CA2110190A1 - Control of treatment dosage to papermill broke - Google Patents

Abstract

Abstract of the Invention An on-line automated control system for use in drainage and retention chemical programs used in tissue, paper and board mills is disclosed. Utilizing strategically placed sensors in the wet end of a paper machine, the fiber charge and turbidity, prior to the headbox as well as in the white water loop, are constantly measured. Based on these readings, the addition of a chemical coagulant and a chemical drainage and retention aid are controlled via a computer system.

Description

~- 2 ~
Backgr¢und of the Inventio 1. Field of ~he l~tion The in~tention relates to a mechanism for controlling chemical treatment dosage to a papermachine broke system. - ~ ;

2. Description of the P~ior ~t Deposit forrnation in papermaking systems reduces machine efficiency, leads to poorer quality products, and ultimately results in a loss of pro~lts for the m~ll.
Deposits can form in the pulp mill and throughout the wet end, ~elts, dryer cans, and other areas of the paper rnill. The location of the deposits is somewhat dependent on the origin of the depositing materials, which can be biological, inorganic, or organic. - ~-Papermakers use terms like pitch, stickies, and tackies to describe organic substances which can lead to deposit formation. Pitch can be loosely categor~zed ~ i as either naturally occurring organics produced as by-products of the pulping process ("wood pitch") or s~ tic organics associated with chemicals used in ; -paper coatings ("white pitch").
White pitch is similar to wood pitch with respect to the types of problems it can cau~.e i~ papermaking, but the substances which charac~erize each are chemically different. Therefore, the conditions under which each type of pitch `~
forms deposits may not be the same.

2 ~ 3 ~

Si~ificant research has been perfo~ned in the area cf wood pitch but less appears to be known about white pitch. Information contained in the published literature includes descriptions of white pitch, descriptions of the problems associated with white pitch and the circumstances leading to its formation, testmethods used to evaluate white pitch treatment programs, and methods for treating white pitch problems.
Based on the information available in the literature, the white pitch depositionprocess does not appear to be well understood. A variety of circumstances are claimed to promote white pitch formation and deposition, but few if any are substantiated with data Four laboratoly tests which measure some aspect of the white pitch ~ormation process have been identified.
1. Hemacytometer method ;
2. Foil Deposition Test 3. Tape test 4. Vacuum Filtration Test None of the tests appear to be preferred over the rest for measuring white pitch related phenomena, and this may be due to a lacl~ of understanding of the white pitch depositionprocess.
Several white pitch treatment programs have been identified In the literature~
1. Talc (Detacldfier) ~:
:

2. Pulp Aids (Wetting Agent) 3. Coagulants (Retention Aids) The fact that the method of treatment for each of these programs varies, and that none appears to be preferred over the others, again indicates a lack of S understanding of the white pitch formation process. ~ `
Information contained in published literature regarding white pitch can be ~ ;
categorized as follows~
1. Description of problems associated with white pitch and circumstances :~. . ~, ;: . .;:.
Ieadingto its formation. ;
2. Description of White Pitch 3. Test methods used to evaluate white pitch treatment programs.
4. Methods for treating white pitch problems.
Desc~tion olr Problems Assoç1at~d ~ith White pit~h ~::
The term "pitch" is used by papermakers to desc~ibe a class of chernicals which tend to be sticly and can combine with themselves and other papermaking -~
components. This agglomera~ed material can then, under the right conditions, .~ -deposit on machille surfaces and in paper products.
~, ,;', ,.

~ .
' ~:~;~'''' - 4 ~

2 ~

T~BLE 1 ~ . ~ .
White Pitch Deposit Locations ~nd Associ~ted Problems ~:
I:)e~ositl~a~iQn ~mmQ~E~
d . : :~
Wet-en: .
Foil Wire Wear Flatbox Lower production rate, downtime to clean wet-end Wet-end breaks ".
Press Sectlon: ~-~
Press Felts Crushirlg, sheet qualicy Uhle Box High energy costs, frequent felt changes, frequent batch cleaning excessive downtimej breaks Dryer Section~
Dryer Can Poor coater nunnability Dryer Fabric Sheet spots, shee~ holes, frequent d~yer fabric cleaning : ~

~; 5 ; . i~

.,...,,~.,~, ",, ~' 2 1 1~

Table 1 lists various locations for deposit formation and includes con~non problems resulting from deposits. The deposit problems listed on this table are associated with white pitch although wood pitch can cause similar problems. The difference between the two types of pitch are the sources and chemistry of the sticky components.
Wood pitch has it's own origins in the chemicals liberated as by-pro~ucts from certain pulping operations. White pitch is characteri~ed by the presence of synthetic coating binders which are brought into the papermaking system through the use of coated broke as a furnish component. Synthetic binders are cherIucally similar to polymers used for adhesives, and can become sticly under certain conditions. These binders can be viewed as the "glue" which binds ~he white pitch complex together. ~ ~ `
A number of paperma!~ng operations havc been claimed to increase the probabili~ of white pitch problems. These can be fou~d in Tiable 2.
" "

.

~ 2 ~ 0 TW3LE 2 `
Conditions E~fectingWhite Pitch ~rmatiola '`'."~ '"' "`~
1. Closing White-Water Systems (Ref. 10) 2. High Machine Speeds (Re 10) 3. High Levels of Coated Broke (Ref. 10) 4. High Temperatures (E~ef. 11) ~ .

