ERIK ABSTRACT
LINDSTRM, VILHELM MATHIAS. Integrating Black Liquor Gasification with Pulping
Process Simulation, Economics and Potential Benefits. (Under the direction of Dr. Hasan Jameel and Dr.
Adrianna G. Kirkman.)
Gasification of black liquor could drastically increase the flexibility and improve the profit potential of
a mature industry. The continuous efforts made in the area of black liquor gasification (BLG) are bringing this
technology closer to commercial realization and potential wide-spread implementation. Research exploring the
integration of BLG into the kraft process and the potential of BLG enabled modified pulping technologies on
modern pulping operations is important to support this effort. The following effort is focused on such research,
utilizing laboratory pulping experiments and process simulation. The separation of sodium and sulfur achieved
through gasification of recovered black liquor can be utilized in processes like modified continuous cooking,
split sulfidity and green liquor pretreatment pulping, and polysulfide-anthraquinone pulping to improve pulp
yield and properties.
Laboratory pulping protocols have been developed for these modified pulping
technologies and different process options evaluated. The process simulation work around BLG has led to the
development of a WinGEMS module for the low temperature MTCI steam reforming process, and case studies
comparing a simulated conventional kraft process to different process options built around the implementation
of a BLG unit operation into the kraft recovery cycle. The implementation of gasification, functioning as the
core of wood pulping recovery operations in a biorefinery, would enable the application of modified pulping
technologies while creating a synthetic product gas that could be utilized in the production of value added
products in addition to wood pulp. The evaluated modified pulping technologies have indicated the potential of
yield increases of 1-3% points with improved product quality, and the potential for capital and operating cost
savings relative to the conventional kraft process. Process simulation work has shown that the net variable
operating cost for a pulping process using BLGCC is highly dependent on the cost of lime kiln fuel and the
selling price of green power to the grid. Under the initial assumptions taken in the performed case study, the
BLGCC process combined with split sulfidity or PSAQ pulping operations had net variable operating cost 2-4%
greater than the kraft reference. When comparing the BLG cases to the MCC reference, the net variable
operating cost break even point based on lime kiln fuel cost is about $47/barrel for the split sulfidity and lower
charge polysulfide processes, and about $38/barrel for the higher polysulfide charge process.
This is
significantly lower than assumed kiln fuel price of $60/barrel used in this work. If the sales price for power to
the grid could be increased through green power credits from 3.5 to 6 /KWh cost savings of about $40/ODtP
could be realized in the investigated BLG processes. Other alternatives to improve the process economics
around BLG would be to modify or eliminate the lime kiln unit operations, utilizing high sulfidity green liquor
pretreatment, PSAQ with auto-causticization, or converting the process to mini-sulfide sulfite-AQ.
INTEGRATING BLACK LIQUOR GASIFICATION WITH PULPING PROCESS
SIMULATION, ECONOMICS AND POTENTIAL BENEFITS
by
ERIK VILHELM MATHIAS LINDSTRM
A dissertation submitted to the Graduate Faculty of
North Carolina State University
in partial fulfillment of the
requirements for the Degree of
Doctor of Philosophy
WOOD AND PAPER SCIENCE
Raleigh, North Carolina
2007
APPROVED BY:
_________________________
Dr. Michael Overcash
__________________________
Dr. Hasan Jameel
Chair of Advisory Committee
_________________________
Dr. Hou-min Chang
_________________________
Dr. Adrianna Kirkman
Co-chair of Advisory Committee
For my wife, my mother and father, and for all those that have believed in me
change is good
ii
BIOGRAPHY
The author was born in Stockholm, Sweden in 1975. He graduated from Nacka Gymnasium in 1995,
after having spent one year (1992-1993) in Flint, MI, as an AFS exchange student. While in Flint he attended
Flint Northern HS, participating in the Magnet Program and lettering in 4 sports. Upon completion of the
Swedish Studenten, he returned to the US, matriculating at Wittenberg University in Springfield, OH. He
received a Bachelor of Arts in chemistry with a minor in music theory and composition in 1999. He then went
on to Miami University, Oxford, OH, where he completed a Master of Science in Paper Science and
Engineering, working with Dr. Bill Scott. He now lives with his wife Amy Joyce in Garner, NC and has
accepted a position with MeadWestvaco in Raleigh, NC, where he will begin his career upon completion of this
degree.
iii
ACKNOWLEDGMENTS
The following work is the result of my own relentless persistence and invaluable assistance from a
group of very important people. First and foremost, I would like to thank my wife Amy Joyce, for not
abandoning me during this long process, and for her endless support! I also would like to express my deepest
regards and gratitude for my advisors, Dr. Hasan Jameel and Dr. Adrianna Kirkman. Without their guidance,
efforts and great patience, this undertaking would never have succeeded! They have helped me grow as a
problem solver, researcher, and person beyond what I could ever have expected! I would also like to thank my
other committee members, members of the faculty and staff, and all my new found friends, for their support and
optimism; especially Dr. Sunkyu Park, and Dr. Ved Naithani and Jim McMurray without whom the pulping
effort would have fallen short. I would also like to acknowledge TRI, MTCI and the US Department of Energy,
for their financial and technical support of this project that allowed me to learn so much about so many things.
E.V.M.L., Raleigh, 2007
iv
TABLE OF CONTENTS
LIST OF TABLES..............................................................................................................................................viii
LIST OF FIGURES...............................................................................................................................................x
LIST OF ABBREVIATIONS............................................................................................................................xiv
1 INTRODUCTION...............................................................................................................................................1
1.1 Review of black liquor gasification development.........................................................................................2
1.1.1 Low Temperature Gasifier/Steam Reformer.................................................................................................3
1.1.2 High Temperature Gasifier............................................................................................................................4
1.1.3 Current Status of BLG Technologies.............................................................................................................5
1.2 Review of kraft pulping chemistry.................................................................................................................6
1.2.1 Lignin structure and kraft pulping reactions.................................................................................................7
1.2.2 Carbohydrate structure and kraft pulping reactions...................................................................................11
1.3 Review of modified kraft pulping processes...............................................................................................15
1.3.1 Modified continuous cooking.......................................................................................................................15
1.3.2 Green liquor pretreatment...........................................................................................................................16
1.3.3 Split sulfidity pulping...................................................................................................................................16
1.3.4 Kraft polysulfide pulping with anthraquinone.............................................................................................18
2 STATEMENT OF OBJECTIVES..................................................................................................................19
2.1 Objectives for laboratory pulping work.....................................................................................................20
2.2 Objectives for process simulation work......................................................................................................20
3 SUMMARY, CONCLUSIONS AND FUTURE WORK..............................................................................20
v
3.1 Review of laboratory pulping results..........................................................................................................20
3.1.1 Green liquor pretreatment...........................................................................................................................20
3.1.2 Split sulfidity pulping...................................................................................................................................21
3.1.3 Polysulfide-anthraquinone pulping..............................................................................................................22
3.2 Process simulation.........................................................................................................................................23
3.2.1 Development of BLG model for WinGEMS..................................................................................................24
3.2.2 Simulation of the effects of BLG integration on kraft mill operations.........................................................25
3.3 Conclusions....................................................................................................................................................25
3.3.1 Conclusions from split sulfidity pulping.......................................................................................................25
3.3.2 Conclusions from green liquor pretreatment pulping..................................................................................25
3.3.3 Conclusions from polysulfide pulping with anthraquinone.........................................................................26
3.3.4 Conclusions from WinGEMS simulation work.............................................................................................26
3.4 Future work...................................................................................................................................................27
4 REFERENCES.................................................................................................................................................28
5 ECONOMICS OF INTEGRATING BLACK LIQUOR GASIFICATION WITH PULPING: PART I
EFFECT OF SULFUR PROFILING.........................................................................................................29
6 EFFECTS ON PULP YIELD AND PROPERTIES USING MODIFIED PULPING PROCEDURES
INVOLVING SULFUR PROFILING AND GREEN LIQUOR PRETREATMENT...........................44
7 THE EFFECT OF INTEGRATING POLYSULFIDE PULPING AND BLACK LIQUOR
GASIFICATION ON PULP YIELD AND PROPERTIES......................................................................61
8 THE EFFECT OF INTEGRATING POLYSULFIDE PULPING AND BLACK LIQUOR
GASIFICATION ON PULP YIELD AND DELIGNIFICATION..........................................................79
9 THE DEVELOPMENT AND VALIDATION OF A LOW-TEMPERATURE BLACK LIQUOR
GASIFIER MODEL FOR USE IN WINGEMS.....................................................................................91
10. WINGEMS SIMULATION OF NET PROCESS VARIABLE OPERATING COSTS RESULTING
FROM BLG INTEGRATION WITH SPLIT SULFIDITY AND POLYSULFIDE PULPING.........100
11 INTEGRATING BLACK LIQUOR GASIFICATION AND PULPING AND A REVIEW OF
CURRENT TECHNOLOGY....................................................................................................................111
12 APPENDICES..............................................................................................................................................132
vi
12.1 Notes on pump calibration and operation..............................................................................................133
12.2 Protocol for simulated MCC pulping......................................................................................................134
12.3 Protocol for Polysulfide generation.........................................................................................................135
vii
LIST OF TABLES
Table 1.1 Demonstration of sulfur utilization as Na2S (kraft) or Na2S/PS (kraft-PS) and the system sulfur
availability for PS generation......................................................................................................................19
Table 3.1 Total yield improvement from PS procedures compared to 25% S MCC pulp...........................23
Table 5.1 Effect of Black Liquor Gasification and the H2S to CO2 Co-absorption on the Process............38
Table 5.2 Effect of Black Liquor Gasification and the H2S to CO2 Co-absorption on Cost........................39
Table 5.3 Effect of Split Sulfidity and Yield Increase on the Process.............................................................40
Table 5.4 Effect of Split Sulfidity and Yield Increase on Cost.........................................................................41
Table 5.5 Effect of using Na2CO3 for Pretreatment and %EA Use on the Process......................................42
Table 5.6 Effect of using Na2CO3 for Pretreatment and %EA Use on Cost.................................................43
Table 6.1 Parameters for MCC protocol...........................................................................................................47
Table 6.2 Outline of collection scheme for digester liquor samples................................................................47
Table 6.3 Parameters for Split Sulfidity protocols, high initial and low initial alkali (SS_HIA/SS_LIA)...48
Table 6.4 Parameters for Green Liquor Pretreatment protocols, high initial and low initial alkali
(GLPT_HIA/GLPT_LIA)............................................................................................................................48
Table 6.5 MCC baseline pulp yield, kappa and viscosity.................................................................................51
Table 6.6 Split Sulfidity baseline pulp yield, kappa and viscosity...................................................................52
Table 6.7 Green liquor pretreatment baseline pulp yield, kappa and viscosity.............................................54
Table 6.8 Pulp strength properties for low kappa range cooks.......................................................................57
Table 6.9 Pulp strength properties for high kappa range cooks.....................................................................57
Table 7.1 Parameters for MCC protocol...........................................................................................................63
Table 7.2 Demonstration of sulfur utilization and system availability for PS generation............................64
Table 7.3 Parameters for Polysulfide-Anthraquinone (PSAQ) cooks.............................................................64
Table 7.4 Parameters for medium (MIA) and high (HIA) initial alkali PSAQ cooks...................................64
Table 7.5 Summary of cooks performed using 25% sulfidity..........................................................................65
Table 7.6 Summary of cooks performed using 40% sulfidity..........................................................................67
Table 7.7 Pulp strength properties for 25% sulfidity cooks............................................................................72
viii
Table 7.8 Pulp strength properties for 40% sulfidity cooks............................................................................75
Table 8.1 Parameters for MCC protocol...........................................................................................................81
Table 8.2 Outline of cooks performed using 25% and 40% sulfidity.............................................................82
Table 8.3 Outline of cooks performed at zero initial alkali exploring the effect of alkali profiling.............82
Table 9.1 BLG user defined block parameters.................................................................................................95
Table 9.2 Algorithm for BLG material balances..............................................................................................96
Table 9.3 Predicted BLG process streams and comparisons to MTCI process data....................................98
Table 10.1 Input parameters and key assumptions for WinGEMS case study...........................................104
Table 10.2 Comparison of fiber line process parameters for each simulated case......................................105
Table 10.3 Comparison of chemical recovery process parameters for each simulated case......................105
Table 10.4 Comparison of power and steam process variables for each simulated case............................107
Table 10.5 Comparison of case study variable operating cost parameters and net cost per ODtP produced
......................................................................................................................................................................107
Table 11.1 Demonstration of sulfur utilization as Na2S (kraft) or Na2S/PS (kraft-PS) and the system
sulfur availability for PS generation.........................................................................................................122
Table 11.2 Total yield improvement from PS procedures compared to 25% S MCC pulp.......................122
Table 11.3 Chemical Requirements for Selected Options for production of 1 ODtP (all chemicals as
Na2O) converted to kg/ton.........................................................................................................................124
Table 11.4 Estimated Cost Comparisons of Tomlinson and BLGCC Power/Recovery Systems Relative the
Tomlinson BASE (index = 100), (68).........................................................................................................126
Table 11.5 General Comparison of BLG Enabled Pulping Technologies...................................................127
ix
LIST OF FIGURES
Figure 1.1 Simplified representation of BLGCC power/recovery systems......................................................3
Figure 1.2 Schematic of MTCI Steam Reformer................................................................................................4
Figure 1.3 Formation of the Quinone-Methide Intermediate (QMI) under alkaline conditions and outline
of lignin degradation reactions (xiii).............................................................................................................8
Figure 1.4 Nucleophilic (HS-) addition to the QMI and cleavage of the -aryl ether bond (xiii)...................9
Figure 1.5 Formaldehyde elimination from the QMI generating an enol ether structure and subsequent
condensation between formaldehyde and enol ether resulting in diaryl-methane structures (xiii)........9
Figure 1.6 Electron-transfer reaction reducing the QMI with cleavage of the -aryl ether bond leading to
the generation of coniferyl-type structures (xiii).......................................................................................10
Figure 1.7 Alkaline degradation of non-phenolic lignin structures with epoxide ring opening through
nucleophilic attack generating stable lignin-carbohydrate complexes (LCC) (xiii)...............................10
Figure 1.8 Outline of the alkaline carbohydrate end-wise peeling reaction (xiii)..........................................13
Figure 1.9 Outline of the carbohydrate stopping reaction (xiii)......................................................................13
Figure 1.10 Oxidative peeling reaction through -alkoxy elimination at the C2 position (xiii)...................14
Figure 1.11 Alkaline (Random) hydrolysis of carbohydrate glycosidic bonds generating a new reducing
end group or levoglucosan (xiii)..................................................................................................................15
Figure 1.12 Outline of process using green liquor pretreatment.....................................................................16
Figure 1.13 Schematic of unit operations in split sulfidity pulping.................................................................17
Figure 1.14 Outline of process using polysulfide...............................................................................................18
Figure 3.1 Obtained results for pulping using green liquor pretreatment procedure..................................21
Figure 3.2 Delignification and yield results for split sulfidity pulping............................................................22
Figure 3.1 Input and output stream structure for BLG model.......................................................................24
Figure 5.2 Schematic of a MTCI Steam Reformer...........................................................................................32
Figure 5.3 Schematic for the Production of Conventional White Liquor with BLG....................................34
Figure 5.4 Schematic for the Production of Sulfide Lean and Sulfide Rich White Liquor with BLG........34
Figure 5.5 Schematic for using NaHS and Na2CO3 in Pretreatment with BLG...........................................36
Figure 6.1 Residual Effective Alkali (REA) profiles for MCC procedure......................................................49
Figure 6.2 Residual Effective Alkali (REA) profiles for SS_HIA procedure.................................................50
x
Figure 6.3 Residual Effective Alkali (REA) profiles for SS_LIA procedure..................................................50
Figure 6.4 Residual Effective Alkali (REA) profiles for GLPT modified procedure....................................51
Figure 6.5 Kappa number versus final h factor for the MCC baseline, SS_HIA and SS_LIA cooks..........52
Figure 6.6 Total yield versus kappa for the MCC baseline, SS_HIA and SS_LIA cooks.............................53
Figure 6.7 Viscosity versus kappa for the MCC baseline, SS_HIA and SS_LIA cooks................................53
Figure 6.8. Kappa number versus final h factor for the MCC baseline, GLPT_HIA, GLPT_LIA and
GLPT modified cooks...................................................................................................................................54
Figure 6.9 Total yield versus kappa for the MCC baseline, GLPT_HIA,
GLPT_LIA and GLPT
modified cooks...............................................................................................................................................55
Figure 6.10 Viscosity versus kappa for the MCC baseline, GLPT_HIA,
GLPT_LIA and GLPT
modified cooks...............................................................................................................................................55
Figure 6.11 Refining response for low kappa pulps..........................................................................................56
Figure 6.12 Refining response for high kappa pulps........................................................................................56
Figure 6.13 Tensile index versus tear index for low kappa pulps...................................................................58
Figure 6.14 Burst index versus tear index for low kappa pulps......................................................................58
Figure 6.15 Tensile index versus tear index for high kappa pulps..................................................................59
Figure 6.16 Burst index versus tear index for high kappa pulps.....................................................................59
Figure 7.1 Kappa number versus AA charge for the MCC baseline and PS cooks at 25% sulfidity..........66
Figure 7.2 Total yield versus kappa for the MCC baseline and PS cooks at 25% sulfidity..........................66
Figure 7.3 Viscosity versus kappa for the MCC baseline and PS cooks at 25% sulfidity.............................67
Figure 7.4 Kappa number versus AA charge for 25% and 40% sulfidity cooks...........................................68
Figure7.5 Total yield versus kappa for 25% and 40% sulfidity cooks...........................................................68
Figure 7.6 Viscosity versus kappa for 25% and 40% sulfidity cooks.............................................................69
Figure 7.7 Kappa number versus AA charge for the MCC baseline and PS cooks at 40% sulfidity..........69
Figure 7.8 Total yield versus kappa for the MCC baseline and PS cooks at 40% sulfidity..........................70
Figure 7.9 Viscosity versus kappa for the MCC baseline and PS cooks at 40% sulfidity.............................70
Figure 7.10 Refining response for 25% sulfidity pulps....................................................................................71
Figure 7.11 Refining response for 40% sulfidity pulps....................................................................................71
Figure 7.12 Tensile index versus tear index 25% sulfidity cooks....................................................................72
xi
Figure 7.13 Burst index versus tear index for 25% sulfidity cooks.................................................................74
Figure 7.14 Tensile index versus tear index for 40% sulfidity cooks..............................................................76
Figure 7.15 Burst index versus tear index for 40% sulfidity cooks.................................................................76
Figure 8.1 Kappa number versus initial alkali charge for 25% sulfidity pulps.............................................83
Figure 8.2 Normalized total yield versus initial alkali charge for 25% sulfidity pulps.................................83
Figure 8.3 Kappa number versus initial alkali charge for 40% sulfidity pulps.............................................84
Figure 8.4 Normalized total yield versus initial alkali charge for 40% sulfidity pulps.................................84
Figure 8.5 Comparing the effect of alkali profiling on delignification rate and pulp yield for zero initial
alkali PSAQ cooks (25% and 40% sulfidity).............................................................................................85
Figure 8.6 The effect of alkali profiling on residual effective alkali for MCC at 25% S..............................86
Figure 8.7 The effect of alkali profiling on residual effective alkali for PSAQ at 25% S.............................86
Figure 8.8 The effect of alkali profiling on REA using for PSAQ 25% S at zero initial alkali charge........87
Figure 8.9 The effect of alkali profiling on residual effective alkali for MCC at 40% S..............................87
Figure 8.10 The effect of alkali profiling on residual effective alkali for PSAQ at 40% S...........................88
Figure 8.11 The effect of alkali profiling on REA using for PSAQ 40% S at zero initial alkali charge......88
Figure 9.1 Schematic of a MTCI Steam Reformer...........................................................................................94
Figure 9.2 Stream structure for BLG model.....................................................................................................94
Figure 9.3 WinGEMS simulation using integrated BLG block model...........................................................97
Figure 10.1 Schematic of a MTCI Steam Reformer.......................................................................................102
Figure 10.2 Stream structure for BLG model.................................................................................................103
Figure 10.3 The effect of lime kiln fuel price on net variable operating cost, keeping all other cost factors
constant (assuming green power sales price at 3.5/KWh).....................................................................108
Figure 10.4 The effect of green power sales price on net variable operating cost, keeping all other cost
factors constant (assuming lime kiln fuel price at $60/barrel)...............................................................109
Figure 11.1 Simplified representation of BLGCC power/recovery systems................................................114
Figure 11.2 Schematic of MTCI Steam Reformer..........................................................................................115
Figure 11.3 Schematic of unit operations in split sulfidity pulping...............................................................118
Figure 11.4 Delignification and yield results for split sulfidity pulping........................................................119
Figure11.5 Outline of process using green liquor pretreatment....................................................................120
xii
Figure 11.6 Obtained results for pulping using green liquor pretreatment.................................................120
Figure 11.7 Outline of process using polysulfide............................................................................................121
Figure 11.7 Total pulp yield and ISO Brightness versus kappa for Kraft, AS-AQ and MSS-AQ.............124
Figure 11.8 Schematic of the alkaline sulfite pulping processes with the RTI absorber.............................126
xiii
LIST OF ABBREVIATIONS
AQ
BAR
BL
BC
BLG
BLGCC
CC
DP
CPS
GL
HIA
HRSG
KFURN
LIA
MCC
MIA
OD
ODtP
PS
PSAQ
SS
WINGEMS
ZAP
anthraquinone
benzylic acid rearrangement
black liquor
base case (reference for case study comparison)
black liquor gasification, and BL gasifier module in WinGEMS
black liquor gasification combined cycle
combined cycle (power production)
degree of polymerization
centipoises (unit)
green liquor
high initial alkali charge
heat recovery steam generator
kraft recovery furnace module in WinGEMS
low initial alkali charge
modified continuous cooking
medium initial alkali charge
oven dry
oven dry metric ton pulp
polysulfide
polysulfide anthraquinone (pulping)
split sulfidity (pulping)
windows version of general energy and mass balance system (software name)
zero effective alkali in pretreatment
xiv
1
INTRODUCTION
The kraft or sulfate process is the dominant chemical pulping technology employed in the paper
industry today. The competitive advantage that has led to its position is the capability to convert most wood
species to high strength pulp combined with an efficient chemical recovery based around the Tomlinson
recovery boiler. There are approximately 280 Tomlinson recovery boilers in operation in North America,
according to the Black Liquor Recovery Boiler Advisory Committee of the American Forest and Paper
Association (AF&PA). Many of these boilers are nearing the end of their useful life, and will either need to be
upgraded or replaced in the near future. Black liquor gasification (BLG) is one technology that could be
implemented into the kraft process to replace a mature technology, while creating new opportunities in process
operations and the potential for enhanced competitiveness.
The underlying fundamental for implementation of any new technology is the impact it will have on
overall process economics. Some deciding factors that will influence the implementation of BLG involve the
cost-benefits associated with power generation and other high-value products that can be derived from the
syngas.
Another area of importance is the potential cost-savings that can be realized through process
modifications and optimization. The effect on wood, chemical and fuel demand from changes in the pulping
process can have a significant effect on the variable operating cost, capital investment and maintenance costs.
Therefore, research exploring the impact of BLG on pulping technologies will be of great importance for the
eventual implementation of this technology. The following effort explores the potential benefits realizable
through the implementation of BLG integrated with the modified pulping technologies that it enables.
Presently, in a typical chemical pulp mill the black liquor is concentrated to greater than 65% dissolved
solids and burned in a recovery boiler. The pulping chemicals are recovered in the smelt and the heat energy is
converted to steam, which is used in a steam turbine generator to produce electricity. The typical thermal
efficiency of a recovery boiler is generally 65-70%, and the thermal efficiency of the Rankine cycle for the
conversion of steam to electricity varies from 30-38%, depending on the temperature and pressures of the
different streams in the cycle. These values result in an overall system thermal efficiency of about 23% ( i). On
the other hand, if the black liquor is gasified, the syngas can after cleanup be combusted in a combined cycle for
production of electricity. Combined cycle power generation entails the sequential utilization of a gas turbine
followed by a steam turbine. The fuel gas is first burned in a gas turbine to produce electricity. The hot exhaust
gas from the turbine is then passed through a heat exchanger to produce steam which is then used in a powerproducing steam turbine. Implementing a gasifier with combined cycle cogeneration of power will increase the
electricity production of the mill. A conventional steam cycle produces about 120-180 kWh/ton of steam, but a
gasifier along with combined cycle power generation has the potential to generate 600-1000 kWh/ton of steam
(ii). Such a production of power would turn a pulp mill into a net exporter of electricity, and this potential is the
main motivation for the implementation of black liquor gasification.
1
In addition to the increased energy efficiency, gasification of black liquor has several other benefits
relative to the traditional combustion recovery process. The BLG process operation is inherently very stable
and flexible with regard to feedstock and load requirements. It is possible to process almost any biomass
material and stable operation can be maintained despite upset feedstock flows, even complete interruptions.
BLG has the potential to revolutionize the chemical recovery cycle and, through the separation of sodium and
sulfur, enable the utilization of modified pulping technologies such as:
Green Liquor Pretreatment
Split Sulfidity Pulping
Polysulfide Pulping
These pulping technologies will increase yield or reduce wood demand, improve product quality,
decrease chemical usage and more importantly simplify the chemical recovery process. A simplification of the
chemical recovery process will decrease the operating and capital costs for recovery. BLG would also decrease
the malodor associated with the kraft process. Besides power generation, the resulting syngas can be used to
generate bio-derived liquid fuels, bio-derived chemicals for the synthetic chemical and pharmaceutical
industries, as well as H2 for use in fuel cells.
Despite these benefits and opportunities, high capital cost and risk associated with new process
implementation are impeding the implementation of BLG technologies in the industry. However, the synergy
between BLG as an increased energy generator and as an enabler of advanced pulping processes should increase
the financial attractiveness of these new process concepts.
1.1 Review of black liquor gasification development
Figure 1.1 shows the typical process elements included in the gasification of black liquor. The black
liquor is initially introduced into a process vessel, the black liquor gasifier, which can either be pressurized or
operate under atmospheric pressure. In general terms, the process involves the conversion of hydrocarbons and
oxygen to hydrogen and carbon monoxide while forming separate solid and gaseous product streams.
The inorganic material, including all sodium salts, leaves as a bed solid or smelt depending on the
gasifier operating temperature. The bed solids or smelt is then slaked and recausticized to form a caustic
solution. The volatiles, including most of the reduced sulfur species, leave as a syngas of medium BTU value.
The major components of the syngas are H2S, CO2, CO, H2O, and H2. To prepare the syngas for other
applications and to regenerate the pulping liquor, all sulfur must be separated from the syngas, and dissolved
into the caustic solution prepared from the bed solids. The clean product gas is burned in a gas turbine and the
hot flue gases are combined and used to generate steam in heat recovery steam generators (HRSGs). This steam
is then used in a steam turbine and other process applications.
2
Figure 1.1
Simplified representation of BLGCC power/recovery systems
Whitty and Verrill has given a review of the development of alternative recovery technologies to the
Tomlinson recovery boiler (iii). The following discussion will focus on the gasification processes currently in
commercial operation. As suggested by Stigsson, black liquor gasification technologies can be classified by the
operating temperature (iv). High temperature gasifiers operate at about 1000oC and low temperature gasifiers
operate at less than 700oC. In the high temperature gasifier, the inorganic material forms a smelt and leaves in
the molten form, while in the low temperature system, they leave as solids. The fuel value of the syngas
produced is also dependent on the gasifying technology. Typically, gasification produces a fuel gas with
heating values of 3-4 MJ/Nm3 using air and 89 MJ/Nm3 using oxygen (v).
1.1.1
Low Temperature Gasifier/Steam Reformer
The development of low temperature fluidized bed gasifiers is being pursued by ThermoChem
Recovery International (TRI) in the USA and by ABB in Sweden. The TRI system uses steam reforming to
generate the product gases.
As opposed to exothermic incineration or combustion technologies, steam
reforming is an endothermic process. The steam reforming vessel operates at atmospheric pressure and at a
medium temperature. The organics are exposed to steam in a fluidized bed in the absence of air or oxygen with
the following reaction:
H2O + C + Heat = H2 + CO
The carbon monoxide produced in this first reaction then reacts with steam to produce more hydrogen and
carbon dioxide.
CO + H2O = H2 + CO2
The result is a synthesis gas made up of about 65% hydrogen.
3
The TRI Steam Reformer technology, as shown in Figure 1.2, consists of a fluidized bed reactor that is
indirectly heated by multiple resonance tubes of one or more pulse combustion modules. Black liquor is
directly fed to the reactor, which is fluidized with superheated steam. The black liquor uniformly coats the bed
solids, producing a char and volatile pyrolysis products which are steam cracked and reformed to produce a
medium BTU gas. The residual char retained in the bed is more slowly gasified by reaction with steam. The
sulfur and sodium are separated in that the sulfur becomes part of the gas stream and the sodium stays in solid
form.
Bed temperatures are maintained at 605-610 C, thereby avoiding liquid smelt formation and the
associated smelt-water explosion hazards.
Figure 1.2
Schematic of MTCI Steam Reformer
Product gases are routed through a cyclone to remove the bulk of the entrained particulate matter and
are subsequently quenched and scrubbed in a Venturi scrubber. A portion of the medium-Btu product gases can
be supplied to the pulse combustion modules, and the combustion of these gases provides the heat necessary for
the indirect gasification process. Low temperature gasification leads to complete separation of the sulfur and
sodium in kraft black liquor to the gas and solid phases, respectively. Bed solids are continuously removed and
mixed with water to form a carbonate solution. The inorganic chemical in the bed solids as well as the sulfur
from the gas stream are recovered and used as cooking liquors for the mill. The product gas residence time in
the fluid bed is about 15 seconds because of the deep bed (20 ft) used, while the solids residence time is about
50 hrs. These conditions promote extensive tar cracking and carbon conversion. In summary the steam
reforming reactor vessel has three inputs; fluidizing steam, black liquor, and heat, and has three outputs; bed
solids, hydrogen rich product gas, and flue gas (vi).
1.1.2
High Temperature Gasifier
High temperature gasification stems from work initiated by SKF in the 1970s. The original patent for
the technology was issued in 1987, and it has since been developed through a sequence of demonstration
4
projects.
The gasifier, as developed by Chemrec, is a refractory-lined entrained-flow reactor.
In high
temperature gasification (900-1000 C), concentrated black liquor is atomized, fed to the reactor and
decomposed under reducing conditions using air or oxygen as the oxidant. The initial chemical reactions
involve char gasification and combustion, and are influenced by physical factors like droplet size, heating rate,
swelling, and the sodium and sulfur release phenomena.
The resulting products, smelt droplets and a
combustible gas, are then brought into direct contact with a cooling liquid in a quench dissolver. The two
phases are separated as the smelt droplets dissolve in the cooling liquid, forming green liquor. The exiting
product gas is subsequently scrubbed and cooled for use in other unit operations. The split of sodium and sulfur
between the smelt and gas phase is dependent on the process conditions. Typically, most of the sulfur leaves
with the product gas and essentially all of the sodium with the smelt (vii,viii,ix).
1.1.3
Current Status of BLG Technologies
The TRI steam reformer has been installed in two locations in North America, at the Norampac Mill at
Trenton, New Jersey and the Georgia Pacific Mill at Big Island, Virginia. The Trenton mill produces 500 tpd of
corrugating medium using a sodium carbonate based pulping process.
Prior to the start-up of the low-
temperature black liquor gasifier in September 2003, the mill had no chemical recovery system. For over forty
years the mills spent liquor was sold to local counties for use as a binder and dust suppressant on gravel roads.
This practice was discontinued in 2002. The capacity of the spent liquor gasification system is 115 tpd of black
liquor solids, and the syngas is burned in an auxiliary boiler (vi).
Georgia-Pacifics mill at Big Island, produces 900 tpd of linerboard from OCC and 600 tpd of
corrugating medium from mixed hardwoods semi-chemical pulp. Like the Trenton mill, the Big Island mill
uses a sodium carbonate process. In the past, the semi-chemical liquor was burned in two smelters providing
chemical recovery but no energy recovery. Instead of replacing the smelters with a traditional recovery boiler
Georgia-Pacific decided to install a low temperature black liquor gasification process. One difference between
the two systems is that unlike Trenton, Big Island burns the generated product syngas in the pulsed combustors,
so the product gas exiting the reformer vessel is cleaned prior to combustion (vi). The Big Island BLG project
was terminated in the fall of 2006.
The evolution of the high temperature gasifier has taken the technology from an air-blown process near
atmospheric pressure to a high pressure (near 30 atm.) oxygen-blown process. Benefits realized through high
pressure oxygen-blown operation are higher efficiencies, higher black liquor throughput and improved
compatibility with down stream unit operations such as combined cycle power generation.
An air-blown pilot plant at Hofors, Sweden, was developed to verify the possibility of gasifying black
liquor using an entrained-flow reactor operating at 900-1000 C. The project showed that green liquor of
acceptable quality could be generated; and the plant was dismantled in 1990. The Fr vi, Sweden plant was
5
designed as a capacity booster for the AssiDomn facility and was operated from 1991 to 1996, demonstrating
the potential for black liquor gasification at a commercial scale. During its operation several technical problems
were encountered and addressed. The identification of a suitable material for the refractory lining remained a
problem. A subsequent commercial project was initiated in 1996 at the Weyerhauser plant in New Bern, North
Carolina. The black liquor gasifier was more or less a scale-up of the Frvi plant, designed for a capacity of
300 tons of dissolved solids/day. In 1999 the process maintained greater than 85% availability. However, over
the course of the project the plant experienced several technical problems, mainly related to the refractory
lining, and it was shut down after cracks in the reactor vessel were discovered in 2000. After detailed studies
and re-engineering, the gasifier operation at New Bern was resumed in the summer of 2003 ( vii). During the
rebuild, it was retrofitted with spinel refractory materials developed at Oakridge National Labs in cooperation
with other partners. The refractory material is in its second year of operation. The gasifier can burn up to
730,000 lb/day of solids or about 20% of the mill production ( x). The syngas generated in the gasifier is
currently burned in a boiler.
A pressurized air-blown demonstration project was established at the Stora Enso plant at Skoghall,
Sweden, in 1994. The project showed the capability of a pressurized system to generate acceptable quality
green liquor while maintaining high carbon conversion ratios. The process was converted to an oxygen-blown
operation in 1997 resulting in a capacity increase of more than 60%. A second pressurized demonstration plant
was completed in Pite, Sweden, in 2005. The purpose of the project is to demonstrate high pressure operation
(near 30 atm.) with associated gas cooling and sulfur handling unit operations required for a full-scale BLG
process. Funding has been obtained for a scale-up project of the Pite facility. The plant is designed for a
capacity of 275-550 tDS/day and encompasses all the required unit operations, including the power island, for a
BLG process with combined-cycle power generation (vii).
1.2 Review of kraft pulping chemistry
To liberate the cellulose fiber contained in wood, lignin, the glue which holds the wood together,
must be degraded and solubilized into the pulping liquor. In kraft pulping, this is achieved by heating the wood
in the presence of an alkaline pulping liquor containing nucleophiles that attack the lignin polymer. The active
chemical agents in kraft pulping chemistry are the hydroxide (OH-) and hydrosulfide (HS-) anions. While these
chemicals act to degrade wood lignin, the alkaline conditions and elevated temperatures also degrade the wood
carbohydrates, cellulose and hemicelluloses, resulting in overall pulp yield losses. Several reviews of kraft
pulping chemistry and its effects on wood component degradation and pulp yields have been given (xi,xii,xiii).
Delignification in kraft pulping has three distinct phases: a rapid initial phase, followed by the bulk phase where
most of the lignin is degraded and solubilized, and the final residual phase ( xiv). In the initial phase a substantial
amount of hemicelluloses undergo deacetylation and dissolution resulting in significant yield losses ( xv).
Phenolic lignin structures also undergo some degradation while the effect on cellulose is minor. In the bulk
phase about 70% of the lignin is degraded, while the carbohydrates undergo further degradation through peeling
6
and alkaline hydrolysis reactions.
Methanol and hexenuronic acid are formed during this phase.
Delignification reaches the residual phase when about 90% of the lignin has been removed. At this point,
delignification reactions slow down as reactive lignin moieties have been depleted, and the remaining alkali
generates rapid carbohydrate degradation (xvi). Thus, there is a state of diminishing returns for lignin removal
from wood in kraft pulping relative to overall yield losses from carbohydrate degradation. A great deal of
research has been performed discerning the reactivities of lignin, cellulose and hemicellulose, generating
knowledge that can be applied to improve lignin degradation while protecting the cellulose and hemicellulose
and improving pulp yield and properties. The following is a brief review of the chemistries involved in lignin
and carbohydrate degradation during the kraft process.
1.2.1
Lignin structure and kraft pulping reactions
The lignin macromolecule has been shown to consist of a complex three-dimensional network of 9-
carbon phenylpropane subunits, mainly p-hydroxyphenylpropane, guiacylpropane and syringylpropane. The
corresponding hydoxycinnamyl alcohols, or monolignols, involved in lignin biosynthesis are p-coumaryl (4hydroxy-cinnamyl), coniferyl (3-methoxy-4-hydroxy-cinnamyl) and sinapyl (3,5-dimethoxy-4-hydroxy
cinnamyl) (xvii,xviii,xix,xx). These lignin precursors are synthesized through the phenylpropanoid pathway (xxi). In
lignin biosynthesis these subunits are polymerized through radical coupling reactions forming a globular
polylignol macromolecule.
The relative proportions of guiacylpropane (G), syringylpropane (S) and p-
hydroxyphenyl propane (H), vary between different wood species, different tissue types and also within the cell
wall layers. This variation results in very different reactivities of lignin during chemical pulping of different
types of wood. The major component in softwood lignin is guiacyl (G-lignin), whereas hardwood lignin
contains as much as 50% syringyl (S-lignin). Softwood lignin is generally described as fairly uniform, and
values for the ratios of G, S and H lignin (G:S:H) in softwood have been reported for pine (Pinus taeda) (95:1:5)
and spruce (Picea abies) (86:2:13). Hardwood lignin is described as having a greater variability in lignin
composition with syringyl levels varying from 20 to 60%. The relative composition of lignin in Beech (Fagus
sylvatica) has been reported as (56:40:4), indicating a much higher level of syringyl groups relative to softwood
lignin. Also involved in lignin biosynthesis is the addition of water and carbohydrates to quinone-methide
intermediates. The combination of lignin and carbohydrates form a nonphenolic lignin-carbohydrate complex,
which is more difficult to degrade during chemical pulping (xiii).
Lignin degradation is achieved by cleaving the linkages that bind lignin subunits within the lignin
macromolecule. The reactivity of these different lignin moieties is dependent on their structure and chemistry.
The majority of these inter-lignin bonds are -O-4 and -0-4 ether linkages between different phenylpropane
units. Other linkage types are the 4-O-5 ether linkage between phenyl carbons, and carbon-carbon bonds like 55, -1, -5, and -. The basic reaction sequence involved in lignin degradation during kraft pulping is initial
ionization of a free phenolic hydroxyl group, followed by cleavage of the -O-4 and -0-4 ether linkage and the
liberation of additional free phenolic groups (xvi). Under typical kraft pulping conditions, which are strongly
7
alkaline (pH 11-14), ionization of the lignin C4 free phenolic hydroxyl group (pKa 10.9) readily takes place.
If the -carbon of the formed phenolate anion is not directly linked to another carbon atom, the charged species
undergoes -elimination of the aryl substituent at the -position forming a para-quinone methide intermediate.
This structure is the key-intermediate in lignin degradation reactions (xxii). After formation of the quinone
methide several types of reactions may occur. These reaction pathways can be divided into nucleophilic
addition reactions, elimination reactions, and electron transfer reactions.
The general outline of lignin
degradation is outlined in Figure 1.3. Competing with these lignin degrading reaction pathways are lignin
Figure 1.3
Formation of the Quinone-Methide Intermediate (QMI) under alkaline conditions and
outline of lignin degradation reactions (xiii)
condensation reactions where carbon-carbon bonds are formed between the quinone methide and
reactive carbon species present in the pulping liquor, like the phenoxide anion and formaldehyde (xvi).
Cleavage of the - and -O-4 ether linkages by nucleophilic addition reactions has been proposed as
the major reaction pathway during the initial phase of the cook (xvi). In nucleophilic addition reactions, all
nucleophiles present in the pulping liquor compete for the available quinone methide. In kraft pulping the
strongest available nucleophile is HS-, followed by OH- and weaker nucleophiles like phenoxide anions and
other anionic species originating from carbohydrates. The reaction pathway is shown in Figure 1.4. After HS addition to the quinone methide -position, the intermediate undergoes intramolecular attack at the -carbon
forming a thiirane intermediate (episulfide). The episulfide intermediate is decomposed forming elemental
sulfur, which reacts with HS- to form small amounts of polysulfide, and unsaturated coniferyl-type structures
(xiii).
8
Figure 1.4
Nucleophilic (HS-) addition to the QMI and cleavage of the -aryl ether bond (xiii)
Elimination of the -hydroxymethyl group from the quinone methide intermediate produces
formaldehyde and an enol-ether structure without significant cleavage of the -aryl ether bond. This reaction
pathway, shown in Figure 1.5, is predominant in soda pulping, where HS - is not present, but has been observed
at the beginning of the bulk phase in kraft pulping (xxiii). The resulting -aroxy styrene-type structures and
diaryl-methane-type structures formed from condensation with liberated formaldehyde are very stable in alkali
solutions. The -hydrogen can also undergo base-induced elimination reactions. The presence of HS - also
causes partial demethylation of lignin methoxyl groups. The resulting methyl mercaptans reacts further as
nucleophiles combining with an additional methoxyl group to form dimethyl mercaptans, which is extremely
volatile and produces the odor typically associated with kraft mills.
Figure 1.5
Formaldehyde elimination from the QMI generating an enol ether structure and
subsequent condensation between formaldehyde and enol ether resulting in diarylmethane structures (xiii)
The quinone methide can also undergo electron-transfer reactions with reducing compounds present in
the pulping liquor, like carbohydrates and anthrahydroquinone when used as an additive. The proposed reaction
mechanism, outlined in Figure 1.6, involves single-electron transfer and radical intermediates, resulting in
cleavage of the -aryl linkage and formation of coniferyl-type structures (xxiv).
