MECHANISMS OF PHOSPHORUS REMOVAL FROM WASTEWATER

2 downloads 0 Views 2MB Size Report
conventional alum, but rather that a mixed aluminum hydroxide phosphate precipitate is formed. Jar tests performed with full-scale alum sludge and wastewater ...
1+1

National Library of Canada

Bibliothèque nationale du Canada

Acquisitions and Bibliographie Services Branch

Direction des acquisitions el des services bibliographiques

395 Wellington Street Ottawa, Ontario

395, rue Wellington Ottawa (Ontario)

K1AON4

K1AON4

NOTICE

AVIS

The quality of this microform is heavily dependent upon the quality of the original thesis submitted for microfilming. Every effort has been made to ensure the highest quality of reproduction possible.

La qualité de cette microforme dépend grandement de la qualité de la thèse soumise au microfilmage. Nous avons tout fait pour assurer une qualité supérieure de reproduction.

If pages are missing, contact the university which granted the degree.

S'il manque des pages, veuillez communiquer avec l'université qui a conféré le grade.

Sorne pages may have indistinct print especially if the original pages were typed with a poor typewriter ribbon or if the university sent us an inferior photocopy.

La qualité d'impression de certaines pages peut laisser à désirer, surtout si les pages originales ont été dactylographiées à l'aide d'un ruban usé ou si l'université nous a fait parvenir une photocopie de qualité inférieure.

Reproduction in full or in part of this microform is governed by the Canadian Copyright Act, R.S.C. 1970, c. C-30, and subsequent amendments.

La reproduction, même partielle, de cette microforme est soumise à la Loi canadienne sur le droit d'auteur, SRC 1970, c. C-30, et ses amendements subséquents.

Canada

• MECHANISMS OF PHOSPHORUS REMOVAL FROM WASTEWATER BY ALUMINUM

by Elisabeth Galarneau Research Supervisor: Professor R. Gehr



October 1995 A thesis submitted to the Faculty of Graduate Studies and Research in partial fulfilment of the requirements of the degree of Master of Engineering

;

...• "

~"" ~

Department of Civil Engineering and Applied Mechanics McGiII University, Montreal



Cil

Elisabeth Galarneau, 1995

1+1

National Library of Canada

Bibliothèque nationale du Canada

Acquisitions and Bibliographie Services Branch

Direction des acquisitions et des services bibliographiques

395 Wellington Street Ottawa, Ontario

39~, rue Wellington Oltawa (Onlario)

K1A ON4

K1A ON4

l'OUI "Il!

V,J'lt' rl'fo\ll"'Ct'

The author has granted an irrevocable non-exclusive licence allowing the National Library of Canada to reproduce, loan, distribute or sell copies of hisjher thesis by any means and in any form or format, making this thesis available to interested persons.

L'auteur a accordé une licence irrévocable et non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de sa thèse de quelque manière et sous quelque forme que ce soit pour mettre des exemplaires de cette thèse à la disposition des personnes intéressées.

The author retains ownership of the copyright in hisjher thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without hisjher permission.

L'auteur conserve la propriété du droit d'auteur qui protège sa thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimés ou autrement reproduits sans son autorisation.

ISBN

Canada

0-612-07977-5

Abstract



Within the scope of a project funded by the Qucbec Ministry of the Environment and FalIna (M EF), alternative wastewater phosphorus removai methods to alum or ferric chloride coagulation were assessed. The most promising technologies were found to be enhanccd biological phospholUS removal, treatment plant process optimization and on-line proccss control. The use of alum siudges From potable water treatment plants appeared promising although it has not yet been widely practised or studied. The MEF therefore decided to retain the process as one requiring further study. Through batch sorption tests with synthetic solutions, it was found that aluminum hydroxide has a significant sorptive capacity for orthophosphate, condensed phosphate and organic phosphate. The removal of these phosphates was independent of pH between pH 3 and pH 6. The solubility of the aluminum hydroxide was highly pHdependent. In ail the tests, except that with organic phosphate at pH 3.6, the measured soluble aluminum was consistent with solutions saturated with solid-phase AI(OH)J' With organic phosphate at pH 3.6, an aluminum-organic phosphate appears to have formed. A theoretical analysis of AI(OH)J and AI(PO,) precipitation showed that phosphate removal is not achieved through AI(PO,) precipitation when dosing with conventional alum, but rather that a mixed aluminum hydroxide phosphate precipitate is formed.



Jar tests performed with full-scale alum sludge and wastewater showed that phosphate is removed by the particulate fraction of the sludge. The removal of reactive phosphate (orthophosphate) decreased with an increase in the storage time of the alum sludge. This decrease was not seen with the non-reactive phosphate. It was therefore suggested that reactive phosphate removal is carried out by adsorption and that non-reactive phosphate removal is performed by a sweep-floc mechanism. Therefore, the use of alum sludges from potable water treatment plants does seem to present an effective and inexpensive means of removing phosphates From wastewaters.



Résumé



Dans le cadre d'un projet financé par le Ministère de l'environnement ct de la faune du Québec (MEF), une analyse des nouvelles méthodes de déphosphatation des e.ulx usées a été réalisée. Les procédés les plus prometteurs sont la déphosphatation biologique, l'optimisation des procédés, ct le contr81e en-ligne. L'utilisation des boucs d',tlan provenant des usines de filtration semble prommeteuse même si elle n'a p,tS encore été étudiée de façon rigoureuse. Le MEF a décidé de retenir le procédé pour une étude plus approfondie. Des expériences à petite échelle avec des solutions synthéthiques ont été réalisées ct il a été confirmé que l'hydroxyde d'aluminium a une capacité pour l'adsorption des orthophosphates, des phosphates condensés et des phosphate organiques. L'enlèvement de ces phosphates était indépendant du pH des solutions entre pH 3 ct pH 6. p.u· contre, la solubilité de l'hydroxyde d'aluminium était fortement relié au pH. Pour' toutes les expériences avec les solutions synthéthiques, sauf celle avec le phosphate organique à pH 3.6, les concentrations d'aluminium soluble supportait la saturation des solutions avec le AI(OH)J' Avec le phosphate organique à pI-! 3.6, un précipité aluminium-phosphate organique s'est formé. Une analyse théorique de la précipitation du AI(OH)J et du AI(P04) a démontré que la déphosphatation n'est pas effectué par une précipitation du AI(P0 4) quand l'alun commercial est utilisé, mais qu'un précipité hydroxyde-aluminium-phosp:,ate est plut8t formé.



Des essais en béchers ("jar tests") ont été réalisés avec des véritables boues d'alun et eaux usées, et ces essais ont démontrés que les boues d'alun ont une capacité déphosphatame. L'enlèvement du phosphore réactif (orthophosphate) a diminué avec le temps d'entreposage des boues, et cette diminution ne s'est pas présentée pour l'enlèvement du phosphore non-réactif. Il est donc suggeré que les orthophosphates sont enlevés par adsorption et que les phosphates non-réactifs sont enlevés par séquestration.

À la lumière des informations ci-haut mentionnées, l'utilisation des boues d'alun semble efficace et peu coûtet'.se pour la déphosphatation des eaux usées.



11

Acknowledgements



The work described in thi, thesis was funded by the Quebec Ministry of the Environment under Project FRDT-E-PREE #92-4 (Volet II) and by Professor Ronald Gehr's research grant from the Naticnal Science and Engineering Research Council. Thanks must also go to McGill University's Department of Civil Engineering and Applied Mechanics for funding granted through teaching assistantships, as weil as for general administrative support. My laboratùry work would not have been possible without the guidance and extensive knowledge of Dr. Eva Waldron, who, in my opinion, runs the safest laboratory at McGill. She will be greatly missed upon her retirement. Many thanks must also go to Matthew Paterson and Jeff Schoolcraft who collected the jar test data for the experiments describing wastewater phosphate removal with alum sludges. Their contribution to this work is immeasurable. The help of Louis Houle at the Physical Sciences and Engineering Library was also invaluable.



Throughout the project's duration, guidance and support have been received from many people in industry. They include Howard Brown and Diane M. Landry at Dessau Inc., as weil as Charles Meunier at John Meunier Inc. The information regarding non·chemical alternative phosphorus removal methods described in Chapter 2 was collected in large part by Jean-François Paradis of John Meunier, Inc. Professor Yves Comeau at the École Polytechnique in Montreal and Professor Jim :Nice1l at McGill have contributed to this work by wa)' of their guidance and suggestions regarding the scope of the project as weil as the analysis of data. 1can not thank my supervisor, Professor Ronald Gehr, enough. He was insistent upon proper planning of the work, as weil as being re1entless in forcing me ta defend my findings and conclusions. Furthermore, without hanging over my shoulder or spoonfeeding me, he was always available for discussion. 1 would finally like to thank Gal Se1a for ail his patience, encouragement and technical help. 1 guarantee that he is the world's most knowledgeable electrical engineer when it cornes to environmental aqueous aluminum-phosphate chemistry. His help with software, as weil as his patience and insights when 1 wouId say (for the hundredth time) "1 just found something in my results... ", were invaluable.



