166 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
Gerdemann et al., 2007) but the downside is that they are
significantly less common than olivine.
Carbonation–Grind Size Relationship
A large number of papers[reference reviews?] assessed reac-
tion rates under a range of conditions, making it difficult to
do proper like-for-like comparisons of the effect of particle
size on carbonation. Nonetheless, a generic meta-analysis
comparing the mass% conversion for roughly similar con-
ditions (high pressure temperature and NaHCO3 addition)
gives a fairly consistent picture of 50–80 mass% olivine
conversion below a grind size of 50 µm (see Figure 2). The
Summers et al. (2005) and Gerdemann et al. (2007) data
may be outliers for a variety of reasons including material
preparation methods, carbonation conditions or difference
in feed composition. That said, there are plausible linear fits
(r2 0.8) that can be achieved for these individual datasets
so they may indicate differences in grind size-dependency
of the carbonation process. The aim of this paper is not
to assess the correctness of these carbonation trends, but
the differences do have implications for grind size opti-
misation, which will be explored later in this paper. In
that context, Summers et al. (2005) data is considered a
pessimistic scenario (little or no carbonation at P80 50
µm), Gerdemann et al. (2007) data represents an interme-
diate scenario (carbonation below P80 ~100 µm and the
continuation of the general 60 µm with the Garcia et al.
(2010) data, indicating carbonation up to P80s 250 µm, is
an optimistic scenario.
Carbonation Opportunities in Mining
Olivine is a common constituent of ultramafic Nickle and
Platinum Group Metal (PGM) ores. Milling of the ore is
driven by the need for metal recovery, but the tailings may
be suitable for carbonation as well. Examples of mining
projects considering this route include FPX Nickel and its
subsidiary CO2 Lock Corp. (FPX Nickel, 2022) as well as
Canada Nickel (2023). The presence of other minerals such
as brucite (Mg(OH)2 targeted by CO2 Lock Corp.), ser-
pentine, wollastonite and Mg-rich clays (Werdann, 2023)
in ores also warrant investigation from a carbon capture
perspective. Smelting slags often also contain Mg silicates
analogous to olivine, although these slags would still require
milling.
Another scenario the author has come across is exten-
sive mobilisation of Mg into solution during an Ni heap
0
10
20
30
40
50
60
70
80
90
100
0 50 100 150 200 250
P80 (μm)
Gadikota et al. (2014) Garcia et al. (2010)
Gerdemann et al. (2007) O'Connor et al (2005)
Summers et al. (2005) Overall 60μm trend
Summers et al. trend (pessimistic) Gerdemann et al. trend (intermediate)
Garcia et al./general trend (optimistic)
Figure 2. Compilation of olivine conversion data as a function of P
80 reported in various publications
%
Olivi
conversi

Previous Page Next Page

Extracted Text (may have errors)

