Assessment of Accelerated Carbonation Processes for CO2 Storage

Article pubs.acs.org/IECR

Assessment of Accelerated Carbonation Processes for CO2 Storage Using Alkaline Industrial Residues Daniela Zingaretti, Giulia Costa, and Renato Baciocchi* Laboratory of Environmental Engineering, Department of Civil Engineering and Computer Science Engineering, University of Rome “Tor Vergata”, Via del Politecnico 1, 00133 Rome, Italy ABSTRACT: An assessment of accelerated carbonation processes applied to industrial residues for CO2 storage was performed with the aim of gaining insight on the feasibility of this process at larger scale, identifying also the key factors for its implementation. Specifically, the material and energy requirements of two types of direct carbonation routes, the slurry phase process and the wet route, were evaluated, and the influence of the properties of the residues (i.e., main mineralogy and particles size) and of the reaction routes and operating conditions applied (i.e., temperature, CO2 pressure, and liquid-to-solid ratio) was analyzed. In addition, the possibility of directly treating diluted sources of CO2, thus avoiding the capture step, was assessed for the wet route by performing lab-scale tests with gas mixtures containing 40% or 9% CO2.

1. INTRODUCTION Mineral carbonation is one of the processes currently under investigation for storing anthropogenic carbon dioxide emissions, commonly identified as one of the main causes for the variations observed in the global climate.1 The basic concept behind mineral CO2 sequestration is to mimic natural weathering processes by which calcium- or magnesiumcontaining minerals are converted into the corresponding carbonates reacting with gaseous carbon dioxide:2

conditions for the carbonation of minerals are typically energy intensive; for example, direct carbonation of minerals is found to be effective at a temperature of 373−473 K and pressure of 1−10 MPa.5,7 An alternative feedstock suitable for mineral carbonation is represented by some types of alkaline industrial residues characterized by high calcium or magnesium (hydr)oxide or silicate contents. The use of alkaline industrial residues instead of minerals in the carbonation process may present in fact some advantages. These materials prove to be generally more reactive compared to natural minerals, hence, suggesting that high CO2 uptakes could be in principle achieved applying less energy intensive operating conditions. In addition, in many cases these materials present a particle size already suitable for carbonation allowing to reduce or avoid the need of a grinding step that represents one of the most energy intensive stages of the process. Furthermore, several studies have demonstrated that carbonation may prove to be an effective treatment for reducing the release of specific metals from alkaline waste materials, thus making their reuse or disposal easier.8−10 Although the availability of residues with suitable properties for use as feedstock for accelerated carbonation processes is considered too limited to substantially reduce CO2 emissions on a global scale, carbonation processes based on these types of materials could help to establish the mineral carbon sequestration technology, since they could be seen as a stepping-stone toward the development of CCS processes making use of natural minerals.11 Besides, these residues are typically available at CO2 point-source emission locations, such as waste to energy facilities and steel and cement manufacturing plants and could allow storage for at least part of the CO2 emissions of the same industries from which they are produced. By far the most

(Ca, Mg)SiO3(s) + CO2(g) → (Ca, Mg)CO3(s) + SiO2(s) (1)

Differently from other storage techniques based on the injection of carbon dioxide in geological formations or depleted oil fields, mineral carbonation does not require long-term monitoring, since CO2 is incorporated into a chemically and thermodynamically stable mineral phase, avoiding any possibility of CO2 release.3 To apply carbonation as an ex situ storage option for fixing large amounts of CO2, this reaction, which occurs spontaneously in nature for geological timeframes, has to be accelerated by the application of specific operating conditions. In the last decades several studies have investigated the possibility of storing CO2 by accelerated carbonation of natural minerals, since, considering the abundance on the earth crust of minerals such as olivine, wollastonite and serpentine, the storage of large amounts of CO2 could be anticipated.4,5 However, in order to achieve a significant CO2 uptake in shorter timeframes (hours instead of thousands of years), the surface of the minerals needs to be activated by physical pretreatments, including size reduction, magnetic separation, and thermal or steam treatment (to eliminate chemically bound water). Upon activation, the mineral may be carbonated in the aqueous phase by two different routes: (a) the indirect route, where the alkaline metal is first extracted from the silicate matrix and then precipitated as carbonate in a separate step, in which CO2 is also dissolved; (b) the direct carbonation route, where CO2 and alkaline metal dissolution and carbonation occur in the same step.6 Operating © 2013 American Chemical Society

Special Issue: Massimo Morbidelli Festschrift Received: Revised: Accepted: Published: 9311

October 31, 2013 December 20, 2013 December 20, 2013 December 20, 2013 dx.doi.org/10.1021/ie403692h | Ind. Eng. Chem. Res. 2014, 53, 9311−9324

Industrial & Engineering Chemistry Research

Article

reaction routes, experiments have been mostly carried out in batch mode employing 100% CO2 flows, that is, assuming to apply the carbonation process after a capture step aimed at obtaining a concentrated CO2 stream from flue gas, syngas, or other diluted CO2 sources. Slurry phase carbonation experiments have been performed under enhanced conditions on several types of residues (steel slag in particular) to assess the maximum CO2 uptakes achievable, as well as the reaction yield of the material and the influence of operating conditions on process kinetics.2,20,21 A maximum CO 2 uptake of about 180 g CO 2 /kg, corresponding to a conversion of 74% of the Ca content of the slag, was achieved in 30 min at 1.9 MPa CO2 pressure and 373 K for basic oxygen furnace (BOF) steel slag with a particle size < 38 μm.2 Recently, very significant CO2 uptakes (between 200 and 400 g CO2/kg) were obtained in experiments carried out at 373 K, 1 MPa CO2 pressure, and a L/S ration of 5 L/kg for reaction times between 1 and 24 h on two types of BOF steel slag and on electric arc furnace (EAF) stainless steel slag, milled below 150 μm.20 As for the wet route carbonation experiments, these have been performed on a wide range of residues, generally applying L/S ratios between 0.1 and 0.4 L/kg and relatively mild operating conditions (ambient to 323 K and 0.1−1 MPa) to investigate the potentially achievable CO2 uptakes as well as the effects on the leaching behavior resulting for these materials. One of the analyzed residues that exhibited the highest uptake at mild operating conditions (303 K, 0.3 MPa CO2 and L/S = 0.2 L/kg) were APC residues (around 220 g CO2/kg after 1 h).9 The CO2 uptakes achieved for steel slag appear to vary depending on the type of analyzed slag; for example for treatment times of 1 h values of 100−120 g CO2/kg were reported for BOF slag and of 140 g CO2/kg for EAF slag (milled < 150 μm, at 323 K, 1 MPa CO2 and L/S = 0.3 L/ kg);20 significantly higher uptakes were reported instead for AOD slag, 240 g CO2/kg for (at 323 K, 1 MPa CO2 and L/S = 0.4 L/kg),10 as well as for a mixture of stainless steel slag 180 g CO2/kg (milled < 125 μm, at ambient T, 0.3 MPa CO2 and L/ S = 0.125 L/kg).22 On the basis of the results of lab-scale studies a few assessments regarding the energy penalties of slurry phase carbonation of industrial residues were published.23,25 Huijgen and co-workers23 assessed the energy penalties of slurry phase carbonation of steel slags concluding that the minimum energy penalty in terms of CO2 emissions was 17%, (at L/S = 5 kg/kg, t = 15 min, T = 473 K, pCO2 = 2 MPa, d < 38 μm), making reference to a power plant characterized by a heat release of 18 GJ/tCO2 emitted. In the study of Kelly et al.24 a preliminary energy balance at industrial scale for two CO2 mineralization pathways (i.e., using industrial caustics and industrial wastes) was performed. The resulting energy penalties for the CO2 carbonation process based on steelmaking slag proposed by Huijgen et al.,23 associated to a coal-fired power plant characterized by a heat release of around 10 GJ/tCO2 emitted, were estimated to be greater than 100% making this process unlikely feasible for these assumptions. Finally, Kirchofer et al.25 estimated the energy and material requirements of different aqueous carbonation processes in order to compare their environmental performance by life cycle assessment (LCA). The minimum energy penalty expressed in terms of percent CO2 emitted per CO2 sequestered, considering natural gas as the energy source, was obtained for cement kiln dust at

