Note: Descriptions are shown in the official language in which they were submitted.
CA 02305201 2000-04-04
WO 99/24815 PCT/CA98/OlOZ4
1
DETERMINATION OF ANIONIC SPECIES
CONCENTRATION BY NEAR INFRARED SPECTROSCOPY
FIELD OF THE INVENTION
This invention relates generally to a method for determining anionic species
in
aqueous solution, particularly pulp process liquors of cellulosic pulp
manufacturing
processes, by near infrared spectrophotometry and more particularly to the use
of an on-
line method for determining concentration parameters of said process liquors,
and
subsequent control of said cellulosic pulp manufacturing process by use of
said
determined parameters.
BACKGROUND OF THE I1WENTION
Kraft pulping is performed by cooking wood chips in a highly alkaline liquor
which
selectively dissolves lignin and releases the cellulosic fibers from their
wooden matrix. The
two major active chemicals in the liquor are sodium hydroxide and sodium
sulfide. Sodium
sulfide, which is a strong allcali, readily hydrolyses in water to produce one
mole of sodium
hydroxide for each mole of sodium sulfide. The term "sulfidity" is the amount
of sodium
sulfide in solution, divided by the total amount of sodium sulfide and sodium
hydroxide
and is usually expressed as a percentage (% S) which varies between 20 and 30
percent in
typical pulping liquors. The total amount of sodium hydroxide in solution,
which includes
the sodium hydroxide produced as the hydrolysis product of sodium sulfide, is
called either
"effective allcali" (EA}, expressed as sodium oxide, Na20 before pulping, or
residual
effective alkali (REA) after pulping. Timely knowledge of these parameters
would enable
good control of the pulping process.
At the beginning of the kraft process, "white liquor" is fed to a digester.
This white
liquor contains a high amount of effective alkali up to 90 g/L, as Na20. At
intermediate
points in the digester, spent liquor, or "black liquor," is extn~cted from the
digester. This
CA 02305201 2000-04-04
WO 99/24815 PCT/CA98/O10?,4
2
spent liquor contains low levels of effective alkali - less than 30 g/L, as
NazO and also
contains large amours of organic compounds which, generally, are burned in a
recovery
furnace. Resultant inorganic residue, called smelt, is then dissolved to form
"green liquor"
which has a low concentration of effective alkali and a high concemrafion of
sodium
carbonate - up to 80 g/L, as Na20. White liquor is regenerated finm the green
liquor by
causticizing the carbonate through the addition of lime. After the
recausticizing operation, a
small residual amount of sodium carbonate is left in the white liquor. The
combined
amount of sodium hydroxide, sodium sulfide and sodium carbonate is called
total titratable
alkali (TTA). The causticizing e~ciency (CE) is usually defined as the
difference in the
amounts, as Na20 of sodium hydroxide between the white and green liquors,
divided by the
amount, as Na20 of sodium carbonate in the green liquor. Sodium sulfate,
sodium
carbonate and sodium chloride represent a dead load in the liquor recycling
system. The
reduction efficiency (RE) is defined as the amount, as Na20 of green-liquor
sodium sulfide,
divided by the combined amounts, as Na20, of sodium sulfide, sodium sulfate,
sodium
I 5 thiosulfate and sodium sulfite in either green liquor or the smelt.
The timely knowledge of the white-liquor charge of EA and of black-liquor EA
would close the control loop in the digester and optimise for example,
production and
product quality and chemical utilization, of alkali and lime consumption. The
control of
sodium sulfide, TTA and of non-process electrolytes, such as sodium chloride
and
potassium chloride would also have a beneficial impact on closed-cycle kraft-
mill
operations. For example, environmentally-driven reduction of sulfur losses
generally
increases liquor sulfidity, thereby creating a sodiumaulfur imbalance that
needs to be made
up through the addition of caustic soda. Another important need is the control
of TTA in
green liquor, which is most easily done by adding weak wash to a smelt
dissolving tank.
The value of the green-liquor TTA is important because it is desirable to
maintain the TTA
at an optimal and stable level so as to avoid excess scaling while obtaining a
high and
stable white liquor strength The ongoing development of modern chemical
pulping
Processes has thus underscored the need for better control over all aspects of
kraft-mill
operations and more efficient use of all the chemicals involved in the process
by knowledge
of the concentration of aforesaid species in the liquors.
CA 02305201 2000-04-04
WO 99/24815 PCT/CA98/01024
3
Sodium carbonate is difficult to characterize and quantify in situ because of
a
current lack of on-line sensors which can tolerate long-term immersion in
highly alkaline
liquors. important economic benefits could result from causticizing control
with a reliable
sensor for sodium carbonate. Accurate ca~c~;on is critical for the uniform
production
of high-strength white liquor in that adding too much lime to the green liquor
produces a
liquor with poorly settling lime mud, whereas adding too little produces a
liquor of weak
~'~8~~ ~rmining the relative quantities of EA and carbonate in green and white
liquor
is thus important for controlling the causticizing process.
Various methods of on-line measurements of either EA or sodium hydroxide have
been proposed. The use of conductivity methods for green and white liquors is
well-
established as a pulp and paper technology. Unfortunately, conductivity probes
are prone to
drift due to scaling, as well as interferences from other ionic species.
Therefore, these
devices require frequent maintenance and re-calibration. An early example of
s~h
measurements describes a method that can determine the EA by neutralizing
hydroxide
ions with carbon dioxide ( 1 ). The conductivity of the solution is measured
before and after
treatment. The difference in conductivities is proportional to the hydroxide
ion
concentration of the liquor. High levels of sodium hydroxide, however, will
increase the
neutralizing time. In white liquors, this time is too long for ei~ective
process control
purposes. Chowdhry (2) describes an analysis of kraft liquors that uses
differences in
conductivity before and after precipitation of carbonates using BaCl2, an
approach which is
not practical.
However, even though conductivity probes may not be suitable for on-line
measurements of EA in white or green liquors, this kind of sensor is also used
with the
liquor produced during the early stages of the pulping in upper-recirculation
digester lines.
