-
- J p T S ~ - - ~ 3 J T /- EWSPRINT DEINKIN
UHIT OPERATIOMS STUDIES OF FLOTATION-WASH DEIMKING
A comparison o f commercial and p rop r ie ta ry surfactants
BY J.K. BORCHARDT AND D.W. MATALAMAKI
ILOT-MILL DEINKING tests can be use- ful for both research
studies and P commercial development of
improved deinking surfactants. Deink- ing surfactant performance
can be assessed under conditions more similar to commercial mill
operations than obtainable on a laboratory scale. Impor- tant
factors such as shear forces leading to foaming are very difficult
to model accurately in the laboratory.
One major limitation of many pilot- mill trials is that process
water is not recycled [ 11. Only one North American pilot facility
now has permanently installed process water recycling capabil-
ities. Also, pilot-mill trials are usually too short to optimize
deinking surfactant use levels [ 11.
OESCRIPTION Deinking surfactant chemistry: The per- formance of
a series of proprietary sur- factants was compared with that of a
deinking agent widely used in North American flotation-wash
deinking mills.
This competitive surfactant is considered to be a
dispersant-collector.
Dispersant-collectors are thought to render ink particles
hydrophobic in the flotation step promoting their aggrega- tion and
adsorption onto the surface of air bubbles [2]. In the wash step,
disper- sant-collectors are believed to render ink particles
hydrophilic. This promotes ink particle dispersion in the washing
step.
This theory of how dispersant-collec- tors function is a
rationalization of observed patterns of ink removal as a function
of ink particle size. Flotation more effectively removes larger ink
par- ticles while washing more efficiently removes smaller ink
particles.
However, there is no direct evidence to support the proposed
dispersant-col- lector deinking mechanism. Accord- ingly, the
emphasis in this report is on comparative results rather than the
mechanism by which the test surfactants function in deinking.
The proprietary deinking surfactants included both
single-component prod-
ucts and formulated products. Deinking surfactants for the
pilot-scale tests were chosen based on performance in a labo-
ratory flotation-wash deinking test. The furnish for the laboratory
scale tests was 70% Houston Chronicle and 30% Time magazine.
Furnish used in pilot-mill studies: The furnish used in the
pilot-mill studies was a 70:30 mixture of newspapers and mag-
azines. Both newspapers and magazines were post-consumer, not
publishers’ overruns. Examination of the bales indi- cated the
newsprint was a mixture of approximately 60% Houston Chronicle, 25%
The Houston Post, and 15% newspa- pers from Oklahoma, mainly The
Daily Oklahoman.
These newspapers are printed with letterpress and No. 1 offset
processes. However, experience at the pilot mill with this type of
furnish indicates that the ink behaves much like flexographic ink.
The low pulper pad brightness sup- ports this contention.
Other workers have noted that the
Water 10% Con. 0.7% Con. Recirculating Flow For 6 Pass
Fine y o & $ F d PDM* 500 PDM* 500 - -
I U u z Uniflow* Posiflow* Dynamic
Washer 1
24 PULP & PAPER CANADA 95:lO (1994) 4 T374
-
decreases in the same properties in Fig. 2. This holds true for
both the scattering coefficient and the sheet density. The decrease
in density can be attributed to increases in fibre length and
coarseness.
Only the tear strength of the spruce/fir TMP yields a
bell-shaped curve showing a 45% limit for the LFF content. Beyond
this maximum, an increase in LFF no longer contributes to the
develop ment of the tear strength of the hand- sheet. On the
contrary, the increase in fibre coarseness, accompanied by signifi-
cant decreases in sheet density and in the short-span BL, tends to
reduce the tear strength of the handsheets.
With respect to the species effect, Fig. 2 shows a clear
separation between the two types of pulp; the spruce/fir furnish
tends to concentrate on the right, while the birch TMP occupies the
left. This shows that the white birch TMP contains a significantly
lower percentage of LFF than the spruce/fir TMP. This is con-
firmed by the consistently longer fibres in the spruce/fir TMP than
in the birch furnish at any level of CSF, Table I.
For example, at 100 and 300 mL CSF, the spruce/fir fibres are
1.5 and 1.8 times longer than the white birch fibres, respectively.
