial solution chemical composition. Forinstance, previous research has centred on presenting the potential of these techniques withlittle discussion of how these parameters would affect overall effluent quality and operatingcosts.With regard to controlled struvite formation, several researchers have completed detailedstudies on systems with post-dosing Mg:P and N:P molar ratios above 1 (Dastur, 2001; Adnan,2002; Britton, 2002; Huang, 2003; Fattah, 2004). However, there is a lack of literature whichevaluates the potential for combined ammonia and orthophosphate removal at Mg:P and N:PCHAPTER 4: LITERATURE REVIEW39molar ratios of 1 and below. The present study works to fill in some of this missing informationand act as a foundation for the selection of various technological features of the proposedammonia recovery system.CHAPTER 5: MATERIALS AND METHODS40CHAPTER 5: MATERIALS AND METHODS5.1 Description of performed batch testsMultiple sets of batch tests were employed to examine the rate and mechanisms of newberyite-to-struvite conversion in the presence of ammonium. Twenty-four experiments were conductedto represent the fundamental system of synthetic newberyite in a simple ammonia solution. Abroad range of experimental conditions were provided in this phase of the study to ?cast a widenet? in identifying a potential optimal region of operation for the proposed technology. Thistechnology relies on the dissolution of newberyite to provide the magnesium andorthophosphate required to induce struvite formation and, hence, ammonia removal.However, complete conversion of newberyite to struvite can only be accomplished through themaintenance of newberyite and struvite supersaturation. To adjust SN and SS, variouscombinations of pH control and temperature were selected. These ranges were meant torepresent the potential temperatures of municipal post-digestion streams and the caustic dosesthat might be required to promote struvite formation. Further, a newberyite dose range thatwould result in a total suspension Mg:N:P molar ratio between 1:1:1 and 1:1.4:1 was employedto observe how excess ammonia relative to the stoichiometry of struvite formation affects ratesand mechanisms. These doses were selected to target high newberyite-to-struvite conversionefficiency, while maintaining low magnesium and orthophosphate residuals. Eight subsequentexperiments were performed in an ?optimal? range of conditions for synthetic solutions intendedto represent the basic characteristics of specific wastewaters. Each combination of parameterswas represented by duplicate batch tests. Table 7 outlines the value of specific parameterswhich were combined to construct experimental matrices (refer to Sections 5.2 and 5.3.2 forammonia solutions and synthetic newberyite compositions respectively).Table 7 ? Summary of experimental parameters for duplicate batch testsSolution No. ofExperiments pH ControlTemperature(? C)Mg:N:P MolarRatioAmmonia solution 24 None, 7, 8, 9 10, 25, 35 1:1.1:1, 1:1.4:1Synthetic crystallizer effluent 4 7, 8 10, 25 1:1:1Synthetic centrate 4 7, 8 10, 25 1:1:1CHAPTER 5: MATERIALS AND METHODS415.2 Ammonia solutions ? Synthetic wastewatersThe compositions of the simple ammonia solution and synthetic wastewaters allocated topreviously described batch tests are compared to typical dewatering centrate from AnnacisIsland WWTP in Table 8. The ammonia solution was prepared to be around 700 mg/L NH4-N torepresent the typical ammonia concentration in the effluent of a pilot-scale struvite crystallizationprocess (Huang, 2002; Fattah, 2004). It should be noted that Annacis Island centrate ischaracterized by high organic matter, atotal suspended solids content greater than 2000 mg/Land alkalinity above 6000 mg/L as CaCO3. A typical anaerobic digester supernatant carriesalkalinity at 2000 to 5000 mg/L (Tchobanoglous et al., 2003). To minimize synthetic wastewatersalinity, an alkalinity of 2000 mg/L, typical of Lulu Island WWTP digester supernatant, wasadopted for these solutions (Huang, 2003). However, some alkalinity losses as carbon dioxideoccurred during preparation due to the low initial pH of nutrient solutions.Synthetic wastewaters were prepared using reagent grade chemicals and distilled water. Thisincluded ammonium phosphate monobasic (NH4H2PO4) and ammonium chloride (NH4Cl).Magnesium was added as magnesium chloride hexahydrate (MgCl2?6H2O). Alkalinity wasprovided as sodium bicarbonate (NaHCO3). Synthetic crystallizer effluent was prepared by abatch procedure simulating conventional struvite recovery. An equimolar quantity of magnesiumwith respect to initial orthophosphate content was mixed with 3 L of synthetic centrate in abeaker on a stir plate. Temperature was maintained constant at 35? C using an aquariumheating rod, while pH was maintained at 8.5 using a 2 M caustic solution made from sodiumhydroxide (NaOH) pellets. Once pH was stabilized, mixing and heating were continued overnight and the resulting suspension was filtered using Whatman 5 qualitative 12.5 cm diameterfilters and a vacuum apparatus. All feed solutions were stored in 8 L Nalgene containers.CHAPTER 5: MATERIALS AND METHODS42Table 8 ? Annacis Island WWTP centrate compared to studied solutionsFeed pH Mg(mg/L)NH4-N(mg/L)PO4-P(mg/L)Alkalinity(mg/L as CaCO3)Annacis Island centrate 8.5 1 1000 150 > 6000Ammonia solution 4.5-5.5 0 737 0 0Synthetic crystallizer effluent 8.1-8.4 16 919 19 1470Synthetic centrate 7.4-7.9 0 1008 147 14755.3 Synthetic newberyite5.3.1 SynthesisThe synthetic newberyite used as a surrogate for thermally decomposed struvite pellets wasprepared in the laboratory. Reagent grade 71.3 g/L sodium phosphate dibasic (Na2HPO4) and544 g/L MgCl2?6H2O solutions were prepared using distilled water. The pH of theorthophosphate solution was adjusted to around 8.3 using several drops of concentratedhydrochloric acid (HCl). 800 mL of the orthophosphate solution and 150 mL of the magnesiumsolution were mixed in a 1 L beaker on a heated stir plate. Temperature was maintained above25? C and concentrated HCl was added dropwise to maintain a pH below 6.5 for an hour. Theresulting suspension was filtered using Whatman 5 qualitative 12.5 cm diameter filters and avacuum apparatus. The retained solids were washed several times with distilled water andreagent alcohol. The synthetic newberyite was then dried in an oven at 90? C overnight toevaporate any residual water and alcohol. This procedure yielded about 70 g newberyite andthis was stored in a closed plastic sample bottle.5.3.2 AnalysisThe purity of synthesized newberyite was evaluated by XRD and chemical analyses. XRDanalyses showed that no other solid phases were formed while preparing synthetic newberyite.XRD output graphs for this newberyite product is provided in Appendix C. A known mass ofnewberyite was dissolved in distilled water with the addition of concentrated hydrochloric acid toreduce pH to below 2. This solution was then diluted and analyzed for orthophosphate using theflow injection method outlined in Section 5.6.2. Table 9 presents the results of chemical analysisCHAPTER 5: MATERIALS AND METHODS43and the batch experiments of which they were used in. Magnesium and orthophosphate contentwith mass was not the same for each batch of newberyite due to variations in its water ofcrystallization (theoretically there are 3 moles of water per mole of newberyite). This variation inmolecular weight and composition is believed to be due to small differences between batches inthe method of synthesis, such as crystallization temperature and pH, as well as drying andstorage time.Table 9 ? Synthetic newberyite chemical composition and concerned experimentsDate prepared Mg and PO4-P content(mmoles/g newberyite)Experiments used in(Section)Theoretical 5.73 -May 2012 5.34 6.2July 2012 5.47 6.3September 2012 5.44 6.5November 2012 5.95 6.45.4 Materials and equipment5.4.1 Batch test method and apparatusDuplicate batch tests were performed in two glass jacketed containers sitting atop stir plates.Temperature control was accomplished using a cooled/heated water bath that providedcontinuous flow of water through the jackets. Prior to each set of experiments, the containerswere cleaned using a 5% hydrochloric acid solution to dissolve any residual newberyite orstruvite that had adhered to the glass during previous experiments. This was followed by threerinses with distilled water. 500 mL of synthetic wastewater feed was added to each reactor andthe water bath and stir plates were turned on to initiate temperature adjustment. Temperaturewas monitored using the temperature sensor feature of two handheld pH probes (see Section5.4.2). Once the feed had stabilized with respect to the desired reaction temperature,premeasured amounts of newberyite were added simultaneously to both apparatuses. Thesynthetic centrate tests required supplementation of magnesium which was provided usingpipette-measured volumes of a 0.5 M MgCl2 solution. The time was noted and the apparatuseswere covered with glass caps to reduce heat loss/gains. However, batch tests were open to theCHAPTER 5: MATERIALS AND METHODS44atmosphere. Finally, caustic burettes and pH probes were positioned for pH control andmonitoring. pH was controlled for the duration of study through the dropwise addition of a 2 MNaOH solution using two 50 mL burettes. At every sample time, pH and caustic consumptionwere recorded. For two sets of experiments, an ammonium selective electrode and conductivityprobe were positioned in Reactor 1 (see Sections 5.4.3 and 5.4.4).5.4.2 pH monitoringpH was monitored in both apparatuses using two Oakton pH 11 Series meters complete withATC probes. pH meters were calibrated prior to each set of experiments using pH 4, 7 and 10standard buffer solutions that were heated in the a water bath to 25? C. The probes weresubmerged in the test suspension by insertion into openings separate from that used for causticaddition.5.4.3 Conductivity monitoringConductivity was monitored in one of the two apparatuses along with online ammoniummeasurement. For this purpose a specialized foam cap was designed to hold and submerge aconductivity, ammonium, and pH probe collectively. The ATC conductivity probe was connectedto an Oakton CON 110 meter and calibrated prior to each experiment using potassium chloride(KCl) standard solutions of similar ionic strengths to that expected in the suspensions. Thesestandards were also heated/cooled to the same temperature as the suspension for eachexperiment. The meter was connected via an analog-USB cord to a laptop PC and conductivitydata was stored every 5 seconds using the software provided with the meter.5.4.4 Ammonium monitoringAmmonium activity was monitored in one of the two reactors along with online conductivitymeasurement. The Cole-Parmer ammonium selective electrode was connected to an OaktonpH 2100 bench meter and calibrated prior to each experiment using standard solutions thatspanned the expected range of ammonium concentrations and background orthophosphate,alkalinity, and ionic strength conditions. These standards were also heated/cooled to the sameCHAPTER 5: MATERIALS AND METHODS45temperature as the feed for each experiment. The meter was connected via an analog-USBcord to a laptop PC and ammonium activity data was stored every 5 seconds using the softwareprovided with the meter.5.4.5 PHREEQC-2 chemical equilibrium modelA chemical equilibrium model was constructed for this study using PHREEQC Version 2 topredict the equilibrium liquid and solid phase compositions that would result from specific sets ofinitial and constant conditions. The experimental parameters outlined in Section 5.1 weresimulated and model-generated outputs were compared with batch test results to identifydiscrepancies between observed and theoretical compositions and to evaluate the model as atool for future studies. PHREEQC is a software provided by the United States GeologicalSurvey, capable of modelling low temperature aqueous chemical reactions. PHREEQC standsfor a pH-REdox-Equilibrium program written in C language. Environmental and geochemicalreactions of interest involving magnesium, ammonium, and orthophosphate were added to acustomized model from a database file (thermo.com.V8.R6.230) containing data compiled byJim Johnson of Lawrence Livermore National Laboratory. The attributed constants of thesereactions, such solubility products and negative enthalpy changes for reactions, were collectedfrom relevant literature; those of special interest, including struvite, newberyite, and bobbieritewere determined experimentally in a previous study at UBC (Lobanov et al., 2013). With theinput of initial composition of dissolved elements and solid complexes, the model will outputinitial batch suspension pH and saturation with respect to the considered solid phases, as wellas elemental and solid phase concentrations after equilibrium has been reached. pH control,through caustic addition, can be simulated and caustic consumption may be estimated. Overall,it is a powerful tool for estimating activity coefficients, supersaturation ratios, and correctedsolubility products for suspensions with controlled experimental conditions. Initial suspensioncomposition model inputs and equilibrium outputs are reported in Sections 6.2, 6.3, 6.4, 6.5 andAppendix D.5.5 Sample collection and preservationSamples were collected from both apparatuses at 1, 3, 6, 9, and 12 hours after initiation forbatch tests with synthetic newberyite in ammonia solution. With synthetic crystallizer effluentCHAPTER 5: MATERIALS AND METHODS46and centrate, samples were collected at 10 minutes, 1, 3, and 4 hours after dosing. This wasaccomplished using two 60 mL syringes with tube extensions allowing for sampling at the centreof the suspension column. Samples were pushed through a Millipore 47 mm diameter 0.45?mnylon membrane filter into 50 mL centrifuge tubes for storage. Two drops of concentrated HClwere added immediately after collection to induce undersaturation in samples with respect toboth newberyite and struvite to prevent further formation of solid phases. The same syringeswere used to force two full volumes of distilled water and one volume of reagent alcohol throughthe filter to wash and partially dry the retained solid phase. The filter and retained solids wereallowed to dry overnight and were stored in individual sealed sample bags.5.6 Analytical methodsAll synthetic wastewater feeds were analyzed prior to experiments for magnesium, ammonia,orthophosphate, and inorganic carbon where applicable. All liquid samples from the batch testswere analyzed for ammonia and orthophosphate. Ammonia and orthophosphate measurementswere also involved in determining N:P molar ratios of solid phase samples dissolved in a weakHCl solution. Liquid samples from tests with crystallizer effluent and centrate were alsoanalyzed for magnesium in order to estimate the newberyite and struvite supersaturation ratio ateach sample time. Each sample was prepared in triplicate and all analyses were undertaken atthe UBC Environmental Engineering Laboratory, unless otherwise specified.5.6.1 MagnesiumSamples and calibration standard solutions were diluted in 25 mL glass tubes at a 1:10volumetric ratio with a 20 g/L lanthanum solution prepared from reagent grade lanthanum nitrate(La(NO3)3) and distilled water. This was followed by addition of three drops of concentrated nitricacid (HNO3) to each tube and agitation using a vortex mixer. This background matrix reducesthe interference of other ionic species during analysis for magnesium using a Varian Inc.SpectrAA220 Fast Sequential Atomic Absorption Spectrophotometer. Prior to each set ofanalyses, fresh deionized water was provided for autosampler rinsing and the magnesium lampwas optimized and warmed up for at least 30 minutes (see Appendix A for instrumentoperational settings).CHAPTER 5: MATERIALS AND METHODS475.6.2 Ammonia and orthophosphateAmmonia and orthophosphate were measured in samples by flow injection analysis on a LachatQuikChem 8000 using Method 4500-NH3 H and 4500-P G from Standard Methods for theExamination of Water and Wastewater (APHA, AWWA, WEF, 2012). Calibration standardsolutions were composed of reagent grade potassium phosphate monobasic (KH2PO4) andammonium chloride in distilled water (see Appendix A for method details and instrumentsettings).5.6.3 Carbonate alkalinityTotal inorganic carbon was measured as a surrogate for carbonate alkalinity since the onlycarbon that synthetic feeds contained was that resulting from the addition of NaHCO3 anddissolution of carbon dioxide from the air. This was accomplished using a Lachat IL550 TOC-TNAnalyzer for Method 5310-B from Standard Methods for the Examination of Water andWastewater (APHA, AWWA, WEF, 2012). Calibration standard solutions were composed ofsodium carbonate (Na2CO3) and NaHCO3.5.6.4 XRD identification of solid phasesCrystalline phases were identified in the solids retained during filtration of suspension samples.This was accomplished with a Bruker D8 Advance X-ray diffractometer using CuK? radiation.XRD output peak patterns were identified using the powder diffraction database file, PDF-2,provided by the International Center for Diffraction Data. This instrument was located in theUBC Department of Chemistry (see Appendix A for pattern database details and instrumentsettings).5.6.5 Crystal morphologyCrystal morphology was observed using a Motic B3 Professional Series microscope and imageswere captured using Motic Images Plus software. Small quantities of the solid samples wereCHAPTER 5: MATERIALS AND METHODS48placed on glass slides and spread using a drop of reagent alcohol. Crystals were viewed at x10and x40 magnifications.5.7 StatisticsComparisons of samples from different combinations of experimental conditions were madebased on the error of the mean resulting from two measurements from each of the duplicatebatch tests. This error was calculated using an unpaired two-tailed t-test at a confidence intervalof 90%. In example, if the errors of the mean of two measurements from different sets ofexperiments overlapped these two values were deemed not statistically different. Eqns. 42, 43,and 44 demonstrate how standard deviation (s), standard error, and error of the mean werecalculated where x, n (n = 2), and t (t1,0.05 = 6.314) represent the measurement value, thenumber of samples compared, and the student?s-t value respectively (Berthouex and Brown,2002).s = ? (x ? x) (42)standard error = ? (43)? error = standard error ? t , . (44)5.8 Terminology5.8.1 Molar ratioThe term molar ratio is used with respect to both N:P and Mg:N:P. In the context of newberyitedosing, these molar ratios consider the total of each element in the initial suspensionimmediately after commencing the batch test. This includes dissolved elements and thatcontained in the solid phase reagent, newberyite. Eqns. 45 and 46 explain the concept of initialN:P and Mg:N:P molar ratios respectively.CHAPTER 5: MATERIALS AND METHODS49[NH4+(aq)] : ([newberyite(s)-PO4] + [PO43-(aq)]) (45)([newberyite(s)-Mg] + [Mg2+(aq)]) : [NH4+(aq)] : ([newberyite(s)-PO4] + [PO43-(aq)]) (46)In the context of solid phase sample composition, N:P refers to the molar ratio of the ammoniumcontent in struvite to the orthophosphate content in the mixed solid phase of newberyite andstruvite. Since newberyite contains no ammonium, this ratio quantifies the fraction or molarpercent of struvite in the solid phase. Eqn. 47 demonstrates this convention.