|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 PROCESS DEVELOPMENT FOR RECOVERY OF NUTRIENTS AS STRUVITE AND STRUVITE BASED PRODUCTS By SACHIN M. GADEKAR A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011
2 2011 Sachin M. Gadekar
3 Dedicated to my beloved mother, brothers and late father
4 ACKNOWLEDGMENTS First of all, I would like to thank my academic advisor supervisor Dr. Pratap Pullammanappallil for his supervision and valuable guidance from the early stage of the shaping my professional career by providing an excellen t work environment and a very healthy group culture. I would like to acknowledge Dr. Lewis Johns, Dr. Bruce Welt, Dr. Ben Koopman and Dr. Amir Varshovi, my research committee members, for their advice and important contribution. I would like to thank Dr. Valentine Cracian from Major Analytical Instrumentation Center (MAIC), at UF for support in solid samples analysis. I am very thankful to the staff at Dairy Research Unit of IFAS at UF for help with my work on dairy wastewater. I want to thank Mr. Abhishek Dhoble and the staff of GreenTechnologies, LLC for their valuable help with pilot plant set up and operation. Many thanks to Mr. David Walker, President of Benchmark Design, LLC, for co operation with economic analysis. I would also like to thank Dr. Art Teixera for his support in analysis of mechanical properties of struvite. I thank my friends David (Brett) Walker, Eric Layton, Gayathri Ram Mohan, Samruthi Buxy and Billy (Huleo) Duckworth who have contributed to make this project a success and my gradu ate experience so enjoyable. I am deeply grateful to my mother and all other family members; without them, this would never have been possible.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 9 LIST OF FIGURES ................................ ................................ ................................ ....................... 10 LIST OF ABBREVIATIONS ................................ ................................ ................................ ........ 12 ABSTRACT ................................ ................................ ................................ ................................ ... 13 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 15 1.1 Project Focus ................................ ................................ ................................ ................. 15 1.1.1 Sources of Phosphorus in Water Streams ................................ ......................... 15 1.1.2 Industry ................................ ................................ ................................ ............. 16 1.1.3 Livestock ................................ ................................ ................................ ........... 16 1.1.4 Hum an Source ................................ ................................ ................................ ... 16 1.1.5 Fertilizers ................................ ................................ ................................ .......... 17 1.2 Nutrient Removal in a Wastewater Treatment Plant ................................ ..................... 17 1.3 Phosphorus in Animal Manure ................................ ................................ ...................... 18 1.4 Current Phosphorus Removal Techniques from Wastewater ................................ ....... 20 1.5 Hypothesis for the Present Research ................................ ................................ ............. 24 1.6 Goals and Research Approach ................................ ................................ ...................... 24 1.7 Outline of Dissertation ................................ ................................ ................................ .. 25 1.7.1 Development and Experimental Validation of a Mathematical Model of Struvite Precipitation ................................ ................................ ......................... 25 1.7.2 Sequential Batch Reactor for Recovery of Nitrogen and Phosphorus as Struvite from Sewage Sludge Centrate and Dairy Wastewater ........................ 25 1.7.3 Development of Product Formulation Techniques U sing Struvite Solids ........ 26 1.7.4 Concept to Commercialization (C2C) ................................ ............................... 26 1.7.5 Conclusions and Recommendati ons ................................ ................................ 26 2 DEVELOPMENT AND EXPERIMENTAL VALIDATION OF A MATHEMATICAL MODEL FOR STRUVITE PRECIPITATION ................................ ................................ ...... 29 2.1 Struvite Modeling ................................ ................................ ................................ ......... 29 2.2 Review of Struvite Precipitation Models ................................ ................................ ...... 29 2.2.1 Struvite Modeling Approaches ................................ ................................ ......... 29 2.2.2 Solution Approaches to Models ................................ ................................ ........ 31 2.3 Model Formulation ................................ ................................ ................................ ....... 33 2.3.1 Magnesium Phosphate Species ................................ ................................ ......... 34 2.3.2 Calcium Phosphate Species ................................ ................................ .............. 34
6 2.3.4 Other Salts ................................ ................................ ................................ ......... 35 2.3.5 Equ ations ................................ ................................ ................................ ........... 35 2.4 Results and Discussion ................................ ................................ ................................ .. 36 2.4.1 Model Validation ................................ ................................ .............................. 37 2.4. 2 Effect of pH ................................ ................................ ................................ ....... 41 2.4.3 Effect of Initial Magnesium, Phosphate and Ammonium Ratio ...................... 43 2.4.4 Determining Exion Concentration for Real Wastewater ............................... 45 2.4.6 Effect of pH on Struvite Purity and Phosphorus Removal in Real Wastewater ................................ ................................ ................................ ........ 46 2.5 Finding s ................................ ................................ ................................ ......................... 48 3 SEQUENTIAL BATCH REACTOR FOR RECOVERY OF NITROGEN AND PHOSPHORUS AS STRUVITE FROM SEWAGE SLUDGE CENTRATE AND DAIRY WASTEWATER ................................ ................................ ................................ ....... 60 3.1 Remocal of Nutrients from Wastewater ................................ ................................ ........ 60 3.1.1 Factors Affecting Struvite Formation : Concentration and Nature of Constituents ................................ ................................ ................................ ....... 61 3.1.2 Effect of pH ................................ ................................ ................................ ....... 62 3.1.3 Effect of Temperature ................................ ................................ ....................... 63 3.1.4 Hydraulic Retention Time (HRT) ................................ ................................ ..... 64 3.2 Review of Reactor Designs for Struvite Production ................................ ..................... 64 3.2.1 Stirred Tank Reactors ................................ ................................ ........................ 66 3.2.2 Fluidized Bed Reactor (FBR) ................................ ................................ ........... 67 3.3 Overview of Sequencing Batch Reactor (SBR) Technology ................................ ........ 69 3.3.1 Introduction ................................ ................................ ................................ ....... 69 3.3.2 Overview of F ull scale A pplications of SBR Technology ............................... 69 3.3.3 Characteristics of SBR Technology ................................ ................................ .. 71 3.4 Objectives ................................ ................................ ................................ ...................... 72 3.5 Materials and Methods ................................ ................................ ................................ .. 72 3. 6 Results and Discussion ................................ ................................ ................................ .. 75 3.6.1 Determination of Magnesium Needed for P Recovery as Struvite from Centrate ................................ ................................ ................................ ............. 75 3.6.2 Struvite Precipitation from Synthetic Solution in the SBR ............................... 76 3.6.3 Phosphorus Recovery from Centrate ................................ ................................ 77 3.6.4 Experiments on Complete Nitrogen and Phosphorus Recovery from Centrate ................................ ................................ ................................ ............. 78 3.6.5 Effect of Aeration on Centrate ................................ ................................ .......... 79 3.6.6 Struvite Recovery from Dairy Manure ................................ ............................. 80 3.6.7 Settleability of Struvite in the SBR ................................ ................................ ... 82 3.7 Outcomes ................................ ................................ ................................ ...................... 84
7 4 DEVELOPMENT OF PRODUCT FORMULATION TECHNIQUES USING STRUVITE SOLIDS ................................ ................................ ................................ .............. 93 4.1 Struvite as a Fertilizer ................................ ................................ ................................ ... 93 4.1.1 Recovery of Ammonia from Wastewater ................................ ......................... 95 4.1.2 Uses of Struvite ................................ ................................ ................................ 97 4.2 Objectives ................................ ................................ ................................ ...................... 99 4.2.1 Co cr ystallization of Struvite with KMAG ................................ ....................... 99 4.2.2 Study Absorption of Ammonia in Acidic KMAG Solution ........................... 100 4.2.3 Investigate Use of Agglomerates to Make Struvite Pellets ............................. 100 4.3 Materials and Methods ................................ ................................ ................................ 101 4.3.1 Struvite Precipitation ................................ ................................ ....................... 101 4.3.2 Preparation of KMAG Solution ................................ ................................ ...... 101 4.3.3 Agglomeration of Struvite ................................ ................................ .............. 102 4.3.4 Crystallization of Struvite with KMAG ................................ .......................... 102 4.3.5 Ammonia Removal from Centrate and Synthetic Solution ............................. 103 4.4 Results and Discussion ................................ ................................ ................................ 103 4.4.1 Effect of the Polymer A ddition on Struvite Filtration ................................ .... 103 4.4.2 Crystallization of Struvite with KMAG ................................ .......................... 104 126.96.36.199 Batch Crystallization ................................ ................................ ........ 104 188.8.131.52 Semi continuous Crystallization of KMAG with Stuvite ................. 106 4.4.3 Absorption of Ammonia in KMAG ................................ ................................ 106 4.5 Outcomes ................................ ................................ ................................ .................... 107 5 CONCEPT TO COMMERCIALIZATION (C2C) ................................ .............................. 129 5.1 Introduction ................................ ................................ ................................ ................. 129 5.2 Materials and Methods ................................ ................................ ................................ 134 5.3 Results and Discussion ................................ ................................ ................................ 135 5.3.1 Trial 1 ................................ ................................ ................................ .............. 135 184.108.40.206 Treatment Step 1 ................................ ................................ ............... 136 5. 3.1.2 Treatment Step 2 ................................ ................................ ............... 137 5.3.2 Trial 2 ................................ ................................ ................................ .............. 138 5.3.3 Comparison between Laboratory and the Pilot Scale Runs ............................ 139 5.4 Scale up Considerations in Precipitation ................................ ................................ .... 140 5.5 Discussion of Feasibility on Full scale ................................ ................................ ....... 141 5.6 Economics of Struvite Recovery ................................ ................................ ................. 141 5.7 Struvite Recovery as a Business ................................ ................................ ................. 144 5.8 Outcomes ................................ ................................ ................................ .................... 144 6 CONCLUSIONS AND FUTURE WORK ................................ ................................ ........... 151 6.1 Overall Findings ................................ ................................ ................................ .......... 151 6.1.1 Mathematical Model of Struvite Precipitation ................................ ................ 151 6.1.2 Sequential Batch Reactor Operation for Recovery of Nitrogen and Phosphorus as Struvite from Sewage Sludge Centrate and Dairy Wastewater. ................................ ................................ ................................ ..... 152
8 6.1.3 Fertilizer Product Formulation U sing Struvite Solids ................................ ..... 152 6.1.4 Pilot Scale Study of the Sequential Batch Operation for Struvite Precipitation ................................ ................................ ................................ .... 153 6.2 Future Work ................................ ................................ ................................ ................ 153 6.2.1 Automation of the Pilot Plant ................................ ................................ .......... 153 6.2.2 Biological Struvite Formation with E xisting Enhanced Biological Phosphorus Removal (EBPR) Processes ................................ ........................ 154 APPENDIX: POLYMATH MODEL FOR STRUVITE PRECIPITATION .............................. 1 56 LIST OF REFERENCES ................................ ................................ ................................ ............. 159 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 168
9 LIST OF TABLES Table page 2 1 List of solids included in the comprehensive model ................................ .......................... 50 2 2 Equilibrium expressions with equilibrium and solubility constants ................................ .. 51 2 3 Input parameters to the model for simulations on effect of pH ................................ ......... 52 2 4 Validation of the chemical equilibrium mode ................................ ................................ .. 52 3 1 Optimum pH values used in struvite formation ................................ ................................ 87 3 2 Effect of magnesium concentration on nitrogen and phosphorus removal from the centrate ................................ ................................ ................................ ............................... 87 3 3 Phosphorus r ecovery from the c entrate ................................ ................................ .............. 86 3 4 Recovery of nitrogen and phosphorus from centrate ................................ ......................... 86 3 5 Sequential b atch o peration: Process d etails ................................ ................................ ....... 89 3 6 Struvite r ecov ery from d airy m anure ................................ ................................ ............... 8 87 4 1 KMAG c hemical a nalysis ................................ ................................ ................................ 109 4 2 Solids considered in x ray diffraction analysis as standards ................................ ........... 110 4 3 Chemical added for s truvite f ormation for p olymer e ffect e xperiments .......................... 110 4 4 Effect of the p olymer a ddition on f iltration ................................ ................................ ..... 111 4 5 Crystallization of synthetic struvite with the acidic solution of KMAG ......................... 111 4 6 X ray diffraction results ................................ ................................ ................................ ... 111 4 7 Absorption of a mmonia in KMAG ................................ ................................ .................. 112 5 1 Details of equipment used in the pilot plant ................................ ................................ .... 146 5 2 Composition of struvite ge nerated in the pilot scale runs ................................ ................ 146 5 3 Chemical r equirement for the p ilot p lant r uns ................................ ................................ 146 5 4 Equipment c ost of the SBR p ilot p lant ................................ ................................ ............ 147 5 5 Capital c ost of f ull s cale s truvite p lant; Capacity: 700,000 GPD ................................ .... 148
10 LIST OF FIGURES Figure page 1 1 Secondary wastewater treatment plant at JEA ................................ ................................ ... 27 1 2 Proposed location for struvite production ................................ ................................ .......... 28 2 1 Effect of pH on struvite purity for ammonium, magnesium and phosphate ratios of 1:1:1 (at 10 mM and 100 mM) and 10 mM : 1.7 mM : 3.4 mM. ................................ ....... 55 2 2 E ffect of molar ratio of magnesium: phosphate on % struvite and total solids for two different molar ratios of ammonium (A) and phosphate (P) at pH = 8.7 ................... 56 2 3 Titration curve for anaerobically digested wastewater from dairy operations .................. 57 2 4 Effect of pH on struvite purity and phosphorus removal in centrate at an Mg:P ratio of 1:1 ................................ ................................ ................................ ................................ .. 58 2 5 Effect of pH on struvite purity and phosphorus removal in centrate at an Mg:P ratio of 0.5:1 ................................ ................................ ................................ ............................... 58 2 6 Effect of pH on struvite purit y and phosphorus removal in dairy flushwater at an Mg:P ratio of 1:1 ................................ ................................ ................................ ................ 59 2 7 Effect of pH on struvite purity and phosphorus removal in dairy flushwater at an Mg:P ratio of 0.5:1 ................................ ................................ ................................ ............. 59 3 1 Sequencing b atch r eactor (SBR) p rocess ................................ ................................ ........... 88 3 2 Sequential b atch r eactor for s truvite p recipitation ................................ ............................. 88 3 3 X ray diffraction patterns for struvite produced from the centrate ................................ .... 89 3 4 X ray diffraction patterns for struvite produced ................................ ................................ 90 3 5 Settleability of s olids in s truvite f ormation ................................ ................................ ........ 91 3 6 Solids s ettling in the SBR with d airy m anure: Settling time = 15 min .............................. 91 3 7 Solids s ettling in the SBR with d airy m anure: Settling time = 30 min .............................. 92 3 8 ( A ) Raw d airy m anure ; ( B ) Settled s olids after s truvite p recipitation ............................... 92 4 1 Struvite recovered from experiments ................................ ................................ ............... 113 4 2 Continuous c rystallization of KMAG with s truvite ................................ ......................... 114
11 4 3 Crystallization of KM AG with s truvite ................................ ................................ ........... 114 4 4 Absorption of a mmonia in KMAG ................................ ................................ .................. 115 4 5 KMAG with struvite: pH adjusted to 2.2 at commencement of crystallization ............... 115 4 6 KMAG from saturated solution ................................ ................................ ....................... 116 4 7 X ray diffraction results of solids from mixing KMAG with synthetic struvite ............. 119 4 8 X ray diffraction of r aw KMAG with struvite from the pilot scale run .......................... 121 4 9 X ray diffraction of solids from mixing KMAG solution with struvite from the pilot scale run ................................ ................................ ................................ ........................... 124 4 10 Change in potassium concentration in semi continuous crystallization of KMA G with struvite ................................ ................................ ................................ ..................... 125 4 11 Change in phosphorus concentration in semi continuous crystallization of KMA G with struvite ................................ ................................ ................................ ..................... 125 4 12 X ray diffraction analysis of solids from the semi continuous run ................................ 127 4 13 Recovery of n itrogen and p hosphorus from c entrate ................................ ....................... 128 5 1 Process l ayout ................................ ................................ ................................ .................. 149 5 2 Process f low d iagram ................................ ................................ ................................ ....... 150
12 LIST OF ABBREVIATION S BNR b iological nutrient removal C2C c oncept to commercialization DCP m onenite EBPR e nhanced biological phosphorus removal EPA environmental protection agency HAP h ydroxyapatite HRT hydraulic retention t ime FBR fluidi zed bed r eactor JEA Jacksonville Electric A uthority KMAG potassium magnesium sulphate MAP magnesium ammonium phosphate (s truvite ) OCP octacalcium phosphate SBR sequential batch reactor XRD x ray diffraction
13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PROCESS DEVELOPMENT FOR RECOVERY OF NUTRIENTS AS STRUVITE AND STRUVITE BASED PRODUCTS By Sachin M. Ga dekar May 2011 Chair: Pratap C. Pullammanappallil Major: Agricultural and Biological Engineering Formation of struvite (NH 4 MgPO 4 6H 2 O) in pipelines and inner surfaces of wastewater treatment processes has been found to make operation of the plant i nefficient and costly. Struvite is also identified as a route to recover nitrogen and phospho rus as a marketable fertilizer. Controlled crystallization of struvite is possible to prevent its spontaneous precipitation. E xisting struvite recovery techniques are found to have limitations such need for large reactor volumes complicated designs, higher capital and operating & maintenance costs The existing systems have been able to achieve only 90% removal of phosphate which does not necessarily meet strict EP A standards In this study a novel struvite recovery process is developed which eliminates the limitations of existing phosphorus recovery techniques discussed above S truvite precipitation reaction wa s found to be rapid therefore if it wa s possible to de couple the precipitation process from agglomeration/crystalliza tion process reactor volumes could be considerably reduced. The precipitation reaction was accomplished using a simple sequential batch reactor operation; this operation is similar to that cur rently employed by wastewater treatment plants thereby making it more operators friendly. A salable product
14 crystallized using potassium magnesium sulphate was produced in a separate unit operation using the reduced volume of settled sludge rather than ent ire wastewater. Predicting struvite precipitation potential, yield and purity is important for designers and operators of reactors for struvite precipitation. In this study a mathematical model of struvite precipitation process wa s developed for a closed system using physicochemical equilibrium expressions, mass balance equations for nitrogen, phosphorus and magnesium, and charge balance. The model simulations were validated against our experimental data using synthetic wastewater and some data from litera ture. The model satisfactorily predicted all data Bench scale studies showed that the pH level 8.4 8.7 allow for rapid struvite nucleation which caused the precipitation of fines. Aeration of centrate was found to increase the pH to 8.4. This was useful for minimizing the caustic requirement for stuvite formation and also to reduce carbonate concentration. This also improved quality of struvite by reducing solids with carbonate. Complete nitrogen and phosphorus removal was gained from centrate in a seque ntial batch reactor. Settling of precipitated struvite was found to be rapid with a settling time of 10 min. The recovery of phosphorus from dairy manure as a struvite containing precipitate was successfully demonstrated using the SBR mode of operation. P ilot scale trials conducted in an SBR of 500 g trial were found to recover 93 % of PO 4 3 P from centrate in a single step. Successful a mmonia recovery was found feasible by addition of phosphoric acid to centrate if additional nitrogen recovery is desired. It was found that the recovered struvite contained almost no hazardous materials and exhibit ed equivalent or better fertilizer effectiveness than conventional chemical fertilizers. Economic analysis of the process revealed that f or a plant treating 750,000 g al per day of centrate, capital investment of $1.5 million i s needed and a net profit of $ 452,000/year is predicted
15 CHAPTER 1 INTRODUCTION 1.1 Project Focus Removal and recovery of nutrients such as, phosphorus, nitrogen and potassium from different types of waste waters is of both ecological and economic importance. Currently mined phosphate represents the main source of phosphorus used by t he fertilizer, detergent or insecticide industries (Sten, 2004). Recovery of phosphorus is of particular importance because of the rapid depletion of mineral phosphate deposits all over the world (Driver et al.,1999) Phosphorus recycling from wastewater effluents has also gained importance, due to stringent regulations of phosphorus levels in the effluents in wastewater discharged to sensible areas such as lakes, rivers and reservoirs (UWWTD, 1991). Therefore, as an alternative to traditional phosphorus removal techniques, recycling of phosphorus in the form of valuable and usable chemicals is gaining importance. Phosphorus can be recovered from effluents generated in a sewage treatment plant, animal wastes and agro processing industries. The studies pres ented in this thesis will focus on recovering phosphorus from sewage treatment plant effluents and dairy wastewater. A brief review of various sources of phosphorus is presented in next section. 1.1.1 Sources of P hosphorus in W ater S treams Phosphorus in m unicipal, industrial, and agricultural wastewater may come from a variety of sources listed below. The Southern Cooperative Series (1998) reports that phosphorus naturally enters the soil solution by the following means: 1. D issolution of primary minerals, 2. D issolution of secondary minerals, 3. D esorption of phosphorus from clays, oxides, and minerals, and 4. B iological conversion of phosphorus by mineralization. In addition to these, other anthropogenic inputs to surface water are substantial and comprise the remaining balance of phosphorus inputs to surface waters. While many sources have been
16 decreased through managed use or treatment in past decades, livestock inputs have actually increased. 1.1.2 Industry The application of phosphorus is wide in many aspects of industries. Phosphorus is used in some softened waters for stabilization of calcium carbonate to eliminate the need of recarbonation (Sawyer et al., 19 94). Polyphosphates are also used in public water systems in order to control corrosion as well as in steam power plants to control scaling in the boilers. If phosphate complexes are used, these rapidly hydrolyze to orthophosphate at high temperatures in t he power plants (Sawyer et al., 1994). Many industries release high phosphorus concentration wastewater, for example food and dairy processing, and other processes that use phosphorus, phosphoric acid or phosphates also contribute. 1.1.3 Livestock Phospho rus is an essential element required for livestock. Animal feeding operations can provide a significant source of nutrients for crops through manure. The manure produced from the livestock is applied to the land, and often the ratio of nitrogen to phosphor us is unbalanced in a manure thus causing over application of phosphorus to the land. However, supplying manure that is nutrient balanced for nitrogen and phosphorus requires reducing the phosphorus content of manure, without compromising the performance o f the livestock (ICM, 2000; Burns et al., 2002). 1.1.4 Human S ource After elimination of phosphate based detergents in the 1980s, much of the inorganic phosphorus is now contributed by human wastes as a result of the metabolic breakdown of proteins and e limination of the liberated phosphates in the urine. The amount of phosphorus released is a function of protein intake. An average person in United States releases 1.5 g/day of
17 phosphorus (CEEP, 2003). The per capita contribution from human populations, af ter sewage treatment, was estimated at 0.62 kg total phosphorus/person/year in Morse et al., 1998, whereas, the report presented by CEEP 2007 estimated 0.43 kg total phosphorus/person/year. In a wastewater treatment plant at Metamorphosis, Greece, the aver age total phosphorus influent value is 30.6 mg/L (Sotirakou et al., 1999). In 1992, Point Loma Wastewater Treatment Plant, City of San Diego discharged 954 metric tonnes of total phosphorus (SCCWRP, 1992). 1.1.5 Fertilizers Phosphorus is applied to the lan d as a fertilizer providing nutrient to the crops. Phosphate is extracted from rocks containing apatite. Phosphate fertilizers are produced by adding sulfuric acid to the phosphate rock which is 16 21 % as phosphorus pentoxide (Pollution Prevention and Ab atement Handbook, 1998). Part of phosphorus applied to the land is taken up by the crops and is accumulated in them, whereas, the remaining fraction of phosphorus dissolves in the rain and can be transported to the nearby water body or infiltrates in to th e groundwater (ICM, 2000). 1.2 Nutrient Removal in a Wastewater Treatment Plant Point sources of anthropogenic phosphorus like municipal sewage can be treated before disposal. The objective of a secondary wastewater treatment plant for municipal sewage is to produce an effluent water and sludge stream suitable for discharge or reuse. A large amount of sludge is generated throughout the sewage treatment process. Primary sludge which is the material that settles out during primary treatment often has a strong odor and requires treatment prior to disposal. Secondary sludge is the excess microorganisms from the biological treatment processes. Untreated sludge is about 97 % water. The goals of sludge treatment are to dewater, stabilize the sludge and reduce odors, decompose some of the organic matter, reduce volume, kill disease causing organisms and disinfect the sludge. Sludge treatment involves aerobic or anaerobic digestion followed by dewatering. C entrate or filtrate produced in dewatering
18 operations contains nutrients like ammonia and phosphate and can create discharge problems for treatment plant. sludge from primary and secondary treatment as well as biosolids brought in by o ther waste dewatered using centrifuges and gravity belt thickening. Operational stages of this wastewater treatment plant are depicted in Figure 1 1. This produces cen trate and filtrate streams containing 750 800 ppm of ammonia N and 130 160 ppm of phosphate P. Due to the nutrient content, environmental regulations prevent JEA from discharging these streams into St. Johns River. Therefore these streams are returned to t he plant headworks where it adds to the wastewater burden. 1.3 Phosphorus in Animal Manure Phosphorus is one of the vital elements needed for animal growth and milk production. The functions include in metabolic activities in soft tissues, the maintenance of appetite, optimal growth, fertility, bone development and the prevention of bone diseases. The daily nutritional requirements for dairy cattle and beef cattle have been stated as 86 95 g/day and 35 40 g/day respectively (CEEP, 2003). The mechanisms of phosphorus digestion and metabolism differ substantially between ruminant and monogastic animals. In many cases excess phosphorus is used in order to maximize the production of livestock. However, feeding excess phosphorus increases ph osphorus levels in an imal waste streams Only 14 % of phosphorus in corn and 31 % of soybean meal phosphorus can be digested by cattle. Because a large percentage of phosphorus is unavailable, much of it is excreted (ICM, 2000). To overcome the limited availability excess phosph orus is fed to animals. The waste stream is therefore very rich in phosphorus.
