Internal Chloride Binding of OPC Pastes: Modelling Using Binding Isotherms Dr. Dipl. Eng. M.V.A. (Miruna) Florea L.C. (Lara) Quaas, MSc Prof. dr. ir. H.J.H. (Jos) Brouwers
Sustainable energy in the Netherlands 10 Other biomass combustion 8 Co ‐ firing biomass in power plants Energy production (%) Municipal solid waste incineration 6 Solar power 4 Biogas 2 Wind energy Hydropower 0 1990 1995 2000 2005 2010* Year • Biomass fly ashes • MSWI by-products
Environmental legislation Shaped Non-shaped building building IBC Contaminant materials materials materials (mg/m 2 ) (mg/kg) (mg/kg) Sb 8.7 0.32 0.7 As 260 0.9 2 Ba 1500 22 100 Cd 3.8 0.04 0.06 Cr 120 0.63 7 Co 60 0.54 2.4 Cu 98 0.9 10 Hg 1.5 0.02 0.08 Pb 81 2.3 2.1 Mo 144 1 15 Ni 400 0.44 8.3 Se 4.8 0.15 3 Sn 50 0.4 2.3 V 320 1.8 20 Zn 800 4.5 14 Br - 670 20 34 Cl - 110000 616 8800 F - 2500 55 1500 SO 4 2- 165000 1730 20000
MSWI by-product leaching • By-products often exceed the leaching limits of the Dutch environmental legislation for more than one contaminant (mostly heavy metals and anions) • Contaminants can be bound in cement-based mixtures • Chloride leaching is one of the largest challenges • It is often challenging to identify the source of chlorides in the by-products SCOPE • To study the individual chloride binding capacity of cement pastes, using various parameters
Chloride binding in cement pastes • Internal chloride binding experiments • Cement type: CEM I 42.5N from ENCI; w/b=0.5 • Chloride type: NaCl, CaCl 2 , SbCl 3 , CuCl 2 • Chloride concentrations: 0, 1, 3 and 5% • Static leaching with daily stirring, up to 42 days
Chloride leaching 12000 10000 1 d Leached Cl ‐ (mg/L) 5 d 8000 14 d 6000 21 d 4000 28 d 2000 35 d 42 d 0 Ref. 1% NaCl 3% NaCl 5% NaCl Added salt 12000 10000 1 d Leached Cl ‐ (mg/L) 5 d 8000 14 d 21 d 6000 28 d 4000 35 d 42 d 2000 0 Ref. 1% CaCl ₂ 3% CaCl ₂ 5% CaCl ₂
Chloride leaching 12000 10000 1 d 5 d Leached Cl ‐ (mg/L) 8000 14 d 21 d 6000 28 d 4000 35 d 42 d 2000 0 Ref. 1% SbCl ₃ 3% SbCl ₃ 5% SbCl ₃ Added salt 12000 10000 1 d 5 d Leached Cl ‐ (mg/L) 8000 14 d 21 d 6000 28 d 4000 35 d 42 d 2000 0 Ref. 1% CuCl ₂ 3% CuCl ₂ 5% CuCl ₂
The Paste Model The binding capacity of a cement paste is m w a function of: m w -type and ammount of hydration product -ion concentration in m CSH the pore solution -age and porosity of m hp the sample m CH m AFm C C C m b b b, CSH b, AFm … m b H.J.H. Brouwers, Cem. Concr. Res 34, 1697-1716, 2004 H.J.H. Brouwers, Cem. Concr. Res 35, 1922-1936, 2005
Hydration reactions: OPC C S+4.5H C SH +1.3CH 3 1.7 3.2 C S+3.5H C SH +0.3CH 2 1.7 3.2 C A + CS+14H C ASH 3 4 14 C A+CH+21H C AH 3 4 22 C A +3CS+36H C AS H 3 6 3 36 C AF+2C S+22H C AFS H +4CH 4 3 6 2 18 C AF+2C S+20H C AFS H +2CH 4 2 6 2 18 H.J.H. Brouwers, Cem. Concr. Res 34, 1697-1716, 2004 H.J.H. Brouwers, Cem. Concr. Res 35, 1922-1936, 2005
