How to Select a Chemical Coagulant and Flocculant. - SSWM 1997 How to... · Colour Solids Contact Clarifier Industrial Application General Use Ion Exchange Temperature Dissolved Air

Alberta Water & Wastewater Operators Association22th Annual Seminar March 11- 14, 1997

- 1 -

Various pesticides, organiccompounds and metals,cyanide,dioxins and furans, fluoride,nitrilotriacetic acid (NTA), totaldissolved solids, trichloroacetic acid,trihalomethanes (THMs).

How to Select a Chemical Coagulant

and Flocculant.Anthony S. Greville.

Easy Treat Environmental.

Abstract

In many water treatment processes theselection of the chemical regime is ofcritical importance. The mechanicalequipment will remove watercontaminants to a reasonable level, but tomeet the increasingly stringent Federaland Provincial licensing requirementschemical coagulation, flocculation, anddisinfection are necessary. This paperwill address several topics that will helpthe water treatment plant operator selectthe most appropriate chemical treatmentprogramme for the needs of thecommunity that the plant services.

Why do we Chemically Treat Water?

Water is essential for life as we presentlyknow it and in North America we havebecome accustomed to receiving goodquality water at a reasonable cost (onworld wide terms Canadian drinking wateris provided at an extremely low cost). Intoday's increasingly complex society thedemands of the consumer, the medicaland scientific communities, and thereforethe Municipal, Provincial and Federalregulators, have caused the qualityguidelines for safe drinking water to be

reviewed. In many cases what wasconsidered acceptable by all segments ofsociety just a decade ago would now bethought of as unsafe. In an era ofchanging regulations and guidelines it isdifficult to define what is considered to be"good" drinking water, but reference tothe Guidelines for Canadian DrinkingWater Quality (Sixth Edition) as well asthe Provincial Licensing Authority willallow for a decision to be made. At thepresent time it can be anticipated thatchanges will be made with respect to thefollowing parameters [1];

Once the community at large has madethe decision as to the delivered quality ofthe finished water that will be enjoyed bythe consumer, the practicalconsiderations have to addressed.


- 2 -

Invariably mechanical means will beemployed to help achieve the treatmentobjectives, but since no mechanicalprocess is 100% efficient, chemicalenhancement will often be necessary.

Another reason for appraising the utility ofchemical treatment is when capital costsare being balanced against operatingcosts. Either the size/expense of a pieceof equipment can be reduced if efficiencycan be increased by implementingchemical treatment, or else physicalspace limitations do not allow for theinstallation of a large process additionand therefore optimization of existingequipment has to be considered first.

Chemicals typically find utility in theremoval of suspended, colloidal anddissolved solids from water, includingcalcium and magnesium hardness,mineral turbidity, organic colour and otherorganic substances, and undesirablemicrobiological species that can cause

health concerns in humans. The fourbroad categories of chemicals used arelime for precipitation softening,coagulants and flocculants for theremoval of suspended and colloidalsolids, powdered activated carbon fortaste and odour, and disinfectants for theremoval of pathogens.

Selection of Chemical Species.

There are three fundamental variables inwater treatment, all three of which willhave a significant influence on the type ofchemical that could be usefully employedin a particular application. The threevariables are;

1) Raw Water Quality.2) Process Equipment.3) Treatment Objectives.

These three variables can be furthercategorized as shown in the table below;

Raw Water Quality Process Equipment Treatment Objectives

Alkalinity Settling LagoonPotable Application

Partial SofteningFull Softening

pH Direct Filtration

Turbidity Sedimentation + Filtration

Colour Solids Contact Clarifier

Industrial ApplicationGeneral UseIon Exchange

Temperature Dissolved Air Flotation

Hardness Mixing Intensity

Taste and Odour Sludge Disposal

If the above variables are reviewed priorto embarking on the coagulant and

flocculant selection process aconsiderable amount of time and


- 3 -

needless effort can be saved. Anunderstanding of how the variables effectwater chemistry will allow the operator tomake sensible pre-screening decisionsand let him/her focus on optimizing theprocess to achieve the treatment goals.

Raw Water Quality

Clearly the quality of the raw water andthe contaminant classification, has tohave a significant impact on the type ofchemicals used for liquid-solidsseparation. There are however severalfactors to consider;

1) The amount of alkalinity present inthe water may eliminate somecoagulants from consideration.

2) The amount of turbidity present mayonly determine the amount ofcoagulant that may be required.

One also has to be aware of how the rawwater quality will change as a function ofthe time of the year.

Alkalinity

Alkalinity is of critical importance whenselecting a metal salt coagulant such aspolyhydroxy aluminum chloride (PACl),aluminum sulphate (alum), or ferricsulphate. All these materials need somealkalinity to drive the hydrolysis reactionsthat allow the coagulants to function. Ifthe water has a low alkalinity, below 50mgL-1, then the use of some of the moreacidic metal salts may be precluded. Inthese instances there are two options,either add supplemental alkalinity (asNaOH, Ca(OH)2 or Na2CO3), or use a

high basicity coagulant (>50% basicity)such as PACl or ACH. If the water to betreated carries a very low alkalinityloading then the use of artificial alkalinitywill always be necessary. In such a caseit might be useful to try a combination ofacidic and basic aluminum salts, PACl,ACH or alum, together with sodiumaluminate. Should the alkalinity be >50mgL-1 then, in general, there will besufficient present to drive mostcoagulation reactions. However, if thecoagulant dosage has to be higher (by afactor of two) than the raw water alkalinityit may be necessary add some alkalinityto drive the hydrolysis reactions tocompletion.

