J-FOR - PAPTAC · J-FOR FOR THE ADVANCEMENT OF THE FOREST INDUSTRY A PAPTAC JOURNAL TABLE OF...

J-FORA PAPTAC JOURNAL

JOURNAL OF SCIENCE & TECHNOLOGY FOR FOREST PRODUCTS AND PROCESSES VOL. 2 , NO. 2, 2012

FOR THE ADVANCEMENT OF THE FOREST INDUSTRY

FEATURING:

Spotlight on the Canadian FIBRE Network Initiative

A wide range of papers from bleaching, co-generation and chemical recovery to sludge characterization and paper machine

www.paptac.ca

Building for the NewPulp and Paper

Community

www.paptac.ca

Join your Canadian technical community and experience new ideas

CONNECTING PEOPLE PAPTAC’s Technical Communities play an essential role in exchanging information on a variety of issues relatedto operation optimization, energy, environment, processand much more. E-mail discussion groups, on-line forums,conferences: a wealth of information accessible to allPAPTAC members.

WHY JOIN?

• Sharing information on specific topics & challenges facing the Canadian pulp and paper industry.• Accessing an exclusive Canadian technical pulp and paper network.• Continuing to learn from your peers, identifying and developping new problem-solving solutions.• Being aware of the latest technological advancements and innovations.• Greater value derived from participating in PAPTAC events (PaperWeek, PACWEST, conferences, webinars, etc.)

To learn more about the Technical Communities, visit the Technical Communities Section on www.paptac.ca or contact Thomas Perichaud at 514.392.6956 or [email protected]

PAPTAC COMMUNITIES

www.paptac.ca

Microsites and On-line Forums have been recently launched for: . Energy . Maintenance . Paper Machine Technology . Paperboard Packaging . Process Control

NEW

Others to come shortly!

PAPTA

C

Alkaline Pulping

Energy

Standard Methods

Bleaching

Research

Environment

Biorefining

Maintenance

Mechanical Pulping

Paperboard Packaging

Steam & Steam Power

Process Control

Paper Machine Technology

Building for the NewPulp and Paper

Community

FOR THE ADVANCEMENT OF THE FOREST INDUSTRYJ-FORA PAPTAC JOURNAL

TABLE OF CONTENTSTECHNICAL PAPERS

30 Comparison of water removal around the forming roll with refined and unrefined furnishes.

29 Creping blade unit. Creping unit in operation.

Ultrasound-assisted Hydrogen Peroxide Bleaching of a Softwood Thermomechanical Pulp by Eric Loranger, Marie Cantagrel, Céline Leduc, Claude Daneault

The Fate of Vanadium after Being Burned with Petroleum Coke in Lime Kilnsby Xiaofei Fan, Honghi Tran, Chris Dietel

The Use of Paper Mill Biotreatment Residue as Furnish or as a Bonding Agent in the Manufacture of Fibre-based Boardsby Adil Zerhouni, Talat Mahmood, Ahmed Koubaa

Development of Pilot Tissue Machineat FPInnovationsby Jimmy Jong, Francis Fournier, Stephan Larivière

Benchmarking Paper MachineInfluence on Linting and Pilingby Joseph Aspler, Jimmy Jong, Tony Manfred

Adding a Biomass-fired Cogeneration Power Plantto a Natural Gas Processing Plantby Derek McCann

6

13

19

25

32

40

4 EDITORIAL

Greg M. HayPAPTAC Executive Director

The FIBRE Network: An Exciting CanadianInititative for Forest Sector Transformation

Steve Goldstein Production SupervisorNorampac - Mississauga

With over 30 years of experience in the industry, Steve completed studies at York University in Toronto ComputerScience in 1982.

24th ISO/TC6, SC2 and SC5 Plenary and WorkingGroup Meetings will be hosted by CanadaNanotechnology Leaders Met in MontrealA Successful 2012 PACWEST ConferencePAPTAC Launches New Microsites and On-lineForums for Technical Communities

5 PAPTAC NEWS UPDATE

5 MILL CHAMPION’S PROFILE

For enquiries, please contact:PAPTAC740 Notre-Dame St. W., suite 1070Montreal (Quebec) H3C 3X6CANADAPhone: (514) 392-0265

PAPTACPulp and Paper Technical Association of Canada

PUBLISHER:

A Word from the Editor-in-Chief

Honghi Tran, PhD, P. Eng.,Editor-in-Chief

J-FOR Journal of Science & Technology for Forest Products and Processes: VOL.2, NO.2, 2012 3

4 J-FOR Journal of Science & Technology for Forest Products and Processes: VOL.2, NO.2, 2012


J-FOR: Bridging research and industry. Leading with

science and technology

EDITORIAL

The FIBRE Network (Forest Innovation by Research and Educa-tion) was launched earlier this year, leading an important initiative for the advancement of the forest sector through transformation. Supported by four solid organizations, it includes 8 R&D networks in specific fields and offers a unique platform of knowledge-shar-ing, innovation and synergies through a solid partnership of net-works working toward a common goal: developing the key compo-nents to draw new outputs, products and markets from the forestry sector. With the priority research areas being defined as energy and chemicals from forest biomass, integrated value maximization,

After four successful J-FOR issues covering many topics related to emerging science and technology, including a special issue dedicated to Forest Biorefinery, we are pleased to offer our readers a different approach for this issue. The present issue covers a wide range of traditional topics: from bleaching to chemical recovery, co-generation, sludge characteriza-tion, and paper machine, addressing key issues and challenges with day-to-day operations.

The FIBRE Network: An Exciting Canadian Initiative for Forest Sector Transformation

next generation building solutions, next generation pulps and papers and novel bioprod-ucts from forest biomass, the individual R&D Networks comprised in FIBRE are: For-ValueNet Network, Value Chain Optimization Network, Bioconversion Network, Ligno-works, Sentinel Bioactive Paper Network, Newbuilds, ArboraNano, and the Innovative Green Wood Fibre Products Network.

The initiative is supported by four pillar organizations: FPInnovations: Among the world’s largest private, not-for-profit forest research centres with more than 600 employees spread across Canada and develops solutions through in-novation in the utilization of forest resources.NSERC (Natural Sciences and Engineering Research Council of Canada): Sup-ports and promotes discovery research and university students in their advanced studies, and fosters innovation by encouraging Canadian companies to participate and invest in research projects.Forest Products Association of Canada (FPAC): Represents the largest Canadian producers of forest products. The Association is the voice of the Canadian wood, pulp and paper producers in government, trade and environmental affairs.Natural Resources Canada: A department of the Government of Canada that works to enhance the responsible development and use of Canada’s natural resources and the competitiveness of Canada’s natural resources products.

It is with great enthusiasm that the industry welcomes FIBRE and we are pleased to an-nounce that PAPTAC will be hosting FIBRE Day in conjunction with PaperWeek Canada 2013 (www.paperweekcanada.ca). We look forward to seeing the progress and developments of this initiative as the many challenges that the industry has faced in recent years now makes way for new and exciting ideas & opportunities.

EDITORIALEditor-in-ChiefHonghi Tran, PhD, P. Eng.Frank Dottori Professor of Pulp & Paper EngineeringDirector - Pulp & Paper CentreUniversity of Toronto, Canada

Deputy EditorPaul R. Stuart, PhD, Eng.Professor - Chemical Engineering DepartmentChairholder - NSERC Design Engineering ChairEcole Polytechnique - Montreal, Canada

ASSOCIATE EDITORS

Thore BerntssonChalmers Institute of Technology (SWEDEN)Virginie ChambostEnVertis Inc. (CANADA)Christine ChiratGrenoble INP – Pagora (FRANCE)Jorge Luiz ColodetteFederal Univerisity of Viçosa (BRAZIL)Ron CrotoginoArboraNano (CANADA)Sophie D’AmoursUniversité Laval (CANADA)Robert DekkerLakehead University (CANADA) Gilles DorrisFPInnovations (CANADA)Paul EarlPaul Earl Consulting Inc. (CANADA)Martin FairbankNatural Resources Canada (CANADA)W. James FrederickTable Mountain Consulting (USA)Ramin FarnoodUniversity of Toronto (CANADA)Gil GarnierAustralian Pulp and Paper Institute (AUSTRALIA)Eldon GunnDalhousie University (CANADA)Ali HarlinVTT (FINLAND)Mikko HupaÅbo Akademi University (FINLAND)David McDonaldJDMcD Consulting Inc. (CANADA)Glen MurphyOrgeon State University (USA)Yonghao NiUniversity of New Brunswick (CANADA) Reyhaneh ShenassaMetso Power (USA)Trevor StuthridgeScion (NEW ZEALAND)

Traditional, and quite topical

Honghi TranEditor-in-chief

Greg M. HayExecutive Director

PAPTAC

J-FOR Journal of Science & Technology for Forest Products and Processes: VOL.2, NO.2, 2012 5


PAPTAC News�Update

Forum Participants: Jean Hamel, James Olson, Tom Johnstone, Andrew Casey, Wade Chute, Martin Pudlas with Pacwest Executives, Al Parsons, Bill Adams and Randy Reimer.

24th ISO/TC6, SC2 and SC5 Plenary and Working

Group Meetings will be hosted by Canada

Every 18 months, ISO/TC6 (ISO Technical Committee 6 on paper, board and pulps) holds a week of meetings to discuss and re-solve technical issues related to the Standards development, and develop new ISO standards and review existing ones. The main ob-jective is to ensure that ISO Stan-dards comply with WTO (World Trade Organization) guidelines for global relevance and, in particular, that they have no adverse effects on fair international trade. The next meeting of TC6 will be hosted by Canada from October 15-19, 2012 and will be taking place at FPInnovations in Pointe-Claire, QC. These meetings are typically attended by about 100 scientific/technical experts coming from 15 to 20 different countries. There are currently 177 standards under TC6 covering all aspects of pulp, paper, tissue and board testing in areas such as environmental, chemical, optical, and physical properties. PAPTAC and its Standard Methods Committee are pleased to actively participate in this important meet-ing and platform for Standards de-velopment.

Nanotechnology Leaders Met in Montreal

The International Conference on Nanotechnology for Renewable Materials, took place place June 4-7, 2012 in Montreal, QC Cana-da, featuring a tour of CelluForce, the world’s first NanoCrystalline

Cellulose (NCC) plant in Windsor, QC and a series of keynotes and presentations, bringing together over 210 experts and leaders in Nanotechnology. Collaborating with TAPPI for the presentation of this conference in Montreal, PAP-TAC will be publishing a special is-sue of J-FOR later this year featur-ing some of the bast papers that were presented at the conference.

A Successful 2012 PACWEST Conference

With close to 300 in attendance, the 2012 edition of the PACWEST Conference, under the theme “Sustainability through People and Technology” was a great success. Taking place May 30 - June 2 in Jasper AB, the four-day conference featured a full program of Meet-ings; Technical Sessions; Panel Dis-cussions; Roundtables and Short Courses. Visit www.pacwestcon.net for details and future updates.

PAPTAC Launches New Microsites and On-line Forums for Technical

Communities

PAPTAC is pleased to announce the official launch of the first 5 microsites dedicated to Technical

Communities. Others still under construction will follow. The Paper Machine Technology, Energy, Maintenance, Paperboard Packaging and Process Control Communities will have the oppor-tunity to improve networking and communication among members through this new platform of infor-mation exchange. Each site offers an in-depth look at the commu-nities, their objectives and what they do.

Technical On-line Forumsfor Microsites

PAPTAC launched 5 on-line tech-nical forums with the microsites. These forums are part of the ben-efits of participating in the PAPTAC Technical Communities and have been added to the 5 new TC Web-sites that were launched recently. Members of the Energy, Mainte-nance, Paper Machine Technology, Paperboard Packaging and Process Control Communities now have the opportunity to ask specific questions and post/answer topics addressing challenges and oppor-tunities related to their operations and field of expertise by connect-ing to the forum with their user-name and password. As PAPTAC develops Microsites for its other communities in the com-ing months, forums for these com-munities will also become avail-able. Members who have not yet joined a Community are welcomed to visit the microsites and register as part of their member benefits, and start exchanging new ideas and opinions! www.paptac.caErratum: Please note that there was an erratum in the previous issue (Vol. 2, No. 1) on p. 41. The Equation in column 2, line 44, should read: CO + O2 → CO2 + O We apologize for any inconve-nience this may have caused.

With over 30 years of experience in the industry, Steve completed studies at York University in To-ronto Computer Science in 1982. He began working in the pulp and paper industry (formerly Domtar) in 1982 as production operator and was promoted to Production Supervisor in 1996. He Served as Technical Supervisor (production) one year, worked in Continuous Improvement for two years and in R&D for one year. Steve earned a technical certificate from Sault College in 2001 and served as Instructor for Cascades Training Centre teaching Pulp and Paper Science for new entrants into the paper industry. Involved in several external mill operations projects for optimization (Toronto, Trenton, Versailles Ct.), Steve is currently employed by Norampac Inc. Mississauga Division as Pro-duction Supervisor.

Steve is currently the Chairman of the PAPTAC Paper Machine Tech-nology Community (PMTC) and has always benefitted strongly from actively participating in its activities which connect paper-makers and provide a platform to share “problems and solutions” through training and forums. In an economic climate where the papermaker faces challenges of high operating costs, narrow margins, competition for produc-tion resources, national economic effects and obstacles from strict government regulations, PAPTAC and the PMTC service the paper industry in ways which are prac-tical, feasible and conducive to improved efficiency with immedi-ate benefits in productivity and or product quality.

SteveGoldsteinProduction

Supervisor

Norampac -

Mississauga


ABST

RACT This research has evaluated the impact of ultrasound irradiation on the bleaching of TMP with hydrogen peroxide. More precisely, the influ-

ences of irradiation power, ultrasound frequency, and reaction time on pulp brightness have been examined. The results indicated that high brightness could be achieved with a dominant ultrasonic mechanical effect generated at low power (500 W) and low frequency (68 kHz) along with a relatively short bleaching time of 30 min. The reverse results occurred when higher power (e.g., 1000 W), higher frequency (e.g., 170 kHz), and longer reaction time (e.g., 90 min) were used because the chemical effect, which promotes cellulose and lignin degra-dation, became dominant under these conditions. Statistical analysis showed that up to 74.4% of the brightness response was accounted for by ultrasonic frequency, ultrasonic power, and reaction time with 95% confidence.

ERIC LORANGER*, MARIE CANTAGREL, CÉLINE LEDUC, CLAUDE DANEAULT

ULTRASOUND-ASSISTED HYDROGEN PEROXIDE BLEACHING OF A SOFTWOOD THERMOMECHANICAL PULP

Thermo-mechanical pulp (TMP) is used in a wide array of paper products, includ-ing some value-added paper grades. Be-cause TMP is a high-yield pulp (≈95%), its inclusion in papermaking can reduce production cost and improve printing properties, but it has relatively low bright-ness (50–60% ISO) due to its high lignin content. This low brightness could hinder its potential use in value-added grades un-less its brightness is greatly enhanced [1]. It is known that alkaline hydrogen per-oxide oxidation is the most effective and

practical means of improving the bright-ness of TMP [2]. As reported, a chromo-phore removal ceiling on TMP is always encountered when using hydrogen perox-ide oxidation [3–6].

It has been reported that hydrogen peroxide can be generated by water sonol-ysis [7]. The in-situ creation of hydrogen peroxide could be efficient and economi-cal for TMP bleaching. The mechanism of creating hydrogen peroxide by sonolysis can be briefly described as follows. Ultra-sound is acoustic energy at a frequency

higher than the upper normal human hearing limit (20 kHz). When the energy from acoustic waves travels in an aqueous medium, the average distance between the molecules that make up water will oscil-late from their mean position. Depending on the ultrasonic conditions (frequency, power, etc.), the gap will eventually be larger than the minimum molecular dis-tance, and the water molecules will break down. This resulting process is called cavitation, which, depending on the acous-tic energy intensity, will lead to stable or

INTRODUCTION

ERIC LORANGERLignocellulosic Materials Research Center, Université du Québec à Trois-Rivières, 3351 boul. des Forges, C.P. 500, Trois-Rivières QC Canada, G9A 5H7*Contact: [email protected]

MARIE CANTAGRELLignocellulosic Materials Research Center, Université du Québec à Trois-Rivières, 3351 boul. des Forges, C.P. 500, Trois-Rivières QC Canada, G9A 5H7

CÉLINE LEDUCLignocellulosic Materials Research Center, Université du Québec à Trois-Rivières, 3351 boul. des Forges, C.P. 500, Trois-Rivières QC Canada, G9A 5H7

CLAUDE DANEAULTCanada Research Chair in Value-added PaperLignocellulosic Materials Research Center, Université du Québec à Trois-Rivières, 3351 boul. des Forges, C.P. 500, Trois-Rivières QC Canada, G9A 5H7

7J-FOR Journal of Science & Technology for Forest Products and Processes: VOL.2, NO.2, 2012

TRADITIONAL AREA CONTRIBUTIONS

transient cavitation bubbles. Stable cavi-tation bubbles will oscillate in diameter for many acoustic cycles, while transient bubbles will rapidly enlarge to the point of collapse after a few energy cycles. The collapsing of transient bubbles creates a violent jet of matter and energy that is re-sponsible for most of ultrasound’s effects. At this critical point, the temperature rises (e.g., to 5000°K) and the pressure increas-es (e.g., to 500 atm) within the cavitation void, resulting in water sonolysis [8] and creating radical species as shown in Fig. 1.

As shown in Fig. 1, hydrogen peroxide is created from the reaction of two hydroxyl radicals. The in-situ creation of hydrogen peroxide could be useful for improving TMP brightness. Moreover, the jet of mat-ter created by the collapse of a cavitation bubble has significant mechanical effects on the statistical probability of contact between reagents, improving the velocity and efficiency of reaction [9,10]. Mistik et al. have successfully bleached cotton fibres by hydrogen peroxide under ultrasound waves [11]. They reported a higher final brightness value and some diminution of the reaction time required to achieve this maximum brightness. A literature survey indicated that there is, however, little in-formation available on the bleaching of wood fibres using ultrasound-assisted hy-drogen peroxide oxidation. The objective of this study is to evaluate the impact of ultrasound irradiation on the bleaching of

TMP with hydrogen peroxide. The result would be useful for determining the po-tential use of ultrasound technology in a pulp bleaching plant.

