See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/290562539

Chloride and potassium removal processes for kraft pulp mills: A technical
review

Article · January 2004

CITATIONS READS

12 3,727

2 authors, including:

Honghi Tran
University of Toronto
241 PUBLICATIONS   2,333 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Honghi Tran on 27 January 2016.

The user has requested enhancement of the downloaded file.

CHLORIDE AND POTASSIUM REMOVAL PROCESSES FOR KRAFT PULP MILLS:
A TECHNICAL REVIEW

Honghi Tran and Paul F. Earl

Pulp & Paper Centre and
Department of Chemical Engineering & Applied Chemistry
University of Toronto
Toronto, ON, CANADA

ABSTRACT

Chloride and potassium are known to have a negative impact on the kraft chemical recovery process. This has led to
renewed interest in chloride and potassium removal. In addition to purging precipitator dust, there are six processes
commercially available for the selective removal of chloride and potassium from the recovery cycle, employing four
distinctly different technologies. This review has been prepared to assist pulp mills in evaluating the different technology
options available to them, and to provide examples comparing the impact of each process on the kraft liquor cycle. The
principles of operation, the advantages and disadvantages, the impact on the liquor cycle and the potential benefits of each
process are discussed.

KEY WORDS

Kraft Mill, Recovery Boiler, Chloride, Potassium, Removal Process, Precipitator Dust, Technical Review

INTRODUCTION

Chlorine (Cl), often referred to as “chloride”, and potassium (K) are known to have a negative impact on the operation of
the kraft chemical recovery process. These elements, despite their small quantities in black liquor, can drastically lower the
melting temperature of fly ash deposits, and contribute to severe fouling and corrosion of heat transfer tubes in recovery
boilers.

Chloride and potassium enter the mill liquor cycle with wood and makeup chemicals. Depending on the wood species and
how they are transported to the mill, and the amount and type of makeup chemicals, Cl and K inputs may vary from 1.2 to
14 kg/adt as Cl, and from 0.7 to 4 kg/adt as K [1]. Once in the liquor cycle, Cl and K continue to accumulate until they
reach a steady-state concentration which is determined by their inputs, the degree of mill closure, and the liquor sulphidity.
For inland mills, the Cl content in the as-fired black liquor typically varies from 0.2 to 0.6 Cl wt% on dry solids, and higher,
1 to 2 wt% Cl, for mills that use caustic make-up contaminated with NaCl or for mills that partially recover bleach plant
effluents. For coastal mills where seaborne logs are used, the Cl content is much higher, 3 to 5 wt% Cl. The potassium
content in the as-fired black liquor, on the other hand, varies typically from 0.8 to 1.5 wt% K on dry solids for softwood
mills, and higher, 2 to 5 wt% for hardwood mills.

As pulp mills have tightened their liquor cycle in recent years to improve spill control and decrease chemical losses, Cl and
K concentrations in the mill liquor have increased, causing problems in recovery boiler operation. This has led to renewed
interest in chloride and potassium removal. Presently, there are six processes commercially available for the selective
removal of Cl and K from the recovery cycle, employing four distinctly different technologies. This paper reviews the
principles of these technologies and provides examples comparing the impact of each technology on the kraft liquor cycle.

HISTORY OF Cl AND K REMOVAL

Due to their highly volatile nature at high temperatures, Cl and K compounds (NaCl and KCl) vaporize from the recovery
boiler char bed, and become enriched in the precipitator ash. Efforts to remove Cl and K from the recovery cycle have
therefore been directed at the precipitator ash.

International Chemical Recovery Conference 1

Charleston, SC, Tappi/PAPTAC,June 2004.
For years, pulp mills have purged a portion of precipitator ash to control Cl and K levels. However, although Cl and K are
concentrated in the ash, they account for only 4 to 20 wt %. The remainder of the material being purged with the ash is
sodium (Na), sulfate (SO4), and carbonate (CO3). This means that more make-up sodium and sulfur must be added to the
liquor cycle when precipitator ash is purged.