III

Table of Contents







Abstract Résl.lrr6 Acknowledl;ements o

i ii

•••••••••••••••••••••••••••••••••••••••••••••••••••

III

List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. List of Tables

VII

1. Introduction

.

VI

2. Literature Review 2.1 A Survey of Phosphorus Removal Methods 2.2 Phosphorus Removal with Aluminum Salts 2.3 The Aqueous Chemistry of Aluminum 2.4 Adsorption of Phosphate on Aluminum I-lydroxidc 2.5 Phosphorus Removal with Alum Sludges

3 3 7 7 B 9

3. Preliminary Assessment of Phosphorus Removal Methods 3.1 The Analytic Hierarchy Process 3.2 Process Evaluations 3.3 Conclusions

13 13 17 19

4. Materials and Methods 24 4.1 Batch Sorption Tests 24 4.2 Jar Tests 25 4.3 Analytical Methods 26 4.3.1 Maintenance of Glassware 26 4.3.2 Phosphate Determinations 26 4.3.3 Aluminum Determinations . . . . . . . . . . . . . . . . . . . . . . . 26 4.3.4 Other Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 5. Results .................................................. 5.1 Phosphate Sorption on Aluminum Hydroxide: Synthetic Solutions 5.2 Removal of Wastewater Phosphates with Alum Sludges 5.2.1 Removal by Sludge Components 5.2.2 Dose·Response for Raw Wastewater 5.2.3 Dose·Response for Mixed Liquor . . . . . . . . . . . . . . . . . . . 5.2.4 Comparison of Raw Wastewater and Mixed Liquor Phosphorus Removals 5.3 Effect of Alum Sludge Age on Phosphate Removal 5.3.1 Raw Wastewater 5.3.2 Mixed Liquor IV

28 28 34

35 36 38 40 41

42 44

Table of Contents (continued)



6. Discussion 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8

Phosphate Removal with Aluminum Hydroxide Precipitation Adsorption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solid Phase in the Synthetic Solution Sorption Experiments Investigation of Hydroxyl-Phosphate Ion Exchange The Effect of Alllm Sludge Addition on Wastewater Treatment " The Decrease in Phosphate Removal with Aillm Sllldge Age .... Logistical and Economic Factors .. . . . . . . . . . . . . . . . . . . . . ..

7. Conclusions

47 47 47 48 49 56 57 58 59

61

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64





Appendices 70 Appendix A: Batch Sorption Test Data 70 . . . . . . . . . . . . . . . . . . 74 Appendix B: Jar Test Data 78 Appendix C: Derivation of AI-Hp-PO, Diagram

v

List of Figures







Figure 3.1: Figure 3.2: Figure 3.3:

A Generalized Decision T ree 13 Phosphorus Removal Decision Trees 15 Response Distributions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 16

Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure

5.1: 5.2:

Phosphate Removal with Aluminum Hydroxide Phosphate Removals at Different Initial pH Levcls Residual Aluminum Concentrations Synthetic Solution pH Values . . . . . . . . . . . . . . . . . . . . . . . . . . Phosphate Partitioning between the Solid and Soluble Phases Phosphorus Removal Capacities of Sludge Components Phosphate Removal from Raw Wastewater using Ahun Sludge Supematant Alkalinity and Suspended Solids Phosphate Removal from Jv!ixed Liquor using Alum Sludge Supematant Alkalinity and Suspended Solids Total Supematant Phosphate in the Wastewater and Mixed Liquor Total Sorbable Phosphate Removal Reactive Phosphate Removal Non-reactive Phosphate Removal Mixed Liquor Total Phosphate Removal Mixed Liquor Reactive Phosphate Removal

Figure Figure Figure Figure Figure

6.1:

5.3: 5.4: 5.5:

5.6: 5.7: 5.8: 5.9: 5.10:

5.11: 5.12:

5.13: 5.14: 5.15: 5.16:

6.2: 6.3: 6.4: 6.5:

28 2') 30 31 32 35 37 38 39 3') 41 4.' 43 44 45 45

Theoretical Aluminum Hydroxide Solubility 50 Theoretical and Measured Aluminum Concentrations 51 Minimum Aluminum Solubility in the AI-Hp-PO., System . . . . . 53 Total Aluminum Solubility at pH 5, 6 and 7 54 Effect of Alum Sludge Dosing on Wastewater pH 57

VI

List of Tables

• -



'l'able 2.1:

Typical Alum Sludge Characteristics

Table 'l'able 'l'able 'l'able 'l'able 'l'able Table Table Table Table Table

Criteria and Sub·Criteria Definitions . . . . . . . . . . . . . . . . . . . .. 14 Phosphorus Removal Processes and Likely Applications 17 Translation of Evaluation Smle . . . . . . . . . . . . . . . . . . . . . . . .. 18 Process Evaluation for Small Activared Sludge Plants 20 Process Evaluation for Small Biodisc Plants 20 21 Process Evaluation for Small Aerated Lagoons Process Evaluation for Large Activated Sludge Plants 21 Process Evaluation for Large Aerated Lagoons 22 Process Evaluation for Large Biofilters 22 Process Evaluation for Large Physicochemical Plants 23 Process Evaluation for Large Sequencing Batch Reactors 23

3.1: 3.2: 3.3: 3.4: 3.5: 3.6: 3.7: 3.8: 3.9: 3.10: 3.11:

'T'able 5.1: Table 5.2:

Alum Sllldge and Raw Wastewater Characteristics Alum Sllldge and Mixed Liquor Characteristics

Table 6.1:

Aluminum Concentrations for AI(PO,) and Al(OH)) Precipitation 55

Table Table Table Table Table

Barch Batch Batch Batch Batch

Al: A2: A3: A4: A5:

Table BI: Table Table T"ble T"ble Table

B2: B3: B4: B5: B6:

Table Cl:



10

Sorption Sorption Sorption Sorption Sorption

Test: Test: Test: Test: Test:

Orthophosphate Olt pH 4.8 Orthophosphate Olt pH 5.7 Condensed Phosphate Olt pH 5.7 Organic Phosphate Olt pH 3.6 Organic Phosphate Olt pH 5.7

Jar Test Data: Raw Wastewater with Alllm Sludge and Aillm Sludge Sllpernatant Jar Test Data: Raw Wastewater with Fresh Alllm Sllldge Jar Test Data: Raw Wastewater with I.Day·Old Alum Sludge Jar Test Data: Raw Wastewater with 5-Day-Old Alllm Sludge Jar Test Data: Mixed Liquor with Fresh Alum Sludge Jar Test Data: Mixed Liquor with 1-Day-Old Alum Sludge Stability Constallts From Kotrly and Sucha (1985)

vu

42 44

71 71 72 72 73

75 75 76 76

77 77 81

1. Introduction



Phosphorus is usually considered to be the limiting nutnent with respect w the eutrophication of water bodies. Since human W,lstes '1I1d detergents contaill sigllil"ic'lIll amounts of phosphorus, wastewater discharges to waterbodies contribute to eutrophication.

For this reason, wastewater treatment plants must oftell 111L'et

maximum phosphorus diseharge limits set by the governmelll.

These limits are

determined on a plant-by-plant basis, but a typical effluent phosphorus target is between 0.8 and 1.0 mg P IL.

In Quebec, the majority of wastewater treatment plants with an effluent discharge limit remove phosphate by chemical addition. The most common chemicals used ill the province are alum and ferric chloride. However, research spanning the last twenty years has uncovered many alternative phosphorus removal methods. These alternatives include the use of novel coagulating chemicals, as well as physical processes. Sinet' these alternative methods have not yet been widcly implemented in the province, the



Quebec Ministry of the Environment and Fauna (MEF) initiated a study ta identiry and evaluate these alternative processes. The evaluation stage of the study was the first goal of this thesis.

An objective rating system, in which relevant criteria were categorized and assigned relative importances, was devised using the Analytical Hierarchy Process (AI-IP). Input to the AHP decision models was obtained from a process identification study which was carried out through severalliterature searches and through interaction with experts in the field. In the AHP analysis, enhanced biological phosphate removal performed consistently well, as did operational modifications such as chemical dosing optimization and on-line process control.

Several wastewater treatment plants in Quebec receive alum sludges from potable water



treatment plants along with the sewage and industrial waste. In sorne of these plants, 1

il has bccn found that effluent phosphorus targets have been met without chemical



addition or en~anccd biological phosphorus removal.

However, since alum sludge

addition ta wastewater has not been studied extensively, it received a low score in the AHP analysis. The MEF remained interested in the process and it moved to initiate a delailed study of the effieiency and relevant mechanisms of phosphate removal induced by alum sludges. This study is presented as the second part of this thesis.