5. Long Stock Residence Time (Ref. 8) 6. Inefficient Repulping (Re 11) :

7. Synthetic Bi~derType (Ref. 11) 8. Calcium Carbonate (Re 11) ~;

-7- :~

The trend these days in papermaking is to reduce effluent by closing white ;~ ~;
water systems. Unless coated broke material is carried out in the sheet, it willcirculate for longer periods of time in the process waters which will favor agglomeration and white pitch formation. High machine speeds and high levels of S coated broke are related and translate to more binder brought into the system per unit time. High temperatures lead to conditions where the polymeric binders are raised above their glass transition, where they can become sticly. Inefficient pulping leads to ineffecti~e treatment of the individual coating components by chenucals used in white pitch treatment programs. ~ -Phvsi~llendçal D~scri~tion o~ Pitch :~
White pitch has been desc~ibed as a sticly, light gray substance which discolors ` -to a black deposit upon carbonization. It usually contains a relatively high concentration of pigment and synthetic binders common to the coating used in themill, and is often combined with resinous wood pitch and other organic mater~al ;
contained in both the wet end and the coating systems.
The composition of white pitch samples can be de~ermined by first extracting the organic portion ~nth a nonpolar solvent like toluene, chloroform, or methylene chloride. Hexane should not be used as it is not polar enough to solubilize mostwhite pitch components. Identification of the organic components can be performed by analytical techni~ues like IR or NMR. Table 3 (taken from reference 4) lists analyses of white pitch deposits taken from various locations in ,,~

2 ~
,`~.;~,`
the papermachine. Dryer scale deposits can contain up to 21% synthetic binder -material.

. " '.

^;~' ' ~ ,, .: ~ -. . :~ .
.
~'.'','.~',.
g ~ . ` ~ . `

, ,.

~ 2110~ sa T~BII E 3 White Pitch 12e~Qsit Sam~l~Anal~es % Solvent Solvent Ext. i ~L~ ~ ~om~ounda ~h ~ ~
Wet End ~ ~;
Foil Fat~y Ester ;
17~0 SBR, PVAC 25-60 -Flat Box Fatty Ester, 15-34 SBR, PVAC 33"58 Press Section . .
SBR, Fatty Acid Press Felt ~1 Fatty Ester ~6 Uhle Box 8-20 Fat~r Ester 18~0 :

Dryer Section Dryer Scale 1-12 PVAC, SBR 19-39 "~ ~';'.'`

'"'~` 2~l0~a The organic portion of whi~e pitch has been identified to contain some of the following:

- Fatty acids - Fatty esters - Starch - Protein - Polyvinyl acetate (PVAC) - Styrene butadiene (SBR) Fatty acids and esters, starch, and protein are indicative of wood resins, defoamers, biological substances, and natural coating binders. It can be difficult to establish which of these sources are actually contributing to a white pitch problem. PVAC and SBR on the other hand are generally indicative of synthetic cuating binder~ and can be traced directly to the use of coated broke as a furn~sh component.
When white pitch deposits or coated broke s~nples are analyzed at Nalco, the synthetic coating binders will be identified as polyvinyl acetate or styrene butadiene (See Table 4).

- 11 ~ ' '-'.~
. . ' ~." ~:

2 ~ 0 Test Results Acid/Ether Extract (~) <0.1 ; ;~
Methylene Choloride Extract (~o) 0.3 ... . ..

. . .

I~Qmments:
IR analysis of the acid-ether ex~ract indicates presence of: fa~ acid, carboxylic acid ester, aliphatic hydrocarbon oil, and styrene-butadiene ~ . .
copolymer.
IR of the methylene chlonde extract indicates presence of: s~rene-butadiene copolymer, aliphatic hydrocarbon oil, and catbo~rlic acid.
: `~ ~';'.,' . ~
IRD = CS~7446 .
~"

. . ~

- 12~
~ . ..

2 1 ~ O

Generalized structures of these binders can be found in Figure 1. Synthetîc coating binders can vary in polymer composition, molecular weight, and monomer types among other things.
In SBR, Styrene is considered the "hard" monomer and butadiene is the "soft" monomer which internally plasticizes the hard monomer so that filmformation and pigment binding are more readily achieved. The ratios of the two monomers can be varied to achieve different coating and printing characteristics.
SBR used for paper coatings contains 50% to 60~o styrene, although for products with special purposes (e.g. for high gloss development) SBR is manufactured with styrene levels outside this range 15.
SBR may be further modified through the introduction of vinyl acids into the polymer chain.
These acid modifiers are added to improve latex emulsion stabili~r, binding power, and coater performance, and can account for as much as 10% (weight basis) of binder.
Similar to SBR, PVAC can have vanable composition9 (e.g. incorporation of ~-vinyl acrylics into the polymer chain) 16 to alter coating characteristics.
Clearly, wi~h the broad range of variability possible with synthetic coating binders, it does not seem plausible to accept the generally held belief that PVAC
is more prone to white pitch Eormation than SBR. This claim is probably only true in the most general sense, as it is conceivable that there is some formulation I

. ~;;.

2 ~

of SBR which could be more problematic to wh;te pitch formation than a particular type of PVAC. : ~
The correlation of white pitch formation to coated broke usage seem ~:
understandable when looking at typical coating components and formulas. A list of typical components can be found in Table 5.

. :~ ~

~,, , .

2 ~

TABLIE~ 5 l~pical Paper Coating Chemicals Binders .
Styrene butadiene latex (SBR) S Polyvinyl Acetate Latex (PVAC) Acrylic Latexes :~
Polyvinyl alcohol (P~A) Casein Pi~n~nt~
Kaolin Clay Calcium carbonate .
Polystyrene pi~nent Titanium dioxide Additive~ ~;
Glyoxal (waterproofing) -Strearates (plas~icizer) Wax emulsion (plasticize~
Cellulose derivatives (thickener) Polyphosphates (dispersants) Lignosulphonate ((dispersants) - l ~y~s " '""'' ' -~,''"'''`''','',' ,' -, ` ' ..,:

,~

~'.-'',,''''' 2 ~

Basically, the binders form a film which holds the pigments to the basesheet.
The additives are present in the formulation to improve coating characteristics,ease of application, as preservatives, and as dyes. A number of the components found on this list (e.g. SBR, PVAC) are major components of white pi~ch and are S most commonly introduced to the wet end through the use of coated broke as furnish.
Coatings formulations can ~rary in amount and type of the components present.
Pigments make up most of the solids in any formulation, but whether clay, calcium carbonate, or titanium dioxide are present is dependent on the desired ~ 6 characteristics of the end-product. Binders can make up approximately 20~
(weight %) of the coating solids, but sirnilar to the pigment, the type of binder (or binders) present can vary. The additives generally make up a much smaller percentage of the coat~ng formulation (<5~o). Most coating forrnulations are proprietary, but an example of a published ~ormulation is provided in Table 6.

