Tankers. A new price category?

[jongoff]

That's certainly true for precious cargo, like aerospace hardware or humans. But for bulk materials it's not inconceivable that mass produced, unreliable and expendable launchers would turn out to be cheaper. Given a market for selling propellant in orbit the market would discover what the most economical approach was.

Why plan to build "unreliable" launchers?  I'm not sure how making a launch vehicle less reliable would save money.  I'm not even sure how to save money by making a launch vehicle less reliable! 

I'm with you Ed.  I used to be in the low-cost expendable camp (having done my undergrad studies in manufacturing engineering), but I'm now firmly in the high-flight-rate reusable camp when it comes to getting costs down.  When you have a market like bulk cargo and propellants, I can't think of a better possible starter market for RLVs.  High divisible cargoes, no mission-specific engineering, no payload insurance needed, payload integration becomes very repetitive, launch licensing becomes a lot easier, the ability to easily get up into the 50ish flights per year needed to get prices down... 

I could very well be wrong--so I'm quite open to seeing how the market goes once cargo/propellant needs like that get better established, but my intuition tells me that propellants and bulk cargoes will be delivered by reusable vehicles, not by "unreliable" mass-produced expendables.

~Jon

[mmeijeri]

I used to be in the low-cost expendable camp (having done my undergrad studies in manufacturing engineering), but I'm now firmly in the high-flight-rate reusable camp when it comes to getting costs down.

So am I, at least in the long run, but Ed asked for an explanation how sacrificing reliability could potentially buy you reduced prices and then went on to ignore the explanation.

  When you have a market like bulk cargo and propellants, I can't think of a better possible starter market for RLVs.  High divisible cargoes, no mission-specific engineering, no payload insurance needed, payload integration becomes very repetitive, launch licensing becomes a lot easier, the ability to easily get up into the 50ish flights per year needed to get prices down... 

That argument actually goes the other way, it shows how bulk cargo would be ideal for RLVs, not how RLVs would be ideal for bulk cargo. After all bulk cargo makes for ideal payloads for mass-produced expendables or HLVs too. It's no coincidence that SDLV proponents want an exploration program. They too know that propellant is the only affordable payload they could ever hope to have.

[mmeijeri]

Because they don't reenter the atmosphere at Mach 23 during each flight.

Of course not, but it does explain the difference in economics.

[baldusi]

I could very well be wrong--so I'm quite open to seeing how the market goes once cargo/propellant needs like that get better established, but my intuition tells me that propellants and bulk cargoes will be delivered by reusable vehicles, not by "unreliable" mass-produced expendables.

The fact is that if you want an RLV not only will your fmr be bad, but you can only go to LEO. Ideally, low LEO. Thus, any reusable infrastructure would need a heavy use of tugs and depots. Orbital mechanics are easy for GSO and EML1/2. May be for any deep space. But LEO/MEO/SSO and such, get killed because nor only for the plane change, but for the nodal change cost. And I'm talking about the tugs and depots. The Van Allen Belt transits don't help, particularly with live cargo and SEPs.
So you sort of always will need some disposable parts in the foreseeable future. But as I stated before, you should worry about how to lower the cost of missions rather than launches. Think about avionics, propulsion modules, power distribution, integration, environmental testing, MMC, etc. Can you lower each of those to 50% of what currently cost? To 25%? To 10%?

[mmeijeri]

But LEO/MEO/SSO and such, get killed because nor only for the plane change, but for the nodal change cost.

We don't really need RLVs for launching large payloads to any of these orbits, EELV class launchers are fine for that. We do really need RLVs for transporting people to LEO. They would be excellent for propellant and supplies too, even if the RLVs are small.

[tegla]

The fact is that if you want an RLV not only will your fmr be bad, but you can only go to LEO. Ideally, low LEO.

What's wrong with RLV doing GTO?

[mmeijeri]

What's wrong with RLV doing GTO?

It would be much harder.

[A_M_Swallow]

The fact is that if you want an RLV not only will your fmr be bad, but you can only go to LEO. Ideally, low LEO.

What's wrong with RLV doing GTO?

RLV have a small payload.  On the same LV GTO payloads are smaller than LEO payloads, so that is a double reduction in size.