In addition, the diameter of a birch fibre (18 pm) is only half
that of a spruce/fir fibre (35 pm), while its wall thickness (3.8
pm) is more than double that of a typical spruce/fir (1.5 pm) fibre
[2]. Thus, the relatively thick-walled birch fibres tend to
maintain their cylin- drical shape and form a rather loose fibrous
network with low sheet density and, consequently, inferior overall
strength.
In comparison, the spruce/fir fibres are thin-walled elements
which tend to collapse during refining and thus form a denser
fibrous network possessing higher general strength. For example, at
30% LFF, the spruce/fir handsheet is 1.8 times denser and 10 times
stronger (BL) than the birch handsheet. Furthermore, the birch
handsheet is three points lower in brightness than the spruce/fir
hand- sheet. This can be attributed to the pres- ence of gum in
birch as reported previ- ously [ 3 ] . However, the three-point
loss in brightness is compensated for by a four-point gain in
opacity.
When the sheet properties were plot- ted against the individual
fibre fractions (14, 28, 48, etc.), highly irregular curves were
obtained which crisscrossed one another in a zig-zag manner. A
detailed analysis of these figures yielded no con- sistent
conclusions. In contrast, the curves in Figs. 1 and 2 are nearly
parallel with one another, and even the experi- mental points tend
to fall on the lines.
-
OONCLUSlONS The introduction of the new SFF and
LFF parameters make much more com- prehensive the use and
interpretation of the Bauer-McNett fibre classification data.
OEFERENCES 1. TASMAN, J.E. The fibre length of Bauer-McNett
screen fractions. TappiJ, 55(1), p. 136-138 (1972). 2. KORAN, Z.
Different processes can be used to enhance hardwood pulping
quality. Pulp Paper Can, 90(2), p.18-20 (1989). 3. KORAN, Z., YANG,
K.C. Gum distribution in yel- low birch. WoodScz., 5(2), p. 95-101
( 1972) .
OCMNOWLEDGEMENT The author thanks the National Sci-
ence and Engineering Research Council of Canada for its
financial support.
I
Circle Reader Service No. 112
National Research Conseil national I*! Council Canada de
recherches Canada
If you sometimes think international competition is cutting into
your markets, a little new growth may be in order. And the National
Research Council can help.
We have top-notch facilities, expert staff and a track record of
working with Canadian firms to boost their competitive edge.
Major companies have already taken advantage of our proven
capabilities in areas such as automation and control, environmental
technology, machinery technology, advanced materials, thermal and
fluids engineering, and computer- ized design, simulation and
decision support.
Call (613) 990-9546 [Fax (613) 952-4569], Our expertise is
deep-rooted.
We’ll help you grow.
PULP & PAPER CANADA 9 5 : l O (1994) 23
-
softwood and alum content of southern US furnishes such as the
Houston Chroni- &makes both wash and flotation deink- ing
particularly sensitive to deinking chemistry and process conditions
[3].
The newsprint was primarily one to four months old. Many of the
magazines were much older. Perhaps 20% dated back to the 1970s and
a few to the 1960s. While 60 to 70% were lightweight, a sub-
stantial percentage, 30 to 40%, were printed on heavier weight,
more heavily coated paper. Review of the deinkiig process design:
The test procedures are summarized in Appendix 1. Figure 1 is the
process flow chart for this simplified deinking process. The
sequence of unit operations was: 1. Pulping at 10% consistency, T =
120" to 125"F, pH 9 to 10; 2. Dilution to 5% consistency followed
by screening using a 0.010-in. (0.254") screen; 3. Dilution to 0.7%
consistency; 4. Two-stage flotation in Beloit Model 500 Pressurized
Deinking Modules (PDMs) at 0.7% consistency. The accepts passed
through the cells an average of six times.
The rejects were discarded, i.e., no effort was made to recover
any fibre in the reject stream and return it to the flotation cell.
Thus yield losses are probably hgher than in a real mill where the
rejects are sub- jected to additional processing. This pro- cessing
recovers some of the cellulose fibre permanently lost in laboratory
and
determined by measuring pulp consis- tency before and after
flotation. These flotation cells are described in [4]. 5.
Through-flow cleaning using a bank of three 3-in. (7.62-cm) Beloit
UniflowTM cleaners [5]. They are situated after flotation to take
advantage of the entrained and dissolved air still in the
pilot 'mill experiments. Meld loss was pulp after flotation
[6].
