Iran's past covert enrichment activities—especially the revelation of Natanz in 2002 and Fordow in 2009—have raised concerns about whether other covert enrichment facilities exist in Iran today. In August 2010, then (and current) Iranian nuclear chief Ali Akbar Salehi claimed that the construction of an additional enrichment facility had started. However, operating a covert enrichment program is not an easy task. Several important decisions must be made which could have severe consequences for the program's survivability. For an external actor, there are three main tasks that would need to be undertaken in order to destroy a nuclear facility: detecting the facility, characterizing the facility and neutralizing it. Each of these tasks can be obstructed by different means.

Detecting and characterizing a nuclear facility can be done in several ways: with satellite surveillance; utilizing technologies that can detect radiation, heat concentrations, sound or vibrations associated with operating nuclear facilities; or by human intelligence (agents or informants). One method that is used by inspectors from the International Atomic Energy Agency (IAEA) is environmental sampling, that is, collecting samples of soil, air, water, vegetation or smears at the site of inspection and measure the isotopic composition between the uranium isotope U-235 (fissile material) and the isotope U-238 (non-fissionable material). In its natural form, the composition would be about 0.7:99.3. If enrichment has taken place the composition would have a relatively higher concentration of U-235. In general, the less enrichment that goes on (both in scale and time), the lesser the chances of detection. Therefore, the size of the facility is crucial in a covert enrichment operation, and the smaller the better. In order to miniaturize the covert facility, one should aim for installing as few centrifuges as possible. In order to reduce the number of enriching centrifuges to a minimum, one should seek to increase the centrifuge quality, that is, improve their effectiveness when it comes to separating the isotope U-235 from U-238. The Iranian centrifuge type called IR-2m is considered to be about three to five times more effective than the IR-1. Iran could limit the number of centrifuges necessary—and thus the size of the facility—by only installing the IR-2m in a covert facility.

The IAEA can only collect environmental samples at facilities declared by the country in question or at undeclared sites when the country is a signatory to the Additional Protocol of the Nonproliferation Treaty. As long as this is not the case, IAEA inspectors do not pose a serious risk to a country operating hidden facilities. However, external intelligence agencies could conduct similar missions. For example, prior to the alleged Israeli operation against a Syrian reactor in 2007, a special unit of the Israeli military was reportedly transported to the site of the reactor for the purpose of collecting environmental samples.

On-site sampling can be done only when specific information about a potential nuclear site exists. In most cases, however, one would need to look for a needle in the haystack (without actually knowing whether the needle exists). The most effective way to do this by environmental sampling is to use instruments that can measure the isotopic composition in air samples. Such instruments can have different detection range. In general, instruments with smaller ranges would have greater probabilities to detect enrichment activities. The problem with smaller ranges, however, is that for a large country such as Iran, a great many such instruments would be needed to measure every potential area. It is therefore a tradeoff between range and probability of detection. Moreover, topographical and metrological factors also affect such instruments' detection probability. Iran's topographical conditions, with hills and valleys, make effective air sampling difficult. Iran can therefore deduce that such methods are not likely to be undertaken by Western intelligence agencies unless specific information about a site is obtained.

A country seeking to hide nuclear facilities would therefore concentrate on obstructing two other potential dangers: satellite imagery and human intelligence. Satellites can help answer questions regarding whether construction is taking place under the surface, what kind of materials are used in the construction process and how deep one is digging, thus contributing to both the detection of a potential nuclear facility, its characterization and the planning of a military operation against the facility. However, given the large size of Iran, and the somewhat limited satellite assets of most countries, searching without leads for clandestine facilities are a challenging task. Again, specific site information could be necessary. However, wide-area satellite searches are still undertaken by Western intelligence agencies. One problem for a proliferator would thus be how to deceive satellites and their operators. There are several ways to camouflage a nuclear facility. First, one can camouflage a nuclear facility by pretending it is something else. The Syrians attempted to hide their reactor by constructing a fake exterior, making the building look like an old Byzantine fortress commonly found in Syria. Second, one can construct underground facilities without relevant functions in order to attract foreign intelligence agencies' attention and exhaust their resources, thus raising the probability that a real facility remains hidden. Some analysts believe that this was part of Hezbollah's deception tactics when it in 2006 conducted two digging operations in the village of Al-Hiam and near Ras Biada without attempting to hide the effort. Third, one can attempt to incorporate the facility into an existing industrial site such as an oil refinery or a milling plant, making it difficult for an outside watcher to identify nonindustrial nuclear activities. However, one would then have to accept a higher risk of detection from human intelligence since workers related to nonnuclear activities would also be present at the site. One problem experienced by Iran is that some of its scientists in the know have defected to the West. It is for this reason that Iran in 2008 chose to establish a counter-intelligence agency named Oghab-2, tasked with preventing sabotage and leaks as well as keeping an eye on Iranian scientists travelling abroad.

There is also the option of choosing a strategy for deniability in the case one is caught red-handed. The way to do this is to remove or reduce incriminating evidence before detection. One important consideration in this regard would be the configuration of the centrifuge cascades. There are several ways to configure cascades, some more suited for the production of weapons-grade uranium. If one would like to maintain at least some credibility in the case the facility is exposed, then one of the less effective configurations should be chosen (for example using single or tandem cascade orientations, which is deployed at Natanz and Fordow). The price one pays for this additional safeguard against detection is that one would need additional time to produce sufficient quantity of weapons-grade uranium (since the configuration is less optimal). There is therefore a tradeoff between deploying the additional safeguard and increasing the time available for detection.

Deciding on the location of the facility is also a major concern. Syria chose to locate its reactor in a desert region, hoping that it would not attract the interest of foreign intelligence agencies. Iran chose the same strategy with its Fordow facility. One can also choose to minimize on-site security in order to reduce signatures that could alert foreign parties. On-site security includes fences, security gates, guard posts, and antiaircraft positions. The latter is of particular importance. Antiaircraft positions are relatively easy to spot on satellite images. Therefore, using them for protection of a hidden facility would only alert foreign intelligence that something of value is located within the antiaircraft battery's action range. One is therefore left with two choices: either to construct the facility in an area already covered by antiaircraft batteries (in Iran's case, for example, in the area around Tehran) or to avoid such defensive measures altogether. However, after exposure one could rapidly deploy batteries to protect it. Foreign intelligence would thus prefer to keep mum on its knowledge of an attempted hidden facility, so as not to induce the country in question to increase its on-site defensive capabilities or to construct another hidden facility.

The last method deployed by a would-be-proliferator is to reduce an external actor's ability to destroy the facility. This can be done by locate the facility deeper underground, thereby making it less vulnerable to air strikes. One can also use building materials more suited for absorbing a strike, such as concrete or steel.

There is a never-ending cat-and-mouse game going on between would-be proliferators and foreign intelligence agencies seeking to obstruct them. They are both faced with several dilemmas. For a proliferator, the most important ones would be to decide the location of the facility and what defensive measures should be implemented (if any). For external intelligence agencies, the main problem would be choosing how to use their limited assets in the most efficient way and whether to keep knowledge of a site secret or expose it to world media.

Thomas Saether is a Norwegian security analyst specializing in Middle Eastern affairs. He is a post-graduate of the MA program in Security Studies at Tel Aviv University.