[struvite(s)-NH4] : ([newberyite(s)-PO4] + [struvite(s)-PO4]) (47)5.8.2 Solubility product (Ksp-N and Ksp-S)The newberyite and struvite solubility products are referred to in this study as Ksp-N and Ksp-Srespectively. Ksp is the product of the activities of ionic components concerned in the formationof a solid phase. Eqns. 48 and 49 demonstrate the calculation of Ksp-N and Ksp-S based onreagent activities (S?hnel and Garside, 1992).Ksp-N = {Mg2+}{HPO42-} (48)Ksp-S = {Mg2+}{NH4+}{PO43-} (49)Section 4.1.2 lists reported and experimentally determined solubility products for a variety ofphases that could also form in solutions supersaturated with respect to struvite. From previousexperiments by Dr. Sergey Lobanov at UBC, Ksp-N and Ksp-S at 25? C are taken as 10-5.88 and10-13.47 respectively (Lobanov et al., 2013).5.8.3 Supersaturation ratioSupersaturation ratio is defined as the square root of the product of reagent activities divided bythe concerned solid phase solubility product for newberyite and cubed root of that for struvite.CHAPTER 5: MATERIALS AND METHODS50Newberyite and struvite supersaturation ratios are referred to in this study as SN and SS and arerepresented by Eqns. 50 and 51 respectively (S?hnel and Garside, 1992).= { } (50)= { } (51)Supersaturation ratio is a measure of the crystallization potential of a solution. For instance,struvite will readily crystallize in a solution with a SS greater than 1, while it will dissolve in asolution with a SS less than 1. In this study, supersaturation ratio of samples was estimated fromknown elemental concentrations and model-generated parameters, including activity coefficientsand temperature-corrected solubility products.5.8.4 Removal efficiencyIn the context of ammonia, removal efficiency is defined as the percent of total ammonia (asconverted to ammonium) removed through struvite formation with respect to the total ammoniainitially in solution. Essentially, this parameter quantifies the ammonium displaced to the solidphase of the suspension at various sample times during batch tests. Eqn. 52 demonstrates thisrelationship between initial ammonia concentration and that at sample time, x.= [ ] [ ][ ] ? 100% (52)CHAPTER 6: RESULTS AND DISCUSSION51CHAPTER 6: RESULTS AND DISCUSSION6.1 Fundamentals of newberyite dissolution-struvite formationmechanism in the presence of ammoniumThe first set of batch tests performed represents the most fundamental cases of this study. Theyare intended to determine the ranges of pH, temperature and newberyite dose, which promotehigh ammonia removal and high newberyite-to-struvite conversion efficiency, while limitingmagnesium and orthophosphate losses to solution. The following sections answer the followingquestions:? How does the composition of liquid and solid phases change when synthetic newberyiteis added to a simple ammonia solution?? How fast do these chemical reactions occur?? How does the liquid and solid composition compare to that of the model-generatedequilibrium?6.2 Transformation of newberyite into struvite in ammonia solution:Phase 1 ? Suspension Mg:N:P molar ratio of 1:1.1:1The rates and mechanisms of ammonia removal and orthophosphate solubilization wereobserved during 12 batch tests combining synthetic newberyite and a simple ammonia solution.Table 10 outlines the average suspension characteristics immediately after newberyite is addedto the solution. This can be considered time zero and Mg:N:P molar ratio represents theproportions of magnesium and orthophosphate contributed by newberyite and the ammoniainitially in solution. This set of experiments was intended to represent a Mg:N:P molar ratio of1:1:1. However, it was determined later, that one mole of synthetic newberyite does not weighexactly 174.3 g. As reported in Section 5.3.2, one mole of synthetic newberyite may containslightly less or more than three moles of water of crystallization. Therefore, the following resultsare for a Mg:P:NH4 molar ratio of 1:1.1:1, indicating that ammonia was in slight excess.CHAPTER 6: RESULTS AND DISCUSSION52Table 10 ? Suspension characteristics at t = 0 h for Mg:N:P molar ratio 1:1.1:1 newberyitedose batch testsReagents added as solid newberyite Initial solution characteristics Mg:N:PMolar RatioNewberyiteadded (g/L)Mg(mM)PO4-P(mM)Mg(mM)NH4-N(mM)PO4-P(mM)8.7 46.7 46.7 0 52.6 0 1:1.1:16.2.1 pH effect on rate and efficiency of ammonia removalBy increasing the pH to 7 and above, undersaturation with respect to newberyite is induced(Boistelle and Abbona, 1983; Kontrec et al., 2005; Babi?-Ivan?i? et al., 2006). Further,maintenance of neutral to alkaline conditions during struvite crystallization remains the simplestmethod of controlling struvite crystallization (Dastur, 2001; Adnan, 2002; Britton, 2002; Huang2003; Fattah, 2004). In a solution initially containing newberyite and aqueous ammonia, pHplays a significant role in both the rate of ammonia removal and how much remains asequilibrium is approached.Similar trends were observed with regard to the average ammonia removal rate for eachcombination of pH and temperature as illustrated in Figure 7. In all cases, ammonia removaloccurred rapidly within the first hour. This indicates an enhanced rate of struvite formation that isdriven by the high ammonium activity and, therefore, high SS at the start of the test. Initialreaction rates were highest at pH 8 and 9. Further, the ammonia residual nears equilibrium inonly three hours. For the batch tests without pH control, pH increased from approximately 5 to6.6, due to the partial dissolution of newberyite, and gradually decreased as struvite formationbegins to dominate. These conditions allowed for an ammonia removal of only 11% to 15% after12 hours for the three temperatures. Ammonia removal was far greater in the cases where pHwas maintained at 7 and above. At pH 7, the ammonia removal was boosted to between 71%and 79% by the end of the tests. The ranges of removal for pH 8 and 9 were slightly higher at80% to 84% and 77% to 85% respectively. However, the final ammonia residuals for pH 7, 8and 9 were statistically similar for 10? and 25? C, while a slight difference between batch testswith respect pH was exhibited for residuals at 35? C.CHAPTER 6: RESULTS AND DISCUSSION53Overall, the data is in agreement with the results obtained by Sugiyama and colleagues (2005).They added newberyite to a 114 mg/L NH4-N solution to provide a suspension Mg:N:P molarratio of 1:1.1:1. In the present study, comparable ammonia removal efficiencies were obtainedat pH 8, with a more concentrated solution. Interestingly, similar removals were observed at afar lower suspension N:P molar ratio, when compared to analogous tests by Novotny (2011)with an amorphous product of struvite thermal decomposition. This suggests that newberyite isa more efficient ammonia removal agent. At a newberyite dose of 1:1.1:1 Mg:N:P, it is apparentthat 71% to 84% of ammonia can be removed from this simple solution by maintaining a pHbetween 7 and 8. However, the question remains: How much orthophosphate is released and,hence, lost during this processCHAPTER 6: RESULTS AND DISCUSSION54Figure 7 ? NH4-N removed for various pH conditions at (a) 10?, (b) 25?, and (c) 35? C01002003004005006007008000 2 4 6 8 10 12NH4-N (mg/L)Time (h)No control(pH~6.3)pH 7pH 8pH 901002003004005006007008000 2 4 6 8 10 12NH4-N (mg/L)Time (h)No control(pH~6.5)pH 7pH 8pH 901002003004005006007008000 2 4 6 8 10 12NH4-N (mg/L)Time (h)No control(pH~6.6)pH 7pH 8pH 9abcCHAPTER 6: RESULTS AND DISCUSSION556.2.2 pH effect on rate and extent of orthophosphate solubilizationOrthophosphate residuals originate primarily from the dissolution of newberyite. On the contrary,higher supersaturation, with respect to struvite, results in lower orthophosphate residuals due toenhanced ammonia removals via struvite crystallization. The following section explores theeffect of pH on orthophosphate solubilization.Unless magnesium phosphates other than newberyite and struvite form during ammoniaremoval, the residual magnesium molar concentration is analogous to that of orthophosphate.As discussed in subsequent Section 6.2.5, no undesirable solid phases were detected in any ofthe samples analyzed by XRD. Figure 8 plots average orthophsophate residuals with time forvarious combinations of pH and temperature. Similarly to ammonia removal, orthophosphatesolubilization rates are highest in the first hour. This indicates that newberyite dissolves rapidlyduring this time and this is probably enhanced by the maintenance of low SN by simultaneousstruvite formation. Orthophosphate concentration also appears to approach a state ofequilibrium by 3 hours, especially when pH was controlled. The 12 hour orthophosphateresiduals were highest in tests with no control of pH ranging from 199 to 352 mg/L PO4-P.Residuals were significantly lower when pH was maintained above 7. This is expected to be dueto the higher SS induced at increased pH. After 12 hours, the orthophosphate concentration wassignificantly lower at pH 7 than with no pH control for all temperatures ranging from 115 to 119mg/L. The lowest final residuals were measured at pH 8 and 9. Orthophosphate was in therange of 25 to 33 mg/L and 11 to 20 mg/L for pH 8 and 9 respectively. However, there was nosignificant difference between orthophosphate residuals at pH 8 and 9 for all cases.The observed solubilization of orthophosphate confirms that a solid phase must first dissolve inorder to transform into another, regardless of how chemically similar they are (Boistelle andAbbona, 1983; Kontrec et al., 2005; Babi?-Ivan?i? et al., 2006; Novotny, 2011). It should also benoted that the residuals measured in the present study compare well to similar tests by Novotny(2011), in which considerably lower suspension N:P molar ratios were examined using thermallydecomposed struvite pellets. Orthophosphate in the liquid phase, following ammonia removalconstitutes a loss of the material that was added as newberyite. This is a concern because theeffluent of the proposed ammonia recovery system must be returned to mainstream biologicalprocesses. If the effluent of this system is higher in orthophosphate than that in the influent, itCHAPTER 6: RESULTS AND DISCUSSION56could make ammonia recovery less attractive, as the returned residual would increasephosphorus loading to secondary treatment operations. This also means more struvite will beconsumed during thermal decomposition operations to make up for this loss. Theorthophosphate residual originating from newberyite is accompanied by an equimolarmagnesium residual. If these residuals are high enough, they may promote the formation ofstruvite scale with their return, which defeats one of the functions of struvite recovery.Nevertheless, the similarity between orthophosphate residuals at pH 8 and 9, suggests that apH higher than 8 may not be necessary to maintain low orthophosphate residuals and thiswould represent a caustic savings. Further, a pH of 8 could provide high ammonia removal,while maintaining an orthophosphate concentration considerably lower than that of AnnacisIsland dewatering centrate.CHAPTER 6: RESULTS AND DISCUSSION57Figure 8 ? Residual PO4-P for various pH conditions at (a) 10?, (b) 25?, and (c) 35? C0501001502002503003504000 2 4 6 8 10 12PO4-P (mg/L)Time (h)No control(pH~6.3)pH 7pH 8pH 90501001502002503003504000 2 4 6 8 10 12PO4-P (mg/L)Time (h)No control(pH~6.5)pH 7pH 8pH 90501001502002503003504000 2 4 6 8 10 12PO4-P (mg/L)Time (h)No control(pH~6.6)pH 7pH 8pH 9abcCHAPTER 6: RESULTS AND DISCUSSION586.2.3 Temperature effect on rate and efficiency of ammonia removalTemperature affects many suspension characteristics including elemental activity, solid phasesupersaturation, and solubility, as well as reaction kinetics. For instance, newberyite is morethermodynamically stable than struvite above 25? C if pH is below 6, while struvite is lesssoluble at lower temperatures (Boistelle and Abbona, 1983; Kontrec et al., 2005; Babi?-Ivan?i?et al., 2006; Lobanov et al., 2013). Further, solid phase dissolution is enhanced at elevatedtemperatures due to the higher kinetic energy of water molecules. Temperature is expected toaffect both the rate of ammonia utilization for struvite formation and near-equilibriumcompositions.Figure 9 compares average ammonia removals over time with respect to temperature. Contraryto expectations, temperature did not appear to affect the rate of ammonia utilization. At the endof the test period, the mean ammonia removal efficiencies at 10? C were highest in most cases.This is expected to be due to lower struvite solubility in these tests. However, there was nostatistically significant difference between final ammonia residuals, with respect to temperaturefor all pH conditions.Evidently, ammonia residual is less dependent on temperature than on pH. This suggests thatfluctuations in wastewater temperature may not significantly affect ammonia removal efficiency.From a practical standpoint, the heat of biosolids digestate could be allocated to operations thatwould benefit from it prior to the ammonia recovery system to gain some process efficiency.CHAPTER 6: RESULTS AND DISCUSSION59Figure 9 ? NH4-N removed for various temperatures for (a) no pH control, (b) pH 7, (c) pH8, and (d) pH 901002003004005006007008000 2 4 6 8 10 12NH4-N (mg/L)Time (h)0 2 4 6 8 10 12Time (h)01002003004005006007008000 2 4 6 8 10 12NH4-N (mg/L)Time (h)T = 10?C T = 25?C T = 35?C0 2 4 6 8 10 12Time (h)a bc dCHAPTER 6: RESULTS AND DISCUSSION606.2.4 Temperature effect on rate and extent of orthophosphatesolubilizationNewberyite dissolution and struvite crystallization are separate mechanisms occurringsimultaneously in this system. However, they are connected because as one reactionprogresses, it affects the solution?s supersaturation with respect to the coexisting solid phase.There is the potential for temperature to play a greater role in newberyite dissolution than it doesin formation of struvite. This section examines the rates of orthophosphate solubilization andresulting residuals as affected by temperature.Figure 10 illustrates the effect of temperature on average orthophosphate residuals for variouspH conditions. Similarly to ammonia removal, the orthophosphate solubilization resulting fromnewberyite dissolution is not significantly affected by fluctuations in temperature between 10? Cand 35? C at a pH of 7 and above. The exception is in tests with no pH control. There was asignificant difference between final orthophosphate residuals for the three temperatures, despiteammonia removals that were statistically similar.Struvite crystallization is enhanced at lower temperatures while newberyite dissolution isincreased. These results reveal that wastewater cooling could provide better conversionefficiency between newberyite and struvite.CHAPTER 6: RESULTS AND DISCUSSION61Figure 10 ? Residual PO4-P for various temperatures for (a) no pH control, (b) pH 7, (c) pH8, and (d) pH 90501001502002503003504000 2 4 6 8 10 12PO4-P (mg/L)Time (h)0 2 4 6 8 10 12Time (h)0501001502002503003504000 2 4 6 8 10 12PO4-P (mg/L)Time (h)T = 10?C T = 25?C T = 35?C0 2 4 6 8 10 12Time (h)a bc dCHAPTER 6: RESULTS AND DISCUSSION626.2.5 XRD analysis of solid phase mixturesThe parallel formation of compounds other than struvite and newberyite might reduce theamount of magnesium and orthophosphate available for ammonia removal. To search forpotential contamination by these compounds, select samples of the solid mixture producedduring ammonia-NH4 batch tests were analyzed using XRD.Bobbierite is known to form at high temperature and/or high pH (Taylor et al., 1963b;K?nigsberger and K?nigsberger, 2006; Lobanov et al., 2013). Therefore, 12 hour solid samplesfrom pH 8 and 9 tests at 35?C were analyzed. Several other samples were examined, includingone from each pH condition. No compounds other than newberyite or struvite were detected.Further, newberyite was detected in every sample even at a pH of 9. Nonetheless, struvitepeaks were more well-pronouced for samples taken from systems at higher pH, whilenewberyite peaks get weaker as pH decreases. Hence, dominance of struvite peakscorresponds to increased struvite formation and, therefore, lower newberyite residuals (outputgraphs identifying solid phase XRD patterns for select samples can be found in Appendix C).The transformation of newberyite and struvite into other solid phases is undesirable becausethey may contain no ammonium or have a Mg:P molar ratio more than 1:1. This is a concernwith respect to the proposed ammonia recovery process. For instance, Eqn. 53 demonstratesthe conversion of struvite to bobbierite.3MgNH4PO4?6H2O(s) + XH2O? Mg3(PO4)2?8H2O(s) + 2NH4+ + NH3 + HPO42- + (10+X)H2O (53)In this case, ammonia removal is reduced because some of the magnesium andorthophosphate is used to form bobbierite rather than struvite. This also results in a higherorthophosphate residual because bobbierite formation requires three moles of magnesium forevery two moles of orthophosphate. In the event that it is stable during the ammonia removalstage, bobbierite could