The Reaction Engines Skylon Master Thread (1)

[bearshrimp]

Given that there were 23 total launches of commercial LEO payloads in 2010 (per the FAA’s 2010 Commercial Space Transportation Review), and that projected growth is 8%
....
However, if we assume SKYLON came on line in 2022 (10 year gestation period) the demand for LEO payloads would only justify 40 launches in the 1st year of operation given an 8% growth rate.  Over a 10 year period, given a growth rate of 8%, a total of 554 payloads would be available for transport to LEO. 

Two thirds of the commercial launches were to GTO not LEO in 2010 which you've failed to take account of in your extrapolation. Skylon addresses GEO either through the Skylon Upper Stage or the Fluyt OTV. Every Fluyt mission takes 5.2 Skylon launches, I believe.  So if half the GEO launches are large enough to require a Fluyt then that's 1331 Skylon launches over 10 years, or an average of 133 launches a year by your numbers or nearly twice the 70 flights a year REL use as their baseline.

Yes, I know that the majority of missions are to GTO rather than LEO, but I haven’t seen any cost projections regarding the Skylon Upper Stage or the Fluyt OTV so I did not include them in my analysis.  The ESA had the following statement on page 17 of the Assessment:

“One point to be made is that ESA considers that the SKYLON Upper Stage (SUS) which is potentially required for GTO missions may need to be included in the overall development costs. This is because if telecoms spacecraft customers have to pay to develop a GTO stage on top of the launch price then this may push the cost to orbit to a point where the SKYLON becomes less competitive. ESA recommends that the development cost model of the vehicle be re-assessed to account for the additional cost of developing the SUS.”

So, it’s difficult to estimate the cost to GTO provided by Skylon when compared to other launch systems. 

In a way my earlier estimate was probably optimistic in that only 33 payloads delivered to any orbit in 2010 were commercial.  So if we assume that SKYLON will be responsible for delivering all commercial payloads to orbit, and the market expands 8% per year.  The total number of commercial launches to any orbit would be 73 in 2022 rising to 150 in 2032.  The total number of commercial launches would be 1,176.  Assuming each SKYLON would be able to deliver 200 payloads over its 10 year service life that means that there would be a market able to support 6 SKYLON spaceplanes.

The issue is still that in the economic analysis the development costs are amortized over 30 SKYLON craft.  I just don’t see that kind of market demand.  In a way it’s a catch-22.  To bring down the operation and manufacturing cost on a per launch basis each SKYLON needs to be capable of performing a large number of missions.  On the other hand, the ability to have each SKYLON perform a large number of missions will reduce the number required to satisfy the market.  This in tern leads to a smaller number of SKYLONs constructed and increases the amortized development cost on a “per vehicle” basis.

To justify a production run of 30 SKYLONs between now and 2032 the demand for commercial launches would need to increase an average of 17.7% per year.  I think we would only see double digit growth in the commercial space sector if there were considerably more money to be made in orbit.  Perhaps if the Japanese really do develop their 1GW orbital power station (requiring between 300-400 orbital trips) Skylon would find its market, but that’s a big if.

[baldusi]

The current Skylon baseline is the D4, which has 20tonnes to LEO. they specifically increased it's size because they needed to address the 4.6mx14m payload envelope and needed the 20 tonnes to have the necessary GTO performance. Not unlike the Shuttle's IUS.

[Robotbeat]

The current Skylon baseline is the D4, which has 20tonnes to LEO. they specifically increased it's size because they needed to address the 4.6mx14m payload envelope and needed the 20 tonnes to have the necessary GTO performance. Not unlike the Shuttle's IUS.

Hopefully it won't suffer the same fate.

[bearshrimp]

Here is a consideration that I haven't seen raised for Skylon costs before, and I for one have no idea of its potential impact on Skylon launch prices. I wonder if anyone has a more quantitative notion.

Skylon, being horizontal take-off, has, at least in principle, anytime-abort. Granted that with a take-off speed of around 0.5 Mach, the abort options may be few.

