Investigation of green lime mud at Harmac mill - Pulp and Paper ...

LIME KILNS
Investigation of
green lime mud
at Harmac mill
By K. Taylor and D. Bossons
Abstract: Over the past several years, the Harmac pulp mill has experienced operating periods
of very poor lime mud quality and upset operation of the recaust and lime kiln areas. These upsets
coincided with the occurrence of green lime mud. This study investigates the chemistry of non-process
elements (NPEs) in the recaust/lime kiln cycle, evaluates the effect of secondary sludge burning
in the recovery boiler, identifies a possible cause of green lime mud, and makes recommendations
to reduce lime mud filter plugging and maximize sludge burning.
H
ARMAC LIME MUD was typically very
white in colour until 2001. Since then,
lime mud colour has varied from
nearly white to dark green. Recaust
upsets such as plugging of pressure
and rotary lime mud filters seemed to occur when
the lime mud was dark green in colour. These
incidents of poor lime mud quality were serious
enough to reduce mill output due to insufficient
white liquor production. There was a concern
that these process upsets were caused by non-process
elements (NPEs) introduced during the
burning of secondary treatment sludge in the
recovery boiler, which also began in 2001. As a
result, sludge burning was discontinued in June
2004. However, a serious recaust upset with green
lime mud occurred from July to October 2004,
casting doubt on sludge burning as the cause of
the problem. A strong economic incentive exists
to burn sludge in the recovery cycle at Harmac,
since it reduces operating costs by approximately
$2 million a year.
The Harmac pulp mill, located on the east
coast of Vancouver Island, is operated by Pope &
Talbot. The mill was built in three distinct phases
between 1950 and 1963, and has undergone considerable
modernization over the past 20 years.
The mill currently operates with a capacity of
1,120 air-dried tonnes (adt/d) of fully bleached
market kraft pulp. The recaust was modernized in
the mid-80s and combined into a single line. Harmac
disposes of the waste secondary sludge from
its effluent-treatment process through three
means: dewatering and burning in the power
boiler, dewatering and sending to landfill, or
direct mixing with the weak black liquor and
burning in the recovery boilers.
Incineration of secondary sludge in the recovery
boiler has been of interest to the pulp industry
for some time [1]. Economics are favourable,
but the sludge is known to contain mill-specific
NPEs that can affect mill operation [2].
Extensive literature searches and discussion
with industry experts did not explain the occurrence
of green lime mud. Extensive reviews of the
effect of non-process elements in kraft pulp mills
are available [3-6]. However, none of them was able
to shed light on the problems occurring at Harmac.
Because of the favourable economic incentive
of sludge burning, and the negative impact of
recaust upsets, a research project was initiated to
address the following objectives:
1. Identify the cause of green lime mud at Harmac;
2. Understand the role of NPEs in the Harmac
process cycle;
3. Evaluate the effect of sludge burning on the
Harmac process cycle; and
4. Make recommendations to reduce lime mud
filter plugging and maximize sludge burning.
METHODS, MATERIALS
Metal concentrations were measured by
Econotech Services Ltd., Delta, BC. After digestion,
samples were analyzed by inductively coupled
plasma (ICP) spectroscopy. Total metals content
was measured in liquid samples.
Step-scan X-ray powder-diffraction data
(XRD) using the quantitative Rietveld method
were collected on a standard Siemens (Bruker)
D5000 Bragg-Brentano diffractometer. These
measurements were performed at UBC Department
of Earth and Ocean Sciences by M. Raudsepp
and E. Pani.
Visible light spectroscopy of liquid samples was
performed with a Carey 5 UV/VIS/NIR spectrometer
while diffuse reflectance measurements
of solids were made with a Carey 1 UV/VIS/NIR
spectrometer.
Brightness of lime mud was measured with a
Technibrite Micro TB-1C (Technidyne Corporation,
USA).
Kaolin clay was obtained from Feldspar Corp.
(Florida) as EPK Kaolin and contained 97 wt%
kaolinite. Surface area was 24 m 2 /g and average
particle size 1.4 microns. Synthetic green liquor
contained 1.1M sodium carbonate (ACS grade)
and 0.5M sodium sulphide (ACS grade).
Samples for the sludge-burning trial were collected
in acid-washed polyethylene sample containers
to eliminate possible metal contamination.
Samples were collected for four days with
normal mill operation then for seven days burning
sludge at 120 US gal/min. One set of samples
was collected seven days after termination of
sludge burning.
K. TAYLOR,
Taylor Industrial Research,
Victoria, BC
kevin.taylor@industrialresearch.ca
D. BOSSONS,
Pope & Talbot Ltd.,
Harmac Pulp Operations,
Nanaimo, BC
Pulp & Paper Canada T 63 107:3 (2006) ❘❘❘ 37

LIME KILNS
TABLE I. Harmac sludge composition.
