lunar water resource demonstration (lwrd) test results · 2017-01-10 · american institute of...

American Institute of Aeronautics and Astronautics1

Lunar Water Resource Demonstration (LWRD) Test Results

Anthony C. Muscatello1, Janine E. Captain2, and Jacqueline W. Quinn3

National Aeronautics and Space Administration, Kennedy Space Center, FL, 32899

and

Tracy L. Gibson4, Stephen A. Perusich5, and Kyle H. Weis6,ASRC Aerospace Corporation, Kennedy Space Center, FL, 32899

In cooperation with the Canadian Space Agency, the Northern Centre for AdvancedTechnology, Inc., the Carnegie-Mellon University, JPL, and NEPTEC, NASA hasundertaken the In-Situ Resource Utilization (ISRU) project called RESOLVE. This projectis a ground demonstration of a system that would be sent to explore permanently shadowedpolar lunar craters, where it would drill into the regolith, determine what volatiles arepresent, and quantify them in addition to recovering oxygen by hydrogen reduction. TheLunar Prospector has determined these craters contain enhanced hydrogen concentrationsaveraging about 0.1%. If the hydrogen is in the form of water, the water concentrationwould be around 1%, which would translate into billions of tons of water on the Moon, atremendous resource. The Lunar Water Resource Demonstration (LWRD) is a part ofRESOLVE designed to capture lunar water and hydrogen and quantify them as a backup togas chromatography analysis. LWRD was designed 1) capture up to 6 g of water perregolith/soil core sample on a water absorber, (2) quantify up to 20 g of water on the samesample using relative humidity measurements, (3) capture and quantify up to 0.10 g ofhydrogen from the core sample, and (4) quantify the water and hydrogen within ±20%accuracy. Laboratory and analog field testing of the subsystem showed that it met its goalssuccessfully. RESOLVE was integrated with the Scarab rover from CMU and the wholesystem was successfully demonstrated on Mauna Kea on Hawaii in November 2008.Subsequent laboratory testing showed the hydrogen capture/quantification technique wasalso successful. Electrolysis of lunar water could provide large amounts of liquid oxygen inLEO, leading to lower costs for travel to other destinations, in addition to being very usefulat a lunar outpost.

I. Introduction

1 Chemist, Applied Sciences Division, KT-D3, AIAA Member Grade MB.2 Chemist, Applied Sciences Division, KT-D1, Non-Member.3 Environmental Engineer, RESOLVE Project Manager, Applied Sciences Division, KT-D3, Non-Member.4 Senior Principal Investigator, USTDC, ASRC-24, AIAA Member Grade MB.5 Principal Investigator, USTDC, ASRC-24, Non-Member.6 Chemist I, USTDC, ASRC-15, Non-Member.

AIAA SPACE 2009 Conference & Exposition14 - 17 September 2009, Pasadena, California

AIAA 2009-6474

This material is declared a work of the U.S. Government and is not subject to copyright protection in the United States.


ESOLVE is a drilling and chemistry plant packaged onto a medium-sized rover that analyzes collected soil forvolatile components prior to heating the soil and reducing it at high temperatures in the presence of hydrogen to

produce water. RESOLVE stands for Regolith and Environment Science & Oxygen and Lunar Volatile Extraction.The RESOLVE project is a ground demonstration of equipment that would be sent to a permanently shadowed polarlunar crater to determine the form of the hydrogen indicated by lunar orbiters as well as extract oxygen usinghydrogen reduction.

As described by us previously,1,2 the RESOLVEEngineering Breadboard 2 (EBU2) consists of EBRC(Excavation and Bulk Regolith Characterization, i.e. aDrill and a Crusher), a Reactor, the Regolith VolatileCharacterization (RVC) subsystem, the Lunar WaterResource Demonstration (LWRD) subsystem, theRegolith Oxygen Extraction (ROE) subsystem, and theGround Support Equipment (GSE) cart. The RESOLVEEBU2 processing module was mounted inside theScarab rover supplied by the Carnegie-MellonUniversity Robotics Institute, who directed the rover todrilling sites. Figure 1 shows a photo of the RESOLVEchemistry module (RVC, LWRD, and ROE) afterassembly at the Kennedy Space Center.

