mostly bound in lignin. Nitrogen-containing organic
compounds, like several amino acids, pyrroles,
pyridines, pyrimidines, and indoles have been found in
black liquors [1,3,4]. In addition, Martin et al. [4] have
reported a significant amount of nitrogen (up to one
third of the total black liquor nitrogen), to be in the
inorganic form, mostly as nitrate, in a southern pine
black liquor. On the other hand, Martin and Malcolm
[5] have found later, based on an analysis of lignin that
the liquor nitrogen is mainly associated with the lignin,
further suggesting little to no nitrogen to be associated
with the inorganics. In these studies on the liquor
nitrogen composition it has not been clearly mentioned
whether a weak or a strong black liquor has been
analyzed or, further, if the liquors analyzed were
oxidized or not.
THE FATE OF NITROGEN IN THE CHEMICAL
RECOVERY PROCESS IN A KRAFT PULP
MILL: Part 1: A general view
M.Kymäläinen, M.Forssén and M.Hupa
Åbo Akademi University, Turku, Finland
Published in “Journal of Pulp and Paper Science 25
(12): 410-417, 1999
ABSTRACT
About a third of the black liquor nitrogen remains in
the smelt in a recovery boiler and continues in the rest
of the recovery cycle. The smelt nitrogen was found to
be in a form which slowly converted to ammonia
under the process conditions: the green liquor entering
the slaker contained clearly more ammonia than the
green liquor leaving the dissolver , about 70 % and 20
% of their total nitrogen, respectively. The nitrogen of
weak wash entering the smelt dissolver was mostly in
the form of NH3, and at least part of it was released in
the vent stack gases of the dissolver. The green liquor
nitrogen, originating both in smelt and weak wash,
corresponded to almost 40 % of the black liquor
nitrogen. The major part, about 60 %, of the green
liquor nitrogen entered the cooking process with the
white liquor. A minor part, less than 20 %, continued
to the weak wash, thus returning to the smelt dissolver.
The rest, about 20 % of the green liquor nitrogen, was
assumed to exit the process as gaseous NH3 around the
slaker and the causticizers. The nitrogen in the white
liquor was mostly NH3. It was released mainly during
the black liquor evaporation, and to a minor extent
during the cooking. This NH3 was mostly dissolved
into the foul and secondary condensates. The stripping
of the foul condensates was the main exit point for
ammonia in the recovery process. The amount of this
ammonia represented about two thirds of a typical
NOx emission of a recovery boiler.
The behavior of the liquor nitrogen in the recovery
boiler has been extensively studied during recent years,
by laboratory tests and NOx measurements in the
boiler flue gases, as well as by kinetic modelling. In
previous experimental work at Åbo Akademi
University (ÅAU) by Aho et al. [2] and by Forssén et
al. [6], the fate of nitrogen during black liquor burning
has been clarified and the key routes for nitrogen in
black liquor combustion have been suggested as shown
in Figure 1 [6]. These studies showed that about one
third of the liquor nitrogen will be bound in the char
residue and may remain in the salt residue leaving the
furnace along with the smelt. The present work is a
continuation of the research at ÅAU on the behavior of
black liquor nitrogen. In this work the main focus was
to elucidate the fate of the smelt nitrogen further on in
the recovery process.
N2
Npyrolysis
Nblack liquor
NTRODUCTION
Nchar
Reactive nitrogen enters the pulping process and the
chemical recovery system of a kraft pulp mill mainly
with the wood chips. During digesting the wood
nitrogen has been found to transfer primarily to black
liquor [1]. The nitrogen content of black liquor is
typically 0.05-0.15 % by weight, and it varies
somewhat depending on wood species, being slightly
higher in birch than in pine liquors [2]. There is hardly
any quantitative data on the form of nitrogen in black
liquor and the chemical behaviour of it is therefore not
fully understood. The nitrogen has been found to
occur mostly in organic compounds, both in straight
chains and in heterocyclic ring structures, presumably
NH3
NO
N2
NO
Nsmelt
NO
Ngreen liquor
Figure 1. Suggested fuel nitrogen pathways in black
liquor combustion [6].
