China's Nuclear Arsenal

Source: Yang Zheng
16 March, 1996
National University of Singapore
An anonymous poster ( sent an internal document of the Chinese Defense Ministry to the Hong Kong magazine The Trend (Dong Xiang). This document reveals that China at present has a total of 2,350 nuclear warheads. This number i s about 8 times larger than the 300 generally cited in the Western media. Among the 2,350 warheads are about 550 tactical nukes and 1,800 strategic nukes. The document also reveals that the annual production of warheads was about 110-120 i n the 1980's and about 140-150 at present.

Those figures are reasonable. According to China Built The Bomb (a book published in U.S. during the late-1980's), data from various U.S. intelligence agencies show that, in the mid-1980's, China was producing at least 800 kilograms of U-235 and 4 00 kilograms of Pu-239 per year:

  1. Lanzhou Gaseous Diffusion Plant 400kg U-235/year
  2. Helanshan Centrifuge I 400kg U-235/year
  3. Helanshan Centrifuge II ???
  4. Yumen Breeder Reactor 250kg Pu-239/year
  5. Baotou Breeder Reactor 150kg Pu-239/year
  6. Guangyuan Breeder Reactor ???

A typical fission-only nuclear device (N-bomb) in U.S. and Russian nuclear arsenals uses an average of 15 kilograms of U-235 or 5 kilograms of Pu-239, with a typical yield in tens of kiloton TNT. In fact, a fission nuclear device can be built with just 1 .8 kilogram of Pu-239 by using the neutron reflection/multiplying technique with U-238/Beryllium tamper in addition to the normal implosion technique, but the yield will be barely kiloton TNT or ever sub-kiloton TNT. However, this will be enough as the f ission trigger in the thermonuclear device (the so-called "Hydrogen-bomb").

Except for the fission-trigger, the thermonuclear device could be built without the additional U-235/Pu-239, just the Lithium6-Deuteride fusion core plus the U-238 pusher and U-238 jacket in the normal fission-fusion-fission H-bomb, or the fusion core wit h the Tungsten pusher and jacket in the fission-fusion Neutron-bomb. Given the amount of fissionable material (U-235/Pu-239) being produced, it is very likely that China is making 140-150 nuclear warheads a year and she has accumulated 2,350 nuclear warh eads so far.

However, this is just a tiny fraction of the 30,000+ nuclear warheads in U.S. and Russia. Even if the START-II Treaty is implemented by 2010, there will still be more than 20,000 nuclear warheads in U.S. or Russia. So China will never be able to acquire a comparable level of nuclear arsenal of US or Russia.

In discussing the U.S. military involvement in the Taiwan Strait recently, an American war hawk screamed "we should nuke the Chinamen/Chinks" and "we don't have to worry about China's nuclear weapons since they can only reach Hawaii and maybe California" . What a callous ignorant blood-thirsty racist (he's on the eastern coast)!

China does have ICBMs, SS18-class DF-5 with a range of 8000 miles, which can reach virtually every corner of every nuclear power (US, Russia, Great Britain and France). It has the same boosters (CZ-2C) as those used for Chinese spy satellites with a la unch success rate of 100% (no failure out of 18 launches). The failed satellite launch in February of 1996 used a different CZ-3B rocket with 4 strapped-on liquid boosters (the February launch was the first flight of CZ-3B).

China's DF-5 ICBM entered into the service in 1980 after extensive flight tests - at least 5 in 1979 alone. According to Jane's Strategic Weapons, its throw-weight is said to be 3200kg, which I think is much lower than its actual capacity. Jane's Strate gic Weapons Year Book had given wrong specifications of Chinese missiles until 1991, although China had published the complete specifications of CZ-2/3 rocket series in 1985, which is listed in Jane's Space-flight Directory. I think the editors of Jane's Strategic Weapons are good only at creating rumors about China's missile sales. They may have certain motives in doing so and I'll deliberate on this matter later.

