ISS Commercial Resupply Services 2 (CRS2) 2017-2030

[meekGee]

after a failure an attempt is made to fix the root cause, so there's a smaller chance of failure after the fix.

Prove that?

Why is an implementation of the fixed design less likely to fail than the implementation of the original?

A TV analogy:  A good design is analogous to the prize, a failure is analogous to the goat, and reality is analogous to Monty Hall.   By showing you were the goat is, the odds of a re-design getting the prize increase.    Yes?

Another try:
A man falls down the stairs.  Upon investigation, it turns out that he couldn't see the warning strip because he is color blind.  As a result, a new warning strip is redesigned.  Also, as a result of increased scrutiny, the handrail was improved.   Why couldn't that have been done before the accident?  because there are a very large number of possible faults, and it took reality to show us the issue with color-blind people, and now there's one less thing to worry about.  The odds of people falling down these stairs have decreased.

[llanitedave]

this thread has gone way off track

There was a prior chance of 75.4% of that.

[freakdog]

after a failure an attempt is made to fix the root cause, so there's a smaller chance of failure after the fix.

Prove that?

Why is an implementation of the fixed design less likely to fail than the implementation of the original?

A TV analogy:  A good design is analogous to the prize, a failure is analogous to the goat, and reality is analogous to Monty Hall.   By showing you were the goat is, the odds of a re-design getting the prize increase.    Yes?

Another try:
A man falls down the stairs.  Upon investigation, it turns out that he couldn't see the warning strip because he is color blind.  As a result, a new warning strip is redesigned.  Also, as a result of increased scrutiny, the handrail was improved.   Why couldn't that have been done before the accident?  because there are a very large number of possible faults, and it took reality to show us the issue with color-blind people, and now there's one less thing to worry about.  The odds of people falling down these stairs have decreased.

The entire point of FMECA along with FDIR are to determine, based on past experiences and analysis of systems, what are the possible failure causes and the effects and to find ways to mitigate them, both through robust design and by software detection.  The engineers on a space program have access to all the ways that all of the spacecraft that were ever launched, by pretty much anyone, had failed or nearly failed.

From my reading of the report of the Orbital ATK failure, it seems they lacked a robust FDIR, specifically wrt to their engine controls, which seems to me to be much less mature than the facilities on Atlas 5.  The overall complaint looks more like a systematic failure to adopt practices to ensure robust engineering in the context of FMECA.  If that matched reality, it's more like a hole is patched in one place and the water leaks out another hole.

[HIP2BSQRE]

is it possible that the reason that we are waiting is that NASA wants to give both SpaceX and OribitalATK a chance to RTF before announcing the winners of the contract?

[deadman719]

I had a thought regarding a potential influence in the number of contract awards that materialize for this effort.  It's my opinion three contracts will be awarded.  Apart from the capabilities each provider brings to the table, the reason for three awards is redundancy.  As we have seen, the two existing providers have had mission failures.  Up until recently, a single or double mission failure was not as critical because the Progress, ATV and HTV provided plenty of capability.  Unfortunately Progress had a failure along with the commercial providers leaving the HTV as the sole provider since ATV had been retired.

Although failures have been rare, the potential for them to occur is a reality.  A third commercial provider would provide similar redundancy to that which existed while the ATV flew.  In addition to redundant supply vehicles, redundant launch vehicles would also exist.

Thoughts?

[deltaV]

I had a thought regarding a potential influence in the number of contract awards that materialize for this effort.  It's my opinion three contracts will be awarded.  Apart from the capabilities each provider brings to the table, the reason for three awards is redundancy.  As we have seen, the two existing providers have had mission failures.  Up until recently, a single or double mission failure was not as critical because the Progress, ATV and HTV provided plenty of capability.  Unfortunately Progress had a failure along with the commercial providers leaving the HTV as the sole provider since ATV had been retired.

Although failures have been rare, the potential for them to occur is a reality.  A third commercial provider would provide similar redundancy to that which existed while the ATV flew.  In addition to redundant supply vehicles, redundant launch vehicles would also exist.

Thoughts?

Each additional provider probably increases costs by circa $1B over the life of the contract due to increased fixed costs. Redundancy is nice, but NASA has better things to do with its limited funds. I would go with 1-2 awards.