19 Similar concerns with nutrients are observed in the animal manure treatment operations. The management, treatment and disposal of liquid and solid manure at dairy operations are receiving increased attention. In Florida, governmental regulation of waste disposal activities at dairies is cited as having a negative impact on dairy producers (Tefertiller et al., 1998). Florida dairies use large volumes of water for barn flushing, resulting in large amounts of dilute wastewater. The most common manure management system utilizes short term holding ponds for flushed manure wastewater storage, with subsequent pumping to sprayfields to supply fertilizer nutrients and irrigation water fo r production of forage crops. Generally, nutrients are conserved and recycled on individual farms. But, due to increased animal operations, the quantities of nutrients produced exceed the recommended agronomic application rates on the available land. Histo rically, anaerobic lagoons are commonly used for stabilization and storage of livestock wastes and wastewaters. Some nutrient removal is obtained by settling of solids and volatilization of ammonia. However, their use is declining because of issues with od ors, groundwater contamination, overflow, and dike failures during heavy rainfall linked to tropical weather events. As the size of livestock operations increase, the quantities of nutrients in the waste or wastewater increases, and available cropland beco mes limiting. Dairy farms in Florida have traditionally used flushing systems in milking parlors, feeding barns, and free stall barns (Reid and Horwath, 1980). It becomes increasingly difficult to apply the wastewater to available cropland at agronomic rat es to prevent environmental problems from excess nutrients. A need exists for waste and wastewater management systems on production enterprises which capture, process and export nutrients from the farming business. Land area requirements would be minimized odors and other gaseous emissions could be controlled, surface and groundwater pollution problems could be eliminated, and the nutrients
20 could be converted into value added products and exported from the farm. Although technologies exist for accomplishin g these goals, it has been difficult to implement them in a manner which is economically viable and reliable. Drivers that could potentially provide positive economics for the recovery of phosphorus from animal waste in the United States include; 1) a dram atic increase in the cost of inorganic phosphorus, and 2) the implementation of enforceable regulations that require the land application of animal manures on a phosphorus basis nationally. 1.4 Current Phosphorus Removal Techniques from Wastewater Mainly, ammonia reduction techniques are well established and widely accepted. This is typically accomplished by nitrification and denitrification operations which biologically convert ammonia into nitrogen gas. Unlike nitrogen, the phosphorus contained in wastew ater effluent is transformed into a solid form to be removed (Heinzmann, 2004). Traditional phosphorus removal processes work then by fixing the phosphorus into the sludge: either chemically by precipitation of soluble phosphorus with aluminum or iron sal ts into insoluble phosphate compounds. Separation of the precipitates is then achieved by sedimentation, filtration or flotation (Parsons and Berry, 2004) or biologically i.e. Biological Nutrient Removal (BNR), Enhanced Biological Phosphorus Removal (EBPR) using the ability of some micro organisms to accumulate phosphates as polyphosphates for their own metabolism (Driver et al., 1999; Mulkerrins et al., 2004) These proce sses are efficient in the sense that they can reduce phosphorus concentrations in waste water treated to less than 1 ppm ( Booker et al., 1999; Tebbut, 1998), but they present some disadvantages. Among them: the accumulation of nutrients (N and some P) in the resulting sludge [Parsons and Doyle, 2004) sludge production especially when using chemical processes where P is removed by the formation of P rich sludge ( i.e. iron phosphates and iron hydroxides precipitation) leading to significant increases in sludge management costs (Woods et al., 1999)
21 phosphorus precipitates generated by these me thods are not recyclable for reuse by the industry (de Bashan and Bashan, 2004). Aluminum phosphates are toxic to human, animals and aquatic life. It is reported as a skin irritant (Berkowitz et al., 2005) Another approach to remove phosphorus from wastewa ter effluents is through the crystallization of phosphate compounds such as calcium phosphate or struvite. Struvite is an ammonium magnesium phosphate mineral with the chemical formula (NH 4 )MgPO 4 6H 2 O. Struvite precipitation is accomplished by addition of magnesium and adjustment of pH. This approach has gained particular interest since both struvite and calcium phosphates have been identified as marketable fertilizers (Gaterell et al., 2000). The general equation describing struvite formation is as follows : Struvite is primarily known by wastewater companies as a hard scale occurring at points of high turbulence along the water treatment process and resulting in major breakdowns and pipe blockages. Struvite as a source of scale deposits in wastewater treatmen t plants was confirmed by Borgerding (1972) when it occurred on the walls of an anaerobic digestion system at the Hyperion treatment plant, Los Angeles, 1963. Indeed struvite naturally occurs under favorable condition of pH and mixing energy in specific ar eas of wastewater treatment plants (e.g. pipes, heat exchangers) when concentrations in magnesium, phosphate and ammonium reach an equimolar ratio 1:1:1. Struvite is frequently formed in recycle flush animal waste management systems. Formation of struvite in flushwater recycle pipes has been problematic in liquid manure handling systems because it creates blockages (Buchanan, 1994; Doyle and Parson, 2002). Most of the ould be of economic interest. Until management strategies were developed, the problem of struvite
22 precipitation caused a serious setback to the recycle flush approach in confinement waste management (Safley et al., 1982). Therefore, a large portion of str uvite research has been directed towards removal and prevention of struvite formation rather than towards its precipitation from solution for recovery and reuse. But, studies by Bridger et al. (1962) have confirmed the excellent agronomic properties of struvite. While only slightly soluble in water and soil solutions, struvite is found to be a highly effective source of phosphorus, nitrogen and magnesium for plants through both foliar and soil application. The release of nutrients is enhanced by a biolog ical nitrification mechanism, with the nutrients being released at a controlled rate over an extended period of time. When properly granulated, it can be applied to soil at rates greatly exceeding those of conventional fertilizers without danger of burning plant roots. Other than use as a premium grade slow releasing fertilizer, struvite also finds uses as a raw material in the phosphate industry, for making fire resistant panels and as a binding material in cements ( Sarkar, 1990 Schuiling and Andrade, 199 9). Struvite precipitation can be adopted to: 1. Lower the phosphorus concentration of the effluent by reducing the phosphorus load of the retur n liquors from sludge treatment 2. Reduce the amount of Polyaluminium Chloride (PAC l) and poly ferric sulfate used, wh ich in turn will reduc e the volume of sludge produced 3. Produce a phosphorus rich material, which could be recycled by the fertilizer industry, providing extra re venue from the sale of struvite 4. Offer material, energy and transportation cost savings for the w astewater treatment plants resulting from reduced sludge transport 5. Save operational costs by preventing clogging of pipes from struvite crystals in the pipeline. For a typical wastewater treatment plant serving one million people, the annual cost of dealin g with the struvite problem ranges from $160,000 to $800,000 (Parsons and Berry, 2004)
23 Several laboratory and pilot scale studies have been carried out to assess the potential of struvite recovery methods in removing and recovering phosphorus as a reusable product, and a few of them have been tested at full scale in The Netherlands (Giesen 1999) and Italy (Battistoni et al., 2005a; Battistoni et al., 2005b). In Europe and Japan, large municipal sewage handling facilities have already embraced phosphorus rec overy technology (Batistoni et al., 2001; Gaterell et al., 2000; Liberti et al., 2001; Piekema and Giesen, 2001; Ueno and Fujii, 2001). However Japan and Canada are the only countries where 90% P removal and recovery from anaerobically digested sludge liqu ors as struvite has been implemented and the resulting product sold to fertilizer companies (Gaterell et al., 2000; Ueno and Fujii 2001). A schematic diagram of wastewater treatment plant after the introduction of the struvite production f acility is depict ed in Figure 1 2. The c entrate/filtrate recovered after sludge dewatering and supernatant liquor from gravity belt thickening, which are rich in nitrogen and phosphorus, can be subjected to struvite production. This will reduce the nutrient load on wastewater treatment plant bec ause the water after struvite recover can be discharged as effluent after chlorination as shown in Figure 1 1 The recovery technologies currently tested are based on the crystallization of phosphorus as hydroxyapatite (HAP) or struvite (MAP) and the major ity uses sludge liquors generated from anaerobic digesters as their influent. The technologies can be classified in three main categories: Selective ion exchange ( i.e. RIM NUT process, (Liberti et al., 1986) Precipitation in a stirred reactor (Laridi et al., 2005; Mangin and Klein 2004; Seco et al., 2004; Stratful et al., 2004) Precipitation in fluidized bed reactors or air agitated reactors (Battistoni et al., 2005a; Battistoni et al., 2005b; Ueno and Fujii, 2001; Von Mnch and Barr, 2001) Limitations o f existing techniques are:
24 It relies on in situ crystallization and agglomeration which results in large reactor volumes to ensure appropriate residence times for successful agglomeration. Formation of crystals is favored as it reduces downstream processi n g cost to make salable product Overall reaction rate (consequently reactor volume) is controlled by crystallization pr ocess rather than precipitation Capital costs are high for these systems Reactors used in the commercial processes are fluidized bed or membrane reactors w hich incur high operating costs These systems have been able to achiev e only 90% removal of phosphate 1.5 Hypothesis for the Present Research There is a need to develop a struvite recovery process which eliminates the limitations of exis ting phosphorus recovery techniques discussed as follows. Struvite precipitation reaction is rapid therefore if it is possible to decouple the precipitation process from agglomeration/crystallization process reactor volum es can be considerably reduced The precipitation reaction can be accomplished in reactor designs similar to that currently employed by wastewater treatment plants thereby ma king it more operator friendly A salable product can be produced in a separate unit operation using the reduced volum e of settled sludg e rather than entire wastewater 1.6 Goals and Research Approach The goal of this project is to develop and implement an efficient and cost effective process to recover N & P nutrients in wastewater as struvite. The nature of effluent from different wastewater unit operations is unique. The optimal conditions like pH and magnesium concentration for struvite precipitation should be determined for different streams (Uludag Demirer et al., 2005). This approach is particularly important for str uvite research because the complex chemical principles, which govern its precipitation, are readily altered by slight changes in the chemical composition. A novel approach of struvite precipitation in an sequential batch reactor (SBR) is developed and vali dated experimentally. A comprehensive model is utilized to
25 compute chemical quantities necessary for the reaction. It is also used to accurately predict the purity and amount of struvite sludge produced. Struvite produced in the process is utilized to dev elop different products with multiple nutrients. Struvite precipitation technique developed in the laboratory was validated in a pilot plant utilizing an SBR. To summarize, the research approach included: 1. Development and validation of a mathematical model to calculate chemical requirements for struvite precipitation 2. Development and validation of an SBR mode of operation for struvite precipitation in a laboratory scale reactor 3. Development of product formulation techniques using settled struvite sludge for s alable agglomerates 4. Pilot scale validation and implementation of the struvite precipitation and settling process 1.7 Outline of Dissertation An outline of subsequent chapters in this thesis is given below. 1.7.1 Development and Experimental Validation of a Mathematical Model of Struvite Precipitation This chapter describes the mathematical approach to modeling struvite precipitation. A review of existing equilibrium chemistry models is carried out. A detailed discussion of m odeling approach and its validation along with simulation results is presented. 1.7.2 Sequential Batch Reactor for Recovery of Nitrogen and Phosphorus as Struvite from Sewage Sludge Centrate and Dairy Wastewater This chapter describes results of sequential batch reactor experiments with synthetic wastewater, centrate and dairy manure. Experiments on struvite formation for chemical requirements with aeration of centrate are discussed. Effect of pH and concentration of struvite on settling of solids in the conical bottom reactor are provided in this chapter.
26 1.7.3 Development of Product Formulation Techniques U sing Struvite Solids Product formulation techniques using synthetic struvite as well as struvite from cent rate are described in this chapter. It includes solids analysis using chemical methods, x r ay diffraction ( XRD ) and other physical parameters like yield strength, particle size distribution etc. 1.7.4 Concept to Commercialization (C2C) Pilot plant testi ng details are provided for the sequential batch operation to recover phosphate, in the form of struvite, from centrate obtained from Buckman Wastewater Treatment Plant, City of Jacksonville, Florida. Study of the variations in concentration of soluble and dissolved species at various depths of pilot scale reactor as a function of settling time are discussed. An analysis of the plant design and operating procedures is provided to reduce costs and meet constraints, with an emphasis on improving efficiency an d increasing profitability. 1.7.5 Conclusions and Recommendations This chapter summarizes the outcomes of this research including results from laboratory, pilot scale and model simulations. It also lists recommendations for future research and process d evelopment for full scale operation.
27 Figure 1 1. Secondary wastewater treatment p lant at J acksonville Electric Authority
28 Figure 1 2. Proposed location for struvite p roduction
29 CHAPTER 2 DEVELOPMENT AND EXPE RIMENTAL VALIDATION OF A MATHEMATICAL MO DEL FOR STRUVITE PRECIPI TATION 2.1 Struvite Modeling Predicting struvite precipitation potential is important to designers and operators for design, development and operation of reactors for stru vite precipitation. For process control, it is essential to know the conditions under which struvite precipitation is likely to occur. Upon mixing salts of magnesium, ammonium and phosphate several ionic and dissolved species are formed in addition to pr ecipitates including struvite. Modeling approaches used in struvite formation studies incorporate solution chemistry and thermodynamics, growth kinetic s and process description of the recovery system. 2.2 Review of Struvite Precipitation Models 2.2.1 St ruvite Modeling Approaches Various researchers have developed equilibrium chemistry models for struvite precipitation (Harada et al., 2006; Loewenthal et al., 1994; Ohlinger et al., 1998 ) These models are based on the physico chemical equilibrium of the v arious ionic, dissolved and solid species. A struvite precipitation model at least requires the incorporation of concentrations of ionic species NH 4 + PO 4 3 and Mg 2+ dissolved species NH 3 and H 3 PO 4 and solid species MgNH 4 PO 4 However, a number of othe r ionic species (e.g. HPO 4 2 H 2 PO 4 2 MgOH + MgPO 4 MgH 2 PO 4 + dissolved species (e.g ., H 3 PO 4 MgHPO 4(dissolved) ) and solid species (e.g. Mg 3 (PO 4 ) 2 .8H 2 O, Mg 3 (PO 4 ) 2 .22H 2 O, Mg(OH) 2 MgHPO 4 (solid) ) exist in equilibrium. The complexity of models depends on the number of soluble and solid species considered. Loewenthal et al. (1994) predicted struvite precipitation potential in synthetically prepared solutions that mimicked anaerobic digester effluents. This was a simple equilibrium
30 model that considered str uvite as the only solid species ; the ionic species considered were Mg 2+ NH 4 + PO 4 3 HPO 4 2 H 2 PO 4 and the dissolved species were NH 3 and H 3 PO 4 In addition to the previous H 2 CO 3 CH 3 COO CH 3 COOH, carbonate and bicarbonate were also considered. Ohlin ger et al. (1998) also considered only struvite as the solid species in their model. But, in addition to species modeled by Loewenthal et al. (1994), Ohlinger et al. (1998) also included MgH 2 PO 4 + and MgPO 4 a s these complexes exert a strong influence on equilibrium conditions. The model also included ionic strength effects. Wang et al. (2006) included the formation of Mg(OH) 2 precipitate in addition to struvite and they also considered MgHPO 4 as an additional dissolved species. Scott et al. (1991) modeled five solid species : struvite, Mg 3 (PO 4 ) 2 .8H 2 O, Mg 3 (PO 4 ) 2 .22H 2 O, Mg(OH) 2 .6H 2 O and MgHPO 4 and all the dissolved and ionic species considered by Wang et al. (2006). Harada et al. (2006) considered eight solid species : calcium precipitates: Ca 3 (PO 4 ) 2 CaHPO 4 Ca(OH) 2 CaCO 3 and CaMg(CO 3 ) 2 and precipitates containing Mg, namely struvite, Mg(OH) 2 and MgCO 3 Wrigley (1999) developed a computer model to describe struvite solubility chemistry which included the mass balance and electro neutrality equation s in the input components. The model was improved by incorporating dissolved magnesium hydrogen phosphate and an activity coefficient. Several researchers have employed MINTEQA2 and/or Visual Minteq to study struvite equilibrium chemistry. The MINTEQA2 is an aquatic chemistry equilibrium model This program is the Windows version of MINTEQA2 originally developed by the U.S. Environmental Protection Agency This model accepts inputs of magnesium, ammonium, and phosphate concentrations, pH, temperature, and ionic strength of a solution and it computes the degree of saturation with respect to struvite. Visual Minteq is a chemical equilibrium computer program
31 with a thermodynamic database that allows for the calculation of speciation, solubility, and equilibrium of solid and dissolved phases of minerals in an aqueous solution Modeling work by Battistoni et al. (2002) was able to link pH with saturation on the precipitation with seed materials. This additional proposed model allows predictions on the precipitat ion efficiency of an FBR with a given concentration of constituents at a set pH for example, at pH 8.5 a precipitation efficiency of 70% can be achieved with a contact time of 0.2h. Though this is not the full story, the conversion of phosphorus to struv ite is higher than the nucleation efficiency on the seed material which implies the production of fines. This researcher concluded that to avoid the production of fines a higher contact time within the reactor is needed. 2.2.2 Solution Approaches to Models As the number of the considered solid and soluble species increases, the complexity of the model also increases, so analytical solutions are no longer possible and a numerical solution is therefore needed. Lowenthal et al. (1994) and Ohlinger et a l. (1998) employed an iterative technique to converge one concentration value to an experimentally measured value while the other concentrations were calculated from equilibrium expressions. Scott et al. (1991) did not explicitly determine the concentrati ons of dissolved, ionic and solid species but they compiled operating curves that related pH with total concentrations of magnesium, nitrogen and phosphorus in liquid under struvite recovery conditions. Only the mole fractions of the solids were comput ed. Harada et al. (2006) simplified the solution procedure by limiting the formation of solid mixtures to 14 patterns that were generated a priori by including or excluding various solid species. The concentrations of dissolved and ionic species were calc ulated for each pattern and the concentrations that gave reasonable values were chosen.
32 Scott and Venkitachalam (1996) used the symbolic codes to calculate the amount of struvite expected to precipitate in an inorganic system. They used the Maple prog ram with the primary calculating variable of NH 4 They used a quadratic expression with five constants which produced the symbolic codes in the FORTRAN program. The calculations in this model proceeded which included a method of convergence. They used checks to assure that the solution obeyed the mass and electroneutrality criteria. They also stated it is possible that the concentrations became too high for validity or the iterations with activity coefficients calculated unreasonable values. They accep ted the limitations of the codes and that the algorithm required that a solid phase be present ; the lack of a solution implied no solid exists. They picked pH and pMg values that were close to the nearest values and also a weighted estimate (interpolated). This gave an estimate if the answer was within the selected set and it was only useful during a single cycle. Also they used the value from the qua dratic fit (extrapolated); this was only effective when close to the correct value. T he last was then an a verage of the last two estimates. In different nutrient or excess non reactive ion concentrations, different estimates were found to be useful. Fast convergence occurred when close to a solution. They illustrated the convergence for the case when approxima tely equal molar quantities of P Mg and N were added at 0.01 molar ( P = 310 mg/litre, M = 243 mg/litre, and N = 140 mg/litre), with no non reactive ions. Wrigley (1999) specified two input variables in the Maple file which had the total molar concentra tion of magnesium in the liquid phase and pH. Maple gave an output which was a series of FORTRAN codes that were algebraic equations with variables (ionic species with N, P and Mg) and equilibrium constants. From the output data file, groups of variables were selected for graphical presentation. The a mount of total solids or struvite precipitated was not reported here. Plots of variation of dissolved concentrations of Mg, P and N with pH were presented.
33 The more complex models previously described are n ot easily amenable for developing dynamic models to predict struvite precipitation in dedicated reactors for the following reasons: Present models cannot be used to determine chemical requirement (magnesium and caustic) to operate a dedicated struvite prec ipitation reactor Models cannot be used directly in the process control of a struvite reactor without modifications Struvite is not included in the database of commercial software such as MINTEQA2 and Visual Minteq Total number of solids considered in t he models is not complete. Some of the solids (calcium species) which are eliminated in the models previously discussed may be formed in the system if the residence times are of the order of four to five days Model reduction is done to solve it in the b eginning by eliminating solid species which may be formed under certain conditions In this chapter a comprehensive model is developed and used to determine the concentrations of all species (dissolved, ionic and solid) to enable investigation of the purity and yield of struvite for various operating conditions (pH and ratios of initial NH 4 + Mg 2+ nd PO 4 3 ). It uses mass and charge balances in addition to the physico chemical equilibrium equations. The model described here considers 15 different soli d species which is the maximum number reported in the literature. The model was validated by comparing it to experimental data from the literature and data obtained from these experiments. 2.3 M odel Formulation The model describes the evolution of an op en system in the presence of ionic species such as ammonium (ammonium chloride), magnesium (magnesium chloride) and phosphate (potassium phosphate). A complete list of solids included in the model is presented in Table 2 1 and discussed as follows.
34 2.3 .1 Magnesium Phosphate Species Five possible magnesium phosphate species can crystallize in the presence of Mg 2+ NH 4 + PO 4 3 species: 1) magnesium ammonium phosphate or struvite (MgNH 4 PO 4 6H 2 O); 2) magnesium hydrogen phosphate or newberyite, MHP (MgHPO 4 ); 3) Mg 3 (PO 4 ) 2 .22H 2 O, MP22; 4) bobierrite, Mg 3 (PO 4 ) 2 8H 2 O, MP8 and 5) magnesium hydroxide or Brucite (Mg(OH) 2 ) Struvite precipitates in the presence of Mg 2+ NH 4 + and PO 4 3 when the pH range is between 7 and 11 and at Mg/C a molar ratios greater than 0.6 Newberyite precipitates at high concentrations of Mg 2+ and P but it precipitates at lower pH (< 6.0). Previous research demonstrated that MgCl 2 .6H 2 O could be added to the solution to force the precipitation of struvite (Munch and Barr, 2001). 2.3.2 Calc ium Phosphate Species Five calcium phosphate crystalline species can precipitate from solutions containing Ca and P : 1) hydroxyapatite [HAP, Ca 5 (PO 4 ) 3 OH], 2) tricalcium phosphate (whitlockite) [TCP, Ca 3 (PO 4 ) 2 ], 3) octacalcium phosphate [OCP, Ca 8 (HPO 4 ) 2 (PO 4 ) 4 .5H 2 O], 4)monenite (DCP, CaHPO 4 ) and 5) dicalcium phosphate dihydrate (brushite) (DCPD, CaHPO 4 .2H 2 O). Although formation of HAP and TCP are thermodynamically favored, the kinetics of this process is extremely slow ( Ferguson and McCarty, 19 71 ). It is als o reported that the magnesium ion kinetically hinders the nucleation and subsequent growth of HAP and OCP ( Salimi et al., 1985 ; Abbona 1990 ; ). Also OCP is formed by the hydrolysis of DCPD in solutions of pH 5 6. It is found that DCP is a thermodynamically stable species. Salimi et al. (1984) reported that the presence of Mg 2+ has no detectable effect on the rate of DCPD crystallization. None of these species was removed from the model because the formation of these may be favored at high calcium concentrat ions (> 200 ppm) and higher residence times in the systems such as pipes or recirculation lines.
35 2.3.4 Other Salts Two forms of magnesium carbonates may also precipitate, magnesite (MgCO 3 ) and nesquehonite (MgCO 3 .3H 2 O). Both species were kept in the model Two mixed carbonates of Ca 2+ and Mg 2+ were included in the model, dolomite [CaMg(CO 3 ) 2 ] and huntite [CaMg 3 (CO 3 ) 4 ]. Calcium carbonate CaCO 3 was also added to the model. The model was thus formulated based on the following assumptions: a) Dissolved and ionic s pecies present in the system are NH 3 NH 4 + PO 4 3 HPO 4 2 H 2 PO 4 H 3 PO 4, MgOH + MgH 2 PO 4 + MgPO 4 MgHPO 4 (dissolved), CO 3 2 HCO 3 2 H 2 CO 3 Ca 2+ Na + K + Mg 2+ H + Cl and OH b) Fifteen different precipitates are produced which are listed in Table 2 1 and are previously discussed c) The pH was kept constant by addition of NaOH or HCl. This was simulated by adjusting the variable exions which were set equal to [cations modeled] [anions modeled] d) Reactions are at equilibrium e) Reactions proceed in a n open system or a batch reactor f) Reactions o ccur at room temperature (25C) g) Activity coeffi cients were assumed to be unity h) Effect of ionic strength on activity was neglected. 2.3.5 Equations The model includes overall mass balance for magnesium, nitroge n and phosphorus, electro neutrality, and physico chemical and solubility equilibrium equations to describe the system. Values of equilibrium constants and solubility products a t 25C are shown in Table 2 2 (Scott 2001; Harada et al., 2006; Moon et al., 2 007). Of particular importance is the solubility product for struvite. Various values have been reported ranging between 12.6 and 13.26 ( Ohlinger et al., 1998) In the present model, a more commonly used value of 12.7 was used unless s tated otherwise.
36 Polymath Educational Version 6.1 was used to solve the model. Initial conditions which include pH, total concentrations of nitrogen, magnesium, phosphorus, inorganic carbon and calcium were input along with all equilibrium constants. Initial guesses for Mg 2+ NH 4 + PO 4 3 CO 3 2 and Ca 2+ were provided. The Polymath program solved the expressions and gave concentrations of dissolved and ionic species and concentrations of solid components. These expressions were included as mass balance equations for total magnesium (Mg t ), calcium (Ca t ), ammonia N (N t ), inorganic carbon (TIC), phosphorus (P t ) as shown in Appendix A. The expressions were written as functions of corresponding ionic species concentrations in molar quantities. Expressions for calculating resi dual concentration in the liquid form were expressed for nitrogen, phosphorus and magnesium to get an idea of recovery of these species after struvite precipitation. Using the charge balance equation the appropriate exion concentration was determined. This gave the acid or base requirement for the given pH. The concept of exion species is discussed in detail in a later section. The f ollowing charge balance equation applies to the system : [NH 4 + ] + [K + ] + 2[Mg 2+ ] + 2[Ca 2+ ] + [Na + ] + [H + ] + [ MgH 2 PO 4 + ] + [MgOH + ] = 3[ PO 4 3 ] + 2[ HPO 4 2 ] + 2[CO 3 2 ] + [HCO 3 ] + [H 2 PO 4 ] + [MgPO 4 ] + [ Cl ] + [OH ] + [Exions] Exions = [other cations] [other anions]. 2.4 Results and Discussion The p olymath program solves systems of nonlinear algebraic equations which can be both nonlinear simultaneous (implicit) and auxiliary (explicit) equations. Only real roots (non complex) are found. All equations are che cked for correct syntax and other errors upon entry. An implicit nonlinear equation is entered in the form of a function [f(x)]. An explicit equation is
37 written in the form: x = an expression where the expression may contain constants, implicit variables, and explicit variables. The solution algorithms require specification of initial estimates for all the variables in the implicit nonlinear equations. Closer initial estimates have a better chance of converging to the desired solution. M ultiple solutions may occur to a particular set of equations which can be identified by starting at different sets of initial guesses. Solution s to the model equations in the the solution method. F our algorithms are available in the Polymath program for solution s of nonlinear equations, and all of them are based on the Newton Raphson (NR) method. This method uses a truncated Taylor series estimate of the function values to obtain better estimates of the unknowns. In each iteration the function values and the matrix of partial derivatives are calculated. Numerical perturbation is used for calculating the matrix of partial derivatives. The iterative solution stops if either the sum of the magnitudes of the functions is less than 10 7 or the sum of the absolute values of the corrections to the unknowns is less than10 7 The maximum number of iterations used in this all simulation was 300. For a system of equations containing constrained variables (some or all the varia bles must be positive throughout the solution process or only at the solution), an algorithm which combines the step length restricted NR method with a continuation type method is used. In the step length restricted NR method the progress in the NR dire ction is restricted so that none of the constrained variables become s zero or negative. In the continuation method, the original problem is converted into a sequence of easier to solve problems where the solution is closer to the initial estima te than in t he original problem. 2.4.1 Model Validation T he model was validated by comparing its predictions to 12 struvite precipitation studies collected from literature that were carried out using synthetic and real wastewater at different pH
38 values and initial concentrations of magnesium, phosphate and ammonium. The initial concentrations of these ions were equal to the total concentrations of magnesium (Mg T ), nitrogen (N T ) and phosphorus (P T ) species i n the system. Data from model simulations are presented i n Table 2 4. Solids containing magnesium are listed in T able 2 4, but the model considers a total of 15 solids as mentioned earlier. The model was found to satisfactorily predict struvite formation for all literature data. Five of these datasets marked b y in Table 2 4 need additional explanation. Closer inspection of experimental methods as described in these references revealed that the struvite concentration reported was actually total solids concentration. For example, Loewenthal et al. (1994) estimated struvite concentration from differences in average molar concentration of Mg 2+ NH 4 + and PO 4 3 between initial and final values in solution of these species. Since Mg 2+ and PO 4 3 can form insoluble compounds other than struvite, for example MP8 MP22, MHP and Mg(OH) 2 difference s in initial and final concentrations do not truly represent concentration s of struvite but represent concentration of solids. Using the experimental conditions of Loewenthal et al. (1994), the model predicted total soli ds concentration of 733 mg/L of which struvite was the only mg/L with MHP being the other dominant solid. Experimentally reported struvite was 601 mg/L which in reality could be the total solids concentration and was in agreement to the total solids predicted by the model. This researcher is reporting no data on calcium concentration. Harada et al. (2006) reported that most phosphate in their experiments was precipitated in the form of struvite. Initial total calcium concentration reported was 1.21 mmol/L which was input for validation. Experimental data indicated that 12.3 mM of PO 4 3 was precipitated as struvite (1685 mg/L). This value, however, was approximately equal to the total solids predicted
39 (1706.2 m g/L) by the model. Concentrations of NH 4 + and Mg 2+ used here were higher than concentration of phosphate. The model showed that other significant components which could make up the solid phase included MP8 (88 mg/L), MHP (222.9 mg/L) CaHPO 4 (15.6 mg/L ) and CaHPO 4 .5H 2 O (7 mg/L). Hence under phosphate limiting conditions, struvite is not the only component in the precipitate. T nay et al. (1997) used a 1:1:1 stoichiometric proportion of ammonia: magnesium phosphate for experiments. At the conclusion of the experiments the solution was filtered and concentrations of total ammonia and phosphate were measured in the filtrate. The difference in the initial and final concentrations of ammonia was assumed to have been precipitated as struvite. In an experim ent with initial pH of 9 and magnesium, total ammonia and phosphate concentrations of 14.26 mM each, 1714 mg/L struvite was precipitated. However, this concentration corresponded to the concentration of total solids predicted by the model. Struvite produc ed was only 1017.7 mg/L about 59.37% of the total solids produced. Formation of two calcium containing solids Ca 3 (PO 4 ) 2 (495.8 mg/L) and CaHPO 4 (384.4 mg/L) was observed. A 1:1:1 molar ratio of ammonia, phosphate and magnesium does not yield a predomin antly struvite precipitate. Similar observations are applicable to data from Stratful et al. (2001). The error in solids concentration which was predicted by the model and that measured in the experiments was mostly less than 10%. But a comparison of model predictions to T nay et al. (1997) and Lowenthal et al. (1994) gave the largest error which was more than 10%. Given the complexity of real wastewaters this researcher observed that the error in model prediction for the case of a mixture of landfi ll leachate and domestic wastewater was only 5%. Depending on the experimental conditions such as pH and initial ratios of magnesium, ammonium and phosphate, the solid phase can consist of several components not just struvite.
