IAEA: Most PMD Claims Groundless

 by Robert Kelley

As readers of this blog know, I have long been skeptical of key details of the International Atomic Energy Agency (IAEA) November 2011 report, in particular specific allegations contained in the report’s Annex, which was entitled “Possible Military Dimension of Iran’s Nuclear Programme.” In addition to my articles published for the Stockholm International Peace Research Institute and for IHS Janes, LobeLog has published five posts on the subject over the past year.

I had hoped that the IAEA’s long-awaited final report on the PMD, which was leaked to the press on December 2, would clear up in the most definitive manner possible specific allegations made in the 2011 Annex. These included, most notably, the existence of a “large explosives containment vessel in which to conduct hydrodynamic experiments” at the sprawling Parchin military factory south of Tehran (paragraph 49) and the assertion that “large scale high explosive experiments were conducted by Iran in the region of Marivan” (paragraph 43), which is located in northwestern Iran close to the Iraqi border. Both allegations were based on reports to the IAEA by “member states” that were never identified by the Agency but are widely believed to have been long been hostile toward Iran, including Israel. In my opinion, and as I have tried to make clear in my analyses for LobeLog and elsewhere, the 2011 report, including those specific allegations, was deeply flawed.

Although the final report is good news in that it will finally close the so-called “PMD file” related to Iran’s past nuclear activities (and permit the JCPOA to go forward), it is also very disappointing in its virtually total failure to disclose any new information about those alleged activities, including at Marivan and Parchin. Also disappointing is the fact that it has failed to report how the IAEA evaluated the reliability of the more questionable information provided by those anonymous “member states.” And, as will be shown below, its technical analysis regarding issues—like detonators used in nuclear weapons that exceed the Agency’s mandate of monitoring nuclear materials to ensure against their diversion for military use—leaves much to be desired.

In fact, the final report appears designed as much to deflect criticism of the 2011 report as to provide any new information substantiating that report’s allegations. In that respect, it should be seen as more of a political document than a technical one. Moreover, it has made new allegations—specifically, that Iran had conducted “some activities” related to designing a nuclear weapon as late as 2009—without providing persuasive evidence or identifying a single identifiable source behind that assertion. In my opinion, repeating the old charges without serious additional substantiation and airing new ones that were certain to make headlines and provide ammunition to those forces that are staunchly opposed to the JCPOA, again without offering specific evidence or sourcing to back them up, is irresponsible.

The Mystery of Marivan

In the 2011 Annex, the IAEA offered a detailed description of a very complex experiment to test a Multi-Point Initiation System (MPI) and asserted that this system was tested once “in the vicinity of Marivan.” The final report amends this statement by asserting that the experiment was performed at “a location called Marivan” (paragraph 41), without any clarification as to the significance, if any, of this “new” version.

Nonetheless, the original allegation is a very serious one, because, if true, it would establish that Iran was indeed performing tests that were directly relevant to a nuclear weapons program. Last year, Iran invited the IAEA to visit Marivan, but the Agency rejected the invitation, insisting that other sites, including Parchin, warranted precedence. The decision not to take advantage of the offer at the time was baffling in itself given the importance of the alleged experiments carried out there, but the fact that the final report omitted any mention of the Iranian offer constitutes a material failure to inform readers of relevant facts. (In a similar omission, the final report failed to mention Vyacheslav Danilenko’s testimony that he had designed the chamber described in his book for the production of nanodiamonds, not for hydrodynamic experiments.) Explaining the failure to visit Marivan would also be instructive since it indicates that the Agency itself may have had doubts about the credibility of the source information about the alleged tests. If so, that would bear directly on other information provided by the same source(s). It may be possible that an IAEA team did, in fact, visit “a location called Marivan” while trying to wrap up its PMD investigation. If so, however, one would reasonably expect that the report would mention such a visit and its findings.

The failure to credit Iran with any cooperation it has offered or provided to the IAEA is not the only instance in a report that repeatedly depicts Tehran as unresponsive and/or obstructionist. Paragraph 33, for example, states that “the Agency identified a 15-page document” during a 2005 meeting with Iranian officials that related to the conversion of uranium compounds to uranium metal. Iran gave the document to the IAEA voluntarily but the wording implies that IAEA somehow discovered it against Tehran’s wishes. I have no doubt that Iran could and should have been more forthcoming in its dealings with the Agency than it has been. But there is no reason for the IAEA to cast Iran’s behavior in such a seemingly gratuitously negative light.