[rcoppola]

Agreed. For cost savings, you need some economies of scale. So 2 providers max. I don't see anyone bidding lower then SpaceX, especially if they merge the Cargo and Crew lines allowing cargo to have propulsive landings or propulsive assisted landings, facilitating real reuse. And Dragon w/ trunk is the only vehicle capable of large unpressurized upmass, which I think will be critical in the years to come as the station ages. So DC vs Cygnus. And I have no idea which way they'll go but I suspect if DC has a close enough bid and they can mitigate development, testing, schedule risk, they could pull it off.   

[pathfinder_01]

I would go with 2-3 for redundancy and to prevent “single tracking”. It would also grow the supplier chain a bit. The problem with one is not only will disaster shut that one down, but delay would cause back up. Not good.

If you have 2-3, while one provider is down or delayed another could be up and running.  In the time Orbital was down two flights of Space X occurred then Space X went bang in June, but Orbital is aiming for December.
It is also useful for growing commercial markets. Bigelow or others can have more than one choice when it comes to Cargo and NASA will have more leverage when it comes to negotiating a CRSIII contract if ever that happens (and such a contract need not be tied to the ISS.).

I don’t think economies of scale apply much here. Space X and Orbital are using their workforces for other company projects and I suspect so will Serra Nevada.  The economies of scale for Space X is by having Falcon 9 have more than NASA as an customer and by having both crew and cargo Dragon. For Cygnus is it in having as much of Cygnus be made of parts from other space craft and missions (an efficient source of what is already available). The trouble with an single supplier is that even if they reduce their own internal prices by purchasing more, they have little reason to pass on the savings to the customer (i.e. NASA). With more suppliers there is competition for missions and lowing your price can increase the odds of being selected.

[yg1968]

I had a thought regarding a potential influence in the number of contract awards that materialize for this effort.  It's my opinion three contracts will be awarded.  Apart from the capabilities each provider brings to the table, the reason for three awards is redundancy.  As we have seen, the two existing providers have had mission failures.  Up until recently, a single or double mission failure was not as critical because the Progress, ATV and HTV provided plenty of capability.  Unfortunately Progress had a failure along with the commercial providers leaving the HTV as the sole provider since ATV had been retired.

Although failures have been rare, the potential for them to occur is a reality.  A third commercial provider would provide similar redundancy to that which existed while the ATV flew.  In addition to redundant supply vehicles, redundant launch vehicles would also exist.

Thoughts?

Each additional provider probably increases costs by circa $1B over the life of the contract due to increased fixed costs. Redundancy is nice, but NASA has better things to do with its limited funds. I would go with 1-2 awards.

I doubt it's that much. The whole point of commercial crew and cargo is that NASA is not supposed to be their only client (at least for the LV part).

[guckyfan]

I see one big problem for Sierra Nevada. They need to recover their development cost over the number of flights. If they get only few, that increases cost per flight by a lot. Or they gamble on follow up contracts which is a risk.

[Prober]

I see one big problem for Sierra Nevada. They need to recover their development cost over the number of flights. If they get only few, that increases cost per flight by a lot. Or they gamble on follow up contracts which is a risk.

add to that thought, and remember SNC is not a startup company.

[Bynaus]

I see one big problem for Sierra Nevada. They need to recover their development cost over the number of flights. If they get only few, that increases cost per flight by a lot. Or they gamble on follow up contracts which is a risk.

OTOH, once the DreamChaser actually flies, I can imagine it finding customers besides NASA. SNC has been very actively pursuing opportunities outside the US as well. In Europe, perhaps, or Japan? A Bigelow station? Lower Ge's on re-entry might be good for those frail millionaires' bones...

[joek]

I see one big problem for Sierra Nevada. They need to recover their development cost over the number of flights. If they get only few, that increases cost per flight by a lot. Or they gamble on follow up contracts which is a risk.

An initial CRS-2 award has a guaranteed minimum of six missions.  That helps set a lower bound on the number of missions for cost recovery.  However...

Each proposal must include not-to-exceed (NTE) pricing for 1-5 missions for each calendar year 2018-2024, for each of up to 4 standard mission types.  Those prices must pass a" price reasonableness" test.

Gambling on additional missions by artificially reducing the cost of a lower number of missions below what would be considered prudent would fail the price reasonableness test and would be rejected.

There is an initial or base ISS integration line item which is separate from pre-mission pricing, and which is where the primary costs for certification of a new vehicle belong.

