Wednesday, July 24, 2013

Steel is Cheap and Air is Free

I see this claim being made repeatedly during discussions about ship design and construction.  It's a common belief among people commenting on design and construction but being common doesn’t make it true.  On a relative basis, it may have an element of truth to it.  A quantity of steel may be cheaper than an Aegis array, for example.  However, steel is not free and neither is air – not by a longshot.  If they were, we'd just build every ship to be the size of a battleship and give them all two feet of armor.  Steel has a significant cost and so does air.  The Navy and the shipbuilders recognize this even if the blogosphere does not.

Consider the cost of the LCS.  The LCS, because of the purchase contract, gives us a unique insight into the cost of the bare hull (the steel and air portion of the ship).  I’ve covered this repeatedly so you should know what’s coming.  The contracted cost of the LCS, currently around $500M per ship, is just for the bare hull.  All the electronics, sensors, radars, weapons, computers, various fittings, etc. are supplied from a separate government account line that is not included in the contracted amount.  While the government supplied equipment costs have never been released, a seemingly reasonable estimate would be $200M-$300M.  That pushes the cost of the LCS to $700M-$800M and that’s without a module.  A module will push the ship cost to up near $1B.  Thus, we see that the bare hull, mostly just cheap steel and free air, represents a half to two-thirds of the total cost – hardly cheap and free!

Remember that there are secondary costs associated with steel, as well.  Every pound of steel must be propelled through the water which means that every additional pound of steel requires additional horsepower which means bigger engines which, in turn, means more bunkerage.  More crew is needed to maintain the larger engines and more crew requires more berthing, mess, laundry, etc.

Before we wrap this up, let’s touch on the “air is free” part of the saying.  Now, I understand that it makes a catchy slogan but we need to remember that air is bounded by steel bulkheads, decks, and overheads which cost money.  Hence, air is not free.  Further, that “air” must be heated, cooled, lighted, cleaned, maintained, manned, powered, and propelled through the water (bigger engines for every cubic foot of “air”).  The air above deck may be free but the air below deck costs money!

I’ve seen proposals to enlarge the LCS, adding 20% - 50% length, to accommodate more weapons and make it a “warship”.  People brush off the resulting costs with the “steel is cheap and air is free” comment.  Well, 20% - 50% more hull would result in a 20% - 50% increase in the hull price of $500M which would give a hull cost of $600M - $750M.  That’s not cheap or free. 

Many use the “cheap and free” argument to justify frigates the size of Burkes;  in fact, many want to use the Burke hull as the basis for a frigate because the “steel is cheap and air is free”.  We’ve just demonstrated that steel is not cheap and air is not free.  A Burke size frigate would be unaffordable in the numbers required for an effective frigate class.

As we continue to discuss ships costs, let’s bear in mind that steel is far from cheap and air is nowhere near free.  Factor that in to your proposals.  Carry on!


  1. Good points. I hadn't thought to attack that assumption.

    I do wonder though, if this is just a US Navy issue.

    From this website:

    I got this:

    "The AUD 100 million Benchijigua Express, which was placed into service in the Canary Islands, also would be called the largest aluminum ship ever built upon its launching in September 2004. Its trimaran design would be the basis for a new fast vessel for the U.S. military."

    It cost $100 million in Austrailian Dollars in '04? Now 9 years later a ship that is within, IIRC, a meter, costs the Navy $500 million for just the bloody hull? Inflation doesn't cover that cost increase.

    Something doesn't add up. I don't know what it is, but I wonder what sort of inefficiencies are there? Is this added cost due to concurrence? If they went in with a straight design and build it will it get cheaper?

    Now I'm *very* curious to find out what the simple hull cost of a Burke is. Of course, that's not *quite* fair. I think you can roughly compare an LCS build to commercial standards and a little more to a similarly sized ship upon which it is based. There is no such similar comparison for a lvl III warship, other than to other nations lvl III warships.

    1. Jim, that's why this blog exists - to analyze and question and to apply logic to issues rather than assumptions and sayings!

      I don't know anything about the Express nor do I know what the relative purchasing power of an AUS dollar is. That aside, even the LCS which was built to a commercial-ish standard had much more in the way of compartmentation, redundancy, firefighting, electronics, big honkin' engines, etc. than would be found in a typical commercial vessel. Consider that as you compare commercial and naval semi-commercial designs.