Parchin Puzzle Left Unresolved

I summarized some of the mysteries and anomalies surrounding the alleged containment vessel in several posts over the past year, beginning with “The Parchin Puzzle” and, more recently, “Putting Parchin to Rest,” which was written shortly after IAEA Director General Yukiya Amano himself visited the suspect building at Parchin on September 20.

At the time, Amano said that he and his chief deputy found no equipment in the building alleged to have housed the explosive containment chamber. The final report contains an elaboration on their observations:

[T]they did not observe a chamber or any associated equipment inside the building. They did observe, inter alia, recent signs of internal refurbishment, a floor with an “unusual cross-section” and a ventilation system which appeared incomplete.

Of course, it is not clear what is meant by “an unusual cross-section,” but the bigger issue is what was not observed. At least until now, the IAEA’s theory regarding this chamber hinged on six sentences in a book by a Russian/Ukrainian scientist, Vyacheslav Danilenko, who wrote (and apparently later told the IAEA in an interview) that he indeed designed a large chamber for nanodiamond compression experiments in 1999 and 2000. He did not indicate that the chamber was actually built or that it was built in Iran. According to the description in his book, the chamber would be as large as a London double-decker bus and many times heavier. It would also be partially encased by a 700-ton block of steel-reinforced concrete that would be much more massive than the chamber itself.


This is an artist’s conception of the chamber and the concrete block and its relative size as described in Danilenko’s book

According to its final report (paragraph 48), the IAEA had obtained information (published in its 2011 report), including satellite imagery, that “indicated that Iran made and installed a large cylinder at the Parchin military complex in 2000” and then constructed a building around it. “[O]ther information,” it said, indicated that the object corresponded to the parameters described by Danilenko in his book. The building, it said, was in use until late 2003. The final report further noted (in paragraph 56) that, in just the last two months, the Agency had “acquired new satellite imagery from different sources, including a commercial source, which supported previous indications of the presence of a large cylindrical object at the location of interest to the Agency at the Parchin site in the summer of 2000.” (Emphasis added)

This account begs a number of questions, not the least of which is the notion that Iran could break ground on this project, complete its construction, and then erect a building around it all within the space of a few months. The report is notably silent about whether the IAEA had attempted to visit the factory that allegedly manufactured the chamber. Nor did it apparently conduct any analysis of how much time, money, and engineering would have been required to design, manufacture, and install it. A related question, which I address here, is how such an enormous structure, including the massive cement block, could be so comprehensively dismantled and disposed of—an operation that would have required scores of workers and convoys of dump trucks over many days—between 2003 and September 20, 2015, without being detected by commercial or spy satellites. Of course, the easiest way to accomplish that would have been to blow up the building and its contents. But the building, albeit partially refurbished, has remained in place since 2003.

Aside from its manufacture, installation, demolition, and removal, however, a number of other obvious questions remain:

  1. What explains the remarkable lapse in the Agency’s own independent efforts to obtain commercially available photographs of the “large cylindrical object” between its 2011 report and late 2015?
  2. If the new imagery is indeed available commercially, what prevents the IAEA from publishing the photographs (or at least the date of the image and the vendor), particularly since no photographic or physical evidence of the alleged chamber has ever reached the public domain? Publication of the evidence would at least permit independent verification of its assertions.
  3. Why, if armed with this additional evidence, didn’t the IAEA in its final report explicitly confirm its previous claim that the cylindrical object matched the dimensions of Danilenko’s design, particularly since it went to some lengths to affirm that assertion in its 2011 report (paragraph 49)?
  4. Although the IAEA says in its 2011 report that the chamber matched Danilenko’s          design, that leaves the question of whether it observed the 700-ton concrete block that was part of Danilenko’s design?
  5. Is there an actual image of the cylinder itself? In the 2011 report, the reference was to “indications” that a building was being “constructed around a large cylindrical object at a location at the Parchin military complex” that was consistent with anonymous information. The new report states that the new imagery “supported previous indications of the presence of a large cylindrical object at the location of the Agency at the Parchin site…” The wording in both reports, particularly the use of the “indications,” is intentionally ambiguous and begs for clarification.