40 Therefore, equating total solids concentration to struvite can lead to erroneous results. The usual approach to accurately determining the struvite component in the solid phase involves redissolving the solid fraction after filtration by digesting in acid and measuring the ammonia concentration ( elen et al., 2007; Munch and Barr 2001; Wilsenach et al., 2007; Yoshino et al., 2003). In this approach, it is assumed that ammonia is precipitated only as struvite and the molar concentration of ammonia in sol ids is equal to the molar concentration of struvite. The model agreed well with experimental struvite data presented using this approach T he error in model predictions rang ed from 1.6% to 9%. The error in the prediction of struvite was much less than th e error in the prediction of total solids. Experiments using defined solutions were also carried out in the research laboratory to validate the model. A solution of 10mM of PO 4 3 was prepared by dissolving 1320 mg of ammonium phosphate (NH 4 ) 2 PO 4 in 1 lite r of distilled water. The total ammonia concentration was made up to 50mM by adding 1605 mg of ammonium chloride. A 200 mM magnesium chloride (MgCl 2 .6H 2 O) stock solution was used as a magnesium source. Precipitation experiments were carried out in 500 ml E rlenmeyer flask in which 475 ml of (NH 4 ) 2 PO 4 and NH 4 Cl solution were taken. To this solution 25 ml of magnesium chloride were added giving an initial NH 4 + :Mg 2+ : PO 4 3 ratio of 48.6:5: 9.5 or ~ 5:0.5:1. A magnetic stirrer was used for mixing. The pH of t he solution was continuously monitored with a pH probe. The pH of the solution was adjusted to 9.6 by adding 10 N NaOH. Precipitation was found to occur instantaneously after the addition of magnesium chloride solution which increased with the addition of NaOH. After reaching a pH of 9.6 the solution was stirred for another 5 min. Then the entire 500 ml solution was filtered using 0.45 m Whatman filter paper to recover the precipitate. The p recipitate was dried overnight in an oven at 104 o C and weighed. The
41 e xperiment was repeated three times. The average concentration of the precipitate measured in the experiments was 653 mg/L +/ 55 mg/L. The m odel predicted total solids of 682 mg/l. The error in the model prediction was about 4% which was consistent with errors in the model predictions for data in Table 2 4. The model was then applied to study the effect of various experimental conditions on struvite concentration and purity. 2.4.2 Effect of pH The pH is an important factor which m ainly determines the form in which ionic species exist in solutions. The model was used to investigate the effect of pH on the yield and purity of struvite (or struvite fraction of the precipitate). Chemically, since struvite contains 1 mole each of ammo nium, magnesium and phosphate most of the struvite experiments reported in the literature have been carried out using approximately 1:1:1 initial molar ratio of magnesium, phosphate and ammonia ( Altinbas et al., 2002; Jaffer et al., 2002; Miles and Elli s 2001; Stratful et al., 2001; Tnay et al., 1997; Uludag Demirer et al., 2005; Yoshino et al., 2003) At the equimolar ratio two simulations were performed first at 10 mM and a second one at 100 mM initial concentration of ammonium, magnesium or phosph ate keeping all the other species concentrations zero. The m aximum molar struvite yield in these experiments was assumed to be equal to the initial molar ammonium (or phosphate or magnesium) concentration used in the experiment. For the 10 mM solution t he maximum yield of struvite will be 10 millimole (i.e. 1370 mg/L). The total precipitate produced increased from 613 mg/L to 1247 mg/L as the pH increased from 5 to 10 and dropped to 1037 mg/L at pH 11. The mass fraction of struvite in this precipita te was only 0.6% at pH of 5, increasing to 89% at pH of 9.2 and then decreased thereafter. The concentration of struvite was the highest at pH 9.0 and was equal to 1122.3 mg/L. At pH of 9.2, the concentration of struvite was 1231 mg/L. The yield of struv ite was
42 0.25% at pH 5, 79 % at pH 9.0 and 57% at pH 11. At pH 9 and below the primary precipitate produced was MHP decreasing from 99.4% of solids at pH 5 to 21% of the solids at pH 9. The MHP decreased to 5.5% of the solids at pH 10. Above pH of 9 the predominant precipitate was MP8 and Mg(OH) 2 The simulations were repeated at a higher initial concentration of 100 mM. The struvite fraction trend was similar. The optimum pH for maximum struvite fraction was 9.2 and its concentration at this pH w as 12,670 mg/L. The total precipitate concentration increased with increasing pH up to pH 10 and then decreased. The total precipitate was 13,410 mg/L at pH 10. The effect of pH on struvite purity follows a different trend when the molar ratios of ammoni um, magnesium and phosphate are set to values that yield struvite as the dominant precipitate. On such a condition is a molar ratio of 10:1.7:3.4 mM of ammonium, magnesium and phosphate. For this mixture the precipitate is made up of struvite and MHP with small amounts of Mg(OH) 2 appearing at pH above 9. The fraction of Mg(OH) 2 is only 0.45% even at pH 10. At pH 7, struvite is about 71% of the total precipitate with MHP being the rest. At pH of 8 the struvite fraction is 82% increases to 98.3% at pH 9.6 and then drops to 97.8% at pH 10. The total solids increased with increasing pH, peaking at pH of 9.8 and then dropping slightly. The effect of pH on the struvite fraction is plotted in Figure 2 1. It has been generally reported that the optim um pH for struvite precipitation is between 8.1 and 9.6 without reference to the conditions of precipitation ( Altinbas et al., 2002; Battistoni et al., 1998; Jaffer et al., 2002; Munch and Barr, 2001 ; Wang et al., 2006) The simulations here showed that t he pH that maximizes the struvite fraction in the precipitate is dependent on the initial molar ratio of ammonia, magnesium and phosphate. At the equimolar ratio the optimum
43 pH is 9.2. However, the total precipitate concentration was highest at pH 10. For conditions at which ammonia is in excess, the optimum pH is 9.6 and the precipitate is predominantly struvite. 2.4.3 Effect of I nitial M agnesium, P hosphate and A mmonium R atio The effect of the initial ratio of ammonium, magnesium and phosphate on struvite purit y was then simulated. Figure 2 2 shows the effect of changing the magnesium to phosphate molar ratio at two different ammonia to phosphate molar ratios. The simulations were performed at a base concentration of 10 mM, that is, an equimolar mixture contain ed 10 mM each of ammonia, magnesium and phosphate. A pH value of 8.7 was used for the simulations as it is close to typical pH values of wastewaters which range from 7.5 to 8.5. Moreover, since pH 8.7 was below the optim um value determined in prior simulations, studies at this pH provided insight into manipulating the molar ratios to maximize struvite purity. This researcher found that for a given ammonium to phosphate ratio, increasing magnesium decreases the fraction o f struvite in the precipitate. The mass concentration of the total precipitate increases reflecting the high magnesium dosage. For an equimolar concentration of ammonium and phosphate, extremely limiting concentrations of magnesium produced precipitate w ith about 89% struvite. This purity decreases rapidly as magnesium dosage is increased. The struvite purity is the lowest at 27% when magnesium is of the same concentration (i.e. 1:1:1 ratio). When magnesium is increased further the struvite purity i ncreases to 50% when the magnesium to phosphate ratio is 1.5 and then decreases slowly thereafter. At the pH value used, the other dominant compound in the precipitate was MHP when the magnesium to phosphate ratio is less than 1.0 decreasing from 70% to 13% as struvite fraction increases. At the magnesium to phosphate ratios greater than 1.0, MHP decreased from 27% to 16% and MP8 increased from 19% to 33%.
44 At conditions where ammonia is in excess compared to phosphate, for example, the ammonia to phosp hate molar ratio of 2.94, the struvite purity decreased very slowly from 87% at magnesium limiting conditions to 75% when magnesium and ammonia were at equal concentration (i.e., magnesium to phosphate of 3). This researcher noted a slight dip occurred in the trend line at equimolar concentrations of ammonia, magnesium and phosphate. The MHP was the other major component in the precipitate, increasing from 13% to 16% as the magnesium concentration increased. At high magnesium concentrations MP8 and Mg(OH ) 2 were also precipitated. Struvite fraction in the precipitate is enhanced when magnesium is limiting as compared to ammonia N concentration. Struvite precipitation is enhanced by an increase in magnesium dosage but it should not be in excess as compared to the ammonia N concentration. When the magnesium concentration was increased (still keeping ammonia N concentration in excess) then the struvite precipitation reaction was favored and it was driven toward right for solids formation. When t he concentration of magnesium increases beyond a certain ratio of Mg:N, the formation of precipitates other than struvite is induced which decreases struvite fraction in the solids. This ratio of Mg:N needs to be determined for a specific wastewater syste m by running simulations. Even at non optimal pH values struvite fraction can be increased by dosing magnesium at low concentrations. If ammonia is in excess when compared to phosphate, as is the usual case, adding magnesium at a concentration equal to that of phosphate may produce a precipitate with at least 82% purity. This will ensure removal of most phosphate. Any remaining ammonia can be removed by further dosing with phosphate and magnesium.
45 2.4.4 Determining E xion C oncentration for R eal W astewater which is a necessary step in designing a pH control system. The titration curve was used in the pr esent stud y to determine exion concentration used in the model. Exion concentration is a measure of difference between cations and anions which were not accounted for in the model simulations. Exion concentration is a necessary parameter for determining t he amount of acid/base required to achieve a specific pH. An average value of exion concentration was determined using the model from the known amount of acid/base required for pH adjustment. The average exion concentration was then used to calculate the a mount of acid/base required. In the present studies, the amount of base required was determined for increasing the initial pH of wastewater to a specific pH (8.7 in most of the cases) for struvite formation. This quantity of base was input to the model as a sodium concentration when sodium hydroxide is used for pH adjustment. The initial concentrations of total Mg, N, Ca, P, K and total inorganic carbon (TIC) were also input to the model. The value of TIC was taken from the literature (Lowenthal et al., 199 4). The model solution computed equilibrium concentrations of ionic and solids species. It also determined the value of exion concentration. The e xion concentration input to the model depended on the pH at which the struvite precipitation experiments were carried out. An average value of exion concentration was determined by performing simulations for different ranges of pH. Using the average value of exion, simulations were performed for determining the acid/base requirement ; a titration curve was generate d as shown in Figure 2 3. Experimentally, the titration curve was obtained by adding small amounts of 0.1M NaOH or 0.1M HCl to 50 ml of wastewater sample, and measuring the pH after each addition. The t itration curve is presented and discussed here for a naerobically treated wastewater from dairy
46 which had initial pH of 7.25. A s imilar approach could be taken for validating exion determination for other wastewater. The titration curve had a typical S shape, and it can be transformed into a buffer capacity profile with an appropriate mathematical algorithm. The S shape curve suggested that the pH does not change at a constant rate with the addition of a strong base. reag ent. The equivalen t point was found to be at pH 6.5 which defined the level at which pH steepest. This researcher assumed that in the titration curve measur ements, each pH measurement was a result of a chemical equilibrium. The model determined an average concentration of exions as 0.002566 M with a standard deviation of 0.00059M. The values of exion concentration in this case ranged between 0.001776M and 0.0 02937M. A titration curve was generated using the model and was found to be close to the experimental titra tion curve as shown in Figure 2 3. It is observed that for pH values between 2 and 6, the error in the model predictions of the amount of caustic ne eded for pH adjustment ranged from 2 % to 11.3 % The m odel predicted more caustic requirement s for pH values higher than 6 the error in model predictions being 4 % to 12.3%. 2.4.6 Effect of pH on S truvite P urity and P hosphorus R emoval in R eal W astewater The e ffect of pH on struvite purity and phosphorus removal was simulated for real wastewater (centrate and dairy flushwater). The input concentration values to the model are shown in Table 2 3. Alkalinity values reported in the literature were used (Loewentha l et al., 1994; Battistoni et al., 2002 ). Alkalinity was reported in mg CaCO 3 /L Alkalinity was converted in the model to total inorganic carbon (TIC) by calculating corresponding carbonate and bicarbonate concentrations at a given pH. Simulations were car ried out at two different
47 concentration ratios of Mg:P : 1:1 and 0.5:1. Ammonia N was in excess in all the simulations. Total concentrations of nitrogen, phosphorus, calcium and potassium were determined in the laboratory. Results of the simulations are sh own in Figures 2 4, 2 5, 2 6 and 2 7. When the Mg:P ratio of 1:1 was used for struvite formation in the centrate, 87 % removal of phosphorus was observed at pH 8.5 as shown in Figure 2 3. At this ratio, total solids increased from 339.66 mg/L to 605.35 m g/L as pH increased from 5 to 8.4. The o ptimum pH of 8.4 was observed S truvite purity at this pH was found to be 72.8 % although the purity increases to 72.91 at pH 9. In acidic pH range, concentrations of CaHPO 4 and MHP were dominant. The c oncentration o f CaHPO 4 was found to be maximum (70.14 mg/L) at pH 5 and decreases to 10.16 at pH 8.4. The c oncentration of CaMgCO 3 was found to increase from 65.56 mg/L at pH 5 to 145.36 mg/L at pH 8.4 and increase thereafter to 160 mg/L at pH 11. These two solid speci es were not included in the database o f models by elen et al. (2007 ) and Harada et al. (2006) due to their slow rate of formation. By ignoring these solids in the present model struvite purity could increase to 95 % This will be verified experimentally i n C hapter 3. Simulations were also carried out using the Mg:P ratio of 1:1 on dairy f lushwater as shown in Figure 2 5. Initial concentrations of various ionic s pecies were as shown in Table 2 3. Initial calcium concentration was found to be higher than that for centrate. Complete removal of phosphorus was observed in acidic pH range. Phosphorus was recovered in the form of CaHPO4. Struvite content in the solids was found to be negligible. A max imum struvite purity of 5.45% was observed at pH of 9.2. Phosphorus removal decreased with pH increase. Phosphorus removal decreased from 100 % to 88.5 % as pH increased from 5 to 9.2. The c oncentration of CaMgCO 3 was found to be maximum at pH 5 and decrease d thereafter for basic pH. The CaMgCO 3 was the dominant solid species for all pH values.
48 Both types of wastewater streams were subject to simulations with the Mg:P ratio of 0.5:1 and compared as follows. This researcher observed that phosphorus removal f rom both types of wastewater was higher for 1:1 ratio of Mg:P as compared to the Mg:P ratio of 0.5:1. Struvite purity was also found higher in the case of 1:1 ratio. Struvite purity was found highest at pH 8 for 0.5:1 ratio for centrate as shown in Figur e 2 4. In this case, total solids decreased as the Mg:P ratio was decreased to 0.5:1. At the Mg:P ratio of 0.5:1, prominent solid species were CaMgCO 3 and CaHPO 4 Changes in these solids were similar as explained for the case of the Mg:P ratio 1:1. Simula tions with the Mg:P ratio of 0.5:1 showed similar changes in concentrations of CaMgCO 3 and CaHPO 4 Struvite formation and purity were negligible for dairy flushwater as well at the ratio of 0.5:1 as shown in Figure 2 6. It was therefore clear that with e xcess ammonia N concentrations and an Mg:P ratio of 0.5:1 this was not suitable for struvite formation in both types of wastewater because precipitation resulted in mainly calcium phosphate species. 2.5 Findings A model was developed for predicting precipitation in closed systems containing solutions of ammonium, magnesium and phosphate. The model incorporates 15 different precipitates and explicitly solves for precipitate, residual ion and dissolved species concentrations using mass balance equat ions for magnesium, phosphorus and nitrogen along with chemical equilibri um and charge balance equations Using the Polymath program and a solution procedure that involves converging the residual phosphate concentration to within tolerance limits, the co ncentration of the different precipitates, residual ion and dissolved species concentrations was determined The model was validated against data collected from literature for synthetic and real wastewaters. The model was able to predict struvite to with in 1.6% to 9% and total precipitate prediction errors ranged from 1 % to 24.5% It found that for solutions containing Mg, NH 3 and PO 4 the optimal pH for struvite concentration depends on the initial ratio of ammonia, magnesium and phosphate. A pH of 9.0 optimizes struvite concentration when the ratio is 1:1:1 and a higher pH of 9.8 when magn esium and phosphate are limiting
49 The e quimolar stoichiometric ratio of magnesium, ammonium and phosphate (i.e, the ratio of their occurr ence in struvite) was not ideal for struvite precipitation. To obtain pure struvite it was necessary to have excess ammonia in the solution with magnes ium being the limiting nutrient A titration curve was generated by using the exions concentration and was validated using an experimental curve. The m odel determined an exion concentration of 0.002566 M in anaerobically treated wastewater from dairy operations. This species concentration should be considered while calculating the amount of c austic needed f or pH adjustment The m odel accurately predicted the amount of base needed for pH adjustment with an error less than 12.3% as compared to the experimental quantities A p resence of calcium and carbonate species was found to dec rease struvite purity In the case of centrate, excess ammonia concentration and an Mg:P ratio of 1:1 were predicted to be suitable for struvite formation considering the purity of struvite of 95% at this ratio In the case of centrate and dairy flushwater, struvite precipitation was not suitable for the Mg:P ratio of 0.5:1 with excess ammonia because phosphate was mainly removed as calcium phosphate and calcium magnesium carbonate.
50 Table 2 1. List of solids included in the comprehensive model
51 Table 2 2 Equilibrium expressi ons with equilibrium and solubility constants Equilibrium equation pK 9.0 12.3 7.2 2.14 6.3 10.2 14.0 2.56 4.8 5.7 0.45 23.98 22.89 0.15 10.7 12.7 25 6.57 16.7 5.3 5.0 4.67 18.995 36.16 36 46.97
52 Table 2 3 Input parameters to the model for simulations on the effect of pH Component (mg/L) Centrate Dairy Flushwater Nitrogen 750 740.88 Phosphorus 120 50.66 Calcium 35 140.5 Total Potassium 110 357.78 Alkalinity as mg/L of CaCO 3 4750 1270
53 Table 2 4 Validation of the c hemical e quilibrium m odel Ref. Type of Wastewater Initial Concentrations mM Exptl Struvite (mg/L) Model Predictions Predicted Struvite fraction in solids Error $ in Model Prediction of struvite pH Mg T P T N T Struvite (mg/L) MP8 (mg/L) MP22 (mg/L) Mg(OH) 2 (mg/L) MHP (mg/L) Total solids (mg/L) (%) (%) Lowenthal et al. (1994) Solutions prepared by adding NH 4 Cl, K 2 HPO 4 MgCl 2 carbonate and acetate 6.8 8.23 12.9 21.43 601* 300.2 0.001 0 0 432.9 733 40.95 18 Harada et al. (2006) Synthetic urine containing PO 4 NH 4 Na, Mg, K, Ca, Cl, citrate, carbonate 8 20 13.45 20.18 1685* 1317 88 11.54 17.72 222.9 1687.7 78 1.23 Wilsenach et al. (2007) Synthetic urine containing PO 4 NH 4 Na, Mg, K, Ca, Cl, citrate, carbonate 9.4 7.415 14.83 18.7 1045 987.07 0.001 0 0.061 14.75 1001.89 98.52 5.87 Wilsenach et al. (2007) Synthetic urine containing PO 4 NH 4 Na, Mg, K, Ca, Cl, citrate, carbonate 9.4 14.83 14.83 18.7 2011 1845.03 7.03 0.922 11.625 80.25 1944.86 94.87 9 Celen et al. (2007) Liquid swine manure 8.5 2.39 5.51 80 338 322.2 0 0 0 2.7 324.9 99.17 4.9 Much and Barr (2001) Supernatant from anaerobically digested sludge dewatering centrifuge 8.5 1.51 1.9677 43.88 195 200.86 0 0 0.003 3.08 203.95 98.49 2.92 Yoshono et al. (2003) Anaerobic digester effluent supernatant 8.5 7.025 6.387 24.5 805 818.23 1.006 0.132 12.157 29.68 883.02 92.66 1.62 Tnay et al. (1997) Synthetic samples prepared by using MgCl 2 NaH 2 PO 4 NH 4 Cl 9 14.26 14.26 14.26 1714* 1017.7 44.44 5.82 284.9 4.75 2273 44.75 24.5 Table 2 4. Continued
54 Ref. Type of Wastewater Initial Concentrations mM Exptl Struvite (mg/L) Model Predictions Predicted Struvite fraction in solids Error $ in Model Prediction of struvite pH Mg T P T N T Struvite (mg/L) MP8 (mg/L) MP22 (mg/L) Mg(OH) 2 (mg/L) MHP (mg/L) Total solids (mg/L) (%) (%) Altinbas et al. (2002) Domestic Wastewater + 2% landfill leachate (DWL3 sample) 9.2 7.785 7.785 7.785 1420 1495 7.83 1.02 7.7 104.1 1616.1 92.5 5.01 Battistoni et al. (1998) Supernatant from sludge centrifuges in a biological nutrient removal plant 8.12 1.54 2 44.5 210.98 198.21 0 0 0.001 6.56 204.77 96.79 6.44 Burns et al. (2001) Swine Waste 9 9.736 6.085 12 758.6 705.47 44.963 5.896 161.05 54.53 971.91 72.59 7.53 Stratful et al. (2001) Deionized water with varying concentration of Mg 2+ NH 4 + and PO 4 3 10 7.692 7.81 14.77 1629 1757 0.67 0.04 17.45 15 1790.5 98 7.2 $ Error in struvite concentration predicted by the model and that measured experimentally For these data the error between solids concentration as predicted by the model and struvite reported in the experiments was calculated.
55 Figure 2 1 Effect of pH on struvite purity for ammonium, magnesium and phosphate ratios of 1:1:1 (at 10 mM and 100 mM) and 10 mM : 1.7 mM : 3.4 mM. 0 10 20 30 40 50 60 70 80 90 100 4 6 8 10 12 14 Struvite in Solids (%) pH 10:10:10 100:100:100 10:1.7:3.4
56 Figure 2 2 E ffect of molar ratio of magnesium: phosphate on % struvite and total solids for two different mola r ratios of ammonium (A) and phosphate (P) at pH = 8.7
57 Figure 2 3. Titration c urve for anaerobically digested wastewater from dairy operations 0 2 4 6 8 10 12 14 0 20 40 60 80 100 120 pH ml of 0.1 N NaOH Experimental Model
58 Figure 2 4 Effect of pH on struvite purity and phosphorus removal in centrate at an Mg:P ratio of 1:1 Figure 2 5 Effect of pH on struvite purity and phosphorus removal in centrate at an Mg:P ratio of 0.5:1 0 100 200 300 400 500 600 700 0 10 20 30 40 50 60 70 80 90 100 4 6 8 10 12 Total Solids (mg/L) % pH Struvite (%) P Removal (%) Total Solids 0 50 100 150 200 250 300 350 400 0 10 20 30 40 50 60 70 4 6 8 10 12 Total Solids (mg/L) % pH Struvite (%) P Removal (%) Total Solids
59 Figure 2 6 Effect of pH on struvite purity and phosphorus removal in dairy flushwater at an Mg:P ratio of 1:1 Figure 2 7 Effect of pH on struvite purity and phosphorus removal in dairy flushwater at an Mg:P ratio of 0.5:1 0 200 400 600 800 1000 1200 1400 1600 0.00E+00 2.00E+01 4.00E+01 6.00E+01 8.00E+01 1.00E+02 1.20E+02 4 5 6 7 8 9 10 11 12 Total Solids (mg/L) % pH Struvite (%) P Removal (%) 420 430 440 450 460 470 480 490 500 0.00E+00 2.00E+01 4.00E+01 6.00E+01 8.00E+01 1.00E+02 1.20E+02 4 5 6 7 8 9 10 11 12 Total Solids (mg/L) % pH Struvite (%) P Removal (%)
60 CHAPTER 3 S EQUENTIAL BATCH REAC TOR FOR RECOVERY OF NITROGEN AND PHOSPHO RUS AS STRUVITE FROM SEW AGE SLUDGE CENTRATE AND DAIRY WASTEWATER 3.1 Remo val of Nutrient s from Wastewater The review presented in this chapter discusses the principles of struvite crystallization and also examines the techniques and processes in experimentation to date by researchers The experiments maximize phosphorus removal and reuse as struvite obtained from wastewater effluents. Untreated nutrient rich wastewater is problematic and causes eutrophication of receiving waters. Removal of nutrients becomes challenging as regulatory authorities tight en discharge standards. Significant costs are associated with the extra treatment processes required to meet the discharge standards (Giesen, 1999). The most widely used technologies for nutrient removal which are adopted by wastewater treatment plants i nclude biological nitrification/denitrification for nitrogen removal and metal salt precipitation or biological treatment for phosphorus removal. A s ignificant amount of sludge is generated in the wastewater treatment plant which is anaerobically digested and dewatered using centrifuges and gravity belt thickening. This produces centrate and filtrate streams containing 750 to 1000 ppm of ammonia N and 100 to 200 ppm of phosphate P. Most wastewater treatment plants return these streams to the plant headwork s where it adds to the wastewater burden. Currently, an approach to remove nitrogen and phosphorus from wastewater effluents which is gaining attention is through the crystallization of phosphate compounds such as calcium phosphate or struvite (MgNH 4 PO 4 .6H 2 O) (Doyle and Parsons, 2002). This approach can be seen as resource recovery and may contribute positively to the economics of wastewater treatment.
61 3.1.1 Factors A ffecting S truvite F ormatio n: Concentration and N ature of C onstituents The concentratio n of constituents in the liquor must exceed the solubility product of struvite for crystal growth to begin (Taylor et al., 1963; Ohlinger et al., 1998). Usually magnesium is a limiting constituent for struvite precipitation. The a mount of magnesium added t o the influent is varied to achieve an appropriate ratio of magnesium to phosphate within the influent. Some reactors raise the magnesium concentration to achieve a 1:1 ratio with phosphate for example, the Shimane and Osaka plants in Japan ; others raise it above a 1:1 ratio. Burns et al. (2001) showed that the magnesium to phosphorus ratio of 1.6:1 is ideal because not all magnesium will be available for precipitation due to the formation of other complexes within the solution. Magnesium is a lso added to counteract the influence of calcium ions within the influent. The calcium to magnesium ratio cannot be greater than 1:1 due to the influence of calcium species on struvite precipitation. Generally calcium is in excess of phosphate within liqu ors so a ratio of greater than 1:1 Mg:Ca should be set within a recovery operation to stop the detrimental effects of calcium on struvite precipitation. Two forms of magnesium are typically used. Magnesium chloride is the preferred choice within the litera ture though magnesium hydroxide has also been trialed (Munch and Barr, 2000; Ueno and Fujii, 2001; Miles and Ellis, 2001). Magnesium hydroxide is an attractive choice for reactors because it will raise the pH of the solution, which would lower the cost of raising the pH with the addition of a base but the pH cannot be altered independently of magnesium addition. Ueno and Fujii (2001) use d sodium hydroxide to alter the pH to offer finer control. Magnesium chloride is cheaper than magnesium hydroxide and is readily soluble in water. Due to these reasons the use of magnesium chloride is more popular.
62 3.1.2 Effect of pH The most important condition within a reactor is the pH of the liquor. The pH choice plays an extensive role in the speciation of reactants, the speed of reaction, and the nature of the product. By far the easiest method to alter pH is by the addition of a base such as sodium hydroxide. The addition of a base is an expensive process and methods to avoid this procedure have been well researc hed. A brief review of pH values used in struvite precip itation is presented in Table 3 1. The Treviso and Geesterambacht treatment plants both employ an air stripping unit for the removal of carbon dioxide from the liquor. Carbon dioxide within the liquor reversibly forms carbonic acid (H 2 CO 3 ) thereby lowering the pH. The removal of this CO 2 by injection of air into the liquor raises the p H without the addition of a base (Battistoni et al. 1997). This attractive method requires a substantial reactor volume for the removal of CO 2 from the liquor which makes it an expensive investment. Ohlinger et al. (1999) show ed that vigorous stirring of solutions raise d the pH from 8 to 8.5 but this took a mixing speed of 500 rpm The economics of such a reactor should be taken into account. Chemical speciation within the liquor has been shown to depend on pH (Ohlinger et al. 1998). Phosphate and ammo nia exist in different forms at different pH values. The solubility of struvite was linked by Ohlinger et al. (1998) to the speciation of these components. Ohlinger et al. (1998) show ed that the presence of magnesium phosphates within the solution lowered the solubility of struvite by the removal of these vital components, with their presence prevailing at pH 8.5. The lowest point of struvite solubility had been shown to be at pH 9. A rise above pH 9 saw the stripping of ammonia from the liquor and an incre ase of phosphate concentration (Booker et al. 1999), which has been linked to the over production of fines due to the stripping of ammonia from the liquors.
