xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200397Ka 4500
controlfield tag 001 002068294
007 cr mnu|||uuuuu
008 100412s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003127
Opportunities for nutrient recovery in post digestion sludge handling :
b analysis and feasibility study using municipal scale aerobic and anaerobic digesters
h [electronic resource] /
by David Starman.
[Tampa, Fla] :
University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 142 pages.
Thesis (M.S.C.E.)--University of South Florida, 2009.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
ABSTRACT: The wastewater treatment process has developed with the primary goals of protecting receiving water ecosystems and human health. Over time, there have been continuous innovations in process efficiencies, energy recovery, and nutrient removal. Wastewater offers opportunity for recovery of resources of various economic values, and recent research aims at process innovation to optimize resource recovery while still achieving the primary goals of the treatment process. The objective of this study is to assess the logistical and economic feasibility of recovery of nitrogen and phosphorus at two municipal treatment plants in the Tampa Bay area, one employing aerobic digestion and the other anaerobic digestion. The study is conducted using literature review of applicable processes, mass balance on the fate of nutrients (N and P) through the treatment plants and special attention to sludge handling. Based on the whole-plant mass balance conducted at the facilities, it is estimated that over 80% of the nutrient influent is routed to the solids handling side of the plant, warranting special attention to this area for nutrient recovery. Sludge digested through anaerobic and anaerobic processes have distinctly different characteristics and opportunities for resource recovery are specific to each process. Mass balances for nitrogen in the anaerobic digestion process show a high concentration of dissolved ammonia. The feasibility of struvite precipitation by addition of phosphate and magnesium compounds is evaluated through batch reaction using anaerobic sludge filtrate. Aerobic sludge contains most of the nutrient resources in the solid phase, ready for recovery if handled properly. Phosphorus release is a potential concern and specific phosphorus release rates are evaluated for a municipal scale aerobic digester.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Advisor: Daniel Yeh, Ph.D.
x Civil Engineering
t USF Electronic Theses and Dissertations.
Opportunities for Nutrient Recovery from Post-Digesti on Sludge Handling: Analysis and Feasibility Study Using Municipal Scale Aerobic and Anaerobic Digesters by David Starman A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Civil Engineering Department of Civil and Environmental Engineering College of Engineering University of South Florida Major Professor: Daniel Yeh, Ph.D. Jeffrey Cunningham, Ph.D. John Wolan, Ph.D. Date of Approval: June 23, 2009 Keywords: wastewater, ammonia re covery, phosphate recovery, phosphate release, struvite Copyright 2009, David Starman
i Table of Contents List of T ables ........................................................................................................ iii List of Fi gures .......................................................................................................iv Abstract ................................................................................................................vi 1Introduction .................................................................................................... 1 2Background ................................................................................................... 32.1Nitrogen and Phosphorus Resour ces Â– Discussion of Value ................. 32.1.1Nitrogen .......................................................................................... 32.1.2Phosphor us ..................................................................................... 52.1.3Summary of Res ource Va lues......................................................... 72.2Resource Partitioning in the Treatment Plant ......................................... 82.3Fate of Nutrients in Digester Nit rogen ................................................ 112.3.1Aerobic Diges tion .......................................................................... 132.3.2Anaerobic Di gestion ...................................................................... 162.4Fate of Nutrients in Digester Â– Phosphorus .......................................... 192.5Current Nitrogen Remova l Technol ogy ................................................ 212.5.1Biological Nitr ogen Remova l ......................................................... 212.5.2BNR Nitrification/ Denitrificat ion Efficiency Im provements ............ 232.5.3Struvite .......................................................................................... 242.6Current Phosphorus Removal Technology ........................................... 252.6.1Metal Salt Prec ipitatio n ................................................................. 252.6.2Biological Phosphorus Removal (BPR) ......................................... 262.6.3Struvite .......................................................................................... 28 3Study Treatment Plants: Pr eliminary Eval uation ........................................ 333.1Process Descriptions ............................................................................ 343.1.1Howard Curren Treat ment Pl ant ................................................... 343.1.2Largo ............................................................................................. 363.2Further Plant Specific Background Â–LG AWTP ..................................... 383.2.1Potential for Phosphorus Release from Sludg e ............................ 383.2.2Rate of Phosphor us Rel ease ........................................................ 393.2.3Correlation Between P-Releas e and P-Uptake Rates ................... 423.2.4Phosphate Release Duri ng Aerobic Dig estion .............................. 433.2.5Summary of Literature Review on Phosphorus Release ............... 43
ii 3.3Further Plant Specific Background Â– HFCAWT P ................................. 443.3.1Struvite Potential for Ammonia Reco very ...................................... 443.3.2Estimates of Struvite Recovery Costs ........................................... 463.3.3Summary of Struvite Potent ial Literature Review .......................... 47 4Rationale and Outline of A ssessment and Experimental Plan ..................... 494.1Evaluation of Resource Content in Solids vs. Liquids .......................... 50 4.2Mass Balance: Nitrogen and Ph osphorus in Di gesters ....................... 514.3Evaluation of N and P Recovery by Struvite ......................................... 534.4Financial Analysis of St ruvite Produc tion ............................................. 544.5Aeration for Struvite pH Adjust ment ..................................................... 544.6Evaluation of Specific Phosphorus Release Rate Â– LGAWTP ............. 55 5Materials and Methods ................................................................................ 565.1Analytical Methods ............................................................................... 565.2Experimental Methods .......................................................................... 585.2.1Preparation of Struvite by Batch R eaction .................................... 585.2.2Phosphorus Release and Sludge Se ttling ..................................... 58 6Results and Di scussion ............................................................................... 606.1General ................................................................................................ 606.2Resource Partitioning Between So lid and Liquid Streams ................... 606.3Fate of Resources in Aerobi c and Anaerobic Digesters ....................... 616.3.1Nitrogen ........................................................................................ 676.3.2Phosphor us ................................................................................... 696.4Whole Plant Mass Balancin g ................................................................ 696.5Phosphorus Release and Slu dge Settling vs. Time ............................. 716.6Struvite Precipitation from AnD F iltrate: Phosphorus Recovery .......... 816.7Evaluation of Struvite Production Us ing Aeration as pH Control .......... 836.8Financial Analysis of Struvite ............................................................... 95 7Conclusi ons ............................................................................................... 102 Referenc es ....................................................................................................... 105 Appendice s ....................................................................................................... 108Appendix A-1: Mass Ba lance Wor ksheets ................................................... 109Appendix B-1: Struvite Batch Reacti on Plots ............................................... 114Appendix C-1: Financial Analysis Wor ksheets ............................................. 129Appendix D-1: Calib ration Curv es ................................................................ 131Appendix E-1: Bioenergetic St oichiometry De terminati on ............................ 135Appendix E-2: Partitioning Based on Stoich iometry ..................................... 138Appendix E-3: Mass Trans fer Calculat ions .................................................. 140
iii List of Tables Table 1: Summary of Resour ce Partitioni ng Results ......................................... 61Table 2: Cumulative Analytical Data for Anaerobic Digester at HFCAWTP ....... 63Table 3: Cumulative Analytical Data for Aerobic Digeste r at LGAWTP ............. 64Table 4: Summary of Phos phorus Releas e Trials ............................................. 78Table 5: Analysis of Phos phate Release Si gnificance ....................................... 78Table 6: Summary of Phos phate Removal Results ........................................... 84Table 7: Summary of Struvite Precipitation Analysis ......................................... 89Table 8: Input Parameters for Financial Analysis ............................................ 100Table 9: Parameters for Mass Transfer Co mparison ....................................... 142
iv List of Figures Figure 1: Nitrogen C ycle in Na ture ...................................................................... 4Figure 2: Concept of Liquids and Solids Treatm ent Trains .................................. 9Figure 3: Theoretical Aerobic Degradation of Pr imary Sludge ........................... 14Figure 4: Theoretical Aer obic Degradation of WAS ........................................... 15Figure 5: Theoretical Anaerobic Degradation of Primary Sludge ....................... 18Figure 6: Theoretical Anaer obic Degradatio n of W AS ....................................... 18Figure 7: Nitrogen Cycle in WWTP .................................................................... 22Figure 8: Nitrogen Cycle Shor tcut in Enhan ced BN R ........................................ 24Figure 9: Struvite Conditional Solubility Curve ................................................... 30Figure 10: Howard F. Curren Advanc ed Wastewater Trea tment Plant .............. 34Figure 11: Largo Advanced Wast ewater Treatme nt Plan t ................................. 37Figure 12: Phosphorus Release Curve .............................................................. 42Figure 13: Circumnavigat e the Nitrogen Cycle .................................................. 45Figure 14: Nitrogen Balance at Anaerobic Digester ........................................... 66Figure 15: Phosphate Balanc e at Anaerobic Digeste r ....................................... 66Figure 16: Nitrogen Balance at Aerobic Digester ............................................... 68Figure 17: Phosphate Balanc e at Aerobic Digester ........................................... 68Figure 18: Nutrient Mass Bal ance for Treatm ent Plant s .................................... 72Figure 19: Trial 1 Phosphorus Release at Largo AWTP .................................... 73Figure 20: Trial 2 Phosphorus Release at Largo AWTP .................................... 73Figure 21: Trial 2 DO A nalysis at Lar go AWTP ................................................. 74Figure 22: Trial 3 Phosphorus Release at Largo AWTP .................................... 74Figure 23: Trial 3 DO A nalysis at Lar go AWTP ................................................. 75Figure 24: Trial 4 Phosphorus Release at Largo AWTP .................................... 75Figure 25: Trial 4 DO A nalysis at Lar go AWTP ................................................. 76Figure 26: Trial 5 Phosphorus Release at Largo AWTP .................................... 76Figure 27: Trial 6 Phosphorus Release at Largo AWTP .................................... 77Figure 28: Phosphate Concentrations During Struvite Pr ecipitation .................. 85Figure 29: SEM Image of Crystallized Pr oduct .................................................. 85Figure 30: Solid Product Analysis by EDS (courtesy of Ru ssel Ferlit a) ............. 86Figure 31: Prediction of Mass Transfe r for Carbon Dioxide and Ammonia ........ 86Figure 32: Acidity Analysis fo r Struvite Aera tion Test s ...................................... 90Figure 33: Alkaline Addition and A mmonia Reduction Analysis ........................ 90Figure 34: Titration Curves fo r Struvite Aera tion Test s ...................................... 92Figure 35: Analysis of Increasing Commodi ty Prices on Struvite Feasibility ...... 96Figure 36: Assessment of Struvite Production Costs ......................................... 97Figure 37: Assessment of pH C hemical Reducti on Effect s ............................... 98Figure 38: Rising Commodity Pr ices with Reduced pH Cost ............................. 99
v Figure 39: Resulting Financial Analysis from th is Study .................................. 101Figure 40: Struvite #14 pH and Ammonium vs. Time ...................................... 114Figure 41: Struvite #14 Ammonium vs. pH ...................................................... 115Figure 42: Struvite #16 pH and Ammonium vs. Time ...................................... 116Figure 43: Struvite #16 Ammonium vs. pH ...................................................... 117Figure 44: Struvite #17 pH and Ammonium vs. Time ...................................... 118Figure 45: Struvite #17 Ammonium vs. pH ...................................................... 119Figure 46: Struvite #18 pH and Ammonium vs. Time ...................................... 120Figure 47: Struvite #18 Ammonium vs. pH ...................................................... 121Figure 48: Struvite #19 pH and Ammonium vs. Time ...................................... 122Figure 49: Struvite #19 Ammonium vs. pH ...................................................... 123Figure 50: Struvite #20 pH and Ammonium vs. Time ...................................... 124Figure 51: Struvite #20 Ammonium vs. pH ...................................................... 125Figure 52: Struvite #21 pH and Ammonium vs. Time ...................................... 126Figure 53: Struvite #21 Ammonium vs. pH ...................................................... 127Figure 54: Struvite #22 pH and Ammonium vs. Time ...................................... 128Figure 55: Typical Calibration Curve for Total N by TO C-V ............................. 131Figure 56: Typical Calibrati on Curve for Amm onia Prob e ................................ 132Figure 57: Typical Calibration Curve for Total Phosphat e ............................... 133Figure 58: Typical Calibrati on Curve for Orth o-Phosphat e .............................. 134
vi Opportunities for Nutrient Recovery fr om Post-Digestion Sludge Handling: Analysis and Feasibility Study Us ing Municipal Scale Aerobic and Anaerobic Digesters David Starman ABSTRACT The wastewater treatment process has developed with the primary goals of protecting receiving water ecosystems and human health. Over time, there have been continuous innovations in process efficiencies, energy recovery, and nutrient removal. Wastewat er offers opportunity for re covery of resources of various economic values, and recent res earch aims at process innovation to optimize resource recovery while st ill achieving the primary goals of the treatment process. The objective of this study is to assess the logistical and economic feasibility of recovery of nitrogen and phosphorus at two municipal treatment plants in the Tampa Bay ar ea, one employing aerobic digestion and the other anaerobic digestion. The study is conducted using literature review of applicable processes, mass balance on the fa te of nutrients (N and P) through the treatment plants and special attention to sludge handling.
vii Based on the whole-plant mass balance conduc ted at the facilities, it is estimated that over 80% of the nutrient influent is routed to the solids handling side of the plant, warranting special attention to th is area for nutrient recovery. Sludge digested through anaerobic and anaerobic proc esses have distinctly different characteristics and opportunities for res ource recovery are specific to each process. Mass balances for nitrogen in the anaerobic digestion process show a high concentration of dissolved ammonia. The feasibility of struvite precipitation by addition of phosphate and magnesium co mpounds is evaluated through batch reaction using anaerobic sludge filtrate. Aerobic sludge cont ains most of the nutrient resources in the solid phase, r eady for recovery if handled properly. Phosphorus release is a potential conc ern and specific phosphorus release rates are evaluated for a municipal scale aerobic digester.
1 1 Introduction The term sewer comes from the English Â“seawardÂ”. Intercepting this seaward flow for public health is the original impetus behind the development of wastewater treatment Protection of rivers, lakes and the sea from eutrophication has brought about another level of wast ewater treatment regulations and technology. The early history of wastew ater treatment was a development of technology to accomplish removal: re moval of oxygen demand, removal of nutrients which can cause eutrophicati on, and removal of pathogens. It has been recognized that resources of va lue exist in wastewater. The following is excerpted from a sewage treatment text book from the 1950Â’s: Â” It is true that there are recoverable constituents in sewa ge, but, like the extraction of gold from seawater, the process of recovery is more costly than the value of the recovered constituents.Â” (Babbitt 1953) Recovery of resources has been steadily growing in the wastewater treatment industry. Biosolids are land applied and reclaimed water is piped throughout many municipalities. Methane recove ry for energy production is a common practice at anaerobic digestion facilit ies throughout the developed world.
2 Most Â“recoveryÂ” efforts result from conv enient byproducts of the removal process, and are not the focus of technology develop ment. But, with rising energy costs, depletion of mineral reserves, increas ing fertilizer co sts, and increasing population stress on resources, is Babbitt still correct? Are focused efforts to recover resources such as energy, nitrogen, and phosphorus now becoming worth the investment? Are we on the cusp of a paradigm switch where recovery of resources from wastewater makes sense? This thesis investigates the value of nitrogen and phosphorus as resources in wastewater and the feasibility of recovery.
3 2 Background 2.1 Nitrogen and Phosphorus Resources Â– Discussion of Value 2.1.1 Nitrogen Nitrogen is the most abundant element in the earthÂ’s atmosphere and is an element crucial to many biological processes. Nitrogen cycles from the atmosphere into biota, journeying through the biosphere as an essential nutrient passing between living systems and inorgan ic forms and eventually back to the atmosphere in its elemental form, nitrogen gas. The Â“nitrogen cycleÂ” is studied in a beginning biology curriculum and is mo st easily underst ood with a graphic representation of the various forms of nitrogen and how they transfer from one form to another.