9
Figure 1.6
Electron-transfer reaction reducing the QMI with cleavage of the -aryl ether bond
leading to the generation of coniferyl-type structures (xiii)
In addition to the degradation of phenolic lignin structures previously described, non-phenolic lignin
structures also undergo degradation in kraft pulping, but demand higher temperatures and alkalinity. This
means that most of the non-phenolic lignin degradation takes place in the bulk phase of the cook. The proposed
reaction mechanism, outlined in Figure 1.7, involves ionization of the lignin -hydroxyl group followed by
formation of an oxirane (epoxide) intermediate and cleavage of the -aryl ether bond through internal
nucleophilic substitution (SN1). The epoxide can then be opened by nucleophilic attack (SN2) by HS-, OH-, or
hydroxyl groups of carbohydrates present in the liquor. The reaction with carbohydrates leads to the formation
of lignin-carbohydrate ether bonds that are somewhat stable during bulk delignification conditions (xxv).
Figure 1.7
Alkaline degradation of non-phenolic lignin structures with epoxide ring opening
through nucleophilic attack generating stable lignin-carbohydrate complexes (LCC)
(xiii)
In practice kraft pulping operations are typically interrupted during the bulk phase of delignification to
prevent excessive degradation of carbohydrates, which would result in overall yield loss and poor pulp quality.
10
The residual lignin, typically about 4-5% for softwood and 3% for hardwood, can be removed during
subsequent oxygen delignification, or addressed during bleaching operations (xvi).
1.2.2
Carbohydrate structure and kraft pulping reactions
The carbohydrates in wood consist of cellulose and hemicellulose. The structures and chemistry of
cellulose and hemicelluloses in wood has been established and studied in detail. The following discussion is a
brief synopsis based on several reviews creating a foundation for ensuing discussions regarding modified
pulping technologies aimed at yield improvement (xii,xiii,xvi,xxvi).
Cellulose is a linear homopolymer consisting of -1,4-glucosidic linked D-glucopyranose units. The
cellobiose disaccharide, consisting of two glucose molecules, is the basic repeating unit in the polymer chain,
which in typical papermaking fibers has a weight averaged degree of polymerization of 600-1500. The terminal
C1 group of the cellulose polymer is present in the form of a hemiacetal generating a reducing hydroxyl group
at the C1 position. The other terminal C4 hydroxyl is nonreducing. Although at first look a very simple
polymer, the supramolecular structure of cellulose is a complex matrix containing parallel homoglucan chains
that exhibit intermolecular hydrogen bonds.
These intermolecular bonds cause a varying degree of
organization, where highly ordered domains exhibit a distinct X-ray pattern.
These domains are called
crystallites or crystalline regions, and the less-ordered domains amorphous or noncrystalline regions. The
crystalline regions are more stable than the amorphous regions to alkaline degradation. Cellulose encountered
in wood consists of microfibrils, which are rod-like structures of parallel homoglucan chains exhibiting a twofold screw symmetry around the chain axis. The helical shape is due to the repeating -1,4-glucosidic linkages
between the glucose molecules in the polymer (xxvi).
Wood hemicelluloses are in general terms short linear heteropolymers of different monosaccharides
with typical DP of 50 to 200. The polymer backbone can be made up of one repeating sugar unit (e.g. the xylan
homopolymer) or with two or more (e.g. the glucomannan heteropolymer). Various substituents can be linked
to these polymer chains creating a variety of different types of hemicelluloses. The relative quantity, structure
and composition of xylan and glucomannan hemicelluloses encountered in softwoods are different than those
encountered in hardwoods. O-acetylgalactoglucomannan is the major constituent in softwood hemicelluloses,
making up as much as 18% of the wood, and arabino-4-O-methylglucoronoxylan is the minor constituent,
representing about 10% of the wood. In hardwoods, however, O-acetyl-4-O-methylglucoronoxylan is the most
prevalent hemicellulose, making up 20-35% of the wood, with glucomannan a minor constituent of 2-4% (xiii).
Due to the prevalence of side chain substituents, xylans are more stable to alkaline degradation than
glucomannans. However, compared to cellulose, hemicelluloses are more susceptible to peeling reactions at
low temperatures. Combined with their small DP, this leads to very high losses of hemicelluloses in kraft
cooking, especially during the initial stages of the cook where little degradation of lignin and cellulose takes
place.
11
During the course of a kraft cook, wood carbohydrates undergo various reactions that lead to polymer
degradation and dissolution which results in overall pulp yield loss and decreased pulp viscosity. As wood
carbohydrates are degraded during the cook, a substantial amount of carboxyl groups are liberated which
consumes available alkali resulting in a lower pH. Thus, the process environment is somewhat dynamic with
regard to alkali concentration and temperature, resulting in a complex environment where alkaline lignin
degradation competes with less desirable carbohydrate degradation. Focusing on carbohydrate reactions, the
strongly alkaline environment and elevated temperatures generate a progression of reactions beginning with
deacetylation of hemicelluloses at temperatures below 70 C. As the temperature is increased, reactions leading
to degradation of both hemicellulose and cellulose are initiated. There are three primary reaction types: endwise peeling of the polymer reducing end, oxidative peeling which cleaves the polymer backbone randomly,
and alkaline hydrolysis, or secondary peeling, which also involves direct cleavage of the polymer chain but
requires temperatures above 140 C (xiii,xvi). As the normal pulping temperatures are well above these levels,
the reactions may take place simultaneously and their relative rates are dependent on the polymeric structures of
available carbohydrates, alkalinity and temperature. Competing with these degradation reactions are so called
stopping reactions which stabilize carbohydrate polymers against peeling through oxidation. Other reactions
that take place in kraft pulping are fragmentation reactions, dissolution of hemicellulose, the elimination of
methanol from 4-O- methylglucuronic acid and the generation of hexenuronic acid. In the final stages of the
cook, some dissolved hemicelluloses also re-precipitate on the fiber surface, improving the pulp yield and
altering the fiber mechanical properties (xxvii).
The end-wise peeling reaction, shown in Figure 1.8, is initiated by an alkali catalyzed rearrangement of
the carbohydrate nonreducing end group, forming an enediol anion intermediate. The intermediate undergoes
elimination of the cellulose chain in the -position, producing a cleaved dicarbonyl species. This dicarbonyl is
very unstable under alkaline conditions, and undergoes further degradation reactions resulting mainly in the
formation of isosaccharinic acid or 2,5-dihydroxypentanoic acid. The initial rearrangement of the carbohydrate
reducing end group is the rate-limiting step for the endwise peeling reaction. The enolization is promoted by
higher OH- concentrations (xv).
12
Figure 1.8
Outline of the alkaline carbohydrate end-wise peeling reaction (xiii)
Competing with the peeling reaction is the stopping reaction, displayed in Figure 1.9. In the stopping
reaction, the reducing end group can be stabilized either through oxidation forming its corresponding aldonic
acid, or through conversion to metasaccharinic acid or 2-hydroxy-2-methyl-3-alkoxy-propanioc acid. The
relative reaction rates of peeling and stopping reactions leads to peeling of about 50 to 60 monosaccharides
prior to stabilization by the stopping reaction (xiii).
Figure 1.9
Outline of the carbohydrate stopping reaction (xiii)
The oxidative peeling reaction is facilitated by the presence of oxidized groups contained in the
polymer chain. Keto groups or aldehyde groups in the C2, C3, or C6 position of the carbohydrate monomers
can undergo alkali catalyzed -alkoxy elimination, resulting in cleavage of the polymer chain. In addition,
elimination reactions induced by C2 keto groups and aldehydes at the C6 or anomeric carbon leads to
generation of new reducing end groups. These can undergo further end-wise peeling reactions, resulting in
13
increased carbohydrate degradation. Figure 1.10 shows the mechanism for oxidative peeling through -alkoxy
elimination at the C2 position (xiii,xvi).
Figure 1.10
Oxidative peeling reaction through -alkoxy elimination at the C2 position (xiii)
Alkaline hydrolysis reactions, also called random hydrolysis, take place at temperatures above 140 C.
They involve the cleavage of the glycosidic linkage of the carbohydrate polymer backbone, resulting in new
reducing end groups that can undergo further degradation through (secondary) peeling.
Although this
degradation reaction has little effect on overall pulp yield, it is responsible for the majority of cellulose
degradation in the kraft process, leading to lower degrees of polymerization, and associated losses in fiber
strength. The reaction, outlined in Figure 1.11, is initiated by alkali induced ionization of the C2 hydroxyl
group. The high temperature promotes a conformational change in the reacting monomer from 4C1 to 1C4
structure.
This change results in the rearrangement from equatorial to axial positions of all monomer
substituents, leading to internal nucleophilic displacement (SN1) where the ionized C2 hydroxyl group forms an
oxirane (1,2-epoxide) with elimination of the cellulose chain from C1. The 1,2-epoxide intermediate formed
can undergo further reactions resulting in a new reducing end group, either through reaction with OH-, or
through further rearrangement with the ionized C6 hydroxyl group producing levoglucosan (xiii).
14
Figure 1.11
Alkaline (Random) hydrolysis of carbohydrate glycosidic bonds generating a new
reducing end group or levoglucosan (xiii)
The carbohydrate degradation reactions that occur during heat-up and the early stages of the cook
liberate or produce a substantial amount of uronic, acetic and sugar acids. As a consequence, by the time the
cook reaches cooking temperature and the bulk delignification phase, a substantial amount of the available
alkali has already been consumed to neutralize the acids formed. This means that a much smaller amount of
alkali is available for lignin degradation than what was initially charged. Kraft process modifications that
would optimize the efficiency of alkali utilization could thus lead to improved delignification rates, while
decreasing the negative effects of alkali promoted carbohydrate degradation (xvi).
1.3 Review of modified kraft pulping processes
The major areas within the kraft process that drive efforts for process improvement are the following:
Low pulp yield relative other pulping processes
High capital requirements necessary to build new pulp mills and to rebuild or maintain
current mill operations
Environmental concerns stemming from the use of sulfur and chlorinated compounds
during pulping and bleaching operations.
Process modifications that work within the pre-existing process unit operations and equipment can
significantly improve process economics or be devised to meet new environmental regulations without the need
for additional capital expenditure.
1.3.1
Modified continuous cooking
Modified kraft pulping processes have gained widespread acceptance, because they can be used either
to extend delignification or to enhance the yield and pulp properties at a given kappa number. The basic
principles of modified extended delignification consist of a level alkali concentration throughout the cook, a
high initial sulfide concentration, low concentrations of lignin and Na+ in the final stage of the cook, and lower
15
temperature in the initial and final stages of the cook (xxviii). This process approach has been developed by
Kamyr around their continuous digester. Using conventional kraft recovery operations the generated white
liquor is split into different feed streams applied during feed, transfer circulation and a countercurrent cooking
zone. The process modifications have resulted in the potential for higher yield or improved delignification with
improved viscosity and bleachability properties.
1.3.2
Green liquor pretreatment
One alternative to avoid the increase in causticization requirements would be to pre-treat wood with
green liquor. Previous work has demonstrated the feasibility of using green liquor in the impregnation stage,
without increasing overall chemical usage (xxix,xxx). It has also been shown that the amount of sulfur adsorbed
during the pretreatment decreases with higher [OH-] (xxxi). By impregnating chips with high sulfidity, low pH
liquor, a mill may enhance yield and further decrease the causticizing load. Figure 1.12 outlines the unit
operations for a possible green liquor pretreatment process in conjunction with BLG.
Comparing conventional kraft pulping with the green liquor pretreatment described above, the greatest
relative cost-benefit from a decrease in causticization using green liquor pretreatment would be achieved in a
situation where the level of TTA was the same in both processes. This requires that similar pulp kappa numbers
must be attainable through both processes at the same TTA charge.
Figure 1.12
1.3.3
Outline of process using green liquor pretreatment
Split sulfidity pulping
BLG would enable a mill to generate a high sulfidity liquor which can be used to provide a high
sulfide concentration during the initial phase of the cook. In split sulfidity pulping, it would be necessary to
generate two streams of white liquor one that is sulfide-rich and another that is sulfide-lean. Sulfur profiling
would be the lowest capital cost process to implement to modify the pulping process especially for mills with a
modified continuous or batch pulping process. Figure 1.13 shows the basic concept design for generating
liquors of different sulfide concentrations.
16
Figure 1.13
Schematic of unit operations in split sulfidity pulping
The concept of sulfur profiling, or split sulfidity pulping, employing a sulfur-rich stream in the rapid
initial phase, followed by a sulfur-lean stream in the bulk and residual phase, has been investigated as a method
for extending delignification or increasing yield (xxix,xxxii,xxxiii,xxxiv,xxxv). Compared to conventional kraft cooks
of similar H-factor, split sulfidity pulping has been shown to enhance selectivity of the pulping reactions,
resulting in increases in both lignin removal and pulp viscosity. Moreover, split sulfidity pulping has been
shown to increase pulp yield and strength properties (xxxvi,xxxvii,xxx).
The effects of multiple stage cooking using sulfur profiling, has also been studied. The process
showed a significant improvement in selectivity (xxxviii,xxxix). Increased sulfide sorption resulted in both higher
lignin-free yields and increased viscosities. At 30% overall sulfidity, the lignin-free yield was 0.6 to 0.9%
higher and viscosity 8.89 to 10.4 mPa higher than conventional kraft. At increasing overall sulfidities, the yield
advantage was reduced. Screened yield increased only slightly with higher sulfidity levels during impregnation.
Similar findings were reported in subsequent work (xl). Pulping work conducted at STFI found that sorption of
sulfide increases with increasing hydrosulfide concentration, time, temperature and concentration of positive
ions, but decreases with an increasing concentration of hydroxide ions (xli). The potential for modifying
softwood kraft pulping by sulfur profiling was also investigated. When all of the sulfide was added to the
beginning of the cook, a high hydrosulfide concentration could be maintained both in the initial phase and near
the transition point from the initial to the bulk delignification phase (xlii).
The work described above is difficult to implement in a mill that utilizes conventional recovery
technologies.
However, BLG generates separate streams of sulfur and sodium, which will allow for
independent sulfur and alkali profiling. Thus, the alkali profile can be adjusted independent of the sulfur
concentration at any point in the cook. These opportunities were investigated at NC State University, exploring
split sulfidity pulping of southern pine with different initial alkali concentrations.
Based on a modified
continuous cooking (MCC) laboratory procedure, different approaches were devised to explore split sulfidity
and different initial alkali profiles (xliii,xliv). Two levels of initial alkali were investigated where a fraction of the
17
available hydroxide was charged in the initial stage. The low initial alkali procedure used 11% of the alkali;
the corresponding value for the high initial alkali procedure was 33%.
1.3.4
Kraft polysulfide pulping with anthraquinone
The effect on pulping chemistry of polysulfide (PS), often in conjunction with anthraquinone (AQ) as
additives to the Kraft process, has been explored for some time (xlv,xlvi,xlvii,xlviii,xlix,l). Its effectiveness has been
established, and it is typically reported that each percent of PS added increases the pulp yield by one percent
(li,lii). However, efficiently generating high concentrations of PS within the Kraft chemical recovery cycle is
difficult.
There are currently three primary competing processes available for PS generation: Chiyoda,
MOXY and Paprilox (liii).
These processes, in general terms, produce pulping liquors with PS
concentrations of five to eight grams per liter and PS selectivities ranging from 60 to 90 percent (liv,lv,lvi,lvii). This
results in a PS limit of about 1% PS charge on oven dry wood for a mill operating at 25% sulfidity. However, a
chemical recovery system based around BLG would allow for different pathways to generate PS liquors which
would enable higher charges of polysulfide. In addition, the separation of sodium and sulfur would allow for
alkali profiling in conjunction with PS utilization. Figure 1.14 shows a schematic of PS process unit operations
with BLG.
Figure 1.14
Outline of process using polysulfide
Research efforts in the area of PS have generally been in one of two major areas; work on PS pulping
associated with PS utilization in Kraft process operations and/or associated PS generation technologies
(liv,lv,lvi,lvii,lviii,lix,lx,lxi) and work investigating optimum parameters for PS pulping ( lxii,lxiii,lxiv). A smaller area of
work has been based around the potential implementation of BLG and the opportunities created by the
unrestricted management of sulfur and sodium as separate entities. The splitting of sulfur and sodium enables
the application of polysulfide, sodium sulfide and sodium hydroxide independently of each other.
Two
processes have been described based on this concept. The ZAP process (Zero effective alkali in pretreatment),
entails a two-stage pulping procedure, where the sulfur-containing cooking chemicals (Na2S and PS) are
charged along with AQ to the wood in the pretreatment stage (lxv). In the subsequent cooking stage, NaOH is
added and the temperature increased. The obtained results using PS without AQ, indicate a potential yield
18
benefit of 1% relative to conventional PS pulping at kappa 30. With the addition of AQ, the yield benefit was
increased to 1.5-2%. Additional results indicate even greater yield benefits at kappa 90. The other process is
called hyperalkaline polysulfide pulping (lxvi). The process utilizes two pretreatment stages followed by a
cooking stage. In the first stage, alkali is charged to the wood at elevated concentrations, neutralizing the acids
formed during the temperature elevation. PS is then charged in the second stage, followed by the cooking stage.
The process resulted in a higher delignification rate, increased pulp viscosities and yield improvements of 1.5%
as compared to modified pulping without PS. Worth noting is that the bleachability and measured tear strength
of the hyperalkaline PS pulps were similar to those of the Kraft reference pulp.
In addition to the capability for independent profiling of NaOH and PS, the implementation of BLG
would allow for the conversion of all the sulfur in the cooking liquor to PS. This would enable higher PS
charges than available through conventional technologies. The total amount of sulfur that is available is
dependent on the sulfidity of the pulping liquor. Table 1 illustrates this balance displaying two examples of the
partitioning of the total sulfur available in the system at 25 and 40% sulfidity. As seen in the table, PS charges
slightly exceeding 2 % on wood are possible at 19.5% AA with 25% sulfidity. To enable higher PS charges the
sulfidity must be increased, and as shown in the table the corresponding value at 40% sulfidity is between 3 and
4% PS on wood.
Table 1.1
Demonstration of sulfur utilization as Na2S (kraft) or Na2S/PS (kraft-PS) and the system
sulfur availability for PS generation
25% Sulfidity
Tot. avail.
40% Sulfidity
S avail.
Tot. avail.
Sulfur (S)
for PS
as Na2S
(kg/ton)
Cook Procedure
S req.
S req.
S avail.
(kg/ton)
(kg/ton)
Sulfur (S)
for PS
as Na2S
(kg/ton)
(kg/ton)
(kg/ton)
MCC
25.2
0
25.2
40.3
0
40.3
1% PS
25.2
10.0
15.2
40.3
10.0
30.3
2% PS
25.2
20.0
5.2
40.3
20.0
20.26
3% PS
25.2
30.0
- 4.8
40.3
30.0
10.3
4% PS
25.2
40.0
- 14.8
40.3
40.0
0.3
2 STATEMENT OF OBJECTIVES
The focus of the work explored in this project was the effects that the implementation of black liquor
gasification might have on conventional pulping operations.
A two pronged approach was pursued: to
investigate modified pulping technologies that BLG implementation would enable, and to explore potential
process modifications to a kraft mill through process simulation. The first approach entailed laboratory pulping
experiments exploring different modified or advanced pulping technologies based on the separation of sodium
and sulfur in chemical recovery. The second entailed the generation of a BLG model, its integration into the
19
WinGEMS software package, and process simulation exploring the combined effects of BLG integrated with
the different pulping technologies outlined above on process economics.
2.1 Objectives for laboratory pulping work
Different opportunities arise in the application of pulping chemicals to wood during pretreatment and
digestion when sulfur and sodium are separated during chemical recovery operations.
The capability to
independently charge sulfur and sodium was explored through laboratory pulping using green liquor
pretreatment, split sulfidity pulping, and polysulfide pulping with and without anthraquinone. The objective for
the pulping work was to establish the potential benefits with regard to total pulp yield, delignification, pulp
viscosity and physical properties as compared to pulps generated using conventional kraft pulping methods.
2.2 Objectives for process simulation work
When considering the implementation of new technologies, it is of great benefit to use process
simulation as a tool to predict the potential effects of process modifications on overall process operation and
economics. To make possible WinGEMS process simulation of the integration of BLG and BLG enabled
modified pulping operations into kraft pulping, it was necessary to generate a BLG WinGEMS model. The
objective of the simulation work was to initially generate a BLG model. This model was then to be integrated
into the WinGEMS software package, and used to explore the effects of retrofitting BLG into kraft pulping on
process operation and economics.
3 SUMMARY, CONCLUSIONS AND FUTURE WORK
The results obtained throughout this project have been reported in several articles and conference
proceedings, manuscripts of which have been included in sections 8.1 8.5. The ensuing discussion is a review
of the obtained results outlining the efforts undertaken in laboratory pulping as well as the work performed to
generate a BLG model and its application in process simulation.
Laboratory procedures for pulping
experiments have been described in the manuscripts, and brief outlines have been included in the appendices.
3.1 Review of laboratory pulping results
The following sections discuss the results obtained through laboratory pulping experiments exploring
green liquor pretreatment, split sulfidity pulping and polysulfide pulping with anthraquinone.
3.1.1
Green liquor pretreatment
Green liquor pretreatment pulping has been investigated and presented (section 6.2). Pulping results
show that green liquor pretreatment would return pulps of higher kappa at the same level of total titratable
alkali, as shown in Figure 3.1. However, as shown in the figure it would be possible to achieve similar kappa
number at the same level of system TTA by pulping to a higher H factor. Another option for decreasing the
kappa number would be to increase the system TTA. Experiments performed with a 10% increase in TTA,
20
labeled Hi-TTA, also resulted in a higher kappa number than the MCC baseline. In this study, if the TTA is
increased by 20% the active alkali is the same in both processes, and the no causticizing benefits exist. The
resulting pulp yield did not show any improvement with green liquor pretreatment, but the green liquor
pretreated pulps had higher viscosity. These results did not show the yield benefit reported elsewhere, and may
be the result from differences in pulping procedures (xxx).
kappa
80
60
40
20
600
Figure 3.1
3.1.2
1100
1600
h factor
Green Liq PT (HIA)
Green Liq PT (Hi_TTA)
MCC Baseline
58
Total yield (%)
Green Liq PT (HIA)
Green Liq PT (Hi_TTA)
MCC Baseline
100
2100
54
50
46
42
20
40
60
kappa
80
100
Obtained results for pulping using green liquor pretreatment procedure
Split sulfidity pulping
The effects of split sulfidity and different levels of initial alkali on delignification and total pulp yield
are presented in Figure 3.2 (section 6.2). As shown, split sulfidity pulping produced lower kappa pulps at
similar H factors relative the MCC procedure. The high initial alkali cooks generated pulps of lower kappa
number compared to those of low initial alkali. The split sulfidity procedures produced pulp yields 1-2%
greater than the MCC procedure, and the difference is more pronounced at higher kappa. Since the high initial
alkali approach produced higher yields and lower kappa numbers than the low initial alkali approach, this would
be the preferred option. At similar kappa numbers the split sulfidity pulps had viscosities 5 to 10 cps greater
than those of the MCC pulps. The high initial alkali pulps produced higher tensile and burst index values
relative the MCC pulps at a similar tear index. The MCC pulps were slightly easier to refine relative to the split
sulfidity pulps.
21
MCC Baseline
kappa
60
40
20
600
Figure 3.2
54
Total Yield (%)
SS_HIA
SS_LIA
80
50
SS_HIA
SS_LIA
46
MCC Baseline
42
1100
1600
h factor
2100
20
40
60
kappa
80
Delignification and yield results for split sulfidity pulping
The co-absorption of H2S and CO2 during the scrubbing in sulfur recovery, results in the production of
NaHCO3. During recausticization all sodium exiting the gasifier will be converted to NaOH. The conversion of
NaHCO3 to NaOH requires twice as much lime compared to the conversion of NaCO3 to NaOH. Thus, there is
a two-fold increase in the amount of lime required to produce an equivalent amount of NaOH, and as a result,
BLG will increase the overall causticization load.
The potential for in-situ causticization within the gasifier could dramatically affect the load on the
recaust cycle and lime kiln. In current recovery operations, the sodium carbonate obtained from the slaking of
the boiler smelt is converted to sodium hydroxide using calcium oxide. The byproduct calcium carbonate is
then calcined in large rotating kilns to regenerate the calcium oxide. A 1000 ton per day pulp mill will use
about 100,000 barrels of fuel oil per year to fire its lime kiln. Through novel chemistries it may be possible to
carry out the causticization reactions directly within a black liquor gasifier. This could potentially eliminate the
need for the lime cycle and the associated fuel costs (xliv).
3.1.3
Polysulfide-anthraquinone pulping
The effects of pulping southern pine with higher PS charges and alkali profiling have been evaluated
(section 6.3 and 6.4). Table 3.1 shows a summary of the yield increases that were measured at various PS
charges and sulfidities. At 40% sulfidity the impact of alkali profiling in the initial stage was also evaluated.
Three different levels of alkali were investigated. In the low initial alkali (LIA) cook, 56% of the alkali was
charged to the impregnation stage. The corresponding values for the medium initial alkali (MIA) was 65% and
for the high initial alkali (HIA) 75%. The flexibility to optimize the alkali profile and PS use would only be
possible in combination with BLG (lxvii).
22
Table 3.1
Total yield improvement from PS procedures compared to 25% S MCC pulp
Cook ID
Estimated Yield at 30
kappa
Yield Improvement (% pts.)
Average Viscosity (cps)
MCC 25% S
45.2
n.a.
39.1
1% PS 25% S
47.7
2.4
38.6
2% PS 25% S
48.1
2.9
38.0
1% PS 40% S LIA
46.7
1.5
49.5
2% PS 40% S LIA
47.3
2.1
51.6
3% PS 40% S LIA
47.4
2.2
55.7
3% PS 40% S MIA
48.9
3.7
47.7
3% PS 40% S HIA
45.8
0.6
46.7
At 25% sulfidity the kappa numbers with PSAQ were comparable to the MCC reference. The yield
benefit was about 2% for a 1% PS charge, and about 3% for a 2% PS charge. At 40% sulfidity the kappa
number decreased with increasing PS charge, and increasing levels initial alkali. The level of initial alkali had a
significant effect on the yield. There is indication of an optimum condition for initial alkali charge, where too
great or too low of an initial hydroxide concentration negatively affects the pulp yield. The work exploring
MIA and HIA indicates that there exists a maximum in yield benefit as a function of initial alkali concentration.
Comparing these results to the ZAP process using PSAQ a yield benefit of about 6% at 30 kappa was reported,
where the initial alkali charge in the PS pretreatment stage was zero ( lxiv). Olm et al. also described a
minimum yield condition which was shown to exists at a hydroxide concentration of about 0.3 mol/l. At
hydroxide concentrations lower or greater than 0.3 mol/l, higher yields could be achieved. The results indicate
that there also may exists a maximum yield benefit at higher levels of initial alkali. The LIA condition in our
procedure corresponds to an initial hydroxide concentration of about 0.6 mol/l, which is greater than the
concentration reported for the yield minimum in the ZAP process. The maximum yield benefit was found
around an initial hydroxide concentration of 0.9 mol/l. The effect of initial alkali on pulp yield should be
further investigated to optimize the benefits of PS pulping.
3.2 Process simulation
Computer process simulation is a commonly used engineering tool with a wide range of applications in
the solution of problems. In the area of chemical and manufacture process engineering, it is frequently utilized
to evaluate the feasibility of new technologies or to determine operating parameters for process improvement
and optimization. WinGEMS is a software simulation tool designed specifically for applications in the pulp and
paper industry. The computer software is built on a modular design, where different unit operations specific to
pulp and paper processes can be linked to represent process segments or whole plants. The overall model is
then solved in a sequential order through multiple iterations until the specified convergence criteria are met.
23
The software was originally developed at the University of Idaho in the early 1970s and is now marketed and
supported by MetsoAutomation (lxviii,lxix).
Despite continuous evolution, WinGEMS does not have a block model for the BLG process. Initial
simulation work around BLG integration into the kraft process was completed using preexisting WinGEMS
blocks, achieving estimates of what effects could be expected in process variable operating costs. This work is
described in the manuscripts (section 5). In order to better simulate the integration of BLG into kraft pulping
operations and the effects of BLG enabled modified pulping technologies on process parameters, it was
necessary to develop such a model.
The following discussion briefly describes the development of the
WinGEMS BLG block, and then its application in full mill simulations of modified kraft pulping operations.
3.2.1
Development of BLG model for WinGEMS
The purpose of constructing the BLG model was the desire to incorporate a BLG block into a full mill
WinGEMS simulation, where its interaction with other mill unit operations could be evaluated in terms of
process variables of interest to the user. The model developed for this application bases the material and energy
balances around the steam reformer on empirical relationships rather than first principles. This approach allows
for the prediction of model output streams based on the given BL input stream and process parameters, as well
as the back-calculation of the required amounts of bed fluidizing steam and energy supplied through the pulsed
heaters to sustain the endothermic steam reforming reactions. An outline of the unit operation input and output
streams is shown in Figure 3.1.
BL
BED SOLIDS
STEAM
FUEL GAS
STEAM
REFORMER
PRODUCT GAS
FLUE GAS
XS AIR
Figure 3.1
Input and output stream structure for BLG model
To simplify the BLG model and its control a number of parameters were identified and utilized for
user input and reaction constraints internal to the model. These parameters were based on process data provided
by MTCI. As a result, the range of operating conditions that can be used with the model in process simulations
is limited by the original data. The development and validation of the WinGEMS BLG block model has been
described in the manuscripts (section 9).
24
3.2.2
Simulation of the effects of BLG integration on kraft mill operations
Case studies were performed in WinGEMS comparing full mill simulations of a kraft process using
modified continuous cooking to BLG processes with split sulfidity and polysulfide-anthraquinone pulping. A
detailed discussion of the findings has been included in the manuscripts (section 10). Input assumptions for the
different simulations were based on available literature, industry contacts and laboratory results generated
throughout this dissertation project. Process unit operations were comparatively analyzed among the different
cases. Noteworthy results include the increased load on the lime kiln cycle due largely to the much greater
generation of carbonate salts in BLG as compared to conventional recovery boiler smelt production. Combined
with this fact is the amount of carbonate created from carbon dioxide which is co-absorbed with H 2S during
syngas scrubbing with amine systems and green liquor. In addition, the BLG cases also generated less steam
from the processed black liquor than conventional recovery boiler operations. To make up the steam demand
difference additional hog fuel was combusted in a power boiler, resulting in increased costs. The unit operation
variables predicted through the simulation were used to calculate cost factors for variable process cost items.
The net variable operating costs for each case were determined from these cost factors. The obtained results
indicate that implementing BLG with the investigated modified pulping technologies would lead to increases in
the net variable operating cost per oven dry ton pulp produced ranging from 1.8 to 3.9% compared to the kraftMCC base case. It was also found that the cost of kiln fuel and the price of power sales to the grid drive the
relative cost performance of the evaluated cases. When comparing the BLG cases to the MCC reference, the
net variable operating cost break even point based on lime kiln fuel cost ranges from $47 to $38 per barrel of
kiln fuel depending on the BLG process, which is much lower than currently available prices.
When
performing similar comparisons of net variable operating costs based on variable power sales price, significant
cost savings could be realized in all BLG processes at prices for power sold to the grid above 5 /KWh.
3.3 Conclusions
The following section outlines the overall conclusions drawn from the project.
3.3.1
Conclusions from split sulfidity pulping
Compared to the kraft MCC reference cooks, the investigated split sulfidity pulping procedure returned
pulps with one to two percent greater yield at a higher rate of delignification. The SS pulps were found to have
greater viscosity and improved tensile and burst strength.
3.3.2
Conclusions from green liquor pretreatment pulping
Compared to the kraft MCC reference cooks, the investigated green liquor pretreatment pulping
procedure generated pulps of comparable yield and viscosity. The resulting pulp strength was also somewhat
improved. However, the rate of delignification was lower than in the MCC reference procedure, and this could
not be overcome by increasing the levels of charged TTA. The greatest benefit from GL pretreatment would be
25
realized through the possibility of decreasing the load on the lime kiln, creating an opportunity for debottlenecking or cost savings from a decrease in lime kiln fuel demand.
3.3.3
Conclusions from polysulfide pulping with anthraquinone
The investigated PSAQ procedure resulted in increased levels of delignification compared to the MCC
reference. This benefit increased with increasing levels of initial alkali charge. The obtained pulp yields
suggest possible yield improvements of up to 3%, with an approximate yield benefit of 1% total yield per % PS
charged on OD wood. This result is in line with literature data. The performed work has suggested that there
exists an optimum condition for PSAQ pulping with alkali profiling. The best results obtained in this work with
respect to delignification rate, yield benefit and even alkali profile was achieved using 3% PS on wood at 40%
sulfidity and an alkali profile where the total available alkali was added to the cook in three separate stages as
fractions 45/31/24 (Stage1/Stage2/Stage3) of the total.
3.3.4
Conclusions from WinGEMS simulation work
Initial simulation work around the integration of BLG into the kraft process combined with modified
pulping technologies showed that the net variable process operating cost would increase due to increased load
on the lime kiln. This cost could be offset through implementation of modified pulping technologies that
allowed for greater pulp yield, lower the overall process demand for wood, chemicals and fuel. The work also
indicated the desirability of having a BLG WinGEMS module available for further simulation.
Further
simulation work has resulted in the generation of such a WinGEMS module for the MTCI low temperature
steam reformer, or black liquor gasifier. The module has been utilized in simulation case studies exploring SS
and PSAQ pulping using BLG as the central unit operation in chemical recovery. The obtained results indicated
that the net process variable operating cost was driven by two major factors, lime kiln fuel and power sales
price. Under the initial assumptions, where lime kiln fuel was set to $50/barrel and power sales price at
$0.35/kWhr, the net variable operating costs for the BLG cases showed a cost increase of about 3% compared to
the MCC reference case. The overall conclusion from this simulation case study is that BLG integration into
the kraft process is highly dependent on the price of lime kiln fuel and the price and amount of power produced
from BLGCC conversion of syngas. Under the initial assumptions made in this case study, the significantly
increased load on the lime kiln is not overcome by the benefits realized in pulping operations through the
introduction of modified pulping technologies, nor by the additional revenue generated from the generation and
sale of green power. However, if modifications could be made to the recausticizing unit operations, such as
high sulfidity green liquor pretreatment, offloading the slaker and resulting load on the lime cycle, this would
change. Another such approach would be auto-causticization in the white liquor stream. Also, as indicated in
figures 10.3 and 10.4, if the lime kiln fuel cost is decreased or the power sale price increased, the BLGCC kraft
process with split sulfidity or PSAQ pulping becomes a more economically favorable alternative than the
conventional reference kraft process, based on variable operating costs. When comparing the BLG cases to the
MCC reference, the net variable operating cost break even point based on lime kiln fuel cost is about $47/barrel
26
for the SS and PS2% processes, and about $38/barrel for the PS2% process. This is significantly lower than
assumed kiln fuel price of $60/barrel used in this work. If the sales price for power to the grid was increased
from 3.5 to 6 /KWh cost savings of about $40/ODtP could be realized in all BLG processes.
3.4 Future work
Future work in the area of pulping integrated with BLG should be focused on technologies or
processes with BLG that allow for decreased load on the lime kiln, such as auto-causticizing or alkaline-sulfite
pulping. Of these, mini-sulfide sulfite pulping with anthraquinone, MSSAQ, may be the most intriguing, as it
would likely permit the elimination of the lime kiln cycle entirely. Pulping work to investigate and optimize
MSSAQ pulping and simulation work showing the overall process effects of BLG implementation integrated
with MSSAQ pulping should be performed.
Future work in the area of simulation should be focused on the processes and process modifications
suggested for future pulping work. Of special interest would be MSS-AQ simulation based on a BLG recovery
system. In addition, a continuous effort for improvement of the BLG block module as more data become
available, to increase the robustness and predictive capability of the model with regards to varying inputs and
operational parameters, is desirable. Finally, the development of a validated combined cycle power plant within
the WinGEMS software package would be most valuable.
27
4 REFERENCES
28
5
ECONOMICS OF INTEGRATING BLACK LIQUOR GASIFICATION WITH PULPING: PART I
EFFECT OF SULFUR PROFILING
Lindstrom, M.E., Kirkman, A., Jameel, H. et al. Economics of Integrating Black Liquor
Gasification With Pulping: Part I Effect Of Sulfur Profiling. Proceedings Tappi Fall Technical
Conference, Tappi Press, Atlanta, GA, (2002)
29
ECONOMICS OF INTEGRATING BLACK LIQUOR GASIFICATION WITH PULPING:
PART I EFFECT OF SULFUR PROFILING
Mathias Lindstrom, Adrianna Kirkman, Hasan Jameel
Julie Cheng, Christy Huggins, Brandon Bray
Department of Wood and Paper Science
North Carolina State University
Raleigh, NC 27695-8005
USA
ABSTRACT
Using black liquor gasification (BLG) the recovered sodium and sulfur can be split into two separate fractions
with varying degrees of separation dependent on the operating conditions and the technology used. This
separation creates some opportunities in the pulping process, due to its potential to increase the yield or to
extend delignification. There may be an increase in the causticizing load. Of the various pulping options with
black liquor gasification, sulfur profiling would be the lowest capital cost process to implement especially for
mills with a modified continuous or batch process. In this paper, the process changes that are necessary to
implement sulfur profiling are examined along with the economics. The results show that BLG increases the
pulp production cost from $140.23/ODtP to $143.31/ODtP at H 2S/CO2 co-absorption ratio of 10:1. Most of the
cost increase is due to the increase in the lime kiln fuel and the lime makeup. An increase in yield of 1%
changes the pulp production cost from $140.23/ODtP to $140.45/ODtP for the Tomlinson vs. BLG with split
sulfidity pulping. If the yield increase is 2%, the cost decreases to $138.91. It may also possible to use the
NaHS and Na2CO3 stream in the impregnation phase without affecting the overall alkali usage, since sulfur
absorption is more effective at the lower pH. At the same EA charge of 19%, the cost increased from $140.23/
ODtP for the Tomlinson case versus $141.15/OdtP for the BLG case with Na 2CO3 pretreatment. But at the
same TTA usage the %EA is 15% and the cost is decreased to $136.09/ODtP.
INTRODUCTION
In the conventional kraft recovery cycle, the sodium and the sulfur ratio is fixed by the sulfidity of the liquor.
Using black liquor gasification the sodium and sulfur can be split into two separate fractions. This separation
creates some opportunities in the pulping process, which can lead to production cost savings or improved
operations.
The implementation of black liquor gasification can increase the efficiency of combined heat and power
generation in the pulp mill. It may be possible to produce twice as much electricity per ton of black liquor solid
as compared to a modern Tomlinson recovery boiler.
The Tomlinson Recovery Boiler has the disadvantages of low thermal efficiency, low power to heat ratio and
the risk of smelt/water explosions. Gasification technology addresses some of these concerns. Black liquor
gasification technologies can be classified by the operating temperature (1). High temperature gasifiers operate
at about 1000oC (Chemrec) and low temperature gasifiers operate at less than 700 oC (MTCI). In the high
temperature gasifier the smelt leaves in the molten form and in the low temperature system the inorganics leave
as solids.
A schematic of the Chemrec black liquor gasification system is shown in Figure 1. The fuel gas stream contains
most of the sulfur as hydrogen sulfide and carbonyl sulfide. A solvent-based regenerative system is used to
clean up the raw fuel gas and form a sulfur rich solvent stream. The solvent is then regenerated to liberate the
sulfur rich gas stream.
A simplified schematic of the MTCI system is shown in Figure 2. It consists of a fluidized bed reactor that is
indirectly heated by multiple resonance tubes of one or more pulse combustion modules (2). Feedstock such as
spent liquor is fed to the reactor, which is fluidized with superheated steam from a waste heat recovery boiler.
30
Gasifier
Gas
Turbine
Fume
Collection
Black
Liquor
HRSG
Clean Fuel
Gas
Steam
Oxygen
Smelt
Quencher
Sulfur
Chemicals
Raw
Fuel
Gas
Sulfur
Removal
Sulfur
Conv.
Slaker
Weak
Wash
Sulfide
Lean
White Liquor
Sulfide
Lean
Green Liquor
Figure 5.1
Schematic of a High Pressure Black Liquor Gasification System
The organic material injected into the bed undergoes a rapid sequence of vaporization and pyrolysis reactions.
Higher hydrocarbons released among the pyrolysis products are steam cracked and partially reformed to
produce low molecular weight species. Residual char retained in the bed is more slowly gasified by reaction
with steam. The sulfur and sodium are separated in that the sulfur leaves mostly with the gas stream and the
sodium stays in solid form. Product gases are routed through a cyclone to remove the bulk of the entrained
particulate matter and then quenched and scrubbed in a venturi scrubber. A portion of the medium-Btu product
gases is supplied to the pulse combustion modules and combustion of these gases provides the heat necessary
for the indirect gasification process. The inorganic chemical in the feedstock is recovered and recombined with
sulfur from the gas stream to recycle the product to the mill.
The products of combustion exit from the resonance tubes completely segregated from the reformate product
gases. Hot flue gases from the steam reformer are used to generate steam and to preheat the pulsed heater
combustion air. Excess fuel gas is exported for use in a boiler, gas turbine or fuel cell. The process uses only a
single reactor; it does not require solids recirculation and handling equipment and it can be easily controlled by
varying the gas-firing rate.
31
Product
Gas
Clean
Flue Gas
Feedstock
Product
Gas/Air
Pulsed
Heaters
Bed Solids
Figure 5.2
Fluidizing
Steam
Schematic of a MTCI Steam Reformer
INTEGRATION OF BLG INTO PULPING
One of the advantages of the black liquor gasification over the Tomlinson boiler is that the sodium and the
sulfur are split into separate streams which can then be utilized in the pulping process as desired. The splitting
of the sodium and the sulfur can be taken advantages of with the following pulping technologies:
Split Sulfidity Pulping
Polysulfide Pulping
Alkaline Sulfite Pulping : Alkaline Sulfite- AQ (AS-AQ) or Mini-Sulfite Sulfide AQ Process
(MSSAQ)
In split sulfidity pulping, it would be necessary to generate two streams of white liquor one that is sulfide rich
and another that is sulfide lean. In polysulfide pulping, a portion of the sulfur stream would be oxidized to
elemental sulfur, which would be used mixed with the sulfur lean white liquor to produce polysulfide liquor. In
the ASAQ or MSSAQ process, the sulfur gas stream would be burnt to produce SO 2, which would then be
absorbed into the Na2CO3 stream from the gasifier. The decision between AS-AQ and MSS-AQ, will depend
on the amount of Na2S in the liquor from the gasifier.
