49th International Conference on Environmental Systems ICES-2019-174 7-11 July 2019, Boston, Massachusetts

A Thermally-Regenerated Solid Amine CO2 Removal System Incorporating Water Vapor Recovery and Ullage Air Recovery

Holden Ranz1, Steven Dionne2 Collins Aerospace, Windsor Locks, CT, 06096-1010

and

John Garr3 National Aeronautics and Space Administration/Johnson Space Center, Houston TX, 77058

The Thermal Amine Scrubber (TAS) flight experiment was developed under contract with NASA/JSC with the goal to demonstrate a Technology Readiness Level 5 (TRL5) prototype advanced (CO2) removal system based on a thermally-regenerated solid amine adsorbent. The TAS was designed to fit within two double lockers within an ISS Express Rack, consisting of a Water Save subsystem (H2O Locker), a CO2 Removal subsystem (CO2 Locker), and an ISIS drawer for the system controller. The H2O Locker incorporates a passive water save desiccant canister technology capable of recovering ~90% of the incoming present in the process air stream. A supplemental desiccant wheel boosts the overall water recovery to ~97%. The CO2 Locker receives dry air from the H2O Locker and splits flow between two pairs of CO2 removal beds containing the solid amine. Each pair of beds is restricted to one adsorbing bed and one desorbing bed, the latter of which is isolated from the process air and exposed to vacuum during thermal regeneration. A valve assembly redirects flow back to the regenerated bed at a regular interval, but first ~96% of the ullage air from the previous adsorbing bed is evacuated using a scroll in an intermediary valve state. This design allows for a removal rate of ~3.7 kg/day of CO2 at 2 mmHg partial pressure CO2 in the process air, which corresponds to approximately a 4 crewmember equivalent.

The TAS will provide an additional means of removing CO2 from the ISS, potentially providing additional capacity for time periods with increased crew size. Addition of a vacuum compressor, residual water separator, and accumulator downstream of TAS will enable integration with a CO2 reduction system, provided the desorbed gas has acceptable CO2 purity, and is directly applicable to exploration missions requiring oxygen recovery.

Nomenclature AC = alternating current AEL = Advanced Engineering Laboratory at Collins Aerospace CD01 = Thermal Amine Scrubber carbon dioxide sensor CDRA = Carbon Dioxide Removal Assembly CO2 = carbon dioxide CO2 Locker = CO2 Removal subsystem DPIN = process air inlet dew point ECLSS = Environmental Control and Life Support Systems EMI = Electromagnetic interference EXPRESS = EXpedite the PRocessing of Experiments to Space Station Gr/lbm = grains water per pound of dry air

1 Advanced Technology Engineer, Space & Sea Systems, 1 Hamilton Rd, Windsor Locks, CT 06096, M/S 3-2-5F 2 Preliminary Design Engineer, Space & Sea Systems, 1 Hamilton Rd, Windsor Locks, CT 06096, M/S 3-2-5E 3 Exploration ECLSS Engineer, OB/ISS Vehicle Office, 2101 Nasa Parkway, Mail Code OB 311

Copyright © 2019 Collins Aerospace

H2 = H2O = water H2O Locker = Water Save subsystem ISIS = International Subrack Interface Standard ISS = International Space Station JSC = Johnson Space Center kg/day = kilograms per day kW = kilowatts lbs = pounds LTL = International Space Station Low Temperature Loop mmHg = millimeters of mercury MSFC = Marshall Space Flight Center MTL = International Space Station Moderate Temperature Loop NASA = National Aeronautics and Space Administration PCO2 = partial pressure of carbon dioxide PGW = propylene glycol/water RCA = Rapid Cycle Amine TAS = Thermal Amine Scrubber TRL5 = Technology Readiness Level 5 USB = Universal Serial Bus VAC = voltage alternating current VDC = voltage direct current VES = International Space Station Vacuum Exhaust System XIN = process air inlet humidity ratio

I. Introduction

Using a solid amine to remove CO2 from spacecraft cabin air has been an established method for over 30 years. Current units like the Amine Swingbed Payload1-4 and the RCA5-6 have successfully demonstrated the ability to control CO2 and humidity levels for spacecraft and spacesuits, respectively. Both units remove CO2 and water vapor from the air and dump them overboard to space vacuum. For long duration missions, recovery of concentrated CO2 and subsequent reduction with hydrogen (H2) to recover water (H2O) decreases resupply penalties associated with water for the electrolysis process (oxygen generation).