63 The rate of formation of struvite has been shown to be controlled by the pH F or a commercial reco very operation this is important. It was show ed that by increasing the pH to 9 nucleation could be achieved within a few minutes. One problem with such a high pH is the production of fines. Fines have been a problem for a number of authors (Mangin, 2004; Le Corre et al., 2006). Due to the production of fines at high pH levels, Battistoni et al. (2005) were able to show that by setting the pH to 8.5 the precipitation of struvite could be changed from a homogeneous to heterogeneous process with the use of seed material. The selection of the pH will be dependent on the conditions required within the reactor, that is a seeded or non seeded reactor. 3.1.3 Effect of Temperature Literature on the effect of temperature on struvite formation gives a different explanation in every study. Borgerding (1972) studied the influence of temperature on struvite solubility in the anaerobic digestion of a wastewater treatment plant. He concluded that when the temperature increased from 0 C to 20 C, struvite solu bility increased. However, above this temperature solubility was found to decline steadily with an increase in temperature. Webb and Ho (1992) assumed 30 C to be the temperature of maximum struvite solubility in experiments designed to define thermodynami c parameters for its crystallization. On the other hand, Webb (1988) measured the solubility of struvite in various salt solutions at temperatures ranging from 0 to 80 C. Webb (1988) concluded that increasing temperature causes a rise in struvite solubili ty, especially for more concentrated solutions. elen and Trker (2001) reported that the reaction temperature did not have an influence on struvite solubility in an anaerobic digester effluent with temperatures 25 to 40C. The e ffect of temperature on st ruvite formation is not studied here. All studies on struvite precipitation were carried out at room temperature (25C) for typical wastewater temperatures.
64 3.1.4 Hydraulic Retention Time (HRT) The hydraulic retention time ( HRT ) of a reactor is of particul ar importance to the design. Ohlinger et al. (2000) reported that struvite removal efficiency exceeded 80% when the HRT exceeded one hour. Munch and Barr (2001) found that after one to two hours, HRT ha d no effect on the effluent otho phosphate (OP) concen tration. However, elen and Trker (2001) concluded that struvite precipitation is very fast and completed in minutes. Therefore, they assumed a 40 minute reaction time for process equilibrium. To achieve larger crystal sizes a high HRT (10 days) is coup led with a recycling unit in the Phosnix process. Mamais et al. (1994) found that an HRT of 14 days was most suitable and achieved a 91% removal of phosphate, though little information is available on product size or quality. Other authors have found that shorter HRTs have yielded good results. Suzuki et al (2005) found that an HRT of 22 h was suitable for the accumulation reactor Short HRTs are more attractive due to lower costs W ays of achieving this should be further investigated T his is one of the important considerations of the present research. 3.2 Review of Reactor Designs for Struvite Production Several laboratory and pilot scale studies have been carried out to assess the potential of struvite recovery methods in removing and recovering phosphorus as a reusable product A few studies have been tested at full scale in The Netherlands (Giesen, 1999) and Italy (Battistoni et al., 2005a; Battistoni et al., 2005b). However Japan and Canada are the only countries where 90% of the p hosphorus removal and recovery from anaerobically digested sludge liquors as struvite ha ve been implemented with the resulting product sold to fertilizer companies (Gaterell et al., 2000; Ueno and Fujii, 2001). C urrently five full scale phosphorus recov ery operations exist three in Japan and two in Europe. All five operations share a similar design philosophy with the only differences being in
65 seed material used and the way the desired pH is achieved. The Shimane plant reactor has two internal compartme nts for the reaction and settling of struvite. The inner section of the reactor is the nucleation zone and the outer section of the reactor is the separation zone. The separation zone allows the settling of struvite crystals with collection happening abo ut every 10 days by discharge from the bottom of the reactor. The smaller struvite crystals, fines, are returned to the reactor to provide new seed material for the reactor to continue operating. With the Crystalactor a similar reactor design is seen but instead of recycled struvite acting as seed material for the growth of the product, qu artz sand is used to grow HAP The influent is pumped in an upward direction at high velocity (30 50 m/h) to enable a complete fluidization. Pellets o f HAP sand grow and as their mass increases there is not enough pressure to keep them fluidized and so they settle at the bottom the same design as the Crystalactor with pellets grown on sand and removed in the same manner. The design philosophy for an FBR reactor in the literature is quite standardized. T wo main options exist which are varied in all the operations : how to achieve operation pH and collection of the product. Aeration has been demonstrated as a good method for the raising of the pH and has been successfully demonstrated at Treviso and Geesterambacht. The addition of a base though is the far easier method that does not require the installation and ope ration of a further unit The NaOH is a relatively cheap chemical to purchase, but amounts required would vary depending on the alkalinity of the liquors. The other choice when designing a recovery operation is the method of recovering the product. Settlin g within the reactor is considered the easiest method available and is achieved by the use of seed material at Shimane, Treviso and Geesterambacht. T wo other options are available for the recovery of the final product : accumulation device s and coagulating polymers. Accumulation devices such as steel mesh as
66 reported by Suzuki et al (2005) and Le Corre et al (2006) have both shown great results for the collection of struvite and initial results from this project have confirmed these results. One probl em with this method however, is how to transfer it to a full scale operation. Unfortunately it is difficult to set a machine to clean the mesh and so it would need to be done by hand, adding costs to the operation. Struvite crystallization is typically carried out using chemical additions. Three elements are typically altered within the reactor to aid precipitation of the desired product 1. Chemical constituents are altered to achieve supersaturation or metastable conditions 2. The influent pH is altered commonly by the addition of a base or the use of aeration to strip carbon dioxide so that the nucl eation step can be made faster 3. The a ddition of seed material allows for the growth of struvite crystals on the surface of seed mat erial which in turn allows for the growth of larger particles, which facilitates easi er collection from the reactor Typical designs involve the use of stirred tank reactors and fluidized bed reactors which are next review ed 3.2.1 Stirred Tank Reactors Stirred tank reactors allow for even distribution of chemicals throughout the system thus preventing any limitations on chemicals at crystal growth sites (Ali and Schneider, 2006). The nature of crystal growth requires a constant source of chemicals for continued growth (Ohlinger et al., 1999). Key unit operations can be varied using a continuously stirred tank to reach specified results thus using the system for modeling. The stirring is generally carried out 1) by a mechanical paddle that operates at a constant speed, 2) by recirculation of the reactant through the reactor (Ali and Schneider, 2006) or 3) by the aeration of the solution with air Large scale stirred reactors for the production of struvite are not widely reported in the literature. Burn s et al. (2001) used a 140,000 liter tank for the production of struvite from animal
67 waste material with the addition of 2000 L of magnesium chloride; a 90% removal of phosphate was achieved. The recovered product was used as a fertilizer on nearby fields but unfortunately no results were given on how these crops were affected A t 13,000 L the struvite reactor was trialed by Jaffer et al. (2001) at Slough S ewage Treatment Works in England with mixing achieved by aeration. With a relatively short residenc e time of one hour within the reactor, the addition of magnesium chloride and sodium hydroxide for pH elevation achieved a phosphate removal of 64%. Along with good phosphate removal large quantities of struvite were produced which were subsequently tria led successfully as a fertilizer when blended with a potassium source (Johnson et al., 2004). Mangin et al. (2004) reported the effect of seeding a stirred reactor. The authors found that seeding the 21L reactor with sand had no effect on changing the precipitation of struvite from homogeneous primary nucleation to heterogeneous primary nucleation. The reactor produced a large quantity of fines which was linked to the over supersaturation of the solution. Le Corre et al. (2006) saw t his lack of nucleat ion in atomic force microscopy (AFM) studies on different types of seed material. Le Corre et al. (2006) also observed that sand had the second lowest affinity to stick to struvite. Mangin et al (2004) trialed struvite as seed material in additional exper iments. Struvite was shown to be ineffective also as seed material for the stirred tank reactor due the production of a large amount of fines. Mangin et al. (2004) reported that even with dilution of the influent over production of fines still occurred wi thin the reactor. 3.2.2 Fluidized Bed Reactor (FBR) In the fluidized bed reactor (FBR) the liquor passes upward through a bed, generally made up of small particulate matter, at sufficient enough velocity to fluidize the media (Thomas, 2007). This process is used in all full scale phosphate recovery processes that is, Geesterambacht, Trevis o and Shimane. At Treviso (Battistoni et al., 2000) and Geesterambacht
68 (Van Dijk and Wilms, 1991) the bed is quartz sand and at Shimane (Phosnix process) (Ueno and Fujii, 2001) the process uses recycled struvite fines. The fluidization process increase s the surface area of the particulate matter because all faces of the solid are now accessible as reaction sites [NIR] (Metcalf and Eddy, 2001). By using a high enough air inflow the solid material is able to overcome any aerodynamic drag and gravity to b ecome suspended within the liquid. The fluidization of the particles is used to achieve a greater level of mixing within the reactor thus achieving a uniform mixing of reactants. A greater mixing of reactants increases the nucleation rate and lower s the c hances of local supersaturation levels within the liquor which can lead to the over production of fines (Ohlinger et al., 1999). The FBR process for recovery of phosphate is generally divided into two phases : the air stripping phase and the reaction phase (Ueno and Fujii, 2001; Battistoni et al., 2005). The air stripping phase of the process is used to increase the pH of the solution by removal of carbon dioxide from the liquor. The air stripping phase can have a significant effect on the rate of precipitat ion where Battistoni et al. (1998) reported the reaction period can be reduced from 8 days to 100 min. Air stripping allows for the phosphate recovery process to proceed with a reduction in the amount of a base needed to raise the pH of the influent. Thou gh FBR reactors seem to be the most popular type stirred reactors have been trialed at two sites along with many bench scale tests. Stirred reactors have their own problems but one big advantage is that they are relatively simple to operate. The reactor design of Slough Sewage Treatment Works was designed with the same principles as that of the Phosnix process that is, two zones within the reactor : one for nucleation and another for settling. Limitations of existing struvite recovery techniques include :
69 Reliance on in situ crystallization and agglomeration which results in large reactor volumes to ensure appropriate residence times for successful agglomeration. Formation of crystals is favored as it reduces downstream processing cost s to make salable pr oduct s. Overall reaction rate (consequently reactor volume) controlled by the crystallization process rather t han precipitation High c apital c osts for these systems Reactors used in the commercial processes are fluidized bed or membrane reactors w hich in cur high operating costs These systems achiev ing only 90% removal of phosphate 3.3 Overview of Sequencing Batch Reactor (SBR) Technology 3.3.1 Concept of Sequential Batch Reactors (SBR) The sequential batch reactor ( SBR ) technology is considered to be an a lternative to conventional processes such as a stirred tank reactor for nutrient removal from wastewater. This configuration has a higher flexibility and controllability, allowing more rapid adjustment to changing influent characteristics (Baetens, 2000). L ower investment and a recurrent cost occur because secondary settling tanks and sludge return systems are not required (Novak and Lindtner, 2003). Furthermore, it is especially appropriate for places where there is significant flow and load variability ( Metcalf and Eddy, 2003) or where space problems become a restriction. The SBRs are widely and commonly used in biological wastewater treatment (Mace and Mata Alvarez, 2002). The SBR technology has been successfully applied in wasterwater treatment plants that treat urban and industrial wastewater (Puig et al., 2005). 3.3.2 Overview of F ull scale A pplications of SBR T echnology Alleman and Irvine (1980) demonstrated the potential of the SBR technology to maintain combined organic carbon oxidation and nitrifi cation. Irvine et al. (1983) concluded t hat the SBR technology is a viable alternative to conventional continuous flow activated sludge treatment of domestic wastewater, nitrification, denitrification and chemical precipitation of phosphorus in
70 the first SBR demonstration site. Norcross (1992) considered the mechanical, process and control aspects for the design of an SBR. Ketchum Jr. (1997) described the SBR physical system and explained approaches used to develop the bases of design needed to meet many different treatment objectives especially for the feeding and reaction periods. Arora et al. (1985) evaluated the SBR technology at several plants in the United States, suggesting that equalization, ideal settling, simple operation, compact layout and cost savings are the major advantages of SBR systems versus continuous flow systems. Okada and Sudo (1985) studied the simultaneous removal of phosphorus in a lab SBR. Rim et al. (1997) carried out successful full scale tests for biological nutrient re moval ( BNR ) purposes using the SBR technology treating between 47.3 and 52.8 m 3 /day of sewage. Helmreich et al. (2000) investigated the performance of SBR plants in operation in Bavaria, Germany. The sizes of SBR plants in Bavaria range from 400 to 25 000 Population Equivalents (PE). Teichgraber et al. (2001) achieved complete biological nitrogen removal at full scale in a single tank SBR treating 850 m 3 /d of domestic wastewater. Torrijos et al. (2001) concluded that SBR technology is, from a technical po int of view, perfectly adapted to treating cheese production wastewater. Steinmetz et al. (2002) evaluate d i n view of their effluent quality, treatment efficiency and energy deman d f our SBR wastewater treatment plants, which were designed for approximatel y 5 000, 8 000, 15 000, 25 000 PE. The study proved that the SBR technology is a good alternative for municipal sewage plants and can help to save investment costs. Two full scale SBR plants in Australia processing between 2000 and 2500 m 3 /d of wastewater which carried out enhanced conventional SBR plant, currently under construction in Malaysia for 1.2 million PE.
71 3.3 .3 Characteristics of SBR T echnology Operational steps in a typical SB R are illustrated in Figure 3 1. The SBR technology is a fill and draw activated sludge system for wastewater treatment. While in continuous systems the reaction and settling occur in different reactors With the SBR technology, all the processes are cond ucted in a single reactor following a sequence of fill, reaction, settling and draw phases. The cycle configuration depends on the wastewater characteristics and legal requirements. The fill phase can be static, mixed or aerated depending on treatment o bjectives. A static fill results in minimum energy input and high substrate concentration at the end of the fill phase. A mixed fill results in denitrification if nitrates are present, and provides anaerobic conditions required for biological phosphorus re moval. A n a erated fill result s in the beginning of aerobic reactions keeping substrate concentration low, which may be of importance if biodegradable constituents exist that are toxic at high concentrations (Ketchum Jr., 1997). The feeding phase could be single or multiple feeding d epending on treatment objective. During the react phase, usually under mixing conditions, the biomass consumes the substrate under controlled environmental conditions (aerobic, anoxic or anaerobic) depending on the wastewater treatment. In the aerobic react phases, the organic matter oxidation and nitrification take place. The c lassical heterotrophic denitrification process and the phosphorus uptake require anoxic conditions. During the anaerobic phase, phosphate is released into the liquid phase. Sludge wasting is another important step in the SBR operation that greatly affects performance (M e tcalf and Eddy, 2003). Its target is the regulation of the sludge solids concentration in the reactor. Sludge wasting could b e done at the end of the reaction phase or during the settling phase.
72 In the present study, a sequential batch reactor (SBR) is utilized for producing struvite. As prevously discussed, such a type of operation is becoming increasingly common in wastewater treatment plant s This operation can offer operating flexibility, better control and cost effectiveness. Operational details for an SBR are established by carrying out various experiments and pilot plant experience. Details of an SBR operation for struvi te formation are presented in this chapter. 3.4 Objectives The main purpose of this study was to test the effectiveness of the sequential batch reactor (SBR) operation to recover struvite. Experimental studies were carried out with the following five obje ctives: To d etermine quantity of magnesium chloride and sodium hydroxide required for struvite precipitation in synthetic and real wastewater To investigate operating details for complete nitrogen and phosphorus recovery in an SBR To test the possibility o f using aeration as a method to increase pH for struvite precipitation To d etermine the nature of the precipitate obtained after struvite formation To s tudy settling characteristics of the solid sludge generated after struvite precipitation 3.5 Materials and M ethods Experiments were performed on t he following samples fresh from different wastewater streams: 1) c 2) s ynthetic solutions containing nitrogen and phosphorus 3) d airy flu shwater collected after grit removal 4) d airy wastewater from an anaerobic lagoon and 5) d airy effluent from an aerobic lagoon Dairy flushwater was collected after mechanical separation and sedimentation. Primary treatment (screening, sedimentation, or both) of flushed dairy manure is widely practiced in the
73 dairy industry since it is required to improve the operation of som e wastewater irrigation systems In addition, screening and sedimentation to remove solids are useful in reducing the organic loading rate to anaerobic lagoons to extend their capacity and reduce the frequency of sludge removal. A conical bottom fluidized bed sequential batch reactor with a total volume 10.6 L was utilized ( Figure 3 2). T he dimensions and locations of ports on a conical bottom reactor were as follows: Diameter: 6 Height of cylindrical section: 1 7 Height of conical section: 1 Ports at following locations from top: 2@ 3 ; 1@ 9 ; 1@ 17.5 1 on conical section @ 2 1@ 29 .5 Lid with four equidistant ports of 1/4 inside the diameter: 2 ports for exhaust, one for air inlet and one for sampling The v olume of the cylindrical section and conical section were 2.5 L and 8 L respectively. Centrate was fed from the top of the r eactor. Magnesium chloride and sodium hydroxide were added to the reactor from the top. The r eactor was fluidized by air through a sparger at the bottom. This provided mixing which ws required for the chemicals. The d esired pH was set using 10 N NaOH. Mag nesium was dosed to the reactor in the form of magnesium chloride solution, to ensure the desired Mg:P molar ratio in the reactor. The reaction of struvite formation start ed instantly and it was carried out for 30 min. The o peration was carried out in a sequential batch mode in steps to remove the phosphorus completely. The s truvite precipitate was allowed to settle for 30 min and then harvested at the end of the batch. Ammonia nitrogen and phosphorus were analyzed according to Standard Methods (APHA, 1992). Potassium and calcium were analyzed using i on selective electrodes purchased
74 from Hach. Standard methods were adopted as suggested in the APHA manual titled Standard Models for the Examination of Water and Wast ewater. Solids were recovered by opening the ball valve at the bottom of the SBR. A part of these solids was dried to find the moisture content. The p recipitate was dried overnight in an oven at 104C and weighed. Solids were analyzed for nitrogen, phospho rus, calcium and magnesium. Solids were also analyzed using x ray diffraction (XRD), using a Philips x scanning rate of 2.0 2 min The s truvite procured from Fisher Scientific was used as a stand ard for comparison. For struvite precipitation, wastewater was filled in the reactor. After filling, the precipitation reaction was carried out by feeding appropriate amounts of caustic and magnesium chloride calculated by a chemical equilibrium model des cribed later. The reaction time was approximately 30 min. After the reaction precipitated struvite was allowed to settle for about 5 min. Centrate was decanted from the top and struvite was drawn from the bottom of the reactor. The settled struvite sludg e was taken through other processing steps to make the final product. Experiments were also carried out to study struvite settling characteristics. Anaerobically digested dairy manure was also subjected to struvite formation and settling. Solids present in the dairy manure were allowed to settle and solid samples were taken at different heights of the reactor to quantify struvite accumulation at specific time. Samples were centrifuged at 5000 rpm for 15 min then decanted dried and weighed to determine solid content. The s equential batch reactor was filled with anaerobically treated dairy manure with a total volume of 9.2 L. Mixing was provided with air fluidized from the bottom of the reactor. Samples were taken when the solid s were uniformly distributed. Raw dairy manure was allowed to settle and samples were taken at 15 min and 30 min after mixing was stopped. Samples were taken at four different
75 depths of the reactor (5 17 24 and 31 ). A 100 ml of sample was centrifuge d and decanted. A s olid pellet in the centrifuge tuve was resuspended in 20 ml deionized water which was then dried at 105C and weighed. The h eight of the liquid level from the top of the reactor was found to be at 5 T he t otal height of the reactor inc luding was 31". The d airy manure was subjected to struvite formation by adding 1.3 g/L of MgCl 2 .6H 2 O at pH of 8.7. The e xperiments were carried out at room temperature (25C). The r eaction was carried out for 30 min in the sequential batch reactor. 3.6 Res ults and Discussion 3.6.1 Determination of M agnesium N eeded for P hosphorus R ecovery as S truvite from C entrate The c entrate was found to have the following species concentrations: NH 4 + N, 710 mg/L; PO 4 3 P, 166 mg/L with pH, 7.4. The centrate was treated for phosphorus recovery in the form struvite using magnesium chloride. Different concentrations of magnesium were adjusted using MgCl 2 .6H 2 O. As discussed in previous sections, magnesium hydroxide has been used by some researchers but magnesium chloride was found to be a better source of magnesium. Some authors have used magnesium oxide (MgO), which has the additional benefit of raising the solution pH and thus promo ting struvite precipitation (Munch and Barr 200 1; Schuiling and Andrade, 1999; elen and Trker (2001) studied NH 3 removal as struvite from molasses based industrial wastewater. They found that MgCl 2 was a better Mg 2+ source than MgO. Using swine wastewater, Burns et al. (2001) and Beal et al. (1999) had the same results. Less NaOH w as used for pH adjustment when MgO was used; however, less NH 3 was removed as struvite. In the present experiments, pH was adjusted to 8.4 using 5N NaOH and the samples were kept on a shaker for 24 hours in glass bottles with a volume of 500 mL. The pH v alue of
76 8.4 was chosen from model simulation results for the highest stuvite purity of 98 % Final residual concentrations of NH 4 + N and PO 4 3 P were analyzed. Experimental results and model predictions are shown in Table 3 2. This researcher found that wh en magnesium was not added to the centrate and the pH adjusted to 8.4, reduction in phosphorus and ammonia nitrogen occurred This indicated the presence of minerals such as magnesium and calcium in the original centrate which cause d precipitation. This researcher observed that 2.2 mM of ammonia N and 2 mM of phosphorus could be precipitated without the addition of magnesium in the centrate. The c oncentration of calcium and magnesium in the centrate was reported to be in the range of 30 to 50 ppm and 50 to 100 ppm respectively Model simulations were carried out using the following concentrations of ionic species which were input to the model: NH 4 + N, 710 mg/L; PO 4 3 P, 166 mg/L; Ca, 30 ppm; pH, 8.4 with magnesium con centrations as shown in Table 3 2 in addition to 35 ppm. Experimentally, the optimum magnesium required was found to be 225 mg/L for centrate for the concentrations of NH 4 + N and PO 4 3 P, as mentioned earlier. The residual NH 4 + N concentration was found to be very high which can be re covered by adding more phosphate to form struvite or by ammonia stripping. 3.6.2 Struvite Precipitation from Synthetic Solution in the SBR Experiments using defined solutions were carried out in the laboratory to study characteristics of the sequential ba tch reactor operation. The data obtained in the experiments were used to validate the model described in C hapter 2. A solution of 6 g/L of PO 4 3 was prepared by mixing 118.8 ml of 85% phosphoric acid in 9 L of distilled water. The total ammonia concentrati on was made to 3.8 g/L by adding 130.66 g of ammonium chloride. Model simulations were performed to get the highest struvite purity of 99%. The m odel predicted that a magnesium concentration of 4.5 g/L was suitable to get more than 99% purity. Magnesium
77 ch loride (MgCl 2 .6H 2 O) of 338.7 g was used as a magnesium source. This gave an initial NH 4 + :Mg 2+ : PO 4 3 ratio of 27.14: 18.5: 0.1935. In these experiments, a hydraulic retention time of 30 min was selected. In the sequential batch mode of the reactor operation, the settling of precipitated solids was studied against the mixing time (hydraulic retention time of the reaction). It was determined that a minimum of 30 min mixing was required for the solids to start to settle. In less than 30 min, th e precipitation started, but solids formation was not observed but the solution in the reactor became translucent in the initial 30 min of the reaction which indicated commencement of struvite formation. Precipitation was found to occur instantaneously a fter the addition of magnesium chloride solution which increased with the addition of NaOH. After reaching a pH of 8.4 the solution was fluidized for 30 min. Precipitated struvite was allowed to settle in the reactor. This researcher noted that it takes about 10 min for the solids to settle. This researcher observed that the solids recovered had 80% moisture. Solids recovered from the SBR were filtered using 0.45 m Whatman filter paper. The solids were dried and this researcher observed that filtration could recover solids which had 63% moisture. Total dry solids of 218 g were recovered in the experiment. This is in good agreement with the model results which predicted 225 g of solids. 3.6.3 Phosphorus R ecovery from Centrate Struvite formation was carr ied out in sequential batch mode in two steps. In every step, a hydraulic retention time of 30 min was used. In S tep 1 the pH of centrate was adjusted to 8.7. Results are shown in Table 3 3. After pH adjustment, the phosphorus concentration was found to d rop from 103 mg/L to 64 mg/L. In S tep 2 414 mg/L of MgCl 2 .6H 2 O was added for further phosphorus recovery. This was found to recover 99.6% phosphorus in the form of struvite at a pH of 8.7. Experimental results were validated using the comprehensive model as previously
78 explained. The m odel validation was also done for residual concentrations of nitrogen and phosphorus as well as total solids. This is in agreement with the model predictions as shown in Table 3 3. Model predictions on the caustic requireme nt do not match with the experiment because data on ionic species other than Mg 2+ PO 4 3 and NH 4 + were not available. The exion concentration was adjusted in the model to get accurate predictions on the caustic requirement. The exion concentration was fou nd to be 0.003751M which gave a close prediction to the experimental value. 3.6.4 Experiments on C omplete N itrogen and P hosphorus R ecovery from C entrate From known concentrations of ammonia and phosphate in centrate, the model was used to calculate the am ount of magnesium chloride required. The reaction of struvite formation started instantly and it was carried out for 30 min. This time was chosen for the solids to form completion of precipitation so that the solids could settle quickly. The r eaction was carried out in a sequential batch mode in steps to remove nitrogen completely. Struvite precipitate was harvested at the end of the batch after all the steps. The p recipitation reaction in every step was carried out at the pH of 8.4. Solids were harvested at the end of S tep 4. In S tep 1 the concentration of magnesium was adjusted to 1.1 g/L in the reactor. In this step, the phosphorus concentration was reduced from 88.97 mg/L to 2.8 mg/L recovering 97% phosphorus in the solid form. In this step, ammonia d ecreased from 588.4 mg/L to 440 mg/L. Ammonia was further recovered by adding phosphorus in the form of phosphoric acid in S tep 2 The m odel was run with 440 mg/L of ammonia and various concentrations of phosphorus and magnesium chloride to ensure the high est purity of struvite. As illustrated in C hapter 2, it is best to have ammonia in excess to get the higher struvite purity. As interpreted from the model simulation results, for Step 2 the concentration of phosphorus and magnesium were adjusted to 477.4 mg/L and 3.34 g/L respectively. A s imilar chemical addition was carried out in Steps 3 and 4 to reduce
79 the ammonia concentration to 45 ppm. The a mount of chemicals added is presented in Table 3 4. Solids collected were analyzed for nitrogen and phosphorus at the end of S tep 4. Solids were found to contain : N, 5.03%; P, 23.14%, Ca, 0.55%; and Mg, 11.23%. Solids were also analyzed using x ray d iffraction (XRD) to ensure formation of struvite. Struvite procured from Fisher Scientific was used as a standard f or comparison. The XRD results are shown in Figure 3 3. The intensity peaks match well with the standard struvite suggesting formation of struvite as the only solid. This was also in agreement with the chemical analysis of the solids. Results from the exp eriments on struvite formation lead to the process details as illustrated in Table 3 5. 3.6.5 Effect of A eration on C entrate The pH of wastewater increases with aeration because of CO 2 stripping ( Battistoni et al. (1997) devised a phosphate rem oval technique by exclusively using the chemical physical properties of anaerobic supernatants without any addition of chemicals This technique reached the operative pH only by aeration and obtaining a quantitative removal of phosphate through nucleation on sand quartz by CO 2 stripping with air. It was previously demonstrated that the wastewater pH increased by aeration, and the concentrations of total P and soluble PO 4 P were reduced by a struvite crystallization reaction induced under a high pH condition Saidou et al. (2009) used the dissolved CO 2 degasification technique for struvite precipitation. The y found that struvite precipitation occurred in alkaline solutions were pH > 8.1. Significant phosphorus removal efficiency through struvite precipitation was observed for experiments carried out with airflow rates between 10 and 25 L min Compared to traditional techniques of struvite precipitation such as stirring and aeration, the dissolved CO 2 degasification technique is promising since a high amount of phosphorus could be removed in a relatively short experiment time.