Mills in coastal British Columbia have had to deal with extremely high chloride levels in their liquor cycle for many
decades, due to the practice of transporting logs in seawater. This led to the development of ash leaching treatments,
practiced at MacMillan Bloedel’s Harmac mill from 1973 to 1976 [2,3], and also at Champion’s Mogi Guaçu mill in Brazil
[4]. Leaching takes advantage of the relatively low solubility of Na2SO4 and K2SO4 compared to that of NaCl and KCl.
Precipitator ash was treated with a limited amount of filtrate to preferentially dissolve NaCl and KCl and leave the
remainder of the ash intact. The Harmac process was technically successful, but because of the equipment layout (three
recovery boilers feeding one leaching unit), the mill experienced frequent plugging of the long slurry feed pipelines [3]. As
well, there was extensive corrosion of the leaching equipment, due the practice of adding sulfuric acid to the ash slurry
improve sodium recovery by converting Na2CO3 to Na2SO4, which can be filtered more easily.

In the 1970’s, Rapson and Reeve pioneered the development and construction of the Effluent-Free Bleached Kraft Pulp
Mill (closed cycle mill) at the Great Lakes Forest Products (now Bowater Canada) mill in Thunder Bay, Ontario [5]. In this
process, bleach plant filtrate was recycled counter-currently back into the brownstock washing line, and thus into the
recovery cycle. With the chlorination stage operating with 70% ClO2 substitution, the concentration of Cl in the recovery
cycle rivaled that in coastal mills in British Columbia. The Rapson-Reeve’s Salt Recovery Process was an essential part of
the closed cycle mill, evaporating white liquor to crystallize NaCl, which was then returned as a feed chemical to the mill’s
R3 ClO2 generator. At the same time, Ahlström also developed a Salt Removal Process which removed Cl from green
liquor via cooling crystallization [6], although the process was never commercialized.

During the 1980’s, some mills in Sweden operated their recovery boilers at high sulfidity levels to remove chloride by
converting it into HCl, which is then purged from the system through the boiler stack gas, according to the following
reaction [7]:

2 NaCl + SO2(g) + ½ O2(g)+ H2O(g) → Na2SO4 + 2 HCl(g)

In the 1990’s there was renewed interest in the removal of chloride ions from the recovery cycle, with the development of
Champion’s BFR™ “Bleach Filtrate Recycle” process [4]. The BRF process revived the concept of the closed-cycle mill
with a new approach to recycle the bleach plant effluents. Champion initially examined an ash leaching process, similar to
that used at their Mogi Guaçu mill, but due to process issues (the amount of ash which needed to be treated, poor
filterability, and difficulties in obtaining sufficient Cl removal without excessive loss of Na2SO4) they selected an
evaporation/crystallization process, developed in partnership with Sterling Pulp Chemicals (now ERCO Worldwide) [8,9].

Several other technologies have been proposed in the past for Cl and K removal including electrodialysis [10] and liquid-
liquid extraction [11], although none of these has progressed past the laboratory or pilot-plant scale.

In recent years, many mills have also tried to decrease chloride input to the recovery cycle, by changing their make-up
caustic from the usual diaphragm-grade (which contains about 5000 ppm Cl) to membrane-grade caustic (which contains
less than 50 ppm Cl). Although there is a premium associated with membrane-grade caustic, the net return can be positive,
in terms of improved recovery boiler performance and/or a decrease in the amount of precipitator ash that needs to be
sewered.

PROCESSES FOR CHLORIDE AND POTASSIUM REMOVAL

All of the commercially-available Cl and K removal processes treat precipitator ash from recovery boilers. They fall into
four main categories: i) leaching, ii) evaporation/crystallization, iii) freeze crystallization, and iv) ion exchange.

The first three processes take advantage of differences in solubility between the various salts, primarily alkali sulfates
(Na2SO4 and K2SO4) and alkali chlorides (NaCl and KCl), while the ion exchange process uses a proprietary resin which
selectively adsorbs NaCl. All of the processes remove the Cl and K in a purge stream, and return the more purified alkali
sulfate to the recovery liquor cycle.

International Chemical Recovery Conference 2

Charleston, SC, Tappi/PAPTAC,June 2004.
Processes that rely on solubility difference must deal with a complex aqueous solution of alkali salts. In pure solutions, the
solubilities of Na2SO4 and NaCl are relatively similar, particularly in the temperature range 35-100°C [12]. However, in
solutions of mixed salts, the presence of NaCl dramatically decreases the solubility of Na2SO4, due a common ion effect.
This allows more NaCl to be dissolved from a slurry of precipitator ash without dissolving Na2SO4 (as in leaching), or
allows more Na2SO4 to be re-crystallized from a solution of precipitator ash without re-crystallizing NaCl.