Percent ink particle removal Particle Size Range Deinking
surfactant
(microns) SDA-3 1 SDA-30 os- 1 2.000 - 6.250 47.7 36.1 68.6
6.250 - 19.600 82.5 81.2 85.8
19.600 - 44.200 89.2 90.8 90.8 44.200 - 78.500 91.1 91.7 90.6
78.500 - 157.000 90.4 92.2 93.0
157.000 - 380.000 91.2 91.3 92.1 380.000 - 707.000 90.6 92.0
88.8 707.000 - 1257.00 86.6 84.0 81.3 1257.00 - 1964.00 63.1 18.1
84.8
All Particles 80.3 79.7 85.8
1 2
1 1
10
g
8
7
6 5
4
3 2
1
0 0 1 2 9 4 5 6 7 8 9 1 0
AFTER PROCESS STEP
SDA-31 SDA-30 A OS-1
z y t z
2 0 1 2 3 4 5 6 7 8 9 2 10 I
AFTER PROCESS STEP
A SDA-SO + SDA-31 os-1
AFTER PROCESS STEP
A SDA-SO + S D A - 3 1 0s-r
* o 2 4 6 8 10
AFTER PROCESS STEP
0 SDA-30 A SDA-31 OS-1
T 375 4 PULP & PAPER CANADA 9 5 : l O (1994) 25
-
6. Two stages of forward cleaning using 3-in. (7.62-cm) Beloit
PosiflowTM cleaners [5].
7. Washing using a Beloit Dynamic Washer equippedwith a
0.012-in. (0.305- mm) basket. For part of the test, a 0.006- inch
(0.152-mm) basket was used. Deinking process water clarification:
Government regulations and environ- mental concerns are leading
deinking mills to recycle more of their process water. Efficient
process water clarifica- tion is necessary for this recycling. If
excessive amounts of dispersed ink remain in the clarified water,
this ink redeposits on fibres when using recycled water. Process
water samples were taken after the wash step. Cytec workers clari-
fied these samples using commercial MagniflocTM water clarification
polymers or blends of these polymers.
OESULTS In the figures, the process steps are
referred to as follows: Step 1 - Pulping; Step 2 -Dilution to 5%
consistency and screening; Step 3 - Dilution to 0.7% consistency;
Step 4 - First flotation stage; Stcp 5 - Second flotation stage;
Step 6 - Through-flow cleaners; Step 7 - First bank of forward
cleaners; Step 8 - Second bank of forward clean- ers; Step 9
-Washing.
The overall brightness gain through the entire deinking process
was similar for the three test surfactants, Fig. 2. The brightness
gain is measured relative to the pulper pad brightness. Reported
val- ues are averages based on readings taken on both sides of the
sheet. Pulper pad brightness values were relatively low, 42 to 43.
The largest brightness gains were observed in the first flotation
stage and in the washing step.
The entire deinking process reduced the ink particle surface
area 75 to 80% relative to the pulper pad, Fig. 3. Ink par- ticle
area reduction was similar for the three test surfactants. The
largest reduc- tions in ink particle area occurred in the first
flotation stage and the washing step. The mean and number median
ink par- ticle diameters were similar for the three deinking
surfactants at each stage of the deinking process, Table I.
The largest decreases in particle size occurred in the first
flotation stage and in the washing step, Fig. 4. Smaller ink par-
ticles, less than about 50 microns in diam- eter, have the greatest
effect in reducing deinked sheet brightness [7]. The effi- ciency
of removal of these small ink par- ticles, 65 to 70%, was similar
for the three test surfactants, Fig. 5. The greatest degree of ink
removal occurred in the first flotation stage and the washing
step.
The greatest yield loss occurred dur- ing the washing step, Fig.
6. Significant yield loss also occurred in the first flota- tion
stage. There was very little additional yield loss in the second
flotation stage. SDA 30 and SDA 31 exhibited lower yield loss than
the competitive surfactant. Most of this difference originated in
the first flotation stage and carried over to the subsequent
deinking process. The flotation step: SDA 30 produced a slightly
higher brightness gain in the two-stage flotation step, Fig. 7. SDA
31 and OS-1 produced equivalent bright- nesses. SDA 30 produced the
largest reduction in residual ink particle area in
the two-stage flotation, Fig. 8. Despite producing an equivalent
brightness gain, SDA 31 produced a greater reduction in ink
particle area than did OS-1.
