In reading through the blogosphere over the last year whenever a discussion of Navy shipbuilding costs pops up, I note the comments are sometimes filled with inaccurate estimates. Yesterday in proposing an idea for a $100 million littoral ship less than 1000 tons on my blog, I didn’t mention that I hope that if we built such a ship, we could build the first in class for less than $200 million. I also didn’t mention how much I hope the $100 million cost cap is achievable. The key word is hope. Why do I need hope?

Because I study history, and shipbuilding costs is a confusing subject with discussions often not rooted in cost reality. Design costs and construction costs towards lessons learned in new classes of ships are never cheap. There is a dirty little secret why so many people are concerned DDG-1000 and DDG-1001 will blow out the shipbuilding budget.

Of the nine first in class ships previous to LCS, four had overruns of greater than 100% (MCM-1, MHC 51, DDG 51, LPD-17), three had overruns between 40-60% (FFG 7, CG 47, LSD 41). Only two had overruns less than 20% (LHD 1 and VA 774/75). NONE came in lower than expected. The LCS-1 $631 million /LCS-2 $634+ million first in class ship cost overrun with the artificially low $220 million seaframe target is around 188%, which makes it the worst first of class overrun in 20 years. If you compare the cost overrun with the realistic but discarded target of $297 million, the overrun was still 113%. While 113% would make the LCS cost overrun comparable to the DDG 51’s overrun of 110%, it is still over 100%.

This means of the last 10 ships built by the US Navy, five of the last 10 ships have had a cost overrun of greater than 100%. In other words, the DDG-1000 has a 50-50 chance of having a 100% or greater cost overrun on each of the first two ships, and a 20% chance of having an overrun of 155-160 percent. Said another way: there is an even chance the first two DDG-1000s will have overruns totaling a minimum of $6.4+ billion.

First in class ships are only part of understanding shipbuilding costs. A lot of people assume the SAR reports account for all the costs of a shipbuilding program. It doesn’t always. In many cases individual systems programs intended for the ships are not calculated as part of the SAR report, but have to be included in the total ship cost because the ship requires the equipment that is specific to that ship. Some of this has changed, we now fund a lot of systems as part of a ship cost included in the SAR, particularly with SPY and VLS, but that wasn’t always the case with rail launchers and older radar systems.

Earlier this year I got into a debate regarding LCS costs with Bob Work of CSBA. In those debates some numbers were provided for cost context, numbers produced by Eric Labs of CBO. CBO has all the historic cost data, and Eric isn’t just a good source for the final tally of ship class costs, he is the best source for those figures and CBO is probably the only organization other than the Navy who has the exact numbers. Consider the average cost per ship of the last three small vessels built for the US Navy:

FFG-7 4000 tons cost $690 million in FY09 dollars = $172.5 million per 1000 tons

MCH 900 tons costs $251 million in FY09 dollars = $279 million per 1000 tons

MCM 1320 tons cost $262 million in FY09 dollars = $198 million per 1000 tons

Now compare to the LCS expected costs.

LCS 3100 tons cost $608 million in FY09 = $196 million per 1000 tons

$608 million = $550 million hull + $58 million module. $58 million is the average cost estimate of the combined 24 ASuW, 24 MIW, and 16 ASW modules costs / 64.

While the cost of the LCS is much maligned, it is an interesting detail that the LCS represents a more cost effective investment per ton than both of the minesweepers it is replacing, plus it can self deploy and conduct other missions besides mine warfare when forward deployed. However, using these numbers the LCS costs nearly $24 million more per 1000 tons than the Perry class frigate, and is neither built to the same survivability standard nor armed even remotely as well as the Perry was. It is going to be very interesting to see what the LCS frigate version costs Israel and Saudi Arabia should either purchase that model, because if either country sacrifices speed for weapons and survivability and is able to obtain a cost more comparable to the Perry class, the resulting discussion is going to be fascinating to observe (and participate in).

For comparison purposes, the Coast Guard is acquiring the new 350 ton Sentinel class FRC which will have a program cost around $44 million each. At that price, the Sentinel class cost per ton will be around $125 million per 1000 tons. Note the Sentinel class is a cutter, not a Navy ship, and is built to operate 2500 hours a year with a top speed over 26 knots. These metrics are excellent for a Coast Guard ship, but a Navy ship of similar size would require at least 5000 hours a year if it intends to make 6 month deployments.

So for those retired folks out there who love the FFG-7, to build a Navy littoral ship at the same cost standard of the Perry class at $100 million would result in a ~580 ton vessel. To build a 1000 ton ship as is being discussed in various places by the Navy and Marines, the cost would be around $172.5 million per vessel. Welcome to the expensive world of shipbuilding, where the historical record and actual costs of previous ship classes puts current programs into perspective.