:

2~10~
, T~LE 6 .
~pical (:~oating Formul~t~on :
No. 1 Web Ofrset Coating Grades .
, l~rc~i~ms Parts/1OQ Pi~nt No 1. Clay 7 CACO3 2~0 `
TI02 10-0 ~; .
Dispersant 0.1 i-: :.,,:,'' Lubricant 1.0 , ~ .: . .
Starch 3 Styrene-Butadiene Latex 13 (Small Particle size Type) ~:

Total Solids, ~o 63 pH 8 ; ` ;~
Viscosi~ (CPS 20RPM) 2000 .,''~' :; ,''.''.'`'"-,'' :
:~ ., . ,:, - 17 ~

::

. . . . , . . ,,,- , . , , . " - i .. , , , , , " . . .-, . . .

```~ 2~01~0 ~
~` ~ `` .

Test Methods Used to Evaluate White Pitcb Treatment Programs Several approaches have been taken to measure white pitch potential on a laboratory scale. One approach uses a method originally developed to measure wood pitch potential, and with the belief that wood pitch and white pitch components behave similarly. This method has been adapted to measure white pitch potential. Other methods examine th~ sticly characteristics of various components or measure the ability of coagulants to retain pulped coatingparticles.
1. Hemacytometer Method ~ `
A hemacytometer method for counting wood pitch particles to include counting ;~
colloidally dispersed particles associated with pulped coated broke. The procedure involves counting the number of colloidally dispersed particles in a ` -given volume of iiquid using a hemacytometer counting cell under a 600 to 900 ` ~``
power microscope. This techni~ue was originally developed for counting the `
spherical particles associated with wood pitch. An improvement on this method involves counting dispersed spherical particles generated when base sheets made ~ ;
with and without groundwood is repulped. This data was compared to counts of dispersed spherical particles generated from repulped coated sheets made w~th ~ ~
and without groundwood. The data indicated that the coated sheets produced ~ ;
higher counts than the basesheets. l'he claim is that some fraction of the . 18~

-~` 2L1 0~90 ~, synthetic coating binder is "soluble" and rnay produce spherical particles sirnilar to ;
wood pitch, and that the soluble material could be related to white pitch.
The basis for this claim is not clear since common coating binders like PVAC
and SBR do not have the ability to form spherical m;celles like typical wood pitch components. Synthetic coating binders form plastic-lilce filrns after application, ~ ~
which after repulping rnight be expected to remain attached to pigment and fiber ~;
to form fragmented (non-spherical) particles.
, :,. .
Aside from lack of clarity as to which of the particles viewed on the ~ ~
hemacytometer cell to include in the count of white pitch particles, this test also ;;
appears to be fairly tedious to per~orm (counting many small particles) and ;~
requires relatively expensive equiprnent (a 600 to 900 power rnicroscope).
2. Foil Deposition Test A second test has been developed to measure the conditions under which synthetic binders will form white pitch deposits. The test involves repulping coated sheets and using a Noble and Wood sheet mold to form hand sheets. These ~ ~
sheets are then pressed, placed betweerl two pieces of foil, and this "sandwich" ~ ~ `
is then heat-sealed. After cooling, the foil is separated from the sheet, and the ~ ~
amount of deposited material on the ~oil is compared to a chart of standards. ~ -The cha~t shows various levels of deposition, with rankings of for little or no adhesion, to for large amounts of deposition. -~9-1 9 ~

The test was used to evaluate parameters such as binder type (PVAC, SBR, acrylic), polymer variables (molecular weight, surfactants, catalysts), binder quantity in the coating formulation, pigments, and process variables (pulping temperature) for their influence on deposition.
This test appears to be lacking in subjective treatment of results by using a chart of standards to evaluate deposition levels. Also, the standards have not been correlated to various levels of white pitch deposition occurring on paperrnachines.
A more quantitative method of determining deposition potential by thei foil testhas been described by the literature. Here a peel strength (g/cm width) is measured as the force required to peel the foil from the paper in the heat-sealed "package" mentioned earlier. What this improvement gains in quantitative data, it must lose in time required to run the test.
3. Tapç Test A third test has been developed to study the sticky nature of material found in secondary fiber. The test utilizes adhesive-backed tapes made from s~rene - ;~
butadiene and vinylic esters as stickies coupons. A second coupon is made from polyethylene terephthalate film in order to simulate the type of material used to manufacture forming fabrics and felts. The tape coupon arld the polyester couponare exposed to chemical treatments and then pressed together. An Instron tensile -tester is used to measure the peel strength of the bond formed between the . 20 -^ -" 21`:~19~ ~

separately, since wood pitch interacts with latex binders, making it more likely to leave deposits. It is claimed that the small talc-coated particles move through the `~
paper machine without leaving any deposits. ~;
Blair 12 has reported that a particular mill has successfully used talc to reduce white pitch deposits on dryer cans. ;
The use of talc to treat white pitch is an extension of its use as a wood pitch treatment. The goal is tbe same in each application; to detackify the sticly surfaces of the pitch particles. But if the talc-covered pitch particles are notremoved from the papermaking system, either by retention in the sheet or by boiling out, these particles will circulate and lead to abrasion of machine and fabric surfaces or may eventually deposit themselves.
2. Pul~ing~ids A second method of white pitch control involves adding a "latex control product" to the broke pulper at a dosage of 3~ Ibs/ton broke. The latex control product is an unidentified chemical which helps wet out coated broke particles thereby preventing them from reagglomerating and forrning deposits on equipment surfaces. ~ ~
l'he propensity for white pitch particles to deposit may be related to th~ size of ;
the particles. It is believed that a direct link exists between white pitch deposition and the efflcienar of brealdng up coated broke particles during repulping. In aaprocess, a 1:1 ratio of tetrafunctional alkoxylated diamine and an anionic ~ ~ , ," ` "," ~ " ,, "~

~ 21101~0 ~

compound con~aining a phosphate or phosphonate functionali~ are added ~ ~ ;
separately to a coated broke repulper. This program is used in conjunction with other additives, like retention aids to carry the treated broke particles out of the ;
paper machine.
S ~QB~ ' ~ third approach to solving white pitch problems is to retain the coated broke particles in the sheet by early coagulation with a cationic polyelectrolyte. Themost widely used coagulant is an arnmonia crosslinked epichlorohydrin/dimethylarnine copolymer. Some success has been reported when ~;:
using this product to reduce white pitch related problems. The coagulant is fed at -. . . ~.