Results summarized in Table I1 indi- cate that SDA 30, producing
the largest flotation brightness gain, was also the most effective
surfactant in removing ink particles less than 44 microns in diame-
ter, the particles having the greatest adverse effect on brightness
[7].
SDA 31 and OS-1 do not differ greatly in their efficiency in
removing ink parti- cles less than 44 microns in diameter. However,
SDA 31 was more efficient in removing larger ink particles (380
to
Particle Size Range (microns)
2.0000 - 6.250 6.250 - 19.600
19.600 - 44.200 44.200 - 78.500
78.500 - 157.000 157.000 - 380.000 380.000 - 707.000 707.000 -
1257.00 1257.00 - 1964.00
All Particles
Percent Ink Particle Removal Deinking Surfactant
SDA-3 1 SDA-30
0 4.2 11.6 9.0 7.8 25.7
10.0 26.0 16.4 29.6 28.7 37.3 43.4 49.6 66.1 59.5 41.9 50.0 12.1
21.6
os- 1 0 2.5
11.4 17.8 18.0 28.8 36.8 27.5 25.0 11.7
Yield loss (% by weight) Process step SDA-3 1 SDA-30 os- 1
Two Step Flotation 1.91 1.53 3.29 Throughflow Cleaner 1.9 1.8
1.9 Forward Cleaners (Two Stage) 21.6 25.9 24.0
I I
t I I I 1
0 5 6 8 9 AFTER PROCESS STEP
A SDA-30 SDA-31 os-1
26 PULP & PAPER CANADA 95:lO (1994) 4 T376
-
1964 microns in diameter). This greater efficiency was
consistent with the larger reduction in residual ink particle
surface area produced by SDA 31, Fig. 8. SDA 30 was also more
efficient than OS-1 in removing these larger ink particles.
Mechanical cleaning steps: The yield losses in the primary
through-flow and forward cleaners are substantial, Fig. 6 and Table
111. However, these yield losses can be reduced in the mill by
send- ing rejects from these stages to sec- ondary cleaners for
fibre recovery. Changing the deinking surfactant had little effect
on yield loss in either the through-flow or forward cleaners. The
only possible exception to this general- ization is the forward
cleaner yield loss using SDA 31. The washing step: All three test
surfac- tants produced higher brightness gains in the washing step,
6 to 8 points, than in the two-stage flotation step, 3 to 4 points,
Fig. 9 and Fig. 7 respectively. Both SDA 30 and SDA 31 provided
larger washing step brightness gains than did OS-1. The sur-
factant producing the highest brightness gain, SDA 31, provided the
largest reduc- tion in ink particle area, Fig. 10. While
SDA 30 produced a higher brightness than OS-1, Fig. 8, it
produced a slightly lower reduction in ink particle area.
Despite the differences in washing step brightness gain, Fig. 9,
the three test surfactants did not differ greatly in removal
effectiveness for ink particles less than 44 microns in diameter,
Table IV. Despite producing a slightly lower reduction in ink
particle area, SDA 30 appeared as efficient as OS-1 in remov- ing
large ink particles, those greater than 380 microns in diameter,
Table IV. SDA 30 produced the largest reduction in ink particle
area during washing. However, it was not noticeably more efficient
than the other surfactants in removing the large ink particles.
The residual ink particle size distribu- tion after washing,
Fig. 11, approximates that of the “no mat” case for this washer
[8]. This suggests that the washer was operating efficiently.
AB DEINKING Comparison with pilot-scale tests: A series of
laboratory deinking tests served to qualify deinking surfactants
for the
pilot-scale study. The flotation-wash pro- cess design is
detailed in the Appendix. A single newsprint, the Houston
Chronicle, and a single magazine, Time, were used to assure furnish
uniformity.
Thus the furnish in the laboratory and pilot-mill tests were
similar but not identical. Many of the magazines used in the
pilot-mill tests were quite old. These differences could be one of
the reasons for brightness and yield loss differences between the
laboratory and pilot-mill tests. Brightness values and image analy-
sis results are reported in Table V.