MATERIALS AND METHODS

Raw materials and chemicals A thermo-mechanical pulp (TMP) made mainly from spruce and balsam fir was ob-tained from an eastern Canadian pulp mill. It had a Canadian standard freeness (CSF) of approximately 100 mL. The commer-cial sequestration agent, diethylene tri-amine pentaacetic acid (DTPA), was also provided by the pulp mill. All other chem-icals used were of A.C.S. reagent grade. Hydrogen peroxide was supplied by Labo-ratoire MAT at a minimum concentration of 33%. Sodium hydroxide, potassium io-dide, ammonium molybdate tetrahydrate, sulphuric acid, sodium thiosulphate, and sodium metabisulphite were supplied by Fisher Scientific. Sodium silicate was sup-plied by Sigma-Aldrich. All chemical solu-tions were prepared with deionized water (conductivity < 0.8 µS cm-1 at 25°C) using a reverse osmosis system equipped with activated charcoal and a UV lamp.

Transition metal ion sequestration pre-treatment methodologyTo stabilize and protect the hydrogen per-oxide from catalytic decomposition by

transition metals ions like iron, manganese, and copper, a pulp pre-treatment using DTPA was conducted before bleaching [13]. The pre-treatment was performed by soaking the pulp sample in deionized water containing 0.2% (dry weight basis) of DTPA at 3% consistency for 15 min at 60°C. The pre-treated pulp was thickened to approximately 20% by filtration on a Reeve Angel grade 202 filter paper.

Bleaching methodologyHydrogen peroxide bleaching was con-ducted at 12% consistency and 70°C, which is a typical industrial condition. Each bleaching experiment (70 g o.d. weight pulp) was performed in a poly-ethylene bag which was the same size as the bottom of the ultrasonic bath (0.23 m²). To prepare the bleaching liquor, the following bleaching agents were added in consecutive order (in % of o.d. pulp weight): 3% sodium silicate (Na2SiO3), water, 2.36% sodium hydroxide (NaOH), water, 3% hydrogen peroxide (H2O2). Residual water was added to achieve the desired pulp consistency. The total alkali ratio, defined as the ratio of total hydrox-ide ions to hydrogen peroxide molar con-centration, was 0.9. The freshly prepared bleaching liquor was poured into the bag containing the pulp sample. The mixture was kneaded manually for 2 min before measuring the initial pH. The mixture was

Fig. 1 - Radical species produced by sonolysis of water.

Fig. 2 - Schematic diagram of the ultrasonic bath.


that of the blank obtained with bleaching times of 30 and 60 min, but was slightly better than that for 90 min. As the applied power became greater than 500 W, the final brightness decreased with increas-ing power and bleaching time. These re-sponses are due to the degradation effect of ultrasound at high power. Aliyu et al. [18] reported that ultrasound treatment increased the enzymatic degradation of

cellulose, while Gadhe et al. [19] attributed this phenomenon to the cleavage of lig-nin. With increased power, more lignin fragments are dissolved, which in turn react with the hydrogen peroxide, thus reducing the effective amount of hydro-gen peroxide available for the destruction of chromophoric groups. When cellulose degrades, by-products such as organic extractives or metal ions are produced

kneaded again for 30 sec and then heated in a microwave oven for approximately 1 min 30 sec at maximum power to raise the temperature of the mixture closer to that of the bath. The bag was then immersed in the ultrasonic bath under various ex-perimental conditions. At the end of the bleaching time, the bag was withdrawn from the bath, and the pulp suspension was diluted to 1% consistency and was then neutralized with sodium metabisul-phite to pH 5.5. The residual hydrogen peroxide in the undiluted bleaching liquor was measured by iodometric dosage [14].

Ultrasonic conditionsBleaching experiments were performed using various ultrasonic power levels (0, 250, 500, 750, and 1000 W) and two fre-quencies (68 and 170 kHz). The bleaching times were 30, 60, and 90 min. The ultra-sonic bath was maintained at 70°C by an integrated thermostatic heater. The exper-imental design was based on the fractional design presented in Table 1. There were 28 conditions plus 5 repetitions to estimate the experimental error.

MeasurementsTo measure the optical properties of the pulp, handsheets of 4 g basis weight were prepared according to the PAPTAC (Pulp and Paper Technical Association of Canada) standard test method [15] and were conditioned at 23°C and 50% rela-tive humidity according to [16]. The opti-cal properties (ISO brightness, L*, a*, b*) were measured by means of a Technidyne ColorTouch PC apparatus (model CTP-ISO). This instrument complies with the PAPTAC standard test method for mea-suring brightness [17].

RESULTS AND DISCUSSION

Brightness vs. ultrasonic conditionsThe bleaching results (ISO brightness) ob-tained at a frequency of 68 kHz are shown in Fig. 3. Note that two distinct trends appeared, with a threshold point at 500 W. When the power was less than 500 W, the brightness of the pulp was similar to

TABLE 1 Experimental design for ultrasonic bleaching.

*Statistical repetitions

Trial Time (min) Power (W) Frequency (kHz)112112*

3030

00

170170

111111*

3030

00

6868

152151

3030

10001000

17068

122121

3030

250250

17068

121*132

3030

250500

68170

131142

3030

500750

68170

141212

3060

7500

68170

211*252

6060

01000

68170

251222

6060

1000250

68170

221232

6060

250500

68170

231242

6060

500750

68170

241312

6090

7500

68170

311*352

9090

01000

68170

351322

9090

1000250

68170

321332

9090

250500

68170

331342

9090

500750

68170

341 90 750 68



or liberated and will also participate in the reaction process, resulting in unde-sired consumption of hydrogen peroxide. However, when the experimental variation observed (Fig. 3) was taken into account, the effect of 68-kHz ultrasound on the final brightness was not statistically sig-nificant, with the exception of the experi-ment conducted at 500 W with a bleaching time of 30 min. In the latter case, the final brightness obtained after 30 min was com-parable to that of the blank after 90 min, amounting to a significant reduction in bleaching time (i.e., 60 min). This finding needs to be elucidated in further investiga-tions. However, the present study seemed to support the hypothesis that ultrasound treatment can speed up bleaching reac-tions.

The ISO brightness values obtained at 170 kHz are presented in Fig. 4. Note that no maximum value in brightness was observed, in contrast to the results ob-tained in the experiments conducted at 68 kHz (Fig. 3). With the exception of 250 W at 60 and 90 min, the brightness of pulp significantly decreased with increasing power and bleaching time. The cellulose degradation and lignin fractionation men-tioned earlier are the most likely cause of such responses.

As seen in Figs. 3 and 4, the optimal bleaching condition under ultrasound irra-diation is obtained using 500 W at 68 kHz, while the worst condition is encountered with 1000 W at 170 kHz. These conditions will serve as a baseline for comparing the two frequencies used.

Properties vs. frequencies The effect of frequency on brightness is shown in Fig. 5, which illustrates that the brightness values for 30 and 60 min ex-hibited a similar trend. A slight increase in brightness was found with 500 W at 68 kHz and 30 min, while a common value was found for the other conditions. The common value of 71.5% ISO for a bleaching time of 30 min was not statisti-cally different from that for the blank at 30 min (control reference), but the value for 60 min (71.6% ISO) was significantly dif-ferent from the blank value at 60 min. The

phenomenon was amplified when a long reaction time of 90 min was used. Note that the experiments conducted with 500 W at 170 kHz and with 1000 W at 68 kHz gave a similar brightness of 71.7% ISO,

while those performed with 1000 W at 170 kHz yielded 70.9% ISO.

This comparison highlights the relationship among frequency, power, and final brightness. Compared to lower

Fig. 3 - ISO brightness obtained with a frequency of 68 kHz.

Fig. 4 - ISO brightness at a frequency of 170 kHz.


frequencies, higher frequencies have a higher sonochemical or radical produc-tion ability [20]. Mason et al. have clearly demonstrated that mechanical effects are

dominant at low frequency, while radical production is maximal at high frequency [21]. They also showed that there is an inverse dependence between mechanical

effects (low frequency) and chemical ef-fects (high frequency) and that both effects can coexist at mid-frequencies. The lowest final brightness was found when the ex-periments were conducted at high power (1000 W), high frequency (170 kHz), and long reaction time (90 min). However, such conditions produce the most radi-cals, indicating that in-situ production of hydrogen peroxide (Fig. 1) is counter-balanced and ultimately over-balanced by the increased degradation of lignin and cellulose. Hence, the highest brightness was achieved with low power (500 W), and low frequency (68 kHz), conditions which produce predominantly mechanical ef-fects. This finding is in agreement with the hypothesis that the statistical probability of contact between reagents is enhanced by the mechanical shear generated by ul-trasonic irradiation.

In the last analysis, the hypothesis of increased statistical probability of contact between reagents because of mechanical shear, which increases reaction velocity, is the most plausible, while in-situ hydrogen peroxide production was not found to be beneficial, as postulated in the study by Xing et al., due to the balance between hy-drogen peroxide production and lignin or cellulose degradation [22].

Figure 6 shows the hydrogen perox-ide residual in the spent bleaching liquor as a function of the base conditions. The quantity of residual peroxide was fairly stable for all conditions with a short re-action time of 30 min (higher statistical probability of contact between reagents by mechanical shear), but it decreased slightly when high power (1000 W) and high frequency (170 kHz) were used along with longer bleaching times (60 and 90 min). This means that the hydrogen per-oxide produced in the in-situ system was most likely consumed by oxidation of or-ganic elements, transition metals liberated during high-frequency ultrasonic treat-ment, or both.

The b* color coordinate of a pulp, which gives a good indication of alkaline reversion due to O2 production or alka-linity conditions, is directly related to the chromophore content. The b* values for

Fig. 5 - ISO brightness for 68 and 170 kHz at 0, 500, and 1000 W.

Fig. 6 - Hydrogen peroxide residual for 68 and 170 kHz at 0, 500, and 1000 W.



selected experimental conditions are sum-marized in Fig. 7.

Note that the pulps that had low brightness had high chromophore con-tent, and vice versa. Compared to the blank, chromophore content dropped by approximately 8%, while brightness was maintained for bleaching performed using low power (500 W), low ultrasound fre-quency (68 kHz), and short reaction time (30 min). In contrast, when high power (1000 W) and high frequency (170 kHz) were used along with long bleaching time (90 min), the chromophore content in-creased by 6%, decreasing the final bright-ness by 2%. These results suggest that the increase in chromophore content is direct-ly associated with the presence of radicals. It is well known that hydroxyl or carboxyl groups from cellulose or lignin can be oxi-dized to carbonyl chromophore groups. The radicals produced by ultrasound can, it is hypothesized, promote such oxidation processes. For a given brightness value, a pulp that has more chromophores will re-quire more hydrogen peroxide to achieve the same final value. Consequently, as ob-served in this study, for a fixed hydrogen peroxide addition rate, the chromophoric

groups created by ultrasound irradiation will reduce the brightness of pulp.

Although the exact mechanism relat-ing brightness to radical formation during oxidation remains to be clarified, it is clear that in ultrasound-assisted hydrogen per-oxide bleaching, low frequency (68 kHz) is more advantageous than high frequency (170 kHz).

Statistical analysis of the experimental designTo estimate the statistical significance of the data measured, an analysis was carried out using commercial software (SAS JMP 9.0). Table 2 summarizes the statistical rel-evance (95% confidence interval), in order of importance, of different variables or variable interactions on the brightness b* and the hydrogen peroxide residual.

From the results shown in Table 2, the output variables were successfully modeled up to 74.4% of the response. As can be seen in Figs. 5 to 7, the frequency does indeed influence the end result, while Table 2 confirms its statistical relevance. This study indicates that frequency is the most important variable attributable to ultrasonic conditions. The influence of

reaction time on the final output is a mix of ultrasonic effects and normal bleach-ing operation. It is therefore difficult to isolate the ultrasound effect. However, the effect of power or of any combination of power levels can be attributed to the pres-ence of ultrasonic radiation. Ultrasound is then responsible for the majority of the effects (positive or negative) on the output variables, and any variations observed are not due to chance.

Even if the brightness values seem to be close to each other, the statistical analy-sis has confirmed that they are significant-ly different at the 95% confidence level. The effect detected can then be modelled and validated as originating from the ex-perimental conditions studied.

CONCLUSIONS

The purpose of this work was to deter-mine the impact of ultrasound irradiation on the bleaching of TMP with hydrogen peroxide. Even if the increase in final brightness value is only slightly significant, the ultrasonic mechanical shear is of in-terest. Further work at lower consistency, thus increasing the mechanical effects for the same power and frequency by reduc-ing ultrasonic attenuation will be required to fully understand the phenomenon [22]. However, the followings are the major findings of this study:

1. The highest brightness is achieved with low power (500 W), low frequency (68 kHz), and short reaction time (30 min) because this ultrasonic condition produces predominantly mechanical effects.2. The mechanical effects of ultra-sonic irradiation, which increase the statistical probability of contact be-tween reagents, improve the reaction velocity.3. The experiments conducted at high power (1000 W), high frequency (170 kHz), and long bleaching time (90 min) yielded the lowest brightness be-cause of the formation of radicals in this environment. 4. Chromophore creation by radical Fig. 7 - b* color coordinates for 68 and 170 kHz at 0, 500, and 1000 W.


oxidation of lignin and cellulose deg-radation is responsible for the decrease in brightness.5. Differences between the output variables are statistically significant at the 95% confidence level, and up to 74.4% of the response can be ac-counted for by frequency, power, and time.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge finan-cial support from the Canada Research Chair in Value-added Paper and the Nat-ural Sciences and Engineering Research Council of Canada.

REFERENCES

Liu, Z., Ni, Y., Li, Z., Court, G, “Per-oxide Bleaching of Low-Freeness TMP”, Pulp and Paper Canada, 106:34-37 (2005).Liu, S., “Chemical Kinetics of Alka-line Peroxide Brightening of Mechan-ical Pulps”, Chemical Engineering Science, 58:2229-2244 (2003).Laperrière, L., Leduc, C., Daneault, C., Bédard, P., “Chip Properties Anal-ysis for Predicting Bleaching Agent Requirement for TMP Pulps”, Tappi Journal, 3:23-27 (2004).Leduc, C. and Daneault, C., “ Impact of Mechanical Pulp Fines on the Effi-ciency of Peroxide Bleaching of TMP Pulp”, Cellulose Chemistry and Tech-nology, 41:399-404 (2008).

1.

2.

3.

4.

sonics, 43:811-814 (2005).Goodson, J.M., “Ultrasonic Trans-ducer”, Crest Ultrasonic Corporation (USA), U.S. patent 5748566 (issued May 9, 1996).Lapierre, L., Bouchard, J., Berry, R.M., Van Lierop, B., “Chelation Prior to Hydrogen Peroxide Bleaching of Kraft Pulps: An Overview”, Journal of Pulp and Paper Science, 21: J268-J273 (1995).“Analysis of Peroxides”, PAPTAC Standard Testing Methods J:16P (2003).“Forming Handsheets for Optical Tests of Pulp (British Sheet Machine Method)”. PAPTAC Standard Testing Methods C.5 (2006).“Conditioning Pulp Handsheets, Pa-per, or Paperboard for Testing in a Standard Atmosphere”. PAPTAC Standard Testing Methods A.4 (2003).“Brightness of Pulp, Paper, and Pa-perboard”. PAPTAC Standard Test-ing Methods E.1 (1990).Aliyu, M. and Hepher, M.J., “Effects of Ultrasound Energy on Degrada-tion of Cellulose Material”, Ultrason-ics Sonochemistry, 7:265-268 (2000).Gadhe, J.B., Gupta, R.B., Elder, T., “Surface Modification of Lignocel-lulosic Fibers Using High-Frequen-cy Ultrasound”, Cellulose, 13:9-22 (2006).Gogate, P.R., Mujumdar, S., Pandit, A., “Large-Scale Sonochemical Reac-tors for Process Intensification: De-sign and Experimental Validation”, Journal of Chemical Technology & Biotechnology, 78:685-693 (2003).Mason, T.J., Cobley, A.J., Graves, J.E., Morgan, D., “New Evidence for the Inverse Dependence of Mechanical and Chemical Effects on the Fre-quency of Ultrasound”, Ultrasonics Sonochemistry, 18:226-230 (2011).Xing, M., Yao, S., Zhou, S.-K., Zhao, Q., Lin, J.-H., Pu, J.-W., “The Influ-ence of Ultrasonic Treatment on the Bleaching of CMP Revealed by Surface and Chemical Structural Analyses”, Bioresources, 5:1353-1365 (2010).

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

Leduc, C., Martel, J., Daneault, C., “Efficiency and Effluent Character-istics from Mg(OH)2-Based Peroxide Bleaching of High-Yield Pulps and Deinked Pulp”, Cellulose Chemistry and Technology, 44:271-276 (2010). Zhao, Y. and Deng, Y., “Improve-ment of Peroxide Bleaching Yield and Efficiency of TMP Using Glyoxal Crosslink Agents”, Industrial & Engi-neering Chemistry Research, 45:5813-5818 (2006). Santos, H.M., Lodeiro, C., Capelo-Martinez, J.-L., Ultrasound in Chem-istry: Analytical Applications, Wiley-VCH, Weinheim, Germany (2009).De la Rochebrochard d’Auzay, S., Blais, J.-F., Naffrechoux, E., “Com-parison of Characterization Methods in High Frequency Sonochemical Re-actors of Differing Configurations”, Ultrasonics Sonochemistry, 17:547-554 (2010).Mohod, A.V. and Gogate, P.R., “Ul-trasonic Degradation of Polymers: Effect of Operating Parameters and Intensification Using Additives for Carboxymethyl Cellulose (CMC) and Polyvinyl Alcohol (PVA)”, Ultrason-ics Sonochemistry, 18:727-734 (2011).Katohgi, M. and Togo, H., “Oxida-tively Sonochemical Dealkylation of Various N-Alkylsulphonamides to Free Sulphonamides and Aldehydes”, Tetrahedron, 57:7481-7486 (2001).Mistik, S.I. and Yükseloglu, S.M., “Hydrogen Peroxide Bleaching of Cotton in Ultrasonic Energy”, Ultra-

5.