4 NO2 N2 NH3 NO2 NO3 Nitrification Â– Ammonia Oxidation Nitrogen Fixation Â– (microbial or synthetic) Denitrification Â– Nitrite Reduction Denitrification Nitrate Reduction Nitrification Â–Nitrite Oxidation Figure 1: Nitrogen Cycle in Nature Ammonia is the primary ni trogen compound used to make fe rtilizer for the worldÂ’s agriculture. The world makes its ammoni a and drives its agricultural economy by the energy intensive Haber Bosch pr ocess, producing anhydrous ammonia. Under sufficient temperatur e, methane in natural gas will reform into hydrogen and carbon monoxide. The carbon monoxide is separated and the hydrogen reacts with atmospheric nitrogen to form ammonia. This process uses large quantities of natural gas. The natural gas is used as the fuel to create high temperature and pressure and also as the feedstock for hydrogen. Production of nitrogen fertilizer constitu tes approximately one percent of global energy expenditures (Worrell et al. 2000). T he US is a net importer of ammonia, mostly from countries with abundant natur al gas such as Trinidad and Canada (Worrell et al. 2000).
5 2.1.2 Phosphorus Like nitrogen, phosphorus is a crucial el ement to biological processes. It is involved in the cellular energy currency, ATP, and is part of the polymer bonding of the backbone of DNA. However, unlike nitrogen, phosphorus is in limited supply as a resource to the biota. The only source of phosphorus is the weathering of phosphorus-containing rocks. Prior to mining of phosphate rock agriculturists made use of many strat egies to conserve this resource, understanding that it was a limiting factor to agricul tural productivity. Thus plants, animals and man were in competit ion for this limited resource and those that could locate, store and recycle phosphor us were successful (Driver et al. 1999). Modern man has enjoyed a relative abundance of phosphorus by mining operations which accelerate the natural weathering and release of available phosphorus to the biosphere. The following is quoted from Driver et al. (1999) and summarizes the abuse of the phosphorus resource by our society. Â“Only modern man, far removed from t he process of primary production, has forgotten the importance of conserving and re-using this precious resource.Â” Mining efficiency and the resulting pr ice of the phosphate product depends on the ease of access to the rock and t he quality or percentage of phosphate
6 available in the rock. T here are limited reserves of phosphate around the world. Florida and Idaho represent the bulk of the USAÂ’s production and the US, China and Morocco represent approximately two Â–thirds of the worldÂ’s production (Steen 1998). Currently the world extrac ts approximately 40 million tons of phosphate per year from 140 million tons of rock (Steen 1998). Morocco contains over 50% of the worl dÂ’s reserves of phosphate rock. Projections of the extent of phosphorus reserves went from 160 years in 1996 to 90 years in 2001. US reserves are only projected at 25 y ears and the US will soon be a major importer ra ther than a major exporter. (Doyle and Parsons 2002) Phosphate rock quality is on the declin e as the highest grade resources are being depleted (Steen 1998). Phosphate quality is also adversely affected by increasing concentration of heavy metals in the rock. Interestingly, the phosphate rock is typically formed in seabe ds and has a high affinity for metals and easily absorb cadmium, uranium, nickel, chromium and copper from seawater during the geological formation millennia ago (Driver et al. 1999). Projections for worldwide phosphate res ources look dismal with significant depletion of known reserves expected in the current centur y. Total phosphate consumption is driven by agriculture with approximately 80% of mined phosphate routed towards this use (Steen 1998). Most models show world population,
7 agricultural production and fertilizer cont inuing a steep increase over the next century. With increasing demand expected and depleting supply and quality, it is expected that phosphate costs will increas e over the next century and the value of phosphate as a resource will also increase. 2.1.3 Summary of Resource Values The objective of this study is to evaluate both the technical and financial feasibility for recovery of phosphor us and nitrogen from treatment plant operations. Based on comparison to equi valent products and market prices the value of nitrogen can be estimated at $1000 per ton of ammonia (compare to anhydrous ammonia) and the value of phosphate can be estimated at $1000 per ton of phosphate (compare to di ammonium phosphate fertilizer). Typical influent wastewater contains between 20-85 mg/L of Total Nitrogen and between 4-15 mg/L of Total Phosphorus. To gain an idea of the mass and value of nitrogen and phosphorus which pass thro ugh wastewater treatment plants on a daily basis, we can use median values of these ranges and plant flow data from the Howard F. Curren Advanced Wastew ater Treatment Plant (HFCAWTP) in Tampa, FL which handles approximately 50 million gallons of wa stewater per day
8 or 189 million liters per day. There are approximately 10,868 kg of nitrogen and 1,781 kg of phosphorus passing through a pl ant of this size on a daily basis. It is estimated that the world uses appr oximately 95 million tons of fertilizer nitrogen per year to support a population of 6 billion peo ple, resulting in a per capita usage of 0.015 tons per person per year. If Howard Curren AWTP handles 10 tons/ day of nitrogen 365 days per year serving a population of 515,780 (according to plant operators), then we can estimate this influent at 0.007 tons of N/ person per year. Thus if all N were recovered from treatment plants (assuming all the world has treatment plants with similar characteristics to HFCAWTP) and the N were in a fertilizer form, something like 47% of the worlds fertilizer demand could be extracted from wast ewater. This figure is hypothetical assuming both that all of the nitrogen could be recovered and that all of the worldÂ’s population was connec ted to treatment plants. 2.2 Resource Partitioning in the Treatment Plant In considering the recovery of resour ces from municipal-scale wastewater treatment, a preliminary step must be taken to quant ify how the resources partition amongst the different treatment processes in the plant. In a simplistic model, the treatment plant can be divided into two treatment trains, solids and liquids. The two treatment trains from a typical wastewater treatment plant are shown in Figure 3.
9 Figure 2: Concept of Liquids and Solids Treatment Trains Typical values of municipal wastewater characterization, pr imary sedimentation efficiencies, and primary sludge solids content have been taken from Tchobanoglous and Burton (1991) and an esti mation of the portioning of solids between the liquid and solid trains of the tr eatment plant can be calculated. It is assumed for this preliminary calculati on that partitioning of resources will generally follow the partitioning of solids. Values used for estimation include a medium strength wastewater with 500 mg/L TDS and 220 mg/L TSS with a primary clarifier that is 60% efficient in removal of TSS and a resulting sludge with solids content of 6% by mass.
10 Using these parameters and balancing t he mass of total solids around the primary clarifier we can show that approx imately 18% of Total Solids (dissolved + suspended) is routed to the solids side of th e treatment plant in 0.2% of the flow volume. The liquid side of the plant handles approximately 99.8% of the flow and 82% of the total solids. A large portion of the solids initially diverted to the li quid side of the plant is biologically incorporated and sent to the solids side at a second clarifying step, where Waste Activated Sludge (WAS) is s eparated and sent to solids. Thus the percentage of total solids sent to the digester increas es from 18%, with little increase in the total flow percentage. Although the total solids (TS) sent to the solids side is less than half of the total influent TS, the high concentration of solids in relatively low flow volume warrants focusing attention to the solids side of the plant for resource recovery. While resources of particular interest resour ces of particular in terest (nitrogen, phosphorus, carbon, energy) may not parti tion in exact accordance with total solids, solids partitioning analysis should give a general idea of resource partitioning and encourage further study. Fu rther assessment of the partitioning between the solids and liquid side of typica l treatment plants is discussed in more detail below.
11 2.3 Fate of Nutrients in Digester Nitrogen If we can establish that the solids side of the plant should offer the greatest opportunity for resource recovery, the fate of the resources in the solids digestion process should be understood in order to evaluate potential for resource recovery. Wastewater treat ment facilities typically employ one of two solids digestion strategies, aerobic or anaerobic. Bioenergetic half-reaction m odeling is an approach which can be used to provide a stoichiometric representation of a micr obial reaction. Development of these stoichiometric representations is c onducted by selecting and combining halfreactions for an electron donor, electron a cceptor, and cell synthesis reaction, each reduced to single electron equivalen t. The half reactions are combined using an energetic partitioni ng coefficient which is specific to how a particular microbial reaction partitions the electr ons between growth of new biomass and cellular metabolism. These energetic coefficients, fs (synthesis) and fe (energy) are specific to the various microbial reac tions (or consortium of reactions which occur in the digestion process) and are developed empirically through monitoring of biomass growth. Development of a stoichiometric repres entation of the digestion process also requires that the complicated influent streams used as carbon source and electron donors be approximated as a single compound. Influent streams into
12 the digester systems are prim ary sludge, represented as C10H19O3N and waste activated sludge, represented as C5H7O2N. The following sections provide the resulti ng stoichiometric representations for the aerobic and anaerobic digestion of primary sludge and wa ste activated sludge, using half-reactions and energetic partitioni ng coefficients provided by Rittman and McCarty (2001). Before discussing the results, several signi ficant limitations to this approach should be noted. Reduction of the complex mixture of suspended and dissolved, organic and nonorganic compounds into one formula deemphas izes the complex disintegration and hydrolysis reactions necessary to make compounds biologically available for the microbially catalyzed reaction. The estimated fs values used to determine the partitioning cell growth versus ce ll maintenance energy have a significant effect on the partitioning of the resources, and these fs values were taken from literature describing general classes of organisms. Th e reaction as written in equation 2-1 proceeds to completion, where a ll primary sludge is fully digested to cells and carbon dioxide, when in reality we expect many compounds to leave the reactor in various stages of br eakdown and intermediat e products.
13 However, understanding these limitations on this approach, it is also understood that the fs values and characterization of primary sludge are the result of empirical studies and following through with this analysis should provide a valuable initial approximation on the partitioning of resource s through the reactor. There is an enormous difference in the typical fs values for aerobic and anaerobic digestion processes, spanning more than an order of magnitude. High values for the aerobic process indicate a build ing of cell mass incorporating influent material into cell mass while low values for the anaerobic process indicate a slow growth rate and a breakdown of influent products into compounds other than incorporated cellular material. Phosphorus is a small portion of the over all mass of primary and waste activated sludges and is not typically tracked in the half reaction methodology described above. Thus the methodology allows for an estimation of the fate of the nitrogen, but not phosphorus. The fate of phos phorus is discussed in the following section. 2.3.1 Aerobic Digestion Using an fs value of 0.6 and ammonia as the nitrogen source for cell synthesis, equation 2-1 is developed. The deriv ation of equation 2-1 is included as an example calculation in Appendix E-1.
14 O H CO N O H C HCO O NH N O H C2 2 2 7 5 3 2 4 3 19 1025 5 2 75 1 75 0 75 3 75 0 Where: N O H C3 19 10 : represents typical primary sludge N O H C2 7 5: represents new bacteria cell mass Equation 2-1: Aerobic D egradation of Pr imary Sludge Appendix E-2 shows calculat ions in determining the partitioning percentages of the resources from Equation 2-1 and re sults for nitrogen partitioning in the aerobic digestion of primary sludg e are displayed in Figure 4. Figure 3: Theoretical Aerobi c Degradation of Primary Sludge
15 Figure 4: Theoretical Ae robic Degradation of WAS An equation representing the aerobic digestion of Waste Activated Sludge (Equation 2-2) has been created according to similar methodology presented in Appendix E-1 from data ta ken from Rittman and McCart y (2001). WAS is represented by a typical formula for cells. Since the formula for cells (C5H7O2N) would be the same on the left and right hand side of the equation, the right hand side has been modified to read Â“new cellsÂ”. NewCells HCO NH CO O H O N O H C 6 0 4 0 4 0 6 1 4 0 23 4 2 2 2 2 7 5 Where: N O H C2 7 5: represents bacteria cell mass degraded New Cells ( N O H C2 7 5): cells growing from the digestion of WAS Equation 2-2: Aerobi c Degradation of WAS
16 According to this analysis, aerobic di gestion of WAS should release ammonia nitrogen and reduce the total cell mass to approximately 60%. Graphical representations of the di stribution of the nitrogen are shown in Figure 5. Aerobic digestion of primary sludge and WA S at the treatment plant often occur together simultaneously and in the same reacto r. Therefore, th e partitioning of the compounds in the total digestion proces s at a typical treat ment plant would be expected as a weighted combination of the two analyses above, considering the relative contributions of primary sludge and WAS. 2.3.2 Anaerobic Digestion Using the same methodology of estimati on of stoichiometry for the microbial catalyzed digestion reaction and then calc ulating partitioning percentages, the fate of the nitrogen resource in anaer obic digestion of primary sludge and WAS has been calculated. An estimation of the anaerobic digestion of primary sludge has been estimated using half reactions provided by Rittman and McCarty, an fs value of 0.05, ammonia as the nitrogen source for cell synthesis, and the same representation for primary sludge as us ed for the aerobic estimation. The sample calculation Appendix E1 provides the basic strategy for determining the following equation:
17 Where: N O H C3 19 10 : represents typical primary sludge N O H C2 7 5: represents new bacteria cell mass Equation 2-3: Anaerobic D egradation of Primary Sludge Again, it should be noted that there are si gnificant limitations to the power of prediction of the resource partitioning us ing this equation, even more significant than in the case of the aerobic digestion. In this case, usi ng the carbon dioxide as the terminal electron acceptor and th is general formula for primary sludge as the donor, there are many steps ignored in the breakdown of organic material to acetate, including disint egration, hydrolysis, acidogenesis, and finally acetogenesis. However, again, the fs values and characterization of primary sludge were determined from empirica l data and this methodology should provide a valuable prelim inary predictive tool. Fi gure 6 depicts the nitrogen resource partitioning using simila r methodology to Appendix E-2. Equation 2-4 has been prepared using the methodology of the sample calculation in Appendix E-1. 3 4 2 7 5 2 4 2 3 19 1086 0 86 0 14 0 54 2 91 5 95 4 HCO NH N O H C CO CH O H N O H C
18 Figure 5: Theoretical Anaerobic Degradatio n of Primary Sludge Figure 6: Theoretical A naerobic Degradation of WAS
19 3 4 2 4 2 2 7 59 0 9 0 1 0 35 1 25 2 6 3 HCO NH NewCells O H CH O H N O H C Where: N O H C2 7 5: represents bacteria cell mass degraded New Cells ( N O H C2 7 5): cells growing from the digestion of WAS Equation 2-4: Anaerobi c Degradation of WAS Partitioning of resources based on Equati on 2-4 has been calculated similarly to example Appendix E-2 and is graphi cally represented in Figure 7. 2.4 Fate of Nutrients in Digester Â– Phosphorus Phosphorus is typically not tracked in the half reaction methodology used above to track nitrogen. However, using the same stoichiometric equations developed above, we can do some very rudimentary prediction of the fa te of the phosphorus in the two digestion systems by using known content of phosphorus in typical wastewater influent and typical cellular mate rial. Cell material typically contains 2-3% P by dry weight and th is can be represented as C5H7O2NP0.1. Typical wastewater influent or pr imary sludge will typically c ontain approximately 0.5-2% P and this can be represented by adding a P term to our previous representation of primary sludge, as C10H19O3NP0.07. Then, using the molar ratios of products and reactants developed above with the bi oenergetics method, we can estimate
20 a percentage of influent phosphorus that is incorporated into cellular material. For example, equation 2-5 modifies equation to 2-3 to include phosphorus. Where: 07 0 3 19 10NP O H C : represents typical primary sludge 1 0 2 7 5NP O H C : represents new bacteria cell mass Equation 2-5: Anaerobic Degr adation of Primary Sludge Using a similar methodology to the ca lculation of nitrogen partitioning in Appendix E-2, we see that there are 2.16 grams of phosphorus per mole of influent primary sludge and 0.43 grams of phosphorus incorporated into new cells per mole of influent primary sl udge. This represents 20% of influent phosphorus incorporated into solids. The significant limitation to this esti mation is that non cellular phosphorus, typically present in the ortho-Phosphate forms is reactive and there is no methodology presented here which can esti mate whether the remaining 80% is complexed into solids or remains as liquids. But, continuing with this estimation pr ocedure we find the following results. Aerobic digestion of primar y sludge will result in 100% incorporation of influent P 3 4 1 0 2 7 5 2 4 2 07 0 3 19 1086 0 86 0 14 0 54 2 91 5 95 4 HCO NH NP O H C CO CH O H NP O H C
21 into cellular material, if all material is digested. Aerobic digestion of WAS will result in 60% incorporation of influent P into cellular material. Anaerobic digestion of primarily sludge will incorporate 20% of infl uent P into cellular matter and anaerobic digestion of W AS will incorporate 10%. Again, as was evident in the nitrogen ana lysis, the fate of phosphorus in the digester is directly tied to the energetic par titioning coefficients for the processes. The high fs of the aerobic process yields hi gher masses of cellular materials, incorporating the phosphorus into the biomass while the low fs in the anaerobic process yields a slower microbial growth, a reducing environment, and a breakdown of influent material without si gnificant build-up of biomass which sequester nutrient into a so lid, recoverable form. 2.5 Current Nitrogen Removal Technology 2.5.1 Biological Nitrogen Removal Biological Nitrogen Removal (BNR) is a well established technology. The process involves several steps to oxidiz e ammonia to nitrite then nitrate and then reduce nitrate to nitrite and then to nitr ogen gas. Figure 8 illustrates the nitrogen cycle which is encouraged by the wastewat er treatment plant for removal of nitrogen from the liquid train effluent.
