Lawmakers Seek To Curb EELV Block Buy

[FinalFrontier]

If SpaceX had:

A. A proven track record

B. Didn't blow up engines in flight.

I might be ok with this. But because of that, and because of the far longer and more extensive record of the EELV vehicles this strikes me as nothing more then politics at their worst intruding into things......again.

Seems like an obama gambit to put "new space" ahead again. I don't mind competition, especially when SpaceX is involved, but in this case they just don't have the credentials to warrant this, IMO.


We'll just have to see how this works out.

[john smith 19]

If SpaceX had:

A. A proven track record

B. Didn't blow up engines in flight.

I might be ok with this. But because of that, and because of the far longer and more extensive record of the EELV vehicles this strikes me as nothing more then politics at their worst intruding into things......again.

Seems like an obama gambit to put "new space" ahead again. I don't mind competition, especially when SpaceX is involved, but in this case they just don't have the credentials to warrant this, IMO.


We'll just have to see how this works out.

I'd suggest you read GAO-11-641

That block buy is based on "evidence" of an erosion of the US launch vehicle infrastructure that is *highly* questionable, over a period which people cannot *explain* their reasons for selecting.

As for Spacex's track record they got 3 F9's where they had to go and that *partial* engine failure probably worries people who *buy* launches (which is not the same group of people who *need* the launches) but probably impressed the users (who would be primary payload users).

Then there is the topic of "Mission assurance". DoD maintain it is *absolutely* needed to maintain their record of launch success, but ULA cannot explain what it *is*, it is too expensive to break out its costs from the rest of the mfg process but it's going up (a lot).

Does this sounds *pretty* suspicious to you?

[edit. I'd also suggest GAO-12-822 which notes the current policy is called "Buy 3" and breaks the contract into a fixed price section for the hardware and a cost plus section for this "mission assurance" and other stuff. IE *whatever* it costs you'll pay it.]

Even if DoD/NRO *began* the process of putting F9 on the selection process it would likely be at least a year before it would be accepted for use as an LV. By the time it would have passed analysis *either* there will be a substantial track record of successful F9 launches *or* their launch record will be so patchy as to make the idea of using them laughable.

[Robert Thompson]

Just curious about what the military is really buying when they are paying for "can't fail" levels of reliability in launchers. Obviously the launchers can in fact fail no matter how many dump truck loads of cash they burn to propitiate the rocket gods.

[Jim]

Not a set of standards, but the ability to see all test data and analyze it independently.  Not just for the vehicles that they are flying but for the whole fleet.

I think that's a better description than the GAO got out of ULA on the subject. 

It sounds like what you're mostly buying is an assurance that a) the supplier will take detailed records (of both tests but also parts mfg and assembly) and b) Storing all that data in case you want to analyze it later

Which sounds a pretty expensive bill for things the business *should* be doing to improve its products anyway + archiving it just in case.

Back when the cutting edge of document storage was manually copied microfilm and test results really *were* spools of papers from chart recorders that *may* have been defensible. 

But 10s of $m ? Either ULA are doing *very* sophisticated processing (lots of human intervention and processing time) on this data before putting it in storage or the mfg/test of an LV generates literally TB of data.

Either (both) is possible but they just seem *very* unlikely.

No, it is also the additional human interaction with the customer.  It takes many addtional MTS to deal with the customer requests.

[john smith 19]

No, it is also the additional human interaction with the customer.  It takes many addtional MTS to deal with the customer requests.

This just seems even more contradictory. The *impression* is of a structured, *mature* process which is carried out throughout the the procurement pretty much up to launch (at least).

Which suggests *standardized* meeting and data formats and probably support tools to assist them. Also as this a FAR process (mission assurance is a cost plus element of the contract according to the GOA) doesn't the govt have substantial access to the contractors database by default already?

Your description makes it sound *much* more ad hoc, with govt officials rounding up people and asking them for data more or less at random, rather than getting it themselves given their access to the contractor systems.

That does not sound like a programme that could be classed as in "sustainment." I'm not being snippy I'm genuinely fascinated because its costs seem to be seriously above inflation and rising.

