Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Comparing the effectiveness of CO₂ strippers to reduce operating costs for struvite formation Sabrina, Nandini 2007

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

Item Metadata

Download

Media
831-ubc_2007-0577.pdf [ 8.57MB ]
Metadata
JSON: 831-1.0063218.json
JSON-LD: 831-1.0063218-ld.json
RDF/XML (Pretty): 831-1.0063218-rdf.xml
RDF/JSON: 831-1.0063218-rdf.json
Turtle: 831-1.0063218-turtle.txt
N-Triples: 831-1.0063218-rdf-ntriples.txt
Original Record: 831-1.0063218-source.json
Full Text
831-1.0063218-fulltext.txt
Citation
831-1.0063218.ris

Full Text

COMPARING THE EFFECTIVENESS OF C 0 2 STRIPPERS TO REDUCE OPERATING COSTS FOR STRUVITE FORMATION B y NANDINI SABRINA B . S c . ( C i v i l Engineer ing) , Bangladesh U n i v e r s i t y o f Eng inee r ing and Techno logy , D h a k a , Bangladesh, 2003 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F MASTER OF APPLIED SCIENCE i n T H E F A C U L T Y O F G R A D U A T E S T U D I E S ( C I V I L E N G I N E E R I N G ) T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A M a y 2007 © N a n d i n i Sabrina, 2007 ABSTRACT Sta iv i te crys ta l l iza t ion offers the potential of r emov ing phosphorus from wastewater and recover ing it in a form that can be used as a fertil izer. In 1999, the Department o f C i v i l Engineer ing at the Univers i ty o f Br i t i sh C o l u m b i a ( U B C ) started a phosphorus recovery project. The U B C struvite group is now researching ways to reduce the operating costs o f struvite product ion. One o f the major operational costs o f struvite product ion is the cost o f caustic chemicals that is added to obtain a desired level o f operative p H . The m a i n object ive o f this research was to introduce two types o f C 0 2 strippers into the struvite c rys ta l l iza t ion process and determine their effectiveness in reducing caustic chemical use, thereby he lp ing to reduce the operational costs o f struvite product ion. In this study, two C O ? strippers were used - (i) compact media stripper and ( i i ) cascade stripper. The strippers were connected to two identical struvite crystal l izers . The reactors were placed at the L u l u Island Wastewater Treatment Plant ( L I W W T P ) . The strippers were tested under different operating condi t ions, and their effectiveness in reducing caustic use was compared. Throughout the project, a h igh percentage o f phosphorus removal was achieved under each condi t ion , by both the reactors/strippers. M o s t o f the t ime, the phosphorus remova l rate was around 9 0 % . The compact media stripper fai led to save any amount o f caustic, regardless o f the operating condi t ions . Instead, more caustic was required once the stripper was introduced. O n e o f the reasons was that the stripper b locked the passage o f stripped off C O 2 , since it was mounted on top o f the clarif ier . Another reason was the suscept ibi l i ty o f stripper 's pack ing media to become frequently c logged, w h i c h also resulted i n b l o c k i n g the movement o f C O 2 through the s t r ipping tower. O n the other hand, the cascade stripper was very effective in saving caustic. The amount o f caustic saved by this stripper ranged from 3 5 % to 86%, depending on the i i operating condit ions. B o t h strippers showed very poor performance regarding ammonia str ipping, w i th the compact media stripper being s l igh t ly better i n s t r ipping ammonia . The harvested struvite pellets f rom both the reactors were composed o f nearly pure struvite (94% by mass), w i th a smal l amount o f c a l c i u m and traces o f i ron and potassium. Different operating condi t ions d i d not have any affect o n the qual i ty o f harvested struvite. u i TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS... iv LIST OF TABLES viii LIST OF FIGURES ix LIST OF ABBREVIATIONS AND SYMBOLS xi ACKNOWLEDGEMENTS xii CHAPTER ONE: INTRODUCTION 1 1.1 Background 1 1.2 Research Objectives 4 CHAPTER TWO: LITERATURE REVIEW 5 2.1 Phosphorus Removal From Wastewater 5 2.1.1 Chemical phosphorus precipitation 5 2.1.2 B iological nutrient removal 5 2.2 Phosphorus Recovery As Struvite 6 2.2.1 Struvite •• 6 2.2.2 Benefits of phosphorus recovery as struvite 7 2.2.3 Struvite recovery technologies 8 2.3 Struvite Chemistry 9 2.3.1 Solubility product (K s p ) 9 2.3.2 Conditional solubility product (Ps) 10 2.3.3 Supersaturation ratio (SSR) 10 2.4 Parameters Of Struvite Crystallization 11 2.4.1 pH 11 2.4.2 Magnesium to phosphorus molar ratio 12 2.4.3 Ammonia to phosphorus molar ratio 13 2.4.4 Temperature 13 iv 2.4.5 M i x i n g energy (or turbulence) 13 2.4.6 Presence o f foreign ions 14 2.4.7 Ini t ial reactor seeding 14 2.5 Operat ing Costs F o r Struvite Produc t ion 14 2.6 Methods O f R a i s i n g The p H V a l u e O f Wastewater 15 2.6.1 Caus t ic chemica l addi t ion '. 16 2.6.2 Aera t ion or CO2 s tr ipping 17 2.7 Mechan i sms O f p H Increase B y St r ipp ing C O ? 19 2.8 A m m o n i a St r ipping 20 CHAPTER THREE: METHODS AND MATERIALS 22 3.1 L u l u Island Wastewater Treatment Plant 22 3.2 Mater ia l s A n d Equ ipmen t 22 3.2.1 Struvite Crys ta l l i ze r 22 Chemica l s , Storage Tanks A n d Pumps 24 3.2.2 C o m p a c t M e d i a Stripper 26 3.2.3 Cascade Stripper 29 3.3 Exper imenta l D e s i g n 32 3.4 Sample C o l l e c t i o n , Storage A n d Preservation 33 3.5 A n a l y t i c a l Methods 34 3.5.1 M a g n e s i u m 34 3.5.2 Ortho-phosphate and ammonia 35 3.5.3 C a l c i u m , a luminum, i ron and potassium 35 3.5.4 Caus t ic analysis 35 3.5.5 D i s s o l v e d C 0 2 36 3.6 Pel let Qua l i t y Determinat ion 36 CHAPTER FOUR: RESULTS AND DISCUSSION 37 4.1 Centrate Characterist ics 37 4.2 Performance O f The Reactors (Without Stripper) 39 S u m m a r y o f results 43 4.3 Performance O f The Strippers 44 v 4.3.1 R u n N o . 1 44 Nutr ient r emova l 45 Caust ic use 46 Ca rbon d iox ide str ipping 48 A m m o n i a str ipping 48 Summary o f results 50 4.3.2 R u n N o . 2 50 Nutrient removal 51 Caus t ic use 52 C a r b o n d iox ide str ipping 53 A m m o n i a s t r ipping 54 S u m m a r y o f results 55 4.3.3 R u n N o . 3 56 Nutrient removal 56 Caus t ic use 58 Ca rbon d iox ide str ipping 59 A m m o n i a str ipping 60 S u m m a r y o f results 61 4.4 C o m p a r i s o n O f Stripper Performance 61 4.5 Qua l i t y O f Harvested Struvite 65 4.6 Operat ional Problems 66 4.6.1 C l o g g i n g o f the compact media stripper 66 4.6.2 p H control ler p rob lem 68 4.6.3 F l o w fluctuation 69 4.6.4 P l u g g i n g o f tubing 69 4.6.5 Reactor fou l ing 70 4.6.6 Centrate supply 70 4.6.7 Suspended solids 70 CHAPTER FIVE: CONCLUSIONS AND RECOMMENDATIONS.. . 72 5.1 Conc lus ions 72 v i 5.2 Recommendat ions 73 CHAPTER SIX: REFERENCES 75 APPENDIX A: INSTRUMENT OPERATIONAL PARAMETERS 81 APPENDIX B: OPERATIONAL DATA 83 APPENDIX C: CHEMICAL ANALYSIS OF STRUVITE PELLETS.. 94 APPENDIX D: C 0 2 CALIBRATION CURVE 95 vii LIST OF TABLES Table 3.1: D imens ions o f the reactor 24 Tab le 3.2: Test condi t ion (for both reactors) 32 Tab le 4.1: Characterist ics o f Centrate (22 A u g - 13 D e c , 2006) 37 Table 4.2: Opera t ing condi t ions (25 Sep-6 Oct , 2006) 39 Tab le 4.3: Opera t ing condit ions for the 1 s t run (13-18 N o v , 2006) 44 Tab le 4.4: Opera t ing condit ions for the 2 n d run (20 N o v -1 D e c , 2006) 50 Tab le 4.5: Opera t ing condit ions for the 3 r d run (4-12 D e c , 2006) 56 Table 4.6: S u m m a r y o f findings from four tests 62 Table 4.7: S u m m a r y o f struvite pellet compos i t ion 65 Table 4.8: Impuri ty contents o f struvite pellet 66 v i i i LIST OF FIGURES Figure 3.1: Bas i c f low diagram o f the struvite crystal l izer process 23 Figure 3.2: The set up o f struvite crystal l izers (wi th strippers) at the L I W W T P 26 Figure 3.3: P a c k i n g material o f the compact media stripper 27 Figure 3.4: Compac t media stripper 28 Figure 3.5: Compac t media stripper connected to the R # l at the L I W W T P 29 Figure 3.6: Deta i led design o f the baffle 30 Figure 3.7: Front v i e w o f the cascade stripper 31 Figure 3.8: Cascade stripper connected to the R#2 at the L I W W T P 31 Figure 4 .1 : p H of centrate (22 A u g - 13 D e c , 2006) 37 Figure 4.2: M a g n e s i u m , phosphate and ammon ia concentration in centrate (22 A u g - 13 Dec , 2006) 38 Figure 4.3: M g : P and N : P ratio o f centrate (22 A u g - 13 D e c , 2006) 38 Figure 4.4: M a g n e s i u m removal rates in R # l and R#2 (25 Sep-6 Oct , 2006) 40 Figure 4.5: A m m o n i a removal rates in R # l and R#2 (25 Sep-6 Oct , 2006) 40 Figure 4.6: Phosphorus removal rates in R # l and R#2 (25 Sep-6 Oct , 2006) 41 Figure 4.7: D a i l y caustic use in R # l and R#2 (25 Sep-6 Oct , 2006) 41 Figure 4.8: Caus t ic use and p H increase in R # l and R#2 (25 Sep-6 Oct , 2006) 42 Figure 4.9: M o l a r P removal and caustic use i n R # l and R#2 (25 Sep-6 Oct , 2006) 42 Figure 4.10: M a g n e s i u m removal rates in R # l and R#2 (13-18 N o v , 2006) 45 Figure 4 .11: A m m o n i a removal rates in R # l and R#2 (13-18 N o v , 2006) 45 Figure 4.12: Phosphorus removal rates in R # l and R#2 (13-18 N o v , 2006) 46 Figure 4.13: D a i l y caustic use in R # l and R#2 (13-18 N o v , 2006) 46 Figure 4.14: M o l a r P removal and caustic use in R # l and R#2 (13-18 N o v , 2006) 47 Figure 4.15: C 0 2 r emoval in R # l and R#2 (13-18 N o v , 2006) 48 Figure 4.16: A m m o n i a s tr ipping in R # l and R#2 (13-18 N o v , 2006) 49 Figure 4.17: A m m o n i a s tr ipping i n R # l and R#2 (13-18 N o v , 2006) 49 Figure 4.18: M a g n e s i u m removal rates in R # l and R#2 (20 N o v -1 Dec , 2006) 51 ix Figure 4.19: A m m o n i a removal rates in R # l and R#2 (20 N o v -1 Dec , 2006) 52 Figure 4.20: Phosphorus removal rates in R # l and R#2 (20 N o v -1 Dec , 2006) 52 Figure 4 .21: D a i l y caustic use i n R # l and R#2 (20 N o v -1 D e c , 2006) 53 Figure 4.22: M o l a r P removal and caustic use i n R # l and R#2 (20 N o v -1 D e c , 2006). . . 53 Figure 4.23: C 0 2 r emoval in R # l and R#2 (20 N o v -1 Dec , 2006) 54 Figure 4.24: A m m o n i a str ipping in R # l and R#2 (20 N o v -1 D e c , 2006) 55 Figure 4.25: A m m o n i a str ipping in R # l and R#2 (20 N o v -1 D e c , 2006) 55 Figure 4.26: M a g n e s i u m removal rates in R # l and R#2 (4-12 D e c , 2006) 57 Figure 4.27: A m m o n i a removal rates in R # l and R#2 (4-12 D e c , 2006) 57 Figure 4.28: Phosphorus removal rates i n R # l and R#2 (4-12 D e c , 2006) 58 Figure 4.29: D a i l y caustic use i n R # l and R#2 (4-12 Dec, 2006) 58 Figure 4.30: M o l a r P removal and caustic use i n R # l and R#2 (4-12 D e c , 2006) 59 Figure 4 .31: C 0 2 removal in R # l and R#2 (4-12 D e c , 2006) 60 Figure 4.32: A m m o n i a str ipping i n R # l and R#2 (4-12 Dec, 2006) 60 Figure 4.33: A m m o n i a str ipping in R # l and R#2 (4-12 Dec , 2006) 61 Figure 4.34: O l d support ing plate 67 Figure 4.35: N e w supporting plate 67 Figure 4.36: H a n g i n g packing materials 68 x LIST OF ABBREVIATIONS AND SYMBOLS A A S A t o m i c Absorp t ion Spectrophotometer B N R B i o l o g i c a l Nutrient R e m o v a l E B N R Enhanced B i o l o g i c a l Nutr ient R e m o v a l E B P R Enhanced B i o l o g i c a l Phosphorus R e m o v a l E C A Externa l Cont inuous Aera t ion E G A Externa l Gradua l Ae ra t i on F B R F l u i d i z e d B e d Reactor G V R D Greater Vancouver Reg iona l Dis t r ic t K S p S o l u b i l i t y Product L I W W T P L u l u Island Wastewater Treatment Plant M A P M a g n e s i u m A m m o n i u m Phosphate P A O Phosphorus A c c u m u l a t i n g O r g a n i s m P s C o n d i t i o n a l So lub i l i t y Product R R R e c y c l e Ra t io S S R Supersaturation Ra t io U B C Unive r s i ty o f Br i t i sh C o l u m b i a x i ACKNOWLEDGEMENTS I w o u l d l ike to express m y sincere gratitude to many people who assisted, supported and encouraged me throughout the course o f this research. F i r s t ly , I w o u l d l ike to thank my supervisor, D r . D . S . M a v i n i c , for p rov id ing me the opportunity to work wi th h i m , and for his guidance and support throughout a l l stages o f this work . K a z i Parvez Fattah, m y fe l low graduate student i n this research group, for his immense help, guidance and support from day one, w h i c h continues even today. H e trained me o n the reactor set up at the L I W W T P , drove me to the plant throughout the sampl ing per iod and helped me in mainta in ing the reactors. Frederic K o c h and Y i n g Zhang , for freely sharing their ideas and experiences, and helping me out at the L I W W T P . Paula Park inson and Susan Harper o f the U B C Env i ronmenta l Engineer ing Laboratory, for their assistance w i t h analyt ical works and logis t ic support at the laboratory. Spec ia l thanks to Raphael Fugere, m y friend and mentor, and Syed Z a k i A b d u l l a h , m y best friend, for their helpful suggestions in technical matters and mental support to deal wi th stress. M y parents, for their uncondi t ional love and support throughout m y graduate studies at the U B C . The people at the L I W W T P , for a l l o w i n g me to conduct m y research at the plant and for their assistance a l l a long. N S E R C , G V R D , Stantec L t d . and Ostara Inc., for their generous funding o f this research project. x u CHAPTER ONE INTRODUCTION 1.1 Background Phosphorus is an essential part o f the nutrient cyc le i n nature. It is an irreplaceable element in many phys io log ica l and b iochemica l processes in plants and animals . Due to its h igh reactivity, phosphorus is never found free in nature, but it is w ide ly distr ibuted in many different minerals. Phosphorus compounds are present in mun ic ipa l wastewater, or ig inat ing from detergents as w e l l as from metabol i sm processes, diffuse runoff from agricultural lands and input from the air. Phosphorus based compounds are w ide ly used in modern industry. The most important commerc ia l use o f phosphorus based chemicals is the product ion o f fer t i l izers .Global demand for fertilizers has led to large increases in phosphate ( P O 4 3 ) product ion i n the second half o f the 2 0 t h century. Unfortunately, phosphoais is a non-renewable resource, w i t h finite reserves g loba l ly . It is estimated that there are 7,000 m i l l i o n tons o f phosphate rocks as P2O5 remain ing in those reserves, that cou ld be economica l ly mined, and another around 11,000 m i l l i o n tons o f phosphate rocks that cannot be processed economica l ly at present. The human populat ion consumes 40 m i l l i o n tons o f phosphorus as P2O5 each year (Jas inski , 1999) and its demand w i l l increase by 1.5% each year (Steen, 1998). It is predicted that this resource cou ld be exhausted w i t h i n next 100-250 years (Shu et al., 2006). Hence, industry and the populat ion, in general, are now seeking alternative, sustainable ways o f recover ing phosphoais f rom different sources. M u n i c i p a l wastewater has a great potential to become a source o f phosphorus recovery. Phosphoais is generally removed from wastewater either by chemica l treatment or by u t i l i z ing enhanced b io log i ca l phosphorus removal ( E B P R ) processes, general ly referred to as a b io log ica l nutrient removal ( B N R ) process. The B N R process is preferable in most instances, since, in addi t ion to removal o f nitrogen compounds, phosphorus removed from 1 wastewaters through this process remains in a bio-avai lable form, hence m a k i n g it easier for subsequent phosphorus recovery. However , in some cases, phosphorus removal may prove to be di f f icul t in B N R processes, since phosphorus is released back into l iqu id phase dur ing the sludge handl ing process, especia l ly i f the process involves anaerobic digestion. It is estimated that as much as 8 0 - 9 0 % o f phosphorus removed dur ing treatment may be released and reintroduced to the process f rom the digester supernatants and this can lead to potential system failure (Niedbala , 1995; M a v i n i c et al., 1998). Under certain condi t ions, the elevated levels o f magnesium, a m m o n i u m , and phosphate ions present in the anaerobic digester supernatant, can combine to fo rm 'struvite' (magnesium a m m o n i u m phosphate hexahydrate, M g N H ^ P C u . b ^ O ) . Unin ten t iona l struvite accumulat ion is considered to be a nuisance in wastewater treatment plants. T h e struvite deposits are hard and often diff icul t to dis lodge, and sometimes require replacement o f encrusted parts. It causes damage to pumping systems, reduces the plant f l ow capacity and contributes to major p lugging o f p ip ing . Therefore, unti l recently, a large port ion o f struvite research has been directed towards removal and prevention o f struvite formation, rather than towards forced precipitat ion o f it f rom solut ion. Research shows that more than 9 0 % of d isso lved phosphorus can be removed f rom anaerobic digester supernatant in the form of struvite precipitat ion, w i th 50-80% o f removed phosphorus recovered as harvestable struvite pellets (Batt istoni et al., 1997, 2001; M i i n c h and Barr , 2001 ; Ueno and Fu j i i , 2001). Struvite has been found to be a good plant nutrient source for nitrogen and phosphorus, since it releases these nutrients s l o w l y and has non-burning features because o f its l ow so lub i l i ty in water (Gaterel l et al., 2000; S h u et al., 2006). In this way , struvite can actually generate revenues for wastewater treatment plants. It has been estimated that the payback per iod o f a struvite plant processing 55,000 m 3 / d o f waste stream c o u l d be less than five years (Shu et al., 2006). In 1999, the Department o f C i v i l Engineer ing at The Unive r s i ty o f B r i t i s h C o l u m b i a ( U B C ) started a phosphorus recovery project, in col laborat ion wi th Br i t i sh C o l u m b i a H y d r o . Ini t ia l ly , a "crysta l l izer model" , developed by the C i v i l Engineer ing Department o f U B C , 2 was tested wi th synthetic wastewater at the bench-scale. After the in i t ia l bench-scale work, the crystal l izer model had been modi f ied to resolve problems associated wi th smal l reactors, and then was successfully scaled-up from the bench-scale (2.51 L ) into the pilot-scale (about 90 L ) . The crystal l izer has subsequently been tested wi th synthetic, and as w e l l as w i t h real anaerobic digester supernatant f rom both the L u l u Island Wastewater Treatment Plant ( L I W W T P ) and the C i t y o f Pent ic ton A d v a n c e d Wastewater Treatment Plant . Studies, conducted by the U B C struvite group, show that this crystal l izer process is capable o f r emoving more than 9 0 % o f ortho-phosphate f rom waste stream, wi th more than 8 5 % o f the removed phosphorus recovered as harvestable struvite pellets (Fattah, 2004). Af ter much research on thermodynamics and process control o f struvite crys ta l l iza t ion process, the U B C struvite group is now concentrating o n f ind ing methods to reduce operating costs o f struvite precipitat ion. A s w i l l be discussed i n the next chapter, operational costs o f struvite ma in ly depends on two factors - costs o f chemicals to be injected in the process and energy requirements for pumping . In their study, Jaffer et al. (2002) showed that, compared to cost o f chemicals , energy requirement costs are quite insignif icant and 9 7 % o f the total chemica l cost was due to the addit ion o f caustic, i n order to achieve a desirable operating p H leve l . Therefore, by reducing caustic use, a large fraction o f struvite crysta l l izat ion operational costs can be reduced. Research carried out by Bat t is toni et al. (1997, 1998, 2001) showed that the operative p H can be obtained by app ly ing air s t r ipping o f CO2 on ly , without any addi t ion o f caustic chemicals . Inspired by these f indings , Z h a n g (2006) developed a cascade CO2 stripper and tested it at the L I W W T P . The stripper was able to reduce 4 6 % to 6 5 % o f caustic chemica l addit ion, depending o n the operating condi t ions . 