Jeffrey LewisHow Capable is the 094?

Behold, the second installment of the series on China’s SSBNs. This took a while, mainly because I wanted to reread Tom Stefanick’s Strategic Antisubmarine Warfare and Naval Strategy. It’s a classic.

Anyway, the bottom line is that I doubt China’s boomers would really be more survivable than the land-based leg of their triad. That reflects three factors: the noise level of the submarine, the range of the JL-2 and China’s geographic location.

How Quiet Is It?

The most common measure of capability is the amount of noise that the submarine makes. A noisy submarine is a dead submarine.

I see a lot of comparisons thrown around about how quiet Chin’s 093 and 094 submarines are, with comparisons to the Victor III and Los Angeles-class (USN 688) submarines.

Data for that is hard to come by, but in 1997, the Office of Naval Intelligence released a chart comparing the performance of China’s new nuclear powered attack submarine (SSN), the 093, on which the 094 is said to share a design heritage.

Obviously, ONI didn’t release any numbers. But thanks to Tom Stefanick’s classic Strategic Antisubmarine Warfare and Naval Strategy, we can place China’s SSBN between 130-150 decibels (with a lot of specific qualifications outlined on pp272-279).

ONI’s estimate was done before we actually saw the submarine, however. An interesting comparison would be to compare the volume of the 093 to the Victor III. Modern quieting involves sound isolating mounts; Stefanick observes a relationship between quieting and larger submarines.

Sadly, I haven’t seen really convincing estimates of the surfaced displacement of the 093 or credible dimensions on the submarine to make a comparison with the Victor III. It would be nice if we had a picture of an 093 sitting at the dock.

Could China Conduct Deterrent Patrols?

The limited range of the JL-2 is a second major constraint. Various official US sources place the range as 4500 miles, 4000 nautical miles or 8000 kilometers.

NASIC is fond of saying the JL-2 will “for the first time, allow Chinese SSBNs to target portions of the United States from operating areas located near the Chinese coast.”

The operative word here is portions … as in Alaska. I may not be willing to trade Los Angeles for Taipei, as Xiong Guangkai (or whoever) suggested, but Anchorage is another story. Or, at least, the Chinese have to worry about that.

To target cities in CONUS – say our latte sipping friends in Seattle, for instance — Chinese boomers would have to patrol the deep waters of the North Pacific – a very long round-trip. Indeed, ONI’s suggestion that five submarines would be necessary to keep only a “near continuous” deterrent at sea suggests ONI’s analysts have reached similar conclusions about the distance of patrolling areas from China’s submarine bases.

(More on that below).

Another operational concern that points to North Pacific patrols is the need to hold at risk both Moscow and populations centers in the Continental United States. The actual operating area might be slightly larger if China’s leaders were willing to have submarines patrol out of range of some targets until receiving the order to fire.

To reach the North Pacific, a Chinese boomer would have to transit the narrow gaps between South Korea, Japan, and Taiwan – a kind of Asian equivalent of the GIUK gap — to reach areas in the North Pacific from which the JL-2 could target significant portions of the United States.

This island chain offers a number of opportunities for fixed sensor emplacements, similar to the SOSUS system, that should provide ample opportunity for US attack submarines and other US and Japanese ASW platforms to pick up and trail the boomer on patrol. Although I wasn’t cleared the see the good stuff during my stint on the Japan desk at OSD, a review of press reporting suggests a high level of knowledge about Chinese submarine operations in and around Japanese waters.

Once out in the deep water, Chinese SSNs and SSBNs would be extremely vulnerable because deep water propagates sound very effectively. Stefanick estimates the Los Angeles-class submarine has a 25-100 nm detection advantage over the Victor III, which is probably quieter than the 093 and 094 class submarines. That reminds of a line in that silly e-mail about Chuck Norris: If you can see Chuck Norris, Chuck Norris can see you. If you can’t see Chuck Norris, you may be only seconds away from death. Replace “see” and “Chuck Norris” with “hear” and “Los Angeles-class” and now “Virginia-class” submarines.

