New White Water Management

8/9/2019 New White Water Management

1/7

S

i

BY J.T. WEARING, S. HUANG, A. PIGGOTT,

M.D.

OUCH1AND A. WONG

ew

strategies limit the

of paper

ne white-water

N A T Y P I C A L newsprint mill, large

amounts

of

effluent, rangin g from 10

I o

150

cubic metres per oven

dry

of

newsprint are dis-

a

la rge quant i ty

of

of

the effluent

riginates from points in the paper

iachine where fresh water

is

used

roue

It is. however, the pulping processes,

mecha nical o r chemica , which are mainly

responsible for the production

of

the

pollu tants . Du e to a

high

degree of overall

integration of pulp and white-water flows

in the newsprint mills, these poilutants

are typically dispersed throughout the

entire

mill

an d are eventually discharged

as

excess white-water, in dilute concentra-

tions.

A t these low concentrations, a ny form

of recovery of the dissolved material for

heat or chemical value is completely

impractical.

As a

result, pollution abate-

ment could be practised economically

membrane filtration and ot her mea ns for

ultimate disposal of dissolced mnterids.

Development of new white-water con-

figurations which are designed to niini-

mize dissolved solids concentrations at

the paper machine. Detailed computer

simulation of these new configurations.

For simplicity, this work was based on

a hypothetical integrated

mill

producing

newsprint from 100 TM P fu rn i sh .

Experimental

The behavior of dicsold

solids

in

nque-

ous pulp

slurric\: Primary- and second-

a r y - s t a g e t h e r m o m e c h a n i c a l p u l p s

only

by ex-plant bioIogical treatment

(Th lP) ,bb ta ined

from

various mill and

a t 60° .’. amples

or to tebting for

prop er opera tion of the paper machine. Technical Section methods wcre used.

By rcnroving some of the extraneous Eqidibritun

siztr l ) t

q f ’d i sdvcr l JoMs i u p d p

material

from

the system, it may be

slurries. A

study

of

a p p r e n i s o r p t i o n

possible to red uce effluent volume with- equilibria

was

done

o n

se\c ml samples of

out unacceptable buildup of dissolved pulp

to cxmiiiie

the effects of the

solids.

Such an approach is already Concentration

of

the liquor

on

the niate-

practised

in

chemical pulp processing. rial loss during dilution. TNO experimen-

Prior

todelivery to the paper machine, the tal techniques were used.

pu lp

is

thoroughly washed t o remove

and

In

method no

1

pulp

was

diluted with

recover val uab le dissolved materials. In deionized water

to

various consistencies

this report, we will demonstrate that this ranging from 1 to that prevailing at the

appronch m ight also

be

used for mrc han- refiner discharge. Th e sam ple\ \\ere

ical pulp of the newsprint furnish. placed in plastic bags, sealed an d held i n a

water bath at

60°C.

T h e

bags of

pulp were

ation, kneaded periodically and allowed to

e following equilibrate

for 24 or 48

hours. The

dewatering or pressing temperature was

In method

no.

2, the required pulp

sample was diluted

to

1% consistency and

mixed gently

60

rpm) using a large

To

assess

the likely effect

of

such

solids maintained at

60°C.

,

OT-

sfer

Further concentration of effluents by

TI39

PULP

&

PAPER CANADA

86:

5 (1985)


2/7

POLLUTION CONTROL

propeller. The pulp was maintained at a

temperature of 60°C for 30 minutes and

then dewatered. T he filtrate was sampled

for analysis, mixed with a small qu ant ity

of m akeup water, and then recycled to the

wash ing step together with the nex t fresh

batch of pulp.

Measuremenrs of rates mass transfer.

The moist pulp was very finely fluffed and

added to preheated, deionized water to

make a

IS

consistency slurry at 60°C.

Gentle stirring (60 rpm) was provided by

a large propeller. Sam ples were removed

at various intervals and dewatered at

60°C for subsequent liquor analysis.