However, Na2SO4 and NaCl are not the only constituents in the precipitator ash. K2SO4, KCl, Na2CO3 and K2CO3 are also
present, and this further complicates the solubility of the system [12-14]. In both leaching and crystallization, the conditions
which remove NaCl from Na2SO4 also tend to remove Na2CO3, which represents a loss of valuable sodium, and hence an
increased need for make-up caustic. The problem can be addressed by acidifying the leaching slurry [2] or the
crystallization mother liquor [8] with sulfuric acid to convert the carbonate to more recoverable sulfate. However, this adds
more sulfur into the system and affects the mill sulfur balance. Acidification also leads to increased risk of corrosion, and
hence, to higher equipment costs associated with the use of acid-resistant materials.

Potassium removal is more complicated than Cl removal because of the different forms potassium can take. Since mixtures
of KCl and NaCl are more soluble than either of the pure salts [14], the presence of KCl does not pose a problem. However,
potassium can also be present as K2SO4, or the double salt glaserite (3K2SO4·Na2SO4) after dissolving and re-crystallization
[13,14]. These compounds tend to behave in a similar manner as Na2SO4 in the system, i.e. the solubilities of K2SO4 and
3K2SO4·Na2SO4 decrease as the concentration of K and/or Cl increases. This hinders the potassium removal in both the
leaching and evaporation/crystallization processes, as potassium compounds remain insoluble during leaching, and re-
precipitate during re-crystallization. In these situations, potassium can no longer removed from the liquor cycle, and the
concentration will continue to increase. When this is the case, the most effective solution may be to simply purge
precipitator dust until the potassium concentration drops into a range where glaserite is no longer formed, and then begin
operation of the removal system.

Leaching

Leaching takes advantage of the high solubility of NaCl and KCl, compared to the relatively insoluble Na2SO4, especially
in an NaCl-saturated solution [12]. The leaching process consists of two main unit operations. In the first stage, precipitator
ash is combined with water (and recycled leachate) to form a slurry or paste. The amount of water/leachate used must be
carefully controlled to promote the dissolution of chloride and potassium salts without dissolving sodium sulfate. A higher
temperature also improves the selectivity of the process, as more Cl salts and less sulfate salts dissolve in the system. After
mixing in a stirred tank, the slurry passes to the separation stage, typically a filter or centrifuge. The solids (primarily
sulfate) are returned to the kraft liquor cycle, while the leachate (rich in Cl and K) is recycled back to the first step. A
portion of the leachate is purged from the system as the means of Cl and K removal.

Figure 1: Kvaerner Ash Leaching System

International Chemical Recovery Conference 3

Charleston, SC, Tappi/PAPTAC,June 2004.
A Precipitator Ash Leaching system is available from Kvaerner Power (Figure 1). The solid-liquid separation is achieved
by means of a centrifuge. The first unit was started-up at the Aracruz mill in Brazil in 2002 [15]. The Cl and K removal
efficiency was reported to be about 70%, with 80% recovery of Na and 85% recovery of S.

Evaporation/Crystallization

Evaporation/crystallization processes also take advantage of the relatively low solubility of Na2SO4, as compared to the
NaCl and KCl, but in the opposite manner to leaching. The precipitator dust is first dissolved in water (or recycled process
condensate); the solution is then evaporated. Because of its lower solubility, Na2SO4 crystallizes first. The presence of Cl
ions helps to decrease the solubility of Na2SO4 further. The purified Na2SO4 is filtered and returned as a solid to the liquor
cycle. Most of the mother liquor (rich in Cl and K) is returned to the crystallizer, but a small portion is sewered to purge the
Cl and K.

There are three evaporation/crystallization processes available commercially: i) HPD CRP “Chloride Removal Process”, ii)
Andritz ARC “Ash Re-Crystallization” process, and iii) Eka Chemicals PDR “Precipitator Dust Recovery”. Although each
uses a different evaporator design, the principle of operation is similar. All the processes use steam-driven evaporator/
crystallizers, which may be single- or multi-effect, depending upon the design and the needs and integration with the pulp
mill.

Figure 2 shows a process schematic of the HPD CRP. The design has been updated considerably since its first installation
in 1995, designed for the Champion BFR “Bleach Filtrate Recycle” project in Canton, FL [8,9,16,27]. The evaporator
operates under vacuum with an external reboiler. Single- and multi-effect systems are now in use, depending upon the
quantity of ash being processed and the integration of the CRP into the black liquor evaporator train. Instead of the original
saltcake filter, the CRP now uses an internal wash column for solid-liquid separation.

