Report by the US Geological Survey
UNCLASSIFIED REPORT NO. 01-28
Engineering and geological constraints regarding the feasibility of underground nuclear tests in large cavities with explosion attenuation (decoupling).
Dr. William Leith, US Geological Survey, Reston, Virginia 20192
United States Department of the Interior
Analysis of geological issues
Until very recently, this report could be downloaded from the publications section of the USGS website at:
http://geology.er.usgs.gov/eespteam/pdf/USGSOFR0128.pdf
This report no longer exists, even in the USGS databases (http://search.usgs.gov/).
A backup copy can be found here: http://membres.lycos.fr/atar/Archives/Report01_28.pdf
We owe this fortunate backup to the foresight of Christophe Giudicci, who discovered its existence. It is a key piece of evidence in the “clandestine nuclear tests” case file. This file was submitted to the judge during the appeal trial in early 2003, in which I was convicted of defamation following a dispute with Antoine Giudicelli. I commented extensively on it during the hearing when the judge asked that “we get to the heart of the matter.” It was not mentioned in the judgment. See also the comments on the judgment.
Below is a brief overview of this report’s contents.
Preliminary work
Over the past 40 years, the US Geological Survey has worked to monitor nuclear tests worldwide and ensure compliance with relevant treaties.
- Systems for detecting nuclear explosions
- Effects on the Earth’s surface and environment, mainly concerning tests conducted abroad; location of testing sites
- Assessments related to so-called “peaceful nuclear tests”
- Comparative studies of seismic effects caused by nuclear explosions versus those due to natural seismic activity or mining-related explosions
- Participation in drafting treaties limiting nuclear testing
- Development of seismological databases to facilitate detection of nuclear explosions
- Distinguishing between nuclear explosions and earthquake effects
- Studies on wave attenuation within the Earth’s crust
- Research on the natural attenuation capacity of porous soils at various depths (Matzko, 1995)
- Natural cavities suitable for installing attenuation devices, focusing on salt domes or layers where salt occurs, as well as areas conducive to creating large underground caverns.
An attenuation scenario (decoupling)
One of the main challenges of the Comprehensive Nuclear Test Ban Treaty (CNTB) is determining whether countries might conduct clandestine tests—that is, tests capable of evading standard detection systems. Possible scenarios include:
- Detonating a nuclear charge in space
- During an earthquake
- In an environment offering natural attenuation
- In a remote marine environment
- In the atmosphere under heavy cloud cover
- Avoiding detection by detonating charges inside sufficiently large cavities at appropriate depths.
All these methods have been thoroughly studied; numerous articles have been written on the subject (Herbst and Werth 1980, Glenn and Goldstein 1994, Sykes, 1995, Linger et al. 1995).
Since Albert Latter proposed the method of detonating a charge inside a cavity in 1959 (Latter et al. 1961), considerable research has been done to theoretically model this phenomenon. The USA and USSR conducted tests involving signal attenuation, as described in papers by Springer et al. 1968, Murphy & Al. 1995, Reinke, 1995.
In 1988 it was concluded that monitoring could be performed for explosions exceeding 10 kilotons; however, no reliable methods existed to eliminate the associated seismic signals.
For explosions below one to two kilotons, it was determined that such tests could be conducted in violation of treaties in environments like granite, alluvial deposits, or salt formations, and that current technology offered no reliable means to detect them.
Between these two ranges (above 10 kilotons or below 1 kiloton), there exists a zone where detection remains problematic even when attenuation techniques are employed.
The purpose of this paper is to review attenuation techniques involving detonation inside cavities (Sykes, 2000). This review focuses on cavities filled with air. Other attenuation methods have also been considered, including porous, cellular materials capable of collapsing during an explosion while absorbing energy. It has been demonstrated that underground nuclear tests are feasible in porous rocks with porosity levels ranging from 5% to 20%. In regions such as the Kalahari, porosity can exceed 20%.
The following criteria determine whether covert nuclear explosions can be carried out:
- The seismic signal must remain below the detection threshold of monitoring instruments.
