...—. 6Ai\iTAfinN (IR

; LiBRARY, INTERNATIONAL r̂ FI:-. C5.NTRE FOR COMMUNITY WA.E;

AMD SANITATION (IRC)/ ?.0. box ^3iyc, 2509 AHJ The i-ic. "=!. ;37C.; 0 ;'*•;•(? oxt. ?4:/:.4?I,

? RK: fe^

Dcpi

GF.REj ractei

2526Q HI

(196'iher. 77i<? American City

Private commtmica-

High-Rate Sedimentation in Water TreatmentWorks

—Gordon Cu/p, Sigurd Hansen and Gordon Richardson—A paper presented on Jun. 4, 1968, at the Annual Conference, Cleve-land, by Gordon Culp, Research Mgr., Sigurd Hansen, ResearchEngr., & Gordon Richardson, Lab. Mgr., all of Neptune MicroFLOC,Inc., Corvallis, Ore.

IT has long been recognized that asettling basin should be as shallow

as possible and that detention timesof only a few minutes can be used invery shallow basins. For example, aparticle settling at a rate of 1 in./minrequires .120 min to fall to the bottomof a conventional clarifier of 10 ftdepth. If the basin were 2 in. deep,this particle would fall to the bottomin only 2 min. In 1904, Hazen J pre-sented his argument that settling ba-sin efficiency is dependent primarilyupon basin depth and overflow rate,and is independent of detention time.He proposed basin depths of as littleas 1 in. Over 20 years ago, Camp 2

proposed settling basin depths of 6 in.and total settling basin detention timesof 10 min.

A detailed literature review by theauthors 3 showed that there have beenmany attempts at applying the shallowdepth sedimentation principles pro-posed by Hazen and Camp. Wide,shallow trays were generally insertedwithin basins of conventional design.These attempts met with only limitedsuccess because of two major prob-lems : (a) the unstable hydraulic con-ditions encountered with very wide,shallow trays, and (b) the minimumtray spacing was limited by the vertical

clearance required for mechanicalsludge removal equipment. The au-thors have overcome both problems byusing very small diameter tubes ratherthan wide, shallow trays. Longitudi-nal flow through tubes with a di-ameter of a few inches offers theo-retically optimum hydraulic conditionsfor sedimentation and overcomes thehydraulic problems associated withtray settling basins. Such tubes havea large wetted perimeter relative tothe wetted area and thereby providelaminar flow conditions, as evidencedby very low Reynold numbers. Fisch-erstrom * felt a Reynolds number ofless than 500 in settling basins wouldbe most beneficial to the settling proc-ess. A 1 in. diameter tube, 4 ft long,through which water is passed at arate of 10 gpm/sq ft of cross-sectionalarea has a Reynolds number of only24, while providing an equivalent sur-face overflow rate of 235 gpd/sq ft.The 3 min detention time of such atube settling device under these condi-tions certainly makes the cost and spacesaving potential apparent. The authorsnow have tube settling devices in oper-ation in many water treatment plantsthat are providing excellent clarifica-tion with settling detention times ofless than 10 min. The authors recently

681

682 GULP, HANSEN, & RICHARDSON four. AW IV A June 1968

made detailed presentations of their re-search on techniques for applying shal-low depth tubes as sedimentation de-vices.5 The purpose of this article isto present operating experiences withapplications in water treatemnt plants,as well as additional research data.

Basic Tube ConfigurationsThe authors have described 3i B two

basic tube configurations shown in Fig.1: (a) essentially horizontal and (b)steeply inclined. The operation of theessentially horizontal tube settlers 3 iscoordinated with that of the filter usedto clarify the tube settler effluent.

Each time the filter backwashes, thetube settler is drained completely.The falling water surface scours thesludge deposits from the tubes and car-ries them to waste. The water drainedfrom the tubes is replaced with the lastportion of the filter backwash water.The tubes are inclined only slightly inthe direction of flow (5 deg) to pro-mote the drainage of sludge during thebackwash cycle. If the inclination ofthe tubes is increased sharply (45-60 deg), continuous gravity drainageof the settleable material from the tubescan be achieved.8 The incoming solidssettle to the tube bottom and then exit

Chemical Coagulants

Raw Water

/a) First Portion ofBW to Waste

Tube Contents DrainedDuring Filter BW

Chemical Coagulants

Raw Water

BW to_Waste__>.

Sludge Dtawoff

Fig. 1. Basic Tube Settler Configurations

(A) is essentially horizontal tube settler; (B) is steeply inclined tube settler.

from the tubes by salong the tube bottomis established in whiitling to the tube bottta downward flowingtrated solids. This c<of solids aids in agglointo larger, heavier pagainst the velocityflowing liquid. Theremoval achieved inclined tubes eliminadrainage or backfillsfor sludge removal.

Horizontal Tube SeThe essentially hor

has been used prima000 gpd) to mediunwater treatment planination of operator aremoval from the clcant benefit. By c'.