6.

7.

8.

9.

10.

11.

TABLE 2 Results of the statistical analysis.

b*

Brightness

Signifi cant variables

TimeFrequencyPower • FrequencyPower

0.744

FrequencyPowerTime • PowerPower • FrequencyTimeFrequencyTimePowerPower • Frequency

Adjusted correlation coeffi cient (R²)

Hydrogenperoxideresidual

0.640

0.697

ABST

RACT



Petroleum coke (petcoke) has been burned at a number of kraft pulp mills in the southern United States as a partial substitute for natural gas and fuel oil in lime kilns. Because of the high vanadium (V) content of the petcoke, there has been some concern about the impact of V compounds on kiln and chemical recovery operations. A laboratory study has been performed to examine the fate of V in lime kilns and the chemical recovery cycle. The results suggest that V introduced with petcoke can react quickly with lime in the kiln to form calcium vana-dates, mostly 3CaO•V2O5. In the slaker and causticizers, calcium vanadates react with Na2CO3 in the green liquor to form sodium vanadate (NaVO3) and lime mud (CaCO3). Because of its high solubility, NaVO3 dissolves in the liquor circulating around the chemical recovery sys-tem. As with water-soluble chloride (Cl) and potassium (K), V accumulates in the liquor cycle, reaching a steady-state concentration that is linearly proportional to the rate of V input with petcoke and the rate of mill soda loss. For a typical kraft mill where petcoke with 1500 ppm V is burned in the lime kiln at a 50% substitution rate, the steady-state V concentrations are approximately 100 ppm in white liquor and 230 ppm in as-fired black liquor dry solids. V does not accumulate in lime mud if the mud is well washed and dewatered.

XIAOFEI FAN, HONGHI TRAN*, CHRIS DIETEL

THE FATE OF VANADIUM AFTER BEING BURNED WITH PETROLEUM COKE IN LIME KILNS

In kraft pulp mills, lime mud (mainly CaCO3) from the causticizing plant is washed, dewatered and calcined in a lime kiln to produce lime (CaO) for reuse in the causticizing process. The calcination pro-cess and the lime kiln operation require a large amount of heat, which is typically provided by burning natural gas or fuel oil. Because of high energy costs in recent years, many mills, particularly those in the southern United States, have been replac-ing their traditional fossil fuels with petro-leum coke (petcoke), a by-product from crude oil refining on the Gulf Coast.

Petcoke for lime kiln use is typi-cally in pulverized form so that it can be burned more efficiently and completely. Table 1 summarizes a typical composition of petcoke. It consists of mainly carbon (C), some sulphur (S), hydrogen (H), and small amounts of nitrogen (N), oxygen (O), and ash. The high C, S, and H con-tents of petcoke (> 95 wt%) give rise to its relatively high heating value, approxi-mately 35 MJ/kg (15 KBtu/lb). Figure 1 shows the main impurities in petcoke from

various sources. Of these, vanadium (V) is by far the most important, with an aver-age concentration of approximately 1160 ppm, but which may range from 350 to 2500 ppm [2,3].

At present, about twenty lime kilns in the United States burn petcoke. Of these, three have been doing so for more than 20 years [4]. Despite the economic advantage, the high concentrations of S and V in pet-coke have been of concern for mills [2,4].

The fate of S introduced with pet-coke is relatively well known. In the lime

kiln, S is oxidized to SO2 and SO3, which subsequently react with lime to form calcium sulfate (CaSO4) [5]. In the causticizing plant, the resulting CaSO4 reacts with Na2CO3 in the green liquor to form CaCO3 (lime mud) and Na2SO4, which becomes part of the sulphate dead load that circulates in the liquor cycle until it is reduced in the recovery boiler to form Na2S. High S content in petcoke could reduce lime availability, increase SO2 emissions from the kiln stack, alter the S balance, increase liquor

INTRODUCTION

CHRIS DIETELDTE Energy Services / DTE PetCoke, LLCAlpharetta, GA, USA, 30004

*Contact: [email protected]

XIAOFEI FAN Pulp & Paper CentreUniversity of TorontoToronto ON, Canada

HONGHI TRANPulp & Paper CentreUniversity of TorontoToronto ON, Canada


sulphidity, and potentially contribute to ring formation in the kiln [5].

The fate of V introduced with pet-coke, on the other hand, is not well un-derstood, although there have been ex-tensive studies of the thermal behaviours of V-containing fuels and their ashes in other combustion systems. Bacci et al., for example, found that V was enriched in particulates collected from the stack of an oil-fired boiler and that NaVO3 was the dominant vanadate in the ash [6]. In boil-ers where high S-containing crude oil was burned, vanadyl sulphate (VO•SO4•xH2O) was found to be the major vanadium spe-cies in particulates [7]. Chen and Lu re-ported that in a circulating fluidized-bed reactor that fired petcoke and used lime-stone to minimize SO2 emissions, all the vanadium from the petcoke remained in the ash and was equally distributed be-tween the bottom ash and the fly ash [8]. A similar study by Jia et al. showed that di-calcium vanadate (2CaO•V2O5) was the main species in the ash [9].

In lime kilns where calcium com-pounds (CaO, CaCO3, CaSO4, etc.) are abundant and the temperature is high, it is not known what happens to V after it has been burned with petcoke, how quick-ly it can react with lime, and what impact it may have on the kiln and chemical re-covery operations. A laboratory study was therefore performed to examine possible chemical reactions involving V in the lime kiln and causticizing plant environments. This paper first describes the experimental

procedures and the results obtained and then postulates the fate of V in the chemi-cal recovery system after it has been intro-duced with petcoke to the lime kiln.

EXPERIMENTAL PROCEDURES

SamplesTwo samples of pulverized (-200 mesh) petcoke were obtained from a petcoke supplier. The samples were dry and con-tained 6.5 wt% S and 1500 ppm V. Lime mud samples were obtained from three different kraft pulp mills, and their com-positions are shown in Table 2. Analytical grade Na2CO3, CaCO3, and V2O5 with a purity of >99.9% on a dry basis were also used.

Thermal AnalysisThe thermal behaviours of petcoke, lime mud, and their mixtures were studied us-ing a thermo-gravimetric analysis/ differ-ential scanning calorimetry (TGA/DSC) analyzer. The equipment continuously measured the sample weight and the heat

flow to and from the sample simultane-ously as it was heated to 950°C in air at a rate of 20°C/min.

Heat Treatment ExperimentsHeat treatment experiments were per-formed to investigate possible reactions between vanadium (V) in petcoke and lime. Mixtures of petcoke and lime mud were first prepared in various ratios. In each experiment, these mixtures were housed in three separate alumina crucibles that were placed in a mullite tube mount-ed horizontally in the center of a tubular furnace, as shown in Fig. 2. Compressed air was passed through the mullite tube from one end at a constant flow rate of 200 mL/min to flush out the combustion gas from the petcoke and the CO2 released from the lime mud. The exhaust gas was bubbled through two caustic scrubbers in series containing 0.01N NaOH solution to remove gaseous S and V compounds, par-ticulates, and condensed matter before it was vented to atmosphere through a fume hood. In all experiments, the temperature in the mullite tube was controlled at 950°C and the reaction time was 5 hours.

Causticizing Experiments The residue from the heat treatment ex-periments described above was essentially a mixture of ash resulting from the com-bustion of petcoke and lime from the calcination of lime mud. This was used as a source of lime in the causticizing experiment. The residue was mixed with an aqueous solution of 200 g/L sodium carbonate (Na2CO3) in a sealed PFA (per-fluoroalkoxy) container placed in a water bath controlled at 90°C. The slurry was constantly agitated for 2 hours and then filtered. The filtered solids were complete-ly dried at 110°C.

Experiments with Pure CompoundsBecause the amount of V in the petcoke was much smaller than that of Ca in the lime mud, it was difficult to identify and to analyze accurately the possible vanadium compounds formed in the samples during the heat treatment and causticizing experi-ments. In this study, both heat treatment

Fig. 1 - Impurities in petcoke [2].

TABLE 1 Typical composition of petcoke [1]

C, wt% ds

Average86

S, wt% ds

H, wt% ds

N, wt% ds

O, wt% ds

HHV, MJ/kg

Range

Volatiles, wt% ds

Ash, wt% ds

5.5

3.6

1.8

81-892-7

2-5

1-41.7 0-3

111.2

5-160-6

35 33-35



and causticizing experiments were also performed on pure CaCO3 and on two mixtures of analytical-grade vanadium pentoxide (V2O5) and CaCO3 contain-ing 10 wt% and 20 wt% V2O5. The large amount, high purity, and high V content of the mixtures made the analysis and identification of the reaction products easier and more accurate.

Analytical TechniquesThe residues from the heat treatment ex-periments and the filtered solids from the causticizing experiments were analyzed using inductively coupled plasma mass spectrometry (ICP-MS) and X-ray fluores-cence spectroscopy (XRF). The scrubber solutions and the filtrates from these tests were collected and analyzed for Ca, V, Na, K, Ni, Fe, and other elements by means of inductively coupled plasma-atomic

emission spectroscopy (ICP-AES).


Thermal Behaviour of Petcoke Figure 3 shows the TGA profiles of the two petcoke samples used in this study. Both samples exhibited similar ther-mal behaviour. They were stable with no weight change up to 300oC and then showed a slight weight gain of approxi-mately 2 wt% between 300°C and 400°C. This weight gain was presumably due to the oxidization of S to SO2/SO3 and the subsequent absorption of some of these sulphur gases by other minerals in the pet-coke to form sulphates (NiSO4, CaSO4, FeSO4, etc.). Above 400°C, the samples burned and lost weight rapidly. At 650°C, only 0.4% of the original weight, which is essentially the ash content of petcoke, remained. This result suggests that the

petcoke used in this study has low ash content and can burn readily in air.

Thermal Behaviour of Vanadium Pentoxide (V2O5)Figure 4 shows the TGA/DSC results in air for the pure V2O5 used in this study. The sample contained 0.5% moisture and melted at approximately 670°C. The rela-tively constant weight of the sample at temperatures below 670°C and the rapid decrease in weight above 670°C suggest that V2O5 is stable at temperatures be-low its melting temperature but becomes

TABLE 2 Composition of lime mud from kraft pulp mills.

CaCO3

Mill A Mudwt%

95.1

MgO

P2O5

Na2O

SiO2

Others

Al2O3

SO3

1.191.38

0.86

95.9

1.321.45

0.820.35 0.15

0.150.55

0.150.11

0.18 0.10

Mill B Mudwt%

Mill C Mudwt%

Fe2O3

TOTAL

0.23 0.03

100 100

96.3

2.050.62

0.560.14

0.030.12

0.13

0.05

100

Fig. 2 - Experimental setup for burning petcoke and lime mud mixtures.

Fig. 3 - TGA profiles of two petcoke samples.

Fig. 4 - TGA/DSC profile for vanadium pentoxide (V2O5) in air.

Fig. 5 - Weight loss profiles for pure CaCO3containing various amounts of V2O5.


volatile once molten.

Reactions between V2O5 and CaCO3Figure 5 shows the TGA profiles in air for pure CaCO3 and for mixtures of V2O5 and CaCO3. Pure CaCO3 (0 wt% V2O5) start-ed to decompose in air at about 650°C, eventually losing 44% of its weight as the temperature exceeded 830°C. Mixtures

containing 10 wt% and 20 wt% V2O5 abruptly lost 2.5% and 5% respectively of their weight at approximately 620°C. As shown in Fig. 6, this temperature is almost the same as the eutectic tempera-ture of V2O5 and CaO•V2O5 mixtures in the phase diagram of the V2O5-CaO system [10]. The weight losses at this temperature were close to the respective theoretical values of 2.42% and 4.84% cal-culated assuming that the reaction prod-uct was CaO•V2O5 (calcium vanadate), i.e.:

V2O5 + CaCO3 → CaO•V2O5 + CO2 (Reaction 1)

Figure 6 also shows that below 778°C, a mixture of CaO and V2O5 would react to form CaO•V2O5. This, along with the TGA/DSC results, suggests that V2O5 immediately reacts with CaCO3 to form CaO•V2O5 at its eutectic melting tempera-ture with CaO to form a thermally more stable compound, CaO•V2O5. The reaction proceeds further at higher temperatures and in the presence of a large amount of CaO, forming 2CaO•V2O5 and 3CaO•V2O5 according to the following reactions:

CaO•V2O5 + CaO → 2CaO•V2O5 (Reaction 2)

2CaO•V2O5 + CaO → 3CaO•V2O5 (Reaction 3)

Note that the CaO-to-V2O5 molar ratios of the mixtures containing 10 wt% and 20 wt% V2O5 are 16.4 and 7.3 respec-tively. These ratios are much greater than 3, meaning that there is much more CaO in the mixtures than is needed to form any of the three known calcium vanadates: CaO•V2O5, 2CaO•V2O5, and 3CaO•V2O5. These compounds are thermally more stable than V2O5, which explains the con-stant weight of the samples observed at temperatures above 830°C, as shown in Fig. 5. Of these compounds, 3CaO•V2O5 is the most likely to form due to the high CaO-to-V2O5 molar ratio of the mixtures.

X-Ray diffraction analysis performed on the residues confirms the presence of 3CaO•V2O5, along with unreacted CaO and a small amount of Ca(OH)2 which was formed as a result of the hydration of CaO by moisture in the air. No V2O5, CaO•V2O5, or 2CaO•V2O was detected. These results are consistent with those ob-tained by Slobodin et al. [11], who showed,

using TGA/DTA (differential thermal analysis), that 3CaO•V2O5 is the most stable compound of the three and that it is the final product of the heat treatment of CaO/V2O5 and CaCO3/V2O5 mixtures with 3:1 molar ratio.

Reactions Involving Vanadium in Petcoke and LimeFigure 7 shows the V content in the resi-due after each heat treatment experiment at 950°C for 5 hours as a function of the petcoke content in the petcoke-mud mix-ture before the experiment. For all three lime mud samples used in this study, the amount of V in the residue increased with increasing petcoke content in the mixture. No significant difference between mills was found, particularly for mixtures that contained less than 20 wt% petcoke.

Assuming that all V compounds in the petcoke stayed in the residue after the heat treatment, the amount of V in the res-idue can be calculated based on the V con-tent of the petcoke, the petcoke content in the petcoke-mud mixture, the amount of the mixture used, and the amount of residue obtained. Figure 8 compares the calculated values to the measured V val-ues. The dotted line represents the case in which the actual value is exactly the same as the calculated value. The measured val-ues fall close to the dotted line, particularly when the V content in the residue is less than 0.2 wt%.

ICP-AES analysis of the solutions taken from two caustic scrubbers after the heat treatment failed to detect V in the solutions. These results suggest that V2O5 melts and becomes volatile at 670°C. In the presence of lime, however, it reacts quickly with lime at temperatures above 620°C to form calcium vanadates, which are stable and do not vaporize at tempera-tures up to at least 950°C.

Causticizing Na2CO3 Solution with V-containing LimeAs mentioned in the experimental proce-dure section, two V-containing lime sam-ples were prepared by heating mixtures of CaCO3 and V2O5 (10 wt% and 20 wt% V2O5) in a muffle furnace at 950°C for

Fig. 6 - The V2O5-CaO system [10].

Fig. 7 - Amounts of V in the residues obtained after heat Treatment experiments at 950°C for 5 hours.

Fig. 8 - Relationship between calculated and measured amounts of V in residues after heat treatment experiments at 950°Cfor 5 hours.



5 hours. These samples had V/Ca molar ratios of 0.122 and 0.275 respectively and were used to causticize a solution of 200 g/L Na2CO3 (equivalent TTA of 20 g/L Na2O) at 90°C for 2 hours.

Figure 9 compares the V/Ca mo-lar ratio of the lime samples before the causticizing reaction to that of the mud precipitated after the reaction. The V/Ca molar of the mud was almost the same in both cases, approximately 0.008, which is much lower than that of the lime be-fore the causticizing reaction. This low V content in the mud implies that most of the V2O5 that formed calcium vanadates with the lime mud must have reacted with Na2CO3 in solution to form water-soluble compounds and consequently must have been transferred to the causticized solu-tion. Because in these experiments the precipitated mud was obtained by vacuum filtering only, without washing with water, it is possible that a small amount of V might still remain in the residual solution in the mud.

IMPLICATIONS

Based on the results obtained from this study, the fate of V introduced with pet-coke to the lime kiln may be postulated. As shown in Fig. 10, as petcoke is burned in the kiln, V is oxidized to V2O5. The majority of the V2O5 thus formed is re-tained in the combustion residue (pet-coke ash) and becomes well mixed with

lime in the kiln bed. It then reacts with lime to form calcium vanadates, mostly 3CaO•V2O5, due to the abundance of lime and the high temperature near the kiln front end. A small amount of V2O5 may fail to take part in the formation of petcoke ash, remaining as vapour in the kiln gas. Some of this gaseous V2O5 may react with CaO to form calcium vanadates if it is in contact with the kiln bed, with suspended lime dust particles, or both; an-other portion may condense to form fine V2O5 dust particles as the temperature de-creases. As a result, the lime dust can be expected to contain calcium vanadates and unreacted V2O5.

For kilns that recycle dust directly from the electrostatic precipitator, the unreacted V2O5, if any, will be well mixed with the feed mud, will form thermally stable calcium vanadates in the kiln, and will exit the kiln with the lime. Therefore, no significant accumulation of V in the lime dust is expected. For kilns that use scrubbers to scrub lime dust before feed-ing it back to the kiln through a precoat filter, the unreacted V2O5 is expected to be washed off along with other water-soluble compounds and to become part of the scrubber solution.