pH

The pH of the water could alsodetermine/eliminate many treatmentoptions. If the pH is higher than 8.5 andDissolved Organic Carbon (DOC), oftenreferred to as colour, has to be removeda highly acidic coagulant that will drive thepH down to ± 7.0 will have to beconsidered. It may be necessary to addsome soda ash in order to bring theLanglier Stability Index back to zero aftersuch treatment. If the pH is acidic greatcare will have to be taken to ensure thatthe chemical reactions occur as desiredand that the finished water is stable,removal of colour will be easy. Ferricsalts often perform well in acidicconditions. The most challengingconditions occur when colour has to beremoved from a water that has a high pHand a low alkalinity. Careful depressionof pH without alkalinity destruction can berealized if gaseous CO2 and Ca(OH)2 areadded together. The choice of coagulantwill determine the extent to which pH has


- 4 -

to be depressed. This is a somewhatsophisticated approach and would not berecommended for a smaller communitywith a restricted capital budget.

Turbidity

The precipitation of mineral turbidity bythe classic coagulation and flocculationprocess is well defined and reasonablystraight forward. Turbidity can beclassified as being anionically chargedsilica particles. Often the effect thatturbidity has is dependent on the amountpresent rather than the classification. Inlow turbidity waters (<10 NTU) anorganic polyelectrolyte should not beconsidered. The choice of inorganiccoagulant should be one that quicklygenerates the Al(OH)3 sweep floc andwill form a stable sludge bed. Inmoderate turbidity waters (<100 NTU)the use of a general purpose inorganicsalt is preferred, and most will besuccessful if the other conditions areright. In high turbidity situations, or inthose instances where surface waterturbidity can increase very rapidly, aPACl blended with a polyepiamine isoften the best choice. Sludge bedheight, sludge volume, dewateringefficiency, and pH depression are allreasons to consider the PACl blend overlarge additions of alum or ferric sulphate. Organic polyelectrolyte on its own will beeffective, but the cost is high, and thereis the potential to blind downstreamfilters with a high dosage of epiamine orpDADMAC (> 4.0 mgL-1).

Colour

Dissolved Organic Carbon, DOC, colour,is the parameter around which a chemicaltreatment regime is built. Hydrophilliccolour is invariably more difficult toseparate from water than is hydrophobicmineral turbidity. The complexing ofcolour is dependent on the pH of thewater, the classification of the colourcolloid, and the ability of the coagulant tobreak the hydrogen bonds present. Thechoice of chemicals must be one that willcreate a water in which the colour will beleast stable (usually at a pH between 5.5and 7.0), the alkalinity will be preservedfor turbidity precipitation, and the finishedwater will be neither corrosive or scaling.

In many applications it is difficult for onematerial to be completely successful byitself, especially the inorganic metal salts. In these instances cost performanceeconomics dictate that a small amount ofan organic short chain polymer, usuallyfrom the pDADMAC family be utilized. Ifalum or ferric sulphate is the primarycoagulant, then the supplemental additionof pDADMAC will have to be via aseparate feed system. If any of the PAClpreparations are used, a one productblend can be selected.

Temperature

Temperature can affect the performanceof the inorganic metal salts that rely on achemical reaction. The coldertemperatures (<50 C) have a profoundeffect on alum and iron salts to the extentthat performance is often unacceptableduring the winter. It is not unusual for awater plant to have to heat the raw waterto a minimum of 80 C in winter to maintainadequate finished water quality. In


- 5 -

industrial applications carryover ofalumina or iron flocs can cause processnon-conformities and off specificationproduction. The non-sulphatedpolyhydroxy aluminum chloride choicedoes not appear to be as temperaturesensitive and is therefore a good firstchoice coagulant for cold waterapplications. Almost all the coagulantswill perform well in warmer waters, 100 C� T � 250 C.

Hardness

Calcium and magnesium hardness arepresent in all waters to some degree oranother. The amount of CaH and theend use of the water will determine thestrategy required to handle the presenceof these minerals. Typically lime isadded to allow for the precipitationsoftening process to take place. Limesludges are dense and will tend to settle,however, it is recommended that 10 mgL-

1 of an alumina coagulant be added tocapture the lime fines. It should bestressed that the coagulant is presentonly to capture the lime fines and not tocoagulate raw water turbidity. Limesludges cannot be returned to theenvironment, so dewatering or lagoonstorage is required, all coagulants shouldbe evaluated with respect to their abilityto dewater on the equipment in the waterplant.

The major criteria for efficient limesoftening is pH control, pH should bemaintained at 10.0 ± 0.2. A metal basedcoagulant will consume alkalinity,especially in a well buffered high pHwater, which could compromise thesoftening process. The best coagulant is

therefore a pre-hydrolysed species with ahigh basicity. PACl has been found to bevery suitable for lime softeningapplications. A flocculant is seldomneeded, but filtering is alwaysrecommended.

The only major problem encountered witha lime softening programme is if there isa needto soften at a high pH and removeorganic colour at a low pH. The only realsolution is to make a capital investment intwo clarifiers, arranged in series. Initiallythe raw water is treated in a conventionalway, at pH 7.0, a low basicity coagulantshould be added to ensure that a goodsludge bed is maintained for the strainingand filtering action. This is followed, in aseparate clarifier, by the lime softeningstep, at pH 10.0 with a high basicitycoagulant.

In municipal applications, where therequirement is to reduce hardness to<100 mgL-1 flocculants are notrecommended, however the water shouldalways be filtered. In those industrialapplications where the water is sent to anion exchange stage, hardness is reducedto �40 mgL-1 and a flocculant is alwaysused. Filtering is still required prior to theion exchange equipment. The advantageof a low basicity coagulant is even morepronounced in full softening applications,and the non-sulphated PACl is thecoagulant of choice.

