Friday, September 6, 2013

VLS - Are More Better?

When people describe ships that they’d like to see designed, one of the most common characteristics is more VLS cells.  Seems reasonable, doesn’t it?  More cells means more offensive or defensive power, right? 

Well, there’s a few things wrong with that concept. 

For starters, there’s the matter of inventory.  Our current fleet has around 9000 VLS cells.  What’s our inventory of Standards and Tomahawks that could, potentially, fill those cells?  No one knows, or, at least, I don’t.  However, going back over purchase announcements for the last several years suggests that our inventory of Tomahawks is around 3000 or so.  If the Standard inventory is similar, that would suggest around 6000 missiles to fill 9000 cells.  Hmm …  We would seem to be a bit short on missiles compared to cells.  If true, that suggests that the desire to put more cells in ships may not be a worthwhile design goal.  Now, I readily admit that I don’t have definitive numbers for the missile inventory but I have done my homework on what’s available in the public domain.  I don’t think I’m too far off one way or the other.  I’m making some allowance for usage, both combat (we seem to periodically shoot off a hundred or so missiles in various conflicts) and training, as well as shelf life limits and electronic diagnostic failures. 

Of course, there’s no requirement to be able to fill every VLS cell at the same time.  Even in a war a certain percentage of ships would not be deployed at any given moment.  That said, the difference between an inventory of 6000 and a capacity of 9000 is well beyond the vagaries of deployment schedules.

VLS - More or Less?

 The example of Harpoon missiles is instructive.  While we technically have an inventory of thousands of Harpoons, relatively few of them are serviceable.  They’ve all exceeded their official shelf lives and are being rotated into storage as they fail their diagnostic checks.  We have very few usable Harpoons left.  I don’t know the situation as it applies to Tomahawks but I suspect that shelf life is a serious issue.

Another problem with larger numbers of VLS cells is the impact of combat loss of a ship.  Historically, most (all?) ships that are sunk do so with the majority of their magazines unexpended.  Thus, the more loaded cells there are on a ship, the more unexpended missiles we’ll lose when a ship is sunk.  This goes directly back to the inventory issue.  To take an extreme example for illustrative purposes, if a ship with 1000 loaded VLS cells is sunk, we’d probably lose 95% of the missiles still unfired.  That’s quite an inventory loss for little gain.  On a related note, this is one of the problems with the old Arsenal Ship concept.  It’s a lot of eggs in one basket.

The only time we’d really need large numbers of VLS cells on a single ship is for high volume Tomahawk land attacks or, to an extent, saturation or long duration AAW scenarios.  Even the saturation AAW scenario probably doesn’t justify too large a number of cells.  Remember, only a relatively few missiles can be in the air at any given time.  Thus, having a hundred or two hundred cells worth of Standards won’t get more than a small portion of them into the fight before the attack is over – incoming high subsonic or supersonic missiles don’t allow for a very long engagement window.  Land attack would benefit from larger numbers of VLS cells (the Arsenal Ship concept), however, it also suffers from the “all your eggs in one basket” risk.

Most people seem to think that the more VLS cells a ship has, the better its ability to conduct AAW and defeat incoming missile attacks.  The reality is that a single ship with a million VLS cells is no better than a single ship with six VLS cells in a single missile engagement.  This is due to the illuminator limitation issue (you can only guide a few missiles at a time even with terminal-only guidance) combined with the very short engagement window for a modern anti-ship missile. 

The only benefit that additional VLS cells offers in an AAW engagement is that the ship can conduct AAW for a longer period of time.  Again, though, remember that even a saturation attack won’t last very long.

These issues suggest that there is an optimum balance of number of cells versus risk.  I don’t have the actual performance data to be able to state what the optimum number of cells should be but I suspect that around 100 cells is about right – basically a Burke.  Much more than that and you run the risk of losing too much inventory for too little gain.

The exception, or modification, to the risk issue is the SSGN.  Due to the inherent stealth of the SSGN it can carry more cells for the same relative amount of risk.  Indeed, the 150 or so cells the SSGN carries represents a potent offensive strike while minimizing the risk.

Now, if the Navy ever develops a new anti-ship missile that is VLS capable that will change the equation.  Additional cells will be needed to field a credible surface strike.  That may change the ideal number of cells but it won’t change the concept that there is an optimum balance point.

All that’s fine but what if we want to accept the risk?  What’s the problem with having more cells?  Well, the answer is cost.  VLS isn’t free.  Aside from the obvious cost of the VLS module itself, each module consumes precious ship’s internal volume and deck space, which translates to larger ships and bigger engines to move the ship.  That in turn requires more fuel, the storage of which requires a larger ship.  In addition, the VLS modules use the ship’s utilities (power, water, etc.). 

We see, then, that VLS cells should be added only if absolutely required to meet the ship’s mission.


  1. Don't forget Quad packed ESSM and ASROC.

    There is an effort underway to move other systems to a quad pack like Nulka and RAM.

    There was also an effort to put MLRS and ATACMS in VLS tubes that could resurface.

    1. SWP, quite right. Additional vertical launch weapons would increase the optimum number of cells. I'm not aware that we have any significant number of ESSM or ASROC, at the moment. Do you have any info on that?

    2. For ESSM, pretty much every ship with VLS carries ~32 ESSM as intermediate point defense. Though that's only 8 cells.

      And LRASM is coming to give USN ships back an AShM capability via VLS.

      And while I agree we shouldn't go overboard on VLS cells, having an extra block or 2 (8-16) for wartime surge capability is a good idea. And yes, while empty VLS cells do have cost, the costs aren't that high.

    3. ats, do you have a source for the ESSM deployment numbers? I've wondered about the extent of ESSM usage but I've never found anything about it.

      If we have use (ESSM or LRASM, for instance) for them, a few more VLS cells is completely warranted.

      The problem with with statements like "the costs aren't that high" is that all the "not that high" costs eventually add up. We can make a ship a little bigger because the added steel costs aren't that high. We can increase the magazine because the cost isn't that high. We can add another sensor because the cost isn't that high. We can put a bigger gun on the ship because the incremental cost isn't that high. We can add more UAVs because the cost isn't that high. We can ... Well, you get the idea. Individually, the statements are somewhat true (though there is a strong tendency among observers to trivialize the actual costs!) but by the time every minor addition is accounted for you've got a major cost problem.

      In this era of severe budget constraints, I look at ship design as being a case of adding nothing to a ship that isn't absolutely required to meet the bare minimum mission requirements. Sure, more would be nice but we don't have the budget for "nice" any more.

    4. Deployed numbers? No. I do know that Raytheon has produced 2000+ missiles to date from the production line and at least some Burke-class ships are listed as carrying it per the navy website. I would assume that the first priority would be replacing RIM-7 SSMs on ships without longer range AAM capability first. Apparently the Carriers themselves were among the first US ships upgraded. I assume they are upgrading the burkes as availability and time allows. It certainly ups the burke capacity with 4 per VLS cells vs 1 RIM-7.

      And I certainly agree with your viewpoint on "the costs aren't that high". But I think the VLS numbers for something like a burke should be pretty much a priority since it is an entire ship build and designed around VLS. At a minimum, the ability to upgrade to more VLS cells should be designed in. Even just having an area that can be retrofitted with 16 SDLS length Mk41 VLS cells provides a lot of capability if we need it.

      Part of the issue is that the threat requires more AAM and that offensive capability also required for cells for new missiles such as LRASM.

      I look at it as almost a basic military requirement. A reasonable Alpha strike from a small fleet can effectively mission kill a CSG given the current VLS capability which is very bad. If we ever have to get into a naval war, the current number of VLS cells is insufficient.

    5. ats, are those production numbers for the total world-wide market? My [vague!] impression is that the USN was not the major buyer for some period of time. Maybe 30% of the production total is US purchases?

    6. 2000 was the produced number as of Aug 2012 from a Raytheon press release. Production doesn't seem to be more than ~200 or so per year currently.

  2. Remember that it's much better to have plenty of space to put weapons that you have yet to buy, than no place to put a weapon after you buy it.

    1. While I agree in concept, the reality is that empty cells cost money, as I described. If there were no fiscal constraints, we could build every ship with 10000 cells. The reality is that if there's no realistic expectation of filling them at a reasonable cost/risk ratio then they're just a waste of money.

  3. From the perspective of a single engagement, you're right. More VLS cells doesn't buy you much.

    The problem is, in a campaign against a near-peer adversary it's not just about just single engagements. You have to consider multiple engagements over an extended period of time, originating from ships, submarines, aircraft and ballistic missiles. That's when increased inventory pays off, because it translates into increased time you can keep your carriers in striking range before being forced to withdraw them. In an A2AD environment, you're in a race against time. Eventually, something will get through your defenses. Defensive weapons only delay defeat. The trick is to be able to continue to deal offense with your strike capability while shielding yourself against the enemy's blows longer than the other guy can do the same.

    1. Quite right. As I stated, the "benefit that additional VLS cells offers in an AAW engagement is that the ship can conduct AAW for a longer period of time."

    2. How many AShM could the average Carrier group defend against at one time before the defenses are saturated completely?

    3. TR, well that depends completely on your assumptions about effectiveness of the AAW weapons (I assume you mean without the aircraft contribution!). My best assessment of Standard missile effectiveness is a 10%-20% probability of hit for a single missile against a single attacking missile. The Navy and manufacturer would have you believe 80%-90%. So, depending on the assumptions, the answer is somewhere between the group can't stop hardly any to the group is nearly invulnerable!

      The unfortunate part is that I don't think even the Navy knows how Standards will perform. Their testing is so contrived and unrealistic that their results are strongly skewed. Only actual combat will give us the real answer and that answer is likely to be very, very disappointing, I'm afraid. I've covered this in previous posts and shown that all the available data strongly suggests that AAW performance will be poor.

    4. TR, the rules of thumb I've been using at 4 AShM required per kill on average without active AA intercept for kill/mission kill. For AAM, 4-6 AAM per threat for kill/effective kill, but that is realistically a combination of area defense and point defense AAM.

      On average, a combatant vessel will have 8-16 AShM (this is basically from FAC to destroyer/cruiser) currently. So assuming you are against an equal size force, for every Area defense ship, you will need (W x (8 or 16) x (4 or 6)) / (number of area defense ships). Where W is the fleet size.

      A US carrier strike group (CSG) generally consists of ~4 CG/DDG plus the carrier and a support ship. So for a rough calculation it would be:

      (6 x 8 x 4) / 4 to (6 x 16 x 6) / 4
      so 48 to 144 area defense AAM per CSG.

      Obviously if the enemy fleet is packing 16 AShM per ship, you are going to need some heavy aviation support and a good bit of luck.

      Against an advesary like China, the likely strike force again a USN CSG is going to be 2 squadrons of 4 Type 022 (8 AShM each), two Type 056 corvettes (4 AShM each), and probably a total of 2 Type 052/054 destroyers/frigates (~8 AShM each). Total AShM against would be 88 missiles. We'll assume that the Carrier has 8 SHs either in the air or on ready alert and that each SH can get being generous 2 kills. That leave 72 missile that the AAW CG/DDG will have to handle. Which will require close to 288 AAW missiles (at a 4-1 ratio). We'll assume that the ratio of MR/LR AAM to SR AAM is ~2:1. So that's 192 standards and 96 ESSM. Or, ~48 SM-2/3 and 24 ESSM per DDG/CG. Even if no AShM hit the CSG, that would be an effective mission kill against the CSG as it would have to withdraw and resupply. That would deplete over half the VLS cells in the CSG, almost all the ESSM, and the vast majority of the SM-2/3s carried by the CG/DDG.