With the exception of SDA 31, yield losses were significantly
greater in the laboratory flotation experiment. How- ever, the
trend in yield losses were similar in the laboratory and
pilot-scale deink- ing experiments. SDA 30 and SDA 31 provided
significantly lower yield loss than did OS-1.
The brightness values in the pilot- scale deinking tests showed
no correla- tion with yield loss. However, in labora- tory
experiments increased brightness was accompanied by increased yield
loss. The trend is clear despite some scatter in the data, Fig. 12.
This behavior is consis-
SDA-31 SDA-30 OS-1
T 377 4
I l-i! SDA-31 SDA-SO pm 05-1
t; 60
8 a w
5 0
I ‘O
5 10
30
z 2 2 0 Q
P a s o
I SDA-3 1 SDA-30 OS-1
PULP & PAPER CANADA 95:lO (1994) 27
-
tent with that observed in several series of laboratory deinking
experiments done under different conditions and using different
deinking surfactants [9].
The brightness values in the labora- tory deinking experiments
were greater than in the pilot-scale deinking tests. Laboratory
brightness values for SDA 30 and SDA 31 were 2 to 3 points higher
than pilot- mill brightness values. The difference for OS-1 was 5
brightness points. The high laboratory brightness values for OS-1
appear to be due to the very high yield losses associated with the
laboratory deinking tests using this sur- factant, Table V.
Evaluation of two differ- ent production lots of OS-1 obtained from
different regions of the US con- firmed the high laboratory yield
losses associated with this surfactant. Flotation step yield losses
in laboratory tests, Table V, were considerably higher than those
obtained in pilot-scale tests, Table 111.
Image analysis for the laboratory deinking tests was done using
a different instrument and different sample prepa- ration
techniques. Therefore, image analysis results could not be directly
compared to those reported for the pilot-mill studies. Despite some
scatter in the data, the ink particle count for parti- cles less
than 0.04 mm in area appeared to be correlated with brightness,
Fig. 13. Except for the lowest brightness samples, entries 2 and 3
of Table V, ink particle counts and ink surface area for particles
greater than 40 microns in diameter did not appear to be correlated
with deinked sheet brightness or yield loss.
Process water clarification: Process water clarification and
recycling are essential in many deinking mills. The process water
containing the most ink particles, the washing step rejects, was
chosen for water clarification studies. Using commercial water
treatment chemicals from a single supplier, wash
5 7
water rejects could be efficiently clari- fied. The most
effective water clarifica- tion polymer system was dependent on the
deinking surfactant. Blends of com- mercial water treatment
polymers often provided improved results compared to
single-componen t systems.
~
e
a I I
OlSCUSSlON These tests were intended to provide
comparative results demonstrating deinking surfactant
effectiveness in flota- tion and combination flotation-wash
deinking processes. Operating mills usu- ally present deinking
surfactant suppli- ers with a given set of operating condi- tions,
particularly temperature, pH and process chemicals used in the
pulper. These can vary significantly from mill to mill even for
mills processing similar fur- nish. Therefore, simple pulper chem-
istry was used and no effort was made to optimize pulper chemistry
and process conditions for each product.
The geometry of the brightness gain curves, Fig. 2 , percentage
reduction in ink particle area curves, Fig. 3, and the number of
residual ink particles, Fig. 4, are very similar. This similarity
empha- sizes the close relationship between the brightness and
residual ink content of sheets prepared from deinked pulp [7] .
Earlier results indicated fairly good correlation of laboratory
deinking brightness results obtained using a Den- ver flotation
cell and pilot-scale tests using a one-atmosphere Voith flotation
cell [9]. However, this correlation was not observed in the current
series of tests.
This lack of correlation may be due to a greater variation
between the labora- tory and pilot-scale furnishes in the pre- sent
test series. Another factor may be the greater difference in the
mode of operation of the laboratory Denver flota- tion cell and the
PDM flotation cell [4]
used in the pilot-scale tests. The different mode of operation
of
the ,pilot-scale flotation cell and the lab- oratory flotation
cell is also a likely factor in the lower flotation step yield
losses observed in the pilot-scale tests. While differences were
not great, the surfactant producing the highest bright- ness gain
and the greatest reduction in ink particle area, SDA 30, produced
the lowest yield loss. This indicates that it is possible to
improve ink removal effi- ciency without increasing flotation step
yield loss.