Vapor
Vapor
Body
Hot
Condensate
Steam
Ash
Recovered Heater
Saltcake Steam
Mix Purge Condensate
Tank
Feed Solution Recirculation
Pump

Figure 2. HPD CRP “Chloride Removal Process”

There are six CRPs operating worldwide: 3 in North America, 2 in South America, and 1 in Australia. Published data from
the Canton installation shows a chloride removal efficiency of 95% and 80% sulfate recovery [16]. Potassium removal can
vary considerably, from 50-85%, depending upon the amount of K in the precipitator dust and the desired balance between
K removal, and Na and S recovery [8,16]. Historical trendlines for chloride and potassium removal from the Canton mill
are shown in Figure 3.

Figure 4 shows a process schematic of the Andritz ARC (Ash Re-Crystallization) process. The ARC uses falling-film-type
evaporators to concentrate the ash solution. Sodium sulfate crystals are removed using a rotary-drum filter, while a portion

International Chemical Recovery Conference 4

Charleston, SC, Tappi/PAPTAC,June 2004.
of the chloride- and potassium-rich solution is purged. A pilot plant was operated at the UPM Kymmene mill in Pietarsaari,
with results indicating 90% chloride removal efficiency and 80% recovery of alkali [18]; as with the CRP, potassium
removal is inversely dependent upon the degree of Na recovery. A full-scale ARC is expected to be in operation at a kraft
mill in Portugal in 2004 [19].

4.0
Bleach Filtrate
Recycle begins
3.0
NaCl
CRPstart-up
2.0

1.0
K+

0.0
Sep 94 Apr 95 Oct 95 May 96 Dec 96 Jun 97 Jan 98

K, % in BL solids NaCl, g/L in WL

Figure 3. Impact of HPD CRP on Chloride and Potassium Concentrations [16]

Figure 4. Andritz ARC “Ash Re-Crystallization Process”

International Chemical Recovery Conference 5

Charleston, SC, Tappi/PAPTAC,June 2004.
The Eka Chemicals’ PDR process is also a recirculating evaporator-crystallizer, operating under vacuum with an external
reboiler (Figure 5). The PDR has not yet been commercialized, but laboratory and pilot-scale testing indicates a chloride
and potassium removal efficiency of 90%, with 80% recovery of sulfate and carbonate [20].

HP Steam CRYSTALLIZER
CONDENSER
CRYSTALLIZER
ESP EJECTOR
DUST FILTER EJECTOR
Cooling Water
CRYSTALLIZER

DUST
AFTERCOOLER
DISSOLVING
TANK
Cooling Water
Vent

Process Water

Hot Water

FILTRATE
STANDPIPE SULFATE FILTER

DUST SOLUTION
PUMP
PURGE LP Steam

LP Steam SULFATE HOT WATER

RECOVERY TANK PUMP

REBOILER BLACK LIQUOR

CONDENSATE
TANK FILTRATE PUMP

TO RECOVERY

Condensate CIRCULATION
PUMP

CONDENSATE FILTER FEED SULFATE

PUMP PUMP RECOVERY PUMP

Figure 5. Eka Chemicals’ Precipitator Dust Recovery Process

Freeze Crystallization

Freeze crystallization relies on the dramatic decrease in solubility of sodium sulfate decahydrate (Na2SO4·10H2O) at low
operating temperatures. Below about 30°C, Na2SO4 suddenly precipitates as large Na2SO4·10H2O crystals, leaving Cl and
K in solution. The two phases can then be separated, with the purified sodium sulfate crystals returned to the kraft liquor
cycle, and the solution purged to remove Cl and K.

There are six Mitsubishi Freeze Crystallization systems in operation in pulp mills in Japan [21,22]. As shown in Figure 6,
the precipitator ash is mixed with a sufficient amount of water to dissolve the chloride and potassium salts, along with a
portion of the sodium sulfate. Acid may be added to improve sodium recovery, as described previously. The slurry is then
transferred to a precipitation tank, where ice is added to decrease the temperature to 15°C, causing Na2SO4 in solution to re-
crystallize as Na2SO4·10H2O. The slurry is then decanted to separate and recover the decahydrate crystals, along with any
undissolved Na2SO4, while the decanted liquor is purged. The removal efficiency was reported to be 90% for Cl and 75%
for K with 70% sodium recovery. Figure 7 shows the performance data at an installation in Japan.