Chloride binding of the AFm phase • The ion exchange mechanism • The dissolution and precipitation mechanism Hydroxy-AFm: Monocarboaluminate: C 3 A.Ca(OH) 2 .nH 2 O C 3 A.CaCO 3 .nH 2 O Monosulfoaluminate: Hemicarboaluminate: C 3 A .½ CaCO 3 .½ Ca(OH) 2 .nH 2 O C 3 A.CaSO 4 .nH 2 O Friedel salt: Hemisulfoaluminate: C 3 A.CaCl 2 .nH 2 O C 3 A .½CaSO 4 .½Ca(OH) 2 .nH 2 O C C C b, AFm b, SO AFm b, HO AFm 4
Chloride binding isotherms 12 HCP1 10 bound Cl, mg/ g sample 8 6 4 HO ‐ AFm SO4 ‐ AFm 2 AFm 0 0 0.5 1 1.5 2 2.5 3 external Cl (mol/l) U.A. Birnin-Yauri and F.P. Glasser, Cem. Concr. Res. , V.28, N.12, p.1713-1723, 1998 H. Hirao, K.Yamada, H. Takahashi, H. Zibara, Vol.3, No.1, 77-84, 2005
XRD data: reference paste 900 CH 800 700 600 Intensity [a.u] CH 500 400 CaCO ₃ Etr CH 300 Etr 200 100 0 5 15 25 35 45 55 Position [ ᵒ Theta]
XRD data: samples a) 1% Cl, b) 3% Cl and c) 5% Cl a) 1% Cl, b) 3% Cl and c) 5% Cl from NaCl from CaCl 2
XRD data: samples a) 1% Cl, b) 3% Cl and c) 5% Cl a) 1% Cl, b) 3% Cl and c) 5% Cl from SbCl 3 from CuCl 2
Model adjustments 30 25 Bound Cl [mg/ g d.s.] 20 Experimental Cb_external 15 CSH-Z SO ₄ -AFm 10 HO-AFm 5 0 0 0,5 1 1,5 2 2,5 3 Chloride concentration [M] from NaCl • The degree of hydration was adjusted for CaCl 2 and CuCl 2 • HO- AFm is considered not to form in the presence of Cl - • NaCl samples leached a lot more sulphates than the others, so the amount of SO 4 -AFm in the sample was lowered • Not enough Cl - were available at lower chloride concentrations (0.28 M) to transform all SO 4 -AFm into Friedel’s salt.
Model adjustments 12 10 8 Bound Cl [mg/ g d.s.] Experimental 6 Cb internal CSH-Z SO ₄ -AFm 4 2 0 0 0,5 1 1,5 2 2,5 3 Chloride concentration [M] from NaCl
Model results Cb_final Experimental Cb_final Experimental 12 12 10 10 Bound Cl ‐ [mg/ g d.s.] Bound Cl ‐ [mg/ g d.s.] 8 8 6 6 4 4 2 2 0 0 0,28 0,82 1,33 0,28 0,82 1,33 Chloride concentration [M] from NaCl Chloride concentration [M] from CaCl 2
Model results Cb_final Experimental Cb_final Experimental 12 12 10 10 Bound Cl ‐ [mg/ g d.s.] Bound Cl ‐ [mg/ g d.s.] 8 8 6 6 4 4 2 2 0 0 0,28 0,82 1,33 0,28 0,82 1,33 Chloride concentration [M] from CuCl 2 Chloride concentration [M] from SbCl 3
Conclusions and further steps • The OPC paste recipes that contain NaCl leach more chlorides than the paste recipes that contain other chlorides. This is true for all chloride concentrations. • Recipes containing SbCl 3 result in a high leaching at an early stage for OPC which decreases over time again. • When the XRD results of samples containing 1% Cl from CaCl 2 and CuCl 2 are compared, it becomes clear that less portlandite is formed in the copper chloride samples. However, C 3 S peaks are visible due to the fact that there is still unreacted cement. • A Rietveld refinement will be attempted in order to validate the model • An internal chloride binding model was proposed, based on an external binding model and the XRD results from this study. • The final chloride binding model correctly predicts the bound chlorides from the four sources employed. • The binding of the associated cation will be further investigated and correlated with the chloride binding.
Questions, thoughts, ideas… Thank you!
Recommend
More recommend