Discussions between Collins Aerospace and NASA in 2016 lead to the idea of launching a TRL5 flight experiment to the International Space Station (ISS) which would further develop solid amine CO2 removal technology while simultaneously mitigating potential CO2 levels with the pending arrival of additional crewmembers when the commercial crew vehicles begin their flights7. This led to a contract being issued in early 2017 between Collins Aerospace and NASA for the design, manufacture, test, and delivery of a Thermal Amine 1-6 based CO2 removal flight experiment utilizing the legacy solid amine . The top level performance requirements are provided in Table 1.

Table 1. Flight Experiment Performance Requirements Shall Requirement Should Requirement CO2 Removal (2 mmHg inlet & 760 mmHg cabin) 3.5 kg/day 4.16 kg/day Water Save 96% 99% Air Save 95% 97%

One of the findings with long term habitation on ISS is that to provide the crewmembers with optimal living conditions, the partial pressure of CO2 (PCO2) should be lower than previously thought. Currently in the US side of ISS the daily average PCO2 limit is 3.1 mmHg and with this limit crewmembers experience some adverse effects that are attributed to higher than desired PCO2. This TAS flight experiment, working in conjunction with ISS ECLSS, is 8 expected to maintain PCO2 at 2 mmHg for up to seven crewmembers . Based on the desire to operate future long duration spacecraft at lower CO2 levels the baseline inlet condition for this flight experiment is 2.0 mmHg. Most

2 International Conference on Environmental Systems

previous testing had been performed at higher inlet CO2 partial pressures. This paper discusses one of multiple next 9 generation CO2 removal technology demonstrations that are planned for evaluation on ISS .

This flight experiment has been named “Thermal Amine Scrubber” or by acronym, “TAS” and differs from the Amine Swingbed Payload in two significant ways. First, in an effort to improve CO2 removal the amine beds are actively cooled during CO2 absorption and actively heated during regeneration. Secondly, to minimize water loss, a bulk water save canister has been added to absorb the majority of the water and the desiccant wheel is limited to capturing the residual amount that gets through the bulk canister.

II. System Description TAS is packaged into two double EXPRESS lockers and an ISIS drawer as shown in Figure 1. The water save components are primarily in the locker on the right (H2O Locker) and the CO2 and air save components are primarily in the locker on the left (CO2 Locker). The majority of the control electronics are housed within the ISIS drawer.

Left: CO2 Locker / Right: H2O Locker TAS ISIS Drawer

Figure 1. Thermal Amine Scrubber Flight Experiment

The TAS schematic is provided in Figure 2. The cycling of a four bed amine based CO2 system is described thoroughly in an earlier ICES paper10 and the operation of this system is consistent with that description. The cycle sequence and timing was refined throughout development testing of the TRL4 system described in the earlier ICES paper, which was approximately ½-scale relative to the quantity of solid amine and desiccant used in the current system.

3 International Conference on Environmental Systems

49th International Conference on Environmental Systems ICES-2019-174 7-11 July 2019, Boston, Massachusetts

U.S. Export Classification EAR 9E515.a xClass ID: 7716499 Thermal Amine Scrubber – Payload Experiment

Double Locker #2 Cooling Cooling Express Stowage Drawer Bed A Cooling Return Interface Fluid Fluid Bed B Cooling Return Interface Outlet Inlet Case Case Fan Bed C Cooling Return Interface CO Removal Bed with with 2 Cooling Fluid Source Bed D Cooling Return Interface Screen Fn04 Screen Fn02 Bed A Cooling Supply Interface Accumulator Fill/Drain Bed B Cooling Supply Interface Tk02 Port Coolant Bed C Cooling Supply Interface Ht01 PC01 Pump QD07 Bed D Cooling Supply Interface Heater Pp01 Air Save Pump CO Removal Motor Controller 2 Motor HX01A LTL Interface HX - A Bd02 Bed A Controller Process Controller Heater QD08 Ht02 ISM QD09 LTL Interface HX - B HX01B Valve Valve Module AC01 x Module Valve Actuator Controller Vv02 Vv02 Ht03 AC02 Fill/Drain Valve Actuator Controller Heater LTL Orifice Or03 Port QD06 AC03 Valve Actuator Controller CO2 Removal

Bd03 Bed B QD ISS Low Temperature Loop QD QD01 QD02 Heater Ht04 QD03 QD04 QD ISS Moderate Temperature Loop QD Cp02 Double Locker #1 Cooling Cooling MTL DW01 x Or04 Fluid Fluid Orifice Outlet Inlet Case Fan HX1 Air Cabin with HX04 Save Screen Inlet Fn03 Pump Air HX2 Ht05 Heater HX03 x CO2 Removal Air Blower Motor Bd04 Bed C Motor Heater Motor Controller CD01 Desiccant Ht06 Bed A Controller CO2 Valve Valve Valve x Valve Module Module Module Module Vv09 Vv01 Desiccant Vv01 Vv03 Vv03 Bed B Ht07 ΔP Heater

PS11 Bd01 Variable CO2 Removal Ht09 Bypass Bed D Mf01 Bd05 To Orifice Heater Heater Or01 Cabin Acoustic Ht08 Muffler

Or02 Regen Orifice

Locker #1 Bump-out QD05 QD Cabin Inlet Air Cooling Fluid Outlet Cooling Fluid Inlet To Space Vacuum Figure 2. Thermal Amine Scrubber Schematic

Copyright © 2019 Collins Aerospace 49th International Conference on Environmental Systems ICES-2019-174 7-11 July 2019, Boston, Massachusetts

A. Water Save The cabin air intake is on the left side of the schematic. The system fan is on the return leg and draws the air flow through the system and as such the air circuit is at a slightly negative pressure. The intake air is filtered and then enters the bulk water save canister (schematic Bd01) where a desiccant removes and holds the majority of the water vapor. The bulk water save beds are thermally linked so that active heating or cooling is not required. Since the water adsorption process is exothermic and the regeneration process is endothermic, the thermally linked beds provide a means of passive between the beds. Air exiting the bulk water save canister has been dried to approximately 7°F dewpoint and subsequently flows through a section of the Desiccant Wheel (DW01) where additional humidity is removed. After the desiccant wheel the process air then flows over to the CO2 locker. Upon return from the CO2 locker the air is very dry, with a dewpoint of approximately -30 to -35°F. A fraction of this air flows through the loaded portion of the desiccant wheel and then reunites with the remainder of the dry air. The entire reunited air stream then travels through the desorbing bed in the bulk water save canister to be humidified to very close to intake conditions.

B. CO2 Adsorption After being dehumidified in the H2O save locker the very dry air enters the CO2 locker and is distributed among two, two-bed CO2 removal subsystems (Bd02/03 and Bd04/05) operating in parallel. A valve assembly associated with a pair of Solid Amine beds distributes the air flow to the adsorbing beds. While a bed is removing CO2, it is being actively cooled to remove the heat of adsorption and maintain lower bed temperature which enhances the amine CO2 loading. This bed cooling is accomplished with a pumped propylene glycol/water (PGW) loop that picks up heat from the amine beds and transfers it to the ISS Low Temperature Loop (LTL) via a liquid-to-liquid . Normaly open solenoid valves are cycled by the TAS controller to only allow flow through adsorbing beds.

C. Air Save As the first part of a bed state transition, when one of a paired set of beds is fully loaded with CO2 and the other has just completed a vacuum regeneration, the two beds will be connected and allowed to pressure equalize, which will result in each bed being at approximately ½ atmosphere in total pressure. The bed that had just been regenerated will then cycle to adsorption and its air volume will be returned to the process stream. Concurrently, an air save pump will remove air from the paired bed, which is CO2 loaded and headed toward desorption, and that air to the process stream. Any air not removed by the air save pump will be lost as this bed is exposed to vacuum for regeneration.

