Monday, December 30, 2013

Virginia SSN Review

The Virginia class SSN has been around for some time now and it’s worth a quick review.  The program was begun in the 1990’s as a replacement for the Los Angeles class and as a cheaper alternative to the Seawolf class.  Construction began in the late 1990’s with around 16 subs completed so far.  Construction is proceeding at a rate of two per year.

The SSN force peaked at 98 SSNs in 1987 and has been declining since.  The current force goal is 48 SSNs, however, the Navy’s various 30 year plan projections show a dip to 42-43 in the late 2020’s and a sustained shortfall below the 48 goal from 2022 – 2034.

Here’s a quick dimensional comparison of the classes.  Note the trend towards bigger boats just as we’ve observed with the SSBN(X) Ohio replacement sub which is bigger than the Ohio despite having several fewer missile tubes.  I’ve never seen a documented reason why the Virginia needs to be bigger than the preceeding classes.

Los Angeles  362’ x 33’                   4 tubes / 26 torpedoes
Seawolf          353’ x 40’                   8 tubes / 50 torpedoes
Virginia           377’ x 34’                   4 tubes / 27 torpedoes

The Virginia construction program was beset by cost overruns early on.  To be fair, that’s a characteristic of almost every Navy construction program.  The Virginia’s were envisioned to be low cost alternatives to the Seawolf but it hasn’t turned out that way with the subs running around $2.5B in today’s dollars.  Proceedings offers an article with the early cost history and states that the total obligation authority increased from $60B in the mid-1990’s to $90B+ in the mid 2000’s (4).  Here’s a partial cost history and planned budget amounts (1).

1998 – 2011  14        $2.6B (average cost for the period)
2012                 2        $2.6B
2013                 2        $2.5B
2014                 2        $2.7B
2015                 2        $2.7B
2016                 2        $2.8B
2017                 2        $2.9B
2018                 2        $3.0B
2019                 2        ?  VPM will be incorporated from this point on

CRS reports (3) that construction in FY14-FY18 will be performed under a Multi-Year Procurement (MYP) arrangement which is estimated by the Navy to produce 14% savings.  However, looking at the planned construction costs versus previous years, there is no evidence of cost savings.  Serial construction cost savings and MYP should be combined in FY14-FY18 for significant savings but the projected costs are not only not reduced, they’re anticipated to rise by $300M.

The Virginia Payload Module (VPM) is planned to be incorporated starting in 2019.  The module consists of 4 tubes x 7 Tomahawk missiles = 28 missiles per sub.  The VPM is the Navy’s response to losing the SSGNs which carry over 150 Tomahawks each.  According to the Navy, it will require 15-20 Virginias to replace the Tomahawk capability of the 4 current SSGNs (yeah, the arithmetic suggests around 16 but Navy articles suggest around 20 – not sure what the exact thinking is).  According to the CRS report (3), VPM will increase the Virginia construction cost by 15%-20% ($375M - $500M).

While the Virginia class is generally considered to be a success, there are some issues, to be sure.  From the DOT&E report (2), we note comments about the Navy’s reluctance to conduct realistic testing.

“Because Navy security rules prevent collection of useful operational test data from Virginia when conducting exercises with foreign ASW capable platforms, the Navy finished IOT&E and recent FOT&E without testing the Virginia class submarine against one of its primary threats, the foreign diesel electric submarine (SSK).”

No surprise there.  We’ve noted in previous posts that realistic testing, in general,  is a Navy shortcoming.

The overall assessment comes across as neutral or a slight improvement over the 688 class.

“Although Virginia was not effective for some of the missions tested, it remains an effective replacement for the Los Angeles class submarine, providing similar mission performance and improved covertness.”

The report notes specific problems with the Wide Aperture Arrays.

“After completion of operational testing, the Navy issued software changes intended to address the severe performance problems observed with the Wide Aperture Array.”

“… the Navy should re-evaluate operational effectiveness on a submarine with a repaired Wide Aperture Array.”

