How the AP Got the Iran Inspections Story Wrong


A new controversy over a small facility on an Iranian military base at Parchin now threatens to blow up the support President Obama needs to get the Iranian nuclear agreement past Congress. Unfortunately, the controversy is the result of shoddy reporting and a poor understanding of nuclear inspections.

The Parchin site, which is southeast of Tehran, has been the subject of scrutiny for a number of years. Iran may have carried out experiments there that would help it design a nuclear weapon. In connection with the recently negotiated Iran deal — the Joint Comprehensive Plan of Action (JCPOA) — the International Atomic Energy Agency (IAEA) and Iran have agreed on a plan to investigate allegations of work that might be related to nuclear weapons design (the IAEA term for this is possible military dimensions, or PMD). Parchin is a military site, not a declared nuclear site under regular IAEA surveillance.

The controversy arose from an article by George Jahn of the Associated Press, which revealed information from what he said was a leaked, draft agreement between Iran and the IAEA on sampling at Parchin. The AP subsequently deleted and then restored parts of the first version of the story. The AP has also published what it says is a hand copy of the document the story was based on. Jahn painted a picture of a dubious, secret deal between the IAEA and Tehran that allows only Iran to conduct inspections of Parchin. Deal skeptics are pointing to this episode as evidence that neither the IAEA or Iran can be trusted and that, therefore, a nuclear deal is a bad move.

Questions have been raised, by me on Twitter and by others, whether that document is an accurate representation of the final agreement between the IAEA and Iran. The reason for my skepticism is that the document presented by AP is missing key elements that should appear in agreements of this type.

Having worked on nuclear issues for years, including environmental sampling at Los Alamos National Laboratory and elsewhere, I expect the agreement between the IAEA and Iran to include a plan for sampling at Parchin. I’ll explain what that might look like.  All quotes below from the AP article are from the version online at the time I wrote this article.

The two IAEA–Iran agreements

Jahn links his document to what others have called “secret side deals” between Iran and the IAEA. He ominously proclaims that the document

is labeled “separate arrangement II,” indicating there is another confidential agreement between Iran and the IAEA governing the agency’s probe of the nuclear weapons allegations.

On July 14, the IAEA released a joint statement with Iran that describes the two agreements. One is called the “Roadmap for the Clarification of Past & Present Outstanding Issues regarding Iran’s Nuclear Program.” The other is for sampling the Parchin site. The details of the agreements are confidential, as are all the IAEA’s agreements with member countries.

The sampling crew

The allegation getting most of the attention is that Iranian technicians will do their own sampling at Parchin, and IAEA inspectors will not be allowed on the site. The original article said that IAEA staff might not be allowed on the Parchin site. A later version stated:

The document suggests that instead of carrying out their own probe, IAEA staff will monitor Iranian personnel as they inspect the Parchin nuclear site.

The current version of the AP story, as I write, states:

Iran is to provide agency experts with photos and videos of locations the IAEA says are linked to the alleged weapons work, “taking into account military concerns.”

That wording suggests that — beyond being barred from physically visiting the site — the agency won’t get photo or video information from areas Iran says are off-limits because they have military significance.

While the document says the IAEA “will ensure the technical authenticity” of Iran’s inspection, it does not say how.

The fact that Jahn’s hand-copied IAEA document does not say how is a key indicator that it is unlikely to be a final draft. The middle paragraph represents only Jahn’s or his editors’ tendentious interpretation.

IAEA monitoring of an Iranian sampling crew should not present a problem. If the IAEA devises the sampling plan and is responsible for the chain of custody — and the agency has an extensive record of being able to do this reliably — it doesn’t matter whose hands take the samples. Jahn has surmised too much from a document that says too little.

The sampling plan

What is the purpose of sampling at Parchin?  The IAEA is looking for residues of experiments that are alleged to have been done to develop a design for a nuclear weapon. They were most likely carried out before 2003, when U.S. intelligence estimates find that Iranian nuclear weapons work ended, although the IAEA has expressed concern that experiments may have been carried out after 2003.