166 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
Gerdemann et al., 2007) but the downside is that they are
significantly less common than olivine.
Carbonation–Grind Size Relationship
A large number of papers[reference reviews?] assessed reac-
tion rates under a range of conditions, making it difficult to
do proper like-for-like comparisons of the effect of particle
size on carbonation. Nonetheless, a generic meta-analysis
comparing the mass% conversion for roughly similar con-
ditions (high pressure temperature and NaHCO3 addition)
gives a fairly consistent picture of 50–80 mass% olivine
conversion below a grind size of 50 µm (see Figure 2). The
Summers et al. (2005) and Gerdemann et al. (2007) data
may be outliers for a variety of reasons including material
preparation methods, carbonation conditions or difference
in feed composition. That said, there are plausible linear fits
(r2 0.8) that can be achieved for these individual datasets
so they may indicate differences in grind size-dependency
of the carbonation process. The aim of this paper is not
to assess the correctness of these carbonation trends, but
the differences do have implications for grind size opti-
misation, which will be explored later in this paper. In
that context, Summers et al. (2005) data is considered a
pessimistic scenario (little or no carbonation at P80 50
µm), Gerdemann et al. (2007) data represents an interme-
diate scenario (carbonation below P80 ~100 µm and the
continuation of the general 60 µm with the Garcia et al.
(2010) data, indicating carbonation up to P80s 250 µm, is
an optimistic scenario.
Carbonation Opportunities in Mining
Olivine is a common constituent of ultramafic Nickle and
Platinum Group Metal (PGM) ores. Milling of the ore is
driven by the need for metal recovery, but the tailings may
be suitable for carbonation as well. Examples of mining
projects considering this route include FPX Nickel and its
subsidiary CO2 Lock Corp. (FPX Nickel, 2022) as well as
Canada Nickel (2023). The presence of other minerals such
as brucite (Mg(OH)2 targeted by CO2 Lock Corp.), ser-
pentine, wollastonite and Mg-rich clays (Werdann, 2023)
in ores also warrant investigation from a carbon capture
perspective. Smelting slags often also contain Mg silicates
analogous to olivine, although these slags would still require
milling.
Another scenario the author has come across is exten-
sive mobilisation of Mg into solution during an Ni heap
0
10
20
30
40
50
60
70
80
90
100
0 50 100 150 200 250
P80 (μm)
Gadikota et al. (2014) Garcia et al. (2010)
Gerdemann et al. (2007) O'Connor et al (2005)
Summers et al. (2005) Overall 60μm trend
Summers et al. trend (pessimistic) Gerdemann et al. trend (intermediate)
Garcia et al./general trend (optimistic)
Figure 2. Compilation of olivine conversion data as a function of P
80 reported in various publications
%
Olivi
conversi

Help

Destination page number Search scope Search Text

loading

XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 167
leaching process. In effect, this process has carried out
reaction (2) in the pursuit of Ni dissolution from a later-
ite ore, providing a residual process stream that is suitable
for carbonation. Likewise, Mg and Ca represent major
constituents of the waste stream from both salar (Li) and
geothermal brines. Concentrations are much lower than in
their equivalent ores, but nonetheless these waste streams
may represent future opportunities for carbonation. These
sorts of opportunities might serve to improve project eco-
nomics (depending on carbon tax for a given project) or,
more likely, improve ESG credentials of a mining operation.
METHODOLOGY
Construction Emissions
Aside from energy emissions during processing, the usage
of concrete and steel in process plant construction are major
CO2 emitters. A review of concrete and steel-related carbon
emissions (see Tables 1 and 2) produced average emissions
of 0.146 t CO2/t concrete and 1.97 t CO2/t steel.
The range of concrete and steel usage for process plant
construction is dependent on plant layout, mill configura-
tion and also the design philosophy and can be as low as
0.02 t steel/kW installed and 0.19 t concrete/kW installed
(Ausenco averages) but often are around 0.08 t steel/kW
installed and 0.4 t concrete/kW installed (Lane et al., 2016).
The approach taken in this methodology is to ‘amortise’ the
construction emissions over the financial payback period of
the plant to convert these values to a t CO2/t ore.
Operational Emissions
For the purposes of this methodology, emissions from raw
materials extraction can broadly be attributed to mining
activities, ore processing and concentrate transportation. A
study by NRC Canada (2005) reported an energy consump-
tion of 5.8 kWh/t ore for the mining process. Assuming
this is fully delivered by burning diesel (11.0 kWh/L) at
30% efficiency and emissions of 2.7 kg CO2/L diesel that
leads to emissions of 4.8 kg CO2/t ore.
Emissions from ore processing are dependent on the
type of process, but generally arise largely from the con-
sumption of energy and grinding media (Ballantyne et al.,
2012) in the comminution process. Both factors are impor-
tant considerations in plant design and well monitored/
reported during operation. The embodied emissions in
steel production provide a direct conversion from steel con-
sumption in comminution to CO2 emissions and depend-
ing on energy source, the specific energy consumption can
be translated directly to CO2 emissions. Table 3 lists several
quoted figures for CO2 emissions related to the production
of ceramics, giving an average of 0.288 t CO2/t ceramic. No
reference could be found specifically for ceramic grinding
media but the tile/refractory emissions values in Table 3 are
considered comparable to ceramic media. As a side-note,
these emissions are significantly lower than those associated
with steel. Given the lower wear rates typically seen with
ceramic media in stirred milling applications, this would
imply that stirred milling is significantly better from a car-
bon emissions better but a detailed trade-off between these
two milling methods is outside the scope of this conference
paper.
For CO2 sequestration purposes it is assumed that a
bulk olivine ore (i.e., dunite) is mined and that no flotation
or other upgrading is required. Emissions associated with
maintaining the high pressure and temperature required
for the carbonation reaction are not included in this cal-
culation. This paper aims to characterise the CO2 ‘budget’
before starting this reaction to provide an envelope that this
reaction would have to work within to allow the reaction
to proceed.
CARBON BALANCE
Based on the methodology and benchmarks outlined in
this paper, a carbon balance was set up for a 100 t/h plant
processing grinding a dunite feedstock (90% olivine) to
Table 1. Concrete production CO
2 emissions
Data Source
Emissions
(t CO
2 /t concrete)
Samarin (1999) 0.180
Hammond and Jones (2011) 0.150
Nazari and Sanjayan (2017) 0.144
Heidelberg Cement 0.132
Lafarge Holcim 0.129
Cemex 0.139
Table 2. Steel production CO2 emissions
Data Source
Emissions
(t CO2/t steel)
Bernstein et al. (2007) 2.00
Hammond and Jones (2011) 2.03
World Steel Association (2020) 1.89
Table 3. Ceramics production CO2 emissions
Data Source
Emissions
(t CO
2 /t ceramic)
Ceramic Sector Report, 2019 0.29
Ecofys, 2009 (refractory products) 0.23– 0.34
Ecofys, 2009 (tiles) 0.30
Mezquita et al., 2009 (tiles) 0.265

