jump to navigation

Lesson 13 Presentation Material 9-23-10 September 20, 2010

Posted by drspaceshow in Uncategorized.

Lesson 13 Presentation Material 9-23-10

Propellant Depot Discussion

Please note that the copyright for each .pdf document posted here belongs to the author.  To use or reference  this material, you must obtain permission from either Dallas Bienhoff or Dan Adamo. Thank you for respecting this copyright requirement.


1.  DRAv5.0commentaryR2   This .pdf document is by Dan Adamo and titled Comments on Human Exploration of Mars Design Reference Architecture 5.0.

2. CircEPOdepartR0  This .pdf document is by Dan Adamo and titled Interplanetary  Departure From Circular Earth Parking Orbits (EPQs).

3.  100527_Cislunar_ISDC10_R2_NoVideo  This .pdf document is by Dallas Bienhoff and titled Propellant Depots and a Reusable Cislunar Transportation Architecture.

4.  070526_NSS_ISDC_Depot_Insfrastructure_Bienhoff  This .pdf document is by Dallas Bienhoff and titled A Propellant Depot System Concept for Outgoing Exploration.

5.  090414_Depot_for_Constellation_Update_Bienhoff  This .pdf document is by Dallas Bienhoff and titled LEO Propellant Depot: A Viable Opportunity?

6.  070720_Beyond_COTS_Panel_R1_Bienhoff  This .pdf document is by Dallas Bienhoff and titled LEO Propellant Depot:  A Future Opportunity?



1. Trent Waddington - September 21, 2010

Dan has said a number of times that you can’t place a propellant depot into a particular inclination (and RAAN) as orbital perturbation will cause it to progress. Dallas seems to disagree, often referring to LEO depots in particular orbits as if they were fixed.

The ISS seems to stay in a particular inclination by station keeping, begging the question: how much delta-v is required to negate orbital perturbation? I’ve read the claim that LH2 boiloff (after being used to cool the LOX so it has zero boiloff) would produce enough for cold gas thrusters to station keep a depot in LEO, but I’ve never seen the math. Dallas, is there a reference somewhere you could recommend? Has Dan seen it?


Trent Waddington - September 21, 2010

I guess that should be “prompting the question”.. gotta love it when I aggravate my own pet peeves.

Dallas Bienhoff - September 21, 2010


Dan is correct about orbit precession, but to clarify, the inclination with the equator doesn’t change but the orientation (right ascension of ascending node) does. The difference between the rate of precession, determined by orbit inclination and altitude, and the Moon’s orbit period is what determines how often the minimum delta V opportunity occurs for missions from the depot orbit to the Moon, or EML1. Some adjustment in departure time may be possible by increasing or decreasing the departing vehicle’s orbit altitude after loading up at the depot.

As for orbit maintenance, just like ISS, a depot, and any other spacecraft in LEO must expend propellant to maintain its altitude or it will eventually deorbit. It happened to Skylab; it can happen to ISS; Hubble too (just not as fast due to its higher altitude). How much delta V it will take to maintain a depot in orbit will depend upon its orbit altitude, configuration, and average propellant load. Ideally, the propellant temperatures will be maintained using efficient insulation and cryocoolers (refrigeration) so there is no boil-off. If there is boil-off, then, yes, it should be used (captured and stored in accumulator tanks) for attitude control and/or orbit maintenance. Gaseous oxygen and hydrogen should be used for attitude control and, due to the higher thrust likely required, liquid oxygen and hydrogen, may be used for orbit altitude maintenance. Any propellant used for attitude control and orbit maintenance or lost will be an operating expense factored into on-orbit propellant sales price.

I hope that helps.

2. Trent Waddington - September 21, 2010

Dan, it clarifies it for me. I guess the only question is whether Dan agrees 🙂

3. Dan Adamo - September 22, 2010

The drift in right ascension of the ascending node (RAAN) is something you don’t want to fight. In LEO, it’s a perennial acceleration on the order of 1 cm/s^2 and would require a compensating Dv of nearly 1 km/s/day. As Dallas points out, maintaining LEO altitude is a more manageable proposition. For ISS, currently near 350 km orbit height, the cost is approximately 1 m/s/month.