In the slaker and causticizers, calci-um vanadates in the lime will react with Na2CO3 in the green liquor to form so-dium vanadates (NaVO3) and to precipi-tate CaCO3 (lime mud) according to the following reactions:

CaO•V2O5 + Na2CO3 → CaCO3 + 2 NaVO3 (Reaction 4)

2CaO•V2O5 + 2Na2CO3 + H2O → 2 CaCO3 + 2 NaOH + 2 NaVO3 (Reaction 5)

3CaO•V2O5 + 3Na2CO3 + 2 H2O → 3 CaCO3 + 4 NaOH + 2 NaVO3 (Reaction 6)

Because NaVO3 has high solubility in water, 255 g/L at 50°C [12], it dissolves in the liquor, leading to vanadium build-up in the liquor cycle. Moreover, because V follows the liquor, it does not accumu-late in the mud if the mud is well washed. The net effect is that V is transferred from petcoke to lime in the kiln and from lime to liquor in the causticizing plant. As pet-coke is continuously burned in the kiln, the concentration of V in the liquor in-creases. Eventually a steady-state concen-tration is reached at which the amount of V introduced with petcoke to the lime kiln is equal to the amount lost from the sys-tem with dregs, grits, mud disposal, liquor spills, etc.

As in the case of chloride (Cl) and potassium (K), the extent to which V can accumulate in the liquor cycle at a given mill depends strongly on the amount that is input to the system. Because the amount of V entering the system with wood and with sources other than petcoke is negli-gible, <1 g/ADT (air dry ton), the V input must come from petcoke. Therefore, the steady-state concentration of V in the li-quor depends on the amount of petcoke

Fig. 9 - V/Ca ratios of the lime product before the causticizing reaction and the precipitated material after the causticizing reaction.

Fig. 10 - The fate of V introduced by petcoke in the lime kiln.


burned in the kiln (percentage substitu-tion), the V content of the petcoke, and the degree of mill closure (total chemical loss or makeup).

Figure 11 shows the steady-state V concentrations in white and black liquors for a typical kraft pulp mill as a function of percentage petcoke substitution. Impor-tant assumptions made in this calculation are: the lime kiln burns natural gas with a heat rate of 7.2 GJ/t CaO; the petcoke contains 1500 ppm V; the Na inventory-to-white liquor ratio is 2.5; and the total Na loss (or makeup) is 14 kg Na2O/ADT.

The results suggest that the steady-state V concentration is linearly propor-tional to the percentage of petcoke sub-stitution. At a typical petcoke substitution rate of 50%, the steady-state V concentra-tion is approximately 100 ppm in white li-quor and 230 ppm in as-fired black liquor dry solids. This concentration is fairly low, only one-fifteenth to one-twentieth of the typical Cl concentration in the liquor.

SUMMARY

A laboratory study was performed to ex-amine the fate of V in lime kilns, causticiz-ing plants, evaporators, and recovery boil-ers. The results suggest that:

• V introduced with petcoke can react quickly with lime in the kiln to form calcium vanadates, mostly 3CaO•V2O5. • In the slaker and causticizers, the formed calcium vanadates react with Na2CO3 in the green liquor to form CaCO3, which is essentially lime mud, and NaVO3. • NaVO3 is highly soluble in water, meaning that once introduced into the system, V will circulate around the sys-tem with the liquor in the form of VO3-. • V does not accumulate in the lime mud if the mud is well washed and dewatered.• As with other water-soluble chloride (Cl) and potassium (K) com-pounds, V accumulates and reaches a steady-state concentration in the liquor that is linearly proportional to the rate of its input with petcoke and the rateof mill soda loss. For a typical kraft

mill where petcoke is burned in the lime kiln at a 50% substitution rate, the steady-state V concentration is ap-proximately 100 ppm in white liquor and 230 ppm in as-fired black liquor dry solids.

ACKNOWLEDGEMENTS

This work was part of the research pro-gram on “Alternative Fuels for Lime Kilns”, which was jointly supported by the Natural Sciences and Engineering Re-search Council of Canada (NSERC), in-dustrial members of the research consor-tium on “Increasing Energy and Chemical Recovery Efficiency in the Kraft Process”, and the following companies: Alberta Pa-cific, Alberta Research Council/Alberta Forestry Research Institute, DTE Petcoke, FPInnovations, Jammbco, Kiln Flame Systems.The material in this paper has been pre-sented at the TAPPI PEERS Conference, Norfork, VA, October, 2010.

REFERENCES

Commandré, J.M. and Salvador, S., “Lack of Correlation Between the Properties of a Petroleum Coke and its Behaviour During Combus-tion”, Fuel Processing Technology, 86(7):795–808 (2005).Angevine, P. and Parton, S., “Con-siderations in the Use of Petroleum Coke as a Supplemental Fuel in Lime Reburning Kilns”, Company Report, The Parton Group, Inc., Lilburn GA (2003).DTE Energy, “DTE PetCoke LLC – Material Safety Data Sheet”, accessed

1.

2.

3.

September 2009. Francey, S., Tran, H.N., Jones, A.K., “Current Status of Alternative Fuel Use in Lime Kilns”, TAPPI Journal, October, 33–39 (2009).Francey, S. and Tran, H.N., “Impact of Burning High-Sulphur Fuels in Lime Kilns”, Proceedings, TAPPI Engineering, Pulping and Environ-mental Conference, Memphis TN, October 11–14 (2009).Bacci, P., Del Monte, M., Longhetto, A., “Characterization of the Particu-late Emissions by a Large Oil-Fired Power Plant”, Journal of Aerosol Sci-ence, 14(3):252–253 (1983).Huffman, G.P., Huggins, F.E., Shah, N., Huggins, R., Linak, W.P., Mill-er, C.A., Pugmire, R.J., Meuzelaar, H.L.C., Seehra, M.S., Manivannan, A., “Characterization of Fine Particulate Matter Produced by Combustion of Residual Fuel Oil”, Journal of the Air and Waste Management Association, 50(7):1106–1114 (2000).Chen, J. and Lu, X., “Progress of Petroleum Coke Combusting in Cir-culating Fluidized-Bed Boilers: A Review and Future Perspectives”, Re-sources, Conservation, and Recycling, 49(3):203–216 (2007).Jia, L., Anthony, E.J., Charland, J.P., “Investigation of Vanadium Com-pounds in Ashes from a CFBC Firing 100% Petroleum Coke”, Energy and Fuels, 16(2):397–403 (2002).Morozov, A.N., Metallurg, 13(12) 24 (1938), in Levin, E.M., et al. (eds.), Phase Diagrams for Ceramists, p. 109, Fig. 251: CaO-V2O5 System, Ameri-can Ceramic Society, Columbus OH (1964).Slobodin, B.V., Zhilaev, V.A., Fotiev, A.A., Arapova, I.A., Tugova, N.P., “A Thermoanalytical Study of the Inter-action of Vanadium Oxide (V) with Calcium Oxide and Calcium Carbon-ate”, Journal of Thermal Analysis, 15(2):197–206 (1979).Trypuc, M. and Kiełkowska, U., “Sol-ubility in the NaVO3 + NH4VO3 + H2O System”, Journal of Chemical and Engineering Data, 42(3):523–525 (1997).

4.

5.

6.

7.

8.

9.

10.

11.

12.

Fig. 11 - Effect of petcoke substitution on the steady-state concentration of V in white liquor and as-fired black liquor.

ABST

RACT



Using paper mill effluent treatment residues as furnish or as a bonding agent for manufacturing fibre-based boards is one of the potential re-use alternatives for this material. However, thorough characterization of the material is a major requirement to develop this alternative. Key characteristics of primary sludge (PS) and secondary sludge (SS) from three pulping processes (TMP, CTMP, and Kraft) were assessed. Standard handsheets were made with PS and SS mixed in three proportions, dried at four temperature levels, and tested for various prop-erties. It was found that morphological and chemical properties varied with the pulping process and sludge type. PS had longer fibre and lower fines content than SS. Also, PS was higher in carbohydrates while SS exhibited higher lignin content. Handsheets made from TMP sludge showed the highest specific bond strength (SBS) as compared to those made with Kraft and CTMP sludge. SBS increased with SS content and drying temperature. Therefore, incorporating SS would likely improve the internal bond strength of the produced fibreboard.

ADIL ZERHOUNI, TALAT MAHMOOD*, AHMED KOUBAA

THE USE OF PAPER MILL BIOTREATMENT RESIDUE AS FURNISH OR AS A BONDING AGENT IN THE MANUFACTURE OF FIBRE-BASED BOARDS

Primary and secondary sludges are the main residues generated from pulp and paper mills. These residues result from effluent treatment. Primary sludge (PS) is composed mainly of waste fibres, fines, and fillers [1]. Secondary sludge (SS) is biological in nature and contains dispersed or flocculated micro-organisms, predomi-nantly bacterial cells, and other organic and inorganic materials [2].

Paper mill sludge is generally inciner-ated or landfilled. Incineration can require substantial supplemental fuel depending upon the feed rate and characteristics of the sludge and the combustion tempera-ture [3]. Landfilling is a very costly opera-tion and results in the emission of green-house gases such as CO2 and CH4 [4]. For these reasons, the number of landfills has been declining in recent years [5].

Sludge has been used as a soil amend-ment and soil fertiliser [6,7]. However, the low nitrogen and phosphorus content of PS [8], the contaminants contained in sludge [9], frequently high transportation costs [10], and future liability concerns are limiting factors for this application.

Several studies have investigated

additional alternatives for beneficial-use applications of sludge, including con-crete [11] and cement [12] manufactur-ing. However, such investigations did not demonstrate promising results. Geng et al. [13] evaluated the potential of sludge for medium-density fibreboard (MDF) manu-facturing. They showed that sludge is a low-cost fibre-rich material that adheres easily to wood fibres and could make it possible to reduce adhesive requirements in the manufacturing process.

Recently, a series of investigations was conducted by Migneault et al. to as-sess the potential of PS and SS from three

pulping processes for MDF manufacturing [14–16]. The performance of binderless MDF made with PS and SS from three different pulping process was investigated [15]. Results from this study indicated that the internal bond strength (IB) of MDF panels increased by up to 90% with increasing SS content (from 10% to 30%) of the panel mass, a result which demonstrated the adhesive properties of SS. The bonding mechanism of the sludge was attributed to proteins and increased lignin content on the fibre surface [15]. However, bonding efficiency could be decreased by fibre surface contamination.

INTRODUCTION

AHMED KOUBAAUniversité du Québec en Abitibi-Témiscamingue,Rouyn-Noranda, QCCanada


ADIL ZERHOUNIUniversité du Québec en Abitibi-Témiscamingue,Rouyn-Noranda, QCCanada

TALAT MAHMOODFPInnovations,Pointe-Claire, QCCanada


When used in combination with urea-formaldehyde (UF) resin, the SS enabled a substantial reduction in resin proportions, but its bonding effect was lower than ex-pected because of its high pH [14,16]. At 25% sludge content, all panels from all the investigated pulping processes met ANSI’s quality requirements for MDF used for in-terior applications [14,16].

The objective of this study was to evaluate the morphological, chemical, and bonding properties of primary and secondary sludge from three commercial pulp mills.

MATERIALS AND METHODS

Primary and secondary sludge samples were obtained from a TMP, a CTMP, and a Kraft mill. The three mills used the con-ventional activated sludge process for ef-fluent treatment. All samples were passed through a Somerville screen and thickened before making handsheets. It is notewor-thy that compared to pulp, the sludge handsheets had a dark color and a rough surface due to the presence of ash and other non-fibrous materials. For each pro-cess, a pulp sample was used as a control. The pulp samples were obtained from the same mill as the sludge sample. The CSF values for the sampled pulps were 160 mL, 420 mL and 690 mL for the TMP, CTMP, and Kraft pulps respectively. Pulp, PS, and SS from TMP, CTMP, and Kraft mills were used to make 60-g/m2 Standard handsheets (C1. PAPTAC Standard meth-od) at three different ratios: 100% PS; 80% PS–20% SS; and 60% PS–40 % SS.

Morphological properties of pulp, PS, SS, and handsheets, including fibre length and width and fines content, were measured using the OPTEST fibre qual-ity analyser (FQA). Fines retention was

measured indirectly by the difference be-tween the fines content of the pulp sus-pension before handsheet making and the fines content of the handsheets. The dried handsheets were then disintegrated and the fibre length and fines content mea-sured using the FQA. In all cases, fines retention was superior to 90%.

Cellulose, hemicellulose, and lignin contents were measured according to TAPPI T-203 om-83 and TAPPI T-222 om-88 standard methods. Ash content was measured according to PAPTAC G.10. The nitrogen content was estimat-ed on the handsheet surface using X-ray photoelectron spectroscopy (XPS). XPS was performed using an Axis Ultra HSA (Kratos Analytical, Ltd., U.K.). The X-ray source was a monochromatic Al with an 800 x 400 micron analysis area. Pressure during analysis was in the 10–8 torr range. Survey scans were recorded at 160eV pass energy and 1eV step size. The survey scans were used for elemental analysis and to calculate apparent element concentrations. Detailed high-resolution C1s spectra were recorded at 20eV pass energy and 0.025eV step size. Five repetitions were conducted for each XPS measurement.

To study the bonding mechanism of the sludge fibres as a function of tempera-ture, five drying temperatures were used: 23°C, 60°C, 100°C, 140°C, and 180°C. The 23°C temperature corresponds to the specifications of the C1 PAPTAC Stan-dard test. The 60°C value corresponds to the temperature used in paper drying. The 100°C and 140°C values correspond to the flow temperature of the hemicellulose and cellulose components of the fibre, and the 180°C value corresponds to the flow temperature of the lignin component of the fibre [17-20]. This last temperature corresponds also to the drying tempera-ture used in MDF hot pressing [14]. Dur-ing drying, the handsheets were pressed at 100 psi. For each temperature, the drying time (Table 1) to reach a final moisture content of approximately 8% was deter-mined experimentally. These drying times correspond to these reported previously for the same process [18,19]. Handsheets were then conditioned according to the

D1 PAPTAC standard method and tested according to TAPPI and PAPTAC Stan-dard methods. Properties measured in-cluded apparent density ( ), Z-directional tensile strength (ZDTS), and light-scatter-ing coefficient (LSC). The apparent den-sity ( ) and the light-scattering coefficient were used to evaluate the bonded area. The specific bond strength (SBS) was de-termined according to the following equa-tion [18,19]:


Morphological propertiesResults of fibre length and fines content variations as a function of pulping pro-cess (control) and effluent treatment are shown in Figs. 1 and 2 respectively. The fibre lengths of the Kraft and TMP pulp samples were greater than those of the CTMP samples. This variation is explained by the fact that the TMP and Kraft sam-ples were from softwood species, while the CTMP sample was from a hardwood species (white birch). It is interesting to note that the fibre-length variation in the PS and SS samples followed the same ten-dency as that in the control samples. The TMP and Kraft sludge samples contained longer fibres than the CTMP sample. In general, the PS samples had shorter fibres than the pulp samples. Similarly, the SS samples had even shorter fibres than the PS and the pulp samples (Figure 1).

The shorter average fibre length of the sludge samples, compared to the pulp samples, is due to the higher fine particu-late content of the sludge (Fig. 1). In gen-eral, the SS had a fine particulate content higher than that of pulp or PS, and the values for all three pulping processes were well above 60%. Similarly, the fines con-tent of the PS sample was much higher than that of the pulp samples. Biological activity in waste-water processes upstream of SS production could also affect the fibre length and fines content of sludge. The variation in fines content among the pulping processes can be explained main-ly by the damaging action on fibres of

TABLE 1 Drying times as a function of temperature.

24 hours23Drying temperature, °C Drying time

10060

180140

20 minutes10 minutes1 minute30 seconds



mechanical action during the pulping pro-cess.

Chemical propertiesLignin content - Results for total, soluble, and insoluble lignin contents are shown in Fig. 2. As expected, the TMP pulp showed the highest total lignin content, followed by the CTMP and Kraft pulps (Fig. 2a). The total lignin content of the Kraft pulp was negligible because the chemical pulping process removes lignin from the middle lamella of the fibre. Similar trends were apparent for the acid-insoluble (Fig. 2b) and acid-soluble (Fig. 2c) lignin con-tents.

The total lignin contents of the PS and SS sludge samples were higher than those of the pulps (Fig. 2a). Surprisingly, the total lignin content of the Kraft PS and SS was not much different from that of the TMP and CTMP sludges. Similar trends were apparent for the acid-insolu-ble (Fig. 2b) lignin contents of the PS and SS. For the PS, this result could be attrib-uted to the fact that Kraft PS is composed mainly of undefibred materials such as shives and chips. Lignin content was high-er in SS than in PS for the three pulping processes. This could be explained by the fact that the chemical tests detected lignin from the fibre cell wall and also microbial and chemical by-products such as polyphe-nols. These results are in good agreement with those reported by Migneault et al. [14].

Carbohydrate content - The total car-bohydrate content varied with the pulping process as expected. The chemical Kraft pulp had the highest carbohydrate content (Fig. 2d). The carbohydrate content of PS was lower than that of pulps, but did not vary substantially among the three pulp-ing processes. The carbohydrate content of the secondary sludge was very low (less than 5%) compared to that of PS and the three pulp types. It did not vary signifi-cantly among the three pulping processes. For the TMP and CTMP, most polysac-charides showed the same trends of varia-tion as the total carbohydrate contents (not shown). However, galactan, arabinan, and mannan contents were lower in the Kraft pulp than in the PS. The difference could be explained by the possibility that these polysaccharides might have been solubilized during the pulping process. Non-fibrous materials - Compared to pulp, the ash content in PS and SS from the three pulping processes was relatively high (Fig. 3a). Kraft PS showed the high-est ash content because of the chemicals used in the pulping process. The pres-ence of ash in the sludge from the three pulping processes could be explained by the various minerals used in the differ-ent paper production operations, which included filled and coated products. For the three pulping processes, the ash con-tent in SS was much higher than that

in PS. This result suggests that most min-erals are removed with secondary sludge [21]. However, this result is contradictory to that reported by Migneault et al. [14–15] who found higher ash contents in the PS. They suggested that some non-fibrous substances such as suspended solids were removed in primary treatment. Differ-ences in the manufacturing processes and effluent treatment systems are among the plausible explanations for this contradic-tion.

The nitrogen contents of pulp, PS, and SS from the three pulping processes are shown in Fig. 3b. The nitrogen con-tent of the three pulps and the PS from the three processes was negligible. In good agreement with previous findings [14,22], SS had high nitrogen content, suggest-ing the presence of proteins. Proteins are used in wood adhesive formulations [23]. Therefore, the presence of proteins in SS can be expected to have a positive effect on bonding.