Prior to the present work, little information has been
published concerning the nitrogen elsewhere in the
recovery cycle than in black liquor burning. In general,
the odour of NH3 has been noticed in the causticizing
plant in many mills, but little work has been done to
explain these emissions. Some recent studies by
4.6-1
collected into plastic containers (0.5-1 liter), which
were filled to the top and sealed immediately after
sampling. Smelt samples from the recovery boiler
smelt spouts were taken in two different ways: either
with a steel mug or with a quartz glass tube. When
sampling with a mug (about 100 cm3), the sample was
wrapped in aluminum foil and kept in a plastic bag.
The sample for the analysis was taken from the middle
of the original sample piece. Contact of the smelt
samples with air was minimized, and an unoxidized
part of the sample piece was used for the analysis.
Tarpey et al. [7,8] have focused on the cause of the
ammonia release, especially in particulates, from a
smelt dissolving tank. Tarpey [7] has also published
nitrogen and ammonia analysis results of several liquor
samples from the recausticizing plant, but no
explanation for these results nor any mass balance for
nitrogen has been given. On the other hand, Thompson
et al. [9] have reported, based on a mass balance of a
recovery furnace, that only about 9 % of the black
liquor nitrogen entering the boiler leaves with the
smelt. This amount of smelt nitrogen was found to be
far too low to explain the nitrogen content of the green
liquor leaving the smelt dissolver, and the reason for
this large difference was unknown.
Most of the samples were analyzed immediately at the
mill. Some of the samples were also stored for a
certain period of time at room temperature and
analyzed in order to compare the results with the
results obtained from the samples analyzed promptly.
The sample containers were kept closed during storage.
The time elapsed between sampling and analysis
varied from a few hours to a few months. As shown in
Table 1, the Kjeldahl nitrogen content of the liquors
was found to remain constant during storage, within
the accuracy limits of the method of analysis.
The purpose of the present work was to clarify the fate
of the smelt nitrogen in green liquor and in the rest of
the recovery cycle. It was of particular interest to find
out in which forms nitrogen is found in the process and
in which forms it finally leaves the recovery cycle. The
potential main exit points for nitrogen in the process
were also of interest.
This paper is the first part in a series of papers
concerning the fate of nitrogen in the chemical
recovery process. A general view of the nitrogen flows
in one kraft pulp mill is given in this paper, and in the
second part the effect of some process variables on the
fate of nitrogen in another pulp mill is described.
Table 1. Storage of two liquor samples, and their
Kjeldahl nitrogen content.
Storage time
Kjeldahl
before analysis
nitrogen
mg/l
A few hours
108
Green Liquor
3 months
109
A few hours
84
White Liquor
6 weeks
82
EXPERIMENTAL
Sampling
The measurements were conducted at a Finnish kraft
pulp mill producing 2000 tons pulp/day. The mill
utilizes a batch digesting process, and both hardwood
and softwood kraft pulps are produced.
The purpose of the sampling campaign during 1997
was to study possible variations of the nitrogen level in
the recovery cycle during different sampling periods
(Table 2). Here, the effects of varying process
conditions were not studied in detail, but that has been
studied during one sampling campaign in 1998 and
will be reported in our second paper of this work.
A simplified general flow sheet of the recovery
process, excluding lime mud dewatering and calcining,
is shown in Figure 2. The recovery cycle consists of a
one-line evaporation and a causticizing plant, a
recovery boiler with a design capacity of 2800 tds/d.
Noncondensible Gases (NCG) from the recovery cycle
are collected and burned: high concentration, low
volume (HCLV) gases from digesting and evaporation,
including stripper gas as well as methanol are burned
in a separate incinerator. Low concentration, high
volume (LCHV) gases from the evaporation plant are
also conducted there, whereas LCHV gases from the
causticizing plant are burned in the lime kiln.
Table 2. Sampling periods.
Period Month Boiler load
Hardwood
#
in '97
tds/d
% in cooking
1
Jan.
2610
63
2
Feb.
2550*)
60
3
March
2560
60
4
Sep.
2760
60
5
Oct.
2110
100
*) in the beginning of the period, 2120 tds/d
The main sampling points (1-16) are shown in Figure 2,
and the corresponding process streams are specified in
Table 3 in the Results section. Liquid samples, like
black, green and white liquors, and condensates, were
4.6-2
Analyses
Some details and comparisons of different analysis
methods for total nitrogen are provided in our earlier
The nitrogen of the foul condensate, a combined
stream from evaporators and batch digesters to the
stripper, was solely in the form of ammonia. This
ammonia was efficiently removed through steam
stripping of the condensate, resulting in a treated
condensate free of nitrogen. The stripper gases,
including ammonia, are taken to the methanol
condenser.
papers [6,10]. In the present work, the Kjeldahl
method was used because the appropriate analysis
facilities were available at the mill, thus making it
possible to analyze the samples directly after the
sampling. Initial tests showed that the Kjeldahl method
gave analysis results comparable with the other
methods [10]. The reproducibility of the Kjeldahl
method determined by duplicate analyses for the
studied samples was found to be good, within some 4
%.
