Automation Issues of the LLOX (Lunar Liquid Oxygen) Production Facility
30 April 1990

Go back to Computer Futures' Space Publications home page
1.0 Introduction
One of the most promising results of establishing a permanent manned outpost on the moon is the capability of producing sufficient oxygen to be used as a fuel to power spacecraft to Mars [33]. The capability of producing oxygen on the moon is also essential to achieving the ultimate goal of establishing a self-sufficient base on the moon by providing oxygen for breathing and for fueling reusable landers. Approximately 40% by weight of lunar surface material is oxygen, in the form of metal oxides of aluminum, iron, titanium, and silicon. This subreport includes this introduction (1.0), Task Overview (2.0), Issues Concerning Automation of the LLOX Production Facility (3.0), Conclusions (4.0), and References (5.0).
2.0 Task Overview
The lunar surface is a rich source of many minerals, from which many elements can be isolated. Appendix E considers the general issues of lunar soil exploitation. "The ability to delineate, mine, and process materials found on the lunar surface is an essential factor in the establishment of man's long-term presence in space.... The act of mining, or removing material (ore) from its natural environment, is a process which may be accomplished in many different ways" [24]. Criswell et. al. [5] suggested that one could reasonably attempt to recover by chemical means any of the 7 "major elements" in lunar material (oxygen, aluminum, calcium, iron, magnesium, silicon, and titanium), and that 6 "minor elements" (chromium, manganese, sodium, potassium, sulfur, and phosphorus) could be recovered in the course of major element processing, but that it would not be practical to recover the minor elements by chemical means solely for their own content. Of the trace elements, they only considered recovery by vacuum pyrolysis of the four light solar wind elements (hydrogen, helium, carbon, and nitrogen). More detailed considerations of the possibility of magnetic extraction of lunar free metals combined with a proposed electrorefining process for an iron-nickel-cobalt alloy suggests that all three metals are potentially recoverable. This report concentrates on the extraction, storage, and processing of oxygen, with a view to assessing the required degree of automation. Electrochemical experiments [10] have demonstrated the feasibility of extracting iron and oxygen from simulated lunar magma (molten rock). This is probably a required step in several of the reducing liquid-liquid extraction methods of chemical beneficiation (see Appendix E) to achieve reasonable enrichments of other trace elements without excessive dilution by molten iron. For this purpose it may be essential that the iron be removed in the solid rather than the molten state, which may impose constraints on the level of eutectic forming elements such as silicon or aluminum in the cathode and the melting temperature of the residual magma, especially as the iron content is deleted. Of course, material at or near the temperature of molten iron is hazardous for human beings, and therefore automation requirements are imposed by the very nature of high-temperature processing. Currently twenty-seven processes are receiving major consideration for use in the extraction of oxygen from lunar regolith. Some of the leading candidates are ilmenite reduction [29], magma partial oxidation [23], and magma electrolysis [28]. Since a decision has not been made between these processes, this section will avoid, as much as possible, paying special attention to any of these processes [30]. Some examples of differences between these processes, by way of illustration, are: (1) Ilmenite reduction imposes a heavy requirement on crushing and grinding (see Appendix E discussion of preconditioning and beneficiation). (2) Magma electrolysis can utilize regolith scooped up from immediately around the process plant, and so there is a reduced requirement for ore processing or transport (see 2.7 for an alternative to regolith haulers). (3) In general, materials handling, preconditioning, and beneficiation requirements are functions of the process chosen for oxygen production. The large scale production of oxygen on the moon will require well organized activities of several components. Ore will be mined and loaded into a regolith hauler (but see 2.7 for an alternative). The hauler will transport the ore to a pre-processing facility which will crush and screen the ore then load it onto a feeder belt or hopper. This feeder will load the ore into the processing facility which will extract the oxygen and any other components of value from the ore. The tailings will be removed from the processing facility and transported to a dump site. The oxygen will be liquified and stored in tanks. The majority of these tasks are covered in other sections of this document and will not be covered here. Why do we consider only the large-scale production of oxygen? One reason is that the adoption of chemical beneficiation steps (other than vacuum pyrolysis) is predicated [5] on the establishment of a substantial mass handling facility for native or physically beneficiated products manufacturing. Economies of scale come into effect only when large amounts of materials are processed, and thus automation requirements are imposed by the need to handle more material than a person can reasonably lift and carry unaided. The Eagle report assumes that the operation of this plant will require full time lunar base coverage. A production rate of 26 metric tons of liquid oxygen per month will require the mining of about 2500 metric tons of feedstock assuming 10 percent ilmenite in the feedstock material. This figure, by the way, is turning out to be over-optimistic. Current projects show ilmenite recovery of under 1%, thus imposing additional beneficiation requirements to concentrate the ore [12, 22]. The Eagle figure requires over 180 loads/month using an 8 cubic meter capacity regolith hauler. Along with the system monitoring, response to alarms and plant problems and maintenance, operations will require 4 full time IVA crew working in shifts. During some short term plant problem situations, more lunar base personnel may be required to provide a 2-man EVA team to resolve the problem while the normal plant operator monitors the situation and assists via lunar base teleoperation. During the day, sunlight can be used indirectly as a source of photovoltaic electricity. If nuclear power is a viable alternative, reliance on solar power can be minimized. Recent studies [12] suggest that sunlight cannot be used directly as a heat source by the use of sun-tracking mirrors, due to problems of mirror and window contamination. Specific tasks that the LLOX production facility will perform are the following: (1) Refine/Preprocess Ore (Crush/Screen) (2) Process Ore (3) Liquify & Store Oxygen (4) Remove Tailings (5) Perform Fault Detection & Prediction (6) Maintain/Repair/Replace Parts What follows is a brief task description discussion on each one of these tasks.