So what does this do for insurance rates? These are a sizable fraction of current launch costs. Might Skylon's abort advantages lower insurance rates and therefore also contribute to lower total cost to orbit?

I did see that issue brought up somewhere but I can’t recall where.  I am no actuary, but if we are talking a .75% failure rate (half that of the Shuttle), a manufacture cost of $405M + $2B in development costs for an insured value of $2.405B (minus payload of course) you would probably see an insurance charge of around $9m per launch.  Any actuaries out there may want to chime in on this one.

[baldusi]

Skylon will behave more like an aircraft. That means that the reliability should be on the 99.999%s or better. Particularly, it would allow for a type certificate process, which requires something like 100 flights. It's obviously not applicable to a disposable LV, but for the Skylon it might be reasonable. If we assume three crafts (800M x 3) plus 100 flights (say, 9M x 100), that's a 3.3B cost. But you could then use the crafts with some refurbishment. And your insurance costs would be really low.
To get an idea, current F9 is about 15%, Ariane 5 is lower than 10%. A reusable Skylon should be much, but much lower. In fact, it they can get 99.99%, insurance costs should be close to 0.05%.

[bearshrimp]

Of course SpaceX are claiming even lower long term figures for (a) re-usable launcher(s), but whilst that too is an amazing prospect, SpaceX seem to me much less concerned about being conservative in their presentation of the future!

Yes, but the Falcon and Falcon-9 have flown and you can go out and contract with SpaceX to put your satellite into orbit.  Right now they are booked until 2013 I believe.  Given that they are already operating at budget, I tend to lend more credence to their statements than those made by companies still 10 years out from a viable product.

Oh, BTW I am not a SpaceX plant I just noticed that a lot of my posts sound like a cheering section for SpaceX. 

[bearshrimp]

Skylon will behave more like an aircraft. That means that the reliability should be on the 99.999%s or better. Particularly, it would allow for a type certificate process, which requires something like 100 flights. It's obviously not applicable to a disposable LV, but for the Skylon it might be reasonable. If we assume three crafts (800M x 3) plus 100 flights (say, 9M x 100), that's a 3.3B cost. But you could then use the crafts with some refurbishment. And your insurance costs would be really low.
To get an idea, current F9 is about 15%, Ariane 5 is lower than 10%. A reusable Skylon should be much, but much lower. In fact, it they can get 99.99%, insurance costs should be close to 0.05%.

I don’t think that’s a great analogy, the SKYLON is predicted to have an expected life of 200 cycles.  Per section 410 of the FAA regs (1.410(a)(1)(i) CFR14) the duty cycle of an Airbus A300 is 36,000 cycles.  While 200 cycles is much better than one (disposable rockets) it still is far away from the expected duty cycle of a commercial airliner.

[baldusi]

Skylon will behave more like an aircraft. That means that the reliability should be on the 99.999%s or better. Particularly, it would allow for a type certificate process, which requires something like 100 flights. It's obviously not applicable to a disposable LV, but for the Skylon it might be reasonable. If we assume three crafts (800M x 3) plus 100 flights (say, 9M x 100), that's a 3.3B cost. But you could then use the crafts with some refurbishment. And your insurance costs would be really low.
To get an idea, current F9 is about 15%, Ariane 5 is lower than 10%. A reusable Skylon should be much, but much lower. In fact, it they can get 99.99%, insurance costs should be close to 0.05%.

I don’t think that’s a great analogy, the SKYLON is predicted to have an expected life of 200 cycles.  Per section 410 of the FAA regs (1.410(a)(1)(i) CFR14) the duty cycle of an Airbus A300 is 36,000 cycles.  While 200 cycles is much better than one (disposable rockets) it still is far away from the expected duty cycle of a commercial airliner.

