The importance of fibre flocculation in flotation deinking P. Huber#, E. Zeno#, B. Fabry#, X. Rousset#, M.C. Angelier#, D. Beneventi*, T. Vazhure& #:CTP, *: LGP2, &: Aylesford Newsprint COST FP1005 “Fibre suspension flow modeling” 24-26 Oct. 2012, Trondheim, Norway Background • Influence of pulp concentration on flotation efficiency • (pilot flotation trials – VOITH facility) Tot. (%) conc. flot. cell # Fibres (%) At same brightness At same cleanliness Losses (%) Accepts brightness concentration Concentration (%) The higher the concentration, the lower the ink removal and the better the yield Britz, H., Peschl, A. (1994) , Wochenblatt für Papierfabrikation, n°15: 603-608, 1994. COST FP1005 – Trondheim 24-26 oct 2012 3 Background • Influence of concentration on air content • (industrial flotation trials) Air content Ink removal Flotation efficiency of free ink (%) 90 TABLE III. Regression model and t-test statistics of a regression model of the form: Air content = a 0 + a1 • consistency + a2 • feed brightness. 88 Regression Coefficient R2 = 0.751 86 Variables Range Intercept [a0] 84 Coefficient σ t-Student 8.574 0.216 39.76 Total effect Consistency (%) [a1] 0.964-1.15 -7.216 0.089 -80.82 -2.56 Feed brightness (%) [a2] 40.6-48.1 0.1729 0.0044 38.83 1.29 82 80 12 13 14 15 16 17 Air content in primary cells (%) Figure 4. Flotation efficiency of free ink in three parallel flotation lines each operated at different air contents. The higher the air content, the better the ink removal The higher the concentration, the lower the air content Dorris, G.M., Pagé, N., Gendron, S., Murray, T. & Ben, Y. (2006) Prog. Pap. Recycling, 16 (1), pp.31-40. COST FP1005 – Trondheim 24-26 oct 2012 4 Mechanisms • Hypothesis (Dorris et al. 2006) • High concentration • (flocculation ?) • heterogeneous fibre suspension chanelling • air bubbles can travel faster, coalescence and rise faster to the top of the cell • Decrease of relative residence time air/pulp • air content is reduced • ink removal is impaired COST FP1005 – Trondheim 24-26 oct 2012 5 Background • Influence of concentration on fibre flocculation • (laboratory trials with various pulps) Example with TMP fibres •Same results with BKP (HW, SW, mix), DIP, etc. The higher the concentration, the higher the fibre flocculation Huber, P., Carré, B., and Petit-Conil, M. (2008). BioRes. 3(4), 1218-1227. COST FP1005 – Trondheim 24-26 oct 2012 6 Background • Influence of crowding factor on pulp flocculation • (flocculation varied by changing concentration and pulp mixtures HW/SW) crowding sphere Kerekes (1985) crowding • N= nb fibres in the crowding sphere (non dimensional concentration) fibre crowding determines fibre flocculation (at given turbulence) Huber P., Roux J.C., Mauret E., Belgacem N., and Pierre C. (2003), J. Pulp & Pap. Sci. 29(3):77-85. COST FP1005 – Trondheim 24-26 oct 2012 7 Background • Influence of crowding factor on gas hold-up • (Column bubbling of virgin pulp) Gas hold-up (crowding) The higher the fibre crowding, the lower the air content Tang, C. & Heindel, T.J. (2006) The Canadian Journal of Chemical Engineering, 84 (2), pp.198-208. COST FP1005 – Trondheim 24-26 oct 2012 8 Motivations & objectives influence of pulp flocculation on flotation efficiency ? How to vary flocculation ? By changing concentration 1.2 flocculation (relative) • By adding dispersants Non surface-active fibre dispersants : 1 0.8 0.6 0.4 • Guar gum • CMC 0.2 0 8 10 12 14 16 concentration (g/L) COST FP1005 – Trondheim 24-26 oct 2012 9 Outline • • • Background Methods : flocculation sensor, gas hold-up sensor… Results • Mechanisms • Effect of concentration • Effect of dispersants COST FP1005 – Trondheim 24-26 oct 2012 10 Materials and Methods Flocculation measurement Assessment of fibre flocculation • • • Pulp circulation on the flocculation pilot loop • Fibre flocculation testing with the CTP FlocSens (image analysis) • Constant flow speed : equivalent shear rate = 690 s-1 (medium speed) Flocculation sensor (+overflow) installed on Recycled fibres pilot plant, at flotation inlet Pilot flow-loop On-line thickness = 3.5 mm overflow tank Floc. sensor IMAGE ANALYSIS flow meter FLASH TRANSPARENT CHANNEL CCD pulp tank 1 m3 surface • General methods CCD CAMERA • Si Di … Flocculation index: FI diameter COST FP1005 – Trondheim 24-26 oct 2012 11 Flocculation sensor principles Flocculation index : • surface • binary morphology floc size distribution • Flocculation index: Si Di … diameter COST FP1005 – Trondheim 24-26 oct 2012 12 Reduced sensitivity to light diffusion • • Problem : filler diffuse light Fibres only Fibres+20% filler RMS = 0.323 FI = 2.63 mm² RMS = 0.069 FI = 2.65 mm² (no filler) Even in presence of light diffusing filler : Fibre flocculation measurement is possible (independently on filler flocculation level) Huber P., Roux J.C., Mauret E. and Carré B. (2006), APPITA Journal 59(1):37-43 COST FP1005 – Trondheim 24-26 oct 2012 13 Flotation monitoring : Assessment of pulp aeration • Air content ≠ Air ratio • Air ratio is a mechanical parameter only • Air content is a true measurement of pulp aeration Includes both hydraulic and physico-chemical effects relevant parameter that affects flotation efficiency Air content = %air / unit volume collection flotation rate efficiency constant Fpulp Fair Air ratio = Fair/Fpulp bubble surface area flux Ek Sb k 4 Sb is proportional to air content Leichtle (1998) COST FP1005 – Trondheim 24-26 oct 2012 14 Flotation monitoring : On-line measurement of air content • Dorris et al. (2006) Paprican sensor • Based on pressure difference between immersed gauges • Apparent pulp density varies with air content P Figure 1. Installation of air content probe in a Voith cell. • Installation • Installed on reject side, across the hatch, at an angle of 60° • probes installed in pre-flotation 1ry COST FP1005 – Trondheim 24-26 oct 2012 15 Materials and Methods ii) Bubble size measurement • Bubble collection via a sampling pipe and visualization in a glass window • Automated bubble count using a CCD camera and image analysis software (Sherlock 7) Viewing chamber CCD camera Halogen light source PC for image analysis Bubble size distribution D.Beneventi, Pilot verticell COST FP1005 – Trondheim 24-26 oct 2012 18 Flocculation Effect of concentration • Effect on fibre flocculation (flotation cell inlet) flocculation index (mm²) 8 8 g/L 7 16 g/L 6 5 4 3 2 Fibre fraction = 56% 1 0 0 5 10 15 20 concentration (g/L) • Pulp flocculation increases when increasing pulp concentration (8-16 g/L) • Higher crowding • More fibres interacting with each other (mainly governed by fibre concentration) • stronger flocculation COST FP1005 – Trondheim 24-26 oct 2012 19 Effect of concentration: industrial trials at Aylesford Air content 100 • Large variations of air content over time (in 1ry) • Higher air content contributes to better ink removal air % 95 yield index (%) efficiency • Higher air impairs flotation yield • Higher air content is caused by concentration decrease 90 conc. 85 80 % 75 70 52 54 56 58 60 35 85 80 25 75 20 70 15 10 65 Air 1ry, 3rd cell 5 Huber, P., Rousset, X., Zeno, E. and Vazhure, T. (2011) Ind. Eng. Chem. Res. 50(7) :4021-4028 60 55 ink removal 01/12/2008 08/12/2008 15/12/2008 40 35 air content (%) 30 ink removal (%) air content (%) 30 0 24/11/2008 (But take care, it is a