Sulfur profiling would be the lowest capital cost process to implement especially for mills with a modified
continuous or batch pulping process. It may also be possible to get some of the benefits of split sulfidity
pulping in conventional indirect-heated batch digesters. In this paper, the process changes necessary to
implement sulfur profiling will be examined along with the economics. Previously, a cost benefit analysis for
polysulfide pulping has been done for low polysulfide charges (3).
Black liquor gasification will potentially increase the causticizing load due to the release of H2S. In
conventional recovery the sulfur ends up as Na2S in the green liquor stream, while in gasification for each mole
of H2S the sodium now becomes available as Na2CO3. A higher amount of Na2CO3 will have to be converted to
NaOH. In addition, the concurrent capture of CO2 in a downstream H2S scrubbing process will also increase the
causticizing load.
32
CASE 1: Production of Conventional White Liquor
The first case studied was for the production of conventional white liquor as shown in Figure 3. All the sulfur
chemicals along with some CO2 is absorbed in the sulfide lean green liquor (Na2CO3) and then combined before
the slaker.
Na2CO3 + H2S = NaHCO3 + NaHS
Na2CO3 + CO2 + H2O = 2NaHCO3
In this case the white liquor produced will be similar to that from a conventional boiler and split sulfidity
pulping cannot be implemented. The absorption of H2S and CO2 in Na2CO3 results in the production of
NaHCO3, which will result in an increase in the causticization load. The equations below show that it takes
twice as much lime to produce an equivalent amount of NaOH from NaHCO3.
2NaHCO3 + 2Ca(OH)2 = 2NaOH + 2CaCO3 +2H2O
Na2CO3 + Ca(OH)2 = 2NaOH + CaCO3
The economic impact of this process was evaluated by simulating the system using WinGEMS. The base case
consisted of a Tomlinson boiler with a continuous digester. The gasifier was simulated using a combustor block
with appropriate reaction blocks to represent the absorption steps and the handling of sodium. From the blocks
representing the combustor, the following streams were predicted:
Gas stream containing Carbon Dioxide, Hydrogen Sulfide, Oxygen, Nitrogen and Water Vapor
Chemical stream containing Na2CO3
Loss streams containing sodium and sulfur
The Na2CO3 solution was formed using weak wash from the lime cycle. The H2S absorption with the Na2CO3
liquor and the subsequent causticizing of the carbonate and bicarbonate components was performed with a
series of Reaction and Slaking/Causticizing blocks.
These reactions were based on the stoichiometry
represented by the equations above.
Ga
sif
ier
The results for the base case and for the production of conventional white liquor are shown in Tables 1 and 2.
The effect of a change in the amount of CO 2 that is co-absorbed along with the H2S is also shown. The
expected the mole ratio of H2S to CO2 that will be co-absorbed is 10:1. The economics of the power cycle was
not accounted for in this simulation. The energy generated from the combustion of black liquor in the black
liquor gasifier case (BLG) was significantly higher than that for the Tomlinson boiler because a combuster was
used to simulate the gasifier. For our cost calculations, it was assumed that the BTU from the Tomlinson and
the BLG are the same (this assumption will have to be need to be adjusted in a further study). Any changes in
the amount of energy available was valued as an incremental change in BTU from the Tomlinson case.
Clean Fuel
Gas
Black
Liquor
Weak
Wash
Sulfur
Conv.
Raw
Fuel
Gas
Smelt
Sulfur
Chemicals
Sulfur
Removal
Sulfide
Lean
Green Liquor
33
Slaker
White Liquor
Figure 5.3
Schematic for the Production of Conventional White Liquor with BLG
The results show that BLG increases the pulp production cost from $140.23/ODtP to $143.31/ODtP at a coabsorption ratio of 10:1. Most of the cost increase is due to the increase in the lime kiln fuel and the lime
makeup. The impact of CO2 co-absorption was quite low. The increase in causticizing cost and lime kiln
production bottlenecks demonstrates the need to evaluate other technologies such as auto-causticizing.
CASE 2: Production of Split Sulfidity Liquor
Black
Liquor
Ga
sif
ier
Clean Fuel
Gas
Sulfur
Conv.
Raw
Fuel
Gas
Sulfur
Chemicals
Sulfur
Removal
Smelt
Weak
Wash
Sulfide Lean
White Liquor
Sulfide Lean
Green Liquor
Figure 5.4
Slaker
Slaker
Schematic for the Production of Sulfide Lean and Sulfide Rich White Liquor with BLG
34
The schematic for the production of split sulfidity liquor is shown in Figure 4. In this scheme a second slaker
was necessary to keep the sulfide land and rich liquors separate. It should be noted that the sulfur gases were
absorbed into Na2CO3 instead of NaOH to minimize the co-absorption of CO2 in the liquor. The production of
two different white liquor streams will enable one to take advantage of split sulfidity pulping.
The application of modified delignification in the kraft process has been investigated with respect to selectivity,
pulp yield and strength properties. The basic principles of modified extended delignification consist of level
alkali concentration throughout the cook, a high initial sulfide concentration, low concentrations of lignin and
Na+ in the final stage of the cook, and lower temperature in the initial and final stages of the cook (4). The
concept of sulfur profiling, or split sulfidity pulping, employing a sulfur rich stream in the rapid initial phase,
followed by a sulfur lean stream in the bulk and residual phase, has been investigated (5-9). Compared to
conventional kraft cooks of similar H-factor, split sulfidity pulping to has been shown to enhance selectivity of
the pulping reactions resulting in higher lignin removal and pulp viscosity. Split sulfidity pulping, has been
shown to increase pulp yield and strength properties (9-16). The basic approaches consist of an initial
presteaming or pretreatment stage, followed by one or two cooking stages. A sulfur rich stream was charged in
the pretreatment or first cooking stage followed by a sulfur lean stream in the second stage. Compared to
conventional reference cooks the results indicate a yield increase of about 1-2 % and higher viscosity pulp for
cooks to kappa less than 20.
In the modified batch systems commercially available the requirement for initial high sulfide concentration is
accomplished by using the spent cooking liquor containing high sulfide for pre-treatment (17). Pre-treatment of
the chips using black liquor before digesting with the white liquor allows a significant reduction of H-factor or
alkali charge to reach the same kappa level as conventional batch system (18). The implementation of the
modified batch system into an existing batch mill is very capital intensive. It may be possible to get some of the
benefits of a modified batch pulping system by using the liquors with different sulfidities during the heat up
phase of a indirect-heated batch digester, thereby decreasing the capital required to attain some of the benefits
of extended delignification. In the modified continuous cooking system, the white liquor is added in multiple
steps to level out the alkali concentration and there are multiple extractions to keep the dissolved lignin low
towards the end of the cook. However, there is no good method to take advantage of the high sulfidity that is
beneficial during the initial phase of the cook. Black liquor gasification with its ability to split the sodium and
the sulfur will enable us to charge a higher amount of sulfur in the initial phase of the cook.
The benefit of split sulfidity pulping can be taken advantage of either be taken in a yield increase or by lowering
of the kappa number to the bleach plant at the same yield. The results of a yield increase on the overall pulping
and causticizing economics are shown in Tables 3 and 4 for three different yield increases. An increase in yield
of 1% changes the pulp production cost from $140.23/OtTP to $140.45/ODtP for the Tomlinson vs. BLG with
split sulfidity pulping. At this yield increase, the cost for Tomlinson and the BLG are comparable. If the yield
increase is 2%, the cost decreases to $138.91ODtP.
CASE 3: Use of Na2CO3 in the Pretreatment Stage
Previous work (9-10) has demonstrated the feasibility of using green liquor in the impregnation stage, without
increasing overall chemical usage. It has also been shown (19) that the amount of sulfur adsorbed during the
pretreatment decreases with higher [OH-]. By impregnating chips with high sulfidity, low pH liquor, a mill may
enhance yield and further decrease the causticizing load. Since the black liquor process increases the
causticizing load, it would be of interest to evaluate the option where the pretreatment is done using a mixture
of NaOH, Na2CO3 and NaHS. This option is shown in Figure 5. The need for another slaking operation is not
necessary in this process.
The results are shown in Tables 5 and 6 for different cases of effective alkali charge. It is expected that the EA
requirement will decrease because of the pretreatment with the sulfide rich liquor. Assuming a yield increase of
1.5% at the same EA charge of 19%, the use of Na2CO3 and NaHS in the pretreatment stage changes the cost
from $140.23/ODtP for the Tomlinson case versus $141.15/OdtP for the BLG case with Na2CO3 pretreatment.
It has been shown in previous work (9) that pretreating with this liquor will decrease the %EA charge. It has
been speculated that the total amount of TTA used may not change significantly.
35
Ga
sif
ier
Black
Liquor
Clean Fuel
Gas
Sulfur
Conv.
Raw
Fuel
Gas
Smelt
Sulfur
Chemicals
Sulfur
Removal
Weak
Wash
Sulfide Lean
Green Liquor
Figure 5.5
Slaker
Sulfide Lean
White Liquor
Schematic for using NaHS and Na2CO3 in Pretreatment with BLG
But at the same TTA usage the %EA is 15% and the cost is decreased to $136.09/ODtP. It is expected that we
may be able to operate at the same TTA usage.
CONCLUSIONS
Using black liquor gasification (BLG) the sodium and the sulfur can be split into two separate fractions with
varying degrees of separation dependent on the operating conditions and the technology used. The low
temperature process (MTCI Steam Reformer) is capable of splitting the sulfur and the sodium completely. This
separation of the sodium and the sulfur creates some opportunities in the pulping process, which can lead to
production cost savings or lower environmental impact. As a result, integration of the gasification process into
the overall mill will affect the economics of gasification. The splitting of the sulfur and caustic has the potential
to increase the pulping yield, but the negative aspects of the increase in the causticizing load also needs to be
taken into account. The splitting of the sodium and the sulfur can be taken advantages of with the following
technologies:
Sulfur Profiling
Polysulfide Pulping
Mini-Sulfite Sulfide AQ Process (MSSAQ)
Sulfur profiling would be the lowest capital cost process to implement especially for mill with a modified
continuous or batch process. The use of liquors with varying sulfidity will be of greatest interest to bleached
mills that have a continuous digester with capability for MCC. After black liquor gasification it is possible to
generate streams of NaHS and Na2CO3 and a stream of NaOH. The relative amounts of these streams will
depend on the sulfidity of the liquor and the causticizing load. High sulfidity in the initial phase is very
important to improve the yield and that a shortage of sulfide ions in the transition phase of delignification can
lower the yield by approximately 1.5%. Black liquor gasification enables us to charge a higher amount of sulfur
in the initial phase of the cook. This increased the yield by about 1-2%, or it decreased the kappa number by 4
units without a change in the yield. It is also possible to use the NaHS and Na 2CO3 stream in the impregnation
phase without affecting the overall alkali usage, since sulfur absorption is more effective at the lower pH. The
use of the NaHS and Na2CO3 stream for impregnation minimizes the impact of gasification on causticizing.
The above processes were simulated using WinGEMS to quantify the impacts on the various mill unit
operations. The results show that BLG increases the pulp production cost from $140.23/ODtP to $143.31/ODtP
at H2S/CO2 co-absorption ratio of 10:1. Most of the cost increase is due to the increase in the lime kiln fuel and
the lime makeup. The co-absorption ratio did not have a big impact on the cost. An increase in yield of 1%
changes the pulp production cost from $140.23/ODtP to $140.45/ODtP for the Tomlinson vs. BLG with split
sulfidity pulping. If the yield increase is 2%, the cost decreases to $138.91. It may also possible to use the
NaHS and Na2CO3 stream before slaking in the impregnation phase without affecting the overall alkali usage,
36
since sulfur absorption is more effective at the lower pH. At the same EA charge of 19%, the cost increased
from $140.23/ODtP for the Tomlinson case versus $141.15/ODtP for the BLG case with Na 2CO3 pretreatment.
But at the same TTA usage the %EA is 15% and the cost is decreased to $136.09/ODtP.
In Part II and III, of this work the integration of black liquor gasification into a mill with polysulfide use
especially at high charges and sulfite pulping will be discussed.
REFERENCES
1. Stigsson, L., Proceedings 1998 International Chemical Recovery Conference, p663
(1998)
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
Rockvam, L.N., Thermochem Recovery International, Personal communications (2001)
Larson, E.D, Yang, W., Iisa, I., Malcolm, E., McDonald, G., Fredrick, J., Kreutz, T., Brown, C.,
International Chemical Recovery Conference Proceedings, p 1.(1998)
Johansson, B., Mjoberg, J., Sandstrom, P., Teder, A., Svensk Papperstid. 87(10):30 (1984).
Herschmiller, D.W., Breaking the yield barrier, p. 59 TAPPI PRESS, 1998
LeMon, S. and Teder, A., Svensk Papperstidning, 79(11):407 (1973)
Olm, L. and Tisdat, G., Svensk Papperstidning, 82(15):458 (1979)
Olm, L. and Teder, A., Paperi ja Puu, 63(4a):315 (1981)
Andrews, E.K., dissertation, North Carolina State University, 1982
Andrews, E.K., Chang, H-m., Kirkman, A.G., Eckert, R.C., Extending delignification in kraft and
kraft/oxygen pulping of softwood by treatment with sodium sulfur liquors, Japan Tappi Symposium on
Wood pulp chemistry, 1982.
Jiang, J.E., Crofut, K.R., Jones, D.B., Proceedings of the TAPPI pulping conference, p. 799, TAPPI
Press, Atlanta, GA (1993)
Lownertz, P.P.H. and Herschmiller, D.W., Proceedings of the TAPPI pulping conference, p. 1217,
TAPPI Press, Atlanta, GA (1994)
Jiang, J.E. and Herschmiller, D.W., Proceedings of the TAPPI pulping conference, p. 317, TAPPI
Press, Atlanta, GA (1996)
Mao, B. and Hartler, N., Paperi ja Puu, 74(6):491 (1992)
Mao, B. and Hartler, N., Nordic Pulp and Paper Research J., 7(4):168 (1992)
Mao, B. and Hartler, N., Paperi ja Puu, 77(6-7):419 (1995)
Tormund, D., Teder, A., A new finding on sulfide chemistry in kraft pulping, Tappi pulping
conference p 247 (1989).
Abuhasan, J., PhD thesis North Carolina State University, p 82 (1994).
I. Lopez, H-m Chang, H. Jameel and W. Wizani, Tappi Pulping Conference Proceedings, p135
Orlando, Florida October 1999
AUTHORS
Hasan Jameel and Adrianna Kirkman are Professors in the Department of Wood and Paper Science
Mathias Lindstrom is a graduate student working on integrating black liquor gasification and pulping
Julie Cheng, Christy Huggins and Brandon Bray are undergraduate students who worked on modifying
WinGEMS to simulate the gasifier and slaking reactions as part of a senior project.
37
Table 5.1
Effect of Black Liquor Gasification and the H2S to CO2 Co-absorption on the Process
Fiber Line
Unit
ODtC/day
ODtP/day
kg/ODtP
Kraft Base
2409.24
993.77
1720.29
BLG (5:1)
2409.24
993.77
1739.69
%
41.25
41.25
41.25
41.25
mt/hr
g/L
%
178.51
151.28
24.97
186.16
151.27
24.36
184.00
151.26
24.66
182.92
151.25
24.81
Recovery
Evaporator
Flow to evaporator
Solids to evaporator
Dissolved solids to evaporator
Steam used (evaporated)
Steam used (concentrated)
mt/hr
%
kg/ODtP
kg/ODtP
kg/ODtP
476.98
15.99
1360.49
1581.26
553.60
483.85
15.81
1847.19
1613.59
555.10
481.91
15.85
1844.98
1604.82
554.43
480.93
15.88
1841.25
1600.41
554.11
Boiler / Gasifier
Flow to unit
Solids fired
Dissolved solids to unit
Amount Inorganic
Amount Organic
Energy available for Steam
Gas to Turbines, mt/hr
mt/hr
%
kg/ODtP
kg/ODtP
kg/ODtP
Mcal/day
mt/hr
118.20
66.26
1314.80
4.88
1314.80
4.71*106
x
100.28
60.12
1456.15
1.40
1316.17
5.66*106
541.84
100.17
60.13
1844.98
1.39
1315.51
5.65*106
541.58
100.12
60.13
1841.25
1.39
1315.20
5.65*106
541.45
Slaker
Flow
Reburned lime
Fresh lime
Make-up lime
mt/hr
kg/ODtP
kg/ODtP
mt/day
231.38
322.84
8.70
8.65
265.18
473.65
13.66
12.49
261.13
462.18
13.32
12.18
259.10
456.45
13.15
12.02
kg/ODtP
kg/ODtP
kg/ODtP
22.05
8.64
0.00
18.94
19.82
0.00
18.93
19.72
0.00
18.91
19.64
0.00
ODt/day
571.59
847.87
827.16
816.82
Chips
Brown Pulp Produced
Digester Steam
Yield
Brown Pulp
White Liquor (synthetic)
Flow
TTA as NaOH
Sulfidity
White Liquor make-up
NaOH (make-up)
Na2SO4
Elemental Sulfur
Table 5.1 continued
Kiln
Throughput
38
BLG (10:1) BLG (20:1)
2409.24
2409.24
993.77
993.77
1734.35
1731.68
Fuel consumed
kg/day
56542
72668
70874
69978
mt/hr
mt/hr
mt/hr
mt/hr
mt/hr
143.41
117.24
22.55
283.69
566.90
141.73
110.82
22.55
283.69
558.79
139.56
112.56
22.55
283.69
558.36
138.47
113.44
22.55
283.69
558.15
kg/ODtP
55.50
83.68
83.93
84.04
Water Usage
Scrubber Mill Water
Brownstock Washer Shwr Make-up
Mud Filter Mill Water
Mud Tank Cons. Reg. Mill Water
Total
Green Liqour Heater Steam
Table 5.2
Effect of Black Liquor Gasification and the H2S to CO2 Co-absorption on Cost
Cost analysis $/ODtP
Chips/kg
NaOH/lb
Na2SO4/lb
Fresh Lime/lb
Steam used/kg
Kiln fuel/barrel
Water/1000 gal
Total
Energy available/ODTP
Net Cost
Cost,
USD
0.062
0.141
0.05
0.032
0.006
20
0.072
x
6
x
Kraft
%
%
%
Basecase BLG(5:1) DELTA BLG(10:1) DELTA BLG(20:1) DELTA
150.31 150.31
0.00
150.31
0.00
150.31
0.00
6.84
5.87
-16.42
5.87
-16.48
5.87
-16.56
0.95
2.18
56.42
2.17
56.20
2.16
56.03
0.61
0.96
36.30
0.94
34.68
0.93
33.83
23.46
23.95
2.04
23.87
1.68
23.82
1.50
8.35
10.73
22.19
10.47
20.22
10.34
19.20
0.99
0.97
-1.45
0.97
-1.53
0.97
-1.57
191.51
51.28
140.23
194.98
61.63
133.35
39
1.78
16.79
-5.15
194.59
61.56
133.03
1.58
16.69
-5.41
194.39
61.53
132.86
1.48
16.65
-5.54
Table 5.3
Effect of Split Sulfidity and Yield Increase on the Process
Fiber Line
Unit
ODtC/day
ODtP/day
kg/ODtP
Kraft
Base
2409.24
993.77
1720.29
BLG
1% Yield
2353.18
993.77
1659.61
BLG
1.5% Yield
2331.67
993.77
1653.06
BLG
2%Yield
2306.38
993.77
1645.24
%
41.25
42.23
42.62
43.09
mt/hr
g/L
%
178.51
151.28
24.97
179.87
151.26
24.66
178.21
151.27
24.66
176.27
151.25
24.66
Recovery
Evaporator
Flow to evaporator
Solids to evaporator
Dissolved solids to evaporator
Steam used (evaporated)
Steam used (concentrated)
mt/hr
%
kg/ODtP
kg/ODtP
kg/ODtP
476.98
15.99
1360.49
1581.26
553.60
473.69
15.49
1772.00
1595.87
532.52
470.48
15.36
1745.70
1591.37
524.62
466.70
15.21
1731.69
1586.20
515.27
Boiler / Gasifier
Flow to unit
Solids fired
Dissolved solids to unit
Amount Inorganic
Amount Organic
Energy available for Steam
Gas to Turbines, mt/hr
mt/hr
%
kg/ODtP
kg/ODtP
kg/ODtP
Mcal/day
mt/hr
118.20
66.26
1314.80
4.88
1314.80
4.71*106
x
95.99
60.04
1391.77
1.42
1255.35
5.42*106
519.08
94.49
60.00
1745.70
1.43
1234.21
5.33*106
510.99
92.72
59.96
1731.69
1.44
1209.23
5.23*106
501.43
Slaker
Flow
Reburned lime
Fresh lime
Make-up lime
mt/hr
kg/ODtP
kg/ODtP
mt/day
231.38
322.84
8.70
8.65
200.03
451.52
13.01
11.90
198.15
447.23
12.90
11.79
195.94
442.17
12.74
11.65
kg/ODtP
kg/ODtP
kg/ODtP
22.05
8.64
0.00
18.70
19.44
0.00
18.40
19.40
0.00
18.11
19.40
0.00
Chips
Brown Pulp Produced
Digester Steam
Yield
Brown Pulp
White Liquor (synthetic)
Flow
TTA as NaOH
Sulfidity
White Liquor make-up
NaOH (make-up)
Na2SO4
Elemental Sulfur
Table 5.3 continued
Kiln
40
Throughput
Fuel consumed
ODt/day
kg/day
571.59
56542
807.94
69209
800.28
68555
791.24
67780
mt/hr
mt/hr
mt/hr
mt/hr
mt/hr
143.41
117.24
22.55
283.69
566.90
135.92
114.34
22.55
283.69
556.50
134.47
115.23
22.55
283.69
555.95
132.78
116.26
22.55
283.69
555.28
kg/ODtP
55.50
67.78
67.15
66.43
Water Usage
Scrubber Mill Water
Brownstock Washer Shwr Make-up
Mud Filter Mill Water
Mud Tank Cons. Reg. Mill Water
Total
Green Liqour Heater Steam
Table 5.4
Effect of Split Sulfidity and Yield Increase on Cost
Cost analysis $/ODtP
Chips/kg
NaOH/lb
Na2SO4/lb
Fresh Lime/lb
Steam used/kg
Kiln fuel/barrel
Water/1000 gal
Total
Energy available/ODTP
Net Cost
Cost,
USD
0.062
0.141
0.05
0.032
0.006
20
0.072
x
6
x
Kraft
BLG
%
BLG
%
BLG
%
Basecase 1%Yield DELTA 1.5%Yield DELTA 2%Yield DELTA
150.31
146.81
-2.38
145.47
-3.33
143.89
-4.46
6.84
5.80
-17.92
5.71
-19.84
5.62
-21.75
0.95
2.14
55.56
2.13
55.47
2.13
55.48
0.61
0.92
33.15
0.91
32.56
0.90
31.74
23.46
23.13
-1.42
23.02
-1.94
22.88
-2.56
8.35
10.22
18.30
10.13
17.52
10.01
16.58
0.99
0.97
-1.87
0.97
-1.97
0.97
-2.09
191.51
51.28
140.23
189.99
48.71
141.28
41
-0.80
-5.28
0.74
188.33
47.79
140.54
-1.69
-7.32
0.22
186.40
46.70
139.70
-2.74
-9.83
-0.38
Table 5.5
Effect of using Na2CO3 for Pretreatment and %EA Use on the Process
Fiber Line
Unit
ODtC/day
ODtP/day
kg/ODtP
Kraft Base
2409.24
993.77
1720.29
BLG
15%EA
2331.77
993.77
1697.14
13.00
BLG
17%EA
2331.78
993.77
1753.06
15.00
BLG
19%EA
2331.76
993.77
1809.05
19.00
%
41.25
42.62
42.62
42.62
mt/hr
g/L
%
178.51
151.28
24.97
175.82
151.35
24.66
199.30
151.34
24.66
222.82
151.31
24.66
Recovery
Evaporator
Flow to evaporator
Solids to evaporator
Dissolved solids to evaporator
Steam used (evaporated)
Steam used (concentrated)
mt/hr
%
kg/ODtP
kg/ODtP
kg/ODtP
476.98
15.99
1360.49
1581.26
553.60
468.51
15.87
1796.09
1558.43
539.99
489.80
15.82
1871.76
1631.54
562.86
511.12
15.77
1953.22
1704.89
585.66
Boiler / Gasifier
Flow to unit
Solids fired
Dissolved solids to unit
Amount Inorganic
Amount Organic
Energy available for Steam
Gas to Turbines, mt/hr
mt/hr
%
kg/ODtP
kg/ODtP
kg/ODtP
Mcal/day
mt/hr
118.20
66.26
1314.79
4.88
1314.80
4,71*106
x
Slaker
Flow
Reburned lime
Fresh lime
Make-up lime
mt/hr
kg/ODtP
kg/ODtP
mt/day
231.38
322.84
8.70
8.65
195.43
347.61
10.04
9.18
221.61
394.18
11.37
10.40
247.85
440.53
12.71
11.62
kg/ODtP
kg/ODtP
kg/ODtP
22.05
8.64
0.00
11.13
20.82
0.00
10.71
23.50
0.00
9.93
26.42
0.00
Chips
Brown Pulp Produced
Digester Steam
Yield
Brown Pulp
White Liquor (synthetic)
Flow
TTA as NaOH
Sulfidity
White Liquor make-up
NaOH (make-up)
Na2SO4
Elemental Sulfur
Table 5.5 continued
Kiln
42
97.74
100.61
103.45
60.18
59.66
59.17
1420.48
1871.76
1953.22
1.95
2.14
2.31
1229.70
1234.41
1238.98
5.35*106est. 5.33*106est. 5.30*106est.
523.87
539.07
554.15
Throughput
Fuel consumed
ODt/day
kg/day
Water Usage
Scrubber Mill Water
Brownstock Washer Shwr Makeup
Mud Filter Mill Water
Mud Tank Cons. Reg. Mill Water
Total
622.16
53530
705.52
60655
788.50
67766
mt/hr
143.41
131.45
151.94
172.55
mt/hr
mt/hr
mt/hr
mt/hr
117.24
22.55
283.69
566.90
121.77
22.55
283.69
559.46
107.24
22.55
283.69
565.42
92.67
22.55
283.69
571.46
kg/ODtP
Green Liqour Heater Steam
Table 5.6
571.59
56542
55.50
83.70
95.22
106.90
Effect of using Na2CO3 for Pretreatment and %EA Use on Cost
Cost analysis $/ODtP
Chips/kg
NaOH/lb
Na2SO4/lb
Fresh Lime/lb
Steam used/kg
Kiln fuel/barrel
Water/1000 gal
Total
Energy available/ODTP
Net Cost
Cost,
USD
0.062
0.141
0.05
0.032
0.006
20
0.072
x
6
x
Kraft
BLG
%
BLG
%
BLG
%
Basecase 15%EA DELTA 17%EA DELTA 19%EA DELTA
150.31 145.48
-3.32
145.48
-3.32
145.48
-3.32
6.84
3.45
-98.01
3.32
-105.91
3.08
-122.10
0.95
2.29
58.51
2.59
63.25
2.91
67.30
0.61
0.71
13.37
0.80
23.52
0.89
31.55
23.46
23.28
-0.81
24.26
3.27
25.24
7.03
8.35
7.91
-5.63
8.96
6.78
10.01
16.56
0.99
0.97
-1.33
0.98
-0.26
0.99
0.80
191.51
51.28
140.23
184.08
47.99
136.09
43
-4.04
-6.87
-3.04
186.38
47.77
138.61
-2.75
-7.36
-1.17
188.60
47.44
141.15
-1.55
-8.10
0.66
6
EFFECTS ON PULP YIELD AND PROPERTIES USING MODIFIED PULPING PROCEDURES
INVOLVING SULFUR PROFILING AND GREEN LIQUOR PRETREATMENT
Lindstrom, M.; Naithani, V.; Kirkman, A.; Jameel, H.; Effects on Pulp Yield and Properties
Using Modified Pulping Procedures Involving Sulfur Profiling and Green Liquor Pretreatment;
Presented at 2004 Tappi Fall Technical Conference, Atlanta, GA, 2004
44
EFFECTS ON PULP YIELD AND PROPERTIES USING MODIFIED PULPING PROCEDURES
INVOLVING SULFUR PROFILING AND GREEN LIQUOR PRETREATMENT
Mathias Lindstrom,Ved Naithani, Adrianna Kirkman, Hasan Jameel
Department of Wood and Paper Science
North Carolina State University
Raleigh, NC 27695-8005
USA
ABSTRACT
Using black liquor gasification (BLG), the recovered entities of sodium and sulfur can be split into two separate
fractions with varying degrees of separation dependent on the operating conditions and the technology used.
The separation of these chemicals creates some opportunities in the pulping process where chemical profiling
may be employed to increase the pulp yield or to extend delignification. This benefit may be offset by an
increase in the causticizing load. Of the various pulping options with black liquor gasification, sulfur profiling
would be the lowest capital cost process to implement, especially for mills with a modified continuous or batch
process. Previous process simulation work showed that BLG implementation alone would likely increase the
cost of pulping operations due to the higher costs associated with limekiln operation. However, it was also
found that by combining BLG implementation with pulping process modifications like split sulfidity pulping or
green liquor pretreatment, the overall operational costs associated with pulping and chemical recovery could be
substantially decreased. The work described here explores the possibilities made available for modified
continuous cooking through the implementation of black liquor gasification. Laboratory pulping protocols were
developed for modified continuous cooking, split sulfidity and green liquor pretreatment pulping. The obtained
results indicate that the split sulfidity protocol can potentially be used to produce kraft pulps with higher yield,
viscosity and strength properties relative the simulated MCC protocol; whereas the green liquor pretreatment
protocol can be used to produce pulps with yields similar to the MCC protocol, but with higher viscosity and
strength properties. Moreover, significant cost savings can be realized by using the green liquor pretreatment
approach in place of the MCC protocol, as the load on the recaust system and particularly the limekiln would be
reduced.
INTRODUCTION
The implementation of black liquor gasification, BLG, into a pulp mill chemical recovery system will allow for
the splitting of sodium and sulfur into separate streams which can then be utilized in the pulping process as
desired. This separation creates some opportunities in the pulping process, which can lead to production cost
savings or improved operations. The following modified pulping technologies can advantageously be used in
combination with black liquor gasification to realize these potential benefits.
Split Sulfidity Pulping
Polysulfide Pulping
Alkaline Sulfite Pulping: Alkaline Sulfite- AQ (AS-AQ)
or Mini-Sulfite Sulfide AQ Process (MSSAQ)
The basic principles of modified extended delignification consist of level alkali concentration throughout the
cook, a high initial sulfide concentration, low concentrations of lignin and Na + in the final stage of the cook, and
lower temperature in the initial and final stages of the cook (1). Two technologies that can accommodate these
principles, split sulfidity pulping and green liquor pretreatment, will be considered in this paper.
The concept of sulfur profiling, or split sulfidity pulping, employing a sulfur rich stream in the rapid initial
phase, followed by a sulfur lean stream in the bulk and residual phase, has been investigated (2-6). Compared
to conventional kraft cooks of similar H-factor, split sulfidity pulping has been shown to enhance selectivity of
the pulping reactions resulting in higher lignin removal and pulp viscosity. Moreover, split sulfidity pulping,
has been shown to increase pulp yield and strength properties (6-14). The basic approaches consist of an initial
45
presteaming or pretreatment stage, followed by one or two cooking stages. A sulfur rich stream was charged in
the pretreatment or first cooking stage followed by a sulfur lean stream in the second stage. Compared to
conventional reference cooks the results indicate a yield increase of about 1-2 % and higher viscosity pulp for
cooks to kappa less than 20.
In the modified batch systems commercially available the requirement for initial high sulfide concentration is
accomplished by using the spent cooking liquor containing high sulfide for pre-treatment (15). Pre-treatment of
the chips using black liquor before digesting with the white liquor allows a significant reduction of H-factor or
alkali charge to reach the same kappa level as conventional batch system (16). Modifications to enable these
methods in an existing batch mill are very capital intensive. A different alternative would be to use liquors with
different sulfidities during the heat-up phase of an indirect-heated batch digester, thereby decreasing the capital
required to attain some of the benefits of extended delignification. In the modified continuous cooking system,
white liquor is added in multiple steps to level out the alkali concentration and there are multiple extractions to
keep the dissolved lignin low towards the end of the cook. However, there is no good method to take advantage
of the high sulfidity that is beneficial during the initial phase of the cook. The implementation of black liquor
gasification would overcome these issues, allowing for the generation of separate sulfide and sodium streams
that could be applied in either batch of continuous pulping systems to take full advantage of the modified
extended delignification.
Black liquor gasification will potentially increase the causticizing load due to t he release of H2S. In
conventional recovery the sulfur ends up as Na2S in the green liquor stream, while in gasification for each mole
of H2S the sodium now becomes available as Na2CO3. A higher amount of Na2CO3 will have to be converted to
NaOH. In addition, the concurrent capture of CO2 in a downstream H2S scrubbing process will also increase the
causticizing load. One pulping technology that could potentially offset the increased demands on the recaust
cycle is green liquor pretreatment. Using green liquor to pretreat the pulp in a first cooking stage would
decrease the amount of Na2CO3 that requires conversion to NaOH for the production of traditional white liquor.
Previous work (6,7) has demonstrated the feasibility of using green liquor in the impregnation stage, without
increasing overall chemical usage. It has also been shown (17) that the amount of sulfur adsorbed during the
pretreatment decreases with higher [OH-]. The chemical sorption profiles and effect of green liquor
pretreatment in kraft pulping have been demonstrated (18,19). By impregnating chips with high sulfidity, low
pH liquor, a mill may enhance yield and further decrease the causticizing load.
Split sulfidity and green liquor pretreatment pulping processes have previously been simulated using
WinGEMS to quantify the impacts on the various mill unit operations. The results showed that BLG
implementation might increase the pulp production operational cost by $3 per oven dry ton pulp. Most of the
cost increase is due to the increase in the limekiln fuel and the lime makeup. However, the cost increase could
be overcome and turned into a cost savings of $2 per oven dry ton pulp, by a 2 % increase in pulp yield. Further
cost reduction may be realized by reducing the load on the limekiln. Increasing the pulp yield or lowering of
the kappa number to the bleach plant at the same yield could potentially be achieved by employing split
sulfidity and/or green liquor pretreatment pulping; the latter would also reduce the load on the limekiln (20).
The following experimental program explores the pulping opportunities that were demonstrated in WinGEMS
regarding pulping using split sulfidity and green liquor pretreatment. The developed cooking protocols for
these two modified pulping technologies utilize the very high initial sulfide liquor concentrations that would be
enabled by the implementation of black liquor gasification and the option of not converting all of the Na2CO3 to
NaOH in the recaust cycle. Process diagrams outlining how these liquors could be generated have been
presented in previous work (20).
46
EXPERIMENTAL
Wood
Screened mixed southern softwood chips were used throughout the laboratory cooking. For each cook, 800
oven dry grams of chips were soaked in water over night. The chips were drained for 30 minutes and placed in
the M&K batch digester prior to chemical addition.
Development of Modified Continuous Cooking procedure (21)
A laboratory cooking procedure was developed to simulate modified continuous cooking, MCC . The
cooking protocol was divided into three stages with addition of active chemical at the end of the first and
second stage. The total cooking time from time 0 (time = 0 at T = 100 C) was fixed at 240 minutes, while
the cooking temperature was varied to produce the desired h factor (153-166 C). The active alkali, AA was set
to 19.5 % on OD pulp and the sulfidity to 25 %. In addition, Na2CO3 was added as 15 % of the total chemical
on wood, simulating a system dead load. The resulting total titratable alkali, TTA was divided for addition
between the three stages. In the first stage, the chips were placed in the digester with 65 % of the TTA and
water, to bring the liquor to wood ratio, L/W, to 3.5. The cook was then brought to 120 C and held at
temperature for 15 minutes. For the second stage white liquor heated to 125 C was added equaling 20 % of the
TTA, bringing the cumulative added TTA to 85 %. The cook was then brought to the desired temperature and
held there until 105 minutes had elapsed from time 0. For the third stage, white liquor heated to 125 C
equaling the final 15 % of the TTA was added, bringing the L/W to about 4.5. The cook was then held at
temperature until complete. The M&K batch digester system used in the experimental work was fitted with
valves to enable liquor addition and extraction. An American LEWA Inc. type EK2 pump was used to feed
heated liquor under pressure into the lab digester. The parameters for the MCC procedure are outlined in Table
1.
Table 6.1
Parameters for MCC protocol
Cumulative % TTA
L/W
Stage Temperature (C)
Time at Temperature (min)
65
3.5
120
15
Stage II
85
4.1
153-166
~ 70
Stage III
100
4.5
153-166
120
Stage I
Digester liquor samples were collected at predetermined intervals throughout the cook and titrated for residual
effective alkali, using a procedure modified from SCAN-N 33:94. The sample collection scheme is outlined in
Table 2. The developed MCC procedure was employed to generate kraft baseline data for comparison to later
cooks, exploring split sulfidity and green liquor pretreatment.
Table 6.2
Outline of collection scheme for digester liquor samples
Sample
Time elapsed from Time 0
1
~ 25 minutes
Comments
At the completion of stage I, prior to liquor addition
2
60 minutes
After the first liquor addition, having reached final T and sufficient mixing
3
105 minutes
At the completion of stage II, prior to liquor addition
4
120 minutes
After the second liquor addition and sufficient mixing
5
240 minutes
At the completion of stage III, during the blow of the cook
47
Development of Split Sulfidity and Green Liquor Pretreatment procedure
The split sulfidity cooking protocols, displayed in Table 3, closely resembled that of the MCC baseline. The
main difference is the liquor profile used. The total titratable alkali was kept constant throughout both the MCC
baseline and split sulfidity cooks, but the addition scheme of the Na2S and NaOH was altered. To maximize the
sulfidity effect on the pulp, the total amount of Na2S used was added in Stage 1. The NaOH was then split
between the three different stages using two different approaches: Split Sulfidity High Initial Alkali and Split
Sulfidity Low Initial Alkali (SS_HIA and SS_LIA). The Na 2CO3 was evenly added to each stage to simulate
system dead load. In the first approach (SS_HIA) the NaOH was added evenly, one third to each stage,
resulting in a high initial alkali concentration followed by uniform increases in alkali through the liquor
additions. In the second approach (SS_LIA), the amount of NaOH added to stage I was limited to a small
amount (about 11 % of the total NaOH added), resulting in a low initial alkali concentration, but adequate to
promote a sufficiently high level of pH during the first stage.
Table 6.3
Parameters for Split Sulfidity protocols, high initial and low initial alkali
(SS_HIA/SS_LIA)
SS_HIA
% of Total for each chemical added
Cum. % TTA
Na2S
NaOH
Na2CO3
L/W
Stage I
47.6
100
33.3
33.3
3.5
120
15
Stage II
73.9
0
33.3
33.3
4.1
153-166
~70
0
33.3
33.3
4.5
153-166
120
Stage Temp. (C)
Time at Temp. (min)
Stage III
100
SS_LIA
Stage Temp. (C)
Time at Temp. (min)
% of Total for each chemical added
Cum. % TTA
Na2S
NaOH
Na2CO3
L/W
Stage I
33.28
100
10.9
33.3
3.5
120
15
Stage II
66.76
0
44.5
33.3
4.1
153-166
~70
Stage III
100
0
44.5
33.3
4.5
153-166
120
The green liquor pretreatment protocol is analogous to the one described for split sulfidity. To generate a high
concentration Na2CO3 green liquor, simulating a decreased level of recaustization in the recovery loop, 25 % of
the NaOH used previously was replaced by 25 % Na2CO3 maintaining a constant system TTA. The concept of
high and low initial alkali was again employed and the resulting parameters for the green liquor pretreatment
cooks are shown in Table 4. In order to produce pulps of lower kappa a series of cooks was additionally
performed where the system TTA was increased by 10 %, using the same TTA component breakdown as the
high initial alkali protocol. The modified green liquor pretreatment procedure was labeled High TTA.
Table 6.4
Parameters for Green Liquor Pretreatment protocols, high initial and low initial alkali
(GLPT_HIA/GLPT_LIA)
GLPT_HIA
% of Total for each chemical added
Cum. % TTA
Na2S
NaOH
Na2CO3
L/W
Stage Temp. (C)
Time at Temp. (min)
Stage I
57.33
100
33.3
65.0
3.5
120
15
Stage II
78.7
0
33.3
17.5
4.1
153-166
~70
Stage III
100
0
33.3
17.5
4.5
153-166
120
Stage Temp. (C)
Time at Temp. (min)
GLPT_LIA
% of Total for each chemical added
Cum. % TTA
Na2S
NaOH
Na2CO3
L/W
Stage I
46.6
100
10.9
65.0
3.5
120
15
Stage II
73.3
0
44.5
17.5
4.1
153-166
~70
Stage III
100
0
44.5
17.5
4.5
153-166
120
48
All obtained pulps were thoroughly washed with water, disintegrated using an impeller mixer and screened.
The screen accepts were fluffed prior to yield determination and refrigerated for storage. The total yield is
given as the sum of the oven dry accepts and rejects. Pulp kappa numbers and viscosities were determined for
all screen accept samples. Samples with kappa number exceeding 50 were treated with standard chlorite
delignification to remove undissolved lignin. A total of six screened pulp samples, representing each procedure
in the low and high kappa range, were refined using a PFI mill according to CPPA Standard C.7. Handsheets
were made according to TAPPI T 205, and tested following T 220. The resulting sheet caliper, tear, tensile and
burst, were determined according to TAPPI T 411, T 414, T 494, and T 403 respectively.