3 1.2 Research Objectives The objectives o f this research program were designed: > T o introduce two types o f C O ? strippers into the struvite c rys ta l l iza t ion process and determine their effectiveness in reducing operating costs o f struvite precipi tat ion. > T o test the two C C h strippers under different condi t ions and compare their effectiveness in reducing caustic chemica l use. > T o investigate the qual i ty o f harvested struvite. > A s a minor objective, to compare the strippers' a m m o n i a s t r ipping abi l i ty . 4 CHAPTER TWO LITERATURE REVIEW 2.1 Phosphorus Removal From Wastewater The technologies for phosphorus removal f rom wastewater started to develop back i n the 1950s, in response to the p rob lem associated w i t h eutrophication i n water bodies and the subsequent need to reduce the levels o f phosphorus before discharging it to surface waters (Morse et al, 1998). The current discharge l i m i t on total phosphorus i n N o r t h A m e r i c a ranges f rom 2 to 0.1 m g / L (Tchobanoglous et al, 2003). Presently, several technologies are avai lable for r emov ing phosphorus f rom wastewater, i n order to meet the discharge guidel ines. A m o n g those, the two most w i d e l y used technologies are chemica l precipi tat ion and b io log i ca l nutrient removal ( B N R ) . 2.1.1 Chemical phosphorus precipitation C h e m i c a l precipi ta t ion is the oldest technology o f phosphorus remova l . It is a s imple and rel iable method, and therefore, s t i l l remains as the leading technology i n r emov ing phosphorus. The chemica l precipi tat ion o f phosphorus is brought about by the addi t ion o f a divalent or trivalent metal salt to wastewater, causing precipi ta t ion o f an insoluble metal phosphate that is settled out by sedimentation. The metal ions used most c o m m o n l y are c a l c i u m [Ca(II)], a l u m i n u m [Al(III)] and i ron [Fe(III)]. Po lymers have been used together w i t h a l um and l ime as f locculent aids. Howeve r , this process offers quite l o w phosphorus recovery as metal-bound phosphorus makes subsequent r ecyc l ing diff icul t . 2.1.2 Biological nutrient removal B i o l o g i c a l phosphorus removal is achieved i n the activated sludge process by u t i l i z ing the abi l i ty o f phosphorus accumulat ing organisms ( P A O s ) to accumulate phosphates as polyphosphates, for their o w n metabol ism. T h i s enhanced b io log i ca l nutrient r emova l 5 ( E B N R ) process also provides simultaneous nitrogen removal . O v e r a l l , phosphorus removal rates o f 80-90% can be achieved through this process (Morse et al, 1998). The m a i n advantages o f this technology over chemica l precipi tat ion are avo id ing the use o f chemicals and producing less sludge. A l s o , f rom a phosphorus recovery perspective, E B N R is better than chemica l precipitat ion, since b i o l o g i c a l l y bound phosphorus is more recyclable . O n the other hand, this technology requires more complex plant configurations and operations. 2.2 Phosphorus Recovery As Struvite 2.2.1 Struvite M a g n e s i u m a m m o n i u m phosphate ( M A P ) hexahydrate ( M g N H 4 P 0 4 . 6 H 2 0 ) , more c o m m o n l y k n o w n as 'struvite', is a minera l that is composed o f magnes ium, a m m o n i u m , and phosphate i n equal molar concentrations. It belongs to the group o f the orthophosphates. Struvite crystal l izes i n the or thorhombic system as white to y e l l o w i s h or brownish-whi te , py ramida l crystals, or i n platey mica - l ike forms. It is a soft minera l w i th a M o h s hardness o f 1.5 to 2 and has a l o w specif ic gravity o f 1.7. It is sparingly soluble in neutral and alkal ine condi t ions, but is readi ly soluble i n ac id (Internetl) . Struvite crystals occur spontaneously i n various b io log ica l media . Fo r instance, it has been found i n rotting organic material such as guano deposits and c o w manure (Omar et al, 1994). It has also been studied i n the medica l f ie ld , as it occurs as crystallites i n urine and as a type o f k idney stone (urolith) (Coe et al, 2005); lately, it has been investigated i n so i l sciences, as a way to entrap nitrogen i n compost (Jeong and H w a n g , 2005). In the wastewater treatment area, struvite is w e l l k n o w n as a "scale p rob lem" . A c c u m u l a t i o n o f struvite o n pipe wal ls and other parts o f treatment plants causes p lugg ing problems. Struvite deposits are hard, often diff icul t to dis lodge and sometimes require replacement o f encrusted parts. Struvite was first d iscovered i n medieva l sewer systems i n H a m b u r g G e r m a n y i n 1845 (Internetl) . Borge rd ing (1972) first reported struvite as a source 6 o f scale deposits i n wastewater treatment plants, when it occurred o n the wal l s o f an anaerobic digest ion system at the H y p e r i o n wastewater treatment plant i n L o s Ange les i n 1963. S ince then, several studies have been carried out on struvite as a scal ing agent ( D o y l e and Parsons, 2002), but most o f the authors have considered struvite as a nuisance and not as a product w h i c h cou ld be o f economic interest. 2.2.2 Benefits of phosphorus recovery as struvite A l t h o u g h , i n most cases, unintentional struvite accumulat ion is k n o w n to be a serious p rob lem i n wastewater treatment plants, struvite precipi tat ion can, i n fact, serve as a process for r emoving and recover ing phosphorus f rom wastewater. T h e interests i n r emov ing and recover ing phosphorus v i a struvite are stated be low. > S ince supplies o f phosphorus and the quali ty o f phosphate-bearing rock are decreasing (Shu et al, 2006), people are now seeking an alternative sustainable source o f phosphorus. Struvite precipi ta t ion offers an excellent sustainable w a y o f recover ing phosphorus. > Struvite precipi tat ion leads to simultaneous removal o f nitrogen and phosphorus. B o t h o f these nutrients are responsible for eutrophication. > Signif icant reduct ion (8% to 31%) i n sludge vo lume can be achieved by implement ing phosphorus recovery b y crys ta l l iza t ion (Woods et al., 1999). > Struvite is k n o w n to be a good qual i ty fertil izer. It represents a h igh ly effective source o f nutrients (P, N an M g ) for plants ( L i and Zhao , 2003). Its l o w solubi l i ty in water also presents the advantage o f p ro longing the release o f nutrients, without the danger o f burn ing roots o f crops treated (Gaterel l et al, 2000; S h u et al, 2006). Ano the r advantage o f struvite as a fert i l izer is that struvite contains l ow amounts o f heavy metals, compare to phosphate bearing rocks that are m i n e d and suppl ied to fert i l izer industries (Dr ive r et al, 1999). 7 Several laboratory and pi lo t scale studies have been carried out to determine the potential o f r emov ing and recover ing phosphorus as struvite. O n l y a few o f them have been tested at fu l l scale ma in ly i n T h e Netherlands (Giesen , 1999) and Italy (Bat t is toni et al, 2005a/b). Howeve r , to date, Japan is the on ly country where complete phosphorus remova l and recovery, as struvite, has been implemented and the result ing product ion is so ld to fert i l izer companies (Gaterel l et al, 2000; U e n o and F u j i i , 2001). It should be mentioned that struvite recovery is not on ly confined to mun ic ipa l wastewater treatment systems; other waste streams also offer prospects o f struvite recovery, depending o n the chemistry o f these waste l iquors meeting the requirements o f struvite precipi tat ion. So far, struvite has been recovered f rom landf i l l leachate, swine wastewaters, piggery effluent and an imal manure ( D o y l e and Parsons, 2002). Research has also been carried out on the poss ib i l i ty o f struvite precipi tat ion f rom urine ( L i n d et al, 2000; W i l s e n a c h and van Loosdrecht , 2003; T i l l e y , 2006). 2.2.3 Struvite recovery technologies T h e technologies used to recover phosphorus as struvite can be c lass i f ied i n three m a i n categories. These are as fo l lows : 1. Select ive i o n exchange 2. Precipi ta t ion i n a stirred reactor 3. Precipi ta t ion i n f lu id ized bed reactors ( F B R ) or air-agitated reactors A m o n g these three processes, F B R is the most c o m m o n l y used and w i d e l y investigated technology to crysta l l ize struvite f rom wastewater. In F B R systems, struvite particles precipitate spontaneously f rom supernatants, f o l l o w i n g the addi t ion o f chemicals to reach the molar ratio o f M g : P : N to 1:1:1. Once the nucleation o f the first particle starts, the growth takes place either b y interaction o f smal l struvite particles together or b y contact o n in i t ia l seed materials. Suspension o f particles inside the reactor is maintained by either l i q u i d f l o w rates (Cecch i et al, 2003) or an upf low c i rcula t ion o f air ( S u z u k i et al, 2002); thus, the 8 particles inside the reactor are i n continuous mot ion , and behave l ike a dense f l u id . In an F B R system, feed solutions enter f rom the bot tom of the reactor and the ve loc i ty o f f l ow decreases w i t h increasing height o f the reactor. 2.3 Struvite Chemistry The crys ta l l iza t ion o f struvite is a h igh ly dynamic and complex phenomenon control led by a number o f factors. The precipi tat ion o f struvite involves a number o f reactions. E x c l u d i n g the side reactions, the formation o f struvite can be expla ined by the f o l l o w i n g general ized equation. M g 2 + + N H 4 + + PO4 3" + 6 H 2 0 M g N H 4 P 0 4 . 6 H 2 0 ( E q . 1) Ano the r formula has been proposed by Sh imamura et al. (2003). A c c o r d i n g to that formula , struvite is precipitated f o l l o w i n g the equation g iven be low. M g 2 + + N H 4 + + HPO4 2 " + O H " + 5 H 2 0 M g N H 4 P 0 4 . 6 H 2 0 (Eq.2) The difference between the above two equations is that, in the second equation, HPO4 " has been used instead o f PO4 " and O H " has been introduced. T h e reason for us ing different phosphate ion is that HPO4"" is more dominant than PO4 " in the normal operating p H o f struvite formation. 2.3.1 Solubility product (Ksp) The so lub i l i ty product ( K s p ) is defined as the equ i l i b r i um constant o f a reaction i n v o l v i n g a precipitate and its constituent ions. The so lubi l i ty product ( K s p ) can be used to describe the rate o f reaction at w h i c h struvite forms and dissolves i n l i q u i d solut ion. Several studies have been carried out to determine the so lubi l i ty product o f struvite (Buchanan et al, 1994; A a g e et al, 1997; Ohl inger et al, 1998; B h u i y a n , 2007). Unfortunately, there is no 9 universal agreement on the value o f struvite so lub i l i ty product. Andrade and S c h u i l i n g (2001) mentioned four reasons w h i c h cause this discrepancy. > The so lub i l i ty product m a y be der ived by us ing approximate solu t ion equ i l ib r ium. > The effects o f ion ic strengths are often neglected. ^ M a s s balance and electroneutrality equations are not always used. > Different chemica l species are selected for calculat ions. H o w e v e r , there is a general agreement on the fact that the value o f the so lub i l i ty product o f struvite decreases w i t h increasing p H o f l i q u i d (Ohl inger , 1999; D o y l e and Parsons, 2002). 2.3.2 Conditional solubility product (Ps) In order to avo id the complexi t ies associated w i t h the ca lcula t ion o f so lub i l i ty product o f struvite, the condi t ional so lub i l i ty product (P s ) is use for pract ical purposes. The condi t ional so lubi l i ty product is determined by measuring o n l y concentrations o f total magnes ium, a m m o n i u m and phosphate present i n solut ion, as shown by the f o l l o w i n g equation. Ps = [ M g + 2 ] t o t a l . [NH4 -N ] t o t a l . [P0 4 -P]totai ( E q . 3) Howeve r , ca lcula t ion o f the condi t ional so lub i l i ty product takes no account o f p H , ionic act ivi ty and ion ic strength, and it is on ly accurate for a specif ic p H value; whereas, a so lubi l i ty product can be appl ied at any p H ( D o y l e and Parsons, 2002; A d n a n et al, 2003a). 2.3.3 Supersaturation ratio (SSR) The supersaturation ratio ( S S R ) o f a solut ion is an indica t ion o f the potential o f struvite precipi tat ion. The S S R can be defined by the f o l l o w i n g equation. 10 S S R = P S / P S _ , ( E q . 4) Where , P s = Cond i t i ona l so lub i l i ty product o f struvite in a solut ion, Ps-eq = P s value under the equ i l i b r ium condi t ion . Genera l ly , S S R > 1 impl ies that the solut ion is supersaturated and precipi ta t ion o f struvite is possible; S S R = 1 indicates that the system is in equ i l ib r ium; wh i l e S S R < 1 means that precipi tat ion is not possible and the system is undersaturated. A d n a n et al. (2003a) reported that, when the in-reactor supersaturation ratio was maintained between 1 and 5, the system's eff iciency and quali ty o f harvested product were at their best. 2.4 Parameters Of Struvite Crystallization The mechanisms o f occurrence and development o f struvite crystals f o l l o w two chemica l stages: (i) nucleat ion (crystal birth) and (ii) crystal growth (development o f crystals unt i l equ i l i b r ium is reached). Predic t ing or con t ro l l ing these reactions is complex , since these are control led by several phys ica l -chemica l parameters. These parameters are described be low. 2.4.1 pH The p H at w h i c h struvite may precipitate is one o f the m a i n factors inf luencing the crys ta l l iza t ion process. It has been found that struvite 's so lub i l i ty decreases wi th increasing p H , unt i l it reaches the m i n i m u m value. The p H , at w h i c h m i n i m u m solubi l i ty o f struvite occurs, varies w i th the characteristics o f wastewater. Buchanan et al. (1994) reported that the m i n i m u m solubi l i ty o f struvite occurred at the p H o f 9.0; wh i l e Oh l inge r et al. (1998) found this value at a p H o f 10.3. A saturated condi t ion o f a so lu t ion is a prerequisite for any crys ta l l iza t ion process. In case o f struvite, the solut ion can be saturated either b y increasing the struvite constituent ions 11 or by increasing the p H . Increasing the p H seems to be more preferable i n running the process. The op t imum operational p H value for different wastewater varies greatly depending on the characteristics o f particular waste stream (Fujimoto et al, 1991; M t i n c h and Bar r , 2001; Stratful et al, 2001). R e c o m m e n d e d values o f p H for struvite crys ta l l iza t ion generally ranges f rom 8.0 to 9.0. Howeve r , a h igh in i t ia l p H can be l i m i t i n g in the sense that it causes the transformation o f N H 4 + ions into gaseous ammonia , thereby reducing the ni trogen concentration and affecting the molar ratio o f M g : N : P that is necessary to fo rm struvite. A l t h o u g h most o f the literatures ci ted p H values between 8.2 and 9.0, to ensure higher (above 80%) phosphorus removals (Batt is toni et al. 2001; M u n c h and Barr , 2001 ; Stratful et al. 2001 ; Jaffer et al. 2002), A d n a n et al. (2003a) were able to achieve over 9 0 % phosphorus removal at a p H o f 7.3. Fattah (2004) also reported over 9 0 % phosphorus removal at a p H o f 7.5. A d n a n et al. (2003a) conc luded that this f inding indicates that p H is not the on ly factor that can cause the process f l u id to be supersaturated; the concentration o f struvite constituent ions also plays a role. 2.4.2 Magnesium to phosphorus molar ratio Theoret ica l ly , struvite format ion requires a M g : N : P molar ratio o f 1:1:1. In most cases wi th mun ic ipa l wastewater, magnes ium is the l i m i t i n g element and hence requires an external source o f magnes ium supplementation. A d n a n et al (2003b) reported that, at a f ixed N : P molar ratio and a g iven p H value, the average phosphorus remova l increased almost l inear ly w i th an increase i n the M g : P molar ratio. Katsuura (1998) also found that phosphorus removal increased wi th an increase i n M g : P and the increase was more pronounced at l o w p H values ( p H at 8.0) compared to h igh p H values ( p H at 9.0). Howeve r , Fattah (2004) reported no significant affect o f the M g : P ratio, regarding phosphorus remova l and highl ighted problems o f keeping a constant M g : P inside the reactor. A l t h o u g h , a M g : P ratio o f 1:1 is required for struvite formation, Jaffer et al. (2002) suggested a ratio o f 1.3:1 at fu l l scale l eve l , i n order to out-compete c a l c i u m ions present i n the centrate. 12 2.4.3 Ammonia to phosphorus molar ratio M u n i c i p a l wastewaters usual ly contain higher molar concentrat ion o f a m m o n i u m ions than magnes ium and phosphorus. S o m e studies have reported that phosphorus removal increases wi th increase i n ammon ia concentrations ( M u n c h and Barr , 2001 ; Katsuura , 1998). Stratful et al. (2001) found that excess a m m o n i u m ion is h igh ly beneficial for struvite precipi tat ion and tends to fo rm a relat ively pure struvite. 2.4.4 Temperature Temperature affects so lub i l i ty and the reaction rate o f struvite formation. A a g e et al. (1997) showed an increase i n the so lubi l i ty product o f struvite w i t h an increase i n temperature, between 10°C and 5 0 ° C . These results agreed w i t h previous f indings. A d n a n et al. (2004) found s imi la r results; their data showed that lower temperature was more beneficial for actual struvite formation, w i th the product c o m i n g out o f so lu t ion more efficiently. 