An example of a successful commercial version of an automatic titrator (3)
involves
titrating alkali with sulfuric acid until no change in conductivity is
observed. This
determination is straightforward and works very well for the impregnation and
early stages
of the cook, but not for the extraction stage. With extraction liquors, a more
complex
pattern is observed when significant quantities of organic acids and black-
liquor solids
appear in the liquor, and the end-point determination becomes more difficult
near the end
CA 02305201 2000-04-04
WO 99/24815 PCT/CA98J01024
4
of the cook. On-line titration methods used in pulp mills suffer from
fi~equent maintenance
problems. Thus, most mill-site measurements still rely on standard laboratory
methods.
At present, control of digesters is performed by keeping the chip and white
liquor
feeds at preset levels. These levels are determined by the overall production
rate, and
control is achieved by adjusting the temperature profile of the cook and
determining the
resultant blow line kappa number. The philosophy behind this strategy is that
alkali
consumption during the removal of lignin is proportional to chip feed at a
given kappa
number. Alkali not consumed in the impregnation phase is then available for
the bulk
removal of lignin that occurs in the pulping zone. This is usually performed
by predicting
the pulp yield with the H-factor (4). The disadvantage of this method is that
it assumes
uniform chip moisture content, pH and density, as well as digester
temperature, etc. Since
the pulp must be analysed in the laboratory for lignin content, this makes it
difficult to close
the control loop in a timely manner. Ideally, a much better way of controlling
digester
operations would be to measure the EA concentration in black liquor directly
on-line at an
appropriate time in the cooking process on both the upper and lower (main)
recirculation
loops in the digester, as well as the ILEA concentration on the extraction
line at the end of
the cook. An on-line method that would give a direct measurement of the EA
throughout a
cook is therefore needed.
Methods relying on spectroscopic methods have been proposed because of the
limitations of titration and conductivity methods for liquor analysis. It is
known that
hydrosulfide ions absorb very strongly in the ultraviolet at 214 nm (5, 6, 7).
However, this
absorption is so strong that a very small pathlength, i.e. less than 10
microns is needed to
get a measurable signal which yields a linear catibration curve (8}. A cell
with such a small
optical path is prone to plugging and, hence, not practical for on-line
applications.
Extensive 1:1 x 103 or 104 dilution is practiced, which results in inaccurate
results and
increases the risk of sulfide being oxidized.
The dilution approach has also been used in techniques such as capillary zone
electrophoresis which use UV detectors (9, 10). Errors in sulfidity
measurements
exceeding 50% were reported. Accordingly, a method which does not need
dilution is
needed.
CA 02305201 2000-04-04
WO 99/Z4815 PCT/CA98/01024
Infiared spectroscopy can distinguish between the inorganic and organic
components of liquors and a number of infrared methods have been proposed.
Faix et al
( 11 ) propose a method for organic compounds in black liquor, based upon on-
line infrared
a~refl~~nce (ATR) measurements between 1400 and 1550 clri l A similar
5 method for kappa number determination ( 12) correlates the increase in the
integrated band
intensity at 1_118 clri ~ with decreasing kappa number. Neither of these
methods can be
used for process control because of interferences fi~om carbohydrates and
uncertainties in
the value of process variables such as liquor-to-wood ratio. Leclerc et al.
(13, 14, 15, 16)
teach that one can measure EA and dead-load components in kraft liquors with
FT-IR
ATR, and that one can use these measurements to control the operations of
important
process units involved in the manufacture of kraft pulp such as the digester,
recausticizers
and recovery boiler. However, ATR optical reflecting elements immersed in very
alkaline
liquors, and/or acidic or oxidizing cleaning solutions, are prone to be
vulnerable to etching
and/or scaling of their surface, which necessitates frequent replacement, re-
polishing and
re-calibration of the elements. Materials that are resistant to caustic,
acidic, or oxidizing
environments are few and cannot be used for ATR measurements in the mid-
infi~ared
region of interest due to infrared absorption of the material itself. ATR
elements have also
slightly differing optical paths and surface properties that exhibit memory,
which makes the
transfer to other instnunents of calibrations developed on one instrument very
difficult to
achieve without substantial expenditures of time and labour.
Recent advances in FT IR iush~mentation and software have made possible the
more widespread use of the near-infi~ared region of the spectrum for
determining aqueous
components such as dissolved elecholytes. Each ionic species causes a unique
and
measurable modification to the water bands that is proportional to its
concentration.
Advantages over previous techniques include: no sample preparation, short
measurement
times, relatively long optical paths and the possibility of using fiber-optic
technology for
real-time, in situ measurements. Also, temperature effects and interferences
by other
canons and anions can be modeled in this spectral region through the use of
partial least-
squares (PLS) mufti-component calibration techniques. PLS is a well-larown
multi-
component calibration method (17, 18). This method enables one to build a
spech~al model
which assumes that the absorbance produced by a species is linearly
proportional to its
concentration. This has been shown by (19, 20, 21, 22, 23). However, because
of its
CA 02305201 2000-04-04
WO 99/24815 PCT/CA98101024
6
relatively intense water bands, the spectral region situated from 4000 to 8000
cai 1 is only
suitable for optical paths ranging from 0.5 to 1.5 mm, a limitation which
precludes the
~c~e determination of weakly absorbing electrolytes such as carbonate, sulfide
and
chloride. Sodium hydroxide, on the other hand, generates a strong signal that
is easily
detectable in this region (24, 25, 26). The concentration of dissolved
electrolytes, such- as
sodium hydroxide, carbonate and chloride concentrations in aqueous streams,
such as
seawater or white liquor have been measured. Accurate results were obtained
for
hydroxide but not for the other ions. Similar results were obtains more
recently (27) with
a PLS calibration. The correlation data obtained for sulfide and carbonate are
not reliable,
and cannot be used as a basis towards developing a method for co~rol]ing the
manufitcture
of cellulosic pulp. A near-infiared PLS method, which can measure sodium
sulfide and
TTA with an accuracy of I to 2 g/L has been described (28). The calibration
method,
however, could not distingwish between sodium carbonate and sodium hydroxide
because
of the similar spectral signatures produced by these two ions, as well as the
relative
wealaiess of the carbonate spectrum. Reference 24 through 28 demonstrate that
hydroxide
is easy to measure in the range 4000 to 8000 ciri 1, while other components
such as
carbonate and sulphide are not. The results obtained (27, 28) strongly suggest
that a control
method for a pulp manufacturing process based on the simultaneous and separate
determination of hydroxide, carbonate and sulfide would be very diffcult with
the small-
bore flow cell used for their work. This type of flow cell would also be
susceptible to
Plugging by suspended solids and fibers, thereby rendering the method
unworkable. The
spectral region situated from 8000 to 12000 ciri' is more amenable to the use
of longer
optical paths ranging from 3 to 20 mm, which makes it much easier to couple a
wide-bore
flow cell to any system of pipes used in the mill. For example, (23, 29) a PLS
calibration
has been used to resolve the hydroxide and chloride ion spechvm near 10500
cm'. In both
cases, however, the range of concentration was extremely wide (0 to 5
moles/L), the
spectra were somewhat noisy, and the precision was no better than 5 g/L for
both species.