A major goal of the RESOLVE project is to performa field demonstration of the process of locating a drillsite, drilling into the soil, taking core samples, crushing them into 1 mm particles, delivering them to the Reactor,

heating a quarter core sample and driving off volatiles,analyzing the volatiles, capturing the water evolved, andextracting oxygen by hydrogen reduction. These taskswere accomplished1 at site at an altitude of 2740 m (9000ft) on the Mauna Kea volcano on the Big Island of Hawaiias a lunar analog. The site on Mauna Kea consists of afield of fine-grained volcanic tephra with scattered lavarocks in a valley next to large cinder cones (Fig. 2). Toensure consistency of the soil sampled, sufficient tephra tofill several >1 m long plastic tubes was prepared by dryingthe native tephra at 110ºC for several hours. This processdecreased the water content from ~20%, which was muchtoo high compared to expected lunar samples, to about1.25%. The field demonstration occurred as planned. Thedesign, construction, and field testing of the LWRD aredetailed in reference 1. Here we will briefly summarizethese details and elaborate on laboratory testing since thefield test.

II. Lunar Water Resource Demonstration (LWRD) DescriptionThe Lunar Water Resource Demonstration (LWRD) is designed to support the objectives of the RESOLVE

project by capturing and quantifying water and hydrogen released by polar lunar regolith from permanentlyshadowed craters upon heating. This task shows the feasibility of using both materials for the production of waterfor astronaut use and shielding, for the production of rocket and vehicle propellant by electrolysis of the water, andthe use of hydrogen as a fuel in rockets and fuel cells. Efficient capture of the water and hydrogen is necessary tomake these uses possible.

LWRD (Fig. 3) consists of a surge tank to accept large gas samples from the Reactor, a rotary valve, a hightemperature relative humidity sensor, thermocouples, high temperature pressure transducers, several latching

R

Figure 2. Lunar-like terrain of the field test site onMauna Kea, Hawaii.

Figure 1. Photo of the RESOLVE chemistry module.


solenoid valves, two water capture beds, arecirculation/pressurization pump, and ahydrogen capture bed. The rotary valvesfor RVC and LWRD are contained in asmall aluminum box and the valves areheated to 150ºC to prevent condensationof water vapor from the Reactor. Theremainder of LWRD is mostly containedin a large aluminum box (see Fig. 1)which is held at 130ºC for the samepurpose. Only the LWRD hydrogencapture bed is outside the 130ºC box,mounted beneath the support plate. The150ºC box is mounted inside the 130ºCbox for thermal efficiency and tominimize volume. The purpose of LWRDis to (1) capture up to 6 g of water perregolith/soil core sample and quantify upto 20 g of water (as a backup to GCmeasurements), (2) capture and quantifyup to 0.10 g of hydrogen from the samecore sample (as a backup to GC

measurements), and (3) quantify the water and hydrogen within 20% accuracy. Fig. 4 shows a diagram of theLWRD components and their connections in the context of the full chemistry package.

Vent

Crusher

Ne

SurgeTank

H2O Capture

H2 Capture

GC

Vacuum

Pump

4 port rotary valve (Vx)

Rotary valve for restrictors

Solenoid valve (SVx)

Diaphragm Pump (Px)

Regulator (Rx) or Hand Regulator(HRxx)

Pressure Transducer (PTx)

Pressure Gauge (Gx)

Flexhose (FHxx)

EBU2 Layout

H2O

P

PP

P

T

T

P

TT

P1*

T (Hot Box Temp.)

