Table of Contents Page
Executive Summary 2
Introduction 2
Mission 4
Conclusion 15
References 15
Table of Figures
Figure 1 GSV 5
Figure 2 Refueler and Depot 7
Figure 3 Refueling Scenario 8
Figure 4 RSV 14
Table 1 Distribution of active satellites 4
Table 2 GEO Servicing opportunities 4
Table 3 GSV mass breakdown 6
Table 4 Power Budget 9
Table 5 Cost of Refueler/Depot 11

Executive Summary Launching a satellite into space with a mission to service other satellites is a plausible idea. Servicing a satellite means to somehow make it so a satellite’s design life is lengthened. This can be accomplished by refueling, assisting in orbit placement, and/or performing mechanical repairs. Satellite servicing has been done already in the past during manned space missions. The problem arises when the attempt is to be made on satellites outside of human intervention like those in High Earth Orbits (HEO). The ESA and d NASA have both put forth their ideas on how to service satellites in space. The Refueler/Depot idea suggested by NASA has the plan to launch a large satellite into space in two parts. One part is the Refueler, which is agile and movable and can dock with cooperating satellites. The other half is the Depot which stores the hydrazine fuel but remains in a constant orbit above the GEO belt. ESA’s idea is of a single satellite that will do it all: refueling, mechanical manipulation of satellites, and orbit manipulation. This idea may take longer to catch on because of the extra challenge of remotely operating a satellite to fix another satellite.
Introduction
The remoteness of satellites after launch is what makes them susceptible to failures before their designated end of life (EOL). Historically, data shows that there is a 9% failure rate of satellites during operational lifetime.1 Add that to the 4–5% failure rate of launch vehicles and this calculates to almost one out of seven satellites failing before EOL.2
In an effort to limit these failures industry designers are obligated to use what is called legacy hardware on space systems. This means that the hardware has to be proven in the environment of space. NASA is one example how systems can become legacy. They use technology readiness levels (TRL) for measuring the maturity of a system.3 They use a scale of 1–9 with most spacecraft designs needing a TRL of at least 8 to insure “flight qualified” hardware.4 These techniques help to reduce risk, it also negatively effects satellite performance and stalls improvement.5
Another way to limit risk along with using legacy hardware is to utilize on-orbit servicing. There have been a number of on-orbit servicing starting in 1984 (the Solar Maximum Mission Satellite) 6. Probably the most famous on-orbit servicing was the Hubble Space Telescope. Many of the manned space operations have included satellite servicing. These missions have proved that it is possible to service satellites after launch. However, they were all made by human intervention. The benefits of on-orbit services have not been offered to satellites outside of human reach. The question arises as to whether or not it is possible to service satellites without human presence. Research has proven that it is technically possible and much cheaper if done in geostationary orbit (GEO).7
The mission of the On-orbit Servicing Satellite (OSS) will be to relocate and restore a failing satellite. An example of relocation would be taking a dead satellite into a disposal orbit or stationing a satellite into a mission orbit after launch vehicle failure. An example of restoring a satellite would be refueling or proximity operations to discover how a satellite failed. An example of this might have been a rescue of Milstar 3 which failed to reach its operational slot in GEO due to an upper stage failure. This on-orbit servicing might have saved $1.2 billion8.
Refueling operations could be done for the purpose of lifetime extension, performing extra maneuvering of a spacecraft, and station-keeping. In October 1984, the crew of Challenger on STS-41G performed on-orbit refueling by transferring 60 kg of hydrazine between two pallet-mounted tanks9.
Table 1 General distribution of active satellites10
LEO MEO GEO Elliptical Total
Astronomy 53 0 1 17 71
Communications 203 2 318 11 534
Earth observation 87 0 23 2 112
Navigation 9 44 3 0 56
Total 352 46 345 30

Table 1 shows the distribution of satellites in different orbits. The vast majority of active satellites are found in LEO and GEO. This would make the conclusion that the most likely place to put a servicing satellite would be either LEO or GEO. With the plethora of space junk in LEO it would be more reasonable to put the OSS in GEO.
Table 2 Annual number of GEO servicing opportunities11
Service Average annual Average GEO
Opportunities opportunities
Refuel 20.0 8.9
ORU replacement 4.4 2.0
General repair 3.8 1.7
Relocation in GEO 13.0 13.0
Deployment assistance 0.3 0.1