For the spectral information to be useful for process control engineers, the
correlation data
must be accurate to within one percent and the level of precision, in the
range of 0.5 to 1
gfL. The level of precision reported is, thus, inadequate for process control.
A recent publication (30) broadly discloses a method of controlling the
causticizing
reaction for producing a white liquor having multiple white liquor components
from a
CA 02305201 2000-04-04
wo ~nasis pc~ricwssroioza
7
green liquor having multiple green liquor components, comprising the steps of
measuring a
characteristic of each of said glen liquor components; measuring a
characteristic of each
of said white liquor components; evaluating said green liquor component
characteristics
and said white liquor component characteristics in a non-linear, application
adaptable
controller to produce a causticizing control signal; and controlling said
causticizing reaction
responsive to said causticization control signal to produce white liquor
wherein the
characteristics are generally measured by near infi~ared or polarographic
measurement
devices and evaluating the characteristics in a non-linear, application
adaptable controller
to produce a causticizing control signal for controlling the amount of time to
a shaker.
However, the specific multiple component liquid process analyzer of use in the
disclosed
process would require a pathlength of less than 3 mm at 1100 to 2200 nm to
avoid
complete saturation of the incident light beam by water molecules in the
sample.
There is, therefore, a need for the rapid determination of effective alkali,
residual
alkali, sodium sulfide and sodium carbonate, particularly, in pulping process
liquor by
spectrophotometric means which provide for a process liquor pathlength of
greater than 3
mm without saturation of the incident radiation beam by water molecules of the
sample.
LIST OF PUBLICATIONS
1. U.S. Patent No. 3,553,075 - Rivers
2. U.S. Patent No. 3,607,083 - Chowdhry
3. U.S. Patent No. 3,886,034 - Noreus
4. K. E.Vroom, Pulp paper Mag.Can., 1957, 58(3), 228
5. U.S. Patent No. 5,582,684 - Holinquist and Jonsson
6. D. Peramunage, F. Forouzan, S. Litch. Anal. Chem., 1994, 66, 378-383
7. Paulonis et al. PCT Application WO 91/17305. Liquid Composition Analyser
and
Method
8. Paulonis et Krishnagopalan. Kraft White and Green Liquor Composition
Analysis.
Part I: Discrete Sample Analyser. J. Pulp paper Sci., 1994, 20(9), J254-J258
i
CA 02305201 2002-09-09
WO 99/24815 PCT/CA98/01024
8
9. Salomon, D.R., Romano, J.P. Applications of Capillary Ion Analysis in the
Pulp
and Paper Industry. J. Chromatogr., 1992, 602(1-2), 219-25
10. Rapid Ion Monitoring of Kzaft Process Liquors by Capillary
Electrophoresis.
Process Control Qual., 1992, 3(1-4), 219-271.
11. U.S. Patent No. 4,743,339. Faix et al.
12. Michell. Tappi J., 1990, 73(4), 235.
13. Lecierc et al. J. Pulp Paper Sci., 1995, 21(7), 231
14. U.S. Patent No. 5,282,931- Lecierc et al.
15. U.S. Patent No. 5,364,502 - Leclerc et al.
1.0 16. U.S. Patent No. 5,378,320 - Leclerc et al.
17. Haaland, D.M. and Thomas, E.V. Anal. Chem., 60(10): 1193-1202 (1988)
18. Haaland, D.M. and Thomas, E.V. Anal. Chem., 60(10): 1202-1208 (1988)
19. Lin and Brown. Appl. Spectrosc. 1992, 46(12), 1809-15
20. Lin and Brown. Enrriron. Sci. Technol. 1993, 27(8), 1611-6
21. Lin and Brown. Anal.Chem.,1993 , 65(3), 287-92
22. Lin and Brown. Appl. Spectrosc. 1993, 4?( 1 ), 62-8
23. Lin and Brown. Appl. Spectrosc. 1993, 47(2), 239-41
24. Watson and Baughman. Spectroscopy, 1987, 2(1), 44
25. Hirschfeld. Appl. Spectrosc., 1985, 39(4), 740-1
26. Grant et al. Analyst., 1989, 114(7), 819-22
27. Vanchinathan, S., Ph.D. Thesis. Modeling and control of kraft pulping
based on
cooking liquor analysis, Auburn University, 1995. Tappi 3., 1996, 79(10):187-
191
CA 02305201 2000-04-04
WO 99/24815 PGT/CA98101024
9
28. U.S. Patent No. 5,616,214. Leclerc
29. Phelan et al. Anal, Chem., 1989, 61(3), 1419-24
30. W098/10137 - Fisher Rosemont Systems, Inc.; March 12, 1998.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a rapid method for
determining
the concentration of OH', C03' and HS' species in aqueous solution,
particularly in
solutions containing all three species.
It is a further object to provide a rapid method for determining the
concentration
of organic species present in a puiping process liquor, particularly, in the
presence of at
least one of the species selected from OH', C03- and HS'.
It is a further object to provide a rapid method for determining the
concentration
of effective alkali, residual alkali, sodium sulfide, sodium carbonate and
dead-load
components such as chloride and dissolved organic species in pulp liquors.
It is a yet further object to provide said rapid process which does not need
frequent equipment maintenance, sample pretreatment or chemical reagents.
It is a still yet further object to provide said method which, optionally,
allows a
plurality of puip liquor process streams to be multiplexed to a single
analyser in a fibre-
optic network.
It is a further object to provide apparatus for effecting said methods.