LWRD

ROE

RVC

T

T

T

RH

Vent

FM

RH

Capture

P T

Hot BoxBlower/Heater

Thermocouple (Tx)

Relative Humidity Probe (RHx)

Flow Meter (FM35)

Hand Valve (HVxx) for refills

Gas Filter (Fx)

Pressure Relief Valve (PRx)

Quick Disconnect (QDx)

T

RH

FM

F

F

PR

PR

130oC Box

150oC Box

P

P

P

PR

PR

VentReactor

Vent

Vent

T

T

H2

P

T

H2Supply

PRAr

Supply

PR

AirCompressor

GSE

T

T

T (Scarab Temp.)Vent

Crusher

Ne

SurgeTank

H2O Capture

H2 Capture

GC

Vacuum

Pump

4 port rotary valve (Vx)

Rotary valve for restrictors

Solenoid valve (SVx)

Diaphragm Pump (Px)

Regulator (Rx) or Hand Regulator(HRxx)

Pressure Transducer (PTx)

Pressure Gauge (Gx)

Flexhose (FHxx)

EBU2 Layout

H2O

P

PP

P

T

T

P

TT

P1*

T (Hot Box Temp.)

LWRD

ROE

RVC

T

T

T

RH

Vent

FM

RH

Capture

P T

Hot BoxBlower/Heater

Thermocouple (Tx)

Relative Humidity Probe (RHx)

Flow Meter (FM35)

Hand Valve (HVxx) for refills

Gas Filter (Fx)

Pressure Relief Valve (PRx)

Quick Disconnect (QDx)

T

RH

FM

F

F

PR

PR

130oC Box

150oC Box

P

P

P

PR

PR

VentReactor

Vent

Vent

T

T

H2

P

T

H2Supply

H2Supply

PRAr

SupplyAr

Supply

PR

AirCompressor

GSE

T

T

T (Scarab Temp.)

Figure 4. LWRD schematic.

Insulated Water Beds

Surge Tank

Blower

Hydrogen Bed (Below Plate)

Figure 3. LWRD during construction.


III. Laboratory and Field Demonstration ResultsTesting at CMU allowed establishment of a baseline for

comparison to the field tests on Mauna Kea (Fig. 5). Thesections below show the results of these tests and comparethem to the requirements established for LWRD prior tothe field demonstration. Figure 6 shows an example of thedata collected during a full RVC/LWRD/ROE operationcycle during a test performed on 25 September 2008 atCMU. The RH and pressure inside the Surge Tank are usedto calculate water content.

A. Water Quantification and AbsorptionThe results for the quantification of water in both the

CMU and Mauna Kea tests are shown in Table 1. Testingwith pre-dried tephra showed that the amount of wateravailable was much less than the 1.25% measured in thelaboratory in a drying oven open to the atmosphere. Thereason for this is because the water vapor was being

evolved into a closed system in the reactor, thus setting up equilibrium for water to return to the tephra. Also, thetephra holds some water tightly, as shown during ROE tests during which not all the water was released until thetemperatures reached several hundred degrees in an essentially open system that used drying beds to capture waterand return the dry argon to the Reactor. The water beds in ROE were equipped with calibrated capacitance sensorsto measure the amount of water captured. The RH probe in the ROE subsystem was usually saturated during thedrying process and does not yield usable data. Except for the test on 9/25/08, the total of the % Water from LWRD

0.1

1

10

100

1000

12:00:00 PM 12:30:00 PM 1:00:00 PM 1:30:00 PM 2:00:00 PM 2:30:00 PM 3:00:00 PM 3:30:00 PM

Time, PM

RESOLVE CMU Generator Powered Run"Day 1 Ops" - 9/25/08

T2 (°C)

T5 (°C)

T13 (°C)

T1 (°C)

T9 (°C)

PT2 (psia)

PT2 (psia)

PT4 (psia)

PT52 (psig)

PT53 (psig)

T51 (°C)

T56 (°C)

CS3 (g H2O)

CS4 (g H2O)

RH34 (%)

T34 (°C)

FM35 (l/min)

PT31 (psia)

PT34 (psia)

Surge Tank RH,%

First SurgeTank Transfer

Final SurgeTank Transfer

2nd Ar Purge(840C)

1st ArPurge(150C)

3rd Ar Purge -H2 Purge (900C)

H2 Off - Ar Purge -Cool Reactor

RVCHeat Up

Auger Temp., C

ROE H2OBed-2, g

ROE H2OBed-1, g

Surge Tank RH, %

Reactor P, psig

Dump Reactor

ROELWRD

ROE P, psia

ROE T, C

ROE RH, %

Reactor Wall Temp., C

H2 Tank P, psia

Surge Tank P, psia

ROE Flow Meter, SLPM

150C Box Temp., C

Surge Tank Temp., C

Scarab Ai r Temp., C

Figure 6. Example of a complete RVC/LWRD/ROE operation.