2.4.5 Mixing energy (or turbulence) M i x i n g energy or turbulence influences struvite accumulat ion. In their study, Ohl inger et al. (1999) showed that crystal size and shape were inf luenced by m i x i n g energy as they found that elongated crystals were formed i n a semiquiescent environment; o n the other hand, h igh m i x i n g energy resulted in the format ion o f less elongated and more t ight ly formed crystals. In an addi t ion to that, Ohl inger et al. (1999) demonstrated that the growth rate was also influenced by m i x i n g energy as the lowest growth rates were found i n the quiescent zone o f the process, and the highest ones were found in h igh m i x i n g environments. Howeve r , too m u c h turbulence m a y hinder growth by increasing the c o l l i s i o n among pellets and thereby breaking the structure (Durrant et al, 1999). 13 2.4.6 Presence of foreign ions The presence o f ions other than M g + 2 , N H 4 + and PO4 3" can increase the so lub i l i ty o f struvite by reacting w i t h any o f these three species. Koutsoukos et al. (2003) showed that the presence o f c a l c i u m ions affects the growth rate negat ively and can lengthen the induct ion t ime. C a l c i u m ions can also interact w i t h phosphate, to f o r m c a l c i u m phosphates. 2.4.7 Initial reactor seeding Struvite precipi tat ion requires a nucleus. Hence , the reactor has to be seeded at the start-up o f the process and the operating process w i l l eventually become self-seeding ( M u n c h and Barr , 2001). Types o f seeding materials have an impact on the react ion rate and removal efficiencies. W a n g et al. (2006) tested three different seeding materials (quartz sand, granite and struvite pellets), and found that struvite pellets, as seeding material , produced the best performance, regarding phosphorus removal . 2.5 Operating Costs For Struvite Production Phosphorus recovery, through struvite product ion, offers both environmental and economica l benefits. H o w e v e r , the success o f in t roducing struvite c rys ta l l iza t ion processes i n wastewater treatment plants w i l l mos t ly depend o n its economica l sustainabil i ty. P roduc t ion costs o f struvite ma in ly depend on two factors, the costs o f chemicals to be injected i n the process and energy requirements for pumping . In their study, Jaffer et al. (2002) showed that compared to cost o f chemicals , the energy requirement cost is quite insignificant . 14 Bas i ca l l y , two types o f chemicals are used i n a struvite crys ta l l iza t ion process -(i) a source o f magnesium, usual ly magnes ium chlor ide ( M g C ^ ) , to obtain a suitable M g : N : P ratio for struvite precipitat ion, and (ii) a caustic chemica l , most ly sod ium hydroxide ( N a O H ) , to raise/adjust p H o f wastewater so that saturated condi t ion for struvite formation prevails i n solut ion. Jaffer et al. (2002) estimated product ion costs o f struvite precipi tat ion for a pi lot scale project treating 400 m 3 / d o f centrate and found that 9 7 % o f the total chemica l cost was due to the addi t ion o f N a O H . Th i s result shows that, by reducing caustic use, the overa l l product ion costs o f struvite can be reduced by a large fraction. In their study at the T r e v i s o wastewater treatment plant, Italy, C e c c h i et al. (2003) used air s tr ipping to raise p H and they were able to obtain a p H value o f 8.5. The results demonstrated that air s tr ipping c o u l d be an opt ion to reduce costs associated w i t h caustic addit ion, depending o n the type o f air s t r ipping process. 2.6 Methods Of Raising The p H Value Of Wastewater A s ment ioned i n Sect ion 2.4.1, a struvite crys ta l l iza t ion process requires a saturated cond i t ion o f so lu t ion , w h i c h can be attained either b y increasing concentrations o f struvite constituent ions (e.g. M g 2 + , N H 4 + and PO4 3" ) or by increasing the solut ion 's p H . L o g i c a l l y , increasing constituent ions ' concentration is not feasible or desirable for running the process; hence, increasing the solut ion 's p H value is more appropriate. Struvite is soluble i n acidic condi t ion and becomes more and more insoluble under a basic environment, unt i l it reaches its m i n i m u m so lubi l i ty (which depends o n several factors). The p H required for struvite format ion usual ly ranges i n between 8.0 and 9.0 ( M u n c h and Barr , 2001 ; W a n g et al, 2005), depending on loca l condi t ions. Howeve r , i n most cases, the p H of m u n i c i p a l wastewater is lower than that required. Hence, it is usual ly necessary to raise the p H value o f wastewater i n order to precipitate struvite. U s u a l l y two methods are employed - these are caustic chemica l addi t ion and aeration/C02 s tr ipping. 15 2.6.1 Caustic chemical addition Genera l ly , the operating p H leve l o f struvite precipi tat ion is achieved and maintained by adding a caustic chemica l . U s u a l l y N a O H , M g ( O H ) 2 , M g O and C a ( O H ) 2 are used to raise the p H . The addi t ion o f M g ( O H ) ? and M g O resulted i n increasing both the p H value and magnes ium content o f the solut ion. B e a l et al. (1999) tested four caustics ( N a O H , N a H C 0 3 , C a ( O H ) 2 and M g O ) as a p H adjuster and found M g O to be the most cost effective, based both on cost and the fact that it p rov ided magnes ium for the reaction. The M g O proved to be effective at increasing the p H of the swine waste slurry to approximately 8.5. However , due to l imi ted so lubi l i ty , it was diff icul t to reach a p H greater than 8.5 w i t h M g O . T o test the effect o f p H above 8.5, they had to add N a O H , i n addi t ion to M g O . A cheaper poss ib i l i ty for increasing the p H o f wastewater is the addi t ion o f l ime (Ca(OH )2 ) . H o w e v e r , c a l c i u m addi t ion means an excess o f c a l c i u m ions for precipi tat ion and not o n l y fo rming c a l c i u m phosphate, but also c a l c i u m carbonate, thus reducing the value o f the recovered product. Furthermore, l i m e interferes w i t h the precipi tat ion o f struvite ( K o h e n and K i r c h m a n n , 2004; L e Cor re et al., 2005). The most c o m m o n l y used caustic is N a O H . A d d i n g l ime introduces impur i ty ions i n the fo rm o f C a 2 + . O n the other hand, adding M g ( O H ) 2 and M g O means either magnes ium concentration or the p H value cannot be op t imized independently; moreover, these chemicals dissociates rather s l o w l y , thus requir ing longer hydrau l ic retention t ime and consequently a larger reactor. Therefore, al though N a O H is relat ively more expensive than the above-ment ioned options, it proves to be more suitable i n ra is ing the p H o f the wastewater (Huang , 2003). T h e a lka l in i ty i n wastewater is generally h igh , thus impos ing a large buffering capacity. In order to raise wastewaters' p H , this buffering capacity needs to be overcome, generally ended up i n us ing large amount o f caustic (Batt is toni et al., 1997). F o r a p i lo t scale 16 experiment, Jaffer et al. (2002) calculated that the da i ly cost o f N a O H addi t ion w o u l d be £ 1 3 9 or £ 5 0 , 7 3 5 per year. Hence , for a fu l l scale plant, the cost required for p H adjustment by adding caustic chemicals , needs to be addressed seriously, before design and costruction. 2.6.2 Aeration or CO2 stripping In wastewater treatment plants, struvite precipi tat ion usual ly occurs at locations where CO2 is stripped f rom the solut ion, w h i c h is l i nked w i t h a corresponding increase i n p H . The f o l l o w i n g equation explains h o w CO2 s tr ipping raises the p H o f a solut ion. HCO3 - * CO2T + O H " ( E q . 5) Areas o f h igh turbulence, such as pipe e lbows, mixe r blades, valves, and pumps are ma in locations o f struvite deposits (Neeth l ing and Ben i s ch , 2004). In these locations, a reduction o f partial pressure o f CO2 takes place. Hence , Loewentha l et al. (1994) concluded that partial pressure o f CO2 is one o f the d r iv ing forces for struvite precipi tat ion. In their study, P i tman et al. (1991) demonstrated the poss ib i l i ty o f increasing p H w i t h aeration. Th i s phenomenon was attributed to CO2 s tr ipping. In a different study, Loewen tha l et al. (1994) showed that the partial pressure o f CO2 controls struvite precipi tat ion inside an anaerobic digester. Bat t is toni et al. (1997) carried out experiments on real anaerobic supernatant, to investigate struvite crys ta l l iza t ion without adding any chemicals to raise the p H . T h e y used two modes o f air strippers, namely external gradual aeration ( E G A ) and external continuous aeration ( E C A ) . The upf low rate ranged f rom 1.8 L / m i n to 5 L / m i n and the a i r f low rate was kept at 15 L / m i n . A i r s tr ipping increased the p H f rom 7.9 to 8.3-8.6. However , the E G A condi t ion (air s t r ipping for 5 6 % o f the total time) proved to be insufficient to obtain rapid phosphate remova l , w h i l e the E C A condi t ion gave faster phosphorus remova l . T h e y were able to obtain up to 8 0 % o f phosphorus removal , when an E C A condi t ion was adopted. 17 In separate study, us ing anaerobic supernatant f rom a centrifugation station wi th an average p H o f 7.7, Bat t is toni et al. (1998) were once again able to achieve the supersaturation p H for struvite cryatal l izat ion, app ly ing on ly air s tr ipping. In 2001, Bat t is toni et al. investigated a struvite crys ta l l iza t ion process i n a fu l l scale plant i n Trev i so (Italy). The average p H of the anaerobic supernatant o f the plant was 7.5.They used the air s t r ipping o f CO2 as the o n l y means to obtain the operative p H ( p H 8.3-8.7), without any addi t ion o f chemicals . The operative p H ranged f rom 8.1 to 8.7 according to the a i r f low rate and hydraul ic head employed , respectively, f rom 20 to 40 m 3 / h and f rom 1.7 to 2.7 m . T h e i r results showed that, i n order to obtain an operative p H > 8.3, a higher head level (2.7 m) must be used at a m i n i m u m ai r f low rate (10 m /h) or lower head one (1.7 or 2.2 m) at higher a i r f low rate (40 m 3 / h ) . Th i s means that higher hydrau l ic head can save energy, as it requires a lower a i r f low rate to achieve the desirable/operative p H . H i r o y u k i and T o r u (2003) used 1 L bench scale reactors to demonstrate the affect o f aeration on phosphorus precipitat ion. A l o n g wi th aeration, N a O H solut ion was also fed cont inuously to increase the p H . T h e y found that, by increasing the aeration intensity f rom 2.1 m g / L to 10.5 m g / L , the rate o f phosphorus removal was also increased. The authors concluded that aeration influenced the quantities o f CO2 in solut ion, w h i c h helped i n ra is ing p H and resulted i n the increased rate o f phosphorus removal . Z h a n g (2006) carried out an experiment us ing real anaerobic supernatant at the L u l u Island Wastewater treatment Plant ( L I W W T P ) . She used a cascade stripper to increase the centrate's p H . Z h a n g (2006) reported a 4 6 % to 6 5 % reduct ion i n caustic use (depending o n operating condi t ions) , by s t r ipping CO2 f rom the centrate, without a sacrifice i n struvite product ion or quality. A l l the above-mentioned findings indicate that air/C02 s t r ipping presents an alternative solut ion for increasing the p H o f wastewater and thereby reducing the use o f caustic chemicals , for struvite precipitat ion; this in turn, can reduce overa l l product ion costs 18 of the struvite crys ta l l iza t ion process, depending on the energy consumpt ion used i n the str ipping process. 2.7 Mechanisms Of pH Increase By Stripping CO2 Increasing the p H o f water by the removal o f CO2, is a w e l l - k n o w n reaction ( E q . 5). Th i s reaction takes place, naturally, i n aquatic environments through the uptake o f CO2 by algae. It has been shown that photosynthetic C O ? uptake by algae can induce a p H increase up to a level o f 10.5 ( K o h e n and K i r c h m a n n , 2004). A s described in Sec t ion 2.6.2, s tr ipping C 0 2 f rom wastewater, us ing other methods (usually stripper) than photosynthetic uptake, has the potential to increase the p H to the same h igh leve l . The carbonate system for s imple aqueous solutions can be descr ibed by the interdependent nature o f s ix solute components, namely C 0 2 , H2CO3, H C O 3 " , CO3 2", H + , and O F T , using a set o f equations g iven be low ( S t u m m and M o r g a n , 1996). Where , K i and K 2 = Firs t and second acidi ty constants respectively; dependent on temperature and presence o f other salts i n the l i q u i d solut ion. K w = Di s soc ia t ion constant for water = [H + ] [OH~] = 10" 1 4 A t neutral p H , the majority o f the inorganic carbon present i n water is i n the fo rm o f HCO3", w i th the rest o f the inorganic carbon present ma in ly as H2CO3* (CO2 + H 2 C 0 3 ) . The equ i l i b r ium for the reaction: H 2 0 + C02(aq) = H 2 C 0 3 (aq) lies rather far to the left, and the greater fraction o f un ion ized CO2 is present i n the fo rm o f d isso lved C02(aq) ( S t u m m and M o r g a n , 1996). R e m o v i n g CO2 f rom the system w i l l induce an increase i n the p H , according to the f o l l o w i n g mechanisms ( K o h e n and K i r c h m a n n , 2004). [H2CO3*] = [COa (aq)] + [ H 2 C 0 3 ] [ H + ] [ H C 0 3 ] / [ H 2 C 0 3 * ] = K 1 [ H + ] [ C 0 3 2 " ] / [ H C 0 3 " ] = K 2 [ H + ] [ O H ] = K w ( E q . 6) ( E q . 7) ( E q . 8) ( E q . 9) 19 A t l o w p H , when CO2 is removed f rom the system, new CO2 w i l l be formed ma in ly by the dissocia t ion o f HCO3" to C 0 2 , i n order to mainta in the equ i l ib r ium. HCO3" + H + -> CO2T + H 2 0 ( E q . 10) A t h igh p H , the concentration o f H + is l o w , and C 0 2 is formed f rom HCO3" through the reaction w i t h water. In both cases, due to either deplet ion o f H + ( E q . 10) or generation o f O H " ( E q . 11), the p H value o f the l i q u i d is elevated. 2.8 Ammonia Stripping The air s t r ipping o f ammon ia f rom wastewater requires that the ammon ia be present as a gas. A m m o n i u m ions i n wastewater exist i n equ i l ib r ium w i t h gaseous ammonia . A s the p H o f the wastewater is increased above 7, the equ i l i b r i um is shifted to the left and a m m o n i u m ion is converted to ammonia , w h i c h may then be removed by air/gas str ipping. H o w e v e r , H e n r y ' s l aw constant o f ammon ia is on ly 0.75 a tm ( m o l H ^ O / m o l air) , w h i c h makes this compound marg ina l ly strippable (Tchobanoglous , 2003). The result obtained by M u s v o t o et al. (2000) was i n agreement wi th this fact. T h e y compared the str ipping rates o f ammonia and C 0 2 , and found that the s tr ipping rate for CO2 was higher by two orders o f magnitude than that for ammonia . T h i s happens as H e n r y ' s law constant for ammonia is m u c h lower than that for CO2 (the dimensionless H e n r y ' s l aw constant for ammonia and CO2 are 0.011 and 0.95 respectively). HCO3" + H 2 0 H2CO3 + O H " ( E q . 11) ( E q . 12) H 2 C O 3 -> CO2T + H 2 0 N H 4 + «-> N H 3 + H + ( E q . 13) 20 The rate o f ammonia str ipping is inf luenced by p H , temperature, relative ammonia concentrations, and agitation o f air-water interface. Theore t ica l ly , the greater these values are, the more efficient the str ipping w i l l be. U s u a l l y , as the temperature decreases, the amount o f air required increases s ignif icant ly for the same degree o f r emova l (Tchobanoglous, 2003). A g a i n , care should be taken as too h igh an air to water f l o w rate can result i n a c o o l i n g effect ( L i a o et al., 1995). C h e u n g et al. (1997) investigated the effectiveness o f ammon ia s t r ipping f rom leachate at different air f l o w rates (0, 1 and 5 L / m i n ) and l i m e dosages (0 and 10,000 m g / L C a ( O H ) 2 ) i n aeration tanks i n a laboratory. T h e y found the ammonia remova l at 2 0 ° C , after one day, was 7 0 % for 0 L / m i n , 8 1 % for 1 L / m i n and 9 0 % for 5 L / m i n , regardless o f the or ig in o f leachate. The p H o f aeration tanks was raised to above 11, by adding 10,000 m g / L C a ( O H ) 2 before str ipping. B o n m a t i and Flotats (2003) studied the effect o f p i g slurry waste type, fresh or anaerobical ly digested, and the effect o f in i t ia l p H on ammonia air s t r ipping f rom p i g s lurry waste at h igh temperature ( 8 0 ° C ) . The p H o f fresh and anaerobical ly digested s lurry was 7.5 and 8.4, respectively. F o r str ipping purposes, the p H was adjusted us ing C a ( O H ) 2 . A n isothermal wet w a l l glass c o l u m n was used for air s tr ipping tests. A i r was suppl ied by an air b lower . A i r and l i q u i d f lows were set at 20 and 0.266 m L / m i n , respectively. W h e n fresh slurry was used, a f inal (after 4 h) ammon ia str ipping eff ic iency o f 65, 69 and 98 .8% was recorded for non-modi f ied p H (i.e. pH=7.5) , in i t i a l pH=9.5 and in i t ia l pH=11.5 , respectively. O n the other hand, when anaerobical ly digested slurry was used, f inal ammon ia removal efficiencies above 9 6 % were reported i n a l l experiments, regardless o f the in i t ia l p H (modi f ied or non-modif ied) . In their experiment, Jaffer et al. (2002) found that the molar removal o f ammonia exceeds the molar removal o f phosphorus and the molar usage o f magnes ium. The removal was, i n fact, greater than that needed for struvite formation. T h e authors be l ieved that the surplus ammon ia was being removed f rom the reactor by air s tr ipping. 21 CHAPTER THREE METHODS AND MATERIALS 3.1 Lulu Island Wastewater Treatment Plant This study was carr ied out at the L u l u Island Wastewater Treatment Plant ( L I W W T P ) , R i c h m o n d , B C . The L I W W T P is a secondary wastewater treatment plant that is operated by the Greater Vancouve r Reg iona l Dis t r ic t ( G V R D ) o f the province o f B r i t i s h C o l u m b i a . The plant started as a pr imary treatment plant i n 1973 and was later be ing upgraded to a secondary system i n 1999, b y the addi t ion o f a T r i c k l i n g F i l t e r /So l ids Contac t process. Presently, it has a f l ow capacity o f 155 M L D and serves around 170,000 residents o f R i c h m o n d . The plant consists o f phys ica l treatment that includes screening, grit r emova l and pr imary sedimentation; and b io log ica l treatment that includes t r i ck l ing filter, solids contact and secondary c lar i f ica t ion. The sludge produced is thickened by us ing both gravi ty thickener and diffused air floatation tanks. These two thickened sludge streams are m i x e d and then fed to the anaerobic digesters. Af te r 32 days retention t ime i n the digesters, the digested sludge has been reduced f rom ~ 4.0 % vola t i le solids to - 1 . 7 % sol ids . The digested sludge is then fed to centrifuges where the solids are further thickened to ~ 25 % (Internet2). The l i q u i d centrate generated i n the dewatering process was used as the i n f low to the two struvite crystal l izers used i n this study. The characteristics o f the centrate dur ing the study per iod are g iven i n the next chapter. 