space-request+@ANDREW.CMU.EDU (11/10/90)
This message was originally submitted by space-request+%ANDREW.CMU.EDU@CARNEGIE
to the SPACE list at UGA. If you simply forward it back to the list, it will be
distributed with the paragraph you are now reading being automatically removed.
If you edit the contributions you receive into a digest, you will need to
remove this paragraph before mailing the result to the list. Finally, if you
need more information from the author of this message, you should be able to do
so by simply replying to this note.
----------------- Message requiring your approval (477 lines) -----------------
SPACE Digest Volume 12 : Issue 520
Today's Topics:
Re: You Can't Expect a Space Station to be Cheap
Administrivia:
Submissions to the SPACE Digest/sci.space should be mailed to
space+@andrew.cmu.edu. Other mail, esp. [un]subscription notices,
should be sent to space-request+@andrew.cmu.edu, or, if urgent, to
tm2b+@andrew.cmu.edu
----------------------------------------------------------------------
Date: Thu, 1 Nov 90 22:42:03 -0500
From: "Allen W. Sherzer" <aws@iti.org>
Subject: Re: You Can't Expect a Space Station to be Cheap
Newsgroups: sci.space
Cc:
Couple of things:
1. Using the LLNL as a Freedom replacement is NOT part of the
Great Exploration. I am extrapolating how it may be done and
as such I may solve a problem differently than they do. I think
these solutions will work but they may not be the solution LLNL
is looking into.
2. If people are getting bored with this, speak up and we can take it
to email or perhaps the space-tech list.
In article <6930@hub.ucsb.edu>:
>>Boeing has a long track record of building highly successful
> >aircraft. LLNL does not have any track record at all in large
> >space systems.
>+ First of all, that is irrelevant.
>It is not irrelevant, track record is the SOLE reason Boeing
>receives these kinds of contracts. Therefore your analogy
>there cannot be applied to LLNL.
As Henry pointed out, that is not true with the 707. However, that
is not what I ment. Your line of arguement was:
1. The price of large new engineering projects cannot be predicted
in advance.
2. LLNL is a large engineering program
therefore:
3. The price of LLNL cannot be predicted.
You backed this up with reference to several fixed price programs
which didn't work.
I point out that many large engineering programs DO work and thus
that cannot be held against LLNL. The failure of a large scale
engineering program nor more shows LLNL to be invalid than it goes
to show Freedom being invalid.
>+Second of all, the LLNL contractors on this DO have a good track
>+record in large aerospace systems.
>
>So do the Freedom contractors.......
So they you should agree that the LLNL approach is just as valid
and likely to work as Freedom?
>Don't you think LLNL will go through similar cost increases as
>reality starts to dawn, and the magnitude of the task is truly
>appreciated ?
Do I think it will cost more? Wouldn't suprise me. Will they
be off by a factor of five like Freedom? No I don't think so.
If they are, then their concept won't work.
>And in the 2 years I have been working Freedom,
>first element launch has been set at March 1995, and has not
>slipped one day. And the costs have been reduced, not
>increased.
Again, according to the original schedule, Freedom was to be
permanently crewed in 91 (next year) at a cost of $8B. Maybe the
schedules havn't changed, but that doesn't mean they can be met.
Example: Freedom will use 27 Shuttle flights for assembly over
five years. You don't seriously think that can be done do you?
If not, why don't your schedules reflect serious numbers? When will
Freedom management slip the schedule to reflect realistic launch rates?
Indeed under the current schedule Freedom cannot be built. Fisher
Price say things will start breaking so fast that all the Shuttle
flights will be needed just for repair. They won't be able to
build anything new.
You man not have changed your schedules in two years, that doesn't
make your schedules realistic.
> +If LLNL doesn't have an Earth Station in orbit four years after
>+ start costing under $4 billion I will call for it to be killed.
>+ How's that?
>
>That's OK provided Freedom does not get cancelled in the interim,
>then we get nothing.
Nope. If they can't do it within a factor of two of their estimates
then they need to be cut off and start over. Otherwise we will have
what is currently going on: slipping real schedules and a program
which is slowly dying but will take years to die. We can't afoard to
let that happen again.
*THAT* will get us nothing.