Previous Page Next Page

Extracted Text (may have errors)

XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 167
leaching process. In effect, this process has carried out
reaction (2) in the pursuit of Ni dissolution from a later-
ite ore, providing a residual process stream that is suitable
for carbonation. Likewise, Mg and Ca represent major
constituents of the waste stream from both salar (Li) and
geothermal brines. Concentrations are much lower than in
their equivalent ores, but nonetheless these waste streams
may represent future opportunities for carbonation. These
sorts of opportunities might serve to improve project eco-
nomics (depending on carbon tax for a given project) or,
more likely, improve ESG credentials of a mining operation.
METHODOLOGY
Construction Emissions
Aside from energy emissions during processing, the usage
of concrete and steel in process plant construction are major
CO2 emitters. A review of concrete and steel-related carbon
emissions (see Tables 1 and 2) produced average emissions
of 0.146 t CO2/t concrete and 1.97 t CO2/t steel.
The range of concrete and steel usage for process plant
construction is dependent on plant layout, mill configura-
tion and also the design philosophy and can be as low as
0.02 t steel/kW installed and 0.19 t concrete/kW installed
(Ausenco averages) but often are around 0.08 t steel/kW
installed and 0.4 t concrete/kW installed (Lane et al., 2016).
The approach taken in this methodology is to ‘amortise’ the
construction emissions over the financial payback period of
the plant to convert these values to a t CO2/t ore.
Operational Emissions
For the purposes of this methodology, emissions from raw
materials extraction can broadly be attributed to mining
activities, ore processing and concentrate transportation. A
study by NRC Canada (2005) reported an energy consump-
tion of 5.8 kWh/t ore for the mining process. Assuming
this is fully delivered by burning diesel (11.0 kWh/L) at
30% efficiency and emissions of 2.7 kg CO2/L diesel that
leads to emissions of 4.8 kg CO2/t ore.
Emissions from ore processing are dependent on the
type of process, but generally arise largely from the con-
sumption of energy and grinding media (Ballantyne et al.,
2012) in the comminution process. Both factors are impor-
tant considerations in plant design and well monitored/
reported during operation. The embodied emissions in
steel production provide a direct conversion from steel con-
sumption in comminution to CO2 emissions and depend-
ing on energy source, the specific energy consumption can
be translated directly to CO2 emissions. Table 3 lists several
quoted figures for CO2 emissions related to the production
of ceramics, giving an average of 0.288 t CO2/t ceramic. No
reference could be found specifically for ceramic grinding
media but the tile/refractory emissions values in Table 3 are
considered comparable to ceramic media. As a side-note,
these emissions are significantly lower than those associated
with steel. Given the lower wear rates typically seen with
ceramic media in stirred milling applications, this would
imply that stirred milling is significantly better from a car-
bon emissions better but a detailed trade-off between these
two milling methods is outside the scope of this conference
paper.
For CO2 sequestration purposes it is assumed that a
bulk olivine ore (i.e., dunite) is mined and that no flotation
or other upgrading is required. Emissions associated with
maintaining the high pressure and temperature required
for the carbonation reaction are not included in this cal-
culation. This paper aims to characterise the CO2 ‘budget’