Analysis results of weak and strong black liquor, smelt,
and green liquor samples taken during different
sampling periods are given in Figure 4. Every result
here is an average value of two or more parallel
determinations. Minor variation in the sample nitrogen
content from one sampling period to another can be
seen.
The ammonia (NH3) content was determined by
titration, where the NH3 is first liberated from
alkalized samples by steam distillation and then
absorbed into an acid solution followed by titration for
the NH3.
A conversion of the nitrogen in green liquor into the
form of ammonia is evident (Figure 3). So far, the
origin of this ammonia has been unknown. In this
work, laboratory measurements were done to clarify
the formation of NH3 in green liquor. First, the
temperature dependence of the NH3 formation in mill
green liquor was studied. One result of those tests is
presented below. Secondly, the key reactions
responsible for ammonia formation were studied by
studying the NH3 formation with well controlled
additions of certain nitrogen compounds to both mill
and synthetic green liquor. Those results and the
description of the batch reactor system used in the
experiments will be published in near future.
The dry solids content of black liquor was determined
by weighing the sample before and after drying with
sand at 105°C for 18 to 24 hours (SCAN-N 22:77).
RESULTS AND DISCUSSION
Kjeldahl Nitrogen and Ammonia Results
Kjeldahl nitrogen results of the samples (Figure 2)
taken during one sampling period are given in Table 3,
to provide data for mass balance calculations. In
addition, ammonia results and the ratio of ammonia
nitrogen to Kjeldahl nitrogen are given. The results are
also shown in Figure 3.
Ammonia Formation in Green Liquor
Hardly any ammonia nitrogen was found in the smelt
samples (Figure 3). The nitrogen of weak wash coming
to the smelt dissolver was mostly in the form of
ammonia. The green liquor entering the slaker
contained clearly more ammonia than the green liquor
leaving the dissolver, about 70 % and 20 % of the
Kjeldahl nitrogen, respectively. Based on these data, it
seems evident that the nitrogen coming with the smelt
to the dissolver and continuing further with the green
liquor to the causticizing converts to ammonia in the
green liquor. Conversion takes place mainly during
green liquor processing, in surge tanks and in clarifiers.
Hardly any conversion takes place in the dissolver. A
few laboratory tests, briefly described below, also
support these findings.
Strong black liquor was practically free from ammonia.
Weak black liquor, instead, contained a significant
amount of ammonia. This indicates that the ammonia
nitrogen of weak black liquor volatilizes during the
evaporation process.
4.6-3
A mill green liquor (the sample taken right after the
dissolver) was kept at a certain temperature in a
laboratory reactor and the NH3 concentration of the
liquor was followed as a function of time (Figure 5a).
The formation rate of NH3 in the studied mill green
liquor was found to be fairly slow: a residence time of
several hours at the process conditions is required for
this green liquor to increase its ammonia content to
correspond to the content of the green liquor entering
the slaker, i.e. around 70 mgN/l. Furthermore, as
shown in Figure 5b, the NH3 formation reaction is
suggested to be a first order reaction with respect to
the NH3-forming compound, based on the following
rate data analysis and the following assumptions:
- Compound A is the only NH3-forming compound
in the liquor, and the initial concentration of A
equals the difference between the total nitrogen
and the NH3 nitrogen.
- Compound A selectively forms NH3 as a sole
nitrogen-containing product. In other words, rA =
rNH3 = k CA
tank and the clarifier, whereas it was around 700 in the
vent gas from the slaker. Because of the difficulties
encountered in gas flow measurements, no reliable
mass balance data could be obtained. Thus, all the
nitrogen flow values of the vent gases presented in
Figure 6 are estimated based on the balance around
each process unit.
Based on a literature survey done in this work, one
potential nitrogen compound which forms ammonia in
alkaline solutions is a cyanate (OCN-) ion, the NH3
formation reaction kinetics of which has been studied,
for instance, by Jensen [11]. Our comparison between
experiments measuring NH3 formation rates in liquors
containing no aditional OCN- and in liquors containing
additional OCN- supports that the NH3-forming ion in
the green liquor would be OCN-. These experimental
results will be published in near future.