Figure 5. Operating the Chemistry Package insidethe SCARAB rover at the field test site on MaunaKea, Hawaii.


and the ROE drying process is fairly close to the expected value of 1.25%. Because it was not possible to preventexposure of the tephra to water vapor in the air once the container was opened for drilling, the high value for the %water on 9/25/09 is not unexpected. A comparison of the LWRD results to the RVC results for water content of thetephra was thoroughly discussed in our previous report1 and was found to be satisfactory.

B. Hydrogen Capture and QuantificationTo simulate lunar applications, samples of hydrogen and water were generated in the reactor in the laboratory at

KSC, and the system was tested with these simulated samples after the field test. Because the average hydrogencontent of shaded polar lunar craters is estimated to be 0.11%, we decided to test hydrogen capture, absorption, andquantification with the following masses of hydrogen: 0.0264 g, 0.044 g, and 0.0704 g. These masses correspond to30%, 50%, and 80% of an average 0.088 g of hydrogen expected to be found in an 80 g sample of regolith in aquarter core sample. Tests were run with the same 83.2 g sample of tephra to avoid loading and unloading theReactor. Water driven off during RVC/LWRD was replaced by adding 1.0-1.5 g of water back into the tephra in theReactor before each run after the first one. ROE was not performed so the tephra maintained most of its originalqualities.

The automated operating procedure used during the Mauna Kea demonstration was modified to include thehydrogen transfer, pressurization, absorption, and quantification steps needed. RVC was calibrated for hydrogen toallow a comparison between the two systems. The operating procedure was to first load the sample into the reactor.This consisted of either adding the tephra for the first run or adding in the lost water during the subsequent runs. Thereactor was then purged with helium to remove air. The reactor was next purged with hydrogen to ensure the gas inthe reactor was initially 100% hydrogen. After the purge, the reactor valves were closed, and the reactor waspressurized to a predetermined value that corresponds approximately to one of the three original masses of hydrogenin the test plan. Excess hydrogen was added to account for the leak rate of the reactor as well as the gases lost duringGC sampling. From here the system was operated as it was on Mauna Kea. The reactor was heated and GC samplesrecorded the evolution of water and the changing gas composition in the reactor. The gas samples were transferred at150°C or when the pressure was 85 psig. After the relative humidity and pressure were recorded in the Surge Tank,the gas was recirculated to capture the water, followed by transfer and pressurization of the hydrogen capture bed.With the water and hydrogen now absorbed, the inert unabsorbed gases were vented. The procedure was repeatedfor an additional transfer to the Surge Tank leaving less than 3 psia in the reactor. To quantify the hydrogencaptured, the hydrogen bed was heated to 300-428°C and opened to the Surge Tank where the resulting pressure ofhydrogen was measured. Finally, the system was vented and cooled.

Date Total WaterTransferred

Mass of Tephra % Water % Waterfrom ROE

Total %Water,

ROE+LWRD9/24/08 0.27g 85g 0.31 NA -

9/24/08 0.08g 90g 0.09 NA -

9/24/08 0.35g 85g 0.41 NA -

9/25/08 0.33g 85g 0.39 4.4 4.8

9/25/08 0.29g 85g 0.34 NA -

11/4/08 0.11g 66g 0.17 1.4 1.6

11/5/08 0.13g 72g 0.18 0.60 0.78

11/6/08 0.16g 92g 0.17 NA -

11/8/08 0.13g 71g 0.19 0.90 1.09

11/9/08 0.11g 76g 0.15 NA -

11/10/08 0.19g 90g (LN2) 0.22* 1.2 1.4

Table 1. LWRD Laboratory and Field Test Water Results.