RESULTS AND DISCUSSION
MCC, Split Sulfidity and Green Liquor Pretreatment cooks
The significant difference between the cooking protocols can be found in their respective residual effective
alkali (REA) profiles. The profiles were generated from cooking liquor samples collected according to the
outline in Table 2. The points were selected at intervals from time 0, representing the end of the different
stages and points following a liquor addition having allowed for temperature adjustment and mixing. The
results, shown in Figures 1 through 4, illustrate how the liquor profiling results in significantly different levels
of cooking liquor REA between the different protocols and how they vary for each stage within each cook. As
shown, the MCC baseline cooks have a high initial REA, which then decreases over time as the cook
progresses. There is a slight increase in REA after the second liquor addition, but the goal of the procedure, a
smooth REA profile is achieved.
BMCC_H750
REA (g/L as Na2O)
22.0
BMCC_H1000
BMCC_H1500
18.0
BMCC_H1800
BMCC_H2100
14.0
10.0
6.0
0
50
100
150
200
250
Tim e (m in)
Figure 6.1
Residual Effective Alkali (REA) profiles for MCC procedure
When looking at the split sulfidity protocols the situation is very different. Here, the goal is to minimize the
initial REA (Stage I) to then increase the level and maintain a high REA throughout the second and third stages.
The initial REA level for the SS_HIA (High Initial Alkali) cooks are as expected higher than those of the
SS_LIA (Low Initial Alkali) cooks, but reach similar levels during the second and third stage. Noteworthy is
that the final REA is very similar among all three protocols. The initial data points at t = 30 minutes for the
cooks MCCSS_H1500 and H1800, are higher than expected. However, when looking at the subsequent REA
data points there is good agreement between all four cooks.
49
REA (g/L as Na2O)
12.0
MCCSS_H750
8.0
MCCSS_H1000
MCCSS_H1500
MCCSS_H1800
4.0
0
50
100
150
200
250
Tim e (m in)
Figure 6.2
Residual Effective Alkali (REA) profiles for SS_HIA procedure
REA (g/L as Na2O)
16.0
12.0
MCCSS_B_H750
8.0
MCCSS_B_H1000
4.0
MCCSS_B_H1500
MCCSS_B_H1800
0.0
0
50
100
150
Tim e (m in)
Figure 6.3
Residual Effective Alkali (REA) profiles for SS_LIA procedure
50
200
250
REA (g/L as Na2O)
12.0
8.0
GLPT_M1_H750
GLPT_M1_H1000
4.0
GLPT_M1_H1500
GLPT_M1_H1800
0.0
0
50
100
150
200
250
Tim e (m in)
Figure 6.4
Residual Effective Alkali (REA) profiles for GLPT modified procedure
Pulp yield, kappa and viscosity for MCC and Split Sulfidity cooks
Three series of baseline cooks were performed according to the given MCC protocol. The average values for
yield, kappa, and viscosities are reported in Table 5. Two series of cooks were performed using the split
sulfidity high initial alkali (SS_HIA) and low initial alkali (SS_HIA). The obtained average values for the split
sulfidity cook yield, kappa, and viscosity are displayed in Table 6. From these data the following relationships
were explored: kappa versus h factor, total yield versus kappa, and viscosity versus kappa.
Table 6.5
MCC baseline pulp yield, kappa and viscosity
MCC Baseline average values
target
actual
Total Yield (%)
kappa
Viscosity (cps)
750
726
51.4
75.9
67.6
1000
1001
48.6
58.5
41.9
1500
1527
45.3
36.8
37.0
1800
1794
44.0
33.4
32.8
2100
2044
44.0
30.8
27.8
51
Table 6.6
Split Sulfidity baseline pulp yield, kappa and viscosity
target
750
1000
1500
1800
actual
729
1015
1542
1844
SS_HIA average values
Total Yield (%)
kappa
Viscosity (cps)
52.4
64.4
63.2
49.6
55.4
55.2
46.2
36.0
42.2
43.4
27.4
30.4
target
750
1000
1500
1800
actual
711
1019
1553
1843
SS_LIA average values
Total Yield (%)
kappa
Viscosity (cps)
53.2
71.5
72.3
49.4
59.2
58.4
44.7
35.5
40.9
44.5
30.8
34.9
Figure 5 shows how the split sulfidity procedures compare to the MCC procedure in regards of lignin removal
as a function of h factor. As shown, the split sulfidity procedures produce lower kappa pulps at similar h factors
relative the MCC procedure. Also, for the split sulfidity cooks, the high initial alkali approach produced pulps
of lower kappa than the low initial alkali approach.
80.0
SS_HIA
70.0
SS_LIA
kappa
60.0
MCC Baseline
50.0
40.0
30.0
20.0
600
Figure 6.5
850
1100
1350
h factor
1600
1850
2100
Kappa number versus final h factor for the MCC baseline, SS_HIA and SS_LIA cooks
The total pulp yield as a function of kappa was compared for the split sulfidity and MCC procedures. The
resulting graph shown in Figure 6, indicates that the split sulfidity procedures produce pulp yields one to two
percent greater than the MCC procedures. The difference is more pronounced at higher kappa. Also, the split
sulfidity high initial alkali approach produced higher yields than the low initial alkali approach.
52
Total Yield (%)
54.0
52.0
50.0
48.0
SS_HIA
46.0
SS_LIA
44.0
MCC Baseline
42.0
20.0
Figure 6.6
30.0
40.0
50.0
kappa
60.0
70.0
80.0
Total yield versus kappa for the MCC baseline, SS_HIA and SS_LIA cooks
The pulp viscosity as a function of kappa was similarly compared for the different procedures. The resulting
graph is shown in Figure 7. As shown in the figure, the split sulfidity procedures produce pulps of viscosities in
the range of 5 to 10 cps greater than those of the MCC pulps at similar kappa.
Viscosity (cps)
80.0
70.0
60.0
50.0
40.0
SS_HIA
30.0
SS_LIA
20.0
20.0
MCC Baseline
30.0
40.0
50.0
60.0
70.0
80.0
kappa
Figure 6.7
Viscosity versus kappa for the MCC baseline, SS_HIA and SS_LIA cooks
Pulp yield, kappa and viscosity for MCC and Green Liquor Pretreatment cooks
The obtained yield, kappa and viscosity values for the green liquor pretreatment procedures are displayed in
Table 7. The high and low initial alkali procedures did not produce pulps of sufficiently low kappa number. To
overcome this, the system TTA was increased by 10 % and employed in a modified procedure based on the high
initial alkali approach. The obtained data is also shown in Table 7.
The obtained values for the green liquor pretreatment cooks were compared to the MCC data in a similar
manner to that for the split sulfidity cooks. Figure 8 shows the resulting pulp kappa as a function of h factor.
As seen in the figure, all of the green liquor pretreatment procedures produce pulps of kappa higher than those
of the MCC procedure. The modified green liquor pretreatment (High TTA) generated the lowest kappa pulp.
The total yields as a function of kappa, were compared and the resulting graph is shown in Figure 9. The figure
indicates that the green liquor pretreatment procedures do not produce an appreciable yield increase, producing
pulp yields similar to those of the MCC procedure at comparable levels of kappa.
53
Table 6.7
Green liquor pretreatment baseline pulp yield, kappa and viscosity
Figure 6.8.
Total Yield (%)
56.0
53.2
49.2
47.9
GLPT_HIA values
kappa
Viscosity (cps)
97.5
x
86.4
67.5
65.6
59.9
54.2
56.7
actual
726
1001
1527
1794
Total Yield (%)
55.3
x
x
51.1
GLPT_LIA values
kappa
Viscosity (cps)
98.6
53.4
x
x
x
x
70.7
50.6
target
750
1000
1500
1800
100.0
90.0
80.0
70.0
60.0
50.0
40.0
30.0
20.0
600
actual
726
1001
1527
1794
target
750
1000
1500
1800
kappa
target
750
1000
1500
1800
actual
719
1002
1424
1807
GLPT_HIA_Modified values
Total Yield (%)
kappa
Viscosity (cps)
53.7
93.7
67.1
50.4
71.8
67.3
48.5
60.6
53.9
46.1
39.5
49.5
Green Liq PT (HIA)
Green Liq PT (LIA)
Green Liq PT (Hi_TTA)
MCC Baseline
850
1100
1350
1600
h factor
1850
2100
Kappa number versus final h factor for the MCC baseline, GLPT_HIA, GLPT_LIA and
GLPT modified cooks
58.0
Total yield (%)
56.0
54.0
52.0
50.0
48.0
Green Liq PT (HIA)
46.0
44.0
Green Liq PT (Hi_TTA)
42.0
20.0
Green Liq PT (LIA)
MCC Baseline
30.0
40.0
50.0
60.0
kappa
54
70.0
80.0
90.0
100.0
Figure 6.9
Total yield versus kappa for the MCC baseline, GLPT_HIA,
modified cooks
GLPT_LIA and GLPT
Figure 10 displays the obtained viscosity values as a function of kappa. As seen in the figure, the modified
green liquor pretreatment procedure produced pulps of similar viscosity to those of the MCC procedure.
Viscosity (cps)
80.0
70.0
60.0
50.0
Green Liq PT (HIA)
Green Liq PT (LIA)
Green Liq PT (Hi_TTA)
MCC Baseline
40.0
30.0
20.0
20.0
Figure 6.10
30.0
40.0
50.0
60.0
kappa
70.0
80.0
Viscosity versus kappa for the MCC baseline, GLPT_HIA,
modified cooks
90.0
100.0
GLPT_LIA and GLPT
Pulp strength properties for low and high kappa
As previously described, pulp samples representing each of the three procedures in the low and high kappa
range, target h factor 1800 and 750, respectively, were used in the strength study. For the green liquor
pretreatment cooks, both pulps were from the modified high TTA procedure. For the MCC baseline the high
kappa pulp was selected from the second series, and the low kappa pulp from the third. For the split sulfidity
cooks, both pulps were selected from the high initial alkali procedure; the low kappa pulp from the first series
and the high kappa pulp from the second series. All pulps were refined using a PFI mill to four different points
of revolutions. The freeness responses to refining of low and high kappa pulps representing the different
cooking procedures are displayed in Figures 11 and 12. As shown in the figures, the MCC pulps refine to a
lower freeness at similar levels of refining compared to the split sulfidity and green liquor pretreatment pulps.
Also, the high kappa pulps have a significantly slower refining response than the low kappa pulps.
Split Sulfidity
Freeness (CSF)
800
GL Pretreatment
700
MCC Basline
600
500
400
300
200
0
1000
2000
3000
4000
5000
PFI (re volutions)
55
6000
7000
8000
Figure 6.11
Refining response for low kappa pulps
Freeness (CSF)
800
Split Sulfidity
GL Pretreatment
MCC Basline
700
600
500
400
300
200
0
Figure 6.12
2000
4000
6000
PFI (revolutions)
8000
10000
12000
Refining response for high kappa pulps
The values obtained through the hand sheet study are displayed in Tables 8 and 9. The low (Table 8) and high
(Table 9) kappa pulps have been treated and analyzed separately.
56
Table 6.8
Pulp strength properties for low kappa range cooks
Pulp ID
BMCC2_H1800
PFI (revolutions)
1000
3000
5000
7000
Basis Wt. (g/m2)
65.0
65.1
63.2
62.7
Apparent density
(kg/m3)
132.2
143.1
152.8
157.3
Tear Index
(mN*m2/g)
19.5
12.3
11.4
11.1
Tensile Index
(N*m/g)
66.5
74.5
77.8
85.7
Burst Index
((kPa*m2/g)
4.96
5.80
6.62
6.98
Pulp ID
1000
3000
5000
7000
SS1_H1800
64.9
63.3
63.2
62.7
131.8
144.7
150.3
156.4
16.8
14.8
14.1
13.4
65.6
76.0
78.0
77.6
5.53
6.78
6.91
7.14
Pulp ID
1000
3000
5000
7000
GLPT_M1_H1800
66.2
64.6
65.0
64.1
126.7
140.9
146.6
153.9
19.0
17.2
13.9
13.0
64.4
74.7
79.8
82.7
5.69
6.64
7.02
7.61
Table 6.9
Average Values
Pulp strength properties for high kappa range cooks
Pulp ID
BMCC3_H750
Average Values
PFI (revolutions)
2000
5000
8000
11000
Basis Wt. (g/m2)
67.2
64.8
65.2
64.3
Apparent density
(kg/m3)
119.9
134.3
142.2
151.2
Tear Index
(mN*m2/g)
19.9
12.8
10.9
11.2
Tensile Index
(N*m/g)
61.3
69.6
69.9
68.3
Burst Index
((kPa*m2/g)
4.38
5.50
6.15
6.71
Pulp ID
2000
5000
8000
11000
SS2_H750
64.5
66.9
64.1
64.0
128.0
135.9
148.6
153.9
22.9
14.5
13.8
13.1
62.8
70.8
79.6
85.6
4.98
6.65
7.19
7.00
Pulp ID
2000
5000
8000
11000
GLPT_M1_H750
66.2
64.6
65.0
64.1
131.0
145.7
151.6
159.1
19.0
17.2
13.9
13.0
62.7
67.4
70.0
72.0
5.69
6.64
7.02
7.61
The relationship between the sheet tear, tensile and burst values were investigated by plotting the tensile index
versus the tear index and the burst index versus the tear index. Figures 13 and 14 show these relationships for
the low kappa pulps, Figures 15 and 16 for the high kappa pulps.
57
Tensile Index (N*m/g)
95.0
MCC Baseline
90.0
Split Sulfidity
85.0
GL pretreatment
80.0
75.0
70.0
65.0
60.0
10.0
12.0
14.0
16.0
18.0
20.0
22.0
Tear Index (mN*m2/g)
Figure 6.13
Tensile index versus tear index for low kappa pulps
As shown in Figure 13, in the case of the more highly refined low kappa pulps the tensile index is greater for
the split sulfidity and green liquor pretreatment pulps compared to the MCC pulps. This difference is less
pronounced at higher levels of tear index. Also, as shown in Figure 14, the burst index is higher for both the
split sulfidity and green liquor pretreatment pulps relative the MCC pulps. These observations indicate that the
split sulfidity and green liquor pretreatment procedures generated low kappa pulps that are stronger than those
obtained through the MCC procedure. At a tear index of 14 mN*m 2/g the split sulfidity and the green liquor
pretreatment pulps respectively held a tensile index of 2.5 and 5 units greater than the MCC pulp. Similarly, at
the same tear index their burst index values were about 1 unit greater than that of the MCC pulp.
Split Sulfidity
Burst Index (kPa*m2/g)
9.0
MCC Baseline
8.0
GL pretreatment
7.0
6.0
5.0
4.0
10.0
12.0
14.0
16.0
18.0
20.0
22.0
Tear Index (mN*m2/g)
Figure 6.14
Burst index versus tear index for low kappa pulps
The same holds true for the high kappa pulps, as displayed in Figures 15 and 16. Also, at a tear index of 14
mN*m2/g the split sulfidity the green liquor pretreatment pulps respectively held a tensile index of 12 and 2.5
units greater than the MCC pulp. Here, at the same tear index their burst index values were about 1.5 units
greater than that of the MCC pulp. Overall, the split sulfidity produced the highest tensile index values, while
its values for tear index and burst index were similar to those of the green liquor pretreatment pulps. The MCC
pulps produced the lowest values in all three indexes. However, the green liquor pretreatment pulps were of a
kappa slightly higher than those of obtained using the MCC procedure.
58
Tensile Index (N*m/g)
90.0
85.0
80.0
75.0
70.0
65.0
60.0
55.0
50.0
10.0
Figure 6.15
Split Sulfidity
GL pretreatment
MCC Baseline
12.5
15.0
17.5
20.0
Tear Index (mN*m2/g)
22.5
25.0
27.5
Tensile index versus tear index for high kappa pulps
Burst Index (kPa*m2/g)
9.0
Split Sulfidity
8.0
GL pretreatment
7.0
MCC Baseline
6.0
5.0
4.0
3.0
10.0
Figure 6.16
12.5
15.0
17.5
20.0
Tear Index (mN*m2/g)
22.5
25.0
27.5
Burst index versus tear index for high kappa pulps
CONCLUSIONS
The split sulfidity procedures generated pulps of lower kappa than the MCC procedure at similar h factors. At
similar kappa, the split sulfidity cooks produced total pulp yields one to two percent greater than those achieved
through the MCC procedure, with significantly higher pulp viscosities. The split sulfidity high initial alkali
approach produced pulps of lower kappa and higher yields than the low initial alkali approach. The high initial
alkali pulps produced higher tensile and burst index values relative the MCC pulps at similar tear index.
The green liquor pretreatment procedures generated pulps of higher kappa than the MCC procedure at similar h
factors. Increasing the green liquor pretreatment system TTA by 10 % lowered the pulp kappa, but it still
remained higher than those of the MCC pulps. The pulp yields and viscosities were similar to those of the
MCC procedure. The high TTA green liquor pretreatment pulps produced higher tensile and burst index values
relative the MCC pulps at similar tear index.
The MCC pulps were slightly easier to refine relative to the split sulfidity and green liquor pretreatment pulps.
The obtained results indicate that the split sulfidity protocol can potentially be used to produce pulps with
higher yield, viscosity and strength properties relative the simulated MCC protocol; whereas the green liquor
59
pretreatment protocol produced pulps with yields and viscosities similar to the MCC protocol, but with higher
strength properties. Moreover, significant cost savings can be realized by using the green liquor pretreatment
approach in place of the MCC protocol, as the load on the recaust system and particularly the limekiln would be
reduced.
FUTURE WORK
The area of green liquor pretreatment will be further explored to determine the required level of system TTA to
generate pulps of similar kappa to those produced using the MCC protocol. Further studies will explore the
application of polysulfide in the first stage of the cooking sequence.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
Johansson, B., Mjoberg, J., Sandstrom, P., Teder, A., Svensk Papperstid. 87(10):30 (1984).
Herschmiller, D.W., Breaking the yield barrier, p. 59 TAPPI PRESS, 1998
LeMon, S. and Teder, A., Svensk Papperstidning, 79(11):407 (1973)
Olm, L. and Tisdat, G., Svensk Papperstidning, 82(15):458 (1979)
Olm, L. and Teder, A., Paperi ja Puu, 63(4a):315 (1981)
Andrews, E.K., dissertation, North Carolina State University, 1982
Andrews, E.K., Chang, H-m., Kirkman, A.G., Eckert, R.C., Extending delignification in kraft and
kraft/oxygen pulping of softwood by treatment with sodium sulfur liquors, Japan Tappi Symposium on
Wood pulp chemistry, 1982.
Jiang, J.E., Crofut, K.R., Jones, D.B., Proceedings of the TAPPI pulping conference, p. 799, TAPPI
Press, Atlanta, GA (1993)
Lownertz, P.P.H. and Herschmiller, D.W., Proceedings of the TAPPI pulping conference, p. 1217,
TAPPI Press, Atlanta, GA (1994)
Jiang, J.E. and Herschmiller, D.W., Proceedings of the TAPPI pulping conference, p. 317, TAPPI
Press, Atlanta, GA (1996)
Mao, B. and Hartler, N., Paperi ja Puu, 74(6):491 (1992)
Mao, B. and Hartler, N., Nordic Pulp and Paper Research J., 7(4):168 (1992)
Mao, B. and Hartler, N., Paperi ja Puu, 77(6-7):419 (1995)
Olm, L.; Tormund, D.; Jensen, Nord. Pulp Pap. Res. J., 15:62 (2000)
Tormund, D., Teder, A., A new finding on sulfide chemistry in kraft pulping, Tappi pulping
conference p 247 (1989).
Abuhasan, J., PhD thesis North Carolina State University, p 82 (1994).
I. Lopez, H-m Chang, H. Jameel and W. Wizani, Tappi Pulping Conference Proceedings, p135
Orlando, Florida October 1999
Ban, W., Lucia, L. A., Ind. Eng. Chem. Res. 42, no. 3: 646-652 (2003)
Lucia, L. A., Ban, W., Ragauskas, A. J., Pap Age 118, no. 8: 24-26 (2003)
Lindstrom, M., Kirkman, A., Jameel, H. et al. Economics of Integrating Black Liquor Gasification
With Pulping: Part I Effect of Sulfur Profiling. Proceedings, 2002 TAPPI Fall Technical
Conference, TAPPI Press, Atlanta, GA, (2002) CD-ROM
Andritz Inc., Personal communication, October, 2003.
60
7
THE EFFECT OF INTEGRATING POLYSULFIDE PULPING AND BLACK LIQUOR
GASIFICATION ON PULP YIELD AND PROPERTIES
Lindstrom, M.; Naithani, V.; Kirkman, A.; Jameel, H. The Effect of Integrating Polysulfide
Pulping and Black Liquor Gasification on Pulp Yield and Properties; Proc. 2005 TAPPI
Engineering, Pulping & Environmental Conference; TAPPI Press: Atlanta, GA, 2005.
61
THE EFFECT OF INTEGRATING POLYSULFIDE PULPING AND BLACK
LIQUOR GASIFICATION ON PULP YIELD AND PROPERTIES
Mathias Lindstrm,Ved Naithani, Adrianna Kirkman, Hasan Jameel
Department of Wood and Paper Science
North Carolina State University
Raleigh, NC 27695-8005
USA
ABSTRACT
The implementation of black liquor gasification (BLG) into the Kraft recovery cycle would present several
opportunities and potential benefits regarding pulp mill operation and process economics. In a mill using BLG
the recovered entities of sodium and sulfur could be split into two separate fractions. The separation of these
chemicals would enable the application of modified pulping technologies to increase the pulp yield or extend
delignification. BLG combined with polysulfide (PS) and/or anthraquinone (AQ) addition, could substantially
decrease the overall operational costs associated with Kraft pulping and chemical recovery. The effects of Kraft
cooking with polysulfide and anthraquinone addition were explored through a multi-stage laboratory cooking
protocol simulating modified continuous cooking (MCC). Investigated variables include the pulping liquor
concentrations of polysulfide, HS-, OH-, and CO3-2 and the resulting level of charged stage and system total
titratable alkali (TTA). Results showed yield increases of 1% per % PS for the investigated range of 1 to 3% PS
on wood. The PS pulps refined more quickly relative the MCC reference, and had a tear index penalty at fixed
levels of tensile and burst index comparable to literature data. The effect of initial active alkali charge was
found to be influential relative obtained pulp yields at higher levels of PS charge.
INTRODUCTION
The implementation of black liquor gasification, BLG, into a pulp mill chemical recovery system will allow for
the splitting of sodium and sulfur into separate streams. This separation will enable several modified pulping
technologies which can be used to increase pulp yield and improve pulp properties. The utilization of
polysulfide (PS), often in conjunction with anthraquinone (AQ), as additives to the Kraft process, their effect on
pulping chemistry and kinetics has been explored for some time (1-6). Its effectiveness has been established,
and is typically reported as increasing the pulp yield by one percent for each percent of PS added to the pulping
liquor (7-8). However, efficiently generating high concentrations of PS within the Kraft chemical recovery
cycle has proved a challenge. There are currently three primary competing processes available for PS
generation, Chiyoda, MOXY and Paprilox (9). These processes, in general terms, produce pulping liquors
with PS concentrations of five to eight grams per liter and PS selectivities ranging from 60 to 90 percent (1013). A chemical recovery system based around BLG would allow for different pathways to generate PS liquors
at potentially higher concentrations and selectivities.
Research efforts in the area of PS have generally been in one of two major areas; work investigating optimum
parameters for PS pulping (14-17), or PS pulping as it relates to PS utilization in Kraft process operations
and/or associated PS generation technologies (10-13,18-20). A smaller area of work has been based around the
potential implementation of BLG and the opportunities created by the unrestricted management of sulfur and
sodium as separate entities (21). This paper is aimed at evaluating the potential of PS pulping in conjunction
with gasification and novel recovery operations and their implications on the modern Kraft process.
In addition to making chemical recovery and pulping liquor regeneration feasible for several modified pulping
technologies, such as Kraft split sulfidity and Kraft PSAQ pulping, as well as sulfite- AQ and mini-sulfite
sulfide-AQ, the separation of sulfur and sodium can advantageously be applied in modified delignification. The
basic principles of modified extended delignification consist of level alkali concentration throughout the cook, a
high initial sulfide concentration, low concentrations of lignin and Na+ in the final stage of the cook, and lower
temperature in the initial and final stages of the cook (22). These principles could straightforwardly be
implemented into an existing continuous Kraft or super-batch process, along with the possibility generating and
applying high concentrations of PS.
62
Some limitations to the generation and utilization of pulping liquors with high PS concentrations exist. The
most fundamental will be the amount of sulfur available in the pulping system, and how this sulfur most
efficiently should be applied to the wood, as a balance between Na2S and PS. In addition the amount of NaOH
that can be added to each stage can also be modified. The work presented outlines some of the basic
considerations of Kraft pulping with PSAQ as they relate to the pulping system chemical charge and the effects
on pulp yield and properties.
EXPERIMENTAL
Mixed southern softwood chips were used throughout the laboratory pulping. The chips were air-dried to
greater than 90 percent solids content and screened collecting the four to ten mm fraction. The chips were
soaked in water over night, and then drained for 30 minutes before being placed in the M&K batch digester
prior to chemical addition. 800 oven dry grams of wood were used in each cook. Polysulfide liquor was
generated by dissolving elemental sulfur in Na2S. The mixture was heated at 60 C under an N 2 atmosphere
until completely dissolved.
Reference pulps were generated using a simulated modified continuous cooking procedure (23). The cooking
protocol was divided into three stages with addition of active chemical at the end of the first and second stage.
The active alkali, AA, charge was varied from 17.0 to 22.5% on OD pulp to produce pulps in the 25 to 40 kappa
range. Two different sulfidity conditions were explored, 25 and 40%. In addition, Na2CO3 was added as 15%
of the total chemical on wood, simulating a system dead load. The resulting total titratable alkali, TTA was
divided for addition between the three stages. In the first stage, the chips were placed in the digester with 65%
of the TTA and water, to bring the liquor to wood ratio, L/W, to 3.5. The cook was then brought to 120 C and
held at temperature for 15 minutes. For the second stage white liquor heated to 125 C was added equaling 20%
of the TTA, bringing the cumulative added TTA to 85%. The cook was then brought to the desired temperature
and held there until 105 minutes had elapsed from time 0. For the third stage, white liquor heated to 125 C
equaling the final 15% of the TTA was added, bringing the L/W to about 4.5. The cook was then held at
temperature until complete. The total cooking time from time 0 (time = 0 at T = 100 C) was fixed at 240
minutes, while the cooking temperature was held at 164 C. This resulted in cooks of h factor slightly above
1800. The parameters for the MCC procedure are outlined in Table 1.
Table 7.1
Parameters for MCC protocol
Cumulative % TTA
L/W
Stage Temperature (C)
Time at Temperature (min)
65
3.5
120
15
Stage II
85
4.1
164
~ 70
Stage III
100
4.5
164
120
Stage I
Kraft cooks employing PS were performed according to the outlined MCC procedure with some modifications.
The main differences were the utilization of sulfur added as a combination of PS and Na2S, the profile of NaOH
and Na2CO3 addition, and the addition of anthraquinone, AQ. AQ was added to each PS cook at a charge of
0.1% on oven dry wood to enhance the delignification rate. Furthermore, to maximize the effect of PS and
system sulfidity, according to the principles of extended delignification, the total amount of Na2S used was
added in Stage 1. The original procedure employed a chemical charge of 19.5% AA and 25% sulfidity. As this
chemical charge was converted for the PSAQ procedure, the total amount of available sulfur was determined.
Based on 800 oven dry grams of wood, this equates a total of 20.13 grams sulfur. For a one percent PS charge
on oven dry wood, 8 grams of sulfur was required as PS. The remainder was used as Na2S in the generation of
the PS liquor. Any remaining sulfur was added as Na2S. As the desired PS charge on wood is increased, more
sulfur is required in the system. This can be achieved by either increasing the active alkali charge or system
sulfidity, or a combination thereof. Table 2 illustrates this balance displaying two examples of the partitioning
of the total sulfur available in the system at 25 and 40% sulfidity. The sulfur division between PS and Na 2S is
dictated by the amount required for PS, the balance being Na2S. As seen in the table, PS charges slightly
63
exceeding 2 % is possible at 19.5% AA with 25% sulfidity. The corresponding value at 40% sulfidity is about
4% PS. To enable higher PS charges either the system AA charge or the sulfidity must be increased.
Table 7.2
Demonstration of sulfur utilization and system availability for PS generation
25% Sulfidity
Cook
Procedure
Total S
available (g)
40% Sulfidity
S required
for PS (g)
S available as
Na2S (g)
Total S
available (g)
S required
for PS (g)
S available as
Na2S (g)
MCC
20.13
0
20.13
32.21
0
32.21
1% PS
20.13
8.0
12.13
32.21
8.0
24.21
2% PS
20.13
16.0
4.13
32.21
16.0
16.21
3% PS
20.13
24.0
- 3.87
32.21
24.0
8.21
4% PS
20.13
32.0
- 11.87
32.21
32.0
0.21
To offset the loss of TTA added to Stage 2 and 3 resulting from the addition of the entire amount of sulfur
species in Stage 1, the profile for alkali addition was altered accordingly. The lost balance of TTA in Stage 2
and 3 was made up by decreasing the addition of NaOH and Na 2CO3 to Stage 1 in amounts equal to that of the
added sulfur species, and charging the alkali to Stage 2 and 3 maintaining the 65/20/15 ratio employed in the
MCC baseline cooks. The values for the 40% sulfidity system are given in parenthesis. The resulting splits for
chemical addition in the PSAQ cooks are showed in Table 3.
Table 7.3
Parameters for Polysulfide-Anthraquinone (PSAQ) cooks
PSAQ
% of Total for each chemical added
Cum. % TTA
Na2S
NaOH
Na2CO3
L/W
Stage Temp. (C)
Time at Temp. (min)
65
100
56 (47)
56 (47)
3.5
120
15
Stage II
85
0
25 (30)
25 (30)
4.1
164
~70
Stage III
100
0
19 (23)
19 (23)
4.5
164
120
Stage I
In addition to the PSAQ liquor parameters outlined in Table 3, two more cooking protocols were investigated at
3% PS charge, where the alkali charge in Stage 1 was increased. These conditions were aimed at exploring the
effect of higher initial alkali charge on the effectiveness of PS relative the delignification rate and pulp
properties. The initial PSAQ protocol employed at 3% PS charged was labeled low initial alkali (LIA), which
was then followed by a medium initial alkali charge (MIA) and a high initial alkali charge (HIA). The liquor
parameters for the PSAQ-MIA and HIA protocols are displayed in Table 4.
Table 7.4
Parameters for medium (MIA) and high (HIA) initial alkali PSAQ cooks
PSAQ-MIA
% of Total for each chemical added
Cum. % TTA
Na2S
NaOH
Na2CO3
L/W
Stage Temp. (C)
Time at Temp. (min)
Stage I
77
100
65
65
3.5
120
15
Stage II
90
0
20
20
4.1
164
~70
Stage III
100
0
15
15
4.5
164
120
PSAQ-HIA
% of Total for each chemical added
Cum. % TTA
Na2S
NaOH
Na2CO3
L/W
Stage Temp. (C)
Time at Temp. (min)
Stage I
84
100
75
75
3.5
120
15
Stage II
94
0
15
15
4.1
164
~70
Stage III
100
0
10
10
4.5
164
120
64
All obtained pulps were thoroughly washed with water, disintegrated using an impeller mixer and screened.
The screen accepts were fluffed prior to yield determination and refrigerated for storage. The total yield is
given as the sum of the oven dry accepts and rejects. Pulp kappa numbers and viscosities were determined for
all screen accept samples. Viscosities for samples with kappa number exceeding 50 were treated not
determined. A total of nine screened pulp samples representing each procedure, were refined using a PFI mill
according to CPPA Standard C.7. Handsheets were made according to TAPPI T 205, and tested following T
220. The resulting sheet caliper, tear, tensile and burst, were determined according to TAPPI T 411, T 414, T
494, and T 403 respectively.
RESULTS AND DISCUSSION
The obtained results will be discussed in two separate parts. The first segment addresses the pulping results
generated using 25% and 40% sulfidity respectively, followed by a second segment describing the pulp
strengths.
Pulp yield, kappa and viscosity for 25% sulfidity pulps
The laboratory cooking parameters and the yield, kappa, and viscosities obtained for the 25% sulfidity cooks are
shown in Table 5. The first three cooks in the table represent the reference MCC pulps, followed by the 1 and
2% PS pulps. As seen in the table, the system AA charge was varied to generate circa kappa 30 pulps. The 2%
PS procedure required a higher AA charge to accomplish this. From these data the following relationships were
explored: kappa versus AA charge, total yield versus kappa, and viscosity versus kappa. The resulting graphs
are shown in Figures 1 through 3.
Table 7.5
Summary of cooks performed using 25% sulfidity
h factor
% AA on
OD wood
Sulfidity
(%)
TTA
(gpl as Na2O)
% PS on
OD wood
25-MCC1
1842
19.5
25
183.5
0
25-MCC2
1849
20.5
25
192.9
0
25-MCC3
1849
21.5
25
202.4
0
kappa
Tot.
Yield
(%)
Viscosity
(cps)
0.01
37.7
46.4
47.8
0.01
31.7
45.7
37.2
0.01
26.0
44.5
32.4
0
COOK
0
0
% AQ on
OD wood
25-PSAQ1
1843
19.5
25
220.6
1
0.01
32.5
48.4
41.8
25-PSAQ2
1841
20.5
25
192.9
1
0.01
29.8
47.4
38.5
25-PSAQ3
1847
21.5
25
202.4
1
0.01
25.5
46.7
35.5
25-PSAQ4
1847
20.5
25
192.9
2
0.01
38.6
49.1
41.9
25-PSAQ5*
1845
20.5
25
163.9
2
0.01
41.5
50.5
49.1
25-PSAQ6
1847
21.5
25
202.4
2
0.01
32.3
48.4
36.9
25-PSAQ7*
1842
21.5
25
172.4
2
0.01
37.2
50.4
49.1
25-PSAQ8
1852
22.5
25
211.8
2
0.01
31.7
48.3
35.1
Olm and Tormund (21) reported, in work exploring PS pretreatment at zero initial effective alkali, a strong
negative effect of CO3-2 charged to the pretreatment stage on pulp yield at kappa 30. Based on these findings it
was reasonable to investigate if this would also be the case in PS pretreatments at higher alkali charge. In
addition to the previously described procedures, two cooks were performed to investigate the potential negative
effect of carbonate dead load on pulp yield. Cooks 25-PSAQ5* and 25-PSAQ7* were performed without the
addition of Na2CO3 resulting in lower system TTA. As compared to the other described 2% PS pulps, the
obtained (carbonate-free) pulps were of higher kappa, accordingly having somewhat higher pulp yield and
viscosity. This indicates that the presence of carbonate, when added in typical process dead load amounts, in
PS pretreatments at sufficiently high alkali charge did not seem to have a negative impact on pulp yield at
similar kappa. No further investigation of this topic was made in this work.
65
MCC 25% S
40
kappa
45
1% PS, 25% S
2% PS, 25% S
35
30
25
20
18.5
Figure 7.1
19.5
20.5
21.5
22.5
Active Alkalie Charge (% on OD wood)
23.5
Kappa number versus AA charge for the MCC baseline and PS cooks at 25% sulfidity
Figure shows 1 how the 1 and 2% PS procedures compare to the MCC procedure at 25% sulfidity, in regards of
lignin removal as a function of AA charge. As shown, the 1% PS procedure produced lower kappa pulps at
similar AA charge relative the MCC procedure, whereas the 2% PS procedure produced pulps of somewhat
higher kappa.
The total pulp yield as a function of kappa was compared for the 1 and 2% PS and MCC procedures. The
resulting graph shown in Figure 2, indicates that the PS procedures produced pulp yields about two percent
greater than the MCC procedure.
Total Yield (%)
50
47.5
MCC 25% S
1% PS, 25% S
45
42.5
22.5
Figure 7.2
2% PS, 25% S
27.5
32.5
kappa
37.5
42.5
Total yield versus kappa for the MCC baseline and PS cooks at 25% sulfidity
The pulp viscosity as a function of kappa was similarly compared for the different procedures. The resulting
graph is shown in Figure 3. As shown in the figure, the 1% PS procedure produced pulps of slightly greater
viscosities than the MCC reference, whereas the 2 % PS procedure resulted in slightly lower viscosities at
similar kappa.
66
Viscosity (cps)
55
MCC 25% S
50
1% PS, 25% S
2% PS, 25% S
45
40
35
30
22.5
Figure 7.3
27.5
32.5
kappa
37.5
42.5
Viscosity versus kappa for the MCC baseline and PS cooks at 25% sulfidity
Pulp yield, kappa and viscosity for 40% sulfidity pulps
The laboratory cooking parameters and the yield, kappa, and viscosities obtained for the 40% sulfidity cooks are
reported in Table 6. The first four cooks in the table represent the reference MCC pulps, followed by the 1, 2
and 3% PS pulps. The system AA charge was again varied to generate circa 30 kappa pulps. As seen in the
Table 7.6
Summary of cooks performed using 40% sulfidity
COOK
h factor
% AA on
OD wood
Sulfidity
(%)
TTA
(gpl as Na2O)
% PS on
OD wood
% AQ on
OD wood
kapp
a
Tot.
Yield (%)
Viscosity
(cps)
40-MCC1
1845
19.5
40
183.5
0
0.01
31.1
46.6
52.1
40-MCC2
1847
20.5
40
192.9
0
0.01
27.0
46.1
46.9
40-MCC3
1844
21.5
40
202.4
0
0.01
24.0
45.6
35.8
40-MCC4
1842
17.0
40
160.0
0
0.01
43.4
48.4
56.8
40-PSAQ1
1846
20.5
40
192.4
1
0.01
36.1
49.6
54.7
40-PSAQ2
1841
21.5
40
220.6
1
0.01
27.8
47.9
50.1
40-PSAQ3
1838
22.5
40
211.8
1
0.01
24.7
46.4
43.7
40-PSAQ4
1841
20.5
40
192.9
2
0.01
42.3
51.5
57.9
40-PSAQ5
1846
22.5
40
211.8
2
0.01
36.3
49.1
47.6
40-PSAQ6
1840
23.5
40
220.6
2
0.01
31.2
47.9
49.2
40-PSAQ7
1839
20.5
40
192.4
3
0.01
66.6
56.4
x
40-PSAQ8
1845
23.5
40
220.6
3
0.01
47
51.8
62.6
40-PSAQ9
1844
25.5
40
240
3
0.01
32.4
47.9
48.8
40-PSAQ10
1843
23.5
40
220.6
3
0.01
44.3
51.7
60.7
40-PSAQ11
1836
24.5
40
230.6
3
0.01
33.7
49.2
42.2
40-PSAQ12
1841
25.5
40
240.0
3
0.01
28
48.7
40.3
40-PSAQ13
1839
22.5
40
211.8
3
0.01
42.6
48.5
53.1
40-PSAQ14
1846
23.5
40
220.6
3
0.01
32.4
46.3
42.5
40-PSAQ15
1845
24.5
40
230.6
3
0.01
29
45.6
44.4
table, in the case of the 1% PS procedure this was accomplished using AA charges similar to those of the MCC
baseline pulps around 20.5%. The required AA values for 2% PS were somewhat higher at 23.5% to reach
67
kappa 30. In the case of the 3% PS procedure kappa 30 was not readily attainable and the AA charge was
initially increased to 25.5% resulting in kappa 32.4. As a result it was decided to increase the initial charge of
alkali (NaOH and Na2CO3) to promote delignification rates. The alkali profile for the 1 and 2% PS pulps
followed the balanced TTA approach used throughout the preceding cooks. Two new different alkali profiles
were devised, as outlined in Table 4, where 65% and 75% of the available alkali (NaOH and Na 2CO3) was
charged to Stage 1. The sequence using 65% of the available alkali, here called medium initial alkali (MIA),
uses the same alkali profile relative NaOH and Na2CO3 as the MCC baseline cooks but has the entire amount of
sulfur (Na2S/PS) charged to Stage 1. In the second sequence this amount was increased to 75%, and the cooks
labeled high initial alkali (HIA).
The results from the 1 and % PS procedures at 40% sulfidity were compared to those from the 1 and 2% PS
procedures at 25% sulfidity, and are shown in Figures 4 through 6.
45
MCC 25% S
kappa
40
1% PS, 25% S
35
2% PS, 25% S
MCC 40% S
30
1% PS, 40% S
2% PS, 40% S
25
20
16.5
17.5
18.5
19.5
20.5
21.5
22.5
23.5
24.5
Active Alkalie Charge (% on OD wood)
Figure 7.4
Kappa number versus AA charge for 25% and 40% sulfidity cooks
As seen in Figure 4, the resulting kappa at similar AA charge was lower for the 40% sulfidity MCC procedure
relative to 25% sulfidity. The kappas obtained using the 1 and 2% PS procedures at 40% sulfidity were
somewhat higher than those at 25% sulfidity and also higher than the MCC reference pulps.
Total Yield (%)
52.5
50
MCC 25% S
1% PS, 25% S
47.5
2% PS, 25% S
MCC 40% S
45
42.5
22.5
Figure7.5
1% PS, 40% S
2% PS, 40% S
27.5
32.5
kappa
37.5
42.5
Total yield versus kappa for 25% and 40% sulfidity cooks
The total pulp yield as a function of kappa was similarly compared for the 1 and 2% PS and MCC procedures at
both 25 and 40% sulfidity. As shown in Figure 5, the 40% PS procedures produced pulp yields similar to those
at 25% sulfidity, but only about one percent greater than the MCC baseline pulp at 40% sulfidity.
68
Viscosity (cps)
60
MCC 25% S
50
1% PS, 25% S
2% PS, 25% S
40
MCC 40% S
1% PS, 40% S
2% PS, 40% S
30
22.5
Figure 7.6
27.5
32.5
kappa
37.5
42.5
Viscosity versus kappa for 25% and 40% sulfidity cooks
The comparison of pulp viscosities is shown in Figure 6 indicates that the 40% sulfidity procedures produced
higher viscosity pulp relative those at 25% sulfidity. There is no clear viscosity improvement when comparing
the 1 and 2% PS and MCC procedures at 40% sulfidity.
The pulps generated using 3% PS charge at 40% sulfidity were compared to the MCC baseline at 40% sulfidity
using the same relationships. As seen in Figure 7 the 3% PS procedures, low, medium and high initial alkali
charge, produced pulps of similar kappa to the MCC reference, but required higher AA charges to reach kappa
30. The high initial alkali charge procedure resulted in the lowest pulp kappas, while the low initial alkali
charge procedure resulted in the highest pulp kappas. It was still possible to achieve circa 30 kappa pulp using
the LIA procedure, but the required AA charge was 25.5% on oven dry wood.