Alum sludges consist predominantly of solid aluminum hydroxide. The soil science literalure has shown that aluminum hydroxide is capable of orthophosphate fixation. I-1owever, wastewaters contain phosphorus forms other than orthophosphate, and condensed and organic phosphates represent major contributions to the total phosphate content of a wastewater. Therefore, laboratory-scale batch sorption experiments were carried out with synthetic solutions in order to determine the fixation ability of aluminum hydroxide for orthophosphate, condensed phosphate and organic phosphate, as weB as to identify the mechanisms responsible for these removals. Jar tests were carried out so that the fixation ability of alum sludges for raw wastewater and mixed



liquor phosphates couId be evaluated and compared to the results of the batch sorption tests carried out with synthetic solutions.

The soil science literature also demonstrates that, as aluminum hydroxide solids age, their fixation ability for orthophosphate diminishes. Further jar tests were therefore carried out to confirm the existence, and to determine the extent, of such an aging phenomenon on ail of the phosphate forms found in wastewaters.



2

2. Literature Review



Phosphorus is found in domestic wastewaters almost exclusivcly as phosphatcs (APl-lA et aL, 1992) in concentrations ranging from 3-15 I11g l'IL (Snocyink 'lI1d Jenkins, 1980). These phosphates are c1assified as orthophosphates, condensed phosphates (polymcric phosphate rings and chains) and organic phosphates, and they l11ay bc dissolvcd or bound to solids (APHA et aL, 1992). The removal of phosphates from wastcwatcrs has traditionally been achieved by coagulation and floccubtion with metal salts such as aluminum sulfate (alum) and ferric chIo ride.

2.1: A Survey of Phosphorus Removal Methods

The MEF specified that the goal of the alternative process identification

W,lS

to

determine which processes and/or chemical products could replace alum or ferric chIoride dosing for phosphorus remova1. Such eligible processes would be capablc of



attaining 1 mg P IL in the final effluent and wouId not include tcrtiary trcatment processes which are normally used ta achieve much lower effluent phosphorus Icvcls.

Several chemicals which could be used instead of the traditional alum or fcrric chloride were identified in the literature. These include sodium aluminatc, prcpolymerized aluminum salts, prepolymerized iron salts, lime, lanthanum salts, ferric sulfate, organic polymers such as alginate and chitosan, and cations produced through clectrolysis. Industrial waste products such as ARMS (alumized red mud solids), waste picklc liquor and alum sludges from potable water treatment plants were a1so identified as potential replacement chemicals.

Sodium aluminate generally requires higher doses than alum with no apparent bencfit (Grigoropoulos et aL, 1971; Long et aL, 1971). The sludges produced are of comparable quality (O'Shaughnessy et aL, 1974) to those produced with alum. However, less pH



adjustment is necessary with this sodium aluminate since it consumcs less alkalinity 3

(Hébert, 1994).



When introduced to water, the aluminum ions in alum or sodium aluminate hydrolyse and produce various soluble monomers and polymers. The polymers are of special importance since they are high in charge and should easily remove particulate or colloidal phosphorus according to the Schulze-Hardy Rule. Simply put, the rule states that coagulation potential is proportional to the sixth power of the valence (Montgomery, 1985). Since the development of these polymers is difficult to control, several prepolymerized coagulants have been introduced. These are solutions of aeid metal salts such as aluminum chloride, aluminum sulfate and ferric chIo ride that have been partially neutralised Golicoeur and Haase, 1989; Od(·g.1ard et al., 1990).

The

aluminum salt most commonly used is aluminum chIo ride since the sulfate ion in alum seems ta catalyse the production of solid-phase aluminum hydroxide, thereby inhibiting the production of soluble hydrolysed aluminum polymers (Odegaard et al., 1990). Use of the prepo!ymerized iron salt is rare since the product tends to be unstable (Odegaard et al., 1990). Severai studies (Brière et al., 1990; Milette and Bosisio, 1989; Obermann



et al., 1987; O'Neil et al., 1986; Sammut et al., 1992; Viraraghavan and Wimmer, 1988) have demonstrated the relative effectiveness of prepolymerized aluminum coagulants. The required doses and subsequent sludge productions are quite elevated. Furthermore, the products tend ta work weil for solids removal but their soluble phosphorus removal ability seems ta decrease with increasing polymerization (Odegaard et al., 1990). These products are better in coId water than their unpolymerized counterparts

and they are said to have a larger pH operation range Golicoeur and Haase, 1989).

Lime, due to its excessive sludge production (King et al., 1974; Kumar et Clesceri, 1973) and the need ta adjust the pH of the effluent, is not a cost-effective precipitant. Furthermore, it cannot be used as a simultaneous precipitant since the high pH it induces inhibits bioiogicai activity.



Lmthanum salts are excellent for phosphorus removai (Melnyk, 1974) but their cost 4

is prohibitive and their use leads to toxic residuals in the tre'lted diluent (Recht ct '11.,



1970),

The use of waste pickle liquor (ferrous sulfate or ferrous chio ride) appe.1rs ta W'UT,un further sttldy due ta its phosphorus removal efficiency.

Hawever, the product is

classified as a waste by the MEF and the heavy metaIs present in it make diluent and sludge toxicities a concern (Lewandowski, 1983). Waste pickle liquar, alont; with fctTic sulfate, the refined waste product, perform favorably with respect ta sludge production and dewatering characteristics, espeeially when compared with ahlln (King ct al., 1974). Organic polymers such as alginate and chitosan are interestÎng since they arc biodegradable and less toxic than other polymers. However, their phosphorus remaval efficiency is low even olt high doses (Megharaj et al., 1992; Gao ct al., 1993).

Information related to the use of alumized red mud solids or ARMS, a residue from bauxite processing, is scant (Couillard and Tyagi, 1986; Shannon and Vcrghese, 1976). Such a product is available locally but the presence of heavy metals and other taxics



calls into question the safety of the product's use. Alum sludges from potable water treatment plants are often discharged directly to sewers and molY be contributing ta phosphorus removal (Lavoie and Meloche, 1993). Tbe performance of tbis practicc bas not been studied in a controlled manner and the second part of this tbesis will investigate the governing mechanisms. Electrolytic cation production is an interesting process in that electric power is relatively inexpensive in Quebec. Furthermore, the consumption of alkalinity and the generation of sali nity are avoided with tbis process (pretarius et al., 1991). Potential problems include fouling of the clectrodes and high capital costs associated with plant conversions.

Enhanced biological phosphorus removal is a process by which biomass is subjected to conditions in which they will take up more phosphorus than normally required for growth (Metcalf and Eddy, Inc., 1991). The main advantage of the proccss is the



absence of coagulant metal ions in the resulting sludges.

5

Proccsses which involve the crystallization of phosphate compounds, usually as



hydroxyapatite, also negate the need for aluminum or iron addition (Zoltek, 1974). The processes are stable and producc granules that can be used as agricultural ferrilizers (Janssen et al., 1991). RIM·NUT, a proprietary process, combines ion exchange and crystallization of struvite (Liberti et al., 1979).

The operation of wastewater treatment plants can be improved by on·line control which involves water quality parameter me~surement and feedback mechanisms (Jeppson and Olsson, 1993; Lotter and Pittman, 1993; Thornberg et al., 1993). Furthermore, eXlstmg systems can often benefit from a detailed analysis and optimization of their current operating scheme (Eastwood and Murphy, 1991; Sérodes et al., 1988).

Several novel and effective ways of achieving solid.liquid separation have been developed commercially (Degremont, 1993; Harrington and Smith, 1987; Mirzejewski et al., 1990). Since suspended solids typically contain 3·5% phosphorus (Nutt, 1991),



total effluent phosphorus levcls can be significantly reduced by superior solids removal. New types of filters which combine biologica! treatment and clarification are also showing promise in both soluble and particulate phosphorus removal (Gonçalves and Rogalla,1992; Séguin et al., 1993).

In the process evaluations described in Chapter 3, the removal of wastewater phosphorus through the use of sludges from potable water treatment plants received a relatively low score.

However, this

WolS

largely due to a lack of information

regarding the process since it is not yet commonly practised. The Ministry of the Environment of Quebec has issued new guidelines to potable water treatment plants stating that they must treat at least a portion of the sludges that they produce. Since the use of these sludges for phosphorus removal at wastewater treatment plants presents an interesting treatment alternative, the MEF decided to study the effectiveness of the



method. The results of this investigation are the subject of the remainder of this work. 6

2.2: Phosl'horus Removal with Aluminum Salts



The introduction of an aluminum salt to an orthophosphate solution can produec two possible binary precipitates: Al(OH), and AI(PO,).

When dosing metal S'lits to

wastewaters, plots of orthophosphate residuals versus metal dose produce curves which exhibit two distinct regions: a linear phosphorus decay region at low meta\ dose and a zero-decay region at higher metal dose Qenkins and Hermanowicz, 1991). The linear phosphorus decay has been modelled in the past as aluminum phosphate precipitation (Ferguson and King, 1977; Arvin and Peterson, 1980). Though thermodynamic data suggested otherwise, Ferguson and King (1977) postulated that aluminum hydl"Oxidc was produced only after the stoichiometric requirement for a\uminum phosphate precipitation had been exceeded (Equation 3.1 below from Kotrly and Sucha, 1985).

