LCS Survivability Debate: By the Numbers

Both Doyle Hodges and Chuck Hill have recently commented on Littoral Combat Ship (LCS) survivability in light of World War II damage reports, responding to Steven Wills’ opinion piece at USNI News (which he followed up yesterday).  The genesis of their response appears to be Wills’ claim:

“Small warships are historically unsurvivable in combat. They have a shorter floodable length, reduced reserve buoyancy and more likely to be affected by fire and smoke damage than larger combatants.”

Unfortunately, both authors have interpreted the data differently, and some clarification is in order.

In analysis of survivability, it is important to look at three facets as defined by the US Navy:

  • susceptibility, the capability of the ship to avoid and/or defeat an attack;
  • vulnerability, the capability of the ship to withstand initial damage effects and to continue to perform assigned primary missions;
  • recoverability,  the capability of the ship, after initial damage effects, to take emergency action to contain and control damage, prevent loss of a damaged ship, and restore primary mission capabilities.

The data from World War II damage reports, along with an understanding of the US Navy inventory at the time, allows us to calculate a value for overall susceptibility of the various ship types during the war.  The damage reports also allow us to measure the combined vulnerability and recoverability of those ships that reported significant damage.

Hill calculated an overall measure of survivability based on the number of each ship type existing at the beginning of the war, which Hodges correctly notes is an insufficient baseline due to additional production during the war.  By using data from Table I of the damage summary, he overstates the overall risk of loss in each type of ship.  I offer the following so that an accurate baseline can be used to inform future discussions of survivability.

In the area of susceptibility, the damage summary, combined with the available construction data, shows significant variability by ship type and year.   In the table below, total inventory of DD and larger combatants is shown for each year, in aggregate and by class, as well as the total of newly commissioned ships for that year.  The susceptibility is the number of damage reports (including lost) divided by the total number of ships available in that year.  This still overstates susceptibility a bit, since not all new ships were available for the entire year, but it still serves a useful purpose.

Note that the total of damage and loss reports, not ships damaged, is used below.  Thus the CV susceptibility of 125% is not really a mistake, but an artifact of the method.  In 1942, Enterprise filed three war damage reports, Saratoga two, and Yorktown one in addition to the losses of Lexington, Yorktown, Wasp and Hornet.

Year and Type Starting Inventory Newly Commissioned Total
1942 167 102 91 33.8%
BB 17 0 12 70.6%
CA 18 0 16 88.9%
CL 19 9 9 32.1%
CV 7 1 10 125.0%
CVE 5 14 0 0.0%
CVL 0 0 0 0.0%
DD 101 78 44 24.6%
1943 231 186 45 10.8%
BB 15 2 0 0.0%
CA 13 4 2 11.8%
CL 26 7 9 27.3%
CV 4 6 1 10.0%
CVE 19 25 1 2.3%
CVL 0 9 1 11.1%
DD 154 133 31 10.8%
1944 403 133 110 20.5%
BB 17 2 10 52.6%
CA 16 1 3 17.6%
CL 32 11 12 27.9%
CV 10 10 11 55.0%
CVE 43 33 12 15.8%
CVL 9 0 3 0%
DD 276 76 59 22.7%

The outcome of each damaging event recorded in the summary gives us an estimate of the vulnerability and recoverability for each ship type.  Here we see a clear relationship between size and ability to take punishment.  I’ve separated the traditional surface combatants from the carriers for clarity.

Type Damage Reports Loss Reports Total Probability of Loss | Given Damage
BB 20 2 22 9.1%
CA 15 6 21 28.6%
CL 27 3 30 10.0%
DD 87 47 134 35.1%
DE 9 7 16 43.8%
CV 18 4 22 18.2%
CVL 3 1 4 25.0%
CVE 9 4 13 30.8%

At the surface, there appears to be a disparity between CA and CL loss rates.   However, this is most likely a function of displacement rather than ship type, as all six of the CAs lost through 1944 were Treaty Cruisers of less than 10,000 tons.