Membrane filrrurion of

eflirents.

A 0.5-niZ

laboratory hyperfiltration module mem-

brane was supplied by the Danish Sugar

Company, Nakskov, Denmark. Samples

weie prefiltered using Reeve Angel 934

A H glass fibre filter paper

to

remove

suspend ed solids. Cellulose-acetate mem -

b r m z b (110.

66Sj

w i t h a noniiniii rooiwu-

lar

weight cutoff

of

50

2S°C and 3.6 MPa.

Th e concentra

until the conce

effluertt. Th e concentra te from the hyper-

filtration test was evaporated at 6OoC to

various concentrations using a Roto-

vap@ operating under vacuum .

The

resultant liquors were tested for viscosity

at various temperatures using Cannon-

Fenske viscometers. Heating value, and

carbon and ash contents were determined

on the oven-dried residue.

Results and

discussion

Equilibrium \tudy of dissolved

solids

in

W I durries: Dissolution

of wood

mate-

C = concentration

of

solute, kg/m'

k = empirical constant, m'/kg

The material balance for the present

experime ntal case, including the sorption

term, is:

Material dissolved in pulp washing =

total soluble material

-

sorbed material

The data shown in

Fig

1 were used in

the above material balance and, using a

non-linear regression technique,

values

for the constants T, Q; nd k were

obtained. Figure

2

illustrates the close

agreement of the CO D data to the

suggested form. Good fits were also

obtained

for

BOD and TDS. The behav-

ior of

U V

lignin could not be calculated

with Equation

(2).

The form of the curve

was quite different for UV lignin. It

is used here for empirical purposes only.

True

Langmuir

sorption

behavior, i.e.,

reversible, monolayer adsorption etc.,

may or may not exist.

Dissolution

of

w o o d material during

washing (at

1

consistency and

0.5

hour)

with successive recycling

of

the filtrate

be reduced

on in the

is shown in

Fig. 3

using COD data f rom the same

pulp processed above. Increased data

scatter is evident with me thod

no. 2

since

each result is found by subtracting the

larger inpqt and output COD values.

The Langmuir empirical data fit from

0

I

I

0 10 20 30

Consistency,

. .

rial during dilution of pulp to V SnO U S

o

nlethod

no,

I ~ was

Fig.

1.

W h en diluting TMP, more wood material is d isso lved at lower consistencies.

he greattr at lower stock consistency This

is shown in Fig. 1 for TIMP samp led a t an

eastern Canadian mill.

Th e effect of consistency appe ars unre-

lated to diffusion or reaction rate lim ita-

tions. as 24-h and 48-h equilibrations

produced similar results. I t seems likely

that f o r the parameters monitored,

a

conceiitration-dc.pendent sorption phc-

nomenon controls the amount of dissolu-

tioi;. This obscri e d effect dii be quan t i -

tati\ely dcscribcd.

a t

least

at

equilibration

times greater t h n n 24 hours. by n Lang-

muir-type sorption equation:

where

Q

= quaiitit , sorbed, kg/odt pulp

24 hour equiltbratlon

4 8

hour equtltbratlon

angmuir approxunatlon

Dissolved

Total-

Swbed

70

6

-

5 O -

-

D : T - k O s C / ( l + kC)

C

5

Concentration

'.

COD +-e,

d i s s o l v e d , 40 4-e-..


3/7

ment evists betiveen the two me th o d s of

equilibration. Despite

large

variations in

washing time and consistency used. the

dissolution

of

ivood material during pulp

washing could bc characterized by a

single relationship.

In using the above procedure. sorption

constants were determined for nunierous

pulp samples.

A

pariial list is shoivn

in

Table I A \vide range in sorption

coefficicnts was found

for

the v:irioits

pulps tesred.

N o

consisicnt pattern relat-

ing sorption to

se;ison.

