Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
METHOD TO DETERMINE PULPING YIELD
Document Type and Number:
WIPO Patent Application WO/2003/100166
Kind Code:
A1
Abstract:
Methods are provided for determining the yield and/or hygroscopicity of a pulp from the dimensions of fibers obtained from a sample of the pulp. The yield and/or hygroscopicity of softwoods and hardwoods may be calculated using the methods of the present invention. The methods of the present invention are especially suited to on-line measurements obtained during the pulping process.

Inventors:
Makkonen, Hannu (Sivukatu 42, FIN-87500, FI)
Cöpur, Yalcin (Büyüksehir Uydukenti, A-16-B Block. D.41 Beylikdüzü, Büyükcekmec, Istanbul Turkey, TR)
Granzow, Stephen (2331 Bremen Court, Punta Gorda, FL, 33983, US)
Application Number:
PCT/US2003/016022
Publication Date:
December 04, 2003
Filing Date:
May 20, 2003
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
THE RESEARCH FOUNDATION OF THE STATE UNIVERSITY OF NEW YORK (35 State Street, Albany, NY, 12207, US)
Makkonen, Hannu (Sivukatu 42, FIN-87500, FI)
Cöpur, Yalcin (Büyüksehir Uydukenti, A-16-B Block. D.41 Beylikdüzü, Büyükcekmec, Istanbul Turkey, TR)
Granzow, Stephen (2331 Bremen Court, Punta Gorda, FL, 33983, US)
International Classes:
D21C3/22; (IPC1-7): D21C7/12; D21D5/02
Foreign References:
US6068732A
US4540468A
Attorney, Agent or Firm:
Brew, Michael R. (Dilworth & Barrese, LLP 333 Earle Ovington Blvd, Uniondale NY, 11553, US)
Download PDF:
Claims:
WHAT IS CLAIMED IS:
1. A method for determining the yield of pulp of a wood pulp comprising: a) initiating a pulping reaction; b) obtaining a pulp sample from the pulping reaction; c) fractionating the pulp sample to isolate a fraction of fibers having similar dimensions; d) measuring at least one dimension of the fibers in the isolated fraction; and, e) calculating the yield of the pulp from a predefined calibration curve derived from the at least one dimension of the fibers from the isolated fraction.
2. The method of Claim 1 wherein the wood pulp comprises at least one softwood.
3. The method of Claim 2 wherein the at least one softwood is selected from the group consisting of pine, spruce, hemlock, fir and conifer.
4. The method of Claim 1 wherein the predefined calibration curve is obtained by correlating fiber dimensions from a fraction of a pulp with an initial yield of the pulp calculated using gravimetric methods.
5. The method of Claim 4 wherein the yield is calculated from the predefined calibration curve for an R14 fraction of the pulp obtained by Bauer McNett classification using the formula y = 40.583x68. 128, wherein y is yield and x is arithmetic average fiber length in millimeters.
6. The method of Claim 4 wherein the yield is calculated from the predefined calibration curve for an Ri4 fraction of the pulp obtained by Bauer McNett classification using the formula y =1. 8124x + 62.99, wherein y is yield and x is the length share as a % of total fibers.
7. The method of Claim 4 wherein the yield is calculated from the predefined calibration curve for an R14 fraction of the pulp obtained by Bauer McNett classification using the formula y = 3.420x66. 732, wherein y is yield and x is average fiber width in micrometers.
8. The method of Claim 1 wherein the wood pulp comprises at least one hardwood.
9. The method of Claim 8 wherein the at least one hardwood is selected from the group consisting of maple, oak, eucalyptus, poplar, beech, birch and aspen.
10. The method of Claim 8 wherein the yield is calculated from the predefined calibration curve for an Rl4 fraction of the pulp obtained by Bauer McNett classification using the formula y =56. 2x2 + 640x1768 where y is yield and x is fiber cell wall thickness in micrometers.
11. The method of Claim 1 wherein the results of the measurements are input into a means for calculating the yield.
12. The method of Claim 11 wherein the means for calculating the yield is a computer.
13. A method for determining hygroscopicity of a pulp comprising: a) initiating a pulping reaction; b) obtaining a pulp sample from the pulping reaction; c) fractionating the pulp sample to isolate a fraction of fibers having similar dimensions ; d) measuring at least one dimension of the fibers in the isolated fraction; and, e) calculating the hygroscopicity of the pulp from a predefined calibration curve derived from the at least one dimension of the fibers from the isolated fraction.
14. The method of Claim 13 wherein the wood pulp comprises at least one softwood.
15. The method of Claim 13 wherein the at least one softwood is selected from the group consisting of pine, spruce, hemlock, fir and conifer.
16. The method of Claim 13 wherein the predefined calibration curve is obtained by correlating fiber dimensions from a fraction of a pulp with an initial hygroscopicity of the pulp calculated using gravimetric methods.
17. The method of Claim 16 wherein the hygroscopicity is calculated from the predefined calibration curve for an Rl4 fraction of the pulp obtained by Bauer McNett classification using the formula Y50=0. 2449X2 + 17.129x282. 78, wherein Y50 is moisture in a fiber at 50% relative humidity expressed as yglm and x is average fiber width in micrometers.
18. The method of Claim 13 wherein the results of the measurements are input into a means for calculating the hygroscopicity.
19. The method of Claim 18 wherein the means for calculating the hygroscopicity is a computer.
20. A process for automatic control of a continuous pulping reaction comprising, a) initiating a pulping reaction; b) obtaining a pulp sample from the pulping reaction; c) fractionating the pulp sample to isolate a fraction of fibers having similar dimensions ; d) measuring at least one dimension of the fibers in the isolated fraction; e) outputting the result of the measurement of the at least one dimension of the fibers; f) calculating the yield or hygroscopicity of the pulping reaction from a predefined calibration curve derived from the at least one dimension of the fibers from the isolated fraction; g) adding one or more replenishing components to the pulping reaction if the result of step (f) is inconsistent with a desired product; h) controlling the continuous pulping reaction in real time according to the pulp yield or hygroscopicity of the pulp obtained from the pulping reaction.
21. The method of Claim 20 wherein the one or more replenishing components are added to the pulping reaction as a feedback control measure.
22. The method of Claim 20 wherein the one or more replenishing components are added to the bleaching reaction as a feedforward control measure.
23. The method of Claim 20 wherein the step of controlling the reaction according to the calculated degree of completion comprises controlling the time duration of the reaction.
24. The method of Claim 20 wherein the step of controlling the reaction comprises controlling the temperature of the reaction.
25. The method of Claim 20 wherein the step of controlling the reaction comprises controlling the ratio by weight of the liquor to wood or pulp.
26. The method of Claim 20 wherein the replenishing components are chemicals.
27. The method of Claim 26 wherein the chemicals are selected from the group consisting of polysulfide and anthraquinone.
28. The method of Claim 20 wherein the replenishing components are enzymes.
29. The method of Claim 28 wherein the enzymes are selected from the group consisting of xylanases and aarabinofuranosidase.
Description:
METHOD TO DETERMINE PULPING YIELD CROSS REFERENCE TO RELATED APPLICATIONS This application claims the benefit under 35 U. S. C. § 119 (e) of earlier filed and copending U. S. Provisional Application No. 60/381,892, filed May 20,2002, the contents of which are incorporated by reference herein.

BACKGROUND 1. Technical Field The present invention is directed to methods for determining the yield of a chemical pulping mill from fiber characteristics. The composition of the pulp, i. e., the amount of hardwood and softwood, and their respective yields can be determined utilizing the methods of the present invention. The hygroscopicity of a pulp may also be determined using the methods of the present invention. The present invention is especially suited to on-line measurements for continuous pulping operations which are capable of accurately calculating the pulp yield and/or hygroscopicity as precisely as laboratory measurements.

2. Background of Related Art Pulp is composed of a mixture of heterogeneous fibers and cell types including more than one source, mass, and/or shape. While the most pronounced

difference is between softwood and hardwood fibers in the pulp, other differences in the fibers are based upon the species, the growth factors and genetic variations within the species, variations among the components of the trees (tracheid, ray, vessel elements etc. ), variations between heartwood and sapwood, juvenile and mature wood, early-wood and late-wood fibers, etc. , and combinations thereof.

The sequence of operations in the manufacture of chemical wood pulp includes debarking and chipping of the wood, pulping or bulk delignification of the chipped or fragmented wood, secondary delignification or bleaching, and subsequent pulp washing, screening and cleaning steps. Pulping is fundamental to the manufacture of paper, with chemical pulping processes, especially the kraft process, the dominant industrial technology. Other chemical pulping processes include biokraft, polysulfide and soda methods.

The main purpose of pulping is to degrade, dissolve and remove lignin structures. Many carbohydrates are removed during the process as a result of the complexity of the fibers'structure and the limited selectivity of the chemicals used in the process. The major change in pulping occurs in the cell wall thickness due to the dissolution of lignin and hemicelluloses from the cell walls. Consequently, pulping decreases the mass of the fiber per unit length (coarseness) in proportion to the mass of material dissolved. Chemical pulping removes cell wall material at different rates for different polymers. During a typical kraft cook, approximately 80 % of the lignin, 50 % of the hemicelluloses and 10% of the cellulose is dissolved.