Explosives and Detonation

In addressing issues that go beyond its specific mandate and area of competence, the report suggests a lack of technical expertise about nuclear-weapons development. Certainly the final report authors did not heed available competent technical advice.

In the 2011 Annex (paragraphs 38-40), the IAEA asserted that Iran’s development and acquisition of exploding bridge wire (EBW) detonators was “a matter of concern” because they were an indicator of nuclear weapons development, particularly in combination with the MPI system allegedly tested at Marivan (or “a location called Marivan”). “The Agency recognizes that there exist non-nuclear applications, albeit few, for detonators like EBWs, and of equipment suitable for firing multiple detonators with a high level of simultaneity,” the Annex stated. (Emphasis added.)

The final report, while in no way repudiating its earlier assertions, retreated somewhat with respect to the possible application of EBWs to non-nuclear activities, allowing that

…[e]xplosive [sic] bridgewire (EBW) detonators developed by Iran have characteristics relevant to a nuclear explosive device. The Agency acknowledges that there is a growing use of EBW detonators for civilian and conventional military purposes.” [paragraph 79]

Aside from the incorrect terminology used in the report, the Agency appears surprisingly clueless that the market for EBWs has expanded steadily since the 1940s! Commercial firms have produced millions of EBWs, and none of them, to the best of our knowledge, has been used in nuclear weapons. Along with exploding foil detonators and slapper detonators, EBWs constitute the modern standard for safe, reliable military detonation systems, as well as for commercial mining and drilling operations.


Commercial offering of exploding bridgewire detonators from a US company with 30 years of experience

Consistency of analysis is critical to credibility. In the final report, the IAEA asserted in paragraph 36 that Iran

…in 2002-2003, developed exploding bridgewire (EBW) detonators and a high voltage firing capability which, in combination, enabled several detonators to be fired with less than microsecond simultaneity. [Emphasis added]

Leaked documents from the IAEA had previously pointed out that Iraq had developed MPI systems that required only two detonators, while the notorious multi-point initiation (MPI) experiment at Marivan would, if the IAEA’s description is accurate, require only one detonator. Thus, the conclusion stated in paragraph 36 appears inconsistent with the MPI design that Iran allegedly developed.

Paragraph 66 of the final report also refers back to an assertion in the 2011 Annex that Iran “had conducted a number of practical tests to see whether its EBW detonator firing component would function satisfactorily over a long distance between the firing point and a test device located down a deep shaft,” such as would be used for an underground nuclear test. To any electrical engineer, particularly one who worked with munitions, such a claim is preposterous on its face. A high-voltage firing set is not located hundreds or thousands of meters away from the device it is designed to explode. It is located as close as possible. The high-voltage firing set is triggered by a firing signal that travels over long distances from a manned firing point as part of a detonation sequence plan. But the actual sharp, short high-voltage pulses that fire the detonators will travel to the device over special low-loss cables, a distance of just a few centimeters. The theory underlying the IAEA’s thesis appears to be based on or reflected in questionable speculations published in The New York Times earlier this year. That article included a diagram with a caption entitled “Bomb-Firing Test” that stated:

The underground test of an atom bomb requires long cables that carry high voltage to fire the sphere of detonators. Iran is alleged to have studied long-distance firing.[Emphasis in the original]

It appears that the IAEA may have relied on a newspaper account for this rather bizarre concept.

In section E.9 (paragraphs 63-65)the IAEA describes a routine visit to a university in 2015 to examine a neutron generator under the same heading as “shock compression neutron initiators” for bombs. The suggestion that the university work is for a neutron initiator is misleading and could be used to promote the false conclusion that Iran was pursuing weapons work after 2009. In fact, the two are unrelated, and conflating them shows either a lack of technical expertise or a deliberate effort to suggest that Iran was engaged in nuclear weapons-related work as recently as this year.

Sampling at Parchin

In paragraph 55, the final report refers to sampling performed at the military factory complex. Unfortunately, it does not describe how the samples were taken despite the public controversy surrounding it. Critics of the JCPOA have alleged that IAEA staff were not present when the samples were taken, although this has never been verified. Nonetheless, according to the report account, the samples yielded two results worth noting.