Table 2 shows the number of annual servicing needs that are in all orbits and the number of those servicing opportunities that are in the GEO orbit. These annual needs prove the need for a GEO on-orbit servicing satellite would be well utilized. Another reason for the OSS in a GEO orbit is a relatively inexpensive use of propellant and 24-hour visibility from a single earth station for extended service operations.12
Mission
Gas Station in Space The following includes two possibilities: 1) The Geosynchronous Servicing Vehicle (GSV) is based on ESA research and 2) the Refueler/Depot spacecraft based on NASA research. They both have similar missions and mass proportions. What will be discussed first is the GSV.
The GSV has a dry mass of 1422 kg and a wet mass of 6429kg. This leaves 5007kg of fuel for servicing satellites. The launch vehicle that would be able to launch such a large mass would either be an Ariane-4 or Delta-4 heavy. The satellite would have a lifetime of up to ten years. The mission would include refueling up to 25 satellites, ten inspections, and two dead satellite removals over a period of five to ten years.13
The GSV uses a “conventionally designed bus, an augmented attitude maneuver and transfer capability, a sensor system for rendezvousing, docking, and visual monitoring, and finally tele-operated robotic arms for mechanical manipulation” (for docking purposes).15
The areas in which technology may have to improve before launch are the robotic arm(s), the high resolution camera system, and the sensors and grapple attachments for rendezvous and docking maneuvers.16 Figure 1 ESA’s GSV17

Table 3 GSV Mass Breakdown18
Item Mass (kg)
Structure 20% 285
Power 16% 230
Thermal Control 3% 40
GN&C 9% 80
TT&C/Data Handling 4% 60
Propulsion 5% 348
Manipulator#* 200
Monitoring# 79
Margin 100
GSV dry mass 1422
Propellant & pressurant 5007
GSV Launch mass 6498
*robotic arms
#part of payload package which is 31% together. Table 3 shows the breakdown of mass for the satellite. The figures are rounded up and for the most part coincide with Brown. The structure is a basic satellite based off of Germany’s DFS satellites which are no longer in operation. An S-band antenna is used for TT&C. It is used for relaying State of Health (SOH) data to ground control (at 500bps), and for sending compressed videos at 35 kbs. A ground station antenna of at least 9 meters is estimated for both kinds of data. The same kind of antenna can be used for sending uplinks at 5kbit/second. The power subsystem uses silicon solar cells mounted on three of the six sides of the hexagonal bus. This will provide about 31 watts of dc power to the unregulated (12-18 volt) bus, together with a 15 amp-hour NiCd battery.19
NASA’s research into a gas station in space resulted in one idea of a spacecraft made up of a Refueler and a Depot. The Refueler spacecraft is small and agile, and is designed to carry fuel for five Customer satellites before returning to the Depot for more fuel. The Depot carries enough fuel to service up to 25 satellites. The Depot itself is a very simple satellite equipped with only enough subsystem to stay passively in GEO.20

Figure 2 Refueler and Depot21

Figure 3 shows a picture of what the Refueler and Depot may look like. The Depot has the drums of fuel on the left hand side where the Refueler would dock and collect more fuel before going on more sorties.
The Refueler and Depot are launched together on a Delta IV Heavy. The Delta IV Heavy has a launch capacity of 4,300 kg to 12,980 kg to a Geo Transfer Orbit (GTO) 22. The servicer (Refueler and Deport together) will eventually make it to 100 km above GEO. This is where the Depot will remain. The Refueler will then deliver fuel to up to five satellites. It then returns to the Depot and refuels itself. Since the Depot has no real attitude control, the Refueler will execute whatever maneuvers are needed to maintain the proper orbit for the Depot. Figure 3 shows how a refueling scenario would happen. A mission life of 10 years is assumed.23