Yield is a general term used in any phase of pulping, papermaking, chip screening, bleaching, etc. , indicating the mass (weight) amount of material recovered after a certain process compared to the starting amount of material before the process.

To have meaning, both samples must be compared on an oven-dry (o. d. ) wood mass.

Pulp yield is composed of a blend of the yield of the natural polymers and cell types in wood, which determine the physical and chemical characteristics of the fibers after pulping.

In chemical pulping, the term"yield"refers to the mass of moisture free pulp recovered after the cooking process as a percentage of the mass of moisture free chips charged to the digester. The total yield is counted based on the pulp fibers passing through a screening process (i. e. , screened yield), plus those remaining on the screens as rejects. More specifically, the yield refers to fibers charged as chips and recovered as pulp fibers.

A discriminating yield comparison of pulps is usually made at the same residual lignin content (kappa number), which allows an increased or diminished yield to be determined. Thus any deviations in the values are due to fluctuations in the composition of carbohydrate content (cellulose plus hemicelluloses). Although the residual lignin is determined only from the screened yield, the lignin content of the total yield is accurate where there are low reject contents, i. e. , below 5% of pulp.

Fiber dimensions may be evaluated to determine the grade of wood used in a pulp and evaluate the properties of a given pulp. The use of manual methods to measure fiber dimensions, such as by microscope, projection microscope and screen

classification, suffer from statistical imprecision, measurement difficulties and high cost which limited their use to academic work. Automated methods were developed in the 1980s and have brought these measurements into wider use with improved statistical validity, measurement ease and low cost per measurement. Automated methods have significantly facilitated decision-making in determining grade change between species (e. g. , softwood and hardwood) and in the evaluation of the strength properties of various mechanical and chemical pulps.

Fiber length is one of the most important fundamental fiber characteristics since this property affects both the paper machine operation and the finished product's quality. Most meaningful definitions for average fiber length such as arithmetic, length-weighed, and weight-weighed are used for different purposes.

Arithmetic average fiber length (Ln) is computed by dividing the total length of all fibers present over a large number of fibers and is set forth below as Equation 1.

Ln =E 1/n Eq. 1 where: Ln = arithmetic average fiber length; 1 = individual fiber length; and n= number of fibers.

Particles or pieces of fibers with lengths less than 0.1 or 0.2 mm are excluded from fiber counts and expressed as"fines, "i. e. number of small particles, which are expressed as a percentage of the total number of particles. The determination of the arithmetic average fiber length of pulps is a good measurement if all the fibers are included in the count and they are fairly uniform in length. A length-weighed average fiber length (L) reduces the impact of the shortest particles and can be more

useful to reflect the capabilities of the fibers, especially their reinforcement characteristics, and is set forth below in Equation 2.

L, =E (n 12)/E (n 1) Eq. 2 where: L, = length weighted average fiber length; 1 = individual fiber length; and n= number of fibers.

Fiber coarseness is another important parameter of a pulp. It correlates with the number of fibers in the sample, the strength of individual fibers and the number of bonding sites. Fiber coarseness (C) is defined as the dry fiber mass (w) per unit length (1) and is obtained by dividing the mass of the pulp by the total measured length of the fibers as set forth in Equation 3.

C =w/Fc Ln Eq. 3 where: w = an amount of oven dry mass of the fibers; Fc = the total number of fibers in the mass; and Ln = the arithmetic average fiber length.

Accurate coarseness determination, therefore, is possible only with accurate measurement of pulp mass, length and number of fibers. Modern optical fiber length analyzers can provide statistically accurate measurement of length and fiber counts, but the mass of the fibers can still be a considerable source of error.

Fiber grammage (Fg) results from dividing the coarseness (C) by the fiber width (W) and may be calculated using the following equation:

Fg= C/W Eq. 4 where: C = the fiber coarseness; and W =the fiber width.

The grammage values of fibers range from about 3 to about 10 g/m2. Paper has a layered structure and more layers provide the opportunity for better formation.

This is more evident in low grammage papers, which have few fibers per unit area to provide physical properties. The grammage of the fibers is a key parameter since the number of layers in paper is determined by dividing the sheet grammage by the fiber grammage.

The pulping yield depends to some extent on the length and mass of the fibers (i. e. , coarseness), since the length of the fibers remains almost constant during the chemical pulping and only the mass (weight) varies. An early yield study compared the coarseness of the wood fibers with commercial pulps and observed that there appeared to be a reduction in coarseness in proportion to the pulp yield. Britt, "Determination of Fiber Coarseness in Wood Samples", Tappi, 48 (1) : 7 (1965).

Another study established a linear relationship between pulp yield and fiber coarseness within the yield range of 45-95 %, but the results were not very accurate.

Scallan et al. "The Effect of Pulping upon the Dimensions of Wood Tracheids", Wood and Fiber, 7 (3): 226-233 (1975).

Other studies on the effects on cross-dimensional properties and coarseness of the pulps for the kraft cooking method have shown that fiber coarseness decreased in proportion to yield loss as organic material was removed from the cell wall, and a linear correlation was observed between pulp yield and coarseness even if the

cooking conditions were varied. Paavilainen, L. ,"Importance of Cross-Dimensional Fiber Properties and Coarseness for the Characterization of Softwood Sulphate Pulps", Paperi Puu 75 (5): 343-351 (1993).

Pulp yield can be increased by retaining more hemicelluloses in the fibers.

Hemicellulose retention may be improved by the application of chemicals stabilizing the hemicelluloses against endwise depolymerization reactions. Most notably, polysulfides and/or anthraquinone are used for this purpose.

Frequently, yield is determined subsequent to the pulping operation. The gravimetric determination of yield is a common practice in the laboratory, though it is very labor-intensive and much less accurate in the commercial scale. This is especially true with batch digesters and continuous digesters. The information of yield is important in the sound, economical operation of a pulp mill. The yield information also reveals the papermaking characteristics of the pulps, including their behavior in changing humidity conditions.

Some methods have been developed to calculate yield from on-line measurements. For example, U. S. Patent No. 4,540, 468 discloses gravimetric methods for monitoring the degree of completion and pulp yield from the process liquor in a chemical pulping process. These methods may be used for on-line monitoring of pulping reactions.

Similarly, U. S. Patent No. 5,970, 783 discloses a method and apparatus for determining the extent of liquid absorbency of pulp wood chips, which may be utilized in calculating yield.

Even with the automated methods available for determining fiber characteristics, there is a major gap in available measurements with respect to the determination of pulping yield. Although gravimetric methods permit the determination of yield of some pulps directly from the fibers by the application of a combination of measurements, e. g. , coarseness, as noted above this method usually takes place after the pulping operation, applies only for a single cooking method (e. g. kraft), and does not allow a yield comparison between different pulping methods.

The prior methods are also limited in their ability to come up with meaningful results where multiple sources of raw material are used. For example, if fibers originate from trees harvested from different growth sites, fiber coarseness may provide little information about the pulping yield, even if a single cooking method is used.

It would be desirable to develop an improved method for determining pulp yield estimates from measurements associated with pulp attributes, especially for use in the on-line control of a continuous pulping process.

SUMMARY OF THE INVENTION The present invention is directed to methods for determining the yield of pulp of a wood pulp by a) initiating a pulping reaction; b) obtaining a pulp sample from the pulping reaction; c) fractionating the pulp sample to isolate a fraction of fibers having

similar dimensions; d) measuring at least one dimension of the fibers in the isolated fraction; and, e) calculating the yield of the pulp from a predefined calibration curve derived from the at least one dimension of the fibers from the isolated fraction.

Preferably, the calibration curves can be used to develop specific formulae for calculating yield based upon fiber length, width, or cell wall thickness.

The same fiber dimensions utilized to calculate yield may also be used to determine hygroscopicity of a pulp. In such a case, the method comprises a) initiating a pulping reaction; b) obtaining a pulp sample from the pulping reaction; c) fractionating the pulp sample to isolate a fraction of fibers having similar dimensions ; d) measuring at least one dimension of the fibers in the isolated fraction; and, e) calculating the hygroscopicity of the pulp from a predefined calibration curve derived from the at least one dimension of the fibers from the isolated fraction.

The present invention may be utilized to monitor pulping operations on-line and automate their control. The process for automatic control of a continuous pulping reaction comprises, a) initiating a pulping reaction;

b) obtaining a pulp sample from the pulping reaction; c) fractionating the pulp sample to isolate a fraction of fibers having similar dimensions ; d) measuring at least one dimension of the fibers in the isolated fraction; e) outputting the result of the measurement of the at least one dimension of the fibers; f) calculating the yield or hygroscopicity of the pulping reaction from a predefined calibration curve derived from the at least one dimension of the fibers from the isolated fraction; g) adding one or more replenishing components to the pulping reaction if the result of step (f) is inconsistent with a desired product; h) controlling the continuous pulping reaction in real time according to the pulp yield or hygroscopicity of the pulp obtained from the pulping reaction.

In one embodiment, replenishing components including chemicals are added to the pulping reaction as a feedback control measure. In another embodiment, replenishing components including enzymes are added to the bleaching reaction as a feed-forward control measure.