Element
Concentration, mg/kg
Aluminum 2,800
Iron 3,400
Magnesium 11,300
Manganese 500
Phosphorus 7,750
Silicon 25,000
Sodium 10,800
Typical sludge composition on a dry solids basis is shown in
Table I. Sludge typically contained 2.5 wt% solids.
RESULTS
Harmac NPE Levels Compared with Other Mills: The concentrations
of seven different non-process elements in Harmac lime
mud were compared with values from other Canadian kraft mills
[6,7]. Harmac values are the average of five measurements taken
from July to November 2004 during normal mill operation. Values
from other mills (except phosphorus) are taken from
Richardson [6] for seven Canadian kraft mills. Phosphorus concentration
was obtained from the average values of ten Canadian
kraft mills [7], estimating the concentration in lime mud by
dividing the lime values by 1.78. Error bars represent ± one standard
deviation of the mean value. Results are shown in Fig. 1.
Figure 1 shows that aluminium, phosphorus and silicon concentrations
in Harmac lime mud are significantly higher than
reported values at other mills. Magnesium and potassium concentrations
are higher, but both overlap with the values seen
with other mills. Iron and manganese concentrations are within
the range of values reported by other mills.
Figure 2 shows that both aluminium and magnesium concentrations
have been increasing in Harmac lime mud since
2001. In 2001, values of aluminium were much lower and were
consistent with other mills. Magnesium concentrations in 2001
were much lower and were at the low end of the range of values
of other mills, as shown in Fig. 1.
Figure 3 shows that phosphorus concentration in the lime
mud was high during periods of sludge burning. Silicon concentration
does not show any trend since 2001, although there
is some variability in the 2002 measurements. Iron concentration
varies considerably but does not show a definite trend.
A further item of concern was high aluminium concentration
in Harmac green liquor, as Fig. 4 shows. Aluminium concentration
was significantly higher than values seen in other mills [6].
Silicon concentration was also much higher, with an average of
250 mg/L at Harmac versus 108 mg/L for other mills [6]. Concentrations
of other elements in Harmac green liquor were
much lower than other mills, due to efficient clarifier operation.
As a result of these comparison studies with other Canadian
kraft mills, it was clear that aluminium and silicon concentrations
in Harmac lime mud, lime, white liquor and green liquor
were significantly higher than in other mills during normal mill
operation. Aluminium and magnesium concentrations showed a
significant increasing trend since 2001 when incidents of green
lime mud and filter plugging first occurred. In addition, high
phosphorus concentrations occurred in the lime mud during
periods of sludge burning.
Cause of Green Lime Mud: Quantitative XRD analysis of a lime
sample showed the presence of kaolinite (0.2 wt%) and gehlinite
(0.9 wt%). Kaolinite, Al 2 O 3 · 2SiO 2 · (OH) 4 , is an aluminosilicate
compound with very high surface area and small particle
size. It is the most common of the clay minerals. Magnusson
et al. [3] state that equilibrium calculations show that kaolinite
could be precipitated from white or green liquor at high silicon
concentrations and relatively low (

LIME KILNS
FIG. 2. Al and Mg trend in Harmac lime mud.
FIG. 3. P, Si and Fe trend in Harmac lime mud.
FIG. 4. NPEs in Harmac green liquor.
FIG. 5. Reflected light spectra of green lime mud.
data, NPEs of primary interest included
iron, magnesium, aluminium, silicon and
phosphorus. As a result of this study, the
behaviour of each element in the Harmac
lime plant was understood in much
better detail.
—Iron: Concentrations of iron in Harmac
lime have fluctuated significantly since
2001, as Fig. 3 shows. The green liquor
clarifier at Harmac is highly efficient, producing
green liquor with iron concentrations
of 1 to 2 mg/L during normal operation.
White liquor iron concentration
was very stable at 5 mg/L, suggesting that
iron should be rapidly removed from the
lime plant in the white liquor. A mass balance
of iron in the lime plant showed that
a major source of iron was the lime rock.
Iron concentrations in the lime rock were
found to vary from 300 to more than
1,000 mg/kg between rock shipments.
When iron input to the lime plant exceeded
losses to the white liquor and grits,
iron levels in the lime increased rapidly.
Iron concentrations in the lime rock of
less than approximately 500 mg/kg did
not result in a build-up of iron in the lime
plant and significantly improved recaust
operation. High iron concentrations were
associated with green lime mud and lime
mud filter plugging at Harmac.
—Magnesium and Aluminium: As Fig. 2
shows, trends in magnesium and aluminium
concentrations at Harmac are related.
Investigation of the fluctuations in iron
levels revealed that the magnesium concentration
in the lime rock also varied significantly.
