Mineral carbonation with thermally activated serpentine; the implication of serpentine preheating temperature and heat integration

https://doi.org/10.1016/j.cherd.2021.06.002Get rights and content

Highlights

  • The process requires 5.85 ton of rocks to bind 1 tonne of CO2.

  • Heat demand = 5.0 GJ/tCO2 captured or 9.5 GJ/tCO2 avoided at 670 °C.

  • The process thermal energy efficiency improved by 25% compared to our previous work.

  • The effect of process parameters on the heat integration strategy was investigated.

Abstract

In this paper, heat integration was coupled with serpentine preheating temperature optimization and was for the first time applied to a mineral carbonation process. For this, a process was selected with the aim to minimise its heat demand. Aspen® Energy Analyzer software was used for heat integration with pinch analysis. The mineral carbonation plant considered here processes 100 tonnes of serpentine per hour, with Mg content of 237 kg per tonne of rock, corresponding to 0.88 million tonnes per year to treat 0.5 million tonnes of CO2 emitted from a cement plant. In the base case considered for heat integration, 50% of the serpentine magnesium content was converted to hydromagnesite using 30% of the flue gas CO2 which means that 5.85 tonnes of rocks were required to capture 1 tonne of CO2.

The serpentine preheating temperature, was optimized as 400 °C. Application of heat integration at this solid preheat temperature reduced the process heat demand by 25% compared to our previous study, as 5.0 GJ per tonne of CO2 captured or 9.5 GJ per tonne of CO2 avoided. This corresponds to 0.86 GJ per tonne of ore as the process heat demand was only attributed to the mineral activation. In addition, the impact of process parameters including the solid-liquid ratio and the dissolution reaction extent on the heat integration strategy was evaluated. It was found that solids concentrations as low as 5% substantially reduced the sequestration efficiency as the process waste heat was not sufficient at the optimized solid preheat temperature. Furthermore, the uncertainty behind unknown parameters such as activated serpentine heat capacity and solid heat transfer coefficient was evaluated to validate the study.

Introduction

The major challenge of carbon mineralization with serpentine is to improve the reaction kinetics in order to ensure efficient conversion in a reasonable timeframe (Wang et al., 2018). While gas phase reaction is the most straightforward way to operate the carbonation reaction (Eq. 1), kinetics remain very slow and the conversion limited even at high operating conditions. For example, only 30% of the serpentine Mg content could be carbonated under 300 °C and 340 bars of pCO2 after several days (Lackner et al., 1997). Hence, the interest was shifted towards aqueous phase reactions (Benhelal et al., 2019; Hariharan et al., 2017; Pasquier et al., 2014b; Zevenhoven et al., 2016). Still, the reaction remained slow unless the solution chemistry was modified or the serpentine crystalline form altered or destroyed (O’connor et al., 2005).Mg3Si2O5(OH)4 + CO2 => 3 MgCO3 2 H2O + 2 SiO2 ΔH0r (25°C) = −64 kJ.mol−1

In this regard, several solutions have been proposed. Some authors suggested to convert serpentine to MgSO4 (Erlund et al., 2016; Koivisto et al., 2016; Sanna et al., 2013) or Mg(OH)2 (Fagerlund et al., 2012; Lackner et al., 1997) prior to carbonation, as these forms were more reactive. These led to the development of what became referred to as the ‘two-stage processes’ (Sanna et al., 2014). Traditionally, the MgSO4 route was led by the Nottingham research team (Sanna et al., 2013), while the Mg(OH)2 was especially taken by the Åbo Akademi research team (Fagerlund et al., 2012). Note that a modification of the Åbo Academi route, referred to as the “alternative” Åbo Akademi route was more recently introduced. This can be distinguished from the former considering that the serpentine is converted to MgSO4 instead of Mg(OH)2, while the process still being different from the Nottingham route (Zevenhoven et al., 2016). The description of these processes will not be made here as it has been extensively discussed in literature (Sanna et al., 2014). With the two-stage processes, reaction times shifted down to 1−2 h and conversions rates varying from 50 to 90% were reported (Sanna et al., 2013; Zevenhoven et al., 2016). Inevitably, these multistage processes did require salts usage and eventually regeneration. Often, ammonium sulfates or bisulfates were used. This is because during the conversion stage, ammonia vapors are liberated, which could be recovered and used to adjust the pH prior to carbonates precipitation phase (Fagerlund et al., 2012; Sanna et al., 2013). These could be even used to capture the CO2 from flue gas in a separate absorption column (Sanna et al., 2013). Many other organic acids have also been widely explored (Bonfils et al., 2012; Park and Fan, 2004). Nevertheless, salts and acids were reported to present some environmental burdens and might increase the process costs (Giannoulakis et al., 2014). This is mainly attributed potential leaks and make up and to energy intensive regeneration (Giannoulakis et al., 2014). Besides and in the case where carbonates would require commercialization to offset process costs, salts might degrade the product quality as they reduce its purity.