[a_langwich]

I see one big problem for Sierra Nevada. They need to recover their development cost over the number of flights. If they get only few, that increases cost per flight by a lot. Or they gamble on follow up contracts which is a risk.

add to that thought, and remember SNC is not a startup company.

What about the redesign of DC down to fit inside a payload shroud?  And to put the larger berthing connector on a smaller vehicle?  Is DC still slated to use an Atlas V 552 for launch?  The cargo capacity was listed at 5000kg, which almost certainly assumes/requires an Atlas V 551 or 552.  That is considerably more expensive than an Atlas V 401 like the one Cygnus is using, which adds cost in addition to the unfinished development costs.

My guess is, since SNC is explicitly still in the competition, they generated a low cost bid (they really, really want to win the contract).  The hard question will be whether or not that represents too high a risk profile or not.

[TrevorMonty]

I see one big problem for Sierra Nevada. They need to recover their development cost over the number of flights. If they get only few, that increases cost per flight by a lot. Or they gamble on follow up contracts which is a risk.

add to that thought, and remember SNC is not a startup company.

What about the redesign of DC down to fit inside a payload shroud?  And to put the larger berthing connector on a smaller vehicle?  Is DC still slated to use an Atlas V 552 for launch?  The cargo capacity was listed at 5000kg, which almost certainly assumes/requires an Atlas V 551 or 552.  That is considerably more expensive than an Atlas V 401 like the one Cygnus is using, which adds cost in addition to the unfinished development costs.

My guess is, since SNC is explicitly still in the competition, they generated a low cost bid (they really, really want to win the contract).  The hard question will be whether or not that represents too high a risk profile or not.

On cargo delivered for $/kg the DC may well work out cheaper than Cygnus.
Reusable DC on V552 delivering 5000kg compared to expendable Cygnus on V401 delivering 3500kg.

[baldusi]

It will quite probably be volume limited.

[joek]

It will quite probably be volume limited.

Safe assumption.  Per the RFP:

2.1.1 A minimum usable pressurized cargo density of 65 Cargo Transfer Bag Equivalents (CTBE) per 1000 kg of pressurized cargo shall be used.

That is 291 kg/m³ or 3.44 m³/1000kg usable.  Usable is what actual ISS cargo can fit, not the claimed pressurized volume.  Unfortunately usable numbers are not public (edit: I cannot find a credible reference for usable cargo volume.  Anyone?)

[NealioSpace]

I could see NASA considering to NOT select SpaceX; due to the fact they really don't need the money anymore to survive and be successful. They could select the other two (and also have the work they have done with SpaceX in their back pocket, give them a back-up capability to launch on Dragon should the emergency need arise). In the meantime, perhaps the money is better spent for our nation to develop an additional U.S. spacecraft. I'm a big fan of SpaceX, but realistically they don't have a need for winning this contract.
Thoughts?

[RonM]

I could see NASA considering to NOT select SpaceX; due to the fact they really don't need the money anymore to survive and be successful. They could select the other two (and also have the work they have done with SpaceX in their back pocket, give them a back-up capability to launch on Dragon should the emergency need arise). In the meantime, perhaps the money is better spent for our nation to develop an additional U.S. spacecraft. I'm a big fan of SpaceX, but realistically they don't have a need for winning this contract.
Thoughts?

If SpaceX is going to get to Mars, they need all the money they can get.

Of course, with SpaceX flying crew to ISS, NASA may want to give cargo to Orbital ATK and Sierra Nevada if building the space industry is part of the plan. Including Boeing flying crew, that's four US companies running flights to ISS.

[rcoppola]

I could see NASA considering to NOT select SpaceX; due to the fact they really don't need the money anymore to survive and be successful. They could select the other two (and also have the work they have done with SpaceX in their back pocket, give them a back-up capability to launch on Dragon should the emergency need arise). In the meantime, perhaps the money is better spent for our nation to develop an additional U.S. spacecraft. I'm a big fan of SpaceX, but realistically they don't have a need for winning this contract.
Thoughts?

None of that is part of the selection criteria. NASA needs, wants the best, most relevant services at the best prices possible.

Besides, neither OATK or SNC offers the un-pressurized capability that Dragon's trunk offers. And we do not know what upgrades they plan to include for the next Dragon Cargo Variant. Extended trunk? Propulsive landing? Increased mass and volume for pressurized cargo? Increased power/electricity for additional experiment handling, etc..

Edit: And possibly a robust re-use certification plan, enabling them to offer even lower prices per mission.