potentially accumulate in the material that is harvested for thermaldecomposition; if fresh struvite is not continuously added, the material?s ammonia removalcapacity will decrease after each subsequent cycle. Fortunately, the solid phase mixtureproduced in these systems was simply a mixture of struvite and residual newberyite.CHAPTER 6: RESULTS AND DISCUSSION636.2.6 Chemical composition of solid phase mixturesAlthough the XRD method employed could identify the crystalline compounds in solid phasesamples, it is not capable of determining their composition with respect to newberyite andstruvite content. The following section quantifies the extent of phase transformation fromnewberyite to struvite using results from elemental analyses of solid phase mixtures. Sincenewberyite contains no ammonium, N:P molar ratio essentially represents the molar fraction ofstruvite with respect to the total mixture of struvite and newberyite.Figures 11 and 12 compare the N:P molar ratio for select solid samples at 1, 3 and 12 hourreaction times. The lowest struvite yield was found in the test with no pH control at 10? C. Thesample was approximately 9% and 22% struvite after 1 and 3 hours respectively. Although thesolid compositions are comparable among samples from the same pH, it appears thattemperature plays a noticeable role in the first hour of struvite formation. For both pH 7 and 8,the fraction of struvite was slightly less in samples from 10? C tests, compared to that of 25? C.This may suggest that the effect of increased reaction rate may dominate over the effect ofstruvite?s higher solubility early during the reaction. Further, struvite yields at pH 7 (34% and42%) are considerably lower than that of pH 8 (78% and 92%) in 1 hour samples. The extent ofconversion at pH 8 and 9 is comparable. However, equilibrium is approached after only 1 hourat pH 8 and 25? C. Therefore, it appears that these conditions may be near the optimum forrapid newberyite to struvite conversion. By 3 and 12 hours, pH 7, 8 and 9 samples are within 0.1units of each other, with the exception of pH 7 at 35? C. This lower struvite yield is believed tobe due to struvite?s higher solubility at pH 7 at 35? C and corresponds with a lower ammoniaremoval efficiency.The molar concentration of magnesium was often slightly lower than that of the orthophosphatein each of the samples, as reported in Appendix B. Considering no impurities were detected byXRD, this was deemed to be due to higher instrumental error with atomic adsorption, comparedto flow injection analysis. For each of the selected samples for analysis, only one of theduplicate batch tests was examined. Therefore, error bars were not included for these figures.However, it is posited that, by providing a pH of 8 and moderate temperature, high newberyite-to-struvite conversion efficiency might be achieved, while limiting residuals.CHAPTER 6: RESULTS AND DISCUSSION64Figure 11 ? N:P molar ratio of solid phase mixtures sampled at 1 h (No pH = no pHcontrol)Figure 12 ? N:P molar ratio of solid phase mixtures sampled at 3 and 12 h (No pH = no pHcontrol)00.10.20.30.40.50.60.70.80.9110? C; No pH 10? C; pH 7 25? C; pH 7 10? C; pH 8 25? C; pH 8 25? C; pH 9N:Pmolar ratio00.10.20.30.40.50.60.70.80.9110? C;pH 725? C;pH 710? C;pH 825? C;pH 810? C;No pH35? C;pH 725? C;pH 935? C;pH 9N:P molar ratio3 h12hCHAPTER 6: RESULTS AND DISCUSSION656.2.7 Comparison of experimental results and model predicionsChemical equilibrium modelling can be a powerful tool for identifying the conditions required forstruvite or ammonia recovery from wastewater and for predicting the characteristics ofcrystallizer effluent. The PHREEQC chemical equilibrium model allows the user to input theinitial suspension composition and temperature and predicts the liquid and solid phasecomposition at equilibrium, as well as the caustic consumed when maintaining constant pHconditions. This study presented an opportunity to compare the liquid and solid compositionsobserved in 12 hour batch tests with model-predicted equilibria and, in the process, evaluate themodel?s accuracy.As previously mentioned, newberyite and struvite were the only phases detected in solidsamples. However, the model predicted that bobbierite would form in suspensions maintained at35? C. Therefore, bobbierite was removed from the model. Following this exemption, the modelpredicted newberyite and/or struvite to exist in its place and generated residual ammonia andorthophosphate values that were closer to that found experimentally. Figure 13 and 14 compareaverage ammonia removals and orthophsosphate residuals from 25? C batch tests to theircorresponding equilibrium model outputs (raw modelling results for experiments and graphicsnot included in this section are provided in Appendix D).For all of the tested conditions, the final ammonia residuals were remarkably similar to theequilibrium concentration estimated by the model. However, in several cases, the model-generated value was slightly lower. This is believed to be due to the fact that the suspensionsmay not have reached equilibrium by 12 hours. The orthophosphate concentrations were alsosimilar to their model-generated counterparts with the exception of some of the tests with no pHcontrol. The lack of correlation for these cases may be attributed to the discrepancy betweenthe measured equilibrium pH and that predicted by the model.CHAPTER 6: RESULTS AND DISCUSSION66Figure 13 ? Comparison of real and model-predicted NH4-N for 25? CFigure 14 ? Comparison of real and model-predicted PO4-P for 25? CFigure 15 compares the N:P molar ratio of solid phase samples from two batch tests with that atequilibrium as predicted by the model. The model calculated struvite yields significantly higherthan that which was observed in samples, even though ammonia removal rates suggest that01002003004005006007008000 2 4 6 8 10 12 14NH4-N (mg/L)Time (h)No control(pH~6.5)pH 7pH 8pH 9ModelledEquilibrium0501001502002503000 2 4 6 8 10 12 14PO4-P (mg/L)Time (h)No control(pH~6.5)pH 7pH 8pH 9ModelledEquilibriumCHAPTER 6: RESULTS AND DISCUSSION67these systems are approaching equilibrium. Further, in several cases the model predictedcomplete transformation of newberyite into struvite, while corresponding samples showed thatconsiderable newberyite remains even after 12 hours at conditions where struvite is highlyinsoluble.Figure 15 ? Comparison of real and model-predicted N:P molar ratios of solid phasemixturesDiscrepancy between model-predicted and experimental results is generally expected.Prediction of equilibrium states based on the laws of aqueous chemistry and thermodynamics iscomplicated by factors such as ionic activity, hydrodynamics, solid morphology, physicalenvironment and atmospheric conditions. Further, the model does not take into considerationmechanisms such as crystal growth and agglomeration. Perhaps, the extent of transformation ofnewberyite into struvite is lower than that expected theoretically due to the entrapment of smallnewberyite fragments inside struvite crystals and agglomerates. This is referred to as crystalseeding. Although the model may not be able to predict the actual solid phase composition, ithas proved itself valuable for identifying conditions where full conversion of newberyite tostruvite cannot be achieved.00.10.20.30.40.50.60.70.80.9110? C; No pH 35? C; pH 7 10? C; No pH 35? C; pH 7N:P molar ratio12 hModelEquilibriumCHAPTER 6: RESULTS AND DISCUSSION686.2.8 Initial rates of newberyite dissolution and struvite formationThe previous sections examined the rates of newberyite dissolution and struvite formation in thescale of hours. Successful struvite crystallization pilot trials are characterized by an HRT of 3 to10 mins (Huang, 2003; Fattah, 2004) which is a far shorter reaction time than the 1 to 3 hoursrequired for synthetic newberyite to convert to struvite, in the presence of ammonium. Under theassumption that pelletized struvite could be produced from a mixture of solid and aqueousreagents, it would be valuable to know the time range in which struvite begins to form, given aMg:N:P molar ratio of approximately 1:1.1:1.From previous tests, the slowest reaction kinetics are expected at a temperature of 10? C, withno pH control. Figure 16 plots the variation of pH with time for synthetic newberyite in anammonia solution with an initial concentration of 749 mg/L NH4-N. pH will vary as a result of thedissolution of newberyite and struvite formation. pH increases as newberyite dissolves due tothe release of hydrogen phosphate ions (HPO42-) which consumes hydrogen to form dihydrogenphosphate (H2PO4-). Struvite formation results in a decrease in pH due to the consumption oforthophosphate ions, which results in the release of hydrogen from hyrdrogen phosphate.Evidently, newberyite dissolution and struvite formation occur simultaneously under theseconditions. pH increases to a maximum at 20 minutes, indicating that newberyite dissolutiondominates for this period of time. A solid phase sample was collected at this time and XRDanalysis showed that there are already well pronounced struvite peaks, which confirm theoccurrence of a simultaneous newberyite dissolution-struvite formation mechanism. Followingthis point, pH decreased, suggesting that the rate of struvite formation is higher than that ofnewberyite dissolution.A sample of the liquid phase was analyzed for 20 minutes time having a magnesium, NH4-N andPO4-P concentration of 81, 727 and 103 mg/L respectively. According to the ammoniaconcentration, minimal struvite formation has taken place suggesting that a maximummagnesium and orthophosphate concentration would be found at a time slightly under 20minutes in this case. Therefore, struvite begins to form rapidly at some time between 0 and 20minutes for the examined conditions. However, it is expected that struvite formation wouldinitiate at an earlier time when caustic is added for pH control due to enhanced supersaturationCHAPTER 6: RESULTS AND DISCUSSION69with respect to struvite. Further, the maximum magnesium and orthophosphate concentrationswould be lower for these casesFigure 16 ? pH vs time for no pH control at 10? C6.3 Transformation of newberyite into struvite in ammonia solution:Phase 2 ? Suspension Mg:N:P molar ratio of 1:1.4:1Adding newberyite in excess of ammonia is expected to increase SS overall but may result inhigher solubilization of magnesium and orthophosphate, which is undesirable from a wastewatertreatment perspective. Alternatively, newberyite could be added in lower doses to provide a N:Pmolar ratio higher than 1. This excess ammonia would also result in a general increase in SS.This stage of the study intends to answer the following questions:? Will ammonia removal relative to available magnesium and orthophosphate beenhanced by increasing the N:P molar ratio above unity?? Will this limit magnesium and orthophosphate residuals?20 mins4.004.505.005.506.006.507.007.508.000 10 20 30 40 50 60pHTime (mins)Apparatus 1Apparatus 2[Mg] = 81 mg/L[NH4-N] = 727 mg/L[PO4-P] = 103 mg/LCHAPTER 6: RESULTS AND DISCUSSION70Table 11 outlines the average suspension characteristics immediately after newberyite is addedto the ammonia solution. This can be considered time zero and Mg:N:P molar ratio representsthe proportions of magnesium and orthophosphate contributed by newberyite and the ammoniainitially in solution. Chemical analyses determined an initial Mg:P:N molar ratio of 1:1.4:1,indicating a 40% excess of ammonia relative to available magnesium and orthophosphate.Table 11 ? Suspension characteristics at t = 0 h for Mg:N:P molar ratio 1:1.4:1 newberyitedose batch testsReagents added as solid newberyite Initial solution characteristics Mg:N:PMolar RatioNewberyiteadded (g/L)Mg(mM)PO4-P(mM)Mg(mM)NH4?N(mM)PO4-P(mM)7.0 37.3 37.3 0 52.6 0 1:1.4:16.3.1 Newberyite dose effect on rate and efficiency of ammonia removalAmmonia removal efficiency is reduced if you decrease newberyite dose, simply because thereis less magnesium and orthophosphate available to form struvite. For instance, for each mole ofmagnesium and orthophosphate in the suspension there is approximately 1.4 moles ofammonium and, therefore, the maximum ammonia removal that can be achieved under theseconditions is 71%. However, it is possible that ammonia removal efficiency relative to availablemagnesium and orthophosphate could be higher at reduced newberyite doses.Figure 17 offers a complete account of average ammonia removals for the 12 batch tests at1:1.4:1 Mg:N:P. The 12 hour removals ranged from 12% to 14% for no pH control and 60% to68% for pH 7, 8 and 9. The 20% reduction in newberyite does not significantly affect the kineticsof ammonia removal. In all cases, ammonia concentration approaches equilibrium in 1 to 3hours. Figure 18 compares the ammonia residuals for 1:1.1:1 and 1:1.4:1 Mg:N:P molar ratios.Evidently, ammonia residual does not vary much with newberyite dose when pH is notcontrolled. The ammonia removals for pH 7 to 9 at the 1:1.4:1 dose appear to be reduced by acommon amount, compared to the 1:1.1:1 dose. On average, this reduction of newberyite doseresults in a decrease in ammonia removal efficiency of about 15%. However, if removal iscalculated relative to the maximum amount of ammonium that can be used to form struviteCHAPTER 6: RESULTS AND DISCUSSION71based on the magnesium and orthophosphate available, 12 hour ammonia residuals for the1:1.4:1 dose are comparable to that of 1:1.1:1. For instance, ammonia removals are 7% to 16%higher than the 1:1.1:1 dose when the percent removal for 1:1.4:1 is multiplied by the initialmolar ratio of ammonia in solution to orthophosphate available (ie. 52.6 mM NH4-N/37.3 mMPO4-P). This suggests that relative ammonia removal is slightly enhanced by reducingnewberyite dose by about to 20%.CHAPTER 6: RESULTS AND DISCUSSION72Figure 17 ? NH4-N removed for various pH conditions at (a) 10?, (b) 25?, and (c) 35? C01002003004005006007008000 2 4 6 8 10 12 14NH4-N (mg/L)Time (h)No control(pH~6.3)pH 7pH 8pH 9ModelledEquilibrium01002003004005006007008000 2 4 6 8 10 12 14NH4-N (mg/L)Time (h)No control(pH~6.5)pH 7pH 8pH 9ModelledEquilibrium01002003004005006007008000 2 4 6 8 10 12 14NH4-N (mg/L)Time (h)No control(pH~6.6)pH 7pH 8pH 9ModelledEquilibriumabcCHAPTER 6: RESULTS AND DISCUSSION73Figure 18 ? NH4-N removed at 25? C for Mg:N:P molar ratio of (a) 1:1.1:1 and (b) 1:1.4:16.3.2 Newberyite dose effect on rate and extent of orthophosphatesolubilizationBy decreasing the newberyite dose, a general increase in SS is expected due to the ammoniapresent in excess. Enhanced struvite formation corresponds to a reduction in magnesium andorthophosphate residual. This section intends to identify any reduction in orthophosphatesolubilization that may result from decreasing newberyite dose by 20%.Again, the molar concentration of orthophosphate is believed to be equivalent to that ofmagnesium. Figure 19 plots average orthophosphate residuals with time for the 12 batch testsat a Mg:N:P molar ratio of 1:1.4:1. The 12 hour PO4-P concentrations ranged from 196 to 307mg/L for no pH control and 8 to 95 mg/L for pH 7, 8 and 9. Fig 20 compares the orthophosphateresiduals for 1:1.1:1 and 1:1.4:1 Mg:N:P molar ratios at 25? C. A similar trend is observedbetween the 1:1.1:1 and 1:1.4:1 doses regarding the rate of orthophosphate solubilization.Again, orthophosphate appears to be approaching an equilibrium concentration after about 1 to3 hours. Similarly to the ammonia residual, the 12 hour orthophosphate residual does not seemto be affected by a reduction in newberyite dose when pH is not controlled. For the cases withpH control, this residual does indeed appear to be slightly lower for the 1:1.4:1 dose compared01002003004005006007008000 2 4 6 8 10 12NH4-N (mg/L)Time (h)0 2 4 6 8 10 12Time (h)No control(pH~6.5)pH 7pH 8pH 9a bCHAPTER 6: RESULTS AND DISCUSSION74to that of the 1:1.1:1 dose for some cases. The discrepancies between mean 12 hourorthophosphate residuals, with respect to dose, are less than 25 mg/L PO4-P, with the exceptionof the tests at pH 7 and 35? C. In this case, the orthophosphate residual for the 1:1.4:1 dosewas 78% lower than that of the 1:1.1:1 dose. This suggests that, theoretically, an ammoniaremoval efficiency of about 15% could be sacrificed for a reduction in orthophosphate residual ifthe treated wastewater maintains some heat of digestion.CHAPTER 6: RESULTS AND DISCUSSION75Figure 19 ? Residual PO4-P for various pH conditions at (a) 10?, (b) 25?, and (c) 35? C0501001502002503003504004505000 2 4 6 8 10 12 14PO4-P (mg/L)Time (h)No control(pH~6.3)pH 7pH 8pH 9ModelledEquilibrium0501001502002503003504004505000 2 4 6 8 10 12 14PO4-P (mg/L)Time (h)No control(pH~6.5)pH 7pH 8pH 9ModelledEquilibrium0501001502002503003504004505000 2 4 6 8 10 12 14PO4-P (mg/L)Time (h)No control(pH~6.6)pH 7pH 8pH 9ModelledEquilibriumabcCHAPTER 6: RESULTS AND DISCUSSION76Figure 20 ? PO4-P residual at 25? C for Mg:N:P molar ratio of (a) 1:1.1:1 and (b) 1:1.4:16.4 Transformation of newberyite into struvite in synthetic crystallizereffluentOne objective of this study was to work towards the development of an ammonia recoverysystem by examining the reactions involved at conditions which might represent real wastewaterstreams. The effluent from a struvite crystallizer, treating a municipal post-digestion stream, wasselected as one of the more promising feeds for ammonia removal using thermally decomposedstruvite. Figure 21 is a schematic illustrating a potential process flow for combined recovery ofstruvite and ammonia from dewatering centrate. The outlined continuous or semi-continuoussystem is referred to as Reactor Configuration 1. This scenario involves four main operationsconsisting of struvite thermal decomposition (D) and primary (A) and secondary (C) struvitecrystallization separated by a newberyite dissolution stage (B). Stage A could be considered aconventional struvite crystallizer, which harvests relatively large pellets such as that studied byBritton (2002), Huang (2003), and Fattah (2004) or that commissioned by Ostara NutrientRecovery Technologies Inc. The still ammonia-rich crystallizer effluent is mixed with thenewberyite yielded from thermal decomposition of struvite in Stage B. Caustic may be added inStage B to allow for partial conversion of newberyite