>+Interesting choice of numbers. Freedom was to be ready in about
>+six years with the start of operations next year. We have
>+already spent six years with nothing to show for it (not
>+counting the $8 billion in briefing slides that is).
>
>Progress is greater than you think. Breadboard hardware has
>been built and tested, and we are moving into Engineering
>Development units.
Only in the most preliminary sense. They are also years late,
far over budget, and still subject to redesign.
>Manufacturing drawings are being released,
You haven't even passed PDR yet. How can you release manufacturing
drawings?
>although probably not in their final form. Flight software
>specifications have been written, and coding will start in a
>matter of months (not sure exactly - not my area). Flight
>article metal bending will begin in 1992 and deliveries of flight
>articles will begin in 1993.
How much of this will be tossed out in the current redesign?
>March 1995 first element launch
>date has been unchanged for the last two years.
See above. In the entire six year history of the program this
is the only year that date hasn't slipped. As we have seen,
the date may not have slipped on paper, they have all actually
slipped.
I'm sorry to be blunt, but your schedules aren't worth the paper
they are printed on.
>+ Besides, sources on Freedom say that the budget cuts are the
>+ excuse not the cause of the redesigns.
>
>Redesigns would have been necessary with or without budget cuts,
Then why is station management blaming the redesign on the cuts?
>but the cuts have made them more often and more severe. And the
>fact that the redesigns have been for political and not technical
>reasons has been very bad for morale.
They have been political because NASA didn't listen to Congress
in the design. Congress wanted a microgravity research facility.
NASA smelled pork and started adding everything but the kitchen
sink onto it. It should come as no suprise when Congress says
they don't want to pay for all that other stuff.
Question: Congress is mandating that Freedom's dependance on the Shuttle
and unrealistic launch rates end in the next redesign. Is this a political
or technical reason?
As for their being no technical reasons, look at the stations
power, weight, and repair budget. Compare them with the current
situation. I would call that a technical reason for redesign.
> >dubious qualification prospects and admitting conventional
>> cells woulBd be just as good and less risky;
>
>But the NiH and solar celss are not space qualified,
Of what use is space qualifying? Yes you need to test components
to see how they will work but 'space qualification' is a subjective
estimation at best. Example: the Shuttle is 'man rated'. Yet it
has a failure rate no better than Titan or Delta.
LLNL will qualify their hardware in the exact manner as Apollo
did (and I assume you thought Apollo was OK). Is this more risky?
Absolutely. Exploration and opening up new frontiers always is.
If Columbus and the early new World colonies where run by NASA,
we would never have come over; it would be too expensive.
>+ But again, none of these are serious problems. If they work, we
>+advance the state of the art and save money. If they fail, their
>+weight budgets and schedules allow for it.
>
>No it is worse, if they fail and Freedom is cancelled, we waste
>decades, it will take that long to rebuild a manned space
>program.
We have already wasted a decade. Freedom will take at least
another decade (minimum) to get off the ground. If you use
realistic Shuttle numbers, it will take another decade still
to build it. Why is Freedom worth three decades?
In half the time it took to get Freedom up we went from nothing but
Redstones to a man on the moon. In a quarter of the time needed for
Freedom (and a third the cost) we built Skylab. I'm not saying that
putting a space station in orbit is trivial but I am saying it is a
lot easier than Freedom makes it seem.
>> station cannot go up in the HLV with the crew following in the
>> Shuttle or on a (safer and cheaper) Delta.
>
>Oh come on, the Delta is an unmanned ELV, there are no manned
>spacecraft designed to fly on delta. And developing a new
>manned spacecraft is not cheap. Don't tell me that has been
>factored into the cost too.
Yep it has. Tell me, in 90 $$ how much does an Apollo capsule
cost?
>But again, Titan is expensive, no cheaper than shuttle.
Titan is a hell of a lot cheaper then a Shuttle. An expensive Titan
launch costs maybe $200M. A Shuttle launch is $500M to $1B just for
operational costs. If you want to add in amortization of the Shuttle
development and production costs, tack on anohter $100M or so.
>Don't tell me they have included costs of Titans in their plan. This
>really is getting very far fetched.