[Jim]

Which suggests *standardized* meeting and data formats and probably support tools to assist them. Also as this a FAR process (mission assurance is a cost plus element of the contract according to the GOA) doesn't the govt have substantial access to the contractors database by default already?

Your description makes it sound *much* more ad hoc, with govt officials rounding up people and asking them for data more or less at random, rather than getting it themselves given their access to the contractor systems.

It isn't data in databases nor "govt officials".  The gov't gets all the flight data, standard analyses and participates in standard reviews.  It is more about the off nominal stuff were the engineers asking why did this happen or what are you going to do fix this or prevent this from happening.

Also, still providing access to databases and such still requires manpower.

[john smith 19]

It isn't data in databases nor "govt officials".  The gov't gets all the flight data, standard analyses and participates in standard reviews.  It is more about the off nominal stuff were the engineers asking why did this happen or what are you going to do fix this or prevent this from happening.

Also, still providing access to databases and such still requires manpower.

I'm still have trouble understanding the *scale* of these costs.
An engineer working a 40 hour week is 2080 hrs a year.
The Bureau of Labor Statistics reckons about $50 an hour salary for an aerospace engineer.
http://www.bls.gov/oes/current/oes172011.htm
But IIRC the rule of thumb is all support expenses make their costs *double* their salary.
So 2080hs @ $100/hr gives $208k/yr
so $100m buys 480 *full* time engineers to work on nothing but  "mission assurance"
However if I'm 100% out that $100m would only buy 240 engineers.
Which might be OK except Spacex build 80% of their launch vehicles with 1600 staff *total*.

While I can understand *payloads* are different (possibly radically so) the core *functions* of the LV's (and their major sub assemblies) are essentially the *same*. I would therefor expect that just as the LV's are "modular" in terms of engines, stages, strapons so each major component has a standard chunk of "mission assurance" that is assembled with it as the LV for a particular payload is assembled.

[Jim]

ULA has around 3500 people for two launch sites, 5 pads, 3 launch vehicle families and integration efforts for around 30 different spacecraft.

[Lurker Steve]

It isn't data in databases nor "govt officials".  The gov't gets all the flight data, standard analyses and participates in standard reviews.  It is more about the off nominal stuff were the engineers asking why did this happen or what are you going to do fix this or prevent this from happening.

Also, still providing access to databases and such still requires manpower.

I'm still have trouble understanding the *scale* of these costs.
An engineer working a 40 hour week is 2080 hrs a year.
The Bureau of Labor Statistics reckons about $50 an hour salary for an aerospace engineer.
http://www.bls.gov/oes/current/oes172011.htm
But IIRC the rule of thumb is all support expenses make their costs *double* their salary.
So 2080hs @ $100/hr gives $208k/yr
so $100m buys 480 *full* time engineers to work on nothing but  "mission assurance"
However if I'm 100% out that $100m would only buy 240 engineers.
Which might be OK except Spacex build 80% of their launch vehicles with 1600 staff *total*.

While I can understand *payloads* are different (possibly radically so) the core *functions* of the LV's (and their major sub assemblies) are essentially the *same*. I would therefor expect that just as the LV's are "modular" in terms of engines, stages, strapons so each major component has a standard chunk of "mission assurance" that is assembled with it as the LV for a particular payload is assembled.

Engineers straight out of college may have a salary around 50K/yr, but that can't be the industry average. Besides, that's just their salary. You forget about the benefits cost, plus the standard overhead costs of having an employee, like office space, office equipment (including computers, software, etc), the employers share of payroll taxes, etc.

Besides, just running the quality operation itself is expensive. What do you think it costs every time SpaceX test fires an engine in Texas ? I'm sure it's not gas it up and go. And even with their quality processes, they aren't finding all the manufacturing / assembly flaws. You could argue that SpaceX needs to spend more on mission assurance.

[john smith 19]

I'm still have trouble understanding the *scale* of these costs.
An engineer working a 40 hour week is 2080 hrs a year.

The Bureau of Labor Statistics reckons about $50 an hour salary for an aerospace engineer.

http://www.bls.gov/oes/current/oes172011.htm

But IIRC the rule of thumb is all support expenses make their costs *double* their salary.