before starting this reaction to provide an envelope that this
reaction would have to work within to allow the reaction
to proceed.
CARBON BALANCE
Based on the methodology and benchmarks outlined in
this paper, a carbon balance was set up for a 100 t/h plant
processing grinding a dunite feedstock (90% olivine) to
Table 1. Concrete production CO
2 emissions
Data Source
Emissions
(t CO
2 /t concrete)
Samarin (1999) 0.180
Hammond and Jones (2011) 0.150
Nazari and Sanjayan (2017) 0.144
Heidelberg Cement 0.132
Lafarge Holcim 0.129
Cemex 0.139
Table 2. Steel production CO2 emissions
Data Source
Emissions
(t CO2/t steel)
Bernstein et al. (2007) 2.00
Hammond and Jones (2011) 2.03
World Steel Association (2020) 1.89
Table 3. Ceramics production CO2 emissions
Data Source
Emissions
(t CO
2 /t ceramic)
Ceramic Sector Report, 2019 0.29
Ecofys, 2009 (refractory products) 0.23– 0.34
Ecofys, 2009 (tiles) 0.30
Mezquita et al., 2009 (tiles) 0.265

XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 165
this analysis is comminution flowsheet consideration based
on ore competency and hardness values as well as media
consumption and grind size. Insights regarding feed source,
grind size (i.e., surface area) and mineralogical composition
are also provided in the context of defining a suitable com-
minution flowsheet that maximizes CO2 absorption.
MINERAL CARBONATION
Olivine is the collective term for a silicate solid solu-
tion series ranging from fayalite (Fe2SiO4) to forsterite
(Mg2SiO4). The carbonation of forsterite to yield magne-
sium carbonate (magnesite) and silica is described in equa-
tions 1 to 3 (Haug, 2010).
CO2 +H2O ⇌ H2CO3 ⇌ HCO3– +H+
⇌ CO32– +H+ (1)
Mg2SiO4 +4 H+ ⇌ 2 Mg2+ +H4SiO4 (2)
Mg2+ +CO32– ⇌ MgCO3 (3)
Considering that water has a catalytic function in this reac-
tion, this gives the following net aqueous carbonation reac-
tion (4) (Haug, 2010):
Mg2SiO4 +2 CO2 → 2 MgCO3 +SiO2 (4)
This means that stoichiometrically, for every forsterite mol-
ecule two CO2 molecules can be reacted. Based on the
molar masses of the reactants, this equates to a maximum
theoretical CO2 sequestration capacity of 0.63 t CO2 per
tonne of forsterite. Aside from technical limitations, this
theoretical maximum is unlikely to be achieved because of
the fayalite-forsterite solid solution series (Fe to Mg oliv-
ine) and other impurities in the rock reducing the reactable
(Mg) content. Figure 1 illustrates the theoretical maximum
CO2 sequestration capacity based on forsterite content,
stoichiometry, the net carbonation reaction (equation 4)
and puts this in the perspective of different ultramafic rock
types. Realistically, 95–98% of the theoretical CO2 seques-
tration capacity of olivine would be a good practical maxi-
mum carbon sequestration limit for dunites.
Gerdemann et al. (2007) showed the optimum con-
ditions for olivine carbonation to be around 185 °C and
150 atm PCO2, showing this reaction requires quite extreme
conditions for favourable reaction kinetics. The addition
of 0.64M NaHCO3 increases ionic strength of the solu-
tion and aids dissolution of Si, ultimately preventing for-
mation of a Si-rich passifying layer and addition of 1M
NaCl is also commonly quoted as aiding the carbonation
reaction (O’Connor et al., 2005 Gerdemann et al., 2007
Wang et al., 2019). Other minerals such as serpentine
(Mg3Si2O5(OH)4) and wollastonite (CaSiO3) can also
be reacted but the majority of research is done on olivine.
These minerals require less extreme conditions (e.g., wollas-
tonite optimum carbonation at 100 °C and 40 atm PCO2
Figure 1. Theoretical CO2 sequestration capacity as a function of forsterite content