International Chemical Recovery Conference 6

Charleston, SC, Tappi/PAPTAC,June 2004.
 ICE MAKER

ESP ASH
HOPPER

CLEAR WATER
CLEAR WATER

ASH
SULFURIC DEHYDRATOR
ACID

DILUTE BLACK LIQUOR

WASTE CAKE
LIQUID MAXING
PUMP TANK TANK
SLURRY TANK
PRECIPITATION
TANK
PUMP
WASTE LIQUID RECIRCULATION LINE
TO DILUTE BLACK
PUMP LIQUOR TANK

DISCHARGE

Figure 6. Mitsubishi Freeze Crystallization Process

8
7
Concentration (wt%)

6 K observed value

5 K calculated value

4
3
Removal
2
System Cl observed value
1 Installed Cl calculated value

0 1/1 1/31 3/1 3/31 4/30 5/30 6/29 7/29 8/28 9/27 10/2711/2612/26 1/25 2/24 3/26

DATE 2000 2001

Figure 7. Impact of Mitsubishi Freeze Crystallization on Chloride & Potassium Levels [21]

International Chemical Recovery Conference 7

Charleston, SC, Tappi/PAPTAC,June 2004.
Ion Exchange

Specialized ion exchange resins are commercially available which contain both cation- and anion-exchange groups on each
particle, and so can remove both ions simultaneously. Because chloride ions are the only monovalent anions in the
precipitator duct mixture, these resins will preferentially adsorb sodium chloride over sodium sulfate and carbonate.

Using this resin, an ion exchange treatment system known as PDP (Precipitator Dust Purification) has been developed by
Eco-Tec and Paprican [23,24]. A process schematic is shown in Figure 8. Due to the sensitivity of the resin, the
precipitator dust solution must be filtered beforehand, to minimize resin fouling. Regeneration is easily achieved with
water.

The PDP system has not been commercialized, although a pilot plant unit is in operation at Paprican’s laboratory. Chloride
removal is reported to be 97%, with minimal losses of sodium, sulfate, and carbonate (6%, 1%, and 0.5% respectively).
However, because the sodium concentration exceeds that of potassium by many orders of magnitude, potassium removal is
low, on the order of only 5%. An ion exchange system is therefore better-suited to a situation where chloride removal is
significantly more important than potassium removal.

Water

Pulse
Water Filter SSU

ESP NaCl
Waste
Dust to black
liquor
NPE evaporators
Sludge

Dissolving Feed
Tank Tank

Figure 8. Eco-Tec PDP Ion Exchange System (SSU = Salt Separation Unit)

One difference between the ion exchange process and other processes is that the purified sodium sulfate is returned to the
kraft liquor cycle in the form of solution, and not as a crystal or slurry. If the mill has spare capacity in the black liquor
evaporator train, then this will not be an issue, and in fact the ion exchange process could end up consuming less steam than
an evaporation/crystallization process. However, if there is no additional black liquor evaporation capacity, this may
necessitate an additional pre-evaporation stage, to minimize the water loading on the existing evaporators. The addition
point of the sulfate solution to the black liquor evaporation train is also critical, since the relatively high concentration of
carbonate could promote burkeite formation and scaling. One proposed alternative for the removal of chlorides from the
recovery liquor cycle is to use ion exchange to treat a stream of oxidized white liquor, instead of the precipitator dust [24].

International Chemical Recovery Conference 8

Charleston, SC, Tappi/PAPTAC,June 2004.
Recently, researchers in Brazil investigated the use of amphoteric anion/cation exchange resins to remove potassium as well
as chloride ions from precipitator dust [25]. Preliminary lab results indicate 95% removal of both Cl- and K+, with 98%
sulfate recovery.

Table 1 summarizes the installations and performance data of the commercially-available processes discussed above.

Table 1. Chloride and Potassium Removal Processes for Kraft Pulp Mills
Total Ash
Process Type of No. of Chloride Potassium Sodium Sulfate
Supplier Processed,
Name Process Installations Removal Removal Recovery Recovery
t/d
CRP
US Filter / Evaporation- variable;
(Chloride Removal 6 446 70-85% 80-82% 80%
HPD Crystallization 50-85%
Process)

MPR (Mitsubishi
Freeze
Potassium & Chloride Mitsubishi 6 192 90% 75% 70% 75%
Crystallization
Removal System)

1
Ash Leach Kvaerner Leaching (3 being 56 69% 69% 80% 85%
engineered)

First start-up
ARC (Ash Re- Evaporation- Not Not
Andritz Autumn 167 90% 80%
Crystallization) Crystallization reported reported
2004

PDR
Eka Evaporation-
(Precipitator Dust Pilot plant 90% 90% NR 80%
Chemicals Crystallization
Recovery)

PDP (Precipitator
Eco-Tec Ion Exchange Pilot plant 97% 5% 94% 99%
Dust Purification)

Sewering Precipitator Dust and Other Options

When evaluating the impact of a process to remove chloride and potassium from the recovery cycle, a pulp mill should
always weigh the costs and benefits against simply sewering a portion of the precipitator dust, if such a practice is allowed.
Purging the precipitator dust, however, will significantly increase losses of sodium and sulfur, and therefore the need for
purchased make-up chemicals.