What provides a depot’s altitude maintenance capability is certainly open to consideration, but I find any claim at zero boil-off capability for future propellant storage systems to strain credulity in the context of thermodynamics, particularly when LH2 is involved. So my suspicion is that at least some of the propulsion to maintain a depot’s LEO altitude would come from cryogenic boil-off.

Now, another reason you wouldn’t want to fight RAAN drift in LEO would arise when supporting trans-lunar injection (TLI) opportunities. Because this drift is inertially westward, while the Moon’s geocentric orbit motion is inertially eastward, TLI opportunities arise more frequently with the RAAN drift than without it. In the context of other destinations such as near-Earth objects (NEOs) or Mars, however, RAAN drift in LEO may seriously curtail a launch season compared to departing from Earth’s surface or near the Sun-Earth L2 point. In the case of Mars, such seasons arise every 26 months or so, but a given NEO’s season, particularly for short-duration human roundtrips, may arise only once every few decades.

Finally, believe it or not, there is a cost to maintaining a *fixed* inclination in LEO. Every few years, ISS intentionally includes a small out-of-plane component in one of its altitude maintenance reboosts to also increase inclination slightly. If you think of the north-south component of ISS orbit motion as a simple harmonic oscillator, atmospheric drag acts as a damping acceleration on it that slowly reduces inclination over time if left unchecked by propulsion. Space Shuttle launches can cope with this effect (it only increases ascent performance), but Soyuz/Progress launches can’t because their avionics systems only steer to a fixed launch azimuth equivalent to 51.66 deg inclination. As ISS strays from this optimum value, Russian rendezvous propellant costs increase to compensate for wayward launch steering with respect to the real ISS orbit plane.

Trent Waddington - September 22, 2010

So what’s you’re saying is that it *is* reasonable to talk about a propellant depot in a particular LEO altitude and inclination. It may take some station keeping to stay there but that’s fine. The RAAN processes and that’s fine too.

So how does one calculate the launch season and the frequency of launch opportunities to a particular destination from a particular fixed altitude+inclination?

For example, say I have my depot in LEO at 400km and 28.5 degrees inclination. How many days of the year can I optimally depart from the depot to get to the Moon?

If I have an extra 1km/s of delta-v in margin does that change anything? Could I fly sub-optimal multi-impulse trajectories to get to the Moon on the days that are not optimal?

4. Tom Hill - September 22, 2010

If I’ve interpreted Dan’s and my interactions correctly, a propellant depot will have some (3-5?) near-optimal lunar departure opportunities per month. These would take place when the orbital plane of the depot aligns with the moon at desired arrival time. I’ve understood his concern to be that for on the order of 50% of these opportunities, any specific spot on the moon will be dark, and very rarely will lighting conditions be ideal for a landing or surface operations.

Given an L1 staging area (or additional depot, per Dallas’ presentations), most of these concerns translate into ‘simply’ more loiter time at E/M L1.

Our discussions on NEO missions are kind of stalled at this point. My analysis of non-optimal inclinations for a depot showed that a ten degree out-of-plane maneuver required a ten percent increase in delta-v, and that percentage decreased as the departure delta-v got larger. Dan told me I was missing another point, and I wanted to analyze that, hoping that he’d get a chance to get me the orbital elements (Earth-centered pre-departure burn, Earth-centered post-departure burn, and solar orbit to the asteroid) used in his animation:

I’m at the point where I acknowledge that there are some operational limitations to the use of depots (which hadn’t occurred to me before talking with Dan), and there are some technologies that need to be demonstrated, but still think that the potential to increase launch rate and improve access makes it something we need to look into more.

5. Dan Adamo - September 22, 2010

Yes, it certainly is reasonable to *talk* about propellant depots in LEO. That’s what Thursday’s seance on The Space Show is all about (I guess it’ll be Friday at some locations).

It’s quite another matter to run the numbers, and that’s what answering your good questions is all about. As a proponent of depots in LEO with resources of The Boeing Company at his disposal, we might hope Dallas Bienhoff already has these answers.