Ratio of Nitrogen to Sodium in Various Liquors
The molar ratio of nitrogen to sodium (N/Na) in four
liquors is presented in Table 4. The ratio of N/Na in
the input stream to the causticizing plant (green liquor)
was clearly higher compared to the ratio in the white
liquor leaving the causticizing plant. (The input of
sodium with lime at the causticizing was, as calculated,
negligible.) This means that nitrogen must leave the
liquor more readily than sodium in the causticizing
plant (during slaking and causticizing), the loss, most
likely NH3, being about 15 % of the nitrogen entering
the slaker.
Weak wash, instead, clearly had the highest N/Na ratio.
This was caused by the nitrogen content of the
secondary condensate used for the lime mud washing.
Nitrogen Mass Balance
The mass balance for nitrogen around the chemical
recovery cycle was performed on the basis of the
Kjeldahl analysis results shown above in Table 3 and
the available flow data. Most of the flow data is based
on the process measurements and/or were, in addition,
calculated based on the density or active alkali content
values of the liquor streams. The smelt flow to the
dissolver was estimated by assuming that all sodium
and sulfur in virgin black liquor, 19.0 and 5.4 weight
%, respectively, end up in the smelt, the sulfur
reduction being 96 %. Here, the load to the recovery
boiler was 32 kgds/s, which corresponds to the green
liquor flow of 98 l/s to the causticizing plant. In Figure
6, the amount of nitrogen is expressed as grams per
second. Values estimated but not measured are
marked with an asterisk (*).
Nitrogen enters the recovery cycle with the wood chips.
The amount of the wood chip nitrogen was estimated
based on the balance around the digester system,
assuming that the vent gas nitrogen flow equals the
amount of the nitrogen found in the condensate from
the turpentine recovery. The major part of the white
liquor nitrogen entering the digester is in the form of
ammonia, representing a nitrogen flow of 5.1 gN/s (84
% of its total nitrogen flow). About 20% of the weak
black liquor nitrogen was found to occur in ammonia,
representing a nitrogen flow of 6.2 gN/s. This suggests
that most of the weak black liquor ammonia nitrogen
originates in the white liquor, but some ammonia
nitrogen also originates in the secondary condensate
and the water used for pulp washing. The ammonia
nitrogen of weak black liquor was found to be released
mostly at the beginning of evaporation and it probably
ended up in the condensates, mostly in the foul
condensate, which, in turn, was treated by stripping,
thus releasing the ammonia to the gas phase which,
further on, was subject to methanol stripping.
The stripped condensate and the condensate from the
last stages of the evaporator were combined and
further used for lime mud washing, thus introducing
some ammonia nitrogen to the weak wash. The
secondary condensate from the first stages of the
evaporator, which is “the cleaner side” of the
evaporator, was collected separately from the dirtier
condensates and further used for the pulp washing,
thus introducing a minor quantity of ammonia to the
weak black liquor.
According to these measurements, about 25 % of the
black liquor nitrogen entering the recovery boiler
leaves with the smelt. However, based on the mass
balance around the smelt dissolver, the smelt nitrogen
entering the dissolver was not quite enough to explain
the amount of the green liquor nitrogen leaving the
dissolver. The smelt nitrogen should have been about
one third of the black liquor nitrogen to have a balance
around the smelt dissolver. So far, the reasons for this
difference are unknown, but one explanation could be
that the large variation in the nitrogen content that was
found between different smelt sample pieces. This is
under further investigation.
During one sampling period, a few vent gas
measurements for detection of ammonia were made by
absorbing NH3 (both gaseous and particulates) into an
acid impinger solution (according to the standard
method VDI 2461). The NH3 concentrations were very
high in all measurements: The NH3 content, given
mgNH3/m3n (dry gas), varied between 300-700 in the
vent gases from the smelt dissolver, between 30004000 in the combined vent gas from the green liquor
4.6-4
NH3 may escape from the liquor phase, depending on
the temperature and openness of the process step. This
study suggested, however, that quite vigorous and
thorough treatment of the liquor, for instance with
steam, is needed to strip out all the NH3 from the
liquor. Only the evaporation step and steam stripping
seemed to be effective ways to release all the aqueous
NH3.