(3.1)

Data from the Ferguson and King (1977) mode\ calibration yielded a molar ratio of



aluminum to phosphate removed frnm solution of 1.4, suggesting that aluminum WolS bound ta solids other than an AI(P04) precipitate. Other researchers have lried to improve predictability by hypothesizing the presence of constituents such as calcium, bicarbonate and iron in the precipitate (Arvin and Peterson, 1980). appears that Al(P04)

[n essence, il

precipitate formation is a gross simplification of the

orthophosphate removal process.

Recent research has shown that the competition between hydroxyl groups and other ligands, such as H 2PO;, is the main reaction path of aqueous aluminum (Edzwald,

1993).

2.3: The AÇjueous Chemistr:y of Aluminum



The aqueous chemistry of aluminum is quite complex and has not yet been completcly 7

dcscribcd. [t is gcncrally agrced that the aluminum ion in aqueous solution coordinates



with six watcr molccules yielding the hydrated cation, AI(HP),J+ (Pankow, 1991). Over time, the coordinated water molecules tend to lose hydrogen ions ta solution yielding AI (HPls(OH)2+ . This is documented in Equation 3.2 below (Kotrly and Sucha, 1985) where the HzO molecules have been omitted for the sake of Lrevity. K- [AI(OH)2'][H'] [AI

Duc to this hydrolysis

1055

3

= 10-4.97

']

mol L

(3.2)

of protons, aluminum is classified as an acidic cation. Several other

reactions

occur

AI(HP).(OH)1+,

yielding

AI(H10lJ(OH)J

and

AI(H10lz(OH).. (see Kotrly and Sucha, 1985). Dimeric and polymeric species also exist (Bersillon et al., 1980). The polymerization is achieved through a bridging process which incorporates non-structural hydroxyl groups as structural members (Sims and Ellis, 1983). All of the soluble hydrolysis products and polymers are thought to be intcrmediates in the formation of crystalline aluminum compounds such as AIOOH (boehmitc) and AI(OH)J (gibbsite) (Rubin and Kovac, 1974).



2.4: Adsorption of Phosphate on Aluminum Hydroxide

Adsorption of phosphate on amorphous Al(OH)J also accounts for ligand fixation Qenkins and Hermanowicz, 1991) and this may serve to describe the non-staichiometric behaviour depicted in typical phosphorus removal curves. In fact, the sulfate ion produced by the dissociation of alum appears ta catalyze aluminum hydroxide production (Odegaard et al., 1984).

According to studies performed in the soil sciences, it is known that both amorphous and crystalline aluminum hydroxides readily remove phosphates from solution (Hsu and Rennie, 1962; Hingston et al., 1967; Lijklema, 1980; Sims and Ellis, 1983). Phosphate is specifically adsorbed since its fixation is unaffected by the ionic strength



of the solution (Hingstan et al., 1967) and the adsorption changes the zero point of 8

charge of the solid (Eisenreich et al., 1977). Furthermore, desorbing the phosphol"lls



requires further surface charge depression than that created by its adsorption (I-lingston et al., 1967). Due to the chemical bonds creared in this type of adsorption, the specific surface area of the solid affects phosphorus fixation. This has been demonstnlted with comparative studies that showed that boehmite, with a moderate surface area, is e'lp'lble of adsorbing more phosphorus than low surface area gibbsite (Larson et al., 1986).

The environmental sCience literature also demonstrates the phosphllrus adsorption ability of aluminum hydroxides.

Land application of aium sludges, which are

predominantly aluminum hydroxide (Montgomery Inc., 1985), has been demonstrated to l'l'duce plant yields. These yil'Ids were improved only when phosphorus fenilizers were added (Lucas et al., 1994).

As stated earlier,

aluminum

hydroxide polymerization

occurs

through

the

transformation of non-structural hydroxyl groups into structural bridging components. These structural groups are strongly fixed and it is thereforl' l'xpected that as aging



occurs (more hydroxyls become fixed), less adsorption of phosphorus occurs. Such a decreasl' in phosphorus adsorption with hydroxide age has been documented in the litl'rature (Duffy et al., 1994; Lijklema, 1980).

2.5: Phosphorus Removal with Alum Sludges

Phosphorus removal from wastewaters by use of alum sludge solids from potable water treatment plants has begun to l'l'ceive attention. If this recycling were shown to be effective, it would be environmentally beneficiai since the original alum would be used twice.

Economic advantages would also be realized since aium sludge l'cuse would

allow wastewater trl'atment plants to l'l'duce their operating budgets for coagulants.

This type of phosphate removal would only be efficient if the hydroxides in the sludges



were relatively fresh, since specific surface area decreases with increasing sol id age. An 9

adsorptiLn study performed on one-day old aluminum hydroxide preeipitate showed



that the solids demonstrated signifieant phosphorus fixation ability although the initial eapaeity was lower than that of fresh solids (Lijklema, 1980). If the alum sludge was transported

within such short time frames,

highly disordered solids (with

eorrespondingly high specifie surface areas) wouId predominate at the wastewater plant entry since sueh erystalline speeies as gibbsite take approximately three months ta form (Sims and Ellis, 1983).

Sorne typieal eharaeteristies of alum sludges are presented in the following table.

1



PROPERTY

1

VALUE

BOD

30 - 300 mg/L

COD

30 - 5,000 mg/L

pH

6-8

total solids

0.1-4.0%

solids aluminum content

3.9 - 10.3 %

.:J

Table 2.1: Typieal Alum Siudge Charaeteristies (Source: Montgomery, Ine., 1985)

Several studies have been performed ta determine the effeet of alum sludge addition on biologieal treatment (Duffy et al., 1994; Lavoie and Meloehe, 1993; Power et al., 1992; Hsu, 1973). For a series of bateh experiments, Hsu (1973) prepared synthetie alum sludges by centrifuging waters ta whieh alum had been added. The precipitate was then rinsed and the pH adjusted ta 7.0. This yielded a jelly of amorphous aluminum hydroxide with concentrations as high as 8,000 mg AlIL.

The gel was added ta

different bateh reaetors simulating primary clarification and seeondary aerobie biologieal treatment. Tests were also run on a simulated anaerobic sludge digester. Results indieated that excellent phosphorus removal eould be aehieved when adding the sludge to either the primary or seeondary treatment vessels. With aluminum doses



la

higher than 150 mg Al/L (2% sludge) at the second'lry levcl, the effluent phosphorus



concentration was seen ta be independent of influent phosphorus concentl·,ltion. A stable effluent value of 0.64 mg PIL was obtained with influent values as high as H mg

PIL. The aluminum hydroxide addition increased BaD removal slightly but did not significantly affect sludge settling velocity or sludge volume index at concentrations of less than 300 mg Al/L. However, the digestion and dewaterability of the activ'lted sludge were affected by alum sludge addition. Gas production \Vas retarded 'lt doses greater than 100 mg Al/L but the composition of the gas \Vas the same as that of the controls. The dewatering characteristics of the mixed sludge were superior to those of the controls.

The Acton Vale Wastewater Treatment Plant in Quebec consists of t\Vo aerated lagoons and a facultative polishing lagoon. Ali three lagoons are operated in series. The effect of disposing of the town's water treatment sludge and filter backwash to the \Vaste \Vatel' treatment plant was investigated (Lavoie and Meloche, 1993). Comparisons \Vere made between additions to the first and third lagoons by the use of jar tests. From full·scale



ca\culations, it was determined that the alum sludge would account for 5.7% (the backwash contribution was low enough to he considered negligible) of the total flow to the wastewater treatment plant, with an aluminum concentration of 179 mg Al/L in the sludge. Waters entering the first lagoon had an orthophosphorus concentration of 1.8 mg PIL and this was reduced to 0.3 mg PIL (83% removal) by a volumetric sludge addition of 6%. For the third lagoon, the orthophosphcrus concentration l'cil l'rom 0.8 mg PIL to 0.25 mg PIL (69% removal) with the same concentration of sludge. Since both addition points yielded the same final phosphorus concentration, il was suggested that addition to the third lagoon should be practised in order to avoid any phosphorus deficiency effects on substrate removal in the first lagoon. Full-scale trials performed at a Kingston, Ontario activated sludge plant also showed significant (90%) soluble phosphate removal in the aeration basin where the wastewater was only exposed to recycled waste activated sludge which had been treated with alum prior to



sedimentation (Duffy, 1994). 11

ln South Africa, alum sludge addition to a pilot-scale Modified Ludzack-Ettinger



activated sludge plant was investigated (Power et al., 1992). Excess nitrate was dosed ta the anoxic zone in order ta guarantee the absence of anaerobic conditions that might

cause enhanccd biological phosphorus removal.

It was found that the volatile

suspended mlids in the alum sludge were not biodegradable and that sludge production in the secondary clarifier increased with increasing alum sludge addition. Furthermore, the dewaterability 01 ,he mixed sludge after treatment was superior to that of the alum sludge aJone. This improvement was not found when the two sludges were mixed without treatment so the authors concluded that ligand exchange of PO/" for OH", with a corresponding decrease in the gelatinous character of the solids, was occurring in the activated sludge reactor. At steady-state, it was found that each milligram of non-volatile suspended solids (with the assumption that these solids were composed exclusively of Al,O) since the raw waters were very soft) removed 0.18 mg P.

