- Email: [email protected]
Weathering of a MSW bottom ash heap: a modelling approach
Waste Management 22 (2002) 173–179 www.elsevier.com/locate/wasman
Weathering of a MSW bottom ash heap: a modelling approach Ph. Baranger*, M. Azaroual, Ph. Freyssinet, S. Lanini, P. Piantone BRGM, 3 avenue C.Guillemin, BP 6009, 45100 Orleans Cedex, France
Abstract Establishing plausible predictive scenarios represents a challenge for the long-term evolution of waste such as municipal solid waste bottom ash. These systems are characterized by complex and sometimes poorly understood physico-chemical mechanisms. The long term prediction of the evolution of such systems must be based on a dynamic approach involving their study in space and time. A preliminary outline of a model integrating chemistry and mass transfer is currently being tested by BRGM on the results obtained from a 16-month monitoring survey of a pilot bottom ash heap subjected to meteoric weathering. The model is based on a simplified coupled chemistry-transport approach using mass action laws and Kinetics chemical model (the Networks of Chemical Reactors approach). This modelling approach is used to monitor the evolution in chemical composition of a column of meteoric water percolating through the pilot bottom ash heap. The system is divided into representative elementary volumes (chemical reactors) on the basis of the major chemical and mineralogical zonations identified in the system. # 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
2. Results and discussion
Establishing predictive scenarios represents a difficult challenge for the long-term evolution of waste such as bottom ash. These systems are characterized by complex and sometimes poorly understood physico-chemical mechanisms. Predictive simulations must be based on a dynamic approach that describes the evolution of the bottom ash in space and time. BRGM has developed and tested a preliminary outline of a model integrating chemistry and mass transfer for Municipal Solid Waste (MSW) bottom ash. The model is based on a Network of Chemical Reactors (NCR) approach. It was elaborated on the basis of an experimental approach carried out by BRGM on a pilot MSW bottom ash heap (375 t) submitted to meteoric weathering. The experimental study involved identifying and quantifying the physico-chemical maturation mechanisms. A detailed description of the experiments is given by Freyssinet et al. [1]. Modelling results are compared with experimental data obtained from a 16month monitoring survey of the pilot heap.
2.1. Outlines of the Network of Chemical Reactor (NCR) approach
* Corresponding author. E-mail address: [email protected] (P. Baranger).
The NCR approach used by BRGM is inspired by that of Villermaux [2] and similar to that used in the IMPACT computer code developed by Jauzein et al. [3]. It is based on the perfect mixers assumption classically adopted in process engineering. In the NCR approach the studied medium is considered as an integrated geochemical and hydrodynamical system with well-defined physical boundaries. The NCR is designed on the basis of biogeochemical and hydrodynamical characteristics of the system. Reactors are defined by their positioning within the flow path network, by their volume, and by their initial chemical content (solid and aqueous phases) as indicated in Fig. 1. Each individual reactor represents a conceptual building block related to a specific area of the total system. Because of its flexibility, this simplified coupled modelling approach can be applied to a wide range of field of investigations (flow cell experiments, process engineering, mining pollution,. . .). BRGM has elaborated and tested a specific methodolodgy for NCR design and construction which is based on the modelling software allan. (modeller) and neptunix 4 (solver) distributed by Dassault Data Services. The combined use of these two tools can automatically generate the simulators providing that the processes
0956-053X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved. PII: S0956-053X(01)00066-6
174
Ph. Baranger et al. / Waste Management 22 (2002) 173–179