into struvite, which would, in turn, reducethe HRT required for secondary struvite crystallization. The suspension is mixed with caustic in0501001502002503000 2 4 6 8 10 12PO4-P (mg/L)Time (h)0 2 4 6 8 10 12Time (h)No control(pH~6.5)pH 7pH 8pH 9a bCHAPTER 6: RESULTS AND DISCUSSION77the injector of the secondary crystallizer, Stage C. The secondary crystallizer could potentiallybe a modified version of the conventional that would provide the HRT required for efficientreduction of newberyite and elemental residuals. Large pellets may not be necessary for thethermal decomposition stage; therefore, secondary crystallization might be characterized byshorter than conventional CRTs.This section examines newberyite-to-struvite conversion as a method of removing ammoniafrom a solution representing the primary struvite crystallizer effluent that could be generated atthe Annacis Island WWTP. Hence, the following batch tests simulated the reactions occurring inStage B and C combined using synthetic newberyite and synthetic crystallizer effluent. Acomparison of Annacis Island centrate and the synthetic crystallizer effluent is provided inSection 5.2. Further, Table 12 outlines the average suspension characteristics for time zero ofthese experiments. This solution is characterized to have relatively high ammonia, alkalinity andinitial pH. From the previous tests, a pH range of 7 to 8 and a temperature range of 10? to 25? Cwere selected as near optimal experimental conditions. This selection is based on theassumption that a reduction in caustic use and, therefore, a decrease in chemical costs areSecondary StruviteCrystallizerMixing/NewberyiteDissolution TankStruvite ThermalDecompositionRecoveryNewberyiteStruvite + NewberyiteStruviteNH4 + PO4StruviteMg + OHRecoveryNH4Mg + NH4 + PO4OHOHNH3Head ofBiologicalOperationsADB CFigure 21 ? Reactor Configuration 1: Ammonia recovery from primary crystallizer effluentPrimary StruviteCrystallizerDewateringCentrateCHAPTER 6: RESULTS AND DISCUSSION78preferred over the small reduction in effluent magnesium and orthophosphate residuals thatcomes with operation at a pH above 8. Also, a suspension Mg:N:P molar ratio of 1:1:1 waschosen over lower newberyite doses to target even higher ammonia removal efficiencies at theexpense of slightly higher orthophosphate residuals. The reaction time was limited to 4 hours,as previous tests demonstrated that equilibrium is nearly reached in 3 hours. This section worksto answer the following questions:? How is ammonia removal and orthophosphate residual affected by higher initial aqueousammonia?? How does a higher struvite harvest and higher initial pH and alkalinity impact causticconsumption?Table 12 ? Suspension characteristics at t = 0 h for Mg:N:P molar ratio 1:1:1 newberyitedose in synthetic crystallizer effluent batch testsReagents added as solid newberyite Initial solution characteristics Mg:N:PMolar RatioNewberyiteadded (g/L)Mg(mM)PO4-P(mM)Mg(mM)NH4-N(mM)PO4-P(mM)Alkalinity(mg/L as CaCO3)11.3 65.0 65.0 0.6 65.7 0.6 1470 1:1:16.4.1 Ammonia removalAs expected, average ammonia removal followed a familiar trend for synthetic newberyite insynthetic crystallizer effluent, as illustrated by Figure 22. Residual ammonia appears to beleveling out after a 3 hour reaction time. After 4 hours, mean ammonia removals ranged from73% to 87%. Nevertheless, no statistical difference was found between the examined scenarios.These results also compare well to that observed for the ammonia solution treated at a Mg:N:Pmolar ratio of 1:1.1:1 (see Section 6.2.1). The changes to media composition, with respect toprevious experiments, consist of the increase in ammonia from 740 to 920 mg/L NH4-N and theintroduction of alkalinity. These additional constituents do not significantly affect performancewith respect to ammonia removal.CHAPTER 6: RESULTS AND DISCUSSION79With the presence of carbonates, the model predicted the formation of magnesite, alongsidestruvite, for this range of conditions. Magnesite and bobbierite become less soluble at high pH.To check for the existence of these phases, a 24 hour batch test was performed at roomtemperature and pH 9. Neither magnesite nor bobbierite were detected during XRD analysis ofa solid sample from this test (the XRD output graph for this sample may be found in AppendixC). Therefore, entries for magnesite and bobbierite were excluded from the model phase input.Experimental and model results are compared in Figure 22. The model-generated ammoniaresiduals are considerably lower than that of the 4 hour batch tests. This discrepancy wasexpected as it may require days for these systems to reach chemical equilibrium. The reactionperiod was minimized to simulate a HRT that might be realistically achieved using continuous,side stream, unit processes. Yet, the model remains a powerful tool that can be used toestimate the completeness of newberyite-to-struvite conversion and, therefore, ammoniaremoval.Figure 22 ? Comparison of real and model-predicted NH4-N6.4.2 Orthophosphate residualFigure 23 plots average orthophosphate concentration for the duration of transformation ofnewberyite into struvite in synthetic crystallizer effluent. The solubilization of orthophosphate010020030040050060070080090010000 1 2 3 4 5NH4-N (mg/L)Time (h)pH 7; 10? CpH 7; 25? CpH 8; 10? CpH 8; 25? CModelledEquilibriumCHAPTER 6: RESULTS AND DISCUSSION80follows a similar pattern to that observed in previous experiments, with newberyite in ammoniasolution. As discussed in Section 6.2.8, the orthophosphate peak observed at the 10 minutemarks for some tests confirms the belief that a maximum magnesium and orthophosphateconcentration exists minutes after the addition of newberyite due to its initial rapid dissolution.The 4 hour orthophosphate residuals are higher in pH 7 tests at 155 and 172 mg/L PO4-P for10? and 25? C respectively. Those of tests at pH 8 were significantly lower at around 50 mg/LPO4-P due to increased supersaturation with respect to struvite. The 4 hour orthophosphateresiduals are statistically comparable to that measured under similar conditions at the end of 12hour tests with pure ammonia solution at a Mg:N:P molar ratio of 1:1.1:1. Since carbonates arenot utilized to form magnesite, it appears that the presence of alkalinity in synthetic crystallizereffluent does not dramatically affect newberyite dissolution. However, the results for theseconditions suggest that the use of newberyite as a source of magnesium and orthophosphatefor secondary struvite crystallization may result in the release of at least 40% of the centrateorthophosphate previously removed by the primary struvite crystallizer. Nevertheless, thisresidual may perhaps be minimized with a small increase to caustic dose or by increasing therecycle ratio.Figure 23 also compares measured orthophosphate residuals with outputs from the chemicalequilibrium model. The 4 hour measurements compare well with model-generated residuals forall scenarios. Evidently, the model is more accurate in estimating orthophosphate residuals thanit is in estimating ammonia removals for shorter reaction periods. This may be due to anequalizing effect by simultaneous newberyite dissolution and struvite formation. The solution isundersaturated with respect to newberyite, but supersaturated with respect to struvite for theentire reaction period. Yet, the orthophosphate residual appears to reach an equilibrium valuebefore ammonia in several cases. Newberyite dissolution kinetics may limit the formation ofstruvite. For instance, aqueous orthophosphate is removed via struvite formation rapidly up untila point near its saturation. Thereafter, the orthophosphate released is utilized at a rateessentially determined by newberyite dissolution. Hence, the orthophosphate and magnesiumconcentrations can remain constant, while the ammonia concentration continues to decrease atlow rates. The degree of phase saturation is discussed further in Section 6.4.4.CHAPTER 6: RESULTS AND DISCUSSION81Figure 23 ? Comparison of real and model-predicted PO4-P6.4.3 Chemical composition of solid phase mixturesThe solid samples collected during 4 hour batch tests with synthetic newberyite and syntheticcrystallizer effluent are assumed to contain only newberyite and struvite. Figure 24 comparesthe N:P molar ratios for solid phase samples at 1 and 4 hour reaction times. At the 1 hoursampling time, the lowest struvite yields were found with tests at pH 7 at 45% and 56%. TheN:P molar ratios are higher for suspensions at pH 8 as a result of their higher overall SS. Struvitemakes up between 79% and 85% of the solid phase after an hour at pH 8. Evidently the systemis near equilibrium after an hour at pH 8 and 25? C. Newberyite-to-struvite transformationprogresses after this time for the other cases. The 4 hour struvite yields are 75%, 78%, and91% for pH 7-25? C, pH 7-10? C, and pH 8-10? C respectively. Newberyite is more soluble whilestruvite is less soluble at lower temperatures and, therefore, struvite yields are slightly greater at10? C compared to that of 25? C for the same pH. The solid phase is predicted to be greaterthan 98% struvite at equilibrium, according to model outputs. This suggests that newberyitetheoretically continues to dissolve past 4 hours. However, it is possible that morphology plays arole in equilibrium of real solid-liquid systems. Struvite may potentially be forming as layerssurrounding a newberyite seed, removing it from contact with the liquid phase and resulting insomewhat reduced conversion efficiencies.02550751001251501752000 1 2 3 4 5PO4-P (mg/L)Time (h)pH 7; 10? CpH 7; 25? CpH 8; 10? CpH 8; 25? CModelledEquilibriumCHAPTER 6: RESULTS AND DISCUSSION82Figure 24 ? N:P molar ratio of solid phase sampled at 1 and 4 h6.4.4 Newberyite and struvite supersaturationAmmonia, magnesium and orthophosphate concentrations vary with time as newberyitedissolves and struvite forms in synthetic crystallizer effluent. Therefore, supersaturation withrespect to newberyite and struvite fluctuates accordingly. Since synthetic media was used in thisstudy, the chemical equilibrium model was used to determine SN and SS based on liquid samplecompositions, rather than constructing conditional supersaturation curves. Figures 25 and 26present supersaturation ratios with time for pH 7 and 8 respectively. At time zero, solutions areundersaturated for both newberyite and struvite due to low initial magnesium andorthophosphate concentrations. Once newberyite is added, it begins to dissolve rapidly andsaturation for both species increases with the solubilization of magnesium and orthophosphate.Due to the high initial ammonia concentration, SS is the highest shortly after newberyitedissolution initiates and ammonia is removed through struvite formation. After an hour, struvitecrystallization rates are significantly reduced and SS decreases gradually towards equilibrium,while SN remains somewhat constant. After 4 hours, newberyite has a slightly lower saturationfor 10? C than it is for 25? C due to newberyite?s higher stability at elevated temperatures. Theopposite is the case for struvite, which is more soluble at higher temperature. Therefore, SS isslightly higher for 25? C than it is for 10? C, as a result of higher magnesium, ammonia, and0.000.100.200.300.400.500.600.700.800.901.0010? C; pH 7 25? C; pH 7 10? C; pH 8 25? C; pH 8N:P Molar Ratio1 h4 hCHAPTER 6: RESULTS AND DISCUSSION83orthophosphate residuals. Following an hour, SS is comparable for both the tests at pH 7 and 8.At pH 7, SS ranges from 1.21 to 1.43 and 1.11 to 1.30 for 1 hour and 4 hour samplesrespectively. 1 hour at pH 8 corresponds to a SS between 1.36 and 1.61, while 4 hour values arebetween 1.26 and 1.39. Several studies suggested that supersaturation ratio is one of theprimary control parameters with respect to performance of the UBC struvite crystallizer, with in-reactor SS ranging from 1.0 to 2.2 (Dastur, 2001; Adnan, 2002; Britton, 2002; Huang, 2003;Fattah, 2004). Assuming pelletized struvite can be formed from a feed suspension containingboth liquid and solid phases, it may be possible to recover ammonia using the conventionalcrystallizer by including a newberyite dissolution stage and by tailoring crystallizer parameters,such as recycle ratio, to attain supersaturations that proved successful for struvite pelletization.Figure 25 ? Newberyite and struvite supersaturation ratio at pH 70.000.501.001.502.002.503.000 1 2 3 4Supersaturation RatioTime (h)Newberyite; 10? C Newberyite; 25? CStruvite; 10? C Struvite; 25? CCHAPTER 6: RESULTS AND DISCUSSION84Figure 26 ? Newberyite and struvite supersaturation ratio at pH 8It should be noted that the addition of solid phase, rather than aqueous reagents, has aninteresting effect on supersaturation with respect to struvite. If dissolved magnesium andorthophosphate were added at the same proportions to the ammonia-rich feed under similarconditions, the SS immediately following addition would be considerably higher than thatreported. Figure 27 shows the calculations for initial SS following mixing of liquid reagents.These values for liquid reagent initial SS are significantly higher than that measured after 10minutes with newberyite in synthetic crystallizer effluent. Internal crystallizer recycle flow plays arole in maintaining the low struvite supersaturation ratios that promote struvite pelletization.Assuming crystallizer operation could be optimized to accommodate a feed suspensioncontaining solid and aqueous reagents, recycle ratios are expected to be different, and perhapslower than that with conventional liquid feed.0.000.501.001.502.002.503.000 1 2 3 4Supersaturation RatioTime (h)Newberyite; 10? C Newberyite; 25? CStruvite; 10? C Struvite; 25? CCHAPTER 6: RESULTS AND DISCUSSION85Figure 27 ? Theoretical struvite supersaturation ratio immediately after mixing of liquidreagents6.4.5 Caustic consumptionThe amount of caustic consumed to achieve the previously discussed ammonia removalsprovides insight with regard to the operating costs involved with secondary struvitecrystallization in primary crystallizer effluent. Figure 28 reports the cumulative mass of causticas sodium hydroxide required to maintain a constant pH with time for newberyite in syntheticcrystallizer effluent. The caustic required for these tests equates to about 1.3 to 2.1 kg ofsodium hydroxide per cubic meter of synthetic feed treated. This range is nearly identical to thatreported by Novotny (2011,) who observed similar ammonia removals from an ammoniasolution using thermally decomposed struvite pellets. As expected, caustic consumption washigher to maintain a pH of 8 compared to that at pH 7. Further, temperature did not appear tohave a dramatic effect on the caustic required to reach nearly steady state conditions. Theseresults are generally slightly higher than under similar conditions explored with lower strength,pure-ammonia solutions. Model outputs were higher than that measured experimentally forcaustic consumed in these systems. This is, in part due to the fact that these systems are notyet at true equilibrium at this time. Further, the discrepancy between these values is the result ofdifferences between measured and model-generated initial pH of the synthetic crystallizer0.002.004.006.008.0010.0012.0014.0016.0018.00Supersaturation RatiopH 7; 10? CpH 7; 25? CpH 8; 10? CpH 8; 25? CCHAPTER 6: RESULTS AND DISCUSSION86effluent. Nevertheless, the model?s ability to provide a rough estimate of caustic needs mayprove helpful in planning future bench and pilot-scale studies on ammonia recovery.Figure 28 ? Comparison of real and model-predicted caustic consumption6.4.6 Crystal morphologyMorphology is expected to play a role in the pelletization of struvite crystals. Nevertheless, howmorphology factors in struvite agglomeration is not well documented. Images captured of solidphase samples using a microscope allow for the comparison of various experimental conditions,with respect to crystal morphology. The synthetic newberyite used in these experiments has avery distinctive crystal shape and habit compared to the struvite observed after 1 hour reactionswith synthetic crystallizer effluent, as demonstrated by Figure 29. Newberyite crystals werecharacterized by tubular, rice-like rhombohedral structures. These were present as singlecrystals, X-shaped twinned crystals, star-shaped dendrites, and what appear to be rosette-likeaggregates In acknowledgement of literature reviewed during this study, this type of morphologyhas not yet been reported for newberyite. After 1 hour under the examined conditions, thesestructures have disappeared, leaving behind struvite crystals and small, relatively round shapeswhich could be the remains of dissolving rhombohedral newberyite crystals. If these smallcrystals are newberyite, it is possible that they are preserved as seeds for struvite crystal0.00.51.01.52.02.53.00 1 2 3 4 5kg NaOH/m3 FeedTime (h)pH 7; 10? CpH 7; 25? CpH 8; 10? CpH 8; 25? CModelledEquilibriumCHAPTER 6: RESULTS AND DISCUSSION87growth; this could potentially result in reduced dissolution rates for a fraction of the newberyiteadded. This could, in part, explain the presence of residual newberyite after 12 hours underconditions where it should theoretically dissolve completely. Struvite crystals form with similarmorphologies to newberyite; however, they were generally larger. Struvite crystals werecharacterized by orthorhombic platelets and star-shaped dendrites. In several cases, struvitecrystals appear to be still growing after 1 hour, as suggested by protrusions of small crystalgrowths on the surfaces of well developed crystals; these probably grew at the beginning oftests, while newberyite was dissolving and supersaturation with respect to struvite was lower.The dendritic growth on larger crystals is likely the result of the following high supersaturationstage.Figure 29 ? x40 magnified newberyite (a); 1 hour samples from pH 7-10? C (b), pH 7-25? C(c), and pH 8-25? C (d)a bc dCHAPTER 6: RESULTS AND DISCUSSION88Figure 30 compares the solid phase morphologies at a lower magnification for 4 hour samplesfrom each test. Overall, orthorhombic platelets and dendrites were the dominant structure ofstruvite crystals. No striking difference was observed between crystals formed at pH 7 and 8 