2.1 Crush and Screen Ore
Crush and Screen Ore: This is a purely mechanical process so automation should be the primary approach. Crushing is technically a size modification preconditioning, and is closely related to grinding, which accomplishes a further reduction. Both involve higher pressures than unaided people can exert, and involve mechanical stresses which present a degree of hazard to humans, so automation is clearly required. Screening, or sieving, is a size control/separation preconditioning process, and involves material flow and vibration which are best isolated from human beings by some degree of automation. "Lunar soils or rocks can be mined as sources of ilmenite for producing oxygen. However, seperable crystals of loose ilmenite in lunar soils are rare (less than 2%) and small (less than 200 micrometers); most ilmenite in the regolith is locked together with silicate minerals as rock fragments. If fragmentation of rocks is attempted to win appreciable amounts of ilmenite (roughly 10% or more), selective collection of high-Ti basalt fragments larger than 1 cm may be advantageous over extensive processing of fine lunar soil. Many alternative schemes for fragmenting rocks on the Moon have been proposed; one process which was tested early in the Apollo program successfully disaggregated lunar and terrestrial basalts by passive exposure to low-pressure alkali (K) vapor. This process is worthy of reinvestigation" [22].
2.2 Process Ore
Process Ore: This is a primarily chemical and electrochemical process so automation should be the primary approach. Processing may involve physical or chemical beneficiation (see Appendix E) with pressures, temperatures, corrosives, and poisonous reagents which represent a hazard to humans. Hazardous voltages and currents may be involved in three classes of electrochemical approaches being considered for lunar processing: 1. Direct electrolysis of lunar soil or beneficiated fractions. 2. Electrolysis of lunar soil fractions dissolved in fused salt (fluoride, chloride, or carbonate) solvents. 3. Electrolysis of sodium hydroxide (or chloride) followed by indirect reduction of aluminum, silicon, and other desired constituents as in a proposed flow sheet. Technical problems associated with these methods, and the automation issues raised, include: 1. Direct electrolysis: (a) solving anode corrosion problems, and possibly requiring robots to remove and install heavy electrodes from the process plant; (b) operating with higher viscosity melts and successfully recovering oxygen, but higher viscosity materials run greater risks of pipe and orifice blockage which may need to be cleared by robotic or teleoperated means; (c) developing secondary refining methods to separate constituents from alloys obtained, but this will introduce a greater variety of tasks, leading to either an increased requirement for the generality and flexibility of automation, or a greater number of specialized automated processes and devices. 2. Electrolysis of lunar soil fractions dissolved in fused salt solvents: (a) purification and recycling fused salt solvents, which introduces new processing tasks for automation, and higher overall risk; (b) avoiding carbon formation in reduction of carbonate melts, or else automating the tricky and tedious job of scraping carbon from containers, electrodes, and other surfaces; (c) development of bipolar cells (if needed); (d) attrition of reagents, which means a risk of contamination of human or telerobotic devices by escaping chemicals; (e) developing secondary refining methods for alloys, for which see 1(e) above. 3. Electrolysis of sodium hydroxide or chloride: sodium hydroxide is a powerful alkali, itself hazardous to humans; (a) effectiveness of proposed drying cycle for modified Castner cells used in the recovery of sodium in various recycle loops; but metallic sodium is considered hazardous for humans, due to fire and explosive risk in the presence of oxygen or water; (b) indirect recovery of oxygen with chlorine formed if Downs cell sodium route is chosen; chlorine is corrosive and poisonous to both humans and metallic robots. Electrodialysis is a useful technique in various stages of aqueous processing of lunar materials. Extension to nonaqueous solvents would require additional research and development money which might better be spent on automation and robotics, especially as these some of these options involve high temperatures and/or hazardous chemicals.
