Monday, April 17, 2006

Quantity (stockpile size) versus quality (testing)

A suggestion has recently been made that the Chinese nuclear arsenal is only about 100 weapons as opposed to 400 weapons as claimed by the Non-Proliferation community. While some of us will find this sudden revision of the number of Chinese weapons by the Non-Proliferation community curious, perhaps even lacking in credibility, there is little doubt that this revised number goes some distance in explaining why the Non-Proliferation theologians have been ranting about Chinese fissile material cutoffs. This completely explains the confidence with which Bob Einhorn spoke about a linkage between India imposing a fissile material cutoff and China retaining its moratorium on fissile material production. I think the term "fissile material" here refers only to the weapons usable stockpile, it does not refer to any thing that is set aside for use in a reactor.

Let's assume for the moment that I accept the working distinction between weapons material and reactor material even if the substances are identical in all respects and we make such a distinction in the case of India's stockpile. Now per the logic touted by the Non-Proliferation community, India should also agree to a "fissile material cutoff" as China has, which means that India has to keep its stocks of "weapons use material" fixed, and keep its stock of "reactor use material" separate. The Non-Proliferation community would like the word "separate" to mean "under intrusive surveillance".

So where does this leave things in India?

Well firstly, the "intrusive surveillance" is out. The IAEA agreement will take note of India's sterling record on proliferation. There is a limit to the amount of nonsense India will put up with.

Secondly the stockpile issue has a catch in it and this is something that the people in the US need to grasp. I outline the same below.

If Indian nuclear weapons are to retain the same potency as the Chinese arsenal, then they must demonstrate the same level of effectiveness. If India is to target a Chinese city with nuclear weapons and guarentee its destruction, then India will have to deliver 4 weapons per city. If the Indian weapons have a 50% delivery probability, then to eliminate one Chinese city, India will need 8 weapons. If the Indian weapons have a 50% survival probability against a Chinese First-Strike, then India will need 16 weapons to ensure that it can guarentee a hit on a Chinese city. It may be recalled that China's weapons are better tested and pack a bigger punch. The Chinese have tested a full thermonuclear device with yeild in the megatons. Chinese missiles have a higher survival probability and a higher accuracy. So it will only take 1 Chinese weapon to eliminate an entire Indian city. It follows from this that in order to keep up with China's 100 nuclear weapons, India has to have about 16 times the number of weapons.... if India is to freeze the level of weapons development and not to test anything that might improve the reliability of its delivery platforms.

Alternatively if India improves the accuracy and the survivability by a factor of 10 each (that is; 95% accuracy and 95% survivability) then India could possibly get away with having just 4-5 weapons per Chinese city. This is still 4 times as many weapons as China, so India would need 400 weapons to keep up with China's 100.

Improving the accuracy and survivability of the Indian deterrent in my opinion signifies that the India's development of a submarine based ballistic missile deterrent will have to be expedited. It may be noted that India's present level of development with ground based missiles in inadequate to deter the Chinese as the nearest Chinese target lies beyond the Tibetan plateau - roughly a 1000 miles away. This is at the outer tip of the Agni's tested range. Additionally very little is known about the Agni missile's performance at longer ranges and under battlefield conditions. Placing missiles in the Indian North-East is a risky proposition given the relative lack of depth there and targetting the Chengdu and Guizhou provinces, China's backwaters is not a viable deterrent strategy.

Alternatively if India could carry out the tests required to make a 1 Megaton warhead. Then it would be able to exactly match the effectiveness of the Chinese arsenal and keep the stockpile limited to 100 weapons.

What I am saying is that the Non-Proliferation community would like you to think that "Fissile Material Cut-offs" and "Testing Bans" go together and India can be asked to do both things - this is simply not true.

India can do either one - it can't do both!

It is incumbent upon the non-proliferation community to determine which best suits its needs. Either option is fine with India so long as India has the means to protect its security in a bad neighbourhood.


At 8:44 AM, Blogger cynical nerd said...


Very good analysis on the deterrence vis-a-vis PRC. I agree with you on submarine based deterrence.