      This is part of the reason I believe we need additional VLS cells on the ships so that we can surge more capacity if needed. For something like a Burke-class, having an additional ~16 tactical/Self-Defense length VLS cells would enable the CSG to stay in the fight a lot longer. And then you add in that in an actual conflict we are going to likely want 16 LRASM per DDG/CG, that's another 16 tactical length cells that would be required. So I think it is in our interest to add another 32 VLS cells to future burkes/DDG even if they are only Tactical/Self Defense length. The SDLS lenth Mk 41 VLS is only 17 feet high and we should be able to figure some place to put them.

    5. ats, nice discussion of the requirements for an engagement. A couple of thoughts for you ...

      You're implicitly assuming that every incoming missile can be engaged. If missiles were to arrive simultaneously there is a good likelihood that some would get through unengaged due to the balance of speed of approach versus numbers of illuminators.

      You're also implicitly assuming that we'll have time to engage each missile with the requisite number of AAW missiles. Consider, though, the case of incoming supersonic (say Mach 2) missiles being detected at 50 miles. Even assuming an instantaneous response launch, the engagement window will only be 60-90 seconds - not enough time to get the requisite number of AAW missiles into play. Remember, the engagement sequence is shoot-shoot-look, so as to avoid interference. By the time the first sequence is done and ready to be repeated, the engagement is probably over, one way or the other.

      The overall point is that I suspect we can go several actual engagements because each "battle" will use far fewer missiles than you've calculated. Of course, we may be sort a few ships after each engagement! Obviously, the farther out the incoming attack is detected and the slower and more spaced out the incoming missiles are, the more your description applies.

      My assessment is that we would be very fortunate to get four longer range shots at a single missile and one shorter range shot before the engagement window is closed. More likely is 2+1.

      Unless we're looking at Badgers/Backfires launching from 40,000 ft up and a couple hundred miles out, we're probably going to be fortunate to detect (and track!) incoming missiles 50 miles out which leads to very short engagement windows.

      What do you think?

    6. I somewhat agree. And I certainly don't think the current CSG can do it. I was assuming a couple things...

      Full time airborne radar, E2-D or similar.

      SM6 instead of SM2, so obviously not viable at this time as SM6 isn't deployed yet. SM6 is capable of ARH and hand-off to air based assets for illumination and targeting. So fending off any future attack is going to be highly reliant on SM6 being deployed.

      Detection close to launch location via E2-D, UAV, or SH/F-35.

      Full battle alert, if it isn't squawking the right IFF, its fired upon without delay. See it, shoot it level of alert.

      Rippling out SM6s as fast as possible with intermediate guidance to put them into position with final targeting likely handed off to farther out airborne assets either E2D/SH/F-35.

      Burkes themselves would be handling illumination primarily for ESSM.

      And no I don't think the capabilities to handle such an attack are there yet, but I do believe it is a realistic scenario. Certainly if I was say china, I would gladly trade 8 FAC, 2 Corvettes, and 2 destroyers/frigates to either mission kill or actual kill a CSG. And the most realistic way to do that is an alpha strike from a couple hundred nmi with every AShM on hand. Do as much as possible to overwhelm the defenses and then turn tail and run as fast as possible, cause it would be nice to have a couple ships make it back. If not, its still a good trade-off.

      So I think if you are building ships and systems for the Navy, it needs to be a scenario that has to be handled.

      So I find it a pretty realistic scenario for attack on a CSG, how about you?

    7. ats, I have no problem with your scenario. I think it's on the optimistic side as far as kill probabilities but there's no data to back that up so your guess is as good as mine.

      Presumably (I really, really hope!), the Navy games this stuff out on a regular basis with better data than we have and takes their design cues from the results.

      I have severe doubts about the hand-off targeting ability for several reasons. Hmmm ... Maybe another post topic?

      Anyway, good discussion!

    8. hand-off is actually less complex than it seems. SM2ER IVs already someone function correctly for this. They don't use full time illumination, instead using illumination only for the terminal phase. Prior to that, they rely on inertial navigation.

      The inertial navigation has 2-way uplink capability for mid course correction.

      The hand-off capability is switching the source for the mid-course corrections and the terminal illumination. The only changes required for the missile are a bit of software to enable the missile to switch to or accept multiple coding signatures.