High magnesium concentrations
tended to occur with higher iron
concentrations in the lime rock. Magnesium
concentration in lime rock shipments
varied from a low of approximately
1,000 mg/kg to more than 9,000 mg/kg.
Aluminium concentrations in the lime
rock were only measured in 2004 and
were generally less than 100 mg/kg.
Figure 7 shows a clear relationship
between magnesium and aluminium concentration
in Harmac lime mud back to
2001. These results agree with both Frederick
[13] and Theliander [14] who propose
that the controlling mechanism is
the precipitation of hydrotalcite,
Mg 6 Al 2 CO 3 (OH) 16 , in the lime mud. As
magnesium concentration in the lime
mud increases due to increased input
from lime rock, hydrotalcite formation
may increase in the lime mud.
At Harmac, it appears that the increasing
magnesium concentration in the lime
since 2001 has occurred with a concurrent
increase in aluminium concentration in the
lime and in green and white liquor. This
contrasts with other studies showing that
the addition of magnesium can reduce aluminium
levels in the liquor system [2,13].
—Silicon: The silicon concentration in
lime, lime mud, green liquor and white
liquor was significantly higher at Harmac
than reported values for other mills. Silicon
concentrations were not significantly
affected by sludge burning and were relatively
stable since 2001.
—Phosphorus: Phosphorus accumulated
rapidly in Harmac lime during sludge
burning. At other times, phosphorus levels
were comparable with other mills.
Phosphorus in the lime mud was identified
by XRD as hydroxylapatite,
Ca 5 (PO 4 ) 3 (OH). In the lime kiln, this
compound is converted to tricalcium
phosphate, Ca 3 (PO 4 ) 2 [15]. Lime containing
1 wt% phosphorus will contain 5
wt% tricalcium phosphate, resulting in a
significant dead load in the lime kiln.
High phosphorus concentrations are not
expected to affect lime mud filtration.
Results of Sludge Burning Trial: Burning
secondary sludge increased the NPE input
into the liquor system at Harmac, but did
not result in a net increase of all NPEs in
the recaust and lime cycles.
There was not a noticeable increase in
iron or magnesium concentration in the
recaust or lime cycle. The green liquor
clarifier removed 93% of iron and 99% of
the magnesium.
The clarifier removed 46% of aluminium
and only 5% of the silicon from the
Pulp & Paper Canada T 65 107:3 (2006) ❘❘❘ 39

LIME KILNS
FIG. 6. Visible light spectra of Harmac green liquor.
FIG. 7. Mg versus Al in Harmac lime mud.
raw green liquor. A slight increase in the
concentrations of both elements occurred
in the liquor system.
Very little phosphorus was removed by
the clarifier, and it accumulated significantly
in the lime during sludge burning.
Based on the results of the trial, sludge
burning in the recovery boiler was continued
at a rate of 60 US gal/min.
CONCLUSIONS
1. Green lime mud at Harmac may be
caused by kaolinite or a related aluminosilicate
compound interacting with sulphide
species in the green liquor.
2. Aluminium and silicon concentrations
are higher in Harmac process streams
than in other mills.
3. Purchased lime rock was a major
source of iron and magnesium in the lime
plant.
4. Lime mud filter plugging at Harmac
may be caused by formation of aluminosilicates
in lime mud and was reduced
by lowering the levels of NPEs, especially
from lime rock.
5. Sludge burning significantly increased
phosphorus concentration in lime and
slightly increased levels of aluminium and
silicon in green, white and black liquor.
6. Secondary sludge can be burned in the
recovery boiler at Harmac with good lime
rock quality and regular lime cycle chemical
analysis.
RECOMMENDATIONS
• Obtain consistent lime rock quality to
reduce iron and magnesium concentrations
in the lime cycle.
• Conduct weekly analysis of lime cycle
for Fe, Mg and Al.
ACKNOWLEDGEMENTS
The authors thank Pope & Talbot Ltd. for
permission to present and publish this
paper. They also thank the University of
Victoria Chemistry Department for the
use of their instruments for UV/VIS and
diffuse reflectance measurements. Many
thanks to Arvinder Dhaliwal and Ryan
Glendinning for experimental work.
LITERATURE
1. HARILA, P., KAILA, J. Proven Technology for Secondary
Sludge Disposal in a Recovery Boiler. Proc.,
1995 TAPPI Engineering Conf., Dallas, TX, 293-297
(1995).
2. BJÖRK, M., ENGSTRÖM, I. Handling of Non-Process
Elements for Improved Runnability and Reduced
Environmental Impact. Pulp Paper Can. 103(4):T102-
105 (April 2002).
