TVSA Direct Air Capture with Aspen
Adsorption
Copyright (c) 2024 by Aspen Technology, Inc. All rights reserved.
Aspen Plus, Aspen HYSYS, Aspen Custom Modeler and the aspen leaf logo are trademarks or registered trademarks
of Aspen Technology, Inc., Bedford, MA.
All other brand and product names are trademarks or registered trademarks of their respective companies.
Table of Contents
1 Background .......................................................................................... 1
2 Simulation Details .................................................................................. 1
2.1 Simulation Description....................................................................... 1
2.2 Model in Aspen Adsorption ................................................................. 2
3 Simulation Results ................................................................................. 3
3.1 Breakthrough Curve .......................................................................... 3
3.2 Temperature Vacuum Swing Adsorption (TVSA) Cycle ............................ 4
4 Conclusion ............................................................................................ 6
References ................................................................................................. 7
ii
1 Background
Carbon capture with underground geologic storage is one of the solutions that will
enable the energy transition, while carbon removal is needed to decarbonize hard
to abate sectors and remove historical emissions to get to net zero. One of these
removal technologies is Direct Air Capture (DAC). DAC can address emissions from
distributed sources as well as point sources; it is not location-specific and has the
advantage of not having to face the problem of contaminants (e.g. NOx, SOx).
The concept of DAC was first introduced for climate change mitigation by Klaus
Lackner in the 1990s [1]. In the past decade, more experimental work and
publications have been produced, and academics are shifting their interests to
developing new materials and processes for this technology. Additionally, several
start-up companies are moving this technology from the lab scale to the pilot and
plant scale [2].
The DAC processes can use either solid sorbents or aqueous basic solutions. The
pros and cons of each have been summarized by Keith et al. [3] Solid sorbents
offer the possibility of low energy input and operating costs applicable over a wide
scale. However, the costs of building a very large structure in the field while
allowing the entire structure to be periodically sealed from the ambient air during
the regeneration step are challenging. This is because of the temperature, pressure,
or humidity cycling and the conflicting demands of high sorbent performance, low
cost, and long economic life in impure ambient air. Aqueous sorbents can be
operated continuously by using less expensive cooling-tower technologies with long
contactor lifetimes irrespective of dust and contaminants, but the cost and
complexity of the regeneration system and water loss in dry environments are
challenging and expensive problems.
One process idea to remove CO2 directly from air utilizes the temperature vacuum
swing adsorption (TVSA) technology. This present work models a TVSA DAC
process to bring a new light on process system design and DAC performances under
cyclic operation. [4]
2 Simulation Details
2.1 Simulation Description
Aspen Adsorption is used to perform the simulation of the adsorption process for
capturing CO2. Different regeneration processes such as PSA, TSA, VSA and TVSA
are available in Aspen Adsorption that can be applied over various operating
conditions to model multiple active adsorption and desorption processes. In this
study, a TVSA simulation is carried out by using Aspen adsorption to capture CO2.
The schematic representation of the Aspen adsorption flowsheet is displayed in
Figure 1.
1
Figure 1. Flow diagram adsorption process in Aspen Adsorption
There are four different steps involved in the adsorption and desorption simulation
process: adsorption (AD), blowdown (BD), heating desorption (HD), and cooling. In
the adsorption step, ambient air at T = 25°C containing 400 ppm of CO2 is
introduced into the column for a duration for adsorption. The column walls are
assumed to be at T = 25°C, with the goal of adsorbing CO2. In the blowdown step,
a vacuum removes N2 from the void spaces to achieve high CO2 purities. The
pressure attained at the end of this step is P = 0.04 bar. During the heating
desorption step, performed at 0.06 bar, the walls are heated to a temperature of
125°C while collecting a mixture of CO2 and H2O at the outlet.
2.2 Model in Aspen Adsorption
In this example, a one-dimensional model for gas adsorption in a fixed bed is
developed in Aspen Adsorption. The adsorption isotherm and model assumptions
are described below.
The model consists of mass, momentum, and energy balances of the gas and the
adsorbent phases. The balance equations form a set of partial differential equations
(PDEs) that are solved, achieving a cyclic steady state in Aspen [Link]
process model is based on the following assumptions:
• The gaseous mixture obeys the ideal gas law.
• Only carbon dioxide is adsorbed.
• The resistance of mass transfer in the gas phase is negligible; the kinetics of
mass transfer within a particle is approximated by the linear driving force
(LDF) model.
• Isothermal condition is assumed for the energy balance calculation.
2
• The adsorbent is considered as a homogenous phase and its physical
properties are assumed to be constant.
The first-order upwind differencing scheme (UDS1) spatial discretization method
was selected for its comparatively higher stability. A default number of nodes (20)
was used. The Toth isotherm-based rate equations are used for describing CO2
adsorption kinetics as shown in Equations 1-4. [4]
𝑏(𝑇)𝑝𝐶𝑂2
𝑞𝐶𝑂2 (𝑇, 𝑝𝐶𝑂2 ) = 𝑛𝑠 (𝑇) (1)
𝑡(𝑇) 1/𝑡(𝑇)
[1+(𝑏(𝑇)𝑝𝐶𝑂2 ) ]
𝑇
𝑛𝑠 (𝑇) = 𝑛𝑠0 exp [𝜒 (1 − 𝑇 )] (2)
0
𝛥𝐻0 𝑇0
𝑏(𝑇) = 𝑏0 exp [ ( − 1)] (3)
𝑅𝑇0 𝑇
𝑇
𝑡(𝑇) = 𝑡0 + 𝛼 (1 − 𝑇0 ) (4)
The parameters q, b and t in the Toth isotherm described in equations are the
saturation loading, the affinity of the sorbent to CO2 and the heterogeneity
respectively and each parameter is a function of temperature. These parameters
can be used to calculate CO2 equilibrium loading (qco2) at various CO2 partial
pressures (Pco2) and operating temperatures.
