(711f) Kinetic Study of Carbon Dioxide Absorption By Aqueous Solutions of N-Methyldiethanolamine and Piperazine

Kinetic study of carbon dioxide
absorption by aqueous solutions of N-methyldiethanolamine and piperazine

Alberto
SERVIA^a, Nicolas LALOUE^a, Julien GRANDJEAN^a,
Sabine RODE^b, Christine ROIZARD^b

^a IFP Energies nouvelles, Rond-point de
l'échangeur de Solaize BP3, 69360, Solaize, France

^b LRGP-CNRS Université de
Lorraine, 1 rue Grandville - BP 20451 - 54001 Nancy,
France

1. Introduction

The kinetics of CO₂ absorption
with aqueous blends of N-MethylDiEthanolAmine (MDEA) and PiperaZine (PZ) has
been widely studied (Zhang et al., 2001, Bishnoi and Rochelle, 2002, Edali et
al., 2010, Samanta and Bandyopadhyay, 2011). Nevertheless, some discrepancies
remain concerning the reaction mechanisms proposed in the literature. For
instance, the synergy between both amines is still not completely understood
and proposed kinetics models cannot fully explain the experimental data at very
high loadings. Moreover, the models used in the literature to deduce the
kinetics of CO₂ absorption by aqueous blends of PZ and MDEA from
experiments present some limitations. They all assume a constant CO₂
partial pressure in the gas phase. This hypothesis can lead to noticeable deviations
if high CO₂ quantities are absorbed.

The objective of this work is to
investigate the kinetics of CO₂ absorption by aqueous blends of MDEA
and PZ. Experiments are performed in a wetted wall column (WWC) under various
operating conditions and simulated using a rigorous reactor model.

2. Experimental setup

A WWC was used to obtain experimental data
on the kinetics of CO₂ absorption by aqueous blends of MDEA and PZ. Within
the reactor, the gas phase flows counter-currently with the liquid that
overflows from the inside of a cylinder to form a thin liquid film. The gas
phase, composed of CO₂ and Nitrogen (N₂), is water-saturated
before being in contact with the liquid in the reactor to prevent from water
mass transfer in the reaction zone. The experimental flux is calculated using
the variation of the CO₂ gas concentration between the inlet and the
outlet of the reactor, measured by an in-line infra-red spectrometer.

The WWC validation and hydrodynamic
characterization were presented in a previous work (Servia et al., 2013).

3. Model description

A 2D stationary model has been developed
using COMSOL software to predict the absorption flux of CO₂ in
aqueous blends of PZ and MDEA in the WWC. This model couples hydrodynamics, gas-liquid equilibrium,
mass transfer and chemical reactions (Servia et al., 2013).

The liquid phase velocity profile is
determined by the Navier-Stokes equation for an incompressible fluid associated
with specific boundary conditions. Since CO₂ was transferred from
the gas into the liquid phase, the gas velocity varied within the reactor. This
evolution was modeled through a mass balance on the inert compounds (N₂
and water).

The concentration of each species at
equilibrium conditions were provided by a thermodynamic model accounting for
the deviations from the ideal behavior through an electrolyte NRTL approach
provided by ASPEN. In this model, the CO₂ Henry constant is
determined by the ratio between P_CO2 and the molecular CO₂
concentration at equilibrium provided by the thermodynamic model, which assures
the consistency between the mass transfer and the thermodynamic models.

Amines, water and bicarbonate dissociation
were supposed to be instantaneously equilibrated, while CO₂ chemical
reactions with amines and OH^- were assumed to be kinetically
controlled and reversible.

As CO₂ is absorbed, P_CO2
presents a decreasing profile within the reactor. Consequently, the proposed
model takes into account the evolution of the CO₂ partial pressure
through a plug-flow model combined with the film theory.

4. Results

This model has been validated by comparing
its predictions with experimental data obtained on unloaded aqueous solutions
of PZ ranging from 0,2 to 1 M and at temperatures between 24 and 60 °C (Servia
et al., 2013). The reaction mechanism proposed by Bishnoi and Rochelle, 2002
and their kinetic constants associated to equations 2 and 4 were used to model
the kinetics of the CO₂/PZ system.

 

4.1. CO₂
absorption on unloaded solutions

A first set of experiments were performed
on unloaded solutions presenting MDEA concentrations of 2 and 3 M, PZ
concentrations ranging from 0.2 to 1 M and temperatures from 297 to 328 K. The
experimental results and model predictions were presented in Figure 1. The
increase of MDEA concentrations has a negative effect on the overall CO₂
mass transfer, especially at low temperature. Indeed, the solution viscosity
rises involving a decrease of the CO₂ diffusion coefficient within
the liquid phase. The experimental results were used to regress the kinetic
constant of the synergy chemical reaction (Equation 3), which plays a
significant role at the tested conditions. Indeed, the model systematically
underestimates CO₂ flux with an AAD of 10 %, if this chemical
reaction is not considered in the reaction mechanism.