Other researchers have therefore decided to operate the process only using the weak acidity generated by the CO2 dissolution in solution (Pasquier et al., 2014b; Werner et al., 2014). To accelerate the reaction, the serpentine would require an energy intensive activation operation held at high temperature (Dlugogorski and Balucan, 2014; McKelvy et al., 2004). This dehydroxylates partially the serpentine, which modifies its structure and leads to an amorphous material that is more reactive (Dlugogorski and Balucan, 2014). This strategy was originally inspired by the work performed ten years earlier by the National Energy Research Laboratory (NETL). The NETL research group were the precursor for activated serpentine mineral carbonation but decided to end the project despite very good conversion rates (up to 92% of conversion rate within one hour of reaction time) (O’connor et al., 2005). This is because they estimated that the energy and cost of the mineral processing operations were too much of a burden for an application aiming to mitigate the CO2 (O’connor et al., 2005). Other scientists did not share this perspective (Gadikota, 2016), and the work with activated serpentine was continued by other research teams. This was the case for two research teams from Canada and Switzerland (Pasquier et al., 2014a; Werner et al., 2014). Their major modification to the NETL process is that they choose to operate the carbonation in two stages, allowing the activated serpentine dissolution and carbonates precipitation to be optimized separately. This allowed these two reactions to be optimised independently, under milder operating conditions and without the necessity of slats. The activated serpentine dissolution using the CO2 weak acidity was operated under ambient temperature while higher temperature 40−90 °C was required to precipitate the carbonates. The Canadian team focused on operating the process with diluted flue gas (18%) and on generating high purity carbonate as a marketable product in order to offset the process costs (Pasquier et al., 2014a, 2016). More recently, they validated the process on a pilot scale (Kemache et al., 2016; Mouedhen et al., 2017), and investigated the relationship between heat activation conditions such as grain size, activation temperature and time on the material reactivity (Du Breuil et al., 2019a, b). They have also shown that the dissolution of activated serpentine can be successfully operated in a bubble culumn reactor under ambiant conditions with pCO2 = 0.18 bars (Tebbiche et al., 2020). The research team from Switzerland focused on process condition optimization and did significant work on process modeling, from serpentine dehydroxylation to carbonates precipitation (Hariharan and Mazzotti, 2017; Hariharan et al., 2014, 2016). Another research team from Australia studied the process scale up on a pilot scale and highlighted the positive effect of improved solution mixing on the conversion rate (Benhelal et al., 2018). They also made significant work on understanding reaction mechanism (Farhang et al., 2016, 2017) and on material deactivation that they suggested was partially attributed to the serpentine recrystallization at relatively high operating temperature and pressure (100−150 °C and 20–120 bars of CO2) (Benhelal et al., 2019). Clearly, this can be a very bad deal breaker as serpentine reactivation would result in a double heat penalty. Despite that the reserpentinization is typically reported to occur at extreme reaction conditions, say >300 °C and >1000 bars, with geological time rates of more than 1000 years (Macdonald and Fyfe, 1985). Benhelal et al. (2019) justified their findings based on the much higher reactivity of activated serpentine as compared to natural serpentine for which serpentinization rates are reported by (Macdonald and Fyfe, 1985). Besides, the re-serpentinization phenomena was only recorded for antigorite, while not observed for lizardite. Therefore, the effect of reserpetinization of activated serpentine might be minimized if the reaction is operated at ambient temperature and pressures not exceeding 10 bars (Pasquier et al., 2014b). Given that no additional chemical was used in processes, the conversion rates reported by these research teams generally ranging from 30 to 50% within 2−3 h of reaction time are considered satisfactory.

Still, the serpentine heat activation remained a burden for the process, and its energy demand had to be minimized in order for the process to remain suitable. In this perspective, it was reported that the heat demand of dehydroxylation operation could be brought down to a minimum of 0.57 GJ/trocks if 80% of the sensitive heat of dehydroxylated serpentine leaving the kiln at 680−730 °C could be recovered and used to preheat the inlet serpentine up to 500−550 °C (Balucan et al., 2013). This is about 50% less compared to initial estimate by the NETL research group (O’connor et al., 2005). Still, the estimate made here relays on the use of solid-solid heat exchangers, which commercial application is very limited (Vorrias et al., 2013). In fact, flue gases resulting from fuel burning in the heat activation kiln and steam generated from the dehydroxylation reaction mix upon exit and might be a better candidate for serpentine preheating, given the fact that solid-fluid heat exchangers are more mature technology (Benson et al., 2014; Sutherland, 2015). Besides, serpentine preheating temperature might be subjected to economic consideration, given the investment incurred by the solid-fluid heat exchangers. Yet and to the best of our knowledge, there is no study that focused on this matter. Besides, mineral carbonation processes typically require several cooling and heating operations where additional process waste heat would be generated. In this regard, heat integration is another technique that was proven effective to reduce mineral carbonation heat demand by optimising the use of process waste heat (Romão et al., 2012). Therefore, it appear more promising to consider serpentine preheating temperature optimization together with process heat integration in order to minimize both the thermal activation and the overall process heat demand. Yet and to the best of our knowledge, this has not been considered previously.

This study is first to couple serpentine preheating temperature with process heat integration for mineral carbonation process operated with thermally activated serpentine. The process operated by the Canadian team was selected for the study. This is because this process does not require salts, which as stated above are better avoided to minimize environmental impacts and produce high quality carbonates. Obviously, higher purity carbonate can be obtained when salts are not used in the process, which improves the chances of utilization and commercialization (Erlund and Zevenhoven, 2018; Pasquier et al., 2016). Also, this process operates with activated serpentine, which as stated earlier, has been subject subject to disagreement upon its thermal energy requirement that varies from 0.57 to 1.2 GJ/trocks du to the unavailability of detailed study (Balucan et al., 2013; O’connor et al., 2005). For the selected process in this study, the heat demand was previously estimated to 6.7 GJ/tCO2 captured while power requirement was estimated to be around 1.1 GJ/tCO2 captured (Pasquier et al., 2016). Regarding heat integration, pinch analysis was chosen as a proven tool that can lead significant improvement of energy usage (Ebrahim and Kawari, 2000). To validate the heat integration strategy, a sensitivity analysis regarding some uncertain parameters such as activated serpentine heat capacity and solid-fluid heat transfer coefficient was performed. Besides, the effect of some process parameters such as solid-liquid ratio and activated serpentine dissolution extent on the heat integration strategy was investigated.

Section snippets

Process description

A simplified process flow sheet diagram for the process selected for this study is illustrated in Fig. 1. At this point, heating and cooling utilities are not shown for simplicity. The process can be divided into three separate steps:

Aspen plus model description

Aspen Plus V9.0® was used for process modeling and simulation. The model here is based on stoichiometric reactors where the extent of reactions are defined according to laboratory results while reaction enthalpies are taken from published literature (see Eqs. 3−8 in Table 1). The SOLIDS property package with Soave-Redlich-Kwong equation of state for gas phase calculations is used (Aspen Technology, 2010). The SOLIDS property method does not consider ion speciation and consequently, the

ΔTmin optimization

First of all, simulation results extracted from Aspen Plus® were initially validated against hand calculations. The initial comparison highlighted some errors in vapor-liquid equilibria, arising from the use of the SOLIDS property method, which although recommended for solids processing modeling, did not encompass vapor-liquid thermodynamics (Aspen Technology, 2010). Adjustments to the properties parameters rectified these issues and validated the proposed model for the scope of this study. To

Conclusion

In this study, serpentine preheating temperature optimization coupled with process heat integration was applied for the first time on a mineral carbonation process. The mineral carbonation plant is designed to capture 30% of the 0.5 M tonnes of CO2 emitted per year by a cement plant. For the base case considered, 50% of the Mg content was converted to hydromagnesite considering average results obtained by previous laboratory essays. This corresponds to 5.85 tonnes of rocks required to capture 1

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgment

This work was financially supported by Les Fonds de recherche du Québec - Nature et technologies en équipe de190542. References and screen images from Aspen Plus® are reprinted with permission from Aspen Technology, Inc. AspenTech®, Aspen Plus® Aspen ONE®, and the AspenTech leaf logo are trademarks of Aspen Technology, Inc. All rights reserved.

References (82)

  • S. Giannoulakis et al.

    Life cycle and cost assessment of mineral carbonation for carbon capture and storage in European power generation

    Int. J. Greenh. Gas Control.

    (2014)
  • M. Hänchen et al.

    Precipitation in the Mg-carbonate system—effects of temperature and CO2 pressure

    Chem. Eng. Sci.

    (2008)
  • S. Hariharan et al.

    Dissolution of dehydroxylated lizardite at flue gas conditions: II. Kinetic modeling

    Chem. Eng. J.

    (2014)
  • S. Hariharan et al.

    Dissolution of dehydroxylated lizardite at flue gas conditions: III. Near-equilibrium kinetics

    Chem. Eng. J.

    (2016)
  • S. Hariharan et al.

    A two-step CO2 mineralization process

    Energy Procedia

    (2017)
  • P.J. Heggs

    Minimum temperature difference approach concept in heat exchanger networks

    Heat Recovery Syst. Chp

    (1989)
  • N. Kemache et al.

    Aqueous mineral carbonation of serpentinite on a pilot scale: The effect of liquid recirculation on CO2 sequestration and carbonate precipitation

    Appl. Geochem.

    (2016)
  • E. Koivisto et al.

    Extraction of magnesium from four Finnish magnesium silicate rocks for CO2 mineralisation - Part 1: Thermal solid/solid extraction

    Hydrometallurgy

    (2016)
  • K.S. Lackner et al.

    Progress on binding CO2 in mineral substrates

    Energy Convers. Manage.

    (1997)
  • A.H. Macdonald et al.

    Rate of serpentinization in seafloor environments

    Tectonophysics

    (1985)
  • N.A. Madlool et al.

    A critical review on energy use and savings in the cement industries

    Renewable Sustainable Energy Rev.

    (2011)
  • I. Mouedhen et al.

    Effect of pCO2 on direct flue gas mineral carbonation at pilot scale

    J. Environ. Manage.

    (2017)
  • A.-H.A. Park et al.

    CO2 mineral sequestration: physically activated dissolution of serpentine and pH swing process

    Chem. Eng. Sci.

    (2004)
  • L.-C. Pasquier et al.

    Parameters optimization for direct flue gas CO2 capture and sequestration by aqueous mineral carbonation using activated serpentinite based mining residue

    Appl. Geochem.

    (2014)
  • L.-C. Pasquier et al.

    Technical & economic evaluation of a mineral carbonation process using southern Québec mining wastes for CO2 sequestration of raw flue gas with by-product recovery

    Int. J. Greenh. Gas Control.

    (2016)
  • A. Pearson

    Refrigeration with ammonia

    Int. J. Refrig.

    (2008)
  • I. Romão et al.

    CO2 fixation using magnesium silicate minerals. Part 2: energy efficiency and integration with iron-and steelmaking

    Energy

    (2012)
  • E.S. Rubin et al.

    A proposed methodology for CO2 capture and storage cost estimates

    Int. J. Greenh. Gas Control.

    (2013)
  • A. Sanna et al.

    Carbon dioxide capture and storage by pH swing aqueous mineralisation using a mixture of ammonium salts and antigorite source

    Fuel

    (2013)
  • I. Tebbiche et al.

    Thermally activated serpentine leaching under flue gas conditions in a bubble column reactor operated at ambient pressure and temperature

    Hydrometallurgy

    (2020)
  • S. Valiani et al.

    Optimization of pre-combustion capture for thermal power plants using Pinch Analysis

    Energy

    (2017)
  • P.S. Varbanov et al.

    Total Site targeting with process specific minimum temperature difference (ΔTmin)

    Energy

    (2012)
  • I. Vorrias et al.

    Calcium looping for CO2 capture from a lignite fired power plant

    Fuel

    (2013)
  • S.-G. Yoon et al.

    Heat integration analysis for an industrial ethylbenzene plant using pinch analysis

    Appl. Therm. Eng.

    (2007)
  • R. Zevenhoven et al.

    A comparison of CO2 mineral sequestration processes involving a dry or wet carbonation step

    Energy

    (2016)
  • A. Amiri et al.

    Waste heat recovery power generation systems for cement production process

    IEEE Trans. Ind. Appl.

    (2015)
  • I. Aspen Technology

    Property method descriptions

    Aspen Physical Property System

    (2010)
  • I. Aspen Technology

    Heat integration-pinch analysis

    Aspen Energy Analyzer Reference Guide

    (2015)
  • I. Aspen Technology

    Heat integration project- design level

    Aspen Energy Analyzer Reference Guide

    (2015)
  • I. Aspen Technology

    Size & cost assumptions

    Aspen Energy Analyzer Reference Guide

    (2015)
  • S. Benson et al.

    Evaluation of Carbon Dioxide Capture From Existing Coal Fired Plants by Hybrid Sorption Using Solid Sorbents

    (2014)
  • Cited by (4)

    View full text