BRIEF DESCRIPTION OF THE FIGURES FIG. 1 is a graph of the chemical composition of the pulp yields from the pulps prepared in Example 1.

FIG. 2 is a graph depicting pine pulp yield vs. the yield of Bauer McNett

fraction Rlo.

FIG. 3 is a graph depicting pine pulp yield vs. the yield of Bauer McNett fraction Rua.

FIG. 4 is a graph depicting the normalized fraction yields of pine pulps prepared in Example 1 after being fractionated by a Bauer McNett classifier. The fractions are of Rio, Rl4, and R28. P28 represents pulps passed through screen R28.

FIG. 5 is a graph of the yield versus the arithmetic average fiber length for fraction Rl4.

FIG. 6 is a graph depicting the correlation coefficients for pulp yield at various fiber length population classes of fraction Rl4.

FIG. 7 is a graph depicting the relationship between population distribution and fiber length for the beginning of the distribution slopes.

FIG. 8 is a calibration curve demonstrating the relationship between yield and arithmetic fiber length of fibers obtained from fraction Rl4.

FIG. 9 is a graphical depiction of the relationship between length share and pulp yield at fiber length class <2.16 mm in fraction Rl4.

FIG. 10 is a graphical depiction of the relationship between pulp yield and fiber width on screen R) 4.

FIG. 11 is a graph of the relationship between observed yield and fiber width values of the Rl4 fraction.

FIG. 12 is a graph depicting the dependence of pulping yield on the fiber cell wall thickness of R2oo Bauer McNett fraction of maple pulps.

FIG. 13 is a graphical depiction of the predicted equilibrium moisture content of pulp fibers at different relative humidities based on fiber width of the R, 4 fraction.

DESCRIPTION OF PREFERRED EMBODIMENTS In accordance with the present invention, a method is provided to determine the yield of chemical pulping with improved accuracy and with less time and effort.

The procedure calls for selective isolation of one or more fractions of fibers, from which one or multiple dimensional characteristics of the fiber are measured.

Calibration curves of the fiber dimensions and yield are prepared from which formulae may be derived and utilized to calculate yield based upon the fiber dimensions. These methods are especially suited for on-line monitoring systems for continuous pulping operations which are capable of examining the dimensions of the fibers to calculate the yield and determine the composition of the pulp.

Any pulping process can be monitored in accordance with the methods of the present invention. Preferably, the pulping process is a chemical pulping process including kraft, biokraft, soda, or polysulfide processes. Most preferably, the pulping process is a kraft process. While the pulping process may be batch or semi- continuous, preferably the pulping process is a continuous pulping process.

A wide variety of cellulosic fibers can be employed in the process of the present invention. It is also possible to use mixtures of one or more cellulosic fibers.

Preferably, the cellulosic fiber used is from a wood source. Suitable wood sources include softwood sources such as pines, spruces, hemlocks, firs, conifers, etc. , and

hardwood sources such as maples, oaks, eucalyptuses, poplars, beeches, birches, aspens, etc.

In accordance with the methods of the present invention, a sample from a pulping process is obtained by means of suitable equipment and techniques. Once a sample of a pulp has been obtained, it is subjected to both traditional gravimetric methods to initially determine yield and fractionation.

Traditional gravimetric methods for calculating yield are known to those skilled in the art and are based upon the mass of moisture free pulp recovered after the cooking process as a percentage of the mass of moisture free chips charged to the digester. In accordance with the methods of the present invention, after the yield of an initial pulp has been determined gravimetrically, additional gravimetric calculations of yield for later pulps are not required.

As used herein, the term"fractionation"and variations thereof are generally meant to refer to the separation of a mixture into separate components and, more particularly, to the separation of a cellulosic fiber mixture into separate cellulosic fiber fractions. By fractionating a pulp sample, one can obtain homogeneous fibers having similar shapes and lengths from which fiber dimensions can be calculated and averaged, which may then be used to calculate yield. Fractionation techniques preferably take advantage of the flexibility and flow properties of the fibers that are most representative for yield determination. Methods for fractioning fibers are known to those skilled in the art and include the Clark method, the Bauer-McNett

method, etc. In a preferred embodiment, Bauer McNett fractionator-type screens are utilized to fractionate the cellulosic fibers.

The Bauer McNett fractionator was developed to quantitatively estimate the weight-weighed average fiber length of a pulp sample using screen classification of a sample into different length fractions. The procedure separates the fibers into length classes by exposing a relatively dilute suspension of fibers to four or more wire-mesh screens in sequential compartments with each screen having successively smaller openings. Water flow at a fixed rate and time carries fibers shorter than the screen mesh thorough to the next screen. Fibers that accumulate in each compartment are collected into a fine cloth, dried, and weighed. The Bauer McNett method provides a rough estimate of fiber length, especially when chemical pulps are concerned.

The types of fractionation devices and/or fiber fractions can be adjusted for each mill/raw material application. Thus, in accordance with the present invention only fiber properties from fractions of a pulp sample which possess similar shape, length and/or flexibility characteristics are measured.

Once the pulp sample has been fractionated, desired fiber dimensions are then measured. While manual methods could be utilized to measure these dimensions, the dimensions are preferably measured using commercially available fiber analyzers which can rapidly and accurately measure the fiber characteristics and are more efficient for use in on-line monitoring systems, especially for use in continuous pulping operations. The fibers of the fraction are optically measured to reveal their physical dimensions, most notably the lengths, widths and/or cell wall thicknesses at

each measurement channel.

Examples of suitable equipment for measuring the fiber characteristics/ dimensions include laboratory analyzers such as Kajaani FS-100 fiber length analyzer, Kajaani FS-200, Kajaani FiberLab; as well as online analyzers such as FSA, all commercially available from Metso Automation (Kajaani, Finland). These analyzers take advantage of simple statistical methods to provide data and estimate individual measurement resolution. The analyzers distribute the results to length class categories to create simple population distribution data, from which the distribution curves, the arithmetic length, the length-weighted average length, and the weight-weighted average length of the fibers may be computed. Fibers can be measured by straightening them under capillary vacuum for exposure to a light beam, e. g. , a laser, and a detector to determine different fiber length (FS-100 and FS-200) or by using image analysis to measure the contoured length of both curled and straight fibers (FiberLab). As a result of more sophisticated data handling, the FiberLab is also able to measure the width of the fibers and provide the average width of the fibers.

Once the fiber dimensions of the fractions are obtained, a calibration curve can be constructed. In order to construct the calibration curve, the initial yield calculated using traditional gravimetric methods is compared with fiber dimensions from certain fractions of the pulp and statistical (regression) analyses conducted using methods known to those skilled in the art to determine the relationship between the fiber dimension and yield. In this manner a predefined calibration curve is produced,

which may then be utilized to predict the yield of later pulps produced by the pulping operation on the basis of the fiber dimension alone. Formulae may also be derived from the calibration curve by which the yield of the pulp may be calculated on the basis of the fiber dimension.

Preferably, each wood combination in the pulp is subject to calibration procedures which then permit the development of formulae to calculate yield. For softwoods, such as pine, spruce, hemlock, fir, conifer, etc. , the fiber dimension to calculate yield is preferably the fiber width or fiber length. For hardwoods such as maple, oak, eucalyptus, poplar, beech, birch, aspen, etc. , cell wall thickness is preferably used as the dimension to calculate yield.

In one embodiment of the present invention, the predefined calibration curves to determine the yield of a softwood are developed from the R,4 fraction of a softwood pulp or a mixed softwood/hardwood pulp fractionated utilizing a Bauer McNett classifier, and the fiber dimensions are obtained using a Kajani FS-100.

Suitable equations which may be utilized to calculate yield are as follows: y = 40. 583x-68.128 EQ. 5 where y = Yield as % of wood; x = arithmetic average length of the fibers in millimeters; or y =-1. 8124x + 62. 99 EQ. 6 where y = yield as % of wood; x = the length share as a % of total fibers.

In another embodiment, where a FiberLab instrument is utilized to analyze fiber dimensions, the following formulae have been developed to calculate yield: y= 3. 420x-66.732 EQ. 7 where y = Yield as % of wood; x = fiber width in micrometers; or y= 49. 9x-73.8 EQ. 8 where y = Yield as % of wood; x = arithmetic average length of the fibers in millimeters.

Similarly, yield of a hardwood from the R200 fraction of a hardwood pulp or a mixed softwood/hardwood pulp fractionated utilizing a Bauer McNett classifier and having fiber dimensions determined by a FiberLab analyzer may be calculated using the following equation derived from calibration curves: y=-56. 2x2 + 640x-1768 EQ. 9 where y = Yield as % of wood; x = fiber cell wall thickness in micrometers.

Once the predefined calibration curve has been produced, only the fiber dimension is needed to further calculate yield of later pulps from the pulping operation.

Thus, utilizing the above methods and the predefined calibration curve the yield of a softwood pulp may be calculated on the basis of the fiber width or fiber length of a given pulp fraction, and for a hardwood pulp, yield may be calculated on the basis of cell wall thickness. For a mixed pulp possessing both softwood and

hardwood, the yield of the softwood component of the pulp may be calculated using the above equations with respect to width and/or length of fractions comprising softwood fibers, and the yield of the hardwood component of the pulp may be calculated using the above equation with respect to cell wall thickness of fractions comprising hardwood fibers.

The inventors have discovered that the methods of the present invention may be utilized to not only calculate yield, but also to calculate hygroscopicity of pulp fibers made from a variety of hemicellulose-rich pulps, which may be utilized to regulate fiber/paper hygroscopicity to improve the dimensional stability of papers produced from the pulp. As with yield, the hygroscopicity of an initial pulp may be determined with traditional gravimetric methods for measuring water adsorption and fiber dimensions may be obtained as described above. Once obtained, the hygroscopicity of the initial pulp is compared with fiber dimensions from certain fractions of the pulp and statistical (regression) analyses conducted using methods known to those skilled in the art to determine the relationship between the fiber dimension and hygroscopicity. In this manner, a predefined calibration curve is produced which may then be utilized to predict hygroscopicity of pulps produced by the pulping operation on the basis of the fiber dimension alone. Formulae may also be derived by which the hygroscopicity of the pulp may be calculated on the basis of the fiber dimension.

In one embodiment, the sample is fractionated with a Bauer McNett classifier, hygroscopicity is determined by measuring water adsorption of the fibers at a relative

humidity of 50%, and fiber dimensions are obtained using a FiberLab instrument. In such a case, the calibration curve constructed from such data provides the following equation for the calculation of hygroscopicity : Y5o =-0. 2449x2 + 17.129x-282. 78 EQ. 10 where Yso is the moisture in fiber as 4g/m ; and x is the fiber width in micrometers.

The methods of the present invention thus provide a tool to monitor the pulping yield of a pulping process, preferably a continuous pulping process, to avoid yield losses; to pulp to a higher yield (smaller wood consumption); and to control and reduce the inadvertent hygroscopicity of paper with smaller yield loss than current technology. This new technique may also be utilized to improve the dimensional stability of printing and writing papers, especially if the pulping yield is increased by the addition of chemicals.

The methods of the present disclosure are especially suited for an on-line monitoring system of a continuous pulping operation, which is much more efficient than existing methods which rely upon gravimetric principles and elaborate data procuring methods with respect to a variety of factors ranging from the wood yard to production management. Once the predefined calibration curve has been created, the calibration curve and/or formulae derived therefrom may be utilized to more rapidly calculate yield and or hygroscopicity on the basis of fiber dimensions alone, without requiring additional gravimetric experiments and data procuring methods to determine yield of later pulps.

Thus, the methods of the present invention make it possible to check the yield and hygroscopicity of a pulp from a pulping process and initiate specific corrective measures. The corrective measures may be implemented manually or automatically, depending on whether or not the control system is configured for automatic adjustment of pulping conditions. The corrective measures may be both feedback controls, such as adjusting the chemical additives to the pulp, or feed-forward controls, such as adjusting enzymatic treatments during the bleaching process.

In one embodiment, the methods of the present invention may be employed in a continuous pulping process to monitor the yield and hygroscopicity of the pulp on- line in real time. Using these determinations, which are based upon fiber dimensions alone, the use of chips and chemicals can be modified to enhance yield and hygroscopicity as additional pulp is made in the continuous process. Where the yield calculation is made on a computer, a warning signal or message can be produced if it appears the yield or hygroscopicity will not be consistent with the desired product.

By this method it is possible to prevent the resulting paper from having poor qualities or undesired characteristics. The warning signal may be transmitted to a remote location, so that the appropriate corrective measures may be initiated from that location.

The determinations regarding yield and hygroscopicity may be transmitted continuously, at a specified time interval and/or on request. Depending on the calculations of yield and hygroscopicity, necessary corrective steps may be taken

either automatically via a process-guidance system or as a result of manual intervention.

In this way, control of the pulping process on-line in real time may be achieved. Such control is achieved, for example, by controlling the duration, temperature or liquor to wood ratio of the pulping or delignification reaction. This may be effected, for example, with the aid of metering pumps for the addition of components of the liquor, in a weight controlled manner for the addition of chips, or temperature may be adjusted with temperature controllers.

In addition, the methods of the present invention may also be used to control the cooking variables of the pulping process, most notably the additives used to increase the pulping yield, and to predict the moisture behavior of the paper manufactured from such pulp and to control the charge of chemicals and/or enzymes to the pulping reaction which are designed to remove the hemicellulose components causing the unwanted moisture behavior of paper. Such controls include feedback control measures and feed-forward control measures.

The addition of chemicals to the pulp can be utilized as a feedback control measure. Suitable chemicals which may be added to the pulp include, but are not limited to, polysulfides and/or anthraquinones, which can effect the resulting yield of a pulp.

As a forward control tool the methods of the present invention allow one to predict the paper functionality in terms of hygroscopicity and dimensional stability, and provide means to adjust these properties when necessary by the application of

enzymatic treatments in the bleaching processes. Specific enzymatic treatments may be added to the bleaching reaction to make the fibers less hygroscopic while maintaining the superior yield and quality of the fibers. Suitable enzymatic treatments include, but are not limited to, the use of xylanases and a- arabinofuranosidases, which may be utilized to minimize yield loss depending upon the results of the calculations.

These treatments not only affect the composition of the pulp, but they also affect the characteristics of any paper produced from the pulp. These adjustments may occur automatically if the yield or hygroscopicity of the pulp is outside the acceptable limits for the desired product, or may take place in response to an external request from a remote location regardless of the current determination of pulp yield or hygroscopicity.

In one embodiment, the methods of the present invention include a control system, preferably a computer control system, which is able to receive the fiber dimension measurements from a pulp sample, calculate yield and hygroscopicity, and thus monitor the pulp on-line in real time. The control system is preferably connected with additional components of the pulping system, such as temperature controllers, metering devices for the addition of components of the liquor, metering devices for the addition of chips, chemicals, enzymes, etc. , so that the computer may automatically adjust the conditions of the pulp if the yield or hygroscopicity is determined from the pulp sample to be outside the acceptable limits of the desired product. In this manner, once the desired yield and hygroscopicity for a given pulp

has been entered into the control system, the control system may monitor the dimensions of fibers from fractions of the pulp and make adjustments to the pulp in order to ensure maximum efficiency of the wood utilized therein and adjust the pulping conditions to the extent necessary to ensure the pulp will have the necessary characteristics for the desired paper product.

Deviations from the desired values may be detected at an early stage and corrected before the pulp is impaired to a point that it cannot be utilized for the desired product. The measured data may also be transmitted to a remote location thereby providing operating or supervisory personnel with constant information on the state of the pulping operation when not in the immediate vicinity of the digester.

The effort expended by personnel on checking and regulating pulp conditions may be considerably reduced by this means and the efficiency of wood consumption and chemicals can be both documented and enhanced. By virtue of documentation of the data collected in accordance with the methods of the present invention, quality assurance is assured.

Once installed in an industrial setting, the methods of the present provide instant data for better control of a continuous pulping operation and allow for more reliability in the determination and adjustment of pulp yield on a mill scale. The enhanced process control which results from the methods of the present invention allow one to better define the quality of the pulp, provide better estimation of raw material consumption and thus save raw wood materials, enhance product uniformity, and lessen the environmental impact found with chemical pulping operations.

EXAMPLES The examples below include comparisons of fiber dimensions and yield or hygroscopicity and statistical analyses thereof. All statistical analyses were based on simple regression analyses, the tools for which are in Microsoft Office Excel. First order statistics were used for linear relationships and second order statistics for less obvious relationships. Measured fiber dimensions were used in regression equations and the calculated yield values were obtained and compared with experimentally observed yield values. The difference between calculated and measured yield values was reported as percent of mean error.

A Microsoft office EXCEL program with Data Analysis Tools was used to study the relationships between data sets and the Correlation Tool of EXCEL was used to quantify how the two ranges of data move together. When large values of one set (e. g. , fiber dimensions) were associated with large values of the other set (e. g., yield), a positive correlation existed, and when small values of one set were associated with large values of the other, a negative correlation existed. If values in both sets were unrelated, the correlation neared zero, and if the variables were slightly related, they had a low correlation.

A correlation matrix revealed whether fiber dimensions correlated to the pulping yield. Charts were drawn between yield and each fiber dimension, which correlated significantly to pulp yield, to visually show the relationship between the two and make it easier to compare patterns and trends. Trendlines were also used to graphically display trends in data and to analyze the predictions.

The trend line equations were validated using a cross validation method. With cross validation, one observation was excluded at a time while estimating regression coefficients, which were then used to predict the excluded data point. This procedure was repeated for all data points, so that estimated values could be compared with real values.

EXAMPLE 1 Scots Pine and Sugar Maple were pulped with four different yield levels using the methods leading from the highest to the lowest yields. These methods were polysulfide (P), kraft (K), biokraft (B), and soda (S).