    9. ats, the aspect of hand-off, in the way you've described it, that I have doubts about is the co-ordination among multiple final illuminators. You've suggested that multiple aircraft would be passed the job of final illumination. That's fine in theory but it runs the risk of having multiple aircraft illuminate the same target or, conversely, no aircraft illuminate some targets. The ability to flawlessly handle the illuminating assignments requires a master program to co-ordinate assignments among multiple platforms in absolute real time (no comm delays whatsoever, no jamming effects, etc.) - in other words, CEC on steroids. Currently, even our "simple" Link XX is somewhat hit and miss and is nowhere near real time. Due to both hardware and software problems, I don't believe we'll be capable of do this in the foreseeable future.

      Frankly, we don't even know that our current mid-course guidance and terminal guidance capabilities will actually work in a high ECM/jamming environment since the Navy never tests under those conditions. I can easily imagine that mid-course guidance commands will not be cleanly transmitted and received in an electronically challenged environment. I'm sure there's some software protocol (actually I'm not sure but it's the way I would do it) that does a back and forth command and confirming response with a repeat, if necessary. However, at the speeds and geometry of a missile engagement, it wouldn't take more than a momentary disruption for a missile to miss or fly out of the intercept window.

  4. The theorizing is interesting, but how many missiles has the USN actually used in attacks? Iraq?

    1. If you're asking how many Standards (or any AAW missile, for that matter) has the USN fired in the AAW mode, the answer is zero, to the best of my knowledge.

    2. I'm asking how many missiles (Standards, Tomahawks, and Harpoons) the USN has actually fired in its campaign. So, zero Standards. I'm guessing that the Harpoon number is close to zero as well. How many Tomahawks?

    3. Total, I'm not sure what campaign you're referring to. Do you mean the total ever fired? Do you mean in some specific conflict? As a general answer, a couple Harpoons were fired during Desert Storm. Probably a couple thousand or so Tomahawks have been fired in Desert Storm, the Iraq wars, and various other conflicts. ESSM and RAM have never been fired as far as I know. If you're looking for more information, you're going to have to be more specific in your question.

    4. Current SMs, RIM-7 Sparrow, and RIM-162 ESSM have rarely ever been fired.

      Early block SM1s/SM2s were fire in the 1980s during the Iran-Iraq war both at Iranian F-4s with minimal effect and against Iran Air Flight 655 with lethal effect.

      In 1988 six SM1s were fired at an Iranian Corvette and at least some hit it destroying it.

      Harpoons have been fired rarely and is a quite aged design.

      Tomahawks are fired seemingly on a yearly basis in anger.

    5. ats, I'm unaware of a Standard missile launch in the Iran-Iraq war. Do you have any details about that?

    6. Data is from:

      The cites all appear to check out.

      Iran Air Flight 655 is verified by the US government.

      The Joshan corvette sinking is documented.

      The F4 cite is the weakest as the only citing authority is the Iranian Airforce.

      So at least 8 missiles for fired between the Joshan and Flight 655 incidents, and however many you want to believe from the IIAF accounts.

  5. Some beyond visual range TALOS missiles fired during Vietnam, since then NADA, I believe.

  6. Hi All, Nice article by the way. thanks.
    Input from Britain again I’m afraid ;)
    I think ,and apparently the RN agrees, that the VL question is probably as the article suggests. reload at sea is possible ( in certain sea states ). And until VL Harpoon or equivalent is available the only use for 100's of cells is for a saturation air attack.
    Now Daring class AAW destroyers can apparently field 10 missiles to 10 targets every 30 seconds ( if the newspapers are to be believer ) as fire control illumination is not required. [So in the above situation you might stand a chance].
    but whilst you are needing illumination I cant
    see justification for 100's of VL.

    Is SM6 not autonomous ?

    1. SM6 supports 3 different modes: active, semi-active, and co-op engagement. Basically, any guidance mode that the AIM-120 can do, the SM6 can do. This is because the SM6 is an SM2ER with the seaker/guidance package of the AIM-120C/D.

