Abstract
Bentonite intrusion into a fracture intersecting the canister deposition hole is modelled. The model describes the expansion of the bentonite within the fracture. It accounts for the repulsive electrostatic double-layer forces, the attractive van der Waals forces and friction forces between the particles and the water. The model also takes into account the diffusion of the colloid particles in the smectite sol. Diffusion of a counterion, sodium, is accounted for as this strongly influences the double layer force and the viscosity of the gel/sol. The gel/sol is considered to be a fluid with a varying viscosity that is strongly dependent on the bentonite volume fraction in the gel and the sodium concentration in the water. Two different geometries were modelled; a rectangular and a cylindrical. The rectangular geometry was used to gain experience with the processes and mechanisms and how they interact since the cylindrical geometry was somewhat less stable numerically and more time consuming. In the rectangular geometry a fracture 1 metre long in the flow direction was modelled. In both geometries the fracture size was selected sufficiently large to ensure that the water velocity, near the distant border was nearly the same as the approaching water velocity and
More>>
Moreno, Luis;
Neretnieks, Ivars;
Liu, Longcheng
[1]
- Chemical Engineering and Technology, School of Chemical Science and Engineering, Royal Institute of Technology, Stockholm (Sweden)
Citation Formats
Moreno, Luis, Neretnieks, Ivars, and Liu, Longcheng.
Modelling of erosion of bentonite gel by gel/sol flow.
Sweden: N. p.,
2010.
Web.
Moreno, Luis, Neretnieks, Ivars, & Liu, Longcheng.
Modelling of erosion of bentonite gel by gel/sol flow.
Sweden.
Moreno, Luis, Neretnieks, Ivars, and Liu, Longcheng.
2010.
"Modelling of erosion of bentonite gel by gel/sol flow."
Sweden.
@misc{etde_1004739,
title = {Modelling of erosion of bentonite gel by gel/sol flow}
author = {Moreno, Luis, Neretnieks, Ivars, and Liu, Longcheng}
abstractNote = {Bentonite intrusion into a fracture intersecting the canister deposition hole is modelled. The model describes the expansion of the bentonite within the fracture. It accounts for the repulsive electrostatic double-layer forces, the attractive van der Waals forces and friction forces between the particles and the water. The model also takes into account the diffusion of the colloid particles in the smectite sol. Diffusion of a counterion, sodium, is accounted for as this strongly influences the double layer force and the viscosity of the gel/sol. The gel/sol is considered to be a fluid with a varying viscosity that is strongly dependent on the bentonite volume fraction in the gel and the sodium concentration in the water. Two different geometries were modelled; a rectangular and a cylindrical. The rectangular geometry was used to gain experience with the processes and mechanisms and how they interact since the cylindrical geometry was somewhat less stable numerically and more time consuming. In the rectangular geometry a fracture 1 metre long in the flow direction was modelled. In both geometries the fracture size was selected sufficiently large to ensure that the water velocity, near the distant border was nearly the same as the approaching water velocity and that the smectite concentration there was vanishingly small. It was found that the velocity of the fluid drops considerably where the bentonite volume fraction is larger than 1-2%. This is due to the strong increase in viscosity with increasing bentonite volume fraction. The loss of smectite by the slowly flowing fluid was found to be proportional to the square root of the seeping water velocity for the rectangular geometry. For the cylindrical geometry, the dependence is somewhat lower (exponent about 0.4) since the length of the gel/water interface decreases with increasing water flow rate. The penetration depth of the gel/water interface decreases with increasing water flow rate. For water velocity of the order of a metre per year the gel may penetrate several metres into the fracture when steady state is reached. The simulations were made with only sodium as counterion. Most simulations had sodium concentrations below the critical coagulation concentration, CCC. In the compacted bentonite at the fracture mouth it was 10 mM and 0.1 mM in the approaching water. At these concentrations the gel is expansive and can turn into a sol releasing colloidal particles. The low ion concentration has a strong impact on the fluid viscosity, which increases with decreasing ionic strength. At the same time, however the repulsion forces between the smectite particles increase causing a quicker expansion. Simulations with higher sodium concentrations had a marginal influence on the erosion rate. For the highest water flow rates the smectite loss could be up to 0.3 kg per year for one canister. This is more than one order of magnitude larger than what could be reached by smectite particle diffusion alone if fluid flow was neglected. In experiments in