foreach temperature. However, crystals at 25? C appear to be larger than that of 10? C. Thisdifference in size may be explained by the fact that relatively larger crystals grow at lowersupersaturation ratios. Further, supersaturation decreases with an increase in temperature dueto enhanced struvite solubility.Figure 30 ? x10 magnified 4 hour samples from pH 7-10? C (a), pH 7-25? C (b), pH 8-10? C(c), and pH 8-25? C (d)The difference in struvite crystal size with temperature was also observed at highermagnifications, as illustrated by Figure 31. Struvite dendrites appear to be larger and to havea bc dCHAPTER 6: RESULTS AND DISCUSSION89broader branches at 25? compared to 10? C. These struvite morphologies may be suitable foragglomeration into pellets. The core of struvite pellets have been shown to contain primarilyagglomerated platelet and dendrite structures in previous struvite pelletization studies (Huang,2003; Fattah, 2004)Figure 31 ? x40 magnified 4 hour samples from pH 7-10? C (a), pH 7-25? C (b), pH 8-10? C(c), and pH 8-25? C (d)6.5 Transformation of newberyite into struvite in synthetic centrateAmmonia and struvite could potentially be recovered using a reactor configuration that employsa single struvite crystallizer. Hence, raw dewatering centrate is also a suitable feed for ammoniaremoval using thermally decomposed struvite. Figure 32 provides an overview of Reactora bc dCHAPTER 6: RESULTS AND DISCUSSION90Configuration 2, a potential continuous or semi-continuous system for this purpose. Thisscenario consists of three main unit processes, including a conventional struvite crystallizer(Stage E), a newberyite dosing tank (Stage B), and a struvite thermal decomposition reactor(Stage D). Stage E could potentially be a commercially available struvite crystallizer repurposedto produce struvite pellets suitable for both subsequent thermal decomposition and use as afertilizer. Assuming the centrate initially contains ammonia in excess of orthophosphate, aportion of struvite produced is recovered as a source of revenue, while the remaining portion isrecycled for ammonia removal. Similarly to Reactor Configuration 1, this may require theadoption of nonconventional HRTs to reduce undesirable residuals. Essentially, ReactorConfiguration 2 involves upgrading existing technology to include Stages B and D, which wouldallow for the recycling of a portion of the struvite produced; this would provide the magnesiumand orthophosphate required to recover the excess ammonia contained in the centrate.This section investigated the potential removal of ammonia from a solution simulatingdewatering centrate from Annacis Island WWTP. Hence, the following batch tests represent thecombined reactions occurring in Stage B and E. A comparison of Annacis Island centrate andSecondary StruviteCrystallizerMixing/NewberyiteDissolution TankStruvite ThermalDecompositionRecoveryNewberyiteStruvite + NewberyiteStruviteMg + OH RecoveryMg + NH4 + PO4OHNH3Head ofBiologicalOperationsDB EFigure 32 ? Reactor Configuration 2: Ammonia recovery from dewatering centrateDewateringCentrateNH4 + PO4CHAPTER 6: RESULTS AND DISCUSSION91the synthetic centrate is provided in Section 5.2. The synthetic centrate is distinguished fromsynthetic crystallizer effluent by containing significantly higher initial orthophosphate, as well asmagnesium chloride, which would be supplied alongside newberyite to bring the suspensionMg:N:P molar ratio to 1:1:1. Table 13 outlines the average suspension characteristics for timezero of these experiments. The conditions examined were the same as that of the experimentswith synthetic crystallizer effluent, in order to identify any advantages or drawbacks of ReactorConfiguration 2 compared to 1. This stage of the study intends to answer the followingquestions:? How is ammonia removal and orthophosphate residual affected by higher initial aqueousmagnesium and orthophosphate?? How does the presence of these aqueous reagents impact caustic consumption?Table 13 ? Suspension characteristics at t = 0 h for Mg:N:P molar ratio 1:1.05:1newberyite dose in synthetic centrate batch testsReagents added as solid newberyite Initial solution characteristics Mg:N:PMolar RatioNewberyiteadded (g/L)Mg(mM)PO4-P(mM)Mg(mM)NH4-N(mM)PO4-P(mM)Alkalinity(mg/L as CaCO3)11.7 63.5 63.5 4.8 72.0 4.7 1475 1:1.05:16.5.1 Ammonia removalAmmonia removal from synthetic centrate followed a similar trend to that of synthetic crystallizereffluent, as shown in Figure 33. Again, residual ammonia appears to be approaching anequilibrium value after 1 to 3 hours. By 4 hours, ammonia removals are within 71% to 83%.However, there was no significant difference between experimental results. Further, theseammonia removals are comparable to that observed with synthetic crystallizer effluent (seeSection 6.4). It appears that the presence of high initial aqueous orthophosphate andmagnesium (added in liquid form at the same time as newberyite) did not dramatically affect theammonia removal efficiency, compared to cases where magnesium and orthophosphate areinitially present as newberyite.CHAPTER 6: RESULTS AND DISCUSSION92Similarly to the synthetic crystallizer effluent, the chemical equilibrium model indicated thatmagnesite would form in small amounts during each of the synthetic centrate batch tests.However, patterns for magnesite and bobbierite did not appear during XRD analysis of solidphase samples and, therefore, these phases were not included with model inputs (the XRDoutput graphs for these tests may be found in Appendix C). As illustrated in Figure 33, themodel-generated ammonia residuals are considerably lower than that of the 4 hour batch tests.Again, the discrepancy between theoretical and experimental ammonia residuals is believed tobe due to the non-equilibrium state of the suspensions after 4 hours and, potentially, entrapmentof newberyite residuals by struvite.Figure 33 ? Comparison of real and model-predicted NH4-N6.5.2 Online ammonium residual monitoringThe apparatus used for batch tests limited the frequency of sample collection and, therefore,reduced the resolution of observed trends, especially for reactions occurring immediately afterthe addition of reagents. To verify that no distinct inflection points occur on plots between 10minute and 1 hour reaction times, ammonium activity was measured in ?real time? using anammonium selective electrode. Solution conductivity was logged simultaneously and measuredvalues were used to estimate ionic strength with time from a calibration curve. Another010020030040050060070080090010000 1 2 3 4 5NH4-N (mg/L)Time (h)pH 7; 10? CpH 7; 25? CpH 8; 10? CpH 8; 25? CModelledEquilibriumCHAPTER 6: RESULTS AND DISCUSSION93calibration curve was constructed correlating ionic strength with model-generated ammoniumactivity coefficient. With measured ammonium activities and attributed coefficients, ammoniaconcentration was plotted with time. Figure 34 compares probe and sample measured valuesfor two synthetic centrate batch tests. It should be noted that the probe used was sensitive tochanges in positioning and hydrodynamics, resulting from routine sample collection. Hence,abrupt changes to offsets and a general lack of trend smoothness were occasionally observedwith online data.Background ionic strength varies with the release and uptake of elemental components.Therefore, offsets may also have been caused by discrepancies between estimated and realionic strength. Nevertheless, the probe-measured data followed a similar trend to that observedin collected samples. This data confirmed that no rapid drops in ammonium were experiencedbetween 10 minute and 1 hour sampling times and that the lines fitted to sample measurementsrepresent the examined reactions remarkably well.Figure 34 ? Comparison of probe and sample measured NH4-N0100200300400500600700800900100011000 1 2 3 4NH4-N (mg/L)Time (h)pH 7; 25? C;ProbepH 8;10? C;ProbepH 7; 25? C;SamplepH 8;10? C;SampleCHAPTER 6: RESULTS AND DISCUSSION946.5.3 Orthophosphate residualMagnesium and orthophosphate molar concentrations were assumed to be similar for theduration of batch tests containing newberyite and synthetic centrate, as suggested by theabsence of magnesite or bobbierite patterns on XRD output graphs (see Appendix C). Figure 35plots average orthophosphate residuals over time for newberyite in synthetic centrate. Residualorthophosphate follows a similar trend to that of synthetic crystallizer effluent. The highest finalorthophosphate residuals were 126 and 139 mg/L PO4-P for tests at pH 7. At pH 8, 4 hourresiduals were around 40 mg/L PO4-P. Further, no statistical difference was observed betweenthese measurements and that of synthetic crystallizer effluent. This suggests that the initialpresence of high aqueous magnesium and orthophosphate does not significantly affectequilibrium residuals compared to cases where these reagents originate exclusively fromnewberyite. Further, as a simulation of a single struvite crystallizer process using newberyite asa supplemental reagent, these results suggest that a simultaneous 73% orthophosphate and83% ammonia removal could be possible.Figure 35 also provides a comparison of measured orthophosphate residuals and model-generated values. These data sets are remarkably similar, despite the expected difference inconsidered reaction time. However, magnesium and orthophosphate residuals were predicted tobe lower in synthetic centrate than crystallizer effluent. This is likely due to the synthetic centratetests? slightly higher N:P molar ratio.CHAPTER 6: RESULTS AND DISCUSSION95Figure 35 ? Comparison of real and model-predicted PO4-P6.5.4 Chemical composition of solid phase mixturesAccording to XRD analyses (see Appendix C), solid samples from synthetic centrate areexpected to contain only newberyite and struvite. Figure 36 compares the N:P molar ratiosmeasured for the solid phase at 1 and 4 hour reaction times. After 1 hour, struvite yields rangedfrom 35% to 61% and 71% to 80% for pH 7 and pH 8 respectively. The 4 hour measurementswere higher at around 80% for all cases. The model predicted the solid component of thesesuspensions to be 100% struvite at equilibrium, with the exception of pH 7-25? C (which isexpected to contain residual newberyite at 4%). Again, this suggests incomplete reactions andpotentially newberyite entrapment within struvite crystals.02550751001251501752000 1 2 3 4 5PO4-P (mg/L)Time (h)pH 7; 10? CpH 7; 25? CpH 8; 10? CpH 8; 25? CModelledEquilibriumCHAPTER 6: RESULTS AND DISCUSSION96Figure 36 ? N:P molar ratio of solid phase sampled at 1 and 4 h6.5.5 Newberyite and struvite supersaturationWith measured values for aqueous magnesium, ammonia and orthophosphate, thesupersaturation with respect to newberyite and struvite in synthetic centrate was estimatedusing the chemical equilibrium model. Figures 37 and 38 illustrate SN and SS with time for pH 7and pH 8 batch tests. Initially, newberyite and struvite are very undersaturated in the absenceof magnesium. 10 minutes after dosing the synthetic centrate with newberyite and supplementalmagnesium, SS ranges from 2.15 to 4.40. These values are generally higher than that of similartests with synthetic crystallizer effluent due to higher initial dissolved magnesium andorthophosphate. However, supersaturation ratios in synthetic crystallizer effluent and centrateare comparable after 1 hour. SS appears to level out soon after this time, approaching morenormal ratios between 1.19 and 1.47 after 4 hours. With the assumption that struvite crystals willagglomerate, these supersaturation ratios with respect to struvite are within a preferable rangefor struvite pelletization after an hours? time; thus, thereis the potential for combined struvite andammonia recovery from dewatering centrate using newberyite as an additional source ofmagnesium and orthophosphate.0.000.100.200.300.400.500.600.700.800.9010? C; pH 7 25? C; pH 7 10? C; pH 8 25? C; pH 8N:P Molar Ratio1 h4 hCHAPTER 6: RESULTS AND DISCUSSION97Figure 37 ? Newberyite and struvite supersaturation ratio at pH 7Figure 38 ? Newberyite and struvite supersaturation ratio at pH 80.00.51.01.52.02.53.03.54.04.55.00 1 2 3 4Supersaturation RatioTime (h)Newberyite; 10? C Newberyite; 25? CStruvite; 10? C Struvite; 25? C0.00.51.01.52.02.53.03.54.04.55.00 1 2 3 4Supersaturation RatioTime (h)Newberyite; 10? C Newberyite; 25? CStruvite; 10? C Struvite; 25? CCHAPTER 6: RESULTS AND DISCUSSION986.5.6 Caustic consumptionThe mass of caustic provided to promote combined ammonia and phosphorous removalsrepresents a major component of the Reactor Configuration 2 operating costs. Figure 39 plotsthe cumulative mass of sodium hydroxide consumed with time for the synthetic centrate batchtests. Around 1.6 kg was required to maintain a pH of 7, while a total mass of 2.1 kg was usedto maintain a pH 8. Interestingly, the amount of caustic consumed for simultaneousorthophosphate and ammonia removal from synthetic centrate is about the same as that used toremove only the ammonia from synthetic crystallizer effluent. This is promising from a practicalperspective, as it suggests that combined phosphorus and nitrogen recovery could potentiallybe achieved with a single crystallizer, at a caustic expense similar to that required to remove theammonia alone.Figure 39 ? Comparison of real and model-predicted caustic consumption6.5.7 Crystal morphologyExamining solid phase samples using a microscope allowed for the identification of residualnewberyite, as well as the study of variations in struvite crystal morphology across experimental0.00.51.01.52.02.53.00 1 2 3 4 5kg NaOH/m3 FeedTime (h)pH 7; 10? CpH 7; 25? CpH 8; 10? CpH 8; 25? CModelledEquilibriumCHAPTER 6: RESULTS AND DISCUSSION99conditions. Figure 40 compares raw newberyite crystals with the solid phase after an hour insynthetic centrate at pH 7. The tubular or rice-shaped newberyite crystals were identified in asample from the pH 7-10? C test. However, only struvite platelets and dendrites were observedat 25? C. The residual newberyite, at the lower temperature of 10? C, is likely caused byreduced kinetics of newberyite dissolution. Similarly to pH 7 tests with synthetic crystallizereffluent, small protrusions on struvite platelet surfaces were also observed after 1 hour at pH 7in synthetic centrate; these are believed to be young struvite crystals that grow into dendritebranches or are separated from the surface through attrition.Figure 40 ? x40 magnified synthetic newberyite batches (a and b); 1 hour samples frompH 7-10? C (c), and pH 7-25? C (d)a bc dCHAPTER 6: RESULTS AND DISCUSSION100Figures 41 and 42 show two different magnifications for images of solid phases after 4 hourreaction times. In all cases, only orthorhombic platelets and dendrites characteristic of struviteare apparent. These are similar to those grown in synthetic crystallizer effluent and resemblethat observed in the core of struvite pellets produced by a UBC struvite crystallizer (Huang,2003; Fattah, 2004). However, crystal size did not vary with temperature in the manner that itdid in the crystallizer effluent. Rather, crystal length, relative to width, increases from pH 7 to 8.This is typical of struvite, as it is known to form long, needle-shaped crystals at highersupersaturation ratios (Abbona et al., 1985).Figure 41 ? x10 magnified 4 hour samples from pH 7-10? C (a), pH 7-25? C (b), pH 8-10? C(c), and pH 8-25? C (d)a bc dCHAPTER 6: RESULTS AND DISCUSSION101Figure 42 ? x40 magnified 4 hour samples from pH 7-10? C (a), pH 7-25? C (b), pH 8-10? C(c), and pH 8-25? C (d)a bc dCHAPTER 7: CONCLUSIONS102CHAPTER 7: CONCLUSIONSBased on knowledge gained from experimental and model-generated results regardingammonia recovery from synthetic wastewaters using newberyite, the following conclusions aremade:? pH control by caustic addition is the dominant factor over temperature, with respect toammonia removal and orthophosphate solubilization resulting from simultaneousnewberyite dissolution and struvite crystallization.? All of the explored systems approached equilibrium between 1 and 3 hours reactiontime.? Maintenance of a pH of 7 and above provides ammonia removal efficiencies between77% and 87%, given a newberyite dose providing a suspension Mg:N:P molar ratio ofapproximately 1:1:1. At a pH below 7, considerable orthophosphate residuals result fromnewberyite dissolution; however, residuals may be reduced to as low as 10 mg/L PO4-Pat pH 9.? A decrease in newberyite dose generally allows for only slight reductions inorthophosphate residual.? No bobbierite or magnesite was formed during newberyite-to-struvite conversion even atpH 9 and 35? C. The solid phase mixture produced contains residual newberyiteregardless of pH and reaction time. This suggests that, unless it is pre-dissolved, someresidual could remain in struvite pellets. Nevertheless, newberyite-to-struviteconversions as high as 92% were achieved.? Conditions approaching the optimum were suggested as follows, based on ammoniaremoval, orthophosphate and newberyite residual, and caustic consumption:o pH 7 ? 8o 10? ? 