1: Launch 24 LV Separation Insertion into GEO +100 Solar Array Deployment
2: Commissioning Systems checkout
3: Depot Deployment Deploy solar sail
Refueler separates from Depot
Depot remains until disposal
4: Refueler Rendezvous Maneuvers to GEO Approach ops
5: Prox Ops Survey satellite Approach and Capture
6: Approach and Capture Burn to intercept Maneuver into capture box and capture satellite
7: Refuel satellite
8: 4 more refueling operations
9: Return to depot after 5 refuelings
10: After final refueling; move to disposal orbit
1: Launch 24 LV Separation Insertion into GEO +100 Solar Array Deployment
2: Commissioning Systems checkout
3: Depot Deployment Deploy solar sail
Refueler separates from Depot
Depot remains until disposal
4: Refueler Rendezvous Maneuvers to GEO Approach ops
5: Prox Ops Survey satellite Approach and Capture
6: Approach and Capture Burn to intercept Maneuver into capture box and capture satellite
7: Refuel satellite
8: 4 more refueling operations
9: Return to depot after 5 refuelings
10: After final refueling; move to disposal orbit
10
10
9
9
8
8
7
7
6
6
5
5
4
4
3
3
2
2
1
1 Figure 3 Refueling scenario
The satellites that the Refueler will service (aka Customer Satellite) are assumed to be 3-axis stabilized. The next assumption is that the satellites will not receive more than 20 kg of Hydrazine. As previously stated it is assumed that the customer satellite’s ground control has control of the satellite. However if it is not than the maximum uncontrolled attitude rates cannot be more than 0.25 degrees per second per axis25. It is imperative that the ground control of the customer satellite have a modicum of control because this will prevent the customer spacecraft’s control system from unintentionally attempting to offset the forces and torques put on it by the Refueler while mated to the satellite. Provisions are included that the Refueler will retreat along the approach axis in case of any anomalies during approach and capture operations. A re-approach can happen when anomalies are fixed.26
Refueling mission operations are conducted from the Refueling Mission Operations Center (RMOC). Autonomous Rendezvous and Capture (AR&C) sequences between the Refueler and the Customer satellites are autonomous but supervised. Times will be scheduled for the ground team to assess mission operations and to provide “go/no-go” commands. Refueling activities will be tele-operated from the ground. To do this communications will be continuous and provide constant telemetry and commanding. Only one contact per day between refueling will be necessary.27 Refueler and Depot System Description

The Refueler carries the AR&C system, two robot arms, two toolboxes, and the refueling system/package. The dry mass for the Refueler is about 1,500 kg (includes 30% contingency). The total mass of the AR&C system is 141.1 kg, its peak power draw is 128.9 W, and its peak data rate is 996 Mbps without compression (786 Mbps with compression).28 Manual Entry Variables | | | | | | Variable | Value | Units | | | | | | | Plp | 129 | w | | | | | | | Margin % | 90% | | Brown table 2.11 page 35 | | | | | Efficiency | 29% | | | | | | | | Mission life | 10 | years | | | | | | | Calculations | | | | | | | | | | | | | | Total Pwr | 314 | w | | | | | | | Bus Pwr | 185 | w | | | | | | | TABLE | | Solar Array | | | | | | | | | min th area | 0.23 | m2 | | Subsystem | % | Value | Unit | | Solar pn area | 0.80 | m2 | | Thermal | 48% | 89 | w | | spin sat area (EOL) | 2.52 | m2 | Brown pg 347 | ACS | 19% | 35 | w | | degradation% | 33% | | | Power | 5% | 9 | w | | | | | | CDS | 13% | 24 | w | | | | | | Comm | 15% | 28 | w | | | | | | Propulsion | 0% | 0 | w | | | | | | Mechanisms | 0% | 0 | w | | | | | | Bus Pwr | 100% | 185 | w | | | | | |
Table 4 Power Budget39
After initial separation from the Depot, the Refueler will move to an orbit from GEO + 100 km to GEO + 127 km. The Refueler will drift to the Customer satellites below at GEO. Another assumption is that Customer satellite orbit planes are within ±1° of the equatorial plane. The next step is that once the Refueler is within 300 km of a Customer satellite, the Refueler will lower its orbit to GEO – 30 km. AR&C sequence will then begin. The ΔV budget for the Refueler is 20 m/s for each AR&C sequence.29
The equation for a bipropellant with a specific impulse of 300 s for the GSV, the rocket equation would be:
∆Vtotal = g (Isp) ln Mp+Mf/Mf = 9.81(300) ln (5007+1200/1200) = 4800 m/s14