[93143]

Skylon's upper stage is optionally recoverable and reusable, and capable of putting over six tonnes into GTO in reusable mode.  Fluyt is reusable and can put Skylon's full 15-tonne payload anywhere in cislunar space - but of course it isn't strictly part of the RLV; it's a chemical tug based in LEO.

[baldusi]

The fact is that if you want an RLV not only will your fmr be bad, but you can only go to LEO. Ideally, low LEO.

What's wrong with RLV doing GTO?

The delta-v for GTO is a lot more than LEO. Given the rocket equation, making an RLV to LEO is very difficult. To GTO might as well be impossible with current technology. And I mean a fully returnable and easily launched vehicle.
Any realistic RLV would take two stages to LEO. The idea of a tug and depot system, is that once you have a stage that goes from LEO to GTO/EML1/TLI/TMI, there's no point on actually returning it to the surface. Just refuel it on LEO and send it up again. So you'd have a three stages to GTO, for example, but the third stage would only to LEO-GTO-LEO, never returning. You'd have to refuel it of course.

But LEO/MEO/SSO and such, get killed because nor only for the plane change, but for the nodal change cost.

We don't really need RLVs for launching large payloads to any of these orbits, EELV class launchers are fine for that. We do really need RLVs for transporting people to LEO. They would be excellent for propellant and supplies too, even if the RLVs are small.

The whole point of the thread was if cheaper launchers can be used for "low value" cargo, like fuel/water/oxygen. From there the discussion went to many people stating that they had moved from the "dumb but cheap" to the "expensive but fully returnable", after having more experience in the industry.
I like to point out that the launch cost is (usually) about 20% of the mission cost. So to really multiply the amount of missions (to justify a RLV), you'd have to lower all the areas.
The proposal underlying to both tourism and fuel depots, is that if you can make some parts of bigger missions standardized you'd save on all the mission phases.
Let's say for example tourism. If you have a fixed set of places to visit, and you repeat the mission every time on the very studied launch opportunities, and you have developed an architecture where your crewed vehicle flies many times without having to refurbish it. In other words, if you don't have to do integration testing again, and your payload doesn't need DDE&T (humans/fuel), and your mission planning is restricted to checking the known launch conditions, then you could lower your cost of that seriously (like to 20% of current costs), if you had enough amount of missions.
I state that even if you could lower your cost to 20% of current prices, the potential market is too small to justify the necessary development and construction of such an infrastructure. And that you couldn't go much lower than that.

[mmeijeri]

I like to point out that the launch cost is (usually) about 20% of the mission cost. So to really multiply the amount of missions (to justify a RLV), you'd have to lower all the areas.

Only if you are going after traditional EELV payloads, not for people (assuming you have reusable spacecraft), and not for propellant and supplies (assuming you have tugs). I see EELV class launchers and small RLVs as highly complementary, RLVs for cost-effectiveness and EELVs for "heavy lift" (20-30mT).

I state that even if you could lower your cost to 20% of current prices, the potential market is too small to justify the necessary development and construction of such an infrastructure. And that you couldn't go much lower than that.

Are you talking about the near to medium term? Because in the long term it seems almost certain we'll have RLVs. But in any event what you are saying merely points out why we don't have RLVs, not that they are physically impossible, or economically infeasible excluding development costs, or undesirable.

[jongoff]

The fact is that if you want an RLV not only will your fmr be bad, but you can only go to LEO. Ideally, low LEO. Thus, any reusable infrastructure would need a heavy use of tugs and depots.

Yes, if you're doing small deliveries to orbit, having a destination to deliver them to is definitely a nice to have...

Orbital mechanics are easy for GSO and EML1/2. May be for any deep space. But LEO/MEO/SSO and such, get killed because nor only for the plane change, but for the nodal change cost. And I'm talking about the tugs and depots. The Van Allen Belt transits don't help, particularly with live cargo and SEPs.

I recently figured out a great way to handle deep space departures from a LEO depot that can handle just about any declination, and pays a very minimal delta-V penalty (though it potentially adds a week or two of mission flight time, and an extra set of VAB transits).  Thanks to an old MXER tether article by Kirk Sorensen of all people (ironic because Kirk hasn't seen the light yet on LEO depots...)

And unless you mean transfer tugs, the D2S work we're doing at Altius is aimed at eliminating the need for prox-ops tugs.