Partial Lhl

Locati

AlabamaOhioAlabamaTennesseeOregonVenezuelaManitobaPennsylvania

WyomingMississippiMassachusettsNew MexicoWest VirginiaBritish ColumbiaPennsylvaniaIdahoOregon

Pap.Sub.NucReoRonSch.MmPou

Oil iMmMmRocMmMmPtmM mM u -

Jour. AW WA June 1968 HIGH-KATE SEDIMENTATION 6X3

ickwashes, thed completely.ice scours thetubes and car-water drained

d with the last:k\vash water.nly slightly indeg) to pro-

Ige during theinclination ofsharply (45-vity drainageVom the tubescoming solidsand then exit

First Portion ofBW to Waste

BW to Waste

from the tubes by sliding downwardalong the tube bottom. A flow patternis established in which the solids set-tling to the tube bottom are trapped ina downward flowing stream of concen-trated solids. This countercurrent flowof solids aids in agglomerating particlesinto larger, heavier particles that settleagainst the velocity of the upwardlyflowing liquid. The continuous sludgeremoval achieved in these steeply in-clined tubes eliminates the need fordrainage or backflushing of the tubesfor sludge removal.

Horizontal Tube SettlerThe essentially horizontal tube settler

has been used primarily in small (15.-000 gpd) to medium sized (10 mgd)water treatment plants, where the elim-ination of operator attention for sludgeremoval from the clarifier is a signifi-cant benefit. By draining the tubes

each time the filter backwashes, posi-tive sludge withdrawal from the clari-fier is achieved. The entire back-wash—tube drainage cycle is auto-mated. Thus, no operator judgmenton when or how much sludge to with-draw from the clarifier is required. Aschematic or package water treatmentin which tube settling has been usedis shown in Fig. 2. The detention timewithin the tubes, at design flow, is 6min. The impact of this low residencetime on the plant dimensions is well il-lustrated by the 6 ft high by 6 ft wideby 14 ft length dimensions for a com-plete 100 gpm water treatment plantproviding flocculation, sedimentationand filtration in one rectangular, pre-fabricated, steel plant. As shown inFig. 2, the coagulated raw water mayfirst be passed upwards through afluidized calcite column that automati-cally maintains the pH in the proper

TABLE 1Partial List of Installations oj Water Treatment Plants Using Horizontal

Tube Sdllcrs and Mixed-Media Filtration

settler.

Location

AlabamaOhioAlabamaTennesseeOregonVenezuelaManitobaPennsylvania

Wyoming;MississippiMassachusettsNew MexicoWest VirginiaBritish ColumbiaPennsylvaniaIdahoOregon

Facility Served

Paper millSubdivisionNuclear reactorRecreational areaRecreational areaSchoolMunicipal i tyPower station

Oil field reuseMunicipal i tyMunicipality-pilotRecreational areaMunicipalityMunic ipa l i t yPower StationM u n i c i p a l i t yMunicipality

Plant CapacityKt>">

10020

10020202060

100

2002,000

2020

350350100100500

Treatment Problem

10-100 JTU turbidity10-65 JTU turbidity2-28 JTU turbiditv2-30 FTU turbidity30-150 JTU turbidity, 20 color10-40 JTU turbidi ty10-35 JTU turbidi ty5-15 JTU turbidity, pH 3.5, iron 2.8

mg/l, manganese 1.0 mg/1Oil and suspended solids3-5 mg/1 iron200 color10-20 ITU tnrbiditv10 JTU turbidity4.5 nig/I iron, 160 color25 JTU tnrbiditv, 1 mg/1 iron, 20 color10-1000 JTU tu rb id i ty , 20 color5-10 JTU turbidi ty , 20 color, algae

684 GULP, HANSEN, & RICHARDSON Jour. A W WA

range for good coagulation. If alumis overfed, the calcite will automaticallybuffer the coagulated water. The useof calcite for this purpose eliminatesone possible cause (overdosage of co-agulant) of poor finished water tur-bidity.

The tubes used in these essentiallyhorizontal tube settlers are hexagonalin shape. The hexagonal tubes nesttogether to form a honeycomb pattern(Fig. 3). The tremendous wettedperimeter of such a tube configurationrelative to its wetted area provides ex-tremely low Reynolds numbers, wellwithin the laminar flow range.

After these data were published, con-firming operating data from severalplants with capacities of 30,000 gpd to3 mgd have been obtained from field in-stallations. The partial list of theseplants shown in Table 1 illustrates thewide range of water quality being sub-jected to treatment in plants employingthe basic flow pattern shown in Fig. 2.The majority of these plants are beingused for potable supplies and are oper-ating at total plant detention times ofabout 30 min. Although this detentiontime is much less than that in plantsemploying conventional sedimentationtechniques, data collected in prototype

FlocculantAid

Coagulant —Hypochlorite-

Raw Water

CalciteColumn

MechanicalFlocculator

.Mixed Media'/Filter Bed/

-fBackwashStorage

^1

Clearwell

PipingToUse

Filter Cycle~—••• Backwash

Fig. 2. Package Water Treatment Plant

Tube settling is used and the detention time within the tubes, at design flow, is 6 min.

Tube diameters of 1-2 in. andlengths of 2-4 ft are used in most wa-ter treatment applications. Hydraulicloading rates of 3-5 gpm/sq ft of tubeentrance area are generally used. Datahave been presented 8 showing that thetube settler, mixed-media filter combi-nation in a plant with total detentiontime of less than 30 min provides effi-cient clarification of very turbid waters(1,000 JU), highly colored waters, wa-ters containing filter-clogging algae,waters containing iron and manganese,and raw waters with taste and odor.

studies indicate the plants are actuallyrated conservatively. To illustratethis, Fig. 4 presents data collected dur-ing one of these studies with a plant,as shown in Fig. 2. During this test,each of the plant components was op-erated at rates considerably higherthan the design criteria normally usedfor these plants, that is. tube settlerdetention of 3 min rather than 6 min,filter rates of 8.5 gpm/sq ft rather than5 gpm/sq ft, flocctilation times of 5.4min rather than 10 min. The plant op-erating with an overall detention time