@ONCLUSIONS All process parameters, i.e., deinking
effectiveness, yield loss, surfactant foam- ing properties in
all process steps and process water clarification should be
considered when evaluating deinking surfactants.
Pilot-mill test results indicate that, in commercial deinking
mill operations, certain proprietary deinking surfactants can
provide:
Excellent deinking results; Low yield loss; Acceptable foaming
levels; Good process water clarification.
With advance testing to choose the optimum SDA deinking agent
for the fur- nish, process design, and process condi- tions,
excellent process performance can be expected in flotation, wash,
and com- bination flotation-wash deinking pro- cesses. The
brightness gains provided by SDA 30 and SDA 31 in both flotation
and wash steps indicate that these surfactants can be considered
dispersant-collectors.
OCKNOWLEDGEMENTS The authors to recognize the impor-
tant contributions of Gita Grube ofBeloit Corp. who supervised
the pilot-scale
0 5 10 15 2 0 25 3 0
MEDIAN INK PARTICLE SIZE (microns)
28 PULP & PAPER CANADA 9 5 : l O (1994) 4 T378
-
deinking tests and Nelson Hsu of Cytec Corp. who supervised the
water clarifica- tion studies. Discussions with Ted Faleski of
Shell Chemical Go. were extremely helpful in the course of this
work.
OEFERENCES 1. BROEREN, L.A., Deinking of Secondary Fibers
Acceptance as Technology Evolves Pulp Paper, March 1990, pp. 71-75.
2. HORACEK, R.G., JARREHULT, B. In Paper Recy- cling: Strategies,
Economics, and Technology. RL. Patrick, ed. Miller Freeman Co., San
Francisco, 1991, Chapter 33, pp. 139-141. 3. LICHT, B.H., LEIGHTON,
W.G. A Practical Guide to Deinking Chemistry. Book B, p. B373-378
Pre- sented at the 76th Annual Meeting, Technical Sec- tion, CPPA,
Montreal, Canada (February 1990). 4. JOHANSSON, B., McCOOL, M.A.
Hylte Bruk’s New Deinking Technology Includes Pressurized Flotation
Deinking. 1989 TAPPI Pulping Confer- ence, TAPPI Press, Atlanta,
pp. 49-59. 5. LeBLANC, P.E., McCOOL, M.A. Recycling and Separation
Technology, 1988 TAPPI Pulping Confer- ence, TAF’PI Press, Atlanta,
pp. 661-667. 6. LAPOINTE, M., MARCHILDON L., FETTERLY, N.W., The
Deinking of Fine Paper, PulpPuper Can, 93 ( l l ) , pp. 46-47
(November 1992). 7. WALMSLEY, M., W, CJ., L. SILVERI, L. Effect of
Ink Specks on Brightness of Recycled Paper. 1993 TAPPI Recycling
Symposium, TAPPI Press, Atlanta, pp. 417441. 8. RANGAMANNAR, G.,
GRUBE, G., KARNETH, A., Behavior of Water-Based Flexographic Inks
in Newsprint Deinking. 1992 TAPPI Pulping Confer- ence, TAPPI
Press, Atlanta, pp. 933-939. 9. BORCHARDT, J.K. Effect of Process
Variables in Laboratory Deinking Experiments. 1992 TAPPI Pulping
Conference, TAF’PI Press, Atlanta, pp. 749- 763.
OPPENDIX 1 - T E S T Pilot-mill studies: These were done at
Beloit Corp.’s Research and Develop- ment Centre in Pittsfield, MA.
Water for the test was taken from the city water sup- ply and
heated in the plant’s water heater. Process water was not recycled.
Because waste water was directly dis- posed of, there was no
secondary fibre recovery system. Waste paper fed into
the system was purchased from a deink- ing mill in Texas. Visual
inspection of the opened old newspaper bales indicated they were
composed of about 60% Hous- ton Chronicle, 25% The Houston Post,
and 15% The Daily Oklahoman. The magazine pile was recent issue
magazines. How- ever, some of this furnish dated back to the 1970s
and even the 1960s.