Taste and Odour

Taste and odour can be controlled in avariety of ways, but one of the mostcommon is with the addition of powderedactivated carbon, (PAC). PAC generates


- 6 -

fine particles that have to be wellcoagulated. Most of the coagulants willsettle PAC/organic particles, howeverbecause of the very fine nature of thespecies it is better to choose a coagulantthat generates the densest sludge. Thecorrect choice will result in minimal pinfloc carryover, while the incorrect choicewill be characterized by a lowconsistency sludge bed and observablepin floc carryover.

Process Equipment

The process equipment does have someimpact on the choice of coagulant andflocculant, however the raw water qualityis the predominant factor.

Settling Lagoon

The presence of a pretreatment settlinglagoon will allow the water plant tomaintain, on a year round basis, aconsistent and much better quality of rawwater. A settling lagoon will enable thewater plant to let any temporary poorquality surface water by-pass the plantin-take, while concurrently permitting allthe settleable solids to be naturallyprecipitated from the raw water prior totreatment. The choice of chemicaltreatment should therefore be consistenton a year round basis, as should be theoverall treatment regime. A chemicalprogramme can be tailored to meetspecific needs and can be far lessforgiving than programmes that have tosatisfy fluctuating raw water qualitysituations.

Direct Filtration

A direct filtration plant will require acoagulant/flocculant combination that willreadily generate a robust yet filterable flocwith the raw water contaminants. Thefirst inclination is to believe that the largerthe floc the better, but this is often not thecase. A large floc will be easily filteredfrom the water but all the filtering actionwill take place on the top of the filter bed,leaving the majority of the media unused. A smaller, denser, and more robust flocwill also filter well but will use a greaterpercentage of the filter bed resulting in animproved quality of finished water andlonger run lengths before breakthrough. Those coagulants that hydrolyze quicklyto form a sweep floc will show up well in ajar test, however the sweep floc is veryfragile and may easily break up into finesthat will be difficult to filter. Theprepolymerized inorganic metal salts arethe best option in direct filtration units.

There will be a temptation to consider theuse of organic polyelectrolytes as filteraids, especially the pDADMAC's if thereis colour to remove. The use of theseproducts can be beneficial, but only as anaid to the primary coagulant. Should thedosage of these materials be above 4.0mgL-1 there will be a very real tendency tocause filter blinding and large "mud balls".

Sedimentation and Filtration

The configuration of the settling basinprocess will have to be considered,especially the chemical feed point and theintensity of the flocculators. If there aregood flocculators, or mixing chambers,with several minutes detention time (>5minutes) then the best choice is a noncatalyzed prepolymerized metal salt. Such a chemical regime will generate a


- 7 -

dense floc that will expel water naturally. This floc will settle well, and dependingon the retention time in the main settlingbasin may or may not require aflocculant. If the mixing energy ordetention time is insufficient, the bestoption is to consider a sulphated aluminasalt since the catalyzing characteristic ofthe sulphate will allow for reasonableresults. Since a sweep floc will surfaceadsorb the raw water contaminantswithout any real coagulation there willalways be a need for an organic polymerflocculant and carryover may become aconcern.

Again there may be a temptation toconsider the use of either a pDADMACor a polyepiamine, they will show well ina jar test. As with the direct filtrationprocess care has to be taken since if anyunreacted short chain polymer is carriedover to the filter plant, mud balls and filterblinding could result. There is also thepossibility that the cationic solutionpolymers may cause an overdose ofpositive charge that may result in poorsettling and extra loading on the filters.

Solids Contact Clarifier

A well designed and operated solidscontact clarifier followed by a properlysized filter plant will present fewchallenges to any one chemicaltreatment regime that it will not presentto another. There is typically goodmixing time and intensity in thecentrewell, where there is also theopportunity for floc collisions and growth. A "good" sludge bed will provide astraining and filtering medium, which canbe complimented by the inclined baffle

plates that many clarifiers are nowequipped with. If there is anyunacceptable carryover, the filter plantwill usually separate the pin floc from thewater.

There are several process variables thatcan enhance or detract from theperformance of a chemical treatmentprogramme. Raw water temperature,chemical feed point(s), turbine speed,sludge bed height, and clarifier blowdowncan all be adjusted so as to optimize theperformance. All chemical treatmentprogrammes, if it is the correct choice forthe raw water and the treatment goals,should be able to be optimized in a solidscontact clarifier.

Dissolved Air Flotation

Dissolved Air Flotation (DAF) Units arebecoming increasingly common in watertreatment. In DAF units coagulant isadded to the raw water or wastewater justbefore the flocculators. A robust floc hasto be formed quickly since upon enteringthe DAF unit the raw water is mixed withan air pressurized stream of effluentwater. The drop in pressure allows theair to be released and the bubbles float tothe surface of the flocculator/clarifier. The coagulated particlesabsorb/adsorb/entrap the air bubbles andalso float to the top of the vessel. Athickened sludge forms at the top whichis periodically removed.

The coagulant has to be able to form avery robust sludge that will resist fracturein the flocculating chamber. A limitationof the DAF unit is that any sludgeparticles that fracture will form a very


- 8 -

small, almost colloidal, particle that willpass through the filters and reduce thequality of the finished water. For thisreason it has been found that theprepolymerized coagulants, that are heldtogether with chemical bonds, providethe best sludge particle and give the bestfinished water quality. The alumsludges, that rely on floc collisions tobuild a settleable sludge tend to fractureeasily. For a DAF unit, a sludge stabilitytest is the best indicator of performance. Do not be fooled by large fluffy flocs thatlook good in a jar test. Prepolymerizedsalts are often the best choice andshould be screened first.