      But no ships are currently equipped with SM6 as full rate production just started this year. Current scheduled crossover timeframe is 2015 for full deployment. Currently only 1200 SM6 missiles are on order and a cost per missile of 4.3 mil.

      So when the fleet switches over to SM6, they can all be fire and forget via ARH, or ship illuminated for terminal via SARH, or can be launched and then controlled by other platforms such as F/A-18 and AWACS.

      Now, SM2ER isn't limited by illumination for launch or mid-course but does require target illumination for terminal guidance.

    2. Very Impressive !
      Roll on SM6
      Very wide area air defence, effectivly fleet defence
      At that point stacking 10's on your destroyers definatly becomes a good idea.

      Quite expencive for local and point defence.
      Any plans to make a baby SM6 like Aster15 or sea ceptor ?

    3. For local and point defense the USN is using ESSM and RIM-116 RAM.

      ESSM is a SARH missile with a 27nm operational range.
      RAM is a anti-radiation/infrared dual mode missile with a 6 nmi range.

      ESSM can be quadpacked into a MK41 VLS cell. RAM requires a separate launcher system but is bolt on with no through deck requirements. Apparently they are also working on a VLS Cell pack for RAM.

      But yes, SM6 is expensive, and it would be nice if they eventually ported a ARH capability to the ESSM.

    4. Thanks very much. All of you. This is a great site, for measures sensible debate and information. Fast becoming one of my faves.

    5. ats,

      ARH is planned for ESSM Block 2.

  7. Its a complicated issue.
    Especially for the US.

    For the UK, I think the best "quick win" we could grab, would be to build another 6 T45s, lengthened to fit an 8x8 or 8x16 VLS "module".
    Given the extremely limited real world strike capability (once you strip out CAP) of our carriers, and the relative scarcity, the ability to add another 128 missiles launchable in a few minutes as a can opener to a carrier, or sending it alone to deliver a slap, would be a huge capability boost.

    But I dont see that for the US

    I'm still a proponent of the Arsenal Ship.
    Minimally manned, the UKs point class have a crew of 22.
    Given that a VLS unit is essentially sealed, does it require any crew hours?
    I'd probably stick a marine platoon on there just to stop the ships being hijacked

    They dont need to be fast, they dont go near combat, ever, they certainly dont "keep up with the fleet. In the event they need to "run away" they shoot instead, you've got 500 missiles to hit what evers chasing you, worst case, you sink the ship and await pickup

    1. The main problem with an Arsenal VLS Ship is that while the VLS are cheap, the missiles in the VLS are not so cheap. So if you are going to have a 50-100 VLS cell Arsenal ship, it is going to have to have adequate self protection.

      And the idea of a 500 missile Arsenal ship from a financial perspective scares me to death. Lightly defended 5+ billion dollar barges aren't a good idea.

    2. Think of it as a cheap carrier rather than an expensive cruiser.
      It could be body guarded by a cruiser is needed, but it could strike 500 targets in Tehran whilst anchored in Cyprus, or Beijing from Hiroshima. Once its fired, and it would be a zero hour thing, it sails off for home, or at least the deep waters.

    3. You don't have to stick 500+ missiles on a single ship. The NPS New Navy Fighting Machine proposes a mini-arsenal ship with 50 missiles.

      IMHO much depends on cost and manning.

    4. If the Virginia Payload Module is developed we'll have mini-arsenal ships, I guess. They would put 40 Tomahawks on each sub, if I remember correctly.

    5. The VPM will increase the price of the Virginia class up even further. They are already pushing $2.8+ billion each according to the Navy SCN Justification books.

      One side benefit to a relatively cheap and relatively numerous VLS arsenal ship is that during a prolonged campaign, we could cycle arsenal ships back to base for reloads while not pulling out destroyers, or subs.

  8. Instead of an arsenal ship, what about a heavy guided missile cruiser? It could pack a VLS arsenal of 300-450 missiles, and it can defend itself adequately.