From the literature cited above, it is apparent that the reuse of alum sludges is effective for phosphorus removal.



The following chapters will give the results of

experimentation designed to better understand the mechanisms involved. In Quebec wastewaters, the orthophosphate fraction typically accounts for 30-85% of the total phosphorus (Brown et al., 1995). The remainder of the wastewater phosphorus is polymeric and organic in nature, and may be dissolved or particulate. The studies presented above focused almost exclusively on orthophosphate removal. However, sinr~ other phosphates account for 15% to 70% of the total phosphate in the

wastewaters in Quebec, their behaviour has a significant impact on total phosphorus removal. The experiments described in the body of the present work were designed to better understand the behaviour of polyphosphate and organic phosphate in the presence of aluminum hydroxide.

Furthennore, experiments were carried out to

determine the effieiency of alum sludge at phosphorus removal from a typical local wastewater, as weil as the effect of aging of the alum sludge on phosphorus removal.



12

3. Preliminary Assessment of Phosphorus Removal Methods



3.1 The Analytic Hierarchy Process

Decision making is a process which involves several phases including defining the objective, determining which criteria are relevant and attributing relative importances to those criteria. For complex decisions, the criteria are often difficult to compare. The analytic hierarchy process (AHP) is a tool which allows the user to rigorously organize and weight different criteria

50

tbat each may be considered with respect to

the others (Expert Choice, 1993). For this project, Expert Choicé rM software was used. This package allows the user to construct a decision tree (see Figure 3.1) which has several levels of criteria 'lIld subcriteria. GOAL



L 1. 000 G 1. 000

CRITER_l

CRITER_2

CRITER_J

L 0.333 GO.333

L 0.333 GO.333

L 0.333 G 0.333

~SUB_2_1

:""'SU8_3_1

L 0.333 G 0.111

L O. 3D

L 0.333 G 0.111 f-S UB_2_3 L 0,333 G 0.111

LO.333 G 0.111 r-SUB_3_3 L 0.333 G 0.111

-5U8_1_1 L 0.333 G 0.111 -sua l 2 L 0.333 G 0.111 -SUB_l_3 L 0.333 G 0.111

_ 3_ 2 r SUB_ 2 _ 2 r GSUB0.111

Figure 3.1: A Generalized Decision Tree

The goal (levelO criterion) of the decision is located at the summit of the tree. For this project, the goal is to determine the best phosphorus removal alternative to the traditional use of alum or ferric chio ride.



Directly below the goal lie the leve! 1

criteria and, below those, the level 2 or sub-criteria. 13

Two numbers, identified as Land G, are associated \Vith each criterion. These indicate



the relative weight of the criterion with respect to (i) the criterion immediately above it where L=local, and (ii) the ultimate goal where G=global. The sum of the local weights (L) for each level of criteria is La, and the sum of ail the global weights on a single leveI associated with one branch is the local weight of the criterion above them.

Severai parameters were consiJered to be important ta the assessment of phosphorus removal methods, and these are listed and defined in Table 3.1.

LEVEL 1 CIUTERION 1

$COSTS

1

SCAPITAL EFFLUENT

1

SOPERATE = operating costs

economic costs

=

LEVEL 2 CRITERION

= effluent quality

=

capital costs

rOXEFF = effluent toxicity PHOSPHOR = effluent phosphorus

5S "" effluent suspended solids EOD = effluent BOD



SLUDGE

=

sludge qu,l1ity

DEVELOP "" stage of proccss devclopment

1

SIMPLIC

1

VARADVN

10 000

m'Id

$COSTS SCOSTS EFFLUENT EFFLUENT EFFLUENT EFFLUENT SlUDGE $OPERATE $CAPlTAL TOXEFF PNOSPNOR SS BOD TOXSlDG

.

Al ternatives

1 on-line controL 2 3 4 5

non-trad clarif bio-P optimize in ject pickle liquor

.1319 LOI/MOD LOI/MOD LOI/MOD LOI/MOD LOI/MOD

6 PASS

MOCHI

7 sodium aluminate

MODNI MODNI MODHI MODNI MODHI MOOER

8 PAC

9 ferric sulfate 10 RIM·NUT

11 crystallization 12 ARHS

.0681 MODNl

MODER

MODHI

MODER MOOER

MODNI MODNl MODER MODER MODER MODER MODER

HIGH MODHI MODER

.

. .1344

MODER MODER MODER MODER MODER

MOOER LOII MODER MODHI

.

.0691 LQWMOD MODER

.0320 MODER MODER

.0291 MODER

.0739 MODER MODER

MOOER

MODER

MOOER MOOER

MODER MODER LOI/MOD MODER LOI/MOD MODER LOI/MOD MODER MODER

MODER MODER

MODER MODER

LOIJMOD

MOOER MOCER MODER MODER LOI/MOD MODER MODER

!'100ER LOI/MOD LOI/MOD LOI/MOD

MODER LOI/MOD MODER LOI/MOD MODER MODN!

SLUDGE TREAl

.0526 MODER MODER

MODER MODER

LO\JMOD

MODER

MOOHI

MODER MODER

MODER MODER MODER

MODER LOI/MOD MODER

MODER MOOER

MODER MODER

MODER MODER

MODER MODHI MODNI MODER

MOOHI

.

FUll

MOOER

MOHIADVN

FULL

MODER

MONIADVN

FULL FULL

MODCOMP

MODADVN MQDADVN

FULL FULL

MODER MODER MODER

MODER

FULL FULL FULL

MODER MOOER MODER

MOOHI

PILOT

CQMPLEX

MODNI LOI/MOD

PILOT

MOOCOMP

LAB

MOOER

Table 3.7: Process Evaluation for LarGe Activated Sludge Plants 21

.

