So, I’ve been keeping my powder dry on this story by Sanger and Mazzetti claiming that Israel struck “a partly constructed nuclear reactor, apparently modeled on one North Korea has used to create its stockpile of nuclear weapons fuel …”

I’ve been reading carefully, going back over the historical data, trying to make sense of a story that, frankly, confuses me. Today, the Times runs a correction:

An article on Monday about the refusal of Israeli officials to confirm or deny a report in The Times that the Israeli Air Force had bombed a Syrian facility on Sept. 6 overstated the conclusion Israeli and American intelligence analysts had drawn about the target. While they judged the facility to be a partly constructed nuclear reactor, they said it was of apparent North Korean design; they did not say so definitively.

[A colleague points out that the correction is for a story by Steven Erlanger that asserts Israel “bombed a partly constructed nuclear reactor apparently of North Korean design.”]

Yeah, I was waiting for that — I spent the better part of a couple of days looking into the question of when (and how) the US determined the design of the Yongbyon reactor. As far as I can tell, in the early phases of construction, the ability to correctly identify reactor types is less than a science.

I am willing to accept that the Israelis might believe they knocked out a “Yongbyon next-door” but I am not willing accept their judgment based on the evidence that has been presented in public.

When Would We Know What Kind?

To illustrate my point, here is a timeline showing US intelligence assessments of the North Korean 30 MWth reactor at Yongbyon (sources at the end of the post):

  • 1980: US spy satellites identify reactor components co-located with a large hole dug for the reactor’s foundation.
  • April 1982: US spy satellites spot a reactor vessel at Yongbyon. In July, the CIA incorrectly describes the new reactor was a “copy” of the Soviet-supplied 2-megawatt thermal research reactor “not designed to produce quantities of plutonium needed for a nuclear weapons program.”
  • March 1984: US spy satellites spot a cooling tower under construction, suggesting a much larger reactor. In April, the CIA would report that the reactor “probably will be graphite-moderated and use natural uranium for fuel.”
  • 1986-1987: The 30 MWth Yongbyon reactor goes critical.

If the Syrian “reactor” was years from completion, then the situation is comparable, I think to the 1980-1984 period when the IC was mistaken about the reactor design.

It fairly boggles the mind that one could confidently determine the Syrian facility was a “copy” of the reactor at Yongbyon when, at a comparable stage of construction, we failed to correctly identify the Yongbyon reactor itself.

Of course, we may have learned a trick or two since then — one might imagine that we now look for tell-tale signs of Yongbyon-like reactors (such as the thickness of the concrete “biological shield” that should provide clues to the neutron flux or the layout of the facility).

But the debate over the intelligence, however, suggests otherwise.

Yongbyon Was A Copy, Too

Yongbyon is a sensational reference point from a news perspective, but the North Korean reactor is itself derived from the Calder Hall magnox reactors developed by Britain. Comparing the Syrian reactor to Yongbyon, without mentioning Calder Hall, leaves the reader thinking the reactor design could not have been supplied by any another state and was certainly not indigenous.

But Syria might have chosen the Calder Hall design for the same reasons as North Korea — the reactor is a relatively simple design that is extensively described in the open literature and does not require difficult to acquire heavy water or enriched uranium. It is perfect, in many ways, for the DIY bomb-builder.

A 1991 news article documents all of these important aspects that Sanger and Mazetti omit — the design heritage of Yongbyon, the information that might be in the public domain and the value of not using heavy water:

In the late 1960’s, having already acquired a miniature reactor from the Soviet Union, the North decided on a a design for a 1950’s-era British reactor called “Calder Hall,” still operated by the British Nuclear Fuels Company, according to published accounts in the nuclear trade press and one intelligence official.

There are disagreements about whether enough details of the reactor’s design were available in public documents to give North Korean scientists the outlines of what they needed, or whether classified details leaked. But the reactor uses graphite to modify, or control, the nuclear reaction — and graphite is something the Koreans could produce at home. That would eliminate the need to import heavy water, or deuterium oxide, a highly controlled substance.

The author was, of course, David Sanger.

Sources on Yongbyon: Oberdorfer, The Two Koreas: A Contemporary History, 2002, 250; North Korea and Nuclear Weapons: The Declassified U.S. Record, National Security Archive Electronic Briefing Book No. 87, edited by Robert A. Wampler, April 25, 2003; Wit, Poneman and Gallucci, Going Critical: The First North Korean Nuclear Crisis, 2004, 1; Richelson, Spying on the Bomb: American Nuclear Intelligence from Nazi Germany to Iran and North Korea, 2006, 246-247.