The Samson Option: Israel’s Nuclear Arsenal and American Foreign Policy

The Tunnel

Israelis have done their best work from below.

The huge underground laboratories at Dimona had their precedent in the Jewish struggle after World War II against the British mandatory power in Palestine. The British authorities had angered David Ben-Gurion and his followers by insisting that they adhere to the strict limits on Jewish immigration to Palestine that were set in 1939, after three years of Arab revolts. The British ruling had meant then that hundreds of thousands of Eastern European Jews were unable to escape the Holocaust. And now those who had somehow managed to survive were again being denied a chance to come legally to Palestine. Many were faced with a desperate dilemma: either return to what was left of their prewar homes and prewar life or remain in the dispirited and overcrowded displaced persons (DP) camps scattered across Europe.

The heavily outnumbered and outgunned members of the Hagannah, the Jewish underground, began the inevitable guerrilla war against the British troops with little other than their guile and determination. One of the war’s most imaginative operations involved what seemed to be yet another farming kibbutz that was set up in 1946 about fifteen miles outside Tel Aviv, adjacent to a large British military base. The kibbutz’s administrative building was constructed, seemingly at random, within a half mile of the base.

The underground facility had been “scooped out,” said Feinberg, in twenty-seven days. The men and women who worked underground alternated that work with farming; those who completed a shift in the arms factory were ordered to muddy their shoes and sit under sunlamps so they could appear to the British and others as if they had been innocently tending crops or looking after the kibbutz’s cows and sheep. Over the next two years, British soldiers and officers were constant—and unsuspecting—customers of the kibbutz’s bakery and laundry, which cheerfully offered their services to the military. Feinberg recalled that a few of the British soldiers even made a point of coming to the kibbutz’s Friday-night Shabbat dinners. Today the underground bullet factory is known as the Ayalon Museum, a popular attraction for Israeli schoolchildren.

Vanunu began working as a technican at Dimona in August 1977 and spent much of the next eight years assigned to various tasks inside the reprocessing plant, formally known as Machon 2 (machon means “facility” or “institute” in Hebrew) and informally known as the Tunnel. The reprocessing plant, which was handling materials that were exceedingly “hot”—that is, highly radioactive—was the most sensitive area at Dimona; only 150 of Dimona’s 2,700 employes worked there. A special pass was needed to enter the plant, and all movement inside, even to and from the bathroom, was—in theory—to be closely monitored. Vanunu, once at work in the Tunnel, found that the stringent security existed in theory only. Constantly in trouble for his public pro-Arab views, he had been laid off in mid-1985 as part of a government-wide cutback. Vanunu appealed through his union, powerful as are all unions in Israel, and won back his job. It was at that point that he smuggled a camera into the reprocessing plant during an overnight shift and wandered around undetected for some forty minutes, taking fifty-seven color photographs. A few weeks later he was fired after calling for the formation of a Palestinian state during an Arab rally. Even then, again with help from his union, Vanunu was able to negotiate a settlement from Dimona’s management that gave him severance pay and a letter attesting to his good record.

Vanunu’s Times interview and his photographs of many of the production units in the Tunnel, or Machon 2, provided the American intelligence community with the first extensive evidence of Israeli capability to manufacture fusion, or thermonuclear, weapons. American intelligence also obtained a copy of many of the Sunday Times’s interview notes with Vanunu; those notes, some of which were also made available to the author, provided much more specific detail of the inner workings of Dimona than was published. Senior American officials, including men and women who have worked in nuclear weapons production and nuclear intelligence, uniformly agreed that the unpublished Vanunu notes are highly credible. One intelligence official who has been analyzing Israel’s nuclear capability since the late 1960s depicted Vanunu’s information, which includes a breakdown of the specific function of each unit inside the Tunnel, as stunning: “The scope of this is much more extensive than we thought. This is an enormous operation.”