The vent stack gas from the dissolver was first
scrubbed with weak wash, then led to the flue gas
scrubber of the recovery boiler, which was operated at
a pH of about 7.1. The scrubber solution continued to
the dissolver, but its input to the nitrogen mass balance
around the dissolver was, however, found to be quite
insignificant. The significance of the nitrogen flow of
the scrubber solution from the NCG incinerator was
even less.
Once the ammonia is stripped from the foul
condensate, it may either stay in the gas phase or
dissolve into water/methanol during the methanol
condensing stage. Supposing that there is no methanol
condensing stage, i.e., the ammonia stays in the gas
phase, it is interesting to examine the fate of this
ammonia. In this case, the ammonia follows the
HCLV gases to the incinerator. Assuming that the total
amount of the HCLV gases is around 40 m3n/ton pulp ,
calculations show the NH3 concentration to be about
one volume percent in these HCLV gases. After
burning the HCLV gases, this ammonia would result in
a NOx emission of around 1900 ppm(v)NO2 (dry f.g.,
O2=3%) assuming a 100 % conversion of NH3 to NOx.
Typical measured values of NOx in the flue gases of
the incinerator at the mill have been around 800-1500
ppm(v)NO2 (dry f.g., O2=3%), which indicates that the
conversion of the NH3 to NOx would have been around
40-80 %. The NOx values given here originate in some
earlier measurements during the years 1993-95. The
fact that some ammonia may also be absorbed by the
condensing methanol does not change these
conclusions, because in this case, both the methanol
and the HCLV gases were conducted to the same
incinerator.
The green liquor nitrogen leaving the dissolver
corresponded to almost 40 % of the black liquor
nitrogen introduced to the recovery boiler. Based on
the balance calculations as well as on the N/Na ratio
results, a significant loss of nitrogen was found to
occur around the slaker and the causticizers, the loss
being 15 -20 % of the green liquor nitrogen. This
indicates that some ammonia was released to the gas
phase, which in this case was led as LCHV gases to
the lime kiln. The major part, about 60 %, of the green
liquor nitrogen entered the cooking process with the
white liquor. A minor part, less than 20 %, continued
with the lime mud and further mostly with the weak
wash, and was thus returning to the smelt dissolver.
DISCUSSION
Forssén et al. [6] have recently suggested that the char
nitrogen fixed in the smelt would be the source of the
ammonia earlier reported to be present in the green
liquor and in the vent gases from the dissolver [7,8].
This assumption is still valid, and based on this work
it seems evident that the smelt nitrogen is in a form,
which slowly converts to ammonia under process
conditions. How the ammonia itself behaves under
varied process conditions is briefly discussed below.
CONCLUSIONS
In this paper, the term ammonia has been used to refer
to both ammonium ions (NH4+) and ammonia (NH3)
in a solution. An equilibrium exists between
ammonium ions and un-ionized or free ammonia,
depending on the pH of the solution, according to
Figure 7 (left). In addition to the equilibrium between
ammonium and ammonia in the solution, another
important equilibrium between aqueous and gaseous
NH3 is formed which, in turn, is very sensitive to
temperature. The temperature sensitivity of this
equilibrium can be seen on the right in Figure 8, which
shows the equilibrium concentration of the gaseous
NH3 above the solution containing various amount of
aqueous NH3.
The primary purpose of this work was to study the fate
of the smelt nitrogen in the green liquor and further
down the recovery cycle. Much effort was put on the
sampling at the pulp mill and on the analysis work. In
addition, laboratory measurements were done to study
the ammonia formation in green liquor in detail.
The present nitrogen analysis results of mill samples
support some findings of Tarpey [7] at another mill.
The total nitrogen concentration of several liquors, like
black liquor, green and white liquor and weak wash,
are very much at the same level in both studies. In this
work, the nitrogen mass balance of the recovery cycle
was closed for the first time, and some very interesting
findings were revealed.
Due to the high pH-value of most of the recovery
process liquors, the ammonia exists in these liquors as
aqueous NH3. Assuming that this NH3 is in
equilibrium with the gaseous NH3, various amounts of
The smelt nitrogen was found to be in a form which
slowly converted to ammonia under the process
conditions. The green liquor leaving the dissolver
4.6-5
Based on this work, an overall picture of the fate of
nitrogen in the chemical recovery process of the
studied pulp mill is presented in Figure 8. The
overview figure is based on the values given in Figure
6, and on the assumption that the missing part of the
green liquor nitrogen originates in the smelt, as
explained above. For the picture, a NOx release of
about 100 ppm(v) in the recovery boiler was assumed.