According to Jane's Space-flight Directory, the 1st-generation Chinese spy satellites (FSW-1) weighed about 2,500 kg and launched by the same CZ-2C booster (as for the DF-5 ICBM) into the LEO. This information is confirmed by China's offer to launch the international satellite of 2,000 kg with the CZ-2C booster. The 500 kg difference is due to the fairing/shroud used to protect the satellite from the aerodynamic force during the launch. The FSW-1 spy satellite is recoverable and thus its body is harden ed for the atmosphere reentry, and so no extra fairing/shroud is needed during the launch. As a comparison, the French Ariane-4 has a longer and wider fairing/shroud weighing 900 kg and a 520 kg vehicle equipment bay (VEB) for the flight control and guid ance.

It is reasonable to assume that the CZ-2C's vehicle equipment bay weighs at least 500 kg. Together with the 4,000 kg dry weight of CZ-2C second stage rocket motors, at least 7,000 kg of satellite plus vehicle bay plus motor is accelerated to the 7,900 m/ sec burnout speed needed for the satellite launch into the LEO.

For an 8,000-mile-range ICBM, the burnout speed is about 7,200 m/sec. Since the specific impulse of the CZ-2C 2nd-stage rocket motor YF-22 is rated at 295 second (lower for the veneer motor YF-23), for a delta V of 700 m/sec (7900-7200), it is easy to de termine that the CZ-2C booster can accelerate at least 8,900 kg to the burnout speed of 7,200 m/sec, an 1,900 kg increase in the payload. Therefore the 8,000-mile DF-5 ICBM should have a throw-weight of at least 4,400 kg (2,500+1,900) for the warhead, in which you could put a 10-megaton thermonuclear device (such as the one on SS-19) easily (China's Lop Nor nuclear test site won't be able to handle such a monster, probably having to use just scale-up for the design purposes).

In the same Jane's Space-flight Directory, China offered to launch 4 satellites on one CZ-2C. This means that its twin DF-5 is also MIRV-capable (after all, MIRV is not hi-tech and it involves just weight-balancing and proper-timing of warhead-releasing. If Russia could develop it in 1960's, China should have no problem developing it in 1980's). In fact, the DF-5 was probably designed with the MIRV in mind since the CZ-2C's second-stage has 4 veneer motors for steering and they continue to burn for fur ther 190 seconds after the main motor shuts off - this could produce a LARGE footprint for MIRVs.

For the MIRV-version of DF-5, its throw-weight should be reduced to 3,900 kg as the fairing/shroud is needed to protect the MIRVs. So it's very likely that DF-5 is carrying six 600 kg-weight warheads as on the DF-21 IRBM (the twin of JL-1 SLBM). The war head size (about 1.9m x 0.9m) is right for fitting six nukes in the CZ-2C/DF-5 fairing/shroud.

I think Jane's Strategic Weapons is wrong again on the yield of DF-21 warhead (it says 250-kiloton). The JL-1/DF-21 entered into service in 1982 and all of China's nuclear tests before 1983 were either 10-50 kilotons (fission weapons) or 1-4 megatons (the rmonuclear devices). China conducted a nuclear test in the 200 kiloton range until 1990. It is very unlikely that China would put a 250-kiloton warhead on DF-21 without testing it back then (there was no political/technical restriction on testing a me dium-yield warhead).

I think the 600 kg warhead on DF-21 is in fact an one-megaton thermonuclear device given the shape and density of the warhead on the picture published. As a comparison, the W-47 RV on the U.S. Polaris A-1 SLBM (1961) weighs 408 kg and has a yield of 800- kilotons; the bomb itself weighs only 275 kg. So the 8,000-mile DF-5 ICBM is probably carrying six one-megaton MIRVs. Now the question is how many DF-5's does China have?

The number of DF-5's cited in various publications has been the same throughout the decade - that is only four DF-5 ICBMs in silo. Does anyone ever use the common sense that China has conducted AT LEAST seven flight tests of DF-5 in 1979-1980 alone, whi le deploying only four?