80 In the present studies, the centrate was aerated at a rate of 3.5 L/min using the same reactor to the test effect on the pH an d precipitation. This researcher found that aeration increased the pH of centrate over a period of 30 min and then stabilized to a pH value of 8.5. Solids precipitation was detected when the centrate was aerated for a total of 60 min. The p hosphorus concen tration was found to decrease from 106.7 mg/L to 82 mg/L. This researcher thought that the centrate received from anaerobic digester was rich in carbonate (~3500 ppm) and aeration removes carbonates in the form of carbon dioxide (Battistoni et al., 1997). The r emoval of carbonate improves struvite purity and decreases the caustic requirement for struvite formation. 3.6.6 Struvite Recovery from Dairy Manure Struvite recovery experiments from dairy manure were performed in the SBR as explained in an earli er section. The d airy manure sample volume of 7 L was treated for struvite recovery using magnesium chloride and 2 N sodium hydroxide. Struvite is recovered by gravity settling, decanting, and drying the resulting sludge like material. Results from phospho rus recovery experiments on different streams of dairy was tewater are shown in Table 3 5. The a mount of magnesium and sodium hydroxide needed for struvite recovery were calculated from the model explained in C hapter 2. Concentrations of nitrogen, phosphorus, calcium, potassium as well as pH were input to the model. The m agnesium requirement is determined from the model for the highest purity of struvite in the solids. Model simulations suggested that it was desirable to carry out the phosphate recovery in two steps. Flushwater was found to contain 50.66 ppm of phosphorus and 740 ppm of ammonia as well as 357.78 ppm of potassiu m and 140.5 ppm of calcium. After treating flushwater with 2 g/L of magnesium chloride, phosphorus concentration was found to drop to 22.35 ppm with some
81 precipitation of ammonia N, potassium and calcium. Phosphorus concentration in S tep 2 of struvite pre cipitation was found to decrease to 5.7 ppm. The e xperimental effluent phosphorus concentration on other types of dairy manure was reduced to less than 5 ppm as shown in Table 3 6. This indicated more than 90 % recovery of phosphorus. As presented in Tabl e 3 5, this researcher inferred that complete recovery of phosphorus in a single step was not possible. Additional steps are required for more than 90% recover y of phosphorus. Model simulations suggest formation of precipitates other than struvite. This re searcher also observed that a single step approach was not suitable for effective utilization of magnesium chloride. Experimental precipitation of other minerals such as calcium and potassium was minimized by selecting appropriate pH value (8.7) determin ed from model simulations. Phosphorus precipitate recovered from cattle manure can be characterized by its physical and chemical properties. Chemical analysis of total nitrogen and total phosphorus indicated the recovered materials are value d as a fertil izer. Figures 3 4( A ) and ( B ) show the results of an x ray diffraction analysis performed on struvite recovered from cattle manure. Some of the definite struvite peaks indicate a presence of the mineral NH 4 Mg(PO 3 ) 3 This shows that phosphorus precipitate other than struvite (MgNH 4 PO 4 .6H 2 O) formed that contained the same type of ionic species as in struvite. The p resence of calcium carbonate was also detec ted by x ray diffraction ( XRD ), as shown in Figure 3 4 (b). The precipitate contained N, 4.5%; P, 21. 14% ; Ca 0.97%; and Mg, 12.3%. The phosphorus precipitate produced was not pure struvite, as the mass ratio of pure struvite is 1:0.74:3.9 and the molar ratio is 1:1:1, excluding the hexahydrate. The formed precipitate was enhanced with phosphorus from the formation of other phosphate containing
82 compounds that may have been formed but not yet identified. Since the overall goal was to recover phosphoru s r ather than produce pure struvit e t his was a favorable result. 3.6.7 Settleability of Struvite in the SBR This study aimed to create a model for computing suspended struvite behavior during batch settling It also endeavored to better understand the settling process as well as obtain a more accurate and convenient determination of struvite profiles in different types of wastewater. The method is built on the graphical approach taking into account the effect of concentration of suspended solids and pH on the settling of the surface of su spension. The essential role of hydraulic settling time in the formation of struvite was studied in the sequencing batch reactor. The batch settling test was found to have four periods : free settling zone settling transition and compression. Figure 3 5 shows a settling curve for explaining these four periods. This researcher observed that struvite formation improved the settleablity of solids in the dairy manure and, at the same time recover phosphorus. The s ettling of solids was studied for different concentrations of struvite formed. The s ettling curve was generated for synthetic struvite for concentrations of 31.2 mg/L and 6.24 g/L. Struvite was formed in the centrate sample as described previously. The s truvite concentration found in this case wa s 3.54 mg/L. The s ludge formed was quantified by recovering settled solids from the bottom of the reactor. Precipitation experiments were carried out using method s discussed earlier. Solids formed were found to be struvite as shown earlier. Struvite was d ried overnight at 105C and weighed. After 30 min of a struvite precipitation reaction, the solid layer was visually monitored in the reactor against the height of the height reactor. This researcher observed that settling time for solids was 10.32 min, 11 .14 min and 13.45 min for 31.2, 6.24 and 3.54 g/L of solid concentration respectively.
83 This researcher observed that the solids in raw manure do not settle completely in 30 min. Solids in fresh manure were found to be 0.29 g/L when uniformly mixed. In 30 min the solids content in raw manure decreased to 0.17 g/L. The s truvite formation helps in solids settlin g as shown in Figures 3 8 ( A ) and ( B ) Solids measured in the top of the reactor were found to decrease considerably after struvite settling to 0.08 g/L. As shown in Figure 3 6, in 15 min of struvite settling, the solids content of 2.35 g/L was observed at a depth of 24 from top of the reactor which is a part of the conical section of the reactor. Complete settling of solids was observed in 30 m in after the struvite formation as shown in Figure 3 7. Formation of fines was not detected in the struvite formation. This researcher observed that struvite formation improved settling of suspended solids in dairy man ure as shown in Figures 3 8( A ) and ( B ). Settling of solids in struvite formation was similar to activated sludge settling. The zone settling period is approximated by a linear equation, while the compression period is approximated by the Roberts formula as discussed below. In the zone sett ling period, flocs of solid interfere with each other while affected by coagulation and destruction, and the interface height, H(t), declines at a constant velocity, A. (3.1) Integrating equation (3.2) gives: H(t) = At + B (3.2) where H(t) = ratio of th e sludge interface height versus its initial height (%) at time t A = initial settling velocity (%/min) t = settling time (rain) B = constant (%)
84 After the completion of zone settling, the sludge accumulated at the bottom of the settling cylinder began to dehydrate by the weight of the solids. Then the settling rate decrease d slowly and the compression period was attained. Empirically, this compression period is expressed by the Roberts formula: (3.3) where H = ratio of the final sludge interface height (after about 30 min) versus its initial height (%) k R = Roberts constant (min 1 ) for expressing compression velocity t 0 = compression starting time (min) H c = ratio of the sludge interface height versus its initial height (%) at the compression point 3.7 Outcomes The pH level within the recovery processes varies from 8.2 to 9 depending on the strategy of the recovery process. Within the current work a pH of 8.4 to 8.7 was selected to allow for rapid struvite nucleation, being aware that this can cause the precipitation of fines Aeration of the centrate was found to increase t he pH to 8.4. This is useful for minimizing the caustic requirement for struvite formation and also to reduce car bonate concentration. This improves the quality of struvite by reducing solids with carbonate The NaOH required to adjust the pH of centrate t o 8.7 for complete phosphorus recovery was found to be 432 mg/L without aeration Complete nitrogen and phosphorus removal is possible from centrate in a sequential batch reactor. Settling of precipitated struvite is rapid and it was found to settle in 10 min. Struvite which was separated from bottom of the reactor was found to have 80% moisture. Filtered struvite was found to have 65% moisture The recovery of phosphorus from animal waste as a struvite containing precipitate has been successfully demon strated using the SBR mode of operation. The next step in the growth of this technology is the development of a field scale recovery unit at a commercial animal production unit. The operation of a field scale recovery unit would supply the necessary data t o complete a cost/benefit analysis to investigate the economics of the technology. A cost effective magnesium source of magnesium chloride and a fast, low cost method of pH adjustment using sodium hydroxide can be utilized to successfully implement thi s t echnology at the farm scale
85 Table 3 1. Optimum pH values used in s truvite formation Optimum pH Wastewater Source Reference 9.0 Anaerobic digester effluent used in a bench scale reactor Jaffer et al., (2002) 8.5 Swine wastewater used in a bench scale reactor Burns et al., (2002) 8.5 9.5 Wastewater and sludge Schultze Rettmer (1991) 9.0 Anaerobic digester effluent used in a pilot scale reactor Munch and Barr (2001) 8.5 9.0 Anaerobic digester effluent used in a bench scale reactor elen and Trker (2001) 9.0 Livestock waste Buchanan et al., (1994) 9.0 Digester supernatant used in a lab and pilot scale reactor Siegrist (1996) Table 3 2. Effect of magnesium concentration on nitrogen and phosphorus removal from centrate Experimental Model Predictions pH Mg 2+ added, Final NH 4 + N Final PO 4 3 P NH 4 + N PO 4 3 P mg/L mg/L mg/L mg/L mg/L 0 679.45 100.88 710 166 8.4 100 639.17 16.82 654.25 40.16 150 619 10.17 637.50 2.24 200 578 5.07 637.00 2.23 225 550 1.05 637.00 2.24 250 535 1.13 637.00 2.23
86 Table 3 3 Phosphorus r ecovery from the c entrate Table 3 4 Recovery of nitrogen and phosphorus from the centrate Parameters Untreated Centrate Step 1 After Step 1 Step 2 After Step 2 Step 3 After Step 3 Step 4 After Step 4 NH 4 + N, mg/L 588.44 439.9 1 160 131.4 45 Phosphorus, mg/L 88.97 2.8 477 .4 46 100 14.69 84 4.5 MgCl 2. 6H 2 O added, g/L 0 1.1 3.3 4 1.05 1.5 pH 7.7 8.4 6.7 8.4 8.4 8.4
87 T able 3 5 Sequential b atch o peration: P rocess d etails Parameter Description Raw Material Centrate in a Secondary Wastewater T reatment P lant Struvite Production Potential from Centrate Depends on N and/or P R emoval Mode of Operation Sequential Batch Harvesting Every Batch Reactor Type Conical Bottom Batch Reactor Total Residence Time 60 min Reaction Time 30 min Settling Time 30 min Phosphorus Recovery 99% Ammonia Recovery 95% Ta ble 3 6 Struvite r ecovery from d airy m anure Sample pH NH 4 + ppm PO 4 P ppm Ca ppm K ppm MgCl 2 addition Flushwater 7.21 740.88 50.66 140.5 357.78 Step 1 8.7 513.25 22.35 113.3 322.1 2 g/L Step 2 8.7 420.2 5.7 110.5 306.51 1.1 g/L Sample pH NH 4 + ppm PO 4 P ppm Ca ppm K ppm MgCl 2 addition Anaerobically Treated 6.8 1585.5 40.93 103.62 289.5 Step 1 8.7 896 15.7 102.7 257.6 1.5 g/L Step 2 8.7 782.6 4.5 94.56 242.57 1 g/L Sample pH NH 4 + ppm PO 4 P ppm Ca ppm K ppm MgCl 2 addition Aerobically Treated 7.55 497.7 30.51 106.22 303.89 Step 1 8.7 385.6 9.74 90.47 257.6 1.3 g/L Step 2 8.7 358.78 2.4 85.47 239.13 0.8 g/L
88 Figure 3 1. Sequencing b atch r eactor (SBR) p rocess Figure 3 2 Sequential b atch r eactor for s truvite p recipitation
89 Figure 3 3 X ray dif fraction patterns for struvite produced from centrate
90 Figure 3 4. A). X ray diffraction patterns for struvite from dairy manure Figure 3 4. B) X ray diffraction patterns for struvite from dairy manure showing presence of CaCO 3 Figure 3 4 X ray diffraction patterns for struvite A) From dairy manure. B) Struvite from d airy manure showing presence of CaCO 3 (A) (B)
91 Figure 3 5 Settleability of s olids in s truvite f ormation Figure 3 6 Solids s ettling in an SBR with d airy m anure: s ettling time = 15 m in 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Settliing Velocity (inch/sec) Distance from top of the reactor (inch) 6.24 gm/L 31.2 gm/L 3.54 gm/L 1 2 3 4 0 0.5 1 1.5 2 2.5 3 3.5 0 10 20 30 Settled Solids (g/L) Depth of Reactor (inch) Solids in Fresh Manure Solids in Manure with Struvite
92 Figure 3 7 Solids s ettling in an SBR with d airy m anure: s ettling time = 30 min Fig ure 3 8 Struvite settling. A ) Raw d airy m anure B ) S ettled s olids after s truvite p recipitation 0 1 2 3 4 5 6 0 10 20 30 Settled Solids (g/L) Depth of Reactor (inch) Solids in Fresh Manure Solids in Manure with Struvite (A) (B)
93 CHAPTER 4 D EVELOPMENT OF P RODUCT F ORMULATION T ECHNIQUES USING STRUVITE SOLIDS 4.1 Struvite as a Fertilizer The f ertilizer industry has made a significant contribution to the agricultural sector by raising the fruitfulness of agricultural products by 50% and the efficiency on vegetal production. Research shows that the expenses for fertilizer result in 10.5 times mor e profit at harvest. The main reason for fertilizer application is to supply elementary needs of agricultural plants during their growth, for instance nitrogen, phosphorus, potassium, calcium, magnesium and sulphur. Therefore the selection of an appropr iate fertilizer should be based on the soil type and composition. The potential hazards of fertilizers to the environment have results in stringent limitation to their use. About half of the applied fertilizers, depending on the method of application and s oil condition, are lost to the environment, which results in the contamination of water (Salman, 1989; Salman et al., 1989). Use of conventional fertilizers may lead to concentration levels that are too high for effective action. A high concentration may p roduce undesirable side effects either in the target ed area, which could lead to crop damage, or in the surrounding environment (El Refaie and Sakran, 1996). Slow release fertilizers demonstrate many advantages over the conventional type s: 1) decreased rat e of removal of the fertilizers from the soil by rain or irrigation water ; 2) sustained supply or minerals for a prolonged time ; 3) increased efficiency of the fertilizer ; 4) lower frequency of application in accordance with normal crop requirement ; 5) min imized potential negative effects associated with over dosage ; and 6) reduced toxicity (Byung Su et al., (1996). Despite the advantages of slow and controlled release fertilizers, only about 0.15% of total fertilizer consumption is such products. This is mainly due to the high costs and lack of proper legislation in most parts of the world to restrict the use of soluble fertilizers.
94 Slow release fertilizers may be more expensive than soluble types, but their benefits outweigh their disadvantages. Following advantages are identified with slow release fertilizers: 1. Fertilizer burn is not a problem with slow release fertilizers even at high rates of application 2. Fertilizers are released at a slower rate throughout the season; plants are able to take up most of the fertilizers without waste by leaching 3. Less frequent application is required 4. Uniform particle size allows easier and precise mechanical distribution 5. Flexibility of release periods last from 40 to 360 days at 25 C 6. Reduced capital and labor outla y in horticultural crop production 7. Reduced nutrient loss via leaching and run off 8. Reduced seed or seedling damage from high local concentrations of salts 9. Reduced leaf burn from heavy rates of surface applied fertilizers 10. Improved storage and handling properties of fertilizer materials 11. Product differentiation resulting in improved market potential An augmentation has occurred worldwide on production of fertilizers containing phosphate with an increasing demand for phosphate rock by 1.5% each year ( Ste n, 1998). About 85% of the phosphate is used for the production of fertilizer with 7 billion tons of phosphate rocks as P 2 O 5 remaining in reserves. Considering the 40 million tons of phosphorus as P 2 O 5 consumption each year, the available resources of phosphorus are expected to be exhausted in 100 years. Therefore, finding a cheap method to obtain phosphate is critical. Considering the importance of fertilizer usage and existence of limited resources for its production, especially phosphorus with no su bstitute in nature, a technology which has been developed for the recovery of the materials from waste can provide a break through solution. One of the recent innovations is production and recovery of struvite (MgNH 4 PO 4 6H 2 O) from waste. If the
95 formation of struvite is controlled, then it can be beneficial since its precipitation removes NH 4 + N and PO 4 3 P from the water and the precipitate has the potential use as a fertilizer. To date, struvite has been commercialized only in Japan as a fertilizer for growing rice and vegetables (Ueno and Fujii, 2001). Shu et al. (2006) presented the reasons why struvite is not widely applied as a fertilizer why it has limited availability to farmers, and why its applicability and benefits are not known At one point i n time, W R Grace & Co. a fertilizer company, produced struvite from its individual components. However, this form of production was found to be too cost ly for the value of the product Beal et al. (1999) estimated a value of $206 per dry ton of struvite based on the nutrient value. Typically, the following characteristics are desired for struvite or a struvite blended product for it to sell to a wider range of fertilizer customers: High purity Multiple nutrient in a single pellet Lower weight and transportation cost Long shelf life Concentrated, granular, non sludgy and non odorous pellets Free of pathogens from wastewater Obtained from an organic sourc e n ot dependent on natural gas or open pit minin g and the raw material is free or government subs idized 4.1.1 Recovery of Ammonia from Wastewater Ammonium removal from wastewater has become increasingly important for the worldwide emphasis on the eutrophication problem. Although many methods, such as biological nitrification/denitrification process, a ir stripping, and breakpoint reaction, have been successfully applied for removing ammonium from wastewater, treatment of wastewater containing medium
96 concentrations of ammonium (hundreds mg/l to thousands mg/1) has always been a big challenge. A mmonia in industrial wastewater systems enters the plant. Additional ammonia is released in the aerobic or anaerobic digester as it reduces sludge volumes. This ammonia is concentrated in centrate or filtrate streams as the sludge is separated and consolidated in th e filter press or centrifuge. Many industrial applications use ammonia as a part of the process chemistry thus ammonia finds its way into the industrial wastewater as a contaminant. Typically, 4 0 % to 50% of the total nitrogen in a municipal wastewater treatment plant is found as ammonia in centrate or filtrate streams. The a mmonia removal and recovery process usually consists of a system that combines flash vacuum distillation with ion exchange to remo ve 90% of the ammonia in these streams at a much lower cost than traditional biological removal systems. Initially, the ammonia in industrial wastewater is conditioned so that neither suspended solids nor precipitates can reach the ammonia removal systems. Influent (with 300 ppm ammonia nitrogen or less) is then input to an industrial grade ion exchange resin which selectively adsorbs the ammonia. The adsorption columns are regenerated using either a brine or sulfuric acid. The regeneration solution is use d repeatedly in which the ammonia concentration builds up to several thousand ppm. The spent ammonia laden regeneration solution is stripped of ammonia to produce a commercial grade (about 40%) solution of ammonium sulfate. Ammonia stripping is possible b oth for full scale wastewater treatment and nitrogen rich liquids. largest wastewater treatment plant VEAS, located in Oslo, uses a closed loop ammonia stripping process on the filtrate from presses for sludge with great success It then sells the ammonium nitrate to a fertilizer manufacturer. Struvite can be re utilized to remove ammonium from wastewater after ammonia gas is released from struvite in a heated alkali
97 solution. Struvite is transformed to MgHPO 4 by releasing ammonia under an acidic c ondition. The reaction can be written as follows: MgNH 4 PO 4 2 O(s) + H + 4 2 O(s) + NH 4 + + 5H 2 O At a higher pH, however, MgHPO 4 reacts with ammonium ions to give struvite again. Zhang et al. (2004) developed this process to solve the nitrogen remov al problems by recovering ammonium using acid and repeatedly using struvite residues. Ammonium was efficiently released from struvite with a low N/P ratio at a pH < 5.0 and temperature > 40C. S truvite was mainly transformed to MgHPO 4 .3H 2 O in acid solutio ns, which could be used for ammonium removal again. A novel approach to recover ammonia with potassium magnesium sulfate or langbeinite (commonly known as KMAG) was utilized in the present study. Experiments on additional ammonia recovery after struvite pr ecipitation from synthetic solution are presented here in this section. 4.1.2 Uses of Struvite The agronomic properties of struvite as a fertilizer have been widely discussed. It represents a highly effective source of nutrients ( phosphorus nitrogen, and magnesium ) for plants (Li and Zhao, 2003) Struvite was also found to be as efficient as mono calcium phosphates (MCP) (Johnston and Richards, 2003). low solubility in water which is 0.018g/100ml at 25C (Bridger et al., 1961), also presents the advantage to prolong the release of nutrients during the growing season (Gatterell et al. 2000) without danger of burning the roots of treated crops. In pure form, struvite crystals contain 5.7% elemental N, 12.6% elemental P and 0% elemental K. Although the value of a natural 6 29 0 product is expected to be high, struvite is not currently a common commercial product that has a well defined market value. Struvite is a nutritional soil supplement that prolongs the effectiveness of commercial fertilizers applied in
98 turf, horticultural or crop application. This quality saves time and labor costs in field application of nutrients. One application can last up to eight months. The most obvious use of struvite is as a raw material for t he fertilizer industry (Gaterell et al., 2000). Struvite can be used as a material in fire resistant panels and in cement (Sarkar, 1990; Schuiling and Andrade, 1999). If cheap production methods are developed, it could be used in detergents, cosmetics, and animal feed, all of which use phosphates (Gaterell et al., 2000). The most promising application is as a slow release fertilizer that can be applied in a single high dose without damage to growing plants. The plants suggested are ornamentals, vegetables, forest out plantings, turf, orchard trees, and potted plants (Munch and Barr, 2001; Li and Zhao, 2003; Johnston and Richards, 2003). Highly soluble orthophosphate, serving as the initial phosphorus supply for establishing container plants, could be used to gether with struvite in a mixed fertilizer product. Struvite could also replace the major fertilizer diammonium phosphate, which is produced by neutralizing phosphoric acid with ammonia. Mixing struvite with phosphoric acid might even yield a superior fert ilizer : part slow release MgHPO 4 and part fast release, highly soluble ammonium phosphate (NH 4 ) 2 HPO 4 This might be considerably more cost effective than commonly used diammonium phosphate fertilizers. Another possible product is untreated granular struvit e that can be mixed with peat to serve as a lightweight potting mix (Gaterell et al., 2000). High solubility is not an asset in many fertilizer applications, as in grasslands and forests, where fertilizer is applied once in several years. The presence of m agnesium in struvite makes it attractive as an alternative to contemporary fertilizers for crops such as sugar beets that require magnesium (Gaterell et al., 2000). The phosphate industry uses rock phosphate for detergents, food, and cosmetics by applying a high energy, high temperature industrial process to purify the phosphate. S truvite will
99 not likely be used in the short term because struvite purification technology is unknown S mall amounts of phosphate is needed by these industries, as well as well k nown rock phosphate purification technology. However, as high quality rock phosphate increases in price, and the industry is forced to use low quality rock phosphate, pressure will increase to develop purification methods for struvite. Markets for struvite also include the natural foods and organic industry as well as backyard gardeners interested in environmentally friendly products. The largest potential bulk market for struvite is the turf industry. Turf requires a tremendous amount of magnesium and sup plemental zinc and copper in addition to nitrogen, phosphorus and potassium. The turf grass industry is concentrated primarily in the metropolitan areas. F ertilizer companies use struvite as an additive and it is mixed with other inorganic and organic ma terials to adjust the amount of nitrogen, phosphoric acid and potassium. Two approaches generate revenue from the recovery of struvite. The first approach is to produce a relatively pure crystal and sell it to a fertilizer company or as a stand alone prod uct. The second approach is to blend struvite with harvested solid material and reap the benefit in terms of a higher price for the blended product. 4.2 Objectives 4.2.1 Co crystallization of Struvite with KMAG Potassium magnesium sulphate ( KMAG ) is an excellent source of potassium, magnesium and sulfur for bulk blended fertilizers. It is a naturally occurring mineral known as langbeinite which has 22% K 2 O, 11% Mg (18% MgO) and 22% S. Potassium magnesium sulphate is totally water soluble providing po tassium, magnesium and sulfur in a form readily absorbed by plants. Because of its crystalline nature, KMAG dissolves gradually throughout the growing season, providing a continuous source of nutrients. Potassium, magnesium and sulfur are basic nutrients essential for plant growth and vital to profitable crop production with nitrogen
100 and phosphorus. The se nutrients are continually removed from the soil by high yield crop production and, in the case of sulfur and magnesium, are not replaced by many of toda y s high analysis fertilizers. Potassium is required for the uptake of nitrogen and synthesis of protein and starch. They also help activate more than 60 enzymes, making them available to stimulate other chemical processes within the plant. Adequate potass ium is essential for fruit formation, optimum yields and high quality. Sulfur activates a number of enzymes, is vital to the formation of amino acids, is crucial in the production of protein and is especially important to plants with high oil content. Struvite was utilized in this study to produce a fertilizer product having multiple nutrients with a slow release property using KMAG. 4.2.2 S tudy A bsorption of A mmonia in A cidic KMAG S olution This researcher explored the p ossibility of ammonia stripping a nd consequent absorption in acidic KMAG solution. This would give a product rich in nitrogen with other nutrients such as potassium, magnesium, and sulfur which can be further pelletized with struvite. This would help meet the nitrogen limits along with phosphorus in an integrated struvite process. With ammonia stripping, the wastewater treatment plant would not rely on an external source of phosphorus for additional nitrogen recovery as struvite. 4.2.3 I nvestigate U se of A gglomerates to M ake S truvite P e llets A commonly used cationic high molecular weight polymer used in wastewater treatment plant would help in struvite agglomeration to reduce moisture content in recovered solids hence reducing the energy requirement for drying the product. The polymers are used to flocculate suspended solids particularly for non potable raw water clarification, primary and secondary effluent clarification, oily wastewater clarification and so forth S odium alginate which is the sodium salt of alginic acid was also used for agglomerati on of struvite solids. T he effect of alginate was assessed on struvite agglomerative properties to p roduce strong pellets which
101 withstand the rigors of storage, handling, packaging and shipping. Sodium alginate was chosen for struvite agglomeration becau se of its water solubility. The calcium and magnesium salts of alginic acid do not dissolve in water. When dissolved in water, alginate has an ability to thicken the resulting solution by increasing the viscosity of aqueous solutions. They form gels as wel l as films of sodium or calcium alginate and fibers of calcium alginates. 4.3 Materials and Methods 4.3.1 Struvite Precipitation An SBR was used for precipitation as explained in C hapter 3 Struvite was allowed to settle for 30 min and then be recovered f rom the bottom of the reactor. The c entrate or synthetic solution containing nitrogen, potassium magnesium chloride and sodium hydroxide were added from the top of the reactor. Mixing was provided by air using a sparger from the bottom of the SBR. The pH for struvite precipitation was adjusted using 10 N sodium hydroxide. Liquid samples were analyzed using calcium and magnesium probes Struvite solids were filtered using Ashless 41 micro meter Whatman filter paper under vacuum. Struvite recovered from t he pilot scale run as shown in Figure 4 1( A ) was also used in the crystallization experiments. 4.3.2 Preparation of KMAG S olution Raw KMAG was obtained from The Mosaic Company which suggested the chemical analysis of KMAG as listed in Table 4 1. The fertilizer specification of KMAG was 0 0 22. The reddish brown granules of KMAG used in the p resent study are shown Figure 4 1 (b). For all experiments performed with KMAG, 1kg of KMAG was mixed in 1 L of deionized water. The mixing was carried out in a 2 L Erlenmeyer flask using an electric stirrer from the Arrow Engineering Co. The solution was stirred for 12 hours at room temperature. The solution was allowed to settle for 30 min and decanted. The decanted solution was filtered using Ashless 40 Whatman filter paper. The settled solids were filtered and dried at 105C in an oven. This
102 researcher found that 40% of the weight of the raw KMAG was dissolved in water. Depending on the experimental need, KMAG was mixed multiple times in the same proportions as mentioned earlier. The solution of KMAG was analyzed for potassium and calcium content. The concentration of potassium in the solution was found to be 64 g/L with a standard deviation of 5.3 g/L. The c alcium concentration was found to be 0.41 g/L with a standard deviation of 0.13 g/L. Solids were analyzed by x ray diffraction (XRD) as explained in C hapter 3. Solids under consideration as reference patterns for XRD analysi s are listed in Table 4 2. These solids were chosen from the database for inorganic solids which have a possibility of formation in the system of ionic species which includes minerals such as magnesium, potassium, phosphorus, nitrogen, sulphur, and calcium 4.3.3 Agglomeration of Struvite The cationic polymer was mixed with the soluti on which was subjected to struvite recovery in the SBR to study enhancement of struvite settling. Sodium alginate was used as an agglomerate for settled struvite. A dry weight of 1% of the sodium alginate was mixed with struvite in a 500 ml beaker which formed a paste of struvite solids. The m ixture was extruded in the form of strings using a c lay pottery ceramics extruder Extruded solids were dried at 105C in an oven for 8 hours. Solids were crushed with size distribution carried out using an Octagon2000 shaker with USA Standard Sieves. 4.3.4 Crystallization of S truvite with KMAG A conical bottom separator was used as a cr ystallizer as shown in Figure 4 2. A s chematic of cryst allization is shown in Figure 4 3. A thermometer was inserted from t he top of the crystallizer. Struvite solids tend to settle at the conical bottom S olids were fluidized by feeding recirculated liquid to the bottom of the crystallizer. The f low rate was maintained to make sure that solids are not carried in the feed to t he recirculation line to the oil bath. T he
103 t emperature was maintained at 85C unless otherwise mentioned using an oil bath. The s upernatant solution was recirculated using a peristaltic pump through the oil bath. Air was allowed to flow on top of the cryst allizer to enhance evaporation. Crystallization was carried out using evaporation which concentrated the solution. Crystallization studies were performed with the following combinations of struvite and KMAG: 1. Struvite from the pilot plant and concentrated solution of KMAG 2. Struvite from the pilot plant and raw KMAG granules 3. Struvite produced from synthetic solutions and concentrated solution of KMAG 4. Synthetic struvite and concentrated solution of KMAG in acidic conditions 4.3.5 Ammonia Removal from Centrate and Synthetic Solution Ammonia removal from solutions was carried out using a continuously stirred tank vessel. The ammonia stripped out was captured in an array of absorbers as shown Figure 4 4. A pH of the solution was adjusted to 10.0 using 5 N NaOH. Air was allowed to flow at the bottom of stirred tank through a sparger. In the first absorber the acidic KMAG solution was used followed by another absorber with dilute d sulphuric acid. 4.4 Results and Discussion 4.4.1 Effect of the Polymer A ddi tion on Struvite Filtration Polymers are widely used to remove fines in wastewater treatment operations. The p olymer named polydiallyldimethylammoniumchloride (polyDADMAC) was used by Le Corre et al. (2007) specifically to remove struvite fines. This resea rcher thought that the polymer would agglomerate struvite and improve settling properties as well as improve filtration. Experiments were carried in the laboratory to study the effect of the polymer. Struvite was produced by mix ing chemicals listed in Table 4 3. The proportions of these chemicals were chosen from the model predictions to get about 220 g of struvite. The s truvite produced was found to settle in 5 min after which it was separated and filtered. T he t ime of filtration was kept for 10 min. A s
104 shown in Table 4 4, this researcher found that the polymer improves settling. The p olymer needs to be added after 30 min of struvite reaction to see the effect of the polymer. The p olymer when added after 5 and 10 min from commencement of the reaction was not found to give an agglomeration effect. The s olid content after vacuum filtration was found to be 35% consistently. The s ettling of struvite solids was improved by the polymer. After 30 min of settling, the solids content increased with the polymer addition to 16% solids as compared to solids content without polymer addition which was 12 % solids after settling. 4.4.2 Crystallization of S truvite with KMAG 220.127.116.11 Batch Crystallization Crystallization under acidic conditions Struvite produced from sy nthetic solutions was used in this experiment. In a 250 ml beaker, 6 g of wet struvite and 100 ml of KMAG solution were mixed under acidic conditions at a pH of 2.2 which completely dissolved struvite. Wet struvite solids were found to have 67 % moisture b efore mixing with the KMAG solution. Crystallization was carried out at 55C in the oven. Crystal formation was observed after 28 hr. The s upernatant was collected by decantation and analyzed for NH 4 + N, phosphorus, potassium and calcium concentrations. Crystals were separated from the solution by decantation A total of 41.13 g of c rystals were recovered after 35 hours. A c hemical analysis of the crystals was done by dissolving 5 g of air dried crystals in 200 ml deionized water. This researcher observed that the crystals only partially dissolved. Therefore, 1 ml of 5N H 2 SO 4 was added for dissolution of the remaining crystals for analysis. This solution was analyzed to estimate nitrogen and phosphorus content. Resu lts are shown in the T able 4 5. This rese archer found that crystals collected from mixing of struvite and KMAG were found to retain nitrogen and phosphorus.