Bleached kraft pulp mills represent a special case. With the replacement of Cl2 with ClO2 as a bleaching agent, many mills
are producing more by-product Na2SO4 (saltcake) from their ClO2 dioxide generators than is needed for chemical make-up.
Since the saltcake from a modern sub-atmospheric ClO2 generator contains significantly less Cl than recovery boiler
precipitator dust, and no K, it would be more advantageous to recover all of the “cleaner” ClO2 generator saltcake, and
sewer precipitator dust instead.

This idea has been taken one step further by researchers at Air Products, who suggest recovering all excess generator
saltcake, and then oxidizing the excess white liquor for use as alkali in the bleach plant [26]. Since the bleach plant effluent
is sewered, the mill will be, in essence, purging Cl and K with the excess white liquor, instead of sewering precipitator dust,
and at the same time decreasing operating costs by replacing purchased caustic. While extensive modeling studies have
been carried out, the concept has not yet been commercialized.

International Chemical Recovery Conference 9

Charleston, SC, Tappi/PAPTAC,June 2004.
IMPACT OF Cl AND K REMOVAL ON PULP MILL OPERATION

To illustrate the impact of the above Cl and K removal processes and options, a simple recovery cycle model was applied
using operating data from a kraft mill. In the base case, the mill is not purging any precipitator dust, but instead recovers
only 70% of the saltcake from its ClO2 plant, and sewers the excess saltcake to control liquor sulfidity. An additional 8.1
kg/ODMt caustic is required for make-up sodium in the liquor cycle. Due to high Cl and K inputs from the wood and mill
process water, and a tight liquor cycle with low losses, the concentrations of these elements in the precipitator dust is high:
7.3 mole-% chloride and 16.6 mole-% potassium.

Two purging strategies were evaluated for the four Cl and K removal processes described above:

1. Recovering all ClO2 generator saltcake and sewering precipitator dust to control sulfidity;

2. Recovering all ClO2 generator saltcake and sewering precipitator dust to control sulfidity, and using low-Cl caustic for
make-up

The results are shown in Table 2. In both of the purge options, all of the ClO2 generator saltcake was recovered, and
precipitator dust was sewered to maintain the sulfur balance. Make-up caustic requirements decrease to 6.5 kg/ODMt,
because of the 3:2 Na:S molar ratio in sodium sesquisulfate (NaHSO4.Na2SO4) from the ClO2 plant. The purge of the
precipitator ash decreases the Cl concentration from 3.0 wt% Cl in the baseline case to 2.3 wt% Cl. Combining a
precipitator ash purge with a switch to low-chloride membrane-grade make-up caustic decreases the chloride concentration
further to 1.7 wt% Cl.

Table 2. Modelling of Options to Remove Potassium & Chloride

Make-Up Make-Up ClO2 Precipitator Concentration in Precipitator Dust

Process Generator
NaOH NaOH Dust Cl/(Na+K), K/(Na+K),
Option Saltcake Cl, wt-% K, wt-%
(kg/ODMt) Grade Treated mole-% mole-%
Recovered

Baseline 8.1 Diaphragm 70% --- 3.0% 7.5% 7.3% 16.6%

Purge
6.5 Diaphragm 100% 13% 2.3% 6.5% 5.5% 14.2%
ESP Dust

Purge Dust + Membrane

6.5 100% 13% 1.7% 6.5% 4.0% 14.2%
Low-Cl Caustic (low-Cl)

Leaching 7.1 Diaphragm 100% 80% 0.7% 2.0% 1.5% 4.2%

Evaporation-
6.4 Diaphragm 97% 60% 0.7% 2.9% 1.6% 6.0%
Crystallization

Freeze Membrane
6.8 98% 50% 0.6% 3.4% 1.5% 7.0%
Crystallization (low-Cl)

Ion Exchange 8.5 Diaphragm 72% 60% 0.5% 7.1% 1.2% 15.6%

Note: Process efficiencies for Leaching, Freeze Crystallization, & Ion Exchange were from Table 1;
Evaporation-Crystallization efficiencies averaged from Table 1 as: Na & S recovery = 80%, Cl removal = 85%, K removal = 70%.

