American Institute of Aeronautics and Astronautics
1
2009 Continued Testing of the
Orion Atmosphere Revitalization Technology
Amy B. Button
1
Engineering and Science Contract Group/Jacobs Technology, Houston, Texas, 77058
and
Jeffrey J. Sweterlitsch
2
NASA Johnson Space Center, Houston, Texas, 77058
An amine-based carbon dioxide (CO
2
) and water vapor sorbent in pressure-swing
regenerable beds has been developed by Hamilton Sundstrand and baselined for the Orion
Atmosphere Revitalization System (ARS). In three previous years at this conference,
reports were presented on extensive Johnson Space Center (JSC) testing of this technology
in a sea-level pressure environment, with simulated and real human metabolic loads, in both
open and closed-loop configurations. The test article design was iterated a third time before
the latest series of such tests, which was performed in the first half of 2009. The new design
incorporates a canister configuration modification for overall unit compactness and reduced
pressure drop, as well as a new process flow control valve that incorporates both compressed
gas purge and dual-end vacuum desorption capabilities. This newest test article is very
similar to the flight article designs. Baseline tests of the new unit were performed to
compare its performance to that of the previous test articles. Testing of compressed gas
purge operations helped refine launchpad operating condition recommendations developed
in earlier testing. Operating conditions used in flight program computer models were tested
to validate the model projections. Specific operating conditions that were recommended by
the JSC test team based on past test results were also tested for validation. The effects of
vacuum regeneration line pressure on resulting cabin conditions was studied for high
metabolic load periods, and a maximum pressure is recommended.
Nomenclature
ARS = Atmosphere Revitalization System
Btu/hr = British thermal units per hour
°C = degrees Celsius
CAMRAS = CO
2
And Moisture Removal Amine Swing-Bed
cfm = cubic feet per minute
CO
2
= carbon dioxide
ELS = Exploration Life Support
°F = degrees Fahrenheit
g/min = grams per minute
GAC = Gas Analyzer Console
H
2
O = water
HMS = Human Metabolic Simulator
HSIR = Human-Systems Integration Requirements
JSC = Johnson Space Center
kg = kilograms
1
Project Engineer, Exploration Life Support Atmosphere Revitalization Systems, 2224 Bay Area Boulevard, Mail
Code JE77, AIAA Lifetime Member.
2
Project Engineer, Exploration Life Support Atmosphere Revitalization Systems, 2101 Nasa Parkway, Mail Code
EC3, not an AIAA Member.
American Institute of Aeronautics and Astronautics
2
kPa = kilopascals
lpm = liters per minute
min = minutes
mmHg = millimeters of mercury
Pa = Pascals
ppCO
2
= partial pressure of carbon dioxide
psia = pounds per square inch, absolute
psid = pounds per square inch, differential
I. Introduction
Human beings produce carbon dioxide (CO
2
) when they breathe, but too high a concentration in the atmosphere
around them can quickly become toxic. For this reason, CO
2
control is critical in the closed environment of a
spacecraft. Humans also exhale water vapor and exchange water vapor with the atmosphere through their skin.
Although excessive water (H
2
O) vapor is not dangerous to humans, it can be uncomfortable, and it can be hazardous
to the electronic equipment in a spacecraft cabin, particularly if it condenses in undesired locations.
In the past, spacecraft have typically used separate systems to control CO
2
and humidity. CO
2
control methods
have included sorption by lithium hydroxide or zeolite compounds, and water has typically been collected by
condensing heat exchangers. However, those CO
2
sorption systems have tended to be large and heavy, whether
regenerable or not, and condensate water collection systems require a lower temperature thermal control system with
a large heat capacity.
As an alternative to traditional CO
2
sorption systems, Hamilton Sundstrand has spent many years developing
amine-based vacuum-regenerated adsorption systems. The first major implementation of this type of system, known
as the Regenerative CO
2
Removal System, was tested on the Space Shuttle in the early 1990s. This design and the
associated sorbent amine have since gone through a number of improvement cycles. The current iteration of the
system uses a pair of interleaved-layer beds filled with SA9T, which is a sorbent system comprised of plastic beads
coated with an amine.
SA9T, in addition to being a good CO
2
sorbent, also has a great affinity for water vapor. When water vapor is
removed from the cabin atmosphere with a regenerable sorbent instead of a traditional condensing heat exchanger,
the spacecraft cooling system can be greatly simplified by eliminating a fairly significant heat load as well as the
need for a low-temperature cooling loop. The interleaved bed system also minimizes total cabin heat loads due to
the adsorption and desorption processes. Hamilton Sundstrand studies have shown SA9T to be very stable over long
periods. For these and other reasons, this technology has been baselined as the primary CO
2
and water vapor
removal device for the new Orion spacecraft.
While Hamilton Sundstrand’s technology was already relatively well developed and had undergone subscale and
open-loop testing, NASA’s Exploration Life Support (ELS) and Orion Environmental Control and Life Support
System development groups wanted more details on the performance of a full-scale device in a realistic spacecraft
environment. The ELS Air Revitalization Systems team at Johnson Space Center (JSC) refitted an existing test
chamber to test Hamilton Sundstrand’s technology, which the Air Revitalization team calls the CO
2
And Moisture
Removal Amine Swing-bed, or CAMRAS.