\vood species. pulp

type or ape. or relative proportion of

dissolved components was evident. One

pulp sample. a secondary-stage pulp from

Mill (not listed) has shown no sorpt ion

behavior at all.

Th e relationship between total dissolu-

tion of wood material and specific water

use was investigated recently by JSni ne n

et al

12)

using laboratory techniques

similar to our method

no. 2.

A mill scale

L , x G p k i w a l x ) conducted. Janqn-

e n et

a1

concluded tha t reduced water use

would result i n decreased dissolution

of

wood

materials.

Again,

UV

lignin (and dichlorome-

thane extractives) dissolution

was

noted

to show a much stronger depend ence on

concentration than

COD

and

BOD

Conclusions were not reached

as

to the

general controlling mechanism. The

observed phenomena might be explained

by an increase in concen tration at lower

water use and hence an increase in

quantity sorbed on the pulp.

The quantity could be a large fraction

.

,_._..-_.

COD

Dissolved,

k g l o d t pulp

6o

OD Concentration.

a/L

Fig.

3. Dissolut ion 01 COD dur ing d i lu t ion of TMP is reduced at high f i l t rate

concentrat ion.

I concentratlon.

400

TDS

8 I

i+

iquor

Washing

time, h

TOS COD

k

Standard

Mill PrOCeM kg l t

kg/t

“lkg emr ln

k g l t

kg/t

m’/kg error in

D kglt D kg/t

0.59

1.1 57.3 31 3 0 27 0 70

’

TMP spnrce/fir 38.4 15.1

o TMP spruce/hr/pine 428 16.0

5 3 2 24.9

spruce/ Rrlpine

61.1 234

f i t unsuccessful 66.6 223

0.49

2 1

1.8

1 9

1.2 1 4

T141

PULP

L

PAPER CANADA 86:

5 (1985)


4/7

POLLUTION

CONTROL

*

pulps tested, to be very fast. Most

of

the

material is transferred in less than 0.25

hours.

effluent was concentrated using hyperfil-

tration membranes to establish dissolved

solids retention and permeate flux for

history

of

small fibre nodules and bun-

waF

obtained.

des. These appeared to

disintegrate,

in Gen erally, Cor mixed organ ic solutes ,

the early stages

of

the wash, at a rate

1301)

has a loner retention than COD in

similar to the increase of wash liquor mem brane filtration because BOD -con-

concentration. Therefore, if a cau wl tributing materials are mostly of low Assum ing 120'C to be th e maximum

relationship does indeed exist, the ratc of molecular weight. Nevertheless, BOD pract icable temp erature i n the product

their disintegration

may

repre sent a retention was nearly 93 n

our

tests. Th e end of the evaporator t ra in, the maximum

significant limitation

to

the ra te of fluu

of

permeate and the concentration of concentrations which m ay be obtained in

transfer of dissolvedmaterials

to

the wash dissolved

solids in the feed stream are com mon eq uipmen t types were calcu-

liquor.

shown

in Fig.

5 as

functions

of

time. lated. It

is

assumed

in

the extrapolations

In a separate experime nt, this hypothe- Similar flux dzta were reported

by

that chemical reactions leading t

sis

was

tested by im proving the degree

of

Clmssen [3] for

concentration of spent increased

or

decreased viscosity would

fibre separation prior to dilution.

The

sulphite liquor.

not occur.Thecon centrat ions aregicen i n

W

solids)

r = 0.9995

S.E.

= 0.0204 in log,, (visc.)

= 1.05 in visc.

is very fast and suggests tha t th c degree of

fiberization

of

the pulp would, for all

practical purposes, control the rate

of

transfer.

TABLE 11 Hyperfiltration

of

TMP wash

effluent.

the product end

of

multiple-effect

evapo-

rator systcmr. Because kiscosity measure-

ments at teniperatures near

or

above the

boiling point require specialized equip-

TABLE

l i l

~ ~ ~ ~ ~ ~ timits

f o r

TMP l iquor

in

two t ypes

o f

evaposa-

tor.