      I believe the Burkes (and all Navy ships) are now contracted as the LCS is, meaning that the contract price is for essentially the hull only. I suspect that the published costs for the Burkes are just the hull. If so, WOW! That's expensive steel and air. I'm stopping just short of making that an absolute definitive statement because I haven't seen the contract language or read an absolute authoritative confirmation of it, as I have with the LCS. However, everything I've read suggests that is the case.

    2. Taking account of the difference in the AUD and adjusting to 'today's dollars', the commercial hull would cost $150 million.

      Keep in mind that the Benchijigua Express was powered by 4 diesels and the LCS designs employ 2 gas turbines that individually roughly equal the power of the entire Benchijigua Express engine room, and you can see that massive coin went into the LCS power plant. And that does not account for electric power and hotel services (H2O etc.). Then tack on the physical aviation facilities flight deck etc, and that was another major driver in costs. And even lightly constructed, the addition of USN specific hardware, modifications for the few weapons (mounting the 14,000 kg pop-gun to a deck is no trivial matter).

      I actually see where the basic hull cost growth came from.


    3. GAB, quite right and well summarized regarding the cost differential between the commercial vessel and LCS. There are additional differences that drive the LCS cost up as well. For instance, the LCS has entire sections of the ship devoted to storing, handling, maintaining, and launching/recovering remote unmanned vehicles which the commercial ship probably does not. We can go on listing the things that need to be added to a warship to make it a warship but Jim undoutedly now gets the point about commercial/warship cost comparisons. The difference would be even more pronounced if the LCS was a true warship!

    4. ComNavOps,

      What is really striking is that the Benchijigua Express is powered by four MTU 8000 diesels (32 Kw) and has a reported speed of 42 kts - the Independence retains two of the MTU 8000s, but adds two LM2500s to make 44 knots! I would love to hear the back story behind that decision. The MTU diesel is a good engine (at least the German version is); why on earth did the Navy slap two reliable, but expensive LM2500s on a platform that was already capable of 40+ knots?

      Sadly, the LCS failed to take advantage of commercial deck handling technology, which is generally superior to military counterparts. Active heave compensated cranes and A-frames head the list.


    5. GAB, you make a great observation about the Navy failing to use state of the art commercial cranes and such. Even on a smaller scale, watching the sailors try to launch the small remote vehicles from the LCS is humorous but painful to watch. The launch system appears to have been designed as a Kindergarten class project.

      The Navy could learn a lot from commercial maritime practices.

      Great comment!


    A year old
    249ft long

    Steel is cheap, the LCS is a rip off.

    1. TrT, c'mon now, be fair. That example you cite is essentially a barge, it's a sale price of a used ship built in 2012, it's outfitted for a crew of 15, it has nothing that a warship (even the LCS) would have in the way of electronics, sensors, weapons, redundancy, damage control, etc. That ship doesn't even remotely compare to a warship.

    2. But as you said, the LCS cost is just a bare hull, without fancy radar and engines and such.
      The LCS is bigger, and better built, but its 50% bigger, and its not 50,000% better built!

      The UKs type 45s are rocking in at not much more than a billion dollars each excluding R&D

      I agree that the point is true to a point, steel is a cost, both in buying it, building it and propelling it, but not that much.
      Maybe give a yacht maker who makes steel yachts a ring and see if they would be prepared to comment on thicker steel? Bering Yachts do steel.

  3. The Average FY2014 cost of $448 million is for the whole ship, less the mission modules. Electronics and weapons (MGs, 57mm, RAM, etc) are included. The reason that Mission Modules are separate is that there are not bought on a 1:1 basis.

    Here is the FY2014 USN budget (cost differences between the LM and GD variants are not listed).

    The LCS costs are covered under the Shipbuilding and Conversion, Navy section

    Mission Modules are covered under the Other Procurement, BA 1 section

    1. SpudmanWP, I've covered this and documented it repeatedly. Check the older posts, the CRS, GAO, and similar reports. Look at the Information Dissemination or Defense Daily articles. Reread the Jun 2013 "Government Supplied Equipment" post (though that deals mainly with other ship classes) and you'll see that the Navy standard practice is to supply weapons, electronics, etc. from a separate account line that does not get captured in the usual cost accounting. The FY14 Highlight Book does not go into sufficient detail to see this.

    2. I'll show you mine if you show me yours.

      You obviously did not look at my links as I looked at the detailed pricing, not the "highlights" book. The budget doc clearly states (included in the $448 pricetag):

      "Provides for the design, construction, integration and testing of the Littoral Combat Ship (LCS), including Ordnance, Government Furnished Equipment (GFE), and includes Program Office and change order costs. "

      The detail section of the LCS shows the cost breakdown and includes the electronics, weapons (baseline RAM, MGs, and 57mm), etc.

      If you think that they are hiding costs elsewhere, besides the Mission Modules, please show me the "Line Item" they are covered under.

    3. Spudman, as one example of many references, here's a quote from the Defense Industry Daily LCS summary document. DID is the premier defense industry procurement website.

      "A $715 million contract modification to Lockheed Martin Corporation will build LCS 9 Little Rock and LCS 11 Sioux City at Marinetter Marine Corporation in Marinette, WI. ... Amounts are based on the competitive, LCS dual block buy contracts (vid. Dec 29/10), and factor in approved FY 2010-11 change orders to the designs. Note that these contracts cover just the base sea frames, and installation of separately-purchased "government furnished equipment" like weapons, etc."

      The key phrase is "separately-purchased "government furnished equipment" like weapons, etc." Thus, weapons, sensors, electronics, etc. are purchased from a separate government account line. The manuf simply installs the equipment. Also, the contract amounts are partial and do not include previously funded elements of the contract, advanced procurement funds, post-delivery outfitting and availability, etc.

    4. Okay. I give up. I've skimmed probably 80 pages of .pdf documents and not come close to getting a price schedule for building the DDG 51's. SO I can't figure out current hull costs. I've seen total program costs to date, but that's it. I don't have the time to just keep perusing. Sorry guys!

    5. SpudmanWP, sorry it's taken me awhile to get back to you on this. I've been busy with summer activities! I think you're confusing the Navy’s budget and contract process. Not surprising since it really is confusing. Let’s see if I can provide some understanding. The budget document you’ve referred to is just that – a budget. From that budget various contracts are issued by the Navy, as needed, to various contractors. Specificially, the Navy issues contracts to the two LCS manfs to build the ships. The contracts do equal the full budgeted amounts. For instance, Austal does not manufacture RAMs (I did, indeed misread the dollar amount in my first quick scan – thanks for pointing that out!). A separate contract is issued to another manf for that.
      Here are actual construction contract amounts issued to the two manufacturers that I’ve been able to verify with the budgeted unit cost in parentheses for comparison.

      LCS6 $432M (?)
      LCS7 $376M (?)
      LCS8 $368M (?)
      LCS9 $357M ($458)
      LCS10 $345M ($458)
      LCS11 $357M ($458)
      LCS12 $345M ($458)
      LCS13 $348M ($446)
      LCS14 $341M ($446)
      LCS15 $348M ($446)
      LCS16 $341M ($446)

      You see the difference between contract and budget in the numbers? The contract is just for the “bare” hull, without most electronics, weapons, sensors, etc. Further, even the budget amounts are understated. For example, the 57mm gun and the radars, among other items do not appear to be included and are provided from separate account lines that I’ve never seen.
      The contracts contain clauses that provide for the manf and the Navy to split cost overruns up to a certain dollar amount and then are absorbed by the Navy after that. The contracts typically seem to wind up around $400-$500 with overruns factored in (the earlier contracts certainly had major overruns; I’m far less sure about the more recent ones). Overrun information is quite sketchy and hard to come by so I can’t quote exact numbers, only estimates from the bits and pieces I’ve come across.

      So, remember that the publicly stated contract amounts are for the hull. Separate contracts are issued for the manufacture of computers, electronics, weapons, etc. and some equipment such as the gun, radars, and whatnot are totally unaccounted for in the budget but, again, are handled via separate contracts.

      So, as regards the original premise about steel and air, the LCS contract is, indeed, for a “bare” hull and allows us to draw conclusions about the validity of the steel and air premise.

      Hope this helps!

  4. Well, it will take some research, but even with inflation and compartmentalization, I still have a hard time seeing a $350 million price swing between the LCS and the ship it was based on if we are just going on sea frame. That's alot of steel and redundancy. My un-informed feeling is that there's alot of inefficiency in the system bubbling under the surface.

  5. I just had a disturbing thought. If the extra compartmentalization and redundancy of a commercial design yields a $400 million hull, then the trick may be up. If we can't produce
    Even a moderately capable hull for less than the price some nations are claiming for
    Whole frigates and Corvettes, then the US obviously can't build ships in an affordable manner anymore.

    1. Jim, the major difference is probably not some extra compartmentation or redundancy but, rather, shipbuilding volume as described in this post.

      You have noticed that the US is not the shipbuilding capital of the world? And that other countries generally buy non-US produced ships? And that the US is down to only a small handful of shipyards? Does that answer your implied question about affordable shipbuilding in the US?

    2. The US may not be a ship building powerhouse but nor is Australia. Austal's shipyard in the US is much much larger than the one in Western Australia where BE was built, with about ten times the workforce. The BE is also a unique design, so there is no advantage of scale production. Labour costs in Australia are also quite high, possibly higher than US.
      A better comparison might be the JHSVs being built in the same shipyard as LCS at the same time. They are slighly smaller and cat rather than trimaran but are built closer to commercial standard. The unit cost is US$214 million. This higlights the cost premium associated with a full naval vessel.

    3. I hate to say this, but maybe we need to start thinking of a downsized Navy as the new norm and adjust accordingly?

      Wouldn't it be better to have 7 or 8 carriers instead of 11? (I know we're headed there anyway, but I'm thinking of staying there as a long term option). Then we have the money not spent on building/maintaining those vessels used to build more lower end ships and improving the air wing as much as possible? Maybe we keep on with the Nimitz design. I really don't see much that suggests its a horrible design that has to be replaced. It can carry alot more plans than it has now.

      I understand a number of intruders were put in the boneyard. Could they be brought back just as tankers to extend the strike reach of the Superhornets and Lightnings? Just flying gas tanks with a couple missiles for emergency self defense.

      Or maybe start shifting more towards subs and SSGN's. The Dong Feng isn't going to do a whole lot against an Ohio SSGN. But if the Chinese tried to forcibly take Taiwan those SSGN's could really wreck an invasion fleet.

  6. "Jim undoutedly now gets the point about commercial/warship cost comparisons. The difference would be even more pronounced if the LCS was a true warship!"

    I see. And that is just stuff prepping to support the separate budget issues like weapons.

    My next question is two fold:

    How do we fix this? (The Koreans built a massive shipbuilding industry in about 40 years, from nearly nothing). I'm guessing some government incentives/regulations would have to change? I remember in college Newport News was going to take on building high volume commercial vessels. That must not have worked out.


    How is it that the British can still develop a fully navalized sea frame (isn't their new 'global frigate' supposed to come in at like $300 million?) I don't see their shipbuilding capacity as much better than ours?!

    I love this blog. But sometimes it just bums me out.

    That said... there has to be a way around it. Again, my expertise is from the auto industry. From a strategic perspective it doesn't matter if Ford or Toyota owns a heavy vehicle plant in Indiana. Its still a high tech highly configurable plant on US soil.And in most heavy industries our industrial productivity is amazingly high, so they do attract workers.

    I wonder if we could, at least in the short term, get some of the foreign nations to build their ships here, and invest in the yards?

    Alternatively, why did we close so many government owned yards? And with the four we have left, could we convert them to building naval vessels instead of just de-mil ing them.

    1. Jim, all great questions and I don't have the answers, unfortunately. My vague sense is that over the last several decades we regulated the shipbuilding industry out of business as well as being unable to compete in terms of cost. How and why all that happened, I don't know. I haven't studied it.

      On the military side, we went from a fleet of thousands of ships post-WWII to a fleet of 285 ships, today. We're building around 10 new ships per year. That just isn't going to support many shipyards. I've suggested in previous posts that we should be building more smaller and simpler vessels and upgrading existing ones. That would increase shipyard demand. Unfortunately, the trend by the Navy is to build fewer, more expensive ships and to minimize upgrades so as to make funding available for new construction. Thus, the Navy is headed down the path of fewer and fewer ships which means less work for shipyards and, thus, higher and higher costs for the ships that do get built. It's a vicious spiral leading to a fleet of 220-250 or so in the moderately near future. You've seen the announced ship retirements for next year versus the new construction. It's a large decrease in numbers. That's just one year but the trend is clear.

      As far as being discouraged, the first step in fixing anything is to understand the problem. Don't be discouraged. Instead, learn what you can here, research the issues, and then you'll be prepared to take whatever action you can (voting, interact with your Representatives, etc.). In addition, while the Navy's leadership may be screwed up, you should be inspired by the rank and file of the Navy. The sailors manning the ships are generally outstanding and do amazing work.

      I viewed a NatGeo special on the LCS and watching the sailors work to operate and maintain the functionality of the ship and accomplish their missions is inspiring. That the LCS program has flaws doesn't detract from what the sailors are achieving.

      Hope this helps with your perspective!

    2. It does, thank you.

      I think that the sailors themselves do a fantastic job. My worry is that while the fleet starts to get reduced in size and capability, the desire of the politicians to use it won't. That means in some future scenario a politician might put a ship and its crew in a position that its not ready for.

      Going back to the shipyards...

      I wonder if we could make some other changes to our philosophy. I.E. design a ship with as much COTS as possible and rethink some things. If the Hulls are going to be expensive from the git go take the Zumwalt philosophy with the FFG's. Come up with a mission, taylor the ship primarily to the mission, and plan on building alot of them. Have the design first.

      As a thought expiriment, build a 4000 ton hull to a level II spec (Similar to OHP). Make it as robust as possible to last a long time. Try to make it as modular as you can in a small ship. Its mission would be the same as the current OHP's. Maybe its possible just to by a sea frame from a foreign design and 'fill it in' ourselves? I don't know.

      ASW primarily, secondary anti ship armament.Point defense anti-air.

      You're requirements, IMHO, would be:

      * Towed Array Sonar. Try something simple like was originally put on the OHP's.Let computing power give you the quality of the signals.

      * Some sort of AShM. A quad pack of Harpoons would be fine. VLS, maybe, or maybe not. I'm not sure how much that would cost vs. a deck mount. Plus if its a deck mount you might have the option of easily upgrading to the LRASM if/when it comes into the fleet.

      *RAM pack for CIWS.

      * Helicopter pad and hanger for one. (I know that's not cheap, but for a frigate role I think its essential).

      * for propulsion... why not diesels? Likely you can find a more commercial setup. You don't need an amazing amount of speed (I think the old Knox's were around 27 knots...) and it helps range and fuel economy. They are probably sufficient for 90% of what this

      * ASROC pack.

      * 1 or 2 triple Mark 32 torpedo packs.

      * Automate what you can, but leave that as much as possible to commercial servers in a shock hardened mount. The point being that as things change or upgrades are needed to computing hardware, you can pull out servers and put in new ones.

      * Cannon... do we need one? Maybe. I'm not sure. I'd be tempted to go with the 57mm on the LCS, or maybe just a couple of Phalanx of the recent ilk that can deal with near surface threats, so the swarming speed boat is covered.

      I think that with this you can end up with a Frigate that can do ship protection, ASW work, defend itself against smaller threats, and shoot against some bigger ones. Its upgradeable, and built to last a long time.

      Maybe you start building it at some of the Navy yards too. Once you get a design down, don't deviate. The goal should be production as much as possible.

      Feel free to start tearing it apart. :-)

      If the Burkate doesn't work out due to costs, then I think that this is the only way to go.

    3. The Three main killers of the U.S. ship building industry were (and still are):

      1) Labor
      2) Environmental regulations
      3) Lack of infrastructure

      This is a reality that echos through heavy industry in the U.S. - if you can wrestle these, then you can be profitable. Sadly, the risks and rewards are just not there.

      Japan and Korea followed a very reasonable course to building up their manufacturing sector. First they started in toolmaking: the base for building things. They continued to refine there tool/instrument making industry and started moving into niche markets making large industrial items like locomotives. Then they moved into more general markets like trucks and earthmoving, while continuing to refine the tools, instruments and massive niche markets. Once they started building cars, the expertise in tooling and prescision instruments let them build up superior robots, which in turn led to better cars and consumer products.

      You might think that building smaller consumer items is the way to go, but the problem is that you cannot even think about production unless you have roads, rail lines, ports, and bridges that can support the weight of heavy lathes (now all CNC), transformers and the like. You have to have the electrical grid to support machining, you have to have access to water.

      The ultimate resource though is a well trained and motivated workforce.

      Anyone visited a U.S. tool maker like Starrett? Most of the production methods have remained largely unchanged since the 1950s. Most of our rail lines and ports are crumbling. Our electric grid is a mess. When was the last nuclear power plant we built? When was the last refinery built? Good luck with getting an environment permit to deal with chemicals and other processes needed for heavy production.

      The area of our biggest failing remains in training the highly skilled workers needed to build items. We ran machine shop, welding, and other vocational training programs out of secondary schools long ago. Unlike Europe where the unions are actively and constructively involved in training new skilled labor, our unions have focused on fighting business for wages and benefits.


    4. GAB, a good comment and I absolutely love your final paragraph. That is an astute, brilliant, insightful (I'll stop with just a short list of adjectives but you get the idea) observation. We so emphasize college that we make every high school student feel like a failure if they don't go on to college. Well, not only is college not for everyone but we desperately need a skilled workforce with the vocational skills to actually build and maintain things. GAB, I wish you were in charge of our public school system! Your comment on unions is right on the money.

      A ComNavOps salute to you, GAB!

  7. This comment has been removed by the author.

  8. In response to Jim above,

    If we (the British) can build the new Type 26 Frigates for $300 million each I will eat my hat. The lowest semi-realistic assessment I've seen is that they will cost around $400 million each, though I suspect personally the price will be closer to $520 million each. And even that's considering the fact that the towed sonar array's will be taken off the old Type 23 (Sonar 2087 is basically a world leading unit and will likely remain so for about another 10-20 years, something we actually did right) and the Artisan radar is also already being put onto the Type 23, so that's two big development costs that have already been covered (along with a variety of other equipment that will probably be ported over, like the Harpoons).

    For ComNavOps,
    To give you a rough reference point, the steel for our aircraft carriers was purchased from our domestic supplier (Corus) in a deal worth roughly $104 million for 80,000 tonnes of steel, which is about $1300 per tonne, though this was back in 2008 and steel prices will have moved since then.

    By comparison, just the reverse osmosis kit for the carriers cost about $1.6 million. I think a lot of the cost is in the details.

    And that documentary, if it was called "Inside; 21st Century Warship" then I just finished watching that before checking up here! The story of the crews and officers fighting against their own ship to try and essentially drag it into working order is both a testament to the men and women who serve in your Navy, and a damning indictment of both LCS designs, and the brass who reset a test because the Captain wasn't doing things "their" way. Heaven forbid that US naval Captains should be free thinking, intelligent, daring, iniative taking individuals. That just would not do.

    1. As you said, I think the estimate for the type 26 is around £250-350m. Not including the equipment transferred from the Type 23. However even if we said that transferred equipment would cost $100m, then the total cost of a type 26 is still going to be around $630m at most.

      Now compared to a LCS, a Type 26 will be much larger and better quipped. Also I don't know how many LCS are now planned, but would expect that they are building more of them, than the maximum 13 Type 26 that the Royal Navy will get, so you would expect the cost to be lower with more built.

      The only thing that I can think of that makes the LCS so expensive is the speed requirement. The Freedom class has two Rolls-Royce MT30 engines. I expect that the Type 26 will have the same engine, but possibly only one of them?

      To compare a couple of other Royal Navy ships, the Type 45 cost somewhere around £580-650m each. Reportedly around £200m of that is for the Sea Viper missile system.

      While the Type 23 cost from £92m to £148m. The last one built which was commissioned in 2002, cost £106m

    2. I should have said that cost for the Type 23, were for the hulls, not including Government furnished equipment.

  9. Just as a yardstick, Morocco is getting this for $676million - just a single unit, built in a French (so pricey) yard:


Comments will be moderated for posts older than 7 days in order to reduce spam.