Data from Figs. 2 and 3 suggest that a 100% accurate mass balance could not be developed, especially for PS and SS, because some of the sample components were not reported or not measured in this study. Previously reported data on extrac-tives in PS and SS suggest high contents compared to those in pulp or wood [14]. In addition, the chemical analysis methods used in this study were designed for woody substances, not sludge. Therefore, results

Fig. 1 - Variation of a) average fibre length and b) fines content of pulp, PS, and SS from three pulping processes.

a b


may have been interfered with by other molecules not found in wood, resulting in a mass balance different from 100%.

Bonding propertiesBonded area - The bonded area in a handsheet is closely related to the appar-ent density. An increase in SS proportion was associated with an increase in appar-ent density (Fig. 4a) and consequently with the bonded area due to the increase in fines content (Fig. 1b). Similarly, the light-scattering coefficient was also closely related to the bonded area in a handsheet. An increase in SS proportion decreased the light-scattering coefficient (Fig. 4b),

indicating an increase in the bonded area.

Bond strength - The ZDTS and the SBS were used to evaluate bond strength. Add-ing SS sludge to PS improved the ZDTS (Fig. 4c). This improvement could be par-tially attributed to an increase in the bond-ed area (Figs. 4a; 4b) and consequently to a higher frequency of hydrogen bonding. However, the specific bond strength im-proved as the proportion of SS increased (Fig. 4d). Therefore, hydrogen bonding was not the only mechanism responsible for the improvement in bond strength. Mechanical interlocking or the occurrence of chemical bonds are among the possible

hypotheses that could explain the im-provement in ZDTS and SBS. Proteins could also have played a role in this im-provement. Indeed, data from Fig. 3b show much higher nitrogen content in SS than in PS, suggesting the presence of proteins.

The SBS values for PS and SS from the TMP process are much higher than those from the CTMP and Kraft process-es (Fig. 4d), showing the better bonding ability of the former. An increase in the bonded area is not the only possible ex-planation for the higher bonding ability of TMP sludge fibres. Differences in chemi-cal composition are also among the plau-sible explanations. In fact, the TMP sludge fibres had higher lignin content (Fig. 2a) and lower carbohydrate contents than the CTMP and Kraft sludge fibres (Fig. 2d).

Impact of temperature on bond strength - Variations in drying tempera-ture have a significant impact on the flow and adhesion mechanisms of fibre com-ponents [17–19]. In general, at tempera-tures lower than 100°C, none of the fibre components will flow, judging from the glass transition temperatures of the fibre components, including carbohydrates. At temperatures between 100°C and 140°C, only the glass transition temperature of carbohydrates is exceeded, so that these components are expected to flow. At 180°C, the glass transition temperature of lignin is exceeded, and lignin is expected to flow. From these results, changes in sur-face properties could occur at drying tem-peratures between 100°C and 140°C and also at 180°C.

The variation in bond strength (ZDTS) as a function of drying tempera-ture and SS proportions is shown for the Kraft (Fig. 5a), CTMP (Fig. 5b), and TMP (Fig. 5c) processes respectively. For the sludge from all three processes, increas-ing the SS proportion improved ZDTS at all drying temperatures. This could be ex-plained partially by an increase in bonded area due to the higher fines content of SS and to the higher nitrogen content of SS as discussed above. For all three pro-cesses, increasing temperature from 25°C

Fig. 2 - Chemical composition of pulp, PS, and SS from three pulping processes: a) total lignin; b) acid-insoluble lignin; c) acid-soluble lignin, and d) carbohydrate content.

a

c

Fig. 3 - Variation in a) ash and b) nitrogen contents of pulp, PS, and SS from the three pulping processes.

a b

b

d



to 140°C slightly improved bond strength. However, the improvement at 180°C was substantial for the sludge from all three processes. The improvement is especially noticeable for samples containing higher

proportions of SS. This improvement could be explained by several factors, but mainly by the higher bonded area at higher drying temperatures and the consequently higher frequency of hydrogen bonds.

The variation in SBS as a function of drying temperature and SS proportions is shown for the Kraft (Fig. 6a), CTMP (Fig. 6b), and TMP (Fig. 6c) processes. The vari-ation in SBS with SS proportion and dry-ing temperature is similar to that of bond strength. Therefore, the improvement in bond strength could not be attributed only to a higher frequency of hydrogen bonds. Additional bonds such as other chemical bonds could also have been involved.

The impact of temperature on the specific bond strength at any SS propor-tion improved with drying temperature. This improvement was especially notice-able at 180°C and 40% SS. This tempera-ture is used in fibreboard manufacturing, including particleboard and MDF. There-fore, using sludge as a furnish for fibre-board can be expected to have an impact on its internal bond strength. This impact is beneficial and could make it possible to reduce the adhesive proportion in fibre-board manufacturing. Adhesives represent approximately 40% of the cost of fibre-board. Reducing the adhesive content in the fibreboard should reduce the overall cost of the manufacturing operation and

Fig. 4 - Effect of SS proportion on the a) apparent density; b) light-scattering coefficient; Z-directional tensile strength (ZDTS), and d) specific bond strength (SBS) of sludge handsheets.

a b

c d

Fig. 5 - Variation in Z-directional tensile strength (ZDTS) with drying temperature and SS proportion for a) Kraft; b) CTMP, and c) TMP sludge samples.

a b c

Fig. 6 - Variation in specific bond strength (SBS) with drying temperature and SS proportion for a) Kraft; b) CTMP, and c) TMP sludge samples.

a b c


improve its economic profitability. CONCLUSIONS

This study led to the following conclu-sions:

• Sludge morphological proper-ties varied with the pulping and efflu-ent treatment processes used. SS con-tained shorter fibres and more fines than PS.• The chemical composition of PS was substantially different from that of SS. Compared to pulp and PS, SS had lower carbohydrate content and higher lignin, ash, and nitrogen contents. • SS improves the bond strength between fibres. This improvement can be explained partially by an increase in the frequency of hydrogen bonding and a higher proportion of protein.• The combination of SS and high temperature led to a substantial in-crease in specific bond strength. This increase was attributed to the possible occurrence of chemical bonds be-tween SS and fibres. • Using sludge fibres as a furnish or as an adhesive for fibreboard manu-facturing could lead to reduced adhe-sive requirements and consequently reduced manufacturing costs.

ACKNOWLEDGEMENTS

The authors acknowledge the financial support of the Natural Sciences and En-gineering Research Council of Canada (NSERC), Le Fonds de Recherche sur la Na-ture et les Technologies du Québec (FQRNT) and FPInnovations. The authors also acknowledge the technical and scientific contribution of FPInnovations and three Anonymous pulp mills for supplying the sludge samples.

REFERENCES

Mahmood, T. and Elliott, A., “A Re-view of Secondary Sludge Reduc-tion Technologies for the Pulp and Paper Industry”, Water Research 40(11):2093-2112 (2006).

1.

741 (2006).Migneault, S., Koubaa, A., Nadji, H., Riedl, B., Zhang, S.Y., and Deng, J., “Medium-Density Fiberboard Pro-duced Using Pulp and Paper Sludge from Different Pulping Processes”, Wood Fiber Sci. 42(3):292-303 (2010).Migneault, S., Koubaa, A., Riedl, B., Nadji, H. Deng, J., and Zhang, S.Y., “Binderless Fiberboard Made from Primary and Secondary Pulp and Paper Sludge”, Wood Fiber Sci. 43(2):180-193 (2011).Migneault, S., Koubaa, A., Riedl, B., Nadji, H., and Deng, J., “Potential of Pulp and Paper Sludge as a Formal-dehyde Scavenger Agent in MDF Resins”, Holzforschung, 65:403-409 (2011).Back, E.L., “Thermal Auto-Cross-linking in Cellulose Material”, Pulp and Paper Mag. Can., 68:T165-T171 (1967).Koubaa, A., Amélioration de la résis-tance des liaisons dans le papier et le carton par raffinage et par pressage et séchage simultanés. PhD thesis, Uni-versité de Québec à Trois-Rivières (1996).Koubaa, A. Riedl, B., and Koran, Z., “Surface Analysis of Press-Dried CTMP Paper Samples by Electron Spectroscopy for Chemical Analysis”, J. of Applied Polymer Sci., 61:545-552 (1996).Strenberg, E.L., “Effect of Heat Treatment on the Internal Bonding of Kraft Liner”, Svensk Papperstidning, 81(2):49-54 (1978). Bassompierre, C., Procédé à boues activées pour le traitement d’effluents papetiers : de la conception d’un pi-lote à la validation de modèles. PhD thesis, Institut National Polytechnique de Grenoble, France (2007).Bitton, G., Wastewater Microbiology, 3rd ed., Wiley, Hoboken NJ. 746 pp. (2005).Pizzi, A. and Mittal, K.L., Handbook of Adhesive Technology, 2nd ed., Marcel Dekker, New York. 672 pp. (2003).

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

Metcalf and Eddy, Wastewater Engi-neering—Treatment and Reuse, 4th ed., McGraw-Hill, New York NY (2003).Elouazzani, D.C., Caractérisation physico-chimique et valorisation en bâtiment et travaux publics des cen-dres issues de l’incinération des boues de papeterie. PhD thesis, Institut na-tional des sciences appliquées de Lyon (2005).Corbun, R. and Dolan, G., “Beneficial Use of Paper Mill Sludge”, Biocycle, 36:69 (1995).Simons, H.A., Methodes de gestion des boues générées par les fabriques de pâtes et papiers. Rapport Synthèse, Mars (1994). Boni, M.R., D’Aprile, L., and De Casa, G., “Environmental Quality of Primary Paper Sludge”, Journal of Hazardous Materials, 108,(1-2):125-128 (2004).Beauchamp, C.J., Chrest, M.H., and Gosselin, A., “Examination of Envi-romental Quality of Raw and Com-posting Deinking Paper Sludge”, Che-mosphere, 46(6):887-895 (2002). Marsh, M., Murray, D., and Kleywegt, S., Field and laboratory bioassays on Atlantic packaging biosolids. Final Report, Phytotoxicology and Soil Standards Section, Standards Devel-opment Branch, Rep. SDB-051-3511 (1998). Priesnitz, W. “Sludge on your Supper Table”, Natural Life, 69 (1999). Pickell, J. and Wunderlich, R., “Sludge Disposal: Current Practices and Fu-ture Options”, Pulp and Paper Cana-da 96(9):41-47 (1995). Ahmadi, B. and Al-Khaja, W., “Utili-sation of Paper Waste Sludge in the Building Construction Industry”, Re-source Conservation and Recycling, 32:105-113 (2001).Amrouz, A., Transformation des boues de papeteries en pouzzolanes artificielles. PhD thesis, Lyon: INSA de Lyon (1996). Geng, X., Deng, J., and Zhang, S.Y., “Effects of Hot-Pressing Parameters and Wax Content on the Properties of Fiberboard Made from Paper Mill Sludge”. Wood Fiber Science, 38:736-

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

ABST

RACT



FPInnovations’ pilot paper machine running at commercial speeds was redesigned to add a tissue and towel production capability. The target tissue configuration is based on a twin-wire machine with C-wrap around a forming roll. The stock system can handle two different furnishes. A versatile multi-layer headbox, equipped with dilution and edge flow control, can provide even CD flow distribution. The tissue web is transferred from the former by a press felt to a press roll, a Yankee cylinder section, and a creping blade unit. A conveyor system transfers the fully-dried creped tissue to a dry-end reel. A series of pilot trials have demonstrated that the pilot machine produces high-quality tissue or towel which achieves the required tensile strength, stretch, bulk, softness, and absorbency properties. The machine can be used for various research and product development projects such as furnish optimization, equipment performance assessment, and evaluation of wet-dry strength chemistry, enzymes, fabrics, creping, and Yankee coatings.

JIMMY JONG, FRANCIS FOURNIER*, STEPHAN LARIVIÈRE

DEVELOPMENT OF PILOT TISSUE MACHINE AT FPINNOVATIONS

FPInnovations’ second-generation pilot paper machine, shown in Fig. 1(a), was completely redesigned and rebuilt in 1999 to accommodate new technology and re-search needs in the areas of papermaking and product performance. At the time of the rebuild, the main use of the pilot machine was the production of light- to medium-basis-weight printing and writ-ing grades. The overall design guideline of the pilot paper machine was described in detail by Crotogino et al. [1] The report emphasized the open concept of the ma-chine so that additional equipment, instru-mentation, wet-end chemistry systems, and white-water closure systems could be quickly installed to evaluate the effect of papermaking operations on process vari-ables and paper properties in the future. One of the important aspects of the ma-chine rebuild was the evaluation of the Pa-priDryTM drying concept [2]. Two Yankee dryers were installed in the pilot machine to evaluate their drying performance in relation to conventional cylinder drying when used for high-speed production of low-basis-weight publication grades.

In response to strong interest from tis-sue manufacturers and equipment suppliers

as well as the need for future business po-tential, FPInnovations, with a critical con-tribution from a tissue-producing mem-ber company, carried out a preliminary technical and economic feasibility study in 2007 to determine the tissue trial capa-bility of the pilot machine. Afterwards, it was decided to embark on an engineering project to convert the existing paper trial configuration to a tissue and towel trial configuration, as shown in Fig. 1. Avail-ability of the existing Yankee dryers on the pilot machine was the decisive factor for carrying out such a conversion. Fig-ure 1(a) shows an existing twin-wire roll blade gap former or impingement shoe former, followed by a tri-nip and a fourth

shoe press, and equipped with two Yankee dryers. The second Yankee dryer is inverted to enable the opposite side of the sheet to contact the hot smooth Yankee surface. Figure 1(b) shows the target tissue configuration in a twin-wire machine with C-wrap around a forming roll. Figure 1(c) shows how the pilot machine replicates the target tissue configuration. This was made possible by increasing the wrap angle around the forming roll, limiting the use of drainage elements in the former, completely by-passing the press section, adding creping blade units, and installing a conveyor-belt sheet transfer system over the second Yankee unit. Table 1 compares the technical

INTRODUCTION

STEPHAN LARIVIÈRE FPInnovations,Pointe-Claire, QCCanada


JIMMY JONGFPInnovations,Pointe-Claire, QCCanada

FRANCIS FOURNIERFPInnovations,Pointe-Claire, QCCanada* Now with Kruger Products.


specifications and operating range of the pilot machine in paper configuration and in tissue and towel configuration.

The pilot tissue machine can be used for research and development work in furnish development, chemistry optimiza-tion, process optimization, new product development, and new equipment demon-stration. This report describes the details of the pilot machine sections available in tissue and towel configuration. The report also presents some preliminary results on tissue operating conditions and their ef-fects on tissue properties.

Other Pilot Tissue Machines World-WideThere are several full-scale pilot tissue machines available in the world for re-search, product development, and cus-tomer demonstration. They belong to paper machine builders or tissue produc-ers. For instance, Metso has a pilot tissue machine at the Tissue Technology Cen-ter in Karlstad, Sweden, which has the capability to run conventional DCT (dry creped), STT (structured), NTT (new textured), and TAD (through air-dried) modes, as described in several references

3-7]. Voith has a pilot tissue machine avail-able in Sao Paulo, Brazil [8]. This machine is equipped with an ATMOS (advanced tissue moulding system), which uses a structured fabric to create a wet-shaped tissue structure with lower energy con-sumption than a TAD machine [9,10]. There are also some proprietary pilot tis-sue machines which are used exclusively by tissue producers, but no detailed in-formation is publicly available. Ashland-Hercules operates a pilot creping unit for evaluation of Yankee additives [11].

TABLE 1Technical specifi cations and operating range of the pilot machine in paper confi guration vs. tissue and towel confi guration.

Grades

Paper Confi guration Tissue and Towel Confi guration

Basis Weight

Speed

Stock Preparation

*Typical Furnish

Wet-end Additive System

Continuous Production

Headbox

HB Consistency

Forming

Pressing

Shoe Press

Drying

Creping

Softener & Spray Boom

ReelingPaper Width

On-line Control & Monitoring

* SW:Softwood, HW:Hardwood, TMP:Thermo-Mechanical Pulp, DIP:Deinked Pulp

Newsprint, LWC base sheet, highly fi lled sheet, uncoated wood-free (fi ne paper), board

30–180 g/m2

Design speed: 2500 m/minOperating speed: 600–1400 m/min

Tissue, paper towel

13–30 g/m2

Design speed: 2500 m/minOperating speed: 800–1200 m/min currently, expected to go above 1500 m/min

100 m3 stock reservoir, 2 fan pump lines, off-line refi ning, pressure screens

100 m3 stock reservoir, 2 fan pump lines, off-line refi ning, pressure screens

Re-slushed rolls, DIP, fresh bales of SW, HW, TMP, etc. SW Kraft, HW Kraft, eucalyptus, DIP, etc.

Various points for retention aid, starch, and fi ller injection

Wet, dry, and other strength agents at various points

6 h for newsprint, 4.5 h for fi ne paper with fi llers at 1000 m/min 12 h for tissue at 1000 m/min at former

1–3 layer dilution headbox 1–3 layer dilution headbox

0.6–1.5 % 0.1–0.3%Multi-impingement-shoe forming, roll forming, roll-blade forming with counter blades Twin-wire C-wrap roll forming

Tri-nip, Twinver, bi-nip with shoe press with/without open draw Single nip press on Yankee

Available with various loadings Not used

2 high-intensity dryers Cast-iron Yankee dryer (3-m diameter) with Infi nicoat

Not used Crepe doctor at various creping angles

Not used Coating and release chemicals

Wet-end and dry-end Dry-end30–36 cm (12–14”) 33 cm (13”)Machine drives, process controls, PI connection

Machine drives, process controls, PI connection



DESCRIPTION OF THE FPINNOVATIONS PILOT TISSUE MACHINE

Stock Preparation and Multi-layer HeadboxFigure 2 shows a simple diagram of the stock preparation system. There are three main stock reservoirs (50m3, 30m3, and 20m3) for a total of approximately 100 m3 stock capacity. Furnish can be prepared off-line from softwood or hardwood pulp bales using a low-consistency refiner (not shown in the diagram). A series of refiner plates is available to target a given free-ness level using low-, medium-, or high-intensity refining energy levels. The stock can also be prepared from re-slushed pulp, re-pulped paper rolls, or pulp suspension from a tissue mill. The stock is fed into a machine chest and is brought to a fan pump after being mixed with white water recirculating within the short circulation loop. Excess white water can be treated for fines recovery using a save-all or dis-solved-air flotation unit.