Range is just one of the operational concerns, of course. China would also have to think about secure, survivable communications systems for patrolling boomers. This is long topic, however, and brings up much larger discussion about command and control issues.

Hainan Island?

Although the Boomer was spotted up North, near Dalian, some folks think China will eventually base the SSBN at or near the Yulin Naval Base on Hainan Island. I was skeptical that China would base the SSBN so far south. Indeed, I would have thought that basing the submarine so from from a patrolling area against North American targets would suggest a Russia-oriented role for the 094.

But ONI and others, however, seem to believe that the 094 will make the 9,000 km round-trip between Hainan Island and the patrolling area that I outlined. Assuming that the submarine travels about 10 knots (18-19 km/hour) to avoid cavitation that would compromise the location of the submarine, the round trip from Hainan to the patrolling area takes about 10 days each way (the trip is about 4300 km). Assuming the 094 has an 80 day endurance, that leaves a 60 day patrol.

All of a sudden, that ONI estimate of 5 submarines to maintain a “near continuous” presence makes sense. 5 submarines for a sixty day patrol works out to about 300 days a year. Possibly ONI estimates that the SSBN’s have a little more endurance or that the Skippers are willing to run a few more risks to get on station. But the numbers work out about right.

Someone told Demetri Sevastopulo that basing the 094 on Hainan Island “would give its submarines easier passage and make them harder to track and target.”

Easier isn’t the same thing as easy. The submarine still has to pass through the Luzon Straitm which is relatively narrow (360 kilometers, broken into smaller passages by several islands).

Admittedly, the Luzon Strait is a complex littoral environment with lots of shipping noise in which Chinese boomers could hide and get out to sea. But the Navy can still study, model and line with the Strait with sensors. This fact is not lost on the Office of Naval Research, which seems to be very interested in the “unique oceanography observed in the South China Sea” and seems to have funded a fair amount of research including the Asian Seas International Acoustics Experiment (ASIAEX), Windy Islands Soliton Experiment (WISE) and god knows what else.

And, of course, once through the Luzon Strait, a skipper still has the problem that deep water propagates sound.

We might need a few more boats to get this done (I haven’t done an estimate), but I find it difficult to imagine problems that could not be solved by giving Representative Courtney his extra attack submarine for his district.

On the other hand, if China envisioned the 094 as solely a deterrent against Russia, the operational challenges for the 094 are less daunting. Moscow is in range of shallow offshore areas that are near China’s naval bases and far away from Russia’s submarine and other ASW forces.

The fact that China is building an SSBN with substantial Russian assistance that seems far more better suited to striking Moscow … well that strikes me as a pretty decent indicator of the degree to which technological imperatives – rather than operational requirements – continue to drive the SSBN program.

A Surge Strategy?

Were China to keep a deterrent on patrol in the Northern Pacific, I would expect that the United States Navy would be quite capable of detecting the submarine coming out of port and trailing it on patrol — absent criminal negligence on their part or, more likely, those who fund and equip them.

Last week, I suggested that the jury was still out on whether China will send its deterrent to sea. Although countries have done crazier things, the technological limits of the missile and submarine interact with the unfavorable geography for China’s boomer fleet in such a way to undermine the case in favor of putting the deterrent out to sea.

Of course – as I said last week — bureaucratic imperatives might outweigh solid policy analysis. But I don’t judge the 094 to be very survivable platform, given US intelligence estimates, for deterrent patrols against the United States.

Other strategies are, of course, possible. I am kind of playing around with the idea that China might build a small number of SSBNs that remain in port, but that could “surge” out to sea during the early phases of a conflict. A lucky Captain might slip through if US and Japanese ASW assets are busy with China’s diesel submarines.

That, of course, raises the prospect of a serious miscalculation during a deep crisis, which just happens to be the subject of my next post in the series.


  1. hallo84 (History)

    If the type 093 was launched in 2002 as stated by DOD then what were the basis for ONI to come to a conclusion of 130-150db in 1997?

    It’s quite common that the progress of PLA developments are under estimate. The recent public photos of Type 093 show no external similarities with Victor III but instead seem to be much more close inline with western SSN designs may point to less Russian involvement then expected.