June 1968

of 16 min reduced averturbidity of 1.000 I U toless than 0.1 JU. A l t hrate of 8.5 gpm/sq ft reslively high initial beadle.1-the percentage of backthe end of the 8 hr ruiper cent. If the runtinned to the normal bac'value of 8 ft, the badmerit would have been 1cent. These data wellthe mixed-media filtergarnet) complements tito accomplish what in;classed as "high rate" •

Flow DistributionFlow distribution pro

less severe in the tubeearlier tray settling dc-the major reasons is

Tube Settler Ei

Hydraul ic :Loading,

Settl ing Tubes(sfin/sQ 'I "f

em! :itva)

PolyelectroDuse —I'f.c

3.7 03.7 . 0.14.04.05.0

0.20.50.2

5.05.06.756.758.58.5

0.200

0.20

0.2

Adder! ahead of flocctilato

nr. A W WA June 1968 HIGH-RATE SEDIMENTATION 685

ished, con-"n several000 gpd to>m field in-•t of theseistrates thebeing sub-employing

i in Fig. 2.^ are being1 are oper-n times ofs detention- in plants.imentation

prototype

Clearwell

ToUse

o. is 6 min.

re actuallyillustrate

lectecl dur-th a plant,g this test,ts was op-l)ly highermally usedube settlerhan 6 min,rather thanmes of 5.4:e plant op-•ntion time

of 16 min reduced average raw waterturbidi ty of 1.000 J U to an average ofless than 0.1 JU . Although the f i l terrate of 8.5 gpm/sq ft resulted in a rela-tively high initial headless (2.7 ft H/)).the percentage of backwash water atthe end of the 8 hr run was only 2.5per cent. If the run had been con-tinued to the normal backwash headlessvalue of 8 ft, the backwash require-ment would have been less than 2 percent. These data well illustrate howthe mixed-media filter (coal, sand,garnet) complements the tube settlerto accomplish what may certainly beclassed as "high rate" clarification.

Flow DistributionFlow distribution problems are much

less severe in the tube settler than inearlier tray settling devices. One ofthe major reasons is the extremely

stable hydraulic condition establishedwithin the tubes. As discussed earlier,laminar How conditions are establishedin the tubes. Thus, there are no turbu-lent flow conditions to promote shortcircuiting. Also, it has been foundthat the sludge deposits within thetubes act as flow distribution aides.If one tube is receiving more flow thananother, the more rapid buildup ofsludge in the first tube will cause someflow to be diverted to the second tube.The sludge deposits themselves thusact as a "self-orificing" device in thehorizontal tube settlers. Of course,care must be taken in the design ofthe tube inlet and outlet conditions sothat no great velocity gradients areestablished across either the inlet oroutlet faces of the tube modules.

Flow distribution analyses have beenmade in a 20 gpm plant, utilizing the

TABLE 2Tubc^Selller Efficiency in Preliminary Field Tests (Tube Length—2//,

Tube Diameter — 1.5 In.)

HydraulicLoading

Sett l ing Pulu^(Rpm.'SQ ft of

end area)

PolyelectrolyteDose — mg/l*

Aver. RawT n r h i d i t v

JU

Aver. Tubt;Settlor lift.Tnrbidi t v

76- "

Chat iRc- in Ki l te rHoadloss — 'it.

Waterfh'Kilter RateKpm/sqft

Tubes Inclined at 5°

3.73.74.04.05.0

00.10.20.50.2

250250250230250

37 1.017 ' 3.070if)21

1.750.54.0

6.36.36.86.88.5

Tubes Inclined at 60°

5.05.06.756.758.58.5

0.200

0.20

0.2

1

260290270240250250

62645134514

0.80.80.852.32.01.0

8.58.58.58.5

1010

* Added ahead of flocculator.

686 GULP, HANSEN, & RICHARDSON Jniir. A W WA .him- ]96S

salt tracer technique, as shown in Fig.2.15 A batch addition of a solutioncontaining 50 g/1 of sodium chloridewas dispersed into a stabilized flowof untreated source water in the floccn-lator.

Conductivity analyses of samples col-lected at 5 min intervals from the in-fluent and twelve vertical and lateraleffluent settler locations showed a vari-ation in peak value of ±10 /xmho ontriplicate runs. Comparing the vari-ation in conductivity with the corre-sponding salt as chloride concentrationindicated that sufficient linearity existedto allow the data to be evaluated on thebasis of conductivity.

Distribution of the inlet flow to theindividual tubes can be consideredsatisfactory based upon the minor vari-ation experienced in peak effluent con-ductivity and that the individual tubesamplings were found to have reten-tion times 4 per cent less than the theo-retic sampling retention period.

Analyses of the tube effluent com-posite sample with time indicated aminimum of short circuiting existed inthat the volumetric displacement effi-ciency was found to be 84 per cent as

TABI.K 3Perfvriiuitice of Pilot P/anl id III Steeply

Inclined Tube Modules

l-'lmvR;lto

i'/""/.«J fl

4444666666

Hoi-Tim,-mm

77*7*7*4.54.54.54.5*4,5"4.5

Pnlyi-li-i--trolyti-I lllSHRC

>»*//———————

000.10 20.10.20.20.20o

AVIM-URCRaw

TmbiililyJU

509250545352

23124649

255

Avcnigi-VttU-cl

TurbiilitvJU

18202013272127543551

* Klocculatnr dr i ' ntui not o]X:ru:riL

Tig. 3. Typical Tube Settler Module

This is for use in plants similar to the oneshown in Fig. 2.

compared with that of 63 per centlisted ° for ideal basins.