Note this article by KS. He says that India does not have enough fissile material for a CMD (God knows how much is that) and needs to boost of fissile material production before the eventual signing of FMCT! He then blames the NDA for letting things drift apart - strange because he was part of the National Security Advisory Board then as well!

The definition of CMD itself various with some talking 100 (K Subrahmanyam) to some like G. Parthasarathy wanting to have as much as a middle level power like France (400)

In addition, SSN/SSBN propulsion reactors apparently need HEU to the tune of 95-98%. Have we taken into account all these before talking about FMCT?

On CTBT, do you think the "mythical" Sagarika when it comes to mating oops fruition can be nuke tipped without fireworks testing.


At 8:49 AM, Blogger cynical nerd said...

Bad link in the previous comment. Here is the correct link.


At 1:18 PM, Blogger maverick said...

Cynical Nerd,

Today for all practical purposes we are not in a nuclear standoff with China. None of China's weapons are targetted at us and we do not have the means to deliver nuclear weapons on to Chinese soil by airplane, missile or submarine. Some Chinese short-range ballistic missiles are supposedly stored in Qinghai province but without an actual deployment into Tibet there is no question of these weapons having the range to reach us. Chinese ICBMs and fissile material stocks are mostly aimed at Russia and the US.

The exact numbers that China has earmarked for use against are irrelevant. If the Chinese move their ballistic missiles into the Tibetan plateau, they will place the entire Gangetic plain in range of nuclear weapons. This has the potential to cause upto 200 megadeaths in India. The only thing that is relevant is that we need to have a credible ability to cause 200 megadeaths in China at any moment of our choosing. This will put us on par with China in the calculus of nuclear warfare.

Today if India does indeed decide to confront China on the nuclear front, then as K Subrahmanyam says, we don't have the means to do it. Until adequate stocks of Pu are available for immediate assembly into weapons, all talk of confronting China is without merit.

The NDA was famous for its aggressive tone on security affairs but the NDA was actually quite balanced in the way it did things. The only thing that was out of sync with its otherwise balanced approach was its rhetoric. There were a number of appointments made to the NSAB by the NDA in its first year, quite a few of these were of the noisemaker variety. This was not well recieved in South Block.

I suspect that after the 1999 tests, the biggest priority in the NDA was to ensure the survival of the government and do whatever it takes to avoid a US conventional strike against India's nuclear establishment. The thing about dealing with President Clinton was that you could lie to him, because you knew he was lying to you but then foreign policy was made in a very ad hoc way. It was Jaswant Singh's reassuring and even servile gestures at the White House that warded of the possibility of a US air-strike on India.

Currently the emphasis in Government is to keep the fissile material for military purposes on an accessible level. This means agreeing to segregate the stocks into military and civilian piles and developing indigenous resources to meet the demand on the non-civilian side.

The submarine reactor is speculated to contain a thorium core. While I cannot confirm or deny that the reactor does use thorium, I will point out that the idea that India will use an enriched Uranium reactor for its nuclear submarine was proposed by the US Non-Proliferation community - not by us in India.

I think the Russians at one point of time proposed using their ICBM launching submarines as satellite launchers.

At 9:42 AM, Blogger cynical nerd said...


Thanks much for the explanation on fissile material requirements vis-a-vis China.

btw, 200 megadeaths = 200 x 1 Megaton bomb?

US conventional strikes agaist India? Is this mentioned somewhere? Either in Jaswnth or Talbott's book. Or is it an Atlanticist plan led by Madelein Albright?

I thought the reactor designs will be similar to the Charlie-class SSN which India leased. Thorium reactors for ATV is news to me. Ofcourse, you won't confirm it ;-) Has this been attempted before by others?

Recently an American co. tried to use a Russian SLBM to launch a satellite powered by a large solar fins - the mission turned out to be dud. But it costs a few million dollars - so the US co. will probably try again with the large stocks of Ru ICBMs lying around.


At 3:12 PM, Blogger maverick said...

Cynical Nerd,

A megadeath is one million dead. It is a measure of the killing capability of the weapon. In terms of physical damage, to ensure complete physical obilteration of a square mile you need about 10 kT of yield.