At the start of each test, a Beloit Tri- dyne pulper was filled
with enough 120°F water to dilute the pulp to a 10% consistency.
The chemicals, the 0.25% surfactant, and 1.0% hydrogen peroxide
were poured into the pulper by hand. Concentrations given were
relative to dry furnish weight. The SDA (Synthetic Deinking Agent)
surfactants are avail- able from Shell Chemical Co. (Houston, TX)
as NONATELLm deinking surfac- tants. The pH was adjusted to 9.5 to
9.8 with sodium hydroxide. The pulper rotor was turned to mix the
solution. Then 700 dry pounds of 70% old news- paper and 30%
magazine was dumped into the pulper.
Mixing was begun slowly to convert the paper to pulp without
splashing water out of the pulper. The speed was then increased to
300 rpm at about 100 hp. The paper was pulped for 20 minutes.
During pulping, the pH tended to decline from 9.5 to 9.8 to 8.3
to 8.5. This is due to the high acidity of the newsprint. To
counter this tendency, aqueous caustic solution was added to
maintain the pH at about 9.5.
After pulping, water was added to dilute the stock to 5%
consistency. The equivalent of 300 dry pounds of this stock was
pumped into twin holding tanks. This stock was further diluted to
0.7% consistency.
The stock was then pumped through a screen equipped with a
three-foil rotor and a 0.010-in. (0.254-mm) slotted Ahlstrom
Profile basket. The screen agi-
tator and pump were started before actual screening to raise the
fluid pres- sure across the screen. During this pres- sure-up
process, the stock was recircu- lated back into the twin storage
tanks. To maintain pressure across the screen, the line that
allowed the rejects to flow out was opened intermittently for 5
seconds at a time. Screen rejects were dumped into barrels and then
discarded. The accepts were pumped into a stock chest. The stock
was mixed continuously in this chest.
The next step in the procedure was flotation. Two-stage
flotation was done using Beloit Model 500 Pressurized Deinking
Module (PDM) flotation cells in series. The flow rate was 530 U.S.
gal- lons per minute. Two-thirds of the stock flow was recirculated
back to the inlet end of each cell. This resulted in the stock
passing through the flotation cells an average of six times.
Rejects were col- lected in an agitator tank. The reject rate was
set at about 1.5% by volume.
Accepts from the PDMs were sent through a bank of three 3-in.
(7.62-cm) Beloit UniflowTM through-flow cleaners to remove
lightweight contaminants. The flow rate was about 180 US gallons
per minute (gpm) . The pressure drop across the cleaners was 20
psig.
The stock was then pumped through two successive banks of three
Beloit Posi- flowTM forward cleaners to remove heavy contaminants.
The stock feed rate was about 160 USgpm. The reject rate was 10 to
14% by mass.
The accepts were sent a stock chest and stirred. Rejects from
both the Uni- flow and Posiflow cleaners were collected in an
agitation tank for disposal. The stock was then pumped through a
Beloit Dynamic Washer equipped with a 0.012- in. (0.305-mm) smooth
basket. The washed stock was circulated back to the stock chest
until the total pulp volume
Percent ink particle removal Particle size range Deinking
surfactant
(microns) SDA-31 SDA-30 OS-1
2.oooO - 6.250 19.8 26.2 29.6 6.250 - 19.600 56.8 60.8 59.8
19.600 - 44.200 64.5 63.8 69.9 44.200- 78.500 66.3 66.9 68.7
78.500 - 157.000 67.4 66.5 69.6 157.000 - 380.000 66.9 69.8 71.8
380.000 - 707.000 74.6 65.7 62.8 707.000 - 1257.00 52.6 65.8 68.2
1257.00 - 1964.00 44.4 14.3 15.4
All particles 55.5 57.3 58.3 Total particle count
After washing step 3853 3362 3720
T 379 4
61
60
59
sa
5 1
5 6
-1 \ \
0 300 600 900 1200 1600
NUMBER OF PARTICLES z0.04mm IN AREA
PULP & PAPER CANADA 95:lO (1994) 29
-
Image analysis Particles > 0.04 mm Particles < 0.04 mm
Flotation step
Count0 Area Counta Area Median diameter yield loss Surfactant
Brightness (PPm) ( P P 4 (microns) (Yo wt.)