Date andDesorptionTemperature

Hydrogen byRVC/GC, mg(Mass Transferredto Surge Tank)

LWRDHydrogenCaptured,mg

Captured %Differencefrom RVC/GC

LWRDHydrogen

Desorbed, mg(# Desorptions)

Desorbed %Differencefrom RVC/GC

12/15/2008 6.4 -56%

300oC (1)

12/16/2008 1.7 -82%

300oC (1)

1/27/2009 30.5 -82%

350-400oC (PureHydrogen)

(1)

1/28/2009 113.4 -31%

402oC (PureHydrogen)

(5)

1/29/2009 -8% 124.9 -26%


(6)

2/2/2009 -8% 77.2 -10%


(7)

2/4/2009 -15% 4.8 -80%

400oC (Dry Tephra+ Pure Hydrogen)

(6)

2/5/2009 -13% 49.7 -19%

400-425oC (DryTephra + Pure

Hydrogen)

(20)

2/6/2009 -3% 15.7 -54%

427oC (Dry Tephra+ Pure Hydrogen)

(8)

2/10/2009 -18% 25.5 -30%

428oC (Wet Tephra+ Pure Hydrogen)

(8)

2/11/2009 -3% 20.6 -41%


(7)

2/12/2009 36% 43.6 4%


(8)

2/13/2009 -21% 7.2 -70%


(7)

AverageDifference

-18.3% AverageDifference

-44.3%

167.8

85.8

23.8 18.9

42.1 57.1

35.1 34.0

36.6 30.1

34.3 33.4

61.1 52.9

23.8 20.3

154.2

78.8

89.3 -47%

143.4 -13%

9.2 6.8 -26%

167.4

164.4

14.4 0 -100%

Table 4. Comparison of RVC and LWRD Hydrogen Quantification Results.


Table 2 summarizes the results of the hydrogen capture/quantification tests. The results from RVC GC analysiswere corrected to give the amount of the hydrogen in the Reactor that would be transferred to the Surge Tank(72.2%) based on their volumes (0.175 l for the Reactor and 0.456 l for the Surge Tank). The amount of hydrogenabsorbed is calculated based on the pressure drop after opening the valve to the hydrogen bed and allowingsufficient time for the hydride to form, with the increased volume taken into account. The first attempt atquantifying hydrogen with LWRD (12/15/08) gave no indication of hydrogen absorption while pumping from theSurge Tank to the H2 Bed. However, there was a pressure rise in the Surge Tank of 3.0 psia during desorption of theH2 Bed, corresponding to 6.4 mg of hydrogen, giving a recovery of 44% of the RVC amount of 14.4 mg. After thesystem had cooled and residual gases were removed, addition of high pressures of pure hydrogen directly to the H2

Bed showed that its temperature rose as expected from the exothermic reaction of forming the hydride and that thehydrogen pressure in the bed declined, indicating that the hydrogen absorber was active.

Another RVC/LWRD run was attempted on 12/16/08. During hydrogen absorption, the pressure changeindicated 6.8 mg were captured or 74% of the RVC amount of 9.2 mg. Unfortunately, very little pressure rise wasobserved in the Surge Tank during desorption, corresponding to 1.7 mg or only 18% recovery of the RVC value, but25% of the absorbed amount. It should be noted that the amount of hydrogen initially available to LWRD was muchsmaller than the planned amount of 70 mg because of leaks in the Reactor and less importantly, losses to RVCsampling. Two tests with pure hydrogen in the Surge Tank not shown in Table 4 gave indications that the hydrideformer was active and does absorb hydrogen because the pressure dropped significantly and the bed heated up. Datafrom the Florida Solar Energy Center (FSEC), who made the hydrogen absorber used in this project, show thematerial should have a capacity of at least 0.25% by weight at hydrogen pressures as low as 100 mbar absolute for atleast 50 cycles.

