Nuclear Issues: Pokhran Row

DISCUSSION

Since the days of the Shakti series of Pokhran-II underground nuclear tests, conducted jointly by the Department of Atomic Energy (DAE) and the Defence Research and Development Organisation (DRDO) more than 11 years ago, controversy over the yields of the devices tested, in particular that of the thermonuclear device, or hydrogen bomb (S1), has refused to go away.

The devices of May 11, 1998—S1, S2 and S3—were exploded simultaneously as the shafts S1 and S2 were just one kilometre apart and there was the danger of the shock wave from the first large explosion damaging the neighbouring shaft and the equipment therein. Similarly, the sub-kiloton devices of May 13, too, were exploded simultaneously, apparently for reasons of “convenience and speed”. The thermonuclear design yield was limited to 45 kt to avoid any damage to Khetolai village, located 5 km away, the DAE has stated. In a paper published in 2008 in the journal Atoms for Peace, R. Chidambaram, the former Chairman of the Atomic Energy Commission (AEC) and the leader of the POK2 tests, claimed that thermonuclear weapons of yields up to 200 kt could be confidently designed on the basis of the S1 test.

Soon after the tests, Western analysts, analysing the data of the tests as recorded by seismic stations worldwide, began to doubt the claims of Indian scientists for the combined yields of the May 11 devices and asserted that the actual yields were much lower. These estimates ranged from 10-15 kt to 20-25 kt. However, on the basis of correct interpretations of the regional and global seismic data and on-site measurements of ground accelerations and post-shot radiochemical analysis of the radioactive debris in the shafts, DAE scientists countered these estimates through a number of published papers on the results of the tests that confirms their early estimates (Table 1). While some well-known experts have concurred with the DAE’s claims, the controversy has sort of remained unresolved with many specialists continuing to question the DAE’s analyses and conclusions.

Table 1

*sub-kiloton test

From the DAE’s perspective, the claimed yield values are accurate and these agreed with the estimates of simulations and design values, thus rendering the Shakti campaign successful. DAE scientists also claimed that the tests were sufficient to build a credible minimum deterrent (CMD) and the data gathered in the tests were sufficient to carry out sub-critical tests, if required. In sub-critical tests, the fissile material is prevented from becoming critical and initiating an explosive chain reaction. Such tests will not be forbidden under a verifiable Comprehensive Test Ban Treaty (CTBT) if it should come into force. It was on this basis that the country declared a unilateral moratorium on testing. Continuation of this moratorium is a precondition to India’s civil nuclear cooperation agreement with the United States.

With K. Santhanam, a former DRDO official who was part of the core group associated with the tests, now stating publicly that the thermonuclear test was a “fizzle”, fresh fodder has been added to the controversy. He first made these remarks on August 24 at an in-house meeting of the Institute for Defence Studies and Analyses (IDSA) on the CTBT. He has since reiterated the statement to the media as well. “Based upon the seismic measurements and expert opinion from world over,” Santhanam has been quoted as saying, “it is clear that the yield in the thermonuclear device test was much lower than what was claimed. I think it is well documented and that is why I assert that India should not rush into signing the CTBT.”

Clearly if there is any credibility to this statement, the government’s premise on the claimed CMD posture and the vacation of nuclear threats, its unilateral moratorium on testing as well as its position on the CTBT would seem shaky. In the wake of President Barack Obama’s apparent reversal of the US stand on the CTBT, there could be increasing pressure on India to sign the Treaty. What is important is that Santhanam’s assertion seems to be based on “expert opinion from world over”. Strangely, he does not wish to rely on measurements—seismic as well as other—made by Indian agencies and the rebuttals by DAE scientists of the various external “expert opinions”. His claim would have been more convincing had he presented any scientific counter-evidence to the DAE’s claims or challenged its analyses with an independent set of measurements by the DRDO or by responding to the DAE’s claims in technical terms.

Domestic criticism of the thermonuclear test had come from none other than P.K. Iyengar, the former AEC Chairman, way back in August 2000. He wrote: “[T]he fusion core [probably] burnt only partially, perhaps less than 10 per cent.” This comment has been wrongly interpreted by various media commentators to mean that the thermonuclear weapon had fizzled. A thermonuclear weapon has a primary fission (or fusion-boosted fission) trigger and a secondary fusion containing the solid lithium deuteride (LiD). Neutrons from the fission are absorbed by Li in the LiD to yield tritium and helium. The tritium in turn combines with deuterium in situ and undergoes fusion, releasing large amounts of energy. Even in the most advanced thermonuclear weapons efficiency of the secondary fusion is around 50 per cent.