Of extreme interest to the United States were Vanunu’s photographs of what apparently were full-sized models of Israeli nuclear weapons.^† Copies of those photos were provided to weapons designers at the Los Alamos and Livermore laboratories for evaluation and analysis, and the designers, working from the photographs, constructed replicas of the Israeli weapons, as had been done with Soviet weapons in the past. They concluded that Israel was capable of manufacturing one of the most sophisticated weapons in the nuclear arsenal—a low-yield neutron bomb. Such weapons, which first came into the American stockpile in the mid-1970s, utilize enhanced radiation and minimal blast to kill anything living within a limited range with limited damage to property. The weapon actually is a two-stage thermonuclear device that utilizes tritium and deuterium (both by-products of hydrogen), and not lithium deuteride, to maximize the release of neutrons.

The Vanunu information also helped American intelligence experts date the progress of the Israeli nuclear arsenal. Vanunu revealed, for example, that Unit 92 in the Tunnel had been painstakingly removing tritium from heavy water since the 1960s, indicating that physicists at Dimona—following Levi Eshkol’s 1965 plea for advanced research—had been attempting from the earliest days of Dimona’s production to manufacture “boosted” fission weapons. The United States began experimenting in the early 1950s with boosting, which dramatically increases the yield, or destructiveness, of a single-stage fission device. Boosting is a process in which small quantities (a few grams) of tritium and deuterium are inserted directly into a plutonium warhead and designed to flood the warhead with additional neutrons at the moment of fission—in essence, jump-starting the weapon at the moment of critical mass—producing a bigger kick, or yield, with smaller amounts of plutonium. Vanunu also told the Sunday Times of returning from a vacation in 1980—his first trip abroad since emigrating with his family to Israel in 1963—and being assigned then to work at a new production plant for lithium 6, another essential element of the hydrogen bomb. In 1984, he further reported, a new facility (Unit 93) for large-scale production of tritium was opened. Lithium is irradiated in the reactor, then moved to Unit 93, where it is heated to release tritium in a gas form, along with helium and hydrogen. The gases are then driven under high pressure through an asbestos palladium column and separated. The helium is stored in powdered uranium and can again be released by heating. The opening of Unit 93 suggests that full-scale production of neutron weapons began then, for up to twenty grams of tritium are used in each neutron warhead.

As described by Vanunu (and confirmed by the author in later interviews with Israeli officials), Dimona includes the reactor and at least eight other buildings, or Machons, the most important of which is the chemical reprocessing plant. Each building apparently is self-contained. Machon 1 is the large silver-domed reactor, sixty feet in diameter, that can be clearly seen from the nearby highway. The uranium fuel rods remain for three months in the reactor, which is cooled and moderated by heavy water. The heavy water is itself cooled by ordinary water flowing through a heat exchanger, creating steam, which in a nuclear power plant would drive a turbine and create electricity. Instead, the steam in Machon 1 is vented into the atmosphere, creating a radioactive cloud ^‡ Machon 2 is the chemical reprocessing plant. Machon 3 converts lithium 6 into a solid for insertion into a nuclear warhead and also processes natural uranium for the reactor. Machon 4 contains a waste treatment plant for the radioactive residue from the chemical reprocessing plant in Machon 2. Machon 5 coats the uranium rods (shipped from Machon 3) with aluminum to be consumed in the reactor. The rods, once stacked in the core of the reactor, provide the fuel needed to sustain a chain reaction—and capture weapons-grade isotopes of plutonium. Machon 6 provides basic services and power for Dimona. Machon 8 contains a laboratory for testing samples and experimenting on new manufacturing processes; it also is the site of Special Unit 840, where Israeli scientists have developed a gas centrifuge method of enriching uranium for weapons use. There also is a laser-isotope-reprocessing facility for the enrichment of uranium in Machon 9. Depleted uranium—that is, uranium with very little or no uranium 235 left—is chemically isolated in Machon 10 for eventual shipment to the Israeli Defense Force or sale to arms manufacturers in Europe and elsewhere for use in bullets, armor plating, and artillery and bomb shells. The shells, buttressed by the heavy uranium, which is much denser than lead, can easily penetrate thick armor plating and are a staple in modern arsenals.^§ (There was no Machon 7 in the years he worked at Dimona, Vanunu told the Sunday Times, and he did not know what, if anything, had taken place there.)