applied reaction route tested for the carbonation of residues is the direct aqueous route, since the dissolution of Ca-containing phases and the precipitation of Ca carbonates can be generally efficiently carried out under the same operating conditions in a one-stage operation.12−14 1.1. Short Review of Direct Aqueous Carbonation of Industrial Residues for CO2 Storage. The types of alkaline residues that have been selected and tested so far for CO2 storage by accelerated carbonation include different types of steelmaking slag, residues from thermal treatment processes such as biomass ash, ash from oil shale and lignite combustion, waste incineration residues, such as bottom ash and air pollution control (APC) residues, as well as byproducts from cement and paper manufacturing.11,12,14,15 The Ca content may provide a first indication of the reactivity of a material with CO2, since it has been shown to correlate quite well with the achievable CO2 uptakes.12−14 Cabased phases, in fact, generally present a higher reactivity even under rather mild operating conditions, as opposed to Mg, Fe, and Al ones. Hence, the residues that have shown to yield the highest CO2 sequestration potential are those generated in steel refining processes, such as argon oxygen decarburization (AOD) and ladle furnace (LF) slag, APC residues, and byproducts from cement manufacturing.14 However, besides bulk chemical composition, the reactivity of industrial residues with CO2 depends greatly also on their mineralogy; free oxide and hydroxide phases such as lime and portlandite, typically found in APC residues, oil shale ash, cement kiln dust (CKD), and paper mill waste in fact have been shown to be very reactive with CO2 even at mild operating conditions,9,12,15,16 while the reactivity of many of the different Ca and Mg containing silicates detected, for example, in metallurgical slag, varies depending on the types of crystal phases and of the presence of inclusions of various elements, such as Al, Cr, Fe, etc.14 In general, the reactivity of Ca silicates is indicated to increase for higher Ca/Si ratios;17 silicates with these characteristics are generally retrieved in materials containing cementitious phases, such as concrete demolition waste17,18 or steel manufacturing slag. 10 Particle size, as previously mentioned, is another critical parameter for the reactivity of alkaline residues with CO2. Byproducts of flue gas treatment units, cement kiln dust, and other ashes or residues from industrial refining processes, such as AOD slags or deinking ash from paper manufacturing, present average grain sizes (generally below 100−150 μm) that are already in the optimum range for carbonation, whereas slags and waste from construction and demolition activities present a wider particle size distribution and a significant percentage of coarse particles and consequently typically exhibit a low reactivity without mechanical pretreatment.14 Direct aqueous carbonation of alkaline residues has been tested basically in two different modes: by the slurry phase route at liquid-to-solid (L/S) ratios higher than 2 L/kg2,19,20 and through the wet (or thin-film) route, applying L/S ratios below 1 L/kg.9,10,12,16 This latter route is specifically suitable for the carbonation of industrial residues that present high contents of soluble elements such as alkaline metals, thus avoiding the treatment and disposal of the processing liquid that would be otherwise eluted in the processing liquid of a slurry phase process; in addition it has been shown to favor the dissolution kinetics of hydroxide and silicate phases under mild operating conditions, which may allow to more easily reach the saturation conditions of carbonate mineral phases.21 For both types of 9312

dx.doi.org/10.1021/ie403692h | Ind. Eng. Chem. Res. 2014, 53, 9311−9324


Article

Table 1. Slurry Phase Carbonation Operating Conditions d0

d1

T

pCO2

L/S

t

RCO2

R

EFF

ID

residue

mm

mm

K

MPa

L/kg

h

t/t CO2

RX

t/t CO2

ref

S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19 S20

EAF1 EAF1 EAF1 EAF1 EAF1 EAF1 BOF1 BOF1 BOF1 BOF1 BOF1 BOF1 BOF1 BOF1 BOF1 BOF1 BOF2 BOF2 BOF3 BOF3

1.19 1.19 1.19 1.19 1.19 1.19 20 20 20 20 20 20 20 20 20 20 1.19 1.19 1.19 1.19

0.15 0.15 0.15 0.15 0.15 0.15 0.052 0.052 0.052 0.052 0.052 0.052 0.052 0.052 0.023 0.052 0.15 0.15 0.15 0.15

323 373 423 323 373 373 373 373 373 373 373 373 323 423 473 473 373 373 373 373

1 1 1 1.9 1 1 0.1 0.9 2 2 1.9 2 1.9 1.9 2 1 1 1 1 1

10 10 10 10 5 5 10.0 10.0 5.0 2.0 10 10 10 10 5 10 5 5 5 5

1 1 1 1 0.5 1 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.25 0.5 1 0.5 1 0.5

2.31 2.31 2.31 2.31 2.31 2.31 4.65 4.65 4.65 4.65 4.65 4.65 4.65 4.65 4.65 4.65 2.76 2.76 2.93 2.93

0.16 0.12 0.09 0.09 0.49 0.44 0.53 0.58 0.63 0.66 0.60 0.57 0.55 0.76 0.66 0.49 0.66 0.50 0.73 0.75

14.29 20.00 25.00 25.00 4.76 5.26 8.71 8.01 7.33 7.00 7.77 8.17 8.52 6.14 7.00 9.52 4.17 5.56 4.00 3.92

21 21 21 21 20 20 2 2 2 2 2 2 2 2 2 2 20 20 20 20

Table 2. Wet Carbonation Operating Conditions (u.d. = unpublished data) d0

d1

T

pCO2

L/S

t

RCO2

R

EFF

ID

residue

mm

mm

K

MPa

L/kg

h

t/t CO2

RX

t/t CO2

ref

W1 W2 W3 W4 W5 W6 W7 W8 W9 W10 W11 W12 W13 W14 W15 W16 W17 W18 W19 W20 W21 W22 W23 W24 W25 W26 W27

AOD1 AOD1 AOD1 EAF1 EAF1 EAF1 EAF1 APC1 APC1 APC1 APC1 APC1 APC1 APC1 APC1 APC1 APC2 BOF2 BOF2 BOF3 BOF3 BOF4 BOF4 BOF4 BOF5 BOF5 BOF5

0.15 0.15 0.15 1.19 1.19 1.19 1.19 0.075 0.075 0.075 0.075 0.075 0.075 0.075 0.075 0.075 0.075 1.19 1.19 1.19 1.19 1.19 1.19 1.19 1.19 1.19 1.19

0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.075 0.075 0.075 0.075 0.075 0.075 0.075 0.075 0.075 0.075 0.15 0.15 0.15 0.15 0.125 0.125 0.125 0.125 0.125 0.125

323 323 323 323 323 323 323 303 303 303 303 303 313 323 303 303 323 323 323 323 323 323 323 323 323 323 323

0.1 0.3 1 0.1 0.3 1 1 0.3 0.3 0.3 0.3 0.3 0.3 0.3 1 0.1 0.3 1 1 1 1 0.1 0.3 0.9 0.1 0.3 0.9

0.4 0.4 0.4 0.4 0.4 0.4 0.3 0.02 0.1 0.2 0.4 0.6 0.02 0.02 0.02 0.02 0.2 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3

1 1 1 1 1 1 0.5 1 1 1 1 1 1 1 1 1 1 1 1 1 0.5 1 1 1 1 1 1

2.11 2.11 2.11 2.59 2.59 2.59 2.59 3.22 3.22 3.22 3.22 3.22 3.22 3.22 3.22 3.22 3.75 3.43 3.43 3.93 3.93 3.17 3.17 3.17 2.86 2.86 2.86

0.45 0.43 0.45 0.23 0.39 0.37 0.32 0.26 0.68 0.69 0.72 0.67 0.68 0.77 0.23 0.24 0.70 0.41 0.33 0.39 0.33 0.31 0.36 0.39 0.23 0.28 0.31

4.73 4.97 4.66 11.10 6.72 6.92 8.00 12.50 4.72 4.65 4.44 4.78 4.76 4.17 14.29 13.33 5.38 8.33 10.53 10.00 11.76 10.08 8.84 8.04 12.36 10.05 9.29

10 10 10 10 10 10 10 9 9 9 9 9 9 9 9 9 this study 20 20 20 20 u.d. u.d. u.d. u.d. u.d. u.d.