A value of 100 represents the nitrogen stream for the
black liquor introduced into the recovery boiler, so the
other numbers associated with the arrows indicate the
ratio of each nitrogen stream to that nitrogen input to
the boiler.
contained clearly less ammonia than the green liquor
entering the slaker, about
20 % and 70 % of their Kjeldahl nitrogen, respectively.
The total amount of the green liquor nitrogen,
originating both in the smelt and the weak wash,
corresponded to almost
40 % of the black liquor nitrogen. This means that
about one third of the black liquor nitrogen had to
remain in the smelt after burning. The major part,
about 60 %, of the green liquor nitrogen left the
recovery process and passed into the cooking process
along with the white liquor. The rest of the green
liquor nitrogen was partly released, preferably in the
form of NH3 in the causticizing plant, part of it
continued with the lime mud further into the weak
wash, and thus returning to the smelt dissolver.
The authors wish to emphasize that this work was
performed at one pulp mill only, and the results may
not be directly generalized and transposed on another
mill because of possible process variations between
the mills.
The white liquor ammonia was transferred into the
weak black liquor. It accounted for some 20 % of the
total nitrogen in the weak black liquor. This is based
on the assumption that no significant amount of
ammonia was formed from the wood chips during
cooking. This weak black liquor ammonia was found
to be released from the liquor in the evaporator plant
and it ended up mostly in the foul condensates. A
minor part was found in the secondary condensate
which was used for the lime mud washing. The
nitrogen in this flow thereby passed the recovery boiler
and mainly ended up in the weak wash. Otherwise all
the white liquor nitrogen, which once left the
causticizing never entered it again. So, in general,
nitrogen did not accumulate in the recovery cycle, but
it was effectively removed from the liquor system as
gaseous ammonia. The following potential exit points
for ammonia from the recovery process were revealed:
- First, a significant amount of nitrogen, in the form
of NH3, was bled from the cycle to HCLV gases at
the stripping of the foul condensates. In the mill
studied, the HCLV gases were treated by
condensing the methanol, after which the HCLV
gases as well as the methanol were burned in a
separate incinerator.
- Secondly, another significant NH3 release from
the liquor cycle was found to occur in the
causticizing plant, around the slaker and the
causticizers, to LCHV gases. The amount of the
NH3 release was about half of the above
mentioned release to the HCLV gases in the
stripper. In this mill, the LCHV gases from the
causticizing plant were burned in the lime kiln.
- In addition, a minor amount of NH3 was found to
be released in the smelt dissolver. It was led to the
flue gas scrubber of the recovery boiler.
ACKNOWLEDGEMENTS
The authors wish to extend their special thanks to the
mill staff for their cooperation. M.Sc. Hanna Malm is
acknowledged for doing the laboratory measurements.
This work is part of the Combustion and Gasification
Research Program LIEKKI 2 in Finland. The support
from the Finnish Recovery Boiler Committee,
Ahlstrom Machinery Oy, and Kvaerner Pulping Oy is
gratefully acknowledged.
REFERENCES
4.6-6
1.
VEVERKA, P., NICHOLS, K., HORTON, R.,
ADAMS, T., “On the Form of Nitrogen in Wood
and its Fate During Kraft Pulping”, 1993 TAPPI
Environmental Conference, Boston, MA, 777-780
(1993).
2.
AHO, K., “Nitrogen Oxides Formation in
Recovery Boilers”, Lic. Tech. Thesis, Åbo
Akademi University, 1994.
3.
NIEMELÄ, K., “Low-Molecular Weight Organic
Compounds in Birch Kraft Black Liquors”, Ph.D.
Dissertation,
Technical University of Helsinki,
1990.
4.
4. MARTIN, D., MALCOLM, E., HUPA, M.,
“The Effect of Fuel Composition on Nitrogen
Release During Black Liquor Pyrolysis”, Eastern
States Section, The Combustion Institute, Fall
Technical Meeting Proceedings, Clearwater
Beach, FL, 294-297 1994.
5.
MARTIN, D. and MALCOLM, E.,” The Impact
of Black Liquor Composition on the Release of
Nitrogen in the Kraft Recovery Furnace”, 1995
TAPPI Engineering Conference, Dallas, TX, 833840 (1995).
6.
9.