3.2 Materials And Equipment 3.2.1 Struvite Crystallizer T w o ident ical struvite crystal l izers were used i n this project. The reactors are referred to as the R # l and the R#2, respectively, throughout this chapter and the next two chapters. Th i s struvite crysta l l izer mode l was developed by the Department o f C i v i l Eng ineer ing at the 22 University of British Columbia. The process consists of a reactor, an external clarifier, storage tanks for centrate, magnesium feed and caustic, pumps for feed flow, recycle flow, magnesium and pH controller. In this study, a stripper was incorporated within the system. The basic flow diagram of the crystallizer process (excluding the stripper) is shown in Figure 3.1. Seed hopper 15" Reaction zone 6" Active zone Feed bypass p • pH probe Harvest zone 4" 3" Check valve Recycle • -j bypass Recycle line Clarifier Effluent Recycle Centrate •> Sludge Figure 3.1: Basic flow diagram of the struvite crystallizer process The reactor has four distinct zones, increasing in diameter from bottom to top. For a given upflow velocity, each section has a different flow rate, decreasing from bottom to top. 23 The dimensions o f different zones are g iven i n Tab le 3.1. The var ia t ion o f reactor diameter w i t h height provides a certain degree o f turbulence above each transit ion w h i c h , i n turn, ensures sufficient m i x i n g i n each zone. A s pellets g row i n size, they overcome higher upf low veloci t ies and m o v e d o w n towards the lower sections where h i g h turbulence enhances further crystal growth (Ohl inger , 1999). Thus , on ly the largest pellets i n the reactor are harvested. Table 3.1: D imens ions o f the reactor Section Length (cm) Diameter (cm) Volume (L) Harvest zone 74.9 7.6 3.42 A c t i v e zone 154.9 10.2 12.56 Reac t ion zone 127.0 15.2 23.17 Seed hopper 45.7 38.1 52.12 B e l o w harvest zone 52.1 - -A p H probe is inserted into the top part o f the harvest zone. This probe is used to control the p H control ler system. The control ler is capable o f mainta ining the p H w i t h i n ± 0 . 1 p H units. Ano the r p H probe is kept i n the external clarifier, for countercheck. The feed and caustic so lu t ion come f rom four different streams - centrate tank, recycle f l ow , magnes ium chlor ide ( M g C L ) solu t ion f rom the dos ing pump and s o d i u m hydroxide ( N a O H ) solut ion f rom the p H controller . A n injection port, made o f stainless steel, is bui l t at the bot tom o f the harvest zone, i n order to provide complete m i x i n g o f the feed constituents before entering the reactor. A more detailed descr ipt ion o f the reactor is p rov ided by Fattah (2004). Chemicals, Storage Tanks And Pumps A s stated previously , centrate f rom the L I W W T P was used in this study. A t the beginning o f the project, the centrate was stored i n two 5600 L capacity ho ld ing tanks. O n many occasions, the plant fai led to provide sufficient amount o f centrate on a regular basis, due to operational problems. A s a result, the reactors had reverted to recycle mode 24 frequently, thus hamper ing the struvite crys ta l l iza t ion process. Later on, another tank o f the same capacity was instal led, to provide more storage for centrate. The supplementary magnes ium feed for this project was made f rom commerc i a l grade magnes ium chlor ide ( M g C l 2 . 6 H 2 0 ) . The solut ion was stored in a 1400 L capaci ty ho ld ing tank. A Mas te r F l e x dos ing pump was used to pump the solut ion to the injection ports o f the reactors. The struvite crys ta l l iza t ion process is h igh ly dependent on p H . The format ion o f struvite is associated w i t h a subsequent drop i n the p H o f the system; hence, it is necessary to return the p H to the required value. In this project, the p H was adjusted us ing a caustic solut ion. T h e caustic was made f rom N a O H and was stored i n a 120 L tank. A carbon d iox ide ( C 0 2 ) trap was used to strip off C 0 2 f rom the air, before the air entered the caustic tank. The p H o f the system was moni tored and control led by the p H probe inserted into the top o f the harvest zone. Th i s p H meter and the p H meter connected to the external c lar i f ier were regular ly calibrated by the two point method, us ing standard buffer solutions o f p H 7.0 and p H 10.0. The set up o f the study area is shown in F igure 3.2. 25 Figure 3.2: T h e set up of struvite crystall izers (wi th strippers) at the L I W W T P 3.1.1 Compact Media Stripper T w o strippers were used i n this project. One o f the strippers was a compacted media type o f stripper. T h i s stripper was connected wi th the R # l . The compact media stripper was developed by Ostara Nutr ient Recovery Technologies Inc., Canada. T h e stripper was designed for a m a x i m u m hydraul ic load ing rate o f 30 L / m i n . It consists o f a c i rcular s t r ipping tower, a supporting plate for the pack ing material at the lower part, a l i q u i d distr ibutor system located above the pack ing material and a fan at the bot tom o f the str ipping tower. T h e rate o f air p rov ided by the fan is f ixed . T h e total height o f the str ipping tower is 180 c m , o f w h i c h 120 c m was packed w i t h the pack ing material . One inch diameter h o l l o w plast ic bal ls were used as packing material . 26 Figure 3.3: P a c k i n g material o f the compact med ia stripper A t the beg inn ing o f the project, the plastic bal ls were s imp ly put inside the s t r ipping tower, in-between the l i q u i d distributor and the support ing plate. H o w e v e r , c logg ing was observed after 3 weeks o f running the stripper and as a result, the performance o f the stripper decreased. T h i s led to a modi f ica t ion i n the arrangement o f the pack ing materials. F r o m that point, the pack ing materials were hung inside the tower f rom the l i q u i d distr ibutor system, wi th the help o f strings. The basic p r inc ip le o f a i r /CO? str ipping is the mass o f gas transferred f rom the l i q u i d phase to the gas phase. T h i s transfer is accompl ished by contact ing the l i q u i d conta ining the gas that is to be str ipped of f w i th another gas, usual ly air, w h i c h does not contain the gas or contains at a lower amount than the l i q u i d in i t ia l ly (Tchobanoglous , 2003). T o achieve a h igh rate o f removal , it is required to provide enough air/water contact t ime and enough specific surface area. The h o l l o w plastic ba l l pack ing media provide both the condi t ions for the stripper. In addi t ion, a certain degree o f turbulence is generated w h i l e water passes through the pack ing media , w h i c h , i n turn, enhances the s tr ipping rate. The schematic o f the s t r ipping tower is i l lustrated i n F igure 3.4. 27 30 cm 120 cm Flange and mount to tank as required 30 cm 37 cm 74 cm Welded liquid distributor supports Welded packing retainer supports Fan mounts as required 6 cm Figure 3.4: Compac t media stripper The str ipping tower is mounted on top o f the external clarif ier . It b locks the top o f the clar if ier entirely and thereby makes the whole stripper system almost a sealed vessel, except for the 20 c m diameter opening at the top o f the str ipping tower. T h i s had a negative impact on the stripper's efficiency, w h i c h w i l l be discussed i n the next chapter. 28 Figure 3.5: C o m p a c t media stripper connected to the R # l at the L I W W T P 3.2.3 Cascade Stripper T h e second type o f stripper used in this project was a cascade stripper. Th i s stripper was connected w i t h the R#2.The cascade stripper was developed by Z h a n g (2006) as a part o f her M . A . S c . research carr ied out i n the Department o f C i v i l Engineer ing , U B C . T h e reasons that led Zhang (2006) to develop an external cascade stripper were as fo l lows : > T h i s k i n d o f stripper does not have p lugging problems > It is easy to b u i l d and operate > It is easy to clean, w h i c h i n turn lowers the maintenance costs The cascade stripper was designed for a m a x i m u m hydraul ic loading rate o f 20 L / m i n . A s mentioned previously , the basic pr inciple o f air s tr ipping is gas transfer and hence, the pr imary guidelines for the design o f the cascade stripper were as fo l lows : 29 > T o provide enough specific surface area. Th i s was achieved by adjusting the baffle angle o f the stripper. F o r the stripper, the baffle angle was f ixed at 10 degrees. > T o provide enough air/water contact t ime. Th i s was achieved by p lac ing reasonable number o f baffles. The m a x i m u m number o f baffles that can be placed i n this stripper is 20. F o r this project, a total o f 10 baffles was selected to be placed i n the stripper. The baffle was designed to have an effective surface area o f 15x15 c m 2 . T h e detailed design o f the baffle is shown i n F igure 3.6. 170 mm ~[X 10 mm O -*0 5mm 15 mm 150 mm 150 mm Figure 3.6: De ta i l ed design o f the baffle The cascade stripper was made o f p lexiglass . The dimensions o f the stripper and the front v i e w are shown i n F igure 3.7. 30 Dimensions of the stripper Tota l height 107 c m Externa l d imens ion : 20 c m x 2 0 c m Internal d imens ion : 19 c m x 19 c m I* 1 20cm 1 Figure 3.7: Front v i e w o f the cascade stripper The cascade stripper was incorporated into the crys ta l l izer system just before the external clarif ier . T h e stripper substituted almost 1/3 o f the reactor downpipe . The f low f rom the seed hopper first passed through the stripper and then it entered the clar if ier . F igure 3.8: Cascade stripper connected to the R#2 at the L I W W T P 31 In her study, Zhang (2006) tested the cascade stripper without an external air supply at the LIWWTP. However, in this project the stripper was run under both air and without air supply conditions, and its efficiency in saving caustic under different conditions was measured. The air was supplied through an airflow meter from the top of the stripper. The maximum airflow rate provided by the flow meter was 107.5 L/min. While running without external air supply, the top surface of the stripper was kept open. With the introduction of the airflow meter, the top surface of the stripper was covered with a plexiglass lid. 3.3 Experimental Design The project was originally planned to test one reactor with one of the stripper and use the other reactor as a control, under different conditions and then switch the case and control; then, repeat the same procedure for the second stripper. However, due to time limitation, as well as previous data obtained, this plan was modified and both the reactors were run with strippers. The efficiency of the strippers was determined by comparing the amount of caustic saved by each of them. Before making any comparison between the strippers' efficiency, we had to make certain that the two reactors, with which the strippers were connected, performed identically, so that a direct comparison could be made between the strippers' performance. Therefore, at the beginning of the study, both the reactors were run in parallel without strippers. The reactors were tested under four different conditions. These are tabulated in Table 3.2. The conditions were the same for both the reactors. Table 3.2: Test condition (for both reactors) Stripper Air Recycle Ratio Upflow velocity (cm/min) Run #1 X X •6 400 Run #2 • V V 6 400 Run #3 V X 6 400 Run #4 9 450 32 Throughout the course o f the study, the operating condit ions were selected us ing the Pot ts ' Crys ta l l i ze r M o d e l (Potts, 2002). F r o m k n o w n values o f centrate's magnes ium, ammon ia and phosphate concentrations, temperature, conduct iv i ty , upf low ve loc i ty , feeding/recycle rates, and recycle ratio, this mode l can calculate the amount o f M g O ? to be added into the reactor to obtain a desired M g : P ratio and the operating p H , for a desired reactor supersaturation ratio. T h e m o d e l also predicts the effluent concentrations o f the three m a i n species o f interest (magnesium, ammon ia and phosphate). B e i n g a cont inuat ion o f the previous study conducted at the L I W W T P , the centrate characteristics obtained by Z h a n g (2006) were used to determine the p re l iminary operating condit ions. H o w e v e r , struvite format ion is a h i g h l y dynamic process and change i n any o f the parameters w i l l have a significant affect on the process. Hence , for each run afterwards, the most recent data were used for selecting operating condi t ions, as the centrate characteristics fluctuated cont inuously . The characteristics o f the centrate dur ing the project per iod is g iven i n Table 4.1 and shown graphical ly i n Figures 4.1 to 4.3. Several parameters were moni tored and recorded each day. These include grab samples o f centrate and effluents for determining M g 2 + , N H 4 " and PO43" concentrations; p H , temperature and conduct iv i ty o f the centrate and effluents; feed and recycle f lows were measured da i ly and adjusted ( i f required); caustic samples f rom both tanks were col lected and amount o f caustic used by the strippers were recorded; a sample f rom the magnes ium feed tank, whenever new feed was prepared; and C O 2 samples f rom seed hoppers and effluents. 3.4 Sample Collection, Storage And Preservation Grab samples o f centrate and effluents were col lected f rom the centrate tank and effluent f lows , respectively. F o r magnes ium feed, the sample was col lected f rom the magnes ium storage tank, whenever new feed was prepared. The p H , temperature and conduct iv i ty o f the samples were recorded on-site, us ing a H o r i b a D 5 4 portable meter. 33 F o r measurement o f ions, the samples were pre-filtered through 1.25 u m glass fibre filter papers and f inal f i l t rat ion was done by us ing 0.45 microns membrane filter papers (Standard Methods) . F o r the N H 4 + and PO43" test, one drop o f 5% v /v sulfuric ac id was added to 2 m L o f filtered sample to lower the p H be low 2. F i v e drops o f 5 0 % ni t r ic ac id was added to 15 m L o f fil tered sample to preserve the metal samples. The samples were kept at 4 ° C unt i l analysis. Caus t ic samples were col lec ted f rom the t w o caustic tanks each day. T h e v o l u m e o f caustic used by the two strippers was recorded da i ly . The samples for CO2 measurement were col lec ted f rom the seed hopper and effluent f l ow. Sample bottles were f i l l ed to the b r i m o f the bottles and then the l ids were c losed t ightly, to prevent p ick-up o f CO2 f rom the air. The d isso lved C 0 2 o f the samples was measured immedia te ly up on ar r iv ing at the Env i ronmenta l Engineer ing Laboratory, U B C . 3.5 Analytical Methods A l l analyses were carried out i n the Env i ronmenta l Engineer ing Laboratory, U B C . 3.5.1 Magnesium M a g n e s i u m analysis was performed by the f lame atomic absorption spectrophotometry us ing a V a r i a n Inc. S p e c t r A A 2 2 0 Fast Sequential A t o m i c A b s o r p t i o n Spectrophotometer ( A A S ) . Instrument operational parameter details are g iven in A p p e n d i x A . The samples were digested wi th ni t r ic ac id before running into the A A S . Th i s was done to reduce dissolve organic matters that interferes w i t h analysis. F o r digest ion, 2 m L 5 0 % ni t r ic ac id was added to 10 m L of sample. Diges t ion was carr ied out o n a hot plate unt i l the sample vo lume was reduced to approximately 1 m L and the color o f the sample got cleared. The sample was then brought back to or ig ina l 10 m L b y adding dis t i l led water. D i l u t i o n o f the sample was carr ied out, i f needed. 34 3.5.2 Ortho-phosphate and ammonia Ortho-phosphate and a m m o n i a samples were analyzed us ing f l o w inject ion analysis o n a Lachat Q u i c k C h e m 8000 instrument. Instrument operational parameters are p rov ided i n A p p e n d i x A . 3.5.3 Calcium, aluminum, iron and potassium C a l c i u m , a luminum, i ron and potassium analyses were undertaken to get the compos i t ion o f the harvested struvite pellets. A l l analyses were done by atomic absorption spectrophotometry, us ing the same instrument as for the magnes ium described earlier. Instrument operational parameter details are g iven i n A p p e n d i x A . 3.5.4 Caustic analysis Caus t ic solutions were made f rom N a O H pellets. Caus t ic samples were col lected everyday f rom the two caustic tanks and were analyzed for the sod ium ion present i n the sample. The sod ium i o n present i n the sample gave the indica t ion o f the amount o f hydroxide i o n present i n the solut ion, as one mole o f sod ium ion reacts w i t h one mole o f hydrox ide ion , to produce one mole o f N a O H . T h e caustic so lu t ion was prepared at the project site, us ing hot water suppl ied at the L I W W T P . The reason for us ing hot water instead o f c o l d water was that the hot water at the L I W W T P contained lesser amount o f suspended solids than the c o l d water. The hot water sample was col lected whenever a new solut ion was made and was analyzed for sod ium ion . The sod ium concentration, obtained f rom the water sample, was deducted f rom the concentration obtained f rom the caustic solut ion sample, to obtain the actual amount o f sod ium ion associated wi th the hydroxide i o n i n the caustic solut ion. Th is analysis gave the concentration o f the caustic solut ion. The vo lume o f the caustic used by the reactors was recorded da i ly . F r o m the da i ly vo lume usage and the concentration o f the solut ion, the mass of caustic used each day by the reactors was determined. 35 S o d i u m analysis was done by the a tomic absorption spectrophotometry, us ing the same instalment as mentioned previously . Instrument operational parameter details are g iven in A p p e n d i x A . 3.5.5 Dissolved CO2 Concentra t ion o f d issolved gaseous CO2 i n l i q u i d solut ion was measured by the A c c u m e t Gas-Sens ing C o m b i n a t i o n I S E . The electrode was connected wi th a meter that gave m i l l i v o l t output. T h e direct cal ibrat ion technique was used for measuring purposes, as it requires o n l y one meter reading for measurement. The cal ibra t ion curve was prepared us ing a series o f sod ium bicarbonate ( N a H C O a ) standards namely 10" 4 M , 5 x l 0 " 4 M , 10"3 M , 5 x l O ~ 3 M , 10~2 M , 5 x l 0 " 2 M . The cal ibra t ion curve is g iven i n A p p e n d i x B . A buffer solut ion was added to the standard just before measurement i n order to convert a l l carbonates and bicarbonates to carbon d iox ide . T h e concentration o f d i sso lved C O ? i n samples was determined by compar i son to the standards. S ince it was intended to get on ly the amount o f d isso lved CO2 i n l i q u i d solut ion and not the bicarbonates and carbonates, no buffer solut ion was added to the samples. Standard and sample vo lume used for each measurement was 75 m L . F i n a l results were adjusted for temperature effect. 3.6 Pellet Quality Determination The qual i ty o f harvested pellets was determined by check ing the compos i t ion o f the pellets obtained f rom different runs performed dur ing the study per iod. Grab samples o f the harvested pellets were randomly chosen for analysis. F o r the test, 0.5 g o f struvite was d isso lved i n 50 m L o f 5 0 % ni t r ic ac id and then 50 m L o f d is t i l led water was added to the sample. Samples were analyzed for magnes ium, ammonia , ortho-phosphate, c a l c i u m , a luminum, i ron and potassium. 36 CHAPTER FOUR RESULTS AND DISCUSSION 4.1 Centrate Characteristics The study was carried out at the L u l u Island Wastewater Treatment Plant ( L I W W T P ) f rom 22 A u g u s t 2006 to 13 December 2006. Centrate samples were col lec ted s ix days a week. S u m m a r y o f centrate characteristics during the study per iod is g iven i n Tab le 4.1 and shown graphica l ly i n Figures 4.1 to 4.3. Table 4 .1: Characterist ics o f Centrate (22 A u g - 13 D e c , 2006) pH Temp CO Cond (mS) Mg (mg/L) P04-P (mg/L) NH3-N (mg/L) Molar Ratio Mg:P N:P Minimum 7.2 15.3 4.11 4.11 42.6 500 0.01 18 Maximum 8.1 34 12.48 17.44 100 916 0.39 36 Average 7.6 25 7.01 9.79 77.75 782 0.17 23 8.4 X a, 7.6 7.2 6.8 • 15-Aug 4-Sep 24-Sep 14-Oct Date 3-Nov 23-Nov 13-Dec Figure 4 .1 : p H o f centrate (22 A u g - 13 D e c , 2006) 37 120 1 80 ^ 60 2 40 * 20 0 , f • $ I i "• ^ f l l • — a J» A A A m A dr A V • • • • • V * 15-Aug 4-Sep 1000 800 600 400 200 0 24-Sep 3-Nov 14-Oct Date • M g A P04-P m N H 4 - N 23-Nov 13-Dec Figure 4.2: M a g n e s i u m , phosphate and ammonia concentrat ion i n centrate (22 A u g - 13 D e c , 2006) 1 I z =3 v 15-Aug 4-Sep 24-Sep 14-Oct Date • M g : P • N:P 3-Nov 23-Nov 13-Dec Figure 4.3: M g : P and N : P ratio o f centrate (22 A u g - 13 D e c , 2006) A s expected, the magnes ium quantity i n the centrate was the l i m i t i n g factor for struvite crys ta l l iza t ion. The molar ratio o f M g : P was a lways be low 1, w h i c h is required for struvite precipitat ion. Hence , magnes ium feed was injected into the reactors, to raise the M g : P ratio. The average supersaturation ratio ( S S R ) o f centrate dur ing the study per iod was 0.96. T h i s value was calculated using the Pott 's Crys ta l l i ze r M o d e l (Potts, 2002). The average values g iven i n Tab le 4.1 were used as inputs i n the model . A m i n i m u m S S R o f 1 is required for struvite crys ta l l iza t ion. A s mentioned i n Chapter 2, the S S R can be raised by increasing 38 the solut ion 's p H (more preferable option). In this study, this was done by injecting N a O H solu t ion and by us ing strippers to strip CO2 f rom centrate. H o w e v e r , struvite format ion inside the centrate in f low pipes was not iced o n some occasions. T h i s can be expla ined by the fact that, dur ing those particular cases, centrate was saturated w i t h respect to struvite constituent ions (i.e. M g 2 + , N H / and P043"). 4.2 Performance Of The Reactors (Without Stripper) The two struvite crystal l izers used i n this project were identical i n size and shape. B o t h were seeded w i t h I L o f struvite at the beginning o f the study. It was assumed that both w o u l d perform i n almost an identical manner. Reactors were run f rom 25 September to 6 October, without strippers, to check whether the above stated assumption was correct. T h e operating condit ions dur ing this run are g iven i n Table 4.2. Table 4.2: Operat ing condit ions (25 Sep-6 Oct , 2006) To ta l feed 2.61 L / m i n Centrate f l o w 2.53 L / m i n M g feed f l o w 80 m L / m i n R e c y c l e ratio 6 R e c y c l e f l o w : 15.63 L / m i n To ta l f l o w : 18.24 L / m i n p H : 8.1 Harvest zone upf low ve loc i ty : 400 c m / m i n In this run, the r emova l rates achieved for magnes ium, ammonia and phosphorus were 6 9 % , 88% and 7% respect ively i n the R # l ; and 66%, 9 0 % and 10% respectively, in the R#2. The difference i n r emova l rates for magnes ium and phosphorus were w i t h i n 4%. F o r ammonia , except for one day, the difference was around 13%, although the absolute removals , for both, were re la t ively l o w . The results are shown graphica l ly i n Figures 4.4 to 4.6. 39 Figure 4.4: M a g n e s i u m removal rates in R# l and R#2 (25 Sep-6 Oct , 2006) 20 -r ! I 12 z B 5 4 z 0 23-Sep 25-Sep 27-Sep 29-Sep 1-Oct 3-Oct 5-Oct Date • R#l »R#2 Figure 4.5: A m m o n i a r emova l rates i n R# l and R#2 (25 Sep-6 Oct , 2006) 40 Figure 4.6: Phosphorus removal rates i n R # l and R#2 (25 Sep-6 Oct , 2006) Excep t for one day, more caustic had been used by the R # l than the R#2, dur ing this per iod. T h e average difference o f caustic used by the two reactors was 9 1 % . O n the other hand, the average p H increased was 0.65 for both reactors. T h e R # l needed an average 2.35 kg /d o f caustic to raise the p H value, wh i l e an average 1.23 kg /d o f caustic was used by the R#2, to achieve the same amount o f p H increase. D u r i n g this per iod, both reactors were r emoving almost the same amount o f molar phosphorus f rom the i n f low , w h i l e the R # l was us ing around 9 1 % more molar caustic than the R#2. The da i ly caustic use profiles are shown in Figures 4.7 to 4.9. 4.00 1B 3 0 0 I 2.00 1.00 0.00 23-Sep 25-Sep 27-Sep 29-Sep Date • R#l • R#2 1-Oct 3-Oct 5-Oct F igure 4.7: D a i l y caustic use i n R # l and R#2 (25 Sep-6 Oct , 2006) 41 1 3 u 3.50 3.00 :.50 2.00 1.50 1.00 0.50 0.00 0.9 + o.: o.-o.i 0 0 0. 0. 0 X a < H N a O H l • N a Q H 2 B A p H l H A pH2 F igure 4.8: Caus t ic use and p H increase i n R # l and R#2 (25 Sep-6 Oct , 2006) 3.00E-03 2.50E-03 2.00E-03 1.50E-03 1.00E-03 5.00E-04 0.00E+00 m m 2 3 4 Day H P I • P 2 m N a O H l 0 N a O H 2 100 80 60 -a 40 20 0 ca Figure 4.9: M o l a r P removal and caustic use i n R # l and R#2 (25 Sep-6 Oct , 2006) It can be concluded f rom this phase that both reactors performed almost ident ica l ly , i n the removal o f magnes ium, ammonia and phosphorus f rom the centrate. The major difference was observed i n the case o f caustic use, where the R # l was consistently us ing more caustic than the R#2 to increase the p H to the same level and remove almost the same amount o f phosphorus. Th i s difference might arise due to several reasons. F i r s t ly , it was very hard to keep the centrate and recycle f lows constant to the set point. The f lows fluctuated everyday and the amount o f f luctuation var ied in-between the two reactors. Change i n f lows w i l l have an effect o n the supersaturation ratio, thereby effecting 42 the struvite product ion rate; this, i n turn, w i l l have an impact on caustic use. In order to m i n i m i z e this effect, data hav ing more than ± 1 5 % fluctuation in f lows f rom the set point had been discarded, for analyt ical purposes. Secondly , both strippers were already connected to the respective reactors (compact media stripper w i t h the R # l and cascade stripper w i th the R#2) dur ing the who le test per iod. F l o w was go ing direct ly f rom the seed hopper to the clarif ier , by bypassing the strippers. A s mentioned previously , the compact media stripper was instal led di rect ly over the clar i f ier #1, thereby total ly b l o c k i n g the top o f the clarifier. O n the other hand, the top o f the clar if ier #2 was open as the cascade stripper was installed about 1 ft above the top surface; as a result, stripped of f CO2 cou ld escape through this opening. A s the compact med ia stripper sealed the R # l system, stripped of f CO2 eventual ly d isso lved back into the l i q u i d stream. A s a result o f this, recycle f l ow f rom the clar i f ier #1 carr ied a so lu t ion saturated w i t h CO2 to the R # l . T h e removal rates o f CO2 dur ing this per iod b y the two reactors support the above stated fact. O n l y 9% C 0 2 r emova l was observed i n the R # l ; on the other hand, a 16% CO2 r emoval rate was achieved b y the R#2. T h e difference between the C 0 2 r emova l rates also had an impact on caustic use by the two reactors. It w o u l d be better i f the experiment c o u l d be re-run wi thout p lac ing strippers above clarifiers i.e. keeping top surface o f both clarifiers open to the environment. Howeve r , due to a shortage o f on-site research t ime, the experiment cou ld not be repeated. Summary of the results > B o t h reactors bas ica l ly performed ident ical ly , as assumed at the beginning, and hence, performances obtained f rom the two strippers w o u l d be direct ly comparable to each other. > The on ly major difference observed was i n case o f da i ly caustic usage w h i c h was bas ica l ly related to the way the compact media stripper was designed (i.e. seal ing the crystal l izer system). Th i s was one o f the major disadvantages o f the compact media stripper design. 43 4.3 Performance Of The Strippers The performance o f the strippers was tested under three different condi t ions. The Pott ' s Crys ta l l i ze r M o d e l (Potts, 2002) was used to set the pre l iminary operating condi t ions . A l l parameters, other than the recycle ratio, air f l ow and upf low ve loc i ty , remained constant dur ing each testing per iod. 4.3.1 Run No. 1 This first, fu l l test was carr ied out f rom 13 to 18 November . In this case, the strippers were run w i t h air. The compact media stripper had a bu i l t - in fan mounted at the top o f the clarif ier . The fan p rov ided a f ixed a i r f low for venti lat ion. O n the other hand, the a i r f low rate cou ld be varied i n the cascade stripper, w i t h the help o f the a i r f low meter. D u e to shortage o f t ime at the site (Christmas shut d o w n at the plant), it was decided to run the cascade stripper wi th the m a x i m u m ai r f low provided by the f l ow meter, w h i c h was 107.5 L / m i n . The operating condit ions for this run are g iven i n Table 4.3. Table 4.3: Operat ing condit ions for R u n N o . 1 (13-18 N o v , 2006) Parameter Compact media stripper Cascade stripper Tota l feed 2.61 L / m i n 2.61 L / m i n Centrate f l o w 2.51 L / m i n 2.51 L / m i n M g feed f l o w 100 m L / m i n 100 m L / m i n R e c y c l e ratio 6 6 R e c y c l e f l o w 15.63 L / m i n 15.63 L / m i n To ta l f l ow 18.24 L / m i n 18.24 L / m i n p H 8.1 8.1 Harvest zone upf low ve loc i ty 400 c m / m i n 400 c m / m i n A i r Y e s 107.5 L / m i n 44 Nutrient removal D u r i n g this test per iod, 9 0 % phosphorus removal was achieved by both reactors. T h e average removal rate o f ammonia was 6% and 5% i n the R # l and R#2, respectively. M a g n e s i u m removal rate was almost 7 4 % i n the R # l . In the R#2, except for the last day, this rate was around 7 5 % . O n that day, o n l y 2 2 % of M g remova l was achieved. The results are shown in Figures 4.10 to 4.12. 100 -80 - * 60 -t * emovi 40 -u f 20 - . m n 12-Nov 13-Nov 14-Nov 1 15-Nov ! 16-Nov 17-Nov 18- Mov Date • R#l • R#2 Figure 4.10: M a g n e s i u m removal rates i n R # l and R#2 (13-18 N o v , 2006) 12 -8 • > O i z £ z 4 -• t m 1 • 0 • • • 12-Nov 1 3-Nov 14-Nov i | i 15-Nov 16-Nov 17-Nov 18- Mov Date • R#l • R#2 Figure 4.11 : A m m o n i a remova l rates in R # l and R#2 (13-18 N o v , 2006) 45 93 92 • "3 b> 91 -• remo 90 -• a a-• T 89 m • • — RX -M 12-Nov 13-Nov 14-Nov 15-Nov 16-Nov 17-Nov 18- Mov Date • R#l • R#2 Figure 4.12: Phosphorus removal rates i n R # l and R#2 (13-18 N o v , 2006) Caustic use Throughout the test period, the amount o f caustic used by the R#2 was consistently lower than that was used by the R # l . O n an average, the R#2 was us ing 0.84 kg /d o f caustic, whereas i n case o f the R # l , this amount was 1.41 kg /d . It should be noted that almost the same amount o f mola r phosphorus had been removed by both the reactors dur ing this per iod. The results are shown graphica l ly i n Figures 4.13 and 4.14. § 1.2 •a o.8 y 0.4 o 12-Nov 13-Nov 14-Nov 15-Nov Date • R#l • R#2 16-Nov 17-Nov 18-Nov Figure 4 .13: D a i l y caustic use i n R # l and R#2 (13-18 N o v , 2006) 46 2.58E-03 2.48E-1 2.38E-E „ £ 2.28E-a. | 2.18E-2.08E-03 03 03 03 03 1.98E-03 m 4- 40 50 30 20 10 Day H P I • P2 0 N a O H l H N a O H 2 F igure 4.14: M o l a r P remova l and caustic use i n R # l and R#2 (13-18 N o v , 2006) C o m p a r i n g these results w i t h the results obtained by running the reactors without strippers, it can be seen that, by introducing the cascade stripper into the system, a 4 6 % savings i n caustic use was achieved. O n the other hand, the compact media stripper fai led to save any caustic; i n fact, around 15% extra caustic was used da i ly by the R # l dur ing this per iod. A s mentioned earlier, the compact media stripper b l o c k e d the top o f the clar if ier #1 and turned the who le crystal l izer system into a sealed vessel. The stripped off CO2 had very little open space to escape f rom the system. T o make things worse, throughout the study period, it was found that the compact media stripper was prone to c logg ing . These two factors raised the amount o f da i ly caustic used by the stripper. It should be mentioned that the results obtained by the R#2 system, without a stripper, are be ing used a l l the t ime for compar ison. It was already discussed i n Sec t ion 4.2 that, dur ing that run, the compact media stripper was placed on top o f the external clar if ier permanently w h i c h , i n turn, gave a biased result regarding caustic use. Excep t for caustic use, other results obtained (i.e. nutrient removal) were almost identical for both reactors, conf i rming the assumption o f the two reactors being identical i n performance. Based on this fact, it can be said that the R # l probably w o u l d have used the same amount o f caustic as the R#2, i f the top o f the clar if ier was not b locked by the compact media stripper. Hence , the amount o f caustic (average 1.23 kg/d) used by the R#2 (when the reactors were tested without the strippers) is used for compar i son purposes. 47 Carbon dioxide stripping Three days o f CO2 s t r ipping data were avai lable for this run. T h e cascade stripper proved to be more effective i n s tr ipping CO2, than the compact media stripper. Throughout this run, the cascade stripper removed more CO2 f rom the system, thus l ower ing the da i ly requirement o f caustic. T h e overa l l CO2 r emoval rate for the R # l was 1 1 % , and was 2 0 % for the R#2. The results are shown graphica l ly i n F igure 4.15. 45 36 « 27 o s a i s O U Q 0 13-Nov 14-Nov 15-Nov 16-Nov Date • R#l • R#2 17-Nov 18-Nov Figure 4.15: C Q 2 r emova l i n R # l and R#2 (13-18 N o v , 2006) Ammonia stripping Theoret ica l ly , the format ion o f struvite requires a molar ratio o f M g : N : P o f 1:1:1. D u r i n g some days, the molar removal o f ammon ia exceeded that o f magnes ium and phosphorus. The remova l o f this extra amount o f ammonia f rom the system might be achieved through str ipping. F igure 4.16 shows this extra amount o f molar removal o f ammonia by the two reactors and Figure 4.17 shows the percent r emova l o f ammon ia through str ipping. 48 4.20E-03 fa ® 3.15E-03 a | 2.10E-03 Z u 3 .05E-03 O.OOE+00 Day 0N1 • N2 Figure 4.16: A m m o n i a s t r ipping i n R # l and R#2 (13-18 N o v , 2006) F igure 4.17: A m m o n i a s t r ipping i n R # l and R#2 (13-18 N o v , 2006) The compact media stripper seemed to be s l ight ly better i n the case o f ammonia str ipping, than the cascade stripper. D u r i n g this period, an extra 2 % ammon ia was removed by the R # l , w h i l e the R#2 removed 1% o f extra ammonia . Stat is t ical ly, this difference is probably not significant. Howeve r , the higher amount o f caustic usage by the R # l cou ld be related to the removal o f ammonia , since the p H o f the system w o u l d decrease. 49 Summary of results > 9 0 % phosphoais removal was obtained by both reactors. > The cascade stripper (R#2) consistently used lesser amount o f caustic than the compact media stripper ( R # l ) . The average difference between the usage rates was around 6 8 % . > 4 6 % of caustic was saved by the cascade stripper. O n the other hand, the compact media stripper fa i led to save any caustic; rather it used an extra 13% caustic. > A higher amount o f C O a was stripped by the cascade stripper. The compact media stripper seemed to w o r k better i n the case o f ammon ia removal . B o t h o f these factors p layed a role i n case o f the caustic use pattern by the two reactors. 4.3.2 Run No. 2 T h i s test was conducted f rom 20 N o v e m b e r to 1 December . In this case, the strippers were run without an external air supply. The other operating parameters remained the same as i n the first a i n . The operating condit ions for R u n N o . 2 are g iven i n Tab le 4.4. Table 4.4: Operat ing condit ions for R u n N o . 2 (20 N o v -1 D e c , 2006) Parameter Compact media stripper Cascade stripper Tota l feed 2.61 L / m i n 2.61 L / m i n Centrate f l ow 2.51 L / m i n 2.51 L / m i n M g feed f low l O O m L / m i n 100 m L / m i n R e c y c l e ratio 6 6 R e c y c l e f l ow 15.63 L / m i n 15.63 L / m i n To ta l f l ow 18.24 L / m i n 18.24 L / m i n p H 8.1 8.1 Harvest zone upf low ve loc i ty 400 c m / m i n 400 c m / m i n A i r N o N o 50 Nutrient removal B o t h reactors were able to remove around 9 0 % o f phosphorus dur ing this period. Howeve r , significant differences were not iced i n the case o f both magnes ium and ammon ia remova l rates. Compared to the previous run, the magnes ium removal rates dropped f rom 7 4 % to 4 6 % in case o f the R # l , w h i l e for the R#2 this rate dropped f rom 6 2 % to 39%. O n the other hand, the ammonia removal rate increased i n both reactors. The ammon ia remova l rate was around 18% i n the R # l and around 15% in the R#2, compared to 6% and 5%, respectively. A l l o f these results are shown i n Figures 4.18 to 4.20. D u e to some operational problems, the reactors were set to recycle mode f rom 24 to 29 N o v e m b e r and restarted o n 30 N o v e m b e r 2006. A m i l k y - w h i t e , k i n d o f material was found at the bot tom of the (both) clarif iers. It was suspected this material to be some fo rm o f residual magnesium, al though no formal analysis was done o n this material . T h e clarifiers were cleared before restarting the reactors. S i x days o f continuous recycle mode might have some affect o n the performance o f the crystal l izers . 80 60 • • m > a 40 Vlg reirn 20 • • 0 m 18- Mov 20-Nov i i 22-Nov 24-Nov 26-Nov Date 28-Nov 30-Nov 2-Dec • R#l « R#2 Figure 4.18: M a g n e s i u m removal rates i n R # l and R#2 (20 N o v -1 D e c , 2006) 51 40 £ 30 20 a 10 3-Nov 20-Nov 22-Nov 24-Nov 26-Nov 28-Nov 30-Nov 2-Dec Date • R#l • R#2 Figure 4.19: A m m o n i a removal rates i n R # l and R#2 (20 N o v -1 D e c , 2006) "3 94 92 90 a. O 86 84 18-Nov 20-Nov 22-Nov 24-Nov 26-Nov 28-Nov 30-Nov 2-Dec Date • R#l * R#2 Figure 4.20: Phosphorus removal rates i n R # l and R#2 (20 N o v -1 D e c , 2006) Caustic use A s expected, without an external air supply, the caustic use rate increased in both reactors. A g a i n , the cascade media stripper used lesser amounts o f caustic than the compact media stripper. T h e average difference o f da i ly use rate between these two strippers was around 6 6 % . The mola r caustic used per mole o f phosphorus removed showed the same trend as previous runs. The results are shown i n Figures 4.21 and 4.22. 