70
kappa
60
MCC 40% S
50
3% PS, 40% S, LIA
40
3% PS, 40% S, MIA
30
3% PS, 40% S, HIA
20
10
16.5
17.5
18.5
19.5
20.5
21.5
22.5
23.5
24.5
25.5
26.5
Active Alkalie Charge (% on OD wood)
Figure 7.7
Kappa number versus AA charge for the MCC baseline and PS cooks at 40% sulfidity
The total pulp yield for the different 3 % PS procedures relative the MCC reference is shown in Figure 8. As
shown, the medium initial alkali charge resulted in pulp yields about 3% greater than the MCC baseline at
similar kappa. The low initial alkali charge procedure also produced greater pulp yields than the MCC baseline.
These values were similar to the medium initial alkali in the higher kappa range, but slightly smaller at lower
kappa. The high initial alkali charge produced pulp yields comparable to those of the MCC reference. Worth
noting is that the pulp yield increased with increasing levels of initial AA charge for the low and medium initial
alkali procedures, as compared to the MCC reference. This compares well to findings published by Brannvall
et al., which indicated that too high or too low an alkali charge in the cooking stage following PS pretreatment
will negatively affect the pulp yield (18).
69
Total Yield (%)
57.5
55
52.5
MCC 40% S
50
3% PS, 40% S, LIA
47.5
3% PS, 40% S, MIA
45
3% PS, 40% S, HIA
42.5
20
Figure 7.8
30
40
kappa
50
60
70
Total yield versus kappa for the MCC baseline and PS cooks at 40% sulfidity
The pulp viscosity variation with kappa for the 3% PS pulps is shown in Figure 9. As seen in the figure, there is
no viscosity improvement relative the MCC baseline pulps in the lower kappa range. The low initial alkali
procedure resulted in higher viscosities than the medium and high initial alkali procedures.
Viscosity (cps)
65
60
55
50
MCC 40% S
45
3% PS, 40% S, LIA
40
3% PS, 40% S, MIA
35
3% PS, 40% S, HIA
30
20
Figure 7.9
25
30
35
kappa
40
45
50
Viscosity versus kappa for the MCC baseline and PS cooks at 40% sulfidity
The results displayed in Figures 8 and 9 may be explained by the system TTA profile, as the initial alkali is
increased, the alkali profile is moving away from the principles of extended delignification. In the case of the
simulated MCC procedure only 65% of the system TTA is initially charged to the cook. The corresponding
value for the 3% PS high initial alkali procedure is 84%. This may lead to increased carbohydrate degradation
relative the MCC reference pulp. More work is required to fully optimize procedures utilizing very high PS
charges and the effect of the system alkali profile on pulp yield and properties.
Pulp strength properties for 25% and 40% sulfidity pulps
As previously described, pulp samples of circa 30 kappa generated from each of the procedures were used in the
strength study. All pulps were refined using a PFI mill to four different points of revolutions. The freeness
responses to refining of the pulps 25% and 40% sulfidity cooking procedures are displayed in Figures 10 and 11
respectively. As shown in the figures, the PS pulps refine to a lower freeness at similar levels of refining
70
compared to the MCC reference pulps. Also, there is not a great difference in refining response between the 25
and 40% sulfidity PS pulps.
Freeness (CSF)
800
700
600
MCC 25% S
500
400
25% S 2% PS
25% S 1% PS
300
200
100
0
1000
2000
3000
4000
5000
6000
7000
PFI mill (revolutions)
Figure 7.10
Refining response for 25% sulfidity pulps
800
MCC 40% S
700
Freeness (CSF)
40% S 1% PS
600
40% S 2% PS
500
400
40% S 3% PS LIA
300
40% S 3% PS HIA
40% S 3% PS MIA
200
100
0
Figure 7.11
1000
2000
3000
4000
PFI mill revolutions
5000
6000
7000
Refining response for 40% sulfidity pulps
The handsheets were tested according to TAPPI standards for caliper, tear strength, tensile strength and burst.
The obtained data was used to calculate the sheet apparent density, tear, tensile and burst indices. The results
are displayed in Tables 7 and 8, showing the values for each of the selected samples at four points of refining.
The investigated pulp samples at 25% sulfidity were one MCC baseline pulp (25-MCC2), and one sample each
for the one and two percent PS procedures (25-PSAQ2 and 25-PSAQ8).
71
Table 7.7
Pulp strength properties for 25% sulfidity cooks
Pulp ID
25-MCC2
Average Values
PFI (revolutions)
Basis Wt. (g/m2)
Apparent density
(kg/m3)
Tear Index
(mN*m2/g)
Tensile Index
(N*m/g)
Burst Index
(kPa*m2/)g)
1000
67.9
118.2
24.7
95.0
5.53
2000
64.3
140.4
20.3
111.5
6.95
4000
62.0
142.4
16.6
128.7
7.18
6000
61.7
139.5
16.2
133.4
7.49
(Pulp ID)
25-PSAQ2
1000
64.5
126.3
20.7
103.4
5.14
2000
61.1
136.3
19.6
115.5
6.45
4000
66.2
145.5
17.8
117.0
7.19
6000
61.7
149.5
17.1
138.2
7.18
(Pulp ID)
25-PSAQ8
1000
71.9
130.4
22.2
87.2
5.48
2000
63.6
134.1
18.1
113.8
6.31
4000
69.3
147.9
15.8
115.4
7.41
6000
68.3
147.8
16.1
127.9
7.40
There was a slight increase in apparent density for the PS pulps as compared to the MCC reference, but there
was no significant difference between the 1% and 2% PS charges. The relationship between the sheet tear,
tensile and burst index values were investigated by plotting the tensile index versus the tear index and the burst
index versus the tear index. Figures 12 and 13 show these relationships for the 25% sulfidity pulps.
Tensile Index (N*m/g)
150.0
MCC 25% S
1% PS, 25% S
125.0
2% PS, 25% S
100.0
75.0
10.0
Figure 7.12
15.0
20.0
Tear Index (mN*m2/g)
25.0
30.0
Tensile index versus tear index 25% sulfidity cooks
As shown in Figure 12, the sheet tensile index for the PS pulps is somewhat lower than that of the MCC
reference pulp at similar levels of tear index. The difference is more pronounced at higher levels of tear index.
When considering the relationship of sheet burst index and tear index (Figure 13), the situation is similar. The
72
MCC reference has somewhat greater burst index at similar levels of tear index, with a more pronounced
difference at higher levels of tear index. These results are in line with values reported in the literature (3,8,9).
73
Burst Index (kPa*m2/g)
9.00
MCC 25% S
8.00
1% PS, 25% S
7.00
2% PS, 25% S
6.00
5.00
4.00
10.0
15.0
20.0
25.0
30.0
2
Tear Index (mN*m /g)
Figure 7.13
Burst index versus tear index for 25% sulfidity cooks
The strength data for the selected pulp samples from laboratory cooks performed at 40% sulfidity are shown in
Table 8. One sample each from the MCC baseline (40-MCC2), and the one and two percent PS procedures (40PSAQ2 and 40-PSAQ6), as well as one each from the three percent PS cooks at low, medium and high initial
alkali charge (40-PSAQ9 (LIA), 40-PSAQ12 (MIA), 40-PSAQ15 (HIA)). The apparent density of the MCC
reference pulp at 40% sulfidity is somewhat greater than that of the 25% sulfidity pulp at all levels of refining,
and is comparable to the values for the one to three percent PS pulps generated at 40% sulfidity. The
relationship between the sheet tear, tensile and burst values were investigated in a similar fashion. Figures 14
and 15 show these relationships for the 40% sulfidity pulps.
74
Table 7.8
Pulp strength properties for 40% sulfidity cooks
Pulp ID
40-MCC2
Average Values
PFI (revolutions)
Basis Wt. (g/m2)
Apparent density
(kg/m3)
Tear Index
(mN*m2/g)
Tensile Index
(N*m/g)
Burst Index
(kPa*m2/g)
1000
66.0
128.5
22.0
99.9
5.42
2000
69.8
138.5
21.1
105.7
6.98
4000
67.8
146.0
20.0
118.0
7.51
6000
68.6
152.3
18.1
119.5
7.72
(Pulp ID)
40-PSAQ2
1000
64.5
127.3
18.6
101.8
5.42
2000
63.2
137.1
19.5
121.3
6.17
4000
68.2
146.7
16.7
117.6
7.00
6000
63.6
151.4
14.7
128.5
7.72
(Pulp ID)
40-PSAQ6
1000
70.2
130.3
20.2
92.7
5.57
2000
70.1
137.9
16.8
102.2
6.65
4000
62.7
143.2
16.6
111.6
7.24
6000
68.0
151.4
16.8
104.7
7.74
(Pulp ID)
40-PSAQ9 (LIA)
1000
69.5
127.5
21.7
101.5
5.55
2000
67.9
132.5
18.6
121.5
6.36
4000
69.3
146.7
16.7
113.3
7.42
6000
70.3
152.0
16.3
117.4
8.21
(Pulp ID)
40-PSAQ12 (MIA)
1000
67.7
131.7
18.6
97.1
5.75
2000
70.2
139.3
17.7
114.1
6.59
4000
61.4
144.7
15.5
128.2
7.39
6000
68.7
155.4
15.6
116.6
7.85
(Pulp ID)
40-PSAQ15 (HIA)
1000
64.6
129.8
19.4
109.2
5.62
2000
68.3
140.1
16.7
122.7
6.69
4000
67.8
148.3
16.4
122.3
7.51
6000
62.3
150.4
16.0
134.5
7.60
As shown in Figure 14, the MCC reference pulp has greater values of tear index at similar levels of tensile
index. Also noticeable in the figure is the scattering of the data for all five PS pulps. The intent of this graphical
display is to show that there are no distinct tendencies in the investigated tensile and tear strengths among the
PS pulps. The 3% PS low initial alkali pulp returned the highest tear index value, while the 3% PS high initial
alkali pulp produced the greatest level of tensile index. These observations also hold true for the burst index
level as displayed in Figure 15, where the MCC reference pulp again has an advantage over the PS pulps at
similar levels of tear index. Overall, the results from the evaluation of pulp strength properties compared well
to literature data.
75
The lack of clear trends within the investigated PS pulp strength properties possibly indicate that there is no
great difference between Kraft pulps prepared with increasing levels of PS addition. It also makes evident the
opportunities that exist to optimize this process for improved yield and physical properties.
Tensile Index (N*m/g)
150.0
MCC 40% S
1% PS, 40% S
2% PS, 40% S
125.0
3% PS, 40% S LIA
3% PS, 40% S MIA
3% PS, 40% S HIA
100.0
75.0
12.0
Figure 7.14
14.0
16.0
18.0
20.0
Tear Index (mN*m2/g)
22.0
24.0
26.0
Tensile index versus tear index for 40% sulfidity cooks
Burst Index (kPa*m2/g)
9.00
8.00
MCC 40% S
1% PS, 40% S
7.00
2% PS, 40% S
6.00
3% PS, 40% S LIA
3% PS, 40% S MIA
5.00
3% PS, 40% S HIA
4.00
12.0
14.0
16.0
18.0
20.0
22.0
24.0
26.0
2
Tear Index (mN*m /g)
Figure 7.15
Burst index versus tear index for 40% sulfidity cooks
CONCLUSIONS
When pulping at 25% sulfidity, the 1 and 2% PS procedures produced pulps of about 2% higher total yield and
comparable viscosities compared to the MCC baseline at similar kappa. Relative the MCC procedure, the 1%
PS procedure returned somewhat lower kappa pulps at similar AA charge, while the 2% PS procedure resulted
in somewhat higher pulp kappa. The PS pulps refined faster than the MCC reference pulp. When comparing
the tensile and burst index levels at fixed levels of tear index for the 25% PS pulps to the MCC reference pulp,
there is a distinct tear penalty comparable to literature data.
When pulping at 40% sulfidity, the opportunities for sulfur use, as well as the management of alkali profile,
increase. With higher PS charges in the first stage of the cook, more initial alkali is required to maintain
delignification rates. As a result more attention must be paid to the cook alkali profile. At 40% sulfidity, the 1
76
and 2% PS procedures produced pulps of about similar yield to the PS pulps generated at 25% sulfidity. At
similar kappa the pulp yields were about 1% higher than the reference MCC pulp with comparable viscosities
for both PS charges. Three levels of initial alkali charge were investigated when charging 3% PS, with varying
results on final pulp kappa and total yield. The delignification rate increased with increasing levels of initial
alkali charge (HIA > MIA > LIA). The effect on total pulp yield was also dependent on both system TTA and
initial alkali charge. The LIA and MIA with 3% PS resulted in higher pulp yields relative to the MCC reference
at similar kappa. Of the three investigated levels of initial alkali charge the medium initial alkali procedure
returned the best results relative total pulp yield. It resulted in pulps of about 3% greater total yield with
slightly lower viscosities relative the MCC baseline pulp of similar kappa. The high initial alkali procedure
resulted in the lowest total yield. There clearly exists an optimum condition for initial alkali charge, where too
great or too low of an initial hydroxide concentration negatively affects the pulp yield. The PS pulps refined
faster than the MCC reference pulp with the 3% PS pulps producing the greatest improvement. When
comparing the tensile and burst index levels at fixed levels of tear index for the 40% PS pulps to the MCC
reference pulp, there is a small tear penalty, comparable to literature data.
The obtained results indicate that the investigated PS procedure can be used to increase the Kraft pulp yield by
about 1 percent per percent PS added on wood, while maintaining the pulp viscosity. Although high levels of
tensile and burst index are attainable, there is a distinct tear penalty comparable to literature data. The effect of
the alkali profile used in a PSAQ cook will significantly affect the resulting pulp yield. More work is required
to optimize the effect of alkali profile on the pulping procedure relative pulp yield and properties. The
flexibility to optimize the alkali profile and PS use will only be possible in combination with BLG.
FUTURE WORK
Future work will be directed towards evaluating feasible pulping liquors as they may be envisioned through
chemical recovery operations utilizing BLG. Parameters such as system active alkali charge and profile,
sulfidity, and PS concentrations and their application in laboratory pulping experiments should be further
optimized.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
Vennemark, E., Svensk Papperstidning, 67(5): 157 (1964)
Landmark, P.A., Kleppe, P.J., Johnsen, K., TAPPI J. 58(8):56 (1965)
Jiang, J.E., TAPPI J. 77(2): 120 (1994)
Jiang, J.E., TAPPI J. 78(2): 126 (1995)
Griffin, C.W., Kumar, K.R., Gratzl, J., Jameel, H., TAPPI Pulping Conference Proceedings, Chicago:
1: 19 (1995)
Li, Z., Li, J., Kubes, G.J., JPPS, 28(7): 234 (2002).
Sanyer, N., Laundrie, J.F., TAPPI J. 47(10): 640 (1965)
Kleppe, P.J., Minja, R.J.A., TAPPI Proc. Breaking the yield barrier symposium, Vol. 1: 113 TAPPI
PRESS, Atlanta, GA (1998)
Luthe, C., Berry, R., Pulp. Pap. Can. 106(3): 27 (2005)
Nishijima, H. et al. Review of PS/AQ Pulping to Date in Japanese Kraft Mills and the Impact on
Productivity, 1995 Pulping Conference Proceedings, TAPPI PRESS, Atlanta, GA, 1995, pp. 31-40.
Yamaguchi, A. Operating Experiences with the MOXY Process and Quinoid Compounds,
Anthraquinone Pulping: Anthology of Published Papers 1977-1996, TAPPI PRESS, Atlanta, GA,
1997, pp. 287-291.
Tench, L., Uloth, V., Dorris, G., Hornsey, D., Munro, F., TAPPI J. 87(10): 120 (1999)
Munro, F., Uloth, V., Tench, L., MacLeod, M., Dorris, G., Pulp. Pap. Can. 103(1): 57 (2002)
Mao, B. F.; Hartler, N.; TAPPI J. 77(11): 149 (1994)
Lindstrom, M and Teder, A., Nordic Pulp Paper Res. J. 10(1): 8 (1995)
Gustafsson, R., Freysoldt, J., Teder, A., Paperi Puu 86(3): 169 (2004)
Brannvall, E., Gustafsson, R., Teder, A., Nordic Pulp Paper Res. J. 1(4): 436 (2003)
Olm, L.; Tormund, D., Bernor Gidert E., TAPPI Proc. Breaking the yield barrier symposium, Vol 1: 69
TAPPI PRESS, Atlanta, GA (1998)
77
19.
20.
21.
22.
23.
MacLeod, M., Radiotis, T., Uloth, V., Munro, F., Tench, L., TAPPI J. [New Series] 1(8): 3 (2002)
Hakanen, A., Teder, A., TAPPI J. 80(7): 189 (1965)
Olm, L.; Tormund, D., Nordic Pulp Paper Res. J. 19(1): 6 (2004)
Johansson, B., Mjoberg, J., Sandstrom, P., Teder, A., Svensk Papperstid. 87(10): 30 (1984).
Lindstrom, M., Naithani, V., Kirkman, A., Jameel, H., Effects on pulp yield and properties using
modified pulping procedures involving sulfur profiling and green liquor pretreatment Presented at
TAPPI Fall Technical Conference, Atlanta, GA, (November, 2004)
78
8
THE EFFECT OF INTEGRATING POLYSULFIDE PULPING AND BLACK LIQUOR
GASIFICATION ON PULP YIELD AND DELIGNIFICATION
Lindstrom, M., Naithani, V., Kirkman, A., Jameel, H., The Effect of Integrating Polysulfide
Pulping and Black Liquor Gasification on Pulp Yield and Delignification , Proceedings TAPPI
2006 Engineering, Pulping and Environmental Conference, Atlanta, GA, (2006).
79
THE EFFECT OF INTEGRATING POLYSULFIDE PULPING AND BLACK
LIQUOR GASIFICATION ON PULP YIELD AND DELIGNIFICATION
Mathias Lindstrm,Ved Naithani, Hasan Jameel, Adrianna Kirkman
Department of Wood and Paper Science
North Carolina State University
Raleigh, NC 27695-8005
USA
ABSTRACT
Kraft recovery based around black liquor gasification would present several opportunities and benefits with
regard to process operation and economics. Gasification of black liquor would split sodium and sulfur into two
separate fractions. The separation of these chemicals would enable the application of modified pulping
technologies to increase the pulp yield or extend delignification. The experimental objective of this work was
to explore and compare MCC and polysulfide-anthraquinone cooks with different alkali profiles, to illustrate the
potential effects of polysulfide and anthraquinone addition on delignification and total pulp yield. Laboratory
pulps were prepared using 20.5% active alkali at sulfidity levels of 25 and 40%. When comparing the
polysulfide-anthraquinone cooks to cooks using MCC with anthraquinone, yield benefits in the range of 1-2%
were obtained at 25% sulfidity using a polysulfide charge of 2% on OD wood and 3-3.5% at 40% sulfidity
using a 3% polysulfide charge on OD wood. By controlling active alkali profiles under the fixed pulping
protocol used throughout the project, it was possible to achieve comparable or improved delignification rates
using polysulfide, while maintaining higher pulp yield. Generated residual effective alkali profiles illustrate the
difference between MCC and polysulfide pulping, and indicate the importance of alkali control and its effects
on delignification and pulp yield. A greater benefit from PSAQ pulping can be realized at higher levels of
sulfidity, where it is possible to charge a higher concentration of PS and the effects of alkali profiling can be
used to optimize both delignification rate and resulting pulp yield. The most advantageous results, as compared
to the MCC reference, were obtained using the PSAQ procedure at 40% sulfidity with 3% PS on wood and an
alkali profile of (45/31/24). These values represent the percentages of the total available alkali charged to (stage
1/ stage 2 / stage 3) in the cook.
INTRODUCTION
The effect on pulping chemistry of polysulfide (PS), often in conjunction with anthraquinone (AQ) as additives
to the Kraft process, has been explored for some time (1-6). Its effectiveness has been established, and it is
typically reported that each percent of PS added increases the pulp yield by one percent (7,8). However,
efficiently generating high concentrations of PS within the Kraft chemical recovery cycle is difficult. PS in
combination with anthraquinone (AQ) as polysulfide-anthraquinone (PSAQ) offers a process option that can
improve yield and delignification, creating opportunities for increased production or alternatively decreased
operating costs at fixed output. However, the application of PS in pulping operations requires an economical
method of PS generation (9). Moreover, to fully take advantage of benefits resulting from higher PS addition
levels the sulfidity demand will be greater than in a conventional Kraft process (10). Black liquor gasification
(BLG) offers, through the separation of sulfur and sodium, an exciting opportunity for improved PS processes,
where increased PS charges and more efficient application of PS are possible.
The implementation of BLG into Kraft process chemical recovery operations will allow for more or less
unrestricted management of sulfur and sodium on the pulping side. This separation of chemicals can be
exploited to maximize the PS charge in the impregnation stage of the cook, while at the same time enable
optimization of the alkali profile throughout the cook. The degree of sulfur/sodium separation depends on the
BLG process used. Research conducted at STFI-Packforsk investigating PSAQ pulping built around the
Chemrec oxygen-blown high-temperature BLG has shown the potential for significant yield improvements as
compared to conventional Kraft pulping. A concept called ZAP, or Zero effective Alkali in the Pretreatment
stage, has been presented, capitalizing on the optimum conditions for PS pretreatment; low or zero initial alkali,
appropriate time and temperature, and a low pH at the end of the ZAP stage (11). Resent findings based on
pulping liquors generated from a simulated Chemrec BLG recovery system indicated an attainable PS charge of
80
a 1.9% on wood with a yield improvement of about 2% pts. as compared to conventional Kraft. By a 2.6% on
wood addition of sulfuric acid, to lower the final pH of the ZAP stage, a yield benefit of 4.5% pts was achieved.
This value was comparable to the ideal ZAP process values (12). A different procedure combining PS pulping
with alkali profiling called hyperalkaline polysulfide pulping has also been reported (13). The process utilizes
two pretreatment stages followed by a cooking stage. In the first stage, alkali is charge to the wood at elevated
concentrations, neutralizing the acids formed during the temperature elevation. PS is then charged in the second
stage, followed by the cooking stage. The process resulted in a higher delignification rate, increased pulp
viscosities and yield improvements of 1.5% as compared modified pulping without PS.
The research presented in this paper explores the possibilities of ZAP type PS pulping and the opportunities for
alkali profiling enabled by a recovery system based around the MTCI low temperature BLG. The MTCI
process achieves near complete separation of sulfur and sodium, i.e. allowing for ideal ZAP conditions and the
possibility to optimize the alkali profile during the cook.
EXPERIMENTAL
The experimental pulping procedure used and the methods for PS liquor generation have previously been
outlined in detail (10). Mixed southern softwood chips were used throughout the project. The chips were
screened collecting the four to ten mm fraction and stored under refrigeration. Prior to pulping, the chips were
soaked in water overnight, drained and pulped in an M&K batch digester. All cooks were performed using the
simulated MCC protocol described previously, which involves three stages and two pulping liquor additions at
the end of the first and second stage. The general parameters for the MCC protocol, outlined in Table 1, were
designed to generate a circa 30 kappa pulp cooked to 1800 h factor. In this conventional MCC procedure 65%
of the active alkali (AA) was added to the first stage, followed by liquor additions containing 20% of the AA at
the end of stage 1 and the final 15% of the AA at the end of stage 2. The ratio of (65/20/15) for (stage 1/ stage
2/stage 3) is the standard liquor splits used in previous work. It was decided to maintain the same relative ratio
of alkali between stage 2 and stage 3, (20/15 = 1.3 = 57/43), throughout the cooking sequence. A simulated
carbonate dead load of 15% of the system total titratable alkali (TTA) was used throughout the cooks unless
otherwise noted.
Table 8.1
Parameters for MCC protocol
Cumulative % TTA
L/W
Stage Temperature (C)
Time at Temperature (min)
Stage I
65
3.5
120
15
Stage II
85
4.1
164
~ 70
Stage III
100
4.5
164
120
The experimental objective was to explore MCC type Kraft cooks and PSAQ cooks with different alkali
profiles, to illustrate the potential effects of PS and AQ addition on delignification and total pulp yield to an
MCC process using BLG in the chemical recovery. Cooks were completed using 20.5% AA at sulfidity levels
(S) of 25% and 40%. 0.1% AQ was added to both the MCC and PS cooks. The total AA available under the
stipulated parameters was then split between the three stages. Table 2 outlines the performed cooks and the
alkali profiles, which are given as ratios of 100 split to each stage (Stage 1/ Stage 2/ Stage 3). The PSAQ cooks
at 25% sulfidity had a PS charge of 2% on OD wood; the 40% sulfidity cooks had a 3% PS charge. As shown
in Table 2, the three levels of alkali charged in stage 1, initial alkali, of the MCC procedure match the middle
three levels of initial alkali for the PSAQ procedure. In addition, a zero initial alkali charge (ZAP-type cook),
and a very high initial alkali was explored for PSAQ.
81
Table 8.2
Outline of cooks performed using 25% and 40% sulfidity
COOK
% AA on
OD wood
Sulfidity
(%)
% PS on
OD wood
% AQ on
OD wood
Alkali Profile
MCC1
20.5
25 / 40
0/0
0.1
(20/46/34)
MCC2
20.5
25 / 40
0/0
0.1
(45/31/24)
MCC3
20.5
25 / 40
0/0
0.1
(65/20/15)
PSAQ1
20.5
25 / 40
2/3
0.1
(0/54/46)
PSAQ2
20.5
25 / 40
2/3
0.1
(20/46/34)
PSAQ3
20.5
25 / 40
2/3
0.1
(45/31/24)
PSAQ4
20.5
25 / 40
2/3
0.1
(65/20/15)
PSAQ5
20.5
25 / 40
2/3
0.1
(80/11/9)
Following this first sequence of cooks the effect of the alkali profile for ZAP-type cooks was also explored as
outlined in Table 3. The first cook in the sequence is taken from Table 2 (PSAQ1). In the two additional cooks
the amount of AA charged to stage 2 was increased to improve delignification rate.
Table 8.3
Outline of cooks performed at zero initial alkali exploring the effect of alkali profiling
COOK
% AA on
OD wood
Sulfidity
(%)
% PS on
OD wood
% AQ on
OD wood
Alkali Profile
*PSAQ1
20.5
25 / 40
2/3
0.1
(0/54/46)
PSAQ6
20.5
25 / 40
2/3
0.1
(0/65/35)
PSAQ7
20.5
25 / 40
2/3
0.1
(0/75/25)
RESULTS AND DISCUSSION
The obtained results from the cooks outlined in Table 2 are displayed in Figures 1 through 4. The first two
figures show the data generated from the cooks performed at 25% S, the following two figures the data from the
40% S.
Figure 1 describes the pulp kappa numbers generated at 25% S using different alkali profiles. The obtained
pulp kappa numbers for the MCC pulps using different alkali profiles were very similar, ranging from 27.8 to
29.3. The effect from alkali profiling was much more pronounced for the 25% S PSAQ pulps, where the zero
initial alkali cook resulted in a pulp kappa of 39.0. As the alkali charge in the initial stage was increased, the
pulp kappa decreased significantly. At a charge of 45% of the initial alkali the resulting kappa from PSAQ is
similar to the MCC and, at higher initial alkali charge PSAQ returned lower kappa values. Thus, when
comparing the kappa values from the pulping experiments performed at 25% sulfidity, the results show that
relative to the MCC reference it is possible to achieve similar to lower kappa number using PSAQ.
82
2% PS at 25% S
40
MCC Reference
Kappa
35
30
25
20
0
10
20
30
40
50
60
70
80
90
Initial Alkali Charge
(% of total available alkali)
Figure 8.1
Kappa number versus initial alkali charge for 25% sulfidity pulps
Normalized Total Yield (%)
The obtained total pulp yields were normalized to kappa 30 using the relationship 0.15 % pts. total pulp
yield / kappa unit different from 30 (14). Figure 2 displays the normalized total pulp yields obtained from 25%
S cooks using different levels of initial alkali. As shown the PSAQ procedure resulted in higher pulp yields
than the MCC reference pulp.
Figure 8.2
Normalized Yield (K=30)
47
MCC Reference
46
45
44
43
0
10
20
30
40
50
60
70
80
90
Initial Alkali charge
(% of total available alkali)
Normalized total yield versus initial alkali charge for 25% sulfidity pulps
At zero initial alkali charge the PSAQ normalized yield was 46.7% compared to 44.3% for the conventional
MCC pulp using 65% of the initial alkali in the first stage, indicating a potential yield benefit of 2.4% pts. As
the amount of initial alkali charged in the PSAQ procedure was increased from 0% to 80%, the obtained yield
decreased by about 2% pts. while the kappa decreased by about 15. The effect from increasing levels of initial
alkali on decreasing pulp yields is similar to values reported by Olm (11,12).
The pulps prepared at 40% sulfidity followed similar trends, but the differences in PSAQ delignification rates
and pulp yields were more pronounced.
83
3% PS at 40%S
MCC Reference
Kappa
50
40
30
20
0
10
20
30
40
50
60
70
80
90
Initial Alkali Charge
(% of total available alkali)
Figure 8.3
Kappa number versus initial alkali charge for 40% sulfidity pulps
As the initial alkali charge was increased from zero to 80% of the total available alkali, the obtained kappa
values for the PSAQ procedure decreased from 50.4 to 26.1. Meanwhile the total yield, as above normalized to
kappa 30, decreased from 48.5 to 47.7. The yield benefit from PSAQ pulping at 40% sulfidity was more
pronounced across the investigated range of initial alkali and when comparing zero initial alkali PSAQ to the
standard MCC reference using 65% of the available alkali in the first stage, the yield benefit is about 3% pts.
Normalized Total Yield (%)
Normalized Yield (K=30)
Figure 8.4
MCC Reference
49
48
47
46
45
44
0
10
20
30
40
50
60
70
80
90
Initial Alkali Charge
(% of total available alkali)
Normalized total yield versus initial alkali charge for 40% sulfidity pulps
The delignification rates for the lower initial alkali cooks (0 and 20% of TTA) were not as high as those of the
MCC reference cooks. Additional cooks using zero initial alkali charge with different alkali profiles in the
subsequent stages were performed to improve the overall delignification rate while maintaining the zero initial
alkali condition in the first stage. Thus, the ratio of alkali added between the second and third stages was, as
outlined in Table 3, varied from (0/54/46) to (0/75/25). This approach was investigated at both 25 and 40%
sulfidity and the obtained results are displayed in Figure 5. As shown the 25% S PSAQ kappa numbers were
decreased from about 39 to 35 accompanied by a loss in total yield of 1.3% pts. The corresponding values for
the 40% sulfidity pulps were a decrease in kappa from 50 to 35 with about 2% pts. loss in yield.
84
Pulp Kappa
Kappa number and
Normalized Total Yield (%)
55
Normalized Total Yield
50
M CC 40%
S
M C C 2 5% S
45
40
35
30
25% S
(0/54/46)
Figure 8.5
25% S
(0/65/35)
25% S
(0/75/25)
40% S
(0/54/46)
40% S
(0/65/35)
40% S
(0/75/25)
Comparing the effect of alkali profiling on delignification rate and pulp yield for zero
initial alkali PSAQ cooks (25% and 40% sulfidity)
Based on these results, using ZAP type cooking it is possible to decrease the obtained pulp kappa by adding
more of the available alkali to the second stage of the cook. However, as more alkali was added to the earlier
parts of the cook, the obtained pulp yield decreased. The pulp yield values in Figure 5 represent the total pulp
yield for each cook normalized to kappa 30. The corresponding value for the conventional MCC reference cook
using an alkali profile of (65/20/15) was 44.3% using 25% sulfidity (blue line) and 45.9 using 40% sulfidity
(red line). In comparison the normalized pulp yields for PSAQ at highest delignification rate (0/75/25) were
45.4 using 25% sulfidity and 46.8% using 40% sulfidity. Thus, when using zero initial alkali pulping, the
delignification rate can be improved by varying the alkali profile of the cook, however, this also results in a
decreasing total pulp yield.
To illustrate the effects of alkali profiling on pulping liquor residual effective alkali, samples were obtained
from the digester during the course of the cook. Figures 6 through 8 compare the residual effective alkali for
the cooks performed at 25% sulfidity. Figures 9 through 11 compare the residual effective alkali for the cooks
performed at 40% sulfidity.
85
MCC-25%S (20/46/34)
MCC-25%S (45/31/24)
REA (gpl as Na 2O)
25
MCC-25%S (65/20/15)
20
15
10
5
0
0
50
100
150
200
250
Time (min)
Figure 8.6
The effect of alkali profiling on residual effective alkali for MCC at 25% S
As shown in Figure 6, for MCC pulping at 25% sulfidity the alkali profile charged to the three stages as
45/31/24 gives the most level or smooth residual effective alkali profile. Although the obtained kappa and total
yield are similar for the three different profiles, there is a large difference in final REA values.
PSAQ-25%S (0/57/43)
PSAQ-25%S (20/46/34)
REA (gpl as Na 2O)
25
PSAQ-25%S (45/31/24)
PSAQ-25%S (65/20/15)
20
PSAQ-25%S (80/11/9)
15
10
5
0
0
50
100
150
200
250
Time (min)
Figure 8.7
The effect of alkali profiling on residual effective alkali for PSAQ at 25% S
Similarly for the PSAQ cook REA profiles shown in Figure 7, the 45/31/24 profile gives the most level alkali
profile and resulted in the median values for kappa and total yield returns. There is a small difference in final
REA values between the different profiles.
86
PSAQ-25%S (0/57/43)
REA (gpl as Na 2O)
15
PSAQ-25%S (0/65/35)
PSAQ-25%S (0/75/25)
12
9
6
3
0
0
50
100
150
200
250
Time (min)
Figure 8.8
The effect of alkali profiling on REA using for PSAQ 25% S at zero initial alkali charge
As shown in Figure 8, when comparing the REA profiles for the three PSAQ cooks at 25% sulfidity using zero
initial alkali the trends are similar but the final REA values vary widely, as well as the obtained kappa and total
yield values.
Figure 9 outlines the REA profiles generated from MCC pulping at 40% sulfidity. The obtained values for the
MCC cooks at 40% sulfidity are somewhat similar to those at 25% sulfidity, in that the 45/31/24 profile resulted
in the most even curve and the lowest final REA.
MCC-40%S (20/46/34)
MCC-40%S (45/31/24)
REA (gpl as Na 2O)
20
MCC-40%S (65/20/15)
15
10
5
0
0
50
100
150
200
250
Time (min)
Figure 8.9
The effect of alkali profiling on residual effective alkali for MCC at 40% S
As shown in Figure 10, when pulping at 40 sulfidity using PSAQ, the 0 and 20% initial alkali charge returned
very low REA values during the first part of the cook, while the final REA values were similar to those obtained
through the other initial alkali profiles.
87
PSAQ-40%S (0/57/43)
PSAQ-40%S (20/46/34)
REA (gpl as Na 2O)
20
PSAQ-40%S (45/31/24)
PSAQ-40%S (65/20/15)
15
PSAQ-40%S (80/11/9)
10
5
0
0
50
100
150
200
250
Time (min)
Figure 8.10
The effect of alkali profiling on residual effective alkali for PSAQ at 40% S
As shown in Figure 11, when comparing the different zero initial alkali PSAQ cooks at 40% sulfidity, the REA
values during the cook follow a similar development and the final REA values are very close to one another.
REA (gpl as Na 2O)
15
PSAQ-40%S (0/57/43)
PSAQ-40%S (0/65/35)
12
PSAQ-40%S (0/75/25)
9
6
3
0
0
50
100
150
200
250
Time (min)
Figure 8.11
The effect of alkali profiling on REA using for PSAQ 40% S at zero initial alkali charge
As shown in the preceding figures, the effect of AA profiling has a significant impact on the REA levels in the
pulping liquor during the course of the cook. When relating these findings to the effects in delignification rate
and total pulp yield, the differences between MCC with AQ and PSAQ are significant. In MCC pulping, the
REA profiles explored have little effect on the delignification rate, and some effect on the total pulp yield. The
variation in these values is around 1 unit of measurement for both the 25% S and 40% S cooks. The effect is
much greater in the PSAQ procedure, where the REA varies more greatly. In the pulping procedure used in this
project OH- is consumed as the PS begins to dissociate during the first stage, and the major part when the
temperature is increased after the first stage. Therefore, controlling the REA profile during the cook will be
very important to achieve the necessary delignification rate while maintaining the yield benefit sought from
PSAQ pulping.
CONCLUSIONS
The objective of the work described in this paper was to investigate the effect of alkali profiling on yield and
delignification rate using polysulfide-anthraquinone (PSAQ) pulping. The possibility to implement the pulping
88
procedures explored in this work may require separation of sulfur and sodium in the chemical recovery cycle, as
enabled by BLG.
MCC pulping at 25% and 40% sulfidity
A baseline for comparisons was generated using a simulated MCC procedure. Three different alkali profiles
were explored where the initial alkali was increased from 20% to 65% of the total available alkali. The
resulting alkali profiles are denoted, [20/46/34], [45/31/24] and [65/20/15]. When pulping at 25% sulfidity pulp
kappa numbers varied little, from 27.8 to 29.3. The obtained total pulp yields, normalized to kappa 30, showed
only a small variation from 44.3% to 45.3%. All pulp yields were for purposes of comparison normalized to
kappa 30. The results were similar for the cooks at 40% sulfidity, where kappa varied from 27.8 to 28.8 and the
normalized total pulp yield varied from 45.1 to 45.9. There was no apparent trend in pulp kappa or pulp yield
resulting from the MCC cooks at either sulfidity level.
PSAQ pulping was explored at 25% and 40% sulfidity using five different alkali profiles. The initial alkali was
varied from zero to 80% of the available alkali. The resulting alkali profiles are denoted, [0/54/46], [20/46/34],
[45/31/24], [65/20/15] and [80/11/9], where three profiles were identical to those of the MCC cooks. The zero
initial alkali cook is similar to the ZAP procedure and the 80% similar to the hyperalkaline procedure.
PSAQ pulping at 25% sulfidity with 2% PS charge
The ZAP-type cook had the highest normalized yield at 46.7% as well as the highest kappa, 39.0. As the initial
alkali charge was increased from 0 to 80%, the kappa was lowered from 39.0 to 25.3, while the normalized pulp
yield decreased from 46.7% to 44.7%. When varying the alkali profile for the zero initial alkali cooks it was
possible to decrease kappa to levels comparable to the MCC pulps, but the associated decrease in yield was
significant returning little overall benefits from this approach. When comparing the PSAQ results to the 25%
MCC pulps, the greatest benefits with regard to both kappa and total pulp yield were obtained for the PSAQ
pulps using 20 and 45% initial alkali. For these pulps the obtained kappa was similar to that of the MCC pulp
while maintaining a yield benefit of about 1%.
PSAQ pulping at 40% sulfidity with 3% PS charge
When using PSAQ pulping at 40% sulfidity, the ZAP-type cook again had the highest normalized yield at 48.5,
and a very high kappa, 50.4. When increasing the initial alkali charge, the delignification rate was improved
significantly and the kappa decreased from 50.4 to 26.1. The normalized pulp yield also decreased, but only
from 48.5 to 47.7. When varying the alkali profile for the ZAP-type cook kappa was decreased from 50.4 to
34.7 while the yield decreased from 48.5 to 46.8. When comparing PSAQ pulping at 40% sulfidity to the MCC
reference, a significant improvement in yield is possible while maintaining a similar delignification rate. The
yield benefit from PSAQ pulping at 40% using a PS charge of 3% on wood was 3-3.5%. The approach of
varying the alkali profile for the ZAP-type cook was also more successful than at 25% sulfidity. The PSAQ
procedure using initial alkali levels of 45% provided the greatest benefit in both delignification rate and pulp
yield as compared to the MCC reference.
Overall, the relative benefits from PSAQ pulping in both delignification rate and yield was greater using 3% PS
on wood at 40% sulfidity than 2%PS on wood at 25% sulfidity. The effects from alkali profiling were also
more pronounced at 40% sulfidity. In conclusion, based on the pulping procedures used in this research, a
greater benefit from PSAQ pulping can be realized at higher levels of sulfidity, where it is possible to charge a
higher concentration of PS and the effects of alkali profiling can be used to optimize both delignification rate
and resulting pulp yield. The most advantageous results, as compared to the MCC reference, were obtained
using the PSAQ procedure at 40% sulfidity with 3% PS on wood and a (45/31/24) alkali profile.
FUTURE WORK
Future work will entail bleaching and strength property studies of selected pulps prepared in this project.
REFERENCES
89
1. Li, Z.; Li, J.; Kubes, G.J. Kinetics of Delignification and Cellulose Degradation During Kraft
Pulping with Polysulphide and Anthraquinone; JPPS. 2002, 28(7): 234-239.
2. Griffin, C.W.; Kumar, K.R.; Gratzl, J.; Jameel, H. Effects of Adding Anthraquinone and Polysulfide
to the Modified Continuous Cooking (MCC) Process; Proc. 1995 TAPPI Pulping Conference; TAPPI
Press: Atlanta, GA, 1995; 19-30.
3. Jiang, J.E. Extended Delignification of Southern Pine [Pinus spp.] with Anthraquinone and
Polysulfide; TAPPI J. 1995, 78(2): 126-132.
4. Jiang, J.E. Extended Modified Cooking of Southern Pine [Pinus] with Polysulfide: Effects on Pulp
Yield and Physical Properties; TAPPI J. 1994, 77(2): 120-124.
5. Landmark, P.A.; Kleppe, P.J.; Johnsen, K. Pulp Yield Increasing Process in Polysulfide Kraft
Cooks; Tappi J. 1965, 58(8): 56.
6. Vennemark, E. Some Ideas on Polysulfide Cooking; Svensk Papperstidning. 1964, 67(5): 157.
7. Kleppe, P.J.; Minja, R.J.A. The Possibilities to Apply Polysulfide-AQ in Kraft Mills; Proc. Breaking
the yield barrier symposium; TAPPI Press: Atlanta, GA, 1998, Vol. 1, pp 113.
8. Sanyer, N.; Laundrie, J.F. Factor Affecting Yield Increase and Fiber Quality in Polysulfide Pulping
of Loblolly Pine, Other Softwoods, and Red Oak; Tappi J. 1964, 47(10): 640.