FORSSÉN, M., HUPA, M., PETTERSON, R.,
MARTIN, D., “Nitrogen Oxide Formation During
Black Liquor Char Combustion and Gasification”,
J. Pulp Paper Sci, 23 (9):J439-J446 (1997).
7.
TARPEY, T., “Adressing Ammonia and
Particulate Emissions from a Kraft Smelt Tank”,
1995 TAPPI Environmental Conference, Atlanta,
GA, 917-924 (1995).
8.
TARPEY, T., TRAN, H. and MAO, X.,
“Emissions of Gaseous Ammonia and Particulate
Containing Ammonium Compounds from a Smelt
Dissolving Tank”, J. Pulp Paper Sci, 22 (4), J145J150 (1996).
THOMPSON, L, MARTIN, D., EMPIE, H.,
MALCOLM, E., WOOD, M., “The Fate of
Nitrogen in a Kraft Recovery Furnace”, 1995
TAPPI Chemical Recovery Conference, Toronto,
Canada, B225-229 (1995).
10. KYMÄLÄINEN, M., FORSSÉN, M. and HUPA,
M.,” The Fate of Nitrogen in the Chemical
Recovery Process in a Kraft Pulp Mill”, 1998
TAPPI Chemical Recovery Conference, Tampa,
USA (FL), 19-32 (1998).
11. JENSEN, M., “On the Kinetics of the
Decomposition of Cyanic Acid, II: The Carbonate
Catalysis”, Acta Chem. Scand., 13 (4), 659-664
(1959).
Table 3. Kjeldahl nitrogen and ammonia analysis results.
Sampling
point
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
presented as
Sample
GL1, Green liquor from dissolver
GL2, Green liquor from clarifier to slaker
WL1, White liquor from causticizers
WL2, White liquor to digesters
WW, Weak wash
SS1, Scrubber solution from flue gas scrubber of rec. boiler
SS2, Scrubber solution from flue gas scrubber of NCG-inciner.
WBL, Weak black liquor from digesters
SBL, Strong black liquor, virgin (before mixing tank)
S,
Smelt from a smelt spout
FOUL,Untreated contaminated (“foul”) condensate to stripper
COND, Condensate from digesters
TC, Treated condensate from stripper to storage tank
SC1, Secondary condensate to pulp washing
SC2, Secondary condensate to lime mud washing
WATER. Warm water to pulp washing
*) mg N/kgds
**) mg N/kg
n.a., not analyzed
Kjeldahl-N
mgN/l
108
103
82
83
28
48
13
976 *)
840 *)
448 **)
338
338
3
10
20
4
Table 4. The molar ratio of N/Na in green and white liquors and weak wash.
Green liquor
White liquor
White liquor
to slaker
from causticizers
to digesters
N/Na ratio
1.73
1.47
1.45
(x 10-3)
4.6-7
NH3
mg N/l
18
70
58
70
24
35
0
198 *)
11 *)
9**)
337
341
0
n.a.
19
n.a.
Weak wash
2.55
NH3/
Kjeldahl-N,
%
17
68
70
84
86
73
0
20
1
2
100
101
0
105
NCG to
NCG to
incinerator
incinerator
stripper gasMETHANOL
CONDENSER
STRIPPER 13
MeOH to
incinerator
NCG to
incinerator
12
turpentine
TURPENTINEcondensate
RECOVERY
wood chips
4
white liquor
11
FOUL
COND.
TANK
8
DIGESTING
+
WASHING
secondarycondensate
16
strong
black
9
liquor
MIX
TANK
14
water
pulp
WHITE
LIQUOR
STORAGE
lime
WHITE
LIQUOR
FILTER
2
GREEN LIQ.
CLARIFIER
CAUSTICIZIERS
1
GREEN LIQ.
TANK
green liquor
LIME MUD
FILTER
SMELT
DISSOLVER
scrubber sol.
from NCG-inciner.
dregs
5
LIME
MUD
MIXER
VENT GAS
SCRUBBER
10
TANK
lime mud
6
smelt
3
SLAKER
FLUE GAS
SCRUBBER
RECOVERY
BOILER
LCHV-gas
LCHV-gas
LCHV-gas
SECONDARY
COND. TANK
“dirty” “clean”
EVAPORATION
weak black liquor
7
weak wash
lime mud to
dewateringand calcining
secondarycondensate
15
Figure 2. The main sampling points (1-16) in the recovery process of a kraft pulp mill.