2.3 Liquify & Store Oxygen
Liquify & Store Oxygen: This is a purely mechanical process so automation should be the primary approach. Liquid oxygen is a cryogenic material, whose degree of coldness is a hazard to humans, thereby necessitating some degree of automation. In addition, there is a fire and explosion risk, threatening to both humans and automated systems. "Oxygen liquefaction is a key processing step of proposed lunar oxygen production plants.... An aspect of most all chemical processes is the removal of waste heat. Terrestrial chemical processing plants typically rely on the convection of air. The complexity of this task is greatly increased on the lunar surface due to the lack of an atmosphere and fluids. Heat rejection will be achieved by transferring the process heat to a cooling fluid which in turn flows through a heat rejection device. [Alternatives include] conduction to the lunar regolith and radiation into deep space [31, 32].
2.4 Remove Tailings
Remove Tailings: In most cases, the level of automation for this process should be the same as that required to haul the ore to the pre-processor. However, some of the processes produce molten byproducts, whose removal would require other systems, presumably at the same level of automation.
2.5 Perform Fault Detection & Prediction
Perform Fault Detection & Prediction: This process should be automated as a caution and warning system from the outset. This process will be among the most straightforward to enhance with more autonomous capabilities.
2.6 Maintain/Repair/Replace Parts
Maintain/Repair/Replace Parts: This can be among the most dangerous yet most delicate operations. Human participation will be required to a large degree for the foreseeable future. However, the dangerousrepair jobs would best be done telerobotically. Ultimately, a supervised approach would be desirable. But as this requires the development of sophisticated sensor, motor, and software capabilities, this task will be slower to mature than the above.
2.7 Ballistic Transport
There is an alternative for ore transport which limits the requirements for a regolith hauler. This is the concept of ballistic transport [11]. Due to the lack of a lunar atmosphere and low lunar gravity, it is possible that material can be thrown long distances to a small central collection point. Transport over tens of kilometers seems reasonable. The throwing process is more energy intensive than hauling, ecept for very deep pit mines, but would be quick, very flexible, and minimize the need for rolling stock (i.e. regolith haulers). Because of the velocity of thrown material, there is a degree of hazard for people to manually operate such a system. Teleoperation appears to be an unneccessary use of human operator time. This is a likely candidate for supervised robotic operations. The process would go on for days at a time with no human intervention. A human supervisor would occassionally monitor velocity dispersion, throw trajectory, angle of repose of the pile where material accumulates, and similar parameters. People should not be too near the reception point while material is being thrown there.

3.0 Issues Concerning Automation

Issues concerning automation of the LLOX production facility include: Process Comparison (3.1), Power-Up (3.2), Power Requirements/Operation Duty Cycle (3.3), Safety Issues (3.4), and Configuration Issues (3.5).
3.1 Process Comparison
Process Comparison: - Most processes have not been studied in depth. - Alternative processes to ilmenite reduction and magma electrolysis must be studied. - Studies must consider development costs, total facility capital costs and operating costs. - Initial lift weights must be considered. - Hardware complexity must be considered. - Studies must consider Earth-resupply.
3.2 Power-Up
Power-Up: - The safety issues of high temperature (1350 deg. C), high amperage (420,000 amps), and seal integrity for dangerous material containment favor the use of remote power-up. Automatic and remote control power-up capability should be built into the equipment to minimize the danger to teleoperated devices.
3.3 Power Requirements/Operation Duty Cycle
Power Requirements/Operation Duty Cycle: - Nighttime power requirements vs. power-down/power-up shock and production efficiency must be addressed. Duty cycles as low as 45% will be imposed if sole reliance on solar power is enforced. If the processing plant is powered down at night and powered up at dawn, the plant and the control regime must be designed to accomodate the mechanical shock which could otherwise severely stress components of the plant. Strict control of plant conditions, process evolution, and fault detection and continuous system health monitoring will be necessary to protect against these thermal shock effects. The plant can be made to run continuously (as high as 95% duty cycle) if a nuclear power plant with sufficient power to run the facility is emplaced. However, this raises political and environmental issues which may preclude this approach, and promote alternatives such as hybrid systems with photovoltaic with regenerative fuel cells, or photovoltaic with thermal storage. Alternate methods of storing energy for nighttime use such as gravity mills (in which dust and rocks are dropped into a rille onto a paddle wheel type turbine) may offer a reasonable compromise. Processes which produce molten magma can store a portion of the output to provide thermoelectric power to maintain hot standby conditions, thereby precluding a range of problems. - Power allocation will be a dynamic activity involving planning and resource management. Automation of this activity would free humans for more interesting tasks. This is especially true at night when solar power is unavailable and processes are competing for scarce energy resources thus imposing stringent constraints on task scheduling.