SDA-30 57.7 f 0.4 65 4.7 652 11.0 303.8 7.1
SDA-31 57.5 f 0.1 456 25.0 652 9.47 222.8 0.4
SDA-32 56.4 rt 0.2 326 21.8 1109 23.1 205.5 0.9
SDA-33 56.8 rt 0.4 65 2.6 1435 22.9 197.9 4.2
5.9 SDA-34 58.0 rt 0.4 0 0 195 5.0 0 0 587 11.0
SDA-35 58.5 f 0.2 195 9.7 1435 26.3 198.1 1.8 195 16.6 782
14.2
130 6.1 260 5.3 194.0
130 13.9 326 6.1 276.4
65 3.4 71 7 13.15 1 97.1
65 3.7 1370 22.4 200.0
130 6.1 91 3 13.6 201.3 - -
- os- 1 b 60.0 f 0.2 65 5.3 456 8.2 327.6 17.4
os- 1 b 59.5 f 0.2 195 10.2 260 4.2 21 8.5 16.2 0 0 326 5.8
-
a. Calculated for one square meter of paper. b. Two different
production lots of the same product obtained about one year apart
and from different regions of the U.S.
had passed through the Dynamic washer for a single pass. Some of
this stock was also sent through a 0,006-in. (0.152-mm) basket.
Samples of the accepts from each step in the process were taken and
made into Buchner funnel pads. Lab deinkiig test procedure: The
labora- tory studies were done at the Shell West- hollow Technical
Centre in Houston, TX. The water used in these studies was in-house
deionized water. The furnish was 70% by weight Houston Chronicle
and 30% Time magazine. Each test used 20 grams of this dry furnish.
Pulping consis- tency was 5%.
The following chemicals were added in the following order to
110" F water: 0.4% DTPA (diethylenetriaminepen- taacetic acid);
1.5% sodium silicate; 0.25% test surfactant; and then 1.0% hydrogen
peroxide. Concentrations given are relative to dry furnish.
The paper was then added to the solution. The mixture was placed
in a Hamilton-Beach blender on the low set- ting until the paper
began to pulp, then on the high setting for 30 minutes.
The pulped paper was diluted to 0.7% consistency with 110" F
water and placed in a Denver flotation cell. The Denver cell was
run for 6 minutes at 900 rpm. The foam generated was continu- ously
skimmed off into a catch pan. This lost fibre and ink mixture was
filtered
over a pre-weighed filter paper. It was then dried in an oven
before weighing to determine yield loss.
The pulp solution remaining in the flotation cell was poured
over a 100 US standard mesh screen and pressed out to increase pulp
consistency to about 10%. This pulp was transferred to a beaker
and diluted with 110" F water to a consis- tency of 1.0%. This
slurry was mixed well, poured over the 100-mesh screen and again
pressed to about 10% consis- tency. This pulp was divided into four
portions, each of which was made into a 2- to 2.5-gram paper
handsheet using a TAPPI standard handsheet apparatus.
Abstract: ~@ilot-scale flotation-wash deinking tests were
performed using a Beloit Model 500 pressurized dznking module. The
furnish was a 70:30 mixture of newspaper and magazines. Deinking
efficiency was studied for each process step: screening, flotation,
washing, forward and reverse cleaning, and washing. Compared to a
widely-used commercial product, certain proprietary deinking
surfactants exhibited improved o r comparable performance in ink
removal, yield loss. foaming, and water clarification using
commercially available water treat- nieni cherriicais.
Reference: BORCHARDT, J.K., MATALAMAKI, D.W. Newsprint deinking:
Unit operations studies of flotation-wash deinking. Pu& Paper
Cun95(10): T374-380 (October 1994). Paper pre- sented at the 2nd
Research Forum on Recycling of the Technical Section, CPPA, at
%e.-Adele, QC, on October 5 to 7, 1993. Not to be reproduced
without permission. Manuscript received July 23,1993. Revised
manuscript approved for publication by the Review Panel April
11,1994.
Keyword*$ NEWSPRINT, DEINKING, FLOTATION, WASHING, UNIT
OPERATIONS.
~
I I R&D NEWS from around the world is brought to you monthly
by
I I
30 PULP & PAPER CANADA 95:lO (1994) & T380