52 £• 1.6 * 1.2 s •a 0.8 S a U 0.4 8-Nov 20 -Nov 22 -Nov 24 -Nov 26 -Nov 28 -Nov 30 -Nov 2-Dec Date • R#l • R#2 Figure 4 .21: D a i l y caustic use i n R # l and R#2 (20 N o v -1 D e c , 2006) 2.50E-03 Day a P I • P2 0 N a O H l 0 N a O H 2 F igure 4.22: M o l a r P remova l and caustic use i n R # l and R#2 (20 N o v -1 D e c , 2006) D u r i n g this period, the cascade media stripper saved an average 3 5 % o f caustic chemica l used, compared to 4 6 % i n the previous run. O n the other hand, the compact media stripper used an extra 2 3 % o f caustic, on a da i ly basis. Carbon dioxide stripping Withou t an external air supply, the C 0 2 removal rate was expected to decrease. Surpr i s ing ly , a s l ight ly higher amount o f C O 2 r emoval was achieved by the compact media stripper. The average C O 2 r emoval rate for the R # l was around 14% dur ing this test, whereas this amount was on ly about 1 1 % i n the previous run wi th the air supply. O n the other hand, 53 the C 0 2 r emova l rate for the R#2 decreased s l ight ly , f rom 2 0 % to 17%. The C 0 2 r emoval rate dur ing this run is shown i n F igure 4.23. - N o v 2 0 - N o v 2 2 - N o v 2 4 - N o v 2 6 - N o v 2 8 - N o v 3 0 - N o v 2-Dec Date • R#l • R#2 F igure 4 .23: C 0 2 r emoval i n R # l and R#2 (20 N o v -1 D e c , 2006) A s already mentioned, the clarifiers were cleared before restarting the reactors. W h i l e c leaning the clar i f ier #1, water was passed through the s t r ipping tower o f the compact media stripper and this seemed to clear the c logg ing o f the pack ing media. A s a result, the stripped off C 0 2 more easi ly escaped f rom the system than i n the previous run. The result shown in F igure 4.23 illustrates this point; o n 23 November , the removal rate was less than 10%, whereas the highest removal (18%) was achieved on 30 November . Ammonia stripping T h e a m m o n i a remova l rate increased s ignif icant ly for both reactors dur ing this run. The average amount o f ammonia be ing removed/str ipped off, after fu l f i l l i ng the theoretical requirement o f struvite formation, was 9% and 7% for the R # l and R#2, respectively. In the previous run, w i t h an external air supply, this amount was 2 % and 1% for the R # l and R#2, respectively. The highest amount o f s tr ipping was achieved o n day 3, w h i c h corresponds to 30 November . The results are shown i n Figures 4.24 and 4.25. These results were quite unexpected, since wi thout an external air supply, ammonia s t r ipping was l i k e l y to decrease. Further study is needed to exp la in this result. 54 1.00E-02 ° 7.50E-03 •8 a. a, | 5.00E-03 Z | 2.50E-03 O.OOE+00 Day HR#1 DR#2 Figure 4.24: A m m o n i a s tr ipping i n R # l and R#2 (20 N o v -1 D e c , 2006) F igure 4.25: A m m o n i a s tr ipping i n R # l and R#2 (20 N o v -1 D e c , 2006) Summary of results > The average phosphorus remova l rate was 9 0 % i n both reactors. > T h e reactors were run i n a recycle mode f rom 23 to 30 November , w h i c h might have affected magnes ium and ammonia removal rates. > The strippers performed less effectively without an external air supply. D u r i n g this t ime, 3 5 % caustic was saved by the cascade stripper, wh i l e the compact media stripper used an extra 2 5 % caustic. 55 > The carbon d iox ide str ipping rate was better i n the compact media stripper than the cascade stripper. The str ipping tower was cleaned on 30 November , poss ib ly contr ibut ing to better CO2 s t r ipping by the compact media stripper. > A higher ammonia removal /s t r ipping rate was achieved by both the reactors than the previous run, where external air was supplied. This surprising result requires further investigation and explanat ion. 4.3.3 Run No. 3 This run was carried out f rom 4 to 12 December . In this test, the recycle ratio and the upf low ve loc i ty were set to 9 and 450 c m / m i n , respectively. The external air supply was resumed. The operating condit ions for this run are g iven i n Tab le 4.5. Table 4.5: Operat ing condit ions for R u n N o . 3 (4-12 D e c , 2006) Parameter Compact media stripper Cascade stripper Tota l feed 2.02 L / m i n 2.02 L / m i n Centrate f l ow 1 .94L/min 1.94 L / m i n M g feed f low 80 m L / m i n 80 m L / m i n R e c y c l e ratio 9 9 R e c y c l e f l o w 17.98 L / m i n 17.98 L / m i n To ta l f l ow 20 L / m i n 20 L / m i n p H 8.1 8.1 Harvest zone upf low ve loc i ty 450 c m / m i n 450 c m / m i n A i r Y e s 107.5 L / m i n Nutrient removal In this run, the R # l was able to remove an average 8 9 % o f phosphorus. The R#2 system was s l ight ly better i n r emov ing phosphorus, achieving an average 9 2 % dur ing this test per iod. B o t h reactors showed improvement in magnes ium remova l , compared to the 56 previous run. The magnes ium removal rate for the R# l averaged 5 5 % and for the R#2 unit, it was 4 2 % . Howeve r , the ammon ia r emova l rate decreased i n both the reactors; i n this run, the ammonia remova l rate averaged 9% and 8% i n the R#l and R#2, respectively, whereas it was 18% and 15% for R# l and R#2, respectively, i n the previous run wi thout an external air supply. The results are shown i n Figures 4.26 to 4.28. 73 > o 80 60 40 § 20 -Nov 20-Nov 22-Nov 24-Nov 26-Nov Date • R#l a R#2 28-Nov 30-Nov 2-Dec Figure 4.26: M a g n e s i u m removal rates i n R# l and R#2 (4-12 D e c , 2006) 40 £ 30 33 Z 20 10 • • 0 18-Nov 20-Nov 22-Nov 24-Nov 26-Nov Date • R#l • R#2 28-Nov 30-Nov 2-Dec Figure 4.27: A m m o n i a r emova l rates i n R# l and R#2 (4-12 D e c , 2006) 57 94 g 92 2 90 a. O 86 84 -Nov 20-Nov 22-Nov 24-Nov 26-Nov 28-Nov 30-Nov 2-Dec Date • R#l • R#2 Figure 4.28: Phosphoais removal rates i n R # l and R#2 (4-12 D e c , 2006) Caustic use The higher recycle ratio and upf low veloc i ty proved to have posi t ive impacts o n the performance o f the strippers, regarding da i ly caustic use. The R # l used an average 1.35 k g / d of caustic dur ing this t ime. The improvement was more pronounced i n case o f the R#2, where the average caustic used during this run was on ly 0.66 kg /d . B o t h reactors were r emoving almost same amount o f molar phosphorus f rom the system. L i k e a l l previous runs, the R#2 used less molar caustic to remove the same amount o f molar phosphorus, compared to the R # l . The results are g iven i n Figures 4.29 and 4.30. 2.50 1 2.00 .50 •5 1-00 a U 0.50 0.00 3-Dec 5-Dec 7-Dec Date • R#l • R#2 9-Dec 11-Dec 13-Dec Figure 4.29: D a i l y caustic use i n R # l and R#2 (4-12 D e c , 2006) 58 2.50E-03 2.00E-03 1.50E-03 l.OOE-03 5.00E-04 O.OOE+00 -I my.' 2 3 Day H P 1 • P 2 0 N a O H l N a O H 2 70 + 6 0 | 50 | + 4 0 .a 20 j3 © io s Figure 4.30: M o l a r P removal and caustic use i n R # l and R#2 (4-12 D e c , 2006) C o m p a r i n g these results w i t h the results obtained by running the reactors without the strippers, showed that the amount o f caustic saved by the cascade stripper alone, dur ing this test, was 86%. O n the other hand, the compact media stripper again fa i led to save any caustic; instead, it used an extra 10% caustic. Howeve r , this stripper also showed an improvement under higher recycle ratio and upf low veloc i ty ; in the previous two runs, it used an extra 1 5 % (wi th external air supply) and 2 3 % (without external air supply) caustic. The caustic use pattern o f the R # l showed a gradual increase i n usage rate over t ime. Th i s was an indica t ion o f " c l o g g i n g " o f the pack ing media i n the compact media stripper. It was ment ioned earlier that the str ipping tower was cleaned on 30 November ; therefore, it was relat ively c lean when this run was started o n 4 December and resulted i n better, overa l l performance. H o w e v e r , the performance again decreased as c logg ing started to occur. Carbon dioxide stripping Under a higher recycle ratio and upf low ve loc i ty , both strippers performed better than the previous two runs, i n str ipping CO2. The average CO2 s t r ipping rates achieved by the R # l and R#2 were 14% and 2 1 % , respectively. The higher the amount o f CO2 s t r ipping, the less caustic was used by the reactors. Th i s result is shown i n F igure 4 .31. 59 Figure 4 .31: C 0 2 r emova l i n R # l and R#2 (4-12 D e c , 2006) Ammonia stripping The amount o f ammonia removal , after fu l f i l l i ng the requirement o f theoretical struvite formation, decreased somewhat, relative to the previous run, when the strippers were run without an external air supply. The average ammon ia str ipping achieved by the R # l and R#2 was 6% and 5%, respectively; i n the previous run, this was 9% and 7% for the R # l and R#2, respectively. O n c e again, the compact med ia stripper was s l igh t ly better i n s t r ipping ammonia than the cascade stripper, al though the difference was almost negl ig ib le . T h e results are shown i n Figures 4.32 and 4.33. 6.00E-03 5.00E-03 © "S 4.00E-03 'I 3.00E-03 Z >- 2.00E-03 "o § 1.00E-03 0.00E+00 3 Day E3R#1 DR#2 Figure 4.32: A m m o n i a s t r ipping i n R # l and R#2 (4-12 D e c , 2006) 60 Figure 4.33: A m m o n i a s t r ipping i n R # l and R#2 (4-12 D e c , 2006) Summary of results > T h e average phosphorus removal was 8 9 % and 9 2 % for the R # l and R#2, respectively. > A higher recycle ratio and upf low ve loc i ty had a posi t ive effect, regarding caustic use. D u r i n g this run, the cascade stripper saved around 86% o f caustic usage. > A l t h o u g h the overa l l performance o f the compact media stripper improved under this condi t ion , it s t i l l c o u l d not save any caustic. The stripper suffered f rom frequent c logg ing , thus affecting its performance. > CO2 s t r ipping improved under this condi t ion . Howeve r , ammonia removal decreased somewhat, compare to the previous run, when the strippers were run without an external air supply. 4.4 Comparison Of Stripper Performance D u r i n g the course o f this study, a total four experiments were conducted - one without strippers and rest w i t h strippers under different condi t ions. The first experiment was 61 carried out to check whether the two reactors used i n this study were s imi lar , regarding overa l l performance, so that direct a compar i son c o u l d be made between the two. T h e results showed that they were very s imi l a r and hence, a l l comparisons between two strippers' performances were done direct ly. M a j o r f indings f rom a l l the experiments are summar ized i n Table 4.6. Table 4.6: S u m m a r y o f f indings f rom four tests Without Stripper 1st Run* 2nd Run§ 3rd Run v Compact media stripper Cascade stripper Compact media stripper Cascade stripper Compact media stripper Cascade stripper p removal (%) 90 90 90 90 90 88 92 Caustic use (kg/d) 1.23 1.41 0.84 1.51 0.91 1.35 0.66 Caustic savings0 (%) - 0 (-15) 46 0 (-23) 35 0 (-10) 86 co2 stripping (%) - 11 20 14 17 14 21 NH3-N stripping (%) - 2 1 9 7 6 5 * § e With air, Recycle ratio (RR) = 6, Upflow velocity = 400 cm/min Without air, RR = 6, Upflow velocity = 400 cm/min With air, RR = 9. Upflow velocity = 450 cm/min Caustic savings were calculated comparing the results of the runs (1st, 2nd and 3rd) to the "without stripper" condition test. Negative values indicate extra amount of caustic usage. 62 Throughout this project, a h igh percentage o f phosphorus remova l was achieved under each condi t ion , by both reactors. O v e r a l l , the removal rate was around 90%. The compact media stripper fa i led to save any amount o f caustic; rather, extra caustic was required a l l o f the t ime, since the compact media stripper was incorporated into the system. T h i s resulted f rom lack o f proper vent i lat ion. The stripper was mounted di rect ly above the clarif ier , l eaving on ly a smal l opening at the top o f the s tr ipping tower for the stripped off CO2 to escape f rom the system. The stripped CO2 had to m o v e upward through the pack ing media to reach to the top. Howeve r , the fact that the pack ing media was prone to c logg ing easi ly and frequently, made this CO2 movement through it even harder. A s a result, a part o f the stripped off CO2 probably d i sso lved back into the l i q u i d stream that was carr ied back to the reactor, through the recycle f l ow. Th i s process went o n repeatedly, thereby putting an addi t ional CO2 load into the system to be stripped. The compact med ia stripper was s l ight ly better in s tr ipping ammonia than the cascade stripper. The str ipping o f ammonia is associated w i t h lower ing the p H i n the system. Thus , ammon ia s tr ipping, together w i t h c i rcula t ion o f stripped off C 0 2 into the reactor, resulted i n more caustic be ing u t i l i zed , i n order to raise the p H o f the system, for improved performance. The cascade stripper proved to be effective i n saving caustic usage. The amount o f caustic saved by this stripper ranged from 3 5 % (without an external air supply) to 86% (wi th air, higher recycle ratio and upf low veloci ty) . In the previous study done at the L I W W T P by Z h a n g (2006), the reduct ion i n caustic addi t ion ranged f rom 4 6 % to 6 5 % . E v e n though Z h a n g (2006) never used any external air supply, she was able to obtain a higher amount o f caustic savings than the present study, where on ly 3 5 % caustic was saved when the stripper was run without an external air supply (the amount was 4 6 % wi th an external air supply at the recycle ratio o f 6.0). Howeve r , it should be noted that the operating p H dur ing the previous study was lower (7.9 and 8.0) than the current study, where the operative p H was maintained at 8.1 a l l the time. The higher caustic saving was achieved by Z h a n g (2006) because a lower p H (7.9) was employed for operation. Hence , it can be conc luded that, w i th a lower operating p H ( i f condit ions satisfy a l l cri teria o f struvite format ion and recovery) , the 63 cascade stripper w i l l be more effective i n saving caustic; thus, a lower cost o f struvite product ion can be expected. In this study, the highest amount o f caustic was saved by the cascade stripper, when the reactor was running wi th a higher recycle ratio (9.0). Th i s result was expected, as the p H of the effluent is higher than the in f low; when this effluent is recycled back to the reactor at a higher amount, it resulted in ra is ing the overa l l in-reactor p H , and hence, reduces the need for caustic addi t ion. O n the other hand, to get a higher recycle ratio, the centrate f l ow has to be reduced, i n order to satisfy the m a x i m u m hydraul ic loading rate o f the stripper/reactor. The system w i l l run w i t h a higher recycle f l o w that ul t imately w i l l reduce the in-reactor phosphorus leve l , as the effluent contains a lower amount o f phosphorus than the centrate/inflow. Th i s might have a detrimental effect on the phosphorus removal and recovery process. A n overa l l cost-benefit analysis is ca l led for A s expected, better results regarding caustic saving were obtained when the cascade stripper was run w i t h an external air supply. D u e to a shortage o f on-site research t ime, tests were performed w i t h o n l y one a i r f low rate (107.5 L / m i n ) . It is l i k e l y that, w i th a higher a i r f low rate, the s tr ipping eff ic iency w i l l increase, result ing i n further reduct ion i n caustic addit ion. Howeve r , it should be noted that the supply o f external air w i l l result i n addi t ional operating costs. Thus , there may be a trade off between saving caustic, by increasing the external air supply rate and str ipping more CO2, and v i ce versa. The cascade stripper d i d not w o r k w e l l , regarding str ipping ammonia . The ammon ia str ipping ranged f rom 1% to 7%. In some days, there was no str ipping at a l l . Th i s l o w amount o f ammon ia str ipping, i n fact, contributed to reducing caustic addi t ion. Hence , it is a matter o f choice between whether we want addi t ional ammonia removal /s t r ipping f rom the centrate, or want savings f rom caustic addit ion. 64 4.5 Quality Of Harvested Struvite The quali ty o f harvested struvite was determined by measuring its compos i t ion and purity. F o r this purpose, apart f rom the regular species o f struvite (i.e. magnesium, ammon ia and phosphoais) , four other elements, namely c a l c i u m , a luminum, i ron and potassium, were tested. Grab samples were chosen randomly f rom each experiment carried out on the two reactors. The summary o f the analyses is g iven in Table 4.7. De ta i l ed analyses are g i v e n i n A p p e n d i x C . Table 4.7: S u m m a r y o f struvite pellet compos i t ion Theoretical value (mg) Actual value (mg) Purity M g N P M g N P (%) 1st Run* R#l 48.5 28.5 62.7 99 R#2 48.2 27.9 60.1 95 2 n d Run§ R#l 49 28.6 63 48.5 27.4 59.3 94 R#2 46.9 26.8 59 93 3 r d Run R#l 48.5 28 60.3 95 R#2 44.9 25.4 57.7 89 * : With air, Recycle ratio (RR) = 6, Upflow velocity = 400 cm/min § : Without air, RR = 6, Upflow velocity = 400 cm/min v|/ : With air, RR = 9, Upflow velocity = 450 cm/min The results show that the harvested pellets f rom both reactors were composed of, on average, 9 4 % pure struvite (by mass). Previous analysis o f struvite pellets g r o w n at the L I W W T P was, on average, found to be 9 6 % pure struvite (Fattah, 2004); thus, both results are quite s imi lar . One th ing to be noted - the different operating condit ions d i d not seem to have any affect o n the qual i ty o f harvested struvite. The impur i ty content o f the harvested struvite pellets are g iven i n Table 4.8. The previous two studies at the L I W W T P found potassium to be present be low detection l i m i t i n the struvite (Huang 2003; Fattah, 2004). Howeve r , this t ime, the potassium leve l was 65 detectable; rather a l u m i n u m was found to be be low the detection level . It can be seen that the impur i ty ions were present i n a very smal l quantity, relative to magnes ium. Hence , there were few or almost no phosphate salts o f impur i ty ions. Table 4.8: Impuri ty contents o f struvite pellet Content by mass (mg) Content by mass (%)• Ca Fe K Ca Fe K 1st Run* R#l 0.90 0.69 0.16 0.18 0.14 0.03 R#2 1.10 0.60 0.18 0.22 0.12 0.04 2 n d Run§ R#l 0.85 0.62 0.13 0.17 0.12 0.03 R#2 0.82 0.73 0.13 0.16 0.15 0.03 3 r d Run v R#l 0.82 0.50 0.12 0.16 0.10 0.02 R#2 0.51 0.48 0.10 0.10 0.10 0.02 * § : With air, Recycle ratio (RR) = 6, Upflow velocity = 400 cm/min : Without air, RR = 6, Upflow velocity = 400 cm/min : With air, RR = 9, Upflow velocity = 450 cm/min 4.6 Operational Problems D u r i n g this project, several operational problems arose wh i l e running the crysta l l izer systems at the L I W W T P . Th i s section describes the problems faced and the ways the problems were mit igated and m i n i m i z e d . 