From my point of view, it is easy to see why the Navy has become hesitant with so many new ship classes. LPD-17 was 160% over budget, while the first in class USS Virginia (SSN-774) was only over budget 10%. Considering the media hype regarding how expensive submarines are, that fact is often lost in the details. With a one in two chance the first two DDG-1000s could cost $6.4 billion each, twice as much as the $3.2 billion current price estimate, it looks to me like an even bet the third DDG-1000 funded in FY09 will not be built, and the funds will instead be used to pay for the cost overruns of the first two.

Posted by galrahn in Navy

You can leave a response, or trackback from your own site.

  • SB

    The argument of inflation adjusted ship construction fails to recognize the realities of modern weapons (ASCMs) against which a modern warship must defend itself. Weaponry advancement models are closer to Moore’s Law which in turn requires an exponential investment in defensive technologies vice inflation adjusted prices of simple ship construction. Should we build “warships” which can get whacked by a cheap FL-10 fired from a FAC or a C-802 in the hands of terrorists (Hezbollah already did it!!)?
    There will be some guy who comes along and claims we can just use an Oto Melara 57mm can do NSFS and “emergency air defense.” That guy is smoking crack . . . or does not have the knowledge of the actual threats out there and what measures are required to counter.

  • SeniorD


    CG-47 cost overruns were due, in part, to serious questions about SPY-1 handoff between arrays. All previous AEGIS tests were conducted with a single array and it wasn’t until the “Cornfield Cruiser” got a second array to begin testing. I know, I was there.

    Then again, adding the equivalent of three decks to a Spruance-class hull to support SPY-1 arrays didn’t help ship stability any. That and survivability in those ships was iffy, a major re-design of interior space was needed. Otherwise, a single 40mm shell would put the whole ship out of action.

    To SB’s comment in re: defensive technologies – you’re right but consider the implications. Pretty soon, modern warships will have almost as much defensive technologies and offensive. There isn’t much room on a ship for both, something will need to ‘give’. My position has been and will be the need for armor belts. The attack that cost sailor’s lives on the Cole would never have been so costly if we had decent armor. Thin skin and Kevlar may make for fast ships, but are meaningless when the ship is at rest and vulnerable to a row boat filled with HE.

  • pred

    It may be worth looking at the operational availability of existing USN surface combatants. The European fleets currently are in the 2,500 hours bracket and the step to 5,000 is sold as a major feature(at commensurate cost) for the German Type 125 stabilisation frigates.

  • galrahn


    I think you have put too much emphasis on the current condition set without a reflection of history. Every naval era has presented unique challenges to shipbuilding cost with design trade offs between offensive and defensive capabilities. What you are talking about is nothing new, indeed the first USN frigates were enormously expensive because we decided that our first frigates would be built with live oak.

    The swing between necessity on offense and defensive systems is as old as naval power, and the pendulum has swung both ways sometimes even within the same era. What we are facing today in terms of the balance tilting to favor the offense are challenges that will be solved by new innovation in defense technologies to meet the threat, indeed the initiatives represented in programs like AEGIS Ballistic Missile Defense, the possibility of cage armor on warships (a strength vs stealth discussion), and even the early developments of small nuclear powered laser systems with a point defense intent represent the next counter balance back towards defensive systems.

    • KenB

      And the use of live oak significantly reduced maintenance costs … extended useful life and afforded the ship greater protection … remember one ship nicknamed “Old Ironsides”

  • sid

    plus [the LCS] can self deploy and conduct other missions besides mine warfare when forward deployed.

    You are not calculating the costs of the mission module, nor the costs of changing missions in theater. Furthermore, the idea this ship can self deploy is summed up best by you in the comments here:“The LCS concept is never going to work without a tender. Ever.”

  • galrahn

    I’m not counting operational costs of the mission modules if that is what you mean, but I did add procurement costs of the modules. This data is available on CBOs website, the report is linked both from Springboards blog and my own. I believe it was released in March or April of this year.

    That linked comment is my opinion, but it is only an opinion.

  • Dee Illuminati

    Recently there seems to have been an edge to the C-802’s and RPG’s as we see when a run to the Sistani was last attempted. The cost loss of Merkava’s and naval tonnage implies that that cost ratio must be acknowledged.

    I was moderately impressed at the cost and tonnage of the Visby class.

  • Mike M.

    I think the real question is just how much ship is needed.

    From an operational-level perspective, there is a case to be made for a small corvette. Something with enough capability to conduct stop-and-board operations, maybe even get off a shot in a serious fight, but not a ship for a major battle. The modern equivalent of the topsail cutters 200 years ago.

    Part of the problem is that the program manager needs the ability to push back when every reviewer of the specification loads on their own pet requirement. Right now, the procurement system tends toward gold-plating of systems as the least difficult way to reconcile the demands.

    But a bigger problem may be Congress. The political investment in a ship is not proportional to size or cost…and it is terribly tempting for Congress to substitute cheap corvettes for expensive DDGs on a one-for-one basis, with a tremendous loss in fighting power.

    I think building a truly small combatant at a reasonable price could be done. But it would require the approach that DARPA used when they started the Global Hawk UAV program – a very small, very competent team on the Government side, and a contract with very, very few hard requirements. It can’t be done in the existing acquisition system.