After refueling has been completed the Refueler will release the Customer satellite and then transfers back to a GEO + 127 km orbit and drift to the next Customer satellite. This cycle repeats until five Customers have been supplied with 20 kg of Hydrazine. The Refueler then begins its rendezvous with the Depot by raising its orbit to GEO + 70 km, drifts to within 300 km of the Depot, and executes the same AR&C sequence as with a Customer satellite with the Depot. After attaching to the Depot and taking on propellant for Customer Satellites, the Refueler will perform any necessary maneuvers to correct the Depot’s orbit. It is estimated that the Depot would have drifted by approximately 1° under the influence of natural perturbations. The Refueler will top off its own fuel tanks and repeat all previous steps. As soon as the Refueler and Depot only have enough fuel for disposal, the Refueler will boost the Refueler/ Depot stack to a disposal orbit.30
The two robotic arms on the Refueler are 2 meters long. They are connected to the front/top of the bus on a mounting surface. On one side of the mounting platform is the arms and the other has the electronics. The mounting pallet is hard mounted to the Refueler structure. The arms will make autonomous capture but will be teleoperated during refueling tasks. Command and control of the robot arm is through the CSI hard lines and then through the Command & Data Handling (C&DH) subsystem on the Refueler.31
The Refueler bus and subsystems are broken down thus: 1) structure is composite truss design and made from composite/aluminum honeycomb decks. 2) Thermal subsystem includes Multilayer Insulation (MLI), heaters, thermistors, thermostats, variable conductance heat pipes, and radiators designed for full sun load. 3) Communication subsystem includes S-band Omni, and X-band HGA, capable of 10 Mbps downlink. 4) Power subsystem is made up of bus mounted solar arrays with a total area of 7.2 m2 and uses Tj GaAS cells. There are two 100 Ah Li-Ion JSB batteries. The power system is sized for two 72-minute maximum eclipse seasons twice per year. 5) The ACS is thruster-based and its components include star trackers, IRU, CSS, and GPS. 32
The Depot is designed to be a passive spacecraft that can exist at GEO with minimal support. The dry mass for the Depot is 1,326 kg (includes 30% contingency). The bus is made up the same way as the Refueler with a composite truss design with composite/aluminum honeycomb decks. The thermal system includes MLI, heaters, thermistors and thermostats. The Depot has no communication subsystem and no avionics. The power subsystem is made up of a solar array about 1m2 and has no battery. The ACS system maintains the Depot in a sun-pointing attitude by using the solar sail and a libration damper. The Depot does not have any thrusters. Depot telemetry is read by the Refueler when it is attached to the Depot. The Depot has retroflectors on its surface to aid the Refueler in docking.33
Budget
Table 5 Geo Refueling Mission Current Best ($M)34
AR&C $60
Robotic Arms $240
Payload $80
Bus (Refueler & Depot) $220
LV $300
Operations $80
Miscellaneous $200
TOTAL $1,180

Ground System The ground system will utilize the Air Force Satellite Control Network (AFSCN) for downlink purposes. The AFSCN operational frequency spectrum is 1) L band, 1760-1842 MHz, uplink and 2) S band, 2025-2110 and 2200-2300 MHz, downlink.35 The Gas Station in space will use S band frequencies for downlink to the AFSCN and the RMOC.
The AFSCN communications design is “Node centric”. Node centric means all users access the AFSCN by connecting to one of the two geographically separated nodes located at Schriever AFB, Colorado and Onizuka AFS, California. The RMOC will be connected to both to provide redundancy.36
The AFSCN is composed of three segments. The first segment is the Range Segment which has 15 Remote Tracking Sites (RTS). These sites are what provide the ground to spacecraft communication. The second segment is the Communication Segment. The communication segment connects the RTS to the two Operational Control Nodes (OCN). The third segment is the Network Management Segment. This segment provides network scheduling and network status. The user must provide the command and control system (which is the RMOC), and have a scheduling and communications control interface to the AFSCN. The AFSCN uses 3-axis 13 meter antennas.37 The RMOC will be located at the Goddard Space Flight Center.
AFSCN antennas are located at Kaena Point, HI, Vandenberg AFB, CA, Onizuka AFB, CA, Schriever AFB, CO, Kirtland AFB, NM, Cape Canaveral AFB, FL, New Boston AFG, NH, Thule AB, Greenland, Oakhanger, England, Diego Garcia Island, and Andersen AFB, Guam.38 This worldwide coverage is ideal for the mission needs of a gas station in space. Product Specs40
The docking unit that would be used on the Refueler or GSV is the Rendezvous and Docking Sensor (RVS) developed by Jena-Optronik for the European Space Agency and the Japanese Space Agency. The capabilities are as followed:
• Laser Range Finder combined with a galvanometric scanning system
• High accuracy range and position measurement from several km to docking
• Automatic target acquisition, identification, and tracking
• 3D Imaging capability
Its first time in space was on the “Jules Verne”. The RSV performed the docking sequence to the ISS successfully. This was the first time docking was done autonomously. The RVS performed the docking completely automated.40
Starting around 3 meters the RVS is able to measure the distance and approaching direction of the servicing satellite to the Customer satellite. This is done by the RVS sending laser beams as short light pulses, a reflector system in the RVS channels these light pulses to the Customer satellite.40
Dimensions [mm]
Optical Head 270 x 278 x 196
Electronic Box 315 x 224 x 176
Mass [g]
Optical Head < 6100
E-Box < 7700
Temperature Range [°C]
Operational -35...+65
Non-operational -55...+70
Measurement Accuracy
LOS noise ± 0.1° [3σ] [maximal] Azimuth ± 0.01° [3σ] [typical]Elevation ± 0.02° [3σ] [typical]
LOS bias ± 0.1°
Range noise ± 0.1 m [3σ] [long range] ± 0.01 m [3σ] [short range]
Range bias ± 0.5 m [long range] ± 0.01 m [short range]
Power Consumption [W]
<35 nominal <70 maximal
Field of View
40° x 40°