a point somewhere after the coated broke has been pulped, but before it has beenblended with another stock stream.
Laboratory evaluations indicate that any cationic coagulant is active in this application, but increases in efficiency over other coagulants may be gained by using high molecular weight poly DADMAC or DADMAC/Acrylamide copolymer.
This method of treatment has an advantage over previous methods in that i~
leads to the removal of the source of white pitch from the paper machine v~a thesheet. With the closing of white water systems to reduse eftluent, retention appears to be the most logical choice for treatrnent.

"'''~'' 2 1 1 ~ 1 9 ~

~. ;, `. .
In addition, the use of a coagulant fed to the broke system has the advantage ;
of minimizing ~uctuation in the norrnal retention aid program (fed closer to the ;;;
headbox) caused by pigment coming in with the broke. ~ `

S Finally, one other solution to white pitch problems is to change coating formulations. Mills have been able to identify particular synthetic binders which contribute to white pitch problems. These binders can be replaced with others which will still provide tbe necessa~y binding characteristics and also reduce the white pitch problem.
~mm~L ' ` ~ '' The amount of published literature specifically on white pitch is minimal.
This literature has been reviewed, and it contains inforrnation on a variety of treatment programs and on tests used to assess the performance of treatment strategies.
S The treatment prograrns tried to use talc, pulping aids, or coagulants.
Changing synthetic binders in coating formulations has also been used as a method for soh~ing white pitch problems. At this time, no clear consensus existson the most effective treatment which may be attributable to a lack of understanding of the white pitch fom ation process.
The test methods available^ measure colloidally dispersed particles associated with pulp coated broke, deposition potential of valious coated broke components, -`-``` 2 1 1 0 ~ 9 ~

In addihon~ the use of a coagulant fed to the broke system has the advantage of minimizing nuctuation in the norrnal retention aid program (fed closer to the ~ `~
headbox) caused ~y pigment coming in with the broke.
~Coatin~Forrnulation Selection Finally, one other solution to white pitch problems is to change coating formulations. Mills have been able to identify particular synthetic binders which contribute to white pitch problems. These binders can be replaced with others which will still provide the necessary binding characteristics and also reduce the white pitch problem.
~mm~
Thie amount of published literature specifically on white pitch is minimal.
This literature has been reviewed, and it contains information on a variety of treatment programs and on tests used to assess the perforrnance of treatment strategies.
The treatment programs tried to use talc, pulping aids, or coagulants.
Changing synthetic binders in coating formulations has also been used as a method for soh ing white pitch problems. At this time, no clear consensus existson the most effecti~e treatment which may be attributable to a lack of understanding of the white pitch fonnation process.
The test methods available-measure colloidally dispersed particles associated with pulp coated broke, deposition potential of vanous coated broke components, - 24 - ;

~, "' ~

2~10190 , , bond strength of adhesives, or retention of coated broke particles after coagulation. Again, no consensus exists in the industry on an appropriate method, likely due to a lack of understanding of the factors influencing white pitch ~ormation. ;
S Polymeric coagulants are shown to have signiffcant differences in molecular ~ ~;
structures. Properties shown to differ among common coagulants are: 1) chemistry of the monomeric unit, 2) charge density, 3) molecular weight averages ; ` `~
and distributions, and 4) the de~ee of three dimensionality. The activity of polymeric coagulants in retaining coating materials in repulped coated broke wasexamined in brokes from three paper mills. Different brokes were found to be treated best by polymers with different structures indicating that screening of a range of polymers is necessary to select the best treatment proerarn. Polyrner activity was found to be influenced by average molecular weight and monomer chemistry and not by the amount of cationic charge present. For this application, lS the coagulants were not simply sources of cationic charge.
The amount of published literature specifically on white pitch is rnirlimal. This ~ ;~
literature has been re~iewed, and it contains information on a varie~ of treatment programs and on tests used to assess the performance of treatment strategies.
l he treatment programs tried use talc, pulping aids, or coagulants. Changing synthetic binders in coating formulations has also been used as a method for solving white pitch problems. At this time no clear consensus exists on the most 21~01~0 effective treatment which may be attributable to a lack of understanding of the white pitch formation process.
Much of the literature contains claims of success associated with the testing methodologies or treatment prograrns, but few if any of the clairns are S substantiated w~th data. Until test and treatment results obtained in the laboratory are verified with machine trials, a better understanding of how to ;
predictively reduce white pitch problems will not be forthcorning.
Summaq o~ the In~tlon This invention provides a mechanism to control the chernical treatment dosage ;~ :
to the broke system of a paperrnachine. Dosage control is accomplished by first measuring on-line eitber the streaming potential of the broke or the turbidity of the broke's filtrate. The measurement of the streaming potential of the broke orthe turbidity of the brolce's filtrate is accornplished through sampling of the water ;
discharged from the broke chest by a detector. The detector measures either . , ...~ ~ .
turbidity or streaming potential, and this information is then fed to a controller which automatically adjusts the treatment dosage to achieve a preset value of streaming potential or filtrate turbidi~y. The polyrner is supplied through a pump, which is connected to the controller by electronic signali~g mea~s. The controller varies the amount of polymer being supplied to the broke stream by the pump.

'~\ 211 al ~0 66530-546 : According to one aspect of the present invention there is provided an apparatus for controlling dosage of treatment chemicals for coated broke, the apparatus comprising: a sensor for measuring flow charge of the coated broke; a sensor for measurlng conductivity of the coated broke; a sensor for measuring `; "
turbidity of the coated broke, the sensor having four electrodes;
a data base conneated to a computer terminal; a printer connected to the computer terminal; a control system electronically conneated to the data base; a dosing pump connected to the control system; wherein the sensors are electronically connected to the data base; and whereby the dosage of treatment chemicals from the . ;:
dosing pump to the coated broke is varied by the control system ~ `
according to interpretation of the sensor signals by the data base. ~.
In preferred embodiments the turbidity sensor comprises :~:
a Monitek low consistency measurement system; the apparatus may further comprise a sensor for measuring p~ levels of the coated .~. ~
broke or a sensor for measuring temperature of the coated broke. ~`i . ~

', ~

26a 2 1 1 ~
`. .
The invention is an on-line automated control system for use in drainage and retention chemical programs used in tissue, paper and board mills. Utilizing ` `
. ~:
strategically placed sensors in the wet end of a paper machine, the fiber charge ` ~;
and turbidity, prior to the headbox as well as in the white water loop, are `~ ` `
S constantly measured. Based on these readings, the addition o~ a chemical ; ` ~
coagulant and a chemical drainage and retention aid are controlled via a ` ;`computer system.
Broke is unsalable paper which is pulped and returned to the papermachine as ~ ~ ~
a stock source. It may often contain materials which were not present as ini~ial ;
furnish components but which were introduced at the dry end of the machine as coatings. The coatings contain predominantly rnineral pigments which are glued together with various binders. These binders are sources of '`white pitch" whichcause machine runnability problems and poor quality paper. The pigments provide a continuously va~ng and unknown amount of filler in the final paper lS furnish which upsets the retention program used to retain Sllers into the paper sheet. Chemical treatments can control these problems. ;
. :.
As in any treatment, the best performance and cost-effectiveness is achieved -when the optimum dose, i.e. the lowest dose to achieve the desired result, is applied. Broke stocksourcesvaryuncontrollably in concentration and composition. Current strategies used to feed chemical are 1) a fixed chem~cal feed rate, 2) a chemical feed rate based on stock flow, and 3) a chernical feed - 27 ~

, ' ', ,'::

~ 2 ~ 1 0 1 9 ~

rate based on solids flow as measured, for example, by a "mag" meter. None of these strategies feeds chemical on the basis of the need for treatment. Ihe first two take no account of solids' concentration or composition. The third strategy accounts for concentration but not for composition changes.
S The appropriate chemical dosage to the broke can be deterrnined by measuring a filtrate turbidity or a zeta potential measurement, like streaming potential, in the laboratory. On-line equipment for measuring strearning potential has recently become available from Innomatic and is referred to as th~ FCA (~Iber charge analyzer). This instrument perforrns a filtration as part of the streaming potential ; `
. ~
measurement, and the addition of Monitek's turbidity sensor to the FCA allows the turbidity of the f;ltrate to be measured in a manner sirnilar to that practiced in the laboratory. After selecting an acceptable value of filtrate turbidity or streaming potential corresponding to effective broke treatment, the output of these sensing devices can be used to control the amount of polyrner needed to achieve the optimum treatment level. In this way ths problems of underfeeding and overfeeding are eliminated, and the broke is "stabilized". Using this dosagecontrol strategy, machine runnability and cost-performance of the coagulant treatment are irnproved significantly.

~ .

Bfief DescriDtion of the Dr~wings Figure i is a schematic representation of the control system of the invention as ~ ~
connected to a paper machine. `
~b~ `:
A treatment compound is fed at 1.0 to 1.5 Ibs/ton with the dosage being controlled by the turbidity measurement generated by FCA filtration and the Monitek turbidity unit. In Figure 1, hardware of the control system is given schematically. Se~eral characteristics including turbidity and flow charge may be measured by the apparatus. For example, its has been shown that a strong correlation between conductivity and fiber charge exists. The measurement system for analyzing conductivity is a sensor w ith four electrodes.
A second measurement that is analyzed by the apparatus is turbidity. A ~ ~;
component for measuring turbidity is part of the analyzer. The component is `~
available as the Monitek low consistency measurement system. ;~
The filtrate of FCA is continuously led to a small vessel. In the vessel, there is a sensor for turbidity and a propeller for mLlnng the filtrate. The filtrate is then taken to the bottom of the vessel.
Serlsor I measures the turbidity of the pulses programmed by the users. The sensor is o~ the pipeline taldng the ISltrate out from FCA~
Sensor 2 measures the tot~l filtrate from charge measurement pulses.
','''''`;:

2 ~

Sensors for pH and T may also be installed, because they may often have an effect on the results.
The FCA an~ Turbidity are at ~lrst used as control pararneters. Other signals like conductivity, temperature and pH are only used for data monitoring. This S makes it possible to investigate correlations between parasneters, e.g.
compensating dosages based on interactions between pararneters.
Data acquisition will be handled by a data base connect~d to a computer terminal. This makes it possible to collect relevant data on any given process-related parameter. ;
The control system can be based on FCA and Turbidity systems already present in a given paper mill. The controller, the PC with printer and all necessary sensors will be installed. It is preferable to install the PC and the ;
printer in the paper machine control room.
At first, for collecting data and testing the control s3~stem, the output of theFCA/Turbidity controller is not connected to a dosing pump. The installation will ~;
be on a spare dosing unit. For testing we switch over from the operating unit tothe spare unit.
Before actually hooldng up the FCA/Turbidi~ control S3 stem to the paper -machine a testing period will be introduced ~o collect data and check the functioning of the control systç~ By recording the actual signal to the dosing pump, which in most systesns is controlled by the massflowcontroller, and the - 30 ~
'', ~.

~-` 2 ~ ~ 0 li ~
. ...

signal produced by the FCA/Turbidity controller we ca~ compare the actual ,.
dosing to the "fictive" dosing of the FCA/Turbidity controller. This should lead to a reduction of the chemical dosing, over a long time period.
When the controller is hooked up to the paper machine, we switch dosing ., , ~
control back to the original dosing pump. We then connect the controller signal to the spare pump and switch dosing control back from the original pump to the FCA/Turbidity controlled spare pump. If, for what reason, anything goes wrong with the FCA/Turbidity controller we can easily and rapidly switch back aBain to . .:
the original massflow controlled dosing unit.
..... .
' ' ~'~'''''"''"'''"'`
U~o~FC~/lbrbidilyE~l~ment Trials were run in which the dosage to the broke was controlled by the strean~ng potential measurement and the turbidity measurement. Both measurements thought to be beneficial for controlling dosage, and tracked ;
together for the most part. The turbidity was preferred because it was generallymore stable and more easily interpreted. Broke stabilization was reported to be -~;
substantially increased. Retention levels on the machine were reported to have ~ ~ -improved during the period in which the coagulant dosage to the broke was ~ :
controlled. This was possibly is an additional benefit of the control method. ;During equipment trials, dosages of the treatment compou~ were reported to be reduced by 1/3 when controlled by charge and by about 1/2 when controlled ," ,~" "~ "-~

~ ` 2 ~ 3 ~ ~ :

by turbidity. The set points for control were not the same, i.e. the turbidity selected did not yield the charge value selected for control, making the dosage reductions incomparable. Nonetheless, this information clearly indicates that very significant cost reductions can be achieved by adjusting the polymer dosage to the changes in broke composition. The Turbidity set point was 0.8 AU (Absorbance Units) which would correspond to a % transmittance of 45~o. Although this value would appear low, the path length is not known and the disruptive mat formation sequence may be expected to allow easy passage of particles through the mat. At thls absorbance value, the FCA charge was reported to be -201, and the set pointused for controlling by charge was -275. No problems with plugging the filtration screen have been encountered when utilizing the control method. -CHARACTERISTICS OF POLYMERIC COAGUL~TS
In order to draw any conclusions regarding how polyrner structure affects activity, knowledge of the polymer structure is obviously required. This type of ~ "
information has generally been in short supply. Consequently, progress in identif~ring structure/activity relationships has been retarded. Polymeric ~ ;coagulants can vary in many ways: ~
Table 7 provides a list of structural characteristics which can differ among - ~ `
.
coagulants. Each of these characteristics will be discussed individually, and exemplary data pertinent to the broke activity testing will be presented.
,",'~

- 32 - `~
'" ,'""''`'' 2 '1 ~ v~

TABLE 7 :

1. Chemical composition 2. Charge density ` ~
3. Molecular weight ~`
-- Average -- Distribution 4. Three dimensionali~

r"~

2'1 '1~0 CHEMICAL COMPOSITION
The monomer units found in some of the common polymeric coagulant types include many general monomer unit types, but major commercial types are also represented (e.g., EPI/DMA, PEI, p-DAI)MAC). Idealized linear structures are S provided for the EPI/DMA and I ;DC/NH3 polymers which can be cxpected to exist more comrnonly in complex cross-linked structures.
A very simple difference among the coagulants i~ the presence or absence of pH-dependent charge. The quaternary ammonium containing polymers like EPI/DMA, DAD~AC, and DMAEA-MCQ have cationic charge independent of pH, while the EDC/NH3's charge vill vary with pH because it contains a secondary ammonium structure.
The location of the cationic ammonium site is another structural variation evident among the monomer chemistries shown in Figure 1. The cationic center can be part of the polymeric backbone as in EPI/DMA and EDC/NH3 or it may exist as part of a side chain as in DADMAC and DMAEA-MCQ. The cationic charge of the DADMAC monomeric unit is contained in a unique ring structure.
The DMA~A-MCQ monomeric unit exemplifies a pendant-type structure in which the cationic charge is positioned at the end of a linear side chain. Many such pendant-type monomers exist which differ by pendant length, type of pendantattachment to the backbone (generally ester or amide), and acrylate or methaclylate backbones. Coagulants synthesized with pendant monomers are not ` ~ 2 ~ 4~ ~3 generally cost-effective in most applications because the monomer costs tend to be greater than common non-pendant ~pe monomers. Interaction of the polymers' cationic charge with the negative surface clharge could conceivably depend on the polymers' ability to distribute their charge over the particle S surfaces. This ability is regulated by the degree of freedom or ~lexibility afforded the cationic quatcrnary ammonium site by the monomer unit. Whether this actually has an effect is not known.
The different monomeric units can also be expec~ed to interact with water differently. For example, all the monomeric units except DADMAC are capable of hydrogen bonding. Because of this, DADMAC has been referred to as more hydrophobic than the other monomeric units, and performance improvements `~have been attributed to it on this basis in some applications. Although this hydrophobic hypothesis seems plausible on a chemical basis, unequivocal evidencedemonstrating this effect has not been presented. Clearly, even this small selection of monomerie units contains significant structural varie~ which is potentially capable of influencing coagulant activity.
CHARGE DENSIIY
Polymeric coagulants are frequently referred to as 100% cationic. This statement does not mean that all the coagulants contain tbe same amount of charge but rather that a charge resides on 100% of the monomeric units, i.e., 100 mole% charge. Because the molecular weights of the monomeric units dif~r 2 ~

sign~ficantly, the charge per unit mass of polyrner will di~er among the coagulants as will the amount of charge added to any particular system for ~he same mass ofpolymer.
Table 2 compares the charge density in rneq/g o~ polymer for a selection of S coagulants. The values given are theoretical charge densities, and again, the idealized linear structure and appropriate pH have been assumed for EDC/NH3 and EPI/DMA. The theoretical values can `be veri~led experimentally using the colloid titration when cross-linking is absent. The last entry in Table 2 is a ~;
copolyrner of acrylamide and DADMAC, and it illustrates that consideration of copolymers can broaden the range of potential charge densities appreciably, if desired. Because charge reduction is most frequently cited as a function of coagulants, the polyrner charge density can be expected to strongly influence coagulant activity. ;
MOLECUL~R WEIGHT
Polymer molecular weight is a characteristic capable of extensive variation.
However, literature indicates that for strict charge neutralizatioll, molecular weight should be unimpor~ant. Consequently, if charge neutralization dominates coagulant activity, the ability to vary the molecular weight can be expected to be of little value.
~.
Characterization OI a poly~er by some so~ of single value "molecular weight"
is commonly encoun~ered. Any real sample and particularly, any commercial 2 ~

polymer will be a collection of many molecular weights, i.e., it will have a distribution. The single molecular weight values will be some sort of statistical average whose meaning is subject to all the problems associated with averages.