Not all aircraft are designed for so many cycles. Loot at the original AN-124, for example, with its 7,500hr of airframe life. It was, obviously extended to 15k, 24k and now 40k. Same might happen with the Skylon.
During its lifetime (if it ever starts), much will be learn about the craft, it's operation environment, and the main failure drivers and the effects of fatigue on each part.
At the same time, the simple fact of flying as an aircraft, makes the second and following fights safer, from an engineering point of view. Since most of the integration problems must have crept up.

[lkm]

 

Given that there were 23 total launches of commercial LEO payloads in 2010 (per the FAA’s 2010 Commercial Space Transportation Review), and that projected growth is 8%
....
However, if we assume SKYLON came on line in 2022 (10 year gestation period) the demand for LEO payloads would only justify 40 launches in the 1st year of operation given an 8% growth rate.  Over a 10 year period, given a growth rate of 8%, a total of 554 payloads would be available for transport to LEO. 

Two thirds of the commercial launches were to GTO not LEO in 2010 which you've failed to take account of in your extrapolation. Skylon addresses GEO either through the Skylon Upper Stage or the Fluyt OTV. Every Fluyt mission takes 5.2 Skylon launches, I believe.  So if half the GEO launches are large enough to require a Fluyt then that's 1331 Skylon launches over 10 years, or an average of 133 launches a year by your numbers or nearly twice the 70 flights a year REL use as their baseline.

Yes, I know that the majority of missions are to GTO rather than LEO, but I haven’t seen any cost projections regarding the Skylon Upper Stage or the Fluyt OTV so I did not include them in my analysis.  The ESA had the following statement on page 17 of the Assessment:

“One point to be made is that ESA considers that the SKYLON Upper Stage (SUS) which is potentially required for GTO missions may need to be included in the overall development costs. This is because if telecoms spacecraft customers have to pay to develop a GTO stage on top of the launch price then this may push the cost to orbit to a point where the SKYLON becomes less competitive. ESA recommends that the development cost model of the vehicle be re-assessed to account for the additional cost of developing the SUS.”

So, it’s difficult to estimate the cost to GTO provided by Skylon when compared to other launch systems. 

In a way my earlier estimate was probably optimistic in that only 33 payloads delivered to any orbit in 2010 were commercial.  So if we assume that SKYLON will be responsible for delivering all commercial payloads to orbit, and the market expands 8% per year.  The total number of commercial launches to any orbit would be 73 in 2022 rising to 150 in 2032.  The total number of commercial launches would be 1,176.  Assuming each SKYLON would be able to deliver 200 payloads over its 10 year service life that means that there would be a market able to support 6 SKYLON spaceplanes.

The issue is still that in the economic analysis the development costs are amortized over 30 SKYLON craft.  I just don’t see that kind of market demand.  In a way it’s a catch-22.  To bring down the operation and manufacturing cost on a per launch basis each SKYLON needs to be capable of performing a large number of missions.  On the other hand, the ability to have each SKYLON perform a large number of missions will reduce the number required to satisfy the market.  This in tern leads to a smaller number of SKYLONs constructed and increases the amortized development cost on a “per vehicle” basis.

To justify a production run of 30 SKYLONs between now and 2032 the demand for commercial launches would need to increase an average of 17.7% per year.  I think we would only see double digit growth in the commercial space sector if there were considerably more money to be made in orbit.  Perhaps if the Japanese really do develop their 1GW orbital power station (requiring between 300-400 orbital trips) Skylon would find its market, but that’s a big if.

You are  ignoring the fact that there is no centralized global launch operator procuring them so there will  be a competitive market between commercial Skylon operators and their fleets. Even if in theory every flight could be completed by six Skylon's working flat out the reality would be several competing spaceports vying for the same business with a much larger global Skylon fleet.