Steeply Inclined Tube SettlersThe solids that settle to the bottom

of a tube inclined at a steep angle(greater than 45°) will slide down thetube bottom continuously. This enablessludge removal to be achieved withoutdraining or backflushing the tubes.

Although the benefits of this con-tinual sludge removal phenomenon areobvious, the effects of steeply incliningthe tubes on the path of the particlesas they settle requires more detailedconsideration. The path traced by aparticle settling in a tube is the re-sultant of two vectors: V, the velocityof flow through the tube and vit thesettling velocity of the particle. It canbe seen in Fig. 5 that if the settlingsurfaces are inclined upward in thedirection of flow, the settling path ofthe particle is altered because the com-ponent of the settling velocity thatis parellel to the tube wall, v^. is op-posite in direction to the velocity vcc-

70

60

50

40

30

20

10

0

0-3

0.2

0.1

0

—— .. —— ._

-1

————— Ef-1

,̂ ^—

\ ______

Fig. 4. Open

The plant operating li'itwater turbidity of 'l,000

tiling at 8

tor V. If V is grearequired length of thedecreases as the angl,zero up to about 25-.2.5 •£,'») and then incre;infinity as the angleincreased to 90 deg.tray length continuesincreasing angle.

The research data csteeply inclined tubewater treatment appllpublished.5 The foil'this article present rdata on water treatm,

Laboratory Studies

During a continual1

studies published earwas designed to strtube inclination on

Jim,' 196S HIGH-RATE SEDIMENTATION 6S7

TAI3I.K 3"/ 1'tlnl Plant K-ilh Sl

i'K-il 'J'hbc Mutinies

1 l',,l1 t ru ly l i -

"'K/i

000.10.20.10.20.20.200

' Raw"'T n i h i i l i i y

JU

509250545352

2.M24649

255

ScttlVd'T u r l m l i t y

JU

1820201.?212i2154.5551

i that of 6.3 per cent;1 basins.

ed Tube Settlersint settle to the bottominccl at a sleep angle5°) will slide down thetinuously. This enablesto he achieved without^flushing the tubes.

benefits of this con-inoval phenomenon are•cts of steeply inclininge path of the particlesequires more detailedThe path traced by a

in a tube is the re-actors : V, the velocity

the tube and fs, the>f the particle. It can5 that if the settlinglined upward in the. the settling path of.•red because the coni--ettling velocity thattube wall, vsh. is op-i to the velocity vec-

4 i

Operating Time-hr

Fig. 4. Operational Data From Package Water Treatment Plant

The plant operating with an overall detention time of 16 win reduced average rawwater turbidity of 1,000 J U to an average of less titan 0.1 JU. The filter was oper-

ating at 8.5 gpm/sq ft and the time settler at 5 apm/sq ft.

tor V. If V is greater than z/a, therequired length of the settling surfacedecreases as the angle increases fromxcro up to about 25-30 deg (at V —2.5 i.'.,) and then increases, approachinginfinity as the angle of inclination isincreased to 90 deg. For V < va thetray length continues to decrease withincreasing angle.

The research data on performance ofsteeply inclined tube settlers in wastewater treatment applications have beenpublished.5 The following sections ofthis article present research and fielddata on water treatment applications.

Laboratory Studies

During a continuation of the researchstudies published earlier,3 an apparatuswns designed to study the effects oftube inclination on settling efficiency

(Fig. 6). It was during the operationof this equipment that the "self-clean-ing" phenomenon was first observed.Initial tests were carried out with fiveindividual tubes inclined at angles of0, 5, 20, 45, and 90 deg. The sludgesettling to the bottom of the tube in-clined at 45 deg was observed to bemoving continuously downward andeventually falling into the inlet plenum.

/ I X

Fig. 5. Effect of Tube Inclination onSettling Path of Discrete Particle

688

Some of the data collected on tubesettling efficiency at the various anglesof inclination are summarized in Fig.7. It was noted that tube efficiencyshowed an increase as the angle of in-

GULP, HANSEN, & RICHARDSON Jntir. A W WA

clination was increased to 35-45 degand then began to decrease as the angleof inclination was increased further.Results comparable to those obtainedat 5° inclination, however, were

Flocculated Water

o Inlet Plenum

ZHx—

ID——

- Flow Control Valve

Flexible Plastic Tubingfor Connection

Inlet Plenum

Flocculated Water_Mag-Mixer Used to Prevent

Deposition in Inlet Plenum

Tig. 6. Diagram of Test Apparatus

ElevationTJsed in Evaluating Effects of Tube Inclination

Fig. 1- Effe<

The tube

achieved at angles as st«It appeared that, as th<clination was increasedwhere the settled sludgedown the tube bottom, ;dilation occurred as trsettled and collided witupward moving floe. <the increased efficiervachieved at 5 deg. Acrease in angle eventvthe tube acting as an ihowever, and the advshallow tube depth arcin a decrease in efficien

Following observati<cleaning principle, the ;in Fig. 6 was modified

Jour. A W WA

sed to 35-45 degcrease as the angleincreased further,to those obtained

however, were

ID-"*—

ID-oo——

T

Yof Tube Inclination

3010 40 50

Tube Inclination-deg

Fig. 7. Effect of Tube Inclination on Settling Performance

The tubes used ivere 1 in. in diameter and 4 ft long.

achieved at angles as steep as 60 deg.I t appeared that, as the angle of in-clination was increased to the pointwhere the settled sludge began to movedown the tube bottom, additional floc-culation occurred as the heavier floesettled and collided with the smaller,upward moving floe, contributing tothe increased efficiency over thatachieved at 5 deg. A continuing in-crease in angle eventually results inthe tube acting as an upflow clarifier,however, and the advantages of theshallow tube depth are lost, resultingin a decrease in efficiency.