22 Figure 7: Nitrogen Cycle in WWTP Ammonia-oxidizing bacteria include Nitrosomonas and other genera, and Nitrite oxidizing bacteria include Nitrobacter Nitrospira and others. Nitrification requires energy input for aeration to supply oxygen to engineered systems culturing aerobic autotrophic ammonia-oxidizing bacte ria and nitrite-oxid izing bacteria. Nitrifying bacteria are sl ow growing in comparison to heterotrophic microbes which dominate in carbonaceous BOD removal. As a result of the differences in growth rates of the two aerobic microbi al systems employed in the treatment plant, the nitrification is sometimes separated from BOD removal and conducted in a separate aeration basin with longer solid retention times.
23 Once nitrogen compounds are dominated by nitrate an anoxic denitrification process is undertaken to reduce the nitr ate to nitrogen gas. Denitrification requires a carbon source and electr on donor for the anoxic heterotrophic denitrifying bacteria. Costs to provi de the electron donor as methanol are typically a great expens e to the plant. The end result of the BNR process is re lease of the nitrogen to the atmosphere as nitrogen gas. In the c ontext of resource recovery, this process represents a loss of the nitrogen resource. 2.5.2 BNR Nitrification/ Denitrifi cation Efficiency Improvements It has long been recognized in the field of wastewater treatment that the traditional BNR process includes some appar ent inherent inefficiencies. Energy input is needed to nitrify and then energy inpu t is needed to denitrify. Nitrite is an intermediate product in both processes. Biological nitrogen removal is under constant development for im provement in efficiency an d reduction in operating costs. Several technologies and processes have been developed to reduce energy and material inputs of the nitr ification and denitrification process. Although these improvements still result in a loss of the resource, they are developed to accomplish the removal of the nitrogen at reduced energy expenditure and operating co st. Figure 9 shows the revised nitrogen cycle, employing the advanced biological nitrogen removal technologies. Technologies
24 such as MAUREEN, SHARON, ANNA MOX, DEMON, and STRASS create a short cut in the nitrogen cycle, allowing for significant savings in the nitrification and denitrification process. Figure 8: Nitrogen Cycle S hortcut in Enhanced BNR 2.5.3 Struvite It is possible to precipitate ammonia ni trogen as a compound called struvite. Struvite contains equimolar concentrations of Ammonia, Phosphate and Magnesium. Removal of nitrogen by prec ipitation of struvite represents a
25 recovery of the resource. Struvite is an effective non-burning slow release fertilizer. Because in most wastewater treatment si tuations, ammonia is in molar excess (estimate 8:1) over phosphorus, only a small portion of the ammonia can be easily removed while high percentages of the phosphorus can be removed. Most of the available literature on struvi te treats it as a phosphorus removal technology. Struvite is discussed in mo re detail in the next section as a phosphorus removal technology. 2.6 Current Phosphorus Removal Technology 2.6.1 Metal Salt Precipitation Under appropriate conditions, various me tal cations will precipitate phosphate from solution with iron the most common. Th is is a typical practice at waste water treatment plants which must meet phosphorus limit and do not employ the biological phosphorus remova l scheme. Addition of ir on and precipitation of ferrous phosphate is the most common me tal salt precipitation for phosphorus removal. Ferrous phosphate has applications in ra ilway brake blocks, but its economic value is low. (Driver et al. 1999). The phosphorus industry typically regards iron
26 in phosphate as undesirable, as most valued phosphate end products are difficult to derive from ferrous phosphate (Driver et al. 1999). 2.6.2 Biological Phosphorus Removal (BPR) Bacteria utilize phosphate in typical cellular processes. It is estimated that typical aerobic organisms present in activated sludge contain 2-3% P on a dry weight basis (Rittmann and McCarty 2001). Bacteria in the genera of Acinetobacter Pseudomonas Arthrobacter Nocardia Beyerinkia Ozotobacte r, Aeromonas Microlunatus Rhodocyclus and others have been shown to uptake phosphate in concentrations which exceed a typica l phosphate concentration. These organisms are known as Bio-P organism s and are utilized for biological phosphorus removal (BPR) in engineered systems. The crucial design component to facilitate BPR is a cycling of the cells between aerobic and anaerobic conditions. The Bio-P organisms have the ability to Â“investÂ” in energy storage during the aerobic cycles and use the stored energy during anaerobic cycles to ferment volatile fatty acids and sequester electrons. The energy storage medium is intrac ellular polyphosphate (poly P) and phosphate is uptaken during the aerobic cycles and released during anaerobic cycles.
27 The Bio-P organisms outcom pete organisms which do not have the ability to invest in energy in aerobic conditions to spend during the anaerobic conditions. The cycling between aerobic and anaerobi c phases induces the uptake and release of phosphorus but it also serves to exert ecological pressure to select for the Bio-P organisms. Phosphate uptake occurs in the aer obic stage when electron acceptors oxygen and nitrate are available for synthes is of adenine tri-phosphate (ATP). Polyhydroxybutyrate (PHB) st ored in the cell is hydrolyzed to acetyl coenzyme A (HSCoA) and then oxidized in the TCA cycle. Released electrons from the oxidation are used for ATP synthesis and then ATP is used to synthesize poly P for energy storage. Thus for the forma tion of the poly P, the organism must uptake phosphorus from the environment. In the anaerobic zone, electrons are s equestered in PHB using HSCoA which consumes energy. Energy comes from ATP through hydrolysis of stored poly P. The hydrolysis of poly-P for ATP and energy releases phosphorus. The strategy at BPR treatment plants is to cycle solids between an anaerobic and aerobic zone allowing them to accumu late and release phosphorus from influent wastewater and then waste so lids immediately following the aerobic stage when the intracellu lar polyphosphate and thus solid phase P is at its maximum.
28 2.6.3 Struvite Unintentional precipitation of struvite in treatment plants has long been a problem where reactors, piping and equipment become fouled with the crystallized product. It is estimated, based on meas ured formation rates, that struvite accumulation can bring a 12 inch pipe to 50% capacity within three years. Controlled and intentional precipitation of struvite (magnesium ammonium phosphate (MAP)) is a potential resour ce recovery technology, studied thoroughly on lab and pilot scales and impl emented in a few cases on the scale of municipal treatment plant s. Struvite requires equal molar concentrations of ammonium, phosphate and magnesium. With increased regulation on phosphorus effluent limits, there have been many studies on phosphorus removal through struvite precipitation. Ty pically, in anaerobic digester effluents, ammonium is in molar excess a nd depending on magnesium concentration, phosphorus can be removed with minimal ad dition of chemicals. Struvite reactors have a smaller footprint and have less operational problems than BPR reactors (Wang et al. 2005). O H PO MgNH PO NH Mg2 4 4 3 4 4 26 Equation 2-5: Basic Struvite Formation
29 The key control parameters to facilitate precipitation of st ruvite are solution super saturation and pH (Ali et al. 2005). Es timations for the solubility product for struvite range significantly in the literat ure. Values are reported ranging from 9.4 to 13.26 (Doyle and Parsons 2002). A conditional solubility product is defined for struvite precipitation and the interacti on between a conditional solubility product and pH is modeled to control struvite solubi lity. The conditional solubility product is defined as follows: ] [C ] [C ] [C PsNH4 t PO4 t Mg t Equation 2-6: Definition of Conditional Solubility (Ohlinger et al. 2000) When the solutionÂ’s condition al solubility product is greater than the equilibrium conditional solubility product the solution is in supersaturation and precipitation of struvite is possible. Several inve stigators have developed curves for the equilibrium conditional solubility vs. pH, an example shown below in Figure 10 is adapted from Ohlinger et al. (2000), rela ting the negative log of the conditional solubility to pH.
30 Figure 9: Struvite Condi tional Solubility Curve ( adapted from Ohlinger et al. (2000)) As shown above in the solubility curve, the super saturation zone is achieved at lower reactant concentrations as pH increases. Optimum pH for minimum struvite solubility has been reported in the literature to range between 8.0 and 10.7 (Doyle and Parsons 2002). Control of pH in pilot scale and full scale reactor has been handled in two ways, via aerati on for carbon dioxide stripping or via addition of an alkaline agen t. Aeration and carbon di oxide stripping should be the least expensive pH control measure. Italian investigator s are using aeration only with reported success (Battistoni et al. 1997). However, Japanese investigations report reduction in alkali agent requirement by stripping, but not elimination (Fujim oto et al. 1991).
31 As the pH approaches 11, struvite producti on is inhibited by two factors. Ammonia volatilization will occur at higher pH and Mg(OH)2 may precipitate (Wang et al. 2005). Many investigations use MgO as the magnesium source as it also provides alkalinity and thus reduc es chemical costs for pH control (Booker et al. 1999). Solution chemistry may produce preferent ial precipitation of other compounds and the presence of calcium ions is the pr imary inhibitory ion (Wang et al. 2005). Increasing the magnesium to calcium ion ratio will result in more efficient struvite production (Battistoni et al. 1997). Studi es have shown that an excess molarity of ammonia drives the reaction towards a pure struvite, while an excess molarity of magnesium yields a less pure product (Wang et al. 2005). Fluidized bed reactors (FBR) appear to be the most established technology to facilitate precipitation in pilot scale and full scale systems. At the time of publication in 2001, a treat ment plant in Japan had been operating a FBR for three years successfully producing struvi te and achieving significant phosphorus removal (Ueno and Fuji 2001). The Japanese system used sodium hydroxide for pH control. Italian investigators have developed a treat ment process which accomplishes pH adjustment with carbon dioxide strippi ng and have implemented this on a full scale plant (Battistoni et al 1997). Britton et al. (2005) could consistently recover
32 over 90% of phosphate in a pilot scale pl ant using anaerobic digester filtrate and sodium hydroxide for pH control. Precipitation of phosphorus by struvite r epresents a recovery of the resources. One investigator estimates that wit h only 55% recovery of phosphate and with 50% of the world attached to sewers, 1.6% of the worldÂ’s annual phosphate consumption could be supplied by recovery (Shu et al. 2005)! If 100% of the world were served by wastewater treat ment plants and 100% of phosphate were recovered, phosphate mining could be r educed by 5.75% (Shu et al. 2005).
33 3 Study Treatment Plants: Preliminary Evaluation In Chapter 2, the following points are discussed regarding resource recovery in wastewater treat ment plants: NitrogenÂ’s value as a resource is deriv ed from its biological significance and the energy consumption to create ammonia from atmospheric nitrogen. PhosphorusÂ’s value as a resource is der ived from its biological significance, its finite and limited quantity on earth, and production costs to mine it from rock. Based on a preliminary assessment, it appears that a significant fraction of the nitrogen and phosphorus should end up in the solids handling portion of a typical wastewater treatment plant, if nutrient partitioning follows solids partitioning. Based on a preliminary investigation and literature review, it appears that the fate of the resources is different within the typical aerobic and anaerobic digesters utilized at treatment plants. Various technologies are discussed wh ich have potential for recovery of nitrogen and phosphorus.
34 In order to further evaluate the potential for resource recovery in the wastewater treatment plant, one operating municipal plant utilizing aerobic digestion and one plant utilizing anaerobic digesti on have been chosen to study. 3.1 Process Descriptions 3.1.1 Howard Curren Treatment Plant The Howard F. Curren Advanced Wastew ater Treatment Plant serves the municipality of Tampa, FL as well as several surrounding suburban municipalities. Typical plant influent flow hovers around 50 million gallons per day (MGD). Figure 10: Howard F. Curren Adv anced Wastewater Treatment Plant
35 Carbonaceous BOD removal of the prim ary effluent is accomplished with high purity oxygen supplied to activated sludge. Waste activated sludge is thickened and then sent to anaerobic digesters. Pr imary sludge is collected in primary settling tanks and digested along with the WAS. The facility maintains six digesters and alternates flow between them The facility conducts nitrification using typical aeration basins and operates anaerobic denitrificat ion filters using methanol as the carbon source and elec tron donor. Wasted sludge from the nitrification tanks is pumped to the primar y settling tanks, and thus is indirectly diverted to the anaerobic digesters. Biogas is collected from the anaerobic digesters and sent to a cogeneration facility. This facility burns the biogas in generators for producti on of electricity. Jacket cooling water is diverted to a heat exchanger where the digester sludge is heated. Effluent sludge from the digesters is pum ped to a solids handling facility. Solids are dewatered using a series of belt filter pr esses. Filtrate water is gravity fed to the high purity oxygenation tanks along with filter press wash water. This water then undergoes nitrification and denitrification before di scharge. Currently no phosphorus discharge limit for the HFCA WTP and no removal or recovery of phosphorus is practiced other than the typi cal accumulation of P in biomass. Figure 11 shows a simplified schematic of HFCAWTP.
36 According to plant operators, the plant experiences problems with high nitrogen (ammonia) concentrations in the filtrate water returning to the head of the plant. The plant is currently investigating options to deal with the filtrate in a sidestream operation. As part of this study, it is decided to in clude an investigation into the specific potential of recovering the anaerobic digeste r filtrate ammonia through struvite precipitation. The feasib ility of recovering struvite should be compared financially to the current nitrification and denitrifi cation process that treats the digester filtrate. 3.1.2 Largo The Largo Advanced Wastewater Treatme nt Plant (LGAWTP) services the municipality of Largo, FL and handles approx imately 12 million gallons per day if influent wastewater. Prim ary sludge is collected in primary settling tanks and sent to an Aerobic Digester.
37 Figure 11: Largo Advanced Wa stewater Treatment Plant Primary effluent is sent into an A2O reactor system facili tating biological nitrogen removal through nitrification and denitrif ication with biological phosphorus removal through PAO. Waste Activated Sl udge is also sent to the aerobic digesters. Liquid effluent from the A2O process undergoes furt her denitrification, final filtration and chlorination prior to discharge. Digested sludge is gravity thickened and then s ent to a solids processing facility. Filtrate from the thickener is retu rned to the head of the plant. Figure 12 represents a simplified schematic of the Largo Treatment Plant.
38 According to plant operators, there ar e sometimes problems with achieving the phosphorus discharge limits. One possible explanatio n is P-release in the thickener and high P concentrations in t he filtrate overloading the A2O system. Further investigation into the possibility and rates of P-release in the thickener are incorporated into the study. 3.2 Further Plant Specific Background Â–LGAWTP 3.2.1 Potential for Phosph orus Release from Sludge There are no significant gaseous forms of phosphorus to be considered under wastewater treatment conditions. Ther efore, when considering the fate of phosphorus through the system boundaries of a treatment plant, all incoming P must either be discharged in the liquid state or recovered as a solid. Wasting sludge from the aeration basin effe ctively removes P from the dissolved phase and out of the liquid side of the pl ant, hopefully to a level to achieve its primary goal of meeting the Total P effluent standard (typically 1 mg/L). Wasted sludge from BPR is typically sent to the solids handling side of the plant and at the time of wasting it contains P in a recoverable form, bound in cell mass. However, conditions in solids handling hav e potential to instigate release of the P.