4.6.1 Clogging of the compact media stripper F r o m the very beginning o f this study, the major p rob lem that was faced i n operating the compact media stripper was its susceptibi l i ty to c logg ing . The s t r ipping tower o f the compact media stripper was packed w i t h one inch, h o l l o w plastic balls , w h i c h were used to provide a larger specific surface area and sufficient air/water contact t ime for the passing l i q u i d stream ( in order to achieve better CO2 stripping). A t the beginning o f this study, the 66 plastic balls were s i m p l y put inside the stripping tower i n between the l i q u i d distributor and the support ing plate. Af ter three weeks o f use, it was noticed that the stripper 's performance was decreasing. It started to use more caustic than it p revious ly needed. The amount o f caustic requirement increased over the course o f operating t ime. This probably happened due to c logg ing o f the pack ing materials. The stripping tower was connected to the external clar if ier w i t h a series o f bolts. In order to check the condi t ion inside o f the s t r ipping tower, it was d ismounted f rom the top o f the clarif ier on 27 September. It was found that the pack ing materials were severely plugged due to formation o f struvite encrustation and suspended material on the wa l l s . T h i s layer left almost no space for the stripped C 0 2 to escape f rom the system, consequently affecting the stripper's performance. H o t water was passed through the tower to clean the pack ing media ; but it d i d not work . Af ter that a l l the plastic balls were taken out f rom the tower, c leaning was attempted by us ing a garden hose. Th i s attempt also failed. Th i s demonstrated h o w diff icul t it was to c lean the pack ing material once it got plugged. Hence , it was dec ided to change the arrangement o f the pack ing materials inside the tower, to make it somewhat easier to clean and maintain . T h i s resulted i n hanging the packing media w i t h the help o f a string f rom the l i q u i d distributor inside the str ipping tower; also, the o l d support ing plate was replaced by a new plate w i t h more open space. These changes i n arrangement are i l lustrated i n Figures 4.34 to 4.36. F igure 4.34: O l d support ing plate F igure 4.35: N e w support ing plate 67 Figure 4.36: Hang ing pack ing materials E v e n after these changes were made, the c logg ing p r o b l e m returned and remained unt i l to the end o f the study. The compact media stripper required thorough c leaning at least once a month . H o w e v e r , the change in arrangement o f pack ing mater ia l made it easier to clean, since it c o u l d be cleaned by on ly taking out the l i q u i d distr ibutor f rom top o f the tower. 4.6.2 pH controller problem T h e struvite c rys ta l l iza t ion process is h igh ly p H dependent, and i n this study, the p H of the system was mainta ined by adding caustic and str ipping CO2. T h e p H control ler setup was used to adjust in-reactor p H to the set point, by cont ro l l ing the caustic addi t ion. D u r i n g mid-October , it was observed that the da i ly caustic use rate, for both reactors, was gradual ly increasing. But , as the strippers were running without an external air supply at that moment , this increased amount o f caustic use was l inked wi th the no-air supply cond i t ion (as under no-air condi t ion , the caustic use rate was expected to raise). The air was restarted f rom 23 October . S t i l l , there was no s ign o f improvement regarding caustic use. 68 The research team started searching for the reason for this unusual behavior and discovered that both the p H controllers were broken. T h e controllers kept p u m p i n g caustic even when the in-reactor p H was right on the set point. In order to solve this problem, the broken controllers were replaced w i t h two new controllers on 10 November . S ince it was not k n o w n when this p roblem started, a l l data col lected dur ing 8 October to 10 N o v e m b e r were discarded. 4.6.3 Flow fluctuation A s mentioned previously , it was very hard to keep f lows , especial ly feed and recycle f lows , constant to the set point. Va r i a t i on i n the f l ow o f one o f the components w i l l have an effect on in-reactor supersaturation ratio and, thereby, on the performance o f the reactor. Centrate was provided once a day f rom the treatment plant and stored i n the storage tanks. F r o m the storage tanks, it was suppl ied cont inuously to the reactors. A s a result o f this intermittent f i l l i n g o f storage tanks and continuous feeding o f reactors, there was a significant drop i n pump head that resulted i n f luctuation o f f low. In order to m i n i m i z e this problem, a l l f lows were measured and readjusted ( i f necessary) by adjusting the pump speed da i ly . T o m i n i m i z e this effect, data hav ing more than ± 1 5 % varia t ion f rom the set point were discarded i n this study. Ano the r solut ion was to m i n i m i z e the variat ion o f centrate head i n the ho ld ing tank, w h i c h was not possible i n this pi lot-scale study. Howeve r , i n a ful l -scale operation, it is expected that this p rob lem w o u l d be m i n i m a l , w i th the use o f onl ine f l o w controllers. 4.6.4 Plugging of tubing P l u g g i n g o f tubing was another reason that caused a change i n f lows . T h e centrate and recycle f l ow tubing were often c logged w i t h suspended solids and struvite. Occas iona l ly , the encrusted layer broke off f rom tubing wal l s and c logged the pump and other ports. Th i s p rob lem was so lved by tapping the c logged por t ion wi th the handle o f screwdriver . Th i s process d is lodged the accumulated layer, into smaller pieces, and a l l owed normal f l o w to be resumed. H o t water was passed through the tubing (especial ly the one that car r ied the centrate to reactor) at least once a week. Th i s process softened the encrusted layer, m a k i n g it 69 easier to wash away. T u b i n g was a lways disconnected f rom the reactors w h i l e running hot water through it. 4.6.5 Reactor fouling The reactor wal l s were often coated w i t h a layer o f struvite. L i k e the c leaning process o f tubing, the handle o f screwdriver was used to tap of f the layer f rom the reactor wa l l s . Care has to be taken i n removing the encrusted layer, as the material may drop to the bot tom o f the reactor and p l u g the injection port. U s u a l l y , c leaning was undertaken right before harvest ing. The most susceptible place for struvite encrustation wi th in reactor is the inject ion port, as the highest loca l supersaturation ratio existed there. The injection port was a lways cleaned whenever struvite was harvested. C l e a n i n g was done by scraping of f the struvite w i t h a screwdriver . The caustic and magnes ium entrances were cleaned w i t h a th in wire . 4.6.6 Centrate supply D u e to some operational problems at the treatment plant, the L I W W T P was unable to supply sufficient centrate regularly. A s a result o f this, the reactors were often set to recycle mode. T h i s hampered the reactors performance. A t the beginning o f the study, two storage tanks were used to ho ld centrate. A third tank was instal led at the beginning o f October i n order to increase storage capacity and to use that extra vo lume o f centrate i n case the plant fai led to provide centrate. 4.6.7 Suspended solids Occas iona l ly , the solids content o f the centrate was quite h igh (700 m g / L ) . S ince the residence t ime o f the centrate i n the storage tank was l o w , some solids made their way into the reactor, result ing in solids accumulat ion in the tubing. Th i s also caused c logg ing i n the compact media stripper, thereby affecting the performance o f the stripper. The solids also accumulated at the bot tom o f the storage tanks. Thus , it was necessary to flush the tanks at 70 least once every 45 days. In a ful l-scale operation, it is recommended to instal l a sedimentation tank between the centrifuge and the reactor, to reduce the amount o f suspended solids i n the in f low, thereby reducing the p lugg ing problem. 71 CHAPTER FIVE CONCLUSIONS & RECOMMENDATIONS 5.1 Conclusions Based on the results obtained from this study on the effectiveness o f introducing a C O ? stripper into the struvite crys ta l l iza t ion process, the fo l lowing conclus ions can be drawn: > The performances o f the two struvite crystal l izers (when running without C O ? strippers) were ident ical . Essent ia l ly , the same amount o f magnesium, ammonia and phosphorus removal was achieved by the two reactors, whi le running in parallel . Hence, the results obtained from the two reactors, when running wi th strippers, w o u l d be direct ly comparable. Throughout the project, a h igh percentage o f phosphorus removal was achieved under each condi t ion by both the reactors. M o s t o f the t ime, the phosphorus removal rate was around 9 0 % . > T h e compact media stripper failed to save any amount o f caustic, regardless o f the operating condi t ions . Instead, more caustic was required when the stripper was introduced. One o f the reasons was that this stripper b locked the passage o f stripped CO2, since it was mounted o n top o f the clarif ier . Another reason was the susceptibi l i ty o f the stripper's pack ing media to become frequently c logged, w h i c h also resulted i n b lock ing the movement o f the C O ? through the str ipping tower. It was also found to be diff icul t to clean and maintain the compact media stripper. > T h e cascade stripper, on the other hand, was very effective in saving caustic. T h e amount o f caustic saved by the stripper ranged f rom 3 5 % (without external air supply) to 86% (wi th air, higher recycle ratio and upf low veloci ty) . In a previous study, Zhang (2006) was able to obtain a higher amount (46% to 65%) o f caustic savings, without an external air supply, but at an operating p H of 7.9 and 8.0. The operating p H of the current study was 8.1, higher than that used by Z h a n g (2006). It is bel ieved that, w i th a lower operating p H ( i f 72 condit ions satisfy a l l cri teria o f struvite formation), the cascade stripper w i l l be more effective in saving caustic. > The cascade stripper was more effective in saving caustic w i th higher recycle ratio and an external air supply. U n l i k e the compact media stripper, the cascade stripper never had any p lugg ing problem and it was easy to clean and maintain. B o t h strippers showed very poor performance regarding ammonia str ipping, wi th the compact media stripper s l ight ly better i n s tr ipping ammonia . > T h e harvested struvite pellets f rom both reactors were composed o f nearly pure struvite (94% by mass), w i th a smal l amount o f c a l c i u m and traces o f i ron and potassium. Different operating condit ions d id not have any affect o n the qual i ty o f struvite. 5.2 Recommendations Based on the experience gained f rom this study on the effectiveness o f in t roducing a CO2 stripper into the struvite crys ta l l iza t ion process, the f o l l o w i n g recommendations are made. > In order to achieve better eff ic iency, it is recommended that the centrate should pass through the stripper first. In this way , the in f low to the reactor (from the stripper) w i l l carry a lesser amount o f CO2, w h i c h w o u l d result in having an in f low wi th a higher p H value. > A higher a i r f low rate cou ld be used in strippers, since aeration increases CO2 str ipping eff ic iency and thereby raises the p H value o f wastewater. However , before app ly ing a higher a i r f low, it is recommended to do a cost-benefit evaluat ion, since this w o u l d the increase energy consumpt ion o f the process. > Instead o f air, other gases (e.g. pure oxygen, nitrogen gas, CO2 free air) c o u l d be used for aeration w h i c h , as it was found, cou ld raise the p H o f wastewater around 10.3 73 (Cohen and K i r c h m a n n , 2004). However , it should be noted that these gases are more expensive than that o f air. > The reactor (wi th stripper) should be tested under different p H values. T h e (both) strippers are bel ieved to perform better wi th a lower operating p H . > The strippers should be tested under different recycle ratios, as improvement was noticed, in case o f both strippers, when the recycle ratio was increased. > The compact media stripper should run wi th app ly ing the stripper 's m a x i m u m hydraul ic loading capacity. A higher amount o f l i q u i d f low has a better probabi l i ty to reduce the p lugging o f packing media. > Further study on the strippers' eff ic iency regarding ammon ia s t r ipping is also recommended, and incorporated into a cost4oenefit evaluation. > In a full-scale operation, it is recommended to instal l a sedimentat ion tank between the centrifuge and the reactor, to reduce the amount o f suspended solids in the in f low, thereby reducing the p lugging problem. 74 CHAPTER SIX REFERENCES Aage H . K . , Ande r sen B . L . , B l o m A . and Jensen I. (1997) T h e so lubi l i ty o f struvite. Journal of Radio Analytical and Nuclear Chemistry 223, 213-215. A d n a n A . , K o c h F . A . , and M a v i n i c D . S . (2003a) Pi lot -scale study o f phosphorus recovery through struvite crys ta l l iza t ion - II: A p p l y i n g in-reactor supersaturation ratio as a process control parameter. J. Environ. Eng. Sci. 2(6), 473-483. A d n a n A . , M a v i n i c D . S . and K o c h F . A . (2003b) Pi lot -scale study o f phosphorus recovery through struvite crys ta l l iza t ion — examin ing the process feasibil i ty. J. Environ. Eng. Sci. 2, 315-324. A d n a n A . , Dastur M . , M a v i n i c D . S . and K o c h F . A . (2004) Pre l iminary invest igat ion into factors affecting control led struvite crys ta l l iza t ion at the bench scale. / . Environ. Eng. Sci. 3, 195-202. Andrade A . and S c h u i l i n g R . D . (2001) The chemistry o f struvite crys ta l l iza t ion. Mineralogy Journal (Ukraine) 23, N5/6. A P H A , A W W A and W E F (1998) Standard Methods for the Examination of Water and Wastewater. 2 0 l h E d . , A P H A Publ icat ions . Washington , D . C . Batt istoni P. , F a v a G . , Pavan P . , M u s a c c o A . , C e c c h i F . (1997) Phosphate removal i n anaerobic l iquors by struvite crys ta l l iza t ion wi thout addi t ion o f chemicals : Pre l iminary results. Water Research 31 (11), 2925-2929. Battistoni P. , Pavan P . , C e c c h i F . , M a t a - A l v a r e z J . (1998) Phosphate removal in real anaerobic supernatants: M o d e l l i n g and performance o f a f lu id ized bed reactor. Water Science and Technology 38 (1), 275-283. Batt is toni P., Boccado ro R . , Pavan P . and C e c c h i F . (2001) Struvi te crysta l l i sa t ion in sludge dewatering supernatant us ing air s t r ipping: the new-fu l l scale plant at T r e v i s o (Italy) sewage works . http://www.nhm.ac.uk/research-curation/proiects/phosphate-recovery/Nordwiikerhout/Battistonil 302.doc. Site visi ted: 20 A p r i l 2007. 75 Battistoni P. , D e A n g e l i s A . , Pavan P . , Prisciandaro M . , and C e c c h i F . (2001) Phosphorus removal f rom a real anaerobic supernatant by struvite crystal l izat ion. Water Research 35(9), 2167-2178. Battistoni P. , Boccadoro R. , Fatone F . and Pavan P . (2005a) Auto-nuc lea t ion and crystal growth o f struvite in a demonstrative f lu id ised bed reactor ( F B R ) . Environmental Technology 26, 975-982. Batt is toni P. , Pac i B . , Fatone F . and Pavan P. (2005b) Phosphorus removal f rom supernatants at l ow concentration using packed and f lu id i sed bed reactors. Industrial and Engineering Chemistry Research 44, 6701-6707. Bea l L . J . , Burns R . T . and Stalder K . J . (1999) Effect o f anaerobic digest ion on struvite product ion for nutrient removal from swine waste pr ior to land appl icat ion. Presentation 1999 A S A E A n n u a l International M e e t i n g , Canada. B h u i y a n , M . I . H . , M a v i n i c , D . S . , and Beck ie , R . D . (2007). A so lub i l i ty and T h e r m o d y n a m i c study o f struvite. Environmental Technology (in press). Bonmat f A . and Flotats X . (2003) A i r s tr ipping o f ammon ia from p ig slurry: characterisation and feasibi l i ty as a pre- or post-treatment to mesophi l i c anaerobic digest ion. Waste Management 23, 261-272. Borgerd ing J . (1972) Phosphate deposits in digestion systems. Journal of the Water Pollution Control Federation 44, 813-819. Buchanan J .R. , M o t e C R . and R o b i n s o n R . B . (1994) The rmodynamics o f struvite formation. Transactions of the American Society of Agricultural Engineers 37, 617-621. C e c c h i F . , Bat t is toni P . and Boccadoro R . (2003) Phosphate crystal l isat ion process for P recovery appl ied at Trev i so munic ipa l wastewater treatment plant (Italy). http:/ /www.nhm.ac.uk/research-curation/proiects/phosphate-recoverv/Treviso2003.pdf Si te visi ted: 20 A p r i l , 2007. Cheung K . C . , C h u L . M . and W o n g M . H . (1997) A m m o n i a str ipping as a pretreatment for landf i l l leachate. Water, Air, and Soil Pollution 94, 209-221. C o e F . L . , E v a n A . and Worcester E . (2005) K i d n e y stone diseases. The Journal of Clinical Investigations 115, 2598-2608. D o y l e J . D . and Parsons S . A . (2002) Struvite formation, control and recovery. Water Research 36, 3925-3940. 76 Drive r J , L i j m b a c h D , Steen I. (1999) W h y recover phosphorus for recyc l ing , and how? Environmental Technology 20, 651-662. Durrant A . E . , Scr imshaw M . D . , Stratful I. and Lester J . N . (1999) R e v i e w o f the feasibi l i ty o f recovering phosphate f rom wastewater for use as a raw material by the phosphate industry. Environmental Technology 20 (7), 749-758. Fattah K . P . (2004) Pilot Scale Struvite Recovery Potential from Centrate at Lulu Island Wastewater Treatment Plant. M . A . S c . Thesis , Department o f C i v i l Engineer ing , T h e Univers i ty o f B r i t i s h C o l u m b i a Vancouver , B . C . , Canada. Fuj imoto N . , M i z u o c h i T . , T o g a m i Y . (1991) Phosphorus f ixat ion in the sludge treatment system o f a b io log i ca l phosphorus removal process. Water Science Technology 23, 635-640. Gaterel l M . R . , G a y R . , W i l s o n R. , G o c h i n R . J . and Lester J . N . (2000) A n economic and environmental evaluat ion o f the opportunities for substituting phosphorus recovered from wastewater treatment works i n exis t ing U K fertiliser markets. Environmental Technology 21, 1067-1084. Giesen A . (1999) Crys ta l l i sa t ion process enables environmental f r iendly phosphate removal at l ow cost. Environmental Technology 20, 769-775. H i r o y u k i H . and T o m S. (2003) Ni t rogen and Phosphorus recovery f rom a piggery waste water treatment faci l i ty . Proc. 2003 Third International Symposium on Environmentally Conscious Design and Inverse Manufacturing, Tokeyo , Japan, 422-426. Huang H . (2003) Pilot scale phosphorus recovery from anaerobic digester supernatant. M . A . S c . Thesis , Department o f C i v i l Engineer ing, The Univers i ty o f Br i t i sh C o l u m b i a Vancouver , B . C . Canada. Jaffer Y . , C l a r k T . A . , Pearce P. and Parsons S . A . (2002) Potential phosphorus recovery by struvite formation. Water Research 36, 1834-1842. Jas insk i S . M . , Kramer D . A . , Ober, J . A . and Searls, J .P . (1999) Fertilizers—Sustaining Global Food Supplies. U S G S Fact Sheet F S , 155-199. Jeong Y . K . and H w a n g S .J . (2005) O p t i m u m doses o f M g and P salts for precipi tat ing ammonia into struvite crystals in aerobic compost ing . Bioresource Technology 96, 1-6. 