Why is planning for contingencies far fetched? Two Titan launches as a
backup would add $400M (half the cost of a single Shuttle flight) to
their budget. I don't know if they have planned for it or not but
compared to Freedom cost overruns, this would be minor indeed.
>+ It is only good engineering to have a backup for all the
>+methods and technology you plan to use. I always do.
>
>I agree 100 %, no 1000 %.
Then why did you defend Freedom's refusal to use other launchers which
are cheaper and have a better track record?
>It is the COST I am disputing, not so much the technical approach.
Again, can you be more specific? When McDonnell Douglas says they
can cluster Deltas to lift a 100K pound payload for $500M in three
years on what do you base your belief that it can't be done? When
Martin Marrietta makes similar claims about the Titan, why are they
wrong? What error in costing did they make?
>Wrong again. Titan is no cheaper than Shuttle.
See above. A Shuttle launch costs two to four times as much as a
Titan launch.
>Delta is too
>small too launch the LLNL package, so requires a complete
>redesign which cancels out the launch cost saving.
No it doesn't. They replace one of the module connectors to the
airlock with a docking port. Then put three modules in one Titan
and the other three modules (with the airlock) in another. Launch both
from separate Titan's and have them dock in orbit automatically
(the Soviets do this all the time).
Cost: An extra $400M for the launches and $1M or so to add the docking
port. We will also need to licence the automatic docking technology
from the Soviets but I figure that's in the noise level.
>A single Titan-IV launch gives you a tiny space station. More
>like a MIR than a Freedom.
See above. However, I should point out that there is a LLNL design
which goes up on a single Titan IV. It has only four modules.
>MIR is a lot cheaper than LLNL.
Could be but I would like to see that number justified better.
>The whole point of Freedom is its size and capability. LLNL gives
>all that away.
Nope. The LLNL design has three times the interior room of Freedom
(those inflatables get big). If you factor in room per $$ LLNL
is an even bigger winner. You could launch over 50 LLNL stations
for the cost of one Freedom and you could do it ten years before
Freedom is finished.
> + That means there is about a 50% chance that Freedom will never
> +be built. I consider a 50% chance of failure high risk.
>
>I will not dispute those probabilities, other people arrive at
>different numbers.
Think of how long the Shuttle has spent grounded. Over a third
of it's operational life has been spent unable to fly.
>The point of failure rate predictions like
>that is as an inspiration to work on the causes and fix them.
>That is why the number was generated. You don't just close your
>eyes and pray...! Reliability predictions are design tools, not
>forecasts of doom.
Agreed. What has NASA done to try and fix these numbers? NASA tried
to squash the first Fisher Price study. In fact, Fisher & Price where forced
to do the study 'on the sly' because NASA didn't want them to do it.
They also tried to sit on the second study as well. A source working
on Freedom was almost fired for providing accurate numbers on
weight and power. You yourself implied that you thought you would get
into trouble if you invited Dr. Wood down. When Congress pointed out
that NASA was assuming unrealistic launch rates for the Shuttle NASA
insisted it wasn't a problem. When OTA pointed out that there was a 50%
chance of loosing an orbiter during assembly, NASA shut its eyes and
said no changes where needed.
These problems are being worked. However, they are being worked because
Congress and not NASA is demanding it.
> The point is you are not able to quote probabilities for LLNL
>because there is no meaningful test data and hence no failure
>rate data, so a responsible engineer has to assume high failure
>rates until proven otherwise.
Which is why all the hardware will be tested in the exact
same manner as Apollo.
> For some reason which I do not
>understand, you seem to assume the failure rates for LLNL will be
>low.
I assume failure rates will be low because the design allows for major
failure without loss of crew. I'm sure their testing will find lots of
problems. That's what tests are for.
>Such "ass-backwards" thinking reminds me of Challenger
>"We got away with launching in cold weather twice, so this gives
>us confidence we will get away with it again...." It reflects a
>fundamental failure to understand the statistical basis of
>reliability predictions, and how to use the numbers generated.
Why don't you do a risk analysis of LLNL? Hell, write a white paper
to NASA and propose just that.
> >With one exception, I would like to see an ACRV added to
> >the program to get some independence from the Shuttle.
>+ How does that make them independant of the Shuttle?
>
>It makes the crew in orbit able to get back. It does not of
>course prevent the SSFP from re-entering.
