Space Tugs vs. Space Ferries: A Useful Distinction?

Something that’s been bugging me for some time is the confusion surrounding the term “space tug”. The term’s been used to describe at least two very different ideas for many years now. At NGEC-2, I tried to inject a little clarity into my working group’s discussions by drawing the distinction between “tugs” and what I called “ferries”, and I was wondering if others thought it was a useful distinction (and if anyone had a less snicker-drawing nickname then “ferries”–you would think the conference took place somewhere near San Francisco or something from all the chuckles that term drew…)

Under my proposed classification scheme, a “space tug” would be a spacecraft of some sort that primarily is used for maneuvering target spacecraft/objects in the near vicinity of a space station or another spacecraft. For instance, the CSI and CSI/SSL systems proposed for COTS 1 and COTS 1.5 would both fall under this category (and Orbital Express would also likely fit under this category). A “space ferry” on the other hand is a spacecraft that hauls other spacecraft, cargo, or people from one orbit to another in a reusable fashion. For instance, CSI’s or Space Adventures’ respective “Soyuz-Around-the-Moon” concepts would somewhat be examples of a one-use ferry.

Basically tug == prox ops, ferry == large orbit transfers.

Both are very important capabilities, but while they have some overlap in requirements, many of their requirements lead to very divergent capabilities.

Tugs for instance are explicitly designed for proximity operations in mind. A good tug system implementation would likely have one or more robotic arms for better handling, grappling with, and berthing target spacecraft. A tug likely doesn’t have a huge amount of propellant on board. Enough to move things around between various low earth orbits, and to maneuver around the station, but total delta-V capability is probably in the low-hundreds of m/s range. Tugs want to be very robust. The very low delta-V requirements actually make a tug very mass insensitive. So long as most of the things you’re moving around are an order of magnitude or more bigger than you, even doubling the mass of your tug has only a minor effect on the total propellant used for tug operations.

Ferries on the other hand are high-performance spacecraft. The delta-Vs necessary for a useful space ferry are on the order of 4-8km/s (though those last 4km/s are probably going to be “dead heading” ie. flying the ferry back to LEO with no payload attached). In the case of a chemically fueled ferry, this means it looks very similar to an upper stage–mostly take, one or two big engines, and some hardware on both ends. An inflatable aerobrake might not be a bad idea depending on how much it weighs. It might not really need much in the way of prox ops capabilities, just navigation and rendezvous capabilities. Ferries are typically going to be much bigger than their cargoes, while tugs will typically be much smaller.

Both ideas also provide different benefits.

The key benefit of tugs is that they enable launch vehicles and their cargoes to be much simpler. Instead of having to come up with a “last mile” solution for every new passenger or cargo spacecraft, you can have a standardized tug interface, and have the tug do all the hard work. That means that it becomes easier for launch providers to get involved in station resupply, because they’re now just taking a standardized container, launching it to a specific orbit, and holding attitude until the tug can swing by and pick things up. Right now, most crew or cargo deliveries to the station require a system that uses a complicated service module and prox-ops hardware to actually get to the station, which results in fairly poor launched mass to delivered mass ratios. What tugs allow you to do in the cargo case is to drastically reduce the amount of wasted mass required to deliver a given mass of cargo to a station. Instead of having your cargo vehicle be a fully capable spacecraft, all it is now is a pressure shell, with some tug interface attachment (probably something brutally simple involving a couple of “hand holds”), and a passive CBM adapter on the other end. If you’re launching to a station that’s in a resonant orbit that provides frequent “first or second orbit rendezvous” opportunities, you might even be able to dispense with the need for power, communications, or even much in the way of thermal management. In other words, the cargo container starts becoming a lot more like your dumb intermodal container that you see on earth (just much lighter…). Tugs can also serve an emergency role for spacecraft that do have their own prox-ops capabilities, by serving as a backup in case something breaks (or in case multiple docking attempts need to be made and the visiting vehicle runs out of maneuvering propellant). Tugs are also a critical enabler for propellant depots. For propellant deliveries, the propellant can go through relatively narrow tubes (compared to what a human could fit through for instance), which means that a tug could allow for a very simple and lightweight standardized propellant transfer interface to be developed that could just be welded into the delivery tank. This interface could be 100% passive–just some mechanical attachment points, and the quick disconnect ports for fluid and if necessary power. A tug with robotic arms could then take all of the complexity onto itself for the fluid coupling. Much better than trying to make an automated docking and fluid coupling system that has to fly on each and every propellant delivery.