D. CO2 Regeneration To desorb CO2 from the Solid Amine Beds, the beds are heated and exposed to vacuum through the ISS VES interface. The heating and vacuum pumping are synchronized to minimize the amount of air that is lost and efficiently regenerate the bed.

E. TAS System Control The Bulk Water Save, the Desiccant Wheel, and the Solid Amine Beds are cycled to allow continuous adsorption and regeneration. Transitions between the various states are determined by time durations programmed into the system software and not by any sensor readings. The bulk water save canister cycling is dependant on bed size and water content in the intake air and is independent of other components. Similarily, the desiccant wheel rotates at a very slow constant rate to allow continuous adsorption and desporption of water and is independent of other components. The CO2 beds have multiple modes that consist of CO2 uptake (adsorption), an equilibrate & air save, a heated desorb, a passive desorb, a cooled desorb, and an equilibrate & isolate. The sequencing of these modes has been staged to ensure sufficient CO2 removal capacity and regeneration while respecting hardware limitations and adhering to NASA system performance and safety requirements. The timing of each mode has been determined by ground testing and can be adjusted on-orbit.

Copyright © 2019 Collins Aerospace

III. Packaging One of the contract requirements was that TAS not exceed the volume of one ISIS drawer and two double middeck EXPRESS lockers with bumpout into the cabin not to exceed 10 inches. Interfacing hoses and cables are allowed to exceed this constraint. The largest components were the bulk water save bed and the CO2 beds and they consumed the majority of the volume in the H2O and CO2 lockers respectively. The bumpout on the CO2 locker houses the vacuum isolation valve and the air save compressor. The bumpout on the water locker houses the desiccant wheel and its associated heater. The final weight of the CO2 locker was 285 lbs. and the final weight of the H2O locker was 235 lbs.

CO2 Locker H2O Locker

Figure 3. CAD models with locker shells ghosted to allow view of internal components

The system control electronics are housed within the ISIS drawer. 28 VDC power enters through a standard connector at the rear of the drawer (right side of Figure 4) and EMI filtering is performed within the adjacent enclosure. 120 VAC enters through the front of the drawer and is used exclusively for the CO2 bed heaters. The processor card has a USB connection to interface with NASA’s control network. The circuit cards that are mounted on edge control the various DC motors and monitor associated sensors. The interior layout channels airflow over the various heat sources to maintain component temperatures below rated limits. A small fan draws air through the drawer and exhausts it through the rear of the drawer.

6 International Conference on Environmental Systems

Figure 4. ISIS Drawer with cover removed

IV. Power

The Thermal Amine Scrubber was limited to 1.2 kW total. The sequencing of the CO2 bed regeneration, which involves heating, was established such that only one of the four beds is being heated at any time and AC power is used exclusively for the CO2 bed heaters. This arrangement results is a steady 400 watts of AC power. The balance of the system is powered by the 28 VDC connection at the rear of the ISIS drawer. Several items such as the fan, the coolant pump, and the desiccant wheel run continuously. Other items such as the air save compressor, the motors that cycle the diverter valves between adsorbent beds, and most of the solenoid valves operate intermittently. Test measurements show average 28 VDC power consumption of 514 watts and peak consumption of 526 watts.

V. Test Results NASA imposed compliance with SSP 57000 Rev R which is the ISS Interface Requirements Document for Pressurized Payloads11. Most of these requirements dealt with TAS interfaces with the ISS including, coolant loops, the VES, the ISS cabin, the crew, and the NASA command and control system.

NASA required a 350-hour burn-in test where the TAS hardware cannot exhibit an off nominal shutdown or hardware failure. The final 350 hours of testing was completed over the course of 3 weeks, from 2/1/2019 to 2/18/2019 and the following results are from this test. The TAS operates cyclically, where the sequence described in Section II.E. represents a complete cycle. Performance was evaluated using averages calculated for individual cycles as shown in Figure 6 through Figure 8. Error bars were determined from a propagation analysis originating with instrumentation error of measured variables used in performance calculations.