MODADVN MODADVN

MOHIADVN LOMOADVN MQOADVN MOOADVN MQOADVN MOHIADVN





AERATED LAGDDNS - FlO\l > 10 ODO



m'Id

$C05TS $C0515 EFFLUENT EFFLUENT EFFLUENT EFFLUENT SlUDGE SOPERATE $CAPITAL TOXEFF PHOSPHOR 55 BOO TOXSLDG

_1319

Al ternatives

1 on-Line control 2 optimize in ject 3 pickLe tiquor

lOllMOD LOIIMOD lOI/HOD MODHI MODHI MODHI MODHI MODER

4 PASS

5 sodium aluminate 6 PAC

7 ferric sulfate 8 ARMS

.06Bl MODHI MDOHI MODER MODER MOOER MODER MODER MODER

.1344 MODER MODER MODER MODER MODER MODER MODER MODHI

.0691

.0320

LOI/HOD MODER MODER LOIIMOD MODER LOIIMOD MODER MODER

MODER MODER MODER lOllMOO MODER lOllMOD MODER MODER

.0291 MODER MODER MODER MODER MODER MOOER MODER MODER

.0739

SLUDGE TREAT

.0526

MOOER MOOER MODHI MODER

MODER MODER MODER MODER

MODER

MODER

MOOER MODER MODHI

MODER MODER MODER

SLUDGE

DEVELOP

SIMPLIC

VARADVN

.2275

.0714

.05B9

REUSE

_D511 MODER MODER LOIIMOD MODER MODER MODER MODER lOllMOD

FULL

MODER MODER MODER MODER

MOHIADVN MOOADVN MODADVN MODADVN

FUll FULL LAB

MODER MODER

MOHIADVN LOHOAOVN

MODER MODER

MODADVN

DEVELOP

SIMPLIC

VARADVN

.2275

.0714

.0589

FUll

FULL FULL FULL

HOHIADVN

Total

10.732 :0.706

1°. 695

10.687 :0.679

1°·675 1°·667 1°·443

Table 3.8: Process Evaluations for Large Aerated Lagoons BIOFILTERS - FLOW

> 10 000

m'Id

EFFLUENT EFFLUENT EFFLUENT EFFLUENT SlUDGE SCOSTS SCOSTS PHOSPHOR S5 BOO SOPERATE $CAPITAL TOXEFF TOXSLDG

-

ALternatives

1 on-line control 2 non-trad clarif 3 optimize in ject 4 PASS

5 sodium aLuminate 6 PAC

7 ferric sulfate 8 RIM-NUT

9 optimized biofiLter 10 crystaLLization 11 ARMS

.1319 lOllMOD LOIIMOD LOIIMOD MODHI MODHI MODHI MOOHI MODHI LOIIMOD MODHI MODER

.0681 MODHI MODHI MOOHI MOOER MODER MODER MODER HIGH HIGH HIGH MODER

_1344 MODER MODER MODER MODER MODER MODER MODER LOI/ MODER MODER MODHI

SLUDGE TREAT

SLUDGE

.0526

.0511

.0691

.0320

_0291

.0739

LOI/HOD MODER MODER LOIIMOD MODER LOIIMOD MODER LOIIHOD MOOHI MODER MODER

MODER MODER MODER lOllMOD MODER LOI/HOD MODER lOWHOD MODER MODER MODER

MODER MODER MODER MODER MODER MODER MODER LOWHOD MODER MODER MODER

MODER MODER MODER MODER MODER MODER MODER lOWHOD lOWHOD

MODER MODER MODER MODER MODER MODER MODER MODHI

LOWMOD

MODHI MODER

MOOHI

MODER

MODER MODER

MODER MODER MODER HOOER MODER MODHI MODHI

FULL FULL FULL FULL FULL FULL FULL PILOT

MODHI

PILOT PILOT

LOWHOD

LAB

Table 3.9: Process Evaluations for Large Biofilters 22

-

REUSE

MODER MODER MODER

MODER MODER MODER MODER COHPLEX MDDCOMP

MOHIADVN HOHIADVN MODADVN MODADVN MOHIADVN lOMOADVN

MODADVN MODADVN

MODCOMP

MDHIADVN MODADVN

MODER

MOHIADVN

Total

10.732 10.718 1°·706 1°·687 10 . 679 1°·675

10.667 10 . 565 10.541 10 . 500 10.443







PNYS!COCNEMICAL - FLOW > 10 000 m3/d $COSTS $COSTS EFFLUENT EFFLUENT EFFLUENT EFFLUENT SLUDGE $OPERATE $CAPITAL TOXEFF PHOSPHOR SS BOD TOXSlDG

··

Alternatives 1 on-line control 2 non-trad clarif 3 optimize in ject 4 pickle Liquor 5 PASS

6 sodium aluminate 7 PAC 8 ferric suLfate

9 electrolysis 10 ARMS

.1319 LOWHOD LOWHOD LOWHOD LOWHOD MODN! MODN! MODN! MODN! MODER MODER

.0681 MODNI MODNI MODNI MODER MODER MODER MODER MODER MotNl MODER

SLUDGE TREAT

SLUDGE REUSE

DEVELOP

.0526

.0511

.2275

· · .1344 MODER MODER MODER MODER MODER MODER MODER MOOER MODER MODNI

.0691 LOWHOD MODER MODER MODER LOWHOD MODER LOWHOD MODER MODER MODER

.0320

MODER MODER MODER MODER LOWHO!J MODER LOWHOD MODER

MODER MODER

.0291 MODER MODER MODER MODER MODER MODER

MODER MODER MODER MODER

.0739 MODER MODER MODER MODN! MODER MODER MODER MODER MODER MODNI

MODER MODER

MODER MODER MODER MODER MODER

MODER MODER MODER

.

.

S!MPLlC

.0714

FULL FULL FULL FULL FULL

MODER MODER MODER MODER M':tJER

FULL

MODER MODER MODER

MODER MODER MODER lD\.lMOO MODER MODER MODER MODER MODER

PILOT

lD\JMOD

LAB

FUlL FULL

VARADVN

· · .0589

Total

MOHIADVN

:0.732 :0.718 :0.706 :0.695 :0.687 :0.679 :0.675 :0.667

MON!ADVN MODADVN MODADVN MODADVN

MOHIADVN lOMOADVN

MODADVN MDDCOHP MOHIADVN MOHIADVN MODER

10.516 10.443

Table 3.10: Process Evaluation for Large Physicochemical Plants SBR - FLOW > 10 000 m3/d

$C051S $COSTS EFFLUENT EFFLUENT EFFLUENT EFFLUENT SLUDGE $OPERATE $CAP!TAL TOXEFF PNOSPNOR SS TOXSLDG BOO

AL ternatives 1 bio-P

2 on-line control 3 optimize in ject 4 pickle liquor 5 PASS

6 sodium aluminate 7 PAC

8 ferric sulfate 9 RIM-NUT

10 crystaLlization 11 ARMS

. .

· ·

.1319

.0681

LOWHOD LOWHOD LOWHOD LOWHOD MODN! MODNl MOON! MODN! MODN! MODN! MODER

"100ER MODN! MODN! MODER MODER MODER MODER MODER N!GN MODN! MODER

.

.1344 MODER MODER MODER MODER MODER MODER MODER MODER LOW MODER MODN!

.0691

SLUDGE TREAT

SLUDGE REUSE

DEVElOP

· · .0320

.0291

SIMPLIC

VARADVN

· .0739

.0526

.0511

.2275

.0714

MOOADVN

1:0.746

HOHIADVN

:1°·732

MODER LOWHOO MODER

MODER MODER MODER

MOOER MODER MODER

MODER MODER MODER

MOOHI

FUll

MODER MODER

MODER MODER

FULL FULL

MODER MODER MODER

MOOER

MODER

MOOER

MOOHI

MODER

LO\olMOO

FULL

MODER

MODER

MODER

MODER MODER MODER

MODER MODER

FULL

MODER MODER MODER

MODER MODER

MON!AOVN

LOWHOD LOWHOD

MODN! MODN!

MOOHI

MODER

MODER MODER CQMr:LEX MOOCQHP MODER

lOHOADVN MODADVN MODADV!J MODADVN HOHIADVN

LOWHOD

LOWHOD

MODER

MODER

MODER MODER

LOWHOD MODER LOWHOD MODER MODER

LOWHOD

MODER

MODER

MODER LOWHOO MODER

LOWHOD MODER MODER

MODER

LOWHOD

MODER MODER MODN! MODN! LOWHOD

FULL FULL FULL

PILOi PILOT LAB

Table 3.11: Process Evaluation for Large Sequencing Batch Reactors

23

Total

.0589

MODAOVN

MOOADVN MODADVN

1

:0.706 :0.695 :0.687 :0.679 :0.675 :0.667 :0.565 :0.514 :0.443

4. Materials and Methods



The expenments and analyses described in this chapter were carried out at the Environmental Engineering Laboratories at McGill University's Department of Civil Engineering and Applied Mechanics. Batch sorption tests were carried out at room temperature (25°C ± 3°C) while jar tests were conducted in a constant temperature room at 25°C ± 1°C.

4.1 Batch Sorption Tests

[n order to determine the sorptive capacity of pure aluminum hydroxide for phosphate, batch sorption experiments were carried out. These tests conformed to the ASTM D4646-87 standard (ASTM, 1992) with the following exceptions. The reaction time was reduced to sixty minutes in order to more fully reflect full-scale treatment plant retention times. The samples were shaken on an orbital shaker at 150 rpm for the duration of the mixing phase of the test after which they were filtered on 0,45-!!m



membrane filters (Millipore) which had been soaked in deionized water for 24 h in accordance with Standard Methods (APHA et al., 1992). Borosilicate Erlenmeyer flasks with a capacity of 250 mL were used and these were carefully maintained as described in Section 4.3.1.

Phosphate solutions ranging in concentration from 2 ta 15 mg P IL were used in the test in order to reflect the concentrations found in the municipal wastewaters of the province. Since phosphates occur in different chemical forms, three different types of phosphates were used. These were orthophosphate (Anachemia KH2P04), condensed phosphate (Fluka (NaPOJ)I2_I3'NaO) and organic phosphate (Fluka CIOH14Ns07P·H20). The aluminum hydroxide suspension was prepared by dissolving aluminum sulphate (Fisher AI 2(SO,h'18HP) and adjusting the pH to 7.0 ± 0.1 with IN NaOH (Fisher). This suspension was allowed ta rest for 16 hours before use. The test samples were



composed of 1 part aluminum hydroxide suspension and 20 parts phosphate solution

24

for a total volume of 210 mL. One control flask with zero phosphate content



W,IS

exposed to the test conditions for each run.

4.2 Tar Tests

Tests were also conducted with full-scale wastcwaters exposed ta altlll1 sludg,e. These jar tests, which were carried out on a Phipps & Bird multiple paddle stirrer, conformed to A5TM D2035-S0 (A5TM, 1992) with the following exceptions. The total volume in the 2,000 mL Griffin beakers

WOlS

always 1,500 mL. The samples were mixed for

5 minutes at 100 rpm followed by 30 minutes at 30 rpm. They were then allowed to settle for 30 minutes before the supernatants were collected. A control jar of llndosed wastewater was always tested in arder to assess the phosphorus removal c'lpacity of the physical procedure alone. The jar placements on the multiple stirrer were randomized with a die for each run.

The jar tests using alum sludges as phosphate-removal agents were run with wastewaters



obtained from the Pincourt Wastewater Treatment Plant in Pincourt, Québec. This activated sludge plant receives wastewater with an average influent phosphate concentration of 4 mg PIL.

In order to simulate the addition of alum sludge to two different points ln the treatment train, tests were conducted on both raw wastewater sampled after the grit removal chamber and mixed liquor sampled from the end of the aeration basin. The alum sludge was obtained from the Pointe-Claire Potable Water Treatment Plant in Pointe-Claire, Québec.

The alum sludge was added to the wastewaters in such a manner as to obtain a total sample volume of 1,500 mL. The dosing reflected additions of 0% ta 12% sludge by total jar volume.