In a nutshell, tugs allow you to take all of the most complicated parts of getting people, propellants, and provisions to a station, and offloads it to either the launch vehicle, or to a reusable vehicle that always stays in orbit, doesn’t have to reenter, etc. Why lug all of that hardware with you each and every time if you can leave it at the destination. Why require each and every company that wants to launch stuff to a station to then also have to come up with their own prox-ops solution? Solve the problem once, and then you don’t have to keep solving it again. If your delivered payloads start outgrowing your tug, the right option might be to build more of them and operate them in a group, instead of designing a newer, bigger model. I think tugboats do just that for very large ships here on earth–instead of building a super jumbo tug, they’ll often just use two or three smaller ones.

Ferries provide very different benefits. First off, and most importantly in my opinion is the fact that ferries (when combined with propellant refueling capabilities) allow you to launch a given exo-LEO vehicle on a much smaller, higher-flight rate vehicle. Dave Salt has on many occasions mentioned that an RLV with an 8000-9000lb payload capability could pretty much service the entire GEO satellite market. Most of the mass required in LEO (I know, many GTO launchers don’t even stop in LEO, but it’s still a useful point) to put a satellite into GEO is not the satellite, or its “beginning of life” propellants–it’s the upper stage, its propellants, and the circularization propellants on the satellite. By having a ferry that operates between LEO and GEO, that has refueling capabilities in LEO, you can launch the largest commercial and government exo-LEO missions without requiring anything bigger than a bottom-of-the-line EELV. In fact, you can even launch manned lunar missions using launchers no bigger than an Atlas V 401 or a Falcon IX (a “Phase One” Atlas V might be a little nicer, but not because of the extra payload to LEO, but because the ICES stages envisioned are scalable and potentially much bigger than a stock Centaur stage, and would thus make a great starting point for a passenger transport ferry). For geostationary satellites, ferries can provide an extra service. Because the ferry can deliver things all the way to GEO, the satellite they’re carrying could possibly forgo its “main propulsion system” and circularization propellant tanks in exchange for more station keeping tanks, more transponders, more solar panels or what have you. Or, you could leave the main propulsion system on, but have the capability to retire the satellite to a different, lower-value GEO slot, where it could spend its last few years before moving itself to a final disposal orbit. For instance, by the time a satellite is nearing 15 years on orbit, it may be a bit obsolete for first-world markets, but maybe it would still be useful for a different GEO slot servicing locations in the third world, or sparsely populated areas in the Pacific for instance (much like how passenger jets in the US are often “retired” only to be refurbished a bit and sold to third world countries at a much lower price). Either of these can help you get more revenue out of a given satellite launch. There are probably plenty of other benefits of ferries that I’m not thinking of right now, but those are just some thoughts.

Ferries can be based around either chemical or solar electric propulsion systems. Some cargoes don’t mind a slow spiral out through the van Allen belts, and thus can be shipped by the more mass efficient (and hopefully therefore more cost efficient) solar-electric “slow boat”. Other cargoes (people, cryogens, and possibly GEO satellites) can be shipped via a much faster chemical ferry. Sure, it’s less mass efficient, so you’re going to be paying for launching a lot more material, but the hardware is relatively cheaper, it can make more flights before being retired, and most importantly, you’re not cooking your payload for several weeks in the van Allen belts. For GEO satellites right now, most of their radiation exposure (for their entire 15 year operation timeframe) happens in just one or two passes through the van Allen belts, so minimizing the time spent there might give chemical ferries a leg up (contra conventional wisdom).

Anyhow, what do you guys think? Does drawing this distinction make sense? And does anyone have a term better than “ferry” for a reusable transfer vehicle? Every time I’ve tried to bring up the idea of a “space ferry” there at the conference, the term would draw smirks or chuckles, or comments along the lines of “I guess NASA Ames is close to San Francisco after all”…

The following two tabs change content below.
Jonathan Goff

Jonathan Goff

President/CEO at Altius Space Machines
Jonathan Goff is a space technologist, inventor, and serial space entrepreneur who created the Selenian Boondocks blog. Jon was a co-founder of Masten Space Systems, and is the founder and CEO of Altius Space Machines, a space robotics startup in Broomfield, CO. His family includes his wife, Tiffany, and five boys: Jarom (deceased), Jonathan, James, Peter, and Andrew. Jon has a BS in Manufacturing Engineering (1999) and an MS in Mechanical Engineering (2007) from Brigham Young University, and served an LDS proselytizing mission in Olongapo, Philippines from 2000-2002.
This entry was posted in Launch Vehicles, Space Development, Space Transportation, Technology. Bookmark the permalink.

25 Responses to Space Tugs vs. Space Ferries: A Useful Distinction?

  1. Rand Simberg says:

    Every time I’ve tried to bring up the idea of a “space ferry” there at the conference, the term would draw smirks or chuckles, or comments along the lines of “I guess NASA Ames is close to San Francisco after all”…

    That’s what you get for using a homophone…

  2. Bob Steinke says:

    How about calling them space freighters.

  3. Jon Goff says:

    I guess freighter (or I’ve also heard lorry suggested) could work. It’s just that these transfer vehicles can also be transferring people, not just bulk cargo or propellants…maybe the term for that could be found in the train world…

    Locomotive anyone? Or maybe just leave the long-range one as a “transfer vehicle” and only use tug to describe the prox ops vehicle?

    Anyhow, just some more thoughts.



  4. Paul Breed says:

    How wide are radiation belts? maybe a combination of Solar electric with some chemical propulsion to get by the radiation.

    the key technical issue iMHO is that you need to keep refueling. And if you need 8Km DV then its more like an SSTO that a normal 2nd/3rd stage.

  5. Jon Goff says:

    How wide are radiation belts? maybe a combination of Solar electric with some chemical propulsion to get by the radiation.

    I can’t remember for sure, but there are several bands, and I think they go most of the way out from LEO to GEO. I’m sure that google can pull up some more info though.

    the key technical issue iMHO is that you need to keep refueling.

    Yeah, both ideas need orbital propellant transfer to make sense. For the tug, this will likely be storable propellant transfer, so we’re talking about a known technology. The ferry though would likely use at least one cryogenic propellant (unless it’s a solar-electric idea).

    As I see it, the progression goes like this: tug->depot->ferry

    And if you need 8Km DV then its more like an SSTO that a normal 2nd/3rd stage.

    Yes and no. It’s actually more complicated than that. If you’re just taking stuff one way (and then “dead-heading” the unloaded ferry back to LEO), it’s really more like 4.2ish km/s. Because most of the mass of a ferry/cargo setup at its destination is in the cargo. Once you drop the cargo off, you now only need a little bit more propellant to get back.

    If you’re to the point where you’re needing to send stuff both ways (the long-term goal, no?) you either want aerobraking, or a propellant depot at the other end (or preferably both).

    In space transportation is a bit different from earth-to-orbit transportation as you have a lot more options and stopping places.


  6. Dave Salt says:

    Hi Jon,

    If you go back a couple of decades (circa 1985) you’ll find NASA was developing your “tug” and calling it the Orbital Maneuvering Vehicle (OMV), while what you term a “ferry” was being developed as the Orbital Transfer Vehicle (OTV) – has some nice pages describing each.

    They were (and still are) logical elements in any reusable LEO infrastructure that uses space stations/platforms/depots as transportation nodes to enable sustainable deep-space missions… which is probably why they don’t feature in ESAS 🙂


  7. Hugh E says:

    “The delta-Vs necessary for a useful space tug are on the order of 4-8km/s …”

    Should read:
    “The delta-Vs necessary for a useful space ferry are on the order of 4-8km/s …”

  8. Jon Goff says:

    Good catch.

    Yeah, I was aware of NASA’s previous interest in the concepts. I just didn’t want to come up with a simple one or two word name that describes the general concept of both approaches, as opposed to using the name of a specific implementation of them.


  9. redneck says:

    If nautical names are prefered, it would seem to me that ferry would be the corrct term for a short haul vehicle. Channel or Staten Island ferries. Clipper or liner might work for the medium distance stuff. GEO clipper making its’ weekly run or each of the lunar liners are doing two cruises a month now.

  10. Charles Miller says:


    Yes, I find it to be a useful distinction.

    Obviously, CSI likes the term “tug”. I think “freighter” is the best word I have heard yet for the “OTV”. Keep trying, you may find something even better.

    I see it as a good sign when we start distinguishing between the different kinds of orbital spacecraft to help us think about on-orbit space infrastructure. NOTE: Eskimos have several dozen different words for snow/ice.

    Onwards and upwards,

    – Charles

  11. Snake Oil Baron says:

    The concept of using a massive craft to tow a tether which sub orbital craft could magnetically grab – pulling them to orbital velocity – has interested me since I first heard of the idea. That would count as a space ferry right?

    It would allow the use of efficient but slow ion engines to gradually add the energy needed to move the tether-carrying mass to a higher orbit which would allow it to rapidly transfer that energy to the (formerly) suborbital craft which would save lots of fuel weight that would otherwise be required by a conventional chemical propulsion rocket (sending up lots of ion fuel on a conventional rocket would require one inefficient journey followed by many efficient journeys by suborbital/tether capture).

    For those of you who are familiar with the concept, please forgive the explanation which likely does not do the concept justice. But the idea seems very creative. In addition to using time and more efficient engines to replace fast but wasteful propulsion, it also makes those suborbital space missions that the private sector is pioneering seem far more valuable than simple the simple tourism they are being developed for.

  12. chris-gerrib says:

    Well, from a Navy / ocean perspective, what you’ve just described is the difference between a harbor tug and an ocean-going tug.

    Any vessel that hooks on to and moves another vessel is a “tug.” Harbor tugs work in a harbor – short range stuff – and oceangoing tugs handle longer-range missions.

  13. Anonymous says:

    Hi Jon,
    This is not on this topic, but I have been thinking about a concept and wanted to write it up and send you an e-mail.
    It doesn’t fit in with any of your posts so far, so I wanted to send it directly to you.
    Anyway, my e-mail is timhc at chorus (dot) net. Drop me a line so I know where to send my great idea, well I thought is was ok.. 🙂

  14. robyn-r says:

    Where would the ESA ATV to be launched next week fall in this spectrum of Tug-Ferry?

    It seems to me a terrible waste of all that marvelous technology that makes up the autonomous service module that is the heart of the vehicle. Has there been a consideration of transfer of control of ATVs after they have completed their delivery and off-loading of supplies and waste to the ISS?

    Since these vehicles already have a docking/refueling port they could be resupplied on-orbit (with cooperative redesign before launch or on-orbit modification) and then perform whatever tasks would be within their capabilities (Debris clearing? Orbital refuel and service of satellites? Resupply missions for commercial LEO space stations?)

  15. Martijn Meijering says:

    > Has there been a consideration of transfer of control of ATVs after they have completed their delivery and off-loading of supplies and waste to the ISS?

    ESA has much very ambitious plans for ATV. Ultimately they want it to become something Europe’s Orion and the heart of European plans for manned spaceflight. See

    With ESA’s small budgets they have to work very incrementally, which is a very good idea anyway. They want to add capability for cargo return, then passenger return and then passenger transport up and down. Another interesting extension under consideration would be to add docking ports at both ends so they could be connected like carriages.

  16. Martijn,
    It could possibly work (using ATV as a proxops tug). I’m not sure if it would be the most ideal for the job (since it’s so big), but I don’t see why something like that couldn’t work. In fact, I think LM has looked at that a bit. Not positive though, but I do recall reading something along those lines.


  17. Martijn Meijering says:

    Hi Jon,

    The ATV consists of two modules: a cargo module (currently a pressurised module, although an unpressurised one is under consideration) and a service module. The service module could be used a tug and that configuration has been compared to Russia’s Parom.

    Another interesting use would be as a crew taxi between the ISS and a man tended free flyer module (or even an Orion!). I think ATV could theoretically be used for that today (provided it has enough fuel), because the cargo hold is pressurised. They’re not going to do that of course, unless as a last resort in some weird emergency situation.

  18. Eric Collins says:

    The ATV could probably not serve as a run-about (crewed OTV) without a more advanced on-board life support system and other ‘facilities’ for supporting crews for days or weeks at a time. Still, it’s an interesting idea. If nothing else, it would be nice to be able to start getting some ‘flight experience’ with a real crewed spacecraft (i.e. one that is meant to travel point-to-point in space).

    It would be interesting to try and figure out what kind of range the ATV would have with a full propellant load. If it can do ISS reboost, then it should be able to make a couple of decent orbital transfer burns.

  19. Martijn Meijering says:

    The service module itself holds a little under 7 tons of propellant (MMH/N2O4) and the cargo module holds another 860 kg as well as 840 kg of water and 100 kg of air. I don’t know if the ATV can pump the propellant in the cargo module back to the service module. If not, it should be possible to design a variant that can.

    I like the idea of using something like the ATV to transfer crew from the ISS to a waiting Orion. In the past Jon has suggested that NASA shouldn’t be in the Earth-LEO transport business at all. I agree, but I don’t see that happening anytime soon.

    What I would like to see is for NASA to switch from Ares to DIRECT, but only developing the J-120, so the money otherwise needed for a whole new upper stage could be spent on more useful things.

    Also, I’d like to see NASA barred from using the J-120 to transport crew or propellant or cargo that can be launched on commercial launchers. Further, I’d like to see them barred from using chemical propulsion for cargo transport to LLO and back and from using expendable craft between LEO and LLO.

    For a moon mission, astronauts would then be launched on a commercial launcher to the ISS, then take something like a crew taxi variant of the ATV, dock with the waiting reusable Orion, itself launched on a commercial launcher and then go on their way to the moon. That part of the trip should be run by NASA, because there exists no commercial alternative. Once back in LEO, the astronauts would use the ATV to transfer back to the ISS and then use a commercial capsule to get back to the surface.

    In this way NASA would more or less be forced to develop sensible heavy permanent infrastructure in cislunar space because that is about the only thing they would be allowed to launch on a J-120. I think this would lead to a much more sensible long term program than what is planned today.

  20. Eric Collins says:

    Just for fun, I went ahead and did some BOTE calculations for the delta-v for an ATV. Assuming a fully loaded mass of about 20000 kg and a dry mass of 12500 kg, with an ISP of 273 s, the resulting delta-v would be about 1257 m/s. Using the ISS orbital elements as a starting point, I figure that one could transfer to a similar orbit in a different plane inclined to up to about 112 degrees from the ISS orbital plane. You’d probably want to be able to bring the ATV back from the rendezvous so you would likely only use about half that much, but that’s still a respectable plane change. Unless I screwed up the calculation, then it sould be possible to use an ATV for and Orbital Transfer Vehicle from the ISS to just about any other useful orbit.

    Of course, this is the first time I’ve used the plane change equations, so someone may want to check my numbers.

  21. Jonathan Goff Jonathan Goff says:

    That equation looks like it’s off. The equations I’ve seen are in the form of DV=2Vsin(Di/2), where Di is the inclination change, and V is the orbital velocity. Which makes sense mathmatetically, because a 180 degree plane change (ie turning your orbit completely around) takes 2x your current velocity (one to stop and one to reaccelerate the other direction.

    Plane changes are supposed to be crazy expensive.


  22. Eric Collins says:

    Yeah, Ok. I usually do a better job of sanity checking my answers. That’s what I get for working out the numbers in the dead of night.

    If that formula is correct, then you’d only be able to get about 10 degrees of plane change for the full 1257 m/s delta-v. That does sound more reasonable, but unfortunately that does not get you very far from the 51 degree inclination of the ISS.

    The ATV service module could still serve as a useful ISS servicing tug if it could be detached from the ICC. Then all they would need to do is add refueling capabilities between the tug and the next ICC, which would be launched without a service module attached. You said that hypergolic transfer was a proven technology, right?

    As I mentioned in the comments on another post, I think the ISS would make a great fuel depot demonstrator if they could start storing excess hydrogen, and oxygen and perhaps start generating some methane too. Of course, the ISS servicing tugs would need to have methane burning engines rather than hypergolics. Surely someone is working on that by now (e.g. LM, Armadillo, …).

  23. Martijn Meijering says:

    A NASA research group on ISRU is looking into converting trash into methane. It is meant for use on the moon, but a field trial on the ISS sounds like a good idea and it could generate positive publicity too.

  24. Pingback: My New Space Hero - Out of the Cradle

  25. Spencer says:

    Isn’t the difference between a tug and a ferry just mating with the higher performance (“Upper Stage equivelent”) engines and fuel tanks? So why couldn’t essentially the same module architecture be extensible for lunar/asteroid ventures by mating with the needed propulsion hardware when needed? If there is any docking planned at the far end (i.e. Moon), the same maneuvering and docking capabilities of the “tug” would be needed for any “ferry”.

Leave a Reply

Your email address will not be published. Required fields are marked *