Arguing that the fusion to fission yield ratio in the Pokhran-II test must have been at best 1:1, Iyengar said that while he had no reason to dispute the yield (of 40+ kt) claimed by DAE scientists, he believed that the burn of the secondary fusion core was likely to have been highly inefficient. That is, the amount of LiD used must have been a great deal more than the optimum. He further argued in favour of more thermonuclear tests to improve the fusion efficiency as well as to increase the fusion to fission yield ratio. Iyengar reiterated the argument in a recent article (New Indian Express, September 2). He has further argued that the fusion yield cannot be derived unambiguously from radiochemical analysis as the methodology is complex and subject to large errors. In reality, however, the design ratio was 2:1, with the yield of the boosted fission trigger being 15 kt. According to Chidambaram, detailed radiochemical analysis too had validated this as well as the total yield (Graph 1).

Now, since both Santhanam and Iyengar were privy neither to the design of the weapon nor to the details of the radiochemical analysis and other measurements, their arguments are quite speculative. National Security Adviser M.K. Narayanan, in fact, said in a recent interview (The Hindu, August 30): “First and foremost, DRDO has nothing to do with [this aspect of the] tests… The measurements are not done by DRDO.” And, in any case, unlike Santhanam now and many Western analysts before him, Iyengar has not questioned the yield itself.

Therefore, the question of whether India should conduct more than one thermonuclear test to improve the efficiency of the weapon and to make its nuclear deterrent more credible, particularly in the context of its no first-use policy, and its relevance to India’s stand on the CTBT, is entirely distinct from the need to do more tests if S1 had been a fizzle. It may even be argued that the bogey of a thermonuclear fizzle is now being raised by those who would like India to conduct more tests and not sign the CTBT. Indeed, as Narayanan said, “I think we are going to face pressures from the international community…[It] is going to say that this is one of India’s very devious methods of preparing for a test, that [our] scientists are saying that was a fizzle, therefore India may find it necessary to prove itself once again. This is my worry. I hope it doesn’t happen.”

¨

Irrespective of the unwarranted fallout of the controversy, it is important to know the exact situation with regard to the yield of the
Pokhran-II tests even if the evidence is not enough to settle the issue. The only data pertaining to the tests that are globally available are the seismic signals. On the other hand, data from the other close-in measurements, namely, on-site accelerometer measurements of the ground acceleration, CORRTEX (Continuous Reflectometry for Radius vs Time Experiment) measurement of the two-way transit time (TWTT) of an electric pulse through a coaxial cable (which determines the strength of the advancing shock front from the explosion as a measure of the explosive yield), and the analysis of radioactivity in the explosion debris are available only to the agencies involved in the tests. In fact, the radiochemical data and the capacity to analyse them—considered the most accurate means to calculate the yield—exist with the DAE only. It is reliably learnt that though on May 11, 1998, the DRDO set up its own accelerometer to measure the ground acceleration, the instrument malfunctioned and did not record the associated waveform correctly. An independent internal check in this regard, outside the DAE, would have been possible if this had worked. Much of the controversy with respect to the test yields has, therefore, naturally arisen from the seismic data, which were the first to be recorded over the global seismic networks as signatures of an underground nuclear explosion.

An underground nuclear explosion sends up a shock wave near the points of detonation and a small portion of the total energy released is converted into elastic seismic waves. The efficiency with which the wave energy is transmitted through the medium depends on the nature of the surrounding medium, the source characteristics and the coupling of the medium on the geophysical properties of the rocks in the vicinity of the explosion site. These seismic waves travel through the body of the earth and also along its surface. The former are called body waves, which include both compressional P waves and shear S waves. P waves travel faster than S waves and also their frequency content is greater. At short distances (less than 2000 km) body waves travel through the crust and top portion of the upper mantle, and these waves are called regional seismic waves. Beyond 2000 km, body waves travel through the mantle and the core and are called teleseismic waves. Surface waves include two groups of waves, Rayleigh (R) waves and Love (L) waves. At regional distances, a group of higher mode Rayleigh and Love waves, called Lg waves, arrives at the detector before the fundamental L and R waves.