question of compromise: a too high consistency will induce a too high decrease in ink removal efficiency) 64 Air 2ry, 1st cell To maximise the flotation yield, work at highest possible concentration while maintaining ERIC target 62 ink removal (%) 50 22/12/2008 1.7 1.5 1st stage cleaner feed concentration 1.3 1.1 25 0.9 20 0.7 15 10 0.5 0.3 Air 1ry, 3rd cell 5 0 24/11/2008 0.1 COST FP1005 – Trondheim 24-26 oct 2012 01/12/2008 08/12/2008 15/12/2008 -0.1 22 22/12/2008 concentration (%) • Flocculation Effect of dispersants • •On the fibre flocculation flocculation index (mm²) 4.5 4 Selected dispersants (Guar gum and CMC) effectively deflocculate DIP -14 to 19 % 3.5 +1% guar +2% guar +1% CMC +2% CMC 3 2.5 2 reference 1.5 guar gum 1 CMC 0.5 0 0 1 2 3 % additive (Flotation cell inlet) COST FP1005 – Trondheim 24-26 oct 2012 23 Effect of dispersants : pilot trials • Selected dispersants (guar gum or CMC) clearly improved flotation selectivity Ink removal (%) • Better ink removal + lower losses at the same time 60 58 56 54 52 50 48 46 44 42 40 •Increased air content thanks to pulp de-flocculation (at the fibre level) •Lower entrainment of fine elements thanks to depressing mechanism (from adsorbed dispersants) Guar gum 2% CMC CMC 1% CMC 1% guar 2% guar ref 0 10 20 30 40 50 (high overall losses because of low froth height) Total Losses (%) Zeno, E., Huber, P., Rousset, X., Fabry, B. and Beneventi D. (2010). Ind. Eng. Chem. Res., 2010, 49 (19), pp 9322–9329 COST FP1005 – Trondheim 24-26 oct 2012 25 Effect of dispersants : pilot trials Better selectivity : the link is fibre flocculation flocculation directly influenced the pulp aeration : Gas hold-up when flocculation 6 air content (%) 5 4 3 2 guar gum CMC 1 0 0 1 2 3 4 5 flocculation index (mm²) Zeno, E., Huber, P., Rousset, X., Fabry, B. and Beneventi D. (2010). Ind. Eng. Chem. Res., 2010, 49 (19), pp 9322–9329 COST FP1005 – Trondheim 24-26 oct 2012 26 Effect of dispersants : pilot trials Bubble size ? Rising velocity in Newtonian fluid : Drift flux model : Limited bubble size decrease Not sufficient to explain increased gas hold-up Higher drag on bubbles in de-flocculated pulp Zeno, E., Huber, P., Rousset, X., Fabry, B. and Beneventi D. (2010). Ind. Eng. Chem. Res., 2010, 49 (19), pp 9322–9329 COST FP1005 – Trondheim 24-26 oct 2012 27 Mechanisms • Improved mechanism (this work) • Lower (fibre) concentration or add fibre dispersants • de-flocculation • homogeneous fibre suspension • higher drag on air bubbles rise slowly to the top of the cell • increase of relative residence time air/pulp • air content is increased • ink removal is improved • (limited coalescence (surfactants) limited chanelling) COST FP1005 – Trondheim 24-26 oct 2012 28 Conclusions • Relationship among flocculation, ink removal, turbulence and air content is not simple • Depends on hydraulic regimes in the flotation cells (turbulence pattern) (lab cell ≠ pilot cell ≠ industrial cell ≠ various industrial flotation cells config.) Will affect interactions between air bubbles and pulp flocs • Pulp flocculation does impact flotation efficiency • Mechanisms involved Ink removal : – pulp de-flocculation (at fibre level) homogeneous fibre network higher higher drag higher air content better ink removal Losses : • – With concentration : mechanism not clear – With dispersants : Lower entrainment of fine elements thanks to depressing mechanism (from adsorbed dispersants) Selectivity ? • When increasing concentration : poorer ink removal, lower losses But little impact on selectivity • With added dispersants : clear selectivity increase (at least at lab and pilot scale) COST FP1005 – Trondheim 24-26 oct 2012 29