For each pulp, 400 g of pine and 500 g of maple oven dry chips (o. d. ) were cooked in a 4.7 L M/K-digester at 170 °C and 165 °C for 90 minutes. After reaching the temperature in 90 minutes, the pulps had H-factor of 1,885 and 1,532, respectively. (H-factor is a combination of time and temperature as a single factor (see Vroom, K. E. "The"H"-factor : A means of expressing cooking time and temperature as a single variable"Pulp & Paper Mag. Can. 58 (3): 228-231 (1957). ) As set forth below in Table 1, the cooks were made at two different alkali charges to reach two different kappa numbers; 25-35 for pine, and 15-25 for maple. Four different levels of hemicellulose retention were achieved by using polysulfide, kraft, biokraft, and soda methods. All but soda cooks were done at 30 % sulfidity liquor.

The liquor-to-wood ratio (L/kg) was 4: 1, except during the second pine soda cook (S2), in which it was 5: 1. For the biokraft trials, the wood chips were treated with Cerioporiopsis subvermispora fungi for 14 days at 27 °C and followed by a typical kraft cook. Polysulfide cooks were made by adding 4 % elemental sulfur based on oven dry chip weight to the digester along with the chips.

Table 1 Cooking conditions for pine and maple pulps Pulps Active alkali, % Na20/weight of wood (o. d) Method Sample Pine Maple Polysulfide P1 21 20 P2 24 22 Kraft K1 20 16 K2 23 18 Biokraft B 1 20 16 B2 23 18 Soda S 1 40 27. 5 S2 50 40

Active alkali is composed of sodium hydroxide, NaOH and sodium sulfide Na2S, both of which are expressed in the calculations as sodium oxide, Na2O. The percentage of Na2S as Na20 from the active alkali is"sulfidity".

The pulps were disintegrated, washed with hot tap water (-40 °C) and then screened using a laboratory flat screen with a slot width of. 006" (0.15 mm). The yield contents of the pulps and rejects were determined according to TAPPI Method T-210 by gravimetric measurements in the laboratory environment. The screened yield, Ys was determined according to Equation 11 using triple replicates. Adding the yield of rejects to the screened yield gave total pulp yield, Yt (Equation 12). The kappa number (TAPPI Method T-236 cm-85) and viscosity (TAPPI Method T-230 om-89) of the pulp samples were also determined by double or triple experiments.

The approximate Klason lignin (the total lignin amount present in cell walls) was calculated using Equation 13 (TAPPI Method T-236).

Ys = Psw/Cw * 100 Equation 11 ; Yt = P + R / Cw * 100 Equation 12; Klason lignin = K * 0.15 Equation 13, where Ys is percent of screened yield, Psw is oven dry weight of screened pulp after pulping, Cw is the oven dry weight of chips into the pulping process. Yt is percent of the total yield, P is the percentage of oven dry pulp weight obtained after screening, R is the oven dry weight of rejects, K is the kappa number.

Table 2 below and Figure 1 present the yield for these pulps.

Table 2 Yield data for pulp Method Kappa'Yield" (mL/g) (%) Pine Polysulfide P1 35. 3 48.4 P2 27. 1 45.4 Kraft Kl 34.4 45.2 K2 24.3 42.3 Biokraft Bol 37. 8 43.7 B2 25.4 40. 3 Soda S133. 2 39.1 S2 24. 9 36.5 Maple Polysulfide P1 15. 9 53. 2 P2 14. 8 51. 9 Kraft K1 23.8 52. 4 K2 18.4 50.6 Biokraft Bd 20. 6 49.5 B2 16.7 48.0 Soda si 19. 9 45.4 S2 12. 5 41. 4

a This observation comes from average of two measurements b These observations come from a single experiment; o. d. = oven dry.

As can be seen from the above table and Figure 1, there were 4 higher and 4 lower kappa number pulps. The percentage of cellulose, glucose (assumed as 1/3 of mannose; total glucose is the sum of cellulose and mannose derived glucose), galactose, mannose, xylose, arabinose and lignin of pine pulps is shown. Yields were expressed as % from (unextracted) wood. The polysulfide method gave the highest yield due to the retention of hemicelluloses. The biokraft and the soda methods gave considerably lower yields due to the destruction of hemicelluloses and cellulose.

EXAMPLE 2 The Scots pine and Sugar maple pulps prepared in accordance with Example 1 were fractionated with a modified 8-bay Bauer McNett apparatus (two Bauer McNett classifiers combined together), by using the following: 10,14, 28,35, 48,65, and 200 mesh screens (Tappi T233), designated as Rlo, Ro4, R28, etc. P200 indicates pulps passing through screen R200. The gravimetric fiber retentions, determined as in Example 1 above, at each screen are set forth below in Table 3, where P denotes polysulfide method; K denotes kraft; B denotes biokraft; and S denotes soda methods.

For each method, pulps were targeted at two different residual lignin contents (i. e., kappa numbers). Table 3 includes the kappa and yield for these pulps as determined in Example 1 and set forth in Table 2 above.

Table 3 Yield data for Bauer McNett fractions Method Kappa Yieldb Yield of pulps at Bauer McNett fractions, (mL/g) (%) % of o. d. pulp)'' Rio R14 R28 R35 R48 R65 R200 P200 Pine Polysulfide Pl 35. 3 48. 4 41. 9 30. 0 18. 0 4. 0 2. 2 2. 3 1. 8 0. 0 P2 27.1 45.4 45.3 27.7 17.2 3.8 2.4 2.1 1.8 0.0 Kraft K1 34.4 45.2 39.5 30.0 19.2 4.4 2.7 1.9 1.9 0.5 K2 24.3 42.3 45.4 26.0 17.7 4.2 2.5 2.1 1.9 0.2 Biokraft B1 37.8 43.7 46.4 26.7 17.2 3.7 2.8 1.3 1.8 0. 2 B2 25.4 40.3 48.2 25.0 15.4 3.2 2.1 1. 1 4.3 0.2 Soda S1 33.2 39.1 55.5 19.5 14.1 3.4 2.7 1.2 1.8 1.8 S2 24.9 36.5 58.4 18.3 13.9 3.4 2. 7 1. 3 1.9 0. 1 Maple Polysulfide P1 15. 9 53.2 0.07 0. 07 0. 51 39. 0 24. 1 13. 4 13.1 9. 73 P2 14.8 51.9 0. 11 0.13 0.29 41. 7 25. 0 14.0 14.2 4. 50 Kraft K1 23.8 52.4 0.05 0.22 4.46 42.4 24.2 12.5 11.9 4. 33 K2 18.4 50.6 0.00 0.00 0.46 44.1 24.4 13.5 13.0 4. 51 Biokraft B1 20.6 49. 5 0.00 0.21 2. 05 41.0 25.1 12.9 13.8 5. 05 B2 16.7 48. 0 0.00 0.00 0.35 43.6 24.1 12.9 14.0 5. 12 Soda S1 19.9 45.4 0.00 0.10 1.03 49.0 23.6 11.8 11.9 2. 65 S2 12.5 41. 4 0.00 0.00 0.12 44.8 24.2 12.8 13. 3 4. 81 Mesh, mm 1. 68 1.19. 595. 42. 297. 21. 074-

a This observation comes from average of two measurements (from Example 1) b These observations come from a single experiment; o. d. = oven dry (from Example 1).

As expected, the pine pulps were mostly composed of long fibers, with almost 90 % of the fibers retained on screens Rio to R28. Remarkably, screen Rio retained the highest percentage of fibers in the soda pulps (55-58 % of o. d. pulp), which indicates that soda fibers were longer and/or stiffer.

Most of the maple pulp fibers were retained on Bauer McNett screens R35 to R200, whereas the rest, less than 5 %, were retained on screens Rio to R28. Maple kraft

(K1) and biokraft (B1) pulps contained fiber bundles at kappa numbers of 34.4 and 37.8, and were retained in screens Rio to R28, respectively. Kraft pulp (Kl) gave twice the fiber bundles compared to biokraft pulp (B1), although cooked under the same conditions. As was the case with the pine soda pulps, maple soda pulps also (S 1 and S2 in Table 3) contained the highest percentage of fibers (45-49 %) on the first screen to collect a significant amount, R35.

EXAMPLE 3 The relationship between pine fiber fraction yield on Bauer McNett screens and pulping yield was studied. The fiber fractionation yields from Example 2 were compared with the yield as calculated in accordance with Example 1.

Fiber fractionation yields on screens Rio and Ri4 had the highest relationship to pulping yield (Yield/Yield (Rlo) ; R2(Pine) = 0. 810, Yield/Yield (Rl4) ; Repine) = 0. 880), the results of which are summarized in Table 4 below and in Figures 2 and 3, respectively.

Table 4 Summary of coefficients of determinations between pulp yield and fiber retentions at each Bauer McNett screen for pine pulps R2 Yield Pine Maple 0. 810- R. t4 0. 880 0.752 - 35 - 0. 476 R48 - 0. 259 a R10-200 = yield of pulps retained on screen RIO, to R200

The wood-based yields compared to lignin content of the fractions are set forth in Figure 4 as Rio, Rl4, R28, with P28 representing pulps passed through screen R28. The yield equals the per cent fraction times per cent yield (e. g. , for pine P1 : Yield (48.4%) *Rlo yield (41. 9%) =20. 3%; lignin = Klason lignin; See Equation 13 above).