Performance data from Table 1 was used for the four Cl and K removal processes; the amount of precipitator dust being
treated was adjusted to yield a final Cl content of 0.5-0.6 wt% in the dust. The models were arbitrarily limited so that sulfur
losses did not exceed the amount of make-up sulfur available from the ClO2 generator saltcake.

All processes significantly decreases the Cl content to well below 1 wt% Cl. The leaching and crystallization processes also
decreased the K content, while the ion exchange process had little effect on K levels.

International Chemical Recovery Conference 10

Charleston, SC, Tappi/PAPTAC,June 2004.
Despite the losses of sodium and sulfur, three of the four of the processes require less make-up caustic, compared to the
base case. Because of its high sodium and sulfur recovery efficiency, the ion exchange process would require mills to
continue sewer about 30% of the ClO2 generator saltcake produced in order to maintain sulfidity, which then increases the
need for make-up caustic slightly.

In this model, all of the processes allowed the mill to keep using diaphragm-grade make-up caustic, with the exception of
freeze crystallization. In order to decrease the Cl concentration to the same degree as the other processes, while adhering to
the limit on sulfur losses, the freeze crystallization model required a switch to low-Cl membrane-grade make-up caustic.
Alternatively, more dust could be treated to increase the amount of Cl (and K) being removed and more sodium sulfate
purchased to make-up the additional sulfur losses.

Note that the results are estimated only to illustrate relative performance. In practice, the performance may vary depending
upon the composition of the precipitator ash, especially with respect to potassium and carbonate concentrations.

SUMMARY

Several processes are now available for the selective removal of chloride and potassium from the kraft recovery cycle.
These processes can significantly decrease the Cl concentration in the liquor cycle, allowing for improved recovery boiler
operation. Potassium levels are also decreased, but to a lesser extent.

ACKNOWLEDGEMENTS

The authors wish to thank the equipment vendors who provided diagrams and information regarding their processes:
Andritz, Eco-Tec, Eka Chemicals, Kvaerner, Mitsubishi Heavy Industries, and U.S. Filter/HPD.

REFERENCES

1. Tran, H.N., Barham, D., and Reeve, D.W., “Chloride and Potassium in the Kraft Chemical Recovery Cycle”, Pulp &
Paper Canada 91(5) : 55-62 (1990).

2. Moy, W.A., Joyce, P., & Styan, G.E., “Removal of Sodium Chloride from Kraft Recovery Systems”, Pulp & Paper
Canada 75(4) : 88-90 (1974).

3. Blackwell, B., and Hitzroth, A., "Recycle of Bleach Plant Extraction Stage Effluent to the Kraft Liquor Cycle: A
Theoretical Analysis", Proceedings of the 1992 TAPPI International Chemical Recovery Conference, pp. 329-349
(Seattle, June 7-11, 1992).

4. Maples, G., et al., “BFR™: A New Process Toward Bleach Plant Closure”, Tappi Journal 77(11) : 71 (1994).

5. Reeve, D.W., Lukes, J.A., French, K.A., & Rapson, W.H., “The Effluent-Free Bleached Kraft Pulp Mill, Part IV: The
Salt Recovery Process”, Pulp & Paper Canada 75(8) : T293 (1974).

6. Henricson, K., Kiiskila, E., & Virkola, N., “Studies of Green Liquor Cooling Crystalization and its Application to
NaCl Removal from the Pulping Chemical Recovery Cycle”, Paperi ja Puu 57(10) : 643-652 (1975).

7. Borg, A., Teder, A., Warnquist, B., "Inside a Kraft Recovery Furnace - Studies on the origins of sulphur and sodium
emission", Tappi, 57 (1) 126-129 (1974).

8. Earl, P.F., Dick, P.D., and Patel, J-C., “Removal of Chloride and Potassium from the Kraft Recovery Cycle”, paper
presented at the 1995 CPPA Pacific Coast & Western Branches Technical Conference (Whistler BC, May 18-20,
1995).

9. Fleck, J.A., Earl, P.F., & Fagan, M.J., “Kraft Mill Bleach Filtrate Recycle and the Commercial Demonstration of
Chloride and Potassium Removal”, Proceedings of the 1996 International Environmental Conference (Orlando, May
5-8, 1996).