Following observation of this self-cleaning principle, the apparatus shownin Fig. 6 was modified to better define

the effects of various inclinations onsludge cleaning and on sedimentation

90

•S 80

70

60

* 50

40

30

Experimental ConditionsTube Size_.--.l-in. Diam 4-tt LongAverage Influent Turbidity. ...150 JU

^

—— — 4

k,——— —•' t—— — '

—- — ̂ _

^*

*->>^̂•̂ c^^

35

Tig. 8.

40 45 50Tube Inclination-deg

55 60

Effect of Tube Inclination onSettler Performance

690 GULP, HANSEN, & RICHARDSON Jour. A W WA June 1968

efficiency. The tubes were repositionedat angles of 35, 40, 45, and 60 deg.A slight decrease in efficiency (Fig. 8)was noted as the angle of inclinationapproached 60 deg. The self-cleaningaction, however, was enhanced as theangle was increased from 45-60 deg.To insure adequate sludge removalfrom the tubes, an angle of inclinationof 60 deg was used in the subsequenttests of multi-tube units.

Flocculated Water Inlet Weirv

Plywood Diaphragm-x. Overflow Weirx 10 Filter \

>2x4 Spreaders

Inlet Plenum -

Tubes

the tests with 2 ft tubes inclined at5 deg. The tubes were installed sothat the inlet and outlet conditions andthe total tube entrance area were thesame for both the 5 and 60 degtubes. The same mixed-media (coal,sand, garnet) filter was used to filterthe tube effluent in both cases. Thesurface water being treated was co-agulated with alum and, as noted inTable 2, polyelectrolyte was added in

Flocculated Water Inlet WeirsOverflow Weir

to Filter ~~-

' Outlet Plenum Plywood -

Fig. 9. Apparatus Used in Preliminary Field Evaluation of Steeply Inclined TubeSettler

(a) is section view of plant with tubes installed in unit at 5 deg; ( fc) is sectionview of plant with tubes inclined at 60 deg.

Field Evaluation—Pilot Plant ScaleA plant of the type shown in Fig. 2

was modified for the first field evalu-ations of the steeply inclined tubes.As shown in Fig. 9, the plant wasevaluated with the tubes at a 5 deginclination and at a 60 deg inclination.The tubes used were 2 ft in length andH in. in diameter. Because the tubechamber was originally designed for 4ft long tubes inclined at 5 deg, a por-tion of it was blocked off by the ply-wood diaphragm shown in Fig. 9 for

some cases. The data shown in Table2 show that the water quality producedby the 60 deg tubes at 8.5 gpm/sq ftwas lower in turbidity than that pro-duced by the 5 deg tubes at 5 gpm/sq ft. with 0.2 mg/1 polyelectrolyteused in both cases. The tube effluentquality was compatible with the mixed-media filter in all cases and filter runsto 8 ft of headloss were greater than18 hr, in all cases. Data collected dur-ing one run of the 60 deg tubes isshown in Fig. 10. The effect of poly-

electrolyte on tube setshown clearly by thein tube effluent turbidbeginning of polyelectihr. The filter effluemained less than 0.1the run. The tubetime under the condFig. 10 was 2.3 min.

Modular Tube UnitsBecause of the very

suits obtained in thetests, work was beguof n modular unit 01

Fig. 10. Prelimina

The filter was < > / •

Jour. A W IVA June 1968 HIGH-RATE SEDIMENTATION' 691

t tubes inclined ati were installed soutlet conditions andance area were theie 5 and 60 degmixed-media (coal,

was used to filteri both cases. Theg treated \vas co-i and, as noted in.tlyte was added in

ated Water Inlet Weir

eeply Inclined Tube

leg; (b) is section

ita shown in Table;r quality produceds at 8.5 gpm/sq ftlity than that pro-; tubes at 5 gpm/j/1 polyelectrolyte

The tube effluentile with the mixed-ses and filter runswere greater thanData collected dur-

60 deg tubes isThe effect of poly-

electrolyte on tube settler efficiency isshown clearly by the sudden decreasein tube effluent turbidity following thebeginning of polyelectrolyte feed at 4.2hr. The filter effluent turbidity re-mained less than 0.1 JU throughoutthe run. The tube settler detentiontime under the conditions shown inFig. 10 was 2.3 min.

Modular Tube UnitsBecause of the very encouraging re-

sults obtained in the preliminary fieldtests, work was begun on the designof a modular unit of steeply inclined

tubes that would minimize installationproblems. Following preliminary eval-uation of a great many potential de-signs, this development work resultedin the tube module design shown inFig. 11 (patent pending) in whichthe material of construction is normallyPVC. Extruded PVC channels areinstalled at a 60 deg inclination be-tween thin sheets of PVC. By inclin-ing the tube passageways, rather thaninclining the entire module, the rectan-gular module can be readily mountedin either rectangular or circular basins.By alternating the direction of inclina-

3.00

2.75

2.50

2.25 3

2.00

1.75

3 4Operating Time —hr

Fig. 10. Preliminary Field Test Data From Run of Steeply Inclined 60 deg Tubes

The filter u<as operating at 8.5 gpm/sq ft and the tube settler at 6.5 gpm/sq jt.