taken into account can be described by an ordinary algebro-differential equation system. Assembly and construction of the reactors are carried out graphically in allan. environment. The equations are simultaneously resolved by neptunix 4 at each time step, which makes it possible to simulate a strong coupling between chemistry and transport. Transport is limited to mass transfer between the connected reactors. The chemical part of each reactor is based on the Specific Chemical Simulator (SCS) concept. This concept consists of developing a chemical simulator for each case study. This is also made possible with the use of the allan. and neptunix 4 modelling tools. These tools and the SCS concept are described by Kerve´van et al. [4]. The main step in constructing an SCS is to list the relevant chemical processes to be considered. This work is based on all the mineralogical and chemical data obtained on the system. The SCS only includes reaction mechanisms specific to the studied problem. The size of the generated code is notably reduced compared to that of general chemical modelling software. The SCS is composed of elementary modules which correspond to either aqueous, dissolution/precipitation, gazeous or sorption reactions. All of these modules can then be stored in a model-library for future use. Equations which govern minerals, gazeous and sorption reactions can take into account kinetic laws. Other reactions are based on mass action laws. 2.2. Application of the NCR approach to the MSW bottom ash heap For the MSW bottom ash heap, the studied medium is limited to a reference column of meteoric water per-
colating through the pilot heap (Fig. 2). This column is divided into representative unit volumes (chemical reactors) on the basis of the major chemical zonations identified in the system. As a first approximation, we have divided the bottom ash heap into four chemical reactors, with the first and the last taken to represent the top (upper 10 cm) and the outlet of the heap respectively. Each reactor is considered as a perfect mixer where the solid phases interact with the aqueous phase. Solid/liquid and aqueous reactions are treated in each chemical reactor by the Specific Chemical Simulator of the system. SCS structure is based on the relevant chemical processes, the major chemical species and mineral phases to be considered in the system. The presented SCS is a preliminary version which only takes the major chemical elements and mineralogical phases into consideration. Once tested and validated, it will be completed by introducing the main chemical reactions in which pollutant metals such as Zn, Pb and Cu are involved. The SCS is now composed of 32 aqueous reactions, 16 dissolution/precipitation reactions of minerals and one gazeous reaction (Tables 1 and 2). This choice was based both on macroscopic and microscopic mineralogical observations of the bottom ash heap and on preliminary thermodynamic calculations carried out on the chemical composition of the input water (rainfall water) and of the outlet leachate waters. These preliminary thermodynamic calculations which aimed to reconstitute the complete chemical speciation of the waters were carried out using the geochemical modelling software EQ3/6 [5]. Simulations were carried out assuming a regular rainfall rate with a main value of 0.05 m3/month/m2. The chemical composition of the rain water is that of an
Fig. 1. General structure of a Network of Chemical Reactor (NCR).
Ph. Baranger et al. / Waste Management 22 (2002) 173–179
175
Fig. 2. The experimental design and the Network of Chemical Reactor (NCR) approach.
Table 1 Aqueous and gaseous reactions considered in the specific chemical simulator Aqueous reactions H2O =H++ OH HCO3 = H++ CO3 Ca++ + H++ CO3 =CaHCO+ 3 NaOH + H+ =Na++ H2O NaHSiO3 + H+ = Na+ + SiO2 + H2O KSO4=K+ SO4 KHSO4 = K+ + H+ +SO4 CaCO3 + H+ = Ca++ + HCO3 MgCO3 + H+ = Mg++ + HCO3 HSiO3 + H+ = H2O + SiO2aq (CO2)aq + H2O = H++ HCO3 CaSO4 = Ca++ + SO4 MgSO4 = Mg++ + SO4 ++ MgHCO+ + HCO3 3 = Mg MgCl+ = Mg+++ Cl CaCl+ = Ca+++ Cl Dissolution reaction of CO2 (CO2)g + H2O = HCO3 + H+
NaCl= Na+ + Cl KCl = + Cl CaCl2= Ca++ + 2 Cl Al(OH)4 + 4H+ = Al+++ + 4 H2O Al(OH)3 + 3H+ = Al+++ + 3 H2O NaAl(OH)4 = Al(OH)4 + Na+ KAl(OH)4 = Al(OH)4 + K+ KOH + H+ = K+ + H2O Mg4(OH)++++ + H+ = 4 H2O + 4 Mg++ 4 NaCO3 + H+ = Na+ + HCO3 NaHCO3 = Na+ + HCO3 NaSO4 = Na+ + SO4 H2SiO4 + 2 H+ = 2 H2O + SiO2aq Al(OH)++ + H+ = Al+++ + H2O + Al(OH)+ = Al+++ + 2 H2O 2 + 2H CaOH+ + H+ = Ca++ + H2O
176
Ph. Baranger et al. / Waste Management 22 (2002) 173–179
average rain observed over a year in the Tours region [6]. For reactor Nos. 1 and 4 the CO2 fugacity was supposed to be controled by the atmospher. Activity coefficients were calculated by Davis equations.