Mixing Intensity

The mixing intensity or available mixingtime will have a greater effect on theperformance of the chemical rather thanon the choice. However, if there isinsufficient mixing to optimize thecoagulation process it is recommendedthat a sulphated aluminum orpolyaluminum salt be considered. Asweep floc of aluminum hydroxide will begenerated that will allow for chemicalenmeshment of colloidal particles. Whilea true coagulation process will not havetaken place, with the aid of a goodflocculant, the solids should beseparated from the water. If the mixingregime is satisfactory it is better to moveaway from the catalyzed materials andrely on hydrolysis to generate thecationic species required for coagulationand particle collisions for flocculation. Chemical efficiency will be muchimproved and there may not be a needfor a flocculant.

Sludge Disposal

Sludge disposal is becoming a veryimportant operating parameter for manyutilities. In some jurisdictions it is stillpossible to return aluminum sludges to areceiving water, including those that haveunder gone pH adjustment with Na2CO3

or Ca(OH)2. However, lime softeningsludges that contain increased Ca(OH)2

concentrations and all iron based sludgeshave to be dewatered and sent to a landfill site. Settling lagoons and dredging orelse the transfer to a municipalwastewater plant are also options forthese more hazardous sludges. The typeof equipment available to the utilityoperator, or else the capital moniesallocated to plant improvement will have asignificant impact on the coagulantchoice.

Treatment Objective

The treatment objective will obviouslyhave an impact on the type of treatmentregime that is considered. While theprocesses available to all water plantowners are the same there is a differencein operating philosophy between amunicipal supplier and an industrialfacility. In the municipal segment thewater plant serves the need to producequality water as an end in itself. Due topublic health concerns the municipalwater plant operator will always strive tosupply the best quality water possible.


- 9 -

Alkalinity (measured in mgL-1)pH (measured in pH units)Turbidity (measured in NTU)True Colour (measured in TCU)Total Organic (measured as UV absorbance Carbon or mgL-1 TOC) Temperature (measured in degrees Celsius)Hardness (measured in mgL-1 as CaCO3)Taste and Odour (subjective test)Aluminum (measured in mgL-1)Iron (measured in mgL-1)% Reduction Rates (measured in %)Filter Requirements

An industrial facility will often treat waterfor use in the process of the plant. Quality is determined by the needs of theprocess only and not by public healthconcerns. Economics play a verysignificant role in the selection of atreatment regime since water treatmentis considered a cost of business. Toconfuse the issue, an industrial facilitythat only has to meet process qualityguidelines may not be able to live withpoor quality water for long due to thepotential to manufacture off-specificationproduct.

In summary, a municipal water plant isoften run to meet regulations andcommunity standards while an industrialprocess facility is run by economics (anindustrial potable water plant has to meetall the government regulations).

If a water treatment plant operator isplanning to optimize the coagulationand flocculation functions within theplant the first option is to consider allthe fixed parameters, as shown above. It is quite possible that byunderstanding the process and thechemicals available that many optionswill be eliminated and that a few willstand out as being good selections. Arelatively easy process to select thebest from all the good options can thenbe initiated.

Determination of AnalyticalRequirements

When evaluating a potential chemicaltreatment programme it is essential thatthere be a method to measure theperformance of the regime under

consideration. In most cases there areseveral parameters that have to beconsidered. Following the same processoutlined to select the most likelycoagulant will also give a good indicationas to the analytical requirements. Rawwater quality, process equipment andtreatment objective will determine theimportant operating parameters and alsoidentify the parameters that need to beconsidered in the laboratory.

While each situation is different, there area number of consistencies throughout theindustry. The following list of analyticalprocedures should always be consideredbefore beginning an evaluationprocedure. Without familiarity with thenecessary procedures and equipmentrequired in order to obtain accurate andmeaningful results, time spent on jartesting would be worthless.

Assembly of RequiredEquipment


- 10 -

The most common test method used isthe jar test. This concept has beenavailable to water plant operators forapproximately 50 years and has provento be reliable. Jar tests will tell anoperator which products or combinationof products will work satisfactorily andgive some indication of the approximatedosage (order of magnitude) required. Once a programme has been selectedand optimized in-plant the jar test canprovide some guidelines for operationalchanges to respond to a change inprevailing raw water conditions. Theusefulness of the jar test procedure isdependent on the protocol used duringthe test, therefore it is important that thetest be designed with the plant and thetreatment objectives in mind. Before ameaningful series of jar tests can beconducted the following planning stepsmust be addressed;

1) Selection of equipment.2) Preparation of reagents.3) Preliminary test protocol compatibility

study.4) Development of final test procedure.5) Determination of data requirements

and necessary calculations.

Selection of Equipment

The most important piece of equipmentin a jar test procedure is the physical jaritself. Other equipment considerationswill be the drive mechanism of the stirrer,the design of the impeller, the presenceof a water bath, analytical equipment,and general supplies.

Type of Jar

There are several types of jar availablebut perhaps the two most common arethe 1 litre circular jar and the 2 litresquare jar. The 1 litre circular jar is morepopular, however it is the least efficientand representative of all the jar optionsand should be avoided if possible, [2]. The 2 litre square jar with a sampling port10 cm from the water line is the bestchoice and of course the most expensive!

The limitations with the 1 litre circular jarinclude the following observations;

i/ Small volume of water increases themargin of error when scaling up to fullplant scale.

ii/ The water will rotate with the paddle inthe jar reducing the effective rate ofmixing.

iii/ It is very difficult to develop goodsettling data in a 1 litre jar.

iv/ Very little supernatent water toanalyze if several tests are required.

If a circular jar is to be used then a 2 litrecapacity jar with stators should beconsidered. Figure 1 below shows anarrangement that allows for good mixing,settling and sample extraction. Thisdesign is often referred to as the HudsonJar. The 2 litre square jar (sometimescalled the Gator Jar) (Figure 2) with asampling port 10 cm below the water lineis the best choice as it offers severaladvantages that address the concernsabove;


- 11 -

Figure 1 Figure 2Hudson Jar Gator Jar

ii/ Larger volume reduces the margins oferror.

ii/ Square walls reduce water rotationmaking stators unnecessary.

iii/ Settling velocity can be easilydeveloped with the 10 cm samplingport configuration.

iv/ Plenty of supernatent for analyticalwork.

iv/ Thicker walls and lower thermalconductivity will reduce temperaturechanges during the testing period.

Stirrer Drive Mechanism

There are two basic types of stirrer drivemechanism that can be used to turn theimpeller. The gear driven unit will have 4or 6 impeller shafts run off a singlevariable speed motor. The impellerstherefore all turn at the same speedduring a jar test sequence. The speedcan typically be varied between 5 and250 rpm.

The second option is the magnetic stirreroption. The advantages the magneticstirrer offers is extra space at the top ofthe jar is available for coagulant additionand the stirrers can be operated


- 12 -

independently of each other. In this wayit is possible the determine the effect ofvarying mixing intensities in a side byside comparison.

A modification to the magnetic stirrerconcept is a recent development thatallows for paddles to be electricallyoperated independently of each other. This allows for variations in both mixingintensities and durations. Since eachpaddle can be activated on its own,mixing can begin concurrently withtreatment chemical addition rather thantrying to add six different chemicals at thesame time. A major improvement withjust one limitation, at present this designis only available for the 1 litre circularjars, and not for the 2 litre Gator Jars.

Impeller Design

There are several impeller designsavailable, the most common being theflat paddle, magnetic stir bar and amarine turbine. Each design offersdifferent characteristics, the mostimportant of which is the mixing energy,or velocity gradient, G. In general the flatimpeller offers the highest velocitygradient, in a 2 litre square jar a G of 350S-1 can be obtained with 200 rpm of thepaddle. This is sufficient to duplicatemost treatment plant conditions. Themagnetic stirrer is less intense, up to 200S-1 can be obtained with 200 rpm. Themarine paddle (shaped like a ship'spropeller blade) is the least efficient, at200 rpm a velocity gradient of 70 S-1 canbe generated in a 2 litre Gator Jar. This

makes a marine paddle unsuitable formost water treatment applications.

Water Bath

A water bath is recommended when coldwater (<100 C) is to be evaluated. Manychemical treatment programmes aretemperature sensitive, and fluctuations of5 - 100 C in temperature, when the rawwater temperature is < 20 C areunacceptable. The best way todetermine if a water bath is required is torun a quick jar test with raw water atambient temperature. Put anothersample aside and let it warm up 100 Cand run an identical test. If there is adifference in floc formation rates andsize, then a cold water bath is necessary.

The best design is to fabricate a clearbath that will contain all the jars to abouttheir mid-point height. A sample of rawwater at ambient temperature should beallowed to flow through the bath keepingthe water temperature in the sample jarsconstant. This flow should be maintainedthroughout the test period, including thesettling time.

Analytical Equipment and GeneralSupplies

There should always be a good supply ofbeakers, syringes, pipettes, filter papersand all the analytical equipment requiredto be most efficient. Without thenecessary tools to properly perform thetest and to accurately assess the resultsthe time and effort spent will be wasted.


- 13 -

Determination of Jar TestProtocol

The most important variable in the jar testis the actual test protocol used to performthe evaluation. A proper procedure willallow for good and meaningful results, aninappropriate procedure will render theresults meaningless. Each plant will ofcourse have a different protocol, and itmay take some time to design the bestprocedure, however it should beremembered that the test protocol onlyhas to be defined once!

Preparation of Stock Solutions

The preparation of the stock solutionsused for evaluation of course has asignificant bearing on the outcome of thetest. In general coagulants andflocculants are diluted for test work forthree reasons;

1) Ease of handling.2) Ensure good mixing in the jar.3) In the case of flocculants, to allow for

activation.

There has been a suggestion that theuse of microsyringes (no preparation bydilution of stock solutions is required)offer advantages over dilution andstandard syringes. The main differenceis that the charged coagulants can beginto lose their effectiveness once they havebeen diluted. To a degree this is true,however, if the coagulant stock solutionis discarded after 2 hours there appearsto be no significant reduction inefficiency. PAM based flocculants, whendiluted to 1% maintain their charge for atleast 24 hours, allowing for a full day's

work.

If dilution is required it is best to prepare1% solution of coagulant and 0.1% offlocculant. A complete outline of theprocedures involved in the preparation ofstock solutions if presented in Calculationof Chemicals Dosages for Water andWastewater Treatment [3], prepared byEasy Treat Environmental. A briefsummary is presented below;

Example 1.Coagulants and Solution Polymers.

For products such as alum, PACl, ferricand ferrous salts, lime, NaOH and thesolution cationic polymers, simpledilutions are all that will be required topermit accurate delivery of product. Solution strength and relative densityshould always be factored into youdilution calculations.

In general it is recommended that a 1%solution (by weight) of coagulant beprepared. This will allow for accuratedelivery, good mixing and reasonablevolumes to handle.

A 1% solution is defined as being 1% byweight, i.e 1g (or 1000 mg) of chemical in99g (or 99 ml) of water for a total of100g. 1 ml of this 1% solution willcontain

1 × 1000 mg = 10 mg100

This added to 1 litre is 10 mg per litre, or10 mgL-1 (or 10 ppm)


- 14 -

If you add 1 ml of a 1% solution to a testvolume of 1 litre, the added dosage is 10mgL-1.

If you add 1 ml of a 1% solution to a testvolume of 2 litres, the added dosage is 5mgL-1.

How is a 1% solution prepared? Insimple terms, a 1% solution is preparedby adding 1g of chemical to 99 ml ofwater. Therefore, since it is easy tomeasure 99 ml of water, the only realsource of potential error is in measuring1g of chemical, especially if the relativedensity is different from 1.0.

Polyhydroxy Aluminum Chloride

To prepare a 1% solution of PACl onehas to consider the relative density. 18%PACl has a relative density of 1.37. Thismeans that 1 ml will weigh 1.37g.

Thus 1g = 1 ml = 0.73 ml. 1.37

So a 1% solution of PACl can beprepared by adding 0.73 ml of neatchemical to 99 ml of water.

Aluminium Sulphate

To prepare a 1% solution of alum (drybasis) one has to consider both relativedensity and solids content. Typicallyalum has a relative density of 1.34, and asolids content of 48.5%.

Thus 1g = 1 ml = 0.746 ml 1.34

0.746 ml will weigh 1g. However, sincealum is 48.5% actives content, 1g ofsolution will contain only 0.485g of alum.

Therefore1g of dry alum = 0.746 × 1 = 1.54 ml

0.485

So a 1% solution of alum (dry basis) canbe prepared by adding 1.54 ml of liquidalum (48.5%) to 99 ml of water.

Example 2High Molecular Weight EmulsionPolymers.

For high molecular weight emulsionpolymers, dilution before use serves twopurposes, activation of the chemical andaccurate delivery during evaluation.

A 1% solution, that is reasonable for thecoagulants or solution polymers, is notpractical for the emulsion polymers, thesolution would be too viscous to handle,making it difficult to perform good jartests. Typically a 0.1% or 0.2 % solutionis prepared in two steps. First a 0.5% -1.0% solution is prepared to allow foractivation, then a second dilution to 0.1%- 0.2% is made to allow for ease ofhandling and accurate delivery.

To Prepare a 0.5% Solution :For Activation of the Polymer.

Dry Powder


- 15 -

If you add 0.5 ml of a 0.1% solution to atest volume of 1 litre, the added dosage is0.5 mgL-1.

If you add 1 ml of a 0.1% solution to a testvolume of 1 litre, the added dosage is 1.0mgL-1.

If you add 0.5 ml of a 0.1% solution to atest volume of 2 litres, the added dosage is0.25 mgL-1.

Weight accurately 0.5g and dissolve in99.5 ml of water.

Emulsion :

Measure 0.5 ml and dissolve in 99.5 ml of water. Density

To prepare a 0.1% Solution : For Ease of Handling.

Once activation/dissolution is complete,dilute by volume 20 ml of the 0.5%solution to 100 ml. That is add 20 ml of0.5% solution to 80 ml of water togenerate 100 ml of 0.1% solution.

[Since the polymer has a relative densityof 1.02 and 0.5g has been added to 99.5ml of water, with a relative density of1.00, it is reasonable to assume that the0.5% solution has a relative density of1.00.]

If you remember that for the coagulants,1 ml of a 1% solution added to 1 litregave a 10 mgL-1 dosage, then it followsthat:

Determination of Test Protocol

The actual test procedure has to becarefully determined, to be representativeof the plant and so yield meaningfulresults. There are several criticalparameters that have to be consideredbefore a preliminary test can beperformed. Three importantconsiderations are mixing intensity,mixing duration and settling time. If oneis attempting to duplicate the actions of aclarifier, the critical operating parametersinclude;1) Velocity Gradient.2) Flocculator Detention Time.3) Settling Time.4) Selection of Filter Medium.5) Preliminary Test Protocol

Compatibility Study.

Velocity Gradient

The Velocity Gradient (G measured in S-

1) is a measure of the mixing energy thatis present during both the flash mix andthe flocculation stages of watertreatment. In a jar test, velocity gradientcorresponds to the fast mix and theslow mix. The higher the velocitygradient the more intense the mixing. The best way to determine the velocitygradient is to refer to the operationsmanual or consult with the designengineer or the manufacturer. G is acalculated number, the calculation ofwhich is beyond the scope of this paper. For more information on how todetermine G please refer to [4] and [5].

Once the velocity gradient has been


- 16 -

determined the important step is toduplicate the same mixing intensity in the

jar test. In general the rapid mix G is 200

Figure 3 Figure 4Velocity Gradient for a Gator Jar Velocity Gradient for a Hudosn Jar

- 300 S-1 while the slow mix is at thelower end of the 15 - 150 S-1 range, 40 S-

1 being typical. When the jars to be usedin the jar test apparatus evaluation arepurchased it is normal to find a chart thatcorrelates impeller speed, impeller designand velocity gradient for the jar. Thediagrams for the flat paddle impeller inthe 2 litre Gator Jar (Figure 3) and aHudson Jar (Figure 4) are shown below. It is straight forward to determine the rpmof the paddles once the G forcesexperienced at various locations withinthe plant are known.

Flocculator Detention Time

The calculation of detention time is atonce very simple, yet an accurate time is

almost impossible to determine. Detention time is calculated using asimple formula;

Td = V/Q

where

Td = Theoretical Detention Time (inminutes)

V = Tank Volume (in cubic meters)Q = Flow Rate (in cubic meters per

minute)

The problem lies in the fact that almostevery system has some sort of shortcircuiting inherent in it. It is quitepossible that the real detention time couldbe as much as 50% lower than thetheoretical. The only way to verify the


- 17 -

real detention time is to run some tracertests in the plant. Of course this onlyneed be done once, when the operator issatisfied that the hydraulics of the plantare understood the all jar tests can betailored to the real conditions.

In general, rapid mix is designed for 20seconds to a minute (normallyreasonable) while flocculation is designedfor 30 minutes (the source of mosterrors).

Please note that the velocity gradient andthe detention time for both the rapid mixand the flocculation stage have to bedetermined, there should be 4 numberson which to base the development of thejar test protocol.

Settling Time

Settling time is one of the greatestsources of error when running a jar test. The main error arises because thesettling time is often too long. While along settling time gives very good results,they are often not very meaningful. Given time, all coagulated andflocculated particles will settle to thebottom of a motionless jar. The objectiveis to duplicate the plant process and beable to determine differences betweenchemical treatment regimes. Settlingtime is therefore an operating parameterthat has to be measured carefully.

An example of how extended settling

times can obscure any meaningfulcomparison between treatment regimesis demonstrated in the graph below(Figure 5). If a sample is drawn from thejar after 3 minutes (3 on the SettlingVelocity axis) there are discernabledifferences between the treatmentregimes. If, however, settling is allowedto continue for 30 minutes (0.3 on theSettling Velocity axis) then there are nodifferences at all, the lines converge. The test results provide no meaningfulrelationship that will allow for comparisonbetween treatment regimes.

The best method to duplicate settlingtime is to determine how long it takes fora particle to settle 10 cm in the clarifier. This can be done by calculation andverified by physical examination. Oncethe time taken for a flocculated particle tosettle 10 cm has been determined thesettling time in the jar has also beendetermined. The 2 litre Gator Jars, with atap 10 cm from the water line are veryuseful, and far more reliable than thecircular jar that demands water besyringed from the surface.

To calculate settling time one has todetermine the Surface Loading of theclarifier or sedimentation basin. Oncethe surface loading is known the settlingvelocity can be determined, and then it isstraight forward to calculate the time itwill take to settle 10 cm. The table belowoutlines settling velocity and time to settle10 cm.


- 18 -

Figure 5 Settling Velocity and Sampling Time

Surface Loading Rate gpd/ft2 m/h

SettlingVelocitycm/min

Sampling Time for 10 cmmins

180 0.3 0.5 20

360 0.6 1.0 10

720 1.2 2.0 5

1440 2.4 4.0 2.5

3600 6.0 10.0 1

Note : 1 gpm/ft2 = 1440 gpd/ft2 = 2.4 m/h is equivalent to a settling velocity of 4 cm/min.

The calculation for settling time can beverified by extracting a sample of waterfor the clarifier and measuring the time ittakes for 80% of the particles to fall 10

cm. In general this estimated settlingtime will be close to the calculatedsampling time and can be used as aguideline in the jar test.


- 19 -

Selection of Filter Medium

In almost all instances the settled water ispassed through a filter before it is sent tothe distribution system.. It thereforemakes sense that there should be afiltering step in the jar test procedure. Ifthe jar test has been well run, filtering thesettled sample, drawn form 10 cm down,will provide interesting results. The mostobvious is a distinction between true andapparent colour. Recently the concept of"filterability" has been gaining attention,alumina may be intentionally precipitatedafter coagulation in order to filter it out toreduce residual. Modern DAF units arealso looking at using the filter plant moreefficiently.

There are a large number of filterspapers available from which to chose. Which one should be chosen forchemical treatment evaluation? Asbefore, the best advice is to obtain a filterpaper that best mimics the filter plant. This is easier said than done, similaritiesbetween the laboratory bench and the fullscale plant are few.

The selection of a specific filter andmethod of use depends on theinformation required and the ability of themethod to reproduce a filtered waterquality that is similar to the full scaleplant. A guide to filter performance istabulated below.

Selected Filter Efficiency Practical Use

0.45 µm membrane filter Natural organics must bedissolved to pass through

True vs Apparent ColourReduction in DOC

1.0 µm glass fibre filter General Purpose Filter forWater Treatment Plants

Colour & TOC Reduction,Chlorine Demand Studies

8.0 µm (Whatman 40)filter

Open filter that allows forbroad comparisons

Direct Filtration Plantsand comparison of FilterAids


- 20 -

Preliminary Test ProtocolCompatibility Study

Once the test parameters are known,mixing intensity and duration for both therapid mix and the flocculation mix, andthe settling time have been calculated, apreliminary procedure can be written. Atthis time it is best to perform acompatibility study that compares theresults from the jar test with the full scaleapplication in the plant.

Aliquots of plant settled water and jar testwater are collected and analyzed for theprimary operational parameters ofconcern. Always analyze for turbidity,colour and pH, as well as any other testsdeemed important. A simple bar chartshould be drawn that compares the fullscale plant results with the jar testresults. A correlation of greater than85% should be experienced on allparameters before the preliminary testprotocol can be finalized. Should therebe inconsistencies the items to revieware velocity gradients and detention timecalculations. Do not make the mistake ofadjusting the jar test procedure basedsimply on differences between turbidityresults in the jar and the plant. Figure 6below shows an example of a typicalcompatibility graph. The selection of thefilter can also be analyzed in a similarfashion.

Selection of ChemicalTreatment Regime

The hard work as been done! Byfollowing the process outlined above,much of the frustration with jar tests hasbeen eliminated.

1) The operation of the plant is wellunderstood.

2) Coagulants have been pre-screenedto select only those that will performwell.

3) Analytical requirements have beendefined and prepared for.

4) Laboratory equipment has beenselected.

5) A jar test protocol that givesmeaningful results has beendeveloped.

6) A filter protocol has been defined.

7) A compatibility study has shown thateverything above is in order.

There are now only three questionsremaining;

1) Which coagulant/flocculantcombination provides the bestresults?

2) How much has to be added?

3) How much will it cost?

The first two questions can be answeredrelatively easily.

Selection of Coagulant

The first step is to select the coagulant ofchoice and make an estimation of theoptimum dosage.


- 21 -

Figure 6 Preliminary Jar Test ProtocolCompatibility Study

The selection of a coagulant and itsoptimum dosage is made by acombination of observations andmeasurements. The objective is togenerate a "pin-floc" that coagulates thehydrophobic mineral turbidity and thehydrophillic organic colour. The DOCloading is often the determining factorwith respect to both coagulant choice anddosage [1] [6]. A good starting point is toconsider 5 mgL-1 of Al2O3 or 2 mgL-1 ofFe3+ for every 1.0 mgL-1 of DOC. pHdepression should also be factored intothe initial thoughts.

A suggested procedure is as follows:

1) Set the paddles to 10 rpm.2) Add any oxidants as required.3) Add the coagulant under

consideration at the requireddosage.

4) Add any pH adjustment chemicalsas required.

5) Set the impeller speed to therequired rpm for rapid mix andcontinue for the predeterminedduration.

6) Observe the performance of thecoagulant to reduce colour duringthe flash mix stage.

7) Set the impeller speed to therequired rpm for flocculation andcontinue for the predeterminedduration.


- 22 -

8) Observe the formation of pin flocs,and record the time fordevelopment.

9) Continue to observe the flocs andrecord any developmentalmilestones.

10) Make a note if the flocs developvery quickly but fail to grow afterthe initial formation.

11) Look for the colour of the waterbetween the flocs and record theobservation.

12) Look for the presence of anyfloating materials in the jar.

13) Set the impeller speed to zero andremove from the jar. Begin thesettling period.

14) Classify the flocs with respect tosize.

15) Classify the flocs with respect tosettling rate.

16) Observe and record if any flocsfloat to the surface.

17) Record if the temperature risesand causes a release of oxygen.

18) Observe and record if there areany flocs that "hang-up" in thecentre of the jar.

19) After the predetermined timewithdraw samples at 10 cm fromthe water line.

20) Perform all the required analyticaltests on the samples.

21) Return to the jar and look for signof continued "hang-up".

22) Slowly switch on the impeller andincrease rpm until the sludge bedlifts, record rpm for each chemicalregime.

23) Turn-off the impeller and observethe settling rate of the sludge bed.

24) If an analysis for a contact-time-dependent reaction, such as theformation of THMs, is required, itmay be necessary to set aside thesamples and let sit for the totaldetention time within the plant,most likely several hours. Underthese circumstances several 500ml beakers will be required.

25) Select the best coagulant at it'soptimum dosage.

Selection of a Flocculant

A flocculant may or may not be required. Once the coagulant and it's optimumdosage have been determined the sameprocedure outlined above can befollowed to determine whether or not aflocculant shows some benefits.

Selection of Final Programme

Once the chemical treatment programmehas been selected a final set of testsshould be performed to determine thebenefits of the flocculant. A series ofeight tests should be performed asfollows;

1) Settled water from the plant.

2) Settled water from the jar withoutflocculant.

3) Settled water from the jar withflocculant.

4) Settled water from the jar using


- 23 -

incumbent chemical treatmentprogramme.

5) Filtered water from the plant.

6) Filtered water from the plantwithout flocculant.

7) Filtered water from the jar withflocculant.

8) Filtered water from the jar usingincumbent chemical treatmentprogramme.

At this stage you are ready to make youdecision as to a chemical treatmentprogramme.

The jar test is a very useful tool if theapproach is correct. The key is to thinkabout what your objectives are anddevelop a protocol that will allow you to

achieve the objectives. Having taken thetime to prepare properly, the jar test canprovide the following information;

1) Which coagulants will work in theplant.

2) Which coagulants are unsuitable.

3) A ranking of the coagulantefficiencies.

4) An indication as to the order ofmagnitude for the required dosageto meet the treatment objectives.

5) A reasonable expectation ofanticipated results.

It should be emphasised, however, thatthere is only one jar test that reallycounts and that is in the large jar on theoperating floor!

References

1] Water Treatment Principles and Applications. A Manual for the Production of DrinkingWater. Canadian Water and Wastewater Association, 1993.

2] Operational Control of Coagulation and Filtration Processes, M37. American WaterWorks Association, 1992.

3] Calculation of Chemical Dosages for Water and Wastewater Treatment. Easy TreatCorporation, Calgary, Alberta. 1997.

4] Cornwall, D. A. and Bishop, M. M. Determining Velocity Gradients in Laboratory andFull Scale Systems. Jour. AWWA, 75:9:470-475. 1983.

5] Fair, G. M., Geyer, J. C. and Okun, D. A.. Water and Wastewater Engineering. JohnWiley and Sons, New York, 1968.

6] Greville, A. S. and Nalezyty, J. A. Polyhydroxy Aluminum Chloride - The "Hi-Tech"


- 24 -

Coagulant with a Future. Alberta Water and Wastewater Operators AssociationSeminar Proceedings, 1994.

© Easy Treat CorporationMarch 1997.

Documents

How to Select a Chemical Coagulant and Flocculant. - SSWM 1997 How to... · Colour Solids Contact Clarifier Industrial Application General Use Ion Exchange Temperature Dissolved Air