The TAS was integrated into a performance test setup in the Advanced Engineering Laboratory (AEL) at Collins Aeropsace in Windsor Locks, CT as shown in Figure 5. This setup included an air rig capable of conditioning CO2 partial pressure, dew point, temperature, and flow rate to desired levels for processing and performance evaluation. Additionally, to simulate the Moderate Temperature Loop (MTL) and Low Temperature Loop (LTL) a was designated for each in order to supply coolant at the appropriate temperature, flow rate, and pressure drop to interfaces on the H2O locker and CO2 locker, respectively. Appropriate levels were established based on SSP 57000 Rev R and NASA input. Furthermore, regeneration of the TAS CO2 removal beds is dependent on vacuum and will interface with the VES once installed on orbit. To simulate the VES in the performance test setup, we utilized a vacuum system in the AEL which consists of a roots-type vacuum pump with ducting on the suction side to provide vacuum utility to the air rig at a source level <1 mmHg. To protect the vacuum pump from excess moisture coming off the test article, there is an integrated liquid nitrogen cold trap.

7 International Conference on Environmental Systems

Figure 5. Photo (top) and block diagram (bottom) of TAS test setup in Collins Aerospace Advanced Engineering Lab

8 International Conference on Environmental Systems

A. CO2 Removal Results Prior to TAS system start up the AEL rig was nominally set to condition air to a PCO2 of 2.0 ±0.15 mmHg. This air was delivered to the controlled air volume and then fed to the TAS air intakes. The PCO2 was controlled using non-dispersive CO2 analyzers on the AEL rig which are calibrated daily at and 760 mmHg. PCO2 breakthrough was measured in the airstream returning from the CO2 Locker (via the AEL rig, purple series in Figure 6) and near the outlet by the system CO2 sensor (CD01 in the schematic, blue series in Figure 6). To calculate removal rate, ΔPCO2 (i.e. PCO2 In – PCO2 Out) was converted to a mass flow rate using the ideal gas relationship with the process air inlet flow rate, and scaled linearly according to the contract requirement (i.e. kg/day scaled for fixed PCO2 In of 2 mmHg and 760 mmHg total pressure).

The CO2 removal results show that the shall requirement was always met and the should requirement was met at certain points. The TAS system sensor PCO2 measurements exhibited slightly higher values than the AEL rig measurements likely due to the measurement location. CO2 carryover from bulk water save valve is not accounted for in the AEL rig measurement, but is observed by the TAS CO2 sensor, as reflected in Figure 6.

Figure 6. TAS burn-in test CO2 Removal results without scaling (top) and with scaling (bottom) 9 International Conference on Environmental Systems

B. Water Save Results Similar to CO2, the AEL rig was used to condition air delivered to the TAS intakes to a dew point of approximately 50 °F. This was established by injecting steam into the rig as necessary then routing the moist air through a chiller and reheater to control dew point and dry bulb temperature, respectively. A chilled mirror dew point sensor and T-type thermocouples on the AEL rig were used to verify the process/regen air inlet dew point and temperature, respectively. To understand the performance of the H2O locker, two aluminum oxide probe type dew point sensors were placed at intermediary locations on the TAS. One sensor was inserted into the process air hose going from the H2O locker to the CO2 locker, and the other was inserted into the return hose going from the CO2 locker to the H2O locker. Solving two material balances, one for the TAS process air and one for the CO2 locker, enables calculation of the water save efficiency. As seen in the results water save performance was always above the 96% shall and at times approached the 99% should requirement.

Figure 7. TAS burn-in test Water Save results

C. Air Save Results Ullage air save efficiency was calculated from CO2 bed pressure sensor data from the TAS. For select cycles, air save efficiency was calculated for each bed by taking the ratio of transition to equalization pressure. Equalization pressure refers to the start of the air save period after the CO2 valve turns and equalizes pressure between the respective bed pair. Transition pressure refers to the end of the air save period where the CO2 valve transitions from Air Save/Isolate state to the next Adsorb/Regen state. Air save efficiency is equal to one minus this pressure ratio. The results were fairly stable at 95.5%.