105 Crystals were analyzed by x ray diffraction and the results are shown in Figure 4 4. This researcher observed that formation of hannayite w ith Schoenite (MgK 2 (SO 4 ) 2 .(H 2 O) 6 occurred. Hannayite is (NH 4 ) 2 Mg 3 H 4 (PO 4 ) 4 2 O) which contains the same minerals as in struvite. The rearrangement of minerals in struvite to hannayite could be due to the presence of higher concentrations of magnesium and hydrogen ions. Langbeinite is crystallized as schoenite in the presence of water as shown in following equation: K 2 SO 4 .2MgSO 4 + 13H 2 2 SO 4 .MgSO 4 .6H 2 O + MgSO 4 .7H 2 O Crystallization under neutral condi tions The s ame proportions of struvite and KMAG were used without any pH adjustment as explained in the previous section. Mixing struvite and the KMAG solution gave a solution with a pH of 7.25. The x ray diffr action results showed ( Figure 4 6) that KMAG crystallized to schoenite (MgK 2 (SO 4 ) 2 .6H 2 O, picromerite, K 2 Mg(SO 4 ) 2 6H 2 O. This researcher found that crystallinity of KMAG was improved after crystallization. Transparent crystals of schoenite and picromerite produced from raw KMAG are shown in Figure 4 1(c). In all cases in which the KMAG solution and raw KMAG were mixed with struvite made synthetically from pilot scale run the x ray diffraction detected a presence of struvite and schoenite in the solids which are listed in Table 4 6 and are shown in Figures 4 7, 4 8 and 4 9. Struvite was bound to schoenite after crystallization d ue to the presence of the magnesium species in the solution which acted as a binder S truvite from the pilot plant before the addition to KMAG solution was in the powder f orm as shown in Figure 4 1( A ). A picture of a combined schoenite and struvite is shown Figure 4 1 ( D ). Raw KMAG and struvite were also found to bin d together as shown in Figure 4 1 ( E ). When dried solids of raw KMAG and struvite were crushed, a powder form ation was observed which can be recycled for re crystallization.
106 18.104.22.168 Semi continuous Crystallization of KMAG with Stuvite The crystallization was induced by evaporation of water usi ng the system shown in Figure 4 2. This type of crystallization is usually referred to concentration crystallization where a concentration of ionic species increases continuously to supersaturation and crystallization occurs. The r ate of evaporation was found to be 15 ml/hr. After 2 1 hours, another batch of struvite slurry and KMAG solution was added to the crystallizer. The c oncentration of potassium in the solution decreased with time as shown in Figure 4 10. This can be directly related to the rate of crystallization or potassium fixation in crystals and it was found to be 4.64 g/L hr. The p hosphorus concentration in the solution increased initially and then decreased as shown in Figure 4 11 due to a partial struvite dissolution and re cryst allization after 6 hours. The c alcium concentration did not change substantially; it was found to decrease from 270.4 mg/L to 239 mg/L continuously over 33 hours. The c rystal formation was not visually detected. Samples were taken at diffe rent times as sh own in Figures 4 10 and 4 11. To verify the crystal formation, solids were collected after a decrease in potassium concentration from 46 g/L to 10.45 g/L. This researcher found that nitrogen and phosphorus were retained in solids along with potassium. Thes e solids can be used as 4.5 9 12 fertilizer. The x ray diffraction confirmed the presence of potassium ammonium hydrogen phosphate, schoenite, picromerite and ammonium sulphate as shown in Figures 4 12( A ), ( B ) and ( C ). 4.4.3 Absorption of A mmonia in KMAG In the present stud y ammonia was stripped out from the synthetic solution with an initial concentration of 650 mg/L which resembled the concentration in the centrate. It was observed that under acidic conditions stripped ammonia was completely a bsorbed in the first receiver. In the second receiver the concentration of ammonia was found to be negligible. The r ate of ammonia stripping was estimated to be 12.6 mg/L hr. The r esults of this experiment are
107 pres ented in Table 4 7. This researcher obser ved that a total ammonia N amount of 1296.72 mg was removed from the scrubber. The pH of the solution decreased with the ammonia removal and the pH was adjusted to a value of 10 whenever necessary. The KMAG was found to crystallize in the first receiver over a period of 6 hours. Crystals were collected and analyzed for chemical content and they were found to contain 4.5 % nitrogen which indicated ammonia immobilization with KM AG in the crystalline form. A process flow was suggested from this st udy for absorption of ammonia as shown in Figure 4 13. Experiments on a mmonia stripping with high initial ammonia concentration were performed The a mmonia removal rate in the previous run was found to be slow. Ammonia concentration in the reactor was inc reased to observe the rate of ammonia removal. Experiments were performed with the following conditions: Starting NH 4 + N in stripper = 15 g NH 4 + N in stripper after 2 days = 6.21 g Total NH 4 N lost from stripper = 8.79 g After about 2 hours, in the ac idic KMAG, the concentration of ammonia N was found to be 1.444 g/L. A total 3.61 gm of ammonia N was recovered in the KM AG solution. Some ammonia was also carried to the second receiver with 150 ml of 0.1 M sulfuric acid. In the H 2 SO 4 trap the following concentrations were observed: 3.496 mg/L of NH 4 + N, pH of 2.2. The s olids/crystals were collected from the KMAG column to the investigate nitrogen content. After dissolution the crystals gave a concentration of NH 4 + N as 161.5 mg/L. 4.5 Outcomes A cationic polymer can be used to remove fines ( struvite crystals ) from the reactor. The p olymer should be add ed after 30 min of struvite
108 The s ettling of struvite solids was improved by a cationic polymer commonly used in wastewater treatment plant s With polymer usage, the solids cont ent increased from 12 % to 16% The KMAG was crystallized to schoenite and pi cromerite in the presence of water Struvite was found to recrystallize with saturated KMAG to give a product rich in nitrogen, phosphorus, potassium and magnesium. These crystals were found to be sparingly soluble in water at a neutral pH The s olution of KMAG and struvite under acidic condition was crystallized to hannayite and schoenite The x ray diffraction analysis confirmed that struvite was a gglomerated to KMAG Calcium precipitation was not detected in the crystallization of KMAG and struvite Semi continuous crystallization of struvite with KMAG was found to produce ammonium sulphate, potassium ammonium phos phate, schoenite and picromerite Saturated KMAG was found to absorb ammonia under acidic conditions to give a solid product rich in nitrogen. Crystals formed in the saturated KMAG solution after ammonia absorption were found to retain ammonia in solids.
109 Table 4 1 KMAG c hemical a nalysis Component Symbol Typical (%) Guarantee (%) Potassium Oxide K 2 O 22.1 22.0 min Potassium K 18.3 Calcium Ca 0.2 Sodium Na 0.3 Magnesium Mg 11.3 10.8 min Sulfur S 22.4 22.0 min Chloride Cl 1.5 2.5 max Water Insoluble 1 Moisture (105C) H 2 O 0.05
110 Table 4 2 Solids considered in x ray diffraction analysis as standards Solid Name Formula Database reference Number Ammonium sulphate (NH 4 ) 2 SO 4 15 241 Ammonium magnesium phosphate NH 4 Mg(PO 3 ) 3 21 704 Ammonium magnesium phosphate hydrate NH 4 Mg 2 P 3 O 10 .6H 2 O 38 203 Struvite,syn NH 4 MgPO 4 .6H 2 O 15 0762 Schoenite MgK 2 (SO 4 ) 2 .6H 2 O 74 1064 Picromerite K 2 Mg(SO 4 ) 2 .6H 2 O 21 1400 Periclase MgO 71 1176 Ammonium magnesium sulphate (NH 4 ) 2 Mg 2 (SO 4 ) 3 18 110 Ammonium potassium phosphate NH 4 KPO 4 1 925 Potassium hydrogen sulphate K 3 H 3 (SO 4 ) 4 17 0597 Potassium magnesium sulphate K 2 Mg(SO 4 ) 2 36 1499 Potassium phosphate K 3 PO 4 27 435 Potassium phosphate sulfate K 2 P 2 S 2 O 13 22 846 Potassium ammonium hydrogen phosphate (NH 4 K) 3 HP 2 O 7 .H 2 O 28 82 Potassium ammonium hydrogen sulfate KNH 4 H 2 (SO 4 ) 2 20 850 Hannayite Mg 3 (NH 4 ) 2 H 4 (PO 4 ) 4 .8H 2 O 16 0361 Struvite MgNH 4 PO 4 .6H 2 O 03 0240 Ammonium calcium phosphate (NH 4 ) 2 Ca 3 (P 2 O 7 )2.6H 2 O 44 0758 Magnesium phosphate oxide Mg 2 P 2 O 7 01 0866 Rustamite Ca 10 (Si 2 O 7 ) 2 (SiO 4 )Cl 2 (OH) 2 84 0292 Calcium carbonate CaCO 3 70 0095 Uranyl urea aqua iodide nitrate C 4 H 24 IO.36N 9 .64O 11 .92U 28 1426 2,5 dichlorotropone C 7 H 4 Cl 2 O 20 1622 N Hydroxysuccinimide C 4 H 5 NO 3 36 1688 Potassium magnesium phosphate KMgPO 4 20 0685 Table 4 3 Chemical added for s truvite f ormation for p olymer e ffect e xperiments Chemicals Concentrations(M) Total Mg added using MgCl 2 .6H 2 O 0.185 Total N as NH 4 Cl 0.2714 Total P as 85% H 3 PO 4 0.1935
111 Table 4 4 Effect of the p olymer a ddition on f iltration Weight of P olymer added (g) Wet Weight of S truvite (g) Dry Weight of Solids (g) % solids 0.02 630.9 221 35.03 0.12 620.4 218 35.14 0.22 714 216 30.25 0.44 540 195 36.11 Table 4 5 Crystallization of synthetic struvite with the acidic solution of KMAG Parameter Supernatant after crystallization Solution of 5g of crystals dissolved in 200 ml DI water Crystals collected after drying Nutrients in crystals and supernatant Amount recovered 19 ml -41.13 g NH 4 + N, mg/L 242.29 80.79 P, mg/L 471.52 287.92 K, g/L 89.2 2.9 Ca, mg/L 763.5 35 pH 6.4 4.14 Total NH 4 + N, mg 4.6 129.71 134.31 Total P mg 8.95 485.52 494.47 Total K g 1.69 4.77 6.46 Total Ca mg 14.5 57 71.5 Table 4 6 X ray diffraction results Sample Solids detected in XRD KMAG with s ynthetic s truvite Schoenite [MgK 2 (SO 4 ) 2 .6H 2 O], Picromerite, K 2 Mg(SO 4 ) 2 .6H 2 O, Periclase (MgO) Raw KMAG with struvite from pilot scale run Ammonium magnesium phosphate, schonite and picromeite KMAG with struvite: pH adjusted to 2.2 at commencement of crystallization Hannayite [(NH 4 ) 2 Mg 3 H 4 (PO 4 ) 4 2 O)] Schoenite KMAG from saturated solution Schoenite, Picromerite KMAG with struvite from pilot scale run Struvite, Schoenite, Ammonium Calcium Phosphate Hydrate [(NH 4 ) 2 Ca 3 (P 2 O 7 ) 2 .6H 2 O], Magnesium Phosphate Oxide [Mg 2 P 2 O 7 ]
112 Table 4 7 Absorption of a mmonia in KMAG Ammonia Scrubber, Volume 2L First Receiver Volume 2.7L Time (hrs) NH 4 + N (mg/L) pH pH adjusted to 10 Total NH 4 + N in the reactor(mg) NH 4 + N stripped out (mg) NH 4 + N (mg/L) NH 4 + N collected (mg) 0.0 650.00 10 yes 1300.00 1.8 471.87 9.54 yes 943.75 356.25 0.00 0.00 7.9 446.68 10.5 No 893.36 406.64 -22.5 396.30 10.5 No 792.60 507.40 -46.5 312.33 9.84 yes 624.66 675.34 -69.6 211.57 9.18 yes 423.13 876.87 354.32 956.65 73.5 161.18 8.92 yes 322.37 977.63 354.32 956.65 93.6 144.39 8.6 yes 288.78 1011.22 404.70 1092.68 97.6 135.99 11.1 No 271.99 1028.01 438.29 1183.37 118.6 1.64 9.58 No 3.28 1296.72 471.87 1274.06
113 Figure 4 1 Struvite recovered from experiments A ) Struvite from the p ilot p lant B ) R aw KMAG C ) C rystals recovered from solution of l angbeinite. D ) S truvite from pilot plant crystallized in the solution of KMAG E ) P ellets from mixing struvite and raw KMAG F ) Struvite crystallized with the solution of KMAG in acidic conditions (A) (F) (B) (C) (D) (E)
1 14 Figure 4 2 C ontinuous c rystallization o f KMAG with s truvite Figure 4 3 Crystallization of KM AG with s truvite
115 Figure 4 4 Absorption of a mmonia in KMAG Figure 4 5 KMAG with struvite: pH adjusted to 2.2 at commencement of crystallization
116 Figure 4 6 KMAG from saturated solution
117 ( A ) ( B )
118 ( C ) ( D )
119 Figure 4 7. X ray diffraction results of solids from mixing KMAG with synthetic struvite A ) Struvite. B)Ammonium sulphate. C) Schoenite. D)Picromerite and periclase. E) Ammonium magnesium phosphate hydrate. F) Ammonium magnesium phosphate ( E ) ( F )
120 ( A ) ( B )
121 Figure 4 8 X ray diffraction of r aw KMAG with struvite from the pilot scale run A ) Ammonium magnesium phosphate. B) Ammonium magnesium phosphate hydrate. C ) Schoenite and picromerite ( C )
122 ( A ) ( B )
123 ( C ) ( D )
124 Figure 4 9 X ray diffraction of solids from mixing KMAG solution with struvite from the pilot scale run A ) Struvite. B) Schoenite. C) Ammonium calcium phosphate. D) Ammonium magnesium phosphate. E ) Synthetic struvite. ( E)
125 Figure 4 10 Change in the potassium concentration in the semi continuous crystallization of KMA G with struvite Figure 4 11 Change in the phosphorus concentration in the semi continuous crystallization of KMA G with struvite 0 5 10 15 20 25 30 35 40 45 50 0 5 10 15 20 25 30 35 Potassium g/L Time (hr) 0 0.5 1 1.5 2 2.5 0 5 10 15 20 25 30 35 Phosphorus g/L Time (hr)
126 ( A ) ( B) ) )
127 Figure 4 12 X ray diffraction analysis of solids from the semi continuous run A ) Picromerite and schoenite. B) Ammonium sulphate. C ) Potassium ammonium hydrogen phosphate ( C )
128 Figure 4 13 Recovery of n itrogen and p hosphorus from c entrate
129 CHAPTER 5 C ONCEPT TO C OMMERCIALIZATION (C2C) 5.1 Commercializing Struvite If conditions for struvite precipitation which were predicted by chemical equilibrium model can be duplicated and exploited in a practical engineering pr ocess, struvite may potentially be economical to extract from wastewater in commercial quantities. This might be done by precipitating struvite in a dedicated reactor instead of allowing spontaneous formation (Stratful et al., 2001; Munch and Barr, 2001; H ao and van Loosdrecht, 2003). This spontaneous formation fouls pipes and other inner surfaces of the treatment process, making operation of the plant inefficient and costly because the struvite must be dissolved with sulfuric acid or broken down manually w ith hammer and chisel (Durrant et al., 1999; Stratful et al., 2001; Williams, 1999). Another advantage of a dedicated reactor is that excess ammonium, which is a normal product of anaerobic wastewater digestion, might yield relatively pure struvite. A gro wing interest has arisen of water utilities in preferential struvite crystallization as a way to control the problems of spontaneous deposition of struvite in post digestion treatment lines and limiting phosphorus recycling in their wastes. However, the ap plication of full scale struvite crystallization processes in wastewater treatment plants is limited due to problems associated with quality control and quantity of the product generated. One of the first points that water utilities should take into accoun t, when considering struvite crystallization concerns the selection of the most appropriate waste stream to precipitate and recover struvite. Anaerobically digested sludge liquors where concentrations in phosphorus and ammonium are relatively high seem the most suitable for this because in that type of liquo r, magnesium and pH should remain the only factors to adjust to initiate struvite crystallization.
130 Step 2 is to precisely determine the chemical and physical characteristics of the liquors duri n g the crystallization process. The key parameters such as solution supersaturation, presence of foreign ions, pH and mixing energy could be responsible for dramatic changes in struvite crystal characteristics in terms of size, shape and purity as well as the quantity of the product generated. To illustrate the purity of the formed crystals was found to be affected by the pH at which precipitation occurs and by the presence of calcium ions if the residence time for precipitation and crystallization is high. In this study, t he optimum pH of precipitation was identified to be pH 8.7. Magnesium ions are commonly known to be the limiting ions that prevent struvite precipitation from real liquors M oreover they were found to be of great influence on struvit e precipitation because increasing doses of magnesium could reduce significantly struvite impurity. This researcher concluded from the present study that the magnesium dosing should not be too excessive because this would result in excessive supersaturatio n ratios which would cause rapid nucleation hence production of other magnesium precipitates. The operator would then have to determine the appropriate dose of magnesium (as MgCl 2 ) and the NaOH (for pH control) necessary to form struvite to ensure the s electivity of the process toward struvite and to control the quality of the product formed. This implies that application of full scale struvite crystallization processes would require a regular and accurate monitoring of the liquors chemical characteristics to regulate pH and chemically counterbalance Mg:N:P when needed. Phosphorus can be obtained from sludge only with biological processes because the commonly used chemical precipitation of phosphorus, which involves the addition of iron and aluminum salts, produces a product that cannot be recycled for industrial recovery of phosphorus (Donnert and Salecker, 1999). No common method can recover struvite from biological
131 processes E quipment ranges from complex patented reactors employ ing ion exchangers to simple stirring tanks (Gaterell et al., 2000; Williams, 1999). From the literature, this researcher found that struvite crystallization from digested sludge liquors by means of fluidized bed reactors directly integrated in wastewate r treatment plant lines are the most common processes ( Bhuiyan a et al., 2008) However, if good phosphorus removals are always insured, these techniques still need improvement with regard to quality of the product formed to be applicable as an economicall y valuable route to recover phosphorus. The application of struvite crystallization processes at full scale is not widespread due to : 1) the unknown economical value of the process and the product, 2) the need of pH control, 3) the necessity of long operat ional times to ensure quality of the product and 4) the formation of crystals. Several laboratory and pilot scale studies have been carried out to assess the potential removing and recovering phosphorus as a reusable product, and a few of them have been t ested at full scale in The Netherlands (Giesen, 1999) and Italy (Battistoni et al. 2005a 2005 b). However Japan is the only country where complete phosphorus removal and recovery from anaerobically digested sludge liquors as struvite have been implemente d with the resulting production sold to fertilizer companies (Gaterell et al. 2000; Ueno and Fujii, 2001). Struvite was produced in Japan from the filtrate of anaerobic sludge digestion by adding magnesium hydroxide and adjusting the pH from 8.2 to 8.8 with sodium hydroxide. A retention time of 10 days allowed the growth of pellets 0.5 to 1.0 mm in diameter. The recovered struvite was sold to fertilizer companies (Ueno and Fujii, 2001). Another experimental industrial process provided simultaneous remova l of phosphate and ammonium ions by selective ion exchange and recovery of the product by chemical precipitation in the form of struvite (Liberti et al., 2001).
132 Usually, pilot plants are used to reduce the risk associated with construction of large process plants. It is done in following ways: Evaluating the results of laboratory studies and making product and process corrections and improvements Determining possible salable co products Determining waste streams requiring treatment before discharge Identify ing scale up issues Providing data that can be used in making a decision on whether or not to proceed to a full scale production process I n the case of a positive decision, designing and constructing a full size plant or modifying an existing plant They a re substantially less expensive to build than full scale plants. The business does not put as much capital at risk on a new project for testing. Further, design changes can be made more cheaply at the pilot scale and kinks in the process can be worked out before the large plant i s constructed They provide valuable data for design of the full scale plant. Scientific data about reactions, material properties, and corrosiveness, for instance, may be available, but it is difficult to predict the behavior of a process of any complexity. Engineering data from other process es may be available, but this data cannot always be clearly applied to the process of interest. Designers use data from the pilot plant to refine their design o f the production scale facility T his chapter reports studies on the pilot scale carried out on centrate from the Jacksonville W astewater T reatment P lant. The City of Jacksonville wastewater system collects and treats an average of 240 million g al of municipal sewage every month. The collection transports both storm water and sanitary flows. The wastewater is treated at a treatment plant located on the north side of Jacksonville. Once treated, the effluent flows into Mauvaisterre Creek. Most of Jacksonville sewer system is constructed as a combined sewer. Street inlets are tied into it to convey storm water during storm events. The combined flow from customers and storm inlets is collected from service laterals into the over 90,000 linear
133 feet of sewer lines in J acksonville. They tie into three main trunk lines that convey them to the wastewater treatment plant. The Jacksonville plant is a secondary treatment system for wastewater. The processes include screening aerated grit removal, primary clarification with scum removal, activated sludge, secondary clarification, chlorination and dechlorination and finally post aeration before discharge into Mauvaisterre Creek. The plant also treats sewage and biosolids from surrounding communities. Biosolids are processed through mesophilic anaerobic digesters, followed by centrifuges, gravity belt thickeners and storage tanks with one year of biosolids storage capacity. Filtrate from the dewatering process is then returned to the head of the works. This p resent study illu strates the trial runs of operating a pilot plant with the s equential b atch r eactor (SBR) technology to produce and recover struvite. GreenTechnologies, LLC constructed a pilot plant in Starke, Florida, for processing 1 000 g al of centrate per batch for re covering nitrogen and phosphorus, using technology developed in this research project. The p rimary goal of the pilot project was to prove that phosphorus recovery was possible using the SBR technology on real process fluids on a technical scale. The p ilot plant was operated with the following specific objectives: Test the technology, on real process fluids which validate feasibility of struvite precipitation technique F ind the amount of struvite and volume of the settled solid sludge Determine the quantity of chemicals needed for struvite recovery Carry out process economics and optimization for a large scale struvite plant using the pilot scale operation experience Produc e small quantities of struvite for chemical, x ray diffraction evaluations, packaging, transportation and storage stability studies Develop a business model of struvite recovery
134 5.2 Materials and Methods The key unit operation in the SBR approach is the settling. A literature review found that a typical struvite particle has a 0.004 cm rad ius (Buchanan, 1993) and a 1.7 g/cm 3 density (Sharp, 2002). Using Stokes law for particle settling, this researcher determined that the average settling velocity will be 0.0027 m/s, which is less than the liquid velocity in the tank. The liquid solid separ ation is carried out in the SBR. The length of the conical bottom is based on the particle settling velocity. The dimensions of struvit e tanks are provided in Table 5 1 D imensions of other equipment were not available at the time of the conducted trials. Tankage is required to store the MgCl 2 6H 2 O and NaOH solutions. Details of the pilot plant layout and flow diagram are given in Figures 5 1 and 5 2 with details of equipment listed in Table 5 1. F ollowing standard operating procedure was developed for operating the pilot plant: Prepare 10 M solutions of magnesium chloride, sodium or potassium hydroxide (subject to availability) and 50% phosphoric acid which are stored in chemical tank s CT1, CT2 and CT3 respectively T ransfer the centrate from the st orage tank (ST), open ball valves BV1 ST, BV1 P1, BV2 P2 and start pump P1 D ose magnesium chloride and caustic, open BV1 CT1 BV1 CT2 BV1 CP1 and BV1 CP2 Adjust the flow of chemicals from CT1 and CT2 according to the flow of centrate from the storage tank using dosing pumps DP1 and DP2 Transfer the r eaction mixture to mixing tank (MT) by opening valve BV1 MT Phosphoric acid from CT3 can be fed to the mixing tank depending on the need for additional nitrogen removal Open BV3 MT which is in the overflow line from the mixing tank. The r eaction of struvite formation starts in the static mixer and mixing tank. The e ffect of mixing either by using the static mixer and/or mixing tank is studied Open BV1 SR1 and FILL the struvite tank (SR1) with the centrate already mixed with chemicals. About 500 gal of the reaction mixture is filled in the struvite tank by visual ly monitoring the liquid level
135 Allow t he liquid mixture to REACT and/or SETTLE f or 30 min in t he struvite tank T ransfer t he c entrate with precipitated struvite to the struvite tank (SR2) by repeating the previous procedure Thus, struvite precipitation is carried out in two steps for comple te nitrogen/phosphorus recovery Transfer the settled soli ds, open BV5 SR2 and BV6 SR2 and feed the settled struvite slurr y from SR2 to the filter press Drain t he c entrate free of nitrogen and phosphorus using pump P1 and opening BV1 Dr 5.3 Results and Discussion The p eliminary trial runs were conducted on the pilot scale using centrate from W astewater T reatment P lant. This run validated the model predictions for the caustic requirement, magnesium requirement and residual concentrations of nitrogen and phosporus in the centrate 5.3.1 Trial 1 The c entrate from JEA Wastewater Treatment Plant was found to have the following characteristics: pH : 7.1 NH 4 + N : 780 ppm PO 4 3 P : 135 ppm Ca 2+ : 37 ppm K + : 75 ppm As observed in laboratory scale runs, complete phosphorus and nitrogen recovery from centrate was found to be successful in two steps achieved for keeping struvite purity in the solids maximum (>99%). This concept was verified on the pilot scale which is explained as follows.