9. Tench, L., Wearing, J., Buchner, W. PS/AQ Can Do More Than Just Increase Digester Yields; Proc.
Growing yield from the ground up, TAPPI Press: Atlanta, GA, 2006.
10. Lindstrom, M.; Naithani, V.; Kirkman, A.; Jameel, H. The Effect of Integrating Polysulfide Pulping
and Black Liquor Gasification on Pulp Yield and Properties; Proc. 2005 TAPPI Engineering,
Pulping & Environmental Conference; TAPPI Press: Atlanta, GA, 2005.
11. Olm, L.; Tormund, D., Nordic Pulp Paper Res. J. 19(1): 6 (2004)
12. Olm, L.; Tormund, D., ZAP Cooking with New Liquors from a BLG Recovery System; Proc.
Growing yield from the ground up, TAPPI Press: Atlanta, GA, 2006.
13. Brannvall, E.; Gustafsson, R.; Teder, A.; Properties of Hyperalkaline Polysulphide Pulps; Nordic
Pulp Paper Res. J. 2003, 18(4): 436-440.
14. Luthe, C., Berry, R., Pulp. Pap. Can. 106(3): 27 (2005)
90
9
THE DEVELOPMENT AND VALIDATION OF A LOW-TEMPERATURE BLACK LIQUOR
GASIFIER MODEL FOR USE IN WINGEMS
Lindstrom, M., Naithani, V., Kirkman, A., Jameel, H., The Development and Validation of a
Low-temperature Black Liquor Gasifier Model for Use in WinGEMS , Proceedings TAPPI 2005
Engineering, Pulping and Environmental Conference, Philadelphia, PA, (2005).
Also in: TAPPI J. Vol. 5 No. 2 (2006) p.24
91
THE DEVELOPMENT AND VALIDATION OF A LOW-TEMPERATURE BLACK
LIQUOR GASIFIER MODEL FOR USE IN WINGEMS
Mathias Lindstrm ,Ved Naithani, Adrianna Kirkman, Hasan Jameel
Department of Wood and Paper Science
North Carolina State University
Raleigh, NC 27695-8005
USA
ABSTRACT
The continuous efforts made in the area of black liquor gasification (BLG) are bringing this technology closer
to commercial realization and potential wide-spread implementation. One area of research that could assist in
this process is the investigation of the effects of modified pulping technologies on modern pulping operations.
The separation of sodium and sulfur could be utilized in processes like polysulfide-anthraquinone, alkaline
sulfite-anthraquinone or mini-sulfide sulfite-anthraquinone to dramatically improve pulp yield and overall
process economics. In order to simulate low temperature black liquor gasification in conjunction with these
process modifications and the resulting effects on full mill operations in WinGEMS, the development of a low
temperature BLG model was required. The predictive capability of the developed BLG WinGEMS block was
validated by comparison to process data. The process model generates output and back-calculated input streams
with good agreement to values provided by the manufacturer. Several obstacles were overcome in the creation
of the block, including requirements on user defined stream components. The BLG block is currently being
used in case studies to investigate the effects of varying process carbon conversion ratios. These efforts are
seen as further validation of the predictive capabilities of the model, which is envisioned to be used in full mill
simulations to explore the implications of modified pulping technologies on current pulping operations.
INTRODUCTION
The implementation of black liquor gasification into conventional chemical recovery operations presents many
benefits and opportunities relative to current chemical recovery technologies. The separation of sodium and
sulfur into different streams will allow for the regeneration of a variety of pulping liquors that could be used to
maximize pulp yield, and even design pulp properties according to market demand. In addition to the potential
benefits applicable on the pulping side, BLG would also increase the efficiency of combined heat and power
generation in the mill. The utilization of combined-cycle power generation will create energy management
alternatives driven by the relative costs of steam production and electricity and their day to day impact on
operating margins. The overall net result of BLG implementation could mean improved pulp mill operations
and significant production cost savings (1-3).
With efforts focused on two different process concepts, a high-temperature, pressurized partial-oxidation process
(Chemrec) and a low-temperature indirectly heated steam-reforming process (MTCI), the development of BLG
technologies is approaching commercial realization. Three mill start-ups in North America are currently in
operation, one utilizing the Chemrec process (Weyerhaeuser New Bern mill, NC) and two installations (G-P
Big Island, Va. and Norampac, Trenton, Ont.) utilizing the MTCI process. Notwithstanding the inherent
benefits of BLG relative to chemical recovery and power generation, wide-scale implementation of these
processes faces industry reluctance until successful commercial operations have been proven (4).
The Tomlinson Recovery Boiler, in spite of being a mature proven technology, has the disadvantages of low
thermal efficiency, low power to heat ratio and the risk of smelt/water explosions. Gasification technology
addresses some of these concerns. Moreover, it allows for the separation of sodium and sulfur which in turn
can be exploited to optimize pulping operations using modified pulping technologies such as those listed below.
Split Sulfidity Pulping
Polysulfide Pulping
Alkaline Sulfite Pulping
Alkaline Sulfite- AQ (AS-AQ)
92
Mini-Sulfite Sulfide AQ Process (MSSAQ)
The positive and potentially substantial impact of these technologies on the pulping process could influence
decision makers to more actively pursue a paradigm shift in chemical recovery operations as exemplified by
BLG.
A good approach to evaluate modifications to any process is computer simulation. In order to explore the
effects on the overall process operations by the implementation of BLG, alone or in conjunction with any of the
above modified pulping technologies, a full mill model must be built. One of the standard simulation tools for
the pulp and paper industry is WinGEMS. However, there is currently no BLG block available within the
WinGEMS software, thus rendering the simulation of BLG implementation complex at best (3). A black liquor
gasifier model in the form of a WinGEMS block would be highly desirable. Work utilizing models of black
liquor gasifiers in other software packages have been presented, focusing on the cost benefits of high- and lowtemperature BLG operations in conjunction with the combined-cycle power generation (1,2), as well as fuel and
energy production (5,6). The work presented here describes the development of a model for a low-temperature
MTCI steam reformer in WinGEMS. This WinGEMS block will be employed in subsequent work to
investigate BLG operating parameters and conditions in full pulp mill simulations exploring the effects of
process modifications, while predicting the BLG process output streams.
THE BLG MODEL
The MTCI low-temperature steam reformer
A simplified schematic of the MTCI system is shown in Figure 1. It consists of a fluidized bed reactor that is
indirectly heated by multiple resonance tubes of one or more pulse combustion modules. Feedstock such as
spent liquor is fed to the reactor, which is fluidized with superheated steam from a waste heat recovery boiler.
The organic material injected into the bed undergoes a rapid sequence of vaporization and pyrolysis reactions.
Higher hydrocarbons released among the pyrolysis products are steam cracked and partially reformed to
produce low molecular weight species. Residual char retained in the bed is more slowly gasified by reaction
with steam. The sulfur and sodium are separated as the sulfur leaves mostly with the gas output stream and the
sodium stays in solid form and exits with the bed solids. Product gases are routed through a cyclone to remove
the bulk of the entrained particulate matter and then quenched and scrubbed in a venturi scrubber. The sulfur
species must be recovered, through scrubbing with green liquor or alternative processes, prior to other unit
operations. A portion of the medium-Btu product gas can be supplied to the pulse combustion modules and
combustion of this gas, or alternatively natural gas, provides the heat necessary for the endothermic reactions in
the gasification process. The inorganic chemical in the feedstock is recovered as the bed solids are dissolved,
possibly in recycled weak wash, and recombined with sulfur recovered from the gas stream to regenerate
pulping liquor used in the mill.
The products of combustion exit from the resonance tubes completely segregated from the reformate product
gases. Hot flue gases from the steam reformer are used to generate steam and to preheat the pulsed heater
combustion air. Excess fuel gas is exported for use in a boiler, gas turbine or fuel cell. The process uses only a
single reactor; it does not require solids recirculation and handling equipment and it can be easily controlled by
varying the gas-firing rate.
93
Product
Gas
Clean
Flue Gas
Feedstock
Product
Gas/Air
Pulsed
Heaters
Bed Solids
Fluidizing
Steam
BLG process model inputs and outputs
Figure 9.1
Schematic of a MTCI Steam Reformer
As shown in Figure 1, the BLG process input and output streams are easily visualized, and can be summarized
as shown in Figure 2. There are four separate input streams and three separate output streams. The input
streams consist of the feedstock stream, in this case concentrated black liquor (BL), the bed fluidizing steam,
the fuel gas combusted in the pulsed heaters and the air stream required for pulsed heater combustion. The third
stream, the fuel gas for combustion in the pulsed heaters, can be natural gas or a recycled fraction of the BLG
product gas. The output streams consist of the BLG bed solids, the product gas, and the flue gas from the
pulsed heaters.
BL
BED SOLIDS
STEAM
FUEL GAS
STEAM
REFORMER
PRODUCT GAS
FLUE GAS
XS AIR
Figure 9.2
Stream structure for BLG model
Although the input and output streams are not complex, the compositions of the BL input stream and the
product gas stream certainly are. The black liquor is comprised of all organic and inorganic material recovered
from the pulping process, constituting a complex matrix of lignin fragments, cellulose, hemi-cellulose and
various ionized inorganic species. The BLG product gas consists of a gas matrix containing all volatile species
generated in the steam reforming reactions, as well as minute amounts of dust particles. The remaining streams
are made up of either a single compound or simple mixtures.
94
BLG process model assumptions and parameters
At the outset of any process model construction, the level of detail in the model must be determined relative to
its intended use and predictive capabilities. The purpose of constructing this BLG model was the desire to
incorporate a BLG block into a full mill WinGEMS simulation, where the interaction of the BLG and other mill
unit operations could be evaluated in terms of process variables of interest to the user. The model developed for
this application bases the material and energy balances around the steam reformer on empirical relationships
rather than first principles. This approach allows for the prediction of model output streams based on the given
BL input stream and process parameters, as well as the back-calculation of the required amounts of bed
fluidizing steam and energy supplied through the pulsed heaters to sustain the endothermic steam reforming
reactions.
To simplify the BLG model and its control a number of parameters were identified and utilized for user input
and reaction constraints internal to the model. These parameters, outlined in Table 1, include process operating
variables such as temperature, reduction ratios and conversion efficiencies similar to those found in the
WinGEMS KFURN block (7). Several of the defined parameters and other considerations made regarding the
BLG model were based on process data provided by MTCI. As a result, the range of operating conditions that
can be used with the model in process simulations is limited by the original data. With increasing amounts of
process data available for various pulping operations, this range will be increased. At present simulations
incorporating the BLG model should be constrained to the default values indicated in the block documentation.
In addition to the user defined parameters outlined in Table 1, available process data was used to determine the
selective generation of gaseous components formed by sulfur and carbon, as well as particulate matter in the
product gas.
Table 9.1
BLG user defined block parameters
User Defined Parameters
Gasifier Operating Temperature
Units
(deg C)
Steam Requirement for bed fluidization
Heat of formation for Dissolved Wood Solids
(mol total H2O /
mol total C)
(kJ/kg)
Dissolved Wood elemental fractions
C
fraction
H
fraction
O
fraction
Sulfur reduction ratio
fraction
Chloride reduction ratio
fraction
Total Carbon conversion ratio
fraction
BLG process model algorithm
The material balances were calculated based on the elemental composition of the black liquor fed to the gasifier,
the fluidizing steam, and the parameters outlined in Table 1. The sequence of calculations is outlined in Table
2.
95
Table 9.2
Calculation
order
Algorithm for BLG material balances
Item
Comment
1
Calculate BL elemental composition (wet)
Organic fraction of C, H, O, based on input parameter
2
Calculate required fluidizing steam
Based in input parameter
3
Calculate Sulfur balance
Based on sulfur reduction ratio and sulfurous gas component split
ratios
4
Calculate Chloride balance
Based on chloride reduction ratio
5
Calculate Sodium balance
6
Calculate Potassium balance
11
Calculate Carbon balance
8
Calculate Oxygen balance
Based on total carbon conversion ratio and parameters controlling
the formation of carbonaceous compounds
9
Calculate Hydrogen balance
10
Calculate Solid stream output
Summation of all solids species
11
Calculate Gas stream output
Summation of all gaseous species
The BLG energy balances were calculated based on the Gullichsen total enthalpy method (8). In simple terms,
the net total enthalpy of a system is computed as the difference between the sum of total enthalpies of the
reactants and the corresponding sum of total enthalpies of the products (Eq. 1). In turn, the total enthalpy of a
substance is calculated as the sum of its isothermal heat of formation, the sensible heat above 25 C, the latent
heat of fusion or evaporation, and the heat of solution and dissolution (Eq. 2).
Equation 1.
Equation 2.
Q = H Reactants H Products
H = H f STP +
T2
Cpdt
+ H agr + H s
298 K
The overall BLG system material balance was calculated using the composition of the BL input stream. The
stream components were used to define compounds present in the BL and compute their total enthalpies.
Subsequent summation with the total enthalpy of the fluidizing steam yielded the total enthalpy of the reactants.
The total enthalpy of the products was computed by summation of all the output components total enthalpies
generated in the bed solids and product gas streams. The difference between these two values was assumed to
represent the amount of energy required to sustain the endothermic steam reforming reactions, and as such, used
to back-calculate the amount of energy required for BLG operation. As shown (Eq. 3), this value minus an
assumed system heat loss based on design parameters was used to determine the amount of fuel gas required for
the pulsed heaters.
Equation 3.
BLG Q Required = H Reactants H Products Q BLG Heat Loss
Thermodynamic properties for the involved species were determined using the JANAF tables (9), the handbook
of Chemistry and Physics (10) and the NIST Chemistry Webbook (11).
Integration of BLG block into WinGEMS
The WinGEMS Block Developers Kit was used to integrate the model into the simulation software. The
appropriate code for the block, including the material and energy balance calculations, was generated in
Compaq Visual FORTRAN. Some hurdles to the successful development of this WinGEMS block, were in
particular, related to the material balance. WinGEMS uses a hardwired variable array for stream components
96
that allows for easy information transfer between blocks. This array is not modifiable, and contains most
stream components that would ever be used by a pulp and paper process engineer. However, the product gas
stream, as an example, created through gasification contains several stream components that are not included in
this variable array. When a user defines a new stream component, he/she has the option of assigning a dummy
variable number to the new stream component. These dummy variables are essentially open slots carried in the
stream component variable array, and can be accessed by any block. This capability was exploited in the code
for the BLG block, where several stream components were designated to specific dummy variables. As a result,
when building a new simulation involving the BLG block, the user defined stream components must be
designated as the correct dummy variable for proper functionality. This and other considerations are addressed
in the block reference file for the BLG WinGEMS block.
The completed BLG block was debugged in a simple WinGEMS simulation shown in Figure 3.
Syn-gas Recycle
(Optional)
5
1
SPLIT
Dust
Recap t ure
6
PC Fuel
2
SPLIT
H2S A mine
Scrub b er
1
21
11
2
1
11
Syn Gas Output
4
24
R EA CT
React 14
1
1
2
M IX
M ix13
1
26
3
25
20
1
14
PC Flu Gas
31
14
H2S Liquor Output
13
1
SOLID PHASE
1
3
2
11
1
2
B LG
B L Gasif ier
2
1
15
GAS PHASE
6
BL Feed
2
5
1
3
SPLIT
Gas Recycle
4
1
2
M IX
M ix12
1
2
4
Steam
12
23
1
SDT
Dissolving
Tank
1
CLF
Green Liq Clar
1
18
2
1
1
9
1
3
2
6
2
12
Green Liquor Output
8
2
SPLIT
Split6
10
Weak Wash Recycle
Dregs Removal
Figure 9.3
WinGEMS simulation using integrated BLG block model
The simulation outlined in Figure 3 was used to predict the BLG process outputs based on the BL feed stream
and selected process parameters outlined in Table 1. The values were compared to process data supplied by
MTCI, and the comparison used as validation of the BLG block as a functional tool for the given conditions.
A black liquor feed of 142 metric tons per hour was fed to the BLG block at 80% solids. The parameters
outlined in Table 1 were set to match data provided by MTCI. The predicted output stream components and the
calculated percent error, as compared to MTCI process data, are shown in Table 3.
97
Table 9.3
Predicted BLG process streams and comparisons to MTCI process data
BLG Output
% difference rel.
MTCI process data
Gas Phase output
Solid components
% difference rel.
MTCI process data
Gas Phase dust output
CH4
0.8000
Na2CO3
0.0240
C2H6
0.0020
Na2SO4
0.0404
C2H4
0.0160
NaCl
0.0164
C3H6
0.0018
K2CO3
0.0001
C3H8
0.0000
CaCO3
0.0000
Phenols
0.0790
C
0.0017
Acetone
0.0243
H2S
0.0340
Na2CO3
0.0044
CH3SH
0.0019
Na2SO4
0.0415
(CH3)2S
0.0080
NaCl
0.0235
(CH3)2S2
0.0020
K2CO3
0.0013
CO
1.2000
CaCO3
0.0000
CO2
1.2000
C
0.0018
H2
0.7600
1.1000
N2
0.0000
Fluidizing Steam Input
Fuel Input for Pulsed
Heaters
O2
0.0000
H2O (v)
0.2480
HCl
0.0137
Solid Phase output
1.3000
As shown in the table, there is very good agreement between the predicted values and the MTCI reference data.
The largest discrepancies are found in the components of CH4, CO and CO2, as well as in the values for the
fluidizing steam and fuel requirements for the pulsed heaters. Work is currently underway to improve these
differences. The overall results indicate that the developed model can accurately predict BLG unit operations
under the given conditions. In addition to continuous updating of the process material and energy balances,
work is in progress to explore the BLG model predictive capabilities under a range of typical process
conditions. For instance, carbon conversion rates are critical in black liquor gasification as they affect the
system efficiency and profitability. Low levels of carbon conversion can lead to tar build up and equipment
fouling, as well as negatively affecting the heat value of the BLG product gas (12). A case study to investigate
the effect on the predicted BLG product gas stream at various levels of carbon conversion is underway.
CONCLUSIONS
The predictive capability of the developed BLG WinGEMS block was validated by comparison to
manufacturers process data. The model generates output and back-calculated input streams with good
agreement to values provided by MTCI for the gasification of black liquor.
Several obstacles were overcome in the creation of the block, including requirements on user defined stream
components. Considerations for the employment of the BLG block in WinGEMS simulations are addressed in
the block reference.
FUTURE WORK
The BLG block is currently being used to investigate the effects of varying carbon conversion ratios. These
efforts are seen as further validation of the predictive capabilities of the model, which is envisioned to be used
98
in full mill simulations to explore the implications of modified pulping technologies on current pulping
operations.
ACKNOWLEDGEMENTS
The authors thank the U.S. Department of Energy for the financial support of this work, and also Pacific
Simulation and MTCI and TRI for their assistance and guidance; in particular Barry Malmberg (Pacific
Simulation), Ravi Chandran and Lee Rockvam (MTCI), and Dan Burciaga and Dave Newport (TRI), who made
this work possible.
REFERENCES
1.
2.
Consonni, S., Larson, E.D., Katofsky, R., Proc. ASME Turbo Expo 7: 1 (2004)
Consonni, S., Larson, E.D., Katofsky, R., Proc. ASME Turbo Expo 7: 15 (2004)
3. Lindstrom, M., Kirkman, A., Jameel, H. et al., Proc. TAPPI Fall Technol. Trade Fair pp. 14171429 (2002)
4. Patrick, K., Pap Age 119(7): 30 (2003)
5. Berglin, N., Lindblom, M., Ekbom, T., Proc. TAPPI Fall Technol. Trade Fair pp. 527-537 (2002)
6. Ekbom, T., Berglin, N., Lindblom, M., Int. Chem. Recovery Conf. Vol. 2: 843 (2004)
7. WinGEMS 5.3 help manual:KFURN block reference Pacific Simulation, Moscow, ID (200x)
8. Gullichsen, J., Proc. IUPAC/EUCEPA Symp. Recovery of Pulping Chemicals (Helsinki) p. 211234 (1969)
9. Malcolm W. Chase, Jr., NIST-JANAF thermochemical tables, 4th ed., American Chemical Society;
Woodbury, N.Y., American Institute of Physics for the National Institute of Standards and
Technology, Washington, D.C. (1998)
10. Lide, D.R., CRC Handbook of Chemistry and Physics, 85th ed., Online Edition,
http://www.hbcpnetbase.com/ (2005)
11. NIST Chemistry Webbook., available online http://webbook.nist.gov/chemistry/ (2005)
12. Sricharoenchaikul, V., Frederick, W.J., Grace, T.M., JPPS, 23(8): J394 (1997)
99
10. WINGEMS SIMULATION OF NET PROCESS VARIABLE OPERATING COSTS RESULTING
FROM BLG INTEGRATION WITH SPLIT SULFIDITY AND POLYSULFIDE PULPING
Unpublished
100
WINGEMS SIMULATION OF NET PROCESS VARIABLE OPERATING COST
RESULTING FROM BLG INTEGRATION WITH SPLIT SULFIDITY AND
POLYSULFIDE PULPING
Mathias Lindstrm,Ved Naithani, Hasan Jameel, Adrianna Kirkman
Department of Wood and Paper Science
North Carolina State University
Raleigh, NC 27695-8005
USA
ABSTRACT
Black liquor gasification enables the separation of sodium and sulfur in pulp mill chemical recovery operations,
thus creating a viable recovery system for pulping processes like polysulfide-anthraquinone, alkaline sulfiteanthraquinone or mini-sulfide sulfite-anthraquinone. These technologies could dramatically improve pulp yield
and overall process economics. A previously developed WinGEMS model for low temperature black liquor
gasification has been used to evaluate the integration of BLG combined with modified pulping technologies into
the kraft process. Case studies were performed in WinGEMS comparing full mill simulations of a kraft process
using modified continuous cooking to BLG processes with split sulfidity and polysulfide-anthraquinone
pulping. The net variable operating costs for each case were determined from a set of cost factors. The
obtained results indicate that implementing BLG with the investigated modified pulping technologies would
lead to increases in the net variable operating cost per oven dry ton pulp produced ranging from 1.8 to 3.9%
compared to the kraft-MCC base case. It was also found that the cost of kiln fuel and the price of power sales to
the grid drive the relative cost performance of the evaluated cases. When comparing the BLG cases to the MCC
reference, the net variable operating cost break even point based on lime kiln fuel cost ranges from $47 to $38
per barrel of kiln fuel depending on the BLG process, which is much lower than currently available prices.
When performing similar comparisons of net variable operating costs based on variable power sales price,
significant cost savings could be realized in all BLG processes at prices for power sold to the grid above 5
/KWh.
INTRODUCTION
The implementation of black liquor gasification into conventional chemical recovery operations presents many
benefits and opportunities relative to current chemical recovery technologies. The separation of sodium and
sulfur into different streams will allow for the regeneration of a variety of pulping liquors that could be used to
maximize pulp yield, and even design pulp properties according to market demand. In addition to the potential
benefits applicable on the pulping side, BLG would also increase the efficiency of combined heat and power
generation in the mill. The utilization of combined-cycle power generation will create energy management
alternatives driven by the relative costs of steam production and electricity and their day to day impact on
operating margins. The overall net result of BLG implementation could mean improved pulp mill operations
and significant production cost savings (1-3).
The Tomlinson Recovery Boiler, in spite of being a mature proven technology, has the disadvantages of low
thermal efficiency, low power to heat ratio and the risk of smelt/water explosions. Gasification technology
addresses some of these concerns. Moreover, it allows for the separation of sodium and sulfur which in turn
can be exploited to optimize pulping operations using modified pulping technologies such as those listed below.
Split Sulfidity Pulping
Polysulfide Pulping
Alkaline Sulfite Pulping
Alkaline Sulfite- AQ (ASAQ)
Mini-Sulfite Sulfide AQ Process (MSSAQ)
101
Modifications to the kraft process, such as split sulfidity and polysulfide pulping would demand the smallest
changes to current prevailing chemical recovery and pulping operations, and are the focus in this effort.
To simulate pulping operations using BLG combined with these processes in WinGEMS, a model for black
liquor gasification had to be developed and integrated into the software. This work has previously been
reported (4). The following is a brief discussion of the assumptions made for the simulated unit operation of the
MTCI low temperature steam reformer and the model functionality.
THE MTCI LOW-TEMPERATURE STEAM REFORMER
A simplified schematic of the MTCI system is shown in Figure 1. It consists of a fluidized bed reactor that is
indirectly heated by multiple resonance tubes of one or more pulse combustion modules. Recovered black
liquor is fluidized with superheated steam and injected into the reactor. Sulfur and sodium are separated as the
majority of the sulfur forms volatile compounds and exits with the gas output stream, whereas the sodium forms
various metal salts, mainly carbonates and exits with removed bed solids.
Product
Gas
Clean
Flue Gas
Feedstock
Product
Gas/Air
Pulsed
Heaters
Bed Solids
Figure 10.1
Fluidizing
Steam
Schematic of a MTCI Steam Reformer
As shown in Figure 1, the BLG process input and output streams are easily visualized, and can be summarized
as shown in Figure 2. There are four separate input streams and three separate output streams. The input
streams consist of the feedstock stream, in this case concentrated black liquor (BL), the bed fluidizing steam,
the fuel gas combusted in the pulsed heaters and the air stream required for pulsed heater combustion. The third
stream, the fuel gas for combustion in the pulsed heaters, can be natural gas or a recycled fraction of the BLG
product gas. The output streams consist of the BLG bed solids, the product gas, and the flue gas from the
pulsed heaters.
102
BL
BED SOLIDS
STEAM
FUEL GAS
STEAM
REFORMER
PRODUCT GAS
FLUE GAS
XS AIR
Figure 10.2
Stream structure for BLG model
Although the input and output streams are not complex, the compositions of the BL input stream and the
product gas stream certainly are. The black liquor is comprised of all organic and inorganic material recovered
from the pulping process, constituting a complex matrix of lignin fragments, cellulose, hemi-cellulose and
various ionized inorganic species. The BLG product gas consists of a gas matrix containing all volatile species
generated in the steam reforming reactions, as well as minute amounts of dust particles. The remaining streams
are made up of either a single compound or simple mixtures.
INTEGRATION OF THE BLG MODEL AND CHEMICAL RECOVERY UNIT OPERATIONS IN
WINGEMS
The developed BLG model was used in a full mill simulation case study comparing a conventional kraft process
to BLG processes with modified pulping operations using split sulfidity and polysulfide-anthraquinone pulping.
The reference kraft process simulation was envisioned as a bleached mixed southern softwood market pulp mill,
producing 1000 oven dry metric tons of brown pulp, followed by an ODEopD bleaching sequence. All process
unit operations were assumed to be conventional, with commonly accepted input parameters and efficiencies.
In the simulated BLG cases, the kraft recovery loop has been modified with a BLG unit replacing the recovery
boiler, and subsequent unit operations for syngas clean-up and pulping liquor generation designed for either
split sulfidity or PSAQ pulping operations. Recovered black liquor from the pulping process was concentrated
in conventional multiple effect evaporators and concentrated to 80% solids. Make-up chemical was added to
the concentrated black liquor as salt cake, prior to gasification. In all BLG cases an amine system is used to
capture the syngas H2S, and the generated gasifier bed solids are processed through conventional green liquor
recausticizing unit operations. In the BLG case with split sulfidity a portion of the generated green liquor is
used to scrub absorb the H2S captured from the syngas, generating high sulfidity green liquor. This liquor is
then causticized in a separate slaker unit prior to being returned to the BLG-SS pretreatment vessel. In the
BLG-PSAQ cases, the captured H2S is put through a Klause process to generate elemental sulfur with the
simultaneous release of the major portion of CO2 that was co-absorbed in the amine scrubber. This elemental
sulfur is then added to a portion of the recausticized (now) white liquor, generating very high sulfidity
polysulfide liquor which is returned to the pretreatment vessel. The clean syngas is combusted in a gas turbine
coupled to a condensing steam turbine to produce the process steam demand while generating power.
WINGEMS CASE STUDIES OF KRAFT-MCC AND BLG WITH SS AND PSAQ PULPING
Four chemical pulping processes were evaluated in separate full mill WinGEMS simulations. Each process
represents an individual case in the study, where the kraft-MCC case was used as a reference for comparisons to
the different BLG cases. The pulp yields for the simulated processes were based on generated laboratory data
(5-7). In all BLG cases, the clean syngas was compressed and combusted in a combined cycle power plant for
the production of process steam and power. The simulated unit operations were in general the same in each of
the evaluated cases, except for the modifications made to those down stream from the BLG unit, involving gas
103
clean-up and the regeneration of pulping liquors for split sulfidity and PSAQ pulping. Table 10.1 shows an
outlined comparison of the major design differences and some important assumptions between the 4 different
cases evaluated in this case study. When estimating the change in process steam demand and power production
from the base case to the BLG cases, the following assumptions were made. The base case reference steam
demand was met by the steam generated from the recovery boiler, and the overall process is a net consumer of
power. In the BLG cases, the difference in steam demand between the base case reference value (~282 mt/hr)
and the steam generated from syngas combustion (~19 mt/hr) was overcome through combustion of additional
biomass in a hog fuel boiler. The hog fuel demand in the BLG cases was set to meet the process steam demand,
resulting in additional power production, above and beyond what was generated in the kraft reference
simulation.
Table 10.1
Input parameters and key assumptions for WinGEMS case study
Kraft Base Case
Liquor Parameters
% EA on OD wood
19
19
19
19
Sulfidity
25
25
25
40
Pretreatment
No
Yes (120 C)
Yes (120 C)
Yes (120 C)
Continuous Digester
Yes (170 C)
same
same
same
45%
47%
47%
48%
8.5% of solids flow same
same
same
same
same
same
Base Case Index = 0 same
same
same
282 mt/hr (Met in RB) mt/hr
282
282 mt/hr
282 mt/hr
Base case Index = 0 BLGCC
BLGCC
BLGCC
Controlled to BC
Controlled to BC
Pulping Operations
Yield Assumptions
Pulping
Rejects and Spills
Bleach Plant
5% of solids flow
Power Demand
Overall Process
Steam Demand
(1000 psig)
Power Generation
KWh/ODtP
Steam Generation
(1000 psig)
Equal to demand
Controlled to BC
The following discussion addresses the generated output report from the WinGEMS case study. The values for
the fiber line calculated from WinGEMS simulation data area shown in Table 10.2. As seen in the table, the
total wood feed is altered from case to case to meet the 1000 ODt brown pulp production target. With
increasing pulp yields resulting from the integration of BLG and modified pulping technologies, the total wood
demand is thus decreased. However, with additional pretreatment and divided sulfur rich and lean streams, the
total digester steam demand, as well as total liquor flow is increased. The values for the bleach plant operations
are very similar, which should be expected since all pulps are assumed to respond to the ODED bleaching
sequence in a similar fashion.
104
Table 10.2
Comparison of fiber line process parameters for each simulated case
Unit
Kraft Basecase
BLG-SS
BLG-PS2%
BLG-PS3%
ODtC/day
ODtP/day
kg/ODtP
2424.12
1000.00
1137.33
2326.50
1000.00
1227.63
2326.57
1000.00
1197.14
2278.10
1000.00
1216.28
Pulping Liquor (synthetic)
Flow
TTA as NaOH
Sulfidity
mt/hr
g/L
%
159.63
151.25
24.97
201.91
151.26
21.70
191.83
171.52
21.12
205.51
179.08
31.95
Bleach Plant (ODED)
BP Steam
OWL Flow
ClO2
NaOH
H2O2
mt/hr
mt/hr
mt/hr
mt/hr
mt/hr
30.47
19.44
0.99
0.07
0.69
29.81
18.87
0.99
0.07
0.70
30.21
17.69
0.98
0.07
0.69
30.06
17.70
0.99
0.07
0.70
%
%
41.25
36.52
42.98
38.08
42.98
38.08
43.90
38.89
Digester
Chips
Brown Pulp Produced
Digester Steam
Yield
Brown Pulp
Bleached Pulp
The corresponding values calculated for the chemical recovery process parameters are show in Table 10.3. As
seen in the table, the increased liquor flow in the BLG cases, coupled with a higher BL solids concentration
being fed to the BLG unit, the total steam demand for evaporation and concentration is increased relative to the
kraft base case. However, the total flow, and total solids flow to the gasifier unit is lower in the BLG cases,
leaving room for extra capacity in situations where the mill is recovery boiler limited. Worth noting is that the
total syngas flow from the BLG unit to the gas turbine in the combined cycle power production unit is smaller
for the BLG case with 3% PS charge on OD wood. Higher levels of sulfur in the liquor loop will lead to higher
volumes of H2S produced in during gasification, which in turn increases the load on the sulfur recovery
operations and a smaller amount of syngas available for power production. The greatest effect of BLG
integration on the chemical recovery loop is the significant increase on the slaking load, with related increases
on the lime kiln and the demand for lime kiln fuel.
The values calculated from the assumptions made around process demand and production of power and steam
are shown in Table 10.4. The assumed steam demand for the process was indexed to the total amount of steam
produced in the base case from the recovery boiler. This total energy available for steam production was
determined from the WinGEMS model as 4.96*10^6 Mcal/day, producing about 282 mt/hr of 1000 psig steam
at 825 F. In order to meet this steam demand index, the BLG process would have to make the same amount of
steam from combustion of generated syngas, or from combustion of additional fuel sources. In this case study,
the balance of the required steam was met through combustion of additional hog fuel. The hog fuel demand
was controlled to generate the required reference of 282 mt/hr of high grade steam. In addition, the kraft base
case was assumed to be a neutral producer of power, neither selling nor buying power to or from the grid. The
reference index for power production is thus set to zero, meaning that any power produced in the BLG cases
from combined cycle syngas conversion plus additional power generated in the CC steam turbines from the hog
fuel boiler heat recovery, would be available for sale to the grid. This net amount of power varies in the BLG
cases from 1941.4 to 1834.4 KWh/ODtP produced.
Table 10.3
Comparison of chemical recovery process parameters for each simulated case
Unit
Kraft Basecase
105
BLG-SS
BLG-PS2%
BLG-PS3%
Evaporator
Flow to evaporator
Solids to evaporator
Dissolved solids to evaporator
Steam used (evaporated)
Steam used (concentrated)
mt/hr
%
kg/ODtP
kg/ODtP
kg/ODtP
456.16
17.35
1899.34
1436.77
570.79
484.72
15.83
1841.78
1604.60
811.84
475.19
16.49
1880.54
1539.44
829.11
483.44
16.47
1910.60
1567.54
842.22
Boiler / Gasifier
Flow to unit
Solids fired
Dissolved solids to unit
Amount Inorganic
Amount Organic
Energy available for Steam
Syngas to Gas Turbine
mt/hr
%
kg/ODtP
kg/ODtP
kg/ODtP
Mcal/day
mt/hr
122.73
66.29
1952.53
586.63
1365.91
4.96E+06
na
94.15
80.20
1812.12
562.10
1250.02
2.83E+05
47.71
95.70
80.10
1839.70
588.07
1251.63
2.74E+05
45.47
97.53
80.17
1876.50
670.31
1206.19
2.56E+05
42.47
Slaker
Flow
Reburned lime
Fresh lime
Make-up lime
mt/hr
kg/ODtP
kg/ODtP
mt/day
229.45
313.01
8.90
8.90
302.56
489.44
5.66
5.66
289.90
497.80
4.10
4.10
308.24
532.84
4.37
4.37
White Liquor make-up
NaOH (make-up)
Na2SO4
Elemental Sulfur
kg/ODtP
kg/ODtP
kg/ODtP
32.56
14.07
0.00
25.91
22.26
0.00
30.91
11.49
0.00
27.72
19.56
0.00
Kiln
Throughput
Fuel consumed
ODt/day
kg/day
557.92
55302.59
880.36
88577.27
895.80
90153.01
956.92
95363.40
Losses
Na, as Na
S, as S
kg/ODtP
kg/ODtP
22.48
2.70
21.14
3.36
22.43
2.62
23.22
4.05
mt/hr
mt/hr
mt/hr
mt/hr
mt/hr
mt/hr
mt/hr
116.47
595.89
3.22
22.55
15.50
143.06
753.62
83.31
596.15
3.22
22.55
24.45
175.85
729.68
96.34
596.14
3.22
22.55
24.88
164.47
743.13
90.76
596.16
3.22
22.55
26.58
175.48
739.28
Water Usage
Brownstock Washer Shwr Make-up
BP Wash Water
Dregs Washer
Mud Filter Mill Water
Mud Tank Cons. Reg. Mill Water
Scrubber Mill Water
Total
Dollar values based on best available data were assigned to selected operating variables, and the calculated
values obtained from the comparative economical analysis between the kraft base case and BLG cases are
shown in Table 10.5. The first two columns list the cost variables and their associated values. The remaining
columns list the values calculated from data obtained through WinGEMS simulations of the kraft base case, and
the three BLG case studies. The % DELTA column shows the calculated percent difference for the cost
variables between each simulated BLG case study and the kraft reference. The net cost in the last table row
indicates the variable operating cost in each case to produce one oven dry ton of pulp (1 ODtP). As shown in
the table, this net cost increases from the kraft base case ($169/ODtP) by up to 3.9% for the BLG cases with
PSAQ pulping ($176/ODtP). The driving force behind this effect is the increased load on the lime kiln that
106
Table 10.4
Comparison of power and steam process variables for each simulated case
Unit
Kraft Basecase
BLG-SS
BLG-PS2%
BLG-PS3%
Steam Users
Digester Steam
BP Steam
Green Liqour Heater Steam
MEE (evaporator)
MEE (concentrator)
BLG unit
Total Steam Demand
kg/ODtP
kg/ODtP
kg/ODtP
kg/ODtP
kg/ODtP
kg/ODtP
mt/day
1137.33
30.47
57.71
1436.77
570.79
0.00
3233.07
1227.63
29.81
103.59
1604.60
811.84
352.77
4130.26
1197.14
30.21
81.71
1539.44
829.11
355.92
4033.55
1216.28
30.06
84.62
1567.54
842.22
343.86
4084.59
Steam Demand (RB cap)
Energy Available/Required for Steam
Mcal/day
4.96E+06
2.83E+05
2.74E+05
2.56E+05
mt/day
mt/day
mt/day
mt/day
6750.52
0.00
0.00
6750.52
0.00
6382.69
385.31
6768.00
0.00
6394.36
373.63
6768.00
0.00
6419.36
348.64
6768.00
Mcal/day
Mcal/day
Mcal/day
Mcal/day
kWh/ODt
P
0.00
na
697024.79
0.00
na
1669299.00
697024.79
2366323.79
na
1663039.88
697024.51
2360064.39
na
1577281.69
697024.85
2274306.54
0.00
1941.40
1934.12
1834.39
Steam Production (1000 psig)
Recovery Boiler
Hog Fuel Boiler
Gas Turbine (Syngas)
Total Steam Production
Power Production
Recovery Boiler (Base Case net)
Gas Turbine (TG) (Syngas)
Condensing Steam Turbine
Total Power Produced
Net Power Produced from TG
Table 10.5
Comparison of case study variable operating cost parameters and net cost per ODtP
produced
Cost
(USD)
Kraft
Basecase
BLG-SS
%
DELTA
BLG-PS2%
%
DELTA
BLG-PS3%
Chips/kg
0.062
150.3
NaOH/lb
0.145
10.4
Na2SO4/lb
0.08
2.5
Fresh Lime/lb
0.055
Steam used/kg
0.006
Hog Fuel/kg
0.015
Kiln fuel/barrel
Water/1000 gal
144.2
-4.2
144.2
-4.2
141.2
-6.4
8.3
-25.7
9.9
-5.4
8.8
-17.5
3.9
36.8
2.0
-22.5
3.4
28.1
1.1
0.7
-57.3
0.5
-117.3
0.5
-103.6
19.4
24.8
21.7
24.2
19.8
24.5
20.8
0.0
58.4
100.0
58.5
100.0
58.7
100.0
60
24.4
39.0
37.6
39.7
38.7
42.0
42.0
0.095
1.7
1.7
-3.3
1.7
-1.4
1.7
-1.9
209.7
280.9
25.4
280.7
25.3
280.9
25.4
40.5
40.6
0.3
40.6
0.3
40.6
0.3
Total
Steam Produced/kg
Power Produced
KWh/ODTP
Total
Net Cost
0.006
%
DELTA
0.035
0.0
67.9
100.0
64.2
100.0
64.2
100.0
0
40.5
108.6
62.7
104.8
61.4
104.8
61.4
169.2
172.4
1.8
175.9
3.8
176.1
3.9
BLG will impose on the chemical recovery operations, combined with the assumption that additional hog fuel
must be burned in the BLG cases to meet the process steam demand. These two cost factors, hog fuel and lime
kiln fuel, are under the current assumptions a greater negative contributor than all the gains achieved through
107
BLG integration with respect to decreased wood costs, and the net process power produced from syngas
conversion in the combined cycle which is available for sale to the grid. This result changes dramatically if the
assumed cost of lime kiln fuel and/or the price of green power sold to the grid is altered. The effect of variable
lime kiln fuel (oil) prices has been illustrated in Figure 10.3. The plotted graphs represent the net variable
operating cost per OD ton pulp produced, calculated at different cost levels of lime kiln fuel while keeping all
other cost factors constant. As seen in the figure, the relative effect of increasing fuel price on the process net
variable operating cost is greater in the BLG cases compared to the MCC reference. This is an expected
outcome, as the BLG cases require a greater volume of kiln fuel. More importantly, the graph clearly shows
that the net variable operating cost for the BLG processes are lower than that of the MCC at lower fuel costs.
The break even points for the SS and PS2% processes compared to the MCC process is at a fuel price of about
$47/barrel. The corresponding value for the PS3% process is about $38/barrel. These results suggest that the
negative effects on variable operating costs for BLG processes, stemming from an increased lime kiln load,
could be overcome if a less expensive fuel alternative was available. The value used in the initial analysis was
based on current market price and assumed to be $60/barrel, well above the estimated break even points.