As I describe on my blog, Nuclear Diner, sampling plans start at the desk, when you figure out what you’re looking for. The Parchin document should either be a sampling plan or include one as an annex. Iran’s sensitivities about access to the site and the IAEA’s need to get meaningful samples require that the plan be both detailed and acceptable to both parties.

The sampling plan must be written down, along with plans for analysis of the samples, which may be in a separate document. These documents are essential support for the case that the IAEA will present on potential military dimensions of Iran’s nuclear program.

How would that sampling plan be developed?

It would start from the experiments believed to have been carried out, their location, and what residues might be found.

The nuclear experiments alleged to have been carried out in a containment vessel about 4.5 meters in diameter and 19 meters long at Parchin would have been tests of:

  • Neutron initiators, which start the chain reaction in nuclear weapons, including those containing uranium deuteride;
  • Multipoint detonation systems, which ensure the optimum implosion of nuclear material into a critical mass; and
  • And hydrodynamic properties of materials, which help designers to know how to fit those materials into nuclear weapons.

Over the past decade, Iran has modified the site and the building believed to have held the vessel, including actions that appear to be cleaning out the building. This activity has been observed through satellite photos.

For all three types of experiments, explosives residues may be discovered. A vessel would have been cleaned out between experiments, but swipe samples would show residue. If the experiments were done outdoors, the explosive may have degraded with time and weather.

Tests of multipoint detonation systems are unlikely to leave residues other than explosives. Fiber optics are sometimes used in such experiments, and fragments might remain.

The neutron initiator and hydrodynamic experiments may have scattered uranium. The isotopic composition of uranium used in these experiments probably would have been natural or depleted in U-235.

Other metals besides uranium, like tungsten, may have been used in the hydrodynamic experiments, and they may still be there. The IAEA may know whether the containment vessel still exists. Most likely, it has been cut up for scrap. Swipes from inside the containment vessel would be the most desirable kind of sample. The document published by the AP mentions only “environmental samples,” without specifying whether they are swipes, paint chips, or wall cores.  Chip samples of the paint or core samples of the wall may give more information than swipes.

If you’re interested in how these samples are taken, Tariq Rauf, a former IAEA inspector, recently described the swipe sample process. Soil samples from outside the building may also be taken, according to the document leaked by the AP. I have described how soil samples are taken at Nuclear Diner.

A sampling plan may include how the samples are to be analyzed. International standards for chemical analysis have been worked out and would be specified. The laboratories each side uses may also be specified. As Rauf points out, samples are split between Iran and the IAEA, with some saved for later analysis if there are differences. After the analyses are done, Iran and the IAEA will meet to share results and work out differences, which might require analysis of the reserved samples by an independent laboratory.

Chain of custody and access

A chain of custody, as is used for environmental and other legal sampling, is essential. The IAEA has standard procedures for transferring samples from one person or organization to another. All parties must sign each handoff.

It doesn’t matter whose hands do the sampling. What matters is the verifiability of the sampling, laid out by the sampling and analysis plans as well as the chain of custody.  The IAEA’s practice is to prepare sampling kits containing the necessary equipment, oversee the sampling, and be responsible for the chain of custody.

If the IAEA does not have direct access to the site or buildings, how can it oversee the process?

Verification experts, armed with cameras and ingenuity, have come up with ways to observe from a distance. On the day of the activity, they might start with a shopping trip that includes  IAEA and Iranian personnel. The group goes to a camera store and buys shrink-wrapped memory cards for the still and video cameras. Then they go to, say, a handicrafts store and buy a small and unique object, maybe a toy. The memory cards are inserted into the cameras in the presence of both groups, and the cameras are sealed with a tamper-indicating seal. Every photograph must include the unique toy.

IAEA personnel then go with Iranian personnel to the closest point to the site that Iran allows. IAEA personnel remain there while the Iranians go to the building or soil sampling site. Real-time video may be sent out to the IAEA monitors from the cameras, along with GPS data to make sure samples are being taken in the correct places. Iran completes the activity and hands the samples and cameras to the IAEA, who check the seals. The group may then convene in a meeting room to review the photos or video to make sure they are as expected, with the unique toy ever present.

What would the IAEA learn from Parchin sampling?