On the SEP side of things...I'm actually with you on being a very big anti-fan of SEPs especially for use within cislunar space.  Maybe for missions starting at EML2, but their long transit times getting out of earth's gravity well combined with their very high unit costs make me think they'll end up on the bad side of the marginal cost equation once the price of prop in LEO drops past a certain point.

~Jon

[jongoff]

The delta-v for GTO is a lot more than LEO. Given the rocket equation, making an RLV to LEO is very difficult. To GTO might as well be impossible with current technology. And I mean a fully returnable and easily launched vehicle.

Unless you refuel the RLV upper stage in LEO.  The delta-V from LEO to GTO is much less than your typical RLV sees on its leg to LEO.  And it has a heat shield that could very well serve as an aerobrake.  It's not clear if it's better to do it that way than to build a dedicated GTO transfer vehicle (expendable or reusable), but when we're talking about using RLVs to deliver propellant per orbit, it's pretty much assumed that we'd be talking about a depot-centric architecture, at which point having some of the RLVs use propellant from the depot to boost a payload from LEO to GTO (or even all the way to GEO) isn't necessarily crazy at all.

Any realistic RLV would take two stages to LEO. The idea of a tug and depot system, is that once you have a stage that goes from LEO to GTO/EML1/TLI/TMI, there's no point on actually returning it to the surface. Just refuel it on LEO and send it up again. So you'd have a three stages to GTO, for example, but the third stage would only to LEO-GTO-LEO, never returning. You'd have to refuel it of course.

That's also a perfectly legit way of doing things.  How the trades turn out in any given case is probably highly dependent on assumptions, but having an in-space only tug stage is a good idea.  Depots are really key for allowing you to reuse in-space elements, which as you've pointed out, are some of the major cost drivers in any transportation architecture.

I like to point out that the launch cost is (usually) about 20% of the mission cost. So to really multiply the amount of missions (to justify a RLV), you'd have to lower all the areas.

The thing is that many of the other cost areas (insurance, in-space hardware fabrication, etc) can be impacted by the availability of cheaper, more frequent launch, and propellant depots.  If you don't have to relaunch your mission hardware every time (just new people, propellant, and provisions), you can start cutting overall mission costs rapidly.

The proposal underlying to both tourism and fuel depots, is that if you can make some parts of bigger missions standardized you'd save on all the mission phases.
Let's say for example tourism. If you have a fixed set of places to visit, and you repeat the mission every time on the very studied launch opportunities, and you have developed an architecture where your crewed vehicle flies many times without having to refurbish it. In other words, if you don't have to do integration testing again, and your payload doesn't need DDE&T (humans/fuel), and your mission planning is restricted to checking the known launch conditions, then you could lower your cost of that seriously (like to 20% of current costs), if you had enough amount of missions.

Quite possibly even further with a reusable system.  I'd be really surprised if a fully-reusable gas-and-go style reusable couldn't get a ticket price down to $1M or less to LEO.

I state that even if you could lower your cost to 20% of current prices, the potential market is too small to justify the necessary development and construction of such an infrastructure. And that you couldn't go much lower than that.

I politely disagree.  I think that many of those other areas of cost you talk about (the other 80%) can also be driven down once you have depots and RLVs.

~Jon

[edkyle99]

Hence, particularly for lower stages, he suggested the use of larger margins and heavier materials to minimize cost without affecting reliability.

So, a rocket with a minimum cost design lower stage would look something like this?



 - Ed Kyle

[baldusi]

I like to point out that the launch cost is (usually) about 20% of the mission cost. So to really multiply the amount of missions (to justify a RLV), you'd have to lower all the areas.

Only if you are going after traditional EELV payloads, not for people (assuming you have reusable spacecraft), and not for propellant and supplies (assuming you have tugs). I see EELV class launchers and small RLVs as highly complementary, RLVs for cost-effectiveness and EELVs for "heavy lift" (20-30mT).