Accordingly, the invention provides in one aspect a method for determining the
concentration of anionic species selected from the gmup consisting of OH',
C03a and HS'
in an aqueous sample solution, said method comprising subjecting said solution
to near
infiared radiation at a wavelength region of wave numbers selected from about
7,000 to
14,000 cni 1 through a solution path length of at least 3 mm to obtain
spectral data for
said solution; obtaining comparative spectral data for said anionic species at
known
concentrations in aqueous solutions; and correlating by multivariate
calibration the
relationships between said spectral data of said sample solution and said
comparative
spectral data to determine said concentration of said anionic species in said
sample
solution.
CA 02305201 2000-04-04
WO 99/24815 PCT/CA98/OlOZ4
Preferably, the wavelength is selected from 7,000 to 12,000 crti', and more
preferably, 9,000 to 12,000 cni'.
The spectral data is preferably obtained by transmittance speetrophotometry,
and
more preferably, from a transmission cell. The relationships between the
spectral data of
5 the sample and the comparative spectral data are, preferably, obtained with
a partial-least
squares multivariate calibration.
In a preferred aspect the invention provides a process for controlling the
operation of individual unit operations within a ceLlulosic pulp manufacturing
process,
which comprises the steps of
10 subjecting samples of process liquors to near infrared radiation at a
wavelength
region of wavenumbers from about 7,000 to 14,000 cni' to produce
measurements of said liquor;
recording the spectrum of different mixture solutions of synthetic and process
liquors having known concentration parameters;
correlating by multivariate calibration the relationships between the spectra
of the
process liquor samples and the different mixture solutions of known
concentration
parameters so as to simultaneously determine concentration parameters in the
process liquor samples; and
adjusting the individual unit operations of the cellulosic pulp manufacturing
process as required by controlling at least one process parameter to bring the
final
product of said unit operation to a desired value, wherein said final product
is
determined in part by concentration parameters in said process Liquors, as
determined by the near infrared measurements of said concentration parameters.
Thus, the invention, in a preferred aspect, provides a rapid method for the
control
of a cellulosic pulp manufacturing process via on-line measurement of chemical
concentration parameters in process liquor streams with near infrared
radiation. The
method eliminates the need for (l) manual sampling, (ii) frequent equipment
maintenance,
(iii) a dedicated instrument at each sampling point, (iv) compensation for
instrumental
drift, and, optionally, (v) an environmentally controlled spectrometer housing
near the
sampling location(s). The method includes the steps of (l) withdrawing samples
of a
process liquor stream from a cellulosic pulp manufacturing process, (ii)
subjecting the
samples to near-infrared spectrophotometry over a predetermined range of
wavenumbers
CA 02305201 2000-04-04
WO 99/Z4815 PCT/CA98/OlOZ4
11
so as to produce spectral measurements which determine the concentrations of
different
combinations of chemical components, (iii) correlating by multivariate
calibration the
relationships between the spectral measurements of unknown samples and the
spectral
variations shown by different combinations of chemical components of the
process liquor
so that concentration parameters can be accurately determined for typical
levels of
chemical components present in the process liquor, and (iv) controlling at
least one
process parameter so as to obtain optimal operation of the cellulosic pulp
manufacturing
process.
The method of the present invention uses "wide-bore" near infrared
spectrometry,
i.e. wherein the cell path of the solution subjected to the near infrared
radiation is at least 3
mm, preferably 3-20 mm, and more preferably 5-12 mm. This clearly
distinguishes the
invention over prior art methods (27, 28) which teach the use of "narrow bore"
path lengths
of <2 mm, when measuring the first overtone of the near infi~red
(approximately 4,000
7,000 cni'), or <1 x 10'3 cm when measuring the mid-infi~ared region
(approximately
4,000-400 eiri ~).
The present invention is thus of significant value in providing for the rapid
determination of the alkalinity OH', C03 and HS- levels in pulp liquors, which
contains
inter alia, all three species in varying amounts.
Surprisingly, the invention provides that although signal strengths of the
water
absorption bonds diminish with increasing wavenumber from the infrared to the
visible
spectral range, increasing the sample path length enables suff cient signal
absorption to
occur in mufti anionic species-containing solutions, within the background
noise to enable
enhanced accurate spectral data on each of the anionic species to be obtained.
Such rapid
and accurate anionic species concentration of the order of tl g/L in pulp
liquors allows for
good and beneficial control of pulp liquor concentrations.
Cellulosic pulp cooking Hquor which has been extracted from the cooking
process
at some point after coming into contact with the wood chips is collectively
referred to as
black liquor. The actual composition of any black liquor can vary
substantially with a
strong dependence on the time a~ location of extraction, the original
composition of the
wood and/or liquor upon entering the digester, and the cooking conditions. The
dissolved
substances in black liquor fall into two primary categories: total inorganic
content and total
l ..
CA 02305201 2002-09-09
WO 99/Z48i5 PCT/CA98/01024
12 -
organic content. The inorganic content, which constitutes 25 to 40% of the
dissolved
substances, consists primarily of anionic species such as hydroxide,
hydrosulfide,
carbonate, chloride, sulfate, sulfite and thiosulfate, where sodium is the
primary counter
ion. The organic content, which constitutes the remaining 60 to 75% of the
dissolved
substances, can be further divided into three main categories: lignin -
aromatic organic
compounds (30-45%), carbohydrates - hemicelluloses and cellulose degradation
products
(28-36%), and extractives - fatty and resinous acids (3-5%). These organic
species provide
unique contributions to the overall electromagnetic spectral signature of a
black liquor
sample. Therefore, it is possible to relate the near infrared spectrum of a
black liquor
sample to the total or constituent organic content of that liquor for
calibration purposes. In
this way, it is possible to simultaneously measure, for example, the lignin
and the sodium
hydroxide (or EA) content of a black liquor extracted from a digester. In a
more general
sense, the total organic content and the total inorganic content, as well as
the sum of these
two constituents (i.e., the total dissolved solids) would also be quantifiable
in a similar
IS manner. Surprisingly, the transmission of near infrared radiation through
black liquor is
still great enough to quantify these components even when a pathlength of 10
mm is used.
Thus, the present invention provides a rapid method for determining effective
alkali, residual effective alkali, sodium sulfide, sodium carbonate, and dead-
load
components, such as sodium chloride, sodium sulfite, sodium sulfate, sodium
thiosulfate
and dissolved organic species in process liquors and controlling appropriate
parameters in
the cellulosic pulp manufacturing process based on the determined values. The
proposed
method largely eliminates the need for frequent equipment maintenance, sample
pretreatment and the use of chemical reagents. High sample throughput can also
be
obtained by allowing many process streams to be multiplexed to a single
analyser through
an optional fiber-optic network .