downward facing slits (fractures) it has been found that bentonite releases gel agglomerates much faster than expected. These are released and sediment also under conditions where it is expected that the smectite particles should have separated into individual smectite sheets, which would not noticeably be influenced by gravity. The reasons for this behaviour are not understood. In the modelling it is assumed that there are no other larger non-smectite particles that would be left behind to gradually build up a bed of particles that could act as filter, slowing down or even straining further smectite penetration into the fracture. The modelling results could therefore be highly pessimistic because bentonites contain tens of percent of accessory minerals that do not form colloids and the presence of which may cause the expansion to be slowed down by friction against the fracture walls}
place = {Sweden}
year = {2010}
month = {Nov}
}
title = {Modelling of erosion of bentonite gel by gel/sol flow}
author = {Moreno, Luis, Neretnieks, Ivars, and Liu, Longcheng}
abstractNote = {Bentonite intrusion into a fracture intersecting the canister deposition hole is modelled. The model describes the expansion of the bentonite within the fracture. It accounts for the repulsive electrostatic double-layer forces, the attractive van der Waals forces and friction forces between the particles and the water. The model also takes into account the diffusion of the colloid particles in the smectite sol. Diffusion of a counterion, sodium, is accounted for as this strongly influences the double layer force and the viscosity of the gel/sol. The gel/sol is considered to be a fluid with a varying viscosity that is strongly dependent on the bentonite volume fraction in the gel and the sodium concentration in the water. Two different geometries were modelled; a rectangular and a cylindrical. The rectangular geometry was used to gain experience with the processes and mechanisms and how they interact since the cylindrical geometry was somewhat less stable numerically and more time consuming. In the rectangular geometry a fracture 1 metre long in the flow direction was modelled. In both geometries the fracture size was selected sufficiently large to ensure that the water velocity, near the distant border was nearly the same as the approaching water velocity and that the smectite concentration there was vanishingly small. It was found that the velocity of the fluid drops considerably where the bentonite volume fraction is larger than 1-2%. This is due to the strong increase in viscosity with increasing bentonite volume fraction. The loss of smectite by the slowly flowing fluid was found to be proportional to the square root of the seeping water velocity for the rectangular geometry. For the cylindrical geometry, the dependence is somewhat lower (exponent about 0.4) since the length of the gel/water interface decreases with increasing water flow rate. The penetration depth of the gel/water interface decreases with increasing water flow rate. For water velocity of the order of a metre per year the gel may penetrate several metres into the fracture when steady state is reached. The simulations were made with only sodium as counterion. Most simulations had sodium concentrations below the critical coagulation concentration, CCC. In the compacted bentonite at the fracture mouth it was 10 mM and 0.1 mM in the approaching water. At these concentrations the gel is expansive and can turn into a sol releasing colloidal particles. The low ion concentration has a strong impact on the fluid viscosity, which increases with decreasing ionic strength. At the same time, however the repulsion forces between the smectite particles increase causing a quicker expansion. Simulations with higher sodium concentrations had a marginal influence on the erosion rate. For the highest water flow rates the smectite loss could be up to 0.3 kg per year for one canister. This is more than one order of magnitude larger than what could be reached by smectite particle diffusion alone if fluid flow was neglected. In experiments in downward facing slits (fractures) it has been found that bentonite releases gel agglomerates much faster than expected. These are released and sediment also under conditions where it is expected that the smectite particles should have separated into individual smectite sheets, which would not noticeably be influenced by gravity. The reasons for this behaviour are not understood. In the modelling it is assumed that there are no other larger non-smectite particles that would be left behind to gradually build up a bed of particles that could act as filter, slowing down or even straining further smectite penetration into the fracture. The modelling results could therefore be highly pessimistic because bentonites contain tens of percent of accessory minerals that do not form colloids and the presence of which may cause the expansion to be slowed down by friction against the fracture walls}
place = {Sweden}
year = {2010}
month = {Nov}
}