Additional testing with pure hydrogen at 63 psia in the Surge Tank (1/27/09) determined that the hydride formerwas definitely active, with a capacity of 89.3 mg, which is 71% of the 125 mg capacity expected and 53% of thehydrogen available. Desorption of the hydrogen at 350°C recovered 30.5 mg or 34% of the hydrogen that wasabsorbed, but only 18% of the initial charge in the Surge Tank. Desorption was at 350°C in an attempt to improverecovery. After desorption, heating at 400°C under vacuum was used to further improve activity for the followingrun. A second test with pure hydrogen (1/28/09) gave better results, with the bed having a capacity of 143 mg, 87%of the initial hydrogen present or 0.29 wt%, which is higher than expected from the FSEC results. The highhydrogen pressure of 63 psia may be increasing the hydride former capacity up to this value. To still further improverecovery, desorption of hydrogen was performed at 400°C in a series of five steps, after each of which the SurgeTank was evacuated and reopened to the H2 Bed. The improvement was dramatic, with 113 mg of hydrogen beingreleased or 69% of the original amount or 79% of the absorbed hydrogen, a most gratifying result. Repeateddesorption into the evacuated Surge Tank circumvents any tendency of the hydride to set up an equilibrium withgaseous hydrogen.

A third test with pure hydrogen (1/29/09) gave even better results, absorbing 154 mg of hydrogen or 92% of thatavailable in the Surge Tank, a Hydrogen Bed capacity of 0.31 wt%. Desorption in six steps at 400°C returned 125mg of hydrogen, which is 74% of the initial amount and 81% of the absorbed amount. These results indicate that thehigher desorption and activation temperatures are quite effective in restoring the hydride former to its intendedcapabilities. Furthermore, the absorption and desorption results indicate that it may be possible to achieve thedesired agreement with RVC results within 20%. As seen in Table 2, the last two pure hydrogen tests averaged 89%capture of the hydrogen available and 72% recovery of the starting hydrogen. One more pure hydrogen test wasdesigned to simulate the pressures of hydrogen that would be experienced with amounts of hydrogen similar to thatmay be on the Moon. The 0.11% hydrogen concentration in regolith would generate 33 psia in the Surge Tank aftertransfer from the Reactor so this pressure of pure hydrogen was loaded into the Surge Tank and testing commencedas with the other pure hydrogen tests. The results (2/2/08) were excellent, with absorption of 92% of the hydrogen inthe Surge Tank by the H2 Bed, desorption of 98% of the absorbed hydrogen in seven steps, for an overall recoveryof 90% of the starting 85.8 mg of hydrogen. If translatable to a real run, such results would easily meet therequirement of agreeing with RVC results within 20%.

To test this ability in a more realistic situation, the reactor was loaded with 86 psia of hydrogen and heated to150°C using the standard procedure with the same tephra as used in the 12/15/08 run. The tephra had been driedthoroughly at 150°C. We decided to keep the tephra dry to establish a baseline performance without water present.Although leakage and GC samples depleted more of the hydrogen pressure than was anticipated, the results (2/4/09)


were better than the December tests, with the H2 Bed absorbing 85% of the hydrogen in the Surge Tank. However,only 24% of the hydrogen absorbed was desorbed in six steps or 20% of the available hydrogen. The test wasrepeated with a modification to the sequence of operation to allow recharging hydrogen in the Reactor just prior tothe final GC sample to overcome the increasing leak rate. The results of this test (2/5/09) were very good, with 87%of the hydrogen being captured and 81% being recovered. Desorption was repeated 20 times to extract as muchhydrogen as possible. One more dry tephra test was run on 2/6/09 to determine the performance of the system at alower hydrogen pressure and absorption results were excellent with 97% captured, but recovery was low at only46% in 8 steps.