    But I do agree with you on the operational difficulties of Chinese SSBN. But how many boats do PACOM need in the asian theatre to cover all 5 SSBNs?

  2. Tim H (History)

    But, Jeffrey,As of 1997, the 093 had not even left the dockyards (IIRC the first sea voyage of the 093 was in 2001)

  3. Jeffrey Lewis (History)

    I know, I know … the projections were exactly that … projections. That is why I said I wanted to see a picture of the 093, to see if the submarine was substantially larger than ONI predicted.

    Bob Suettinger prepared a report for then-incoming DNI John Negroponte on China that reviewed how well the IC performed. Although Gertz reported that the review identified “missed” military developments, he didn’t mention the 093 being extra special quiet. Of course, Gertz’ stories need to be treated with caution. (“Analysts missed Chinese buildup,” _The Washington Times_, June 9, 2005, A1)

    In terms of having to deal with five SSBNs at once, that might make sense were China to adopt a “surge” strategy, but ONI seems to be estimating that China would need five SSBNs to maintain one on continuous station (with other boats in transit or being refitted for the next voyage).

  4. hallo84 (History)

    you asked for a picture here it is…

  5. Mark

    Why isn’t the Red October on that chart?

  6. johnwbragg (History)

    Wait, does Gertz have a history of underhyping PRC capabilities anywhere?

  7. Allen Thomson (History)

    — hallo84 contributed

    >you asked for a picture here it is…


    Interesting that there appear to be big plates on the hull over the engineering spaces (reactor, turbines, gearing) aft of the sail. I wonder if they’re removable to facilitate maintenance and overhaul.

  8. Robot Economist (History)

    Dr. J – If there are two things the JMSDF can do, it is demining and ASW—heck, even their DDG Kongos have ASROC launchers on them. You could almost say the 88 Formation is optimized for ASW missions.

    Whether the JMSDF is up to the task of patrolling potentially hostile, sub-infested waters is another question. They may need support from the USN ASW planes to stiffen their backbone a little.

  9. palolo lolo (History)

    Nice pic. Don’t want to know how that happened,tho

  10. Jian Feng (History)

    What if the range of JL-2 is much more than 8000 km? Which is technically more difficult, extending the range of the missiles to 13000km or building 5 boats to patrol? The idea of patrol sounds very silly to most Chinese. Since China practices no-first-use, the best strategy would be to simply let the sea hide the boat/missiles.

    It’s naïve to think that China builds this expense toy just to scare the Russians. Such analysis suffers from the same mistake that a Mr. Butterfield made on May 19, 1980 on the New York Times, the day after the first successful full-range flight test of DF5. He said that the bird was intended for the Russians, but it flew way to far – 10000 km. Someone in the White House probably did not want him to mention that the missile is intended for CONUS targets, for fear that it might spoil the honeymoon between Beijing and Washington at the time. Pundits in the US still like to say similar things, but for a different reason – to warn the Russians of not letting out too many secret weapons to China. But Putin does not seem to mind. He knew that all the good stuff was concocted by the Chinese anyway.

  11. None (History)

    There are a few flaws with equating a boomer’s capability to hide with the amount of noise it generates. Yes, noise is always an important first approximation, but there are many other factors.

    Detailed nowledge of terrain is aonther. Chinese naval forces have neither invested or acquired by other means the detailed knowledge of the seas around the Asian continental shelf, and the deeper oceans they need to transit or operate in. Such information as detailed maps of where thermal and saline layers are, how they shift with the seasons, can only be gained with extensive operating experience and research. SSN-711’s collision with an undersea mountain shows that even the most experienced fleet have gaps in their knowledge that can lead to serious accidents.

    Chinese submariners only have a limited understanding of their potential hunters, primarily from their experience with Taipei and Japan’s fleet, and have not had the experience of playing ‘cat and mouse’ accumulated by the Soviets during the cold war. Moreover, because their experience was gained with largely obsolete equipment rather than with first rate equipment, it is limited and of questionable utility in the current context.

    Then there is the issue of expertise / experience / and doctrine of the crew. A great deal of the amount of noise generated by a submarine can be attenuated (even with the noisiest subs) with crew training and learning to operate around a boat’s limitations. It may be something as simple as tweaking the turbines in a particular way to cut out a certain frequency that propagates well in the particular waters the boat is operating in. These kind of tricks only come with experience that China is short on.

    ASW technologies have changed greatly since the 1980s. Instead of facing largely ‘dumb’ passive and active sonar arrays mounted on conventional ASW platforms and SSNs, (or permanently moored arrays like SOSUS), a Chinese sub is facing an array of underwater RPVs and Robot subs, portable sonar arrays that can be deployed on short notice, and also such devastating weapons as the yet to be made public American versions of the Shkval torpedos. Sensors are now so cheap and plentiful and intelligent that it is possible to virtually populate the entire potential patrol area of Chinese boomers with them. The days when ASW forces can only hunt and clear a small volume of the ocean is over.

    There are also new techniques made possible by advances in computational and sensor capabilities like detecting the slight perturbance in the ocean surface caused by the passage of a large submersible that can quickly give a gross estimate of a sub’s location. Whereas previous gross techniques like MAD can only cover a small area, these techniques can cover vast stretches of the ocean quickly, efficiently, and be used by a navy with a sophisticated command and control system to direct ASW forces to the Chinese subs. Because American and allied assets are being built on a network model, these asset’s capabilities far exceed the platform based systems from the cold war.

    The bottom line: much of the SSBN doctrine that analyst have relied on to understand and explain the threat poised by the 094 subs may be out-of-date. For example, the old rule of thumb that it is easier to build a quieter sub by building a bigger one may no longer apply as that slightly quieter sub is likely to have a much bigger signature detectable by other means. A larger sub also makes for a easier target and greater ease in knocking out a large percentage of the Chinese nuclear deterrent in one stroke.

    Finally, another issue worth considering is, is this whole exercise flawed by an assumption that China has a need to strike at the US mainland? Who is to say that their primary goal is NOT the US, but Taiwan and Japan, and the Chinese nuclear deterrent is in fact primarily aimed at preventing either from winning a war with China?

    Take this hypothetical scenario: Suppose war broke out between Taipei and China and Taipei launched a nuclear armed missile at China which caused millions of casualties. Beijing responded by leveling Taiwan with a barrage of nuclear weapons. Is it clear that war between the US and China would follow in its aftermath? Is it likely that the US would conclude that Taipei brought the war on themselves and decide to stay neutral?

    The 094 submarines is a straw man that raises many more issues than it answers. It is dangerous to read too much into their commissioning prematurely and to draw conclusions based solely on the American cold war experience.

  12. Alex (History)

    That chart of submarine types would carry more credence if its author knew that there is no such boat as a “Vanguard SSN”.

    Vanguards are Trident boomers..

    I’m surprised they don’t base it in, well, Shanghai – with a clear run to the Taiwan/Okinawa gap.

  13. Enoch

    “The operative word here is portions … as in Alaska.”

    Hello, Honolulu? As in, nearly a million people, major naval base, major air base, hub for Pacific power projection? Last I checked, Hawaii was still a portion of the US. How many Presidents would be willing to trade Honolulu for Taipei? The 094 wouldn’t have to get very far into the Philippine Sea to be able to target Hawaii. It is even reachable from portions of the East China Sea.

    What is the assumption about MIRVing here? Can a JL-2 with one RV reach CONUS (if indeed a threat to CONUS is the only thing that deters us) from the South China Sea? If so, that is an interim solution for the Chinese until they have a longer-range missile.

  14. Bob Melley (History)

    Given the “issues” that the PLAN had with the Han class Project 091, may we be guilty of believing that the PLAN has now a modern SSN/SSBN fleet that could stand to a prolonged conflict in the western Pacific…..China loves to deceive. Easy answer is to form a Joint Naval Patrol, composed of the USN, RAN, JMSDF, Indian navy & Singapore’s navy. The JNP would cover the ocean sea lanes from Japan all the way to Strait of Hormuz and back, equal legs…..ships and ASW a/c. Focus on the Strait of Malacca, South China Sea, etc…to help guard againstterrorists….and the PLAN…

  15. Jing (History)

    Speculation on the capabilities of Chinese subs have seemed to hit a fever pitch for the 80th anniversary of the PLA. Here are two additional pictures of what is possibly the 093 class of submarines.