There's a couple of points here. I'd like to separate lots of people to a single point (say, a Med Club at EML1), from fuel. Sending people to a single place in space might get very cheap. But let's analyze the costs. I'm not that experience and there might be cheaper was of doing this.
But as long as you launch rockets, you have to contend with range limitations. I.e. you have to send a small army of boats and crafts to make sure there's nobody in your flight path. You have to use and pay for the telemetry/radar stations/satellites. You have to make constant analysis of trajectories and possible collisions avoidance. You have to check the weather conditions (weather balloons, etc.). All this means lots of money. It's very little when launches are in the hundreds of dollars. But if you want to go towards single digit millions, all this starts to be a significant percentage of costs.
Then you'd have that mythical place in space. You'd need an MCC for that. And constant resupplies. The cost of a passenger to a station is not the crew transfer. Is that, plus the consumables, plus the spare parts, plus the ground controllers, plus the communication costs, etc. So even if sending a single person would cost 1M, it would cost significantly more per person.
But then there's the issue of fuel. Fuel has to be on an useful place. If not, you'd have to move your depot. Since even nodal point change is pretty expensive in delta-v terms, you'd have to launch fuel to move depot, move the depot, second launch to actually fill the depot for the mission, and then the third launch to actually launch the mission. So now you need three launches for one mission. Remember, all the stuff I named before. Those controllers and orbital engineers are not cheap. Nor the range services.
I'm not saying it's not possible to do it. I'm saying that you'd need to have the actual launch costs on the single digits percent of current costs. And that's incredible difficult to achieve. Even a Falcon 9 fuel is about 3% of the total launch cost. So lowering to less than 6% (very few industries have bigger fuel costs than that) of current costs seems plain impossible with current rocket technology.
Then there's a different issue that I would like to touch. Actually it's a sort of contradiction. On the one hand, once of the drivers of launch market is not USD/kg only. I'm sure a 1000tn to GTO launcher would be very cheap on that account. But who will be able to afford such a ticket? On the other hand, if you could launch 1kg for 5.000USD, you could do a lot of very interesting things. But because that is a small ticket. So the market should also be analyzed from the single ticket perspective.
But then there's the other issue of economies of scale above mentioned. I'm sure it's a lot simpler to achieve a low USD/kg for a "big" graft, than a "small" one. So I see that the only way to get low per passenger cost, will be by carrying a lot of people per flight. Ditto for the fuel. So I see the RLV actually being very big.
Do you see the dichotomy? The big market is small total amount of money, and the economies go to big amount of payload in single launch. In that sense I see the nanoracks to ISS as pure genius. But that's part of the problem I see in developing this market.

[charliem]

Economies of scale are a possible way to cut prices down but not the only ones.

Comparisons can help. I propose building and operating big crude oil tanker ships.

It is not uncommon that cargo value surpasses the cost of the ship. The number built every year is small so economies of scale are not possible, and the demand for their services is quite inelastic.

Even so their cost per tonne have been constantly decreasing for decades (partly because they have become bigger, but not only for that, also inside the same tonnage category).

Maybe we could learn one thing or two from shipyards and shipowners.

Listing all the things that have to be payed for (to take something to orbit) gives info about why that's so expensive now, but offers no clue about how hard (or easy) is going to be to cheapen it.

[baldusi]

Economies of scale are a possible way to cut prices down but not the only ones.

Comparisons can help. I propose building and operating big crude oil tanker ships.

It is not uncommon that cargo value surpasses the cost of the ship. The number built every year is small so economies of scale are not possible, and the demand for their services is quite inelastic.

Even so their cost per tonne have been constantly decreasing for decades (partly because they have become bigger, but not only for that, also inside the same tonnage category).

Maybe we could learn one thing or two from shipyards and shipowners.

Listing all the things that have to be payed for (to take something to orbit) gives info about why that's so expensive now, but offers no clue about how hard (or easy) is going to be to cheapen it.