692GULP, HANSEN, & RICHARDSOX

J our. AW IV A June 1968

tion of each row of the channels form-ing the tube passageways, the modulebecomes a self-supporting beam whichneeds support only at its ends. Fol-lowing the development of this module,field tests of its efficiency as a sedi-mentation device were begun. A tubecross-section of 2 X 2 in. and a tubelength of 24 in. was used in the follow-ing tests.

The apparatus (Fig. 12) was set upat the authors' laboratory. The labora-tory ground water supply was used.A mud slurry was mixed with theincoming water to provide variouslevels of raw water turbidity. Alum(40 mg/1) was added as the primary

coagulant with the polyekctrolyte addi-tions made in some tests. Tube load-ing of 4-6 gpm/sq ft were investigated(tube entrance area = 9 sq ft) withraw water turbidities of 50 and 250 JU.The data from these tests are summar-ized in Table 3. In some runs, asnoted, the flocculator drive motor wasturned off to evaluate the tube effi-ciency without prior mechanical floccu-lation.

At the lower rate of 4 gpm/sq ft, theaddition of polvelectrolyte did notmarkedly improve the effluent clarity.When the flow rate was increased to 6gpm/sq ft, however, the higher settlingvelocities imparted by the polyelectro-

Flc . - . Measunr..;'.Yeir

Fig. 11. Module of Steeply Inclined Tubes

rig.77,

lyte were of significa:the flocculator was ibidities were fairly <out the run. When t'tor was not operatedfound that the efflu.creased with time astration beneath the1 his is not surprisingin and beneath the tulsource of flocculationthough it was foundblanket could be est;steeply inclined tubesubjecting the incomichanical flocculation.tened the developmer,After the sludge blantablished, the flocciturned off with no nothe clarified effluent qservation suggests th;ian upflow of newly •

Jour. All''IV A June 1968 HIGH-RATE SEDIMENTATION 693

lyelectrolyte addi-ests. Tube load-were investigated= 9 sq ft) with

of 50 and 250 JU.tests are summar-n some runs, asdrive motor \vas

;te the tube efn-nechanical floccu-

f 4 gpm/sq ft, the:trolyre did note effluent clarity.•as increased to 6he higher settling• the polyelectro-

Fie- - - - .Vieasufir.gX Weir

Sludge Draw Off

Fig. 12. Diagram of rlocculator and Settling Tank

This is used in evaluating tube module designs.

lyte were of significant benefit. Whenthe rlocculator was operated, the tur-bidities were fairly constant through-out the run. When the flocculator mo-tor was not operated, however, it wasfound that the effluent turbidity de-creased with time as the solids concen-tration beneath the tubes increased.This is not surprising as solids contactin and beneath the tubes was the primesource of flocculation in this case. Al-though it was found that the sludgeblanket could be established with thesteeply inclined tube settler withoutsubjecting the incoming water to me-chanical flocculation, flocculation has-tened the development of the blanket.After the sludge blanket was well es-tablished, the flocculator could beturned off with no noticeable effect onthe clarified effluent quality. This ob-servation suggests that by maintainingan upflow of newly coagulated water

through a region of high solids con-centration, the external flocculation re-quirements can be reduced significantly.This principle, of course, is recognizedand capitalized on by solids contactclarifier manufacturers.

These tests indicated the steeply in-clined tube modules shown in Fig. 11performed well as a sedimentation de-vice and were capable of producing

Fig. 13. Newport, Oregon, Water Treat-ment Plant

694 GULP, HANSEN, & RICHARDSON Jour. A W WA June 196ft

settled water turbidities consistent withthe capabilities of the mixed-media fil-ter under all the conditions shown inTable 3.

Field Evaluation—Plant Scale

The next logical step in the develop-ment of the steeply inclined tube set-tling process was a plant scale applica-tion. Fortunately, the city of Newport,Oregon, and their consulting engineerwere faced with a water treatment plantexpansion when the tube settling ex-periments were being completed. Theexisting 1.5 mgd plant (Fig. 13) con-sisted of a circular flocculator-clarifierfollowed by rapid sand filters. The raw-water characteristics are as follows:turbidity—10-20 JU; color—50-130units; pH—7.6 ; iron—4.7 mg/1; tem-perature—66°F; alkalinity—85 mg/1;and hardness—17 mg/1. The plantoperator normally applies about 35mg/1 alum and several mg/1 of acti-vated carbon to the raw water to pro-duce an acceptable finished water qual-ity. Pilot tests were conducted usinga plant of the type shown in Fig. 2. I twas found that an alum dose of 30 mg/1polyelectrolyte feed of 0.3 mg/1, and

Fig, 15. Plan View of Modified NewportClarifier

Fig. 14. Tube Modules Used In NewportClarifler Conversion

The modules are labeled A-D for pur-poses of identification.

1.5 mg/1 chlorine would enable thetube settler-mixed-media filter 'combi-nation to produce a finished water qual-ity of 0.15 JU turbidity. 5 color units.and 0.1 mg/1 iron. The tube settlerand mixed-media filter were both oper-ated at 5 gpm/sq ft in these pilot tests.

Based upon the pilot test results,modification of the full scale plant wasbegun late in 1967 in order to increasethe plant capacity from 1.5 to 3.0 mgdby installation of tube modules in theexisting clarifier, and by conversion ofthe rapid sand filters to mixed-mediabeds. As a first step, tube moduleswere to be installed in the existingclarifier to evaluate their performanceon a plant scale.

Because the available water supplyto the clarifier was to be limited by theexisting 1.5 mgd raw water pump dur-ing the early tests, tubes were installedin only a portion of the basin. Tube

modules of the type :were used. The tubover ,'; of the clarifierin Fig. 15. Tube rrused as support beaimodules, see Fig. 15none of the clarifierdue to a support sradial support beamsPVC pipe to supporiinlet well on one encweir at the other endbrackets, pipe, andtured in Fig. 17 whiuf a support moduleIS. Once the suppoplace, the remainingules were placed in p<tion of the basin ininstalled (Fig. 19) \the remaining part oiplastic barrier attachclarifier on each sidetion. Flow throughwas regulated by cliuf the effluent weirbasin by a galvanize'clamped to the weir ptests quickly led to ctire weir area and i

Fig. 16. Section ofClarili

Jour. A W WA Jinn-1968 HIGH-RATE SEDIMENTATION 695

Modified Newportr

'rf A-D for pur-fication.

mid enable thelia filter conibi-slied water qual-y. r color units.The tube settlerwere both oper-these pilot tests,lot test results.scale plant was

•rcler to increase1.5 to 3.0 mgd

modules in theiv conversion ofto mixed-media. tube modulesin the existing:ir performance

e water supply<i limited by theriter pump dtir-s were installede basin. Tube

modules of the type shown in Fig. 11were used. The tubes were installedover I1, of the clarifier surface, as shownin Fig. 15. Tube modules were alsoused as support beams for the uppermodules, see Fig. 15 and 16, so thatnone of the clarifier surface was lostdue to a support structure. Theseradial support beams were attached byPVC pipe to support brackets on theinlet well on one end and the effluentweir at the other end. The supportingbrackets, pipe, and modules are pic-tured in Fig. 17 while the installationof a support module is shown in Fig.IS. Once the support beams were inplace, the remaining tube settler mod-ules were placed in position. The por-tion of the basin in which tubes wereinstalled (Fig. 19) was isolated fromthe remaining part of the clarifier by aplastic barrier attached radially to theclarifier on each side of the tube sec-tion. Flow through the tube sectionwas regulated by closing off portionsii f the effluent weir in the rest of thebasin by a galvanized sheet steel platedamped to the weir plate. Preliminarytests quickly led to closing off this en-tire weir area and passing the entire

Lrl'( I

Tube Settler/ Modules \

I S 11 L f o*T ^Support Module

——————————

— ——— —————— 1

Fig. 17. Support Module Prior to Instal-lation

Fig. 16. Section of Modified NewportClarifier

plant flow through only the 210 sq ftof the basin covered with tubes. Al-though the nominal plant flow was 1.5mgd, flow measurement during the testperiod indicated the raw water pumpwas actually delivering only 910 gpm.Thus, the hydraulic loading on the tubearea was 4.3 gpm/sq ft.

Evaluation of DataFlow distribution analyses and ef-

fluent quality determinations weremade. Because only the existingperipheral effluent weir was used, less-than-perfect flow distribution was an-ticipated. Although radial collectionweirs would greatly aid in flow distri-bution, it was desired to first evaluatethe performance using only the existingweir.

696 GULP, H A N S E N , & R J C H A R H S O X Ji'in: .-I II' ll-'.-l

The rise rate of hydrochloric acidinjected into each tube module at sev-eral points was used to determine thevelocity in each module. For purposesof identification, the modules werelabeled A. B, C, and D, as shownin Fig. 15. The resulting flow distri-bution data are shown in Fig. 20. Aswas expected, the outer modulesnearest the effluent weir were receivingthe bulk of the flow and were operatingat 6.6 gpm/sq ft as compared to theaverage of 4.3 gpm/sq ft based uponthe entire surface area covered bytubes. Even with this flow distribu-tion, the tubes were performing as ef-ficient sedimentation devices. Asshown in Fig. 20, the tube effluentturbidity increased only slightly as theflow rate increased from 2 gpm/sq ftin module D to 6.6 gpm/sq ft in mod-ule A. The tube effluent contained nosettleable solids while samples col-lected earlier from the existing clarifierindicated that its effluent frequentlycontained 0.2-1 ml/1 settleable solids

Fig. 18. Support Module Being Installed

Fig. 19. Tube Modules in Place

and an average turbidity of 5.1 JL".The fact that the tube modules oper-ating at 4.3 gpm/sq ft (average) wereproducing better effluent than the clari-fier previously did operating at 0.7gpm/sq ft was further confirmed bythe fact that the length of filter runs in-creased from 26 to 60 hr followingthe modification of the clarifier.

At the time of this writing, the finalNewport plant conversion is beingmade. The filter media conversion isunderway with the design rate for themixed-media filter being 5 gpm/sq ft.The final clarifier conversion is beingmade with a ring of tube modules beinginstalled completely around the periph-ery of the basin. The tube ring willoperate at a rate of 5 gpin/sq ft at 3mgd. The peripheral location of themodules was selected to take advantageof the existing effluent collection sys-tem and the resulting flow distribution.If more of the clarifier surface is even-tually covered with tubes to furtherincrease the basin capacity, additionaleffluent collection weirs to better dis-tribute the flow would be needed torealize the full advantage of the ad-ditional tube modules.