In 2010 there were twice as many government launches as commercial and it seems unlikely that there won't be at least some of those payloads on either commercial or government operated Skylons, the later itself increase order numbers for perhaps non=-economic reasons.
SpaceX has at least two clear weaknesses compared to Skylon, firstly it's American and thus ITAR encumbered where as Skylon is ITAR-free and while one would think ITAR problems should be long past in 2020, this is the US government at work so that's perhaps not a safe bet. Secondly Skylon is inherantly cheaper as launch rates increase, and also fundamentally cooler.
A third source of flights over the decade is HSF as if the ISS operational life is extended to 2028 then there will be six years during Skylon could deliver passengers and cargo with a SPLM for either NASA or ESA further the current NASA exploration direction of an L1 gateway station and lunar return mirror well the Skylon Lunar architecture so Skylon and Fluyt would be easily capable of providing commercial cargo or crew to support  a gateway station.

If we assume this 8% growth rate holds ( which I think is actually much too small) from 2022-2032 there will be 2138 government launches and 964 commercial launches ( there were actually 23 not 33 in 2010) if Skylon with 30 vehicles  take 50% of the government market and 70% of the commercial market they would have 1744 launches not counting extra flights for Fluyt operations which is an average of 174 flights a year or every Skylon flying every two months. That seems pretty reasonable for a first generation RLV SSTO.
 

[flymetothemoon]

“One point to be made is that ESA considers that the SKYLON Upper Stage (SUS) which is potentially required for GTO missions may need to be included in the overall development costs. This is because if telecoms spacecraft customers have to pay to develop a GTO stage on top of the launch price then this may push the cost to orbit to a point where the SKYLON becomes less competitive. ESA recommends that the development cost model of the vehicle be re-assessed to account for the additional cost of developing the SUS.”

Jonathan Amos mentions at the end of this ATV article that ESA are considering developing the ATV into "a new European multi-role space tug".

http://www.bbc.co.uk/news/science-environment-17539319

Does anyone know any more about this? Could this potentially take on the Fluyt/SUS role?

More about the ATV:
http://www.nasaspaceflight.com/2012/03/europes-atv-3-spacecraft-deliver-large-cargo-load-iss/

http://www.nasaspaceflight.com/2012/03/atv-3-esa-important-resupply-mission-to-iss/

http://www.esa.int/SPECIALS/ATV/index.html

[bearshrimp]

Given an 8% annual growth rate, here are the number of commercial launches one may reasonably expect annually.  Note that with a duty cycle of at least 200 missions over a 10 year period and a 48 hour turnaround time, there would be little need to ever have more than 2 or 3 SKYLONs operating at any given time.  Even if you assume competing vendors offering services I just don’t see the market for 30 SKYLONs.
 
2010   33
2011   35
2012   37
2013   39
2014   42
2015   45
2016   48
2017   51
2018   55
2019   59
2020   63
2021   68
2022   73
2023   78
2024   84
2025   90
2026   97
2027   104
2028   112
2029   120
2030   129
2031   139
2032   150
Total   1176

If there were say 6 Skylons operational in 2025, they would be substantially underutilized.  The ESA Analysis states that operating costs are based on 70 missions per year per SKYLON.  With just 3 SKYLONs operating in 2025 at a 21% utilization rate, the cost per mission would go up as the fixed annual operating costs would be spread across fewer launches.

My argument here is that I cannot see a valid reason to amortize the development costs over 30 craft unless there is some dramatic (and unprecedented) need to access space.  The development costs could possibly be amortized over 10 vehicles…maybe, but the cost projection would need to be heavily back loaded because of the demand curve, i.e. 3 SKYLONs will serve the world’s commercial needs nicely until about 2029.

[flymetothemoon]

http://www.rinkworks.com/said/predictions.shtml

[flymetothemoon]

Yes, but the Falcon and Falcon-9 have flown and you can go out and contract with SpaceX to put your satellite into orbit.  Right now they are booked until 2013 I believe.  Given that they are already operating at budget, I tend to lend more credence to their statements than those made by companies still 10 years out from a viable product.

Oh, BTW I am not a SpaceX plant I just noticed that a lot of my posts sound like a cheering section for SpaceX. 