So 2080hs @ $100/hr gives $208k/yr

so $100m buys 480 *full* time engineers to work on nothing but  "mission assurance"
However if I'm 100% out that $100m would only buy 240 engineers.
Which might be OK except Spacex build 80% of their launch vehicles with 1600 staff *total*.

While I can understand *payloads* are different (possibly radically so) the core *functions* of the LV's (and their major sub assemblies) are essentially the *same*. I would therefor expect that just as the LV's are "modular" in terms of engines, stages, strapons so each major component has a standard chunk of "mission assurance" that is assembled with it as the LV for a particular payload is assembled.

Engineers straight out of college may have a salary around 50K/yr, but that can't be the industry average. Besides, that's just their salary. You forget about the benefits cost, plus the standard overhead costs of having an employee, like office space, office equipment (including computers, software, etc), the employers share of payroll taxes, etc.

Besides, just running the quality operation itself is expensive. What do you think it costs every time SpaceX test fires an engine in Texas ? I'm sure it's not gas it up and go. And even with their quality processes, they aren't finding all the manufacturing / assembly flaws. You could argue that SpaceX needs to spend more on mission assurance.

Re-read my post. Specifically the source of my figures and my adjustment for benefit costs. Musk has stated that the *total* propellant bill is about $200k to orbit. c0.3% of launch price.

[JohnFornaro]

But IIRC the rule of thumb is all support expenses make their costs *double* their salary.
So 2080hs @ $100/hr gives $208k/yr
so $100m buys 480 *full* time engineers to work on nothing but  "mission assurance"
However if I'm 100% out that $100m would only buy 240 engineers.

Which would be $200/hr.

Engineers straight out of college may have a salary around 50K/yr, but that can't be the industry average. Besides, that's just their salary. You forget about the benefits cost, plus the standard overhead costs of having an employee, like office space, office equipment (including computers, software, etc), the employers share of payroll taxes, etc.

What the problem is?  Seems like he made an effort to suggest an industry average, already including the costs you mention.

[Lurker Steve]

But IIRC the rule of thumb is all support expenses make their costs *double* their salary.
So 2080hs @ $100/hr gives $208k/yr
so $100m buys 480 *full* time engineers to work on nothing but  "mission assurance"
However if I'm 100% out that $100m would only buy 240 engineers.

Which would be $200/hr.

Engineers straight out of college may have a salary around 50K/yr, but that can't be the industry average. Besides, that's just their salary. You forget about the benefits cost, plus the standard overhead costs of having an employee, like office space, office equipment (including computers, software, etc), the employers share of payroll taxes, etc.

What the problem is?  Seems like he made an effort to suggest an industry average, already including the costs you mention.

I don't know the aerospace industry, but you certainly couldn't build a staff of software, electrical, or mechanical engineers on that "average" salary. That's barely a starting salary in the Chicago area. Once those engineers get 5-10 years of experience, they are going to be looking for something in the 75-80K salary range, if not more. I would expect California-based companies to pay even higher salaries, given the higher cost of housing and other items.

My company uses a standard engineering cost of around 125-135/per man hour. That just a fully loaded estimate for covering salary / benefits / and the facility overhead.

[john smith 19]

What the problem is?  Seems like he made an effort to suggest an industry average, already including the costs you mention.

The source of my numbers was

http://www.bls.gov/oes/current/oes172011.htm#top

but if we go with the cost for the *top* 10% of aerospace engineers
$71.06/hr. *doubling* that to account for benefits and support costs gives $142.12. A 40 hr week at that rate gives $295609.6 [edit] per year.

so (assuming you're team is *exclusively* composed of aerospace engineers at the top 10% of the industry pay scale on *nothing* but "mission assurance" $100m gets you 338 engineers a year.

Caveats. I'm not an American so I cannot confirm that the "fully burdened" cost of an employee is 2x their annual salary. I can't remember where I saw that figure as a rule of thumb and I could be wrong. Increasing the multiplier (the direction I would *expect* it to go) would reduce the number of engineers.

but that is still one *hell* of a lot of variance between launches needing *all* this one off attention. 