The JSC team tested a single CAMRAS unit in two test phases in late 2006. The preliminary results of those
tests were presented at this conference in 2007. A second CAMRAS unit of slightly-modified design was added to
the system for the third phase of testing in mid-2007, and those results were presented at this conference in 2008. A
third, significantly redesigned, CAMRAS unit with a new, more flight-like, valve style was tested in the ambient-
pressure portion of a fourth phase of tests during the spring of 2009, and those results are presented in this paper. A
second portion of the fourth phase, reduced-pressure testing of many of the same cases, is planned for mid-2010.
II. Test Rig Description
To evaluate the CAMRAS for Orion use, it was placed in a controllable, well-mixed atmosphere of the
appropriate volume. A motive force for airflow through the amine beds and a vacuum source to simulate a link to
space vacuum was provided, as was a supply of dry pressurized air representing a launchpad regeneration capability.
The effects of humans on the cabin atmosphere were simulated with a Human Metabolic Simulator (HMS), and the
whole test rig was outfitted with various sensors to monitor test conditions and experimental results. Figure 1 shows
a simple diagram of the test rig described in this section.
American Institute of Aeronautics and Astronautics
3
Air flowed from the process loop inlet
past a filter, flow meter, and several sensors
before passing into the CAMRAS air inlet
port on the top of the unit. Air flowed out of
the CAMRAS unit through another line,
where several more sensors were located.
The blower outlet air was passed over
another thermocouple as it returned to the
chamber atmosphere. External gas analyzer
consoles (GACs) monitored sample gas
streams from the chamber atmosphere and
from the CAMRAS inlet and outlet lines,
and all samples were returned to the
chamber.
A. Test Article
In each CAMRAS unit, a valve directs
airflow from the cabin through the adsorbing
bed layers and back to the cabin, while
isolating the desorbing bed layers to a direct
line to space vacuum. The highly porous
plastic beads in this next-generation device
are coated with a liquid amine, which
becomes immobilized in the bead pores. In
this SA9T sorbent, both carbon dioxide and
water are adsorbed simultaneously and
somewhat independently. The CO
2
adsorption reaction generates some heat,
while the desorption reaction consumes heat;
the interleaving of bed layers helps conserve the overall system thermal energy so that no active heating or cooling
of the unit is required.
The previous CAMRAS design used a
spool-type valve to direct gas through the
adsorbing and desorbing beds, equalizing
pressure between them as it transitioned from
one end of its path to the other. A gas purge
could drive desorption for launch pad
operations at the cost of restricting the
vacuum desorption to a single end of each
bed. The new design tested in CAMRAS
Phase 4A, CAMRAS unit 3 (Fig. 2), uses a
linear multiball valve, in which four ball
valves with alternating port orientations are
linked together for actuation by a single
motor. The chain of valves rotates 270°
between two positions, equalizing pressure
between the adsorbing and desorbing beds as
it turns. Figure 3 shows a simple schematic of
these new rotary valve flow paths with
vacuum desorption. This new valve also
allows purge gas to be pushed into one of the
vacuum ports and out the other for launch pad
operations while maintaining the capability to
vacuum desorb the beds from both ends when
the vehicle is on orbit.
vacuum: CO
2
+ H
2
O
CAMRAS
Tes t
Article
GAC1:
O
2
, CO
2
,
H
2
O
GAC3:
CO
2
x2
Air Handling:
heat, cooling,
circulation
T
P,T
airair
air
air: low
CO
2
& H
2
O
air
blower
T
dP
H
2
O,
F,T
P,T
P
T x4
bleed
assembly
T
T x4
P
CAMRAS
purge air
purge gas vent
HMS
air
air +CO
2
+H
2
O
F
T,
H
2
O
T,
H
2
O
T
Figure 1. CAMRAS Phase 4A simplified test rig schematic.
Figure 2. CAMRAS unit 3.
American Institute of Aeronautics and Astronautics
4
B. Test Chamber
The test chamber was a closed and sealed environment directly monitored for temperature and pressure. Air
conditions in the chamber were also analyzed by an external sampling rack (Gas Analyzer Console 1, or GAC1) for
dew point and for CO
2
and O
2
concentrations. Inside the chamber, a condensing heat exchanger with blower was
operated with the coolant loop above condensing temperatures to both control temperature and provide ambient
circulation. The total free volume of the chamber test volume was approximately 16.14 m
3
. The nominal Orion
configuration calls for operation of two CAMRAS units, so for most CAMRAS Phase 4A test cases, which only
used one unit, the chamber free volume was further reduced with airtight space-filling boxes to about 8.05 m
3
, or
half the projected vehicle free volume. The chamber leak rate at the beginning of the Phase 4A testing was
measured at an average 8.9% per day by a CO
2
decay test with all external air loop systems (analyzers, metabolic
simulator) circulating air out from and back into the chamber and the volume fillers installed.