Next, 1.5 g of water was added to the tephra and the test repeated with hydrogen added, resulting in 36.6 mg inthe Surge Tank (2/10/09). The percent hydrogen in the Surge Tank gases was corrected for the removal of watervapor by the Water Beds. The results were fairly good with 82% of the hydrogen as determined by RVC beingcaptured and 70% being recovered in eight steps. Another 1.0 g of water was added and a similar test on 2/11/09gave 97% capture, but only 59% recovery in seven steps. The hydrogen pressure was increased on 2/12/09, yielding42.1 mg in the Surge Tank. Capture was very high at 136% and recovery was a bit high at 104%. The pressure dropmethod to determine hydrogen capture is subject to error if other gases that will react with the hydride former arepresent, such as water vapor or oxygen. However, a GC sample taken from the Surge Tank during one of the wettephra tests showed the Water Beds had removed all the water vapor prior to hydrogen absorption and that only traceamounts of air were present. Consequently, an explanation for the high capture fraction is elusive. A final wet tephratest with 23.8 mg of hydrogen resulted in 79% capture, but only 30% recovery in seven steps.

Considering only the last seven tests in which we have some confidence that the hydrogen absorber was wellactivated and for which we have RVC results, the average absolute difference between the GC results and theLWRD captured results is 15.6%, which meets the goal of being within 20% of the GC results, an excellentachievement given the difficulties of working with such an active hydrogen absorber. The results for the desorptionprocess is not as good, with an average absolute difference of 42.6%. Clearly we need to improve the desorptionprocess somehow. Multiple desorptions helps tremendously in improving recovery, but sometimes somethinginterferes, causing low recoveries.

Visual presentation of the data gives an alternative assessment of the results. Figure 7 shows the three groups ofhydrogen absorption test results plotted vs. the amount of hydrogen initially present in the Surge Tank. A straightline fit of the dry tephra results appears to fit the other data points fairly well. Figure 8 is a similar plot of all thehydrogen desorption results. A linear fit of the wet tephra tests is a fairly good fit of all the data points except for thehigher pure hydrogen tests. The desorption data is expected to be somewhat more scattered because of the varyingnumber of desorptions performed in each test. In both figures, the 1/27/09 data points appear to be fliers, far fromthe scatter of the other data points, perhaps because the hydride former was still in the process of being activated.Deleting these two data points and fitting the results with straight lines gives the results in Figure 9. The R2 valuesfor both lines are remarkably good at 0.978 and 0.957 considering the variety of test conditions encountered. Thelinear fit for the desorption data was forced to have a zero intercept to avoid artificially improving the R2 value andthe slope.

The slope of the line for the absorption tests is equivalent to the average capture of hydrogen compared to theGC or pure hydrogen values and is quite good at 90.9%, meeting the goal of being within 20%. The same value forthe desorption tests is much lower at 73.3%, giving an average of 26.7% lower than the standard value, but muchbetter than the absolute value average of 42.6% different from the standard. This is to be expected because the lowsand the highs tend to cancel out the errors. Consequently, we can say that we have met the requirement to capturehydrogen within 20% of the known amount. On the other hand, we can say that we are close to meeting therequirement of recovering hydrogen within 20% of the known amount, but that improvement is required.


Figure 7. LWRD Hydrogen Absorption Test Results.

Figure 8. LWRD Hydrogen Desorption Test Results.


Addition of a palladium membrane between the recirculation pump and the Hydrogen Bed may improve theperformance of the system by excluding contaminating gases that definitely reduce the capacity of the hydrideformer and may interfere with efficient hydrogen absorption and desorption. Alternatively, the hydride former canbe coated with palladium to exclude these gases as well. Further work would be necessary to confirm thispossibility. If this approach is effective, it could remove the need to quantify hydrogen by desorption althoughefficient recovery of pure hydrogen would still be needed for lunar ISRU operations. Consequently, it is important tohave a hydride former that both absorbs and releases close to 100% of the hydrogen processed.

In all runs, the H2 Bed heated up from the exothermic absorption reaction, with a maximum of 76°C for the1/29/09 test. Plotting the amount of hydrogen absorbed vs. the temperature rise yields a fairly good fit to a straightline, as would be expected. This temperature rise might be quite useful in quantifying hydrogen for LWRD or otherlunar applications. More data points would be needed to reduce the error, but the principle has been demonstrated.

