Elsevier

Fluid Phase Equilibria

Volume 363, 15 February 2014, Pages 149-155
Fluid Phase Equilibria

Effect of impurities in captured CO2 on liquid–vapor equilibrium

https://doi.org/10.1016/j.fluid.2013.11.009Get rights and content

Highlights

  • The impact of impurities on the phase diagram of CO2 was studied and the capabilities of three mixture models were assessed.

  • An expansion in the phase transition envelope was observed in all mixtures compared to pure CO2.

  • The phase transition zone for CO2-impurities mixtures is located at higher pressures in comparison with pure CO2.

  • Looking to the models predictions, simulation deviations were observed close to the critical point.

  • Transporting liquid CO2 in pipelines with the presence of impurities requires higher pipeline pressure.

Abstract

The capture of large amounts of CO2 from power plants and other large CO2 point sources has become relevant within the concept to mitigate CO2 emissions via carbon capture and storage. The objective of this study was to investigate the impact of some major impurities found in the CO2 stream captured from power plants and other large point sources on the liquid–vapor phase change compared to pure CO2. The study contributes to a better understanding of how and within which boundary conditions transmission of impure CO2 from an emission and capture point to a storage location can be realized.

A set up was constructed to provide the experimental data needed to trace the phase diagrams of the CO2-impurities mixtures. The data show an expansion of the liquid–vapor phase envelope and a start of boiling (in liquid) or condensation (in vapor) at relatively higher pressures. The experimental data were later used to assess the capabilities of three mixture models: the Soave–Redlich–Kwong cubic equation of state, the GERG-2008 model, and the EOS-CG model.

Introduction

The capture of CO2 from large fossil fuel and biomass energy facilities with major CO2 emissions, results in certain types and quantities of impurities. The impurities in the captured stream may include nitrogen, argon, oxygen, carbon monoxide, water and some toxic elements like sulfur and nitrogen oxides. The types and quantities of impurities depend on the type of fuel, the type of the capture process and any gas treatment steps either prior or subsequent to the CO2 capture [1].

The presence of impurities in the CO2 captured stream creates many technical challenges. Some impurities like nitrogen and methane lower the critical temperature which would mean that higher strength pipelines should be used to reduce the ductile fracture potential [2]. Other impurities like hydrogen sulfide, sulfur dioxide and carbon monoxide are toxic and should be kept within ppm levels [3]. The presence of water may result in ice or gas-hydrates formation and this causes corrosion if it exists as a free phase [4]. The presence of non condensable components like oxygen, argon and nitrogen should be monitored and limited in order to save compression energy. Having any type of impurities has also an impact on the liquid–vapor phase boundary and may cause expansion of the two phase regime which is unwanted by CO2 pipeline operators who seek to avoid any change in phase in their pipeline grid [4], [5]. The decrease in the impurities levels of the captured CO2 mixture is a complicated process not only due to the technical difficulties but also for cost implications.

There are enough physical data available in the literature on pure CO2 including a clear and precise description of the phase envelope; however, there are very limited data on mixtures of CO2 and impurities. Part of this literature data is given in this paper. Knowing the phase envelope of CO2-impurity mixtures is of a high importance especially for the designers and the operators of the CO2 pipeline transmission systems where phase change has to be avoided.

In the presented study, the impact of the major impurities present in the CO2 stream on the liquid–vapor phase envelope of pure CO2 was studied. An experimental set up having a transparent view cell was used where well-defined mixtures were prepared under controlled pressure and temperature conditions. The body of the cell is completely transparent providing perfect visibility of the phase transition when it takes place.

The experimental data generated were compared to three different mixture models used for the prediction of phase boundaries of CO2 mixtures. Beside the well-established Soave–Redlich–Kwong cubic equation of state [6], two modern mixture models explicit in the Helmholtz free energy were used for comparison. The first is the GERG-2008 model [7] that was developed for the description of natural gas mixtures and is capable to describe the CO2 mixtures studied in this work as well. The second is the EOS-CG model [8] that is based on the same mathematical structure as the GERG-2008, but was developed with special focus on CO2 mixtures and humid gases.