Black liquor and smelt
mg N / kg DS
1200
1000
Green and white liquor system
Condensates
mg N / l
mg N / l
120
400
Kjeldahl-N
NH3-N
100
800
80
600
60
400
40
200
20
0
0
Scrubber solutions
mg N / l
60
350
50
300
40
250
30
200
150
20
100
WBL
*)
SBL
S *)
10
50
0
GL1
GL2
WL1
WL2
WW
0
FOUL
TC
SC
mg N / kg
Figure 3. Kjeldahl nitrogen and ammonia analysis results. (Sample codes as in Table 3.)
4.6-8
SS1
SS2
mg N / kg DS
1200
mg N / kg
600
weak BL
mg N / l
120
smelt
strong BL
green liquor
1000
500
100
800
400
80
600
300
60
400
200
40
200
100
20
0
0
1
2
3
4
5
1
2
3
4
5
0
1
Sampling period
2
3
4
5
1
Sampling period
2
3
4
5
Sampling period
100
1.0
80
0.8
-ln([A]/[A]o)
NH3, mgN/l
Figure 4. Variation of nitrogen concentration in black liquor, smelt and green liquor samples.
60
40
Total-N
NH3
20
0
0
100
200
0.4
Fitted line
0.2
Data
0
300
0
Time, min
Figure 5.
0.6
100
200
Time, min
a. Experimental data on the formation of NH3 in a mill green liquor at 96°C.
b. The first-order reaction test for the data shown in Fig.a.
4.6-9
300
NCG to
incinerator
NCG to
incinerator
turpentine
TURPENTINE
RECOVERY
0.4
FOUL
COND.
condensate TANK
6.1
white liquor
1.6
3.8*
23.9*
wood chips
6.7
DIGESTING
+
WASHING
1.6*
31..2
EVAPORATION
weak black liquor
0.6*
0.6
0.2
secondary condensate
water
0.4*
RECOVERY
BOILER
LCHV-gas
LCHV-gas
6.1
1.9*
0.3*
lime
smelt 6.8
8.2
SLAKER
TANK
CAUSTICIZERS
1.7 lime mud
dregs
2.4
0.4*
LIME
MUD
MIXER
SMELT
DISSOLVER
1.9
scrubber sol.
from NCG-inciner. 0.02
Nitrogen flows in the recovery cycle of a mill. Units: gN/s.
* : estimated (not measured) values
-2
NH3 (0.01 M)
-4
-6
NH4+
-8
-10
-12
6
8
10
12
14
pH
Figure 7.
0.14
1.5*
VENT GAS
0.3*SCRUBBER
secondary condensate
log(ci/M)
0
4
green
liquor
0.2
lime mud to 0.5
dewatering and calcining
LIME MUD
FILTER
Figure 6.
2
10.2
FLUE GAS
SCRUBBER
weak wash
1.6
0
GREEN LIQ.
TANK
Equilibrium concentration of gaseous NH3, ppm
WHITE
LIQUOR
FILTER
10.1
GREEN LIQ.
CLARIFIER
SECONDARY
COND. TANK
“dirty”
“clean”
strong
black
26.9
liquor
MIX
TANK
pulp
WHITE
LIQUOR
STORAGE
NCG to
incinerator
METHANOL
gas 6.7* CONDENSER
STRIPPER
0
MeOH to
incinerator
3000
95°C
2500
90°C
2000
80°C
1500
1000
500
25°C
0
0
50
100
150
Concentration of ammonia, mg/kg H2O
200
Left: NH3 /NH4+-equilibrium as a function of pH in aqueous solution (25°C).
Total concentration 0.01 M.
Right: Equilibrium concentration of gaseous NH3 over aqueous solution as a function of
aqueous NH3 concentration and temperature
4.6-10
DIGESTER EVAPORATION
(+WASHING)
RECOVERY BOILER
SMELT DISSOLVER
CAUSTICIZING
White Liquor
23
NH3
16
NOx
28
N2
Wood
Chips
Weak
Black
Liquor
Virgin
Black
Liquor
100
38
NH 3
6
8
Wa
ter
Smelt
34
2
NH3
h
as
W
Green
Liquor
38
s
ion
lut
o
S
Lime Mud
7
10
Weak Wash
Secondary Condensate
6
Figure 8. An overview of the nitrogen flows in the recovery cycle
4.6-11
White
Liquor
1
NH3
2 Lime Mud