DOT&E notes that problems are propagating into the fleet due to the Navy’s insistence on meeting schedules over identifying and fixing problems.

“DOT&E assesses that the late fix of the array’s deficiencies is a result of the Navy’s schedule-driven development processes, which fields new increments without completing adequate developmental testing.”

Further testing is recommended.

“Repeat the FOT&E event to determine Virginia’s susceptibility to low-frequency active sonar and Virginia’s ability to conduct ASUW in a low-frequency active environment. This testing should include a Los Angeles class submarine operating in the same environment to enable comparison with the Virginia class.”

In summary, the Virginia program has exhibited poor cost performance although not too bad by Navy standards (of course, that’s a pretty low bar!).  With respect to performance, the Virginia appears adequate and probably represents a modest improvement over the Los Angeles class.  To be fair, actual performance data is nearly non-existent in the public domain.  My vague impression is that the Virginia is to the Los Angeles what the JSF is to the Super Hornet – a modest improvement at a hefty price.  Nevertheless, the Virginia is needed and reasonably meets the need at an acceptable, if disappointing, cost.


(1) Dept of Navy FY14 Budget Estimates, Shibuilding and Conversion, Apr 2013

(2) DOT&E FY2012 Annual Report

(3) Congressional Research Service, “Navy Virginia (SSN-774) Class Attack
Submarine Procurement: Background and Issues for Congress”, Ronald O'Rourke, 2-Apr-12

(4) U.S. Naval Institute Proceedings, “The Sweet Smell of Acquisition Success”, RAdm. John Butler (USN, Ret.), June 2011

8 comments:

  1. Kind of makes one wonder if we just shouldn't have stuck with the Seawolf. How much did we really save by going Virginia class? or was it just a political/navy strategy to fund only 3 Seawolfs and go to the "cheaper" model,even if it really wasn't that much cheaper?

    ReplyDelete
  2. ComNavOps,

    Be careful in your criticism of costs on the Virginia boats.

    First, you must remember that each "flight" is an improved variant, with new equipment and systems onboard. In the past these were different ship classes, but alongside the inane Navy ship naming scheme, we seem to have adopted the software industry system of using versions rather than classes.

    Second, you must account for both inflation, and constant year cost comparisons. Accounting for modest inflation, your 2012 to 2018 numbers look reasonable.

    GAB

    ReplyDelete
    Replies
    1. GAB, to be fair to me, I offered little explicit criticism of costs in the post. I noted that the original intent of being a low cost alternative to the Seawolf class was a failure. That's a criticism and seems a fair one. Wiki reports the projected costs of 12 Seawolfs as $2.8B each. Compared to the Virginia's costs of $2.5B, even allowing for inflation, that seems a failed "cheap" alternative. Beyond that, I simply noted the costs.

      I further stated that the serial production and MYP cost savings did not seem to be evident though they may be present but swamped by the flight cost increases, as you note. The 2012 to 2018 cost increases appear to be simple inflation. That's not a criticism, just an observation of the data.

      By Navy standards, which are abysmally low, the Virginia is a stable program with controlled costs. Now that is an implied criticism but it's also a fair statement, I think!

      Was there a specific aspect of costs that you think I've been unfair about?

      Here's an amusing insight into the Navy's thought process... In the Proceedings article, the author notes that the original cost of the Virginias ballooned by $1B per sub. The Navy then embarked on a massive cost cutting exercise and reduced the increase from $1B to $500M. The author then wrote an article praising the Navy's outstanding fiscal management of the program for the highly successful cost cutting (his thoughts, not mine). He seems oblivious to the fact that the net result was a $500M increase. The Navy touts the Virginia program as a shining example of acquisition success despite failing in its stated intent to be a low cost alternative, racking up a half billion dollars of overrun per vessel, and failing to demonstrate any serial production or MYP savings (again, they may be there but swamped by other cost increases). The Navy's idea of success is what most of us would consider failure.

      Delete
  3. Seawolf cost $2.4 billion in *FY 1997 dollars*.

    Virgina cost $2.0 billion in *FY 2005 dollars*.