    Anyone care to make a displacement estimate based on the size of the people or tugboat?

    Also note that there appears to be a second flight of the 093 submarines with the diving plane moved from the sail to the hull. The hull shape is itself even a little different, with a flat back. These photos are the most authentic, being photographs taken from the PLA military museum 80th anniversary expo, the first one actually a snapshot of a running video showing the submarine.

  16. Jeffrey Lewis (History)


    Um, I hate to break this to you buddy, but the world is round.

    Hawaii is substantially closer to the East China Sea than the South China Sea. Mercator maps were great for sail powered navigation, but they are woefully misleading in the age of ballistic missiles.

    No coastal area in China is close enough to allow the DF-31 to target Hawaii. It did occur to me, before posting, to check the ranges. Using Honolulu as the target, the JL-2 cannot reach Honolulu from Chinese coast waters using the 4,000 nautical mile or 4,500 mile range estimate. Although Honolulu is barely within the 8,000 km estimate I mentioned — a nice round number — most recent DOD range for the DF-31 is actually 7,250 km. So, no, the JL-2 probably cannot reach Hawaii.

    That assumes, by the way, a single warhead. The IC seems to assume that the DF-31 RV weighs around 450 kg and is too large to place more than one on DF-31/JL-2. So, that option isn’t available to increase the range.

    Thanks for reading and don’t feel too sheepish about the round/flat world thing. I do it too, which is I always measure distances between places using a globe or (now) software like Google Earth. Those awful grade school Mercator images drilled into our impressionable little minds are hard to shake off even in adulthood.

  17. Jeremy

    This post kicks ass.

  18. Jing (History)

    Mr. Lewis, the JL-2 is not the DF-31. While it is commonly said to be the submarine launched equivalent, the profiles of the missiles are different. According to 2006 NASIC report, the DF-31 is a 3 stage missile where the 3rd stage tapers significantly and is narrow enough to match the diameter of the re-entry vehicle. The JL-2 has a uniform diameter and is approximately the same size as the Russian Bulava missile and is likely to be moderately heavier than the DF-31.

    This is just theorizing on my part, but I feel that while the JL-2 and the DF-31 share technical similarities, they are not the same missile. I feel it is DF-31A variant, also with the uniform missile diameter that is in fact the land-based version of the JL-2 and not the other way around as it is commonly speculated.

  19. Stephen Young (History)

    So, you are saying (or implying) the ONI estimate of five Chinese subs (rather than the magic number four selected by France and the UK) is because the long transit time requires an extra sub?

    And each sub can ONLY make one patrol a year? A quick web scan showed Tridents reaching 50 patrols in 20 years (recognizing U.S. experience probably increases patrol frequency capability).

    Sorry to snark, but this strikes me as a little too much like using the old analysis of who is next to Brezhnev in the photos to divine Soviet intentions.

  20. Jeffrey Lewis (History)

    To: Stephen

    Sorry, I should have been more explicit about the calculation.

    I don’t know how long the Chinese refit period is, although given the number of Chinese submarine patrols we can conclude that the refit period is substantially longer than the 30 days to refit a US boomer. A handy rule of thumb was outlined by David Shambaugh, who noted submarines spend about 3/4 of their time in port.

    So, let’s assume that a submarine with an eighty day endurance spends 3/4 of its time in port, that would work out to just slightly more than one patrol a year.

    To Jing

    That is a very good point about the third stage that I had missed.

    NASIC, though, lists both the DF-31 and JL-2 as a 3 stage missiles with a 4500 mile range. DOD, in recent years, has dropped the range estimate of the DF-31 from 8,000 km in the late 1990s to 7,250 km today. I presume that change reflects the opportunity to observe DF-31 testing. I wouldn’t be surprised if we saw the JL-2 range estimate come down, and the picture evolve, with more extensive flight testing, as well.