ambient temperature and pressure conditions,15 resulting equal to 14%, while the results obtained for the other considered materials were 34% for steel slag2 and 45% for fly ash.19 As can be noted, the energy penalties evaluated in these works differ considerably from one study to the other, even considering the different power plant types used as reference, also when the

evaluation was based on the same experimental data, probably as a consequence of the differences in the assumptions made and in the selected process layout. 1.2. Aims of This Study. As clearly shown by the above overview, many lab-scale studies were performed on several types of industrial residues, applying both direct aqueous 9313



Article

Table 3. Initial Composition (% wt) of the Residues Listed in Tables 1 and 2 Estimated on the Basis of the Data Reported in the Corresponding Reference Ca2SiO4 CaSiO3 Ca3SiO5 SiO2 Ca(OH)2 CaOHCl Mg(OH)2 MgO CaCO3 Fe3O4 CaF2 FeO NaCl Cr2O3 Al2O3 MnO

AOD1

EAF1

92.6

75.5

APC1

APC2

BOF1

BOF2

BOF3

BOF4

BOF5

19.8

17.2

30.9

12.1

56.8 26.3

32.1

2.9 40.0 30.6 2.7 4.6 0.1 0.05

0.06

4.20 3.60 4.60 0.03

24.4 51

19.1

10.8

10.3

13.8

5.1 4.0

6.5

7.7

6.4 0.9

35.5

37.6

28.0

26.6

11.5

17

8.4 34.9

5.9 19.6

17.1

7.9 4.1 22.1

2.4 4.6

and its initial particle size (d0), the particle size of the material treated in the carbonation reactor (d1), the reaction temperature (T), the CO2 pressure (pCO2), the liquid-to-solid (L/S) ratio and the experiment duration (t). Table 1 and Table 2 report the specific carbonation conditions considered for the slurry phase and the wet route, respectively. As can be noted, for the slurry phase route experiments carried out at T = 323− 473 K, pCO2 = 0.1−2 MPa, t = 0.25−1 h and L/S = 2−10 L/kg were considered, while for the wet route the selected operating conditions were T = 303−323 K, pCO2 = 0.1−1 MPa, t = 0.5−1 h and L/S = 0.02−0.6 L/kg. Nine alkaline industrial residues were considered, that is, two samples of APC residues collected from the baghouse section of a hospital waste incineration plant,9 the basic oxygen furnace steel slags (BOF1) tested in carbonation experiments by Huijgen et al.,2 four different samples of basic oxygen furnace slags (BOF2, BOF3, BOF4, and BOF5) collected from an integrated steel manufacturing Italian plant,20 electric arc furnace slags (EAF1), and ladle slags produced at the outlet of argon oxygen decarburization and desulfurization units (AOD1) in a stainless steel making plant.20,21 The main composition of the AOD1, EAF1, APC1, APC2, BOF2, BOF3, BOF4, and BOF5 samples was estimated on the basis of the results obtained by elemental composition measurements, XRD analysis, acid neutralization capacity measurements, and calcimetry analysis,9,10,20,21 whereas the composition reported by Huijgen et al.23 was considered for the BOF1. Table 3 reports the chemical composition assumed for each residue before carbonation. To estimate the chemical composition and the physical properties of the final product obtained after carbonation, it was assumed that in the carbonation reactor one or more of the reactions reported in eq 2−8 take place depending on the composition of the treated residues and on the type of reaction route applied. In particular eqs 7 and 8 were taken into account only for the EAF1, BOF2, and BOF3 slags treated by the slurry phase route for which dolomite was detected in the product.20 The reactivity of iron containing phases, which anyhow was quite limited, was not considered in this assessment.

carbonation routes and various operating conditions. However, few applications of this process at larger scale have been developed and tested up to now. One of the reasons that in our view has so far hindered the scale-up of this technology consists in the uncertainty regarding the feasibility of accelerated carbonation, especially as far as material and energy requirements are concerned. For this reason we decided to perform an assessment of the material and energy requirements of the direct carbonation of alkaline industrial residues. This assessment is based on the experimental results of previously published works analyzing the influence of the properties of the residues (i.e., main mineralogy and particles size) and of the reaction routes and operating conditions (i.e., temperature, CO2 pressure, and liquid-to-solid ratio) applied. In particular, we focused also on the wet carbonation route which has not been evaluated in terms of energy requirements so far. In this context, as the available data set is based on tests performed with pure CO2 only, new lab-scale tests were carried out with gas mixtures simulating the composition of either syngas or flue gas, assuming to operate the carbonation process directly with diluted sources of CO2, thus avoiding the capture step.

2. METHODS 2.1. Mass and Energy Balance. The material and energy balance was evaluated on the basis of the results of selected labscale tests performed on different types of residues (see section 2.1.1) and making reference to the two process layouts reported in section 2.1.2. The total energy requirement of the process (expressed in MJ/tCO2) was estimated summing the specific energy requirements associated to each unit operation (see section 2.1.3), taking into account the conversion efficiency of electrical energy into mechanical energy for each unit and considering a 35% efficiency of conversion of thermal energy into electrical energy. 2.1.1. Carbonation Data Set. The energy consumption associated with the carbonation process applying the slurry phase or the wet route was evaluated considering the results of selected lab scale carbonation tests,2,9,10,20,21 limiting the data set to experiments carried out at residence times lower than 1 h, that was considered the maximum time to allow for a technically feasible carbonation process. For each carbonation test the following parameters were considered: type of residue

Ca 2SiO4 + 2CO2 → 2CaCO3 + SiO2 9314

(2)



Article

Figure 1. Carbonation process layouts considered for the slurry phase (a) and the wet route (b). Adapted with permission from ref 27. Copyright 2013, Elsevier.

CaSiO3 + CO2 → CaCO3 + SiO2

(3)

Ca3SiO5 + 3CO2 → 3CaCO3 + SiO2

(4)

Ca(OH)2 + CO2 → CaCO3 + H 2O

(5)

2CaOHCl + CO2 → CaCO3 + CaCl 2 + H 2O

(6)

these residues compared to the other types of tested materials. However, the effective amount of residues to be used in each case per unit of stored CO2 (REFF) proved to change significantly depending on the specific operating conditions applied; in general the lowest REFF values were obtained for the BOF2 slags in the slurry phase route and for the APC1 and AOD1 slag in the wet route. 2.1.2. Process Layouts. Two different carbonation process routes were taken into account, that is, the slurry phase and the wet route; the corresponding process layouts, already discussed in Zingaretti et al.,27 are reported in Figure 1. In the slurry phase route (see Figure 1a), the industrial residues, after being ground to a specific particle size (unit A), are mixed with water at a given liquid-to-solid ratio (unit B); the resulting slurry is pumped (unit C) to the heat exchanger (unit D) where it is heated to 30 K below the reactor temperature, and then is fed to the carbonation reactor (unit F) after flowing through a heater (unit E) required to achieve the desired reaction temperature. The CO2 is pressurized in a multiple stage compressor (unit G) to reach the desired pressure and fed to the reactor in which the carbonation reaction takes place. The heat of the slurry leaving the carbonation reactor is recovered in unit D; after cooling, the slurry is separated (unit H), producing a solid product and a liquid stream that could be in principle recycled to unit B, although this option was not considered in this work. It is worth noting that the process layout considered for the slurry phase route is similar to the one proposed by O’Connor et al.7 and Huijgen et al.23 The process layout considered for the wet route (see Figure 1b) comprises a lower number of unit operations, and the energy consumption is mainly associated with CO2 compression (unit G), size reduction of the residues (unit A), and the operation of the carbonation reactor (unit K), envisioned for this application as a rotary drum.27 2.1.3. Energy Requirements for Each Unit Operation. 2.1.3.1. Size Reduction. Following the approach of O’Connor et al.7 and Hangx et al.,28 the different crushing and grinding steps needed to reach the desired final particle size were considered. First, in order to achieve a particle size of 50 mm, based on data from the U.S. Bureau of Mining,7 a primary crushing step with an associated energy requirement of around

Mg(OH)2 + 2CO2 + CaO → MgCa(CO3)2 + H 2O (7)

MgO + 2CO2 + CaO → MgCa(CO3)2

(8)