11 Katsuura H . (1998) Phosphorus recovery f rom sewage by granule forming process (full scale struvite recovery from a sewage w o r k at Shimane Prefecture, Japan). In International Conference on Phosphorus Recovery from Sewage and Animal Waste, Warwick University, U.K. K o h e n Y . and K i r c h m a n n H . (2004) Increasing the p H of wastewater to h igh levels w i t h different gases—CO2 s t r ipping. Water, Air, and Soil Pollution 159, 265-275. Koutsoukos P . G . , K o f i n a A . N . and Klepetsanis P . G . (2003) Exp lo ra t ion ofvalternatives for phosphorus recovery f rom wastewater by crystal l isa t ion. W a s i c W o r k s h o p , Istanbul (Turkey) . L e Cor re K . S . , Valsami-Jones E . , Hobbs P., Parsons S . A . (2005) Impact o f ca l c ium on struvite crystal size, shape and purity. / . Crystal Growth 283, 514-522. L i X . Z . and Zhao Q . L (2003) Recovery o f ammonium-ni t rogen from landf i l l leachate as a multi-nutrient fertil izer. Ecological Engineering 20, 171-181. L i a o P . H . , C h e n A . , L o K . V . (1995) R e m o v a l o f nitrogen f rom swine manure wastewaters by ammon ia str ipping. Bioresource Technology 54(1), 17-20. L i n d B . , B a n Z . and B y d e n S. (2000) Nutrient recovery f rom human urine by struvite crys ta l l iza t ion wi th ammonia adsoipt ion on zeoli te and wollastonite. Bioresource Technology 73, 169-174. Loewentha l R . E . , K o r n m i i l l e r U . R . C . and van Heerden E . P . (1994) M o d e l l i n g struvite precipi tat ion i n anaerobic treatment systems. Water Science and Technology 30(12), 107-116. M a v i n i c D . S . , K o c h F . A . , H a l l E . R . , A b r a h a m K . and Niedba la D . (1998) Anae rob ic co-digest ion o f combined sludges from a B N R wastewater treatment plant. Environmental Technology 19, 35-44. M o r s e G . K . , Brett S . W . , G u y J . A . and Lester J . N . (1998) R e v i e w : Phosphorus removal and recovery technologies. The Science of the Total Environment 212, 69-81. M u n c h E . and Bar r K . (2001) Con t ro l l ed struvite crys ta l l iza t ion for r emov ing phosphorus from anaerobic digester sidestreams. Water Research 35 (1), 151-159. M u s v o t o E . V . , E k a m a G . A . , Wen tze l M . C . and Loewentha l R . E . (2000) Ex tens ion and appl ica t ion o f the three-phase weak acid/base k inet ic mode l to the aeration treatment o f anaerobic digester l iquors. Water SA 26(4), 417-438. 78 Neeth l ing J .B and B e n i s c h M . (2004) Struvite control through process and fac i l i ty design as w e l l as operation strategy. Water Science and Technology 49,191-199. Niedba la D . (1995) Pilot-scale Studies of the Anaerobic Digestion of Combined Wastewater Sludges and Mitigation of Phosphorus Release. M . A . S c . Thesis , Department of C i v i l Engineer ing, T h e Univers i ty o f Br i t i sh C o l u m b i a Vancouver , B . C . , Canada . Ohl inger K . N . , Y o u n g T . M . and Schroeder E . D . (1998) Pred ic t ing struvite formation in digestion. Water Research 32, 3607-3614. Ohl inger K . N . , Y o u n g T . M . and Schroeder E . D . (1999) K ine t i c s effects o n preferential struvite accumulat ion i n wastewater. Journal of Environmental Engineering 125, 730-737. O m a r N . B . , Entrena M . and G o n z a l e s - M u h o z M . T . (1994) Effects o f p H and phosphate o n the product ion o f struvite by M y x o c o c c u s xanfhus. Geomicrobiology Journal 12, 81-90. Pi tman A . R . , Deacon S . L . and Alexande r W . V . (1991) The th ickening and treatment o f sewage sludges to m i n i m i z e phosphorus release. Water Research 25(12), 1285-1294. Potts D . (2002) Struvite Crystallization Model. Department o f C i v i l Engineer ing , T h e Univers i ty o f B r i t i s h C o l u m b i a Vancouver , B . C . , Canada. Sh imamura K . , Tanaka T . , M i u r a Y . and Ishikawa H . (2003) Deve lopment o f a h i g h eff iciency phosphorus recovery method us ing a f lu id ized-bed crys ta l l ized phosphorus removal system. Water Science and Technology 48 (1), 163-170. Shu L . , Schneider P . , Jegatheesan V . and Johnson J . A n economic evaluat ion o f phosphorus recovery as struvite from digester supernatant (2006) Bioresource Technology 97, 2211-2216. Steen, I. (1998) Phosphorus avai labi l i ty i n the 2 1 s t century management o f a non-renewable resource. Phosphorus and Potassium 217, 25-31. Stratful I., Sc r imshaw M . D . and Lester J . N . (2001) Condi t ions inf luencing the precipi tat ion of magnes ium a m m o n i u m phosphate. Water Research 35, 4191-4199. S t u m m W . and M o r g a n J .J . (1996) Aquatic chemistry—chemical equilibria and rates in natural waters. John W i l e y & Sons, N e w Y o r k , 148-204. 79 S u z u k i K . , Tanaka Y . , Osada T . and W a k i M . (2002) R e m o v a l o f phosphate, magnes ium and c a l c i u m f rom swine wastewater through crysta l l izat ion enhanced by aeration. Water Research 36, 2991-2998. Tchobanoglous , G . , Bur ton , F . L . and Stensel, H . D . (2003) Wastewater Engineering: Treatment and Reuse. M c G r a w - H i l l , N e w Y o r k , U S A . T i l l e y E . (2006) The Effects of Urine Storage Conditions on Struvite Recovery. M . A . S c . Thesis , Department o f C i v i l Engineer ing, The Unive r s i ty of B r i t i s h C o l u m b i a Vancouver , B . C . , Canada . U e n o Y . and F u j i i M . (2001) Three years experience o f operating and se l l ing recovered struvite f rom full-scale plant. Environmental Technology 22, 1373-1381. W a n g J . , B u r k e n J . G . , Zhang X . Q . , Surampal l i R . (2005) Engineered struvite precipitat ion: Impacts o f component- ion molar ratios and p H . / . Environmental Engineering-ASCE 131(10), 1433-1440. W a n g J . , B u r k e n J . G . and Zhang X . Q . (2006) Effect o f seeding materials and m i x i n g strength on struvite precipitat ion. Water Environment Research 78 (2), 125-132. W i l s e n a c h J . and van Loosdrecht M . (2003) Impact o f separate urine co l lec t ion on wastewater treatment systems. Water Science and Technology 48, 103-110. W o o d s N . C . , S o c k S . M and Daigger G . T . (1999) Phosphorus recovery technology mode l ing and feasibi l i ty evaluat ion for munic ipa l wastewater treatment plants. Environmental Technology 20, 743-748. Zhang Y . (2006) Struvite Crystallization from Digester Supernatant - Reducing Caustic Chemical Addition by CO2 Stripping. M . A . S c . Thesis , Department o f C i v i l Engineer ing , T h e Unive r s i ty o f Br i t i sh C o l u m b i a Vancouver , B . C . , Canada . Internet l i n k l : h t tp: / /en.wikipedia .org/wiki /Struvi te (Site visi ted: 20 A p r i l 2007) Internet l i nk2 : ht tp: / /www.eocp.org/plants- lulu.html (Site vis i ted: 3 M a r c h 2007) 80 APPENDIX A INSTRUMENT OPERATIONAL PARAMETERS FLAME ATOMIC ABSORPTION SPECTROPHOTOMETER Element Analyzed Magnesium Calcium Iron Concentra t ion Uni t s m g / L m g / L m g / L Instrument M o d e Absorbance Absorbance Absorbance S a m p l i n g M o d e A u t o n o r m a l A u t o n o r m a l A u t o n o r m a l Ca l ib ra t ion M o d e Concentra t ion Concentra t ion Concentra t ion Measurement M o d e Integrate Integrate Integrate Replicates Standard 3 3 3 Replicates Sample 3 3 3 Wave leng th 202.6 n m 422.7 n m 248.3 n m Range 0-100 m g / L 0-60 m g / L 0.06-15 m g / L F l a m e T y p e N 2 0 / C 2 H 2 N 2 0 / C ? H 2 N 2 0 / C 2 H 2 Cal ib ra t ion A l g o r i t h m N e w Ra t iona l N e w Rat iona l N e w Ra t iona l L a m p Current 4.0 raA 10 m A 5.0 m A 81 FLAME ATOMIC ABSORPTION SPECTROPHOTOMETER Element Analyzed Aluminum Potassium Sodium Concentra t ion Uni t s m g / L m g / L m g / L Instrument M o d e Absorbance Absorbance E m i s s i o n S a m p l i n g M o d e A u t o n o r m a l A u t o n o r m a l A u t o n o r m a l Ca l ib ra t ion M o d e Concentra t ion Concentrat ion Concent ra t ion Measurement M o d e Integrate Integrate Integrate Replicates Standard 3 3 3 Replicates Sample 3 3 3 Wavelength 309.3 n m 766.5 n m Range 0-20 m g / L 0.03-2 m g / L F lame T y p e N2O/C9H2 N2O/C2H2 N 2 0 / C 2 H 2 Cal ib ra t ion A l g o r i t h m N e w Ra t iona l N e w Rat ional N e w Ra t iona l L a m p Current 10 111A 5.0 m A FLOW INJECTION ANALYSIS Ion Analyzed PO4-P NH3-N Concentra t ion Uni t s m g / L m g / L Range 0-100 m g / L 0-100 m g / L Temperature 6 3 ° C 6 3 ° C M e t h o d A m m o n i u m Molybda t e Phenate Reference 1 2 1. L a C h a t Instrument Methods M a n u a l for Q u i c k C h e m ® Automated Ion A n a l y z e r (1990). Q u i c k C h e m method number 10-115-01-1Z. 2. A P H A , A W W A , W P C F (1995). M e t h o d 4 5 0 0 - N H 3 - F . Phenate M e t h o d . In Standard Methods for the E x a m i n a t i o n o f Water and Wastewater, 19 l h E d i t i o n . A m e r i c a n Pub l i c Heal th Assoc ia t ion , Wash ing ton , D C . 82 APPENDIX B OPERATIONAL DATA CENTRATE Date pH Temp (°C) Cond. (mS) Mg (mg/L) P04-P (mg/L) NH3-N (mg/L) Molar Ratio Mg:P N:P 2 2 - A u g 7.6 31.8 8.17 4.14 60.3 791 0.09 29 2 3 - A u g 7.3 30.5 7.19 7.75 95.2 879 0.10 20 2 4 - A u g 8.1 28.9 6.7 5.19 91.5 848 0.01 21 2 5 - A u g 7.4 30.7 6.2 6.76 89.8 797 0.10 20 2 7 - A u g 7.3 29.8 7.03 8.71 95.1 804 0.12 18 2 8 - A u g 7.3 34 7.7 7.12 100 854 0.09 19 2 9 - A u g 7.4 32 6.18 6.43 90.2 804 0.09 20 5-Sep 7.5 31 6.16 7.23 71 705 0.13 22 6-Sep 7.6 25.3 6.16 5.27 83.4 746 0.08 20 7-Sep 7.7 29.5 6.79 7.98 82.3 796 0.12 21 11-Sep 7.6 25.9 6.9 12.01 96.6 850 0.16 19 12-Sep 7.6 26.5 6.7 10.22 94.4 861 0.14 20 13-Sep 7.7 23.1 6.6 10.56 91.6 849 0.15 21 18-Sep 7.6 25.8 7.29 14.27 78.1 916 0.23 26 19-Sep 7.6 33 7.19 13.42 92.9 888 0.18 21 20-Sep 7.2 27 7.2 11.3 88.8 863 0.16 21 22-Sep 7.5 31 7.19 9.33 96.8 891 0.12 20 23-Sep 7.5 27 7.15 7.01 91.6 881 0.10 21 25-Sep 7.5 27 6.9 5.3 90.4 878 0.07 21 26-Sep 7.6 29.6 7.2 9.38 85.8 838 0.14 22 28-Sep 7.7 32.1 7.5 13.45 81.1 797 0.21 22 29-Sep 7.6 29.6 7.4 12.1 81.6 800 0.19 22 3-Oct 7.6 26 7.36 13.48 82 802 0.21 22 5-Oct 7.6 24.7 7.15 9.56 77.1 785 0.16 23 6-Oct 7.5 28.4 7.17 5.9 76.6 769 0.10 22 8-Oct 7.6 24.5 7.22 8.5 77.1 796 0.14 23 9-Oct 7.5 28.2 6.44 8.32 80.2 846 0.13 23 10-Oct 7.5 22.1 7.66 9.2 85.3 762 0.14 20 11-Oct 7.5 26 7.21 12.92 77.8 783 0.21 22 12-Oct 7.5 27.2 6.93 8.89 70.6 708 0.16 22 13-Oct 7.5 25.4 7.04 8.22 73.8 852 0.14 26 18-Oct 7.6 23.2 6.46 5.74 75.3 712 0.10 21 83 CENTRATE Date pH Temp Cond. Mg P04-P NH3-N Molar Ratio CC) (mS) (mg/L) (mg/L) (mg/L) Mg:P N:P 19-Oct 7.6 24.6 6.55 8.22 73 734 0.14 22 20-Oct 7.4 25.4 7.02 10.86 89.5 768 0.15 19 21-Oct 7.5 24.4 6.21 14.04 72 817 0.25 25 23-Oct 7.3 26 7.23 7.38 74 870 0.13 26 24-Oct 7.4 25 7.76 8.98 72.3 860 0.16 26 25-Oct 7.4 26 6.11 10.98 76.7 854 0.18 25 26-Oct 7.5 26 7.11 13.9 84 869 0.21 23 27-Oct 7.5 26.3 7.27 13.56 89.4 746 0.19 18 28-Oct 7.6 24.5 7.39 13.21 83.7 713 0.20 19 30-Oct 7.7 22.8 7.25 10.76 81 787 0.17 21 31-Oct 7.3 21.4 6.32 8.84 76.8 671 0.15 19 1-Nov 7.4 23.6 7.3 11.16 77.7 694 0.18 20 2 -Nov 7.6 21.4 7.05 11.08 83.2 688 0.17 18 3 -Nov 7.4 28.1 7.03 12.08 78.2 688 0.20 19 4 - N o v 7.5 22.9 6.62 11.4 70.9 658 0.20 21 8-Nov 7.8 23.8 7.33 10.9 81.2 823 0.17 22 9 -Nov 7.8 20.7 6.57 7.86 77.8 729 0.13 21 10-Nov 7.9 21.2 6.95 13.4 76.8 724 0.22 21 13-Nov 7.8 26.3 7.62 11.5 88.3 826 0.17 21 14-Nov 7.8 23 7.44 7.93 78 814 0.13 23 15-Nov 7.8 24.9 7.22 10.15 84.1 845 0.15 22 16-Nov 7.8 23.8 7.04 10.12 68.3 834 0.19 27 17-Nov 8.1 21 7.4 11.23 82.2 775 0.17 21 2 0 - N o v 8.0 21 7.33 10.69 69.3 634 0.20 20 2 3 - N o v 7.8 24 7.16 9.45 50.6 827 0.24 36 3 0 - N o v 8.0 19.3 6.83 9 42.6 698 0.27 36 1-Dec 7.8 25.6 7.15 8.58 63.1 760 0.17 27 4 -Dec 8 16.3 6.99 8.31 67.3 799 0.16 26 5-Dec 8.1 15.3 12.48 13.65 38.2 553 0.39 32 8-Dec 7.9 17.1 5.09 11.09 42.6 575 0.33 30 11-Dec 7.9 18.9 5.99 9.64 49 670 0.25 30 12-Dec 7.9 19.9 5.68 9.36 52.6 696 0.23 29 13-Dec 7.4 19.1 5.3 11.14 54.9 680 0.26 27 84 MAGNESIUM FEED Date Concentration Conductivity (mg/L) (mS/cm) 18-Aug 2276 17.81 2 2 - A u g 1956 16.69 2 5 - A u g 2072 16.93 2 7 - A u g 2443 18.48 2 8 - A u g 2470 17.5 2 9 - A u g 2471 18.88 5-Sep 2443 17.95 6-Sep 2140 18 7-Sep 2250 18.11 18-Sep 2422.5 17.43 20-Sep 2362.5 17.51 26-Sep 2357.5 19.7 29-Sep 2207.5 20.3 3-Oct 2122.5 19.58 10-Oct 2480 19.4 11-Oct 2447.5 18.2 19-Oct 2050 17.79 26-Oct 2207.5 18.23 27-Oct 2170 17.96 7 -Nov 2026 19.3 8-Nov 1846 17.58 16-Nov 1688 17.58 17-Nov 1774 17.33 3 0 - N o v 1754 16.11 1-Dec 1732 15.63 11-Dec 2147 17.56 12-Dec 2150 18.3 85 REACTOR #1 Flows Total Date Centrate Mg Feed Feed Recycle Total RR Mg Inflow (L/min) (L/min) (L/min) (L/min) (L/min) to Reactor (L/min) 25-Sep 1.92 0.08 2 13.36 15.36 6.7 99.39 26-Sep 2.42 0.08 2.5 15.5 18 6.2 84.52 28-Sep 2.42 0.08 2.5 15.5 18 6.2 83.66 29-Sep 2.47 0.08 2.55 15.5 18.05 6.1 80.98 3-Oct 3.52 0.08 3.6 16.8 20.4 4.7 60.35 13-Nov 2.51 0.1 2.61 14.49 17.1 5.6 81.79 14-Nov 2.2 0.1 2.3 16 18.3 7.0 87.85 15-Nov 2.2 0.1 2.3 15.4 17.7 6.7 89.97 17-Nov 2.51 0.1 2.61 15.63 18.24 6.0 78.77 2 0 - N o v 2.51 0.1 2.61 15.63 18.24 6.0 78.25 2 3 - N o v 2.51 0.1 2.61 15.63 18.24 6.0 77.06 3 0 - N o v 2.51 0.1 2.61 15.63 18.24 6.0 75.02 1-Dec 2.6 0.1 2.7 14.4 17.1 5.3 72.41 4 -Dec 1.42 0.08 1.5 17.1 18.6 11.4 100.24 5-Dec 1.97 0.08 2.05 18.47 20.52 9.0 80.71 8-Dec 1.87 0.08 1.95 19.35 21.3 9.9 81.69 11-Dec 2.67 0.08 2.75 17.65 20.4 6.4 59.75 12-Dec 2.52 0.08 2.6 16.9 19.5 6.5 75.22 86 REACTOR #1 Effluent Date p H T Cond. Mg P0 4 - P N H 3 - N (°C) (mS/cm) (mg/L) (mg/L) (mg/L) 25-Sep 8.15 26 6.9 5.53 21 819 26-Sep 8.22 26 7.53 33.22 5.3 810 28-Sep 8.4 29.3 7.27 13.3 10.5 741 29-Sep 8.2 25 7.76 74.88 2.71 736 3-Oct 8.29 26 7.13 2.4 12.3 723 13-Nov 8.29 21.5 7.13 15.24 9.56 740 14-Nov 8.3 28 6.88 21.78 7.51 780 15-Nov 8.41 21.8 6.73 20.49 6.49 791 17-Nov 8.42 21 7.04 30.21 8.45 761 2 0 - N o v 8.51 21 7.21 55.88 7.11 570 2 3 - N o v 8.44 17.2 7.26 26.22 6.87 761 3 0 - N o v 8.49 15.4 5.18 54.42 3.13 573 1-Dec 8.45 22.5 7.09 26.97 4.95 718 4 -Dec 8.42 15.2 7.04 44.4 4.57 696 5-Dec 8.45 14.7 5.25 16.86 4.47 483 8-Dec 8.48 16.4 5.33 44.44 4 .89 525 11-Dec 8.46 18.2 5.94 32.52 6.26 652 12-Dec 8.25 19.2 5.8 38.49 9.03 637 87 R E A C T O R #1 Caustic use Date Na N a O H Caustic use Caustic use (g/L) (g/L) (L/d) (kg/d) 25-Sep 19.42 33.77 86 2.91 26-Sep 28.1 48.87 34 1.66 28-Sep 30.04 52.24 44 2.30 29-Sep 37.04 64.42 30 1.93 3-Oct 40.11 69.76 42 2.93 13-Nov 26.52 46.12 20 0.92 14-Nov 23.52 40 .90 38 1.55 15-Nov 22.08 38.40 43 1.65 17-Nov 28.44 49.46 31 1.53 20-Nov 27.72 48.21 35 1.69 23-Nov 28.48 49.53 33 1.63 30-Nov 32.44 56.42 18 1.02 1-Dec 30.6 53.22 32 1.70 4 -Dec 33.16 57.67 15 0.87 5-Dec 31.72 55.17 19 1.05 8-Dec 36.64 63.72 18 1.15 11-Dec 24.32 42 .30 33 1.40 12-Dec 24.8 43.13 53 2.29 88 REACTOR #1 Dissolved CO2 Date Concentration (ppm as CO2) Inflow to stripper Outflow from stripper 14-Nov 14.62 13.8 15-Nov 20.75 16.34 17-Nov 20.75 19.77 2 3 - N o v 45.65 41.38 3 0 - N o v 48.38 39.78 1-Dec 44.5 38.1 4 -Dec 57 48.38 8-Dec 53.27 47.29 11-Dec 50.5 40.35 12-Dec 51.37 47.32 89 REACTOR #2 Flows Total Date Centrate Mg Feed Feed (L/min) (L/min) (L/min) 25-Sep 2.37 0.08 2.45 26-Sep 2.37 0.08 2.45 28-Sep 2.47 0.08 2.55 29-Sep 2.47 0.08 2.55 3-Oct 2.62 0.08 2.7 13-Nov 2.2 0.1 2.3 14-Nov 2.2 0.1 2.3 15-Nov 2.2 0.1 2.3 17-Nov 2.1 0.1 2.2 2 0 - N o v 2.51 0.1 2.61 2 3 - N o v 2.31 0.1 2.41 3 0 - N o v 2.74 0.1 2.84 1-Dec 2.6 0.1 2.7 4 -Dec 1.97 0.08 2.05 5-Dec 1.97 0.08 2.05 8-Dec 1.82 0.08 1.9 11-Dec 1.92 0.08 2 12-Dec 1.82 0.08 1.9 Recycle Total RR Mg Inflow (L/min) (L/min) to Reactor (L/min) 15.55 18 6.3 82.11 15.55 18 6.3 86.05 15.65 18.2 6.1 82.28 15.65 18.2 6.1 80.98 18.9 21.6 7.0 75.97 13.3 15.6 5.8 91.26 14.8 17.1 6.4 87.85 15.7 18 6.8 89.97 15.5 17.7 7.0 92.00 15.63 18.24 6.0 78.25 15.5 17.91 6.4 82.67 15.85 18.69 5.6 69.67 15.69 18.39 5.8 72.41 18.47 20.52 9.0 75.58 18.47 20.52 9.0 80.71 17.6 19.5 9.3 83.55 15.95 17.95 8.0 78.54 20.7 22.6 10.9 99.49 90 R E A C T O R #2 Effluent Date p H T Cond M g P 0 4 - P N H 3 - N (HC) (mS/cm) (mg/L) (mg/L) (mg/L) 25-Sep 8.16 26 7.11 10 16.2 841 26-Sep 8.3 26 7.81 57.55 3.32 787 28-Sep 8.1 29.3 7.39 24.18 16.1 734 29-Sep 8.3 24.3 8.48 31.23 1.95 649 3-Oct 8.4 25.1 7 18.3 5.85 703 13-Nov 8.2 22.7 7 21.03 9.14 772 14-Nov 8.1 25 7.34 24.33 8.62 790 15-Nov 8.28 21.4 7.11 22.08 7.78 779 17-Nov 8.46 21 7.13 72 7.11 752 2 0 - N o v 8.46 21 8 33.68 6.9 600 2 3 - N o v 8.48 19.1 7.5 76 7.04 772 3 0 - N o v 8.36 16 8.09 46.8 2.42 605 1-Dec 8.3 22.8 7.14 31.65 4.49 713 4-Dec 8.42 15.8 6.98 34.29 4.42 732 5-Dec 8.54 16.3 8.61 53.7 2.47 545 8-Dec 8.23 16.2 8.47 51 2.96 221 11-Dec 8.22 20 6.14 52.60 5.94 653 12-Dec 8.31 20.3 5.94 49.13 5.13 624 91 R E A C T O R #2 Caustic use Date Na N a O H Caustic use Caustic use (g/L) (g/L) (L/d) (kg/d) 25-Sep 15.36 26.71 16 0.43 26-Sep 15.56 27.06 30 0.81 28-Sep 15.32 26.64 62 1.65 29-Sep 28 48.70 44 2.14 3-Oct 21.52 37.43 30 1.12 13-Nov 19.32 33.60 18 0.61 14-Nov 24.12 41.95 8 0.34 15-Nov 24.32 42.30 22 0.93 17-Nov 20.48 35.62 41 1.46 20-Nov 25.08 43.62 20 0.87 23-Nov 27 46.96 20 0.94 30-Nov 28.1 48.85 13 0.64 1-Dec 26.96 46.89 25 1.17 4-Dec 27.32 47.51 13 0.62 5-Dec 25.97 45.17 17 0.77 8-Dec 22.91 39.85 18 0.47 11-Dec 22.83 39.70 13 0.52 12-Dec 22.64 39.37 20 0.79 92 REACTOR #2 Dissolved CO2 Date Concentration (ppm as CO2) Inflow to stripper Outflow from stripper 14-Nov 23 21.6 15-Nov 28.68 17 17-Nov 18.66 16.34 2 3 - N o v 44.95 39.81 3 0 - N o v 48.38 35.84 1-Dec 44.5 37.82 4 -Dec 50 38.5 8-Dec 53.4 45.12 11-Dec 73.72 57.13 12-Dec 65.23 52.42 93 APPENDIX C CHEMICAL ANALYSIS OF STRUVITE PELLETS Sample Mg (mg/L) N (mg/L) P (mg/L) Ca (mg/L) Al (mg/L) Fe (mg/L) K (mg/L) 1st Run R#l 485 285 627 9 0 6.9 1.6 R#2 482 279 601 11 0 6 1.8 2 n d R#l 485 274 593 8.5 0 6.2 1.3 Run§ R#2 469 268 590 8.2 0 7.3 1.3 3 r d Run R#l 485 280 603 8.2 0 5 1.2 R#2 449 254 577 5.1 0 4.8 1 * : With air, Recycle ratio (RR) = 6, Upflow velocity = 400 cm/min § : Without air, RR = 6, Upflow velocity = 400 cm/min v|/ : With air, RR = 9, Uptlow velocity = 450 cm/min 94 APPENDIX D C 0 2 CALIBRATION CURVE 95 

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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0063218/manifest

Comment

Related Items