Figure 4 RSV40

Conclusion
Launching a servicing satellite into space is a real possibility. The advantage of the GSV is it is lighter and less expensive. The advantage of the Refueler//Depot is future applications could include refueling the Depot by just launching a new Depot. This would negate some of the expenses of launching a whole new Refueler/Depot combination.
1 (Long, Richards, & Hastings, 2007, p. 964)
2 (Long, Richards, & Hastings, 2007, p. 964)
3 (Long, Richards, & Hastings, 2007, p. 964)
4 (Long, Richards, & Hastings, 2007, p. 964)
5 (Long, Richards, & Hastings, 2007, p. 964)
6 (Yasaka & Ashford, 1996, p. 9)
7 (Yasaka & Ashford, 1996, p. 9)
8 (Long, Richards, & Hastings, 2007, p. 965)
9 (Long, Richards, & Hastings, 2007, p. 964)
10 (Long, Richards, & Hastings, 2007, p. 967)
11 (Long, Richards, & Hastings, 2007, p. 967)
12 (Yasaka & Ashford, 1996, p. 9)
13 (Yasaka & Ashford, 1996, p. 12)
14 (Long, Richards, & Hastings, 2007, p. 969)
15 (Long, Richards, & Hastings, 2007, p. 964)
16 (Yasaka & Ashford, 1996, p. 11)
17 (Yasaka & Ashford, 1996, p. 13)
18 (Yasaka & Ashford, 1996, p. 13)
19 (Yasaka & Ashford, 1996, p. 15)
20 (On-orbit Satellite Servicing Study, 2010, p. 157)
21 (On-orbit Satellite Servicing Study, 2010, p. 161)
22 (Defense, Space & Security, 1995-2012)
23 (On-orbit Satellite Servicing Study, 2010, p. 158)
24 (On-orbit Satellite Servicing Study, 2010, p. 158)
25 (On-orbit Satellite Servicing Study, 2010, p. 158)
26 (On-orbit Satellite Servicing Study, 2010, p. 159)
27 (On-orbit Satellite Servicing Study, 2010, p. 159)
28 (On-orbit Satellite Servicing Study, 2010, p. 143)
29 (On-orbit Satellite Servicing Study, 2010, p. 159)
30 (On-orbit Satellite Servicing Study, 2010, p. 155)
31 (On-orbit Satellite Servicing Study, 2010, p. 160)
32 (On-orbit Satellite Servicing Study, 2010, p. 160)
33 (On-orbit Satellite Servicing Study, 2010, p. 161)
34 (On-orbit Satellite Servicing Study, 2010, p. 162)
35 (Hodges & Woll, 2008, p. 2)
36 (Hodges & Woll, 2008, p. 3)
37 (Hodges & Woll, 2008, p. 3)
38 (Hodges & Woll, 2008, p. 2)
39 Derived from Power Budget spread sheet
40 (Rendezvous- and Docking Sensor RVS)

Works Cited
Defense, Space & Security. (1995-2012). Retrieved from Boeing: http://www.boeing.com/defense-space/space/delta/delta4/delta4.htm
On-orbit Satellite Servicing Study. (2010, October). Retrieved from Nasa.gov: http://ssco.gsfc.nasa.gov/images/NASA_Satellite%20Servicing_Project_Report_0511.pdf
Hodges, L., & Woll, R. (2008). Air Force Satellite Control Network Support for Operational Responsive Space. AIAA/6th Responsive Space Conference (pp. 1-5). Los Angeles: AIAA.
Long, A. M., Richards, M. G., & Hastings, D. E. (2007). On-Orbit Servicing: A New Value Propostion for Satellite Design and Operation. Journal of Spacecraft and Rockets, 964-976.
Rendezvous- and Docking Sensor RVS. (n.d.). Retrieved from Jenoptik: https://webster-vista.blackboard.com/webct/RelativeResourceManager/sfsid/2425859594021
Yasaka, T., & Ashford, E. W. (1996). GSV: An Approach Toward Space System Servicing. Earth Space Review, 9-17.