10 International Conference on Environmental Systems

Figure 8. TAS burn-in test Air Save results

VI. Next Steps As of this writing the Thermal Amine Scrubber flight experiment has been delivered to NASA and is being processed for launch to ISS on NG-11, which has a launch date in April 2019. The components will be launched soft stowed. Current plans are for NASA to destow TAS and install into EXPRESS Rack 2 as soon after arrival to ISS as crew time permits. A CAD image of the TAS in its anticipated installed state, with interface cables and hoses attached, is shown in Figure 9. Activation and checkout will follow shortly thereafter. Once installed operation is autonomous with no activity needed by the station crew or the ground. Data will be monitored and, as this is a flight experiment, we will compare on-orbit performance with ground performance and look for opportunities for further enhancements.

11 International Conference on Environmental Systems

Figure 9. Thermal Amine Scrubber installed in EXPRESS Rack 2

VII. Conclusions The Thermal Amine Scrubber has been tested rigorously on the ground and verified to meet specified system performance requirements. The flight experiment onboard the ISS will advance the TRL of solid amine sorbent technology. It will provide benefit to current ISS crews and advance progress toward fully closed loop environmental control and life support systems.

Acknowledgements The authors would like to thank NASA/JSC for funding this effort. It took longer than expected to get TAS into the launch preparation phase, however, the engineers and technicians at Collins Aerospace and NASA/JSC & MSFC maintained focus and completed the testing, evaluated the software, verified end-to-end command and control, and completed launch preparations all as a thoroughly integrated team. We look forward to on-orbit installation and beginning the test and evaluation phase of this experiment and hope to author a future ICES paper to share the results and benefits to the ISS crew.

References

1Lin, A., et al., "Testing of an Amine-Based Pressure-Swing System for Carbon Dioxide and Humidity Control," International Conference on Environmental Systems, 2007-01-3156, SAE, Warrendale, PA, 2007. 2Lin, A., Smith, F., Sweterlitsch, J., Nalette, T., and Papale, W., "Further Testing of an Amine-Based Pressure-Swing System for Carbon Dioxide and Humidity Control," International Conference on Environmental Systems, 2008-01-2101, SAE, Warrendale, PA, 2008. 3Button, A. B., and Sweterlitsch, J. J., “2009 Continued Testing of the Orion Atmosphere Revitalization Technology,” International Conference on Environmental Systems, 10.2514/6.2010-6163, AIAA, Reston, VA, 2010. 4Button, A. B., and Sweterlitsch, J. J., “Amine Swingbed Payload Testing on ISS,” ICES-2014-63, 44th International Conference on Environmental Systems, Tuscon, AZ, 2014. 5Chullen, C., Campbell, C., Papale, W., Wichowski, R., “Rapid Cycle Amine 3.0 System Development”, ICES-2015-313, 45th International Conference on Environmental Systems, Bellevue, Washington, July 2015. 6Swickrath, M., Anderson, M., McMillin, S., and Broerman, C., “Simulation and Analysis of Vacuum Swing Adsorption Units for Spacesuit Carbon Dioxide and Humidity Control," AIAA-2011-5243, 41st International Conference on Environmental Systems, Portland, OR, July 2011.

12 International Conference on Environmental Systems

7Nalette, T., “Thermal Amine Carbon Dioxide Removal System Feasibility Study Final Report,” Prepared for NASA JSC by Hamilton Sundstrand Space Systems International, Contract No. NNJ16GK03P, October 2016. 8Garr, J., “Thermal Amine Scrubber Tech Demo Overview,” NASA ECLS AIT Forum Presentation, Sponsored by NASA OB, January 2019. 9Knox, J. C., “Development of Carbon Dioxide Removal Systems for NASA’s Deep Space Human Exploration Missions 2017-2018,” ICES-2018-215, 48th International Conference on Environmental Systems, Albuquerque, NM, July 2018. 10Papale, W., Nalette, T., Heldmann, M. and Hidalgo, J., “An Advanced CO2 Removal System Using Regenerable Solid Amines,” ICES-2016-144, 46th International Conference on Environmental Systems, Vienna, Austria, July 2016. 11“Pressurized Payloads Interface Requirements Document International Space Station Program,” Revision R, NASA JSC, Document Number SSP 57000, October 2015.

13 International Conference on Environmental Systems