If you want to know *how* to run these numbers yourself, that may be beyond the scope of The Space Show Classroom. I’m certainly available for hire as an instructor at the right hourly rate. 😉

6. Tom Hill - September 22, 2010

I work for an aerospace contractor and have done orbital analysis/flight dynamics since college. I have orbital propagation software at my disposal, which I’d like to use and look at the orbits you used for your NEO demo so we’re talking from a common point. I don’t require training, but I’d like to see the element sets.

Trent Waddington - September 22, 2010

Ya know Dan, I’m the kind of guy who takes people up on offers like that 🙂

7. Dan Adamo - September 23, 2010

Tom, I’ve explained this to you one-on-one already, but I need to set the public record straight since you insist on reiterating your request in blogs. The questions you’re trying to answer involve LEO planar dynamics and interplanetary Earth departure direction variations over a 1999 AO10 launch season spanning at least several weeks in the September 2025 timeframe. These questions cannot be answered by disclosing data from the single “point design” trajectory captured in the 1999 AO10 YouTube video you’re citing.

If you indeed possess the experience and design tools you claim to have at your disposal, I invite you to generate the data you require on your own time. My current priorities and work backlog simply don’t permit me to do this for you.

Tom Hill - September 23, 2010

Most of this particular exchange took place on the Lesson 3 blog archive:


As I stated in our one-on-one communication, my orbit analysis software is Earth-centric and does not calculate interplanetary trajectories/ seasons. When you said that you didn’t have time to send me the full data, I asked for the elsets as something easy to do which would let me try and work with the data here, improving my understanding.

As I also stated before, we agree in more areas than we disagree.

Public records set straight. I will not ask you for the elset information again.

Trent Waddington - September 23, 2010

Tom, I can provide you with an Earth-centric dataset for any asteroid you want. Have a look at http://quantumg.net/asteroids/1999_AO10.jpd the data array is J2000 inertial frame positions relative to Earth, one a day, around midnight.. I can also provide velocity vectors if you need.

This is the input data to my Earth-centric asteroid trajectory visualization software: http://quantumg.net/asteroids/skymap.html?obj=1999_AO10

And, of course, I can tell you the recipe I use to generate this stuff if you’re interested.

8. Bill White - September 23, 2010

While listening to the September 7th edition of the Space Show, I found myself in agreement with Dan Adamo’s suggestion that fuel depots should be deployed as close to the final destination as possible, a proposition that supports the lunar surface rendezvous model.

However, what if we do not know what landing sites on the lunar surface are the most interesting and the most valuable?

It seems to me that an EML-1 fuel depot could be used as a central network node to facilitate the deployment of multiple lunar surface depots. Any thoughts on that?

9. Bill White - September 23, 2010

A second question:

How might single impulse ballistic trajectories that can require as much as 100 days to travel from LEO to EML-1 (or LLO) but which substantially decrease the delta V needed accomplish the transit affect the leverage offered by various beyond LEO depot ideas?

See, for example, this AAS paper by Jeffrey Parker:

Click to access AAS%2006-132.pdf

Trent Waddington - September 23, 2010

I think what makes this relevant to this discussion is the argument over which propellant to use. Dallas seems to be primarily interested in LH2/LOX for which I would imagine 100 day transfers wouldn’t be particularly beneficial. Bill has suggested that exploiting these ballistic trajectories would allow one to stockpile a storable propellant. Combined with lunar LOX the storable could just be RP-1, as used by legacy hardware.

Dallas Bienhoff - September 23, 2010


You are correct in that EML1 is a good gateway to any point on the Moon for approximately the same dV as well as a good target location from the Moon that allows departure at any time for the same dV. Going there from the surface does not require waiting for orbit alignment with the surface departure point.

The ballistic trajectories you mention are good for non-human cargos, including propellants. Cryogenic propellant tankers on these tranjectories would need the same thermal management capabilities as a depot. Storables could be sent, such as kerosene, or water could be sent and then split into oxygen and hydrogen once at the depot. The depot would then need to include an electrolysis and liquefaction capability.

10. Trent Waddington - September 23, 2010

What a great show, thanks to Dan and Dallas, John and Jim and, of course, Dave.