There are options to use one or two main fan pump lines. The secondary fan pump line with an extra machine chest can be used to handle two different pulp types simultaneously. The pulp stock goes through a pressure screen to eliminate large

contaminants and to prevent flocculation. Then it enters an air-padded radial dis-tributor, which distributes the stock under equal pressure across the full width of the headbox for each layer.

The headbox, supplied by Johnson-Foils in 2008, is a dilution-enabled three-layer hydraulic headbox equipped with step diffusors for turbulence generation. The width of the headbox is 44.77 cm (17.625”). There is no tapered header be-cause the radial flow distributor provides evenly distributed pulp stock across the full headbox width at each layer. Either one radial distributor can be used for the

entire headbox operation, or a secondary radial distributor can be used to inject an additional pulp stream.

Sheet Forming and TransferThe forming section consists of a suc-tion forming roll and a couch roll with no vacuum applied. Both rolls have the same diameter, 1.37 m. Two identical forming fabrics specially designed for tissue are in-stalled on the conveying and backing sides. The forming fabric width is 66 cm, and the width of the tissue sample after trim-ming is 33 cm.

The headbox is mounted in such a

Fig. 1 - Conversion of pilot paper machine to pilot tissue machine configuration.

Fig. 2 - Layout of stock preparation and multi-layer headbox with two fan pump lines.

Fig. 3 - Schematic of twin-wire tissue forming and transfer sections.


way that it can rotate from a vertical posi-tion to a horizontal position. For tissue or towel operation, the headbox is positioned horizontally to wrap the forming roll fully over 90 degrees, as shown in Fig. 3. In tis-sue operation, no vacuum is applied to the couch roll, which can also induce ad-ditional drainage under tension. The tissue web is trimmed after the couch roll. An inclined pick-up shoe under vacuum is lo-cated inside the press felt loop. As shown in Fig. 3, it can be rotated in the directions of the arrow to optimize the sheet transfer to the press felt from the conveying form-ing fabrics. Side trims from both edges are removed and recirculated back to the couch pit.

Pressing, Drying, Creping, and Spray BoomOnce the tissue web has been transferred by the press felt around a press roll, it is mechanically pressed between the roll and the Yankee cylinder. As shown in Fig. 4, the roll is a two-zone suction press roll with a diameter of 64 cm equipped with a roll cover. The press nip loading can go as high as 120 kN/m (685 lb/in), and press roll vacuum can be applied up to 65 kPa (19.5 inch of Hg). To achieve maximum water removal in the nip, triple doctor blades, supplied by GL&V, are applied against the press roll as shown. The water removed by each doctor blade is collected

into one of three separate compartments, which makes it possible to determine the amount of water removed for perfor-mance characterization. The press felt is 56 cm (22”) wide. A flooded nip shower is installed after the nip for felt cleaning.

Figure 5 provides an overview of the Yankee dryer section, which includes a press roll, a Yankee cylinder roll, two dryer hoods, and a creping unit. The Yankee cyl-inder consists of a 3-m-diameter (10 ft) and 0.76-m-wide (30 inch) cast-iron cylin-der. The sheet wraps at approximately 300 degrees around the Yankee cylinder roll. The outside surface is coated with Infini-cote, a hard material of metallurgic com-position. The Yankee cylinder is equipped with rotary joints, rotary siphons, and con-densate separators. It is rated to a pressure of 1,200 kPa (160 psi), but typically oper-ates between 375 kPa (50 psi) and 1012.5 kPa (135 psi). The differential pressure can go up to 112.5 kPa (15 psi) and the steam capacity up to 450 kg/h (1000 lb/h).

The cross-direction surface tem-perature of the Yankee cylinder tends to have a pronounced U-shaped profile, with both edges higher than the centre where the surface is covered by the wet sheet. To avoid a chattering effect, which results in a pattern of creping chemical buildup around the Yankee surface, edge show-ers can be used to reduce the temperature variation across the full width of the sheet.

The steam-heated Yankee roll is sur-rounded by two hot-air impingement dryer hoods. Air is heated by the combustion of natural gas and impinges on the tissue or towel surface from the dryer hoods. The air impingement velocity and temperature are adjusted according to the desired pro-duction rate. The system is designed for very high impingement velocity and sup-ply air temperature, 350 m/sec (70,000 ft/min) and 815˚C (1500˚F) respectively, al-though significantly less severe conditions are used for typical tissue production of 1000 m/min at Yankee speed. The con-densate is removed by rotary siphons. The split between steam and hot air contribu-tion for overall drying can range between 33% and 66%.

Photos of the creping blade unit are shown in Figs. 6(a) and 6(b). The pneu-matic loading system for the creping blade consists of air cylinders connected to the blade system journals. The blades can be pivoted away from the operating positions of the Yankee cylinder for blade change and cleaning. The creping unit offers a wide range of operating conditions. Blade angles can be set up to vary from 15 to 30 degrees. Because most tissue mills pre-fer to use blades with slightly different blade-tip angles to influence bulk, tensile strength, and softness, a series of blades with different tip angles (square edge to 25-degree angle) is normally used to

Fig. 4 - Press roll against Yankee surface with three doctor blades for water removal.

Fig. 5 - Overview of Yankee dryer: press roll, Yankee roll, two hoods, and creping blade unit.



optimize tissue quality in production. A number of parameters such as furnish types, chemical additives, creping ratio, press roll loadings, refining level, blade thickness, and machine operating condi-tions all influence the optimum blade-tip angle. It is important to optimize the crep-ing blade unit before each trial run. A cut-off cleaning doctor is installed on the Yan-kee cylinder roll after the creping blade unit. It is used to pick up fibre debris from the Yankee surface and to maintain opti-mum coating conditions on the surface.

The spray boom for coating chemi-cals is located under the Yankee cylinder in the area not covered by the sheet. The spray boom system is designed to cover the full width of the Yankee cylinder sur-face and to function under wide ranges of dwell time, spray distance, pressure, and temperature. Up to three different chemi-cals can be injected into the spray boom.

All flows are monitored and controlled from the main DCS.

Conveying to ReelingOnce the sheet is creped by the creping blade unit, it is fed into a plain turning roll and sandwiched between two conveying fabrics over the second Yankee, which is idle for tissue and towel trials. The loca-tion of the turning roll is adjustable to achieve proper creping geometry. Bulk is maintained as the sheet is transported by the conveyor belt unit. Then the sheet is reeled at the dryer-end reel at an optimum crepe ratio by adjusting the dryer-end reel speed. Figure 8 illustrates a typical tissue roll produced in a pilot machine tissue trial.

PILOT TISSUE MACHINE TRIALS

Since the introduction of the tissue con-figuration, a series of start-up and follow-up tissue trials have been carried out to evaluate the overall performance of the

pilot tissue machine operation and also to determine whether the quality of tissue samples is satisfactory.

For the start-up trials, a mixture of 50% NBSK and 50% eucalyptus was used: 1) co-refined to 515 ml CSF, and 2) unre-fined at 575 ml CSF. A matching pair of tissue forming fabrics was installed in the former. Each had a permeability of 570 cfm and a fibre support index (FSI) of 165. A pre-conditioned tissue press felt with a permeability of 15 cfm was used for the trial.

Figure 9 shows that the pilot tissue machine was able to achieve the selected target range of headbox consistency (i.e., 0.14% to 0.17%) during the start-up tri-als, which corresponded to basis weights ranging from 13 g/m2 to 17 g/m2.

Figure 10 compares the water re-moval profile around the forming roll for a refined (a) and an unrefined (b) furnish. Clearly, refining decreases the water re-moval rate at the forming roll, which is the first stage of the drainage process. Figure 11 shows how a required MD/CD tensile strength ratio could be achieved by alter-ing the jet/wire speed ratio.

It was also noted that the current configuration achieved effective transfer of tissue to the press felt. The press roll, Yankee cylinder, and creping blade section all performed properly. The solids level was 10-11% entering the press roll and 40-44% leaving the roll, depending on press roll loadings. The final tissue rolls were produced at 3-4% moisture at the reel. The collected tissue samples satisfied the target specifications of commercial prod-ucts for MD-CD tensile strength, MD-CD stretch, bulk, and softness.

Fig. 6a - Creping blade unit. Photo of the creping unit.

Fig. 6b - Creping blade unit. Creping unit in operation.

Fig. 7 - Schematic of conveying and reeling.

Fig. 8 - Tissue roll.

Fig. 9 - Effect of headbox consistency on basis weight control.


Since the tissue configuration was added, a number of trials have been run to improve the tissue capabilities. For example, a range of creping blades with blade-tip angles varying between 8 and 18 degrees were tried to evaluate the influence of blade-tip angle on bulk and tensile strength. The ma-chine speed at the Yankee was also tested up to 1200 m/min. Several pilot trials with dry-strength resins, softeners, and various strength agents were also carried out. Dif-ferent types of forming fabrics and press felts were installed on the machine to eval-uate the effect of tissue clothing on drain-age performance and tissue properties. A fully-dried paper towel at a basis weight of 26 g/m2 was also produced, and the required towel specifications were met for MD-CD wet and dry strength, MD-CD stretch, absorbency, and bulk.

AREAS FOR RESEARCH AND DEVELOPMENT

FPInnovations’ pilot tissue and towel ma-chine has been successfully used for test-

ing and optimizing:

• Furnish optimizationo Furnish types, blending ratioo Refining (co-refining or separate)o Furnish layeringo New materials (avoiding contam-ination in a commercial machine)

• Wet-end chemistryo Additives (strength resins, en-zymes, etc.)o Points of addition and controlo Softener

• Forming and pressingo Forming fabrics, press felto Operating conditions (HB con-sistency, jet/wire ratio, speed, etc.)o Press loadings, vacuum, doctor

• Dryingo Effect of furnish, refining o Yankee vs. hood drying rate

• Crepingo Creping geometry, blade angle, materials, loadingo Spray boom, coating and releas-ing chemical optimizationo Creping ratio

SUMMARY

FPInnovations’ pilot paper machine has been modified to produce a wide range of paper products. The pilot machine can be

configured to manufacture low- to high-basis-weight tissue, paper, and board prod-ucts ranging from 13 g/m2 to 180 g/m2. In tissue applications, several trials have been conducted to support a wide variety of tissue-related research and product devel-opment projects for tissue producers and suppliers. A stock preparation system can handle bales of virgin and recycled fibres up to 100 m3 pulp capacity. Refining may be done off-line. Two separate furnishes can be handled and simultaneously stored in the stock reservoirs. Two fan pump lines with separate pressure screens and radial distributors are available to feed two different types of pulps into a headbox. The multi-layer headbox with two radial distributors can handle low-basis-weight tissue (as low as 13 g/m2 at production) and towel grades. The headbox does not require a tapered header because of its cir-cular distribution system. The headbox is equipped with dilution and an edge flow control system which provides an even CD flow distribution. The forming roll, coupled with the couch roll, has enough water-drainage capability for low-basis-weight tissue. The tissue web is transferred from the forming section by a press felt to a suction press roll. The wet tissue web is wrapped around the press roll that con-tacts the 3-m-diameter Yankee cylinder under different loadings. The steam-heat-ed Yankee roll is surrounded by two hot-air impingement dryer hoods. Air is heated by the combustion of natural gas and im-pinges on the tissue or towel surface from the dryer hoods. The air impingement velocity and temperature can be adjusted according to the desired production rate. The creping blade unit can handle vari-ous creping angles, blade-tip angles, and take-off angles. The spray boom can ac-commodate numerous coating, releasing, and modifying chemicals. The tissue roll is produced at the dry-end reel.

Start-up and follow-up trials have shown that the produced tissues and towels meet the quality specifications of the market for MD-CD tensile strength, stretch, bulk, softness, and absorbency. The pilot machine can be used for testing

Fig. 10 - Comparison of water removal around the forming roll with refined and unrefined furnishes.

Fig. 11 - Effect of jet/wire ratio on MD/CD tensile ratio.



and optimizing furnish blends, wet-dry strength chemistry, forming, pressing, creping, and Yankee coating. It can also be used to evaluate equipment performance for tissue and towel manufacturing.

ACKNOWLEDGEMENTS

The authors acknowledge the critical con-tribution of Kruger Products to the suc-cessful conversion of the pilot paper ma-chine to a tissue and towel machine. We thank Wladimir Janssen, Paul Patry, and Marc Desaulniers for their valuable tech-nical input and continuous support. The authors also thank JohnsonFoils, Albany, AstenJohnson, and GL&V for supplying the equipment needed for tissue conver-sion.

REFERENCES

Crotogino, R.H., Pikulik, I.I., Lé-ger, F., Daunais, R., Goulet, B., and Hamel, J., “Paprican’s New Pilot Pa-per Machine”, Pulp & Paper Canada, 101(10):48-52 (2000).Pikulik, I.I., Gauthier, G., and Hamel, J., “PapriDryTM—Continuous Op-eration of Two Pilot Units at Papri-can”, Proceedings, Paptac 96th Annual Meeting (2010).Andersson, I., “Pilot Tissue Machine at KMW Development Center Cre-ates New Possibilities for Research on Paper Drying”, Proceedings, Tappi Engineering Conference, pp.457-461 (1985).Andersson, I., “STT Technology for High-End Tissue Products”, Paper Technology, 49(6):27-30 (2008).Klerelid, I. and Thomasson, O., “AdvantageTM NTTTM: low energy,

1.

2.

3.

4.

5.

high quality”, Tissue World, Oct/Nov (2008).Metso, www.metsopaper.com, “Ad-vantageTM NTTTM” (2010).Ivarsson, H.,”Two Full-sized Pi-lot Machines”, International Paper World, 3:16-17 (2005).Voith, www.voithpaper.com, “Pilot Tissue Machine at the PTC Sao Pau-lo” (2010).Derardi, R. and Scherb, T.T., “Struc-turing Tissue”, Tissue World, Aug/Sep (2008).Scherb, T. and Zane, R., “Innovative Technology for Premium Tissue Pro-duction”, CD Proceedings, Tappi Pa-perCon2008, Dallas, May (2008).Patterson, T. and Choi, D., “Predict-ing the Performance of Creping Ad-hesives”, CD Proceedings, Tappi Pa-perCon2010, Atlanta, May (2010).

6.

7.

8.

9.

10.

11.

PaperWeek Canada, the major Canadian gathering for the advancement of the pulp and paper industry, will be taking place February 4-8 at the Queen Elizabeth Hotel in Montreal. Building on the success of the previous edition’s program, we encourage all mill personnel, suppliers, researchers and experts to submit their latest work for presentation in the technical program. PaperWeek Canada has been the top platform in Canada to reach the key players of the industry and raise global awareness of the most recent progress in the technology and innovation of the pulp and paper industry. We invite all technical experts to take advantage of this opportunity and play an active role in making our industry innovative.

Topics for submissions:

Advanced Controls

Bleaching

Emerging Research

Energy Cost Saving

Fibre Supply & Biomass

Forest Biorefinery Symposium

Green Chemistry

Maintenance

Mechanical Pulping

Paperboard Packaging

Paper Machine Technology & PerformanceWaste & Sludge Management

Water Management

Others (please specify)

Please submit your abstract to PAPTAC at [email protected] / Telephone: 514-392-6969.The deadline for abstract submissions is Monday September 3rd, 2012.


ABST

RACT A wide range of Canadian newsprint machines were surveyed for their linting propensity using the long-established FPInnovations Lint

Test (formerly the Paprican Lint Test) for image area linting and the newly developed LP (Linting and Piling) Test for non-image area linting. Results showed that although overall linting performance has been improved, there are significant differences among paper machines. A closer examination of paper machine configuration showed that linting is affected strongly by the former type, as well as by the degree of consolidation in the press section. This was especially evident for cases where similar pulp furnish is supplied to very different machines in the same mill.

JOSEPH ASPLER*, JIMMY JONG, TONY MANFRED

BENCHMARKING PAPER MACHINE INFLUENCE ON LINTING AND PILING

The newsprint industry in North America is in the midst of a severe decline, caused largely by Internet news sources. Accord-ing to the Newspaper Association of America [1], newspaper circulation in the United States peaked in the mid-1980s, when approximately 63 million newspa-pers were sold per day. The most recent figures (to the end of 2009) show that circulation had dropped to 46 million per day, a figure which is still declining. Concurrently, North American newsprint tonnage has shown a sharp decline of ap-proximately 60%, from 13 million tonnes in 2000 to a forecasted 5 million tonnes in 2013.

For that very reason, it is of critical importance for suppliers to do their ut-most to maintain product quality in a mar-ketplace that has also become increasingly quality-sensitive. Furthermore, the decline in the North American newspaper indus-try has not been matched in the rest of the world. In developing economies, newspa-per readership is on the increase. Of the world’s top 100 circulation newspapers, 40 are located in India and China. The world’s top five circulation newspapers are located in Japan. Canadian newsprint manufac-turers who sell overseas must frequently

compete with overseas newsprint mills that are often more modern than those in Canada.

Linting of newsprint – the buildup of fibres and other debris on the offset blanket during printing – is a critical quali-ty issue. Despite substantial improvements in paper quality since the 1960s, the expec-tations of printers have also increased, and therefore linting remains a major problem today. In addition, many mills are trying to reduce product cost by using a wider range of wood species and by reducing refining energy. This creates new challenges for controlling linting.

Linting may be further subdivided into image area linting and non-image area

linting, with different mechanisms and appearance for each. Typical examples of image and non-image area lint – on the same offset blanket – are shown in Fig. 1.

All lint samples contain a large amount of fine material such as ray cells and fibre fragments. Image area linting is further distinguished by the presence of more long fibres and other coarse materi-al. Image area linting tends to be the result of excessive ink tack, poor fibre develop-ment, or a combination of the two.

In offset lithography (the most im-portant printing process), the separation between image and non-image areas is maintained by the aqueous fountain so-lution applied to the non-image area.

INTRODUCTION

JOSEPH ASPLERFPInnovations,Pointe Claire, QC, Canada*Contact: [email protected]

JIMMY JONGFPInnovations,Pointe Claire, QC, Canada

TONY MANFREDFPInnovations,Pointe Claire, QC, Canada



Fountain solution contains between 1% and 2% active ingredients, with the re-mainder being water. As noted below, fountain solution formulation is a highly proprietary field. In general, the active in-gredients include a polysaccharide material such as gum arabic; buffers such as phos-phate or citrate salts; and so-called “anti-linting additives”, which are normally a proprietary mixture of co-solvents such as ethylene glycol and related material. Lint in the non-image area consists of very fine material. This most commonly consists of ray cells, fibre fragments, and (for filled papers) mineral filler. It may also include the piling or milking of coating pigment from coated paper. Frequently, non-image area linting is more serious, with problems that include distortion of text and illustra-tions and damage to the offset blanket. Longer fibres may be seen in non-image area lint, but to a lesser extent than in the image area.

As shown in [2], a considerable amount of work has been done on under-standing and controlling linting, including work related to pulp quality, paper ma-chine operation, printing press conditions, and testing methods. Nevertheless, lint performance is a constantly moving tar-get. Acceptable paper quality from as little as 10 years ago would not be acceptable today.

The great majority of historical lint

tests simulate the action of the high-tack offset ink [2]. As a result, there is virtu-ally no quantitative information available on non-image area lint testing. Therefore, in addition to the image-area lint, the non-image area lint was benchmarked using an instrument, the Linting and Piling (LP) Tester, developed as part of this research and described in more detail below. Both image and non-image area lint values are compared to basic chemical and physi-cal properties of the newsprints. Finally, as noted below, it has been more than 20 years since the last systematic study of the effect of the paper machine on linting, and therefore paper machine de-sign was also taken into consideration.

The Paper Machine as a Variable in LintingIn addition to the considerable amount of time spent in improving mechani-cal pulp quality, there is also a body of knowledge on the influence of the paper machine on linting. More than 20 years ago, Hujala and Ottelin [3] examined a large number of Fourdrinier, top-former, and twin-wire newsprint machines. Their results were further analyzed by Wood et al. [4], who sub-divided machines into “high”, “moderate”, and “low” linting machines according to their forming sec-tion and press section. Given the chang-es that have occurred in papermaking

since this 1987 work, the present lint benchmarking study provided an addition-al opportunity to analyze linting in terms of its machine of origin. This is even more important given the newer tools now avail-able to measure and quantify linting.

Forel et al. [5] carried out trials on a pilot paper machine. They showed that fines distribution – as controlled by choice of forming fabrics and by other wet-end drainage conditions – had a strong influ-ence on the degree of linting. Although a correlation was found between linting and water absorbency, it is far more likely that fines distribution is the key factor and that the relation between water absorbency and linting reflects the lower water absor-bency of fines-rich surfaces.

EXPERIMENTAL

Newsprint SamplesNewsprint samples and their key proper-ties are described in Tables 1 and 2. Sam-ples are coded by mill, by machine, and (where applicable) by production variable on that machine.

These samples were collected in the spring and summer of 2008. Furnish in-formation provided by the mills is shown in Table 1. More detailed information on furnish, such as freeness, was not provid-ed. In particular, several of the samples were the product of paper machine trials involving proprietary chemical additives. All physical properties of paper were measured according to PAPTAC stan-dard methods. The water-drop absorption times were measured with diluted fountain solution at the same concentration used in the LP Tester, rather than with pure water. Fines distribution in the z-direction was characterized using a recently reported sheet-splitting method [6].

FPInnovations Lint Test – Image Area Linting (formerly Paprican Lint Test)As described in a PAPTAC Useful Method under the name of the Paprican Lint Test [7], the image area removal of lint by the tack of the oil-based offset printing ink is modelled by printing under relatively mild

Fig. 1 - A typical lint sample showing both image area lint (long, coarse fibres and mini-shives) and non-image area linting and piling (very fine material, chiefly ray cells and mineral filler). Scanning electron micrograph – note 1-mm scale bar on the bottom of the image.


conditions with an ink varnish (the ink-makers’ term for ink without pigment). The lint is collected and then quantified using the Fibre Quality Analyzer (FQA). Linting is reported as the total fibre length collected per square metre of paper (m/m2).

LP (Linting and Piling) Tester – Non-Image Area LintingAs noted above, the great majority of lint tests simulate the action of the high-tack offset ink [2]. However, non-image area linting caused by the aqueous offset foun-tain solution can be of equal or greater importance. For this reason, the LP (Lint-ing and Piling) Tester has been developed as part of this study to measure linting caused by the fountain solution. This will be described in detail in a forthcoming publication. The paper is run through a nip and moistened with a film of com-mercial offset fountain solution of the ap-propriate thickness. Lint removed by the fountain solution is reported as weight of lint per m2 of paper. A conventional foun-tain solution designed for newspapers was used to remove lint from the paper at a dilution of 2% (by volume). The web-fed LP Tester is based on a previous sheet-fed instrument, as described in [8]. RESULTS AND DISCUSSION

Image and Non-Image Area LintingFigure 2 shows the results for image area linting. There are two median values. For almost 900 samples measured between 2006 and 2008, the median value is 8.5 m/m2. Image area lint test values have shown a large long-term decrease over the last 20 years [2]. The continued reduction in linting propensity is demonstrated by the value of 6.2 m/m2 for 180 samples mea-sured in 2008 alone, even with the outlying samples in Fig. 2.

Figure 3 shows the non-image area lint results as measured using the LP Tes-ter. For 80 surfaces, the median value for newsprint fibre removal by fountain solu-tion is 0.28 mg/m2. The COV (coefficient of variation) for non-image area linting is approximately 8% for test samples that

TABLE 1 Samples used in this study.

Same reel – different positions

A

Mill Sample Code Other notesMachine Furnish

B

C

D

E

F

G

H

I

J

K

L

1 1 – A

“2 1 – B“3 1 – C“4 1 – D

Different day5 1Different day39 1 Different day40 1

6 17 28 142 243 1 9 110 211 1 55% TMP/45% DIP12 2 50% TMP/50% DIP13 3 50% TMP/50% DIP14 4 40% TMP/60% DIP15 1 80% TMP/20% DIP16 2

45 gsm17 1 45 gsm18 248 gsm19 3 74% DIP/26% TMP

20 1 67% DIP/33% TMP21 2 62% DIP/38% TMP

48 gsm22 323 1 24 225 326 2 65% TMP/35% DIP

Chemical trial 127 2Chemical trial 228 2

35 136 2 37 338 429 130 231 332 133 2 34 141 1

100% TMP

100% TMP

100% TMP100% TMP

100% TMP

100% DIP

100% TMP100% TMP100% TMP100% TMP

100% TMP100% TMP100% TMP

100% TMP100% TMP

100% DIP100% DIP100% DIP

100% TMP

100% TMP

100% TMP100% TMP

100% TMP100% TMP100% TMP100% TMP100% DIP

75% TMP/25 % DIP

100% TMP/6% fi ller

100% TMP/3% fi ller100% TMP/3% fi ller100% TMP/3% fi ller



were repeated five times. Note that three of the samples from the lint test described in Table 1 and Fig. 2 were not provided in sufficient quantity to do the LP test.

Depending on the sample and print-ing conditions, linting may occur only in the image area, only in the non-image area, or in all areas. Therefore, there is no a pri-ori reason to expect a correlation between measured image and non-image area lint values. This is shown in Fig. 4, where the coefficient of determination is R2 = 0.36. This is statistically significant for this number of samples, but is not useful for predicting one type of lint behaviour from the other.

This leads to the obvious conclusion that different mechanisms may be at work in image area and non-image area linting. In image area linting, long fibre linting is caused by the “tack” of the ink. In turn, ink tack is caused by the z-directional splitting of the ink and can be related to the fundamental rheological and compo-sitional properties of the ink [9, and refer-ences cited therein].

Non-image area linting is caused by the aqueous fountain solution and is poor-ly understood. Based on the influence of fountain solution formulation on material removal, it has been proposed that non-image area linting results from a combina-tion of factors:

1. The degree to which very fine ma-terials such as ray cells and filler par-ticles are bonded to the paper surface. To some extent at least, this may be related to factors that cause image area lint, for example, the degree of fibre bonding and the degree of fines/fibre retention at the sheet surface.2. The degree of adhesion between the blanket and the paper surface. A smoother blanket will lead to more linting/piling due to the greater force required to release the paper from the blanket surface [e.g., 10,11]. Blanket suppliers provide deliberately rough-ened, “quick-release” blankets recom-mended for lint-sensitive grades (par-ticularly newsprint). These roughened blankets are thought to reduce linting

TABLE 2 Physical properties of paper samples.

A

Mill CodePPS

roughness,μm

Paper outside (nominal bottom side)Fountainsolution

dropsorptiontime, s

B

C

D

E

F

G

H

I

J

K

L

1 5412 5413 5414 5415

3940678

42439

101112131415161718192021222324252627283536373829303132333441

Paper inside (nominal top side)

Airpermeability,

mL/min

PPSroughness,

μm

Fountainsolution

droptime, s

Airpermeability,

mL/min

3.543.543.543.54

3.453.453.453.45

8.48.48.48.4

10101010

535535535535

468576557248

3.743.853.533.84

3.463.393.203.75

9.07.18.518

9.39.19.618

231

344344396420

3.513.103.63

3.273.223.50

168.07.2

158.110

377499

297368348350

5.075.273.824.10

4.864.763.723.98

50592321

50472622

560407322327

380343486169

3.874.003.783.19

4.134.463.773.75

55352124

56382034

380366542183

334308238252

3.343.263.193.59

3.723.163.283.61

46201936

40182229

348279230259

285159188342

3.222.933.504.32

3.303.083.903.61

141811678

152310974

231161204360

276475457

3.814.383.894.13

3.324.203.814.11

70251214

59301413

258

4003.633.71

3.733.81

1913

1914

285166450432

3.473.754.293.59

3.073.093.943.42

128.62016

20152116

450432

256503

243

3.963.964.133.80

3.983.803.873.87

27132215

30161916

256503

243


by contributing localized asperities or rough spots that help to relieve stress in the printing nip. For the same rea-son, worn blankets (or intentionally smooth blankets) are believed to gen-erate more lint because they lack these rough spots.3. Fountain solution formulation and amount. Surprisingly little informa-tion is available because of the highly competitive and proprietary nature of fountain solution formulation. Em-pirical evidence shows that both the formulation and the amount applied to the paper have a major influence on non-image area linting. It is well known that an inappropriate fountain solution can cause major – sometimes catastrophic – increases in linting. Vari-ous “anti-linting additives” such as gly-cols and glycol ethers are known to re-duce linting, although the mechanism is still unclear. The most commonly accepted mechanism is that as water-miscible, low-vapour-pressure liquids, the anti-linting additives act as hu-mectants, preventing the evaporation of the very thin aqueous film in the non-image area. This is the key reason why the LP Tester was developed. Be-cause fountain solution formulation is an almost entirely empirical art, the LP Tester provides a new way to predict fountain solution behaviour.

Relationship between Linting and Other Paper PropertiesThere were no significant correlations between image or non-image area linting and basic paper properties such as rough-ness, air permeability, compressibility, or absorption of either fountain solution or pure water.

Although it has often been said that linting may be related to the water absor-bency of the sheet, as noted previously, the only definitive case was a recent trial in which the fines at the sheet surface were systematically varied [5]. Even here, the causal relationship was with the increased fibre bonding caused by the increased

Fig. 2 - Lint removed for 43 newsprint samples, simulating image area lint generated by the action of a high-tack ink. The median value for the 43 samples here is the same as the overall median value measured in 2008.

Fig. 3 - Lint removed from 40 newsprint samples by the Linting and Piling (LP) Tester, simulating non-image area lint generated by the action of the offset fountain solution. The median value of non-image area lint for these 40 samples (80 surfaces) is 0.28 g/m2.

Fig. 4 - The correlation between image and non-image area linting is poor. Although statistically significant, the correlation has little predictive power.



fines content. The (apparently beneficial) effect of lower water absorbency was al-most certainly a consequence of the in-creased amounts of fines.

Multiple Samples from the Same Mill or MachineAs shown in Tables 1 and 2, several sets of samples were taken from the same mill or machine. In the case of paper from the same machine, the samples may have come either from different production dates, from different positions from the same reel, or from chemical or furnish tri-als.

As noted previously, Mill J provides the most egregious example of three very different machines receiving nominally the same furnish, yet showing very dif-ferent linting performance. Mill C (with two samples from each of its three ma-chines) provides another good example of this effect. Mill H – perhaps the most modern mill in the study – showed a much smaller variation among its three machines. Mill I (four machines) also showed a considerable difference in lint-ing with nominally the same furnish.

In general, within-machine linting variations are much smaller than the range of all the samples in this benchmarking. This can be seen for the single machine from Mill A (four positions from the same reel, and single samples from three other production periods).

Influence of the Forming Section and Press Section on LintingThe paper machines in this trial covered a very wide range in age and design, from older machines with hybrid Fourdrinier/top former sections immediately followed by two-roll straight-through presses, to much newer twin-wire gap formers with three-nip or four-nip presses.

Figure 5 shows image area linting ac-cording to the type of former [12]. The three machines in Mill J – all receiving nominally the same furnish – cover the widest range of linting, from the very low to the very high. All three machines are top-formers: Top-Flyte S (blade-roll), Belform (blade-roll), and Duoformer D(blade). The main difference among the three machines is the press section con-figuration. The paper machine with the poorest linting performance was equipped with an open draw at the couch and two straight-through press sections having only two nips. The configuration of this machine meant that only one side of the sheet touched the press roll surface twice. The other two machines had three-nip press sections.

Figure 6 shows non-image area lint according to the type of former. In general, twin-wire formers tend to have better image area linting performance. However, the effect of press section type should also be considered for overall linting performance in both image and non-image

areas. Non-image area linting covers a much wider range of values, perhaps reflecting greater difficulties in maintaining sheet consolidation in the presence of water. As noted in the Introduction, quantitative measurement of non-image area linting is still a very new field, and therefore further work is required to gain a better understanding of this phenomenon.

Figures 7 and 8 present image and non-image area linting tendencies according to the type of press section. Regardless of former, the machines with two straight-through presses have the poorest linting tendency for both image and non-image area linting. The Tri-Vent press type (the latest design within this benchmarking study) with a fourth press has the best performance in both image area and non-image area linting.

Table 3 describes each press type from this study with respect to which paper side touches the press felt, the total number of press nips, and how often the paper is pressed on the top or bottom side. For example, the Tri-Nip press type has three nips. In the first nip, the paper is pressed by felts from the top and bottom sides at the same time (double-felted). In the second nip, the paper is pressed by the top-side felt only (single-felted from the top side). In the third nip, the paper is again pressed from the top side only (single-felted from the top).

In Fig. 9, non-image area linting is

Fig. 5 - Image area linting shown for different types of forming sections. Fig. 6 - Non-image area linting shown for different types of forming sections.


compared for the different types of press sections. For the Top-Flyte S, the press section with two straight-through presses tends to have poor linting. These two ma-chines with two straight-through presses, shown in Fig. 9 and Table 3, had their bot-tom sides contacting press felts twice. Be-cause the side that touches the press felt tends to have better consolidation of its surface fibres, the bonding on that side would be stronger, and linting is less likely.

Figure 10 shows an example of fines distribution in the z-direction for Sample 19, which was produced by a BelBaie II former with a tri-nip press section. The BelBaie II is well known for fines two-sidedness due to uneven drainage in the forming section. More drainage tends to take place on the backing (top) side in the early part of the former. Therefore, there tend to be more fines in the top side as well. In general, the side with more fines tends to lint less, which is true for this case, as shown in Fig. 2. Similar trends were observed for non-image area linting, as shown in Fig. 3.

SUMMARY AND IMPLICATIONS FOR PAPERMAKERS

For the first time in many years, a wide range of newsprint machines have been surveyed for their linting propensity. This

study used both the long-established FPInnovations Lint Test (formerly the Paprican Lint Test) for image area linting, and the newly developed LP (Linting and Piling) Tester for non-image area linting. A wide range of linting performance was observed for different machines even when they used the same or a similar furnish.

Although it is not unexpected that linting will be strongly affected by the

paper machine former and press section, the results show that the interaction of the three key variables – furnish, former, and press configuration – provides the means for paper manufacturers to focus on the key areas in which linting may be effec-tively reduced. Although not within the scope of this work, the operation of the paper machine drying section is another area where linting performance could be further improved [13].

Fig. 7 - Image area linting shown for different press sections. The number in brackets after the description of each type of press section represents the number of machines with the specified configuration. Error bars represent the variation across the range of specified configurations.

Fig. 8 - Non-image area linting shown for different press sections. The number in brackets after the description of each type of press section represents the number of machines with the specified configuration.

TABLE 3 Press section description.

2 Straight-through

Types

B/B

Felt Side*(B: bottom,T: top,D: Double)1st /2nd /3rd /4th nip

Total Numberof Nips

Numberof Top-felted

NumberofBottom-felted

Total Numberof Felted Sides

Combi pickup + 2 straight-throughTwinver + 3rd pressBiNip + 3rd pressTri-NipTri-VentTri-Nip + 4th PressTri-Vent + 4PB: bottom side of the sheet makes contact with a press feltT: top side of the sheet makes contact with a press feltD: double-felted, both sides of the sheet make contact with press felts

T/B/B

T/T/B

D/T/BD/T/TD/T/TD/T/T/BD/T/T/B

2 0 2 2

3 1 2 3

3 2 1 3

3 2 2 43 2 2 43 2 2 44 3 2 54 3 2 5



ACKNOWLEDGEMENTS

We acknowledge our Member Companies for their support and contribution of paper samples. The financial and material support of Unigraph International (fountain solution supplier) is acknowledged. Laboratory measurements were done by Joëlle Grenon, Jurgen Superga, and Marc-Antoine Brunet.