If, against expectations, the containment vessel itself could be sampled, swipe samples from inside it would give the most information. Explosives, fiber optics, and uranium in those samples would point to some or all of the suspected experiments.

If the vessel isn’t there, swipe samples would be taken of building walls and soil samples from outside the building.  In this case, finding explosives, fiber optics, or uranium in the samples would not necessarily point to the alleged experiments. After a decade or more at a military test site like Parchin, other activities could have contributed them.

It is difficult to associate explosive residues with particular types of explosives, unless an experiment went badly and pieces of explosive that didn’t explode can be found. The three types of experiments might use different types of explosives, so knowing what kind of explosives were used might give insight into which of the experiments were done. The IAEA will probably analyze for isotopes of uranium. In particular, uranium-235 is used in weapons, and finding it in higher than natural concentrations would indicate weapons-related experiments.

If no explosives, fiber optics fragments, or uranium are found, then there are two possibilities: No such experiments were ever done or Iran successfully cleaned up after experiments. There is no way to distinguish between the two.

Information from sampling will be fitted together with the documentary evidence the IAEA already has in hand, which includes information from their own investigations and material contributed by other countries. The full extent of that evidence is not public. With that evidence, IAEA might be able to draw fuller conclusions than I’ve indicated here.

The AP document

The document published by the AP and said to be a hand-written copy of material shown to George Jahn is not a sampling plan and does not refer to a sampling plan of the type I’ve outlined. I would expect the real plan agreed to by the IAEA and Iran either to be a sampling plan or to refer to one as an annex. As I’ve explained, verifiability, which the IAEA knows well, requires a written plan for all activities in detail. Those details must be agreed between the two parties in order to avoid disagreements during the actual sampling. This suggests that the document published by the AP could be a very early draft or extended outline of what the agreement might be.

It’s too bad that Jahn didn’t put the document he copied into the context of standard sampling activities by speaking with nuclear experts on how sampling is done. That might have avoided the erroneous implications that Iran would be doing its own inspections without IAEA oversight.

The larger context

The results from sampling at Parchin are part of the IAEA’s investigation into Iran’s past PMD work. The IAEA has been investigating these issues for several years, and Iran has supplied information, most recently last week. The IAEA will review this information, get back to Iran with questions to be cleared up, and issue a report on its findings by December 15 to the IAEA Board of Governors.

The report will most likely say that Iran has answered the IAEA’s questions from the annex to its November 2011 report on Iran’s program and that the IAEA finds that Iran is not now working on nuclear weapons development. That is not the end of the IAEA’s investigations, however. It will continue to work with Iran to develop a Broader Conclusion, which says that all nuclear materials within Iran are in the scope of peaceful activities. Broader Conclusions take years to develop. The JCPOA specifies that the IAEA’s submission of a report on a Broader Conclusion will take place eight years after the JCPOA comes into effect in October 2015. The Broader Conclusion could be finished sooner than that.

How much do we need to know of the Iranians’ nuclear weapons programs?

Advocates of a full disclosure of all of Iran’s weapons-related activities say that it would help us to know what Iran might do in the future. This is doubtful. It has been 12 years since the weapons program is thought to have ended. Diagnostic equipment, the way Iran runs its programs, and staff have changed. For many reasons, Iran would change the ways it does things if it reconstituted its nuclear weapons program. In this case, the past tells us little about the future. And those advocates might just be looking for another way to condemn Iran.

Other countries have given up nuclear weapons programs without full disclosure, and they have kept to their nonproliferation obligations. South Africa, Brazil, Argentina, Sweden, and Switzerland all had nuclear weapons programs at one time and have given them up. Intensive interactions of the P5+1 nations and the IAEA with Iran under the JCPOA will give more insight into Iran’s past. Most importantly, the JCPOA will prevent Iran from developing a nuclear weapons program now and in the next decade or two.


Cheryl Rofer worked at the Los Alamos National Laboratory for 35 years, where she directed programs in environmental remediation and plutonium storage. She has worked with Kazakhstan and Estonia on environmental issues and now blogs at Nuclear Diner.

Photo credit: IAEA Imagebank