IV. Lessons LearnedAlthough RESOLVE was successful both in the laboratory and in the field campaign, some issues arose during

the design, construction, and testing phases that could be handled better in future projects.• Define and order commercial off-the-shelf (COTS) items much earlier in the process. In many cases,

commercial suppliers do not have items in stock, especially high cost items which are made only afteran order is confirmed. Instruments such as the relative humidity probe, equipment such as therecirculation pump, and even fairly simple items such as cartridge heaters may take weeks or monthsto receive after ordering. In addition, promised delivery dates tend to slip when vendors experiencunexpected problems.

Figure 9. Linear Fits of LWRD Hydrogen Adsorption and Desorption Test Results.


• Determine which items should have backup units on hand as early as possible. The first relativehumidity probe failed close to the shipment to CMU date for RESOLVE and it took a week or so to getit repaired. Testing of LWRD at KSC was incomplete as a result and the repaired unit had to beinstalled at CMU. Similarly, other spares ordered for the field test on Mauna Kea took weeks to obtainand resulted in concerns about not receiving them in time.

• Define and finalize requirements well in advance of the final design. The extreme wetness of the in-situ tephra at the Mauna Kea site resulted in the need for two large water beds. When the decision wasmade to dry the tephra, the water beds were oversized, adding unnecessary mass.

• Although obvious in retrospect, do not anticipate that measurements of water content of the driedtephra in open ovens with no time constraints will predict how much water will be evolved in a closedreactor running on a tight schedule.

In addition to the negatives, other lessons learned confirmed that well known concepts such as teamwork and astaff of talented, committed, experienced, and cooperative personnel are essential to project success. Regularworking meetings and teleconferences make sure issues are identified and resolved as soon as possible. It is alwaysessential that everyone know what they are supposed to be working on and know the changes as they occur. TheLWRD team worked very well together, accepting new leadership and new team members just over a year beforethe field demonstration occurred.

V. ConclusionAll the field test objectives were met, a remarkable achievement. Agreement of the RVC and LWRD water

determinations agreed within the 20% goal and the hydrogen absorption/quantification laboratory tests were alsowithin the required limits. These results confirm that the design and performance of the LWRD was adequate for thedifficult tasks assigned to it, demonstrating the talents of the team that made it.

AcknowledgmentsA.C. Muscatello would like to express his deep appreciation to the LWRD team at the Kennedy Space Center,

without whom none of this work would have been accomplished. Key participants were: Tracy Gibson, StevePerusich, Steve Parks, and Kyle Weis (ASRC); Jeremy Parr, Mark Nurge, Dale Lueck, Janine Captain, CurtisIhlefeld, Tom Moss, Jackie Quinn, and Bill Larson (NASA). We also wish to thank Mike Hicks and Mojib Hasan atthe NASA Glenn Research Center for thermal modeling of LWRD that assisted with heater and insulation designselections. Also, interactions with Julie Kleinhenz (GRC) and Candice Howard (JSC) during integration of LWRDwith the Reactor and ROE subsystems were extremely helpful. Funding for RESOLVE is provided by the NASAExploration Technology Development Program to the ISRU Program which directed by Bill Larson (KSC) withoversight by Jerry Sanders of Lunar Surface Systems.

References1A. Muscatello, J. Captain, C. Ihlefeld, W. Larson, D. Lueck, T. Moss, M. Nurge, J. Parr, J. Quinn, T. Gibson, S.

Parks, S. Perusich, and K. Weis, “Lunar Water Resource Demonstration,” AIAA-2009-1202, 47th AIAA AerospaceSciences Meeting, Orlando, FL, January 5 – 8, 2009.

2Lueck, D. E. , Captain, J. E. , Tracy L. Gibson, T. L. , Peterson, B. V., and Berger, C. M., “Selection, Development andResults for The RESOLVE Regolith Volatiles Characterization Analytical System,” CP969, Space Technology and ApplicationsInternational Forum—STAIF 2008, edited by M. S. El-Genk, American Institute of Physics, Melville, New York, 2008, pp. 149-156.

lunar water resource demonstration (lwrd) test results · 2017-01-10 · american institute of...

Documents