Frequently, coagulant molecular weights are compared using viscosities S determined in a solution of fixed ionic strength. The validity of such a comparison depends on a well-behaved molecular weight distribution and `
identical solvent~polyrner interactions. The latter assumption is guaranteed to be invalid for polymers based on different chernistries.

-2 ~

~ABLE 8 Pol~mer Charge Densities C:harge densl~ Chemlstry(nneq/g polymer) EDC/NH3 12.S8 EPI/DMA 7.27 p-DADMAC 6.19 p-DMAEA-MCQ 4.84 -AcAm/DADMAC 2.59 :~: -Polymer (76/24 mole ratio) ~ :-2'~

Size exclusion chromatography (SEC) is capable of providing the full f h molecular s1ze dlstrlbutlon o t e po ymenc coagulants. If a one-to-one ;
correspondence exists betweeTI molecular weight and size, it provides a molecular weight distribution. Such a correspondence w~ll exist for linear homopolymers but S will break down ~or copolyrners and cross-linlced polyrners to varying degrees depending on the circumstarlces. Since the entire distribution is available fromthis method, averages and median values as well as measures of polydispersity or ~ ~`
distribution broadness can be determined. ; ~ `
Manufacturing processes can lead to different distributions even if the polymers are made from the same starting composition. Such a situation occurs for the cross-linked EPI/DMA polymer. The data suggests that a wide variety in both average molecular weights and distributions can be found among commercial ;~
coagulants. Should coagulant activi~r depend on molecular weight, significant differences in their performances should be expected.
THREE DIMENSIONALITY
This polymer proper~r is probably the least well characterized. Water-soluble polymers are generally thought of as linear chain molecules. However, many factors may cause one chain to bond to another chain or to cross~link. When cross-linldng occurs, the linear structure is lost, and a three dimensional or more globular structure is forrned. Monomer impurities may cause significant cross-linking, or agents may be added intentionally to create cross-linking. Although the cross-linking agents are generally added with the idea of increasing molecular weight or size, they result in the generation of a three dimensional polymer structure as well. The effects of three dimensionality in the polymer structure on activity is not clear but excessive cross-linlcing is generally perceived to be bad ~or activity because charge accessibility and hydrodynamic size may be lost.
This brief examination of the molecular characteristics of some common ~ `
polymeric coagulants was meant to exhibit that a wide range in structural characteristics can be found among these coagulants. The data shown demonstrate that coagulants may differ significantly in ways other than the ; ~-amount of charge available, but do these other factors influence their activity to any extent? This question will be examined in the next section which deals with ;
the coagulant treatment of coated broke. i:
COAGULANT TREATMENT OF COATED BROKE
The use of coated broke as part of a papermaking ~urnish introduces substances which would not normally be part of the base paper furnish. The bulk of these substances is coating pigments and their binders. The latex polymer binders, like the s~rrene-butadiene or polyvinylacetate resins, for example, are the source of the so-called "white pitch" which can sause various sheet defects and runnability problems. A useful strategy to control these detrimental substances is to prevent their accumulation by removing them for the papermalciDg system. An 21~ 0~ ~B ~`

effective way to accomplish rernoval is to retain the substances in the sheet by ~ ;
early coagulation with polymeric coagulant. The addition of the coagulant also has the advanta~e of rninirn~zing ~luctuations in the normal retention aid program ~ `
fed near the headbox caused by the broke pigment. Coagulant testing with brokeisS from three different mills are reported here.
EXM~PLE:
Retention of coating substances was evaluated using a simple Buchner funnel filtration test. Broke samples (200 rnl) were filtered to completion using a very coarse filter paper, e.g., Whatman Reeve Angel 802; and the resulting filtrate turbidity was taken as a measure of retention. Treatment of the broke with coagulant was conducted using a Britt jar n~Lxer at S00 rpm and polymer concentrations of 0.50 weight percent. After treatment, the broke was transferred to the Buchner funnel and vacuum filtered. The reduction in turbidity from the original value was followed as a function of polymer dose. The consistencies of the brokes examined were: 1) Mill A, 2.8 weight ~o; 2) Mill B, 5.3 weight ~o; and 3) Mill C, 2.8 weight %. The brokes from Mills A and C were pulped in the laboratory from dIy broke while that from Mill B was the mill's wet broke.
A reduction in the filtered turbidity for Mill A's broke was shown in for sets of EPI/DMA aIld p-DADMAC coagulants with variable intrinsic viscosities. Two of the EPI/DMA polymers are of the cross-linked type, but they differ in the cross-linker used. Polymer activity data can be evaluated in two ways. The polymers 2 ~

can be judged or compared on the basis of their ability to attain the greatest reduction in turbidity regardless of the dose required; this method of evaluation w~ll be called effectiveness. Generally, many polymers are capable of achieving an acceptable performance level. Under these circumstances, the polymers are S compared on the dose required to achieve some given performance level; this method of evaluation will be referred to as the polymer's efficiency. At a performance level of 80% turbidity reduction, the p-DADMAC (IV = 0.43 dl/g) is considerably more efficient than a cross-linked EPI/DMA, but the latter is equally as effective as the former because it can achieve the same max~mum ~ ;
turbidity reduction. Acceptable performance levels will vary from mill to mill, but turbidity reductions of 80~o or greater are generally acceptable. Consequently, efficiency will be most important for comparing polymer performance.
Inspection of the data reveals some unusual results. The three p-DADMACs exhibit nearly identical performances regardless of intrinsic viscosity, which should be a reasonable indicator OI relative average molecular weight. Although not quite as identical, the three EPI/DMA polymers also show very sim~lar performances. Some differentiation in effectiveness appears to o~ cur at the highest performance level. In general, the p-DADMACs are about twice as efficient as the EPI/DMAs in this broke. Among the EPI/DMAs, the linear polymer is slightly less efficient and effecti~e; but considering large differences in molecular strucblre caused by cross-linking, the differences in per~o~nance are -- 2 ~