25? Co Newberyite dose providing suspension Mg:N:P molar ratio of 1:1:1CHAPTER 7: CONCLUSIONS103? The proposed secondary crystallization process (Reactor Configuration 1) usingnewberyite to treat primary crystallizer effluent may result in considerable solubilizationof orthophosphate? The proposed single stage crystallization process (Reactor Configuration 2) usingnewberyite to treat centrate may provide simultaneous ammonia and orthophosphateremovals as high as 83% and 73% respectively, with a caustic consumption similar tothat of secondary crystallization? For these configurations, SS was between 1.9 and 4.4 after 10 minutes but between 1.2and 1.8 after 1 hour suggesting that a pretreated newberyite-struvite suspension mayhave potential as a feed for a UBC struvite crystallizer? The observed and model-generated liquid phase compositions were comparable.Predicted solid phase compositions were significantly different than that produced inbatch tests, but the model proved to be an excellent tool for identifying conditions wherenewberyite is still stable and for estimating caustic consumption? An unreported newberyite morphology was revealed: rice-shaped, elongatedrhombohedralsCHAPTER 8: RECOMMENDATIONS104CHAPTER 8: RECOMMENDATIONSThe following research tasks are recommended for the development of ammonia recovery viathe struvite formation-thermal decomposition cycle:? Perform pilot studies for the optimization of thermally decomposed struvite compositionand quality? Using this study?s model, explore ammonia removal and orthophosphate solubilizationfor newberyite and MgCl2 combination doses providing N:P ratios below 1 and Mg:Pratios above 1? Perform batch tests using combinations of real thermally decomposed struvite withsynthetic and real wastewaters and delineate the effect of pellet size on newberyitedissolution? Evaluate caustic consumption and struvite quality in two pilot studies using UBC struvitecrystallizers with:1. An acidic feed containing ammonia, orthophosphate, and dissolved newberyite2. A caustic-enhanced feed containing a suspension of ammonia, orthophosphate,and newberyite and struvite fines? Perform bench-scale studies to better understand the fundamentals of struvite crystalagglomeration and pelletization in FBRs for the development of next generation reactordesignsREFERENCES105REFERENCESAbbona, F., & Boistelle, R. (1985). Nucleation of struvite (MgNH4PO4.6H2O) single crystals andaggregates. Crystal Res. Technol., 20(2), 133?40.Abdelrazig, B., & Sharp, J. (1988). Phase changes on heating ammonium magnesiumphosphate hydrates. Thermochim. acta, 129, 197?215.Adnan, A. (2002). Pilot-scale study of phosphorus recovery through struvite crystallization.Aiking, H. (2011). Future protein supply. Trends Food Sci. Tech., 22, 112?20.Allar, A. D., & Beler-Baykal, B. (2013). Phosphorus recovery from source separated humanurine upon processing with clinoptilolite. Proceedings of the WEF/IWA Conference: NutrientRemoval and Recovery.Andrade, A., & Schuiling, R. D. (2001). The chemistry of struvite crystallization. Mineral Journ.,23(5), 37?46.APHA, AWWA, WEF (2012). Standard Methods for the Examination of Water and Wastewater,22nd Edition. American Public Health Association, Washington, D.C.Babi?-Ivan?i?, V., Kontrec, J., Bre?evi?, L., & Kralj, D. (2002). Precipitation Diagrams of Struviteand Dissolution Kinetics of Different Struvite Morphologies. Croat. Chem. Acta, 75(1), 89?106.Babi?-Ivan?i?, V., Kontrec, J., Bre?evi?, L., & Kralj, D. (2006). Kinetics of struvite to newberyitetransformation in the precipitation system MgCl2?NH4H2PO4?NaOH?H2O. Water Res., 40(18),3447?3455.Ba?ak?ilardan-Kabakci, S., ?peko?lu, A. N., & Talinli, I. (2007). Recovery of Ammonia fromHuman Urine by Stripping and Absorption. Environ. Eng. Sci., 24(5), 615?24.REFERENCES106Beler-Baykal, B., Allar, a D., & Bayram, S. (2011). Nitrogen recovery from source-separatedhuman urine using clinoptilolite and preliminary results of its use as fertilizer. Water Sci.Technol., 63(4), 811?7.B?n?zeth, P., Saldi, G. D., Dandurand, J.-L., & Schott, J. (2011). Experimental determination ofthe solubility product of magnesite at 50 to 200?C. Chem. Geol., 286(1-2), 21?31.Berthouex, P. Mac, & Brown, L. C. (2002). Statistics for Environmental Engineers (SecondEdition). Boca Raton, Florida: Lewis Publishers.Bhuiyan, M. I. H., Mavinic, D. S., & Beckie, R. D. (2007). A solubility and thermodynamic studyof struvite. Environ. Technol., 28(9), 1015?26.Bhuiyan, M. I. H., Mavinic, D. S., & Koch, F. A. (2008). Thermal decomposition of struvite and itsphase transition. Chemosphere, 70(8), 1347?56.Boistelle, R., Abbona, F., Madsen, H. E. L., & Massimo, S. (1983). On the transformation ofstruvite into newberyite in aqueous systems. Phys. Chem. Minerals, 9(5), 216?222.Britton, A. T. (2002). Pilot-scale struvite recovery trials from a full-scale anaerobic digestersupernatant at the City of Penticton Advanced Wastewater Treatment Plant.Britton, A., Prasad, R., Baizer, B., Cubbage, L. (2009). Pilot testing and economic evaluation ofstruvite recovery from dewatering centrate at HRSD?s Nansemond WWTP. In Proceedings ofthe International Conference on Nutrient Recovery from Wastewater Streams (pp. 193?202).Christensen, C. H., Johannessen, T., S?rensen, R. Z., & N?rskov, J. K. (2006). Towards anammonia-mediated hydrogen economy? Catal.Today, 111, 140?44.Cilona, A., Nilsson, S., Malpei, F., & Levlin, E. (2009). Gas transfer membrane for ammoniaremoval of condensed flue gas.REFERENCES107Cohen, L. H., & Ribbe, P. H. (1966). Magnesium phosphate mineral replacement at Mono Lake,California. Am. Mineral., 51, 1755?1765.Le Corre, K. S., Valsami-Jones, E., Hobbs, P., & Parsons, S. A. (2009). Phosphorus recoveryfrom wastewater by struvite crystallization: A Review. Crit. Rev. Env. Sci. Technol., 39(6), 433?77.Dastur, M. B. (2001). Investigation into the factors affecting controlled struvite crystallization atthe bench-scale.Elm?e, T. D., S?rensen, R. Z., Quaade, U., Christensen, C. H., N?rskov, J. K., & Johannessen,T. (2006). A high-density ammonia storage/delivery system based on Mg(NH3)6Cl2 for ? invehicles. Chem. Eng. Sci., 61, 2618?25.Elston, J. T., & Karmarkar, D. (2003). Aqueous ammonia stripping technology for SCRapplications. In Proceedings of the Electric Power Conference.Environment Canada. (2001). Priority Substances List Assessment Report: Ammonia in theAquatic Environment.Erisman, J. W., Sutton, M. A., Galloway, J., Klimont, Z., & Winiwarter, W. (2008). How a centuryof ammonia synthesis changed the world. Nature Geosci., 1(10), 636?9.Evans, T. D., & Thompson, A. (2009). Recovering ammonium fertiliser ? an alternative toblowing it away. In Proceedings of the 14th European Biosolids & Organic ResourcesConference.Fassbender, A. G. (2001). ThermoEnergy Ammonia Recovery Process for municipal andagricultural wastes. ScientificWorldJournal, 1(S2), 908?13.Fattah, K. P. (2004). Pilot scale struvite recovery potential from centrate at Lulu IslandWastewater Treatment Plant.REFERENCES108Franke, J., & Mersmann, A. (1995). The influence of the operational conditions on theprecipitation process. Chem. Eng. Sci., 50(11), 1737?53.Frost, R. L., Weier, M. L., & Erickson, K. L. (2004). Thermal decomposition of struvite ?Implications for the decomposition of kidney stones. J. Therm. Anal. Calorim., 76, 1025?1033.Fux, C., & Siegrist, H. (2004). Nitrogen removal from sludge digester liquids bynitrification/denitrification or partial nitritation/anammox: environmental and economicalconsiderations. Water Sci. Technol., 50(10), 19?26.He, S., Zhang, Y., Yang, M., Du, W., & Harada, H. (2007). Repeated use of MAP decompositionresidues for the removal of high ammonium concentration from landfill leachate. Chemosphere,66(11), 2233?2238.Hedstr?m, A. (2006). Reactive filter materials for ammonium and phosphorus sorption in smallscale wastewater treatment.Hejze, T., Besenhard, J. O., Kordesch, K., Cifrain, M., & Aronsson, R. R. (2008). Current statusof combined systems using alkaline fuel cells and ammonia as a hydrogen carrier. J. PowerSources, 176, 490?93.Huang, H. (2003). Pilot-scale phosphorus recovery from anaerobic digester supernatant.Huang, H. M., Xiao, X. M., & Yan, B. (2009). Recycle use of magnesium ammonium phosphateto remove ammonium nitrogen from rare-earth wastewater. Water Sci. Technol., 59(6), 1093?1099.Huang, H., Xu, C., & Zhang, W. (2011). Removal of nutrients from piggery wastewater usingstruvite precipitation and pyrogenation technology. Bioresour. Technol., 102(3), 2523?2528.Huang, H. M., Song, Q. W., & Xu, C. L. (2011). The Mechanism and influence factors of struviteprecipitation for the removal of ammonium nitrogen. Adv. Mater. Res., 189-193, 2613?20.REFERENCES109Jaffer, Y., Clark, T. a, Pearce, P., & Parsons, S. a. (2002). Potential phosphorus recovery bystruvite formation. Water Res., 36(7), 1834?42.Kessler, K. (2010). Analysis of Nitrogen Loading Reductions for Wastewater TreatmentFacilities and Non-Point Sources in the Great Bay Estuary Watershed - Appendix E : Capitaland Operation/Maintenance Costs Associated with Nitrogen Removal at 18 MunicipalWastewater Treatment Facilities Discharging to the Great Bay Estuary.Kiehl, S., & Hardt, H. (1933). The dissociation pressures of magnesium ammonium phosphatehexahydrate and some related substances. VII. J. Am. Chem. Soc., 55(2), 605?618.K?nigsberger, E., & K?nigsberger, L. (2006). Biomineralization ? Medical Aspects of Solubility.West Sussex: John Wiley & Sons, Ltd.Kontrec, J., Babi?-Ivan?i?, V., & Bre?evi?, L. (2005). Formation and morphology of struvite andnewberyite in aqueous solutions at 25 and 37 degrees C. Coll. antropol., 29(1), 289?94.Koutsoukos, P. G., Kofina, A. N., & Kanellopoulou, D. G. (2007). Solubility of salts in water: Keyissue for crystal growth and dissolution processes. Pure Appl. Chem., 79(5), 825?850.Kurtulus, G., & Tas, A. C. (2011). Transformations of neat and heated struvite(MgNH4PO4?6H2O). Mater Lett, 65(19-20), 2883?2886.Lan, R., Irvine, J. T. S., & Tao, S. (2012). Ammonia and related chemicals as potential indirecthydrogen storage materials. Int. J. Hydrogen Energ., 37, 1482?94.Liu, Y., Kumar, S., Kwag, J.-H., & Ra, C. (2013). Magnesium ammonium phosphate formation,recovery and its application as valuable resources: A review. J. Chem. Technol. Biotechnol., 88,181?89.REFERENCES110Lobanov, S., Koch, F. A., & Mavinic, D. S. (2013). The implication of aqueous equilibriummodelling to evaluate the potential for nutrient recovery from wastewater streams. InProceedings of the WEF/IWA Conference: Nutrient Removal and Recovery.Maggio, G., & Cacciola, G. (2012). When will oil, natural gas, and coal peak? Fuel, 98, 111?23.Maxwell, G. R. (2004). Synthetic Nitrogen Products: A Practical Guide to the Products andProcesses. New York: Kluwer Academic/Plenum Publishers.Mohr, S. H., & Evans, G. M. (2011). Long term forecasting of natural gas production. EnergyPolicy, 39, 5550?60.M?nch, E. V, & Barr, K. (2001). Controlled struvite crystallisation for removing phosphorus fromanaerobic digester sidestreams. Water Res., 35(1), 151?9.Novotny, C. (2011). Ammonia removal and recovery using heated struvite as an adsorbent.Ohlinger, K. N., Young, T. M., & Schroeder, E. D. (1998). Predicting struvite formation indigestion. Water Res., 32(12).Ohlinger, K. N., Young, T. M., & Schroeder, E. D. (1999). Kinetics effects on preferential struviteaccumulation in wastewater. J. Environ. Eng., 125(8), 730?737.Orentlicher, M. (2012). Overview of nitrogen removal technologies and application/use ofassociated end products. In Proceedings of: Got Manure? Enhancing Environmental andEconomic Sustainability Conference.Ostara Nutrient Recovery Technologies Inc. (2007). Ostara Nutrient Recovery TechnologiesInc.: Edmonton reveals world?s first industrial scale sewage treatment facility to recycle nutrientsinto environmentally-safe commercial fertilizer. Retrieved April 29, 2013, fromhttp://www.ostara.com/news/news-releases/2007/ostara-nutrient-recovery-technologies-inc-edmonton-reveals-worlds-first-induREFERENCES111Ostara Nutrient Recovery Technologies Inc. (2011). Utility plans expansion of phosphorusrecovery system. Retrieved May 1, 2013, from http://www.ostara.com/content/utility-plans-expansion-phosphorus-recovery-systemPaulik, F., & Paulik, J. (1975a). TG and EGA investigations of the decomposition of magnesiumammonium phosphate hexahydrate by means of the derivatograph under conventional andquasi-isothermal-quasi-isobaric conditions. J. Therm. Anal., 8, 557?566.Paulik, J., & Paulik, F. (1975b). TG and EGA investigations of the decomposition of some metalammonium phosphate monohydrates by means of the derivatograph under conventional andquasi-isothermal-quasi-isobaric conditions. J. Therm. Anal., 8, 567?576.Pokrovsky, O. S., Schott, J., & Thomas, F. (1999). Processes at the magnesium-bearingcarbonates/solution interface . I. A surface speciation model for magnesite. Geochim.Cosmochim. Acta, 63(6), 863?880.Randall, D. J., & Tsui, T. K. N. (2002). Ammonia toxicity in fish. Mar. Pollut. Bull., 45(1-12), 17?23.Reiter, A. J., & Kong, S.-C. (2011). Combustion and emissions characteristics of compression-ignition engine using dual ammonia-diesel fuel. Fuel, 90, 87?97.Ribbe, P. (1969). The decomposition of struvite: further evidence. Mineral. Mag., 37(286), 290?291.Shand, M. A. (2006). The Chemistry and Technology of Magnesia. Hoboken: John Wiley &Sons, Ltd.S?hnel, O., & Garside, J. (1992). Precipitation: Basic Principles and Industrial Applications.Oxford: Butterworth-Heinemann.REFERENCES112Stefanowicz, T., Napieralska-Zagozda, S., & Osittska, M. (1992). Ammonia removal from wastesolutions by precipitation of MgNH4PO4 II . Ammonia removal and recovery with recycling ofregenerate. Resour., Conserv. Recycling, 6, 339?345.Smil, V. (2001). Enriching the Earth: Fritz Haber, Carl Bosch and the Transformation of WorldFood Production.Sugiyama, S., Yokoyama, M., Ishizuka, H., Sotowa, K.-I., Tomida, T., & Shigemoto, N. (2005).Removal of aqueous ammonium with magnesium phosphates obtained from the ammonium-elimination of magnesium ammonium phosphate. J. Colloid Interface Sci., 292(1), 133?8.Sugiyama, S., Yokoyama, M., Fujii, M., Seyama, K., & Sotowa, K.-I. (2007). Recycling of thin-layer of magnesium hydrogen phosphate for removal and recovery of aqueous ammonium. J.Chem. Eng. Jpn., 40, 198?201.Taylor, A. W., Frazier, A. W., & Gurney, E. L. (1963a). Solubility products of magnesiumammonium and magnesium potassium phosphates. Trans. Faraday Soc., 1580?1584.Taylor, A. W., Frazier, A. W., Gurney, E. L., & Smith, J. P. (1963b). Solubility products of di- andtrimagnesium phosphates and the dissociation of magnesium phosphate solutions. Trans.Faraday Soc., 59, 1585.Tchobanoglous, G., Burton, F. L., & Stensel, H. D. (2003). Wastewater engineering: treatmentand reuse. Metcalf & Eddy. Inc. New York: McGraw-Hill.ThermoEnergy Corporation. (2007). Ammonia Recovery Process: Cost benefits to the operationof a typical wastewater treatment plant.T?rker, M., & Celen, I. (2007). Removal of ammonia as struvite from anaerobic digestereffluents and recycling of magnesium and phosphate. Bioresour. Technol., 98(8), 1529?1534.REFERENCES113Ulbricht, M., Schneider, J., Stasiak, M., & Sengupta, A. (2013). Ammonia recovery fromindustrial wastewater by TransMembraneChemiSorption. Chem. Ing. Tech., 85(8), 1259?62.USDA ERS. (2013). Fertilizer use and price. Retrieved from http://www.ers.usda.gov/data-products/fertilizer-use-and-price.aspx#26727USEPA. (2008). Municipal Nutrient Removal Technologies Reference Document: Volume 1 -Technical Report.USGS. (2013). PHREEQC (Version 3) - A Computer Program for Speciation, Batch-Reaction,One-Dimensional Transport, and Inverse Geochemical Calculations: In database file, ?llnl.dat.?Retrieved from http://wwwbrr.cr.usgs.gov/projects/GWC_coupled/phreeqc/Verbeeck, R. M. H., De Bruyne, P. A. M., Driessens, F. C. M., & Verbeek, F. (1984). Solubility ofmagnesium hydrogen phosphate trihydrate and ion-pair formation in the system Mg(OH)2-H3PO4-H2O at 25? C. Inorg. Chem., 23(13), 1922?1926.WERF. (2010). Nutrient recovery: State of the knowledge.Whitaker, A. (1968). The decomposition of struvite. Mineral. Mag., (1966), 820?824.Wood, S., & Cowie, A. (2004). A review of greenhouse gas emission factors for fertiliserproduction. IEA Bioenergy Task, 38.Yin, S. F., Xu, B. Q., Zhou, X. P., & Au, C. T. (2004). A mini-review on ammonia decompositioncatalysts for on-site generation of hydrogen for fuel cell applications. Appl. Catal., A, 277, 1?9.Zamfirescu, C., & Dincer, I. (2008). Using ammonia as a sustainable fuel. J. Power Sources,185, 459?65.REFERENCES114Zamfirescu, C., & Dincer, I. (2009). Ammonia as a green fuel and hydrogen source for vehicularapplications. Fuel Process. Technol., 90, 729?37.Zhang, S., Yao, C., Feng, X., & Yang, M. (2004). Repeated use of MgNH4PO4?6H2O residuesfor ammonia removal by acid dipping. Desalination, 170(1), 27?32.Zhang, T., Ding, L., Ren, H., & Xiong, X. (2009). Ammonium nitrogen removal from cokingwastewater by chemical precipitation recycle technology. Water Res., 43(20), 5209?5215.APPENDIX A: INSTRUMENT OPERATIONAL SETTINGS115APPENDIX A: INSTRUMENT OPERATIONAL SETTINGSTable A.1 ? Settings for magnesium analysis using flame atomic absorptionspectrophotometerParameter SettingMode AbsorbanceMeasurement Mode IntegrationFlame Type Air/C2H2Lamp Current 4.0 mAWavelength 202.6 nmCalibration Range 0-250 mg/LTable A.2 ? Settings for ammonia and orthophosphate analysis using flow injectionanalysisParameter NH4-N PO4-PMethod 4500-NH3 H1 4500-P G1TemperatureCalibration Range63? C0-50 mg/L63? C0-25 mg/L1. APHA, AWWA, WEF (2012). Standard Methods for the Examination of Water andWastewater, 22nd Edition. American Public Health Association, Washington, D.C.APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS116APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONSThe following tables report the mean of triplicate analyses of liquid and solid phase samples.Initial synthetic wastewater volumes were 500 mL and the 2 M sodium hydroxide titrantmeasurements represent cumulative consumption. Solid analyses are presented as Mg:N:Pmolar ratios with respect to orthophosphate concentration.Table B.1 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.1:1, no pH control, 10? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) Mg (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 0 0 4.72 6.11 0 01 683 677 220 223 163 163 6.52 6.57 0 03 645 663 291 299 212 215 6.33 6.39 0 06 633 626 320 324 229 229 6.20 6.25 0 09 637 632 341 338 239 239 6.17 6.22 0 012 622 631 355 349 254 241 6.15 6.20 0 0Table B.2 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.1:1, pH 7, 10? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 4.99 4.55 0 01 538 562 129 136 7.01 7.03 3.7 3.33 219 227 139 140 7.03 7.05 8.2 8.16 150 180 96 136 7.11 7.10 9.2 8.79 150 176 107 141 7.00 6.98 9.2 8.712 137 166 83 119 7.14 7.01 9.3 8.8APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS117Table B.3 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.1:1, pH 8, 10? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 5.37 5.39 0 01 196 213 26 45 8.11 8.03 9.4 8.93 139 137 33 34 8.11 8.01 10.6 10.16 129 129 31 29 8.20 8.07 10.8 10.39 122 123 22 26 8.47 8.26 10.4 10.312 120 120 23 28 8.30 8.16 10.4 10.3Table B.4 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.1:1, pH 9, 10? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 4.73 5.07 0 01 194 211 11 10 9.13 9.13 10.2 10.43 176 190 11 11 9.36 9.41 10.4 10.46 178 186 10 11 9.16 9.23 10.5 10.49 174 185 11 11 9.16 9.27 10.5 10.412 164 179 12 11 9.06 9.20 10.5 10.4Table B.5 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.1:1, no pH control, 25? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 4.88 4.55 0 01 643 682 203 212 6.62 6.66 0 03 650 668 221 213 6.50 6.47 0 06 655 657 234 229 6.45 6.38 0 09 652 663 238 234 6.44 6.35 0 012 645 657 239 236 6.42 6.35 0 0APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS118Table B.6 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.1:1, pH 7, 25? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 5.17 5.20 0 01 435 488 136 138 6.99 6.99 4.7 3.93 179 201 116 118 7.02 7.09 8.8 8.86 167 187 110 112 7.02 7.00 9.0 9.19 163 180 116 119 7.01 6.98 9.0 9.112 159 179 117 122 7.01 7.08 9.0 9.2Table B.7 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.1:1, pH 8, 25? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 5.49 5.03 0 01 148 148 39 30 8.06 8.13 10.3 10.63 144 146 31 39 8.15 7.95 10.3 10.66 143 137 27 40 8.32 8.35 10.6 10.99 135 134 30 30 8.20 8.23 10.7 10.912 135 136 29 32 8.23 8.17 10.7 10.9Table B.8 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.1:1, pH 9, 25? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) Mg (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 0 0 5.07 4.56 0 01 166 169 14 12 6 5 9.28 9.81 11.4 12.23 155 158 14 12 7 5 9.31 9.78 11.4 12.26 145 145 14 14 7 6 9.27 9.69 11.4 12.29 136 135 15 13 8 6 9.22 9.65 11.4 12.212 125 123 16 14 8 7 9.18 9.60 11.4 12.2APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS119Table B.9 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.1:1, no pH control, 35? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 5.06 4.92 0 01 672 675 173 172 6.67 6.67 0 03 660 662 190 188 6.54 6.52 0 06 663 660 196 196 6.50 6.51 0 09 653 660 197 199 6.51 6.53 0 012 652 658 198 199 6.49 6.52 0 0Table B.10 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.1:1, pH 7, 35? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 4.99 4.87 0 01 637 637 141 142 7.02 7.09 1.1 1.23 288 300 119 122 7.10 7.12 6.6 6.66 238 252 126 128 6.96 6.98 7.2 7.29 228 241 130 133 6.97 6.98 7.3 7.312 212 222 116 114 7.07 7.10 7.5 7.5Table B.11 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.1:1, pH 8, 35? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 4.65 4.73 0 01 176 193 30 22 8.11 8.15 10.2 10.13 170 174 37 30 8.03 8.22 10.2 10.16 163 170 41 35 7.98 8.10 10.2 10.19 158 161 44 41 7.92 7.98 10.3 10.112 147 154 30 35 8.24 8.08 10.4 10.2APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS120Table B.12 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.1:1, pH 9, 35? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 4.85 4.97 0 01 179 176 16 13 9.19 9.62 11.7 12.43 160 154 16 14 9.15 9.54 11.7 12.46 139 137 17 16 9.04 9.47 11.7 12.49 124 123 20 17 8.92 9.40 11.7 12.412 109 108 21 18 8.95 9.30 11.8 12.4Table B.13 ? Solid sample analyses for Mg:N:P molar ratio 1:1.1:1, Reactor 2Temperature pH SampleTime (h)Solid Molar RatioMg: NH4 :PO410? CNocontrol1 0.976 0.085 112 0.975 0.223 17 1 0.942 0.336 13 0.971 0.827 18 1 0.942 0.779 13 0.942 0.863 125? C7 1 0.953 0.423 13 0.946 0.832 18 1 0.939 0.917 13 0.939 0.864 19 1 0.954 0.770 112 0.986 0.856 135? C 7 12 0.991 0.770 19 12 0.990 0.911 1Table B.14 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.4:1, no pH control, 10? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 5.30 5.26 0 01 685 692 192 197 6.52 6.48 0 03 640 650 252 255 6.33 6.29 0 06 618 643 276 285 6.26 6.21 0 09 635 635 296 296 6.19 6.15 0 012 630 635 307 307 6.14 6.12 0 0APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS121Table B.15 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.4:1, pH 7, 10? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 4.76 4.89 0 01 538 562 94 98 7.13 7.19 2.4 2.83 300 290 83 82 7.10 7.11 6.4 6.86 251 244 73 74 7.13 7.02 7.1 7.49 254 243 82 81 6.86 6.87 7.1 7.412 244 233 63 58 7.10 7.16 7.3 7.6Table B.16 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.4:1, pH 8, 10? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 4.52 4.68 0 01 312 330 22 30 8.15 8.12 7.0 6.73 244 240 20 22 7.98 7.99 7.9 8.06 246 237 22 18 7.99 8.36 7.9 8.19 240 231 21 19 7.91 8.16 7.9 8.112 241 237 22 20 7.97 8.14 8.0 8.1Table B.17 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.4:1, pH 9, 10? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 4.68 4.96 0 01 303 279 7 9 9.21 9.14 8.8 8.43 282 264 8 9 9.26 9.10 8.9 8.46 275 264 7 9 9.25 9.11 8.9 8.59 278 259 8 9 9.22 9.03 8.9 8.512 268 256 8 9 9.20 9.04 8.9 8.5APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS122Table B.18 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.4:1, no pH control, 25? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 4.92 4.98 0 01 673 678 189 187 6.66 6.71 0 03 640 642 214 210 6.47 6.51 0 06 652 653 235 234 6.40 6.42 0 09 645 645 236 235 6.36 6.38 0 012 643 653 241 241 6.34 6.36 0 0Table B.19 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.4:1, pH 7, 25? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 5.23 5.32 0 01 630 635 147 139 6.99 7.07 1.8 2.13 320 303 102 98 6.96 7.00 6.0 6.56 301 291 92 94 7.08 6.99 6.4 6.59 298 294 98 103 6.94 6.92 6.4 6.512 295 290 95 94 6.98 6.98 6.4 6.6Table B.20 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.4:1, pH 8, 25? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 4.94 4.86 0 01 287 286 23 20 8.00 8.15 7.9 8.23 282 275 23 20 8.09 8.16 8.1 8.26 277 278 17 21 8.28 8.10 8.2 8.29 277 273 17 24 8.26 8.05 8.2 8.212 277 276 18 24 8.25 8.03 8.2 8.2APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS123Table B.21 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.4:1, pH 9, 25? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 5.70 4.76 0 01 316 303 10 11 9.12 9.02 9.6 9.43 300 293 10 11 9.12 8.99 9.6 9.46 293 284 9 12 9.13 9.01 9.8 9.69 285 276 10 12 9.17 8.99 9.9 9.612 277 269 11 12 9.11 8.97 9.9 9.6Table B.22 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.4:1, no pH control, 35? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 5.00 4.90 0 01 670 673 166 166 6.74 6.76 0 03 658 650 183 182 6.55 6.52 0 06 652 642 192 185 6.53 6.49 0 09 643 640 192 191 6.49 6.48 0 012 652 650 196 197 6.48 6.47 0 0Table B.23 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.4:1, pH 7, 35? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 5.03 4.39 0 01 605 647 26 45 7.05 7.14 1.4 1.03 304 318 33 34 7.05 7.05 6.2 5.96 287 295 31 29 7.09 7.04 6.4 6.39 292 300 22 26 7.08 7.01 6.4 6.312 291 296 23 28 7.15 7.08 6.4 6.3APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS124Table B.24 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.4:1, pH 8, 35? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 5.01 4.93 0 01 253 259 11 10 8.09 8.28 8.6 8.93 243 245 11 11 8.08 8.30 8.7 8.96 234 248 11 11 8.15 8.22 8.9 8.99 233 250 11 11 8.09 8.20 9.0 8.912 232 245 12 11 8.06 8.17 9.0 8.9Table B.25 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.4:1, pH 9, 35? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) Mg (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R2 R1 R20 737 737 0 0 0 0 5.02 4.93 0 01 307 276 11 13 7 9 9.09 8.80 10.4 10.13 277 260 11 12 8 9 9.08 9.06 10.5 10.66 260 244 12 14 8 9 9.07 9.01 10.6 10.69 255 253 13 13 9 9 9.04 8.99 10.6 10.612 231 238 13 14 9 9 8.99 8.99 10.6 10.6Table B.26 ? Solid sample analyses for Mg:N:P molar ratio 1:1.4:1, Reactor 2Temperature pH SampleTime (h)Solid Molar RatioMg: NH4 :PO410? C 9 12 0.975 0.774 135? C 9 12 0.963 0.841 1APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS125Table B.27 ? Liquid sample analyses for Mg:N:P molar ratio 1:1:1 in synthetic crystallizereffluent, pH 7, 10? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) Mg (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R2 R1 R20 919 919 19 19 16 16 8.35 8.37 0 00.17 382 395 96 103 71 77 7.28 7.12 0 01 232 215 118 132 88 98 7.00 7.00 4.5 4.83 133 121 136 137 101 102 7.00 7.00 8.0 8.04 122 110 153 157 115 117 6.98 6.98 8.1 8.0Table B.28 ? Liquid sample analyses for Mg:N:P molar ratio 1:1:1 in synthetic crystallizereffluent, pH 8, 10? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) Mg (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R2 R1 R20 919 919 19 19 16 16 8.35 8.41 0 00.17 783 782 62 67 47 52 8.00 8.00 0 01 217 205 46 53 31 36 8.00 8.00 11.5 11.13 132 116 49 53 32 36 7.99 7.97 12.7 12.34 124 109 49 57 33 38 7.99 8.02 12.7 12.4Table B.29 ? Liquid sample analyses for Mg:N:P molar ratio 1:1:1 in synthetic crystallizereffluent, pH 7, 25? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) Mg (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R2 R1 R20 919 919 19 19 16 16 8.09 8.05 0 00.17 799 818 191 161 137 119 7.03 7.45 0 01 423 381 170 160 122 116 7.00 7.00 5.9 7.33 279 247 184 164 130 118 6.98 7.04 7.3 8.64 263 230 181 164 129 117 7.01 7.03 7.9 8.6APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS126Table B.30 ? Liquid sample analyses for Mg:N:P molar ratio 1:1:1 in synthetic crystallizereffluent, pH 8, 25? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) Mg (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R2 R1 R20 919 919 19 19 16 16 8.07 8.09 0 00.17 356 299 85 82 62 59 8.00 8.00 9.0 10.11 150 146 54 52 37 36 8.04 8.07 12.5 12.83 141 137 46 51 32 35 8.11 8.04 12.6 12.84 138 136 46 50 32 36 8.12 8.03 12.6 12.8Table B.31 ? Solid sample analyses for Mg:N:P molar ratio 1:1:1 in synthetic crystallizereffluent, Reactor 1Temperature pH SampleTime (h)Solid Molar RatioMg: NH4 :PO410? C7 1 1.012 0.455 14 1.004 0.778 18 1 1.008 0.794 14 1.024 0.915 125? C7 1 1.003 0.560 14 1.005 0.754 18 1 1.020 0.847 14 0.994 0.846 1Table B.32 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.05:1 in synthetic centrate,pH 7, 10? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) Mg (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R2 R1 R20 1007 1007 147 147 0 0 7.89 7.85 0 00.17 946 963 143 175 92 123 7.00 7.00 0 01 591 685 132 152 89 106 7.00 6.99 3.0 4.23 304 364 125 101 93 75 7.00 7.03 9.2 9.24 271 308 142 109 89 81 7.00 7.04 9.3 9.3APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS127Table B.33 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.05:1 in synthetic centrate,pH 8, 10? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) Mg (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R2 R1 R20 1008 1008 147 147 0 0 7.56 7.51 0 00.17 918 941 106 93 81 85 8.00 8.00 0 01 277 298 47 32 34 21 8.09 8.11 12.8 13.23 169 194 40 33 27 19 8.06 8.22 12.9 13.24 164 185 46 36 32 23 7.91 8.12 12.9 13.2Table B.34 ? NH4 probe readings for Mg:N:P molar ratio 1:1.05:1 in synthetic centrate, pH8, 10? CTime(h)ConductivityReadingIonicStrength(mol/L)ActivityCoefficient,?Probe{NH4-N}(mg/L[NH4-N](mg/L)0.00 18.9 0.0979 0.757 732 9670.08 18.5 0.0958 0.758 751 9900.17 18.2 0.0942 0.760 684 9000.25 18.7 0.0969 0.758 740 9770.33 18.5 0.0958 0.758 664 8750.42 18.2 0.0942 0.760 589 7750.50 17.9 0.0927 0.761 522 6860.58 17.5 0.0906 0.763 492 6450.67 17.2 0.0891 0.765 436 5700.75 17.1 0.0886 0.765 378 4940.83 16.8 0.0870 0.767 326 4250.92 16.6 0.0860 0.768 303 3951.00 16.5 0.0855 0.768 281 3663.00 16.2 0.0840 0.770 171 2224.00 16.2 0.0840 0.770 172 224APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS128Table B.35 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.05:1 in synthetic centrate,pH 7, 25? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) Mg (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R2 R1 R20 1007 1007 147 147 0 0 7.58 7.57 0 00.17 929 940 197 194 128 134 7.00 7.00 1.5 1.81 439 430 133 138 118 98 7.02 7.02 7.5 8.63 271 268 143 143 87 102 6.99 7.12 10.1 10.24 261 254 141 137 85 97 7.00 7.12 10.1 10.2Table B.36 ? NH4 probe readings for Mg:N:P molar ratio 1:1.05:1 in synthetic centrate, pH7, 25? CTime(h)ConductivityReadingIonicStrength(mol/L)ActivityCoefficient,?Probe{NH4-N}(mg/L[NH4-N](mg/L)0.00 18.4 0.1015 0.750 844 11260.08 18.3 0.1009 0.750 838 11170.17 17.8 0.0980 0.753 828 11000.25 17.8 0.0980 0.753 818 10870.33 17.7 0.0974 0.753 745 9890.42 17.4 0.0957 0.755 708 9380.50 17.2 0.0945 0.756 668 8840.58 17.0 0.0934 0.757 613 8100.67 16.8 0.0922 0.758 562 7420.75 16.5 0.0905 0.759 514 6770.83 16.4 0.0900 0.760 471 6200.92 16.2 0.0888 0.761 424 5571.00 16.2 0.0888 0.761 407 5351.50 15.7 0.0861 0.764 305 3992.00 15.3 0.0839 0.766 250 3262.50 15.3 0.0839 0.766 212 2773.00 15.3 0.0839 0.766 205 2683.50 15.1 0.0828 0.767 190 2483.75 14.9 0.0817 0.768 186 2425.00 14.5 0.0795 0.770 170 221APPENDIX B: LIQUID AND SOLID SAMPLE COMPOSITIONS129Table B.37 ? Liquid sample analyses for Mg:N:P molar ratio 1:1.05:1 in synthetic centrate,pH 8, 25? CSampleTime (h)NH4-N (mg/L) PO4-P (mg/L) Mg (mg/L) pH NaOH (mL)R1 R2 R1 R2 R1 R2 R1 R2 R1 R20 1008 1008 147 147 0 0 7.51 7.38 0 00.17 812 825 65 51 65 55 8.00 8.00 7.7 5.61 210 207 39 32 25 21 8.05 8.12 12.8 13.23 198 189 38 39 25 26 8.07 8.03 12.9 13.24 194 188 38 40 25 26 8.07 8.03 12.9 13.2Table B.38 ? Solid sample analyses for Mg:N:P molar ratio 1:1.05:1 in synthetic centrateTemperature pH ReactorSampleTime(h)Solid Molar RatioMg: NH4 :PO410? C71 1 0.929 0.405 14 0.960 0.813 12 1 0.964 0.289 14 0.948 0.753 18 2 1 0.952 0.706 14 0.948 0.818 125? C71 1 0.956 0.642 14 0.945 0.808 12 1 0.954 0.581 14 0.950 0.792 18 2 1 0.949 0.798 14 0.967 0.815 1APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES130APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLESThe following figures are output graphs from XRD analyses of select solid samples. Sampleswere screened for patterns representing various magnesium salts of interest. However, onlynewberyite and struvite were detected. Note that the graph legend varies for each figure.APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES131Figure C.1? Synthetic newberyite prepared September 12th, 2012Newberyite Sept1200 -0 35 -07 80  (*) -  Ne wb eryi te, syn  -  Mg HP O4 ?3 H2O  -  Y: 50 .0 0 %  - d  x b y: 1. - W L : 1.5 40 6 - Ortho rho mbic - a 10 .2 08 30  -  b  1 0.68 4 50  - c 1 0 .0 12 90  -  al ph a 9 0.0 00  - b eta  9 0.00 0 - ga mma 90 .0 00  -  P rim i tiveOp eration s: Imp ortNewbe ryite S ep t1 2 - F ile: N ewb Sep t12 .ra w - Type : 2Th /Th locked  -  S ta rt: 5.00 0 ? -  E nd : 75 .00 1 ? - S tep : 0 .0 19  ? - Step time: 3 6.2  s - T emp.: 2 5 ?C (Room)  -  T i me S tarted: 22  s - 2 -Th eta: 5 .0 00  ? - The ta: 2Lin (Cps)01002003004005002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES132Figure C.2 ? Synthetic newberyite prepared November 12th, 2012Newberyite Nov1200 -0 35 -07 80  (*) -  Ne wb eryi te, syn  -  Mg HP O4 ?3 H2O  -  Y: 50 .0 0 %  - d  x b y: 1. - W L : 1.5 40 6 - Ortho rho mbic - a 10 .2 08 30  -  b  1 0.68 4 50  - c 1 0 .0 12 90  -  al ph a 9 0.0 00  - b eta  9 0.00 0 - ga mma 90 .0 00  -  P rim i tiveOp eration s: Imp ortNewbe ryite N ov12  - F ile : Newb  Nov1 2.raw - Type: 2Th /T h locked  - S tar t: 5 .00 0 ? - En d : 75 .0 01  ? - Step : 0.01 9  ? - S tep ti me: 3 6.2 s -  Te mp.: 2 5 ?C ( Room) - T ime  S ta rte d: 2 1 s -  2 -Th eta: 5.0 00  ? - Theta: 2 .Lin (Cps)01002003004005006002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES133Figure C.3 ? XRD output for Mg:N:P molar ratio 1:1.1:1, no pH control, 10? C, 12 hN-S 10C pH 6 12h R200 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  P00 -0 35 -07 80  (*) -  Ne wb eryi te, syn  -  Mg HP O4 ?3 H2O  -  Y: 50 .0 0 %  - d  x b y: 1. - W L : 1.5 40 6 - Ortho rho mbic - a 10 .2 08 30  -  b  1 0.68 4 50  - c 1 0 .0 12 90  -  al ph a 9 0.0 00  - b eta  9 0.00 0 - ga mma 90 .0 00  -  P rim i tiveOp eration s: Imp ortN-S  1 0C p H 6  1 2h  R2 - Fi le: N -S 10 C pH6  1 2h  R2.raw - Type: 2Th /Th  locked  - Start: 5 .0 00  ? - En d: 75 .0 07  ? - Step : 0.01 9 ? - S te p ti me: 3 8.4 s -  Te mp.: 25  ?C (R oom) - T ime  S tar te d: 2 6 s -  2 -Th eta : 5.00 0Lin (Cps)01002003004005006007008009001000110012001300140015001600170018001900200021002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES134Figure C.4 ? XRD output for Mg:N:P molar ratio 1:1.1:1, pH 8, 10? C, 12 hN-S 10C pH8 12h R101 -0 70 -23 45  (C) - Newber yite , syn - MgHP O4 (H2 O)3  - Y : 50 .0 0 %  -  d  x b y: 1 . - W L : 1 .5 40 6 - Orth orho mbi c - a 10 .2 03 00  - b  1 0 .6 78 00  -  c 10 .0 15 00  -  al ph a 90 .0 00  -  b eta 9 0.00 0  - ga mma 90 .0 00  -  P rim i ti v00 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  POp eration s: Imp ortN-S  1 0C p H8 12 h R1  - F ile : N-S  1 0C p H8 12 h R1 .r aw - Typ e: 2T h/Th locked  - S ta rt: 5.00 0 ? -  E nd : 7 5.00 7 ? -  S tep : 0 .0 19  ? - Step  time: 38 .4  s - Temp .: 2 5 ?C (Room ) -  T i me Started: 24  s - 2-Th eta: 5 .0 00Lin (Cps)01002003004005006007008009001000110012001300140015002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES135Figure C.5 ? XRD output for Mg:N:P molar ratio 1:1.1:1, pH 9, 10? C, 12 hN-S 10C pH9 12h R101 -0 70 -23 45  (C) - Newber yite , syn - MgHP O4 (H2 O)3  - Y : 50 .0 0 %  -  d  x b y: 1 . - W L : 1 .5 40 6 - Orth orho mbi c - a 10 .2 03 00  - b  1 0 .6 78 00  -  c 10 .0 15 00  -  al ph a 90 .0 00  -  b eta 9 0.00 0  - ga mma 90 .0 00  -  P rim i ti v00 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  POp eration s: Imp ortN-S  1 0C p H9 12 h R1  - F ile : N-S  1 0C p H9 12 h R1 .r aw - Typ e: 2T h/Th locked  - S ta rt: 5.00 0 ? -  E nd : 7 5.00 7 ? -  S tep : 0 .0 19  ? - Step  time: 38 .4  s - Temp .: 2 5 ?C (Room ) -  T i me Started: 25  s - 2-Th eta: 5 .0 00Lin (Cps)010020030040050060070080090010002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES136Figure C.6 ? XRD output for Mg:N:P molar ratio 1:1.1:1, no pH control, 25? C, 12 hN-S 25C pH6 12h R100 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  P01 -0 70 -23 45  (C) - Newber yite , syn - MgHP O4 (H2 O)3  - Y : 50 .0 0 %  -  d  x b y: 1 . - W L : 1 .5 40 6 - Orth orho mbi c - a 10 .2 03 00  - b  1 0 .6 78 00  -  c 10 .0 15 00  -  al ph a 90 .0 00  -  b eta 9 0.00 0  - ga mma 90 .0 00  -  P rim i ti vOp eration s: Imp ortN-S  2 5C p H6 12 h R1  - F ile : N-S  2 5C p H6 12 h R1 .r aw - Typ e: 2T h/Th locked  - S ta rt: 5.00 0 ? -  E nd : 7 5.00 7 ? -  S tep : 0 .0 19  ? - Step  time: 38 .4  s - Temp .: 2 5 ?C (Room ) -  T i me Started: 23  s - 2-Th eta: 5 .0 00Lin (Cps)01002003004005002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES137Figure C.7 ? XRD output for Mg:N:P molar ratio 1:1.1:1, pH 9, 25? C, 1 hN-S 25C pH9 1h R101 -0 70 -23 45  (C) - Newber yite , syn - MgHP O4 (H2 O)3  - Y : 50 .0 0 %  -  d  x b y: 1 . - W L : 1 .5 40 6 - Orth orho mbi c - a 10 .2 03 00  - b  1 0 .6 78 00  -  c 10 .0 15 00  -  al ph a 90 .0 00  -  b eta 9 0.00 0  - ga mma 90 .0 00  -  P rim i ti v00 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  POp eration s: Imp ortN-S  2 5C p H9 1h  R 1 - Fi le: N- S 25 C pH9  1 h R1 .ra w - Type : 2T h/Th locked  -  S ta rt: 5.00 0 ? -  E nd : 7 5.00 7 ? -  S tep : 0 .0 19  ? - Step  time: 3 8 .4  s - Temp .: 2 5 ?C (Room ) -  T i me Started: 23  s - 2-Th eta: 5 .0 00  ? -Lin (Cps)01002003004005006007008009001000110012001300140015002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES138Figure C.8 ? XRD output for Mg:N:P molar ratio 1:1.1:1, pH 9, 25? C, 12 hN-S 25C pH9 12h R101 -0 70 -23 45  (C) - Newber yite , syn - MgHP O4 (H2 O)3  - Y : 50 .0 0 %  -  d  x b y: 1 . - W L : 1 .5 40 6 - Orth orho mbi c - a 10 .2 03 00  - b  1 0 .6 78 00  -  c 10 .0 15 00  -  al ph a 90 .0 00  -  b eta 9 0.00 0  - ga mma 90 .0 00  -  P rim i ti v00 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  POp eration s: Imp ortN-S  2 5C p H9 12 h R1  - F ile : N-S  2 5C p H9 12 h R1 .r aw - Typ e: 2T h/Th locked  - S ta rt: 5.00 0 ? -  E nd : 7 5.00 7 ? -  S tep : 0 .0 19  ? - Step  time: 38 .4  s - Temp .: 2 5 ?C (Room ) -  T i me Started: 23  s - 2-Th eta: 5 .0 00Lin (Cps)01002003004005006007008009001000110012001300140015002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES139Figure C.9 ? XRD output for Mg:N:P molar ratio 1:1.1:1, pH 8, 35? C, 12 hN-S 35C pH8 12h R100 -0 35 -07 80  (*) -  Ne wb eryi te, syn  -  Mg HP O4 ?3 H2O  -  Y: 50 .0 0 %  - d  x b y: 1. - W L : 1.5 40 6 - Ortho rho mbic - a 10 .2 08 30  -  b  1 0.68 4 50  - c 1 0 .0 12 90  -  al ph a 9 0.0 00  - b eta  9 0.00 0 - ga mma 90 .0 00  -  P rim i tive00 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  POp eration s: Imp ortN-S  3 5C p H8 12 h R1  - F ile : N-S  3 5C p H8 12 h R1 .r aw - Typ e: 2T h/Th locked  - S ta rt: 5.00 0 ? -  E nd : 7 5.00 7 ? -  S tep : 0 .0 19  ? - Step  time: 38 .4  s - Temp .: 2 5 ?C (Room ) -  T i me Started: 24  s - 2-Th eta: 5 .0 00Lin (Cps)0100200300400500600700800900100011001200130014001500160017002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES140Figure C.10 ? XRD output for Mg:N:P molar ratio 1:1.1:1, pH 9, 35? C, 12 hN-S 35C pH9 12h R101 -0 70 -23 45  (C) - Newber yite , syn - MgHP O4 (H2 O)3  - Y : 50 .0 0 %  -  d  x b y: 1 . - W L : 1 .5 40 6 - Orth orho mbi c - a 10 .2 03 00  - b  1 0 .6 78 00  -  c 10 .0 15 00  -  al ph a 90 .0 00  -  b eta 9 0.00 0  - ga mma 90 .0 00  -  P rim i ti v00 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  POp eration s: Imp ortN-S  3 5C p H9 12 h R1  - F ile : N-S  3 5C p H9 12 h R1 .r aw - Typ e: 2T h/Th locked  - S ta rt: 5.00 0 ? -  E nd : 7 5.00 7 ? -  S tep : 0 .0 19  ? - Step  time: 38 .4  s - Temp .: 2 5 ?C (Room ) -  T i me Started: 24  s - 2-Th eta: 5 .0 00Lin (Cps)010020030040050060070080090010002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES141Figure C.11 ? XRD output for Mg:N:P molar ratio 1:1.4:1, pH 9, 10? C, 12 hN-S 10C pH9 12h 0.8 R101 -0 70 -23 45  (C) - Newber yite , syn - MgHP O4 (H2 O)3  - Y : 50 .0 0 %  -  d  x b y: 1 . - W L : 1 .5 40 6 - Orth orho mbi c - a 10 .2 03 00  - b  1 0 .6 78 00  -  c 10 .0 15 00  -  al ph a 90 .0 00  -  b eta 9 0.00 0  - ga mma 90 .0 00  -  P rim i ti v00 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  POp eration s: Imp ortN-S  1 0C p H9 12 h 0.8  R1  -  F il e: N-S  10 C p H9 1 2h  0 .8  R 1.raw - T yp e: 2 Th /Th  l ocked - Start: 5 .0 00  ? - En d: 7 5 .0 07  ? - Step: 0.01 9 ? -  S te p tim e: 3 8.4 s - Tem p.: 25  ?C  (R oom) - T ime S tar ted : 2 3 s - 2 -The ta:Lin (Cps)01002003004005006007008009002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES142Figure C.12 ? XRD output for Mg:N:P molar ratio 1:1.4:1, pH 7, 25? C, 12 hN-S 25C pH7 12h 08 R101 -0 70 -23 45  (C) - Newber yite , syn - MgHP O4 (H2 O)3  - Y : 50 .0 0 %  -  d  x b y: 1 . - W L : 1 .5 40 6 - Orth orho mbi c - a 10 .2 03 00  - b  1 0 .6 78 00  -  c 10 .0 15 00  -  al ph a 90 .0 00  -  b eta 9 0.00 0  - ga mma 90 .0 00  -  P rim i ti v00 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Prim itive -  POp eration s: Imp ortN-S  2 5C p H7 12 h 08  R1 - F ile: N -S 25 C pH7  1 2 h 08  R1 .raw -  Typ e: 2 Th/Th  l ocke d - Start: 5.0 00  ? - En d: 7 5.00 7 ? - S tep: 0 .01 9 ? - S tep  time : 38 .4  s - Temp .: 25  ?C (Ro om) - T ime Started : 23  s - 2- Theta:Lin (Cps)01002003004005006007008009001000110012001300140015001600170018001900200021002200230024002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES143Figure C.13 ? XRD output for Mg:N:P molar ratio 1:1.4:1, pH 7, 35? C, 12 hN-S 35C pH7 12h 08 R100 -0 35 -07 80  (*) -  Ne wb eryi te, syn  -  Mg HP O4 ?3 H2O  -  Y: 50 .0 0 %  - d  x b y: 1. - W L : 1.5 40 6 - Ortho rho mbic - a 10 .2 08 30  -  b  1 0.68 4 50  - c 1 0 .0 12 90  -  al ph a 9 0.0 00  - b eta  9 0.00 0 - ga mma 90 .0 00  -  P rim i tive00 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  POp eration s: Imp ortN-S  3 5C p H7 12 h 08  R1 - F ile: N -S 35 C pH7  1 2 h 08  R1 .raw -  Typ e: 2 Th/Th  l ocke d - Start: 5.0 00  ? - En d: 7 5.00 7 ? - S tep: 0 .01 9 ? - S tep  time : 38 .4  s - Temp .: 25  ?C (Ro om) - T ime Started : 23  s - 2- Theta:Lin (Cps)01000200030004000500060002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES144Figure C.14 ? XRD output for Mg:N:P molar ratio 1:1.4:1, pH 9, 35? C, 12 hN-S 35C pH9 12h 08 R101 -0 70 -23 45  (C) - Newber yite , syn - MgHP O4 (H2 O)3  - Y : 50 .0 0 %  -  d  x b y: 1 . - W L : 1 .5 40 6 - Orth orho mbi c - a 10 .2 03 00  - b  1 0 .6 78 00  -  c 10 .0 15 00  -  al ph a 90 .0 00  -  b eta 9 0.00 0  - ga mma 90 .0 00  -  P rim i ti v00 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  POp eration s: Imp ortN-S  3 5C p H9 12 h 08  R1 - F ile: N -S 35 C pH9  1 2 h 08  R1 .raw -  Typ e: 2 Th/Th  l ocke d - Start: 5.0 00  ? - En d: 7 5.00 7 ? - S tep: 0 .01 9 ? - S tep  time : 38 .4  s - Temp .: 25  ?C (Ro om) - T ime Started : 24  s - 2- Theta:Lin (Cps)01002003004005006007008009001000110012001300140015001600170018001900200021002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES145Figure C.15 ? XRD output for Mg:N:P molar ratio 1:1.1:1, no pH control, 10? C, 20 minsN-S 10C pH6 20 min R200 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  P00 -0 35 -07 80  (*) -  Ne wb eryi te, syn  -  Mg HP O4 ?3 H2O  -  Y: 50 .0 0 %  - d  x b y: 1. - W L : 1.5 40 6 - Ortho rho mbic - a 10 .2 08 30  -  b  1 0.68 4 50  - c 1 0 .0 12 90  -  al ph a 9 0.0 00  - b eta  9 0.00 0 - ga mma 90 .0 00  -  P rim i tiveOp eration s: Imp ortN-S  1 0C p H6 20  m in R2  - F ile : N-S  1 0C p H6 20 min  R 2.raw -  T yp e: 2 Th/Th  l ocked - Start: 5 .0 00  ? - En d: 7 5.0 07  ? - Step: 0.01 9 ? -  S tep  tim e: 38 .4 s - Temp .: 25  ?C (Ro om) - T ime Started : 2 5 s - 2- The ta: 5Lin (Cps)01002003004005006007002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES146Figure C.16 ? XRD output for Mg:N:P molar ratio 1:1:1 in synthetic crystallizer effluent, pH 9, 18? C, 24 hN-S 18C 24h pH9 synth centrate00 -0 35 -07 80  (*) -  Ne wb eryi te, syn  -  Mg HP O4 ?3 H2O  -  Y: 50 .0 0 %  - d  x b y: 1. - W L : 1.5 40 6 - Ortho rho mbic - a 10 .2 08 30  -  b  1 0.68 4 50  - c 1 0 .0 12 90  -  al ph a 9 0.0 00  - b eta  9 0.00 0 - ga mma 90 .0 00  -  P rim i tive01 -0 77 -23 03  (C) - Struvi te - MgN H4P O4 (H2 O)6  -  Y: 50 .0 0 %  -  d  x b y: 1 . - W L : 1 .5 40 6 - Ortho rho mbi c - a 6.95 50 0 - b 6.1 42 00  - c 11 .2 18 00  -  al ph a 9 0 .0 00  -  b eta  9 0.00 0 - ga mma 90 .0 00  -  P rim i ti ve  - P mnOp eration s: Imp ortN-S  1 8C 2 4h  p H9 synth centra te - F ile: N-S 1 8C 24 h pH 9.raw - T yp e: 2 Th /Th  l ocked - Start: 5 .0 00  ? - En d: 7 5 .0 01  ? - Step: 0.01 0 ? -  S te p tim e: 5 4.3 s - Tem p.: 25  ?C  (R oom) - T ime S tar ted : 2 1 s - 2 -The ta:Lin (Cps)01002003004005006007008009001000110012001300140015001600170018001900200021002200230024002500260027002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES147Figure C.17 ? XRD output for Mg:N:P molar ratio 1:1.05:1 in synthetic centrate, pH 7, 10? C, 4 hN-S 10C 4h pH7 R2 synthetic centrate00-035-0780 (*) - Newberyi te, syn - MgHPO4?3H2O - Y: 50.00 % - d x by: 1. - WL: 1.5406 - Orthorhombic - a 10.20830 - b 10.68450 - c 10.01290 - alpha 90.000 - beta 90.000 - gamma 90.000 - Primitive00-015-0762 (*) - Struvite, syn - NH4MgPO4?6H2O - Y: 50.00 % - d x by: 1. - WL: 1.5406 - Orthorhombic - a 6.94500 - b 11.20800 - c 6.13550 - alpha 90.000 - beta 90.000 - gamma 90.000 - Primitive - POperations: ImportN-S 10C 4h pH7 R2 synthetic centrate - Fi le: N-S 10C 4h pH7.raw - Type: 2Th/Th locked - Start: 5.000 ? - End: 75.001 ? - Step: 0.019 ? - Step time: 36.2 s - Temp.: 25 ?C (Room) - Time Started: 21 s - 2-TLin (Cps)01002003004005006007008009001000110012002-Theta - Scale5 10 20 30 40 50 60 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES148Figure C.18 ? XRD output for Mg:N:P molar ratio 1:1.05:1 in synthetic centrate, pH 8, 10? C, 4 hN-S 10C 4h pH8 R2 synthetic centrate00 -0 35 -07 80  (*) -  Ne wb eryi te, syn  -  Mg HP O4 ?3 H2O  -  Y: 50 .0 0 %  - d  x b y: 1. - W L : 1.5 40 6 - Ortho rho mbic - a 10 .2 08 30  -  b  1 0.68 4 50  - c 1 0 .0 12 90  -  al ph a 9 0.0 00  - b eta  9 0.00 0 - ga mma 90 .0 00  -  P rim i tive00 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  POp eration s: Imp ortN-S  1 0C 4 h pH8  R 2 syn thetic ce ntrate - F i le: N- S 10 C 4 h  p H8.raw - Type: 2Th /T h locked  - Star t: 5 .0 00  ? - En d: 75 .0 01  ? - Step : 0.01 9 ? - S te p ti me: 3 6.2 s -  Te mp.: 2 5 ?C ( Room) - T ime  S tar te d: 2 1 s -  2 -TLin (Cps)0100200300400500600700800900100011001200130014001500160017002-Theta - Scale6 1 0 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES149Figure C.19 ? XRD output for Mg:N:P molar ratio 1:1.05:1 in synthetic centrate, pH 7, 25? C, 4 hN-S 25C 4h pH7 R2 synthetic centrate00 -0 35 -07 80  (*) -  Ne wb eryi te, syn  -  Mg HP O4 ?3 H2O  -  Y: 50 .0 0 %  - d  x b y: 1. - W L : 1.5 40 6 - Ortho rho mbic - a 10 .2 08 30  -  b  1 0.68 4 50  - c 1 0 .0 12 90  -  al ph a 9 0.0 00  - b eta  9 0.00 0 - ga mma 90 .0 00  -  P rim i tive00 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  POp eration s: Imp ortN-S  2 5C 4 h pH7  R 2 syn thetic ce ntrate - F i le: N- S 25 C 4 h  p H7.raw - Type: 2Th /T h locked  - Star t: 5 .0 00  ? - En d: 75 .0 01  ? - Step : 0.01 9 ? - S te p ti me: 3 6.2 s -  Te mp.: 2 5 ?C ( Room) - T ime  S tar te d: 2 1 s -  2 -TLin (Cps)010020030040050060070080090010002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX C: RESULTS OF XRD ANALYSIS OF SOLID SAMPLES150Figure C.20 ? XRD output for Mg:N:P molar ratio 1:1.05:1 in synthetic centrate, pH 8, 25? C, 4 hN-S 25C 4h pH8 R2 synthetic centrate00 -0 35 -07 80  (*) -  Ne wb eryi te, syn  -  Mg HP O4 ?3 H2O  -  Y: 50 .0 0 %  - d  x b y: 1. - W L : 1.5 40 6 - Ortho rho mbic - a 10 .2 08 30  -  b  1 0.68 4 50  - c 1 0 .0 12 90  -  al ph a 9 0.0 00  - b eta  9 0.00 0 - ga mma 90 .0 00  -  P rim i tive00 -0 15 -07 62  (*) -  S tr uvite, syn  - NH4 MgP O4 ?6H2 O - Y: 5 0.00  %  - d  x by: 1 . - W L: 1 .5 40 6  - Or th orh omb ic -  a 6.9 45 00  -  b  1 1.20 80 0 - c 6 .1 3 55 0 - alph a 90 .0 00  - b eta 90 .0 00  -  g amma 90 .00 0 - Pri m itive -  POp eration s: Imp ortN-S  2 5C 4 h pH8  R 2 syn thetic ce ntrate - F i le: N- S 25 C 4 h  p H8.raw - Type: 2Th /T h locked  - Star t: 5 .0 00  ? - En d: 75 .0 01  ? - Step : 0.01 9 ? - S te p ti me: 3 6.2 s -  Te mp.: 2 5 ?C ( Room) - T ime  S tar te d: 2 2 s -  2 -TLin (Cps)01002003004005006007008009001000110012001300140015002-Theta - Scale5 10 20 30 40 50 6 0 70APPENDIX D: MODELLING RESULTS151APPENDIX D: MODELLING RESULTSTable D.1 ? Model equilibrium prediction for newberyite in ammonia solution at 35? C andMg:N:P molar ratio 1:1.1:1 with consideration of bobbieriteConditions Concentration (mg/L)Solid Species Aqueous SpeciesTemp (?C) pH Bobbierite Struvite Mg NH4-N PO4-P357.04 591 8453 48 256 1688.09 232 10339 12 141 579.26 249 10368 2 147 47Table D.2 ? Model equilibrium prediction for newberyite in ammonia solution at Mg:N:Pmolar ratio 1:1.1:1 without consideration of bobbieriteConditions Concentration (mg/L)Solid Species Aqueous SpeciesTemp (?C) Final pH Newberyite Struvite Mg NH4-N PO4-P105.88 3063 3247 380 553 4917.05 0 10707 72 127 938.18 0 11280 16 95 209.21 0 11388 5 88 7256.20 5676 1432 198 657 2567.02 0 10491 93 140 1208.18 0 11236 20 97 269.48 0 11373 6 89 8356.44 6626 766 132 695 1707.04 437 9859 95 176 1228.09 0 11162 27 101 359.26 0 11334 10 91 14APPENDIX D: MODELLING RESULTS152Figure D.1 ? Comparison of real and model-predicted NH4-N for newberyite in ammoniasolution at 10? C and Mg:N:P molar ratio 1:1.1:1Figure D.2 ? Comparison of real and model-predicted NH4-N for newberyite in ammoniasolution at 25? C and Mg:N:P molar ratio 1:1.1:101002003004005006007008000 2 4 6 8 10 12 14NH4-N (mg/L)Time (h)No control(pH~6.3)pH 7pH 8pH 9ModelledEquilibrium01002003004005006007008000 2 4 6 8 10 12 14NH4-N (mg/L)Time (h)No control(pH~6.5)pH 7pH 8pH 9ModelledEquilibriumAPPENDIX D: MODELLING RESULTS153Figure D.3 ? Comparison of real and model-predicted NH4-N for newberyite in ammoniasolution at 35? C and Mg:N:P molar ratio 1:1.1:1Figure D.4 ? Comparison of real and model-predicted PO4-P for newberyite in ammoniasolution at 10? C and Mg:N:P molar ratio 1:1.1:101002003004005006007008000 2 4 6 8 10 12 14NH4-N (mg/L)Time (h)No control(pH~6.6)pH 7pH 8pH 9ModelledEquilibrium0501001502002503003504004505000 2 4 6 8 10 12 14PO4-P (mg/L)Time (h)No control(pH~6.3)pH 7pH 8pH 9ModelledEquilibriumAPPENDIX D: MODELLING RESULTS154Figure D.5 ? Comparison of real and model-predicted PO4-P for newberyite in ammoniasolution at 25? C and Mg:N:P molar ratio 1:1.1:1Figure D.6 ? Comparison of real and model-predicted PO4-P for newberyite in ammoniasolution at 35? C and Mg:N:P molar ratio 1:1.1:10501001502002503003504004505000 2 4 6 8 10 12 14PO4-P (mg/L)Time (h)No control(pH~6.5)pH 7pH 8pH 9ModelledEquilibrium0501001502002503003504004505000 2 4 6 8 10 12 14PO4-P (mg/L)Time (h)No control(pH~6.6)pH 7pH 8pH 9ModelledEquilibriumAPPENDIX D: MODELLING RESULTS155Table D.3 ? Model equilibrium prediction for newberyite in ammonia solution at 35? C andMg:N:P molar ratio 1:1.4:1 with consideration of bobbieriteConditions Concentration (mg/L)Solid Species Aqueous SpeciesTemp (?C) pH Bobbierite Struvite Mg NH4-N PO4-P35 7.07 398 7855 56 290 103Table D.4 ? Model equilibrium prediction for newberyite in ammonia solution at Mg:N:Pmolar ratio 1:1.4:1 without consideration of bobbieriteConditions Concentration (mg/L)Solid Species Aqueous SpeciesTemp (?C) Final pH Newberyite Struvite Mg NH4-N PO4-P105.88 1434 3247 380 553 4917.07 0 8653 49 245 638.08 0 9031 12 223 159.16 0 9114 3 219 4256.20 4047 1432 198 657 2566.99 0 8448 69 256 898.14 0 9006 14 224 189.07 0 9094 5 220 7356.44 4997 766 132 695 1707.07 0 8423 71 258 928.16 0 8982 16 226 219.01 0 9070 8 221 10APPENDIX D: MODELLING RESULTS156Figure D.7 ? Comparison of real and model-predicted NH4-N for newberyite in ammoniasolution at 10? C and Mg:N:P molar ratio 1:1.4:1Figure D.8 ? Comparison of real and model-predicted NH4-N for newberyite in ammoniasolution at 25? C and Mg:N:P molar ratio 1:1.4:101002003004005006007008000 2 4 6 8 10 12 14NH4-N (mg/L)Time (h)No control(pH~6.3)pH 7pH 8pH 9ModelledEquilibrium01002003004005006007008000 2 4 6 8 10 12 14NH4-N (mg/L)Time (h)No control(pH~6.5)pH 7pH 8pH 9ModelledEquilibriumAPPENDIX D: MODELLING RESULTS157Figure D.9 ? Comparison of real and model-predicted NH4-N for newberyite in ammoniasolution at 35? C and Mg:N:P molar ratio 1:1.4:1Figure D.10 ? Comparison of real and model-predicted PO4-P for newberyite in ammoniasolution at 10? C and Mg:N:P molar ratio 1:1.4:101002003004005006007008000 2 4 6 8 10 12 14NH4-N (mg/L)Time (h)No control(pH~6.6)pH 7pH 8pH 9ModelledEquilibrium0501001502002503003504004505000 2 4 6 8 10 12 14PO4-P (mg/L)Time (h)No control(pH~6.3)pH 7pH 8pH 9ModelledEquilibriumAPPENDIX D: MODELLING RESULTS158Figure D.11 ? Comparison of real and model-predicted PO4-P for newberyite in ammoniasolution at 25? C and Mg:N:P molar ratio 1:1.4:1Figure D.12 ? Comparison of real and model-predicted PO4-P for newberyite in ammoniasolution at 35? C and Mg:N:P molar ratio 1:1.4:10501001502002503003504004505000 2 4 6 8 10 12 14PO4-P (mg/L)Time (h)No control(pH~6.5)pH 7pH 8pH 9ModelledEquilibrium0501001502002503003504004505000 2 4 6 8 10 12 14PO4-P (mg/L)Time (h)No control(pH~6.6)pH 7pH 8pH 9ModelledEquilibriumAPPENDIX D: MODELLING RESULTS159Table D.5 ? Model equilibrium prediction for newberyite in synthetic crystallizer effluentat Mg:N:P molar ratio 1:1:1 with consideration of magnesite and bobbieriteConditions Concentration (mg/L)Solid Species Aqueous SpeciesTemp (?C) pH Magnesite Struvite Mg NH4-N PO4-P10 7.00 344 14382 69 102 2167.99 428 14774 7 80 16725 7.01 680 13558 54 149 3218.07 704 13990 5 125 266Table D.6 ? Model equilibrium prediction for newberyite in synthetic crystallizer effluentat Mg:N:P molar ratio 1:1:1 without consideration of magnesite and bobbieriteConditions Concentration (mg/L)Solid Species Aqueous SpeciesTemp (?C) pH Newberyite Struvite Mg NH4-N PO4-P10 7.00 0 14823 125 77 1617.99 0 15626 46 31 5925 7.01 310 14308 132 106 1718.07 0 15562 52 35 67Table D.7 ? Model equilibrium prediction for synthetic newberyite in synthetic centrate atMg:N:P molar ratio 1:1.05:1 with consideration of magnesite and bobbieriteConditions Concentration (mg/L)Solid Species Aqueous SpeciesTemp (?C) pH Magnesite Struvite Mg NH4-N PO4-P10 7.01 271 15347 60 136 1748.10 337 15714 5 115 12825 7.05 620 14504 42 184 2818.06 627 14872 4 163 234APPENDIX D: MODELLING RESULTS160Table D.8 ? Model equilibrium prediction for synthetic newberyite in synthetic centrate atMg:N:P molar ratio 1:1.05:1 without consideration of magnesite and bobbieriteConditions Concentration (mg/L)Solid Species Aqueous SpeciesTemp (?C) pH Newberyite Struvite Mg NH4-N PO4-P10 7.01 0 15734 99 114 1258.10 0 16474 27 71 3125 7.05 592 15577 115 123 1458.06 0 16386 35 76 42

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0103342/manifest

Comment

Related Items