C. Metabolic Simulation
A Human Metabolic Simulator (HMS) was used with the chamber for this testing. It was designed to simulate
human production of CO
2
and exhaled H
2
O vapor. Liquid water was pumped into a hot oil/water heat exchanger at
a metered rate, the resulting steam was controlled to achieve slight pressurization (up to about 69 kPa gauge), and
the steam was then injected directly into a chamber air circulation loop. CO
2
was separately injected into the air
loop from a pressurized and flow-controlled gas source.
The CAMRAS tests were typically run with simulated loads representing four or six people.
*
*
A six-person crew was eliminated from standard Orion operations plans, but this test series was already in progress
when that change was implemented.
Table 1 lists the
metabolic constituent generation rates used in CAMRAS Phase 4A testing. These rates were usually halved for
metabolic loads of two or three people with a single test article. The rates are based on the early 2007 version of
NASA’s Human-Systems Integration Requirements
2
(HSIR) and represent 82-kg males.
Figure 3. CAMRAS rotary linear multiball valve flow paths.
American Institute of Aeronautics and Astronautics
5
Exercise scenarios were run with only four
simulated people, as a crew of six would not
have enough space to exercise in the Orion
capsule. The metabolic loading provided by the
HMS was increased from four at nominal level
to three nominal plus one person exercising for
the time period that it would take for the entire
simulated crew to complete their exercises.
Exercise for each crew member was simulated at
75% of maximum volumetric oxygen use rate
and 5% exercise efficiency for 30 minutes, with
a 15-minute break between active exercise
periods. Water generation rates for each
simulated exerciser remained elevated for
60 minutes of cooldown time, where people
would continue sweating after ceasing the actual exercise. To approximate real-life metabolic loading profiles, the
HMS output rates were manually stepped up and down by prescribed amounts every 7.5 minutes for the nearly four
hours required for all four exercise and cooldown periods.
D. Test Article Air Flow
Airflow through the CAMRAS could be controlled within a range of rates, depending on the experimental
scenario, and it was designed to overcome the pressure drop caused by the plumbing fixtures and the amine beds
themselves. Several sensors, including those measuring temperature, moisture, and airflow rate, were tapped into
this plumbing stream. CO
2
analysis was provided both upstream and downstream of the CAMRAS by external
analyzers in closed sample loops. A cold trap upstream of each CO
2
analyzer minimized the adverse effects of
water vapor on the accuracy of the readings. The sample lines connected to both ends of the CAMRAS were cross-
connected to enable cross-checks of the analyzer readings, and they were also fitted with connections to gas bottles
that allowed the readings of the CO
2
analyzers to be compared to a known standard on a daily basis. Both of these
were intended to minimize unseen errors in the collected data.
E. Test Article Regeneration
In the flight environment of the Orion, the CAMRAS would be plumbed through a hole in the spacecraft shell,
allowing it direct access to space vacuum for desorption of CO
2
and H
2
O from the sorbent beds. Vacuum for the
test was provided by a facility vacuum pump system. The vacuum flow path was fitted with pressure and
temperature sensors to help characterize the CAMRAS performance. The vacuum line pressure near the CAMRAS
unit could be varied within a small range to simulate the effects of long and small versus short and wide flow paths
to space vacuum. This variability should help refine the Orion vacuum plumbing requirements and also allowed
validation of performance models based on different vacuum pressures.
Phase 4A testing included testing of the gas purge option for desorption while the Orion vehicle would be sitting
on the launch pad. A supply of dry compressed air was provided into the chamber, a flow controller allowed testing
at different rates of gas flow, and the gas was vented to ambient pressure outside the building. A three-way valve on
one of the CAMRAS unit's vacuum lines allowed purge gas to be pushed into that line or vacuum to be pulled on it,
depending on the test case.
III. Test Cases and Results
Several different types of test cases were included in the Phase 4A series. At the beginning of every JSC
CAMRAS test series, functional checkouts are run to ensure that the new test rig performs fundamentally the same
as it did in prior test series and that all of the instrumentation and controls respond as expected. When testing a new
CAMRAS unit, a few cases are run simulating vendor pre-delivery tests, to ensure that the unit has not been
damaged in transit.
CAMRAS Phase 4A then tested a series of representative flight operations scenarios. Baseline cases were run at
standard air flow rates and valve cycle time for various metabolic loads to provide direct comparisons to the
performance of the other CAMRAS units in previous test series. Two additional baseline cases were run with the
CAMRAS unit enclosed in an insulation blanket. This testing provided JSC computer modelers with data on the
thermal effects of running a CAMRAS unit without environmental heat exchange, representing operations in an
Table 1. Human-Systems Integration Requirements
metabolic constituent generation rates used in CAMRAS
Phase 4A testing.
Crew Size
& Activity
4 Sleep
4 Normal
3 Rest,
1 Exercise
6 Sleep
6 Normal
Metabolic CO
2
Generation Rate
(g/min)
1.82 2.88
peak
7.15
2.73 4.32
Metabolic H
2
O
Generation Rate
(g/min)
2.52 4.71
peak
21.38
3.78 7.06