The population in urban centers usually climbs to about 250,000 per square mile. If you obliterate about 4 square miles you need 100 kT.

A factor of 10 in the yield should give you another factor of 2.5 in the number of immediate dead. So with a 1 Megaton bomb you can kill about 2.5 million people immediately.

In the zone outside ground zero, you will get casualties due to super-fires. These will kill an estimated 50 percent of the population in the area they affect. This process will take a few days. In a city this area will be roughly 3 times the size of the ground zero area. So you are looking at roughly 3-4 megadeaths within a few days.

Over a longer timescale, say a week, you see deaths due to acute exposure to radiation.This should account for another 1-2 Megadeaths.

Additionally the breakdown of law and order, abortion of foetuses, malnutrition of children and the old, should claim 1-2 Megadeaths within the month.

So I think it is reasonable that a 1 Megaton bomb can cause 10 Megadeaths, i.e. 10 Million dead over the course of a month.

The US airstrikes against India after the first round of tests were a significant possibility. If nothing else they could have stopped the second round of tests and killed scores of strategically significant staff. This has always been a factor in the calculations of when to test.

I cannot offer more comments about the nuclear submarine project.

I cannot discuss the results of the Russian launch experiments in detail, but I will point out that a satellite requires far greater stability in its housing than a nuclear warhead.

At 3:17 PM, Blogger maverick said...

Cynical Nerd,

In the previous calculation, I must point out that this calculation is missing several factors of order unity. It may take 3 Megatons to ensure 10 million dead in a month.

So 1 Megaton is order of magnitude estimate, not an exact number.

I hope you understand what I am saying.

At 12:36 AM, Blogger cynical nerd said...


Thanks for the much detailed response. Much appreciate. Ofcourse, I understand on things you cannot talk about.


At 11:51 PM, Blogger s c r a p s s t u f f s said...

Do you think we need different overkill factors with respect to China and Pakistan? China would require a higher overkill factor when compared to Pakistan.

My random calculation required 320 warheads for Pakistan and China together (overkill of 4 for a Pakistani city and 8 for a Chinese one, with top 20 cities in Pakistan and top 30 cities in China targeted for severe damage) with more required for other comers. I think most Indian estimates for numbers are too low (Kama Sutra has the lowest numbers - understandable as he's completely sold out to the US, G. Parathasarathy's numbers are a good starting point).

On another note I don't agree with the opinion that the most survivable Indian second strike vehicle is a SSBN. With the US having, what?, 30+ LA class boats it would be very easy for them to hunt down a handful of Indian SSBNs simply using sheer numbers. Unless we have SSBNs in the double digits a better option would be road mobile, low footprint, launchers that can launch from anywhere inside India to any target we want to reach.

On yet another note, I cannot understand the fascination with megaton yield weapons, multiple yields in the hundreds of kilotons should suffice and give greater confidence in being able to obliterate a target particularly when the delivery vehicles are improved. What is the basis for your assertion that Chinese missiles are more accurate than the Agni-II series?

If the Indian nuc boat uses thorium based reactor, then can it be assumed that we are more advanced in the thorium cycle than the progress on the AWHR would suggest (still in the design stage)?

At 8:53 AM, Blogger BangaloreGuy said...

The Russians did launch satellites from ICBMs that were to be decommissioned - twice already(about 6 months back). I dont think any SLBMs came into play.


Your calculation doesnt take into account where the nuke explodes - on the ground, in the air(at what height) - all that has a significant bearing on the damage.

I wish our folks skipped the Megatons and went straight to MIRVs - which will act as both a force multiplier and serve better for 2nd strikes. Also, for a population centre, it makes better sense to use a MIRV than one single warhead.

How do you arrive/guesstimate at the population density figure btw? for Mumbai, for eg., avg density is 75,000/ml2 (30,000/km2) or thereabouts.

Our cities top Chinese ones by a factor of 2 or so in population densities, so if we want parity, we need to factor that in as well(again MIRVs rule).

Also almost all our top cities are spread to the South-West, much like the Chinese ones.

On the Agnis - tested range is 2100km - not enough to hit Beijing, but probably can reach some parts of the South, South-Western Chinese regions.

At 8:57 AM, Blogger BangaloreGuy said...

scraps stuffs,

Unless our Agnis approach ICBM range - 6000km+, having road-mobile/rail-mobile Agnis makes little sense (Vs China) cos they will definitely have to be close to our North-Eastern borders!

At 10:32 PM, Blogger s c r a p s s t u f f s said...

bangaloreguy, Agni-III should be able to do the needful. Also, tested range may not always be equal to actual range. Tested range depends on payload and trajectory of launch, both can be changed to improve actual range.

All I'm saying is that unless you have a large number of SSBNs, I don't think they are survivable against an opposition (who has the means) determined to get them. Road/rail mobile missiles with sufficient range have a better chance of disappearing in the large Indian land mass and being undetectable until launch.

All this is IMHO, of course.

At 11:03 AM, Blogger maverick said...

Hi scrapstuff,

It will take less yeild to wipe out a Pakistani population center than it will to take out a Chinese population center due to differences in built-up area. Concrete does not convert to debris or burn easily.

The submarine base ballistic missile deterrent will only guarentee second strike against China as long as China is not able to improve its ability to patrol the Bay of Bengal. I guess this puts the events in Burma in perspective.

The Chinese have tested their missiles more extensively than we have and they have bigger engines and higher payloads. The best way to compensate for poor accuracy is to use a higher yield- provided you are aiming for a specific target. If you simply want to kill lots of people in a built up area then you have to use an MT yeild.

We have been talking about the Thorium cycle for a while, perhaps there is more to us than just talk? who knows? - it is all speculation at this point.

At 1:03 PM, Blogger maverick said...

Hi Bangaloreguy,

I used the population density of Shanghai approximately 150,000 per sq mile at its densest. During peak office hours this goes up by a factor of about 2. As I said earlier there are factors of order unity that this number could be off by.

I agree that the damage per kT of yeild depends on height of detonation (per US published studies), I have assumed that the height of the device's detonation is optimized to achieve complete (>90%) destruction of 25 square miles.

At 5:09 AM, Blogger BangaloreGuy said...


I hope A-3 can do it. From indications it doesnt seem to approach the said ranges - more likely 4.5km.

The tested Agnis were for 2100km, and trajectory calculated for what could be 3000km(or so I read).

Agnis can carry 1-ton warhead only. So if you reduce warhead size, you reduce impact, and hence destruction.

If I may say so, India isnt priority target no 1 for the USA -so all its SSNs, diesels wont be put out to patrol our sub - the USA has enough worries otherwise to be tracking only Indian subs.

The large Indian Mass is towards the peninsular/plains region. To beijing, from Bihar/NE you need a minimal range of at least 4000km to hit it - considering Agnis limited payload, and lack of being MIRVed.

Which is why we need MIRVs and a range which reaches the minimal ranges of ICBMs

At 5:16 AM, Blogger BangaloreGuy said...


Of course a nuke will cause destruction - but its important as to what sort of destruction we are aiming at, right?

A high nuke burst will reduce fall-out, a ground level one will have the highest chance of fall-out travelling to various places - so one would expect a ground-level targeting of Chinese cities and a High-alt targeting of Paki ones.

I was wondering which you assumed - ground/hig alt or medium alt - which I guess you mean to be ground level, from your last comment?

Do the DF-21s have more accuracy? afaik they're 1-ton/1MT payloads, like the Agni - only the DF-31s were 3MT capable.

At 1:19 PM, Blogger maverick said...

Hello bangaloreguy,

I assumed that the detonation height will be optimised to achieve 2.5 Megadeaths in an area with peak population density of about 250,000 per square mile. Using the wrong height of detonation could change the number of immediate dead by a factor of 2-3. This is not a detailed calculation, I am merely interested in projecting the numbers of weapons we would need to kill this many people. Its all a thought exercise.

The second wave of deaths would not be cause by fallout but by super fires spreading through built up areas. This is similar to the effect observed in the fire bombings of Tokyo and Dresden.

The accuracy of the DF11 are not very relevant, if they can hit any spot in the Indo-Gangetic plain they project far higher casualties than possible with a single explosion.


Post a Comment

<< Home