considered to be surprisingly small. This indicates that the three dimensionality of the cross-linked polymers has only a rninor effect on perforrnance. The absence of any performance differences between the p-DADMAC polymers indicates that rnolecular weight does not ir~luence polyrner activity in treating this broke. Also, S inspection of ~able 2 shows that charge densi~ of the EPI/DMA polyrners is greater than that of the p-DADMACs. A given performance level is achieved by the p-DADMACs with considerably less charge. Therefore, charge density would appear not to play a significant role in differentiating polymer performance.
Through this elirn~nation process, the polyrner characteristic which seems to bemost important in differentiating polyrner performance in this broke is ~h~
chemistry of the rnonomeric unit.
Because this observation was considered to be unusual, the testing was repeated and an additional chemistry type (DMAEA-MCQ) was included. Also the range of p^DADMAC IV was expanded to include a value of 1.64 dl/g at the high end. The p-DADMACs again exhibit identical behavior even with the increased range of average molecular weight; and as before, the EPI/DMA
polymers are clustered with performance significantly less than the p-DADMACs.
The new chemistIy (p-DMAEA-MCQ) examined here was also found to cluster, and performance was independent of molecular weigh~ is observation reinforces the notion that the chemistly of the monomeric unit is the factor differentiating polymer a~iYity. Charge does not appear to be impor~ant as the highest charged polymers (EPI/DMAs) per~orm worst, the intermediate charged (p-DADMAC) the best, and the lowest charged (p-DMAEA-MCQ) in the rniddle.
An acrylarnide/DADMAC copolymer chernistry type was compared with the p-DADMAC and a cross-linked EPI/DMA. Even though the copolymer has the lowest charge density by far (see Table 2), it clearly shows the greatest activity among all the polyrners examined. Because the copolymer is expected to be much less hydrophobic than p-DADMAC due to the presence of the acrylamide, attributing the DADMACs activity to its hydrophobicity is inappropriate.
The activities of the p-DADMAC and EPI/DMA polyrner sets examined earlier in Mill A's broke were investigated in a broke from araother ~ill (Mill B).
The pattern of polymer activities in Mill B's broke is very differe~t then that observed in Mill A's broke. Unlike earlier results, inspection of the activity variations among tbe three p-DADMACs reveals a strong efficiency and effectiveness dependence on molecular weight. Again charge density does not appear to play an important role ~n differentiating polymer activity. The cross-linked EPI/DMAs which contain a hi8h molecular wei8ht fraction also exhibit activity as good as the p-DADMAC polymers, indicating the chemistry of the monomeric uDit is not critical.
The polymer with the highest average molecular weight would be predicted to exhibit the best performance. To test this hypo~hes~s, the S0/50 weight %
acrylamide/DADMAC copolymer was examined, as it had the highest average molecular weight arnong the polymers examined. The copolyrner was also with two of the EPI/DMA polymers. As predicted, the copolymer was both the most efficient and most effective.
Test results from the broke of a third mill (Mill C) were compared to the results for coagulants examined in the brokes of Mill A and Mill B. The pattern of polymer activities obtained in Mill C's broke can be considered as a combination of those observed in the brokes ~rom Mills A and B. Relative polymer performance in Mill C's broke appears to be dependent on both average molecular weight and chemistry of the monomeric unit. The DADMACs as a whole do better than the EPI/DMAs; but within each groupl the highest molecular weight performs best. The highest charge density polymers (EPI/D~A) again perform the worst in terms of both efficiency and effectiveness,and the lowest charge density polymer (AcAm/DADMAC copolymer) performs the best, indicating the ur~importance of this factor in differentiating activity.
Even a crude characterization of commonly ~vailable polymeric coagulants demoDstrates that considerable variations in their molecular structure can exlst.
Coagulants were shown to vary in the chemistry of the monomeric Ullit, charge de~si~r, molecular weight properties, and struchlral three-dimensionality. A
number of observations can be made concerning the activity testing of polymeric coagulants for ~he treatmene of coated broke. Different brokes were found to be treated most efficiently and effectively by polyrners with different molecular . :
~5 -~,. ' ' , ' ~ ~ ' '' ', . ' . .

2~1~190 structures. Consequently, the screening of a range of coagulants is necessary toselect the treatment program best suited for a given broke. Coated brokes were identifiled in which the chernistry of the monomeric unit and the molecular weight of the polymer were the molecular features controlling polymer activity. In the S brokes examined, charge density was never found to differentiate the relative perforrnance among polymers.
Changes can be made in the composition, operation and arrangement of the method of the present invention described herein without departing from the concept and scope of the invention as defined in the following claims:

~, :. ~: ~,

Claims (5)

1.An apparatus for controlling dosage of treatment chemicals for coated broke, the apparatus comprising:
a sensor for measuring flow charge of the coated broke;
a sensor for measuring conductivity of the coated broke;
a sensor for measuring turbidity of the coated broke, the sensor having four electrodes;
a data base connected to a computer terminal;
a printer connected to the computer terminal;
a control system electronically connected to the data base;
a dosing pump connected to the control system;
wherein the sensors are electronically connected to the data base; and whereby the dosage of treatment chemicals from the dosing pump to the coated broke is varied by the control system according to interpretation of the sensor signals by the data base.

2. The apparatus of Claim 1, wherein the turbidity sensor comprises a Monitek low consistency measurement system.

3. The apparatus of Claim 1, further comprising a sensor for measuring pH levelsof the coated broke.

4. The apparatus of Claim 1, further comprising a sensor for measuring temperature of the coated broke.

5. An apparatus for controlling dosage of treatment chemicals for coated broke, the apparatus comprising:
a sensor for measuring flow charge of the coated broke;
a sensor for measuring conductivity of the coated broke;
a sensor for measuring turbidity of the coated broke, the sensor having four electrodes;
a data base connected to a computer terminal;
a control system electronically connected to the data base;
a dosing pump connected to the control system;
wherein the sensors are electronically connected to the data base; and whereby the dosage of treatment chemicals from the dosing pump to the coated broke is varied by the control system according to interpretation of the sensor signals by the data base.