Please stop using the oil market as an example. The amount of oil transported is huge. What you consider "few" tankers is actually a huge number. Since shipyards are sized for building one or even less tankers per year. And the tankers have operation lives of more than 20 years. Thus, cheap financing would lower the cost of transport (through it's effect on both amortization and financial burden).
What most people here seems to be failing to grasp, is that RLV/cheap rockets aren't only possible, but highly studied, and we have current examples. Take the Shuttle, or even the Delta IV. At 24 launches/year both LV would have been very cheap. The fact is that there's not enough payloads to reach those rates. Let me strengthen this point:
The LV cost problem is lack of payload quantity elasticity to price
If you want cheap LV you've gotta have orders of magnitude more payloads. As it's now, you could slash by half the cost of LV, and yet you wouldn't have many more payloads. What's worse, most of national defense markets in the world are much more politics sensitive than price sensitive. Russian National payloads will fly on Russian rockets, same for China and the US and, to an extent, India. The European seem to use a mix of Ariane 5 and Soyuz. But since they will have everything on Kourou, it's almost the same situation.
So even if you could launch at 20% of current costs, you'd only get into to "commercial" market. Which is very price inelastic (even though it's very price sensitive). Price sensitive means that a LV operator has to compete every single launch and very small percentages of discount can win or lose a launch contract. Price inelasticity means that even if the whole launches industry made huge reductions in price, there would be very little extra payloads.
So, my conclusion is, don't work on making LV cheaper, work on making payloads cheaper. Even current LV can have huge cost reductions from more payloads.

[charliem]

Please stop using the oil market as an example. The amount of oil transported is huge.

And why should I do that? That's a difference alright, but I see more similarities, and if similarities outnumber differences I don't see why not comparing.

And I think there's a much more relevant difference than the size of the cargo, that all those sea ships are reusable tens or even hundreds of times.

About similarities I see:

1) Quite few oil tankers are built every year. According to the CIA World Factbook the number of them in service (10.000 long tons and above) is around 4,000. If we accept 25 years as their mean life that implies that some 160 are made every year. The number of LV is in the range of dozens. Not that far.

2) For the biggest tankers the value of the cargo for one trip is higher than the cost of the hardware plus the cost of operation also for one trip. Not dissimilar than a launcher.

3) The cost per unit of big tankers is in the same range than LV, from tens to hundreds of millions.

4) The market for oil transport by sea is inelastic. No decrease in its price is going to make the world consumption increase significantly. About LV we don't know for sure.

5) Insurance amount to a significant part of the costs in both cases.

Take the Shuttle, or even the Delta IV. At 24 launches/year both LV would have been very cheap. ...
The LV cost problem is lack of payload quantity elasticity to price

Well, I'm not so sure. I remember reading that the big computer companies of the 50's said the whole world market was no bigger than a few tens in the foreseeable future. Then in the 60's repeated it, just giving the figure another zero, and even during the 70's, when PCs have already appeared some still affirmed that the thing would not sell more than a few thousands.

You can imagine that if todays computers were as expensive as those from the 50's they could have very well been right.

We can't see the future, and THAT is a well established fact.

As it's now, you could slash by half the cost of LV, and yet you wouldn't have many more payloads.

Well, at least this is something we are going to verify soon enough. SpaceX has more than slash it that half (Atlas V 401, 9.8 mt to LEO for $130-150 M, Falcon 9 block 2, 10.4 mt to LEO for $50-60 M). We'll see if they make the global launch market bigger or not.

So even if you could launch at 20% of current costs, you'd only get into to "commercial" market. Which is very price inelastic (even though it's very price sensitive). ...
So, my conclusion is, don't work on making LV cheaper, work on making payloads cheaper. Even current LV can have huge cost reductions from more payloads.

Well, in my opinion current LV can have their cost reduced from more payloads, but "huge" reductions? I'm not so sure. More than just the frequency of flights have to change to see costs cut by ten or more.

And that's exactly were big oil tankers come into play, they have been able to reduce their prices almost continuously for the last 40 years even having to deal with an inelastic market (this one for sure).

[Xplor]

At the start of the international "commercial" launch era in the 1990's the price of launch dropped by a factor of 2.  The dot com boom promised a huge jump in payloads, not because of the reduced launch price but driven by the internet market demand.  Reality showed that the market was basically inelastic, no growth.  All that happened was launch companies around the world loosing a lot of money.

[oldAtlas_Eguy]

At the start of the international "commercial" launch era in the 1990's the price of launch dropped by a factor of 2.  The dot com boom promised a huge jump in payloads, not because of the reduced launch price but driven by the internet market demand.  Reality showed that the market was basically inelastic, no growth.  All that happened was launch companies around the world loosing a lot of money.

The Internet demand did show up but 20 years later. The big comm birds currently going up are related to internet demand. The NEXT Iridium constellation will have a much improved data capability. But alas what is happening is that the traditional comm birds are just being swapped for more powerful data birds, there isn’t a significant count increase just a operational mode change.