One of the topics that wasn’t discussed, but Dallas touched on briefly, was the hardware to do rendezvous with the depot. If I heard correctly, Dallas suggested that something like only 40% of the mass of a Falcon 9 launched payload would be propellant, but that the delivery vehicle would be reusable? That kinda sounds like a Dragon, which is a much more capable vehicle than is likely to be needed, and operating in the reentry/recovery mode of Dragon sounds pretty inefficient too.

Tom Hill - September 24, 2010

Thanks for the offer on the data, Trent. I’ve only worked with my software in elsets so far, but will let you know of any results I come up with.

Enjoyed the start of the show, but needed to get to bed for early work. Will listen to the rest when it’s posted.

11. John Hunt (in Atlanta) - September 24, 2010

It seems that this discussion got underway well before the Sept. 23 Space Show. The following is in response to the classroom show.

The Dan Adamo/Dallas Bienhoff discussion (debate?) on issues relating to orbital propellant depots is loaded with valuable information that should be carefully analyzed by people interested in this important topic. The two experts are both highly knowledgeable about their topics. Dan’s specialty is flight dynamics and current real world operational issues. Dallas brings the expertise on the depot technology itself. This leads to divergent views on the type of depot that is preferred.

There is general agreement that depots main applicability in the next generation of spaceflight is in there applicability to lunar missions. The greatest disagreement is that Dallas favors orbit LEO depots and Dan prefers destination depots prepositioned on the Moon or other final mission objective. The leads Dan to see the requirement for 150 ton class HLVs for human lunar missions and the use of two of these large HLVs to accomplish a lunar mission. Dallas came prepared with studies that show the advantages depots can add to a single Ares V class mission utilizing LEO refueling prior to TLI.

In a sense I think the debate aspect was basically a draw. Since Dallas’s studies were conducted under restrictions to the architecture of the Constellation Program, I think this left his position weakened somewhat. This is because if we actually had an Ares V, Dan has a simpler approach from an operational view point. However, it is unlikely that we will have a very large HLV given the cancellation of the program of record. While we wait a final resolution from Congress, it seems unlikely that Ares V will be developed.

I have in connection with Dan’s last show (9/7/10) posted http://thespaceshowoutsidethebox.blogspot.com/2010/09/dan-adamo-on-propellant-depots-tuesday.html#comments in conception of an alternative that utilize a smaller launcher in the 50 ton class (should we call it a medium lift vehicle (MLV)?). My vehicle in addition incorporates as much reusability as prudent. My current concept is to make the first-stage reusable and the second-stage refuelable. There are some major options here including just recovering engine modules and not whole stages, etc. The point is that this second-stage refueled by a LEO depot has at greater TLI potential than the Ares V would have had. At the same time my proposed 50 ton class launch gives an economy of scale and there significant reusability (at least component-wise) to provide much lower cost per kilogram to LEO to establish and maintain depots than using EELVs and Falcon 9s

One other advantage that I see in the 50-ton class vehicle it that it is sized to be a general workhorse to LEO for a wide variety of missions. The large HLV would not only be intrinsically more expensive but would be produced in much lower quantities resulting in very high costs since it would be use only for very high end missions, i.e. Moon, large object to LEO, etc.

12. Nels Anderson - January 24, 2011

I recently listened to the archived show, which I found very interesting. On the off chance that someone is still reading this thread, let me ask the following question. During the show, Dallas mentioned, if I recall correctly, that an LEO depot would allow 12 (or maybe it was 6) launch windows to the moon per year. Dan mentioned that, absent lighting constraints, there should be about two windows per month, or actually somewhat more on average, thanks to nodal regression. Did anybody ever get to the bottom of why the ULA study was showing relatively infrequent windows? Two-plus per month on average makes sense to me.

drspaceshow - January 24, 2011

Nels, The Classroom series of programs has started up again with the introductory program on Jan. 23, 2011. We hope to have another in-depth depot program in the very near future. Stay tuned to The Space Show for Classroom news and scheduling.


13. hiellebrini - January 25, 2011

Very good post! Your thoughtfulness knows no bounds.

14. doterra essential oils - October 25, 2011

Might be your best article I have read…

Sorry comments are closed for this entry

%d bloggers like this: