Quote from: MarcAlain on 09/21/2015 01:49 amWell, not dozens. They'd likely have to double or triple their manufacturing capacities. Michoud can turn out a max of 4 per year, right?NASA has currently set up the production rate to support building slightly less than two per year, and with some additional funding they can get up to two per year.
Well, not dozens. They'd likely have to double or triple their manufacturing capacities. Michoud can turn out a max of 4 per year, right?
Quote from: MarcAlain on 09/20/2015 05:53 pmIs a hydrogen second stage even necessarily a bad thing when you're using solids as your main thrust/first stage?What would the performance benefits be if they used similar thrust RP1 engines on the main liquid/second stage?RP-1 is the best choice there is for a first stage due to its ISP-DENSITY. RP-1 would have been far better for either an SLS first stage or for its boosters, and Hydrogen is the best choice for an upper stage.
Is a hydrogen second stage even necessarily a bad thing when you're using solids as your main thrust/first stage?What would the performance benefits be if they used similar thrust RP1 engines on the main liquid/second stage?
We can determine fairly closely the marginal cost of a Delta-IV Heavy DOD launch from the 2014 block-buy since this block buy was pure marginal cost. The Engineering, launch operations, and Infrastructure had already been paid for. You had 36 cores purchased for $4.3 Billion, which equals about $120 Million a core so we can infer about $360 Million for the incremental cost of a Delta-IV Heavy.
Quote from: 93143 on 09/20/2015 07:36 pmApples to oranges. Unless you're trying to tell me that the Delta IV Heavy price is pure marginal cost, No, one for one. An additional Delta IV Heavy price would be almost pure marginal cost. Anyways, if it isn't, it still makes the case that it is way cheaper than SLS.
Apples to oranges. Unless you're trying to tell me that the Delta IV Heavy price is pure marginal cost,
Stay on topic guys. It gets really boring really fast when people start comparing SLS to Delta IV.
Quote from: Brovane on 09/21/2015 01:51 amWe can determine fairly closely the marginal cost of a Delta-IV Heavy DOD launch from the 2014 block-buy since this block buy was pure marginal cost. The Engineering, launch operations, and Infrastructure had already been paid for. You had 36 cores purchased for $4.3 Billion, which equals about $120 Million a core so we can infer about $360 Million for the incremental cost of a Delta-IV Heavy.To my knowledge the AF buys launches, not cores.
NASA has currently set up the production rate to support building slightly less than two per year, and with some additional funding they can get up to two per year.
At the full production rate capability of 2 every year the per launch cost would be from $.9B to $1B or $3.6B to $4B every 2 years <$2B more for three more flights in the same period or additional budget required of $600M per flight!!!!
Quote from: Coastal Ron on 09/21/2015 03:32 amNASA has currently set up the production rate to support building slightly less than two per year, and with some additional funding they can get up to two per year.I think that's the interesting "marginal cost" to look at. Assume the anticipated rate is "three every two years." What is the incremental cost to reach "four every two years?" I think this is similar to what oldAtlas_Eguy attempts to estimate, with the result of $600M.
what the effects are of increasing or decreasing production rates.
Can anyone defend talk of a single "marginal cost" value given the actual anticipated flight rate?
This particular contract was able to leverage being a continuation of previous contracts, so the workforce was already in place and stable, the supply chain was mature, and costs were well known.The SLS SRM's, which are 5-segment, and new designs, can only be more expensive, not less than, what the Shuttle program was paying.
QuoteThe efficiency of a high production rate is simply that you get a lot of work out of your existing facilities and personnel. Lowering the rate doesn't increase the marginal cost of a unit very much; what it does is increase that unit's share of the fixed costs.Lowering the production rate can affect costs very much, and what we don't know is where the cost inflection point is for SLS production. It may not be until it reaches something like 4-6 per year (~ Shuttle rate). NASA is quite proud of how efficient the tooling is for the SLS, which is good for touch labor costs, but overhead and other sustaining costs are going to be significant at low production rates, both for Boeing and every other major contractor for SLS.
The efficiency of a high production rate is simply that you get a lot of work out of your existing facilities and personnel. Lowering the rate doesn't increase the marginal cost of a unit very much; what it does is increase that unit's share of the fixed costs.
QuoteBut SLS is being set up for two units per year at steady state; you can only take such a scenario so far before it becomes more expensive than just using the Shuttle infrastructure as it stood, and $1B marginal cost per launch is past that point. I don't see this as a serious possibility, not at the scale you people are talking about.The Shuttle shouldn't enter into this, since it was not a comparable transportation system, despite sharing some design elements.
But SLS is being set up for two units per year at steady state; you can only take such a scenario so far before it becomes more expensive than just using the Shuttle infrastructure as it stood, and $1B marginal cost per launch is past that point. I don't see this as a serious possibility, not at the scale you people are talking about.
The key though is that no upsized transportation system should be built until the existing transportation has been maxed out, and the new system is going to provide increased capacity for a known customer demand.
You guys aren't getting this. You can't calculate marginal cost from expenditure breakouts.
It isn't just the primary contractor that has fixed costs and variable costs. Every subcontractor and part supplier does too. And if you're a substantial fraction of someone else's business, your purchase rate will strongly affect their economies of scale, changing the price of the item(s).