    At least the $$$ are in SSN 774's favor

    You are also trying to compare Virginia block III with Seawolf and LA on the basis of torpedo loadout and not considering other features, to include the VLS capability, which is a 1:1 for torpedoes, or the other capability(ies) of VPM. An this does not begin to address the really important features like sensors. We will not know how well Virginia flights compare with Seawolf (which has also most certainly been updated)

    GAB

    ReplyDelete
    Replies
    1. Whoa, there, GAB! You're reading way more into the post than was intended. The very brief list of torpedo capacities was just a passing reminder and was in no way a capabilities comparison. The post was not a comparison of Seawolf and Virginia in any way other than the mention that Virginia was intended to be a lower cost alternative.

      As far as cost numbers, I'm using RAdm Butler's figures from the Proceedings article that I referenced in the post. Here's the relevant quote:

      "All told, each boat's final cost increased by approximately $1 billion on average and, over the 30-ship program, the total obligation authority went from approximately $60 billion in the mid-1990s to $90-plus billion in the mid-2000s."

      That means that the TOA of $90B over 30 ships = $3B. Subsequent cost reduction programs reduced that to $2.5B per ship.

      RAdm Butler was PEO Submarines. His numbers ought to be as reliable as can be had in the public domain. If you have a better source, I'll be glad to consider it.

      Delete
    2. ComNavOps

      Fair enough on the weapons.

      Reference costs:

      "In FY 2005 dollars, SSN-21 submarines cost between $3.1-3.5 billion each."

      http://www.defenseindustrydaily.com/Virginia-Class-Sub-Program-Wins-Acquisition-Award-05167/

      zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz

      So like I said, back in the 1990s, when the funding decisions were made, Seawolf was much more expensive than the Virginia boats. Again:

      Seawolf cost $2.4 billion in *FY 1997 dollars*.

      Virgina cost $2.0 billion in *FY 2005 dollars*.

      When you adjust the costs to FY 2005 when USS Virginia was commissioned in 2005: Seawolf cost over $3 billion dollars, and the flight one Virginias came in at just over $2 billion per copy in *FY 2005* dollars. Ergo Virginia boats cost 2/3rds the cost of a Seawolf in FY 2005 dollars.

      There is no smoke and mirrors reference costs: you have to adjust for inflation.

      The SSN774 program is now going into its third design, basically improving performance at each stage (well we will see how flight III works), greatly adding weapon capacity (VPM), radically shortening the actual construction time of each boat, and roughly holding costs constant in FY 2005 dollars.

      How good is Virgina versus Seawolf: well that is the question, and one that anyone not intimately connected with the submarines is unlikely to know.

      That said, Seawolf was not without its issues, and got dinged up by GAO for construction problems and issues with its combat systems. GAO/NSIAD-95-4

      GAB

      Delete
    3. GAB,

      A cost comparion between SEAWOLF and VIRGINIA is bit miseleading - or at least incomplete.

      SEAWOLF never went into serial production. We only built three. All the up-front' costs (R&D, spare parts) are levied onto that small denominator. The end result is an artificially unit cost.

      In contrast, we've since built ten VIRGINIA class and have eight more laid down or under contract. All the up-front costs for VIRGINIA get spread out over a larger denominator.

      It's interesting to Imagine if the Navy had stuck SEAWOLF class back in the '90s - and gone into true serial production. The unit cost likely could've gone down quite substantially.

      Matt

      Delete
    4. Matt

      Nonetheless, adjusted to FY 2005 dollars, SSN-774 (also a RDT&E boat built on a cost plus contract) cost ~$2 billion, while SSN-21 cost exceeded $3 billion.

      Also missing is the fact that SSN-774 program includes assistance to get a second shipyard on line for submarine construction as well as modernization of Electric Boat.

      Eighteen Virginia submarines in three flights (essentially three different designs), is not like cranking out Dodge Ram pickups, though the point is taken. The sonar array of the flight three boats is a the most radical design change of a U.S. submarine sonar in ~50-years.

      GAB

      Delete