[aero]

but if we go with the cost for the *top* 10% of aerospace engineers
$71.06/hr. *doubling* that to account for benefits and support costs gives $142.12. A 40 hr week at that rate gives $295609.6

Maybe this was explained above, but my calculator gives $71.06 x 2 x 40 = $5684.80 so I guess you mean, per year.

[oldAtlas_Eguy]

ULA has around 3500 people for two launch sites, 5 pads, 3 launch vehicle families and integration efforts for around 30 different spacecraft.

The comment here is that ULA recieves a yearly $650M launch assurance contract that would cover the cost of all of ULA's employees salaries and overhead. This ensures ULA dosn't have to fire anybody because it isn't launching often enough, or even at all. ULA is basiclly a cash basis operated company, it must maintain a positive profit. There isn't provision for survival through a year that has more costs than revenues. ULA dosn't get to keep and reinvest its profits.

The charges per launch vehicle would be for those purchases for parts and services that ULA pays someone else to provide.

[sdsds]

There isn't provision for survival through a year that has more costs than revenues. ULA dosn't get to keep and reinvest its profits.

You mean there's no automatic provision in such a situation, right? Is it wrong assume that if costs exceeded revenues, ULA's parent companies would decide to finance its operations?

[john smith 19]

The comment here is that ULA recieves a yearly $650M launch assurance contract that would cover the cost of all of ULA's employees salaries and overhead. This ensures ULA dosn't have to fire anybody because it isn't launching often enough, or even at all. ULA is basiclly a cash basis operated company, it must maintain a positive profit. There isn't provision for survival through a year that has more costs than revenues. ULA dosn't get to keep and reinvest its profits.

The charges per launch vehicle would be for those purchases for parts and services that ULA pays someone else to provide.

Note that would be *launch* assurance and is basically to keep the pads open and the staff on the payroll. I think this used to be called the "Assured access" payment or what Elon Musk refers to as the ULA "subsidy." IIRC it's nearer $1Bn a year.

*mission* assurance is the per mission costs to improve mission reliability which is broken out as a separate cost plus contract (apart from the LV purchase, which is fixed price) and whose definition seems somewhat vague. Except it's getting bigger (although the near term manifest is meant to include a number of 2nd and 3rd launches of existing payloads, which *should* be cheaper to do MA on)

[Jim]

*mission* assurance is the per mission costs to improve mission reliability which is broken out as a separate cost plus contract (apart from the LV purchase, which is fixed price)

No, it is included in the price of the launch service contract.

[JohnFornaro]

My company uses a standard engineering cost of around 125-135/per man hour. That just a fully loaded estimate for covering salary / benefits / and the facility overhead.

Fine, but I'm still not getting it.

BTW, when I do this sort of figuring, I use 2000 hours; it includes a two week vacation, and is easier for me to multiply by 2.

Anyhow, John Smith was starting out at $50/hr, which is a hundred grand a year; doubled to two hundred grand, in the ballpark of your two hundred fifty to two hundred seventy grand.

I took his number to roughly mean an average or maybe a median cost, some higher, some lower.  I took his term "all support services" to be roughly equivalent to our term "fully burdened".

But IIRC the rule of thumb is all support expenses make their costs *double* their salary.
So 2080hs @ $100/hr gives $208k/yr

Even then, you think his estimate is too low?

[Lurker Steve]

My company uses a standard engineering cost of around 125-135/per man hour. That just a fully loaded estimate for covering salary / benefits / and the facility overhead.

Fine, but I'm still not getting it.

BTW, when I do this sort of figuring, I use 2000 hours; it includes a two week vacation, and is easier for me to multiply by 2.

Anyhow, John Smith was starting out at $50/hr, which is a hundred grand a year; doubled to two hundred grand, in the ballpark of your two hundred fifty to two hundred seventy grand.

I took his number to roughly mean an average or maybe a median cost, some higher, some lower.  I took his term "all support services" to be roughly equivalent to our term "fully burdened".

But IIRC the rule of thumb is all support expenses make their costs *double* their salary.
So 2080hs @ $100/hr gives $208k/yr

Even then, you think his estimate is too low?

Yes, his estimate is too low, but no need to derail the discussion further.

I'm sure the "mission assurance" funding does more than just pay salaries.