Maximum Maximum

viscosity

concentration

m's

'Y1O6 T O G

___ _____I___

DS

Natural circulation

BOD COD

Feed. mglL 2340 4613 3532 LTV 100 40

Concentrate. mg/L 44500 857

66400

Horizontal tube.

Permeate. mg/L

176 333 190

Retention

92 8 93 1 94 6 spray type 200 4 4

I 1 0

11

0

1

0 5 0 100

Time,

min.

1 2 0 ,

-

I

--__

0

Dissolved solids conc entration,

I

Fig. 5. Typical

f l ux

and disshlved solids concentration during

hyperfiltration of TMP wash liquor.

Fig.

6

Viscosity

of TMP

wash liquor inc rea ses dramatically abovc

30

to ta l

dissolved

solids.

T I

4 2

PULP

&

PAPER CANADA 86: 5 (1985)

65


5/7

Ca/or c

vahre. The

combustion proper-

ties of the dried liquor solids were

determined. Table 1V. For comparati\,e

purposes, typical values for other corn-

mon fuels are shoun .

Dried

TMP

liquor is a fuel of rathe r

loa,

heat value. comparable to bark. I t would

be available only

in small

qusnt i t ies

equivalent to the yield loss i n T M P

rkfining. ca..

20 to

40 kg/odt pulp.

I f

combustion is considered

for

ultiniate

disposal. integration

of

the liquor into

an

existing chemical recovery f urn ace

or

bark boiler would appear

to

be most

practical.

White-water management strategies

for

minimum contaminated effltrent volume:

Based on the equil ibriun~ nd ra te data

'POI-LUTION CONTROL

dissolved solids

machin e white-wa

developed for pulp-water intera ctions we

devise d models

for

white-water strategies.

aimed at reducing the concentration of

dissolved solids in the white-water, thus

allowing low effluent volumes. F or illus-

tration purposes, a newsprint mill usin

100 TMP

furnish was used

as

the stud

model.

The conventional white-water systerr

(A)

is shown in Fig. 7 The system is full)

integrated. White-water from dewatering

processes at the pulp mill and papel

machine enters a common white-water

storagechest. Dilution an d show er waters

in pulping and papermaking are drawn

from this common source.

As a

result of this integration, the

white-water concentration is very uni-

form

throughout the mill. It is desirable,

however, to have low concentrations

of

.. - -

5

e

{

paper machine, it would seem Gossible to

depress the

DS

concentration in the

white-water used at the pap er machine by

arranging for a counter-current flow of

pulp and effluent. Such a co unter-cu rrent

s js tem (B)

is

also shown in Fig. 7.

All paper machine dilution an d shower

needs

are met by

paper m achine and

save-all filtrates and some fresh water.

The

e x e s

paper machine white-water

is

i

routed

to

the

TMP

mill.

TMP

mill

effluent is removed from the pulp at the

thickeners but held back for reuse in the

T M P mill. Excess white-water is sewered

from the

TMP

mill, not the paper mill.

A

further refinement

of

this approach

may be the

use of

inter-stag e washing of

ThlP.

Earlier information developed by

Wong et

al [6] shows

that the majority of

pollutants

are

released in the primary

stage

of

refining. Thus. washing the pulp

a t this

stage

could provide effective

removal of dissolved materials while

taking advantage

of

the fast drainage

characteristics

of

high-freeness primary-

stage pulp.

To

maintain pulp quality in

refining, the washer mus t discharge pulp

to the secondary refiner a t

a

high

con.'stency. Figure

7 shoas

the inte-

g a te d . mill with im proved u.hite-\viiter

design and add ed inter-stage press w ash-

ing (System C ) .

A computerized process simulation

was developed to evaluate Lvater. pulp

and dissolved solids material balance5

and energy ba h c e s for these three cases.