I LOVE what SpaceX are doing. Go Elon Musk! However even Elon said that making the Falcon-9 re-usable will be "super damn hard", but he will give it his best shot. And that's just re-usable Falcon-9. I really wanna see re-usable Falcon Heavy too!

[bearshrimp]

http://www.rinkworks.com/said/predictions.shtml

Don’t get me wrong, I am a supporter of SKYLON and think that the future of orbital interface will be the aerospace plane.  I am sure that the SABRE engine will work and that the technical issues do not stand in the way of developing a fully reusable SSTO craft.  Of course SKYLON will work and I have gone through the previous posts dealing with all the potential engineering issues.

My concern is that the costs presented to the ESA and to Parliament are vastly understated.  In addition it appears IMHO that the commercial justification for the low cost to orbit is circular, i.e. the cost will come down when you can have a 48 hour turn around and that the development costs will be amortized over a large number of aerospace craft.

This is a better solution than the shuttle ever was, but I think it’s more realistic to see SKYLON developed to provide the public goods noted in the ESA Assessment including:

Advanced heat exchanger technology
Advanced materials
Hydrogen aviation
Formulation of new markets
Maintenance and replacement parts industry
Space finance and professional services
Spaceports
Downstream services
E-commerce
Global weather and navigation
Catastrophe management
Space manufacturing and research
Solar power
Space tourism

I want to make sure everyone knows what they are getting into this time.  The shuttle was truly traumatic in that the claims of cost efficiency in the 70’s are basically the same as those being made for SKYLON today.  I am not saying that SKYLON will be the boondoggle that the Shuttle was in the economics field, but that the costs to orbit have already dropped so much that SKYLON will be very hard pressed to turn a profit as a strictly commercial venture.

Do I believe that government should support development of new technology…of course I do. 

Would I be heartbroken if $2 of my taxes when to support SKYLON R&D even if it never turns a profit…no.

[flymetothemoon]

http://www.rinkworks.com/said/predictions.shtml

I shoulda put a wink in that last post! No offence intended

Granted, but you still seem to be saying that the launch rate required for acceptable amortised costs can't be supported based on current launch requirements and rates for throwaway vehicles and that just doesn't make any sense to me. It's surely a non-sequitur. And an oft-repeated one at that.

I think it is a very real concern that any complex engineering project will vastly over-run of course, but I don't feel that this one is vastly more ambitious than another. More ambitious than an A380 perhaps, but not orders of magnitude more ambitious.

I am not a rocket scientist, that much is probably painfully obvious, but they have been working on this for 30 years and it doesn't appear to be in any way made of unobtainium - assuming the pre-cooler tests go well. I guess I am saying it is a novel plane/rocket with a bit of secret sauce, not a ludicrous leap into the exotic.

So sure, building an A380 can get expensive with overruns, but it appears at least to be within the same sort of parameters.

Yes the market has to evolve. I accept that, but it is private enterprise that is expected to take those risks and if they don't it won't happen. If they do, at these costs, I am sure they will have done pretty good homework on the business case and would be imagining the kinds of volumes I was alluding to in previous posts. The cost of launch would tumble if building dozens of Skylons could be justified.

The current Skylon baseline is the D4, which has 20tonnes to LEO. they specifically increased it's size because they needed to address the 4.6mx14m payload envelope and needed the 20 tonnes to have the necessary GTO performance. Not unlike the Shuttle's IUS.

And thanks for this Baldusi. I believe this information had escaped me. Where did you see that? Any more info anywhere?

If they did develop a twenty ton to LEO version that would dwarf the launch capacity of a Falcon-9 with the fuel and bulking up required for re-use. I believe Ariane have only just managed that capacity for the first time in the last month. AFAIK that leaves the only possible competitor we could currently imagine to be a Falcon Heavy re-usable which, with the best will in the world (and I would like to see it tomorrow) has got to be 'quite a few' years away.