39 The potential for P release is well establis hed. It is the cycling of uptake and release of P at increasing capacity that facilitates the BPR process. Anaerobic conditions will cause a release of P. In the case of the La rgo Treatment Plant WAS is sent to the aerobic digester w here there is continuous aeration and should not reach anaerobic conditions unt il a gravity thick ening stage between the digester and the solids processing facili ty. The rate at which this phosphorus release occurs in this thickener wi ll determine the quantity of phosphorus release. 3.2.2 Rate of Phosphorus Release Rates of P-release have been quantified by several studies and using several different units for quantification. In a study on extended aeration times in the aerobic cycle of the BPR process, Brdjanovic et al. (1998) m easure specific P-release rates ranging from 0.059 to 0.092 (mg-P/mg-active biomass)/hour (Brd janovic et al. 1998). The Brdjanovic experiments were conducted in cont rolled sequencing batch reactors with simulated wastewater. In another experiment conduc ted by Brdjanovic et al. (2000), specific P-release rates were calculated at 6 mg Â– P/ g VSS-hour In this study it is interesting to
40 note that the model correlation of P-rel ease was one of the poorest aspects of the model. In a study to quantify the affect of ni trate in the anaerobic zone on P-release rates and subsequent P-uptake and BPR pe rformance, Artan et al. (1998) quantified the P-release rate ranging from 5-37 mg PO4 3-/ g VSS-h. The three studies above quantify P-releas e rates but each of them uses a laboratory scale reactor with a synthesized wastewater consisting of a carbon source and phosphate source. Two of the P-release rates are specific to VSS and one is specific to active biomass. All three studies quantify P-release rates in a cycling system similar to the BPR process, not after P-rich WAS has undergone an extended aeration/ digestion process. Although the rates above provide a starting point, they are derived from conditions distinctly different from those at the Largo Treatment Plant thick ener and application of these rates to our system would be difficult. Kuba et al. (1997) investigates the kinet ics of the phosphorus removal process and how it is affected by shortening t he cycling times. The study discusses numerous factors which affect the P uptak e rate and the overall growth rate of the Poly P organisms. Although the study conducted by Kuba et al. (1997) was primarily concerned with phosphate uptake rate and capacity he does provide Prelease rate data. However the P-release data is specific to acetate uptake rates
41 and would be tough to apply to another system. Most notably, the P release rate appears to rise and fall with the P-uptake ra te and total uptake capacity. We see in the experimental data considerable va riation in P release rates but it appears to show a strong dependency on the upt ake rate (Kuba et al. 1997). Many studies which provide P concent ration vs. time show a similar and characteristic shape to the curve, giving an indication of the P release behavior under anaerobic conditions. Figure 13 shows a typical curve adapted from previous literature (Kuba et al. 1997). From this curve we see a rapid and nearly linear increase in phosphate concentration until it nears the maximum concent ration. It is th is characteristic shape that allows quantificat ion of a P-release rate by approximating the bottom of the curve as linear. From this curve we also see that P-release occurs on the order of minutes, rather than hours once the phosphate accumulating organisms (PAO) are introduced to anaerobic conditions.
42 Figure 12: Phosphorus Release Curve 3.2.3 Correlation Between P-Release and P-Uptake Rates Although Kuba et al (1997) does not discu ss this explicitly, it appears from the experimental data that more effective P-uptake in the aerobic phase results in more rapid release of P in the anaerobic phase. At the Lar go treatment plant, this result may prove somewhat counter productive where effective P-uptake in the cycling A2O system could result in rapid P-release in the digested sludge gravity thickener. Mulkerins et al. (2003) also discuss the correlation between P-uptake and Prelease rates and the strong dependence be tween the two, but restricts the correlation to a temperature dependence at 15-25 C. This study discusses cases
43 where at lower temperatures, P-release rates do not correlate with uptake rates and overall BPR performance is dimini shed (Mulkerins et al. 2003). 3.2.4 Phosphate Release Du ring Aerobic Digestion As discussed, the PAO in the BPR system accumulate phosphate during aerobic cycles and release it during anaerobic. It may be assumed that if the WAS is kept aerobic, that it should hold onto t he P. However, it has been noted by plant operators that over extended aeration periods, the accumulated P in the cells is released. Through experience in Johannesbur g South Africa BPR plants, Pitman states that endogenous hydr olysis of P-rich WAS wil l release P in an aerobic digester (Pitman 1998). Pijuan et al. (2005) further characterize the P-release rate in periods of extended aeration showing P-release rates increasing significantly between day 8 and day 11. 3.2.5 Summary of Literature Review on Phosphorus Release The following summary points are drawn fr om the literature review on P-release as it relates to the sludge handlin g at the Largo Treatment Plant: Phosphorus release rates are quantified in the literature, but no values were found which were directly applicable to the specific situat ion at Largo where aerobically digested P-rich WAS is gravity thickened.
44 Phosphorus release rates appear dire ctly correspond to phosphorus uptake rates, indicating that the more effective phosphate uptake is in the A2O system, the more rapid the P-release should be when the sludge encounters anaerobic conditions. This correspondence may be affected by extended aeration in the digester. Even while maintaining aerobic conditions P-release may be occurring in the digester. Based on these summary points, it is dec ided that a study should be conducted to evaluate the specific phosphorus release rates in the Largo AWTP thickener. Concurrent to this study an evaluation of the settling rate of the sludge can be conducted. Comparison of these tw o studies should provide an optimum residence time in the thickener to minimi ze P-release and maximize thickening. Also, the mass balance of P around the di gester system should reveal whether P is released from solid to liquid during the solids retention time in the digester. 3.3 Further Plant Specific Background Â– HFCAWTP 3.3.1 Struvite Potential for Ammonia Recovery According to Howard F. Curren AWPT plan t operators, digester filtrate contains a high concentration of ammonia nitrogen and is periodically causing problems with nitrogen loading when pumped to the head of the plant. Attempts at ammonia recovery in wastewater treatment thr ough struvite precipitation are scarce,
45 because ammonia is typically in molar excess with respect to the phosphate and magnesium and significant chemical add itions are required. But, when examining the nitrogen cycle which occurs in the treatment plant and combined with the synthetic fixation of ammonia by the Haber Bosch process, it is apparent that society is paying to circumnavigat e the nitrogen cycle, as outlined in Figure 13. Figure 13: Circumnavigate the Nitrogen Cycle Struvite precipitation has the potential to cut out the loop in this process recover the valued nitrogen resource.
46 Celen and Turker (2001) evaluate the potential for full nitrogen removal from digester effluents. Their study uses batch reactors and quantifies costs for chemical additions to achieve full ni trogen removal and show effective ammonia reduction, but the ammonia source is laboratory chemicals simulating effluent concentration. There is no literature found wh ich provides pilot scale or full scale operational data which would be applicable to the Howa rd Curren Plant. There is no study which attempts a full ammonia recovery from a high ammonia concentration waste stream from anaer obic digestion supernatant by addition of phosphate, magnesium and pH control. Phosphate is a limited and expens ive chemical and its discharge is regulated. Addition of phos phate into the wastewater stream is at high quantities is not desirable. 3.3.2 Estimates of St ruvite Recovery Costs Estimates are given in previous literatu re for the material and operating costs to remove struvite. A lower range is $8.50 per kilogram of NH4-N (Celen and Turker 2001). A higher range is estimat ed at $9.72 per kilogram of NH4-N (Siegrist 1996). Doyle and Parsons (2002) provide a review of struvite literature and tabulate various reported costs to produc e and sell struvite. Production costs ranged from $140 $460 per ton. These costs do not include attempting to remove nitrogen, but were developed in systems where ammonia was left in
47 excess of phosphate concentration and re moval of phosphate was the objective. Struvite resale costs varied even more significantly than the production costs, ranging from $198$1885 per ton. Siegrist evaluated struvite precipitation costs versus other nitrogen removal costs in 1996 and concluded that struvite pr ecipitation was more expensive than nitrification/ denitrification by a factor of 4 (Siegrist 1 996). However, the Siegrist evaluation did not include resale pot ential for the recovered struvite. 3.3.3 Summary of Struvite Po tential Literature Review From the literature review on struvite and its potential for a si destream treatment technology for nitrogen and phosphorus recove ry from anaerobic digester filtrate, the following summary points are provided: Recovery of nitrogen and phosphorus should be possible by creating appropriate conditions and providi ng appropriate concentrations of constituents. There is ample data in the literature which provide ranges of operating conditions plus discussion of inhibitory constituents. The variability of wastewater conditions coupled with variability in literature values for optimum pH and solubility product for struvite precipitation suggests that bench and pilot testing is required prior to any system implementation.
48 If ammonia recovery is attempted, repor ted problems associated with the Ca : Mg ratio should be easily avoided as large quantities of magnesium would be added. It does not appear that many investigat ors or treatment plants are pursuing ammonia recovery through struvite, lik ely due to fear of adding a phosphate compound to the wastewater stream which must later be removed. Evaluations of financial aspects to struvite recovery in the literature vary widely. From the literature, it appears that using phosphoric acid as the phosphate source, magnesium oxide as the magnes ium source, and aeration for carbon dioxide stripping and a fluidized bed reac tor would be the most cost effective system for struvite recovery. Based on the summary points above it was dec ided to conduct a study of struvite precipitation using digester effluent filtrate. Bench scale batch tests of precipitation potentia l, required chemical additions and recovery potential of nitrogen and phosphorus were conducted. From this initial investigation, further assessment of financial considerations can be made to evaluate the overall feasibility of struvite as a sidestream f iltrate treatment tec hnology for the Howard Curren treatment plant.
49 4 Rationale and Outline of A ssessment and Experimental Plan Chapters 2 and 3 represent a background discu ssion to serve as a starting point for further analysis of re source recovery potential at the Howard Curren and Largo Treatment Plants. The following basic point s were established: There are resources of value in wastewater The value of the resources in wast ewater will likely increase relative to operating costs, and if this occurs, financ ial benefits of resource recovery may become increasingly attractive. An initial attempt is made to underst and the partitioning of the resources through the wastewater tr eatment plant and from th is understanding it is suggested that the solids side of the plant will be the most effective area to focus resource recovery efforts. We attempt to understand the partitioning of resources through the two common digestion processes, aer obic and anaerobic digestion. We discuss some of the ways that the resources are treated, recovered or removed in various technologies associated with the solids side of the treatment plant.
50 Two operational treatment plants are chosen to evaluate the potential for resource recovery in aerobic and anaer obic digestion systems and specific target areas of investigat ion are identified at each of the two study plants. At the anaerobic digestion system of the Howard Curren Treatment Plant, investigation into the feasibility of st ruvite precipitation is suggested as a means for ammonia removal an d recovery from filtrate. For the aerobic digestion system at t he Largo Treatment Plant, a study of specific phosphorus release kinetics is suggested in order to better design retention time in the gravity thi ckener to maximize sludge settling and minimize phosphorus rel ease to the supernatant. The following sections outline the rational and investigative plan to further evaluate the points above. 4.1 Evaluation of Resource C ontent in Solids vs. Liquids The objective of this study is to provi de an estimate of the partitioning of nitrogen and phosphorus into the solids and liquid side of the treatment plant. Samples from the primary effluent, primary sludge, and waste activated sludge of each plant were collected in order to quantify the total nitrogen and phosphorus in the dissolved and suspended phases. From this data, an estimation of the partitioning between the solids train and liqu id train can be estimated. Analysis
51 of partitioning was conducted on a mass per time basis. Thus, the following equations are used for eac h of the resources. WAS PE liquids WAS PS solidsm m m m m m Where : m = mass of nutrient per time (kg/day) PS = Primary Sludge WAS = Waste Activated Sludge Equation 4-1: Solid Train and Liquid Train Partitioning 4.2 Mass Balance: Nitrogen and Phosphorus in Digesters The objective of this study is to quantif y the fate of the resources through two digestion systems, one anaerobic and the ot her aerobic using the Howard Curren Treatment Plant and Largo Treatment Plant reactors A quantification of resource partitioning through these systems will allow for evaluation of resource recovery potential after the digestion proce ss. A thorough eval uation of resource partitioning may also allow for decision making in technology selection, if resource recovery potential factors in to financial decisions for municipal treatment plants and farm scale treatment operations.
52 An experimental plan was developed in order to det ermine the fate of the resources in an aerobic digestion and an ana erobic digestion system. Samples were collected from two digester influent streams, primary sl udge (PS) and waste activated sludge (WAS). Samples were al so collected from one digester effluent stream, effluent sludge (ES). Solids and liquids were separated from each sludge sample and analyzed for nitrogen and phosphorus. Mass balance calculations were performed according to the following equation, shown as an example for nitrogen mass balance: es gas was psQ Q Q Qes aq es es s es was aq was was s was ps aq ps ps s psTL X TS q Y TL X TS q TL X TS q Equation 4-2: Example Mass Balance Using Nitrogen Each bracketed term in the equation 42 results in a mass per time of the resource. The mass balance equation was developed without accumulation, sink or source. This generalizat ion is made, ignoring the small quantity of ammonia analysis) lab from (data sludge) L liquid/ (L Sludge Primary in Liquids Total TL analysis) lab from (data sludge) L solids/ (mg Sludge Primary in Solids Total TS analysis) lab from (data solid) mg N/ (mg sludge primary in nitrogen solid of ion concentrat q analysis) lab from (data N/L) (mg sludge primary in nitrogen aqueous of ion concentrat X ) operations plant from (data sludge primary of flowrate average Q ) operations plant from (data biogas of flowrate average Q analysis) lab (from biogas) N/L (mg biogas in nitrogen of ion concentrat Y : Whereps ps s ps aq ps ps gas
53 vaporization in the anaerobic di gester. Thus flow in for any resource (nitrogen and phosphorus) should equal the flow out. Following evaluation of the mass balance a bove, the partitioning of total nitrogen into the solid, liquid and gas phases is ca lculated from the diffe rent terms in the above equation. From this data many calculations can be made regarding the fate of the resources, percentages re covered, percentage recoverable and value of the resources. Digester supernatant or filtrate is pum ped to the head of the plants at both of the study plants. The percentage of re source loading resulting from this filtrate is calculated. 4.3 Evaluation of N and P Recovery by Struvite The objective of this study is to determine the percentage of phosphate and nitrogen recovery possible through struvite precipitation in batch reaction, using actual anaerobic digester supernatant from the Howa rd Curren AWTP. This study will also provide specific chemical input requirements for the magnesium, phosphorus, and pH control needed to facilitate struvite precipitation. Specific chemical input requirements will provide data to allow for financial calculations to assess the feasibility of struvite producti on in comparison to more traditional methods.
54 4.4 Financial Analysis of Struvite Production Given the percentage recovery obtained in the above experiments and specific chemical input requirement s, a calculation methodolog y is developed in order to assess the financial feasibility of struvi te precipitation at the Howard Curren Treatment Plant. Actual ammonia and phosphate recovery rates and actual plant data are utilized. Market prices for chem ical additions are used. The production costs of struvite are analyzed and compared to the current method of sidestream nitrogen removal including aeration fo r nitrification and methanol driven denitrification. 4.5 Aeration for Struvite pH Adjustment Based on results and conclusions from the financial analysis, a revised methodology for struvite production was ev aluated. The costs for pH adjustment through chemical addition of a strong base are a significant portion of the total production costs. Therefore investigation into a less expensive pH adjustment is desired and the literature pr ovides examples of pH adjustment for struvite precipitation through aerati on (Battistoni et al. 1997). However, because the aim of the work by Battistoni et al. (1997) is phosphate recovery, the stripping of ammonia is not investigated during the aeration process. Therefore, this study will investigate the adjustm ent of pH through aeration for carbonate stripping while monitoring ammonia stripping as a potential negative side effect, as full
55 ammonia recovery is the desired result. This section includes a theoretical calculation of mass transf er of carbonate and ammoni a through aeration. The experimental objectives are to: observe pH and ammonia concentra tions as a result of aeration quantify a relationship between aeration and reduction in lime addition for pH control Conduct further financial feas ibility study on the struvi te precipitation process using aeration and lime toget her as a pH adjustment. 4.6 Evaluation of Specific Phos phorus Release Rate Â– LGAWTP The objective of the phosphorus rel ease and sludge settling tests were to understand the kinetics of phosphate rel ease (if any) when aeration ceases and quiescent conditions are induced by the plant for sludge settling and dewatering. The tests were conducted in jars test in order to mimic the quiescent conditions in the gravity thickener employed at the Largo treatment plant. An understanding of the kinetics of phosphorus release may assist in plant operation schemes to maximize phosphate recove ry while balancing with the need to dewater sludge prior to enteri ng the belt filt er press.
56 5 Materials and Methods 5.1 Analytical Methods Sludge samples were collected from the two treatment plants on various dates. Samples were collected from various sampling ports or dipped from the digesters. The liquid portion of sludges were separat ed from solids for dissolved constituent analysis. Sludge samples were centrif uged in 50 mL centrifuge tubes at 3500 RPM for 20 minutes. Supernatant was ex tracted with a pipette and then passed through a 0.45 micron glass fiber filter on a vacuum pump assembly. Total Nitrogen in liquid was analyzed us ing the Shimadzu TOC-V with the TNM1, Total Nitrogen Measuring Unit using calibration curves generated with known concentrations of urea. Total Suspended Solids was measured a ccording to Standard Method 2540, subtracting Total Dissolved Solids from Total Solids.
57 Ammonium was analyzed by Ion selectiv e probe manufactured by Cole-Palmer Instrument Co. and a Corning 350 pH/ ion analyzer man. Calibration curves were created according to the probe m anufacturer specification using known concentrations of ammonium chloride. Nitrate was analyzed by Ion selective probe manufactured by Cole-Palmer Instrument Co. and a Corning 350 pH/ i on analyzer. Calibration curves were created according to the probe manufac turer specification using known concentrations of sodium nitrate. Reactive Phosphate (ortho-Phosphate) in liquid was analyzed using the Standard Method 8114 (molybdovanadate method) usi ng a HACH spectrophotometer. Adsorbance is measured at 420 nm. Total Phosphate in liquid and mixed liquo r suspended solids was analyzed by the molybdovanadate method with acid persulfate digestion using a test kit provided by HACH. Potassium persulfate was added to the sample and then heated. After digestion and release of bound phos phate, molybdovanadate was added to affect color change and the sample was ana lyzed using the spectrophotometer at a wavelength of 420 nm. The pH was measured using a pH probe and the Corning 350 pH/ ion analyzer.
58 5.2 Experimental Methods 5.2.1 Preparation of Stru vite by Batch Reaction Initial concentrations of ammonia, ortho-Phosphate, and magnesium were analyzed. Calculations were conducted for addition of chemicals in order to bring all three reactants to equimolar concentrati ons in the solution. Phosphate was added in the form of phosphoric acid and magnesium was added in the form of magnesium sulfate. The solution was stirred and a pH meter measured pH continuously as pH was adjusted with sodi um hydroxide or aer ation or both. When the pH reached 9.75, adjustment was stopped and stirring continued. After approximately 5 minutes, stirring was stopped and solids were allowed to settle. Final concentrations of ammonia and phosphate were measured in the solution. Portions of the solid was colle cted on a filter paper, dried, re-dissolved in deionized water and analyzed for am monium and phosphate concentration for analysis of struvite content. 5.2.2 Phosphorus Release and Sludge Settling Aerobic digester sludge was collected and put in beakers to evaluate sludge settling and collect phosphate release sa mples. As a clear interface between settling sludge and Â“clearÂ” supernatant dev eloped, the volume that the sludge occupies was recorded with time. During settling samples of supernatant are collected at various times with a syringe and filtered using a 0.45 micron glass
59 fiber syringe filter. Supernatant tota l phosphate concentrations are analyzed and dissolved concentration of total phosphate is plotted versus time.
60 6 Results and Discussion 6.1 General Significant variability in samples was obser ved. The inconsistencies in plant influent and operating conditions were apprec iated during this study. It should be noted that the Largo Treat ment Plant and Howard Curren Treatment Plant employ different digestion systems (aer obic vs. anaerobic) but also different activated sludge and nutrient removal systems in the liquid side of the plant. The BPR and BNR system at Largo generates a WAS of different composition than that of Howard Curren which uses a high purity oxygen aeration system. However, collection of primary sludge at each of the plants is very similar. 6.2 Resource Partitioning Betw een Solid and Liquid Streams Worksheet 1 attached in A ppendix A-1 outlines the calculation methodology and input parameters for estimation of the re source partitioning between the solid and liquid streams of the treatm ent plant. Results are shown in Table 2. Only one sample was collected of prim ary effluent for input into this calculation and the results are displayed on Worksheet 1. Mean values were used for input in to Worksheet 1, taken from the summary of analytical data, Tables 3 and 4.
61 Table 1: Summary of Res ource Partitioning Results Partitioning of resources in both the Howard Curren Treatm ent Plant and the Largo treatment plant show that although handling a very small percentage of the flow rate, the solids side of both facili ties handle a significant portion of the nutrient mass flow (see Table 2). Phos phate partitions into the solids stream in higher proportions than nitr ogen, with over 90% of the phosphate in the solids stream and 38%-45% of the nitrogen. The high percentage of nutrients routed to Â“solidsÂ” was expected from literature values and solids flow analysis and only one sample was collected from each treatment plant for confirmation. This result encouraged continued investigations in to the solids side of the plant for evaluation of the resource recovery pot ential of the two digester systems plus opportunities for further resource reco very from post digestion sludge handling. 6.3 Fate of Resources in Aerobic and Anaerobic Digesters Tables 2 and 3 provide a summary of all analytical data collected during the investigation with statistical analysis. From this data and from flow rate data provided by the plant operations, ca lculations and mass balances were PlantStreamFlow %Nitrogen %Phosphate % Solids 2.33893 Liquids97.7627 Solids0.845122 Liquids99.255 22 Howard Curren Largo
62 conducted for the fate of nitrogen and phos phorus at the Howard Curren and Largo Treatment Plants. The initial intent for mass balancing was to collect samples and flow rates in and out of the digesters and conduct a mass balance on a daily basis. However during the investigation, it was learned that flow out of the digesters to solids handling at each of the facilities did not match the daily input. This is most prevalent at Largo, wher e primary sludge (PS) and waste activated sludge (WAS) are pumped to the digester over the weekends, but no effluent sludge (ES) is taken for processi ng into biosolids. During the week, more ES is taken from the digester than PS+WAS put in to make up for the weekends. Therefore, daily flow volumes proved to be an insufficient length of time to ensure equalized flow and no Â“accumulationÂ” term in the mass balance. It was decided to average flow rates and average analyt ical constituent concentrations and conduct one mass balance over the entire study period for each constituents. Worksheets in Appendix A-1 provide the mass balance calculations. All analytical data is taken from mean val ues, provided on Tables 2 and 3. Flow data represents the average daily flow volu me for PS, WAS, and ES provided by plant operations. Figures 15 through 18 in provide graphic representations of the fate of the resources thr ough the digestion system.
63 Table 2: Cumulative Analytical Da ta for Anaerobic Digester at HFCAWTP 17 Dec12 Jan30 Jan25 FebCumulative17 Dec12 Jan30 Jan25 FebCumulative17 Dec12 Jan30 Jan25 FebCumulative mean 4318.534.57342282149.4193.136224715521188.1116022231531 sd3.830.474.616.7237.240.9717.5459513.076.2240.72.42495 cv0.090.030.130.230.540.030.010.090.120.380.010.010.040.000.32 n222242222422224 mean 16641370136512151403.5031003589330029703239.7526992652321028052841.50 sd31.313421.221.21883303880127.3270397313063.6254 cv0.020.100.020.020.130.110.110.000.040.080.150.120.000.020.09 n222242222422224 mean 31.8105100.2579131223.5276.25210218309219.75249 sd0.51NA1.06410.51NA1.7774NANA1.0652 cv0.016038NA0.010.520.003893NA0.0064070.35NANA0.0040.21 n212321231123 mean 166511782385134316432883318048704915396217051715261517931957 sd21995.5NA39535753566NA14110813397.07NA11440 cv0.130.08NA0.030.330.260.18NA0.030.270.200.00NA0.010.23 n122122122 mean 53411602220130522201896318824351096192419761665 sdNANANA852NANANA672NANANA494 cvNANANA0.65NANANA0.28NANANA0.30 n111311131113 mean 7230065000114700840006360043050855006405031100276903240030397 sdNANANA26836NANANA21229NANANA2432 cvNANANA0.32NANANA0.33NANANA0.08 n111311131113 mean 28,95735,89349740381483818530,60341,28539440380503734519,93318,042195502085519595 sd2085132210479778640120996257166246866962201562201171 cv0.070.040.020.030.230.040.020.0010.040.130.030.010.010.010.06 n332343323433234Waste Activated SludgeEffluent Sludge Primary SludgeDissloved Nitrogen (mg/L filtered supernatant) Total Nitrogen (mg/L total sludge) Dissolved Phosphate (mg/L filtered supernatant) Total Phosphate (mg/L total sludge) Dissolved COD (mg/L) Total COD (mg/L) Total Suspended Solids (mg/L)
64 Table 3: Cumulative Analytical Da ta for Aerobic Digester at LGAWTP 17 Dec8 Jan30 Jan23 FebCumulative17 Dec12 Jan29 Jan23 FebCumulative17 Dec12 Jan30 Jan23 FebCumulative mean 3537.132.749.539924.104112.72.924.910 sd0.210.620.9380.110.020.1040.1800.10.1210 cv0.010.020.030.060.190.010.010.020.001.020.020.000.030.001.00 n222242222422224 mean 3641015645670777977848.2960920926758951.4810690802 sd337121.242.4207302.97028.357246584.914.1111 cv0.090.070.030.060.270.030.000.000.030.060.030.070.100.020.14 n222232222422224 mean 8.296.5735914966.5110.5109173101.5207.5161 sd1.5NANA464.8NANA416.3NANA54 cv0.18NANA0.770.03NANA0.380.04NANA0.34 n21141142114 mean 9310419106908801505125119411985167110741223173112001307 sd030NANA17778.527NANA354217NANA290 cv00.03NANA0.200.050.02NANA0.210.020.01NANA0.22 n2113201142114 mean 306503830546148334074167567098 sdNANANA265NANANA64NANANA60 cvNANANA0.48NANANA0.88NANANA0.62 n111311131114 mean 90164510033500393001362012975151801392512112179401857016207 sdNANANA18421NANANA1134NANANA3561 cvNANANA0.47NANANA0.08NANANA0.22 n111211131114 mean 579828,77517930190882193110,497942312950115951111692121243312540900010796 sd2225036369799419123145142131510323189283641954 cv0.040.020.040.050.430.010.020.0010.020.140.040.020.020.010.18 n332333323433234Primary SludgeWaste Activated SludgeEffluent SludgeTotal Suspended Solids (mg/L) Dissloved Nitrogen (mg/L filtered supernatant) Dissolved Phosphate (mg/L filtered supernatant) Total Phosphate (mg/L total sludge) Dissolved COD (mg/L) Total COD (mg/L) Total Nitrogen (mg/L total sludge)
65 The following discussion represents a comp arison of the fate of the various resources through the aerobic and anaerobic digestion processes and an evaluation of the resource recovery pot ential. It is noted that there was significant variability in the data collect ed. The statistical analyses shown on Tables 2-3 show a high coefficient of va riation for several of the analyses. The majority of the variability is attributed to va riations in the plant. Daily fluctuations in influent flow and concentrations ar e a well established factor in sewage treatment. Additionally instantaneous constituent concentrations can also fluctuate significantly based on changes in industrial use inputs. The intention of this study was to provi de a strict accounting of the fate of the resources through the two digester systems in order to evaluate the resource recovery potential. During the data analysis it was realized that a highly accurate mass balance would require an enormous amount of sampling, not feasible for this study. However, despite some of the high coefficients of variations,
66 Figure 14: Nitrogen Balance at Anaerobic Digester Figure 15: Phosphate Bala nce at Anaerobic Digester
67 valuable discussion comparing the fate of resources in the aerobic and anaerobic digester can be facilitated. 6.3.1 Nitrogen The most distinct differences between the aerobic and anaerobic system are in the fate of nitrogen through the digesters. In the aerobic system, there is a net solidification of nitrogen. At Largo, 24 kg/day of dissolved nitrogen enter the digester and only 10 kg/ day of dissolv ed nitrogen leave. The difference is incorporated into solids in the effluent. This contrasts sharply to the nitr ogen balance for the anaerobic system at Howard Curren where dissolved nitrogen influent is approximately 229 kg/ day but 2401 kg/day dissolved nitrogen exits t he reactor. The percentage of solid phase nitrogen drops significantly, from 96% solid entering the reactor to 60% exiting. The majority of the liqui d phase nitrogen is in the ammonia form. This ammonia stream returning to the head of the plant repr esents a signifant percentage (approximately ) of the plantÂ’s influent nitrogen loading estimated between 8,000 and 10,000 kg to tal nitrogen per day.
68 Figure 16: Nitrogen Balance at Aerobic Digester Figure 17: Phosphate Bala nce at Aerobic Digester
69 6.3.2 Phosphorus The fate of phosphate in the two digesti on systems appears to be similar. Both digestion systems lose phosphate from solid to liquid. However, the majority of the phosphorus in both system s enters in solid form (94-95%) and leaves in solid form (88%). 6.4 Whole Plant Mass Balancing By using plant data and the mass balance data discussed in section 6.3 above, we can also gain an understanding of the fa te of the nutrient resources through the treatment plant in order to discuss the current process and the Â“recoverabilityÂ” of nutrients through the treatment plant. Figure 18 shows a material balance with the system boundar y around the entire treatment plant and a general accounting of the partitioning of resources through the various effluent streams in the pl ant: discharge in the liquid effluent, incorporation into solids, and gaseous release. In both plants, a majority of the nitrogen is disc harged through gaseous release to dinitrogen gas. This is accomplish ed through the Biological Nitrogen Removal processes. Nitrogen is lost to t he atmosphere and this loss represents an opportunity for process change and re covery of the resource.
70 The fate of phosphorus through the treatment plant is in sharp contrast to the nitrogen. Because there are no gas eous forms of phosphorus, all influent phosphorus will leave the plant either in the liquid effl uent or in the processed solid. Howard Curren does not utilize any phosphorus removal technology because the plant has a variance and no phosphate discharge limit (due to high background concentration in receiving wa ter). Approxim ately 20% of the phosphorus is discharged in the effluent where at Largo al most all of the phosphorus leaves the plant in a recover ed form, in the processed solids. Largo accomplishes this recovery throug h the phosphate accumulating organisms (PAO) which hyper accumulated phosphat e into cellular compounds. At Largo, there have been problems wit h overloading the BNR system with phosphorus and exceedances of the dischar ge limits. The mass balances shown in Figure 18 do not show internal recycle. Maintaining phosphate in its solid form in PAO through the digestion process is a component to reducing recycle of phosphate from the digester back into the BNR system. So, although most phosphate at Largo should eventually be recover ed as solid even if it is released in the thickening system, the release of phosphate at this stage may contribute to higher percentages of phosphate leaving the plant in liqu id phase and potentially incurring discharge fines on the plant.
71 6.5 Phosphorus Release and Sludge Settling vs. Time Phosphate release and sludge settling were analyzed as aerobically digested sludge from the Largo AWTP was allowed to settle in condition s similar to the gravity thickener employed at the pl ant. Figures 19 through 24 provide data collected during six trials. Phosphorus concentration and sludge settling is plotted vs. time. Dissolved oxygen conc entrations were monitored during three of the six trials, with results displayed in Figures 25 through 27. During trial 1, the effluent sludge was transport ed from the plant to the l aboratory, thus the first phosphate analysis was conducted at 75 minut es. It was realized that phosphate data during the first 75 minutes would be critical and trials 2 through 4 were conducted at the plant so that sample collection could begin immediately. Tables summarizing the phosphate concent rations and calculated release rates for the trials over various time period are included as Table 3. An analysis of the phosphate release significance compar ed to plant and digester phosphate loading is included as Table 4.
72 Figure 18: Nutrient Mass Ba lance for Treatment Plants
73 Figure 19: Trial 1 Phosphor us Release at Largo AWTP Figure 20: Trial 2 Phosphor us Release at Largo AWTP Sludge Volume (Percentage of Total Volume) Total Phosphate (mg/L)Time (minutes)Phosphorus Release and Sludge Settling vs. Time Trial 1 Sludge Volume (Percentage of Total Volume) Total Phosphate (mg/L)Time (minutes)Phosphorus Release and Sludge Settling vs. Time Trial 2
74 Figure 21: Trial 2 DO Analysis at Largo AWTP Figure 22: Trial 3 Phosphor us Release at Largo AWTP Dissolved Oxygen (ppm) Ortho Phosphate (mg/L)Time (minutes)Total Phosphate Concentration and DO vs. Time Trial 2 Sludge Volume (Percentage of Total Volume) Phosphate (mg/L)Time (minutes)Phosphorus Release and Sludge Settling vs. Time Trial 3
75 Figure 23: Trial 3 DO Analysis at Largo AWTP Figure 24: Trial 4 Phosphorus Release at Largo AWTP Dissolved Oxygen (ppm) Phosphate (mg/L)Time (minutes)Total Phosphate Concentration and DO vs. Time Trial 3 Sludge Volume (Percentage of Total Volume) Phosphate (mg/L)Time (minutes)Phosphorus Release and Sludge Settling vs. Time Trial 4
76 Figure 25: Trial 4 DO Analysis at Largo AWTP Figure 26: Trial 5 Phosphor us Release at Largo AWTP Dissolved Oxygen (ppm) Phosphate (mg/L)Time (minutes)Total Phosphate Concentration and DO vs. Time Trial 4 Sludge Volume (Percentage of Total Volume) Total Phosphate (mg/L)Time (minutes)Phosphorus Release and Sludge Settling vs. Time Trial 5
77 Figure 27: Trial 6 Phosphor us Release at Largo AWTP Sludge Volume (Percentage of Total Volume) Total Phosphate (mg/L)Time (minutes)Phosphorus Release and Sludge Settling vs. Time Trial 6
78 Table 4: Summary of Phosphorus Release Trials Table 5: Analysis of Phosphate Release Significance The objective of this portion of study is to quantify the phosphorus release rates in a quiescent clarifier environment relative to the rate of sludge settling in order to further an understanding on how to maximize recovery of phosphate as solid and maximize sludge dewatering through the cl arifier. Additionally, because the Largo Treatment plant only operates the clarifier and solids handling facility during the business week, special attention was paid to differences in the release rates at the end of the week when mean sludge age would be significantly lower than at the beginning of the week. TrialDate Day of Week Starting Concentration Start Time Ending ConcentrationEnd Time Change in Concentration Elapsed TimeLinear Slope (mg/L)minutes(mg/L)minutes(mg/L)minutes(mg/L minute)223 FebMonday1955199.5454.5400.11 327 FebFriday95.25512245*26.75400.67 42 MarMonday203.255212458.75400.22 58 MayFriday24.5541.754217.25370.47 611 MayMonday49.25558.75459.5400.24 extrapolated Average Release Rate for t = 5 45 min Average Starting Concentration Average Concentration after 45 min settling (based on Release Rate) Supernatant Flow Rate (75% of sludge rate) Daily PhosphateMas s Released Daily Return Load (Sent to Head of Plant) Daily Mass Released/ Total Digester Load Daily Return Load/ Total Plant Loading(mg/l minute)(mg/L)(mg/L)(L/day)(kg/day)(kg/day)(%)(%) Monday 0.19149.2157.77826476.7123.40.5%9.0% Friday0.5771.296.778264720.075.71.6%5.5%
79 Three of the trials included evaluation of dissolved oxygen levels after removal from the aeration. The resu lts in the DO testing are consistent, with a very rapid drop in DO concentration, indica ting microbial oxygen utilization. There is significant variability in the init ial concentration of total phosphate and in the rate of release when comparing all th ree Â“end of the weekÂ” trials, Trial 1, 3 and 5. Similarly, the Â“beginning of the weekÂ”, Trials, 2, 4 and 6 trials show significant variability in both initial phos phate concentration and release rates. However, it is noted that in the fi rst 45 minutes the two highest phosphate release rates are on the two Friday sa mples and the lowest release rate. The most useful comparisons to draw anal ysis come from comparison of the two pairs of trials which span a weeke nd. Trials 3 and 4 surrounded a single weekend and Trials 5 and 6 su rrounded a single weekend. In both of these pairs, t he initial phosphate concentration on Friday was lower than on Monday and the initial rate of release on Friday was higher than on Monday. These results are consistent with two expectations gained from the literature. First, PAOÂ’s can rel ease phosphate when exposed to extended aeration and second, PAOÂ’s which have not been exposed to extended aeration will rapidly release phosphat e when stressed for oxygen as an electron acceptor.
80 In the context of the sludge settling rate, it appears that in all trials the sludge reached its maximum settling volume prior to 1 hour. This observation is made by visual inspection of the sludge settling curves. Therefore if residence time in the clarifier can be held to less than 1 hour, the sludge will reach its maximum settling point and analysis of the total phosphate release rate for the time 0-45 minutes should be the most significant. Table 5 provides an analysis of average phosphate release rates for Friday and Monday between t=5 minutes and t=45 minute s with a calculation of the total mass of phosphate released per day at the Largo Treatment Plant plus an estimation of the total mass of phosphat e returned to the head of the plant in digester supernatant. Based on the estimations made in th is investigation, it appears that approximately 75 to 123 kg/day of dissolved total phosphate is returned to the plant headworks per day from diges ter supernatant. This represents approximately 5.5 to 9 perc ent of the total plant daily phosphorus loading, possibly significantly more on Mondays when the solids digestion facility is initiated. In the broader context of aerobic di gestion and phosphate recovery as a resource, the total phosphate mass released during a 45 minute sludge settling time is between 6.7 and 20 kg/ day. This represents 0.5% 1. 6% of the total
81 load of phosphate sent to the digester. In other words, even with the sludge settling release, the digested material retains 98.4% to 99.5% of the total phosphate in the solid form as biosolids. The total load of phosphate sent to the head of the plant fr om the digestion system may be a significant percentage (5.5 %-9%), but little of this phosphate is released during post digestion sludge handling. The A2O system followed by aerobic digestion appears to recove r phosphate at the facility with good efficiency. There is some recycling and Â“loopingÂ” of phosphorus through the system, but closing the majori ty of this loop would requi re a modification to the digestion process rather than to post digestion sludge handling. 6.6 Struvite Precipitation from An D Filtrate: Phosphorus Recovery The initial area of investigation duri ng this study was on the ability to add phosphate to actual filtrate and then reco ver the phosphate in a re-marketable form. The addition of phosphate is necessary to recovery ammo nia, but recovery of the phosphate is crucial to the financ ial and regulatory f easibility of the process. Many trials were conducted in order to quantif y the potential for struvite precipitation using filtrate from the Howard Curren Tr eatment Plant. Results of phosphate analysis are presented in Table 4 and illustrated graphically in Figure 28. A scanning electron micrograph of struvite crystallized during this
82 investigation is shown as Figure 29 and solids analysis by X-ray diffractive analysis is shown in Figure 30. Data in Table 6 analyzes the phosphate removal in several manners each resulting from a comparison of the final concentration of ortho-Phosphate in the liquid phase after precipitation with some initial concentration. The three initial concentrations used are the initial phos phate concentration in the supernatant solution, the total ortho-Phosphate conc entration after addition of phosphate, and the concentration which was added. T able 4 shows promising results for all three analyses. In each of the trials all added phosphate was removed from solution plus there was removal of phosphate originally in the solution. Figure 28 provides a graphical look at the phosphate concentrations during the batch reactions. Large quantities must be a dded to facilitate precipitation, but the final concentration is below the initial in each batch. Analysis of the product by Energy Diffr active Spectrophotometry (EDS) shown in Figure 30 shows a high phosphorus concentration in the solid product.
83 6.7 Evaluation of Struvite Produc tion Using Aeration as pH Control Once it is displayed that phosphate coul d be recovered if added, further financial analysis (discussed in Section 6.7) indica ted that reducing costs for pH control would also be crucial to the feasibility of the struvite process. The objective of aeration is to reduce the acidity of the solution and reduce the chemical input for pH adjustment without stripping ammonia. The potential for ammonia stripping is first analyzed theoretically using mass tr ansfer relationships. Our objective to facilitate carbonic acid stri pping without ammonia stripp ing is assisted due to the fact that ammonia is in equilibrium with the gas phas e in its deprotonated form, abundant at high pH, and carbonic acid is in equilibrium with the vapor phase in its protonated form, abundant at low pH.
84 Table 6: Summary of Phosphate Removal Results Struvite Batch # Initial Phosphate Phosphate after Addition (Total) Dissolved Phosphate after Precipitation (Final) (Initial Final)/ Initial (Total Final)/Total (Total Final)/(Total Initial) #mg/Lmg/Lmg/L%%% 13347.5161523632%85%109% 14250398515538%96%103% 15204224176.263%97%106% 162042610169.217%94%101% 1720419997364%96%107% 1820422409354%96%105% 195054039176.565%96%109% 20440513016163%97%106% 21440424021950%95%106% 22400391517357%96%106%
85 Figure 28: Phosphate Concentrations During Struvite Precipitation Figure 29: SEM Image of Crystallized Product (courtesy of Russell Ferlita) Supernatant ortho Phosphate Concentration (mg/L)Struvite Batch #Dissolved Supernatant ortho Phosphate During Struvite Precipitation Process
86 Figure 30: Solid Product Analysis by EDS (courtesy of Russel Ferlita) Figure 31: Prediction of Mass Trans fer for Carbon Dioxide and Ammonia 00.511.5 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 020406080100120initial concentration carbonic acid (mg/L)mass transfer (mg/ hour)initial concentration ammonia (mg/L)Concentration vs. Predicted Diffused Aeration Mass Transfer Equation from Matter-Muller et al. 1981 ammonia carbon dioxide
87 Our objective is further facilitated by a difference in HenryÂ’s constants between the two compounds of over two orders of magnitude (0.006 and 1.6 for ammonia and carbon dioxide respectively). Thus, if our aeration occurs at a low pH the driving force for carbonic acid stripping should be at a maximum and the driving force for ammonia stripping should be at a minimum, given the total concentrations of each species in our so lution. Some preliminary analysis was conducted using mass transfer relationships taken from literature (Matter-Muller et al. 1981). Typical oxygen transfer rates for diffused bubble aeration were taken also from the literature (Gilot et al. 2005). Example calculations using the Matter-Muller et al. (1981) re lationship is provided in Appendix E-3. Figure 31 shows a plot of mass transfer rates vs. a range of concentrations for ammonia and carbonic acid which are expected. The result cl early shows that aeration should remove carbonic acid far more rapidly than ammonia, given the pH and concentration range expected. Appendix B-1 shows plots ammonium and pH vs. time for each of the struvite batch reactions. Also in Appendix B-1 ar e plots of ammonium vs. pH for each of the batch reactions. These trials each us e of combination of aeration followed by lye addition for pH control. Aeration at a power of 0.036 Watts is conducted first for a set duration and then NaOH added to co mplete the pH adjustment to above 9.75.
88 The ammonia vs. pH plots also contain a plot of ammonium ion vs. pH under theoretical acid/ base speciation conditions. This plot allows for reference of the measured ammonium concentration. Table 7 displays a summary of data co llected and calculated during aeration and lye struvite preparation trials. Â“AcidityÂ” is defined as the concentration in millieqivalents per liter of a strong base (NaOH) required to raise a solution pH to a set point. For this investigation, the set point is pH 9.75, determined by tr ial and error for struvite precipitation. Figure 32 shows a plot of post aeration so lution acidity vs. specific aeration energy for the aeration trials. Directly proportional to the acidity is a fa ctor termed Â“specifi c alkaline additionÂ” defined here as the mass of NaOH per liter of supernatant needed to raise pH to the operational set point (9.75) for struvite precipitation. This specific alkaline addition is a more useful term than acidity for financial analysis and for operational calculations. Figure 33 shows a plot of specific alkaline addition vs. aeration energy and also includes a pl ot of ammonia reduction vs. aeration energy.
89 Table 7: Summary of Struvi te Precipitation Analysis Struvite batch Initial Ammonia to Phoshpate Ratio Ammonia Reduction Phosphate Reduction Aeration pH Adjustment Chemical pH Adustment Calculated Acidity (after Aeration) recovered solids molar ratio Recovered Solid Phosphate "Purity" Recovered Solid Ammonium "Purity" #[NH %% kWh/ L Supernatant g NaOH/ L supernatant mequivalents [OH ] %% 134.7380.3584.0305125not analyzed82%57% 1415.886.3796.5906.721681.356%61% 151183.996.606.931731.0697%43% 1612.893.3193.520.0122.4601.1770%39% 179.7895.5396.350.01051.65412.9440%58% 1810.7891.8595.850.01031.6402.3149%49% 197.990.395.630.00722.88720.8944%66% 2011.6987.8696.860.00823.45861.2745%56% 218.492.4494.830.0242.23570.8563%45% 2211.589.71950.0242.27601.0382%44%
90 Figure 32: Acidity Analysis for Struvite Aeration Tests Figure 33: Alkaline Addition and Ammonia Reduction Analysis 00.0050.010.0150.020.0250.03Acidity vs. Aeration Energy Input 00.010.020.03 Specific Alkaline Addition and Ammonia Reduction vs. Aeration Energy Input
91 Another graphical repres entation of the differences in acidity affected by aeration is shown on Figure 34. A titration curve for each of the trials is plotted on the same graph. Referencing Table 7 for each trial and the specific aeration energy, it can be clearly seen that the trials with no aerat ion (#13-#15) have less steep titration curves than those with extended aeration (#19-#20). The results of the aerati on trials show a relationship between aeration energy input and acidity with very little affect on the reduction of ammonia in the final solution supernatant. Aerati on of the batch reaction solution clearly reduces the acidity and the specific al kaline addition needed to s ubsequently raise the pH to the struvite precipitation target. The acidity vs. aer ation energy curve shown in Figure 32 appears to be moving toward an a symptotic shape. This would make sense as there is a finite mass of carbonic acid (and perhaps other volatile compounds contributing to acidity) to be removed. Thus, some guidance from this study can be gained towards optimizi ng aeration energy to minimize reactor size and aeration electricity while maximi zing acidity reduction. For our small reactor size and tiny aeration power, approximately 0.5 hours of aeration resulting in approximately 0.010 Â– 0.012 kWh/ L of super natant appears optimal. It would be expected that full scale system s would achieve similar mass transfer rates with greater efficiency re sulting from larger blowers.
92 Figure 34: Titration Curves for Struvite Aeration Tests pHspecific alkaline addition (g NaOH/ L Supernatant)pH vs. Specific Alkaline Addition (Titration Curve)
93 While the reduction in acidity thr ough aeration is clearly demonstrated, confirmation of the retention of the amm onia resource has proven more difficult. Examination of the ammonium vs. pH plots clearly show that as the pH is raised, something is occurring other than simp le acid/ base speciation change from ammonium ion to ammonia. While our hope is that the di fference between the observed ammonium vs. pH pl ot is attributed to precip itation and recovery, there is the possibility that vola tilization has occurred. Characterization of the precipitate product can give an indication of the presence of struvite and the relative purity of t he product. Charac terization of the precipitated product was carried out by dissolving a known mass into deioinized (DI) water, reducing pH and m easuring phosphate and ammonium concentrations. Included in Table 7 ar e data which analyze the product. The molar ration of ammonium to phosphate shoul d be 1:1 in a pure struvite product. And, if a known mass of a pure struvite product is dissolved in DI water, the concentrations of ammonium and phosphate should be known. The Â“purityÂ” figures shown in Table 7 are the ratio of the observed concentration of ammonium or phosphate divided by the predicted concentration if the precipitate were pure struvite. The molar ratio of ammonium to phos phate observed in the re-dissolved precipitated solids indicates a significant variability. Several of the samples have
94 a molar ratio of ammonia to phosphate which gives an indication that the ammonia and phosphate present could be fr om struvite, while others do not. However, analysis of the concentrations of ammonia and phosphate in comparison to a Â“pureÂ” struvite product indi cate that there is a significant mass of solids present which are not struvite. The fact that ammonia is present in t he dissolved product is a great indication that ammonia recovery is occurring, but quantification of this recovery has not been accomplished. Inspecting the ammoni um vs. pH curves also gives an indication that ammonium is precipitati ng rather than volati lizing during aeration because pH remains low (near 6) during the aeration phase and the concentration of volatile ammonia (NH3) should be low at this pH. Comparison of the data between the aeration batches (#16-24) and the NaOH only batches (#13-15) does not show significant diffe rences in the ammonia: phosphate ration or the purity factors shown on Table 7. This gives further indication that ammonium is precipitating rather than volatilizing. Although the factors above provide some hope that the desired result, (struvite precipitation rather than ammonia volatilization) is occurring, a strict mass balancing of ammonia species before and after the precipitation would be the best route to quantify recovery. Howeve r, the batch reaction process used for the precipitation creates some physical difficulties in recovering the product.
95 Product was inefficiently recovered thr ough filtration onto glass fiber paper at high energy consumption. The phosphorus recovery struvite efforts explained in the literat ure typically use a fluidized bed reactor to accomplish the precipitation, as this configuration encourages crystal growth and eas e of the physical recovery of the precipitate. Further efforts in a feasibility study for struvite crystallization for ammonia recovery should move towards this configuration. 6.8 Financial Analysis of Struvite There has been a continuing adjustment on se veral levels of financial analysis which have directed the struvite precipitat ion research throughout this study. The first questions involved an analysis of the economic feasibility of adding large quantities of phosphate to the system, if it was recoverable, and if so would it make financial sense. It was demonstrat ed that, given current market prices struvite precipitation could not compete as an ammonia removal technology with the current nitrification/ deni trification process at Howard Curren and would likely be far behind more advanced BNR technologie s. The initial calculation showed that at current market prices struvite production less its resale value would cost the plant $3.07 per kg of influent nitr ogen while the BNR process would cost $1.08 per kg of influent nitrogen.