Of course you are still dependant on Shuttle for assembly. How do
you justify that risk and expense? ACRV by no means makes you
independant of the Shuttle.
>+ If you want to know how much it costs, you do.
>
>I do not understand.
Since we are comparing two options it would be nice to know how much
each costs. The Shuttle is filled with hidden subsidies which makes
it harder to estimate the true cost of Freedom.
> >No, it does not. NASA is prohibited by Executive Order from
> >competing with the commercial ELV industry.
>+Ah but it does. It takes buisness away and costs the taxpayers
>+ billions.
>Are you saying Freedom should be launched on commercial ELV ?
Since it would provide more reliable transport at half the cost
I think it wold be a good idea, yes.
>I have no real problem with that. Would require several dozen
>Titan-3 launches.
But no more Titan IV flights than Shuttle flights. Each load going
up on a Titan will save the taxpayers half a billion $$. Now
if we wanted to add some risk we could build HL Delta. Each load
on this vehicle will save the taxpayers about 1.1 billion $$.
(HL Delta wold be better than Titan V because it would reduce the
risk of having all launchers grounded at the same time).
>The ELV industry does not have that capacity
>right now, would require them to change gear.
We can launch six Titan IV's per year from pad 41 at the Cape.
Each Titan launch can carry as much as the shuttle which only
flies about four times a year. This means we have 50% more
Titan capacity than Shuttle capacity.
Pad 40 could with modification also launch 6 Titan's per year.
giving us three times the Titan capacity as Shuttle capacity.
>But could save
>$$$. Will need to redesign all the Freedom launch packages
>however, big cost and schedule hit at this stage in the program.
It's going to happen. This years Freedom appropriation will force
NASA to do just that.
> + The Earth Station already has an airlock and facilities for
>+EVA.
>
>I do not mean the Earth station, I mean your mysterious "Delta-
>launched" manned spacecraft will have to be designed to support
>EVA in order to assemble the Earth station in the event the Earth
>Station fails to deploy automatically. That makes the Delta-
>spacecraft expensive. More than the entire HLV, to put it in
>perspective.
Do it with a Soyuz. They can be bought for 1/6 the cost of the
ACRV and we can do EVA from it.
>+ They already have. The House Senate Conference put back the DoE
>+ IR&D money. Some of this money will go to the Great Exploration
>+ for A detaild study. After that, there is good reason to expect
>+ them to start in earnest. This probram has the support of the
>+ Space Council (Quale recently told the head of LLNL that he
>+ wanted to see this program move forward).
>
>I am delighted. I just hope Quayle and the congress won't be
>too disappointed when they see the cost projections escalate
>and/or the schedule start slipping, happens to everybody sooner
>or later.
Not everybody (see Boeing). I hope you won't be too disappointed if
it works.
>+I would like to see a reason more detailed. What specific
>+engineering problems are there?
>
>I do not know. That is the problem in itself. They express
>extraordinary optimism about an unknown quantity.
No more optimism than the engineers of Apollo.
> > >>Any of those guys coming to southern California any time soon?
> > >+ Invite them down
> > >Hah ! I value my job too much.
> >+ That's an interesting remark. Why would inviting some
>+ speakers in get you fired?
>>I work for a Freedom contractor. LLNL represents the competition.
>+ Now that is another interesting remark. I routinely visit and
>+look at the work of my competition. I want to know what they are
>+ up to and if they have any good ideas.+
>+ Why are your bosses so afraid of evaluating the competition?
>We are of course VERY interested in the competition. My concern
>is giving competitiors visibility of our status, and providing
>them a forum. Meeting on neutral turf is AOK, which is why I
>ask: are they coming to So Cal.
They already have a forum in the Senate, on the Space Council,
and other places. Besides, why does giving them a forum bother you?
If you are correct in what you say won't you win the arguement?
> Also, my competitors do net give me the same access to their
>facilities as yours do !
Sure they do. Don't you go to trade shows or read your competitors research
papers?
Allen
--
+-------------------------------------------------------------------------+
| Allen W. Sherzer | What should man do but dare? |
| aws@iti.org | - Sir Gawain |
+-------------------------------------------------------------------------+
------------------------------
End of SPACE Digest V12 #520
*******************