Dimona’s most essential facility, of course, is the reprocessing plant in Machon 2, where Vanunu spent most of his career. It is here that plutonium, a by-product of the fission process in the reactor, is extracted by chemical means from the spent uranium rods. The residual uranium is then reprocessed and reconstituted for use in new fuel rods. There are at least thirty-nine separate units in the six underground levels of the Tunnel, the most important of which is the production hall where the spent uranium rods undergo reprocessing. Before reprocessing can begin, however, the rods must be cooled for weeks in water-filled tanks, reducing radioactivity by a factor of several thousand. Even then, the radioactive rods are still lethal and are always handled by remote control and from behind lead shielding. The Tunnel’s production hall dominates levels one through four below ground; work there is monitored by a large control room that includes an observation area known to technicians as “Golda’s Balcony,” a reference to Golda Meir’s frequent visits after she became prime minister in 1969. The end result of the chemical processing, according to Vanunu, is a weekly average of nine “buttons” of pure plutonium whose combined weight is 1.2 kilograms.

Once completed, the weapons parts are moved by convoys of unmarked cars, under armed guard, to another facility to the north—not known to Vanunu—for assembly into warheads. Israeli officials subsequently told me that the final stage of warhead production takes place at a defense plant north of Haifa operated by Rafael, the top-secret Israeli research and manufacturing agency that is responsible for Israel’s most sensitive weaponry.

The Tunnel remained in operation around the clock for thirty-four weeks a year, according to Vanunu, and was shut down from July to November for routine maintenance and repair. American nuclear experts consulted about Vanunu’s story describe the methods used to reprocess the spent uranium at Dimona as essentially routine; the industrial solvents and solutions used by the Israelis are the same as those relied upon at the Savannah River Plant in Aiken, South Carolina, where state-of-the-art heavy-water-production reactors have operated since the mid-1950s.

What especially impressed American experts about Dimona’s reprocessing plant was its location—underground—and its sophistication. “You have to understand,” an American explained, “Machon 2 is very sophisticated because it’s so hot. There’s an extraordinary level of radioactivity. You need three-foot lead walls, all automated; people in suits; robotics. You’re going to have a hell of a time keeping it undetected. So you go very deep.” That, in turn, drives up the price of ventilation shafts, air intakes, and fan systems, as well as all ordinary construction costs.

Going underground also posed enormous engineering risks that could be met only by superb master planning and expert intelligence. For example, the construction teams that initially built the U.S. Atomic Energy Commission’s Savannah River Plant in South Carolina decided to put the thick, lead-shielded doors that protected the work force on custom-made coasters with specially engineered automated motors for opening and closing. “We didn’t move the doors often enough,” the American added, “and the coasters flattened. The doors were too heavy and we had misjudged the physics. We had to stop the process to remove them. We didn’t test this beforehand because we didn’t think of it.”

Another area of great sensitivity involved the science of robotics, whose most important early use in the United States came in the hot weapons laboratories where humans could not work. The precision involved in machining the thin plutonium hemispheres and placing them around the gases needed to create boosted nuclear weapons was achieved only after enormous strides had been made in the use of remote control. It was not an accident that Aharon Katzir (formerly Katchalsky), who became, like Ernst Bergmann, an intellectual force inside the Israeli Atomic Energy Commission, was world-renowned for his research into robotics at the Weizmann Institute. Katzir was even featured with some of his research apparatus on the cover of the December 3, 1966, issue of the Saturday Review; the article was entitled “Man’s First Robot with Muscles.” It reported on Katzir’s pioneering work in converting chemical energy into the energy of motion. Katzir’s team at the Weizmann Institute also was concentrating on the development of artificial muscle tissue for use in robots. His research was heavily funded by the U.S. Air Force’s Office of Scientific Research; the Air Force’s primary interest was in utilizing robotics in outer-space research. The Air Force had no idea that it was also helping to underwrite research for the Israeli nuclear arsenal; nor did it know that Katzir’s main work was being done at Dimona, and not at the Weizmann Institute.