REFERENCES

“Newspaper Circulation Volume”, http://www.naa.org/Trends-and-Numbers/Circulation/Newspaper-Circulation-Volume.aspx, accessed November 29, 2011.Aspler, J.S., “Linting and Surface Contamination: Current Status”, Pro-ceedings, Technical Association of the Graphic Arts, Rochester NY, pp. 375–398 (2003).Hujala, J. and Ottelin, E., “Improv-ing Paper Quality for Offset Printing by Twin-Wire Units”, Preprints, 73rd Annual Meeting of the Technical Sec-tion, CPPA, Montreal, p. A207 (1987).Wood, J., McDonald, J.D., Ferry, P., Short, C.B., and Cronin, D.C., “The

1.

2.

3.

4.

Effect of Paper Machine Forming and Pressing on Offset Linting”, Pulp and Paper Can. 99(10):53 (1998).Forel, F., Jong, J., Orccotoma, J.-A., Fournier, F., McVey, D., and Bernier, M.-A., “Effects of Forming Fabric Design on Drainage, Paper Quality, and Linting – Pilot Machine Trials”, Proceedings, TAPPI 2nd Annual Pa-perCon’09 Conference - Solutions for a Changing World, St. Louis MO, May 31,– June 3, 2009.Forel, F., Orccotoma, J.O., and Bates, J., “A New Method for Quantifying the Z-Direction Distribution of P200 Fines”, Preprints, PAPTAC 2009 An-nual Meeting, Montreal, pp. 589-594.“Paprican Lint Test Using the IGT Printability Tester”, Useful Method L.5U, PAPTAC, Montreal.Aspler, J.S. and Manfred, A., “Foun-tain Solution Effects on Linting and Piling Using the GFL Fluff Tester,” Proceedings, International Printing and Graphic Arts Conference, Atlan-ta, September, TAPPI Press (2006).Aspler, J.S., “Ink-Water-Paper Inter-actions in Printing: A Review”, 9th

5.

6.

7.

8.

9.

TAPPI Advanced Coating Funda-mentals Symposium, Turku, Finland; TAPPI Press, Atlanta, p. 117 (2006).Moller, K., Thomassen, B., Weidem-muller, J., Menzel, P., Walther, K., Fal-ter, K., Sporing, G., Meissner, M., and Axell, O., “Factors Influencing Lint-ing in Offset Printing of Newsprint”, Proceedings, 47th APPITA Confer-ence, Hobart, Australia, p. 115 (1995).Sudarno, A.T., Investigation of the Effect of Press and Paper Variables on Linting During the Offset Print-ing of Newsprint, M.Eng Thesis, Monash University, Melbourne, Aus-tralia (2006).Jong, J., Gooding, R., and Rompre, A., “Survey of Top-Former Operation and Performance”, Pulp and Paper Can. 105:8, T187 (2004).McDonald, J.D. and Tchepel, M., “The Effect of Dryer Section Op-eration on the Linting Propensity of Newsprint”, presented at the 95th Annual Meeting of the Pulp and Pa-per Technical Association of Canada, Montreal, 2009.

10.

11.

12.

13.

Fig. 9 - Non-image area linting for different forming sections and press sections.

Fig. 10 - An example of z-direction fines profile: BelBaie II with Tri-Nip press.


ABST

RACT Natural gas processing plants typically consume natural gas to generate steam for mechanical power and process heating use. In western

Canada, these plants remove hydrogen sulphide and carbon dioxide and extract hydrocarbon liquids from the incoming raw natural gas.These plants are sometimes located remotely in Forest Management Areas. Such locations are key to the concept of installing a biomass power plant at a natural gas processing plant. Harvesting of local wood waste to supply this boiler could be economically viable.

A high-pressure biomass-fired power boiler and a back-pressure steam turbine generator would replace the duty of the natural gas-fired power boilers in the gas processing plant. The back-pressure steam turbine would exhaust into the existing steam system. Internal natural gas consumption used for steam generation would be eliminated, imported power would be reduced, and there would be a net reduction in greenhouse gas emissions.

DEREK McCANN

ADDING A BIOMASS-FIRED COGENERATION POWER PLANT TO A NATURAL GAS PROCESSING PLANT

In western Canada, there are several natu-ral gas processing plants. These plants treat raw or sour natural gas and make it “sweet”. They remove hydrogen sulphide and carbon dioxide and extract hydro-carbon liquids from the raw gas. These processing plants are sometimes located within Forest Management Areas (FMAs).

A typical natural gas processing plant consumes sweet natural gas to generate steam for mechanical-drive turbines and process heat.

It is proposed to install a high-pres-sure biomass-fired power boiler and steam turbine generator that would replace the duties of the natural gas-fired power boil-ers and superheater and reduce imported electrical power in the gas processing plant.

A cogeneration plant is proposed where the natural gas processing plant would be the “steam host”.

PROJECT OBJECTIVES

There are several project objectives:

Use Locally Available Wood Waste - The haulage cost of wood waste is essentially a function of collection radius. With the

natural gas processing plant located within a Forest Management Area, the collection radius would be minimized.

Eliminate In-house Natural Gas Con-sumption - The in-house natural gas-fired power boilers and superheater would no longer be required. The new biomass-fired power boiler would be capable of the full steaming rate on natural gas alone. This would be available in case there was a breakdown of the biomass handling sys-tem.

Reduce Imported Power Consumption - The new back-pressure steam turbine gen-erator would generate most of the electri-cal power required by the natural gas pro-cessing plant.

Reduce Greenhouse Gases Due to Natu-ral Gas Consumption - The biomass-fired power boiler would produce low levels of greenhouse gases (GHGs), and natural gas consumption would be eliminated.

Reduce Greenhouse Gases Associated with Local Imported Power Generation - Reduction of local imported power would reduce its associated GHGs.

Create an Acceptable Return on Invest-ment (ROI) on Project Capital Using All Available Federal and Provincial Gov-ernment Incentives for Greenhouse Gas Reductions - For the project to be eco-nomically viable, all available federal and provincial incentives for GHG reductions would have to be used.

DESCRIPTION OF EXISTING NATU-RAL GAS PROCESSING PLANT STEAM SYSTEM

The existing natural gas processing plant steam system is an example of cogenera-tion in which mechanical rather than elec-trical power is generated. It includes several

INTRODUCTION


DEREK McCANNAMECVancouver, BCCanada



back-pressure mechanical-drive turbines driving pumps and one condensing me-chanical-drive steam turbine driving a compressor.

Saturated high-pressure (HP) steam is generated at 2500 kPa(g) using natural gas-fired power boilers and a sulphur-plant waste heat boiler. Saturated low-pressure (LP) steam is consumed at 380 kPa(g). A natural gas-fired superheater is used to provide the HP steam supply to the condensing steam turbine.

The steam system is shown in Fig. 1.

SELECTION OF INLET STEAM CONDITIONS FOR THE STEAM TURBINE GENERATOR

It is proposed to install a “topping” back-pressure steam turbine generator (STG). In this arrangement, the steam turbine exhaust pressure would essentially match (allowing for pressure drop) the pressure in the processing plant’s HP steam sys-tem. Therefore, the inlet pressure to the steam turbine generator must be some-what higher than 2500 kPa(g). To select the STG inlet conditions, a comparison was performed for three alternative cases; the results are shown in Table 1.

Overall efficiency is the product of the isentropic and the mechanical and electrical efficiencies.

Alternative C was selected because it had the highest power generation output. The economics of scale usually favour the largest plant. However, the cycle efficien-cies of the three alternatives are somewhat similar.

Mechanical-Drive TurbinesAt the natural gas processing plant, there are several mechanical-drive turbines, but no steam turbine generators. In pulp mills, mechanical-drive turbines are often replaced by electric motors because they are inefficient compared to steam tur-bine generators. For this potential project,

replacing the various mechanical-drive turbines (MDTs) in the processing plant was considered. The overall efficiencies of the existing turbines were evaluated; the results are shown in Fig. 2.

It was found that small drives (~1000 HP) had low overall efficiencies (<50%). This is the case with fan drives on power and recovery boilers in pulp mills. How-ever, large drives (>6000 HP) had good overall efficiencies (>70%). (A modern steam turbine generator has an overall ef-ficiency of ~80%, depending on its size). Because there are many large drives in the existing processing plant, converting them would provide little benefit. Based on this, it was decided not to consider changing the mechanical drives.

Cogeneration Gross Cycle EfficiencyCogeneration gross cycle efficiency can be defined as:

Gross Cycle Efficiency = (E+H) x 100 F

where E is gross power output, H is net process heat, F is fuel energy (HHV basis),

and H = QOUT – QIN,

where QOUT is total heat exported

and QIN is total heat returned.Fig. 1 - Natural gas processing plant—existing steam system.

TABLE 1 Comparison of STG inlet steam conditions.

Inlet Steam Pressure Variable / Alternative Units A B C

Inlet Steam Temperature 6,200 8,600 10,340kPa(g)454 482 508°C

Exhaust Steam Pressure Steam Flow Theoretical Steam Rate Isentropic Effi ciency Actual Steam Rate Mechanical & Electrical Effi ciency Overall Effi ciencySTG Output

2,600 2,600 2,600kPa(g)152,000 152,000 152,000kg/h

14.8 10.8 9.3kg/kWh80 80 80%

18.5 13.5 11.6kg/kWh95 95 95%76 76 76%7.8 10.6 12.4MW


Biomass Cogeneration OptionsThere are two biomass cogeneration op-tions:

• Maximum cycle efficiency using a back-pressure turbine generator;• Maximum power using a con-densing steam turbine generator.

The two options were found to have the energy characteristics shown in Table 2.

The biomass fuel is assumed to have a moisture content of 35% (wet basis).

The condensing STG option would permit power export, but at a higher capi-tal cost than the back-pressure STG op-tion. It was decided to pursue the back-pressure STG option.

Proposed Biomass Power PlantThe proposed biomass power plant would include:

• A biomass-fired power boiler op-erating at 10,340 kPa(g) and 508°C• An electrostatic precipitator to limit particulate matter emissions• A biomass handling system• An ash handling system• A back-pressure steam turbine generator• A new de-aerator• A new make-up water treatment plant• Piping tie-ins to existing systems• Electrical tie-ins to existing sys-tems• A new boiler and steam turbine building.

INTEGRATION OF THE NEW BIO-MASS POWER PLANT AND THE EXISTING NATURAL GAS PRO-CESSING PLANT STEAM SYSTEM

The integration of the new biomass power plant and the existing natural gas process-ing plant steam system is shown in Fig. 3.

Energy Impacts of the Proposed Biomass-fired Power PlantThe energy impacts of the proposed bio-mass-fired power plant are shown in Table 3.

Fig. 2 - Mechanical-drive turbines—overall efficiency.

TABLE 2 Energy characteristics.

GenerationSTG Option Units Back- Pressure

Cycle Effi ciency12.4 42.4MW74 24%

Condensing

Fig. 3 - Integration of new biomass power plant into the existing natural gas processing plant.



The plant electrical load will increase due to the biomass power plant auxiliaries. The imported electrical load has dropped significantly. Natural gas consumption has disappeared.

Greenhouse Gas ReductionsThe following greenhouse gas (GHG) emissions were considered:

• Carbon dioxide (CO2)• Methane (CH4)• Nitrous oxide (N2O).

These were totalled as CO2 equivalent using global warming potential (GWP) [1]. The annual reductions are summarized in Table 4.

The imported-power GHG emis-sions intensity was assumed to be for Al-berta (see [1]), which has several coal-fired power plants. The other GHG intensities used can be found in Reference 1.

Economics of Biomass Power PlantsThe total cost of generated electricity

includes the following components:• Capital recovery cost• Fuel cost• Operations and maintenance cost.

These costs are usually evaluated on a $/MWh basis.

The capital recovery cost depends on the amount of capital debt, the inter-est rate, and the repayment period. It is usually expressed as a percentage of the project capital cost. The capital cost of a biomass-fired power plant is significant compared to other types of power plants. Therefore, the capital recovery cost would be significant.

The fuel cost would include the costs of harvesting and transporting the biomass. Biomass fuel costs are strongly influenced by the radius of the collection area. Biomass fuel has high moisture con-tent, which results in low cogeneration cycle efficiency. A greater quantity of wet biomass than of dry biomass is required to produce the same steam flow (because

boiler thermal efficiency drops with in-creasing biomass moisture content).

The operations and maintenance cost would include labour and materials.

Economics of scale would apply; a large plant would have lower total power costs than a small plant. Consequently, it can be difficult to make small biomass power plants economically attractive.

One method to improve the project economics is to obtain as many GHG re-duction grants as possible from the federal and provincial governments. These grants can be of critical importance in deter-mining the viability of a project. It would also help if the generated power could be sold as “biomass” power because biomass power can be sold for more then $100/MWh. This is higher than the price of power from traditional fuels.

Most projects require an internal rate of return (IRR) of 18% to be economi-cally viable. Current low natural gas prices (~$3.00/GJ) mean that it is difficult to make a project viable.

UNIQUE OPPORTUNITY FOR COOPERATION BETWEEN THE OIL AND GAS AND THE FOREST INDUSTRIES

This project could provide a unique op-portunity for cooperation between the oil and gas and the forest industries. A forest industry company could provide the bio-mass, and an oil and gas company could act as steam host. This arrangement could provide economic benefits to both com-panies while significantly reducing GHG emissions for the common good.

REFERENCESCalculation Tools for Estimating Greenhouse Gas Emissions from Pulp and Paper Mills, Version 1.1, July 8, 2005, NCASI.

1.

TABLE 3 Energy impacts of the proposed biomass-fi red power plant.

Electrical Load

Parameter Units ExistingPlant

ExistingPlant with Biomass

PowerPlant

Increment

Generation15.5 16.5 +1.0MW

0 12.4 +12.4MWImported Load Annual Natural Gas Consumption Annual Biomass Consumption

15.5 4.1 -11.4MW106 SCM/a

0 241,500 +241,500BDt/a-1000100

TABLE 4 GHG annual reductions.

Annual Natural Gas GHG EmissionsParameter Units Baseline Future Reduction

Annual Biomass Boiler GHG Emissions

218,585 0 218,585t CO2eq/a

0 7,180 (7,180)

Annual Biomass Harvesting GHG Emissions

Imported Power GHG EmissionsTotal

0 4,735 (4,735)

129,030 34,130347,615 46,045 301,570

t CO2eq/a

t CO2eq/a

t CO2eq/at CO2eq/a

94,900

J-FOR Journal of Science & Technology for Forest Products and Processes contains papers which are the property of the Pulp and Paper Technical Association of Canada (PAPTAC). Papers may not be reprinted or reproduced without permission. These papers have been peer reviewed and approved for publication by the Editorial staff.

J-FOR is published bi-monthly by PAPTAC, 740 Notre Dame West, Suite 1070, Montreal, QC, Canada H3C 3X6, to which all general correspondence should be addressed.

J-FOR Journal of Science & Technology for Forest Products and Processes is indexed and/or abstracted with CAS (Chemical Abstracts Service, a division of the American Chemical Society), Thomson Reuters, and ProQuest.

2012 subscriptions for a printed copy and on-line access are: Canada (includes all applicable taxes) – $850.00 Cdn; USA – $850.00 Cdn; Overseas – $875.00 Cdn.

No part of this Journal may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, without the prior written permission of PAPTAC or, in case of photocopying or other reprographic copying, a licence from Access Copyright (the Canadian Copyright Licensing Agency), 1 Yonge Street, Ste. 1900, Toronto, ON, Canada M5E 1E5.

ISSN 1927-6311 (Print)ISSN 1927-632X (Online)

Printed in Canada



Building for thePulp and Paper

Community

JOURNAL OF SCIENCE & TECHNOLOGY FOR FOREST PRODUCTS AND PROCESSES

Some of the photos in the Journal provided courtesy of FPAC.

Building for the New Pulp and Paper

Community

A PAPTAC JOURNAL

J-FOR is PAPTAC's preeminent flagship journal, and publishes peer-reviewed high-quality articles that address technological and scientific issues that are critical to the forest industry.

J-FOR emphasizes a unique balance between papers that address traditional technology areas, as well as emerging technology areas as our industry transforms to new business models.

These target areas are supported by a distinguished and international Editorial Board comprised of leadingexperts in the field.

Authors are welcome to submit their papers to J-FOR

Traditional Technology and Scientific Target Areas: Emerging Technology and Scientific Target Areas:

• Pulping, bleaching and papermaking fundamentals, processes and technologies• Energy and chemical recovery fundamentals, processes and technologies• Recycled fibre and recycling technology• Development of sensors, analytical methods and process control logics• Mill water and energy usages and optimization• Environmental concerns and their mitigation

• Emerging forest-based products and their chains of added value• Fundamentals of converting forest-based biomass into biofuels and other bioproducts• Nanotechnology and other high added-value processes• Development of chemical, biochemical and thermochemical processes for the forestry industry• Integrating emerging and sustainable processes into the pulp and paper industry• Harvesting and procurement of forest and other biomass feedstocks

Published by:

For further information, please visit www.paptac.ca or contact PAPTAC at 514-392-0265/[email protected].

Publishing establishes ownerships of ideas, enhances professional reputation and opens networking opportunities with colleagues.

J-FOR is published on a bi-monthly basis and distributed in printed format to corporate subscribers and libraries in over 30 countries.

J-FOR is also available on-line in an open access publication format to maximize author impact and visibility.

Journal of Science & Technology for Forest Products and Processes

BUILDING FORTHE NEW PULP & PAPER

COMMUNITY J-FOR

CALL FOR PAPERSBe part of this emergingglobal forestcommunity publication

The PAPTAC staff and Executive Council are actively working on the advancement of PAPTAC's objectives as our industry reshapes itself. We are more than ever committed to providing you and your company with pertinent tools to face today’s challenges and to be your partner in keeping you connected with your industry.

MEMBERSHIPM

BUILDING FOR THE NEW PULP & PAPER COMMUNITY

M TC FW FA PWCMembership Technical

CommunitiesFuture Workforce

ForestAdvancement

PaperWeekCanada

Become a member and experience the sharing of Canadian knowledge and new ideas!

J-FOR is sponsored by

OperationsImprovementsResources

www.paptac.ca

J-FOR - PAPTAC · J-FOR FOR THE ADVANCEMENT OF THE FOREST INDUSTRY A PAPTAC JOURNAL TABLE OF...

Documents

Transcript of J-FOR - PAPTAC · J-FOR FOR THE ADVANCEMENT OF THE FOREST INDUSTRY A PAPTAC JOURNAL TABLE OF...