136 22.214.171.124 Treat ment Step 1 In S tep 1 phosphorus originally present in centrate was recovered. The s tandard procedure as explained earlier was not followed completely because the complete pilot plant was not set up at the time of this trial. The s truvite precipitation reaction was carried out by adding chemicals to 240 g al of centrate in the tote tank. A total of 687.54 gm of magnesium chloride hexahydrate was used for phosphorus recovery. The pH was adjusted to 8.7 which required 1.074 kg of KOH. The chemicals were mi xed using an agitator in a tote tank. After the chemicals mixing, the centrate was transferred to a conical bottom 500 g al tank at a rate of 16 g al /min. The c entrate was allowed to settle for 30 min and a supernatant sample was taken. Nitrogen and phosphorus were recovered in the form of solids. The solids and liquid separation level was visible at the bottom of the struvite tank. The s upernatant and the struvite sludge generated were stored in the conical bottom tank. The s eparation of solids was not done at the end of this step. In the supernatant sample, NH 4 + N was found to be 739.3 ppm and PO 4 3 P to be 9.4 ppm. This gave 93% recovery of PO 4 3 P. The c alcium concentration in the centrate decreased from 37 ppm to 25 ppm. The p otassium concentration in the centrate also dec reased from 135 ppm to 108 ppm. Battistoni et al., ( 2000) demonstrated the operation of a full scale fluidized bed reactor struvite formation from the belt press liquors using aeration t o adjust the pH. They found only 61.7 % to 89.6 % removal of PO 4 3 in the process as compared to 93% removal of PO 4 3 P in the present study Jaffer and Pearce (2004) also reported the operation of an air agitated reactor on full scale for recovery of struvi te from centrifuge liquors where they found only 60 % to 80% recovery of phosphate. The Shimane Prefecture reactor removed 90% of ortho phosphate from the treated supernatant.
137 126.96.36.199 Treatment Step 2 In this step, additional phosphorus in the form of phos phoric acid was used to recover residual NH 4 + N in the centrate previously treated. After about 22 hours of initial phosphorus recovery, NH 4 + N was found to decrease to 699.5 ppm from 739.3 ppm. This could be due to the increased pH (8.7) causing ammonia to volatilize. The c entrate was transferred to the tote tank and the following chemicals were added to it : 85 % Phosphoric Acid : 2.3 kg MgCl 2 .6H 2 O : 3.588 kg KOH : 2.5 kg pH adjusted to : 8.9 The c entrate was transferred to the conical bottom struvite tank. The solids were allowed to settle for 30 min and the supernatant sample was analyzed which w as found to contain 452 ppm of NH 4 + N and 157 ppm of PO 4 3 P. The m odel predicted a supernatant ammonia concentration of 481 ppm but the actual experimental residual concentration was less than 481 ppm perhaps due to ammonia volatilization. The m odel predicted a residual PO 4 3 P concentration of 105 ppm which was less than an experimentally observed concentration of 157 ppm. The m odel predicted a formation of calcium phosphate precipitates in addition to struvite. Step 2 was carried out to test the recovery of additional ammonia but the centrate after this step can be treated further to recover PO 4 3 P to recover more struvite. A total of 6.1 kg of dry solids were recovered from both treatment steps which were close to the expected value of 5.9 kg computed by the model. Struvite solids were analyzed for chemical composition and x ray diffraction (XRD) as explained in Chapter 3. The solids were utilized for struvite product development as explained in Chapter 4. The c omposition of the struvite solids produced in the trial runs is shown in Table 5 1. The magnesium, ammonium, and phosphate content were
138 similar to the theoretical values of struvite. The s olids recovered were found to be mainly struvite which is also inferred from x ray diffraction analysis as shown in Figure 5 3. The s olids produced were found to have a good match to the theoretical int ensity values for struvite. The XRD analysis did not show the p resence of other solids along with struvite which indicated that calcium precipitation did not occur. This is also confirmed by chemical composition of solids where calcium content was o nly 0. 0049% as shown in Table 5 2. 5.3.2 Trial 2 The p revious pilot plant run on struvite recovery incorporated two step operations. In S tep 1 phosphorus originally present in centrate was recovered along with some nitrogen. In Step 2 addition recovery of nitrogen was carried out by adding extra phosphorus in the form of phosphoric acid. Unlike this operation, T rial 2 was carried out by combining the two steps. Initially, the pH of centrate was found to be 7.64 with ammonia N concentration of 740 ppm. The m odel simulations suggested the addition of the following chemicals for 240 g al of centrate: 2.3 kg of 85 % phosphoric acid and 3.5 kg of MgCl 2 .6H 2 O. A total of 3.7 kg of potassium hydroxide was needed for pH adjustment. These chemicals were added to 240 g al of centrate in the tote tank. The solution was mixed using an agitator. At the start of the experiment, the centrate pH was 7.64 The a ddition of phosphoric acid decreased pH to 6.0. Then 3.5 kg of magnesium chloride was added. Potassium hydroxide was a dded gradually F irst 2 kg of KOH was added and the mixture was recirculated for 15 min which increased the pH to 6.7. A n a dditional 1 kg of KOH was mixed in the solution which increased the pH to 7.5. Another 0.7 kg of KOH was added slowly until the pH was set to 8.7. After pH adjustment, the centrate was transferred to a 500 g al conical bottom tank for settling and separation of the solids. The s olids were found to settle in 10 min. A sample was taken to determine solids concentration in supernatant af ter 10 min and 30 min to study the
139 settling of the solids. The s uspended solids concentration at both times in the supernatant was found to be 10 ppm suggesting complete settling of the solids. Samples were taken for chemical analysis after 30 min after t ransfer. The f inal ammonia nitrogen was found to be 240 ppm and phosphorus of 85 ppm. 5.3.3 Comparison between Laboratory and the Pilot Scale Runs The a mount of chemical addition in the pilot scale runs was not calculated by scaling up the laboratory scal e runs but this researcher found to be proportionate to the laboratory runs. The chemical amounts used in the pilot scale runs were determined by model simulations. As discussed in C hapter 3, when 7 L centrate was subjected to struvite recovery the amoun t of chemicals required to recover only phosphorus was as follows: 4.72 g of MgCl 2 .6H 2 O, 3.24 g of NaOH. The amount of sodium hydroxide is equivalent to 5.83 g of KOH for laboratory scale. This scaled up to 757.3 g of KOH required for 240 g. But this resea rcher observed that 1.074 kg of KOH was actually added in the pilot scale as explained in treatment S tep 1 of T rial 1 possibly due to the aeration provided for mixing in the laboratory scale which enhanced the pH increase. The amount of magnesium chloride was found to be proportional to the pilot scale run conducted on 240 g al of centrate. The l aboratory amount of magnesium chloride hexahydrate scaled up to 611.8 gm whereas in the pilot scale, 687.54 gm were added. The s o lids were not separated from the centrate in the pilot scale run so a comparison with the laboratory scale run was not done. In the laboratory scale runs, 99.6% phosphorus recovery was observed which was close to 93% removal of PO 4 3 P observed in T rial 1 treatment S tep 1. In the laboratory scale experiments, magnesium chloride was dosed to the centrate in two steps in the SBR I n the pilot scale run magnesium chloride was dosed in one step, as explained in treatment S tep 1 of T rial 1. When 7 L centrate was subjected to recovery of both nitrogen and phosphorus, the amount of chemicals required was as follows: 39.96 g of MgCl 2 .6H 2 O; 4.29 g of 85% H 3 PO 4
140 which was proportionate to the chemicals added for 240 g al of centrate as explained in T rial 2. Data on the caustic usage in the laboratory scale experiments were not available and not compared to the pilot scale run. 5.4 Scale up Considerations in Precipitation Processes for precipitation of solid products from dissolved reactants are almos t always carried out in stirred reactors. The reactors are operated semi batchwise or continuously. In both operation modes, one reactant is added to a stirred solution containing an excess of other reactant s The functions of a stirrer include : mixing of reactants, suspension of formed solid particles, and promotion of heat transfer to the wall. The primary objective of the bench scale experiments is to find out which phenomena are important in the process. First it is important to decide for the process whether primary or secondary nucleation prevails. When it follows from the lab experiments that meso mixing (mixing of incoming stream with the reactor contents) has a strong influence on particle size, primary nu cleation is probably essential. When particle size is related to slurry concentration then secondary nucleation must be taking place. Second, it is important to find out whether or not growth of particles to their final size is mainly determined by the s urface growth or by agglomeration. Struvite precipitation carried out in the present stud y, takes place with primary nucleation with no agglomeration due to short hydraulic retention time. Therefore, the design of the reactor should be aimed at an effective control of meso mixing and of the mass transfer of the growing crystals. The mass transfer of the growing crystals is mainly determined by specific power consumption of the impeller. At a given power input, meso mixi ng is dependent on the way reactants are introduced into the reactor. However, secondary nucleation may occur on larger scales because the tip speed of the impeller increases with the scale (for constant specific
141 power input). Also, agglomeration may take place on a larger scale, even if does not on a lab scale since the longer circulation times may give agglomeration a better chance. 5.5 Discussion of Feasibility on Full scale For the purpose of estimating the cost of struvite precipitation from centrat e, the treatment of 700,000 gal per day would be needed. For this operation, the same process sequence explained for the pilot plant was used. Table 5 4 illustrates specifications of the equipment and related approximate cost. A low pressure high volume p ump would transfer the centrate into the mixing tank through a static mixer. Within the static mixer, MgCl 2 6H 2 O and NaOH would be mixed with the centrate. An overflow at the top of the reactor would then direct the centrate to the struvite SBR tank. The s upernatant from this tank would be supplied with chemicals for the second step which would tank the centrate to the second struvite tank. Commercial tanks with this volume can be purchased with 45 conical bottoms and cylindrical sides. A mixing tank ment ioned earlier is needed to allow the mixing of wastewater, magnesium and a base, and it provides sufficient hydraulic retention time (HRT) for the precipitation reaction to occur. The p resent research has shown 30 min to be an appropriate HRT. 5.6 Econom ics of Struvite Recovery An estimate of the chemical requirement for the pilot pl ant runs is shown in Table 5 3. The chemical requirement varies with the change in ammonia N concentrations in the centrate. The a mmonia N concentration in the effluent centrate is targeted to 100 ppm. An economic analysis of the pilot plant is ca rried out as shown in Table 5 4 with the following four assumptions: a. 1 000 g al of centrate processed per day (8 h of operation) b. 120 mg/L of PO 4 P available for treatment from centrate c. Effluent concentration of NH 4 + N is allowed to be 100 mg/L d. Selling cost of struvite = $500/ton
142 Since struvite contains phosphorus and nitrogen, removal will affect the content of both elements in the leftover sludge, which is used by farmers as a soil improvement agent and fertilizer. Struvite recovery from wastewater might have a marginal effect on the net content of nitrogen, but a greater impact on the concentration of phosphorus in the sludge. Gaterell et al. (2000) calculate d b ecause sewage has a typical N:P ratio of 8:1 and struvite 1:1 that a theoretical maximum of 12.5% of the nitrogen load could be removed as struvite. The practical limit is lower because not all phosphorus can be recovered as struvite Struvite removal leaves less phosphorus in the sludge. This is beneficial from two perspectives: 1. Sludge applied to fields usually has phosphorus in excess of the needs of plants. Subsequent eutrophication of water bodies from leached phosphorus compounds would be reduced 2. Struvite recovery would help to meet legal requirements imposed on sludge disposal and reduce the area needed for disposal The s truvite recovery technique offers material, energy and transportation cost savi ngs for the wastewater treatment plants. The transportation savings is a result of the reduced sludge transport. Operation cost savings will be in the prevention of clogged pipes from struvite crystals in the pipeline. This unintentional struvite depositio n leads to additional pumping and cleaning costs that can be avoided by the intentional formation and recovery in the reactor. For a typical wastewater treatment plant serving 1 million people, the annual cost of dealing with the struvite problem ranges fr om $160,000 to $800,000. The total nitrogen and phosphorus market for the United States and Puerto Rico is more than 21 million tons. In terms of struvite, 19 million tons annually would be needed to fulfill that demand. However, slow release fertilizers, such as struvite, account for only 4% of the total fertilizer market or roughly 791,000 tons/yr.
143 Because 84 % reactors, struvite has low production costs compared with fertilizers produ ced from mined phosphorus. A s trong demand and higher input costs are rapidly pushing prices for phosphate end products to record highs. Global companies that import phosphate rock are faced with costs nearly five times higher than a year ago, necessitatin g high end product prices. PotashCorp is the world's third largest producer of phosphates and second largest seller of phosphoric acid and with high quality, low cost phosphate rock, this company is well positioned in these unprecedented phosphate market conditions. PotashCorp expects a combined gross margin for nitrogen and phosphate to exceed 2007 levels by 20 25 % (Agri Marketing E Newsletter, 1/25/08. PotashCorp Issues 2008 Outlook). The economics of struvite recovery also become important in final prod uct marketing. The value of the recovered phosphorus in the form of struvite, tends to be higher than that of phosphate ore, due to higher quality (in terms of purity) as well as low heavy metal contamination. The intrinsic value of struvite is also important because it was identified as a non burning, slow release fertilizer as early as the 1960s. The disadvantage is that struvite is more costly to produce, that is, US D $140 to $ 460 per ton, compared to the USD $40 to $ 50 per ton that is required for phosphate ore. This cost differential is offset by the suggested market value of USD $198 to $ 1885 per ton (Ueno and Fujii 2001; Jaffer et al. 2002). Several fluidized bed reactors have been in full scale production in Japan and the recovered struvite has been sold as a fertilizer (Shimamura et al. 2003) at a price of approximately USD $250 per ton in 2001 dollars (Ueno and Fujii 2001). This market analysis demonstrates, if distributed successfully, that wastewater treatment plants can generate revenue to offset operational costs through struvite recovery processes C oncurrently these processes are becoming more sustainable.
144 5.7 Struvite Recovery as a Business Results from the pilot scale runs helped in the demonstration of the commercial potential of the struvite process. An approximate capital cost of a full scale plant of 700,000 gal per day ( GPD ) capacity is presented in Table 5 5. With a selling price of $5 00 per ton the payback period of the plant of 700,000 GPD is about six years. This will decrease if the struvite selling price is increased to up to $750 per ton. The market price of struvite is flexible and this researcher found that some of the compani es sell struvite at a price of $1 200 per ton. The e conomic analysis in the present study is carried out with conservative estimates. The s truvite produced gave a clear idea of the amount of the product recovered as well as quality. The pilot plant is used to reduce the investment risks which are usually encountered in unproven production methods. The pilot plant continue s to operate to test ideas for new products, new feedstocks, and different operating conditions. Alternatively, it is operated as a produc tion facilit y augmenting production from the main plant. A struvite recovery business using the developed technology can offer design, consulting and construction management services for wastewater nutrient recovery systems for primary customers such as local governments and agricultural operations. In addition to these services the recovery process can distribute a superior slow release fertilizer additive in the form of struvite crystals to fertilizer blenders, 5.8 Outcomes The p ilot scale trial recover ed 93 % of PO 4 3 P from the centrate in a single step Ammonia recovery was possible by the addition of phosphoric acid to the centrate if additional nitrogen recovery is desired The amount of magnesium chloride added in the pilot scale was proport ionate to the amount needed in the laboratory scale for phosphorus and nitrogen recovery. The a mount of caustic needed for pH adjustment in the pilot scale was more than the laboratory scale The recovered struvite contains almost no hazardous materials a nd exhibits an equivalent or better fertilizer effectiveness than co nventional chemical fertilizers
145 For a plant treating 750,000 g al per day of centrate, a capital investment of $1.5 million is needed ; a net profit of $452,000/year is possible.
146 Table 5 1. Details of equipment used in the pilot p lant Equipment Capacity Storage tank (ST) 5000 gal Flushwater tank (FT) 1000 gal Chemical tanks (CT1, CT2 and CT3) 50 ga l Mixing tank (MT) 50 gal H 500 gal Major pipes of PVC (i.e. all except pipes from chemical tanks, CT1, CT2, CT3) All ball valves (BV) manually operated. Table 5 2 Composition of struvite generated in the pilot scale runs Component Theoretical Value (%) Struvite from Trial 1 (%) Struvite from Trial 2 (%) magnesium (Mg) 9.8 9.7 9.7 nitrogen (N) 5.7 5.9 5.7 phosphorus (P) 12.7 12.9 12.8 c alcium (Ca) 0.049 0.018 p otassium (K) 0.6 0.51 Table 5 3 Chemical r equirement for the p ilot p lant r uns NH 4 N in Centrate (mg/L) NH 4 N removed (mg/L) Total N removed per day (kg) Total P needed (Kg) 85% H 3 PO 4 needed per day (L) MgCl 2 .6H 2 O needed (Kg) NaOH needed (kg) Struvite Produced per day (kg) 1000 900 3.41 3.91 8.61 23.19 142.36 33.34 950 850 3.22 3.49 7.69 21.90 134.45 31.48 900 800 3.03 3.07 6.77 20.61 126.55 29.63 850 750 2.84 2.66 5.84 19.33 118.64 27.78 800 700 2.65 2.24 4.92 18.04 110.73 25.93 750 650 2.46 1.82 4.00 16.75 102.82 24.08 700 600 2.27 1.40 3.08 15.46 94.91 22.22 650 550 2.08 0.98 2.16 14.17 87.00 20.37
147 Table 5 4 Equipment c ost of the SBR p ilot p lant Unit Number of units Capacity Unit Cost $ Total Cost $ Storage Tank 1 5000 gal 5000 Flushwater Tank 1 1000 gal 1000 Chemical Tanks 3 250 gal 250 Mixing Tank 1 50 gal 50 Struvite Tanks 2 500 gal 500 Major Pipes 100 ft PVC 3 schedule 40 4.31/foot 431 Small connection pipes for chemical tanks 50 ft PVC 2 schedule 80 2.87/foot 143.5 PVC Fittings 30 --200 Ball Valves 20 PVC 3 56.85 1177 Ball Valves 10 PVC 2 9.95 99.5 Static Mixer 1 PVC 3 100 100 Centrifugal Pump 1 1000 Dosing Pumps 3 450 1350 Blower 1 450 Instrumentation 4000 Estimated Total = $15,751
148 Table 5 5 Capital c ost of a f ull s cale s truvite p lant; c apacity: 700,000 GPD Equipment Capacity Approximate Cost 2 x Struvite Tanks 20,000 gal $100,000 Filter Press 2,000 gal $100,000 3 x Fiber Glass Chemical Tanks 6,500 gal $150,000 2 X Pumps 500 gal/min $10,000 3 X Dosing Pumps 10/min $45,000 Storage Tank 50,000 gal $50,000 Mixing Tank 2,000 gal $2,500 Instrumentation and Controls -$35,000 Piping and Fitting 500 ft $5,000 Valves: 15 automatic $7,500 Building 3,000 sq. ft $250,000 Engineering Fees -$100,000 Contractor Fees -$300,000 Drier 40 tons $300,000 Packaging Equipment 1 $50,000 Total $1,505,000
149 Figure 5 1. P rocess l ayout
150 Figure 5 2 Process f low d iagram
151 CHAPTER 6 CONCLUSIONS AND FUTU RE WORK 6.1 Overall Findings The critical findings in the process development study are as follows: 6.1.1 Mathematical Model of Struvite Precipitation A model was developed for predicting precipitation in closed systems containing solut ions of ammonium, magnesium and phosphate. The model incorporates 15 precipitates and explicitly solves for precipitate, residual ion and dissolved species concentrations using mass balance equations for magnesium, phosphorus and nitrogen along with c hemical equilibri um and charge balance equations Using the polymath program and a solution procedure that involves converging the residual phosphate concentration to within tolerance limits, the concentration of the different precipitates and residual ion and dissolved species concentrations were determined The model was validated against data collected from literature for synt hetic and real wastewaters. The model was able to predict struvite to within 1.6% to 9% and the total precipitate prediction errors ranged from 1 % to 24.5% This researcher found that for solutions containing Mg, NH 3 and PO 4 the optimal pH for struvite concentration depends on the initial ratio of ammonia, magnesium and phosphate. A pH of 9.0 optimizes the struvite concentration when the ratio is 1:1:1 and a higher pH of 9.8 when magn esium and phosphate are limiting The e quimolar stoichiometric ratio of magnesium, ammonium and phosphate (i.e the ratio of their occurrence in struvite) was not ideal for struvite precipitation. To obtain pure struvite it was necessary to have excess ammonia in the solution with magnes ium being the limiting nutrient A titration curve was generated by using the exions concentration and was validated using an experimental curve. The m odel determined an exion concentration of 0.002566 M in anaerobically treated wastewater from dairy operations. This species concentration should be considered while calculating the amount of caustic needed for pH adjustment The m odel accurately predicted the amount of a base needed for pH adjustment with an error less than 12.3% as compared to the experimental quantities The p resence of calcium and carbonate species was found to decrease struvite purity In the case of centrate, an excess ammonia concentration and an Mg:P ratio of 1:1 was predicted to be suitable for struvite formation considering the purity o f struvite of 95% at this ratio
152 In the case of centrate and dairy flushwater, struvite precipitation was not suitable for an Mg:P ratio of 0.5:1 with excess ammonia because phosphate was removed mainly as calcium phosphate and calcium magnesium carbonate 6.1.2 Sequential Batch Re actor Operation for Recovery of Nitrogen and Phosphorus as Struvite from Sewage Sludge Centrate and Dairy Wastewater The pH level within the recovery processes varies from 8.2 to 9 depending on the strategy of the recovery process. Within the current work a pH of 8.4 to 8.7 was selected to allow for rapid struvite nucleation, being aware that this can cause the precipitation of fines The a eration of centrate was found to increase the pH to 8.4. This is useful for minimizing the caustic requirement for str uvite formation and also for reducing carbonate concentratio n. This improves the quality of struvite by reducing solids with carbonate The NaOH required to adjust the pH of centrate to 8.7 for complete phosphorus recovery was found to be 432 mg/L without aeration The c omplete nitrogen and phosphorus removal is possible from centrate in a sequential batch reactor. The s ettling of precipitated struvite is rapid and it was found to settle in 10 min. The s truvite separated from the bottom of the reactor was f ound to have 80% moisture. Filtered struvite was found to have 65% moisture The recovery of phosphorus from animal waste as a struvite containing precipitate has been successfully demonstrated using an SBR mode of operation. The next step in the growth of this technology is the development of a field scale recovery unit at a commercial animal production unit. The operation of a field scale recovery unit would supply the necessary data to complete a cost/benefit analysis to investigate the economics of th e technology. A cost effective magnesium source of magnesium chloride and a fast, low cost method of pH adjustment using sodium hydroxide can be utilized to successfully implement thi s technology at the farm scale 6.1.3 Fertilizer Product Formulation U si ng Struvite Solids A cationic polymer can used to remove fines ( struvite crystals ) from the reactor. The p olymer should be add ed after 30 min of struvite The s ettling of struvite solids was improved by a cationic polymer commonly used in wastewater treatm ent plant s With the polymer usage, the solids cont ent increased from 12 % to 16% The KMAG was crystallized to schoenite and pi cromerite in the presence of water Struvite was found to recrystallize with saturated KMAG to give a product rich in nitrogen, phosphorus, potassium, and magnesium These crystals were found to be sparingly soluble in water at a neutral pH
153 The s olution of KMAG and struvite under acidic condition was crystallized to hannayite and schoenite The x ray diffraction analysis confirmed that struvite was agglomerated to KMAG Calcium precipitation was not detected in the crystallization of KMAG and struvite Semi continuous crystallization of struvite with KMAG was found to produce ammonium sulphate, potassium ammonium phosphate, schoenite and picromerite Saturated KMAG was found to absorb ammonia under acidic conditions to give a solid product rich in nitrogen. Crystals formed in the saturated KMAG solution after ammonia absorption were fou nd to retain ammonia in soli ds 6.1.4 Pilot Scale Study of the Sequential Batch Operation for Struvite Precipitation The p ilot scale trial was found to recover 93 % of PO 4 3 P from the centrate in a single step Ammonia recovery was possible by the addition of phosphoric acid to the c entrate if additional nitrogen recovery is desi red The amount of magnesium chloride added in the pilot scale was proportionate to the amount needed in the laboratory scale for phosphorus and nitrogen recovery. The a mount of caustic needed for pH adjustmen t in the pilot scale was more than the laboratory scale The recovered struvite contains almost no hazardous materials and exhibits the equivalent or better fertilizer effectiveness than co nventional chemical fertilizers For a plant treating 750,000 g al per day of centrate, a capital investment of $1.5 million is needed and a net profi t of $452,000/year is possible 6.2 Future Work 6.2.1 Automation of the Pilot Plant The p ilot plant in this study was run by manually feeding centrate and other chemicals. A model based control system can be developed using the model described in Chapter 2. The model can calculate the chemicals required for a recovery of nitrogen and phosphorus and the model can be used to dose appropriate amounts in the reactor. Thus, con trol algorithms can be developed. The main gap in instrumentation is an online phosphorus probe which is not available in the market for commercial application. A number of sensors can measure the total or individual chemical species such as ammonia N, alkalinity and pH. O verall, a quantum advance
154 in application and sophistication of instrumentation and cont rols has occurred and it is an effective option for an improved process loading rate and conversion efficiency. Automation of the pilot plant can be used to test feasibility of the model application with real time controls 6.2.2 Biological Struvite Formation with E xisting Enhanced Biological Phosphorus Removal (EBPR) Processes The formation of extracellular struvite crystals by bacteria was first des cribed by Robinson (1989). Cyrstal formation is induced by the microorganisms by the combination of NH 4 + -released by the metabolism of nitrogenous organic substance s w ith magnesium and phosphate that are present in the medium with concomitant precipitatio n of struvite. Struvite formation is also achieved by the physical chemical process when ammonium, magnesium and phosphate are present in sufficient quantities (Momberg, 1992). Myxococcus xanthus can be utilized for the biological formation of struvite. T his technique is useful in achieving a supersaturation ratio for the struvite formation other than stoichiometric proportions as considered in chemical precipitation. The cell membranes are useful for attracting magnesium, phosphate and ammonium ions on t hem so that supersaturation conditions for the struvite formation are achieved in the bulk around the cell dedris. This increases the specificity of the struvite formation also minimizing the chances of formation of other solids, so it increases the struv ite purity. The cell debris can be useful to achieve proper fluidization velocities in the reactor. Conditions under which biological struvite crystallization occur are similar to that of a chemical formation in terms of pH, temperature, supersaturation ra tio and so forth This researcher looks forward to achiev ing better, faster struvite crystallization using a biological technique. The following scope of the research is identified: 1. To s tudy induction time (nucleation) and crystal growth of st uvite using biological surfaces
155 2. To m odel struvite crystallization incorporating the surface energy of seed material, induction time and so forth 3. To s tudy bacterial surfaces forming struvite and characterize them plus contribut e knowledge to the medical field 4. To r un a continuous crystallizer on b iological formation of struvite 5. To s tudy different stoichiometric proportions of different ionic species i n the biological struvite formation 6. To d evelop a kinetic s model on the bacterial struvite formation
156 APPENDIX POLYMATH MODEL FOR STRUVITE P RECIPITATION Mgt = 50/24310 Nt = 740.88/14000 Pt = 106.7/31000 Alk = 4750 # in mg/L Alk_mole = Alk/100000 Cat = 0/40000 # in mg/L K = 0/39000 Ac =0 Cl = 0/35500 pH = 8.7 TIC = (2*Alk_mole 10^( 14+pH))/(1+2*10^(pH 10.3))+(10^(pH 10.3)) *(2*Alk_mole 10^( 14+pH))/(1+2*10^(pH 10.3)) # includes carbonate and bicarbonate alkalinity H = 10^( pH) OH = 1/100000000000000/H MHPd = 7/10*500000*Mg*H*PO4/0.5011000000e 12 # Molar Concentration MHPds = 120000*7/10*500000*Mg*H*PO4/0.5011000000e 12 # Molar Concentration CaHPO4_5H2O=(Ca*H*PO4)/10^ 18.995 # Molar Concentration MgCO3_3H2O=Mg*CO3/10^ 4.67 # Molar Concentration f(PO4) = Pt (PO4+H*PO4/0.5011000000e 12+H*H*PO4/0.5011000000e 12/0.6340000000e 7+H*H*H*PO4/0.5011000000e 12/0.6340000000e 7/ 0.7110000000e 2+500000*Mg*H*PO4/0.5011000000e 12+2.818489290*Mg*H*H*PO4/0.5011000000e 12/0.6340000000e 7+63131.31313*Mg*PO4+Mg*NH4*PO4/0.1995262315e 12+7/10*500000*Mg*H*PO4/0.5011000000e 12+2*0.9615000000e24*Mg^3*PO4^2+ 79000000000000000000000*Mg^3*PO4^2+2 *(Ca^3*(PO4)^2/ 10^ 25)+Ca*H*PO4/0.4688133821e 10+Ca*(H*PO4/0.5011000000e 12)/10^ 6.57+3*Ca^5*PO4^3*OH/10^ 36+6*Ca^8*(H*PO4/0.5011000000e 12)^2*(PO4)^4/1.06*10^ 47+CaHPO4_5H2O) PO4(0) = 0.0000000001 > 0 PO4(min) = 0 PO4(max) = 1 f(Mg) = Mgt (Mg*NH4*PO4/ 0.1995262315e 12+Mg+63131.31313*Mg*PO4+3*0.9615000000e24*Mg^3*PO4^2+3* 79000000000000000000000*Mg^3*PO4^2+500000*Mg*H*PO4/0.5011000000e 12+2.818489290*Mg*(H*(H*PO4/0.5011000000e 12)/0.6340000000e 7)+50000000000*Mg*OH^2+363.1082062*Mg*OH+7/10*500000*Mg*H*PO 4/0.5011000000
157 e 12+(Mg*CO3/10^ 5)+Ca*Mg*CO3^2/10^ 16.7+3*Ca*Mg^3*CO3^4/6.8*10^ 37+MgCO3_3H2O) Mg(0) = 0.000000000001 > 0 f(CO3) = TIC (CO3+H*CO3/0.4688133821e 10+(Mg*CO3/10^ 5)+2*Ca*Mg*CO3^2/10^ 16.7+Ca*CO3/10^ 5.3+4* Ca*Mg^3*CO3^4/6.8*10^ 37+MgCO3_3H2O) CO3(0) = 0.00000001 > 0 f(Ca) = Cat ((Ca*CO3/10^ 5.3)+(Ca*Mg*CO3^2/10^ 16.7)+5*Ca^5*PO4^3*OH/10^ 36+3*(Ca^3*(PO4)^2/ 10^ 25)+Ca*(H*PO4/0.5011000000e 12)/10^ 6.57+8*Ca^8*(H*PO4/0.5011000000e 12)^2*(PO4)^4/1.06*10^ 47+Ca*Mg^3*CO3^4/6.8*10^ 37+CaHPO4_5H2O) Ca(0) = 0.0000000000001 > 0 f(NH4) = Nt ( 0.5690000000e 9*NH4/H+NH4+Mg*NH4*PO4/0.1995262315e 12) NH4(0) = 0.001 > 0 Residual_N = (0.5690000000e 9*NH4/H+NH4)*14*1000 # Molar Concentration TotalP_soluble = 31000*H*H*PO4/0.5011000000e 12/0.6340000000e 7+31000*H*PO4/0.5011000000e 12+31000*PO4+31000*H*H*H*PO4/0.5011000000e 12/0.6340000000e 7/0.7110000000e 2+31000*2.818489290*Mg*H*H*PO4/0.5011000000e 12/0.6340000000e 7+31000*63131.31313*Mg*PO4+31000*MHPd # Mola r Concentration Mgsolution = 24000*Mg+24000*363.1082062*Mg*OH+24000*63131.31313*Mg*PO4+24000*2.818489290 *Mg*H*H*PO4/0.5011000000e 12/0.6340000000e 7+24000*7/10*500000*Mg*H*PO4/0.5011000000e 12 # Molar Concentration # All solid concentrations below are in mg/L Struvite = Mg*NH4*PO4/1.995E 13*137*1000 MP8s = 406000*0.9615000000e24*Mg^3*PO4^2 MP22s = 648000*79000000000000000000000*Mg^3*PO4^2 MgOH2s = 58000* 50000000000*Mg*OH^2 MHPs = 120000* 500000*Mg*H*PO4/0.5011000000e 12 MgCO3_s= 84310*Mg*CO3/10^ 5 CaCO 3_s = 100000*Ca*CO3/10^ 5.3 CaMgCO3_s = 184.31*1000*Ca*Mg*CO3^2/10^ 16.7 CaMg3CO3_s = 352.93*1000*Ca*Mg^3*CO3^4/6.8*10^ 37 Ca5PO4OH_s = 502*1000*Ca^5*PO4^3*OH/10^ 36 Ca3PO4_s = 310*1000*Ca^3*(PO4)^2/ 10^ 25 Ca8HPO4_H2O = 982*1000*Ca^8*(H*PO4/0.5011000000e 12)^2*(PO4)^4/1.06*10^ 47 CaHPO4_s = 136*1000*Ca*(H*PO4/0.5011000000e 12)/10^ 6.57 CaHPO4_5H2Os=226*1000*(Ca*H*PO4)/10^ 18.995 MgCO3_3H2Os=102310*Mg*CO3/10^ 4.67
158 Total_solids = Struvite+MP8s+MP22s+MgOH2s+MHPs+CaCO3_s+CaMgCO3_s + CaMg3CO3_s + Ca5PO4OH_s + C a3PO4_s + Ca8HPO4_H2O + CaHPO4_s+MgCO3_s+CaHPO4_5H2Os+MgCO3_3H2Os #in mg/L Struvite_purity= 100*Struvite/Total_solids #in % Na = 2*Mg+363.1082062*Mg*OH+2.818489290*Mg*H*H*PO4/(5.011*10^ 13)/(6.34*10^ 8)+K+H+NH4 63131.31313*Mg*PO4 Ac Cl OH H*H*PO4/0.50110 00000e 12/0.6340000000e 7 2*H*PO4/5.011*10^ 13 3*PO4 2*CO3 H*CO3/0.4688133821e 10 # Charge Balance Equation NaOH = 40000*Na # mg/L
159 LIST OF REFERENCES APHA, 1992. Standard Methods for the Examination of Water and Wastewater (18 th edition), APHA Press, Washington, DC. Agri Marketing E Newsletter, 2008. Potash Corp Issues. Outlook. January 25, 2008 Ali, M.I., Schneider, P.A., 2006. A fed batch design approach of struvite system on controlled supersaturation. Chem. Eng. Sci. 62 (12) 3951 3961. Alleman, J.E., Irvine, R.L., 1980. Nitrification in the sequencing batch biological reactor. J. Water Poll Contr Fed 52 (11), 2747 2754 Altinbas, M.J., Yangin, C.L., Ozturk, I.K., 2002. Struvite precipitation from anaerobically treated mun icipal and landfill wastewaters. Water Sci. & Technol. 46, 271 278. Arora, M.L., Barth, E.F., Umphres, M.B., 1985. Technology evaluation of sequencing batch reactors. J. Water Poll. Contr. Fed. 57 (8), 867 875. Baetens, D.K., 2000. Enhanced biological phos phorus removal: M odeling and experimental design. Dissertation Ghent University, Ghent, Belgium Burns, R.T., Moody, L.B., 2002. Phosphorus Recovery from animal manures using optimized struvite precipitation. Proceedings of coagulants and flocculants: Glo bal market and technical opportunities for water treatment chemicals, Chicago. Burns, R.T., Moody, L.B., Walker, F.R., Raman D.R., 2001. Laboratory and in site reductions of soluble phosphorus in swine waste slurries. Environ. Technol. 22 1273 1278 Battistoni, P.B., Boccadoro, R .K., Fatone, F.C., Pavan, P.V., 2005a. Auto nucleation and crystal growth of struvite in a demonstrative fluidised bed reactor (FBR). Environ. Technol. 26, 975 982. Battistoni, P., De Angelis, A., Prisciandaro, M., Boccadoro, R. & Bolzonella, D. (2002). P removal from anaerobic supernatants by struvite crystallization: long term validation and process modeling. Water Research, 36 1927 1938. Battistoni, P.B., Fava, G.F., Pavan, P.V., Musacco, A.K., Cecchi, F.T., 1997. Phosphat e removal in anaerobic liquors by struvite crystallization without addition of chemicals: Preliminary results. Water Res. 31, (11) 2925 2929. Battistoni, P.B., Paci, B.R., Fatone, F.C., Pavan, P.V., 2005b. Phosphorus removal from supernatants at low concen tration using packed and fluidi z ed bed reactors. Industrial and Eng. Chem. Res. 44, 6701 6707. Battistoni, P.B., Pavan, P.V., Cecchi, F.T., Mata Alvarez, J.C., 1998. Phosphate r emoval in r eal a naerobic s upernatants: Modeling and p erformance of a f luidized b ed r eactor. Water Sci. & Technol. 38, 275 283.