The energy of seismic sources —a measure of the yield in the case of explosions – is measured using a logarithmic magnitude scale. Three magnitude scales are used: body wave magnitude m(B), surface wave magnitude m(S) and Lg wave magnitude m(Lg). The yield Y of a nuclear explosion (in kt) is given by an empirical relation m=a +b log Y, where a and b are not universal constants but are site-specific. To arrive at the value of explosive yield, one needs to measure the magnitude and also use site-specific values of the constants a and b. For m(B) in particular, such well-established relations exist only for a few well-known testing sites of nuclear weapon states. While a varies significantly from site to site, b varies in a narrow ragne 0.75-0.85.

According to S.K. Sikka, one of the key DAE scientists involved in the Pokhran-II tests, a major reason for Western analysts giving a lower yield is the arbitary use of an a value of a known site, such as the Russian Shagan river site, for an unknown site such as Pokhran. Owing to the anisotropy and heterogeneities in the earth through which waves travel, m(B) can wary, and given a logarithmic m-Y relation, yield estimates would vary considerably even for small differences in m(B). In practice, assuming that errors in magnitudes arising from differences in propagation characteristics from the source to different seismic stations are random, the magnitude of an event is arrived at by averaging all globally measured m(B) values.

In the case of Pokhran-II, the computation of the average was further complicated because of the simultaneity of the tests, which causes P waves emanating from individual explosions to interfere constructively or destructively depending on the direction of detection with respect to the source geometry. Sikka and associates showed that owing to the interference of P waves from the two large signals S1 and S2, the values of m(B) along the line joining the two shafts (east-west) would be lower compared with m(B) values long north-south. For Pokhran-II, the average m(B) estimates of the networks of the International Data Centre (IDC), Arlington, US, and the US Geological Survey (USGS) are thus smaller, they argued. After making the necessary corrections, they showed that, as compared to m(B)=5.0 and 5.2 respectively for IDC and USGS, the correct average value was 5.4. This gave a combined yield value for the May II tests to be 58+5 kt (Graph 2).

¨

Soon after the Indian announcements of the test yields, Terry Wallace in Seismological Research Letters (SRL) and Brian Barker and associates in Science questioned the Indian yields. In fact, these two papers continue to be cited to challenge the Indian figures. But in their analysis, Sikka and colleagues had also rebutted their conclusions. First, Wallace and Barker used the average USGS and the IDC values of m(B) respectively to calculate the yields, which, according to Sikka, were inaccurate without including interference effects.

Moreover, both used the formula for the Shagan river site for Pokhran, which was inappropriate. DAE scientists pointed out that the Indian plate was different from the Eurasian plate, and in the absence of a site-specific m-Y relation for Pokhran, it was more appropriate to use the formula for the Nevada test site (NTS) (with a=4.05 instead of 4.45 and b=0.77) to calculate the Pokhran yields. Using an m(B) of 5.4, this gives 58 kt (Graph 2).

It must be pointed out that while the seismology community had not accepted the DAE’s argument of interference being significant, there has not been any convincing rebuttal based on detailed analysis either. Wallace’s rejection had been rebutted by DAE scientists who pointed out that his use of USGS stations only amounted to a biased selection as they lay within a narrow angle with respect to Pokhran and interference within them would be negligible. In a 2001 analysis in the journal Current Science, British weapon scientists A. Douglas and others concluded that the effect was small. But they too rejected a number of stations as, according to them, their m(B) measurements were corrupted by the arrival of coincidental earthquakes.

Since there is a great deal of site-specific uncertainty (in a) in the determination of the absolute yield from seismic data and b does not vary significantly in the m-Y relation, the relative yields between two tests for a given site can be evaluated with much greater confidence by using the difference in m(B) values and eliminating a. By measuring the ratio of amplitudes of P waves (see picture) at 13 seismic stations common to both Pokhran-I and II (Table 2), Sikka and others have calculated the average change in m(B) to be 0.45. This, in turn, corresponds to a ratio of 4.46 between the yields of Pokhran-I and II. A Pokhran-I yield value of 12-13 kt gives the Pokhran-II yield to be 54-58 kt.

Clearly, this method of estimating the Pokhran-II yield critically depends on the Pokhran-I yield. It may be recalled that there is controversy over its value as well. On the basis of an apparent statement made by Iyengar that the Pokhran-I yield was 8-10 kt, this is the value that has generally been used by Western analysts instead of the official figure of 12-13 kt. Some, in fact, believe that it was less than 5 kt. A figure of 2 kt has also been stated.