Samples of process liquors are analysed by near-infrared Fourier transform
infrared
(FT-IR) spectrometry. Spectra are collected using a flow-through wide-bore
transmittance
accessory. The absorbance of the liquor is measured over a predetermined
wavelength
region. The absorbance is then correlated through a multivariate regression
method known
in the art as partial least-squares {PLS) with the concentration of the
absorbing compound.
This correlation is made by comparing results previously obtained with
standard samples.
i
CA 02305201 2002-09-09
WO 99/24815 PCT/CA98/01024
13 .
The chemical composition of the liquor is then calculated. The process samples
are also
analysed with either standard CPPA, SCAN or TAPPI analytical methods, to
establish a
correlation with the data obtained by near-infrared spectrometry.
The on-line method for EA and REA may primarily be used for controlling the
operation of either batch or continuous digesters. The blow-line kappa number
can then be
predicted by using its well-known relationship with the REA. The method can
also be used
for controlling carbonate and hydroxide levels in green and white liquors. The
causticizing
e~ciency could also be calculated. In sumnnary, this new sensing and control
method could
IO replace automatic titrators and conductivity sensors. It would also give
previously
unavailable information on the carbonate levels in process liquors, while
improving the
control of scaling in multi-effect evaporators.
In a preferred aspect, the present invention provides a method for measuring
effective alkali in a kraft pulp manufacturing process and controlling the
appropriate
process parameters said method comprising the steps of
subjecting samples of process liquors to near infrared radiation at a
wavelength
region of wavenumbers from about 7,000 to 14,000 cni ~ to produce measurements
of
said liquor;
recording the spectrum of different mixture solutions of synthetic and process
liquors having known EA;
correlating by multivariate calibration the relationships between the spectra
of the
process liquor samples and the different mixture solutions of known EA so as
to
simultaneously determine -EA in the process liquor samples; and
adjusting the cooking conditions selected from time and temperature of the
kraft
pulp manufacturing process by controlling at least one process parameter to
bring said
cooking conditions as determined by said near infrared measurements on the
process
liquor to desired values.
In a further aspect the invention also provides an apparatus for determining
the
concentration of an anionic species selected from the group consisting of OH',
CO3 and
HS' in an aqueous solution, said apparatus comprising sample means for
providing said
sample with a solution path length of not less than 3 mm; FT near-infrared
means for
* Trade-mark
CA 02305201 2000-04-04
WO 99/24815 PCT/CA98/OI024
14
subjecting said solution over said path length to near-infi~ared radiation at
a wavelength
region of wave numbers selected from about 7,000 to 14,000 cni 1; and spectral
recordal
means for recording spectral data of said radiation after subjecting said
solution to said
radiation.
BRIEF DESCRIPTION OF THE DRAWIN S
In order that the invention may be better understood, preferred embodiments
will
now be described by way of example, only, wherein:-
FIG. 1 is a diagrammatic view of the recovery and recausticizing process
system,
complete with sensing and control apparatus according to one embodiment of the
present
invention;
FIG. 2 is a diagrammatic view of a pulp digester, complete with sensing and
control apparatus according to a further embodiment of the present invention;
FIG. 3 is a graph of absorbance versus reciprocal centimeters showing the
change
in near-infrared absorbance with respect to an air reference between 4000 and
14000
wavenumbers for a range of temperatures selected from between 5 and
25°C.;
FIG. 4 is a PLS calibration graph of the predicted versus actual EA
concentration
for the three-component PLS calibration model;
FIG. 5 is a PLS calibration graph of the predicted versus actual sodium
carbonate
concentration for the three-component PLS calibration model;
FIG. 6 is a PLS calibration graph of the predicted versus actual hydrosulfide
concentration for the three-component PLS calibration model;
FIG. 7 is a graph of absorbance versus reciprocal centimetres showing the
change
in near-infrared absorbance for a range of diluted black liquors with respect
to a 10 g/L
EA reference between 4000 and 14000 wave numbers;
FIG. 8 is a graph of absorbance versus percent black liquor added showing the
change in near-infrared absorbance at 11500 cm 1 for a range of diluted black
liquors
with respect to a 10 g/L EA reference;
FIG. 9 is a PLS calibration graph of the predicted versus actual EA
concentration
for the three-component PLS calibration model with sodium chloride added as an
interference;
CA 02305201 2000-04-04
WO 99/24815 PCT/CA98/01024
FIG. 10 is a PLS calibration graph of the predicted versus actual sodium
carbonate concentration for the three-component PLS calibration model with
sodium
chloride added as an interference;
FIG. 11 is a PLS calibration graph of the predicted versus actual sodium
sulfide
5 concentration for the three-component PLS calibration model with sodium
chloride added
as an interference;
FIG. 12 is a diagrammatic view of sensing apparatus of use in the practice of
the
present invention;
FIG. 13 is a plot of the concentration of white liquor being fed into the B
digester
10 at the Bowater, Inc. kraft pulp mill in Thunder Bay, Ontario, over a period
of
approximately nineteen days, as measured by FT-IR and by manual titration;
FIG. 14 is a plot of the concentration of white liquor, upper circulation
black
liquor, lower circulation black liquor, and extraction zone black liquor at
the Bowater,
Inc. kraft pulp mill in Thunder Bay, Ontario, over a period of approximately
four days, as
15 measured by FT-IR and manual titration.
FIG.15 is a calibration graph concenaing effective alkali;
FIG.16 is a calibration graph concerning organic solids; and
FIG.17 is a calibration graph concerning total solids.
DESCRIPTION OF PREFERRED EMBODI1VVIENTS
FIG. 1 is a diagrammatic view of a recovery system, complete with sensing
apparatus, according to one embodiment of the present invention. The sensing
apparatus
shown in FIG.12 is further described, hereinafter.