In reviewing these data as it relates to LCS survivability, I think we have to be somewhat cautious.  The general trend, showing that smaller ships are generally more vulnerable to (or less recoverable from) significant damage, confirms Wills in part.  While smaller warships are indeed, less survivable, they are not unsurvivable.  Because of their large numbers and lesser value as a target, the small destroyers were damaged less frequently than the larger surface combatants.  They relied more heavily than their capital ship brothers on the susceptibility leg of the survivability triangle.  Damaging events that were survivable by bigger ships were often fatal to the destroyers, because the bigger ships were more resilient and possessed a deeper capability for damage control.

The same will be true in the US Navy fleet of the 2020s, where Arleigh Burke destroyers will be much more capable of taking a hit than Littoral Combat Ships.  No modern ship, and especially no small ship, will survive as an effective combatant in the event of a major damaging event, like a torpedo, mine, or cruise missile attack.  Even though the ship might not be a total loss, it will be out of action for significant period of time after that event.  Look no further than USS Chancellorsville if you believe otherwise.

Ken Adams is a former US Navy Surface Warfare Officer with experience in amphibious ships and staffs, a battleship tour in USS Iowa, and as a tactics instructor for the Royal Saudi Navy. He is currently an operations analyst for a large defense contractor, for whom he is not authorized to speak.

13 thoughts on “LCS Survivability Debate: By the Numbers”

  1. I agree that small ships are “survivable” to a point. Since so many folks seem fixated on WW2 numbers, consider that the Japanese Type 93 “Long Lance” is the equivalent of today’s large ASCM. Few ships under 10K tons survived more than one or two hits from the Long Lance in WW2. Even one hit was often a mission kill. Today, a DDG 51 can also be put out of commission by as little as one hit.

    Survivability is complicated by the fact that systems such as today’s 76mm or 57mm gun is not the 5″ 38 of yesteryear. Modern weapons have no crew-served capability, and their electronics are not as easy to restore as older weapons. Modern small ships have significantly fewer people to conduct DC. It all adds up to less survivability for small ships and perhaps less in comparison with the WW2 cousins.

  2. I like it. But I don’t really think the reality that small ships can take less damage than large ones was ever in doubt.

    Adams has done us a favor by breaking down the elements of survivability. Using the three components he sites, my complaint was that fixating on vulnerability and recoverability without considering susceptibility will lead to an undervaluing of small ships, which is apparently endemic in the USN.

    As to my sample size, I would challenge anyone to demonstrate that 233 ships is an inadequate sample size.

    Again I was not arguing in favor of the LCS which I regard as somewhat ill conceived, but in favor of small ships in general. The small boys still have a place.

    During WWII smaller ships were better able to avoid dive bombers and torpedoes, that may no longer apply, but small size does still confer advantages. They are harder to detect. They are harder to distinguish from other targets that may be in the area. They are easier to defend using soft kill methods, Because tradeoffs allow more of them to be built the enemy has a harder time tracking them all.

    Rather than compare one hit on a large ship with one hit on a small ship, perhaps it would be better to compare one hit on a large ship, with one hit on one of four small ships that might be built for the same price.

  3. As to this comment, “At the surface, there appears to be a disparity between CA and CL loss rates. However, this is most likely a function of displacement rather than ship type, as all six of the CAs lost through 1944 were Treaty Cruisers of less than 10,000 tons.”

    I would point out that the displacement of all the Treaty Heavy Cruisers were little different from that of the Brooklyn and Cleveland class Light Cruisers, their were a number of much lighter light cruisers, eight Atlanta class and ten Memphis class, that were significantly lighter than any of the Treaty heavies.

    The explanation has to be something other than displacement.

    1. In the Pacific, pre-war ships took heavier losses in the early battles, before Midway, Guadalcanal, and the submarine campaign really turned the tide and tilted the numbers in favor of the US Navy. In the Atlantic, wartime classes got to serve into 1944-45, when U-boats were being sunk almost as fast as they were built.

  4. I carried Adam’s statistics to their logical conclusion multiplying “Probability of loss/Given Damage” by the average annual susceptibility to get an annual probability of loss. The answers are interesting.

    For Carriers
    CV: 11.5%
    CVL: 1.4%
    CVE: 1.9%

    For other combatants
    BB: 3.7%
    CA: 11.3%
    CL: 3.9%
    DD: 6.8%

    Particularly if you look at carriers it is clear that WWII data does not support the contention that larger vessels are more survivable than smaller ones.