Clarifying this to this correspondent, Iyengar said that local acceleration measurements at Pokhran had given a value of 10 kt, whereas British weapon scientists had measured an m(B) corresponding to 8 kt. “Therefore, we were very happy that our device had worked with an yield in the ballpack we had estimated,” Iyengar said in an e-mail exchange.

Table 2: Comparision of mb values at common stations of 1974 and 1998 Indian explosions

According to Sikka, radiochemical analysis of Pokhran-I had been done and it gave a value of 12 kt. Based on post-shot data such as cavity radius, surface velocity and the extent of rock fracturing, an analysis in 1985 has yielded a value of 12-13 kt. This has been accepted by some Western analysts on the basis of international m(S) measurements (Graph 3). But despite this, people like Wallace continue to use a lower figure for Pokhran-I. Interestingly, however, Wallace himself was a co-author of a report of the IRIS Consortium to the US Senate in 1994 that gives a value of 10-15 kt, according to Sikka.

However, in a post-1998 analysis for the Federation of American Scientists (FAS), Carey Sublette, while gererally agreeing with the arguments of DAE scientists, has pointed out that given Pokhran’s sandstone and shale strata over a water table, the plot of yield versus crater morphology fits better with a Pokhran-I value at 8 kt rather than 13 kt. He then goes on to rely on this value to give a lower estimate of around 30 kt for Pokhran-II. In a comparative analysis similar to that of Sikka and Co., Douglas and associates arrive at 0.37 for the average m(B) difference. This corresponds to a yield ratio of 3.1. With Pokhran-I at 13 kt, this gives a
Pokhran-II yield of 40 kt. They prefer to use a value of 8 kt and arrive at a Pokhran-II value of 25 kt.

Given the uncertainties in dealing body wave magnitudes and the possibility of introducing bias in analysing m(B) values, renowned seismologist Jack Evernden prefers to use long-period surface waves. These show less scatter compared with short period P waves. Being waves of longer wave-length (60 kilometres), they are less influenced by the small-scale in homogeneities as well as interference effects. In fact, the relationship is almost independent of the site. Soon after the Pokhran-II tests, Evernden used USGS’ m(S) value and calculated the yield to be in agreement with the Indian claims. It may, however, be pointed out that for Pokhran-II very few stations reported m(S) values. Using an m(S) formula due to J.R. Murphy, the value of m(S) = 3.56 estimated by DAE scientists yields a value of 49 kt. Similarly, the use of a formula due to Evernden and G.E. Marsh yields a value of 52 kt, both of which are consistent with DAE figures.

Notwithstanding Iyengar’s reservations about the method, the most reliable estimate comes from the post-shot radiochemical analysis. It may be pointed out that the US has always relied on radiochemical analysis for estimating its nuclear test yields, rather than seismic data. In a 1999 analysis, DAE scientists claimed that the post-shot radioactivity from the S1 site had confirmed that the fusion secondary gave the designed yield.

This radioactivity, apart from unburnt fissile and tritium, consists of (a) fission products from the trigger and the fission component of the secondary (if present); and, (b) activation products due to the high-energy (14 MeV) neutrons produced by fusion, such as sodium-22 and manganese-54, which are produced much more in fusion than in fission. Graph 4 shows the gamma radiation peaks due to fission and neutron-activation products, which are much higher in the case of the thermonuclear sample than in the case of pure fission samples (Graph 1).

According to Chidambaram’s Atoms for Peace paper, “a study of this radioactivity and an estimate of the cavity radius, confirmed by drilling operations at positions away from ground zero, the total yield as well as the break-up of the fission and fusion yields could be calculated.” The yield estimate by this method was 50 + 10 kt.

But this too does not seem to satisfy Western analysts. According to Sublette, the radiochemical analysis refers to an entirely different method. He argued that the DAE method had inherent limitations arising from the error in measuring the cavity radius. Values lower than the claimed radius of 40 m would substantially bring down the yield value, he said. The upshot of the ongoing story is that notwithstanding the DAE’s detailed arguments and analyses, doubts continue to persist. But that should not prevent the DAE and the government from carrying out a totally objective internal evaluation of the success or otherwise of Pokhran-II.

Since the present AEC Chairman Anil Kakodkar, who was also part of the Pokhran-II team, has stated categorically that no more tests are needed, the current controversy, one hopes, will not drive the country’s polity towards more nuclear tests.

[Courtesy: Frontline (September 25, 2009)]