Referring to FIG.1, weak black liquor recovered from the digestion process 10
may be temporarily stored in a weak black liquor storage tank 12 before being
concentrated through multiple-effect evaporators 14 to form strong black
liquor which is
stored in a strong black liquor storage tank 16. Line 18 delivers the strong
black liquor
from the strong black liquor storage tank 16 to the recovery furnace ZO to
generate flue
gases 22 and smelt 24. The smelt 24 flows to the smelt dissolving tank 26 to
form green
fiquor. Green liquor samples are taken at sample withdrawal point 28 in line
30 leading
to the green liquor clarifies 32. The samples are fed through a 1.25 cm
diameter conduit
I i1 l
CA 02305201 2002-09-09
WO 99/24815 PCTICA98/01024
16
34, optionally merged with other optional sample streams 36, 38, 40, 42 andlor
44,
through either a transmittance-mode or a reflectance-mode flow-cell 46, well-
known in
the art. Infrared light from an infrared source which is integral to a Fourier
transform
spectrometer 48 is brought to the flow-cell 46 by means of a direct optical
coupling with
mirrors or by a fiber optic cable 50. Some of the infrared light is absorbed
by the liquor
and the residual light is returned to the Fourier transform spectrometer by
means of either
a direct optical coupling with mirrors or by a second fiber optic cable 50.
The
spectrometer 48 records the near-infrared single-beam spectrum of the liquor.
Readings
from the spectrometer 48 are transferred to a computer 52 which calculates the
individual
component concentrations of the liquor, such as, sodium hydroxide, sodium
sulfide,
sodium carbonate, and optionally, sodium chloride with the use of a PLS
multicomponent
calibration model. The concentration parameters of conversion eff ciency
and/or
causticity and/or total titratable alkali (TTA) are calculated from said
concentrations
automatically by the computer 52.
The concentration parameter of TTA is used to automatically control the flow
of
weak wash 54 entering the smelt dissolving tank so as to obtain an optimal
value of TTA
in the unclarif ed green liquor leaving the smelt dissolving tank 2fi through
flow line 30
which transports said liquor to the green liquor clarifies 32.
Liquor in line 56 flows from the green liquor clarifies 32 and enters the
slakes 58
where a variable quantity of calcium oxide is added through line 60 to form
calcium
hydroxide. Trim weak wash 62 is added to line 56 immediately before sample
withdrawal point 64 which transfers a sample through line 44 to the flow cell
46 for
analysis. The concentration parameter TTA is calculated by the computer 52 and
used as
feedback control of the trim weak wash line 62 flow rate, and/or feedforward
control of
the calcium oxide line feed rate 60 to the slakes 58.
Upon leaving the slakes, the liquor flows through a series of three or more
recausticizers 66 which allow most of the sodium carbonate to react with the
calcium
hydroxide to form sodium hydroxide and calcium carbonate. The resulting
suspension
then proceeds to the white liquor clarifies 68. The partially recausticized
white liquor is
sampled from withdrawal point 70 andlor 72 where it is delivered to the flow
cell 46
where the concentrations of sodium hydroxide, sodium sulfide, sodium
carbonate, and
i
CA 02305201 2002-09-09
WO 99/24815 PCT/CA98/01024
17
optionally, sodium chloride, are simultaneously determined. The concentration
parameter of causticity is calculated from these values and used as fast
feedback control
of the feed rate of calcium oxide to the slaker through line 60 if withdrawal
point 70 is
used or slow feedback control of said feed rate if withdrawal point 72 is
used. The
clarified white liquor leaves the white liquor clarifier 68 and flows to the
white liquor
storage tank 74 where it is ready for use in the digestion process through
Line 76. If the
retention time of the white liquor ciarifier 68 is sufficiently short, as in
the case of
pressure or disk filters used for clarifying, withdrawal point 78 may be used
in place of
withdrawal point 72.
FIG. 2, shows a diagrammatic representation of a continuous type Kamy~digester
and of a control system as embodied by the invention. This control system may
be used
to monitor the effective alkali (EA) consumption during the impregnation and
cooking
stages of a continuous cooking pulping operation. EA is a concentration
parameter
defined as the sodium hydroxide plus half of the sodium sulfide (expressed as
Na2(?)
present in a mill liquor. Refezring to FIG. 2, a digester 80 is shown with a
white liquor
supply line 82 from the white liquor storage tank (not shown). The liquor in
the digester
80 is indirectly heated through a transfer line by high pressure steam
supplied through a
steam supply line 84. Black liquor is withdrawn from the digester 80 through
the upper
circulation screen 86 and then sent through an upper heater 88 using a
recirculating loop
90. A second steam Line 92 provides steam to a second recirculation loop 94 in
which the
liquor is withdrawn from the digester 80 through the lower circulation screen
96 and sent
to a lower heater 98.
Chips are fed to the digester 80 through line 100. Samples from the digester
are
withdrawn from the extraction liquor Line 102 at withdrawal point 104. For
other tests,
samples are withdrawn from the sample point 106 in the upper beater loop,
sample point
108 in the lower heater loop, and sample point 110 in the white liquor supply
line 82.
The samples are fed individually through I.25 cm conduits by a means of
valves, and
merged with each other before flowing through either a transmittance-mode or a
reflectance-mode flow-cell 46, for which either mode is well-known in the art.
Infrared
light from an infrared source which is integral to a Fourier transform
spectrometer 48 is
brought to the flow-cell 46 by means of a direct optical coupling with mirrors
or by a
* Trade-mark
I I
CA 02305201 2002-09-09
WO 99/24815 PCT/CA98/01024
18
fiber optic cable 50. Some of the infrared light is absorbed by the liquor and
the residual
light is returned to the Fourier transform spectrometer by means of either a
direct optical
coupling with mirrors or by a second fiber optic cable 50. The spectrometer 48
records
the near-infrared single-beam spectrum of the liquor. Readings from the
spectrometer 48
are transferred to a computer 52 which determines the EA and sulfidity of the
white
liquor, and the EA and total organic content of the black liquor with the use
of a PLS
multicomponent calibration model. The white liquor EA is used to control the
ratio of
EA to wood in the digester by adjusting the feed rate of white liquor. Black
liquor EA is
used to ensure that the residual EA present in the cook zones is sufficient to
ensure
dissolution of the lignin present in wood chips while not exceeding a lower
set-point and
is achieved by adjusting the EA to wood ratio. White liquor sulfidity, black
liquor EA
and total organic content are used as a feedforward signal for kappa or k-
number control
by adjustment of the cooking conditions, such as temperature and time, of the
digester.