The presented experimental work is a part of the Dutch National Program for CCS Research CATO2.

Section snippets

Experimental test set up

The experimental investigation was carried out using a high pressure view cell system in which CO2 mixtures were prepared, and where pressure and temperature were controlled and varied. The view cell is made up of a cylindrical sapphire tube (Fig. 1) that is closed by two lids and covered by a double jacket made up of polycarbonate and thus provided a complete view of the inner view cell volume. The temperature inside the view cell was controlled using water filled refrigerated and heating

Measurement procedure

To detect the liquid vapor transition zone for a certain CO2-impurity mixture, two different sets of experiments were carried out:

In the first set of experiments, the dew-point line of the mixture was detected. In this experiment, a mono-phase vapor mixture of a certain sought composition was prepared inside the cell. The initial pressure was chosen to be 50 bar with a temperature of 20 °C for all trials as these conditions would provide initial mono-phase vapor conditions inside the view cell.

Models

The models used for comparison with the experimental data are the Soave–Redlich–Kwong (SRK) [6], the GERG-2008 [7], and the EOS-CG [8] equations of state. The SRK equation of state is a well-established model and will not be discussed here. It was used with the generalized mixing rules given in the original publication with the adjusted binary mixture parameters of Li and Yan [9].

The GERG-2008 and EOS-CG models are based on the same multi-fluid approach but use different adjusted mixture

Experimental data

The obtained experimental data were used to trace the liquid–vapor phase diagrams of the different studied CO2 mixtures (Fig. 3) with the numerical values given in the paper (Table A1). The uncertainty in the temperature measurement inside the cell is 0.5 °C and that of the pressure is 0.3%. The uncertainty in the mixtures composition measurements using the GC (Gas Chromatography) is 0.01%. The uncertainty in the detection of the dew-point or the boiling temperature is 1 °C.

A polynomial trend

Conclusions

The influence of some major impurities on the liquid–vapor phase change of pure CO2 has been studied. The experimental data and the different models predictions have been analyzed and the following conclusions can be made:

  • An expansion in the phase transition envelope has been observed in all studied mixtures compared to pure CO2. The width of the phase transition zone is wider with high concentrations of impurities.

  • The boiling of liquid CO2-impurities mixtures and the condensation of vapor CO2

Acknowledgement

This project is funded by N.V. Nederlandse Gasunie in the framework of the Dutch National CCS Research Program (CATO2).

References (13)

There are more references available in the full text version of this article.

Cited by (26)

  • The impressive impact of including enthalpy and heat capacity of mixing data when parameterising equations of state. Application to the development of the E-PPR78 (Enhanced-Predictive-Peng-Robinson-78) model.

    2022, Fluid Phase Equilibria
    Citation Excerpt :

    Table S1 in the Supporting Information presents the list of the 204 pure components involved in the development of the E-PPR78 model. For each binary system considered in the development of the E-PPR78 model, Table S2 contains the main features of collected experimental data, extracted from references [38–2068]. Bibliographic references are reported in Table S3.

  • Vapor-liquid equilibria, solid-vapor-liquid equilibria and H<inf>2</inf>S partition coefficient in (CO<inf>2</inf> + CH<inf>4</inf>) at temperatures between (203.96 and 303.15) K at pressures up to 9 MPa

    2020, Fluid Phase Equilibria
    Citation Excerpt :

    These outliers are shown individually and often occur in the critical region where, as noted above, the PR-EoS model does not exhibit the correct scaling. Based on Fig. 11, the most consistent data sets appear to be: Xu et al. [16], Bian and Bian et al. [17,18], Webster and Kidnay [20], Ahmad et al. [22], Petropoulou et al. [24] and the present work. Each of these data sets shows small deviation in both liquid and vapor phases with few outliers.

View all citing articles on Scopus
View full text