When Rio yields were normalized to wood (Rl0wt*Yieldll00), all the normalized RIO fractions of the pulps represented yields of around 20 % of wood.

This means that in lower yield pulps the fiber count on the first screen was much more than for higher yield pulps. The yield increase for the higher yield pulps must, therefore, come from fractions caught on the finer screens.

It was found that the roughest Bauer McNett screen, Rio, retained proportionally more lower-yield fibers than higher-yield fibers; this result is readily apparent in Figure 2. For the next screen, R14, this relationship was reversed, as is readily apparent in Figure 3.

As seen in Figure 4, with Scots pine the fraction R14 was most representative regarding both the yield and the fiber properties and thus most related to the actual pulping yield.

As can be seen from Figure 4, fiber fractionation yields on screens Rio and Ri4 had the highest relationship to pulping yield. The correlation between pulp yield and fiber retention of fraction RIO was negative, while it was positive for Rl4.

EXAMPLE 4 A Kajaani FS-100 fiber length analyzer (Metso Automation, Kajaani, Finland) was used to measure the fiber dimensions from the Bauer McNett pine and maple fractions of Example 2. Coarseness was determined by dividing the pulp mass (o. d.) by the total measured length of the fibers (see Eq. 3 above). These results are set forth below in Tables 5-11 (for all tables, the values in parentheses are error limits at 95% level).

Table 5 Fiber dimensions of pine pulps on Bauer McNett fractions Method Arithmetic average fiber lenth, (mm) R10 R14 R28 R35 R48 R65 R200 Polysulfide PI 3.37 2.83 2.01 1.32 0.938 0.730 0.367 (0. 03) ~0 02) (0. 04) (0. 008) (0. 008) (0. 007) (0. 005) P2 3. 32 2.81 1. 99 1.23 0.741 0.577 0. 335 (~0. 01) (0. 03) (0. 04) (0. 01) (0. 009) (~0. 006) ~0. 004) Kraft K1 3.42 2.77 1.97 1.25 0.807 0.674 0.332 ~0.09) (~0.03) (~0.03) (~0.01) (~0.008) (~0.005) (~0.003 K2 3.24 2.72 1.90 1.19 0.775 0.658 0.348 (0. 09) (0. 02) (0. 03) (0. 005) (0. 007) (0. 007) (0. 001) Biokraft B1 3.30 2.81 2.02 1.34 0.981 0.690 0.367 ~0. 1) (0. 03) (0. 02) (0. 02) (~0. 01) (0. 01) (0. 004) B2 3.29 2.67 1.88 1.22 0.933 0.586 0.27 ~0. 04) ~0. 03) ~0.02) (~0.01) (~0.02) (~0.01) (~0.02 Soda S1 3.19 2.63 1.92 1.29 0.889 0.612 0.345 (~0. 06) (~0.03) (~0.02) (~0.02) (~0.005) (~0.004) (~0.004 S2 3.16 2.59 1.91 1.26 0.923 0.642 0.353 (0. 07) (~0. 03) (0. 01) (0. 02) (0. 01) (~0. 005) (~0. 007) Table 6 Fiber dimensions of pine pulps on Bauer McNett fractions Method Length-weighed average fiber length, (mm) R10 R14 R28 R35 R48 R65 R200 Polysulfide P1 3. 78 3. 15 232 1. 54 1. 11 0.869 0.508 (0. 06 (~0. 04 ~0. 06) (~0. 01) 0. 01) ~0.01) (~0.006 P2 3.68 3.12 2.28 1.44 0.926 0.704 0.474 ~0. 06) ~0. 05) (0. 02) ~0. 03) ~0. 009) ~0. 003) ~0. 006) Kraft K1 3.94 3.13 2.30 1.45 1.02 0.807 0.479 ~0. 08) ~0. 05) ~0. 05) ~0. 02) ~0. 02) ~0. 004) ~0. 001) K2 3.61 3.04 2.19 1. 38 0.974 0.788 0.483 ~0.02) (~0.02) (~0.04) (~0.01) (~0.01) (~0.005) (~0.004 Biokraft B 1 3.63 3.16 2.43 1.67 1.27 0.923 0.565 ~0. 08) ~0. 05) (0. 04) ~0. 04) ~0. 04) ~0. 05) (0. 02) B2 3.66 3.05 2.23 1.52 1.22 0.840 0.477 ~0. 02) (0. 03) ~0. 03) ~0. 02) ~0. 05) ~0. 03) ~0. 02) Soda S1 3.53 2.96 2.24 1.54 1.06 0.736 0.493 ~0. 05) ~0. 07) ~0. 04) ~0. 04 0. 008) ~0. 01) ~0. 01) S2 3.48 2.92 2.20 1.49 1.12 0.785 0.498 (~0. 05) (0. 04) (~0. 03) (~0. 04) (0. 02) (~0. 01) (0. 01)

Table 7 Fiber dimensions of pine pulps on Bauer McNett fractions Method Coarseness, m Rio R14 R28 R35 R48 R65 R200 Polysulfide P1 189 161 160 152 134 142 131 (~10 (+7. 4) ~7. 9) ~6. 0) ~1.1) (~3.3) (~3.4 P2 216 148 132 134 128 128 130 ~11 ~7. 9) (3. 8) ~5. 2) ~3. 1) ~2. 5) (2. 1) Kraft K1 287 205 148 137 141 138 132 (21) ~9. 6) ~5. 5) ~2. 0) ~3.2) (~2.2) (~2.4 K2 249 198 145 129 134 117 143 ~7. 9) (9. 5) (3. 9) (~3.1) (~4.0) (~2. 0) (_4. 5) Biokraft B1 162 150 141 141 117 108 96 ~18 ~7. 1) (6. 6) ~7. 9) ~4. 0) ~2. 5) ~1. 2) B2 160 126 121 123 145 172 373 ~7.5) (~10) (~5.5) (~4.8) (~6.4) (~3.5) (16 Soda S1 143 139 136 107 106 77 113 ~7. 1) ~6.0) (~7.9) (~3.0) (~3.7) (~1.3) (1.7 S2 157 125 110 113 100 101 106 (7. 7) (~4. 2) ~4. 5) (~5. 3) (6. 7) (~1. 4) (3.6) Table 8 Fiber dimensions of pine pulps on Bauer McNett fractions Method Fiber count, (Number of fibers in 200 of pulp) R10 R14 R28 R35 R48 R65 R200 Polysulfide PI 315 440 623 1022 1661 1988 3896 ~17) (~17) (~18) (~43) (~29) (~47) (~94 P2 279 483 764 1242 2146 2643 4329 ~16 (28) (24) ~46 ~43 ~57 ~19 Kraft K1 204 353 687 1197 1813 2165 4311 11) (15) (28) (20) (26) (30) (92 K2 247 373 726 1343 1975 2611 3782 ~5) (~19) (~12) (~33) (~62) (~48) (~115 Biokraft B1 377 475 706 1092 1829 2657 5336 ~38) (~22) (~69) (~71) (~76) (~98) (~36 B2 379 595 881 1370 1552 1969 1924 ~12) (~42) (~34) (~50) (~81) (~42) (~73 Soda S1 439 547 767 1476 2224 4125 4822 ~28) (~26) (~39) (~49) (~69) (~74) (~113) S2 403 618 958 1440 2272 3064 5020 (~14) (20) (39) ~88) (26) (50) (142)

Table 9 Fiber dimensions of pine pulps on Bauer McNett fractions Method Fines, (Fibers length < 0.2 mm) (% of total fibers) Rio R14 R28 R35 R48 R65 R200 Pine Polysulfide PI 3.67 1. 31 1. 74 0.92 1.80 2.70 20.1 ~1.0) (~0.9) (~0. 1) (iO 1 ~0. 1) (0. 3) ~0. 6) P2 3. 83 1.81 0.45 1.32 5.71 5.69 23.7 (t1. 0) (t0. 9) (~0. 5) ~0. 4) (0. 3) (0. 3) (0. 2) Kraft K1 10.9 2.60 2.95 0.99 5.28 3.58 25. 3 (5. 6) (1. 3) (1. 1) (0. 3) 0. 5) (0. 3) (0. 6) K2 3.28 1.05 0.69 1.14 5.54 3.79 21.9 (2. 9) ~0. 4) ~0. 8) ~0.3) (~0.2) (~0.1) (~0.3 Biokraft B1 1.24 1.45 0.99 0.48 1.27 2.94 22.9 (1. 2) (0. 8) ~0. 4) ~0. 1) ~0. 1) (0. 3) ~0. 6) B2 2.38 1.67 1.20 3.02 3.17 11.1 33.8 ~1.0) (~0.8) (~0.2) (~0.3) (~0.5) (~0.7) (~0.5 Soda Sl 1.47 1.30 0.60 0.52 1.33 3.39 24. 2 ~0.7) (~1.1) (~0.2) (~0.1) (~0.1) (~0.07) (~0.3 S2 1.64 0.83 0.78 0.99 1.55 3.75 22.6 (~1. 0) ~0. 2) ~0. 2) ~0. 2) ~0. 3) ~0. 2) ~0. 5) Table 10 Fiber dimensions of maple pulps on Bauer McNett fractions Method Arithmetic average fiber length, (mm) R35 R48 R65 R200 Polysulfide PI 0. 644 0.570 0.491 0.328 (0. 002) (0. 001) (0. 001) (0. 004) P2 0.650 0.575 0.490 0.330 ~0. 002) ~0. 001) ~0. 001 (0. 001) Kraft K1 0.662 0.586 0.498 0.333 (0. 007) ~0. 001 (0. 003) (0. 004) K2 0.637 0.563 0.480 0. 322 ~0.005) (~0.001) (~0.009) (~0.003 Biokraft B 1 0.652 0.566 0.480 0. 320 ~0. 005) (0. 003) ~0. 003) (0. 002) B2 0.625 0.560 0.478 0. 315 ~0. 01) ~0. 003) ~0. 002) (iO. 002 Soda si 0.643 0.572 0.483 0.315 ~0.005) (~0.002) (~0.002) (~0.002 S2 0. 607 0.536 0.455 0.284 (~0. 004) (~0.003) (~0.001) (~0.001)