25

4.3 Analytical Methods



4.3.1 Maintenance of Glassware

Ali the glassware that was used throughout the experimentation described herein was maintaincd in the following manner. After use, glassware was rinsed with tap water and left ta soak overnight in a bath of dilute phosphate-free detergent. It was then repeatedly rinsed with tap water to remove detergent residues, rinsed twice with 10% hydrochloric acid and finally given two rinses with distilled wnter before being left to air-dry on a drying rack.

4.3.2 Phosphate Determinations

Ali phosphate determinations were carried out according te the HACH Test 'N' Tube procedure (I-IACH, 1992) using a Milton-Roy Spectronic 200 spectrophotometer at 890 nm. These methods are US EPA-approved variations on the procedures defined in



Standard Methods (API-IA, 1992). Reactive phosphate was determined by colorimetry after reaction with molybdate and ascorbic acid.

The acid-hydrolysable phosphate

fraction was determined after its conversion to reactive phosphate by boiling the sample at 105°C with sulfuric acid.

Total phosphate was converted to reactive

phosphate by boiling at 105°C with sulfuric acid and potassium persulfate.

The

recovery on ail of the digested phosphates were within 100% ± 10%. The method detection limit was determined by the procedure described in Standard Methods (APHA, 1992) and was found to be 0.01 mg P /L. Ali phosphate determinations were run in triplicate.

4.3.3 Aluminum Determinations



Ali aluminum determinations were carried out by atomic absorption spectrometry with

26

a nitrous-oxide/acetylene flarne on a Perkin-Elmer 3110 FAAS in '1cconbnœ \Vith



Standard Methods (APHA, 1992).

Total alurninum \Vas determined .1fter s.unple

digestion by boiling sampies with nit rie aeid. No ehernieal .1gents \Vere uscd ta control interferenees. The rnethod deteetion lirnit was determined by the method outlined by Perkin-Elrner (Perkin-Elmer, 1982) and was found to be 0.3 mg Al/L.

4.3.4 Other Analyses

Other physieal and ehernieal analyses were earried out; ail of them in aeeordance \Vith Standard Methods (APHA, 1992). A Radiorneter Copenhagen PHM62 with an 1\TC probe was used ta rneasure pH after the probe had been ealibrated with buffers of pH 4 and 7.

Alkalinity deterrninations were made with the same apparatus while the

sarnple was titrated with 0.02N H 2SO,.

Suspended solids were determined by sample filtration on Whatman 934-1\1-1 glass fibre filters and subsequent drying at 103°C-10SoC. Total solids were determined by drying





unfiltered sarnples at the same ternperature. Total volatile solids were measured after heating the sarnples in a rnuffle furnace at SOO°C ± SO°C.

27

5. Results



5.1: Phosphate Sorption on Aluminum Hydroxide: Synthetic Solutions

The literature cited in Chapter 2 deals predominantly with orthophosphate removai. Since the phosphates found in wastewaters also comprise condensed and organic forms, batch sorption experiments were carried out in arder ta determine the sorption behaviour of each form independently. Each 200 mL phosphate solution was dosed with 10 mL of 300 mg Al/L synthetic aluminum hydroxidc~ solution and mixed for one hour. The results of these experimems are shawn in Figure 5.1.

final P (mg PIL)

lS PHOSPHATE

"'ortho 12 + condtMtd *organlc 9

6 ...



3 0

0

,

"\"'

..

3

6

9

12

15

initial P (mg PIL)

Figure 5.1: Phosphate Removal with Aluminum Hydroxide The line plotted at an angle of 45 0 is the line of zero removaI. This line crosses the points through which the effluent phosphate concentration is equal ta that of the influent. Thcrefore, better removal is characterized by increased distance from the zero removal line.

The organic phosphate removal was the lowest of the three with a maximum of 40% rcmoval at a dose of 8 moles of aluminum per mole of phosphorus. The removals of



the condensed and orthophosphate fractions were similar ta each ether; being greater 28

than 95% Olt 8 mol AI/mol P. Typical ah1l11 doses Me bet\Veel1 2 omd -' mol AI/mol Il



(Metcalf and Eddy, 1991). Removals by aluminum hydroxide

Olt -'

mol AI/mol il were

70%, 85% and 15% for orthophosphate, condensed phosplute .md ol'ganic phosph.\tl" re!Spectively.

In order to minimize the effects of ionic strength and competing ions, the pH of the solutions in these initial experiments

WolS

not adjllsted.

I-1owcVel", each re.lgcnt

prodllced solutions with different initial pH values. The average pH values of the dilutions were 4.8, 5.7 and 3.6 for the orthophosphate, condenscd phosphate and organic phosphate, respective1y, ,md these varied by each set of phosphate species solutions.

il

maximum of 0.7 pH units within

Since the solution pH affects a\1I1l1inlll11

solubility, and since this solubilîty determines the amount of aluminum present in the solid phase, it

WOlS

deemed necessary to adjust the initial pH of the solutions so that

they would all be l'quaI.

The selected pH was that of the condensed phosphate

solution (pH 5.7) .lnd the other solutions were adjusred with lN NaOH.

Results

obtained for the altered solutions before and after this adjustment are shown in Figure



5.2.

ftnal P (mg PlI..)

15,.--------r-----------, PHOSPHATE . pH S.7 , 12

... ortho

" ....;:' ' ,

+o~anK

,"~.......

. ,~-

9

6

3

o ,.' o

~

.'

;~~:'\

......-

....

• 11

3

*~""

.'

.,- ,"

.'

pH 3..6

pH •.8

\

.' .

6

9

12

15

inilial P (mg P/L)

Figure 5.2: Phosphate Removals Olt Different Initial pH Lcvels



29

As can be seen in Figure 5.2, in each case the deviations in phosphate concentration for



the two pH levels arc small (less than 10% difference) except at the low initial orthophosphate concentrations where the absolute difference in final concentration was Jess than 0.1 mg l'IL.

This suggests that phosphate removal for these systems is

independent of pH for the pH values studied.

Aluminum solubility, however, is highly dependent on pH. Mechanistically, it is interesting to know whether precipitation or adsorption is responsible for the phosphate

1•• noval.

In either case, a knowledge of the quantities of aluminum and

phosphate in the resulting solid phase is desirable.

The residual aluminum

concentrations were measured in the filtrate of the same sorption test samples as used for Figure 5.2. The results are plotted against the final phosphate concentrations in Figure 5.3.

dose = 14.3 mg AUL

soluble alwninum (mg AIIL)

6~-----------~--=--, .,;fPHOSPHATE

s.



.;,.,..", ,..:

pHS:?.. : 4

.. .. ... .......:.. .>...

,0.",,,,,,.01,Al

fot

. . :. . . . ·~_··~~:~~·t·~~. B.~~:~:.7

3 .

,,'

2

:tt'"

o

.. +condtfUtd '*o'Jlllllc "orfho

i'"

-

~ .. - .. -+- .... -

3

ortlJopbospbatc al

pHS.1

6

..

9

.;f.

12

1S

fmaI soluble phosphate (mg PIL)

Figure 5.3: Residual Aluminum Concentrations

The soluble aluminum concentrations seem surprising at first glance. The aluminum measured after the sorption test with condensed phosphate is the highest of all; this was at the pH c10sest to the minimum solubility of aluminum hydroxide (pH 6.8). However, the pH values noted on the graph are those of the initial phosphate



solutions.

The final values, those that would determine the soluble aluminum

30

concentration, differed significantly From the initial v.llues. For eX.....'..

20

.. ;,......

......

~

. " ..'

SLUDGEAGE

.

-·0 daY$ -1- 1 day

;J';"

·...·-5 dayl

o• o

...... ......

369

12

alum aludge dote ('.lb of total jar volume)

Figure 5.12: Total Sorbable Phosph,\te Removal

When reviewing the removals of the phosphorus species sepanltely (Figures 5.13 and 5.14), it was noted that the major reason for the decrease in total phosphate sorption was the reduced affinity of the sludge for reactive phosphate. The capacity of the ahlln sludge for non-reactive phosphate was fairly constant over time.



pbospborus removed by dudge (% of IOrbable RP)

100,----------,------,-----, 80

....

60·

.•...•.1-...

....................»..........

40

..f 20 .

......

....;~:......... 3

..

.

.... ....

.. .. i

..

SLUDGEAGE

"·0 day, + 1 day

..:.. , . ; / .

+5days

...... o

.. '

.. ,1-'

6

9

12

alum aludge dose (% of total jar volume)

Figure 5.13: Reactive Phosphate Removal



43

phosphorus removed by sludge (% of sorbablc NRP) ;liJ---~~~-----------,

......

..:.~:;.~~.~~.~ .. "." .