96 However, since the struvite process has a marketable product where the others do not, a financial analysis of the potential effect of a general rise in commodity prices. This effect was approximated by simply linearly scaling the current prices of all commodities and examining the effect on struvite feasibility. The results of this exercise showed that the struvite ac tually became less competitive. Figure 35 displays this result. Figure 35: Analysis of Increasing Comm odity Prices on Struvite Feasibility $$1.00 $2.00 $3.00 $4.00 $5.00 $6.00 $7.00 $8.00 020406080100120140Cost per kg-NH4NCommodity Price Increase (% increase from current)Compare Nitrogen Recovery with Removal Costs with Rising Commodity Prices Complete pH Adjustment with Sodium Hydroxide Struvite baseline
97 However, it was observed that the pH adjustment costs comprised a large portion of the struvite production costs as shown in Figure 36. Another level of financial analysis was conducted to evaluate the competitiveness of struvite precipitation if the pH adjustment costs could be reduced through aeration, while keeping a high ammonia recovery for resale. Figure 36: Assessment of Struvite Production Costs This analysis showed significant promise for the technology both in its immediate competitiveness and also for its potentia l to buffer against rising commodity prices (see Figures 37 and 38). Figure 37 compares the price per kg-N at Struvite Chemical Addition Costs for Howard Curren AWTP (Price per kg of influent ammonia with all pH adjustment by NaOH)
98 current prices if the ph adjustment cots can be reduced by varying percentages. Figure 38 shows that if pH can be cut to 40% of the full lye addition cost now, struvite production costs will not increase with rising commodity prices. This analysis encouraged the further experimenta tion into aeration and pH reduction described above. Figure 37: Assessment of pH Chemical Reduction Effects $(2.00) $(1.00) $$1.00 $2.00 $3.00 $4.00 0 0.2 0.4 0.6 0.8 1Cost per kg-NPercentage of Calculated NaOH AdditionCompare Nitrogen Recovery with Removal Costs if pH addition is reduced from Observed Need Struvite baseline
99 Figure 38: Rising Commodity Prices with Reduced pH Cost A final effort at financial analysis has been conducted, incorporating all of the data collected during this investigati on, including optimized aeration energy, specific alkaline addition, Howard Curr en plant flow data, and estimates of phosphate and ammonia recovery percentages. Input parameters are shown in Table 8. The results of this analysis are shown in Figure 39. It should be noted that ammonia and phosphate recovery percentages were estimated from ammonia and phosphate removal rates observed in this study and recovery rates from previous literature. Q uantification of the recovery rates in this study has not been accomplished. The financial feas ibility is highly dependent on high recovery rates. The estimate in Figur e 39 uses 90% recove ry for both ammonia $$0.50 $1.00 $1.50 $2.00 $2.50 $3.00 050100150Cost per kg-NCommodity Price Increase (% increase from current)Compare Nitrogen Recovery with Removal Costs with Rising Commodity Prices pH Adjustment costs reduced to 40% Struvite baseline
100 and phosphate and the financial feasibility of st ruvite recovery would be seriously compromised with significant r eductions in this estimate. Table 8: Input Parameters for Financial Analysis ValueUnit 1000mg/ L 100mg/ L 200mg/ L 1000000L/ day 0.5$/kg Mg 0.5$/kg Phoshpate 0.46$/kg NaOH 1.6g NaOH/ L Supernatant 0.9percentage 0.9percentage 1$/kg N +P 0.11$/kWh 1.5$/ gallon 0.005kWh/ L Supernatant resale value of struvite Electricity Costs Methanol Costs Parameter Aeration Power Requirement Phosphate Unit Cost pH Adjustment Unit Cost pH Adjustment Requirement phosphate recovery ammonia recovery Ammonia Concentration Magnesium Concentration Phosphate Concentration Supernatant Flow rate Magnesium Unit Cost
101 Figure 39: Resulting Financial Analysis from this Study (with Nitrogen and Phosphorus Recovery Values from Literature) Struvite Chemical Addition Costs for Howard Curren AWTP (Price per kg of influent ammonia with pH adjustment by aeration and NaOH)
102 7 Conclusions From the literature review and data colle cted during this in vestigation, the following conclusions are made. As would intuitively be expected, the solids side of the typical waste water treatment plant is the place to look for resource recovery. Given the high flow rates and dilute resource concentrations on the liquid side, it is expected that it will be a long time before resource re covery considerations will compete with current removal technologies. Firm conclusions from the phosphate re lease study at the Largo Treatment Plant are difficult due to the enormous amount of variables which cannot be controlled when analyzing a treatment plant. However some observations can be made. Phosphorus release rates during the end of the week are slightly more rapid than the phosphorus release ra tes at the beginning of the week. This may be attributable to a higher percentage of PAO who have experienced short retention time and have retained the r apid phosphorus release characteristic acquired during the aerobic/ ana erobic cycling in the A2O system
103 As sludge remains in the digester ov er the three day weekend period, it appears that phosphorus is released. This release may be a result of extended aeration of the PAOÂ’s and release of the Poly-P during the extended periods with ample electron acceptors. Total phosphorus release during the anaerobic settling and thickening process during post digestion sludge hand ling at the Largo AWTP does not appear to release a significant perc entage of the total phosphate in the digester sludge. Based on the data collect ed in six trials, 98.4-99.5% of the total phosphate is retained in the solid form during the settling process. The sludge at the Largo Tr eatment plant appears to settle to its fullest extent within 45 minutes to 1 hour. Although phosp horus release rates are relatively slow at this time, phosphate recovery can be maximized by limiting retention time in the settler to a minimum. Aeration has a clear effect on acidit y and therefore the quantity of lye needed to raise the pH of a supernatant bas ed struvite precipitation solution. Phosphate was added to the batch reacti on solutions in order to increase the phosphate molar concentration to equal the ammonium concentration. Based on analysis of dissolved phosphate the conclusion of the batch tests, it appears that the added phosphate was removed and available for recovery as a precipitate. Struvite recovery could become financia lly feasible if the ammonia recovery rate can be kept high, while reducing lye addition for pH contro l. Further, if
104 struvite can be made financially feasible now, it will buffer plant operations against rising commodity pric es in the future. Further study is needed to quantify the ammonia recovery and ammonia volatilization during the aeration process. There were data several indicators collected during this investigation that ammonium is precipitating rather than volatilizing during the batch reaction process, however a strict material balance on all ammonium species would be desirable.
105 References Ali, M. I., Schneider, P. A., and Hudson, N. (2005). "Thermodynamics and Solution Chemistry of Struvite." Indian Institute of Sciences 85, 141-149. Babbitt, H. E. (1953). Sewage and Sewage Treatment John Wiley & Sons, Inc., New York. Battistoni, P., Fava, G., P.Pavan, Musacco, A., and Cecchi, F. (1997). "Phosphate removal in anaerobic liquors by struvite crystallization without addition of chemicals: preliminary results." Water Research 31(11), 29252929. Booker, N. A., Priestley, A. J., and Fraser, I. H. ( 1999). "Struvite Formation in Wastewater Treatment Pl ants: Opportunities for Nutrient Recovery." Environmental Technology 20, 777-782. Brdjanovic, D., Slamet, A., Loosdrecht, M. C. M. V., Hooijmans, C. M., Alaerts, G. J., and Heijnen, J. J. (1998). "Impact of Excessive Aeration on Biological Phosphorus Removal from Wastewater." Water Research 32(1), 200-208. Celen, I., and Turker, M. (2001). "Recovery of Amm onia as Struvite from Anaerobic Digester Effluents." Environmental Technology 22, 1263-1272. Doyle, J. D., and Parsons, S. A. ( 2002). "Struvite formation, control and recovery." Water Research 36, 3925-3940.
106 Driver, J., Lijmbach, D., and Steen, I. (1999). "Why Recover Phosphorus for Recycling and How?" Environmental Technology 20, 651-662. Fujimoto, N., Mizouchi, T., and Togami, Y. (1991). "Phosphorous fixation in the sludge treatment system of a bi ological phosphate removal accomplishments and needs." Water Research 25(12), 1471-1478. Gilot, S., Capela-Marsal, S., Roustan, M., and Heduit, A. (2005). "Predicting oxygen transfer of fine bubble diffu sed aeration systems-model issued from dimensional analysis." Water Research 39, 1379-1387. Kuba, T., Loosdrecht, M. C. M. V., and Murnleitner, E. (1997). "Kinetics and Stoichiometry in the Biological Ph osphorus Removal Process with Short Cycle Times." Water Research 31(4), 918-928. Matter-Muller, C., Gujer, W., and Giger, W. (1981). "Transfer of Volatile Substances from Water to the Atmosphere." Water Research 15, 12711279. Mulkerins, D., Dobson, A. D. W., and Co lleran, E. (2003). "P arameters affecting biological phosphate remova l from wastewaters." Environment International 30, 349-259. Ohlinger, K. N., Young, T. M., and Sc horeder, E. D. (2000). "Post Digestion Struvite Precipitation Usi ng a Fluidized Bed Reactor." Journal of Environmental Engineering 126(4), 361-368. Pitman, A. R. (1998). "Managem ent of Biological Nutrie nt Removal Plant Sludges Change the Paradigms?" Water Research 33(5), 1141-1146.
107 Rittmann, B. E., and McCa rty, P. L. (2001). Environmental Biotechnology: Principles and Applications McGraw-HIll, New York, NY. Shu, L., Schneider, P., Jegatheesan, V ., and Johnson, J. (2005). "An economic evaluation of phosphorus recovery as struvite from digester supernatant." Bioresource Technology 97, 2211-2216. Siegrist, H. (1996). "Nitrogen Removal fr om Digester Supernatant Comparison of Chemical and Biological Methods." Water Science and Technology 34(1-2), 399-406. Steen, I. (1998). "Phosporus availability in the 21st century Management of a non-renewable resource." Phosporus and Potassium (217). Ueno, Y., and Fuji, M. (2001). "Three Year s Experience of Operating and Selling Recovered Struvite from Full-Scale Plant." Environmental Technology 22, 1373-1381. Wang, J., Burken, J. G., Jackie Zhang, and Surampalli R. (2005). "Engineered Struvite Precipitation: Impacts of Component-Ion Molar Ratios and pH." Journal of Environmental Engineering 131(10), 1433-1440. Worrell, E., Phylipsen, D., Einstein, D. and Martin, N. (2000). "Energy Use and Energy Intensity of the U.S. Chemical Industry." U. D. o. Energy, ed., Environmental Energy Technologies Division, Ernest Orlando Lawrence Berkeley National Laboratory, University of California.
Total Plant Flow Total Primary Sludge Flow Total WAS Sludge Flow Total Flow to Solids Total Flow to Liquids Total Plant Flow Total Flow to Solids Total Flow to LiquidsFlow to Solids Flow to Liquids (g al/da y) (g al/da y) (g al/da y) (g al/da y) (g al/da y) ( L/da y) ( L/da y) ( L/da y) %% Howard Curren50,000,000196,923221,170418,09349,581,9071890000001580391.54187419608.50.8%99.2% Largo12,000,000156,964119,103276,06711,723,933453600001043533.2644316466.742.3%97.7% Resource Total Mass per Da y PE Total Mass per Da y PS Total Mass per DayWAS Total Mass/Da y Solids Total Mass/ Day Liquids Resource to Solids Resource to Liquids ( k g /da y) ( k g /da y) ( k g /da y) ( k g /da y) ( k g /da y) %% Nitrogen7216104527083753450845%55% Phosphorus2483122333124535-829122%-22% Nitrogen1839461417878142238%62% Phosphorus84252275312758993%7% Total EffPeff Total (mg/L)(kg/day) Nitrogen 38.5 7216 Phosphorus 13.25 2483 Nitrogen41.51839 Phosporus 19 842 Howard Curren Largo Calculation of Total Mass in Primary EffluentWORKSHEET 1: ResourcePartitioning Between Solids and LiquidHoward Curren Largo
Sample Date December Sample Location Howard Curren Notes Sample Flow Rate(gallons/day) Total Average Flow Rate (liters/ day) Total Suspended Solids (masssolid/ volumesludge) (mg/L) Total Liquid (volumeliquid/ volumesludge) MixedLiquor Total Nitrogen (mg/L) Supernatant Total Nitrogen(mg/L) Total Nitrogen (kg/day) Liquid Total Nitrogen(k g/day) Solid Total Nitrogen (kg/day) PS196,92374436938,1850.9618151404421045301015 WAS221,17083602337,3450.962655324024727091992510 ES423,233159982119,5950.98040528421531454724012145 Total PSIn(kg/day Nitrogen) Total WAS In(kg/day Nitrogen) Total ES Out(kg/day Nitrogen) Total GasOut (kg/dayNitrogen) Total Out/ Total In 1045 2709 4547 0 1.2 kg/daypercentage of influent Total NIn 3754 100% Total Dissolved NIn 229 6% Total Solid NIn 3525 94% Percentage ofEffluent Total NOut 4547 121% Total Dissolved NOut 2401 64% 53% Total Solid NOut 2145.4 57% 47% Total GasNOut 0 0% 0%Worksheet 2NitrogenBalanceatHCAWTPNITROGEN PARTITIONINGColor Scheme PlantData Lab Analysis Calculated ValueCALCULATIONOF NITROGENFLOW RATESSLUDGE SAMPLES GASSAMPLESNITROGEN BALANCES=
Sample Date Decembe r Sample Location Howard Curren Notes Sample Flow Rate(gallons/day) Total Average Flow Rate (liters/ day) Total Suspended Solids (masssolid/ volumesludge) (mg/L) Total Liquid (volumeliquid/ volumesludge) MixedLiquor Total Phosphate (mg/L) Supernatant Total Phosphate (mg/L) Total Phosphate (kg/day) Liquid Total Phosphate (kg/day) Solid Total Phosphate (kg/day) PS 196,923 744369 38,197 0.9618031643 79 1223 57 1166 WAS 221,170 836023 37,109 0.9628913962 210 3312 169 3143 ES 423,233 1599821 19,175 0.9808251957 249 3131 391 2740 Total PSIn(kg/day Phosphate) Total WASIn(kg/day Phosphate) Total ES Out(kg/day PO4) Total GasOut (kg/dayPO4) Total Out/ Total In 1223 3312 3131 0 0.69 kg/daypercentage ofinfluen t Total PO4In 4535 100% Total Dissolved PO4In 226 5% Total Solid PO4In 4310 95% Percentage ofEffluent Total PO4Ou t 3130.849188 69% Total Dissolved PO4Out390.7169002 9% 12% Total Solid PO4Ou t 2740.1 60% 88% Total GasPO4Ou t 00 %0 % Color Scheme PlantData Lab AnalysisWorksheet 3Phosphate BalanceatHCAWTPCalculated ValueGASSAMPLESPHOSPHATEBALANCES=PHOSPHATEPARTITIONING CALCULATIONOF PHOSPHORUSFLOW RATESSLUDGE SAMPLES