To estimate the energy requirement associated to the size reduction unit (S.R.U.) the values of the work index (Wi) to be used for each residue were taken from Perry and Green.26 In particular Wi was assumed to be equal to 12.16 kWh/t for the seven types of steel slags, whereas for the APC residues the work index value reported for cement raw material was used (i.e., 10.57 kWh/t). On the basis of the mass flow rate of CO2 to store (i.e., mCO2), the amount of residues required in the process for each carbonation condition was determined. The mass ratio of residues necessary to convert a unit mass of CO2 into the solid carbonate phase (RCO2) was estimated according to O’Connor et al.7 considering calcium (hydr)oxides and silicates phases and magnesium ones only for EAF1, BOF2, and BOF3 slags in the slurry phase route. Furthermore, the reactivity of the residues at the operating conditions of the carbonation process was expressed in terms of RX that indicates the extent of conversion of the calcium and/or magnesium (hydr)oxides or silicate phases to the corresponding carbonates. In this way we accounted for the kinetics of carbonation, although in a simplified way, assuming the extent of carbonation (Rx) in the carbonation reactor to be equal to the one achieved in the experiments carried out at the same residence time (t) in a batch reactor. On the basis of RCO2 and RX, the effective amount of residues to be used for each set of operating conditions (REFF) was estimated. The values of RCO2, RX and REFF estimated for each carbonation condition are reported in Table 1 and Table 2. As it can be noticed, the AOD slags presented the lowest RCO2 values (i.e., around 2.1 t residues/t CO2, respectively) indicating the higher potential reactivity of 9315



Article

carbonation temperature and to cool the outlet slurry from the carbonation reactor down to 313 K. The supplementary heat required to reach the final carbonation temperature was assumed to be provided in unit E. The overall energy requirement associated to the heating of the slurry was estimated considering the energy required to achieve the final carbonation temperature, the energy recovered in the heat exchanger by cooling the outlet stream from the carbonation reactor, the heat of reaction of the carbonation process, and the heat loss from the carbonation reactor. Namely, the energy required for heating the slurry to the desired temperature was calculated by eq 12, where m and CP are the mass flow rate and the specific heat capacity of the slurry, respectively, and ΔT is the difference between the inlet and the outlet temperature.

2.0 kWh/t residues was assumed. After this step a second grinding step was considered, and the associated energy requirement was determined on the basis of the amount of residues to be treated (REFF) and on the grinding work estimated by the application of Bond’s equation reported in eq 9,26 in which d0 is the original particle size of the residues, d1 is the theoretical sieve size through which 80 wt % of the residues pass, and Wi is the standard Bond’s work index of the residues. ⎛ 1 W = 10Wi ⎜⎜ − ⎝ d1

1 d0

⎞ ⎟⎟ ⎠

(9)

In the case of a final grain size (d1) lower than 70 μm, an empirical multiplying factor was applied to Bond’s equation as suggested by Perry and Green26 leading to the relationship reported in eq 10. ⎛ 1 W = 10Wi ⎜⎜ − ⎝ d1

1 d0

⎞ 10.3 + d 1 ⎟⎟ · 1.145 · d ⎠ 1

E HEAT = (10)

VμG2 mCO2

(12)

The same equation was also used to calculate the energy recovered by cooling the outlet stream from the carbonation reactor in the heat exchanger. The specific heat capacity of the slurry to be heated or cooled was estimated considering the chemical composition of the solid product before and after the carbonation reaction (estimated as described in section 2.1.1), the amount of water fed and the dependence of the specific heat capacity on temperature, according to the equations reported in Table 4.

Finally, to achieve a final particle size lower than 37 μm, a tertiary grinding step was included and an associated energy requirement of 150 kWh/t residues was taken into account according to direct measurements carried out for minerals.7 On the basis of the values of d0 and d1 of the considered residues (see Table 1 and Table 2), in this study the energy requirements associated to the size reduction unit (S.R.U.) were mainly estimated considering the formulation of Bond’s equation reported in eq 9. 2.1.3.2. Liquid−Solid Mixing. The energy required for mixing the residues with water was estimated by the application of eq 11, where V is the total volume of water and residues to be mixed, μ is the dynamic viscosity, and G is the average velocity gradient.

EMIX =

mCP ΔT mCO2

Table 4. Heat Capacity Equations (J/mol K) Used to Estimate the CP of the Residues phase Ca2SiO4 MgO

(11)

CaCO3 Fe3O4

In particular, by setting the dimensions of one mixing unit and considering the total volume to be mixed in each carbonation test, the required number of units was estimated together with the overall energy requirement associated to this unit operation. The value of the average velocity gradient for the mixer (unit B) and the carbonation reactor (unit F) was chosen considering a typical rapid mixing operation (i.e., G = 500 s−1) while for the liquid−solid separation step applied downstream of the carbonation reactor (unit G) a value typical for flocculation processes was selected (i.e., G = 60 s−1). 2.1.3.3. Slurry Pumping. The energy required to pump the slurry was estimated by multiplying the flow rate of the slurry for the total dynamic head applied in each carbonation condition considering the final pressure required in the carbonation reactor and the friction head of the pipeline. Specifically, as already done by Huijgen et al.,23 the total pressure to be reached by the slurry pump was estimated considering the partial CO2 pressure and the water vapor pressure reported in the literature for the specific temperature applied in each carbonation condition. 2.1.3.4. Heating. For the slurry phase route the heat required for reaching the reaction temperature was assumed to be provided in two steps by means of a heat exchanger (unit D) and a heater (unit E). Specifically, in unit D it was assumed to heat the inlet slurry from 293 to 30 K below the final

CaF2 SiO2 Ca(OH)2 Mg(OH)2 CaSiO3 FeO NaCl CaOHCl Ca3SiO5 Cr2O3 MnO Al2O3 CaCl2 MgCa(CO3)2 H2O 9316

heat capacity equation −4

= 2.4871 × 10 − 8.3145 × 10 T − 2.0521 × 103T0.5 − 9.0774 × 104T−2 c0p = 65.211 − 1.2699 × 10−3T −387.24T−0.5 − 4.6185 × 105T−2 c0p = 99.715 + 2.6920 × 10−2T −2.1576 × 106T−2 c0p = −3.5580 × 103 + 3.3473 × 102T0.5 −9.3090T + 2.5388 × 10−3T2 + 1.4273 × 105T−1 0 cp = −24.692 + 5.8095 × 10−2T + 1.8706 × 103T−0.5 −2.8774 × 106T−2 c0p = 2.3306 −7.7765 × 10−2T + 1.9237 × 10−5T2 −3.3753 × 103T−0.5 + 2.6036 × 106T−2 c0p = 1.8667 × 102 − 2.1911 × 10−2T − 1.5998 × 103T−0.5 0 cp = 1.0222 × 102 + 1.5107 × 10−2T − 2.6172 × 106T−2 0 cp = 1.1125 × 102 + 1.4373 × 10−2T + 16.936T−0.5 − 2.7779 × 106T−2 c0p = 67.352 + 3.7580 × 10−3T − 3.8167 × 102T−0.5 + 3.1570 × 105T−2 c0p = 45.151 + 1.7974 × 10−2T c0p = 18.181 + 8.345 × 10−3T + 0.216 × 106T−2 + 0.166 × 10−6T2 0 cp = 161.317 + 0.14234T − 5.9152256 × 10−5T−2 − 2090270T−2 c0p = 1.1902 × 102 + 9.4964 × 10−3T − 3.4045T−0.5 − 1.4419 × 106T−2 0 cp = 59.749 + 3.6 × 10−3T − 2.8265 × 102T−0.5 − 3.1362 × 104T−2 0 cp = 1.5736 × 102 + 7.1899 × 10−4T − 9.8804 × 102T−0.5 − 1.8969 × 106T−2 c0p = 76.846 + 6.6490 × 10−6T2 − 1.2847 × 106T−3 c0p = 5.4788 × 102 − 0.16759T + 7.7076 × 10−5T2 − 6.5479 × 103T−0.5 0 cp = 75.19 c0p

2

ref 29 29 29 29 29 29 29 29 29 29 29 30 31 29 29 29 29 29 29



Article

Figure 2. Effect of temperature on the energy requirements estimated for carbonation of BOF1 (a) and EAF1slags (b). For the specific operating conditions see Table 1.