.. '

SLUDGEAGE -·-0 d~Y1

+1 d~y '... 5 daY1 3

12

9

6

alum sludge dose (% of total jar volume)

Figure 5.14: N on-reactive Phosphate Removal

5.3.2: Mixed Liquor

The same experiment was repeated on mixed liquor with a sludge that was fresh and the same sludge after it had been allowed to age for one day. The characteristics of the



sludge and mixed liquor are shown in Table 5.2.

b

le

Sllldge

1

Parame/cr pH tOIJl sol ids

Mixed Li'!llOr

1 Units

mg/kg

pH

1

Day 0

1

Day 1

5.23

5.83

2,670

2,728

6.79

s~me JS

1

Day 0

suspended sol id!

mg/L

2,893

same as Day 0

tot;Ù phosp horus

mg/kg

53.0

same as Day 0

Table 5.2: Alum Sludge Characteristics

The experiments show that a greater reduction in the alum sludge's phosphate removal C'lp'lcity occurrcd \Vith one clay of alum sludge aging for mixed liquor ("'= 20% phosphate removal) than it clid with raw wastewater ("'= 10% phosphate removal) .



44

phosphorus rcmovcd by sJudgc (% or sorbablc l'Pl IOOr--------------



80

60

..

40

.. ' .. ' o

,.

.. 1"-

, •• l

.. ' ." .. '

'

'

::•.............

.,

20

.a.

.•..

SWDGEAGE

.. ....

.' 0 daYJ "1 day

'

fi

3

9

12

a1um sludge dose (% oftotallar volume)

Figure 5.15: Mixed Liguor Total Phosphate Remov;ll

Due to measurement errors in assessing the non·reactive phosphate removal in the control jar on the day of testing with the aged sludge, dctcrmination of the change in removal capacity for non-reactive phosphate

WOlS

not possible. Howcvcr, the change

in the reactive phosphate was available and it is shawn in Figure 5.16. As can be seCll,



the removals :lre guite similar ta those seen in the total phosphate removal graph above.

pbosphorus rcmovcd by s1udgc (% of somablc RP)

100..------------------, BO·

....... .. , ••••• ./1•••••••

60·

............... 40·

.. .'

.. '

SLUDGEAGE

'

• 0 rJoyJ day

~'1

3

fi

9

)2

a1um aludgc dose (% of lolal jar volume)

Figure 5.16: Mixed Liguor Reactive Phosphate Rcmoval



45

It should be noted that the same sample of mixed liguor was used on both days of



lesling. The phosphorus concentration in the control jar supernatant was four times higher on the second day than on the first. This is thought to be due to a change in the settleability of the solids caused by the mixed liguor being stored. Due ta this unfortunate occurrence, the decrease in the alum sludge's phosphate removal capacity may have been due in part ta the increased unsettleable solids on the second day. However, in light of the results obtained in the raw wastewater experiments, it is unlikely that the decrease in phosphorus removal was due only ta changes in the mixed liguor.

The following chapter attempts ta identify the prevailing mechanism of phosphate removal by aluminum hydroxide. The practical implications of alum sludge use are elaborated upon with respect to both technical and economic factors.





46

6. Discussion



6.1: Phosphate Removal with Aluminum Hydroxide

As discussed in previous chapters, several mechanisms may be responsible for the sorption of phosphates on alum sludges. These include precipitation of phosphates with the soluble aluminum found in the sludge, i.e. as Al(PO,), adsorption of phosphates on to the alum sludge solids, sweep-floc removal of particulate phosphates or any combination of these.

The experiments with synthetic solutions described in Chapter 5 involved phosphate solutions with no particulate P fraction. Hence, for those mixtures, only precipitation and adsorption are likely candidates for removal mechanisms.

6.2: Precipitation



Precipitation occurs when the conditions in a solution yield concentrations of precipitating ions that exceed those supported by the saturated condition. For example, a saturated solution which contains one mole of the precipitate A,Bh is said to be in equilibrium with a moles of ion A and b moles of ion B. This equilibrium is described by Equation 6.1 (adapted from Benefield et al., 1982). (6.1)

The ionic molar concentrations of

A(,q)

and

B(,q),

denoted by [A] and [B], can

theoretically vary in any way that satisfies Equation 6.2 (adapted from Benefield ct al., 1982). K sp = [A] a [EP

(6.2)

K,p is called the solubility product of the precipitate and it is constant under a given



set of thermodynamic conditions. Amorphous aluminum hydroxide (am-AI(OI-l)J), the

47

sorbcnt uscd in the expcriments described in Chapter 5, has a solubility product of



10-32 .3' (Kotrly and Sucha, 1985). Thus (6.3)

which shows that aluminum hydroxide solubility depends directly on pH. Note that ail concentrations in this work are expressed in moles per liter unless otherw:se indicatcd.

The aluminum hydroxide suspensions used in the synthetic phosphate solution sorption cxperiments had initial soluble aluminum concentrations which were very low «

0.1

mg Al/L). However, sorne aluminum may have been solubilized on contact with the phosphate solutions.

Therefore, aluminum ions could have become available for

aluminum precipitates other than aluminum hydroxide (i.e. aluminum phosphate).

When several ions are present in solution, and those ions could form a variety of



different precipitates, solubility theory states that the solid that will form is the one that minimizes the common ion solubility (Benefield et al., 1982). For example, in a solution which contains both aluminum and orthophosphate, two binary precipitates could form: AI(OH)3 and AI(PO,). The solid that will control is the one which minimizes the aluminum solubility.

6.1.2: Adsorption

Precipitation describes r. process by which soluble ions become insoluble. In contrast, adsorption refers to the formation of complexes on the surfaces of precipitates which have already formed. Surface complex formation with anions and weak acids usually involves the exchange of one surface ligand, such as OH, for another, such as H 2PO. (Stumm, 1992).



48

An adsorption equilibrium constant similar te the solllbility prodllct may bc dcfincd,



and this constant is referred te as the surface complex formation const'lnt, K',

A

typical surface complex formation constant expression, and its associatcd stoichiomctric equation, are given in Equations 6.4 and 6,5 (adapted from Swmm, 1992). S rcfcrs lo the surface and S-L refers to the surface and its sorbed ligand, L. ln the cqllations below, the exchangeable ligands are represented by A and B. K

S

=

[S~B] [A] [S-A] [B]

(6.4)

(6 .5)

6.4: The Solid Phase in the Synthetic Solution Sorption Experiments

The synthetic solution sorption experiments described in Chapter 5 were conducted with five groups of solutions, each group consisting of various dilutions of a single



phosphate type at a constant pH value. The groups were: orthophosphate at pH 4.8, orthophosphate at pH 5.7, condensed phosphate at pH 5.7, organic phosphate at pH 3.6 and organic phosphate at pH 5.7.

Soluble aluminum measurements produced detectable results in only three cases: condensed phosphate at pH 5.7, and organic phosphate at pH 3.6 and pH 5.7. Within each of these three groups, the residual soluble aluminum increased with increasing soluble phosphate concentration (see Figure 5.3). This implies that the aluminum and phosphate formed soluhle complexes which increased in abundance with increasing initial phosphate concentration.

As stated in Chapter 3, aluminum ions hydrolyse when exposed to water.

An

expression for the total soluble aluminum concentration which considers only the



monomeric hydrolysis species (taken from Kotrly and Sucha, 1985) is shown in 49

Equation 6.6.



[Al]

T;;

[Al) t] + [Al (OH) 2+] + [Al (OH) ;] + [Al (OH)~] + [Al (OH)~]

(6 . 6)

This expression, at any given pH, yields a minimum aluminum solubility since the inclusion of polymerie species and/or soluble aluminum complexes wouId increase the value of [AllI"

Each of the monomeric species in Equation 6.6 can be replaced by an expression consisting of an equilibrium constant, [AP+] and [H+]. The values of the constants have been determined by several researchers and are therefore well-defined. Ail of the stability and solubility constants used in this thesis were taken from Kotrly and Sucha (1985) and they are listed in Appendix C. Combining Equation 6.6 with Equation 6.3, along with the relevant equilibrium constants, allows for the graphical representation of the minimum possible aluminum hydroxide solubility. This is shown in Figure 6.1



for 25°C and zero ionic strength.

108 {Alh 12,.-----------------, 10 ,

.

2· .. ·· .. ···

..

01---...-----------'7""'-1 .2

..

-4 .... -6 0

l

2

;)

..

5

6

7

8

"

10 11 12 13 14

pH

Figure 6.1: Theoretical Aluminum Hydroxide Solubility



For comparative purposes, Figure 6.2 shows a plot of the soluble aluminum

50

concentrations that were measured during synthetic solution sorption expcrimcnts,



along with the theoretical curve. Three possible cases could occur whcn comparing measured values to those calculated using Equations 6.3 and 6.6.

Case 1: Measured values lie on the theoretical curve This would iOiply that aluminum hydroxide is the controlling solid and th.tt no soluble polymeric or mixed complex species exist.

Case Il: Measured values lie above the theoretical curve Aluminum hydroxide is the controlling solid but soluble aluminum complexes other than monomeric hydrolysis species are present.

Case III: Measured values lie below the theoretical curve A solid other than aluminum hydroxide is controlling smce the measured aluminum values are less than the minimum possible values for aluminum



hydroxide.

log [Al],

6r------------r-----, EXPERIMENT ........ -w-rluJ +tMMIIHlIf1HHJ

4,

:«.,...1Ik (pH UJ



• lJIIPk (pH 1.7)

ol--"