    Still, I take your point that the JL-2 could be substantially modified to raise the range by something on the order of ten percent or so.

  21. Jeffrey Lewis (History)


    That was a really interesting point.

    I did a quick comparison, though. Based on the NASIC drawings, the third stages appear to be approximately the same volume, if a different shape. If anything, the DF-31 third stage looks slightly larger than the JL-2 third stage.

    Seriously, check my math because I could be wrong. But assuming that the DF-31 is 15.5 m tall, I come up with the following dimensions for the DF-31 and JL-2 third stage:

    DF-31: 2.2 m height, .75 m radiusJL-2: .9 m height, 1.1 m radius.

    If volume is equal to pi r-squared times the hight (v=3.14r2h) then the DF-31 stage is about 3.9 cubic meters of propellant, while the JL-2 is 3.4 cubic meters of propellant.

    Propellant volume ought to scale to range, although there may be some efficiencies, related to shape of the stage and the fuel, that I just don’t know about.

    DOUBLE CHECK THESE NUMBERS BEFORE USING. The NASIC drawings are to different scales and I am a little tired, so I could easily have made a mistake either in estimating the inputs or calculating the volume.

  22. palolo lolo (History)

    As a resident of Honolulu,these discussions of range are always important-toput it mildly

  23. Enoch

    Jeffrey, thank you for your condescending response. I was well aware before I posted that the world is round, and I checked the ranges myself with the Google Earth measuring tool before I wrote.

    You will note that my other post did not claim that the JL-2 can reach Honolulu from Chinese “coastal waters”. I observed that the 094 is within strike range of Honolulu from the middle of the Philippine Sea (roughly 135 degrees East) and portions of the East China Sea (specifically, just inside the Ryukyus between Okinawa and Kyushu).

    The point I wished to make, which perhaps was not clear, was that an 094 could reach an operating area that had significant “deterrent value” that was much closer to its potential home bases than the North Pacific / Kuriles area you indicated above. Based at Hainan, an 094 at 10 knots can be in the Philippine Sea, within strike range of Honolulu, in less than 6 days, as opposed to the 10 days you quote above. Based at Dalian, an 094 can reach an area in the East China Sea that is within strike range of Honolulu in less than three days.

    I will now return to coloring my old Mercator maps.

  24. Jeffrey Lewis (History)


    Apologies for being flip. And I wish I had thought of “coloring” mercator maps.

    But the discussion was about whether the SSBN could reach from Chinese coastal waters, or had to traverse a strait and operate in deep water, exposing itself to US and Japanese ASW platforms.

  25. Enoch

    Cheeck this out:

    “China has launched the first of five Type 094 Jin-class ballistic missile submarines. Each of these behemoths will carry 16 JL-2 ballistic missiles with a range of 5,000 miles.

    These missiles, sea-based versions of the land-based DF-31, which can reach Alaska and Hawaii, can be launched against mainland U.S. cities from the relative safety of Chinese coastal waters.”

    Ah, if only they read this blog, they would not be so poorly informed! Not only do they stretch the range, but even with 5,000 mi it could not reach Seattle from coastal waters.

  26. Bob Melley (History)

    This is a great site.Serious folks and good skinny.Question:volume/weight wise, can someone out thererate the JL-2 with newest Trident D-5? The 094 boats can carry 16 missiles vs. 24 in a US boomer. MIRV wise,where does this put USin a face-off vs. PLAN? Without either side actually the trigger…Intimidation factor should be vastly on US side.Comments…

  27. Ben Reilly

    My main question about the sound/timeline chart is about the Vanguard. How on earth did the Brits mess it up so badly, especially if “loudness” decreases with size?

  28. None (History)

    It is not hard to see the Brits mess up.

    Look at the Upholders and the mess they are:

    Diesel engines that are railroad engines that broke under the strain of repeated start/stops in submarine use.

    Improper insulation on power cable that caused the fire in the HMCS Chicoutimi.

    These are just some examples of design faults that suggest Britain have come a long way from being a first rate source of submarine technology.

    Perhaps Britain no longer build / operate the number of boats it needs to keep their skills at world class levels.