Obtained results generally show a good agreement between experimental and calculated data for some of the major chemical elements considered in the system. To illustrate the results of this preliminary modelling
Table 2 Dissolution/precipitation reactions considered in the specific chemical simulator Minerals
Reactions
Calcite Gypsum Portlandite Quartz Gibbsite Gismondine Kaolinite Glass 1 Glass 2 Diopside Ettringite Larnite Prehnite Gehlinite Rankinite
CaCO3 + H+ = Ca++ + HCO3 (CaSO4, 2 H2O) = Ca++ + SO4 + 2H2O 0.5 Ca(OH)2 = 0.5 Ca++ + OH SiO2 = SiO2aq Al(OH)3 + H2O = Al(OH}4 + H+ 0.25 (Ca2AI4Si4O16, 9 H2O) = 0.5 Ca++ + Al(OH)4 + SiO2 + 0.25 H2O 0.5 Al2Si2O5(OH)4 + 1.5 H2O = H+ + SiO2 + Al(OH)4 0.5 Na2SiO3 + H+ = 0.5 SiO2aq + Na+ + 0.5 H2O 0.5 K2SiO3 + H+ = 0.5 SiO2aq + K+ + 0.5 H2O 0.25 CaMgSi2O2 + H+ = 0.5 SiO2aq + 0.25 Mg++ + 0.5 H2O + 0.25 Ca++ 0.25 (Ca6Al2(SO4)3(OH)12, 26 H2O) + H+ = 1.5 Ca++ + 0.75 SO4 + 0.5 Al(OH)4 + 7.5 H2O 0.25 Ca2SiO6 + H+ = 0.25 SiO2aq + 0.5 Ca++ + 0.5 H2O 0.33 Ca2Al2Si3O1O(OH)2 + 0.66 H2O + 0.66 H+ = 0.66 Ca++ + SiO2 + 0.66 Al(OH4) 0.33 Ca2Al2SiO7 + H2O + 0.66 H+ = 0.66 Ca++ + 0.33 SiO2 + 0.66 Al(OH4) 1/6 Ca3Si2O7 + H+ = 0.33 SiO2aq + 0.5 Ca++ +0.5 H2O
Fig. 3. Examples of simulated and measured evolutions of (a) pH and of concentrations of (b) chloride, (c) sulphate, (d) sodium, (e) calcium and (f) potassium at the outlet of the bottom ash heap (reactor No. 4).