A modular simulation technique

was

used. Called GEMS. this is

a

com pute r -

based proces s si m

i

I

a

ti

on

tech

ti

iq

ue

developed specifically

for

pulp

a n d

paper

manufacturing processes

by

Dr.

L.

Edwards and his associates

a i

the Univer-

sity of Idaho, Moscow, Idaho.

The design parameters for the

Th4P

mill

were derived from the material and

W -.tr.

Wood

ChlP.

I S.*..lI

Effluent

A .

Fully integrated white-water system.

1

I

I

P P 1

I

I

I

=

W.l

P M

W W c h e s l

3 .

Segregated,

counter-flow

system.

TABLE JV.

Characterization

of

TMP

liquor solids and selected

commer-

Fig.

7

TMP

white-water management methods.

T143

66

ood

IIpm

w.1

PP1

L - - - A

I

I PY

W W

chest


6/7

energy balances given by Azarn iouch et a1

80

at

an effluent volume

SI. Th e conventional p aper mach ine pulp. Lower effluent volu

white-water configuration c orrespon ds to considered. Wah ren

et

a1

[8

cur ren t Scandinavian opera t ion as out that the

effect

of

s

white-wa ters.

3.7

m'/odt in the

whitewater , the

atioii v effluent

0.06

60°C

white-water schemes. Substantial im-

provement could be achieved s imply by

KMW press washer

segregat ing the pulp and paper mill

white-waters. At constant headbox con-

centration, excess white-water may be

reduced from 10

to

3.7

m3/odt pulp. Th e

addi t ion of an efficient one-stage press

washing would allow

a

further reduct ion

of excess white-water to

1.7 m3/odt

discharge consistency:

t-bden E actor in wash zone:

Consistency in wash zone :

Genera,

-yield

IOSS:

l o

stage:

constants

40%

2.0

15.0%

4.0

3.2

0.8

Q. = 15.5 kg/od

pulp

k =

1

.O

m / k g

DU~D

L

The counter-current arrangement of

wh itew ater flows also results in reduced

t r a n s f a

af

.&axe

heat from rcfining into

the paper machine white-water. This

means that,

to

maintain a given headbo x

temperature (above that of the fresh water

input), niore heat would have to be added

directly to the paper machine white-

ivater.

This coiiiparison is drawn in Fig.

9

w

here heating rcc~uirements re

shown

as

a

funct ion of

excess white-water flow for

the

three

c:iscs

A t reduccd effluent

vnlunie

(;IS in

Systcni

C). there w u i d

he n

sninll s a t i n g in

s t c a ~ n

or

the

heat ing

of

mill \\

hitc-\rater

3s

compared

t o

the

con\ entional case

(Sys tem A).

Using

3 GEX1S

s imulat ion of ; I II

integr;rtcd refiner niechanical ptllp

( K h f

P) mill. Vcnkatesh et

a1 [7]

predicted

t1i:it the incorporation

of

a segrsgated,

countcr-ciirrcnt \vhi te-water arrang cnic nt

I 8 1 1

I \

6

Dissolved 5

solids

concentration

at

headbox

g/L

2

1

Fully

I n t e g ra t e d

(System A )

S e g r e g a t e d [ s y s t e m Sl

I n t e r s l a g s

washing

( S y s t e m C )

I

0

0

4 8 1 2 16 2 0

I_

xcess white water volume, m3/odt pulp

and P ( ~ ~ t - s ~ c ~ ~ ~ 1 ~ % '7cchanicd Pulp Fig. 8. T h e d i s s o l v e d s o l i d s c o n ce n t r a t io n at t h e i i e a d b o x

c a n

be great ly reduced by

u x h i n g \ \ . t T i i l d rcduce the concenlralion s e g r e g a t i n g the pulp mil l and paper mil l white-waters and by the incorporation

o f

of

di>sol\ed 5 c > l i d h

i n

thc pape r inill h y inter-stage washing .