First, the report disclosed that two particles of man-modified uranium were found in “environmental samples.” Unfortunately, as any competent scientist can attest, this is meaningless in the absence of background baseline samples taken in the region to determine how many such particles are found on average in any given sample. The apparent failure to collect baseline samples and conduct statistical analysis reflects poorly on the IAEA. The Agency did visit a different site at Parchin on two occasions in 2005. They reported finding no nuclear materials in samples taken there. Possibly those samples could be used as a baseline for the 2015 visit.

Second, the IAEA claims to have analyzed the environmental samples for explosive and chemical residue. Over the past two decades, the Agency has built up a highly sophisticated capability for analyzing environmental samples, but this capability has been focused—entirely correctly in light of its mission—on detecting radioactive traces. If the IAEA can analyze samples for purposes unrelated to radioactive fissile materials such as commercial industrial secrets or chemical weapons programs, it would appear to be operating far outside its specific mandate. This should be of concern to all states that permit the IAEA to take “environmental samples.”

Iran and the IAEA made confidential arrangements regarding how the sampling would be carried out, and these have not been publicly disclosed. It can only be hoped that Iran agreed to a procedure that permitted the Agency to analyze the samples for any traces of fissile materials and/or chemicals.

Other Technical Problems in the Report

In Paragraph 86, the final report states that safeguards activities were conducted at various locations in Iran, including Parchin. But safeguards activities by definition could not have been carried out at Parchin because that complex is not a declared nuclear site, and no nuclear-related activities have been carried out there. Although this may appear to be a minor technical point or a drafting error, it reflects poorly on the care with which the report was prepared.

In 2011, moreover, the IAEA reported a discrepancy of 19.8 kg of natural uranium waste and scrap at a site called Jabr Ibn Hayan Multipurpose Research Laboratory, suggesting that Iran may have diverted some of it for other purposes. (The Agency has a standard for determining when a discrepancy is significant. For natural uranium, the trigger for concern is 10,000 kilograms, so this discrepancy is not only trivial but unrelated to PMD.) What the Agency failed to report at the time, however, was that the material in question had been under IAEA seal for many years. In fact, no material was missing. The discrepancy was due entirely to a change in the way the Agency itself measured heterogeneous waste in containment drums. The issue was subsequently cited in a few reports to the IAEA board but then eventually faded away. It is thus very odd that the final report resurrected the issue (paragraph 31) but changed the description of the amount of the alleged discrepancy to “several” kilograms. Moreover, instead of forthrightly explaining its own responsibility for the discrepancy, the report stated that it had simply determined that the “amount of natural uranium involved was within the uncertainties associated with nuclear material accountancy and related measurements.”

Finally, in the 2011 Annex, the IAEA claimed that:

Information which the Agency has been provided by Member states, some of which the Agency has been able to examine directly, indicates that Iran has manufactured simulated nuclear explosive components using high density materials such as tungsten.” (paragraph 48)

But, in the final report, the specific reference to tungsten is dropped (paragraph 47). Why? As has been publicly pointed out, tungsten is a poor simulant for uranium in such experiments. But instead of forthrightly acknowledging its mistake, the Agency appears to be trying to finesse the issue. If the Agency can make the case that tungsten should have been and/or was indeed used, it should do so and not hide the facts it asserted previously. The omission of the specific reference to tungsten clearly undermines the credibility of the original allegation and hence its source(s). That the Agency feels obliged to ignore its own responsibility in the “missing” uranium and its mistaken assumptions about the use of tungsten is not only self-serving. It also tends to undermine its own credibility.

In a similar vein (para 59), the Agency accuses Iran of planning equation-of-state measurements of “materials of concern” in the critical period: 2009. (Equation of state is simply a mathematical model of how materials behave under extreme pressure and temperature.) Although the report reassures us that Iran has not performed experiments with uranium, it fails to identify the “materials of concern” to which it was referring, a curious omission especially if they were not nuclear materials covered by Safeguards. Transparency here would improve confidence in the Agency’s assessment that Iran was still doing nuclear weapons work between 2005 and 2009.