3. MAGNUSSON, H., MÖRK, K., WARNQVIST, B.
Non-Process Chemical Elements in the Kraft Recovery
System. Proc. 1979 TAPPI Pulping Conference, Seattle,
WA, 77-83 (1979).
4. JEMAA, N., THOMPSON, R., PALEOLOGOU, M.,
BERRY, R. Inorganic Non-Process Elements in the
Kraft Recovery Cycle: Sources, Levels, Process Effects,
and Options for Removal. MR-348. Pointe Claire:
Paprican (1997).
5. KEITAANNIEMI, O., VIRKOLA, N.E. Undesirable
Elements in Causticizing Systems. Tappi J. 65(7): 89-92
(July 1982).
6. RICHARDSON, B., ULOTH, V., LOWNERTZ, P.,
GLEADOW, P., FORGET, C., HOGIKYAN, R.
Behaviour of Non-Process Elements in the Kraft
Recovery System. Proc., 1998 TAPPI International
Chemical Recovery Conference, Tampa, FL QC, 1025-
1039 (1998).
7. DORRIS, G.M., ALLEN, L.H. The Effect of
Reburned Lime Structure on the Rates of Slaking,
Causticizing and Lime Mud Settling. JPPS 11(4):J89-
J98 (1985). Erratum: 11(6): J188 (1985).
8. EITEL, W. Silicate Science. Ceramics and Hydraulic
Binders. Vol. V. New York: Academic Press, A-73: 68
(1966).
9. CLARK, R.J.H., COBBOLD, D.G. Characterization
of Sulfur Radical Anions in Solutions of Alkali Polysulfides
in Dimethylformamide and Hexamethylphosphoramide
and in the Solid State in Ultramarine
Blue, Green and Red. Inorg. Chem.17(11): 3169-3174
(1978).
10. GIGGENBACH, W. Optical Spectra of Highly
Alkaline Sulfide Solutions and the Second Dissociation
Constant of Hydrogen Sulfide. Inorg. Chem. 10(7):
1333-1338 (1971).
11. GIGGENBACH, W. Equilibria Involving Polysulfide
Ions in Aqueous Sulfide Solutions up to 240°.
Inorg. Chem., 13(7): 1724-1730 (1974).
12. WANNENMACHER, P.N., FREDERICK, W.J.,
HENDRICKSON, K.A., HOLMAN, K.L. Application
of Solubility Data to Predicting the Accumulation of
Aluminum and Silicon in Alkaline Pulp. Proc., TAPPI
Minimum Effluent Mills Symposium Proceedings,
Atlanta, Georgia, 303-308 (1996).
13. FREDERICK, W.J. Jr. Managing Nonprocess Element
Flows in Pulp-Mill Operations. In: The Impact of
Energy and Environmental Concerns on Chemical Engineering
in the Forest Products Industry. AIChE Symposium
Series. 80(239): 21-29 (1984).
14. THELIANDER, H. Si and Al Inputs and Behaviour
in the Recovery Cycle. Proc., TAPPI Minimum Effluent
Mills Symposium, Atlanta, Georgia, 295-300
(1996).
15. ULMGREN, P., RÅDESTRÖM, R. The Build-Up of
Phosphorus in a Kraft Pulp Mill and the Precipitation
of Calcium Phosphate from Green and White Liquors.
Proc., TAPPI International Chemical Recovery Conference,
Toronto, ON, B149-B157 (1995).
Résumé: Au cours des dernières années, certaines périodes d’exploitation à l’usine de pâtes
Harmac ont été marquées par des problèmes de boue de chaux de très faible qualité et un fonctionnement
problématique de l’usine de recaustification et du four à chaux. Ces problèmes coïncident
avec la présence de boue de chaux verte. Cette étude porte sur la chimie des éléments non
reliés aux procédés dans le cycle de recaustification et du four à chaux, évalue l’effet de l’incinération
des boues secondaires dans la chaudière de récupération, identifie une cause possible
à la présence de boue de chaux verte et fait des recommandations en vue de réduire l’obstruction
du filtre de boue de chaux et de maximiser l’incinération des boues.
Reference: TAYLOR, K., BOSSONS, D. Investigation of green lime mud at Harmac mill. Pulp
& Paper Canada 106(3): T63-66 (March, 2006). Paper presented at the 2005 Pacific Coast Branch
in Parksville, BC, on April 22 to 23, 2005. Not to be reproduced without permission of PAPTAC.
Manuscript received on April 25, 2005. Revised manuscript approved for publication by the Review
Panel on July 18, 2005.
Keywords: WHITE LIQUOR MUD, LIME KILNS, CAUSTICIZING, SLUDGE DISPOSAL,
COMBUSTION, RECOVERY FURNACES, PLUGGING.
40 ❘❘❘ 107:3 (2006) T 66 Pulp & Paper Canada