https://doi.org/10.1016/j.egypro.2017.03.1312
“2.1. Rotating Packed Bed (RPB).
This was manufactured from PEEK, polypropylene and 316L stainless steel. The inner diameter of the bed was
80mm and the outer diameter was 300mm giving a radial packed depth of 110mm. The radial packed depth is
equivalent to the packed height of a conventional absorber column. The length of the packed bed along the axis of
rotation was 20mm. This can be thought of as being equivalent to the diameter of a conventional packed column.
The RPB was packed with stainless steel expanded mesh grade 707 supplied by the Expanded Metal Company. The voidage and surface area of the packing were 0.801 and 663 m2 m-3 respectively. Two discs made of PEEK attached
to a stainless steel drive shaft supported the packing. The RPB was housed in a polypropylene case with an internal
diameter of 360mm and was driven by a 1.1 kW synchronous electric motor (ABB M3BP 80MC2) with a maximum
speed of 3000 rpm. The motor speed was controlled by a frequency inverter (ABB ACS150-03E-03A3-4). The
rotating packed bed was driven through a 2:1 pulley and timing belt so that the maximum speed of the RPB was
1500 rpm. A flag mounted on the motor shaft and a stationary proximity sensor measured the speed of the motor.
The motor was supported on two bearings and the power consumption of the motor was measured using a torque
arm attached to the base of the motor. Weights attached to he torque arm balanced the torque generated by the motor
under load.”
“2.2. Solvent Feed System.
The solvent being tested was fed to the RPB from a 300 litre polypropylene tank. The solvent was loaded into the
feed tank from either 210L drums or 25 litre containers that are placed on weighing scales so that the mass of
solvent transferred to the feed tank could be measured. A drum pump (Crest Pumps model number CBP-
TRP0900SEL) is used to transfer the solvent to the feed tank. The solvent was heated to 40ÛC using water
circulated from a Julabo SE-26 heating circulator with an external temperature probe in the solvent feed tank that
automatically maintained the temperature. The solvent was fed to the RPB through 16mm OD polypropylene pipe
using a stainless steel gear pump (Liquiflow Series 4 supplied by Michael Smith Engineers Ltd). The mass flow of
the solvent was measured using a Coriolis meter (Rheonik RHM06 supplied by Bronkhorst UK). The temperatures
of the solvent feed to the RPB and the solvent leaving the RPB were measured using a platinum resistance
thermometer (Pico Technology SE019). The liquid leaving the RPB flows into a sump where it was pumped out
into a receiving tank. The liquid leaving the RPB was sampled so that the CO2 loading could be determined.
Fig. 1 – Flow diagram for the rotating packed bed test rig.”
“2.3. Air/CO2 feed system.
To provide a simulated flue gas, air from the building supply was mixed with CO2 (BOC Industrial Grade CO2)
supplied from a cylinder (BOC VK size). The supply pressure of both gases is 400 kPa abs. The air and CO2 were
fed to gas mass flow controllers (Bronkhorst model D-6383 for the air and Bronkhorst model D-6371 for the CO2).
After the mass flow controllers the gas streams were mixed and fed to the air humidifier. The humidifier was a
polypropylene tank with a 400 L capacity filled to a depth of 700mm with water. The water was heated using hot
water circulated though a stainless steel coil from a Julabo F-34 water bath. The feed gas was bubbled through
porous polypropylene. A temperature probe in the gas line after the humidifier ensured the temperature of the feed
gas was maintained at 40ÛC prior to entering the RPB. The temperature, humidity and CO2 content of the gas are
measured at the RPB inlet and outlet using a GEM Scientific G100 gas analyser. The pressure drop between the gas
inlet and gas outlet was measured using a differential pressure gauge (Omega Engineering DPG409).
Figure 2 – Photo showing the connections to the rotating packed bed.
2.4. Experimental Procedure.
Experimental tests on amine solvents were carried out in two stages. Firstly both the 30 mass% and 90 mass%
solutions he solvent was pre-loaded to 0.1 mol CO2/mol MEA. This was achieved by feeding an air stream at 10 l s-
1 at 20ÛC to humidifier and then into the rotating packed bed. The air stream contained 5 mole% CO2 and was
contacted in counter current flow with 80 kg/hr of MEA. The inlet and outlet concentrations of CO2 in the air were
measured using the gas analyzer. When the amine had been preloaded to the desired level, the CO2 concentration in
the air was raised to 12 mole%. The MEA flow rate was varied for both 30 mass% and 90 mass% MEA solutions to
give MEA:CO2 molar ratios of 3.3, 4.0 and 4.4. This translates into liquid to gas mass flow ratios of 2.8, 3.3 and 3.7
when using 30 mass% MEA solutions and 0.9,1.0 and 1.2 when using 90 mass% MEA solutions. The volume
fraction of CO2 in the inlet and outlet gas was measured using the G100 analyzer and liquid samples were stored to
be analyzed at a later. ”