The heat released by the carbonation reactions taken into account in our process (eq 2−8) was estimated on the basis of the heat of reaction calculated at the specific operating conditions applied and considering the Peng−Robinson’s state equation for carbon dioxide. The heat losses from the carbonation reactor walls to ambient air were estimated assuming the stainless steel reactor to be insulated with a glass wool layer and considering natural convection on the exterior ambient-air side and conduction on the inner side. For the wet carbonation route, instead, the overall energy requirement associated to the heating of the reactants was estimated considering only the energy required to reach the final carbonation temperature according to eq 12. In this case, in fact, the recovery of the energy produced by the exothermic carbonation reaction was neglected since it was considered to be difficult to recover because, differently from the slurry phase process, the outlet stream from the carbonation reactor is basically made up of a solid material. 2.1.3.5. CO2 Compression. The energy requirement associated to CO2 compression was estimated assuming to use a multiple stage compressor to increase the carbon dioxide pressure from 0.1 to 7.38 MPa and a boosting pump to compress CO2 to higher pressure, as reported by McCollum and Ogden.32 Specifically, it was assumed to use a 5-stage compressor, with the energy requirement for each stage (ESTAGE,i) estimated by applying eq 13, where Zs is the average CO2 compressibility for each stage, R is the gas constant, Tin is the CO2 temperature at the compressor inlet, M is the molecular weight of CO2, ηis is the isentropic efficiency of the compressor, ks is the average ratio of the specific heats of CO2 for each individual stage, and CR is the compression ratio of each stage. ⎛ Z RT ⎞⎛ k ⎞ s ESTAGE, i = ⎜⎜ s in ⎟⎟⎜ ⎟[(CR)ks− 1/ ks − 1] ⎝ Mηis ⎠⎝ ks − 1 ⎠

The values of the coefficients to be used for both the compressor and the boosting pump were chosen according to McCollum and Ogden.32 2.1.3.6. Operation of the Wet Carbonation Reactor. To estimate the energy consumption associated to the rotation of the drums used as carbonation reactors in the wet route, the energy requirement of one drum was first estimated considering the resistive torque due to kinetic friction and the constant angular velocity of the drum. The total energy consumption for this unit operation was then estimated calculating the number of drums required in each condition by setting the size of the drum and the volume of residues to be fed in each unit. 2.2. Carbonation Experiments with Diluted Sources of CO2. Wet-route carbonation tests with diluted sources of CO2 were performed by applying the same experimental setup already described in previous works (e.g., Baciocchi et al.9) and using five types of the above-mentioned industrial residues already tested with pure CO2, i.e. APC1, APC2 residues, AOD1 slags, BOF4, and BOF5 slag. In short, the tests were performed in a 150-mL pressurized stainless steel reactor placed in a thermostatic bath, and the reaction temperature was set at 303 K for APC1 residues and at 323 K for the other materials. In each run, three 1 g samples of dry residues were homogeneously humidified with deionized water to adjust the L/S ratio to the desired value, placed in tin foil containers and exposed to the CO2 gas source for set times (from 0.5 to 24 h). Experiments were carried out using a gas mixture containing 9% CO2, 21% O2, and 70% N2 vol. to simulate flue gas from combustion processes, a gas mixture containing 40% CO2 and 60% N2 vol. to simulate a syngas from gasification processes, and 100% CO2 as a comparison. Gas humidity was maintained at 75% by placing a saturated NaCl solution in the reactor, while gas pressure was varied in the range 0.1−1 MPa. After carbonation, the samples were oven-dried at 378 K and analyzed to determine their carbonate content by calcimetry analysis. The extent of carbonation (RX) was calculated as described in section 2.1.1 considering calcium (hydr)oxides or silicate phases as the only reactive phases.

(13)

The energy requirement associated to the boosting pump, instead, was calculated according to eq 14, where Pfin and Pin are the final and initial CO2 pressures, respectively, and ρ is the density of CO2 during pumping. E BOOST.PUMP =

(Pfin − Pin) ρ

3. RESULTS AND DISCUSSION 3.1. Mass and Energy Balance: Slurry Phase Route. 3.1.1. Effect of the Operating Conditions. The effect of temperature, liquid-to-solid ratio, and CO2 pressure on the

(14) 9317



Article

Figure 3. Effect of the liquid to solid ratio on the energy requirements estimated for carbonation of BOF1 (a) and EAF1 slags (b). For the specific operating conditions see Table 1.

Figure 4. Effect of CO2 pressure on the energy requirements estimated for carbonation of BOF1 (a) and EAF1 slags (b). For the specific operating conditions see Table 1.

al.2 The slower reaction kinetics exhibited by the EAF1 slag in comparison to the BOF1 may be related to the differences in particles size (i.e., 150 against 52 μm) as well as to the different composition in terms of reactive phases. As far as the effect of the applied liquid-to-solid ratio is concerned, the energy requirements reported in Figure 3 showed a similar behavior for the two types of residues: an increase of the L/S ratio led in both cases to a significantly higher total energy requirement of the process. This trend can be mainly correlated to the observed decrease in the reactivity of the material when operating at higher L/S ratios, which implies a larger amount of residues and water to be processed in each unit. However, the effect of the liquid-to-solid ratio on the reactivity of the residues was more pronounced for the EAF1 slags than the BOF1 slag tested by Huijgen et al.2 This result may be ascribed not only to the differences in the mineralogical composition of the residues but also to the different type of stirring methods employed in the two studies; in fact Huijgen et al.2 used a hollow turbine to mix the slurry and feed the CO2 hence probably enhancing the carbonation efficiency also at high L/S ratios, whereas in Baciocchi et al.21 only magnetic stirring was adopted, thus the liquid-to-solid ratio may have represented a more critical issue. Figure 4 reports the effect of carbon dioxide partial pressure on the energy requirements observed for the two types of industrial residues. For the BOF1 slags (Figure 4a) the total

energy requirements estimated for the slurry phase carbonation of EAF1 slag (a) and BOF1 slag (b) are reported in Figure 2, Figure 3, and Figure 4, respectively. As can be seen, for the slurry phase route, regardless of the type of residues, heating resulted by far the most energy intensive operation followed by size reduction, in good agreement with the results reported by Huijgen et al.23 Figure 2 shows that the energy requirements resulting for the two types of residues presented different trends with operating temperature. On the one hand for the BOF1 (see Figure 2a) the total energy requirement of the process showed to decrease with increasing temperature as the material proved to be more reactive at a higher temperature. On the other hand the total energy requirement of the process with EAF1 slag (see Figure 2b) proved to be negatively influenced by an increase of the operating temperature from 323 to 423 K due to the decreasing reactivity of the material, as reported in Table 1. This latter result can be explained considering that the evaluation of the energy requirements of the process was carried out on the basis of the results obtained for 1 h lab-scale tests. The extent of carbonation achieved for the EAF1 slags for times shorter than 2 h proved to be rather low and characterized by decreasing Rx values for increasing temperature.21 However, for longer experiment durations, higher RX values were obtained, and the effect of temperature on the extent of carbonation proved to be in good agreement with the results reported by Huijgen et 9318



Article

includes the energy requirements associated to the mixing of the slurry, the carbonation reactor and the clarifier. The main contributions to the total energy requirements of the slurry phase carbonation process are associated to the size reduction of the residues and carbon dioxide compression. It should be noted in fact that, except for case S5, in all of the cases considered in Figure 5 the energy required for heating the slurry could be totally recovered from the heat of reaction. The total energy requirement resulting for this route ranged between 1300 and 2750 MJ/t CO2, with the highest duty associated to the carbonation of BOF1 slags (case S10). As can be seen from Table 1, these residues presented a much higher energy duty associated to grinding, namely 1500 MJ/t against the 350−400 MJ/t estimated for the other residues. This difference is due to the lower particle size of the material treated in the carbonation reactor (d1), but especially to the greater initial particle size (d0 = 20 mm) of the BOF1 slag with respect to other steel slags. However, it should be noted that in many steel making industries the slag is currently ground at the plant for metals recovery thus presenting a particle size similar to that of the EAF1, BOF2, and BOF3 slags. 3.1.2. Comparison with Previous Studies. The total energy requirements of the slurry phase carbonation process estimated following the approach outlined in section 2.1 are compared in Figure 6 with those reported for the same type of residues and operating conditions in previous studies.23−25 In particular, in Figure 6a the energy requirements estimated following the approach presented in this study for carbonation of BOF1 in the S15 case (at 473 K, d1 < 38 μm, pCO2 = 2 MPa, L/S = 5 L/kg, t = 15 min) are reported together with the results of Huijgen et al.23 and Kelly et al.24 In Figure 6b, instead, the energy penalties estimated in the present work and the results reported by Kirchofer et al.25 for the carbonation of BOF1 in the condition S7 (373 K, d1 < 100 μm, pCO2 = 0.1 MPa, L/S = 10 L/kg, t = 30 min) are compared. For completeness and further discussion, Figure 6 panels a and b also report the energy requirements estimated following the approach outlined in this work, but assuming to neglect the recovery of the heat of the outlet stream from the carbonation reactor. Remarkable differences can be noticed between the results obtained in these studies, which can be substantially