There is a photo on page 239 of the 1992 issue of Jane's Space-flight Directory, which shows the final assembly workshop for the CZ-2C/DF-5. In the photo, there are about nine CZ-2C/DF-5's in the various phases of final assembly and the text states that the production rate is about 10--12 per year. Does it really take 9 months to put ONE rocket through the final assembly line? After all, the liquid-fueled rocket is just some aluminum tanks connected with pipelines and combustion chambers!

It has been fifteen years since the DF-5 entered into services, during which China was under the threat of a Soviet invasion (other Chinese missiles could not reach Soviet Europe). During the period, there were no more than 30 satellites launched by Chin a. So there could be between 120 and 150 DF-5 ICBMs in the Chinese arsenal. With six one-megaton warheads on each of them, 700-900 out of 1,800 so-called strategic warheads are accounted for. Now the question is where are the DF-5 ICBMs?

I think the answer lies in the PLA's bible the Art of War written by the famous military strategist Sun Tzu some 2,300 years ago. One of Sun Tzu's lessons is - if you cannot win a battle, do NOT fight it.

China knows that the silos can be easily located by enemy satellites and that China's early-warning systems are inadequate; the launch-on-warning could be very dangerous (despite the popular belief in the newsgroups, all China's liquid-fueled SSMs use the same storable propellants as in SS-18). China recognizes that silo-hardening is just a losing battle and its nuclear submarine technology is generations behind that of US and Russia.

So China avoids the very vulnerable silo-based deployment of strategic ballistic missiles and instead hides her ICBMs in underground tunnels and caves dug in deep canyons and mountains. Those missiles are prepared inside the caves and moved outside to la unch. Command-and-Control will be much easier to maintain for this kind of deployment.

Early 1995, China's media reported that a Great Wall project for China's strategic missile force was completed after 10 years of construction in a "famous" mountain range in North China. Look at the topographic maps and read the news reports caref ully, it can be deduced that the underground tunnel network is in the famous Tai-Hai Mountain Range between Hebei and Shanxi provinces. According to the news reports, "tens of thousands" of Army engineers spent over 10 years there digging tunnels.

Normally, a company of soldiers (about 100 men) can dig about 100 meters of tunnel per month (based on the news reports about railroad tunnel construction) without using any advanced tunnel drilling machinery. So the "tens of thousands" of Army engineers (= hundreds of companies) over the 10-year period would have constructed an underground tunnel network of thousands of kilometers inside the Tai-Hei Mountain Range to hide some of China's strategic missiles. I guess it was called the "Great Wall" projec t not without a reason for the Great Wall is at least 5,000 kilometers long.

Like other known mountain ranges housing underground tunnel networks for China's strategic missiles, the Tai-Hei Mountain Range has many steep cliffs and canyons with large big elevation changes over a short distance between 1,000 and 2,000 meters. So yo u can easily dig tunnel networks with over one kilometer thick earth-cover in mountain ranges.

A typical 500-kiloton nuclear warhead in U.S. or Russian arsenals can `dig' a big hole 70m deep and 300m wide on the ground, and that is more than enough to destroy a missile silo or even an airport. If specially hardened for the earth-penetration purpos e, it may create a huge crater sphere 200 meters in diameter underground. Taken into account of the rupture zone around the crater and the likely penetration depth of warheads, at least three 500-kiloton warheads will have to land on the same spot sequen tially in order to penetrate the 1-kilometer thick earth cover and destroy the tunnel underneath. Even with the monster 20-megeton warhead on Russia's single-warhead SS-18, at least two warheads have to land on the same spot.

Moreover, one would destroy less than 300 meters of a tunnel using three warheads. Assuming the underground tunnel network under the Tai-Hei Mountain Range is only 1,000 kilometer long, one would need to use 10,000 (ten thousand) 500-kiloton warheads in order to make sure the tunnel network is completely destroyed. This is the VERY unlikely case in which you know the exact layout of the entire tunnel network. AND this is just one of several missile sites in China.

China has been digging underground tunnels in the mountains since the mid-1960s. There is no hi-tech needed to dig tunnels - just dynamite and concrete, plus enthusiastic young soldiers who are never in the short supply. China happens to have many huge mountain ranges everywhere. So there are other underground tunnel networks for the strategic missile force in the central and southern China's mountain ranges.