Figure 1 ? CO₂ absorption flux evolution
as a function of PZ concentration for different conditions of MDEA
concentrations and temperatures. Filled symbols: [MDEA] = 2 M; empty symbols:
[MDEA] = 3 M; continuous curves: [MDEA] = 2 M; discontinuous curves: [MDEA] =
3M. Curves represent model predictions whereas symbols represent experimental
data. a ? T = 297 K; b ? T = 319 K; c ? T = 328 K; d ? Comparison between model
predictions without considering the synergy chemical reaction between MDEA and
PZ and experimental data as a function of temperature.

4.2. CO₂ absorption on loaded solutions

A second set of experiments were carried
out on loaded solutions presenting MDEA concentrations of  2 and 3 M with a fixed
PZ concentration of 1 M, temperatures ranging from 297 to 328 K and CO₂
loadings up to 0.25 mol_CO2/mol_amine, at two different P_CO2.
The experimental results and model predictions were presented in Figure 2 .The
model systematically overestimates CO₂ flux at low P_CO2,
even if the deviation between experimental and model data is lower than the
determined experimental error. A methodology based on the experimental data
acquisition at operating conditions allowing the experimental error and the
model limitations impact to be reduced was then proposed and validated. This
methodology is based on the experimental data acquisition at higher P_CO2.
Model predictions are in good agreement with the experimental data at high
solutions loadings indicating that the reaction between MDEA and PZCOO^-
can be neglected at these conditions. Moreover, the model accurately reproduces
the overall CO₂ flux, even in presence of mass transfer limitations.

Figure 2 ? CO₂ absorption flux evolution
as a function of solution loading for different conditions of MDEA
concentrations and temperatures. Filled symbols: [MDEA] = 2 M; empty symbols:
[MDEA] = 3 M; continuous curves: [MDEA] = 2 M; discontinuous curves: [MDEA] =
3M. Curves represent model predictions whereas symbols represent experimental
data. a ? T = 297 K, y_CO2^inlet = 7000 ppm; b ? T = 319 K,
y_CO2^inlet = 7000 ppm; c ? T = 328 K, y_CO2^inlet
= 7000 ppm; d ? Parity chart at y_CO2^inlet = 50000 ppm.

4.3. Sensitivity analysis

A sensitivity analysis (see Figure 3) was
carried out involving all kinetics and diffusion parameters. At low loadings,
the kinetic constant associated to the interaction between PZ, CO₂
and water as a significant impact. An accurate prediction of the reaction
between PZCOO^-, CO₂ and water, as well as the liquid
phase species diffusion are required at higher CO₂ loadings. The
enhancement factor is sensitive to the synergy reaction (Equation 3) at low
loadings, high temperatures, and high MDEA concentrations.

Figure 3 ? Sensitivity analysis of the kinetics
parameters and liquid phase species diffusion coefficient. a ? T = 298 K, y_CO2^inlet
= 7000 ppm; b ? T = 298 K, y_CO2^inlet = 50000 ppm; c ? T =
330 K, y_CO2^inlet = 7000 ppm; d ? T = 330 K, y_CO2^inlet
= 50000.

5. Conclusion

This paper presents an experimental and
numerical investigation on the CO₂ absorption by aqueous blends of
MDEA and PZ. A synergy exists between both amines to enhance CO₂
capture, whereas the chemical reaction between the PZCOO^- and MDEA
to accelerate CO₂ capture can be neglected at high loadings.

A sensitivity analysis on the kinetic and
diffusion parameters indicates that an accurate description of the PZ
interaction with CO₂ is required at low loadings, whereas diffusion
phenomenon and PZCOO^- interactions with CO₂ become
essential at higher loadings.

References

Bishnoi, S., Rochelle, G.T., (2002). Absorption
of carbon dioxide in aqueous piperazine/methyldiethanolamine. Aiche Journal,
48, 2788-2799.

Edali, M. et al.  (2010). 1D and 2D
absorption-rate/kinetic modeling and simulation of carbon dioxide absorption
into mixed aqueous solutions of MDEA and PZ in a laminar jet apparatus.
International Journal of Greenhouse Gas Control, 4, 143-151.

Samanta, A., Bandyopadhyay, S. S.,
(2011). Absorption of carbon dioxide into
piperazine activated aqueous N-methyldiethanolamine . Chemical
Engineering Journal, 171, 3, 734-741.

Servia, A., Laloue, N., Grandjean, J.,
Rode, S. and Roizard, C. (2013). Modeling of the CO₂ absorption in a
wetted wall column by piperazine solutions. Oil & Gas Science and
Technology. DOI: 10.2516/ogst/2013136

Zhang, X. et al.  (2001). A kinetics
study on the absorption of carbon dioxide into a mixed aqueous solution of
methyldiethanolamine and piperazine. Industrial & Engineering Chemistry
Research, 40, 3785-3791.