160 Beal, L.J., Burns, R.T., Stalder, K.J., 1999. Effect of anaerobic digestion on struvite production for nutrient removal from swine waste prior to land application. Presented at the 1999 ASAE Annual International Meeting. Paper No. 994042. ASAE, St. Joseph, MI Bhuiyana, M.I., Mavinica, D.S., Beckie, R.D., 2008. Nucleation and growth kinetics of struvite in a fluidized bed reactor. J. of Crystal Growth 310 1187 1194. Booker, N.A., Priestley, A.J., F raser I.H., 1999. Struvite formation in wastewater treatment plants: opportunities for nutrient recovery Environ. Technol. 20, 777 782. Borgerding, J., 1972 Phosphate deposits in digestion systems. J. Water Poll Contr Fed. 44, 813 81 9. Bridger, G.L., Salutsky, M.L., Starosika, R.S., 1961. Metal ammonium phosphates as fertilizers. 140 th Meeting of the American Chemical Society, 1 19, Chicago. Buchanan, J.R., 1993. Struvite c ontrol in f lush w ater r ecycle c omponents of l ivestock w aste. M aster of Science T hesis, Knoxville University of Tennessee. Burns, R.T., Moody, L.B., Walker, F.R., Raman, D.K., 2001. Laboratory and in situ reductions of soluble phosphorus in swine waste slurries. Environ. Technol. 22 (11) 1273 1278. Byung Su, K.C., Young Sang S.P., Hy un Ku, H.S., 1996. Controlled release of urea from rosin coated fertilizer particles. Industrial Eng. Chem. Res. 35 250 257. CEEP ), 2000. Recovery of phosphorus for recycling. CEFIC (European Chemical Industry Council). Polyphosphates ( CEEP ), 2003. http://www.nhm.ac.uk/mineralogy/phos/. (accessed Nov. 20 2007; verified Dec. 26 2007) elen, I., Buchanan, J.R., Burns, R.T., Robinson, R.B., Raman, D.R., 2007. Using a chemical equilibrium model to predict amendments requir ed to precipitate phosphorus as struvite in liquid swine manure. Water Res. 41 1689 1696. elen, I., Trker, M., 2001. Recovery of ammonia from anaerobic digester effluents. Environ. Technol. 22 (11), 1263 1272. Chen, C., 2009. Can the pilot BOT p roject provide a template for future projects? A case study of the Chengdu No. 6 Water Plant B Project. International J. of Project Management 27, 573 583. de Bashan, L.E., Bashan, Y., 2004. Recent advances in removing phosphorus from wastewater and its us e as fertilizer (1997 2003), Water Res. 38, 4222 4246. 21 October 33.
161 Donnert, D., Salecker, M., 1999. Elimination of p hosphorus from m unicipal and i ndustrial w aste w ater. Water Sci. & Tech nol. 40 ( 4 5 ), 195 202 Doyle, J.D., Parsons, S.A., 2002. Struvite formation, control and recovery. Water Res. 6 3925 3940. Doyle, J.D., Parsons, S.A., 2004. Struvite scale formation and control. Water Sci. & Technol. 49 177 182. Durrant, A.E., Scrimshaw, M.D., Stratful, I., Lester, J.N., 1999. Review of the feasibility of recovering phosphate from wastewater for use as a raw material by the phosphate industry. Environ. Technol. 20 749 758 El Refaie, K., Sakran, A.A., 1996. Controlled release f ormulation of agrochemicals from calcium alginate. Industrial Eng. Chem. Res. 35 (1996) 3726 3729. Gaterell, M.R., Gay, R., Wilson, R., Gochin, R.J., Lester, J.N., 2000. An economic and e nviron. evaluation of the opportunities for substituting phosphorus recovered from wastewater treatment works in existing UK fertiliser markets. Environ. Technol. 21, 1067 1084. Giesen, A., 1999. Crystallisation process enables Environ. friendly phosphate removal at low cost. Environ. Technol. 20, 769 775. Hao, X.D., van Loosdrecht, M.C.M., 2003. A proposed sustainable BNR plant with the emphasis on recovery of COD and phosphate. Water Sci. & Technol. 48 (1), 77 85 Harada, H.S., Shimizu, Y., Miyagoshi, Y.Y., Matsui, S., Matsuda, T., Nagasaka, T.H., 2006. Predicting struv ite formation for phosphorus recovery from human urine using an equilibrium model. Water Sci. & Technol. 54, 247 255. Heinzmann, B.S., 2004. Phosphorus recycling in sewage treatment plants with biological phosphorus removal. In: Proceedings of the International Conference on Struvite: I ts role in phosphorus recovery and reuse, Cranfield England Helmreich, B., Schreff, D., Wilderer, P.A., 2000. Full scale experiences with small sequencing batch reactor plants in Bavaria. Water Sci. & Technol. 41 ( 1 ), 89 96 Integrated Crop Management ( ICM ) 2000. Newsletter article. Soil testing and available phosphorus September 18, 2000 164 166 Irvine, R.L., Ketchum L. H., Breytogle, R., Barth, E.F., 1983. Municipal application of sequencing batch treatment. J. WPCF 55 (5), 484 488 Jaffer, Y., Clark, T.A., Pearce, P., Parsons, S.A., 2002. Potential phosphorus recovery by struvite formation. Water Res. 36 (7) 1834 1842.
162 Jaffer, Y., Pearce, P., 2004. Phosphorus recovery via struvite production at Slough sewage treatment works, UK a case study. In: Phosphorus in Environ. T echnologies. Principles and A pplications Valsami Jones E. (Ed.), pp. 402 427, IWA P ublishing, London. Johnson, A.S., Hammond, R.S., Lewington, M.K., Ratnam, A.P., 2004. Struvite potential as an ingredient in horticultural fertiliser. Struvite: Its role in phosphorus recovery and reuse Cranfield, England Johnston, A.E., Richards, I.R., 2003. Effectiveness of different precipitated phosphates as phosphorus sources fo r plants. Soil Use and Management 19, 45 49. Ketchum Jr., L.H., 1997. Design and physical features of s equencing b atch r eactors. Water Sci. & Technol. 35 (1), 11 18. Laridi, R., Auclair, J.C., Benmoussa, H., 2005. Laboratory and p ilot scale phosphate and ammonium removal by controlled struvite precipitation following coagulation and flocculation of swine wastewater. Environ. Technol. 26, 525 536. Lee, J.J., Choi, C.U., Lee, M.J., Chung, I.H., Kim, D.S., 2004. A study of NH3 N and P refixation by struvite formation in hybrid anaerobic reactor. Water Sci. & Technol. 49 (5 6) 207 214. Le Corre, K.S., Valsami Jones E., Hobbs, P., Parsons, S.A., 2005. Impact of calcium on struvite crystal size, shape and purity. J. of Crystal Growth 2 83 (3 4) 514 522. Le Corre, K.S., Valsami Jon es, E., Hobbs, P., Parsons, S.A., 2006. Impact of reactor operation on the success of struvite precipitation from synthetic liquors. Environ. Technol. 28 (11), 1245 1256. Le Corre, K.S., Valsami Jones, E., Hob bs, P., Jefferson B., Parsons, S.A., 2007. Agglomeration of struvite crystals. Water Res. 41 ( 2 ) 419 425 Li, X.Z., Zhao, Q.L., 2003. Recovery of ammonium nitrogen from landfill leachate as a multi nutrient fertilizer. Ecological Eng. 20 171 181. Liberti, L., Limoni, N., Lopez, A., Passino, R., Boari, G., 1986. The 10 m 3 h 1 Rim Nut demonstration plant at W est Bari for removing and recovering n itrogen and p hosphorus from wastewater. Water Res. 6, 735 739. Liberti, L., Petruzelli, D., De Florio, L., 2001. REM NUT ion exchange plus struvite precipitation process. Environ. Technol. 22 1313 1324. Loewenthal, R.E., Kornmller, U.R.C., van Heerden E.P., 1994. Modeling struvite precipitation in anaerobic treatment systems. Water Sci. & Technol. 30, 107 116. Mac, S.J., Mata Alvarez, J., 2002. Utilization of SBR Technol. for wastewater treatment: An overview Industrial Eng. Chem. Res. 41 5539 5553.
163 Mamais, D., Pitt, P.A., Cheng, Y.W., Loiacono, J., Jenkins, D., 1994. Determination of ferric chloride dose to control struvite precipitation in anaerobic sludge digesters. Water Environ. Res. 66, 912 918 Mangin, D., 2004. Struvite crystallisation reaction in a semi continuous stirred reactor, Struvite: Its role in phosphorus recovery and reuse Cranfield, England. Metcalf and Eddy, 2003 Wastewater e ng. : Treatment and r euse McGraw Hill, New York Miles, A., Ellis, T.G., 2001. Struvite precipitation potential for nutrient recovery from anaerobically treated wastes. Water Sci. & Technol. 43 (11) 259 266. Momberg, H., 1992. Phosphate fixation in waste water by means of controlled struvite formation. WRC Report No 250 /1/93 Division of W ater Technol. CSIR. Momberg, G.A., Oellermann, R.A., 1992. The removal of phosphate by hydroxyapatite and struvite crysta llisation in South Africa Water Sci. & Technol. 26 (5 6) 987 996. Moon, Y.H., Kim, J.G., Ahn, J.S., Lee, G.H., Moon, H., 2007. Phosphate removal using sludge J. of H azardous M aterials 143 (1 2), 41 48. Morse, G.K., Brett, S .W., Guy, J.A., Lester J.N., 1998. Review: phosphorus removal and recovery technologies. The Sci. of Total Environment 212 69 81. Mulkerrins, D., Dobson, A.D.W., Colleran, E., 2004. Parameters affecting biological phosphate removal from wastewaters. Envir onment International 30, 249 259. Munch, E.V., Barr, K., 2001. Controlled Struvite Crystallization for Removing Phosphorus from Anaerobic Digester Sidestreams. Water Res. 35, 151 159. Norcross, K.L., 1992. Sequencing batch reactors; an overview, Water Sci. & Technol. 26 (9 11), 2523 2526. Novak, O., Lindtner, S., 2004. Investment and operating cost of small WWTPs compared to larger plants. In: 6 th Specialist Conference on Small Water and Wastewater Systems. Frementle Western Australia Australia, 11 13 Okada, M., Sudo, R., 1985. Simultaneous removal of phosphorus and nitrogen by sequencing batch reactor activated sludge process. Water Sci. & Technol. 17 (11 120 ) 315 316. Ohlinger, K.N., Young, T.M., Schroeder, E.D., 1998. Predicting struvite formation in digestion. Water Res. 32 3607 3614. Ohlinger, K.N., Young, T.M., Schroeder, E.D., 1999. Kinetics effects on preferential struvite accumulation in wastewater. J. of Environ. Eng. 125 730 737. Ohlinger, K.N., Young, T.M., Schroeder, E.D., 2000. Post digestion struvite precipitation using a fluidized bed reactor. J. of Environ. Eng. 126, 361 368.
164 Reid, R.L., Horvath, D.J., 1980. Soil c hem. and mineral problems in farm livestock, A review, Animal Feed Sci. and Technol. 95 167 Robinson, H., 1989. On th e formation of struvite by microorganisms. Proc. Cambr. Phil. Soc. 6, 360 362. Parsons, S.A., Berry, T.A., 2004. Chemical phosphorus removal. In: Phosphorus in Environ. Technologies, P rinciples and A pplications Valsami Jones, E. (Ed) 260 272 Parsons, S.A., Doyle, J.D., 2004. Struvite scale formation and control. Water Sci. & Technol. 49, 177 182. Pollution Preventi on and Abatement Handbook, 1998 Toward s Cleaner Production. Puig, S., Corominas, L I., Vives, M.T., Colomer, J., Balaguer, M. D., Colprim, J., 2005. Development and implementation of a real time control for nitrogen removal using OUR and ORP as endpoints. Ind. Eng. Chem. Res. 44 (9), 3367 3373. Rim Y.T., Yang H.J., Yoon C.H., Kim Y.S., Seo J.B., Ryu J.K., Shin E.B., 1997. A full scale test of a biological nutrients removal system using the sequencing batch reactor activated sludge process. Water Sci. & Technol. 35 (1), 241 247. Salman, O.A., 1989. Polythylene coated urea. l. Improved storage and handling properties, Ind Eng. Chem. Res. 28 (5), 630 632 Salman, O.A., Hovakeemian, J., Khraishi, N., 1989. Polyethylene coated urea. 2. Urea release as affected by coating material, soil type and temperature. Industrial Eng. Chem. Res. 28 (5), 633 638 Sarkar, A.K., 1990. Phosphate cement based fa st setting binders. J. of American Ceramics Society Bulletin 69 234 238. Sawyer, C.N., McCarty P.L., Parkin, G. F., 1994. Chem. for Environ. Eng. McGraw Hill New York Schuiling, R.D., Andrade, A., 1999. Recovery of struvite from calf manure. Environ. Technol. 20 765 768. Scott, W. D., Wrigley, T. J. & Webb K. M. (1991). A computer model of struvite solution chemistry. Talanta 38, 889 895. Seco, A., Pastor, L., Barat, R., Ferrer, J., Mangin, D., 2004. Phosphate recovery as struvite: factors influencin g the process. In: Proceedings of The International Conference on Struvite: I ts role in phosphorus recovery and reuse, Cranfield England. Sharp, E., 2002. Phosphorus recovery from centrifuge liquors using magnesium hydroxide for controlled struvite precip itation. M aster of Science Thesis, Cranfield University, England.
165 Shimamura, K., Tanak a, T., Miura, Y., Ishikawa, H., 2003. Development of a high efficiency phosphorus recovery method using a fluidized bed crystallized phosphorus removal system. Water Sci. & Technol. 48 163 170. Shu, L., Schneider, P., Jegatheesan, V., Johnson, J., 2006. An economic evaluation of phosphorus recovery as struvite from digester supernatant, Bioresource Technol. 97 (17) 2211 2216 Sotirakou, E., Kladitis, G., Di amantis, N., Grigoropoulou, H., 1999 Ammonia and phosphorus removal in municipal wastewater treatment plant with extended aeration. Southern California Coastal Water Res. Project (SCCWRP) 1992. Characteristics of effluents from large municipal wastewater treatment plants in 1990 and 1991 pp .8 18, In: Southern California Coastal Water Res. Project, Annual Report 1990 91 and 1991 92. Southern California Coastal Water Res. Project, Long Beach Southern Cooperative Series, 1998. Soil testing for P Environ. uses and implications. S. Coop. Ser. Bull. 389. USDA CSREE, U.S. Gov. Print. Office, Washington, DC. Steen, I., 1998. Phosphorus availability in the 21th century management of a non renewable resource. In: Phosphorus & Potassium, Issue No: 217 (September October, 1998) Sten, I., 2004 Phosphorus recovery in the context of industrial use. In: Phosphorus in Environ. Technologies, P rinciples and A pplications Valsami Jones, E. (Ed) 339 354 Steinmetz, H., Wiese, J., Schmitt, T.G., 2002. Efficiency of SBR Technol. in municipal wastewater treatment plants. Water Sci. & Technol. 46 (4 5), 293 299. Stratful, I., Scrimshaw, M.D., Lester, N.J., 2001. Conditions influencing the precipitation of magnesium ammonium phosphate. Water Res. 35 4191 4199. Stratful, I., Scrimshaw, M.D., Lester, N.J ., 2004. Removal of struvite to prevent problems associated with its accumulation in wastewater treatment works. Water Environ. Res. 76, 437 4199. Suzuki, K., Tanaka, Y., Kuroda, K., Hanajima, D., Fukumoto, Y., 2005. Recovery of phosphorus from swine wastewater through crystallization. Bioresource Technol. 96 1544 1550. Taylor, A.W., Frazier, A.W., Gurney, E.L., 1963. Solubility products of magnesium ammonium and magnesium potassium phosphates. Transactions of the Faraday Soci ety 59, 1580 1584. Tebbutt, T.H.Y., 1998. Principles of water quality control. Butterworth Heinemann, Oxford, England Tefertiller, K.R., Jauregui, C., Olexa, M.T., 1998. Impact of the regulatory environment facing Florida dairy farmers. Circular 1208. Fl orida Cooperative Extension Service, University of Florida, Gainesville, FL
166 Teichgrber, B., Screff, D., Ekkerlein, C., Wilderer, P.A., 2001. SBR Technol. in Germany : A n overview. Water Sci. & Technol. 43 (3), 323 330 Thomas, D 2007. Maximising Struvite Recovery from Real Wastewater Sources. Cranfield University School of Appl. Sci. Department of Sustainable Systems Centre for Water Sci. M aster of Science by Research. Torrijos, M., Vuitton, V., Moletta, R., 2001. The SBR process: A n efficient and economic solution for the treatment of wastewater at small cheesemaking dairies in the Jura mountains. Water Sci. & Technol. 43 (3), 373 380 Tnay, O., Kabdasli, I., Orhon, D., Kolak, S., 1997. Ammonia removal by magnesium ammonium phosphate precipitati on in industrial wastewaters. Water Sci. & Technol. 36, 225 228. Von Mnch, E. Barr, K. (2001) Controlled crystallisation for removing phosphorus from anaerobic digester side stream. Water Res. 35, 151 159. Ueno, Y., Fujii, M., 2001. Three years experience of operating and selling recovered struvite from full scale plant. Environ. Technol. 22, 1373 1381. Urban Wa ste Water Treatment Directive (UWWTD), 1991 Council of t he European communities. Council directive of 21 M ay 1991 concerni ng urban waste water treatment (91/271/EEC). Official J. of the Europea Community Series L, 135/40 Van Dijk, J.C., Wilms, D.A., 1991. Water treatment without waste material fundamentals and state of the art pellet softening J. of Water SRT Aqua 40, 263 280. Vives, M.T., Balaguer, M.D., Garca, S., Garca R., Colprim, J., 2003. Textile dyeing wastewater treatment in a sequencing batch reactor system. J. of Environ. Sci. and Health part (A), Toxicological Hazard Substance Environ. Eng. 38 (10), 2089 99. Wang, J.S., Song, Y.H., Yuan, P., Peng J.F. & Fan M.H. (2006). Modeling the crystallization of magnesium ammonium phosphate for phosphorus recovery Chemosphere, 65, 1182 1187. Webb, K.M., 1988. The solubility of struvite and its application to a piggery effluent problem. Bachelor thesis, School of Biological and Environ. Sci. Murdoch University, Perth, Western Australia, Australia. Webb, K.M., Ho, G.E., 1992. Struvite (MgNH 4 PO 4 .6H 2 O) solubility and its application to a piggery effluent problem. Water Sci. & Technol. 26 (9 11) 2229 2232. Williams, S., 1999. Struvite precipitation in the sludge treatment stream at Slough wastewater treatment plant and opportunities for phosphorus recovery. Environ. Technol. 20, 743 747. Wilsenach, J.A., Schuurbiers, C.A.H., van Loosdrecht, M.C.M., 2007. Phosphate and potassium recovery from source separated urine through struvite precipitation. Water Res. 41, 458 466.
167 Grace W.R., Co., 1969. Dearborn Chemical Division, Ultimate disposal of phosphate from wastewater by recove ry as fert ilizer phase I final report. SCOPE Newsletter, 38, 2. Wrigley, T. J. (1999). Water quality improvement of piggery effluent Ph.D. Thesis, Division of Science and Engineering, Murdoch University Perth Yoshino, M., Yao, M., Tsuno, H., Somiya, I., 2003. Removal and recovery of phosphate and ammonium as struvite from supernatant in anaerobic digestion. Water Sci. & Technol. 48, 171 178. Zhang, S., Chonghua Y., Xiaoxi, F., Min Y ., 2 004. Repeated use of MgNH 4 PO 4 .6H 2 0 residues for ammonium removal by a cid dipping. Desalination 170, 27 32
168 BIOGRAPHICAL SKETCH Sachin Gadekar was born in Maharashtra ( India ) He received his Bachelor of Technology in chemical e ngineering from Laxminarayan Institute of Technology, Nagpur (India). Gadekar is a chemical and bioprocess engineer with understanding and practical knowledge of industrial processes in biofuels, waste to energy, and nutrient recovery. He is professionally trained and experienced working in U S India and Canada. He has p rove n ability to use applied research and problem solving skills to support personal and professional development. In October 2005, he came to University of Florida to pursue his Ph.D. degree in the Agricultural and Biological Engineering Department under the supervision of Dr. Pratap Pullammanappallil. He has received prestigious fellowships and awards for development of sustainable technologies. Till date, Gadekar has more than 9 years of experience working in the chemical and biological engineering fields. H e has worked on multiple international projects that bring process and product development expertise. He has given several seminars and technical presentations in national and international conferences. He has written several research articles, and grant p roposals.