PULP &

PAPER CANADA

86:

( 1 9 8 5 )

T144

67


7/7

-

.-

-

-

-

~

-

f

the conventional

fu l ly

integrated sys-

I O

rn'/odt.

This means that \ - s l y small changes

i n

water input or production rate i n the

\vi11 result in large chnngcs i n

te-water concentrat ion,

affecting

uct quality and machine efficiency

i n

This would

be

volumes. Neverthe-

i t

is expected that a process control

water input

to maintain constant

The washing

of

the pulp

ivill

rcsult in a

of chrotnatographing

of

ap er m achine white-water relative to

h littleor nosorpiion. This might

of closure in th e white-

ter system after pulp washing in some

To

etermine theoptinium white-water

is

to specify the concentration of

aterial which can be tolerated

of the

been

developed [SI

to

on

product quality which can

to existing or projected mill

n economical alternative to

estcnial

With conventional

ons, the minimum

al excess white-water volume fro m

be

10

m3/odt

D U ~ D

ue to cons t raints

Piggott. contributed to

t h i s

nn rk u.liile on

co-oI)C'r;"i\,e-\vor~ ternis

f rom

the Uni-

versily of Waicrloo. 'The authors extend

their appreciation

to

J Kersch for

techni-

c d assistance and cxperiineni:il advice.

With such low volume of concentrated

other means

of

disposal of the

t hus

become

For

example, the excess white-

may

be concentrated further by

by

Since the amo unt of material

isvery s m d ,20kg/odt pulp,

might be can ied

out

in

an

or recovery boiler.

If all the mechanical pulping effluent

be

integrated in to an adjacent chem-

lacing fresh water,

sts might be effec-

White water

heating

requirements

GJ/odt pulp

2 0

-

Fully

Integrated

4

12

16

2 0

t : : ~&mtniig&s pour reduire d e facon significative

l es volumes

excddentaires d'eau

blanche tout en maintenant aussi basses que possible les concentration de matihres

so-

lides dissoutes durant le procede de fabrication d u papier.

La

rdduction

d e

ces volumes

peut permettre

d

de telles usines de liberer des effluents qui contiennent un tres fable

pourcentage d e matieres contaminantes, e t cela a des cod t s qui se comparent ceux des

I

installation conventionnelles de traitement biologique des effluents.

~ ~ ~

Reference:

WE ARING.

J.T.. HUANG. S..

PICXOTT.

A..

OUCIII .

A1.D.

i~ndN'OSC;.

A.

New u

hiie-watcr

management strategies

for

integrated new

sprint

mills. PI+ P U ~ W

LJII

O :

T139-145 (May 1985).

Paper

preserited at the 1983 Xlcchanical Pulping Coiifcrcncc

of

the

Tcclinic.al Section.

CPPA. co-sponsored

w i th EUCEPA and TAPPI

ai

Washington.

D.C..

Junc I 4

t o 17. 1983. Not to be reproduced without permission. Manuscript rcccivcd Junc 14.

19x3.

Approvcd

by

Review Panel. Scptcmbcr

20. 1984:

Abstract:

The white-water systcins in integrared

nrusprint m ilk

may bc re-organizctl with

niuch lo\vcr

excess

white-water

volunics

v, hile

still

maintaining low dissol\cd solids

concentrations in the papermaking operation. These lower volumes proniiw to niake i t posziblr

for such

inills

to

operate with effluent that

has a

very

low

degree

of

contaruination. at costs

comparable

to

the operation

of

conventional biological effluent

treatment

Keywords: WHITE-WATER, NEWSPRINT. INTEGRATED

MILLS.

TIIERMOME-

CHANICAL

PULPS,

V.OLUME,

POLLUTION

CONTROL.

T145

PULP

&

PAPER CANADA

86: 5 (1985)

New White Water Management

Documents