    • KenB

      You live in an idealized world where information is not an impediment. Global Hawk cost overruns were SIGNIFICANT as a result of the limited acquisition experience DARP possessed. Their goal was to demonstrate a capability – which they did – but left out the key component of producibility. Just ONE of MANY “oversights” was the inability to use both the optical sensor and radar sensor. Another was the aircraft had to be taken to 50,000 ft to adjust the optical focus from the ground … there are MANY MORE that had to be FIXED to make GH useful
      You do not understand the requirements process … the process in the Pentagon did not come up with a corvette-sized ship to satisfy the OPERATIONAL need. If the requirement could have been satisfied by a corvette THAT is the ship that would have been built

  • Byron

    Yup, if you want to build the right ship for the right cost, you first need to scrap the current aquisition process and start over. But seeing how this process has been mis-managed by Congress since 1792, when they demanded that one of the original six frigates be built by one yard (that had a cozy relationship with someone else)even though it was known that this yard had a less than stellar reputation (sound familiar?) Still, we got six extraordinary vessels that flat tore up the British, French and Tripolitanians that fought them. One was built so well, it’s STILL a registered ship of the US Navy: USS Constitution.

    Now if you could only bring Joshua Humphries back to run the shipbuilding program…

  • pk

    i would solicit you to go to information dissemination and observe the comments of one William Powell (74) [smaller ships could mean more guns] who is handling the fellows singlehandedly with no one to hold his coat or glasses in the matter of size of LCS.

    he seems to apply that golden principle of “EXPERIENCE” to the problem and has pointed out quite a number of items not hitherto masticated in the public arena.


  • doc75


    You use the FRC for a cost comparison with LCS, but wouldn’t the National Security Cutter be a better comparison? What is the cost per tonnage for NSC?

  • gahlran;

    Your frustration on the confusion of naval ship construction costs (in the US) is evidence of the historical problem our USN ship acquisition process has of “under-estimating” ship costs.

    Has this affected the quality of our delivered ships? I would have to say that it has not. The nation has had systems in place to manage the various changes and contract growth (thanks to NAVSEA) required to produce quality ships.

    However, what this historical process of under-estimating ships costs has done is undermine our USN’s credibility with OSD, Congress, and the Nation. (I have spent the last year as the President of one of our Navy League Councils speaking to the public about the needs of our Navy, and the most common response is “the Navy needs to get its house in order on ship costs”. Therefore I attest to our USn losing credibility with that small part of the Nation in my neck of the woods.)

    Regaining credibility is vital. Let’s get our naval ship acquisition cost estimates back to reality.

    While historical US costs serve well, I would recommend comparative cost estimates based on costs of naval ships built by leading international naval shipbuilders. I have studied these costs (my firm has historical records on all free world major international naval ship construction programs enacted since 1992) and you will find the US compares very favorably.



  • Dan

    Very interesting stuff. I enjoyed reading your post. I was wondering if you had any info. regarding foreign Navy cost and time over runs compared to US programs like LCS, LHD 8, DDG 1000, etc. I’m really curious to see if foreign governments/Navys have similar problems in cost and time overruns, or if they manage their shipbuilding programs better Thanks alot.

  • Guy Dixon

    We are now in a process of re-orientation of the global economy away from a US$ centric world, the US economy needs a devalued US$ to help restore its competitiveness. This has major implications for costing of US Navy Shipbuilding programs. The price of all major metals (Fe, Al, Cu. Pb, Zn, Ni, Ti etc) have risen strongly over the last decade, while prices are off their highs they still high in US$ terms, investors buy metals to hedge against as the US$ depreciation. The metals and energy density of major naval shipbuilding means that in the US ship cost inflation will be higher than underlying inflation. Accurate comparisons need to reference a better ship centric inflator that considers metals, energy, skilled labour (e.g skilled aluminum welders are hard to find) Most commodities rise as the US$ weakens, looking at the LCS its engines and jets are also European so there are some big ticket items which are exposed to US$ variations.
    Careful consideration should be given to hedging foreign currency purchases in the selected programs with high foreign input such as the LCS. The Australian navy is currently in major expansion mode, the first of 3 air warfare destroyers is currently being built and is on budget and on time according to recent reports, we are also building large amphibious landing craft (LHD equivalents) which I believe have not run into problems. Recent meetings between US and Australian defense ministers/secretaries reported that the US was interested in our “project at risk” procedures which are a sort of “early warning” systems for major acquisition projects. We have adopted an approach which seeks to be proactive in monitoring and reporting project progress and moving quickly if there is evidence of contract/project underperformance. Interestingly the strong rise in the A$ will support projects which were costed 1-2 years ago when the A$ was in the range US$0.5-.65 it is currently ~US$0.90. I do not know that we are doing things that much better, but there is low tolerance here for major defense projects that do not live up to expectations. This has led us to create a separate Ministerial entity called the Defense Material Organization (DMO) which runs and is responsible for major acquisition programs.

  • Byron

    I take it the Collins was responsible for your learning curve?