Quality of Information

Underlying the entire final report is the sometimes explicit, sometimes implicit assumption that Iran lied whenever it denied a specific allegation or that it was stonewalling whenever it insisted that it had no knowledge of or could not provide an answer to a question. Unfortunately, nothing in the report indicates that, conversely, the IAEA tried to evaluate the reliability of the information it was given by anonymous “member states,” some of which are known to be hostile to Iran. The introductory paragraphs, for example, contain no mention of unconfirmed intelligence reports from unnamed sources. Nor is there any mention of the numerous forged reports provided to the Agency in 1995, 2002, and 2003. (For more on this, click here and search for the report on the implementation on the plan for Ongoing Monitoring and Verification in Iraq under paragraph 12 of UNSCR 687. The document was issued as S/1995/604.) In these cases, the IAEA engaged in extensive forensic analysis of forged documents and built up impressive experience in detecting forgeries. But there is no indication in the final report that the Agency used any of this experience to evaluate documents that were provided to it in the case of Iran.

Indeed, much of the information on which the PMD issue has been based has provoked skepticism among many experts, not least Amano’s immediate two predecessors, Mohamed ElBaradei (in his memoir) and Hans Blix. It would seem that, for the sake of its own credibility, the IAEA should at least have made some attempt to address the very difficult problem it faces when it publishes information or “disinformation,” as Blix has referred to it, provided by member states hostile to Iran and then grants the sources a cloak of anonymity. This apparent failure may help explain, at least to some extent, Iran’s reluctance to provide the IAEA with any more information than absolutely necessary.

Indeed, most of the IAEA’s accusations, especially in the 2011 report, are based on formulations like “Information provided to the Agency by a Member State…” Items like the tungsten simulant or the still-mysterious explosion chamber at Parchin fuel skepticism. The Agency must validate the accuracy of information provided to it by sources whose motives may not be altogether altruistic. Yet there is no description of how the Agency verifies such information. In every case where Iran claims the information is invalid or not correct, the Agency implies that Iran is lying or covering up. Some of Iran’s answers as transmitted by the Agency in its reports do indeed lend themselves to skepticism or suspicion. But many of the IAEA’s own assertions are also incorrect and stretch credulity in the absence of greater transparency.

For that very reason, we can conclude after four years that the concerns raised in the 2011 report have not been fully resolved. That certainly isn’t the IAEA’s responsibility alone. Despite the weaknesses of the final report’s analysis it is reasonable to close the PMD file, at least as it is based on the 2011 Weapons Annex.

An exercise such as this should end with lessons learned and contingency planning that will avoid the mistakes of the past four years. This is not a defense of Iran or a conclusion that there was no PMD. Rather, the IAEA could not justify its accusations in the four intervening years and has left a trail of mistakes. The main lesson in this case is that the IAEA Secretariat incautiously extended its own mandate far beyond its capabilities and institutional competence. The member states, through the IAEA Board of Governors, should review this episode and resolve that this very valuable agency should stick to do what is does best: using modern analytical and scientific tools to verify nuclear materials declarations in member states around the world to reduce the possibility that such materials will be diverted to non-peaceful purposes without timely detection.

Robert Kelley

Robert Kelley is a licensed professional nuclear and mechanical engineer in California. He spent his career in nuclear weapons development activities such as plutonium metallurgy, survivability of U.S. nuclear warheads in ABM intercepts and isotope separation by gas centrifuge and lasers. He later used these hands-on skills to lead intelligence analysis of foreign nuclear weapons systems for the U.S. government and then as a director at the International Atomic Energy Agency, particularly in weapons related non-proliferation analysis in Iraq, South Africa and Libya. Along the way, he has been a research reactor supervisor, a plutonium facility manager and director of the DOE Remote Sensing Laboratory at Nellis Air Force Base. He currently writes on non-proliferation for a number of publications and is an associate research fellow at the Stockholm International Peace Research Institute.



  1. Aw, Robert Kelley, give poor Yukiya Amano a break. It’s hard to shill for Washington and Israel when he’s so hard pressed to find anything he can convincingly hang on Iran.

    Besides, the era of elBaradei and Blix is long past (unfortunately).

  2. Barbara Slavin of the NY Times has declared that the IAEA report proves there was a nuclear weapons program, and that;s all that counts. Or so she thinks.

Comments are closed.