Table 1. Toth isotherm parameters
Toth Unit Value
T0 K 296
ns0 mol•kg-1 2.38
b0 kPa-1 7.07E+05
t0 - 0.4148
ΔH0 kJ•mol -1 57.047
χ - -1.606
α - 0
3 Simulation Results
3.1 Breakthrough Curve
The breakthrough curves at air inlet conditions of 25 °C, at atmospheric pressure,
and 400 ppm on a clean adsorption bed are compared with the results from the
Aspen Adsorption fixed bed model in Figure 1. Table 2 presents modelling
parameters used in the fixed column model.
3
Table 2. Parameters for breakthrough modelling
Value Units Description
Hb 0.5 m Height of adsorbent layer
Db 0.12 m Internal diameter of adsorbent layer
Ei 0.46 m3 void/m3 bed Inter-particle voidage
Ep 0.23 m3 void/m3 bead Intra-particle voidage
RHOs 1580 kg/m3 Bulk solid density of adsorbent
Rp 2.60E-04 m Adsorbent particle radius
SFac 1 n/a Adsorbent shape factor
MTC("CO2") 3.26E-04 1/s Constant mass transfer coefficients
MTC("N2") 1.00E-08 1/s Constant mass transfer coefficients
IP(1,"CO2") 1.43 n/a Isotherm parameter
IP(2,"CO2") 0.403 n/a Isotherm parameter
IP(3,"CO2") 599 n/a Isotherm parameter
Figure 2 shows the breakthrough curve model results with the parameters
presented in Table 2. It shows that it should take about 20000 seconds to reach
peak summit. So, in the adsorption step, it sets 20000 s as the adsorption time.
Figure 2. Fixed bed operations breakthrough curves
3.2 Temperature Vacuum Swing Adsorption (TVSA) Cycle
In order to further validate the reliability and simulation capabilities of the software
tool, a full TVSA cycle was simulated using Aspen Adsorption. The Toth isotherm
parameters needed to be changed with the temperature. In this work, they are
calculated with a formula by hand and entered as constants in different cyclic steps.
Another option is to write Fortran code to edit the isotherm parameters as a
function of temperature if needed. The 1 cycle results are presented in Figure 3.
4
(a) CO2 loading profile (b) pressure profile
Figure 3. Cyclic profiles from Aspen Adsorption during the four steps
The TVSA cycle consists of four steps which are adsorption (25°C, 1 bar, 20000 s),
blowdown (0.04 bar, 300 s), heating and desorption (2000 s), and cooling (25°C,
1800 s) with purge steam. Figure 4 presents a cyclic CO2 loading at the end of the
adsorption bed over a period of ~24 hours. It presents a kilogram scaled DAC
model as a basis to study operating and economic performance.
Figure 4. Cyclic CO2 loading results at the end of column obtained from Aspen Adsorption
Based on the simulation results, 0.0363 kg CO2/cycle could be obtained, leading to
0.109 kg CO2/day. Indeed, as can be seen in Figure 3, a total of 3 cycles could be
achieved within a day. To achieve 1 kg CO2 capture per day, a DAC process with
several parallel fixed bed reactors. In this case, nine parallel reactors are needed to
capture 1 kg of CO2 per day.
For optimization, the step times can be changed for CO2 capture maximization. The
comparison results are shown in Table 3.
5
Table 3. comparison results for optimization
Item Unit Value
Cycle time s 24100 14100 21100
Adsorption time s 20000 10000 10000
Desorption time s 2000 2000 9000
Upper CO2 loading kmol 1.15E-03 1.04E-03 9.91E-04
Lower CO2 loading kmol 3.34E-04 3.02E-04 2.94E-05
CO2 removal kmol 8.18E-04 7.39E-04 9.62E-04
CO2 removal per cycle kgCO2/cycle 3.60E-02 3.25E-02 4.23E-02
CO2 removal per day kgCO2/day 1.08E-01 1.95E-01 1.69E-01
Parallel 9 5 6
Total CO2 removal per day kgCO2/day 9.72E-01 9.75E-01 1.02E+00
With the different step times, different CO2 removal results can be obtained. Here,
when adsorption time is 10000 s and desorption time is 2000 s, the DAC process
can achieve about 1 kg CO2 capture per day with 5 parallel fixed bed reactors.
Other case studies can be seen in Table 3 and further studies can be conducted by
the user.
4 Conclusion
A full TVSA model for DAC is developed to analyze and optimize the cyclic operation
of this carbon removal technology. This example demonstrates a process scaled to
~1 kg CO2 capture per day from air for a fixed-bed DAC process model with the
breakthrough curve of the solid sorbents under DAC conditions in Aspen Adsorption.
This example can be scaled for industrial use and act as a starting point for more
sophisticated process development, economic evaluation, integration, and
optimization.
6
References
[1] Lackner, K. S., Ziock, H., Grimes, P. Carbon Dioxide Extraction from Air: Is It
an Option? In Proceedings of the 24th Annual Technical Conference on Coal
Utilization & Fuel Systems, Clearwater, FL, 1999, 885−896
[2] Sanz-Perez, E. S., Murdock, C. R., Didas, S. A., Jone, C. W. Direct Capture of
CO2 from Ambient Air, Chem. Rev. 2016, 116, 11840−11876
[3] Keith, D. W., Holmes, G., Angelo, D. S., Heidel, K. A Process for Capturing CO2
from the Atmosphere, Joule, 2018
[4] S. Kim, G. L ́eonard, Process modelling of Direct Air Capture (DAC) of CO2 using
solid amine sorbents, Comput. Aided Chem. Eng. 51 (2022) 265–270s