Even if you can eliminate all of them by using tens of thousand of nukes, the problem will be, with so many nuclear warheads detonated on the earth surface, so much earth will be thrown up into the upper atmosphere and spread around the globe in the str atosphere, the sun light will be blocked and we will have a real "nuclear winter", definitely not a pleasant picture even for our friends in America.

Sorry for using anon-service. But I don't want my boss to know I'm wasting his time. Thanks for your understanding.

MA Tuowen

The following information on the Chinese Military was compiled from various sources on the Internet (acknowledgment omitted). This information in no way represents the whole picture of the PLA. PLA guards its secrets quite well and it is extremely diff icult for other countries to conduct spying activities in China (because Chinese are harder to buy with money?)

Chinese ICBMs

Chinese Name West name Range Warhead Notes
DongFeng-1 ? 600 km 1500Kg HE ? No longer in service
DF-2 CSS-1 1250 20KtTNT No longer in service
DF-3 CSS-2 2650 3Mt Nuke
DF-3A CSS-2 2800 Nuke Multi heads
DF-4 CSS-3 4750 2Mt Nuke
DF-5 CSS-4 12000 5Mt Nuke
DF-5A CSS-4 13000 5Mt uke
DF-6 ? 19000 ? Multi heads
DF-7 ? 15000 ? Multi heads
DF-11(M11) CSS-7 300 500Kg HE
DF-15(M9) CSS-6 600 500Kg HE
DF-21 CSS-5 1800 500Kt Nuke
M-7 ? 180 500Kg HE For export only
M-18 ? 1000 ? HE for export only
JuLang-1 CSS-N-3 1700 500Kt Submarine launched
JL-2 CSS-NX-4 8000 ? (multi) Sub Launched

The strength of Chinese strategic force is unknown; the estimated number of warheads is 2,500, with 140-150 produced each year. Most of the ICBMs are hidden in tunnels in the mountains (only 4 DF-5 are in silos - the figured cited in America as being th e total of China's ICBMs!). The estimated length of the underground tunnels about 1,000 metesr deep is 2,000 km. Chinese missile production capacity is also unknown. The Western intelligence agencies estimate that China has only a dozen of DF-3s; note t hat China sold 120 DF-3s to Saudi Arabia, delivered so quickly that U.S. did not have time to react.

The M-9 was the one tested recently. The export version M-9's precision is 300 meters.

Firing sequence is simple:

  1. Check things are functioning
  2. Zero the inertia guidance system
  3. Input target coordinates
  4. Fire (the missile will calculate how to get there)

Chinese anti-ship missiles

Chinese name West name Range(KM) Warhead
ShangYou-1 CSS-N-2 85 400Kg HE
HaiYing-1(land) CSSC-2 85 400Kg HE
SH-1A/HY-2 CSSC-3 95 513Kg HE
HY-3(C-301) CSSC-X-6 130 513Kg HE
HY-4(C-201W) CSSC- 150 500Kg HE
FeiLong-1 CSS-N-1 45 513Kg HE
FL-2 CSS-NX-5 50 365Kg HE
FL-7 ? 30 365Kg HE
YingJian-1(C-801) CSS-N-4 40 165Kg HE
YJ-2(C-802) CSSC-X-8 120 165Kg HE
C-101 ship CSSC-X-5 45 300Kg SemiAP HE
plane CSSC-X-5 45 300Kg SemiAP HE
YJ-6(C-601/611) CAS-1 110/200 513Kg SemiAP HE

The Chinese Navy is only a defensive force. It consists of large number of high-speed boats equipped with anti-ship missiles.

Chinese Nuclear Submarines

  • Han class: 5500 tons, conventional weapons?
  • Xia class: SSBN (type 092, 8,000 tons), armed with 12 JL-1 (CSS-N-3) SLBMs (range 1,700km with a single 1.25MT nuclear warhead).

    Return to: Korea WebWeekly