Table 11 Fiber dimensions of maple pulps on Bauer McNett fractions

Method Arithmetic average fiber length, Length-weighed average fiber length, (mm) (mm) R35 R48 R65 R200 R35 R48 R65 R200 Maple Polysulfide P1 0. 644 0.570 0. 491 0.328 0.722 0.636 0. 561 0. 424 (0. 002) (0. 001) (0. 001) (0. 004) ~0. 003) (0. 003) (~0.002) (~0.004 P2 0.650 0.575 0.490 0.330 0.729 0.640 0.560 0.425 (~0. 002) ~0. 001 (0. 001) (0. 001) ~0. 001) (~0.003) (~0.001) (~0.002 Kraft K1 0.662 0.586 0.498 0.333 0.751 0.659 0.570 0.447 (~0. 007) ~0. 001 (0. 003) (0. 004) (0. 002) ~0. 005) (~0. 005) (0. 004) K2 0.637 0.563 0.480 0.322 0.712 0.631 0.552 0.418 ~0. 005) (0. 001) ~0. 0009) ~0. 003) ~0. 009) ~0. 002) (0. 002) ~0. 004) Biokraft BI 0.652 0.566 0.480 0.320 0.749 0.641 0.561 0.426 ~0. 005) ~0. 003) ~0. 003) (0. 002) ~0. 01) ~0. 003) (0. 005) ~0. 004) B2 0.625 0.560 0.478 0. 315 0.709 0.645 0.569 0. 426 ~0.01) (~0.003) (~0.002) (~0.002) (~0.01) (~0.007) (~0.004) (~0.006 Soda S1 0.643 0.572 0.483 0.315 0.719 0.641 0.553 0.412 ~0.005) (~0.002) (~0.002) (~0.002) (~0.005) (~0.005) (~0.003) (~0.004 S2 0.607 0.536 0.455 0.284 0.694 0.612 0.524 0. 382 (~0.004) (~0.003) (~0. 001) (0. 001) ~0. 01) (~0. 005) ~0. 003) (~0. 003)

EXAMPLE 5 Fiber dimensions from the pine fractions as determined in Example 4 were then compared with the yield for the pulp from Example 1 to see if a correlation between the dimensions and yield existed.

Regression equations were obtained using a subset of data points and the percent of mean error calculated from data not in the subset. The difference between measured and calculated yield was reported as percent of testing error. Then for each regression equation obtained from missing data points, the missed measurement data was used to predict pulping yield and the difference between measured yield and calculated yield was reported as percent training error.

The fraction from screen Rl4 gave the best-fit overall prediction of yield using the arithmetic average fiber length (See Table 12 below and Figures 5 and 8).

Coefficients were statistically significant at the 95 % confidence level. R2 showed goodness of fit, and 91.1 % of time data is captured with the equation (Y = 40. 583 * (arithmetic average fiber length in millimeters) -68. 128 (see Equation 5 above, figure 8, and Table 13).

The model between pulp yield and arithmetic fiber length based on fraction Rl4 was validated using a cross validation method (see Table 14) (i. e. , one pair of values was left out from the equation for each calculation). The training error was found to be 0.77 % and the testing error was 1.09 %. The small variation in R2 in Table 14 and the modest training and testing errors, comparable to the mean error of

the full equation (0.87), indicate that the yield prediction was robust and not dominated by any of the individual data points.

Table 12 Summary of coefficients of determinations between pulp yield and Kajaani FS-100 measurements at each Bauer McNett Screens for pine Yield (Pine) Arithmetic av. Length-weighted av. Coarseness Fiber count Fines fiber L. fiber L. 10 0.766 0.601 R, 4 0. 911 0. 835 0. 511 0.238 R28 0. 664 0. 681 0. 752 0. 211 R35 0. 859 0. 790

Table 13 Summary statistic for yield and length on screen R14 (Yield = a + b*Length) a b t-stat (a) t-stat (b R - 68. 128 40. 583-4. 81 7.84 0.911

Table 14 Model validation for arithmetic average fiber length on screen Rl4 Model R Testing error Training error %) % PS #1 missing 0.922 0. 65 2. 50 PS #2 missing 0.903 0. 82 0. 59 K #1 missing 0.903 0. 78 0. 94 K #2 missing 0.903 0. 89 0. 07 B #1 missing 0.980 0. 51 2. 89 B #2 missing 0.903 0. 88 0. 11 S #1 missing 0.903 0. 81 0. 74 S #2 missing 0.846 0. 81 0. 84 Average 0. 77 1. 09

EXAMPLE 6 Fiber length population percentages of fraction Rl4 for pine were studied more thoroughly in order to evaluate the hypothesis that this fraction may carry yield information. Fiber length and fiber coarseness were also examined and subjected to statistical analysis. As set forth in Figure 6, a significant length dependent correlation between yield and fiber length population percentages was obtained.

The data on the share of lengths can be extracted from the analyzers, since the analyzers also are run by microprocessors and thus allow access to this information.

The fiber populations measured were in the length range of 0.9 to 2.2 mm and showed R2 values from 92 to 98%, as set forth in Figure 6. The best prediction was possible using the fiber length population percentages for the 2.16 mm fiber length fraction, which is set forth in Figure 7. The results of this study show that the population percentages in the ranges from 0.9 to 2.2 mm predict pulping yield best.

This fraction most clearly carries the yield information for pine pulps.

As set forth in Figure 9, the plot of length share (the share of total length of fibers in the 2.16 mm class) as a % of total length versus the yield provided a calibration curve leading to the following equation to predict yield: y =-1.8124* (the length share as a % of total fibers) + 62.99 (see Equation 6 above).

EXAMPLE 7 Bauer McNett fractionated maple pulps were also studied and relationships determined. Fiber dimensions from the fractions as determined in Example 4 were

compared with the yield from Example 1 to see if a correlation between the dimensions and yield existed. Statistical analyses were as described above in Example 5. As can be seen in Table 15 below, with maple a strong relationship existed between yield and the arithmetic average fiber length of fraction R200 (Yield/Length (LnR2oo) ; R2 (Maple) = 0. 893). Coarseness of fibers (Yield/Coarseness (R48) ; R2 (Maple) = 0. 760) on screen R48 and fines content measured with the Kajaani FS-100 on screen R200 of Bauer McNett fractions (Yield/Fines (R200) ; R2 (Maple) = 0. 874) also provided good correlation with yield.

Table 15 Summary of coefficients of determinations between pulp yield and Kajaani FS-100 measurements at each Bauer McNett Screens for maple Yield (Maple) Arithmetic av. fiber Length-weighted av. fiber Coarseness Fiber Fines L. L. count R35 0.624 0. 417 0. 605 0. 648 0. 415 R48 0. 308 0. 399 0. 760 0. 799 0.410 R65 0. 778 0. 608 0. 154 0.399 0.590 woo 0. 893 0. 712 0. 021 0.413 0. 874

Other fiber dimensions, e. g. cell wall thickness, were also measured for the Maple pulps. These measurements were made by Metso Automation, Kajaani, Finland, using a FiberLab instrument. The results are summarized in Figure 12. As is apparent from Figure 12, a correlation was found to exist between yield and cell wall thickness of maple pulps. As can be seen in Figure 12, the calibration curve provided the following equation to calculate yield: y =-56.2* (fiber cell wall thickness in micrometers) 2 + 640* (fiber cell wall thickness in micrometers)-1768 (see Equation 9 above).

EXAMPLE 8 Kajaani FiberLab measurements were made at Metso Automation, Kajaani, Finland from the whole pine pulps and from the Bauer McNett Rio and R14 fractions described above. Fiber grammage values were calculated using Equation 4 above with coarseness and fiber width data. The results are set forth below in Table 16.