Despite such comments, there remains no actual evidence that Israel needed outside help for its nuclear weaponry. Dr. George A. Cowan, who spent more than twenty years designing nuclear weapons at Los Alamos, acknowledged that there always was a close association with Israeli physicists from the Weizmann Institute. “They’ve visited the labs [Los Alamos and Livermore] and probably are treated more openly than other visitors here, but there’s too much emphasis to the notion that there’s a secret that somebody has to tell them,” Cowan said. “The Israelis are smart enough to do their own research. The need for secret information is largely promoted by spy novelists. There’s very much less to it than most people believe.” Like many of the scientists in the American nuclear laboratories, Cowan has a close Israeli friend who was involved with Dimona: “He never asked me anything over the years about the bomb and wouldn’t have.” Similarly, physicist Hans Bethe, the Nobel laureate who helped design the first American nuclear and thermonuclear weapons, recalled three visits to the Weizmann Institute during which his hosts would “take me anywhere and discuss anything with me. They knew I was interested in nuclear power reactors,” he added, “and yet they never offered to take me to Dimona. I found that significant.”

The American experts got that one right, at least. Shimon Peres has confided to friends that during the early construction of Dimona he often signed requisition orders and other technical documents on behalf of the Ben-Gurion government without knowing precisely what he had approved.

* Vanunu described the cooling unit to Frank Barnaby, a nuclear physicist and former employee of Britain’s nuclear weapons installation at Aldermaston. Barnaby spent two days with Vanunu, at the request of the Sunday Times, in a continuing effort to verify his account. He concluded, the Sunday Times said, that Vanunu’s account “is totally convincing.” After leaving government service, Barnaby became director of the Stockholm International Peace Research Institute (SIPRI), an arms control study group funded by the Swedish government.

† Mock-ups are commonly used for training purposes and military briefings in the American nuclear weapons complex, for the obvious reason that no one would want to work next to a fully operational nuclear warhead filled with highly enriched materials. The mock-ups are accurate replicas, in terms of external design and size, of a normal warhead, and American nonproliferation experts assumed that the Israelis’ models were carefully prepared.

‡ Vanunu said that the steam, contaminated to varying degrees by leaks and corrosion, was vented only on those days when the prevailing wind was blowing toward the Jordanian border, about twenty-five miles to the east. It was one of those ventings, apparently, that was photographed by Army Colonel Carmelo Alba in 1965, providing the CIA with the first concrete evidence that Dimona was operational.

§ The American forces who fought in Desert Storm, the 1991 war against Iraq, were equipped with uranium-tipped bullets and antitank munitions. Some of the American tanks also were equipped with uranium armor plating for added defense against Iraqi attacks.

‖ The nuclear fuel cycle is so precise that scientists can compute how much uranium was consumed by Dimona at a given reactor output. According to Vanunu, the average flow rate of dissolved uranium and plutonium through the chemical reprocessing plant was 20.9 liters per hour with a uranium concentration of 450 grams per liter and a plutonium concentration of 170 to 180 milligrams per liter (or 0.39 milligrams of plutonium per gram of uranium). Vanunu said, however, that the actual flow rate in the Tunnel normally exceeded the standard flow rate by 150 to 175 percent, which corresponds to the reprocessing of as much as thirty-seven kilograms of weapons-grade plutonium a year, assuming eight months of continuous operation. Nuclear technicians have further noted that Vanunu claimed that the spent uranium fuel processed at Dimona contained a smaller concentration of plutonium—about 0.30 milligrams per gram instead of 0.39 milligrams per gram—suggesting that as much as 125 metric tons of uranium was needed to operate the reactor, far more than officially estimated. It is impossible to even roughly determine the amount of plutonium that has been produced at Dimona without knowing the power output and operating history of the reactor. That information remains a closely guarded state secret in Israel. The general accuracy and scientific validity of Vanunu’s statistical data added to his credibility with American intelligence officials.