3.4 Safety Issues
Safety Issues: - Due to the high temperatures and volume of the material being processed, the LLOX plant will be placed sufficiently far from the hab module to minimize personnel danger in the event of a major process upset. To minimize the probability of a catastrophic failure, automatic shutdown capability must be built into the plant. These safety features may be effected by a combination of smart (active) and dumb (passive) mechanisms.
3.5 Configuration Issues
Configuration Issues: - Both the magma electrolysis and the ilmenite reduction process will require real-time modification of operational parameters such as electrode position, current and voltage levels, etc. to maximize the efficiency of the processing plant. This can be accomplished with currently available process control strategies.
4.0 Conclusions
The efficient generation of lunar oxygen is critical to the long term goals of the human space exploration initiative. To assure the success of this project, the most economically efficient method of oxygen extraction must be exploited. Also, an efficient automated process must be established from the acquisition of raw material to the disposal of the tailings and the storing of liquified oxygen. Monitoring the process of extracting oxygen from lunar regolith is very tedious. Ancillary activities such as maintenance and repair can be dangerous. The use of between four and six personnel for tasks of this nature is a waste of human resources and indicates the processes to be a prime candidates for automation. The full scale production plant will not be emplaced until flight 23 [1]. By this time, considerable equipment will already be at the lunar site. The regolith haulers will initially be operated telerobotically. However, they too should be designed with upgrade in mind. Since many of the operations of LLOX production are dangerous, automation is essential from the outset. Neither of the two oxygen extraction processes being considered rely on sophisticated sensor or actuator capabilities. Due to the repetitive nature and well defined domain of this task, the LLOX production operation offers perhaps the best platform for the initial implementation of autonomous systems.

5.0 References

[1] "Initial Study Period Results Summary, Planet Surface Systems, Reference Mission -- Option 1 -- Conceptual Design and Development Requirements", NASA, October 1989 [2] Eagle Engineering: Lunar Surface Operations Study, Dec 1, 1987 [3] Space Station Evolutionary Power Technology Study, Task II Report - Lunar Mission Power Impacts, Waldron, R.D. June 20, 1988 [4] Lunar Oxygen Process Design Considerations - briefing charts, Fluor Daniel [5] "Extraction of oxygen from lunar ores", N. Jarrett, S.K. Das, and W.E. Haupin, in "Extraterrestrial Materials Processing & Construction, Final Report", David R. Criswell, Ed., Lunar & Planetary Institute, NASA NSR 09-051-001 Mod. No. 24, Ch.VIII, p.255, 1980 [6] "Extraction processes for the production of aluminum, titanium, iron, magnesium, and oxygen from nonterrestrial sources", D.B. Rao, U.V. Choudary, T.E. Erstfield, R.J. Williams, and Y.A. Chang, in "Space Resources and Space Settlements", ed. J. Billingham and W. Gilbreath, NASA SP-428, p.257, 1979 [7] "Lunar oxygen by vapor phase pyrolysis", W.H. Steurer, in "Space Manufacturing 5", ed. B. Faughnan & G. Maryniak, AIAA, p.123, 1985 [8] "Non-electrolytic route to oxygen and metallic elements from, lunar soil", R.D. Waldron, in "Space Manufacturing", ed. J.D. Burke and A.S. Whitt, Vol.53, Advances in the Astonautical Sciences, AAS, p.297, 1983 [9] "Lunar Manufacturing: A Survey of Products and Processes", R.D. Waldron, Acta Astronautica, Vol.17, No.7, pp.691-708, 1988 [10] D. J. Lindstrom and L. A. Haskin, "Electrochemistry of Lunar Rocks", 4th Princeton/AIAA Conference on Space Manufacturing Facilities, March 1979 [11] R. D. Waldron, Rockwell International, personal communication, 1990 [12] E. M. McCullough, Rockwell International, personal communication, 1990 [13] Stephen