Net Variable Operating Cost ($/ODtP)
190
180
170
MCC
SS
160
PS2%
PS3%
150
20
30
40
50
60
70
80
Lime Kiln Fuel ($/barrel)
Figure 10.3
The effect of lime kiln fuel price on net variable operating cost, keeping all other cost
factors constant (assuming green power sales price at 3.5/KWh)
A similar analysis was performed to explore the effect of sales price of power generated in combined cycle
operations and sold to the grid. The net variable operating costs were calculated by varying the power sales
price and keeping all other cost factors constant. The calculated values for each simulated case are shown in
Figure 4. The basic assumption around power consumption/production for the kraft-MCC process was that the
reference mill was a neutral user/producer of power, and a change in power sales price would have no effect on
the net variable operating cost for the MCC case. Thus, the blue line in the figure, representing the MCC case is
flat over the explore range of power sales prices, whereas the calculated values for the BLG cases are highly
sensitive to changes in the power sales price. The initial assumed price for power sold to the grid (3.5/KWh)
was based on current market conditions, and at this level, the MCC process has a lower calculated net variable
operating cost. However, as the sales price is increased the effect on the BLG cases net variable operating costs
is dramatic and positive. In the current model, if the power sales price is increased from 3.5 to 6 /KWh, the net
variable operating costs would decrease by about $40/ODtP, with even greater cost savings realizable at higher
power sales prices.
108
Net Variable Operating Cost ($/ODtP)
250
200
MCC
150
SS
PS2%
100
PS3%
50
0
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Power Sales Price ($/Kwhr)
Figure 10.4
The effect of green power sales price on net variable operating cost, keeping all other
cost factors constant (assuming lime kiln fuel price at $60/barrel)
CONCLUSIONS
The overall conclusion from this simulation case study is that BLG integration into the kraft process is highly
dependent on the price of lime kiln fuel and the price and amount of power produced from BLGCC conversion
of syngas. Under the initial assumptions made in this case study, the significantly increased load on the lime
kiln is not overcome by the benefits realized in pulping operations through the introduction of modified pulping
technologies, nor by the additional revenue generated from the generation and sale of green power. However, if
modifications could be made to the recausticizing unit operations, such as high sulfidity green liquor
pretreatment, offloading the slaker and resulting load on the lime cycle, this would change. Another such
approach would be auto-causticization in the white liquor stream. Also, as indicated in figures 10.3 and 10.4, if
the lime kiln fuel cost is decreased or the power sale price increased, the BLGCC kraft process with split
sulfidity or PSAQ pulping becomes a more economically favorable alternative than the conventional reference
kraft process, based on variable operating costs. When comparing the BLG cases to the MCC reference, the net
variable operating cost break even point based on lime kiln fuel cost is about $47/barrel for the SS and PS2%
processes, and about $38/barrel for the PS2% process. This is significantly lower than assumed kiln fuel price
of $60/barrel used in this work. If the sales price for power to the grid was increased from 3.5 to 6 /KWh cost
savings of about $40/ODtP could be realized in all BLG processes.
FUTURE WORK
Future work should be focused on exploring the effects of alternative pulping processes such as autocausticization and mini-sulfide sulfite anthraquinone pulping and the overall improvement on net variable
operating costs that could be realized. More work should also be done in the area of combined cycle simulation
in WinGEMS, as well as improvements in the BLG WinGEMS model.
REFERENCES
1)
2)
3)
4)
Lindstrom, M., Kirkman, A., Jameel, H. et al., Proc. TAPPI Fall Technol. Trade Fair pp. 1417-1429 (2002)
Consonni, S., Larson, E.D., Katofsky, R., Proc. ASME Turbo Expo 7: 1 (2004)
Consonni, S., Larson, E.D., Katofsky, R., Proc. ASME Turbo Expo 7: 15 (2004)
Lindstrom, M., Naithani, V., Kirkman, A., Jameel, H., The Development and Validation of a Lowtemperature Black Liquor Gasifier Model for Use in WinGEMS TAPPI J. Vol. 5 No. 2 (2006) p.24
109
5) Lindstrom, M.; Naithani, V.; Kirkman, A.; Jameel, H.; Effects on Pulp Yield and Properties Using
Modified Pulping Procedures Involving Sulfur Profiling and Green Liquor Pretreatment; Presented at
2004 Tappi Fall Technical Conference, Atlanta, GA, 2004
6) Lindstrom, M.; Naithani, V.; Kirkman, A.; Jameel, H. The Effect of Integrating Polysulfide Pulping and
Black Liquor Gasification on Pulp Yield and Properties; Proc. 2005 TAPPI Engineering, Pulping &
Environmental Conference; TAPPI Press: Atlanta, GA, 2005.
7) Lindstrom, M., Naithani, V., Kirkman, A., Jameel, H., The Effect of Integrating Polysulfide Pulping
and Black Liquor Gasification on Pulp Yield and Delignification , Proceedings TAPPI 2006
Engineering, Pulping and Environmental Conference, Atlanta, GA, (2006).
110
11 INTEGRATING BLACK LIQUOR GASIFICATION AND PULPING AND A REVIEW OF
CURRENT TECHNOLOGY
Lindstrm, M; Jameel, H; Naithani, V; Kirkman, A; Renard, J; Integrating Black Liquor Gasification and
Pulping; A Review Of Current Technology; IN. Materials,Chemicals and Energy from Forest Biomass, ACS
Symposium Series, Argyropoulos, D. S., Ed.; Washington, ACS Books, 2006. (in Print, Scheduled for Public
Release Dec 2006)
111
INTEGRATING BLACK LIQUOR GASIFICATION AND PULPING AND A
REVIEW OF CURRENT TECHNOLOGY
Mathias Lindstrm, Hasan Jameel*, Ved Naithani, Adrianna Kirkman, and Jean Renard
Department of Forest Biomaterials Science and Engineering,
North Carolina State University, Raleigh, NC 27695-8005, USA
Gasification of black liquor could increase the flexibility and improve the profit potential of
the paper industry. Its implementation would enable the application of modified pulping
technologies, while creating a synthetic product gas that could be utilized in the production of
value added products or electrical power. B lack liquor gasification produces output streams
that can be used with great benefit in modified pulping operations. S plit sulfidity and
polysulfide modifications to the kraft process lead to yield increases of 1-3% points with
improved product quality. Modified sulfite pulping technologies resulted in yield increases of
5-18% points with much higher brightness and significant capital and operating cost savings.
Introduction
Biomass can be converted to power and/or fuels using a variety of technologies based around direct
combustion, gasification and pyrolysis. In direct combustion, the amount of oxygen provided has to be
sufficient for the efficient conversion of the carbon to carbon dioxide. In gasification the amount of oxygen is
limited in order to produce a medium to low calorific gas, and in pyrolysis the biomass is heated to a very high
temperature with limited oxygen to produce a mixture of gases and liquids with medium heating value.
One source of biomass that has typically been burned in a boiler to produce steam is black liquor, a byproduct from the pulping process. Black liquor gasification (BLG) has some inherent advantages compared to
the traditional combustion process. The efficiency of combustion is dependent on the mixing that occurs
between the combusted material and oxygen, and gases burn more efficiently than either liquids or solids due to
the improved contact between the oxygen and the fuel. In addition, the synthesis gas or syngas generated from
gasification can be burned in a gas turbine. This is advantageous as the gas turbine can convert energy to
electricity much more efficiently than a steam turbine, as used in a conventional chemical recovery system. The
syngas can also be converted to other fuels, chemicals and materials via a wide range of proven chemical
processes.
Presently, in a typical chemical pulp mill the black liquor is concentrated to greater than 65% dissolved
solids and burned in a recovery boiler. The pulping chemicals are recovered in the smelt and the heat energy is
converted to steam, which is used in a steam turbine generator to produce electricity. The typical thermal
efficiency of a recovery boiler is generally 65-70%, and the thermal efficiency of the Rankine cycle for the
conversion of steam to electricity varies from 30-38%, depending on the temperature and pressures of the
different streams in the cycle. These values result in an overall system thermal efficiency of about 23% (1). On
the other hand, if the black liquor is gasified, the syngas can after cleanup be combusted in a combined cycle for
production of electricity. Combined cycle power generation entails the sequential utilization of a gas turbine
followed by a steam turbine. The fuel gas is first burned in a gas turbine to produce electricity. The hot exhaust
gas from the turbine is then passed through a heat exchanger to produce steam which is then used in a powerproducing steam turbine. Implementing a gasifier with combined cycle cogeneration of power will increase the
electricity production of the mill. A conventional steam cycle produces about 120-180 kWh/ton of steam, but a
gasifier along with combined cycle power generation has the potential to generate 600-1000 kWh/ton of steam
(2). Such a production of power would turn a pulp mill into a net exporter of electricity, and this potential is
main motivation for the implementation of black liquor gasification.
In addition to the increased energy efficiency, gasification of black liquor has several other benefits relative
to the traditional combustion recovery process. BLG process operation is inherently very stable and also
flexible with regard to feed stock and load requirements. It is possible to process most any biomass material
and stable operation can be maintained despite upset feed stock flows, even complete interruptions. BLG has
112
the potential to revolutionize the chemical recovery cycle and, through the separation of sodium and sulfur,
enable the utilization of modified pulping technologies. These pulping technologies will increase yield or
reduce wood demand, improve product quality, decrease chemical usage and more importantly simplifying the
chemical recovery process. A simplification of the chemical recovery process will decrease the operating and
capital cost for recovery. BLG would also decrease the malodor associated with the kraft process. Besides
power generation, the resulting syngas can be used to generate bio-derived liquid fuels, bio-derived chemicals
for the synthetic chemical and pharmaceutical industries, as well as H2 for use in fuel cells.
Despite these benefits and opportunities, high capital cost and risk associated with new process
implementation are impeding the implementation of BLG technologies in the industry. The synergy between
BLG as increased energy generator and enabler of advanced pulping processes should increase the financial
attractiveness of the realization of these new process concepts. Based on these observations, this paper will
address the following topics:
Review of BLG technologies and their status
Review of modified pulping technologies enabled by BLG
Split sulfidity pulping
Polysulfide pulping
Alkaline sulfite with anthraquinone (AS-AQ)
Mini-sulfide sulfite pulping with anthraquinone (MSS-AQ)
Effects on process economics of BLG implementation
The effects on overall process economics of enhanced power production and modified pulping technologies
will then be discussed, and how the combined implementation of BLG with these technologies can improve the
financial attractiveness of the BLG technology.
Black Liquor Gasification Processes
Figure 1 describes the typical process elements included in the gasification of black liquor. The black
liquor is initially introduced into a process vessel, the black liquor gasifier, which can either be pressurized or
operate under atmospheric pressure. In general terms, the process involves the conversion of hydrocarbons and
oxygen to hydrogen and carbon monoxide while forming separate solid and gaseous product streams.
113
Figure 11.1
Simplified representation of BLGCC power/recovery systems
The inorganic material, including all sodium salts, leaves as a bed solid or smelt depending on the gasifier
operating temperature. The bed solids or smelt is then slaked and recausticized to form a caustic solution. The
volatiles, including most of the reduced sulfur species, leave as a syngas of medium BTU value. The major
components of the syngas are H2S, CO2, CO, H2O, and H2. To prepare the syngas for other applications and to
regenerate the pulping liquor, all sulfur must be separated from the syngas and then dissolved into the caustic
solution prepared from the bed solids. The clean product gas is burned in a gas turbine and the hot flue gases
are combined and used to generate steam in heat recovery steam generators (HRSGs). This steam is then used
in a steam turbine and other process applications.
A review of the development of alternative recovery technologies to the Tomlinson recovery boiler has
been given (3). The following discussion will focus on the gasification processes currently in commercial
operation. Black liquor gasification technologies can be classified by the operating temperature (4). High
temperature gasifiers operate at about 1000oC and low temperature gasifiers operate at less than 700oC. In the
high temperature gasifier, the inorganic material form a smelt and leave in the molten form, while in the low
temperature system, they leave as solids. The fuel value of the syngas produced is also dependent on the
gasifying technology. Typically, gasification produces a fuel gas with heating values of 3-4 MJ/Nm3 using air
and 89 MJ/Nm3 using oxygen (5).
Low Temperature Gasifier/Steam Reformer
The development of low temperature fluidized bed gasifiers is being pursued by ThermoChem Recovery
International (TRI) in the USA and by ABB in Sweden. The TRI system uses steam reforming to generate the
product gases. As opposed to exothermic incineration or combustion technologies, steam reforming is an
endothermic process. The steam reforming vessel operates at atmospheric pressure and at a medium
temperature. The organics are exposed to steam in a fluidized bed in the absence of air or oxygen with the
following reaction:
H2O + C + Heat = H2 + CO
The carbon monoxide produced in this first reaction then reacts with steam to produce more hydrogen and
carbon dioxide.
CO + H2O = H2 + CO2
The result is a synthesis gas made up of about 65% hydrogen. The TRI Steam Reformer technology, as
shown in Figure 2, consists of a fluidized bed reactor that is indirectly heated by multiple resonance tubes of
one or more pulse combustion modules. Black liquor is directly fed to the reactor, which is fluidized with
superheated steam. The black liquor uniformly coats the bed solids, producing a char and volatile pyrolysis
products which are steam cracked and reformed to produce a medium BTU gas. The residual char retained in
the bed is more slowly gasified by reaction with steam. The sulfur and sodium are separated in that the sulfur
becomes part of the gas stream and the sodium stays in solid form. Bed temperatures are maintained at 605-610
C, thereby avoiding liquid smelt formation and the associated smelt-water explosion hazards.
114
Figure 11.2
Schematic of MTCI Steam Reformer
Product gases are routed through a cyclone to remove the bulk of the entrained particulate matter and are
subsequently quenched and scrubbed in a Venturi scrubber. A portion of the medium-Btu product gases can be
supplied to the pulse combustion modules, and the combustion of these gases provides the heat necessary for
the indirect gasification process. Low temperature gasification leads to complete separation of the sulfur and
sodium in kraft black liquor to the gas and solid phase, respectively . Bed solids are continuously removed and
mixed with water to form a carbonate solution. The inorganic chemical in the bed solids as well as the sulfur
from the gas stream are recovered and used as cooking liquors for the mill. The product gas residence time in
the fluid bed is about 15 seconds because of the deep bed (20 ft) used, while the solids residence time is about
50 hrs. These conditions promote extensive tar cracking and carbon conversion. In summary the steam
reforming reactor vessel has three inputs; fluidizing steam, black liquor, and heat, and has three outputs; bed
solids, hydrogen rich product gas, and flue gas (6).
High Temperature Gasifier
High temperature gasification stems from work initiated by SKF in the 1970s. The original patent for the
technology was issued in 1987 and it has since been developed through a sequence of demonstration projects.
The gasifier, as developed by Chemrec, is a refractory-lined entrained-flow reactor. In high temperature
gasification (900-1000 C), concentrated black liquor is atomized, fed to the reactor and decomposed under
reducing conditions using air or oxygen as the oxidant. The initial chemical reactions involve char gasification
and combustion and are influenced by physical factors like droplet size, heating rate, swelling, and the sodium
and sulfur release phenomena. The resulting products, smelt droplets and a combustible gas, are then brought
into direct contact with a cooling liquid in a quench dissolver. The two phases are separated as the smelt
droplets dissolve in the cooling liquid forming green liquor. The exiting product gas is subsequently scrubbed
and cooled for use in other unit operations. The split of sodium and sulfur between the smelt and gas phase is
dependent on the process conditions. Typically, most of the sulfur leaves with the product gas and essentially
all of the sodium with the smelt (7-9).
Current Status of BLG Technologies
The TRI steam reformer has been installed in two locations in North America, at the Norampac Mill at
Trenton and the Georgia Pacific Mill at Big Island. The Trenton mill produces 500 tpd of corrugating medium
using a sodium carbonate based pulping process. Prior to the start-up of the low-temperature black liquor
gasifier in September 2003, the mill had no chemical recovery system. For over forty years the mills spent
115
liquor was sold to local counties for use as a binder and dust suppressant on gravel roads. This practice was
discontinued in 2002. The capacity of the spent liquor gasification system is 115 tpd of black liquor solids, and
the syngas is burned in an auxiliary boiler (6).
Georgia-Pacifics mill at Big Island, Virginia, produces 900 tpd of linerboard from OCC and 600 tpd of
corrugating medium from mixed hardwoods semi-chemical pulp. Like the Trenton mill, the Big Island mill
uses a sodium carbonate process. In the past, the semi-chemical liquor was burned in two smelters providing
chemical recovery but no energy recovery. Instead of replacing the smelters with a traditional recovery boiler
Georgia-Pacific decided to install a low temperature black liquor gasification process. One difference between
the two systems is that unlike Trenton, Big Island burns the generated product syngas in the pulsed combustors,
so the product gas exiting the reformer vessel is cleaned prior to combustion (6).
The evolution of the high temperature gasifier has taken the technology from an air-blown process near
atmospheric pressure to a high pressure (near 30 atm.) oxygen-blown process. Benefits realized through high
pressure oxygen-blown operation are higher efficiencies, higher black liquor throughput and improved
compatibility with down stream unit operations such as combined cycle power generation.
An air-blown pilot plant at Hofors, Sweden, was developed to verify the possibility of gasifying black
liquor using an entrained-flow reactor operating at 900-1000 C. The project showed that green liquor of
acceptable quality could be generated; and the plant was dismantled in 1990. The Fr vi, Sweden plant was
designed as a capacity booster for the AssiDomn facility and was operated from 1991 to 1996, demonstrating
the potential for black liquor gasification at a commercial scale. During its operation several technical problems
were encountered and addressed. The identification of a suitable material for the refractory lining remained a
problem. A subsequent commercial project was initiated in 1996 at the Weyerhauser plant in New Bern, North
Carolina. The black liquor gasifier was more or less a scale-up of the Frvi plant, designed for a capacity of
300 tons of dissolved solids/day. In 1999 the process maintained greater than 85% availability. However, over
the course of the project the plant experienced several technical problems, mainly related to the refractory
lining, and it was shut down after cracks in the reactor vessel were discovered in 2000. After detailed studies
and re-engineering, the gasifier operation at New Bern was resumed in the summer of 2003 ( 7). During the
rebuild, it was retrofitted with spinel refractory materials developed at Oakridge National Labs in cooperation
with other partners. The refractory material is in its second year of operation. The gasifier can burn up to
730,000 lb/day of solids or about 20% of the mill production (10). The syngas generated in the gasifier is
currently burned in a boiler.
A pressurized air-blown demonstration project was established at the Stora Enso plant at Skoghall, Sweden,
in 1994. The project showed the capability of a pressurized system to generate acceptable quality green liquor
while maintaining high carbon conversion ratios. The process was converted to an oxygen-blown operation in
1997 resulting in a capacity increase of more than 60%. A second pressurized demonstration plant was
completed in Pite, Sweden, in 2005. The purpose of the project is to demonstrate high pressure operation (near
30 atm.) with associated gas cooling and sulphur handling unit operations required for a full-scale BLG process.
Funding has been obtained for a scale-up project of the Pite facility. The plant is designed for a capacity of
275-550 tDS/day and encompasses all the required unit operations, including the power island, for a BLG
process with combined-cycle power generation (7).
BLG The Cornerstone of the Biorefinery
The integrated forest biorefinery is a concept which, if implemented, has the potential to dramatically
change the pulp and paper industry. The conversion of existing pulp mills into biorefineries is a natural
progression when trying to realize the full potentials in the by-product streams from pulp and paper making. It
is also a promising option for increasing the return on investment in an energy- and capital-intensive industry.
The gasification of black liquor and biomass presents the best option for the generation of high value products
from what today is essentially process waste (11). It is an inherently stable yet highly flexible technology that
can be designed and sized according to the needs and requirements of individual mills. It allows for the
recovery of any spent pulping liquor, and enables the generation of a wide array of liquors that can be used to
optimize pulping chemistry, pulp yield and properties. It can be coupled with various other unit operations to
116
generate power or feedstocks for liquid fuels and bio-chemical processes, and even hydrogen for utilization in
fuel cells. It is environmentally superior to current recovery boiler technologies and presents a carbon-neutral
source for power generation and synthetic products. BLG is an enabler of the biorefinery and the technologies
it encompasses (12).
The underlying fundamental for implementation of any new technology is the impact it will have on the
overall process economics. Some deciding factors that will influence the implementation of BLG involve the
cost-benefits associated with power generation and other high-value products that can be derived from the
syngas. Another area of importance is the potential cost-savings that can be realized through process
modifications and optimization. The effect on wood, chemical and fuel demand from changes in the pulping
process can have a significant effect on the variable operating cost, capital investment and maintenance costs.
Therefore, research exploring the impact of BLG on pulping technologies will be of great importance for the
eventual implementation of this technology.
Pulping Technologies Enabled by BLG
The implementation of black liquor gasification into the Kraft recovery cycle would present several
opportunities and potential benefits regarding pulp mill operation and process economics. Using black liquor
gasification, the recovered entities of sodium and sulfur can be split into two separate fractions with varying
degrees of separation dependent on the operating conditions and the technology used. The separation of these
chemicals creates some opportunities in the pulping process which can be employed to increase the pulp yield,
extend delignification and improve product quality. The following modified pulping technologies can be used
in combination with black liquor gasification to realize these potential benefits:
Split Sulfidity Pulping
Polysulfide Pulping
Alkaline Sulfite Pulping
Alkaline Sulfite- AQ (AS-AQ)
Mini-Sulfite Sulfide AQ Process (MSS-AQ)
Split Sulfidity Pulping
Modified kraft pulping processes have gained widespread acceptance, because they can be used to either
extend delignification or to enhance the yield and pulp properties at a given kappa number. The basic principles
of modified extended delignification consist of a level alkali concentration throughout the cook, a high initial
sulfide concentration, low concentrations of lignin and Na+ in the final stage of the cook, and lower temperature
in the initial and final stages of the cook (13). BLG would enable a mill to generate a high sulfidity liquor
which can be used to provide a high sulfide concentration during the initial phase of the cook. Figure 3 shows
the basic concept design for generating liquors of different sulfide concentrations.
117
Figure 11.3
Schematic of unit operations in split sulfidity pulping
In split sulfidity pulping, it would be necessary to generate two streams of white liquor one that is sulfide
rich and another that is sulfide lean. Sulfur profiling would be the lowest capital cost process to implement to
modify the pulping process especially for mills with a modified continuous or batch pulping process.
The concept of sulfur profiling, or split sulfidity pulping, employing a sulfur rich stream in the rapid initial
phase, followed by a sulfur lean stream in the bulk and residual phase, has been investigated as a method for
extending delignification or increasing yield (14-18). Compared to conventional kraft cooks of similar Hfactor, split sulfidity pulping has been shown to enhance selectivity of the pulping reactions, resulting in
increases in both lignin removal and pulp viscosity. Moreover, split sulfidity pulping has been shown to
increase pulp yield and strength properties (19-21).
The effects of multiple stage cooking using sulfur profiling, has also been studied. The process showed a
significant improvement in selectivity (22,23). Increased sulfide sorption resulted in both higher lignin-free
yields and increased viscosities. At 30% overall sulfidity, the lignin-free yield was 0.6 to 0.9% higher and
viscosity 8.89 to 10.4 mPa higher than conventional kraft. At increasing overall sulfidities, the yield advantage
was reduced. Screened yield increased only slightly with higher sulfidity levels during impregnation. Similar
findings were reported in subsequent work (24). Pulping work conducted at STFI found that sorption of sulfide
increases with increasing hydrosulfide concentration, time, temperature and concentration of positive ions, but
decreases with an increasing concentration of hydroxide ions (25). The potential for modifying softwood kraft
pulping, by sulfur profiling has also investigated, where all of the sulfide was added to the beginning of the
cook, a high hydrosulfide concentration could be maintained both in the initial phase and near the transition
point from the initial to the bulk delignification phase (26).
The work described above is difficult to implement in a mill that utilizes conventional recovery
technologies. However, BLG generates separate streams of sulfur and sodium, which will allow for
independent sulfur and alkali profiling. Thus, the alkali profile can be adjusted independent of the sulfur
concentration at any point in the cook. These opportunities were investigated at NC State University, exploring
split sulfidity pulping of southern pine with different initial alkali concentrations. Based on a modified
continuous cooking (MCC) laboratory procedure, different approaches were devised to explore split sulfidity
and different initial alkali profiles (27,28). Two levels of initial alkali were investigated where a fraction of the
available hydroxide was charged in the initial stage. The low initial alkali procedure used 11% of the alkali;
and the corresponding value for the high initial alkali procedure was 33%.
The effects of split sulfidity and different levels of initial alkali on delignification and total pulp yield are
presented in Figure 4. As shown, split sulfidity pulping produced lower kappa pulps at similar H factors
relative the MCC procedure. The high initial alkali cooks generated pulps of lower kappa number compared to
those of low initial alkali. The split sulfidity procedures produced pulp yields 1-2% greater than the MCC
procedure, and the difference is more pronounced at higher kappa. Since the high initial alkali approach
118
produced higher yields and lower kappa numbers than the low initial alkali approach, this would be the
preferred option. At similar kappa numbers the split sulfidity pulps had viscosities 5 to 10 cps greater than
those of the MCC pulps. The high initial alkali pulps produced higher tensile and burst index values relative the
MCC pulps at a similar tear index. The MCC pulps were slightly easier to refine relative to the split sulfidity
pulps.
kappa
60
MCC Baseline
40
20
600
Figure 11.4
SS_HIA
SS_LIA
54
Total Yield (%)
80
50
SS_HIA
SS_LIA
46
MCC Baseline
42
1100
1600
h factor
2100
20
40
60
kappa
80
Delignification and yield results for split sulfidity pulping
The co-absorption of H2S and CO2 during the scrubbing in sulfur recovery, results in the production of
NaHCO3. During recausticization all sodium exiting the gasifier will be converted to NaOH. The conversion
of NaHCO3 to NaOH requires twice as much lime compared to the conversion of NaCO 3 to NaOH. Thus, there
is a two-fold increase in the amount of lime required to produce an equivalent amount of NaOH, and as a result,
BLG will increase the overall causticization load.
The potential for in-situ causticization within the gasifier could dramatically affect the load on the recaust
cycle and lime kiln. In current recovery operations, the sodium carbonate obtained from the slaking of the
boiler smelt is converted to sodium hydroxide using calcium oxide. The byproduct calcium carbonate is then
calcined in large rotating kilns to regenerate the calcium oxide. A 1000 ton per day pulp mill will use about
100,000 barrels of fuel oil per year to fire its lime kiln. Through novel chemistries it may be possible to carry
out the causticization reactions directly within a black liquor gasifier. This could potentially eliminate the need
for the lime cycle and the associated fuel costs (29).
Another alternative to in-situ causticization, avoiding the increase in causticization requirements, would be
to pre-treat wood with green liquor. Previous work has demonstrated the feasibility of using green liquor in the
impregnation stage, without increasing overall chemical usage (15,21). It has also been shown that the amount
of sulfur adsorbed during the pretreatment decreases with higher [OH-] (30). By impregnating chips with high
sulfidity, low pH liquor, a mill may enhance yield and further decrease the causticizing load. Figure 5 outlines
the unit operations for a possible green liquor pretreatment process in conjunction with BLG.
119
Figure11.5
Outline of process using green liquor pretreatment
Comparing green liquor pretreatment and kraft pulping, the greatest relative cost-benefit from a decrease in
causticization using green liquor pretreatment would be achieved in a situation where the level of TTA was the
same in both processes. This requires that similar pulp kappa numbers must be attainable through both
processes at the same TTA charge. Green liquor pretreatment pulping has been investigated (28). Pulping
results show that green liquor pretreatment would return pulps of higher kappa at the same level of total
titratable alkali, as shown in Figure 6. However, as shown in the figure it would be possible to achieve similar
kappa number at the same level of system TTA by pulping to a higher H factor. Another option for decreasing
the kappa number would be to increase the system TTA. Experiments performed with a 10% increase in TTA,
labeled Hi-TTA, also resulted in a higher kappa number than the MCC baseline. In this study, if the TTA is
increased by 20% the active alkali is the same in both processes, and the no causticizing benefits exist. The
resulting pulp yield did not show any improvement with green liquor pretreatment, but the green liquor
pretreated pulps had higher viscosity. These results did not show the yield benefit reported elsewhere, and may
be the result from differences in pulping procedures (21).
kappa
80
Green Liq PT (HIA)
Green Liq PT (Hi_TTA)
MCC Baseline
60
40
20
600
Figure 11.6
1100
1600
h factor
Green Liq PT (HIA)
Green Liq PT (Hi_TTA)
MCC Baseline
58
Total yield (%)
100
2100
54
50
46
42
20
40
60
kappa
80
100
Obtained results for pulping using green liquor pretreatment
Polysulfide Pulping with Anthraquinone
The effect on pulping chemistry of polysulfide (PS), often in conjunction with anthraquinone (AQ) as
additives to the Kraft process, has been explored for some time (31-36). Its effectiveness has been established,
and it is typically reported that each percent of PS added increases the pulp yield by one percent (37,38).
However, efficiently generating high concentrations of PS within the Kraft chemical recovery cycle is difficult.
There are currently three primary competing processes available for PS generation, Chiyoda, MOXY and
Paprilox (39). These processes, in general terms, produce pulping liquors with PS concentrations of five to
120
eight grams per liter and PS selectivities ranging from 60 to 90 percent (40-43). This results in a PS limit of
about 1% PS charge on oven dry wood for a mill operating at 25% sulfidity. However, a chemical recovery
system based around BLG would allow for different pathways to generate PS liquors which would enable for
higher charges of polysulfide. In addition, the separation of sodium and sulfur would allow for alkali profiling
in conjunction with PS utilization. Figure 7 shows a schematic of PS process unit operations with BLG.
Figure 11.7
Outline of process using polysulfide
Research efforts in the area of PS have generally been in one of two major areas; work on PS pulping
associated with PS utilization in Kraft process operations and/or associated PS generation technologies (40-47)
and work investigating optimum parameters for PS pulping (48-50).
A smaller area of work has been based around the potential implementation of BLG and the opportunities
created by the unrestricted management of sulfur and sodium as separate entities. The splitting of sulfur and
sodium enables the application of polysulfide, sodium sulfide and sodium hydroxide independently of each
other. Two processes have been described based on this concept. The ZAP process (Zero effective alkali in
pretreatment), entails a two-stage pulping procedure, where the sulfur containing cooking chemicals (Na 2S and
PS) were charged along with AQ to the wood in the pretreatment stage (51). In the subsequent cooking stage
NaOH was added and the temperature increased. The obtained results using PS without AQ, indicate a
potential yield benefit of 1% relative to conventional PS pulping at kappa 30. With the addition of AQ the yield
benefit was increased to 1.5-2%. Additional results indicate even greater yield benefits at kappa 90. A different
procedure called hyperalkaline polysulfide pulping has been suggested ( 52). The process utilizes two
pretreatment stages followed by a cooking stage. In the first stage, alkali is charge to the wood at elevated
concentrations, neutralizing the acids formed during the temperature elevation. PS is then charged in the second
stage, followed by the cooking stage. The process resulted in a higher delignification rate, increased pulp
viscosities and yield improvements of 1.5% as compared modified pulping without PS. Worth noting is that the
bleachability and measured tear strength of the hyperalkaline PS pulps were similar to those of the Kraft
reference pulp.
In addition to the capability for independent profiling of NaOH and PS, the implementation of BLG would
allow for the conversion of all the sulfur in the cooking liquor to PS. This would enable higher PS charges than
available through conventional technologies. The total amount of sulfur that is available is dependent on the
sulfidity of the pulping liquor. Table 1 illustrates this balance displaying two examples of the partitioning of the
total sulfur available in the system at 25 and 40% sulfidity. As seen in the table, PS charges slightly exceeding
2 % on wood is possible at 19.5% AA with 25% sulfidity. To enable higher PS charges the sulfidity must be
increased and as shown in the table the corresponding value at 40% sulfidity is between 3 and 4% PS on wood.
121
Table 11.1
Demonstration of sulfur utilization as Na2S (kraft) or Na2S/PS (kraft-PS) and the
system sulfur availability for PS generation
25% Sulfidity
40% Sulfidity
Cook Procedure
Tot. avail.
Sulfur (S)
(kg/ton)
S req.
for PS
(kg/ton)
S avail.
as Na2S
(kg/ton)
Tot. avail.
Sulfur (S)
(kg/ton)
S req.
for PS
(kg/ton)
S avail.
as Na2S
(kg/ton)
MCC
25.2
0
25.2
40.3
0
40.3
1% PS
25.2
10.0
15.2
40.3
10.0
30.3
2% PS
25.2
20.0
5.2
40.3
20.0
20.26
3% PS
25.2
30.0
- 4.8
40.3
30.0
10.3
4% PS
25.2
40.0
- 14.8
40.3
40.0
0.3
The effects of pulping southern pine with higher PS charges and alkali profiling was evaluated at NC State
University. Table 2 shows a summary of the yield increases that were measured at various PS charges and
sulfidities. At 40% sulfidity the impact of alkali profiling in the initial stage was also evaluated. Three
different levels of alkali were investigated. In the low initial alkali (LIA) cook, 56% of the alkali was charged
to the impregnation stage. The corresponding values for the medium initial alkali (MIA) was 65% and for the
high initial alkali (HIA) 75%. The flexibility to optimize the alkali profile and PS use would only be possible in
combination with BLG (53).
Table 11.2
Total yield improvement from PS procedures compared to 25% S MCC pulp
Cook ID
Estimated Yield at 30
kappa
Yield Improvement (% pts.)
Average Viscosity (cps)
MCC 25% S
45.2
n.a.
39.1
1% PS 25% S
47.7
2.4
38.6
2% PS 25% S
48.1
2.9
38.0
Cook ID
Estimated Yield at 30
kappa
Yield Improvement (% pts.)
Average Viscosity (cps)
1% PS 40% S LIA
46.7
1.5
49.5
2% PS 40% S LIA
47.3
2.1
51.6
3% PS 40% S LIA
47.4
2.2
55.7
3% PS 40% S MIA
48.9
3.7
47.7
3% PS 40% S HIA
45.8
0.6
46.7
At 25% sulfidity the kappa numbers with PSAQ were comparable to the MCC reference. The yield benefit
was about 2% for a 1% PS charge, and about 3% for a 2% PS charge. At 40% sulfidity the kappa number
decreased with increasing PS charge, and increasing levels initial alkali. The level of initial alkali had a
significant effect on the yield. There is indication of an optimum condition for initial alkali charge, where too
great or too low of an initial hydroxide concentration negatively affects the pulp yield. The work exploring
MIA and HIA indicates that there exists a maximum in yield benefit as a function of initial alkali concentration.
In the ZAP process using PSAQ a yield benefit of about 6% at 30 kappa was reported, where the initial alkali
charge in the PS pretreatment stage was zero (50). They also showed that a minimum yield condition exists at a
hydroxide concentration of about 0.3 mol/l. At hydroxide concentrations lower or greater than 0.3 mol/l, higher
yields could be achieved. The results shown in Table 2 indicate that there also exists a maximum yield benefit
at higher levels of initial alkali. The LIA procedure corresponds to an initial hydroxide concentration of about
122
0.6 mol/l, which is greater than the concentration reported for the yield minimum in the ZAP process. The
maximum yield benefit was found around an initial hydroxide concentration of 0.9 mol/l. The effect of initial
alkali on pulp yield should be further investigated to optimize the benefits of PS pulping.
Alkaline Sulfite Pulping
The use of anthraquinone as an accelerator to alkaline sulfite pulping led to the development of the AS-AQ
process (54). The Mini-Sulfide Sulfite anthraquinone (MSS-AQ) pulping process was investigated in Sweden
in the mid 1980s (55,56). Both the AS-AQ and the MSS-AQ process have some unique features outlined
below (55-60):
Pulp yield for linerboard is 10% pt higher than kraft at equivalent Kappa number and strength
properties.
The brightness of the AS-AQ/MSS-AQ linerboard pulp is considerably higher than the brightness
of the corresponding kraft pulp: 40 % ISO and 18% ISO, respectively. This would be a
considerable advantage for high quality printing on cardboard boxes for advertising.
The need for the caustic room, lime kiln and associated energy usage is eliminated. This would be
a very large capital savings, especially for greenfield mills and mills with major rebuilds.
The yield advantage of AS-AQ/MSS-AQ over Kraft decreases rapidly as the kappa number
decreases. However, because the AS-AQ/MSS-AQ pulps respond very well to oxygen
delignification, it is possible to stop pulping at the defiberizing point (kappa 50) and to continue
the delignification with oxygen and alkali while keeping most of the yield increase achieved at
Kappa 50.
The AS-AQ/MSS-AQ pulps are considerably brighter and also easier to bleach than the
corresponding kraft pulps. ECF bleaching can be accomplished with lower ClO2 usage. High
brightness can be achieved with TCF sequences while producing a bleached pulp with acceptable
pulp strength. If TCF bleaching becomes cost effective, the capital cost for the chlorine dioxide
generators would also be eliminated.
Lower TRS emissions results in a low odor mill.
The difference between the AS-AQ and MSS-AQ processes is the amount is sulfur that exists as sodium
sulfide. In the AS-AQ process, all the sulfur exists as Na2SO3 with no Na2S, while the mini-sulfide sulfite
process uses a mixture of Na2S and Na2SO3. The decision to utilize AS-AQ or MSS-AQ will depend on the
amount of Na2S that can be generated in pulping liquor during recovery operations, and on the potential
improvements in delignification rate and pulp yield in MSS-AQ as compared to AS-AQ.
In the MSS-AQ process the total charge of alkali is about 22% Na 2O on OD wood, which is higher than for
the kraft process (19% Na2O). The delignification rate is slower than for the kraft process, and higher pulping
temperatures (+ 18 F) and anthraquinone addition (0.15% on OD wood) are required to obtain acceptable
pulping rates. Preliminary optimization studies done at STFI ( 55,56) indicate that a sulfide ratio between 0.05
to 0.1 results in the lowest kappa number. The optimum also shifts to a lower value with increasing AQ charge.
If the amount of sodium sulfide is increased above the optimum, the kappa number increases again.
Despite these significant advantages, the development of both the AS-AQ and MSS-AQ processes has not
been pursued more aggressively because of the lack of an attractive chemical recovery process. Black liquor
gasification combined with appropriate gas cleanup/absorber technology would present precisely such an
alternative. The benefits of these two processes were compared for the production of linerboard.
AS-AQ Pulping Work
Two AS-AQ processes have been evaluated with southern pine: the traditional strongly alkaline, SA, and
also the moderately alkaline, MA. It would be more attractive to operate at lower alkali charges since this
123
would minimize the amount of NaOH that would be necessary in the process. When compared to the Kraft
reference pulp, the moderately alkaline AS-AQ procedure produced pulps with a 10% yield benefit and a higher
ISO brightness. The pulp refined somewhat more slowly, had a higher apparent sheet density, a lower tensile
index and burst index at similar tear index ( 61). Similarly, the strongly alkaline AS-AQ procedure returned a
5% yield benefit and a pulp of comparable ISO brightness.
MSS-AQ Pulping Work
When compared to the Kraft-AQ reference pulp, MSS-AQ pulping produced pulps with a 15-18% yield
benefit and ISO brightness that was significantly higher. The yield and brightness of the different pulps are
shown in Figure 7. The pulp refined more easily and at similar levels of tear index the MSS-AQ pulps had a
slightly lower tensile and burst index (61).
ISO Brightness (%)
Total Yield (%)
72
66
60
AS-AQ
MSS-AQ
Kraft
54
48
95
Figure 11.7
100
105
kappa
40
AS-AQ
MSS-AQ
Kraft
30
20
10
95
110
100
105
kappa
110
Total pulp yield and ISO Brightness versus kappa for Kraft, AS-AQ and MSS-AQ
Table 3 shows the chemical balance for the different processes in kg of chemicals required as Na2O per
oven dry metric ton pulp (ODtP). Integration of alkaline sulfite AQ into a mill would require the addition of
sodium hydroxide to adjust the pH. The AS-AQ process operated at strongly alkaline conditions would require
the conversion of 163 kg of NaOH. This would require a recausticization system to be operated at the mill, and
therefore make this process alternative less attractive. In the mildly alkaline AS-AQ process the use of alkali
can be decreased to 35 kg NaOH/ODtP.
Table 11.3
Chemical Requirements for Selected Options for production of 1 ODtP (all chemicals as
Na2O) converted to kg/ton
Yield
(%)
Wood
(OD kg)
TTA
(%)
NaOH
(kg)
Na2S
(kg)
Na2CO3
(kg)
Kraft
50.0
2000
17.6
226
74
53
0
AS-AQ SA
55.4
1805
21.0
163
0
53
163
AS-AQ MA
60.0
1667
21.0
35
0
35
280
MSS-AQ1
66.6
1502
22.0
33
17
33
248
MSS-AQ2
67.3
1486
22.0
0
32
32
261
MSS-AQ3
69.6
1437
22.0
0
15
31
268
124
Na2SO3
(kg)
Some of the additional alkali required could be added as makeup chemical, but the 35 kg NaOH/ODtP is
still greater than that required as makeup in most well operated mills. The MSS-AQ pulping process can be
operated with no caustic, which would eliminate both the causticization process and the lime kiln. When the
sodium sulfide charge is 5% of the total chemical, 17 kg of sodium sulfide will have to be generated by
absorption of the H2S from the syngas. Alkaline sulfite methods generate pulp yields significantly greater than
the Kraft-AQ procedure at similar kappa. The moderately alkaline AS-AQ procedure resulted in an additional
10 % yield benefit. MSS-AQ pulping resulted in yield increases ranging from about 15 to 20% as compared to
the Kraft baseline. MSS-AQ and the moderately alkaline AS-AQ procedures generate pulps of significantly
higher unbleached ISO brightness, ranging from about 25 to 30 %, as compared to the Kraft and strongly
alkaline AS-AQ procedures with ISO brightness around 15 %. The higher pulp brightness indicates that there is
a large potential for cost savings in bleaching operations if bleachable grade pulp was produced.
MSS-AQ with RTI Absorption Technology
Research Triangle Institute (RTI) has developed durable zinc oxide based regenerable desulfurization
sorbents. The primary application for the technology has been removal of reduced sulfur compounds from hot
coal derived synthesis gases and recovery of concentrated SO2 streams. The product gases containing the sulfur
compounds will be scrubbed to remove particulates and then sent to the RTI absorber. The sulfur containing
compounds are absorbed in the ZnO bed. The clean gas can then be fed to the turbine generators for high
efficiency power generation. When the absorber bed is regenerated using air, the sulfur is desorbed as sulfur
dioxide which is the active chemical for the alkaline sulfite pulping processes. Tests were conducted at 260 to
530C and at pressures from 240 to 2000 kPa. In fixed bed sulfidation reactor tests, inlet H 2S levels were
decreased from percent levels to ppm levels. The sorbent capacity at breakthrough was as high as 17% S. The
sorbent was regenerated with 3.5% oxygen at 566C (61).