SummaryShallow tubes are very e

mentation devices. Two bafigurations have been usedtially horizontal and (b)clined. Sludge is retnovtessentially horizontal tubiinatically draining them c:filter backwashes and rewith filter backwash watewater treatment plantsthese horizontal tubes willlion detention times of lesswith capacities of 20-2,0now in operation. In te?in this paper, a plant pro\lation, tube sedimentationmedia filtration, producedter (0.1 JU turbidity)water turbidity of 1,000overall plant detention tinFlow distribution analy?shallow horizontal tubesflow distribution to be rea>

The continuous selfsludge from tubes inclineangle allows sludge re!

achieved without the neecthe tubes. Laboratory ashow that an angle of 60continuous sludge renw\allowing the tube to funcficient sedimentation dplant tests have showninclined tubes to removefficiently at rates as highsqft . These tests led tcment of tube modules wstalled in an existing c'crease its capacity from 1mgd. Analyses of the fulation showed good clarifiof 6.6 gpm/sq ft. Thethe tube modules in an e>and the conversion of ;

June 1968 HIGH-RATE SEDIMENTATION 697

in Place

of 5.1 JU.idulcs oper-cragc) werean the clari-iting at 0.7jnfirmed byliter runs in-ar followingifier.ing, the finalm is being:onversion israte for the

5 gpm/sqft.sion is beingmodules l>eingd the periph-ube ring will•)in/sq ft at 3cation of theike advantageollection sys-

\- distribution.irface is cvcn-es to furtherity. additionalto better dis-be needed to,re of the ad-

SummaryShallow tubes are very efficient sedi-

mentation devices. Two basic tube con-figurations have been used: (a) essen-tially horizontal and (b) steeply in-clined. Sludge is removed from theessentially horizontal tubes by auto-matically draining them each time thefilter backwashes and refilling themwith filter backwash water. Over 20water treatment plants employingthese horizontal tubes with sedimenta-tion detention times of less than 10 minwith capacities of 20-2,000 gpm arenow in operation. In tests describedin this paper, a plant providing floccu-lation, tube sedimentation, and mixed-media filtration, produced potable wa-ter (0.1 JU turbidity) from a rawwater turbidity of 1,000 JU with anoverall plant detention time of 16 min.Flow distribution analyses show theshallow horizontal tubes enable goodflow distributien to be readily achieved.

The continuous self-cleaning ofsludge from tubes inclined at a steepangle allows sludge removal to beachieved without the need for drainingthe tubes. Laboratory and field testsshow that an angle of 60 deg providescontinuous sludge removal while stillallowing the tube to function as an ef-ficient sedimentation device. Pilotplant tests have shown these steeplyinclined tubes to remove alum floe ef-fficiently at rates as high as 8.5 gpm/sq ft. These tests led to the develop-ment of tube modules which were in-stalled in an existing clarifier to in-crease its capacity from 1.5 mgd to 3.0mgd. Analyses of the full scale instal-lation showed good clarification at ratesof 6.6 gpm/sq ft. The installation ofthe tube modules in an existing clarifierand the conversion of the filter to a

B cTube Module

Tig. 20. Flow Distribution and TurbidityData in Initial Newport Tube Module

Installation

mixed-media bed provides plant expan-sion with substantial savings in costand space. The coupling of tubesettlers and mixed-media filters allows areduction in the size and cost of newtreatment facilities. This combina-tion provides new design concepts toachieve efficient treatment plant designproduce a given quality finished waterfrom a given raw water or waste-water.

AcknowledgmentsThe steeply inclined tube modules

shown in Fig. 11 and the techniquesfor installing the modules in the New-port clarifier were devised by CurtMcCann and Stan Aikins, Design En-gineers. Neptune MicroFLOC. Thevaluable contributions of these twogentlemen are gratefully acknowledged.

698

References

CULP, HANSEN, & RICHARDSON J our. AW W A

1. HAZEN, A. On Sedimentation. Trans.Amer. Soc. Civ. Eng., 53:45 (1904).

2. CAMP, T. R. Sedimentation and the De-sign of Settling Tanks. Trans. Amer.Soc. Civ. Eng., I l l :895 (1946).

3. HANSEN, S. P. & CUI.P, G. I.. ApplyingShallow Depth Sedimentation Theory.Jour. AWWA, 59:1134 (Oct. 1967).

4. FISCHERSTKOM, C. N. H. Sedimentationin Rectangular Basins. Proc. Amer.

Soc. Civ. ling., San. Eng. Div. (May,1955).

5. HANSEN, S. P.; GULP, G. L.; & STUKEN-BERG, J. R. Practical Application ofIdealized Sedimentation Theory. Pre-sented at the 1967 Water Pollution Con-trol Federation Conf., New York City(Oct. 1967).

6. Cox, C. R. Operation and Control ofWater Treatment Processes. Mono-graph Series No. 49. WHO, Geneva,Switxerland (1964).

Necessary Me

A paper presentcland, by Mark (Eng. Co., Kansas

HIGH-RATE filtraithe door to econo-

of must, of the nation'?tration plants. This c:made at a small tractrequired for the rapid-

The mapority of raare sound physically a-to lend themselves to tnecessary for high-rateifications, however, m:siderable ingenuity.

The potential capa<high-rate filtration, tplants in the US is irfrom one to two geneisignificant!

ParametersThe companies that

some of the high-raiguarantee rated capamost waters will appnthe old-time rate of 2 s.rapid sand filters. Tlrate of some of these jcreased as much as 3(still produce higher quthat of the original rap

Therefore, the hydr;of the high-rate plantmined with care. Jusin the many hydraulic