L. Gillett, "Lunar Ores from Magmatic Processes: A Speculative Assessment", Proc. Space-90 Engineering, Construction, and Operations in Space, Albuquerque, New Mexico, 22-26 April 1990, pp.88-97 [14] Paul W. Weiblen, Marian J. Murawa, and Kenneth Reid, "Preparation of Simulants for Lunar Surface Materials", Proc. Space-90 Engineering, Construction, and Operations in Space, Albuquerque, New Mexico, 22-26 April 1990, pp.98-106 [15] J. H. Alton, C. Galindo, Jr., and L. A. Watts, "Guide to using Lunar Soil and Simulants for Experimentation", in Lunar Baeses and Space Activities of the 21st Century, W. W. Mendell (ed.), 1985, pp.497-506 [16] A. B. Binder, "On the Origins of Lunar Pristine Crustal Rocks", in Proc. Conf. on the Lunar Highlands Crust, eds. J. J. Papike and R. B. Merrill, 1980, pp.71-79 [17] S. S. Goldich, "Lunar and Terrestrial Ilmenite Basalt", Science, Vol. 171, 1971, pp.1245 ff [18] J. J. Papike, F. N. Hodges, A. E. Bence, M. Cameron, and J. M. Rhodes, "Mare Basalts: Crystal Chemistry, Minerology, and Petrology", Reviews of Geophysics and Space Physics, 14(4)475-540, 1976 [19] H. H. Schmitt, G. Lofgren, G. A. Swann, and G. Simmons, "The Apollo 11 Samples: Introduction", Proc. Apollo 11 Lunar Science Conf., Vol. 1, pp.1-54 [20] G. J. Taylor, "The Environment at the Lunar Surface", in Lunar Base Agriculturw: Soils for Plant Growth, eds. D. W. Ming and D. L. Henninger [21] S. R. Taylor, "Lunar Science: A Post-Apollo View", Pergamon Press, 1975 [22] D. T. Vaniman and G. H. Heiken, "Getting Lunar Ilmenite: From Soils or Rocks?", Proc. Space-90 Engineering, Construction, and Operations in Space, Albuquerque, New Mexico, 22-26 April 1990, pp.107-116 [23] R. D. Waldron, "Magma Partial Oxidation: A New Method for Oxygen Recovery from Lunar Soil", in Space Manufacturing 7, eds. B. Faughnan and G. Maryniak, AIAA, 1989, p.69 [24] William R. Sharp, John P. H. Steele, Benton C. Clark, and Edward R. Kennedy, "Mining and Excavating Systems for a Lunar Environment", Proc. Space-90 Engineering, Construction, and Operations in Space, Albuquerque, New Mexico, 22-26 April 1990, pp.294-304 [25] Leonhard E. Bernold and Shankar Sundareswaran, "Laboratory Research on Lunar Excavation",Proc. Space-90 Engineering, Construction, and Operations in Space, Albuquerque, New Mexico, 22-26 April 1990, pp.305-314 [26] Egons R. Podnieks and Robert L. Schmidt, "Lunar Mining: Surface or Underground?", Proc. Space-90 Engineering, Construction, and Operations in Space, Albuquerque, New Mexico, 22-26 April 1990, pp.315-324 [27] Alan B. Binder, "LLOX-Metal Production via NaOH Electrolysis", Proc. Space-90 Engineering, Construction, and Operations in Space, Albuquerque, New Mexico, 22-26 April 1990, pp.339-346 [28] Edward McCullough and Carl Mariz, "Lunar Oxygen Production via Magma Electrolysis", Proc. Space-90 Engineering, Construction, and Operations in Space, Albuquerque, New Mexico, 22-26 April 1990, pp.347-356 [29] Michael A. Gibson, Christian W. Knudsen, and Anton Roger III, "Development of the Carbotek Process", Proc. Space-90 Engineering, Construction, and Operations in Space, Albuquerque, New Mexico, 22-26 April 1990, pp.357-367 [30] Robert O. Ness, Jr., Brian D. Runge, and Laura L. Sharp, "Process Design Options for Lunar Oxygen Production",Proc. Space-90 Engineering, Construction, and Operations in Space, Albuquerque, New Mexico, 22-26 April 1990, pp. 368-377 [31] E. B. Jenson and J. N. Linsley, "Waste Heat Rejection System for a Lunar Oxygen Production Plant", Proc. Space-90 Engineering, Construction, and Operations in Space, Albuquerque, New Mexico, 22-26 April 1990, pp.378-388 [32] E. B. Jenson and J. N. Linsley, "Rejection of Waste Heat from Oxygen Liquefaction Operations at a Lunar Oxygen Production Plant", 1989 Space Cryogenics Workshop, Pasadena, CA, July 1989 [33] D. S. McKay, "Current status of Technology to Extract Useful Materials from Lunar and Martian Resources", 13th Annual Technical Symposium, Houston, Texas, May 1988
*** the end ***

Go back to Computer Futures' Space Publications home page