    Looking at the other combatants, After Pearl Harbor the relatively large number of pre-1922 battleships were largely kept out of harms way. The same can be said of the ten Omaha class light cruisers in that only two of the class were ever engaged in significant combat and then only for short periods.

    It is time to stop saying that WWII data supports the conclusion that small ships are less survivable than larger ships. It is simply not true. The data does not support the claim.

  5. @ Ken: Great job with the numbers–a very thorough analysis and a well-phrased conclusion.

    @ Chuck: My issue with your sample was not its size, but its non-random nature. By focusing only on ships in commission at the start of the war, you skewed your sample. This would be reflected both in the employment (older, less capable ships would be less likely to be assigned the most dangerous missions) and technology (older ships would not benefit from radar and other improvements that affect survivability). 233 ships randomly chosen–good sample. 233 ships chosen because of their age–not a good sample.

    I’m still concerned about trying to use numbers from WWII to prove a point about LCS today. The issue is not whether small ships are more or less survivable than large ships. The issue is whether LCS is survivable enough for its envisioned employment. I don’t think WWII data will answer that question for us, since it can’t begin to address that question of “how much survivability is enough, and against what threat?”

  6. @Doyle, thanks. You’re right that the match of LCS survivability to its employment is the real issue, but Chuck’s point re: small vs. large ships is still germane. What I hoped to illustrate with this analysis is that we can’t lump all the factors of survivability together. We must talk about it in terms of its individual components, or the full capabilities of the ship will become subject to holy war rather than rational debate.
    Much argument against LCS has centered on its perceived low recoverability and high vulnerability against certain threats, with many references to the toughness of older designs and specifically WWII destroyers. The reality of the situation is now clearly laid out from the data – WWII destroyers were lost 35% of the time they were successfully attacked, and we could further use the damage reports to establish baseline levels against specific threat types. Similarly, we also know now that somewhere between 10% and 25% of DDs were successfully attacked during the war. These two facts give us the ability to have a rational discussion about what the susceptibility, vulnerability, and recoverability performance of a littoral combatant should be in the face of a major conflict against a determined near-peer.

  7. In regards to the disparity between CA and CL figures, I believe that that is primarily due to exposure. The USN heavily committed it’s cruisers in the fighting that occurred around Guadalcanal. A review of the various engagements indicates that the heavy cruisers bore the brunt of the fighting, especially at Savo. Consider, of the allied cruisers present at Savo, only two (HMAS Hobart & USS San Juan) were light cruisers. These ships were with the transports and neither engaged the Japanese. The remaining cruisers were all heavy cruisers and four out of six were sunk in this engagement. The fifth (USS Chicago) was moderately damaged and HMAS Australia did not engage. Replace these ships with light cruisers and the results are probably the same due to the element of surprise. The CAs continued to see the majority of the action which in turn led to the loss of USS Northampton at Tassafaronga and USS Chicago at Rennell Island. This is in addition to USS Minneapolis, USS Pensacola and USS New Orleans all being badly damaged. USS Honolulu, the only light cruiser present, escaped without a scratch. Good luck and good ship handling. More can be written about Cape Esperance and the Battle of Friday the 13th, but I’ve written too much already. The figures posited are simply due to the heavy cruisers being in the thick of action. Also, as noted in a previous post, the Omaha class were mostly kept out of action, further skewing the numbers.

  8. Ken Adams, where did you find your list of fleet strength? Destroyer strength in 1942 does not look right, although the total including newly commissioned, looks about right for the beginning of 1942. I have considerable documentation that there were 171 on Dec. 7, 1941 when the Japanese attacked and I know there were commissionings between then an the beginning of the year

  9. A source of distortion of the results is that while the quality of the opposition changed over time, the number of ships in the various types did not did not change in the same proportion.

    The first year was much more dangerous than the succeeding years. After mid 1943 the Japanese were no longer a near peer.

    Ships that were only completed late in the war, were therefore less susceptible to damage.

    Looking at the force levels at the beginning and end of the war,

    Fleet carriers (CV and CVL) increased by 300%
    Battleships by only 35.3%
    Cruisers by 94.6%
    Destroyers by 120.5%

    CVEs went from one to 71.
    DEs went from none to 361.

    1. This is part of the reason I looked only at ships that were in commission at the beginning of the war. With very few exceptions they served the entire war if they were not sunk.

Leave a Reply