[lkm]

Given an 8% annual growth rate, here are the number of commercial launches one may reasonably expect annually.  Note that with a duty cycle of at least 200 missions over a 10 year period and a 48 hour turnaround time, there would be little need to ever have more than 2 or 3 SKYLONs operating at any given time.  Even if you assume competing vendors offering services I just don’t see the market for 30 SKYLONs.
 
....

If there were say 6 Skylons operational in 2025, they would be substantially underutilized.  The ESA Analysis states that operating costs are based on 70 missions per year per SKYLON.

The London Economics review was based on 70 flights per year, not per Skylon, just  per year. At 70 people make money. With 30 aircraft. With 10 aircraft nobody loses any money. And you're still ignoring government launches.
The 48 hour turn around is a feature allowing for a high flight rate in a busy operationally responsive launch market, it doesn't your going to fly it like that at 70 launches a year.

I want to make sure everyone knows what they are getting into this time.  The shuttle was truly traumatic in that the claims of cost efficiency in the 70’s are basically the same as those being made for SKYLON today. 

The shuttle wasn't a disappointment until the OMB capped its development funding to $1 billion a year and told NASA to trade development  costs for much higher operational costs. Its claims of efficiency largely originate from when they were true, before the OMB gutted them.

The comparison your looking for is Concorde an Anglo-French development that never made enough models to pay for itself, but Concorde made money for its operators and was only limited by politics.

[bearshrimp]

Does anyone know where I can find the “SKYLON System requirement Review: SKYLON Commercial Operation, Alan Bond, Reaction Engines Ltd” listed as RD2 in the ESA Assessment?

I would like to take a look at the actual ROI numbers provided by REL.  Not that I need to be satisfied for SKYLON to happen , but if I can see a reasonable argument where SKYLON will be able to achieve a LEO cost to orbit of $1M per mt (i.e. $1000/kg) then I will be more enthusiastic. 

To get really excited about a new orbital interface system I would like to see a cost to LEO of $500/kg, now that IMHO would trigger a true commercial space race.  Even if the development costs are subsidized I would like to see an analysis of how we can get to $500/kg within the next 25 years. 

The more I think about it the more I am worried about the insurance costs.  Even more than R&D, I see a need for governments to subsidize insurance rates until the SKYLON develops a proven track record.  Without a set record and given the high value/high utilization of the SKYLON I doubt Lloyd’s would allow use of a failure rate less than 1.5%.

For ESA operations maybe the ESA could step in and provide insurance until the SKYLON had an established safety record.   

[Robotbeat]

The easiest way to get launch costs down to $100/kg is just to have the government pay the other $9900/kg.

[bearshrimp]

The easiest way to get launch costs down to $100/kg is just to have the government pay the other $9900/kg.

Indeed...

[lkm]

Does anyone know where I can find the “SKYLON System requirement Review: SKYLON Commercial Operation, Alan Bond, Reaction Engines Ltd” listed as RD2 in the ESA Assessment?

I would like to take a look at the actual ROI numbers provided by REL.  Not that I need to be satisfied for SKYLON to happen , but if I can see a reasonable argument where SKYLON will be able to achieve a LEO cost to orbit of $1M per mt (i.e. $1000/kg) then I will be more enthusiastic. 

To get really excited about a new orbital interface system I would like to see a cost to LEO of $500/kg, now that IMHO would trigger a true commercial space race.  Even if the development costs are subsidized I would like to see an analysis of how we can get to $500/kg within the next 25 years. 

I don't think any of the submission documents to the review have been published yet but I think the paper you might want to read is "A Two Stage To Commercial Market Approach To Reusable Launch Vehicle Development" which was presented at the 17th AIAA International Space Planes and Hypersonic Systems and Technologies Conference last year. You can buy a copy here:

https://www.aiaa.org/IframeTwoColumn.aspx?id=4745

Alternatively the Space solar power study paper has an operator cost breakdown which you might find useful.