Ph. Baranger et al. / Waste Management 22 (2002) 173–179
exercise, we have reported examples of the chemical evolutions of some elements as a function of time in the aqueous phase (Figs. 3 and 4). Most of these chemical evolutions can be related to three main processes leading to the maturation of the MSW bottom ash through time: chloride and sulphate leaching, carbonation. Chloride and sulphate leaching derive from the dissolution of salts present in the heap (NaCl, KCl and CaCl2) and from the dissolution of gypsum, respectively. For mineral processes we have also reported the evolutions of some mineral masses as a function of time. As examples, Fig. 5 shows dissolution of ettringite and gypsum in the first and the fourth chemical reactors (top and bottom of the heap). These results also show precipitation of calcite at the top and the bottom of the heap. Such carbonation process results from the dissolution of minerals such as portlandite, larnite, ettringite, etc. that release large amounts of calcium and produce a high pH (pH> 12). The high pH of the leachates greatly
177
favours the dissolution of atmospheric CO2, which causes the precipitation of carbonates and oversaturation with respect to calcite. This process is corroborated by the observation of calcite at the top of the heap. Results show that this preliminary version of the model is capable of correctly reproducing the main trends observed experimentally for most of the major chemical elements of the system. However, as illustrated in Figs. 3 and 4, the model is not always capable to reproduce correctly the evolution of all chemical elements (for example, see potassium and silica in reactor Nos. 2 and 3). To improve these results, some modelling works are now in progress. They consist of refining the chemical assumptions concerning reaction mechanisms. They also consist in testing the influence of parameters like solid/solution ratios or dissolution/precipitation rate constants of the considered minerals (tests with bibliographical kinetic data and with ‘‘relative kinetics’’). Some simulations are also in progress using more
Fig. 4. Examples of simulated and measured evolutions of (a) calcium, (b) chloride, (c) sulphate, (d) sodium, (e) potassium and (f) silica concentrations in the bottom ash heap (lysimeter Nos. 2; reactor Nos. 2 and 3).
178
Ph. Baranger et al. / Waste Management 22 (2002) 173–179
Fig. 5. Examples of simulated evolutions of calcite, gypsum and ettringite masses in the chemical reactors of the system (g of minerals for 1 litre of H2O).
Ph. Baranger et al. / Waste Management 22 (2002) 173–179
realistic pluviometric data with a variable rate over the experimental period. Once tested and validated on an other MSW bottom ash system, the model will be completed by introducing the main chemical reactions in which pollutant metals such as Zn, Pb and Cu are involved.
3. Conclusions Results of this preliminary modelling exercise show that the main chemical trends of a MSW bottom ash heap submitted to weathering can be correctly reproduced by a simplified coupled chemistry-transport approach. The presented modelling approach is based on the concept of Network of Chemical Reactors (NCR approach) using a mechanistic chemical model based on Kinetics and mass action laws. This preliminary outline of the model only takes into account the major chemical elements of the system. Once tested and validated, the model will be completed by introducing the main chemical reactions in which pollutant metals such as Zn, Pb and Cu are involved. The main difficulty of this type of modelling concerns the high instability of certain solid phases under surface conditions and, above all, the lack of reliable thermodynamic data for such systems.
179
Works are now in progress to improve the presented model by refining and testing hypotheses adopted for reaction mechanisms and for parameters like solid/solution ratios or kinetics of dissolution/precipitation processes. References [1] Freyssinet Ph, Piantone P, Azaroual M, Itard Y, Clozel B, Baubron JC, Hau JM, Guyonnet D, Guillou-Frottier L, Pillard F, Jezequel JC. Evolution chimique et mine´ralogique des maˆchefers d’incine´ration d’ordures me´nage`res au cours de la maturation. Doc. BRGM n 280, 1998. [2] Villermaux J. Ge´nie de la re´action chimique, conception et fonctionnement des re´acteurs. Paris: Lavoisier, 1982. [3] Jauzein M, Andre´ C, Margrita R, Sardin M, Schweich D. A flexible computer code for modelling transport in porous media: IMPACT. Geoderma 1989;44:95–113. [4] Kerve´van C, Baranger Ph, Thie´ry D. SCS: Specific Chemical Simulators dedicated to chemistry-transport coupled modelling: part I. Design and construction of an SCS. Goldschmidt Conference Proceedings, Toulouse, France, September, 771–774, 1998. [5] Wolery TJ. EQ3NR, A computer program for geochemical aqueous speciation-solubility calculations: theoretical manual, user’s guide and related documentation (version 7.0). UCRL-MA110662-PT-I, Lawrence Livermore National Laboratory, 1992. [6] Grosbois C. Geochimie des eaux de la Loire: contributions naturelles et anthropiques, quantification de l’e´rosion, thesis, Tours University, 264, 1998.
Weathering of a MSW bottom ash heap: a modelling approach Ph. Baranger*, M. Azaroual, Ph. Freyssinet, S. Lanini, P. Piantone BRGM, 3 avenue C.Guillemin, BP 6009, 45100 Orleans Cedex, France
Abstract Establishing plausible predictive scenarios represents a challenge for the long-term evolution of waste such as municipal solid waste bottom ash. These systems are characterized by complex and sometimes poorly understood physico-chemical mechanisms. The long term prediction of the evolution of such systems must be based on a dynamic approach involving their study in space and time. A preliminary outline of a model integrating chemistry and mass transfer is currently being tested by BRGM on the results obtained from a 16-month monitoring survey of a pilot bottom ash heap subjected to meteoric weathering. The model is based on a simplified coupled chemistry-transport approach using mass action laws and Kinetics chemical model (the Networks of Chemical Reactors approach). This modelling approach is used to monitor the evolution in chemical composition of a column of meteoric water percolating through the pilot bottom ash heap. The system is divided into representative elementary volumes (chemical reactors) on the basis of the major chemical and mineralogical zonations identified in the system. # 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
2. Results and discussion
Establishing predictive scenarios represents a difficult challenge for the long-term evolution of waste such as bottom ash. These systems are characterized by complex and sometimes poorly understood physico-chemical mechanisms. Predictive simulations must be based on a dynamic approach that describes the evolution of the bottom ash in space and time. BRGM has developed and tested a preliminary outline of a model integrating chemistry and mass transfer for Municipal Solid Waste (MSW) bottom ash. The model is based on a Network of Chemical Reactors (NCR) approach. It was elaborated on the basis of an experimental approach carried out by BRGM on a pilot MSW bottom ash heap (375 t) submitted to meteoric weathering. The experimental study involved identifying and quantifying the physico-chemical maturation mechanisms. A detailed description of the experiments is given by Freyssinet et al. [1]. Modelling results are compared with experimental data obtained from a 16month monitoring survey of the pilot heap.
2.1. Outlines of the Network of Chemical Reactor (NCR) approach
* Corresponding author. E-mail address: [email protected] (P. Baranger).
The NCR approach used by BRGM is inspired by that of Villermaux [2] and similar to that used in the IMPACT computer code developed by Jauzein et al. [3]. It is based on the perfect mixers assumption classically adopted in process engineering. In the NCR approach the studied medium is considered as an integrated geochemical and hydrodynamical system with well-defined physical boundaries. The NCR is designed on the basis of biogeochemical and hydrodynamical characteristics of the system. Reactors are defined by their positioning within the flow path network, by their volume, and by their initial chemical content (solid and aqueous phases) as indicated in Fig. 1. Each individual reactor represents a conceptual building block related to a specific area of the total system. Because of its flexibility, this simplified coupled modelling approach can be applied to a wide range of field of investigations (flow cell experiments, process engineering, mining pollution,. . .). BRGM has elaborated and tested a specific methodolodgy for NCR design and construction which is based on the modelling software allan. (modeller) and neptunix 4 (solver) distributed by Dassault Data Services. The combined use of these two tools can automatically generate the simulators providing that the processes
0956-053X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved. PII: S0956-053X(01)00066-6
174
Ph. Baranger et al. / Waste Management 22 (2002) 173–179
taken into account can be described by an ordinary algebro-differential equation system. Assembly and construction of the reactors are carried out graphically in allan. environment. The equations are simultaneously resolved by neptunix 4 at each time step, which makes it possible to simulate a strong coupling between chemistry and transport. Transport is limited to mass transfer between the connected reactors. The chemical part of each reactor is based on the Specific Chemical Simulator (SCS) concept. This concept consists of developing a chemical simulator for each case study. This is also made possible with the use of the allan. and neptunix 4 modelling tools. These tools and the SCS concept are described by Kerve´van et al. [4]. The main step in constructing an SCS is to list the relevant chemical processes to be considered. This work is based on all the mineralogical and chemical data obtained on the system. The SCS only includes reaction mechanisms specific to the studied problem. The size of the generated code is notably reduced compared to that of general chemical modelling software. The SCS is composed of elementary modules which correspond to either aqueous, dissolution/precipitation, gazeous or sorption reactions. All of these modules can then be stored in a model-library for future use. Equations which govern minerals, gazeous and sorption reactions can take into account kinetic laws. Other reactions are based on mass action laws. 2.2. Application of the NCR approach to the MSW bottom ash heap For the MSW bottom ash heap, the studied medium is limited to a reference column of meteoric water per-
colating through the pilot heap (Fig. 2). This column is divided into representative unit volumes (chemical reactors) on the basis of the major chemical zonations identified in the system. As a first approximation, we have divided the bottom ash heap into four chemical reactors, with the first and the last taken to represent the top (upper 10 cm) and the outlet of the heap respectively. Each reactor is considered as a perfect mixer where the solid phases interact with the aqueous phase. Solid/liquid and aqueous reactions are treated in each chemical reactor by the Specific Chemical Simulator of the system. SCS structure is based on the relevant chemical processes, the major chemical species and mineral phases to be considered in the system. The presented SCS is a preliminary version which only takes the major chemical elements and mineralogical phases into consideration. Once tested and validated, it will be completed by introducing the main chemical reactions in which pollutant metals such as Zn, Pb and Cu are involved. The SCS is now composed of 32 aqueous reactions, 16 dissolution/precipitation reactions of minerals and one gazeous reaction (Tables 1 and 2). This choice was based both on macroscopic and microscopic mineralogical observations of the bottom ash heap and on preliminary thermodynamic calculations carried out on the chemical composition of the input water (rainfall water) and of the outlet leachate waters. These preliminary thermodynamic calculations which aimed to reconstitute the complete chemical speciation of the waters were carried out using the geochemical modelling software EQ3/6 [5]. Simulations were carried out assuming a regular rainfall rate with a main value of 0.05 m3/month/m2. The chemical composition of the rain water is that of an
Fig. 1. General structure of a Network of Chemical Reactor (NCR).
Ph. Baranger et al. / Waste Management 22 (2002) 173–179
175
Fig. 2. The experimental design and the Network of Chemical Reactor (NCR) approach.
Table 1 Aqueous and gaseous reactions considered in the specific chemical simulator Aqueous reactions H2O =H++ OH HCO3 = H++ CO3 Ca++ + H++ CO3 =CaHCO+ 3 NaOH + H+ =Na++ H2O NaHSiO3 + H+ = Na+ + SiO2 + H2O KSO4=K+ SO4 KHSO4 = K+ + H+ +SO4 CaCO3 + H+ = Ca++ + HCO3 MgCO3 + H+ = Mg++ + HCO3 HSiO3 + H+ = H2O + SiO2aq (CO2)aq + H2O = H++ HCO3 CaSO4 = Ca++ + SO4 MgSO4 = Mg++ + SO4 ++ MgHCO+ + HCO3 3 = Mg MgCl+ = Mg+++ Cl CaCl+ = Ca+++ Cl Dissolution reaction of CO2 (CO2)g + H2O = HCO3 + H+
NaCl= Na+ + Cl KCl = + Cl CaCl2= Ca++ + 2 Cl Al(OH)4 + 4H+ = Al+++ + 4 H2O Al(OH)3 + 3H+ = Al+++ + 3 H2O NaAl(OH)4 = Al(OH)4 + Na+ KAl(OH)4 = Al(OH)4 + K+ KOH + H+ = K+ + H2O Mg4(OH)++++ + H+ = 4 H2O + 4 Mg++ 4 NaCO3 + H+ = Na+ + HCO3 NaHCO3 = Na+ + HCO3 NaSO4 = Na+ + SO4 H2SiO4 + 2 H+ = 2 H2O + SiO2aq Al(OH)++ + H+ = Al+++ + H2O + Al(OH)+ = Al+++ + 2 H2O 2 + 2H CaOH+ + H+ = Ca++ + H2O
176
Ph. Baranger et al. / Waste Management 22 (2002) 173–179
average rain observed over a year in the Tours region [6]. For reactor Nos. 1 and 4 the CO2 fugacity was supposed to be controled by the atmospher. Activity coefficients were calculated by Davis equations.
Obtained results generally show a good agreement between experimental and calculated data for some of the major chemical elements considered in the system. To illustrate the results of this preliminary modelling
Table 2 Dissolution/precipitation reactions considered in the specific chemical simulator Minerals
Reactions
Calcite Gypsum Portlandite Quartz Gibbsite Gismondine Kaolinite Glass 1 Glass 2 Diopside Ettringite Larnite Prehnite Gehlinite Rankinite
CaCO3 + H+ = Ca++ + HCO3 (CaSO4, 2 H2O) = Ca++ + SO4 + 2H2O 0.5 Ca(OH)2 = 0.5 Ca++ + OH SiO2 = SiO2aq Al(OH)3 + H2O = Al(OH}4 + H+ 0.25 (Ca2AI4Si4O16, 9 H2O) = 0.5 Ca++ + Al(OH)4 + SiO2 + 0.25 H2O 0.5 Al2Si2O5(OH)4 + 1.5 H2O = H+ + SiO2 + Al(OH)4 0.5 Na2SiO3 + H+ = 0.5 SiO2aq + Na+ + 0.5 H2O 0.5 K2SiO3 + H+ = 0.5 SiO2aq + K+ + 0.5 H2O 0.25 CaMgSi2O2 + H+ = 0.5 SiO2aq + 0.25 Mg++ + 0.5 H2O + 0.25 Ca++ 0.25 (Ca6Al2(SO4)3(OH)12, 26 H2O) + H+ = 1.5 Ca++ + 0.75 SO4 + 0.5 Al(OH)4 + 7.5 H2O 0.25 Ca2SiO6 + H+ = 0.25 SiO2aq + 0.5 Ca++ + 0.5 H2O 0.33 Ca2Al2Si3O1O(OH)2 + 0.66 H2O + 0.66 H+ = 0.66 Ca++ + SiO2 + 0.66 Al(OH4) 0.33 Ca2Al2SiO7 + H2O + 0.66 H+ = 0.66 Ca++ + 0.33 SiO2 + 0.66 Al(OH4) 1/6 Ca3Si2O7 + H+ = 0.33 SiO2aq + 0.5 Ca++ +0.5 H2O
Fig. 3. Examples of simulated and measured evolutions of (a) pH and of concentrations of (b) chloride, (c) sulphate, (d) sodium, (e) calcium and (f) potassium at the outlet of the bottom ash heap (reactor No. 4).
Ph. Baranger et al. / Waste Management 22 (2002) 173–179
exercise, we have reported examples of the chemical evolutions of some elements as a function of time in the aqueous phase (Figs. 3 and 4). Most of these chemical evolutions can be related to three main processes leading to the maturation of the MSW bottom ash through time: chloride and sulphate leaching, carbonation. Chloride and sulphate leaching derive from the dissolution of salts present in the heap (NaCl, KCl and CaCl2) and from the dissolution of gypsum, respectively. For mineral processes we have also reported the evolutions of some mineral masses as a function of time. As examples, Fig. 5 shows dissolution of ettringite and gypsum in the first and the fourth chemical reactors (top and bottom of the heap). These results also show precipitation of calcite at the top and the bottom of the heap. Such carbonation process results from the dissolution of minerals such as portlandite, larnite, ettringite, etc. that release large amounts of calcium and produce a high pH (pH> 12). The high pH of the leachates greatly
177
favours the dissolution of atmospheric CO2, which causes the precipitation of carbonates and oversaturation with respect to calcite. This process is corroborated by the observation of calcite at the top of the heap. Results show that this preliminary version of the model is capable of correctly reproducing the main trends observed experimentally for most of the major chemical elements of the system. However, as illustrated in Figs. 3 and 4, the model is not always capable to reproduce correctly the evolution of all chemical elements (for example, see potassium and silica in reactor Nos. 2 and 3). To improve these results, some modelling works are now in progress. They consist of refining the chemical assumptions concerning reaction mechanisms. They also consist in testing the influence of parameters like solid/solution ratios or dissolution/precipitation rate constants of the considered minerals (tests with bibliographical kinetic data and with ‘‘relative kinetics’’). Some simulations are also in progress using more
Fig. 4. Examples of simulated and measured evolutions of (a) calcium, (b) chloride, (c) sulphate, (d) sodium, (e) potassium and (f) silica concentrations in the bottom ash heap (lysimeter Nos. 2; reactor Nos. 2 and 3).
178
Ph. Baranger et al. / Waste Management 22 (2002) 173–179
Fig. 5. Examples of simulated evolutions of calcite, gypsum and ettringite masses in the chemical reactors of the system (g of minerals for 1 litre of H2O).
Ph. Baranger et al. / Waste Management 22 (2002) 173–179
realistic pluviometric data with a variable rate over the experimental period. Once tested and validated on an other MSW bottom ash system, the model will be completed by introducing the main chemical reactions in which pollutant metals such as Zn, Pb and Cu are involved.
3. Conclusions Results of this preliminary modelling exercise show that the main chemical trends of a MSW bottom ash heap submitted to weathering can be correctly reproduced by a simplified coupled chemistry-transport approach. The presented modelling approach is based on the concept of Network of Chemical Reactors (NCR approach) using a mechanistic chemical model based on Kinetics and mass action laws. This preliminary outline of the model only takes into account the major chemical elements of the system. Once tested and validated, the model will be completed by introducing the main chemical reactions in which pollutant metals such as Zn, Pb and Cu are involved. The main difficulty of this type of modelling concerns the high instability of certain solid phases under surface conditions and, above all, the lack of reliable thermodynamic data for such systems.
179
Works are now in progress to improve the presented model by refining and testing hypotheses adopted for reaction mechanisms and for parameters like solid/solution ratios or kinetics of dissolution/precipitation processes. References [1] Freyssinet Ph, Piantone P, Azaroual M, Itard Y, Clozel B, Baubron JC, Hau JM, Guyonnet D, Guillou-Frottier L, Pillard F, Jezequel JC. Evolution chimique et mine´ralogique des maˆchefers d’incine´ration d’ordures me´nage`res au cours de la maturation. Doc. BRGM n 280, 1998. [2] Villermaux J. Ge´nie de la re´action chimique, conception et fonctionnement des re´acteurs. Paris: Lavoisier, 1982. [3] Jauzein M, Andre´ C, Margrita R, Sardin M, Schweich D. A flexible computer code for modelling transport in porous media: IMPACT. Geoderma 1989;44:95–113. [4] Kerve´van C, Baranger Ph, Thie´ry D. SCS: Specific Chemical Simulators dedicated to chemistry-transport coupled modelling: part I. Design and construction of an SCS. Goldschmidt Conference Proceedings, Toulouse, France, September, 771–774, 1998. [5] Wolery TJ. EQ3NR, A computer program for geochemical aqueous speciation-solubility calculations: theoretical manual, user’s guide and related documentation (version 7.0). UCRL-MA110662-PT-I, Lawrence Livermore National Laboratory, 1992. [6] Grosbois C. Geochimie des eaux de la Loire: contributions naturelles et anthropiques, quantification de l’e´rosion, thesis, Tours University, 264, 1998.