Previous Page Next Page

Extracted Text (may have errors)

XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 165
this analysis is comminution flowsheet consideration based
on ore competency and hardness values as well as media
consumption and grind size. Insights regarding feed source,
grind size (i.e., surface area) and mineralogical composition
are also provided in the context of defining a suitable com-
minution flowsheet that maximizes CO2 absorption.
MINERAL CARBONATION
Olivine is the collective term for a silicate solid solu-
tion series ranging from fayalite (Fe2SiO4) to forsterite
(Mg2SiO4). The carbonation of forsterite to yield magne-
sium carbonate (magnesite) and silica is described in equa-
tions 1 to 3 (Haug, 2010).
CO2 +H2O ⇌ H2CO3 ⇌ HCO3– +H+
⇌ CO32– +H+ (1)
Mg2SiO4 +4 H+ ⇌ 2 Mg2+ +H4SiO4 (2)
Mg2+ +CO32– ⇌ MgCO3 (3)
Considering that water has a catalytic function in this reac-
tion, this gives the following net aqueous carbonation reac-
tion (4) (Haug, 2010):
Mg2SiO4 +2 CO2 → 2 MgCO3 +SiO2 (4)
This means that stoichiometrically, for every forsterite mol-
ecule two CO2 molecules can be reacted. Based on the
molar masses of the reactants, this equates to a maximum
theoretical CO2 sequestration capacity of 0.63 t CO2 per
tonne of forsterite. Aside from technical limitations, this
theoretical maximum is unlikely to be achieved because of
the fayalite-forsterite solid solution series (Fe to Mg oliv-
ine) and other impurities in the rock reducing the reactable
(Mg) content. Figure 1 illustrates the theoretical maximum
CO2 sequestration capacity based on forsterite content,
stoichiometry, the net carbonation reaction (equation 4)
and puts this in the perspective of different ultramafic rock
types. Realistically, 95–98% of the theoretical CO2 seques-
tration capacity of olivine would be a good practical maxi-
mum carbon sequestration limit for dunites.
Gerdemann et al. (2007) showed the optimum con-
ditions for olivine carbonation to be around 185 °C and
150 atm PCO2, showing this reaction requires quite extreme
conditions for favourable reaction kinetics. The addition
of 0.64M NaHCO3 increases ionic strength of the solu-
tion and aids dissolution of Si, ultimately preventing for-
mation of a Si-rich passifying layer and addition of 1M
NaCl is also commonly quoted as aiding the carbonation
reaction (O’Connor et al., 2005 Gerdemann et al., 2007
Wang et al., 2019). Other minerals such as serpentine
(Mg3Si2O5(OH)4) and wollastonite (CaSiO3) can also
be reacted but the majority of research is done on olivine.
These minerals require less extreme conditions (e.g., wollas-
tonite optimum carbonation at 100 °C and 40 atm PCO2
Figure 1. Theoretical CO2 sequestration capacity as a function of forsterite content