This can be done by adjusting the production rate and the temperature of the
upper and/or
lower circulation heaters 88 and 98, respectively. The extraction liquor flows
through
Line 102 to the flash tanks (not shown) on its way to the recovery cycle.
Digested wood
chips exit through the blow line 112 to the blow tank (not shown) before
entering the
brownstock washing stage.
Fig. 12 shows the interface between the liquor sample and the Fourier
transform
spectrophotometer (e.g., Bomem; Hartmann and Braun, WorkIR 160) in greater
detail. A
beam of infrared light 114 leaves the infrared source 116 within the Fourier
transform
spectrometer, 48 and enters an interferometer 118. Light 120 leaving the
interferometer
118 enters an optional fiber-optic extension accessory 122 which includes (l)
an entrance
lens which concentrates the wide incoming beam (perhaps 30 mm) down onto the
0.6 mm
diameter fiber, (ii) a variable length of fiber-optic cable (as much as 300 m
or more), and
(iii) an exit lens which expands the narrow beam of the fiber back to a wide
beam of
similar width to the incoming beam. The spectrometer may also be coupled
directly to
the transmission cell over relatively short distances by eliminating the fiber-
optic
extension accessory. The beam of infrared light 124 leaving the exit lens of
the fiber-
optic extension accessory is focussed through the 316 stainless steel
transmission cell 126
by parabolic mirror 128. The beam 130 passes through two caustic-resistant
windows
132 (e.g. Harrick Scientific, BK-7) which contain the flowing or static liquor
in the
* Trade-mark
i ~~
CA 02305201 2002-09-09
WO 99/24815 PCTICA98/01024
19
transmission cell 126. The liquor arrives in and leaves from the transmission
cell via 316
stainless steel sample conduit 134. The infrared beam 136 is then redirected
back into the
spectrometer and onto the germanium (Ge) detector 138 via route 140 and 142
with the
option of extending this distance with the fiber-optic extension accessory 144
in a similar
way that the beam 120 leaving the interferometer 118 was extended. After a
complete
scan of the wavelength region of interest, the spectrometer transfers the
resulting
interferogram to an acquisition card located in an IBM-compatible personal
computer 52
via serial cable 146. The spectrum can then be computed by the acquisition
card and
several spectra (e.g. 128) can be co-added by the computer software. The
resulting
averaged spectrum can then be used to calculate the individual component
concentrations
of the liquor such as sodium hydroxide, sodium sulfide, sodium carbonate, and
optionally, sodium chloride with the use of a PLS mufti-component calibration
model.
The concentration parameters of conversion efficiency and/or causticity and/or
total
titratable alkali (TTA) are calculated from said concentrations automatically
by the
computer.
EXAMPLE 1
A three-component PLS calibration was performed on the set of synthetic
samples
listed in Table I for the purpose of building a calibration model that is
capable of
predicting 1) effective alkali concentrations 2) sodium sulfide concentrations
and 3)
sodium carbonate concentrations. The spectral region chosen for building the
model was
from 11000 to 7300 wavenumbers (cm t) for all three components. The
calibration graphs
are shown in FIG. 4 (effective alkali), FIG. 5 (carbonate) and FiG. 6
(sulfidity), all of
which demonstrate good agreement between predicted and actual values. The
standard
deviation of the differences between the actual and predicted values are (all
in g/L as
Na20) 0.34 for effective alkali, 1.0 for sulfidity, and 1.1 for carbonate.
From the
predicted concentrations shown herein, it is possible to calculate TTA, %
sulfidity, and
causticity for purposes of control.
* Trade-mark
CA 02305201 2000-04-04
WO 99/4815 PCT/CA98/01024
Table I
Compositions
of synthetic
liquor samples
used for the
three-component
PLS Calibration
Sample Effective AlkaliSodium Sulfide Sadium Carbonate
No. (g/L as Na20) (g/L as Na20) (g/L as Na20)
1 100.2 0 0
2 5.2 0 0
3 102.0 24.6 0
4 103.5 56.8 0
5 101.0 0 42.5
6 100.2 0 82.8
7 100.9 50.9 21.8
8 20.2 40.7 0
79.9 28.3 11.0
10 81.0 29.1 21.2
11 81.9 29.1 31.6
12 81.0 8.5 16.4
13 80.8 16.6 16.3
14 81.1 28.7 15.8
15 81.3 41.1 15.9
16 20.0 0 0
17 81.8 - 0 16.7
EXAMPLE 2
5
The absorbance spectra of samples consisting of various dilutions of a black
liquor sample are shown in FIG. 7. There is clearly a strong correlation
between the
dilution of the black liquor and the absorbance in the region between
wavenumbers
12000 to 9000 (cm'). A calibration graph is shown in FIG. 8 based on the
absorbance at
10 11500 wavenumbers (cni'). The trend is slightly non-linear, and a good fit
is shown by
the second order polynomial trendline.
CA 02305201 2000-04-04
WO 99/24815 PCT/CA98101024
21
EXAMPLE 3
The accuracy of the PLS model calibrated for EA, sodium sulfide, and sodium
carbonate concentrations was investigated to see how it was affected by
varying sodium
chloride concentrations from 0 to 40 g/L (as NaCI). Synthetic solutions were
made up of
fixed concentrations of EA, sodium sulfide, sodium carbonate, and varying
concentrations of sodium chloride. The concentrations of all the components
except
sodium chloride were included in the model, which was generated from the
samples in
Table I (all of which contained no sodium chloride) and Table II
(concentrations as
shown). The model still accurately predicts EA (shown in FIG. 9), sodium
carbonate
{shown in FIG. 10), and sodium sulfide (shown in FIG. 11) for solutions
regardless of
sodium chloride concentration.
Tabte II
Compositions
of synthetic
liquor samples
added to
three-component
PLS Calibration
Sample Effective Sodium SulfideSodium Sodium
No. Alkali
(g/L as Na20)Carbonate C~oride
(g/L as Na20) (g/L as NaCI)
(glL as Na20)
18 79.9 28.3 11.0 0
19 79.9 28.3 11.0 10
79.9 28.3 11.0 20
21 79.9 28.3 11.0 30
22 79.9 28.3 11.0 40
From the above examples it can be seen that different types of process liquors
in
20 the cellulosic pulp manufacturing process can be analyzed and that
concentration
parameters can be simultaneously determined with the use of various types of
partial least
squares (PLS) multivariate calibration which correlate the spectral behavior
for different
concentrations of each chemical component in a calibration sample with their
actual
concentration in that sample. The set of correlations represents a model which
can then
be used to predict the concentration parameters of an unknown sample.
Consequently, by
varying at least one process variable, the process can be controlled so that
optimal
production of desired product is obtained.
CA 02305201 2000-04-04
wo ~n4sis rcTicA9sioioi4
22
EXAMPLE 4
A multi-component PLS model was generated for white liquor using as many as
278 near ii>frared absorbance spectra of synthetic and real white liquor
samples in the
calibration training set. These training samples included variations in the
concentration
of EA, sulphide, carbonate, and chloride, as well as variations in the
temperature of the
sample liquor and the reference water. This model was applied to spectra
collected by an
on-line FT-IR spectrometer (Bomem, Hartmann & Braun, Workir 160) at the
Bowater,
Inc. kraft pulp mill in Thunder Bay, Ontario. FIG.13 is a plot of the EA
concentration of
white liquor being fed into the B digester at this mill over a period of
approximately
nineteen days, as measured by FT-IR and by manual titration with hydrochloric
acid.
A one-component PLS model was generated for black liquor using as many as
457 near infrared absorbance spectra of synthetic and real white and black
liquor samples
in the calibration training set. FIG. 14 is a plot of the concentration of
white liquor,
upper circulation black liquor, lower circulation black liquor, and extraction
zone black
liquor at the Bowater, Inc. Itraft pulp mill in Thunder Bay, Ontario. Data is
shown for a
period of approximately four days, as measured by FT-IR and by manual
titration with
hydrochloric acid. A shorter time period is presented for graphical clarity.
Manual
titration data is only collected by the mill personnel for EA on white liquor
and lower
circulation black liquor. This example demonstrates ( I ) long term
correlation with
manual titration results, (2) no instrumental drift, (3) no optical
degradation, (4) accurate
measurement in the presence of gaseous bubbles and solids, and (5) no plugging
of the
flow cell by solids or fibres since a large pathlength flow cell was used (8
mm) as stated
in the present invention.
Thus, a rapid method is provided for the control of a cellulosic pulp
manufacturing process via on-line measurement of chemical concentration
parameters in
process liquor streams with near infi~ared radiation. The method eliminates
the need for
(i) manual sampling, (ii) frequent equipment maintenance, (iii) a dedicated
instrument at
each sampling point, (iv) compensation for instrumental drift, and (v) an
environmentally
controlled spectrometer housing near the sampling location(s). The method
includes the
steps of (i) withdrawing samples of a process liquor stream from a cellulosic
pulp
manufacturing process, (ii) subjecting the samples to near-infrared
spectrophotometry
over a predetermined range of wavenumbers so as to produce spectral
measurements
I ~I I
CA 02305201 2002-09-09
WO 99/24815 PCT/CA98/010Z4
23
which determine the concentrations of different combinations of chemical
components,
{iii) correlating by multivariate calibration the relationships between the
spectral
measurements of unknown samples and the spectral variations shown by different
combinations of chemical components of the process liquor so that
concentration
parameters can be accurately determined for typical levels of chemical
components
present in the process liquor, and (iv) controlling at least one process
parameter so as to
obtain optimal operation of the ceilulosic pulp manufacturing process.
EXAMPLE 5
A three-component PLS calibration was performed on the infrared spectra of a
set
of nineteen black liquors collected from several kraft pulp mills across
Canada. A
calibration model was constructed that is capable of predicting ( 1 )
effective alkali (EA)
concentrations, (2) organic solids content and (3) total solids content. Table
III lists the
concentrations of the effective alkali (g/L as Na20), organic solids (w/w %),
and total
solids (w/w %) content of ~ese black liquor samples. The EA was determined by
automatic titration with 1.00 N HCl to an endpoint determined by the
inflection of a pH
versus volume of acid added curve between pH 11.0 and 11.5, in the presence of
0.1 M
Na2C03. The total solids content was determined gravimetrically by drying
25.00 mL of
the black liquor sample to a constant weight in a drying oven at 105 ~
2°C. the organic
solids content was also determined gravimetrically by subtracting the mass
obtained by
igniting to a constant weight the remaining dried solids at 550 ~ 25°C
from the total
solids content. The spectra were measured at a constant temperature of
30°C using a
pathlength of 8 mm. The spectral region chosen for building the model was from
11533
to 7382 wavenumbers (crri l) for all three components. A pre-processing step
of
calculating a second derivative function with a 31-point Savitzky-Golay
smoothing
procedure was performed on the spectra prior to running the calibration. A
total of three
PLS factors were used for the predictions. The calibration graphs are shown in
FIG. 15
(effective alkali), FIG. 16 (organic solids) and FIG.17 (total solids), all of
which
demonstrate good agreement between the FT-IR and the reference method values.
Since
total solids content is equal to the sum of the organic solids content and the
inorganic
solids content, the inorganic solids content can be calculated by determining
the values of
CA 02305201 2000-04-04
WO 99/24815 PCT/CA98/OlOZ4
24
the organic and the total solids contents from the liquor. Fmm these results,
it is possible
to calculate effective alkali, organic solids, inorganic solids, and total
solids content.
TABLE III
Compositions of mill black liquor samples used for the three-component PLS
calibration
Sample No. Effective Alkali Organic Solids Total Solids
(g/L as Na20) (w/w %) (w/w/ %)
1 0.3 8.6 17.2
2 20.2 S.1 15.6
3 21.3 5.7 16.4
5.4 6.4 14
2
5 8 8.3 .
16.2
6 7'9 8~ 1 16.3
7 19.6 6.1 17.7
8 4.7 7.7 15.4
20.2 3.9 13
9
10 4.8 6.1 .
12.7
11 17.2 6.1 16.1
12 0.7 8.5 16.8
13 9.8 12.8 23.6
14 10.4 11.0 22.3
15 15.1 5.6 13.8
16 6.4 10.4 19.6
17 14.2 6.5 16.0
18 8.7 7.8 15.0
19 19.7 4.2 14.1
Although this disclosure has described and illustrated certain preferred
embodiments of the invention, it is to be understood that the invention is not
restricted to
those particular embodiments. Rather, the invention includes all embodiments
which are
functional or mechanical equivalents of the specific embodiments and features
that have
been described and illustrated.