Black liquor gasification, RTI Absorption technology and MSS-AQ will have to be integrated into a
chemical pulp mill to take advantage of the yield benefits, energy savings, capital savings and the increased
power generation. The process flow-sheet with the above technologies integrated is shown in Figure 8.
The sulfur dioxide from the RTI absorber will be absorbed in the green liquor (sodium carbonate) to form
the sodium sulfite liquor. An appropriate amount of the sulfur gases will be absorbed as hydrogen sulfide to
produce the required amount of sodium sulfide. The RTI absorber serves the dual purpose of cleaning up the
product gas and also regeneration of the chemicals. As shown in the process schematic, the causticization and
lime kiln operations can be eliminated from the system as they are not needed in MSS-AQ pulping with BLG.
This will result in significant capital and operating cost-savings (61).
125
Figure 11.8
Schematic of the alkaline sulfite pulping processes with the RTI absorber
Economics of BLG Implementation
The overall economics associated with operating a BLG system will be impacted by the pulping process.
The return on investment will also have a significant effect on how rapidly the technology will be implemented.
The return on investment is dependent on the market price of oil, natural gas, and electricity, as well as the
value of products that can be produced from the BLG syngas, and process savings achieved through BLG
implementation. Work on forecasting the economic effects of BLG implementation, using analyses of existing
and potential technologies will therefore be important. Areas of interests for these studies involve the
exploration of markets for and economics of potential products, and the comparative analyses of different
technologies and process options for their manufacture.
The BLG generation of a syngas high in hydrogen and carbon monoxide will allow for the production of a
wide range of high value products currently unattainable with the Tomlinson recovery boiler. With different
options for down-stream technologies following the BLG the capability exists to generate steam, power, or
various bio-fuels and source materials for bio-chemicals processes such as (12).
DME/Methanol
Mixed alcohols
Fischer-Tropsch liquids
Syngas fermentation
Work has been done exploring the potential effects of modified pulping processes on the variable operating
costs of pulp mill operations. To create a wider knowledge-base for future decision making regarding what
products to manufacture and what process to use, more research is needed (11,62-65).
The Energy Group at the Princeton Environmental Institute has been a significant contributor to the work
exploring the cost-benefits of BLG operation and associated technologies. Work has been presented showing
the potential for different alternative products within the framework of the biorefinery concept, focusing on the
combined cycle generation of electrical power (64-68). A comparison of the reported estimated capital
investment required for installation of the different process options, including operations and maintenance costs,
is displayed in Table 4. The values in the table are given as percent changes for each process option compared
to the Tomlinson base case (BASE) indexed at 100.
Table 11.4
Estimated Cost Comparisons of Tomlinson and BLGCC Power/Recovery Systems
Relative the Tomlinson BASE (index = 100), (68)
Tomlinson
BLGCC System
BASE
HERB
Low-Temp
Mill-Scale
Direct Costs
100
130
172
162
202
Non-Direct Costs
100
140
267
151
186
TOTAL INSTALLED CAPITAL
COST
100
132
192
159
199
Annual Operating and
Maintenance Costs
100
154
154
154
162
126
High-Temp
Mill-Scale
High-Temp
Utility-Scale
The different process options include Tomlinson alternatives with either a conventional or a high efficiency
recovery boiler, and mill scale LT and HT BLG, as well as a utility scale HT BLG. The values were based on
Nth plant level of technology maturity and reliability and have an estimated accuracy of 30%. The capital
requirement for the installation of BLG technologies is higher than for the Tomlinson recovery boiler.
However, recovery operations built around a BLGCC system will convert the pulp mill to a significant net
producer of electricity, compared to a mill using a Tomlinson boiler which is a net importer of electricity. The
kraft reference mill used in the study required a net import of 36 MW, compared to a net production and export
of power to the grid of 22 MW and 126 MW, for the low and high temperature BLG processes, respectively.
To meet the same process steam demand, additional fuel or biomass would have to be processed. In detailed
work comparing the kraft reference mill to the LT and HT BLG processes, it was shown that the conversion
efficiency of biomass to power was about 15% for the non-pressurized LT BLG, nearly 50% greater than for the
reference kraft mill. For the two HT BLG processes the same values were about 19% for the smaller utility
design (85% increase) and about 17.5% for the mill-scale design (70% increase). The work also indicates
superior financial performance of the BLG technologies relative the kraft base comparison (66-68).
The internal rate of return (IRR) and net present value (NPV) were calculated for the different process
options. Assuming a 2.5/kWh premium and a tax credit of 1.8/kWh for the production of renewable energy,
the LT BLG process yielded an IRR of 20.9%/yr and a NPV of $73 million compared to the kraft base case.
The corresponding values for the HT BLG mill and utility scale processes were 34.8%/yr and $138.5 million
and 35.1%/yr and $216 million respectively. Also indicated in the work is the distinct window of opportunity
for the installation of the technology that exists in the near future, and that a lag in market penetration could be
assessed as a loss of $9 billion over a 25 year period measured in terms of energy costs and emission reductions
(66-68).
Summary and Conclusions
Biomass can be converted to power and/or fuels using gasification. The gasification of black liquor
produces output streams that can be used with great benefit in modified pulping operations with the additional
production of power and other high value products. BLG with combined cycle power production can generate
more electricity than combustion in a boiler given the same fuel, turning an integrated pulp and paper mill into a
net exporter of electricity. Other options for syngas utilization are the generation of bio-derived liquid fuels,
bio-derived chemicals for the synthetic and pharmaceutical industries, and H2 for use in fuel cells.
As discussed in this paper, the implementation of a BLG system will have a dramatic effect on current pulp
mill operations. In addition to the high value products made possible through the concepts related to the
biorefinery, BLG will enable different modified pulping technologies through the separation of sulfur and
sodium. A comparison of the different technologies enabled by BLG and their effects on the different pulping
processes is given in Table 5.
Table 11.5 General Comparison of BLG Enabled Pulping Technologies
Pulping technology
Effect on kappa
Effect on Yield
Split Sulfidity
Lower kappa at same
chemical charge
1-2% higher
yield
Green Liquor
pretreatment
Higher kappa at same
chemical charge
Polysulfide-AQ
Similar to lower kappa
at same chemical
charge
Effect on viscosity
Effect on properties
Comparable
viscosity
Improved strength properties
No/small
benefit
Increased viscosity
Improved strength properties
1-2% yield
increase per %
PS charged
Increased viscosity
(ZAP slight
decrease)
Similar properties, but lower tear
strength
(Hyperalkaline-PS has comparable
tear)
127
AS-AQ
Higher kappa at same
total chemical charge
5-10% yield
increase
Comparable
viscosity
Somewhat lower strength, much
greater brightness, easy to bleach
MSS-AQ
Higher kappa at same
total chemical charge
15-18% yield
increase
Comparable
viscosity
Somewhat lower strength, much
greater brightness, easy to bleach
As shown, modifications to the kraft process, using split sulfidity, green liquor pretreatment and polysulfide
pulping, can generate increased rates of delignification, improvements in pulp yield, viscosity and strength
properties, but not all at the same time. Further optimization of these technologies could have a significant
effect on the kraft process. Alkaline sulfite pulping can generate significant yield benefits and significantly
higher brightness. A unique opportunity exists in MSS-AQ pulping where it would be possible to eliminate the
causticizing and lime kiln operations, resulting in very significant cost-savings.
The overall economics associated with BLG implementation will have a significant effect on how rapidly
the technology will be implemented. Analyses of the costs and benefits associated the installation of BLGCC
technologies compared to conventional kraft operations using recovery boilers indicate increases of 20.9% in
the internal rate of return and $73 million in the net present value for a mill-scale low temperature BLG process.
The corresponding values for a mill-scale high temperature BLG process are 34.8% (IRR) and $138.5 million
(NPV). Above and beyond these cost-benefits, additional benefits could be realized from modified pulping
operations and the production of other high value products.
References
1. Nilsson, L.J.; Larson, E.D.; Gilbreath, K.R.; Gupta, A. Energy Efficiency and the Pulp and Paper Industry;
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
ACEEE: Washington D.C./Berkeley, CA, 1995.
Jahne, F. Commercial Success of Gasification Technology, Tappi J. 1999, 82(10): 49.
Whitty, K.; Verrill, C.L. A Historical Look at The Development of Alternative Black Liquor Recovery
Technologies and The Evolution of Black Liquor Gasifier Designs, Proc. 2004 TAPPI Int. Chem. Rec.
Conf.; TAPPI Press: Atlanta, GA, 2004; p 13-33.
Stigsson, L., ChemrecTM [Kvaerner Chemrec AB's] Black-Liquor Gasification, Proc. 1998 TAPPI Int.
Chem. Rec. Conf.; TAPPI Press: Atlanta, GA, 1998; p 663-674.
Grace, T.M.; Timmer, W.M. A Comparison of Alternative Black Liquor Recovery Technologies, Proc.
1995 TAPPI Int. Chem. Rec. Conf.; TAPPI Press: Atlanta, GA, 1995.
Rowbottom, B.; Newport, D.; Connor, E. Black Liquor Gasification at Norampac, Proc. 2005 TAPPI
Engineering Pulping & Environmental Conference; TAPPI Press: Atlanta, GA, 2005.
Chemrec ; URL http://www.chemrec.se; 2005.
Whitty, K.; Ekbom, T.; Stigsson, L. Chemrec Gasification of Black Liquor; published on website URL
http://www.chemrec.se/forsta.htm; 2000.
Lindblom, M. An Overview of Chemrec Process Concepts; Presentation at the Colloquium on Black
Liquor Combustion and Gasification; Park City, Utah, May, 2003.
Erickson, D.; Brown, C. Operating Experience with Gasification Pilot Project; Tappi J. 1999, 82(9).
Farmer, M. Adaptable Biorefinery: Some Basic Economic Concepts to Guide Research Selection; Proc.
2005 TAPPI Engineering, Pulping & Environmental Conference; TAPPI Press: Atlanta, GA, 2005.
Burciaga, D. Black Liquor Gasification: The Foundation of The Forest Bio-Refinery New Value
Streams, Proc. 2005 TAPPI Engineering, Pulping & Environmental Conference; TAPPI Press: Atlanta,
GA, 2005.
Johansson, B.; Mjoberg, J.; Sandstrom, P.; Teder, A. Modified Continuous Kraft Pulping - Now a
Reality; Svensk Papperstidning--Nordisk Cellulosa. 1984, 87(10): 30-35.
Herschmiller, D.W. Kraft Cooking with Split Sufidity- A Way to Break the Yield Barrier; Proc. Breaking
the yield barrier symposium; TAPPI Press: Atlanta, GA, 1998; Vol. 1 pp 59.
128
15. Andrews, E.K. Ph.D. thesis, North Carolina State University, Raleigh, NC, 1982.
16. Olm, L.; Tisdat, G. Kinetics of the Initial Stage of Kraft Pulping; Svensk Papperstidning--Nordisk
Cellulosa. 1979, 82(15): 458-464.
17. LeMon, S.; Teder, A. Kinetics of Delignification in Kraft Pulping (1). Bulk Delignification of Pine;
Svensk Papperstidning--Nordisk Cellulosa, 1973, 76(11): 407.
18. Olm, L.; Tormund, D.; Jensen, A. Kraft Pulping with Sulfide Pretreatment-Part-1 Delignification and
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
Carbohydrate Degradation; Nord. Pulp Pap. Res. J. 2000, 15, 62.
Jiang, J.E; Herschmiller, D.W. Sulfide Profiling for Increased Kraft Pulping Selectivity; Proc. 1996
TAPPI Pulping Conference; TAPPI Press: Atlanta, GA, 1996, p. 311-318.
Lownertz, P.P.H.; Herschmiller, D.W. Kraft Cooking with Split White Liquors and High Initial Sulfide
Concentration: Impact on Pulping and Recovery; Proc. 1994 TAPPI Pulping Conference; TAPPI Press:
Atlanta, GA, 1994, p. 1217-1224.
Andrews, E.K.; Chang, H-m.; Kirkman, A.G.; Eckert, R.C. Extending Delignification in Kraft and Kraft/
Oxygen Pulping of Softwood by Treatment with Sodium Sulfur Liquors; Proc. Japan Tappi Symposium on
Wood pulp chemistry; TAPPI Press: Atlanta, GA, 1982.
Mao, B.; Hartler, N. Improved Modified Kraft Cooking. (1). Pretreatment with a Sodium Sulfide
Solution; Paperi ja Puu. 1992, 74(6): 491-494.
Mao, B.; Hartler, N. Improved Modified Kraft Cooking. (2). Modified Cooking Using High Initial Sulfide
Concentration; Nordic Pulp and Paper Research J., 1992, 7(4): 168-173.
Mao, B.; Hartler, N. Improved Modified Kraft Cooking,4: Modified Cooking with Improved Sulfide and
Lignin Profiles; Paperi ja Puu. 1995, 77(6/7): 419-422.
Olm, L.; Backstrom, M.; Tormund, D. Pretreatment of Softwood with Sulfide Containing Liquor Prior to
Kraft Cook; Proc. 1994 TAPPI Pulping Conference; TAPPI Press: Atlanta, GA, 1994; 29.
Jiang, J.E.; Greenwood, B.F.; Phillips, J.R.; Stromberg, C.B. Improved Kraft Pulping by Controlled
Sulfide Additions; Proc. 7th ISWPC Conference; CICCST: Beijing, China, 1993.
Andritz Inc.; Personal communication; Oct, 2003.
Lindstrom, M.; Naithani, V.; Kirkman, A.; Jameel, H.; Effects on Pulp Yield and Properties Using
Modified Pulping Procedures Involving Sulfur Profiling and Green Liquor Pretreatment; Presented at
2004 Tappi Fall Technical Conference, Atlanta, GA, 2004.
Van Heiningen, A.; Schwiderke, E.; Chen, X. Kinetics of the Direct Causticizing Reaction Between Black
Liquor and Titanates During Low Temperature Gasification; Proc. 2005 TAPPI Engineering, Pulping &
Environmental Conference; TAPPI Press: Atlanta, GA, 2005.
Lopez, I.; Chang, H-m.; Jameel H.; Wizani, W. Effect of Sodium Sulfide Pretreatment on Kraft Pulping;
Proc. 1999 TAPPI Pulping Conference; TAPPI Press: Atlanta, GA, 1999; pp135.
Li, Z.; Li, J.; Kubes, G.J. Kinetics of Delignification and Cellulose Degradation During Kraft Pulping
with Polysulphide and Anthraquinone; JPPS. 2002, 28(7): 234-239.
Griffin, C.W.; Kumar, K.R.; Gratzl, J.; Jameel, H. Effects of Adding Anthraquinone and Polysulfide to the
Modified Continuous Cooking (MCC) Process; Proc. 1995 TAPPI Pulping Conference; TAPPI Press:
Atlanta, GA, 1995; 19-30.
Jiang, J.E. Extended Delignification of Southern Pine [Pinus spp.] with Anthraquinone and Polysulfide;
TAPPI J. 1995, 78(2): 126-132.
Jiang, J.E. Extended Modified Cooking of Southern Pine [Pinus] with Polysulfide: Effects on Pulp Yield
and Physical Properties; TAPPI J. 1994, 77(2): 120-124.
Landmark, P.A.; Kleppe, P.J.; Johnsen, K. Pulp Yield Increasing Process in Polysulfide Kraft Cooks;
Tappi J. 1965, 58(8): 56.
Vennemark, E. Some Ideas on Polysulfide Cooking; Svensk Papperstidning. 1964, 67(5): 157.
Kleppe, P.J.; Minja, R.J.A. The Possibilities to Apply Polysulfide-AQ in Kraft Mills; Proc. Breaking the
yield barrier symposium; TAPPI Press: Atlanta, GA, 1998, Vol. 1, pp 113.
Sanyer, N.; Laundrie, J.F. Factor Affecting Yield Increase and Fiber Quality in Polysulfide Pulping of
Loblolly Pine, Other Softwoods, and Red Oak; Tappi J. 1964, 47(10): 640.
129
39. Luthe, C.; Berry, R. Polysulphide Pulping of Western Softwoods: Yield Benefits and Effects on Pulp
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
Properties; Pulp. Pap. Can. 2005, 106(3): 27-33.
Munro, F.; Uloth, V.; Tench, L.; MacLeod, M.; Dorris, G. Mill-Scale Implementation of Paprican's
Process for Polysulphide Liquor Production in Kraft Mill Causticizers - Part 2: Results of Pulp Mill
Production Trials; Pulp. Pap. Can. 2002, 103(1): 57-61.
Tench, L.; Uloth, V.; Dorris, G.; Hornsey, D.; Munro, F. Mill Scale Implementation of Papricans
Process for Polysulfide Liquor Production in Kraft Mill Causticizers. Part-1 Batch Trials and
Optimization; TAPPI J. 1999, 82(10): 120.
Yamaguchi, A. Operating Experiences with the MOXY Process and Quinoid Compounds; In
Anthraquinone Pulping: Anthology of Published Papers 1977-1996; Goyal, G.C., Ed.; TAPPI Press:
Atlanta, GA, 1997; pp 287-291.
Nishijima, H.; et al., Review of PS/AQ Pulping to Date in Japanese Kraft Mills and the Impact on
Productivity; Proc. 1995 TAPPI Pulping Conference; TAPPI Press: Atlanta, GA, 1995; 31-40.
MacLeod, M.; Radiotis, T.; Uloth, V.; Munro, F.; Tench, L. Basket Cases IV: Higher Yield
with Paprilox Polysulfide-AQ Pulping of Hardwoods; TAPPI J. 2002, 1(10): 3-8.
Olm, L.; Tormund, D.; Bernor Gidert E. Possibilities to Increase the Pulp Yield in a Kraft Cook of [the]
ITC-Type; Proc. Breaking the yield barrier symposium; TAPPI Press: Atlanta, GA, 1998; Vol. 1, 69-78.
Hakanen, A.; Teder, A. Modified Kraft Cooking with Polysulfide: Yield, Viscosity, and Physical
Properties; TAPPI J. 1997, 80(7): pp 86,93,100, 189-196.
Jiang, J.E.; Crofut, K.R.; Jones, D.B. Polysulfide Pulping of Southern Hardwood Employing the
MOXYRG and Green-Liquor Crystallization Processes; Proc. 1994 TAPPI Pulping Conference; TAPPI
Press: Atlanta, GA, 1994; pp 799-806.
Gustafsson, R.; Ek, M.; Teder, A. Polysulphide Pretreatment of Softwood for Increased Delignification
and Higher Pulp Viscosity; JPPS. 2004, 30(5): 129-135.
Mao, B. F.; Hartler, N.; Improved Modified Kraft Cooking,3: Modified Vapor-Phase Polysulfide
Cooking; TAPPI J. 1994, 77(11): 149-153.
Lindstrom, M.; Teder, A. Effect of Polysulfide Pretreatment When Kraft Pulping to Very Low Kappa
Number; Nordic Pulp Paper Res. J. 1995, 10(1): 8-11.
Olm, L.; Tormund, D. ZAP Cooking- Increase Yield in PS and PS-AQ Cooking with Zero Effective Alkali
in the Pretreatment Stage; Nordic Pulp Paper Res. J. 2004, 19(1): 6.
Brannvall, E.; Gustafsson, R.; Teder, A.; Properties of Hyperalkaline Polysulphide Pulps; Nordic Pulp
Paper Res. J. 2003, 18(4): 436-440.
Lindstrom, M.; Naithani, V.; Kirkman, A.; Jameel, H. The Effect of Integrating Polysulfide Pulping and
Black Liquor Gasification on Pulp Yield and Properties; Proc. 2005 TAPPI Engineering, Pulping &
Environmental Conference; TAPPI Press: Atlanta, GA, 2005.
Nomura, Y.; Wakai, M.; and Sato, H. Process for Producing Pulp in the Presence of a Quinine
Compound; Japanese Patent 112,903, 1967; Canadian Patent 1079906, 1980.
Dahlbom, J.; Olm, L.; Teder, A. Characteristics of MSS-AQ [Minisulfide-Sulfite-Anthraquinone] Pulping
- A New Pulping Process TAPPI J., 1990, 73(3): 257-261.
Olm, L.; Wiken, J.E.; Olander, K. Alkaline Sulfite Pulp: Competitor to Sulfate Pulp; Svensk
Papperstidning--Nordisk Cellulosa. 1986, 3(4): pp 75-77,85.
Borgards, A.; Patt, R.; Kordsachia, O.; Odermatt, J.; Hunter, W. D. Comparison of ASAM and Kraft
Pulping and ECF/TCF Bleaching of Southern Pine [Pinus spp.] Proc. 1993 TAPPI Pulping Conference;
TAPPI Press: Atlanta, GA, 1993; Vol. 2, pp 629-636.
Patt, R.; Schubert, H.L.; Kordsachia, O.; Oltmann, E.; Krull, M. Chlorine-Free Bleaching of Sulfite and
ASAM Pulps on the Laboratory and Pilot-Plant Scale; Papier. 1991, 45(10A): V8-V16.
Ingruber, O.V.; Stradal M.; Histed, J.A.; Alkaline Sulfite-Anthraquinone Pulping of Eastern Canadian
Woods; Pulp and Paper Canada. 1982, 83(12): 79-84, 87-88.
Virkola,N.E.; Pusa,R.; Kettunen, J.; Neutral sulfite AQ Pulping as an Alternative to Kraft Pulping; Tappi
J. 1981, 64(5): 103-107.
130
61. Jameel, H.; Lindstrom, M.; Naithani, V.; Kirkman, A. Alkaline Sulfite Pulping and its Integration with the
62.
63.
64.
65.
66.
TRI Process; Proc. 2005 TAPPI Engineering, Pulping & Environmental Conference; TAPPI Press:
Atlanta, GA, 2005.
Lindstrom, M.; Kirkman, A.; Jameel, H. et al., Economics of Integrating Black Liquor Gasification With
Pulping: Part I Effect Of Sulfur Profiling; Proc. 2002 TAPPI Engineering, Pulping & Recycling
Conference; TAPPI Press: Atlanta, GA, 2002.
Berglin, N.; Lindblom, M.; Ekbom, T. Preliminary Economics of Black Liquor Gasification with Motor
Fuels Production; Presentation at the Colloquium on Black Liquor Combustion and Gasification; Park
City, Utah, May, 2003.
Larson, E.D.; Haiming, J.; Biomass Conversion to Fischer-Tropsch Liquids: Preliminary Energy
Balances; Proc. 4th Biomass Conference of the Americas; Elsevier Science Ltd: Oxford, UK, 1999.
Williams, R.H.; Larson, E.D.; Katofsky, R.E.; Chen, J. Methanol and Hydrogen from Biomass for
Transportation; Energy for Sustainable Development. 1995, 1(5): 18.
Larson, E.D.; Consonni, S.; Katofsky, R.E. A Cost-Benefit Assessment of Biomass Gasification Power
Generation in the Pulp and Paper Industry Final Report ; prepared with support from the American Forest &
Paper Association and the U.S. Department of Energys Office of Biomass Programs (available from the authors) (Oct.
8, 2003).
67. Consonni, S.; Larson, E.D.; Katofsky, R.E. An Assessment of Black Liquor Gasification Combined Cycles
68.
Part A: Technological Issues And Performance Comparisons; Proc. ASME Turbo Expo 2004 Power for
Land, Sea, and Air; ASME: 2004.
Larson, E.D.; Katofsky, R.E.; Consonni, S. An Assessment of Black Liquor Gasification Combined Cycles
Part B: Emissions, Costs and Macro-Benefits; Proc. ASME Turbo Expo 2004 Power for Land, Sea, and
Air; ASME: 2004.
131
12 APPENDICES
132
12.1 Notes on pump calibration and operation
The cart based American LEWA dual head piston pump, was purchased from Hydrocorp Inc.,
represented by Mr. Gene Prather (Am LEWA representative):
[phone]
[phone]
[e-mail]
The operators manual is contained in a black spiral three ring binder and will be stored with Dr. Hasan
Jameel. The following are some operational notes on how to set up and use the pump.
Note: Never run the pump against a closed line, as the piston will generate enough force to almost
immediately cause the rupture disks in the pump assembly to fail.
The rupture disk performance and
replacement information is contained on the plates attached to the pump assembly.
The pump flow rate is calibrated by stroke length and pump motor speed. The stroke length is adjusted
from zero to 15 mm by turning the wheels on top of each pump head. If the dial is set to zero, it is possible (and
permissible) to run the pump motor without displacing the piston. The pump motor speed is controlled using
the electrical control zero to sixty (0-60) Hz. The pump was during this project operated at a 200 ml/min flow
rate. The pump is wired to match the 3 prong outlet on the south wall of the pulping lab (Biltmore 1103) next to
the hallway door behind the EKA digester. If the pump is moved, appropriate power compatibility must be
ensured!
133
12.2 Protocol for simulated MCC pulping
Protocol for MCC procedure NCSU, January 04
The following procedure was developed from information provided by International Paper, and is aimed at
laboratory simulation of modified continuous cooking. The protocol involves a three stage procedure, involving
two liquor additions (see Liquor addition/transfer protocol) and using the stage II and III temperature to reach
the desired h factor under a fixed total cooking time.
Table 1. Outline of MCC protocol
Cumulative % TTA
Stage I
65
Stage II
85
Stage III
100
L/W
3.5
4.1
4.5
Stage Temperature (C)
120
155-166
155-166
Time at Temperature (min)
15
70
120
Stage I.
The first stage is organized as a traditional lab batch cook using the M&K digesters. 65 % of the TTA is placed
in the digester with the chips and dilution water required to meet a liquor:wood ratio of 3.5:1. The cook is then
heated to 120 and held at temperature for 15 minutes. At this point a BL sample for residual effective alkali
(REA) determination is collected and the first liquor addition is initiated.
Stage II
The second stage follows the first liquor addition. During Stage II the digester is heated to a predetermined
temperature that over the span of the cook will result in the desired final h factor. The cook is held at
temperature for 70 minutes, at which point a BL sample for REA determination is collected and the second
liquor transfer is initiated. (Optional An additional BL sample can be collected when at any point after the
first liquor addition to monitor the REA.)
Stage III
The third stage follows the second liquor addition. In Stage III the digester is held at the predetermined
temperature (see above) for 120 minutes. A sample of BL is collected for analysis at the end of the cook during
the blow. (Optional An additional BL sample can be collected when at any point after the second liquor
addition to monitor the REA.)
The pulp is then defibered using the mixer (chips are evenly distributed in two buckets and each bucket is
mixed for 5 minutes) and screened. Screened yield, rejects and total yield are determined.
Suggested BL sample collection points for REA determination
At the completion of stage I
(optional) Anytime after the first liquor addition and sufficient mixing
At the completion of stage II
(optional) Anytime after the second liquor addition and sufficient mixing
At the completion of stage III, during the blow of the cook
134
12.3 Protocol for Polysulfide generation
Polysulfide (PS) liquors were generated and their concentration determined based on procedures
developed from work published by Dorris and Uloth at Paprican.
PS liquor was generated by dissolving elemental (sublimated) sulfur into a Na2S solution, so as to
generate a 10% PS liquor. This was accomplished by placing 160g of elemental S in 1600 ml of Na 2S solution
(50 gpl as Na2O), and mixing the solution continuously heated to 50-60 deg C under a N 2 atmosphere. The
basic equipment used was a 2000 ml E-flask (reaction vessel) with a modified stopper to allow for a
thermometer and continuous circulation of N2, a stirrer hot-plate and Teflon coated magnetic bar. The mixture
was allowed to react until the visible sulfur particles had gone into solution, turning the solution deep
orange/amber. This takes 4-12 hrs depending on the relative concentrations of elemental sulfur and Na 2S
solution. The PS concentration of the generated PS mixture can then be determined through the Dorris and
Uloth gravimetric procedure. The PS solution should be stored separated from oxygen, by vacuum seal or by
floating a layer of mineral oil on top of the PS solution. The PS solution is stable for some time, depending on
the storage conditions.
135
i
) Nilsson, L.J.; Larson, E.D.; Gilbreath, K.R.; Gupta, A. Energy Efficiency and the Pulp and Paper Industry;
ACEEE: Washington D.C./Berkeley, CA, 1995.
ii
)
iii
) Whitty, K.; Verrill, C.L. A Historical Look at The Development of Alternative Black Liquor Recovery
Technologies and The Evolution of Black Liquor Gasifier Designs, Proc. 2004 TAPPI Int. Chem. Rec. Conf.; TAPPI
Press: Atlanta, GA, 2004; p 13-33.
iv
) Stigsson, L., ChemrecTM [Kvaerner Chemrec AB's] Black-Liquor Gasification, Proc. 1998 TAPPI Int. Chem.
Rec. Conf.; TAPPI Press: Atlanta, GA, 1998; p 663-674.
v
) Grace, T.M.; Timmer, W.M. A Comparison of Alternative Black Liquor Recovery Technologies, Proc. 1995
TAPPI Int. Chem. Rec. Conf.; TAPPI Press: Atlanta, GA, 1995.
vi
) Rowbottom, B.; Newport, D.; Connor, E. Black Liquor Gasification at Norampac, Proc. 2005 TAPPI
Engineering Pulping & Environmental Conference; TAPPI Press: Atlanta, GA, 2005.
vii
)
viii
) Whitty, K.; Ekbom, T.; Stigsson, L. Chemrec Gasification of Black Liquor; published on website URL
http://www.chemrec.se/forsta.htm; 2000.
ix
) Lindblom, M. An Overview of Chemrec Process Concepts; Presentation at the Colloquium on Black Liquor
Combustion and Gasification; Park City, Utah, May, 2003.
x
)
Erickson, D.; Brown, C. Operating Experience with Gasification Pilot Project; Tappi J. 1999, 82(9).
xi
)
Kleppe, P.J., TAPPI 53:1 35-47 (1970)
xii
)
McDonough, T.J., Kraft Pulp Yield Basics, Proc. Breaking the yield barrier symposium; TAPPI Press: Atlanta,
GA, 1998, Vol. 1, p1.
xiii
) Sixta, H. ed., Handbook of Pulp and Paper Vol 1 and 2. Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim,
Germany, 2006 Ch. 4
xiv
)
Aurell, R., Hartler, N., Svensk Papperstidning 68(4):59-68 (1965)
xv
)
Kondo, R., Sarkanen, K.,V., Holzforshung 38:31 (1984)
xvi
) Gellerstedt, G. Pulping Chemistry. In: Wood and Cellulosic Chemistry, 2nd ed., Hon, D.N.S., Shiraishi, N., Eds.,
M. Dekker, New York, Basel, 2001 p.859
xvii
)
Adler, E., Wood Sci. Technol. 11:169 (1977)
xviii
)
Gierrer, J., Holzforschung, 36:43 (1982)
xix
)
Gellerstedt, G., Lindfors, E-L., Nordic Pulp Paper Res. J., 2(2):71 (1987)
xx
)
Gustavsson, C., Sjostrom, K., Wafa Al-Dajani, W., Nordic Pulp Paper Res. J., 14(1):71 (1999)
xxi
)
Nimz, H., H., Holzforschung, 335 16-26 (1981)
xxii
)
Gierrer, J., Svensk Papperstidning 73:571 (1970)
xxiii
)
Gierrer, J., Wood Sci. Technol. 19:1289 (1985)
xxiv
)
Dimmel,.;D.R., Bovee, L.F.., Brogdon, B.N., Wood Chem. Technol., 14:1 (1994)
xxv
)
Gierrer, J., Wannstrom, S., Holzforschung, 40(6):347 (1986)
xxvi
) Ishii, T., Shimizu, K., Chemistry of Cell Wall Polysaccharides. In: Wood and Cellulosic Chemistry, 2nd ed., Hon,
D.N.S., Shiraishi, N., Eds., M. Dekker, New York, Basel, 2001 p.859
xxvii
)
xxviii
) Johansson, B.; Mjoberg, J.; Sandstrom, P.; Teder, A. Modified Continuous Kraft Pulping - Now a Reality;
Svensk Papperstidning--Nordisk Cellulosa. 1984, 87(10): 30-35.
xxix
)
Jahne, F. Commercial Success of Gasification Technology, Tappi J. 1999, 82(10): 49.
Chemrec ; URL http://www.chemrec.se; 2005.
Meller, A., Holzforschung, 19(4):118 (1965)
Andrews, E.K. Ph.D. thesis, North Carolina State University, Raleigh, NC, 1982.
xxx
) Andrews, E.K.; Chang, H-m.; Kirkman, A.G.; Eckert, R.C. Extending Delignification in Kraft and Kraft/Oxygen
Pulping of Softwood by Treatment with Sodium Sulfur Liquors; Proc. Japan Tappi Symposium on Wood pulp
chemistry; TAPPI Press: Atlanta, GA, 1982.
xxxi
) Lopez, I.; Chang, H-m.; Jameel H.; Wizani, W. Effect of Sodium Sulfide Pretreatment on Kraft Pulping; Proc.
1999 TAPPI Pulping Conference; TAPPI Press: Atlanta, GA, 1999; pp135.
xxxii
) Herschmiller, D.W. Kraft Cooking with Split Sufidity- A Way to Break the Yield Barrier; Proc. Breaking the
yield barrier symposium; TAPPI Press: Atlanta, GA, 1998; Vol. 1 pp 59.
xxxiii
)
Olm, L.; Tisdat, G. Kinetics of the Initial Stage of Kraft Pulping; Svensk Papperstidning--Nordisk Cellulosa.
1979, 82(15): 458-464.
xxxiv
LeMon, S.; Teder, A. Kinetics of Delignification in Kraft Pulping (1). Bulk Delignification of Pine; Svensk
Papperstidning--Nordisk Cellulosa, 1973, 76(11): 407.
xxxv
) Olm, L.; Tormund, D.; Jensen, A. Kraft Pulping with Sulfide Pretreatment-Part-1 Delignification and
Carbohydrate Degradation; Nord. Pulp Pap. Res. J. 2000, 15, 62.
xxxvi
) Jiang, J.E; Herschmiller, D.W. Sulfide Profiling for Increased Kraft Pulping Selectivity; Proc. 1996 TAPPI
Pulping Conference; TAPPI Press: Atlanta, GA, 1996, p. 311-318.
xxxvii
) Lownertz, P.P.H.; Herschmiller, D.W. Kraft Cooking with Split White Liquors and High Initial Sulfide
Concentration: Impact on Pulping and Recovery; Proc. 1994 TAPPI Pulping Conference; TAPPI Press: Atlanta, GA,
1994, p. 1217-1224.
xxxviii
) Mao, B.; Hartler, N. Improved Modified Kraft Cooking. (1). Pretreatment with a Sodium Sulfide Solution;
Paperi ja Puu. 1992, 74(6): 491-494.
xxxix
) Mao, B.; Hartler, N. Improved Modified Kraft Cooking. (2). Modified Cooking Using High Initial Sulfide
Concentration; Nordic Pulp and Paper Research J., 1992, 7(4): 168-173.
xl
) Mao, B.; Hartler, N. Improved Modified Kraft Cooking,4: Modified Cooking with Improved Sulfide and Lignin
Profiles; Paperi ja Puu. 1995, 77(6/7): 419-422.
xli
) Olm, L.; Backstrom, M.; Tormund, D. Pretreatment of Softwood with Sulfide Containing Liquor Prior to Kraft
Cook; Proc. 1994 TAPPI Pulping Conference; TAPPI Press: Atlanta, GA, 1994; 29.
xlii
) Jiang, J.E.; Greenwood, B.F.; Phillips, J.R.; Stromberg, C.B. Improved Kraft Pulping by Controlled Sulfide
Additions; Proc. 7th ISWPC Conference; CICCST: Beijing, China, 1993.
xliii
) Lindstrom, M.; Naithani, V.; Kirkman, A.; Jameel, H.; Effects on Pulp Yield and Properties Using Modified
Pulping Procedures Involving Sulfur Profiling and Green Liquor Pretreatment; Presented at 2004 Tappi Fall
Technical Conference, Atlanta, GA, 2004.
xliv
) Van Heiningen, A.; Schwiderke, E.; Chen, X. Kinetics of the Direct Causticizing Reaction Between Black Liquor
and Titanates During Low Temperature Gasification; Proc. 2005 TAPPI Engineering, Pulping & Environmental
Conference; TAPPI Press: Atlanta, GA, 2005.
xlv
) Li, Z.; Li, J.; Kubes, G.J. Kinetics of Delignification and Cellulose Degradation During Kraft Pulping with
Polysulphide and Anthraquinone; JPPS. 2002, 28(7): 234-239.
xlvi
) Griffin, C.W.; Kumar, K.R.; Gratzl, J.; Jameel, H. Effects of Adding Anthraquinone and Polysulfide to the
Modified Continuous Cooking (MCC) Process; Proc. 1995 TAPPI Pulping Conference; TAPPI Press: Atlanta, GA,
1995; 19-30.
xlvii
) Jiang, J.E. Extended Delignification of Southern Pine [Pinus spp.] with Anthraquinone and Polysulfide; TAPPI
J. 1995, 78(2): 126-132.
xlviii
) Jiang, J.E. Extended Modified Cooking of Southern Pine [Pinus] with Polysulfide: Effects on Pulp Yield and
Physical Properties; TAPPI J. 1994, 77(2): 120-124.
xlix
) Landmark, P.A.; Kleppe, P.J.; Johnsen, K. Pulp Yield Increasing Process in Polysulfide Kraft Cooks; Tappi J.
1965, 58(8): 56.
l
)
Vennemark, E. Some Ideas on Polysulfide Cooking; Svensk Papperstidning. 1964, 67(5): 157.
li
) Kleppe, P.J.; Minja, R.J.A. The Possibilities to Apply Polysulfide-AQ in Kraft Mills; Proc. Breaking the yield
barrier symposium; TAPPI Press: Atlanta, GA, 1998, Vol. 1, pp 113.
lii
) Sanyer, N.; Laundrie, J.F. Factor Affecting Yield Increase and Fiber Quality in Polysulfide Pulping of Loblolly
Pine, Other Softwoods, and Red Oak; Tappi J. 1964, 47(10): 640.
liii
) Luthe, C.; Berry, R. Polysulphide Pulping of Western Softwoods: Yield Benefits and Effects on Pulp Properties;
Pulp. Pap. Can. 2005, 106(3): 27-33.
liv
) Munro, F.; Uloth, V.; Tench, L.; MacLeod, M.; Dorris, G. Mill-Scale Implementation of Paprican's Process for
Polysulphide Liquor Production in Kraft Mill Causticizers - Part 2: Results of Pulp Mill Production Trials; Pulp. Pap.
Can. 2002, 103(1): 57-61.
lv
) Tench, L.; Uloth, V.; Dorris, G.; Hornsey, D.; Munro, F. Mill Scale Implementation of Papricans Process for
Polysulfide Liquor Production in Kraft Mill Causticizers. Part-1 Batch Trials and Optimization; TAPPI J. 1999,
82(10): 120.
lvi
) Yamaguchi, A. Operating Experiences with the MOXY Process and Quinoid Compounds; In Anthraquinone
Pulping: Anthology of Published Papers 1977-1996; Goyal, G.C., Ed.; TAPPI Press: Atlanta, GA, 1997; pp 287-291.
lvii
)
Nishijima, H.; et al., Review of PS/AQ Pulping to Date in Japanese Kraft Mills and the Impact on Productivity;
Proc. 1995 TAPPI Pulping Conference; TAPPI Press: Atlanta, GA, 1995; 31-40.
lviii
) MacLeod, M.; Radiotis, T.; Uloth, V.; Munro, F.; Tench, L. Basket Cases IV: Higher Yield with Paprilox
Polysulfide-AQ Pulping of Hardwoods; TAPPI J. 2002, 1(10): 3-8.
lix
) Olm, L.; Tormund, D.; Bernor Gidert E. Possibilities to Increase the Pulp Yield in a Kraft Cook of [the] ITCType; Proc. Breaking the yield barrier symposium; TAPPI Press: Atlanta, GA, 1998; Vol. 1, 69-78.
lx
) Hakanen, A.; Teder, A. Modified Kraft Cooking with Polysulfide: Yield, Viscosity, and Physical Properties;
TAPPI J. 1997, 80(7): pp 86,93,100, 189-196.
lxi
) Jiang, J.E.; Crofut, K.R.; Jones, D.B. Polysulfide Pulping of Southern Hardwood Employing the MOXYRG and
Green-Liquor Crystallization Processes; Proc. 1994 TAPPI Pulping Conference; TAPPI Press: Atlanta, GA, 1994; pp
799-806.
lxii
) Gustafsson, R.; Ek, M.; Teder, A. Polysulphide Pretreatment of Softwood for Increased Delignification and
Higher Pulp Viscosity; JPPS. 2004, 30(5): 129-135.
lxiii
) Mao, B. F.; Hartler, N.; Improved Modified Kraft Cooking,3: Modified Vapor-Phase Polysulfide Cooking; TAPPI
J. 1994, 77(11): 149-153.
lxiv
) Lindstrom, M.; Teder, A. Effect of Polysulfide Pretreatment When Kraft Pulping to Very Low Kappa Number;
Nordic Pulp Paper Res. J. 1995, 10(1): 8-11.
lxv
) Olm, L.; Tormund, D. ZAP Cooking- Increase Yield in PS and PS-AQ Cooking with Zero Effective Alkali in the
Pretreatment Stage; Nordic Pulp Paper Res. J. 2004, 19(1): 6.
lxvi
) Brannvall, E.; Gustafsson, R.; Teder, A.; Properties of Hyperalkaline Polysulphide Pulps; Nordic Pulp Paper
Res. J. 2003, 18(4): 436-440.
lxvii
) Lindstrom, M.; Naithani, V.; Kirkman, A.; Jameel, H. The Effect of Integrating Polysulfide Pulping and Black
Liquor Gasification on Pulp Yield and Properties; Proc. 2005 TAPPI Engineering, Pulping & Environmental
Conference; TAPPI Press: Atlanta, GA, 2005.
lxviii
)
lxix
) Metsoautomation website: http://www.metsoautomation.com/automation/index.nsf/FR?
ReadForm&ATL=/Automation/pp_prod.nsf/WebWID/WTB-061205-2256F-8ECE6
Edwards, L.L., et al. GEMS Users Manual, University of Idaho, Moscow, ID, 1972, 1977, 1983, 1989.