energy requirement of the process seemed to remain almost the same at pressures from 0.1 to 1.9 MPa. This may be due on the one hand to the increase in the reactivity of the material observed for higher pCO2, leading to a slight decrease of the amount of residues to be processed and on the other hand to an increase of the energy requirement associated to gas compression. For the EAF1 slag, instead, as shown in Figure 4b, an increase in the carbon dioxide partial pressure implied a higher total energy requirement of the process due both to the increased energy requirements associated to the compressor and to the reduction of the reactivity of the material with higher pCO2 values. Figure 5 summarizes the lowest energy requirements achievable for each considered residue applying the slurry

Figure 5. Comparison of the lowest energy requirement estimated for each type of residues applying the slurry phase carbonation route. For the specific operating conditions see Table 1.

phase route process, highlighting the contribution of the main unit operations relevant for this route; the term ″other″

Figure 6. Comparison of the energy requirement estimated by applying the present approach and the ones reported in previous studies for the same types of residues and operating conditions: (a) carbonation condition S15 (b) carbonation condition S7. 9319



Article

Figure 7. Effect of the CO2 pressure on the energy requirements estimated for wet carbonation of AOD slags (a), EAF slags (b), APC1 residues (c), and BOF4 slags (d). For the specific operating conditions see Table 2.

taken into account, in fact, as done in this paper and in the assessment carried out by Huijgen et al.,23 the total energy requirement of the process would considerably decrease from around 19000 MJ/t CO2 estimated by Kelly et al.24 to roughly 7000 MJ/t CO2. Regarding the comparison of the results obtained by Kirchofer et al.25 and the energy requirements estimated for the same operating conditions in this paper (see Figure 6b) also in this case, the information included by Kirchofer et al.25 does not clarify whether the recovery of the heat of the outlet stream from the carbonation reactor was considered. Nevertheless, looking at Figure 6b, it can be noticed that the total energy requirement reported by Kirchofer et al.25 proved to be only slightly higher than the one estimated using the approach presented in this paper considering the heat recovery term. This observation leads to the conclusion that these authors have probably assumed to consider (at least partially) the heat recovery term; if the heat recovery term is neglected, in fact, the energy requirement associated to the process is shown to drastically increase from around 14600 to 75200 MJ/t CO2. The comparative analysis discussed in this section highlights the great influence of heat recovery on the energy requirements of the carbonation process. Besides the other hypotheses, the assumption made on this term of the energy balance is

ascribed to the different energy requirements estimated for the heating of the reagents (residues and water). The results obtained in this work including the heat recovery term proved to be in good agreement with those reported by Huijgen et al.23 (see Figure 6a). The layout and assumptions considered in these two studies, in fact, are quite similar and the slight difference that can be noted is mainly ascribed to the different assumption made for the energy balance of the heat exchanger (unit D in Figure 1a). Namely, in the present work it was assumed to use this unit to heat the inlet slurry from 293 to 30 K below the final carbonation temperature, while Huijgen and co-workers opted for heating to 20 K below the carbonation temperature.23 A more relevant difference, instead, can be observed between the energy requirements estimated in this work and the ones reported by Kelly et al.24 The authors did not specifically clarify if they accounted for heat recovery from the outlet stream of the carbonation reactor in the evaluations carried out for industrial residues. However, as shown in Figure 6a, the estimated energy requirement is intermediate between the values calculated in this study assuming to neglect or to consider heat recovery. This seems to suggest that the high energy requirement for steel slags carbonation reported by Kelly et al.,24 which led these authors to conclude that the process is likely unfeasible, is affected by the assumption of partly or totally neglecting heat recovery. If this term were 9320



Article

term “other” includes only the energy required to operate the rotary drum carbonation reactor. The main contributions to the total energy requirements in the wet carbonation process are given by the material size reduction, the heating of the reagents, and the carbon dioxide compression. The minimum energy requirement obtained for the wet carbonation route was observed to range between 550 and 2550 MJ/t CO2 with the lowest values obtained for the APC1 residues and the AOD1 slags, that is, 550 and 744 MJ/t CO2, respectively. These results can be explained considering that these two types of residues presented an initial particle size already suitable for carbonation, allowing to avoid the energy requirements associated with the grinding step. In addition, AOD1 slag and APC1 residues exhibited a higher reactivity with CO2 also at mild operating conditions (i.e., 323 K and 0.1 MPa for AOD1 slag and 303 K and 0.1 MPa for APC1 residues), thus resulting in lower compression and heating requirements. Comparing the results obtained in the wet route for the EAF1, BOF2, and BOF3 slag with those achieved for the same type of residues in the slurry phase process (see Figure 5), it can be observed that the energy requirements for the slurry phase route are generally equal or lower than those estimated for the wet route. This result can be explained considering that a significantly higher reactivity (RX) of these materials was typically observed in the slurry phase route, thus overcoming the duties associated to the more severe operating conditions required. Furthermore, for all the considered wet carbonation conditions (see Figure 8) more than 30% of the total energy requirement was found to be associated to the heating of the residues and water. This result was obtained assuming to neglect the recovery of the heat generated by the exothermic carbonation reaction. This assumption was made since, differently from the slurry phase route where the carbonation reaction takes place in the same medium to be heated, the heat transfer efficiency in the wet route can be hardly predicted and exploited. This choice may penalize the wet route compared to the slurry phase one, since if the heat recovery were taken into account, as for the slurry phase route, the energy required for heating the residues in the wet route could be reduced or even avoided. 3.2.2. Use of Diluted Sources of CO2. To estimate the total energy requirement of the whole CCS chain, the requirements of the wet carbonation route must be coupled with those associated to the capture step needed to get pure carbon dioxide. Assuming a gas fired power plant, the energy requirements associated to capture would equal roughly 4000 MJ/tCO2 captured;33 these should be added to the 550−2550 MJ/tCO2 energy needed for the carbonation of industrial residues, leading to a total of 4550−6550 MJ/tCO2. To reduce the total CCS energy needs, the carbonation step could be directly performed using diluted CO2 sources, thus lumping capture and storage in a single step. As no data of this kind are available to our knowledge for the wet carbonation route, we performed a series of carbonation tests on some industrial residues using different CO2 sources in order to build a database for estimating the associated energy requirements. Figure 9 reports the comparison of the extent of carbonation obtained at 323 K using pure CO2 or a gas mixture with 40% CO2 at similar carbon dioxide partial pressure for AOD1 slag at L/S = 0.4 L/kg (a), APC2 residues at L/S = 0.2 L/kg (b), BOF4 at L/S = 0.3 L/kg (c), and BOF5 slag at L/S = 0.3 L/kg (d). Using the 40% CO2 gas mixture in tests performed with

therefore a key choice and should be clearly stated to avoid misunderstandings in the interpretation of results. 3.2. Mass and Energy Balance: Wet Phase Route. 3.2.1. Effect of the Operating Conditions. Figure 7 reports the energy requirements associated to the wet carbonation route carried out by applying different carbon dioxide pressures and using four types of industrial residues, that is, AOD1 slags (Figure 7a), EAF1 slags (Figure 7b), APC1 residues (Figure 7c) and BOF4 slags (Figure 7d). At increasing CO2 pressure, a higher total energy requirement of the process resulted for the AOD1 slags and APC1 residues. This can be explained considering that the amount of the two residues fed to the process for the three carbonation conditions (see Table 2) is almost the same since the reactivity of these materials proved not to be significantly affected by the increase in the carbon dioxide pressure and only different duties for the CO2 compressor can be observed. The reactivity of the EAF1 and BOF4 slags, instead, seems to be influenced by the applied carbon dioxide pressure, since at increasing CO2 pressure a lower amount of residues is required (i.e., around 7 t EAF1 slag/t CO2 and 8 t BOF4 slag/t CO2 at pCO2 = 1 MPa compared to 11 t EAF1 slag/t CO2 and 10 t BOF4 slag/t CO2 needed at 0.1 MPa). This reduced material requirement entails lower duties associated to the size reduction unit and to the carbonation reactor for higher carbon dioxide pressures, while the energy requirement associated to the compressor is obviously higher. However, for these two types of residues the total energy requirements of the process showed only slight variations with the applied pCO2 (i.e., from 1720 to 2263 MJ/t CO2 for the EAF1 slag and from 2124 to 2300 MJ/t CO2 for BOF4 slags). The differences observed for the analyzed residues can be mainly correlated to their mineralogy, since the dissolution of the silicate phases retrieved in EAF1 and BOF4 slags requires more severe conditions in comparison to the one required for the phases characteristic of APC1 residues and AOD1 slag. Figure 8 shows the lowest energy requirements achievable for each considered residue together with the contribution of the main unit operations. In this case, differently from Figure 5, the

Figure 8. Comparison of the lowest energy requirement estimated for each type of residues applying the wet carbonation route. For the specific operating conditions see Table 2. 9321



Article

Figure 9. Comparison of the extent of carbonation obtained at 323 K using pure CO2 or a gas mixture with 40% CO2 at similar carbon dioxide partial pressure for AOD1 slag at L/S = 0.4 L/kg (a), APC2 residues at L/S = 0.2 L/kg (b), BOF4 at L/S = 0.3 L/kg (c) and BOF5 slag at L/S = 0.3 L/kg (d).

Figure 10. Comparison of the extent of carbonation obtained using pure CO2 or a gas mixture with 9% CO2 at similar carbon dioxide partial pressure for AOD1 slag at at 323 K and L/S = 0.4 L/kg (a), APC1 residues at 303 K and L/S = 0.2 L/kg (b).

APC2 residues, AOD1 slags, and BOF4 slags, a plateau value corresponding to a carbonation extent of around 60−70% was achieved after 24 h. For these residues no significant differences were observed in the carbonation extent resulting for experiments carried out with a 100% CO2 atmosphere, even considering the slight difference in the applied CO2 partial pressure (see Figure 9a−c).

For the BOF5 slags, instead, a slight difference in the extent of carbonation was observed depending on the composition of the gas phase: for example, after 24 h a maximum value of 44% was obtained in the experiments carried out with 40% CO2, while the use of pure carbon dioxide led to a maximum carbonation extent of around 53%. The different behavior observed for BOF5 can be ascribed mainly to the specific 9322



Article

and from 550 to 2600 MJ/t CO2 for the wet one, with an associated demand of residues of roughly 4−7 t/t CO2 and 5− 13 t/t CO2, respectively. Therefore, a clear indication of which could be the most promising route to test at larger scale cannot be directly drawn. However, interesting indications on the applicability of each type of process may be made on the basis of the results of this study. For the slurry phase process, the initial particle size and the mineralogical composition of the residues as well as the applied liquid to solid ratio and operating temperature proved to be the most influential parameters. In particular, the most promising results for slurry phase carbonation were achieved for materials rich in calcium and magnesium containing phases, presenting a limited initial particle size and treated applying a L/S ratio ≤5 L/kg and the minimum temperature required to dissolve the reacting phases. In addition, to reduce the total energy requirement of the process, milling of the residues to a fine particle size (d1 ≤ 50 μm) may prove beneficial since it may significantly improve the reactivity of the material at the timeframes of industrial scale applications (t < 1 h). Furthermore, a crucial aspect to consider for this type of process route was shown to be the possibility of recovering at least part of the heat released from the reaction; hence larger scale tests could be specifically aimed at investigating this issue and quantifying the achievable heat recovery. Concerning the wet route, by far the most significant issue in terms of material and energy requirements appears to be the type of residues treated. Specifically this process appears to be particularly interesting for residues characterized by high contents of calcium hydroxide or reactive Ca−silicate phases (i.e., dicalcium silicate), and fine particle size; the operating parameters in this case showed to play only a secondary role. However, a crucial factor to consider for the applicability of wet carbonation as a CO2 storage technique is the availability of the alkaline feedstock. On the basis of the results of this study in fact the lowest amount of residues to be treated in the process ranges within 5−13 t/t CO2, which is very demanding considering the typical availability of the tested materials. In addition, the possibility of avoiding CO2 capture before carbonation was experimentally tested for the wet route, demonstrating the feasibility of performing carbonation by directly treating diluted CO2 sources characterized by at least 40% CO2 concentrations and 0.3 MPa partial pressure, conditions that are typically met by syngas streams.

mineralogy of this material, characterized by a prevalence of less reactive phases. Hence, it appears that for the tested reaction mode and partial pressure of CO2 the carbonation kinetics of materials such as APC2 residues, AOD1 slags, and BOF4 slags, characterized by a fine particle size and significant content of phases reactive at mild operating conditions, may not be affected by variations in gas composition for CO2 contents ≥ 40%. For residues such as the BOF5 slag, that showed a slower reaction kinetics at the tested operating conditions, the use of CO2 diluted gas mixtures may further hinder the extent of carbonation. In addition, for the AOD1 slags and APC1 residues that showed high extents of carbonation when a gas mixture with 40% CO2 was used instead of pure CO2, some tests were also performed with a gas mixture presenting a lower CO2 content. Figure 10 reports the comparison of the extent of carbonation obtained for the AOD1 slag (a) and APC residues (b) using pure CO2 or a gas mixture with 9% CO2 at similar pCO2 thus simulating a carbonation reaction for treating flue gas. As far as the extent of carbonation of the AOD1 slag is concerned (see Figure 10a), a significant reduction can be observed when the gas mixture is used instead of pure CO2, with a halving of the maximum RX after 24 h (28.3 instead 61.1%). For the APC1 residues, instead, for similar CO2 partial pressures, experiments performed with 9% CO2 flows showed remarkably lower RX values for reaction times up to 4 h for L/S of 0.2 L/kg while for longer experiment durations the extent of carbonation proved to be almost the same. The different behavior of APC1 and AOD1 residues after 24 h reaction time can be attributed to the different mineralogy of the two residues. The dissolution of the silicate phases found in the AOD residues requires lower pH, and thus higher CO2 partial pressure, than the one required to dissolve the calcium hydroxide phases characteristic of the APC residues. The results of these tests indicate that for the experimental setup (static batch reactor with down-flow gas supply) and at the operating conditions employed in this work, the Rx values achieved at 1 h reaction time under 9% CO2 concentration, proved to be too low to be considered feasible for the scale-up of the process employing either type of residues. Given these results, it can be concluded that the wet carbonation route could be effectively performed on the selected residues directly treating a diluted source characterized by a CO2 concentration as low as 40%. On the basis of the carbonation extent data measured at 1 h reaction time, reported in Figure 9, the associated energy requirements proved to be equal to 784−4450 MJ/t CO2 with a clear reduction compared to those estimated for the respective residues assuming to capture and store CO2 in two separate steps.

■

AUTHOR INFORMATION

Corresponding Author

*Tel: +39-0672597022. Fax: +39-0672597021. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

4. CONCLUSIONS In this paper an assessment of accelerated carbonation applied to industrial residues for CO2 storage was performed with the aim of gaining insight on the feasibility of this process at larger scale, identifying also the key factors for its implementation. On the basis of the analysis of the lab-scale tests performed in the last years, it appears that significant CO2 uptakes may be achieved by applying either the slurry phase route or the wet one. Depending on the type of residues used, the lowest total energy requirements for CO2 storage only were estimated to range within 1300−2750 MJ/t CO2 for slurry phase processes

ACKNOWLEDGMENTS The authors would like to acknowledge the students Martina Di Gianfilippo, Alessio Gattaino, Erica Martellato, and Stefano Pantano for performing the new experiments reported in this work.

■

REFERENCES

(1) IPCC, Intergovernmental Panel on Climate Change. Climate Change 2013, the Physical Science Basis; Working Group I Contribution to the IPCC 5th Assessment Report, 2013.

9323



Article

(2) Huijgen, W. J. J.; Witkamp, G. J.; Comans, R. N. J. Mineral CO2 sequestration by steel slag carbonation. Environ. Sci. Technol. 2005, 39, 9676−2682. (3) Lackner, K. S.; Wendt, C. H.; Butt, D. P.; Joyce, E. L.; Sharp, D. H. Carbon dioxide disposal in carbonate minerals. Energy 1995, 20, 1153−1170. (4) Gerdemann, S. J.; O’Connor, W. K.; Dahlin, D. C.; Penner, L. R.; Rush, H. Ex situ aqueous mineral carbonation. Environ. Sci. Technol. 2007, 41, 2587−2593. (5) Sipilä, J.; Teir, S.; Zevenhoven, R. Carbon Dioxide Sequestration by Mineral Carbonation: Literature Review Update 2005−2007; Report VT 2008−1; Åbo Akademi University, Faculty of Technology Heat Engineering Laboratory: Finland, 2008. (6) Costa, G.; Baciocchi, R.; Polettini, A.; Pomi, R.; Hills, C. D.; Carey, P. J. Current status and perspectives of accelerated carbonation processes on municipal waste combustion residues. Environ. Monit. Assess. 2007, 135, 55−75. (7) O’Connor, W. K.; Dahlin, D. C.; Rush, G. E.; Gerdemann, S. J.; Penner, L. R.; Nilsen, D. N. Aqueous Mineral Carbonation: Mineral Availability, Pretreatment, Reaction Parametrics, and Process Studies; Report DOE/ARC-TR-04−002; Albany Research Center: Oregon, 2005. (8) Van Gerven, T.; Van Keer, E.; Arickx, S.; Jaspers, M.; Wauters, G.; Vandecasteele, C. Carbonation of MSWI-bottom ash to decrease heavy metal leaching in view of recycling. Waste Manage. 2005, 25, 291−300. (9) Baciocchi, R.; Costa, G.; Di Bartolomeo, E.; Polettini, A.; Pomi, R. The effects of accelerated carbonation on CO2 uptake and metal release from incineration APC residues. Waste Manage. 2009, 29, 2994−3003. (10) Baciocchi, R.; Costa, G.; Di Bartolomeo, E.; Polettini, A.; Pomi, R. Carbonation of Stainless Steel Slag as a Process for CO2 Storage and Slag Valorization. Waste Biomass Valor. 2010, 1, 467−477. (11) Bobicki, E. R.; Liu, Q.; Xu, Z.; Zeng, H. Carbon capture and storage using alkaline industrial wastes. Prog. Energy Combust. Sci. 2012, 38, 302−320. (12) Gunning, P. J.; Hills, C. D.; Carey, P. J. Accelerated carbonation treatment of industrial residues. Waste Manage. 2010, 30, 1081−1090. (13) Pan, S. Y.; Chang, E. E.; Chiang, P. C. CO2 capture by accelerated carbonation of alkaline wastes: A review on its principles and applications. Aerosol Air Qual. Res. 2012, 12, 770−791. (14) Baciocchi, R.; Costa, G.; Zingaretti, D. Accelerated Carbonation Processes for Carbon Dioxide Capture, Storage, and Utilization. In Transformation and Utilization of Carbon Dioxide, Green Chemistry and Sustainable Technology; Bhanage, B. M., Arai, M., Eds.; Springer-Verlag: Berlin Heidelberg, in press. (15) Uibu, M.; Uus, M.; Kuusik, R. CO2 mineral sequestration in oilshale wastes from Estonian power production. J. Environ. Manage. 2009, 90, 1253−1260. (16) Huntzinger, D. N.; Gierke, J. S.; Kawatra, S. K.; Eisele, T. C.; Sutter, L. L. Carbon dioxide sequestration in cement kiln dust through mineral carbonation. Environ. Sci. Technol. 2009, 43, 1986−1992. (17) Fernández-Bertos, M.; Simons, S. J. R.; Hills, C. D.; Carey, P. J. A review of accelerated carbonation technology in the treatment of cement-based materials and sequestration of CO2. J. Hazard. Mater. 2004, B112, 193−205. (18) Iizuka, A.; Fujii, M.; Yamasaki, A.; Yanagisawa, Y. Development of a new CO2 sequestration process utilizing the carbonation of waste cement. Ind. Eng. Chem. Res. 2004, 43, 7880−7887. (19) Back, M.; Kuehn, M.; Stanjek, H.; Peiffer, S. Reactivity of alkaline lignite fly ashes towards CO2 in water. Environ. Sci. Technol. 2008, 42, 4520−4526. (20) Baciocchi, R.; Costa, G.; Di Gianfilippo, M.; Polettini, A.; Pomi, R.; Stramazzo, A. Evaluation of the CO2 storage capacity of different types of steel slag subjected to accelerated carbonation. In Proceedings of Crete 2012, 3rd International Conference on Industrial and Hazardous Waste Management, 12−14 September 2012, Chania, Crete, Greece.

(21) Baciocchi, R.; Costa, G.; Di Bartolomeo, E.; Polettini, A.; Pomi, R. Wet versus slurry carbonation of EAF steel slag. Greenhouse Gases: Sci. Technol. 2011, 1, 312−319. (22) Johnson, D. C.; Macleod, C. L.; Carey, P. J.; Hills, C. D. Solidification of stainless steel slag by accelerated carbonation. Environ Technol. 2003, 24, 671−678. (23) Huijgen, W. J. J.; Ruijg, G. J.; Comans, R. N. J.; Witkamp, G. J. Energy consumption and Net CO2 sequestration of aqueous mineral carbonation. Ind. Eng. Chem. Res. 2006, 45, 9184−9194. (24) Kelly, K. E.; Silcox, G. D.; Sarofim, A. F.; Pershing, D. W. An evaluation of ex situ, industrial-scale, aqueous CO2 mineralization. Int. J. Greenhouse Gas Control 2011, 5, 1587−1595. (25) Kirchofer, A.; Brandt, A.; Krevor, S.; Prigiobbe, V.; Wilcox, J. Impact of alkalinity sources on the life-cycle energy efficiency of mineral carbonation technologies. Energy Environ. Sci. 2012, 5, 8631− 8641. (26) Perry, R. H.; Green, D. W. Perry’s Chemical Engineers’ Handbook; McGraw-Hill: New York, 1987. (27) Zingaretti, D.; Costa, G.; Baciocchi, R. Assessment of the energy requirements for CO2 storage by carbonation of industrial residues. Part 1: Definition of the process layout. Energy Procedia 2013, 37, 5850−5857. (28) Hangx, S. J. T.; Spiers, C. J. Coastal spreading of olivine to control atmospheric CO2 concentrations: A critical analysis of viability. Int. J. Greenhouse Gas Control. 2009, 3, 757−767. (29) Robie, R. A.; Hemingway, B. S.; Fisher, J. R. Thermodynamic Properties of Minerals and Related Substances at 298.15 K and 1 bar (105 Pascal) Pressure and at Higher Temperatures; United States Government Printing Office: WA, 1978. (30) Allal, K. M.; Dolignier, J. C.; Martin, G. Determination of thermodynamical data of calcium hydroxichloride. Inst. France Pet. J. 1997, 52, 361−368. (31) Schmetterer, C.; Masset, P. J. Heat capacity of compounds in the CaO−SiO2 SystemA review. J. Phase Equilib. Diffus. 2012, 33, 261−275. (32) McCollum, D. L.; Ogden, J. M. Techno-Economic Models for Carbon Dioxide Compression, Transport, and Storage & Correlations for Estimating Carbon Dioxide Density and Viscosity; Institute of Transportation Studies (UCD); UC Davis: CA, 2006. (33) Working Group III of the Intergovernmental Panel on Climate Change. IPCC Special Report on Carbon Dioxide Capture and Storage; Metz, B., Davidson, O., de Coninck, H. C., Loos, M., Meyer, L. A., Eds.; Cambridge University Press: Cambridge, United Kingdom and New York, NY, USA, 2005.

9324


Assessment of Accelerated Carbonation Processes for CO2 Storage

Recommend Documents