Sample Date 1/29/2009 Sample Location Largo Notes Sample Flow Rate(gallons/day) Total Average Flow Rate (liters/ day) Total Suspended Solids (masssolid/ volumesludge) (mg/L) Total Liquid (volumeliquid/ volumesludge) MixedLiquor Total Nitrogen (mg/L) Supernatant Total Nitrogen(mg/L) Total Nitrogen (kg/day) Liquid Total Nitrogen (kg/day) Solid Total Nitrogen (kg/day) PS 15696 4 59332 4 21,931 0.978069777 39 461 23 438 WAS 119103 450209 11,116 0.98888 4 926 4 417 2 415 ES 27606 6 1043529 10,796 0.98920 4 802 10 837 10 827 Total PSIn(kg/dayNitrogen) Total WASIn(kg/day Nitrogen) Total ES Out(kg/day Nitrogen) Total GasOut (kg/dayNitrogen) Total Out/ Total In 461 417 837 0 0.95 kg/da y percentage ofinfluent Total NitrogenIn 878 100% Total Dissolved NitrogenIn 24 3% Total Solid NitrogenIn 853 97% percentage ofeffluent Total NitrogenOut 837 95% Total Dissolved NitrogenOut 10 1% 1.2% Total Solid NitrogenOut 826.6 94% 98.8%WORKSHEET 4NitrogenBalanceatLGAWTPNitrogenPARTITIONINGColor Scheme PlantData Lab Analysis Calculated ValueCALCULATIONOF NITROGENFLOW RATESSLUDGE SAMPLES GASSAMPLESNitrogenBALANCES=
Sample Date 1/29/2009 Sample Location Largo Notes Sample Flow Rate(gallons/day) Total Average Flow Rate (liters/ day) Total Suspended Solids (masssolid/ volumesludge) (mg/L) Total Liquid (volumeliquid/ volumesludge) MixedLiquor Total Phosphate (mg/L) Supernatant Total Phosphate (mg/L) Total Phosphate (kg/day) Liquid Total Phosphate (kg/day) Solid Total Phosphate (kg/day) PS 15696 4 59332 4 21,931 0.978069880 59 522 34 488 WAS 119103 450209 11,116 0.98888 4 1671 109 752 49 704 ES 27606 6 1043529 10,796 0.98920 4 1307 161 1364 166 1198 Total PSIn(kg/dayPhosphate) Total WASIn(kg/day Phosphate) Total ES Out(kg/day Phosphate) Total GasOut (kg/day Phosphate) Total Out/ Total In 522 752 1364 0 1.07 kg/da y percentage ofinfluent Total Phosphate In 1274 100% Total Dissolved Phosphate In 83 6% Total Solid Phosphate In 1192 94% percentage ofeffluent Total Phosphate Out 1364 107% Total Dissolved Phosphate Out 166 13% 12.2% Total Solid Phosphate Out 1198 94% 87.8%Worksheet 5Phosphate BalanceatLGAWTPColor Scheme PlantData Lab Analysis Calculated ValueGASSAMPLESPHOSPHATEBALANCES=PHOSPHATEPARTITIONING CALCULATIONOF PHOSPHORUSFLOW RATESSLUDGE SAMPLES
114 Appendix B-1: Struvite Batch Reaction Plots Figure 40: Struvite #14 pH and Ammonium vs. Time pH Ammonium N (mg/L)Time (Minute)Struvite #14: pH and Ammonium vs. Time (pH adjustment: NaOH addition only)
115 Appendix B-1 (Continued) Figure 41: Struvite #14 Ammonium vs. pH Ammonium N (mg/L)pHStruvite #14: Ammonium vs. pH (pH adjustment: NaOH addition only)
116 Appendix B-1 (Continued) Figure 42: Struvite #16 pH and Ammonium vs. Time pH Ammonium N (mg/L)Time (Minute)Struvite #16: pH and Ammonium vs. Time pH adjustment: 30 minutes aeration, NaOH addition
117 Appendix B-1 (Continued) Figure 43: Struvite #16 Ammonium vs. pH Ammonium, N (mg/L)pHStruvite #16: Ammonium vs. pH pH adjustment: 30 minutes aeration, NaOH addition
118 Appendix B-1 (Continued) Figure 44: Struvite #17 pH and Ammonium vs. Time pH Ammonia N (mg/L)Time (Minute)Struvite #17: pH and Ammonium vs. Time pH adjustment: 30 minutes aeration, NaOH addition
119 Appendix B-1 (Continued) Figure 45: Struvite #17 Ammonium vs. pH Ammonium N (mg/L)pHStruvite #17: Ammonium vs. pH pH adjustment: 30 minutes aeration, NaOH addition
120 Appendix B-1 (Continued) Figure 46: Struvite #18 pH and Ammonium vs. Time pH Ammonium N (mg/L)Time (Minute)Struvite #18: pH and Ammonium vs. Time pH adjustment: 30 minutes aeration, NaOH addition
121 Appendix B-1 (Continued) Figure 47: Struvite #18 Ammonium vs. pH Ammonium N (mg/L)pHStruvite #18: Ammonium vs. pH pH adjustment: 30 minutes aeration, NaOH addition
122 Appendix B-1 (Continued) Figure 48: Struvite #19 pH and Ammonium vs. Time pH Ammonium N (mg/L)Time (Minute)Struvite #19: pH and Ammonium vs. Time pH adjustment: 15 minutes aeration, NaOH addition
123 Appendix B-1 (Continued) Figure 49: Struvite #19 Ammonium vs. pH Ammonium N (mg/L)pHStruvite #19: Ammonium vs. pH pH adjustment: 15 minutes aeration, NaOH addition
124 Appendix B-1 (Continued) Figure 50: Struvite #20 pH and Ammonium vs. Time pH Ammonium N (mg/L)Time (Minute)Struvite #20: pH and Ammonium vs. Time pH adjustment: 15 minutes aeration, NaOH addition
125 Appendix B-1 (Continued) Figure 51: Struvite #20 Ammonium vs. pH Ammonium N(mg/L)pHStruvite #20: Ammonium vs. pH pH adjustment: 15 minutes aeration, NaOH addition
126 Appendix B-1 (Continued) Figure 52: Struvite #21 pH and Ammonium vs. Time pH Ammonium N (mg/L)Time (Minute)Struvite #21: pH and Ammonium vs. Time pH adjustment: 60 minutes aeration, NaOH addition
127 Appendix B-1 (Continued) Figure 53: Struvite #21 Ammonium vs. pH Ammonium N (mg/L)pHStruvite #21: Ammonium vs. pH pH adjustment: 60 minutes aeration, NaOH addition
128 Appendix B-1 (Continued) Figure 54: Struvite #22 pH and Ammonium vs. Time pH Ammonium N (mg/L)Time (Minute)Struvite #22: pH and Ammonium vs. Time pH adjustment: 60 minutes aeration, NaOH addition
1000 mg/L g NaOH/L Superg NaOH/L 2.5 ML 2.5 M NaOH/L Super 100 mg/L 1.6 100 0.016 200 mg/L 1000000 L/ day 0.5$/kg Mg 0.5$/kg Phoshpate 0.46$/kg NaOH Compound Measured Concentration Molar Concentration Set Molarity (to highest) 0.016L 2.5 M NaOH/L (mg/L)(M)(M) 0.9percentageAmmonium10000.0560.056 0.9percentageMagnesium1000.004 1$/kg N +P Phosphate 200 0.002 0.11$/kWh 1.5$/ gallon 0.005kWh/ L Quantity of Compound Needed Quantity of Compound Needed Source CompoundTotal MW of Solid SourceMolarity of Liquid Source Moles Ion/ Mole Source Compound Mass of Solid Source Needed Volume Liquid Source Needed (M) (mg/L) (g/mole) (M) (mg/L) (L/L) 0 0 NH4Cl 53.4 NA 1 0 0 0.0514403291250 MGO 40 NA 1 2.057613169 NA 0.0534502925077.885% H3PO4 NA 15.2 NA NA 0.003516467 Ai Assessment of Howard Curren Anaerobic Digester Sludge Supernatant Assessment of Ammonia FlowpH Adjustment Requirement phosphate recovery ammonia recovery Electricity Costs Methanol Costs Phosphate Concentration Supernatant Flow rate Magnesium Unit Cost Phosphate Unit Cost pH Adjustment Unit CostFinancial Assessment of Struvite Feasibility Worksheet Page 1Aeration Power Convert g NaOH/L to L of 2.5 M NaOH/L SuperAssessment of Required Chemical Additions for Equimolar Concentrations and Struvite Precipitationresale value of struvite DATA INPUT Ammonia Concentration Magnesium Concentration Supernatant Flow Ammonia ConcentrationTotal Ammonia Mass Flow (L/day)(mg/L)(kg/day) 1,000,00010001000 Compound Total Mass NeededProduct Addition CostTotal Cost/ DayTotal Cost/ Year Chemical Addition Cost/ Kg NH4 (kg/day)($/kg)($/day)($/year)($/kg NH4) Ammonium0$0.50$0.00$0.00$0.00 Magnesium2057.613169$0.50$1,028.81$375,514.40$1.03 Phosphate3516.466605$0.50$2,068.51$755,006.07$2.07Assessment of Chemical Addition Costs for Precipitation of Struvite
pH ControlMolarity Addition RatioAdditionAddition MassAddition MassBulk NaOH CostTotal Daily Cost Chemical Addition Cost/ Kg NH4 Source(M)(L NaOH/ L)(Moles NaOH/L)(moles/day)(kg/day)($/lb)($/day)($/kg) NaOH2.50.0160.0440000 1600.0000$0.46$1,619.2$1.62aerationaeration cost $ /day $ /kg NH4 $ /day $ /kg NH4 $ /day $ /kg NH4 $ /day $ /kg NH4 k wH / L $ /kg N $2,068.51$2.07$1,028.81$1.03 $1,619.20$1.62$4,716.52$5.270.0050.55 Supernatant Ammonia Production Cost for StruviteExpenditure Expenditure Cost/ YearStruvite Value Ammonia Recovery Percentage Phosphate Recovery Percentage Valued Struvite Production Recovery valueTotal Cost/ DayTotal Cost/Kg NH4 (kg NH4/ day)($/ kg NH4)($/day)($/year)($/kg A+P)(kg NH4 + PO4/day)($/day)($/day)($/kg) 1000$5.825,816.52 $ 2,123,028.47 $ $190.00%90.00%5085.05,085.00 $ 731.52 $ 0.73 $ Supernatant Required Total Cost/KgAssessment of Total Chemical Addition Costs for Struvite Precipitation at Howard Curren AWTP Assessment of pH Control with Sodium Hydroxide to Facilitate Preciptiation of StruviteFinancial Assessment of Struvite Feasibility Worksheet Page 2Total Struvite Production Chemical Costs Lye Magnesium PhosphateAssess Current Bilogical Nitrogen Removal Process Costs Assess Costs for Production of Struvite with Consideration to Struvite Resale p Ammonia NH4/ Day q Aeration PowerElectricity PriceAeration Cost/ DayMethanol Usage Methanol PriceMethanol Cost/ DayTotal BNR Cost /g NH4 (kg NH4/ day)(KW)($/KW)($/day)(gal/kg NH4)($/gal)($/day)($/day)($/kg) 100067.30.11177.67 $ 0.61.50 $ 900.00 $ 1,077.67 $ 1.08 $ Method$/ kg ammonia N Struvite0.73 $ Baseline (BNR)1.08 $ Comparison
131 Appendix D-1: Calibration Curves Figure 55: Typical Calibrati on Curve for Total N by TOC-V
132 Appendix D-1 (Continued) Figure 56: Typical Calibrat ion Curve for Ammonia Probe
133 Appendix D-1 (Continued) Figure 57: Typical Calibrati on Curve for Total Phosphate
134 Appendix D-1 (Continued) Figure 58: Typical Calibration Curve for Ortho-Phosphate
135 Appendix E-1: Bioenergetic Stoichiometry Determination d c s a eR R f R f R Where: R = overall reaction ef= energetic partitioning coefficient aR= electron accepto r half reaction sf= synthesis partitioning coefficient cR= cell synthesis half reaction dR= electron donor half reaction Equation E-1: Microbial Energetic Stoichiometry For aerobic digestion of primary sludge, us e the following values from Rittman and McCarty (2001). Primar y sludge is represented as N O H C3 19 10 and new bacteria cells are represented asN O H C2 7 5.
136 Appendix E-1 (Continued) sf= 0.6 ef= 0.4 e H HCO NH CO O H N O H C Rd 3 4 2 2 3 19 1050 1 50 1 50 9 25 9 50 1 : O H N O H C e H NH HCO CO Rs c20 9 20 1 20 1 20 1 5 1 :2 7 5 4 3 2 O H e H O Rw a2 1 4 1 :2 Equation E-2: Determination of Ae robic Degradation of Primary Sludge Converting fraction to decimal, multiplyi ng by the appropriate half reactions by the energetic and synthesis coefficients and adding the reactions, we arrive at the following total reaction. N O H C CO O H O HCO NH N O H C R2 7 5 2 2 2 3 4 3 19 1003 0 06 0 11 0 1 0 01 0 01 0 02 0 : Equation E-3: Resulting Aer obic Degradation of Primary Sludge
137 Appendix E-1 (Continued) And then converting to a one molar basis of the influent primary sludge we arrive at the equation presented in the text. N O H C CO O H O HCO NH N O H C R2 7 5 2 2 2 3 4 3 19 1015 0 3 0 5 5 5 5 0 5 0 : Equation E-4: Aerobic Degradation of Primary Sl udge on Single Molar Basis
138 Appendix E-2: Partitioning Based on Stoichiometry Beginning with equation F-5 fr om the previous appendix we observe that the aerobic digestion of primary sludge as represented would require additional nitrogen source, shown as ammonia on the left side of the equatio n, in order to proceed fully. For this example we assu me that sufficient supplemental nitrogen is available and the reaction proceeds. Following N through the equation F-5, we see that ther e are 14 g N / mole of influent primary sludge and 14 g N/ mole of influent supplemental ammonia multiplied by 0.5 moles ammonia per mole influent primary sludge equals 7 grams of ammoniaN / mole of influent sludge. Therefor e, the total influent is 21 g N/ mole of influent primary sludge. Looking at nitrogen on t he effluent side, we see that solid there is 14 g N/ mole of effluent cells multiplied by 1.5 mole of effluent cells per mole influent primary sludge. This N is considered solid, so that solid N equals 21 g N per mole of influent primary sludge. Thus solid N equals 100% the 21 g of influent N. In this example, there is no N on the ri ght side of the equation in a liquid form and liquid effluent N equals 0%. Similarly there is no gaseous N on the right side of equation F-5 and effluent gaseous N equals 0%.
139 Appendix E-2 (Continued) Because all effluent N is in the form of ce llular N, this is c onsidered solid and the partitioning for this example is 100% solid and 0% liquid, 0%gas Were there to be ammonia or any gaseous forms of nitrogen on the right hand side of the equation, the percentages would be calcul ated using the same methodology as above.
140 Appendix E-3: Mass Tran sfer Calculations G y l y OL y L y G yQ H aV K C H Q F, ,exp 1 Where: yF= mass transfer rate (M/T) GQ= gas flow rate (L3/T) y LC,=dimensionless HenryÂ’s Constant y OLK,=liquid concentration of y (M/ L3) a= interfacial are aper unit volume of liquid (L3/ L2) lV= liquid volume (L3) Equation E-5: Mass Transfer for a vola tile compound out of an aeration system (Matter-Muller et al. 1980) Our objective is to compare mass trans fer rates for carbon dioxide and ammonia out of solution in the same aeration system. The overall mass transfer coefficient for each compound is calculated from liquid and gas phase mass transfer coefficients. The liquid phase mass transfer coefficient is estimated from the diffusivi ty of the compounds, the diffusivity of oxygen, and published values for the li quid phase mass transfer coefficient for oxygen in fine bubble aeration conditions. The following relationships are used.
141 Appendix E-3 (Continued) 5 0 , , A L B L A l B lD D k k and y g L y OLH k k K 1 1 1, Where: B lk, = liquid phase mass transfe r coefficient, compound B B LD, = diffusivity of compound B gk= gas phase mass transfer coefficient Equation E-6: Determination of Diffu sion and Mass Transfer Coefficients Gilot et al. (2005) estimate the liquid phase mass transfer rate for oxygen in fine bubble aeration to range from 3.2 Â– 13.4 hour-1. For our purposes, weÂ’ll take the median and use 8.3 hour-1. Carbon dioxide and ammonia have the same diffusivity (2X 10-5), but differing HenryÂ’s constants results in signific antly different overall mass transfer coefficients.
142 Appendix E-3 (Continued) Based on the equations and calculations outlined above, the following values were used for comparison of mass transfer rates using the Matter Muller equation. The gas flow rate, interfacia l area per unit volume, and liquid volume should be the same for both compounds and not affect the result. These values were set arbitrarily. Table 9: Parameters for Mass Transfer Comparison Parameter Ammonia Carbon Dioxide HenryÂ’s Constant 0.0006 1.1 Overall Mass Transfer Rate 0.7/ hour 7.66/ hour Gas Flow Rate 1 m 3 /h 1 m 3 /h Liquid volume 0.15 L 0.15L Interfacial area 1m 1m With the values above, only concentr ation and the mass transfer rate are unknown in the equation and the pl ot in the text shows tr ansfer rate vs. various concentrations of ammonia and carbon di oxide for a visual comparison of different scenarios.