Please note that with the EELV program, it isn't that non-DoD customers are charged incremental cost, however that's calculated. They're charged normally, and the DoD is reimbursed for the fixed cost thus defrayed. And apparently the amount of the reimbursement has been controversial in the past; the DoD was still complaining even after it tripled...
QuoteCan anyone defend talk of a single "marginal cost" value given the actual anticipated flight rate?Sorta, yeah. See the previously-linked chart from DIRECT for J-246. Goes all the way from one flight per year to 16.
But marginal cost itself is not strongly dependent on flight rate.
Quote from: Coastal Ron on 09/21/2015 01:01 amThis particular contract was able to leverage being a continuation of previous contracts, so the workforce was already in place and stable, the supply chain was mature, and costs were well known.The SLS SRM's, which are 5-segment, and new designs, can only be more expensive, not less than, what the Shuttle program was paying.So now you're trying to lump DDT&E into the marginal cost?
The SLS boosters have been reworked for affordability. If I recall correctly they slashed the required man-hours per segment by nearly 40%.
QuoteThe key though is that no upsized transportation system should be built until the existing transportation has been maxed out, and the new system is going to provide increased capacity for a known customer demand.NASA is not a commercial launch provider, that has to respond to customer demands.
I can appreciate the detailed work the DIRECT folks did, but their assumptions were based on the continued use of the Shuttle supply chain. The Shuttle supply chain ended, so the SLS can't leverage that, even for the SRM's. Plus the SLS is a completely different design, which greatly affects projected costs.
There is a unique cost associated with every single part, and it is affected by your consumption rate (i.e. flight rate) and how much (and how often) you are buying for each part. Those are major cost drivers for suppliers, as well as Boeing who is doing fabrication and final assembly.
the marginal cost off the J-246 plot from 2009 is almost identical to the marginal cost I backed out of the ESD Integration SLS estimates from 2011
If you actually plot the total program cost vs. flight rate for a large launcher program with a reasonably well-defined launch schedule, the slope changes fairly slowly. That slope is the marginal cost I'm talking about.
All that consumption rate dependency is due largely (not entirely, but in the main) to fixed cost; you buy less of something, or buy more sporadically, and the supplier's overhead becomes a bigger portion of the total, so the price per unit goes up. This is a particularly strong effect if you're the sole customer for an item, since you have to eat the supplier's entire fixed cost regardless of how many you buy. But it doesn't matter whose fixed cost it is; it's still fixed cost.If you actually plot the total program cost vs. flight rate for a large launcher program with a reasonably well-defined launch schedule, the slope changes fairly slowly. That slope is the marginal cost I'm talking about.
SLS is not a "completely different design". It's an inline Shuttle-derived launch vehicle that's actually quite similar to Jupiter, but a little bigger, with modernizations and cost reductions.
You can't pretend Delta IV is a closer analogue than Jupiter.
I just got through saying exactly that, except that I'm not trying to redefine "marginal cost" to mean "purchase price".
Based on what you are showing in your graphs, the Jupiter-246 showed a marginal cost of ~320 Million and you expect the SLS to be in the same range? Since the SLS is also Shuttle derived hardware.
We'll have to wait until Congress allows NASA to buy SLS production material before we find out what the real prices are.
Quote from: Brovane on 09/22/2015 07:37 pmBased on what you are showing in your graphs, the Jupiter-246 showed a marginal cost of ~320 Million and you expect the SLS to be in the same range? Since the SLS is also Shuttle derived hardware.Basically, yeah.The leaked ESD Integration document from 2011 seems to back me up. Cases #4a and #4b have a few years at the end where development is over and the launch rate is steady (and plainly continues well into the future); aside from the "In Space Elements Wedge", the only difference is that #4a has an extra "70-ton" SLS launch every year. The cost difference, summing SLS and ground systems, is $414M in 2023, $426M in 2024, and $437M in 2025.The inflation rate in NASA New Start 2010 (use in FY11) is 2.6% from 2012 to the end of the table, except for 2013 and 2017 in which it is 2.7%. This matches nicely with the rate of inflation apparent in the ESD Integration data. Using NASA New Start 2010 (use in FY11) to deflate to 2011 and NASA New Start 2014 (use in FY15) to inflate it back to 2015 results in a difference of about $330M.$330M is pretty darn close to $320M considering the differences between the two vehicle designs, and the fact that one is measured off a graph and inflated by six years, while the other is deflated by fourteen years and reinflated by four...Considering how close the deflated fixed costs are to the DIRECT estimate, as well as the fact that the estimates are explicitly said to be based on STS and CxP, I suspect there wasn't much modernization or cost reduction in the numbers. If this is true, the actual vehicle should end up less expensive overall on an ongoing basis, though not necessarily on a marginal basis.
Well, with design differences and tech changes and the optimization of production for a low flight rate, there's probably a good deal of wiggle room in that number.Personally, I think a full-up Block 1B should have a marginal cost to go from one flight per year to two of somewhere between $300M and $400M. Higher and lower are both possible. But I would be very surprised if it got up near $1B; that seems downright unreasonable...