International Chemical Recovery Conference 11

Charleston, SC, Tappi/PAPTAC,June 2004.
 10. Thompson, R., Paleologou, M., and Berry, R.M., “Sodium Recovery with Chloride Removal from Electrostatic
Precipitator Catch at Coastal and/or Closed-Cycle Mills Using Bipolar Membrane Electrodialysis”, Tappi Journal
80(6) : 154-160 (1997).

11. Prakash, T.S., and Moudgil, B.M., “Application of Aqueous-Aqueous Separation Technique for Chloride Removal
from Precipitator Salt Cake”, Proceedings of the 1994 Non-Chlorine Conference, paper 12-4 (Amelia Island FL,
Maarch 6-10, 1994).

12. Seidell, A., and Linke, W.F., “Solubility of Inorganic and Metal Organic Compounds” (Fourth Edition, American
Chemical Society, Washington DC), Vol. 2, pp. 982-983, 1121-1123 (1965).

13. Blasdale, W.C., “Equilibria in Solutions Containing Mixtures of Salts, Part I: The System Water and the Sulfates and
Chlorides of Sodium and Potassium”, Journal of Industrial and Engineering Chemistry 10(5) : 344-347 (1918).

14. Gilbert, A.F., and Rapson, W.H., “Accumulation of Potassium in a Closed Cycle Mill”, Pulp & Paper Canada 81(2) :
T43-47 (1980).

15. Knutsson, M., et al., “Experiences from First Start-Up of an Ash Leaching System”, Proceedings of the 2002 TAPPI
Fall Conference & Trade Fair (San Diego, CA, Sept. 8-12, 2002).

16. Stratton, S.C., and Ferguson, M., “Progress Report on the BFR Technology Demonstration – December 1996”, paper
presented at the 1997 TAPPI Environmental Conference (Minneapolis, May 5-7, 1997).

17. Caron, J.R., and Delaney, G., “Initial Learnings from the BFR Demonstration”, Proceedings of the 1997 TAPPI
Minimum Impact Mill Symposium, pp. 73-78 (San Francisco CA, Oct. 23-24, 1997).

18. Koskiniemi, J., Vakkilainen, E., Paju, R., and Hanninen, E., “Removal of Chlorides from Chemical Circulation in the
Kraft Pulp Mill”, Appita Journal 52(6) : 460-463 (1999).

19. Ferreira, L.M.G.A., et al., “Selective Removal of Chloride and Potassium in Kraft Pulp Mills”, Tappi Journal 2(4):
21-25 (2003).

20. Minday, A.M., Reid, D.W., Burke, M.A., and Brown, C.J., “An Overview of Various Strategies for Balancing
Saltcake, Chloride, and Potassium Levels in an ECF Kraft Mill”, Proceedings of the 1997 TAPPI Minimum Impact
Mill Symposium, pp. 115-130 (San Francisco CA, Oct. 23-24, 1997).

21. Tran, H., & Arakawa, Y., “Recovery Boiler Technology in Japan”, Proceedings of the 2001 TAPPI Engineering
Conference (San Antonio, TX, Dec. 2-6, 2001).

22. Fujisaki, A., et al., “Plugging Prevention of Recovery Boiler by Character Improvement of the Ash which used
Potassium Removal Equipment (Part II)”, Pulp & Paper Canada 104(1) : T1-3 (2003).

23. Brown, C.J., et al., “Ion Exchange Technologies for the Minimum Effluent Kraft Mill”, Pulp & Paper Canada 100(8)
: T251 (1999).

24. Jemaa, N., et al., “Sodium Chloride Removal from Oxidized White Liquor Using a Fixed Resin Bed System”, Pulp &
Paper Canada 103(3) : T67-70 (2002).

25. Jaguaribe, D., et al., “Clean Ion-Exchange for Selective Removal of Chloride and Potassium Using IXISS Effect”,
Proceedings of the TAPPI fall 2003 Technical Conference (Chicago, Oct. 26-30, 2003).

26. Mullen, W.T., “Using White Liquor Oxidized to Thiosulphate in the Bleach Plant can Reduce Caustic Purchases by
20%”, paper presented at the 88th PAPTAC Annual Meeting, Montreal, January (2002).

International Chemical Recovery Conference 12

Charleston, SC, Tappi/PAPTAC,June 2004.

View publication stats