Table 16 Kajaani FiberLab results from screened pulp, fraction Rio and R14 Method Sample Lna Ll Lw C Ci W CWT F Fgb (mm) (mm) (mm) (µg/m) (µm²) (µm) (µm) (%) (g/m²) Pine, Original pulp Polysulfide P1 1.86c 2.80 3.29 150 684 31. 8 10.0 10.2 4.72 (0. 03) (0. 03) (~0. 03) (~3) (~10. 6) (0. 25) (~0. 17) (0. 47) P2 1. 72 2.76 3.27 152 661 31. 2 9. 8 14. 0 4.87 (~0. 02) (~0. 02) (0. 02) (~3) (~18. 4) (0. 42) (0. 15) (~0.38) Kraft K1 1. 61 2.68 3. 18 148 645 30. 6 9. 9 17. 2 4.84 (0. 03) (0. 03) (0. 02) (~5) (~33. 8) (0. 30) (0. 51) (~0.35) K2 1. 78 2.73 3.24 137 607 30. 0 9. 4 10. 8 4.57 (0. 02) (0. 02) (0. 02) (~4) (14. 4) (0. 25) (0. 2) (0. 15) Biokraft B1 1. 83 2.79 3.28 140 637 30. 5 9. 8 11. 3 4.59 (~0.01) (~0.01) (~0.01) (~2) (~9.9) (~0.1) (~0.2) (~0.17) B2 1. 79 2.75 3.25 137 654 30.6 10.2 11.2 4.48 (0. 03) (0. 02) (0. 02) (~4) (~14. 7) (~0. 3) (~0. 17) (~0. 38) Soda S1 1. 91 2.80 3.29 128 586 29. 1 9. 6 8. 9 4. 40 (0. 00) (0. 01) (0. 01) (~2) (~7. 1) (0. 12) (~0. 12) (0. 12) S2 1. 86 2.77 3.24 118 548 28. 0 9. 4 10. 2 4.21 (0. 02) (0. 02) (0. 02) (2) (34. 8) (~0. 38) (0. 51) (0. 31) Bauer McNett Fraction Rio Polysulfide P1 3. 13 3.49 3.75 170 813 33.9 11. 6 1.2 5.01 (0. 01) (0. 01) (~0. 02) (~3) (~14. 5) (~0. 38) (~0. 15) (~0. 06) P2 3.11 3.47 3.75 160 765 33. 1 11. 0 1. 0 4. 83 (~0. 01) (~0. 02) (~0. 03) (~2) (16) (0. 2) (~0. 31) (0. 1) Kraft Kl 3.01 3. 36 3.63 162 763 32.8 11.3 0.9 4.94 (0. 01) (~0.01) (~0.01) (~3) (~33. 8) (0. 62) (0. 36) (~0. 17) K2 2. 97 3. 34 3. 62 155 736 32.3 11.0 0.9 4. 80 (~0. 01) (0. 01) (0. 02) (~6) (38. 6) (0. 65) (0. 51) (~0. 12) Biokraft B1 3. 09 3. 43 3. 70 154 760 32. 5 11. 5 0. 8 4. 74 (0. 02) (0. 02) (0. 02) (~6) (~31.1) (~0. 55) (0. 46) (~0. 15) B2 3. 03 3.39 3. 66 145 747 32. 3 11. 3 0. 8 4. 49 (~0. 01) (0. 01) (0. 01) (~3) (30. 9) (0. 55) (0. 4) (~0. 12) Soda S 1 3. 01 3.34 3. 61 151 652 30. 3 10. 5 0. 8 4. 98 (~0. 01) (0. 01) (0. 01) (~6) (~21. 3) (~0. 52) (~0. 15) (0. 06) S2 2. 92 3.27 3.53 129 597 29. 0 10. OF) 4.45 (~0. 01) (0. 02) (0. 02) (~3) (9. 3) (0. 25) (0. 06) (~0. 12) Bauer McNett Fraction R) 4 Polysulfide P1 2. 76 3.02 2. 96--33. 5 9.7 0. 15 - (0. 03) (~0. 03) (0. 03) (~0.1) (~0.1) (~0. 06) P2 2.75 3.06 3.31 - - 32. 8 9.7 0. 16 - (0. 01) (0. 02) (0. 03) (~0.1) (~0.1) (~0. 08) Kraft K1 2.73 2.99 3.38 - - 32.9 9.7 0.22 - (~0. 02) (~0. 01) (0. 01) (~0.1) (~0.0) (~0.03) K2 2. 66 2.97 3. 29 - - 31.7 9.2 0.18 - (0. 01) (0. 01) (~0. 02) (0. 1) (~0.1) (~0.05) Biokraft B1 2.65 2.92 3. 21 - - 32.3 9.5 0.44 - (~0. 02) (0. 02) (~0. 02) (~0.1) (~0.0) (~0. 06) B2 2. 61 2. 91 3. 20 - - 31.6 9.3 0.46 - (~0. 01) (~0. 01) (~0. 01) (0. 0) (0. 0) (~0. 08) Soda S1 2. 58 2.84 3. 12 - - 30.8 9.1 0. 06 (0. 01) (~0. 01) (~0. 01) (~0.1) (~0.1) (~0.06) S2 2. 56 2.86 3. 16 - - 30.2 9.1 0.14 (~0. 01) (0. 02) (~0.04) (~0.0) (~0.1) (~0.06)

a Ln = number (arithmetic) average fiber length, Ll= length-weighted average fiber length, Lw= weight-weighed average fiber length, C= fiber coarseness, Ci= Coarseness index, W= Width, CWT= cell wall thickness, F= Kajaani fines, Fg= fiber grammage. b Fiber grammage (Fg), g/m2 = C/W, c All of these variables are average of three replicates.

EXAMPLE 9 The pulp yields obtained in Example 1 and the results of the FiberLab measurements from Example 8 were then compared using the statistical analyses described above. Pulp yield could be predicted from R14 fiber length with the equation yield = 49.9* (arithmetic average length of the fibers in millimeters) -73. 8 (see Table 17 and Equation 8 above) from FiberLab data. Model validation resulted in 1.07 % training error and 0.80 % testing error (Table 18).

Table 17 Summary statistic for yield and length on screen R14 (Yield = a + b*Length) a b t-stat(a) t-stat(b) R2 - 73. 8 49. 9 1-5. 12 8. 08 0. 916 Table 18 Model validation for arithmetic fiber length from screen Rl4 Model R2 Testing error Training error % % % PS #1 missing 0.828 0. 64 1. 95 PS #2 missing 0.884 0. 73 0. 73 K #1 missing 0.846 0. 85 0. 67 K #2 missing 0.846 0. 89 0. 41 B #1 missing 0.922 0. 61 2. 14 B #2 missing 0.846 0. 88 0. 29 S #1 missing 0.846 0. 82 2. 11 S #2 missing 0.757 0. 91 31 Average 0. 80 1. 07

The fraction from screen Rl4 gave the best-fit overall prediction of yield, using the fiber width (Table 19 and Figure 11). R2 showed goodness of fit, and 97.7 % of the time data was captured with the equation (Y = 3.420 * (fiber width in micrometers) -66.732 (see Equation 7 above) ) from the calibration curve set forth in

Figure 11. The model between pulp yield and fiber width based on fraction R14 was validated using a cross validation method (Table 20), with the training error was found to be 0.57 % and the testing error 0.44 %. The small variation in R in Table 20 and the modest training and testing errors, comparable to the mean error of the full equation (0.44 %), indicated that the yield prediction was robust and not dominated by any of the individual data points.

Table 19 Summary statistic for yield and fiber width on screen R14 (Yield = a + b*Width) a b t-stat a t-stat R- - 66. 732 3. 420-9. 89 16. 2 0. 977 Table 20 Model validation for fiber width Model R2 Testing error Training error (%) (%) PS #1 missing 0.922 0. 40 1. 07 PS #2 missing 0.980 0. 50 0. 03 K #1 missing 0.980 0. 42 0. 85 K #2 missing 0.960 0. 42 0. 64 B #1 missing 0.960 0. 48 0. 12 B #2 missing 0.960 0. 36 1. 14 S #1 missing 0.960 0. 44 0. 65 S#2missing 0.922 0. 49 0. 10 Average 0. 44 0. 57

EXAMPLE 10 The correlation of hygroscopicity of pine and fiber dimensions at different relative humidities was investigated. Pulp samples were weighed and then placed on open aluminum pans into a chamber, where the temperature and air humidity was controlled within. 5 % changes. The samples were acclimatized for 3 days to reach

the moisture equilibrium with the surrounding air and then weighed again. The weight gain/loss was due to adsorbed/desorbed moisture. The relative humidity of air affected the fiber products in undesirable ways causing cockling, bugling, and other changes in product dimensions.

Statistical analyses of the water adsorption of the pine fibers and the fiber dimensions obtained in Example 8 led to the creation of a calibration curve based on the water adsorption and the widths of the fibers (see Figure 13). As the industry standard is to test the hygroscopicity of papers at 50% relative humidity, Figure 13 provided the following formula for calculating hygroscopicity at 50% relative humidity: Y50=-0. 2449x2 + 17.129x-282. 78 where Y50 is the moisture in fiber as . g/m at 50% humidity, and x is the fiber width in micrometers (see Equation 10 above).

It will be understood that various modifications may be made to the embodiments disclosed herein. Therefore, the above description should not be construed as limiting, but merely as exemplifications of preferred embodiments. For example, combinations of fiber dimensions, such as cell wall thickness, length, width may be used to determine yield of a pulp in accordance with the methods described herein. Those skilled in the art will envision other modifications within the scope and spirit of the claims appended hereto.