- The test depth must be sufficient to contain radioactive products after the explosion, preventing detection systems from distinguishing them from natural signals.
- The test site must be inaccessible to satellite surveillance.
Attenuation in ellipsoidal cavities
Both Americans and Soviets conducted numerous tests using high-power chemical explosives in cavities with an elongation ratio of up to 4:1. In the USSR, these tests took place in Kyrgyzstan in 1960. Similar tests were conducted in Magdalena, New Mexico in 1994. Their aim was to verify predictive models.
Attenuation factors in salt and granite
According to a 1988 OTA report, spherical cavities measuring 25 meters in diameter in salt and 20 meters in granite can provide sufficient attenuation for one-kiloton charges; the detonation depth was set at 825 meters. Sykes (1995) suggests these estimates should be increased slightly, though they remain valid.
Environmental considerations for underground cavity construction
Cavity-based explosion attenuation requires constructing relatively large spaces without support pillars. To minimize detection risks, the following factors must be considered:
- Geological analysis of the environment
- Preliminary data on feasibility in such soil
- Soil stability information
- Attenuation efficiency
- Containment capabilities regarding reaction products
- Operational discretion
- Cost
Constructing cavities in salt
Salt deposits at moderate depths (100–1500 meters) offer ideal properties for both attenuation and containment due to their rheological characteristics. Salt remains compact under short-term loads yet exhibits plasticity and impermeability over time. The USA and USSR conducted low-signature underground nuclear tests in salt cavities reaching powers of 10 kilotons. In 1999, Davis & Sykes concluded that this method was the most practical way for nations to conduct covert underground nuclear tests.
The oil industry and other energy sectors have excavated thousands of such cavities for storing hydrocarbons like crude oil, propane, butane, ethylene, and compressed air. According to Kedrovskiy, 1974, the USSR also created many such cavities via nuclear explosions.
Although salt deposits are widespread globally, only a limited number possess sufficient thickness. These deposits typically occur alongside layers of calcium carbonate (CaCO3), gypsum (Ca5O4*2 H2O), and anhydrous sylvite. Common salt is NaCl (sodium chloride). Salt layers range from several meters to hundreds of meters thick; “salt domes” also exist.
(...)
Excavation techniques. Size of cavities... Stability..
Sites with thick salt deposits:
China – Numerous
France – Numerous
India – Limited to the Kumaun region; depth/thickness unknown
Iran – Numerous
Iraq – Numerous
Israel – Limited sites near the Dead Sea; depth/thickness unknown
Libya – Limited to northeast border with Tunisia; depth/thickness unknown
Croatia – No data
Pakistan – Limited to the Sargodha region
Russia – Numerous
England – Limited
USA – Numerous
Two-million-cubic-meter salt dome cavities are rare; however, artificial cavities of 17 million cubic meters have been constructed.
Below are required diameters for different charge sizes:
Charge | Radius | Volume enabling full attenuation
1 kT | 25 m | 65,500 m³
5 kT | 43 m | 333,000 m³
etc.
Construction costs..
.......
Excavating cavities in hard rock
Numerous spherical cavities without supports, reaching diameters of 30 meters, have been built.
Data on large-scale cavities was published in France by Duffaut in 1987.
Records show maximum dimensions of 70 meters. The largest natural cavern measures 400 meters and is located in Indonesia. France hosts a cavern 230 meters wide with a volume of eleven million cubic meters.
...the following pages discuss feasibility of excavating cavities in various materials. Specialized machinery, known as “Jumbos,” has been developed for such tasks. The report provides data on diameters and volumes required in hard rock to achieve full attenuation.
20 meters for a one-kiloton charge
Subsequent sections address containment of nuclear reaction products and summarize previous nuclear experiments (Table 6), detailing power levels, dates, soil types, and depths.
Approximate formulas are given for determining optimal detonation depths; further considerations regarding containment are discussed.
....
French experiments:
Key data on French experiments is summarized. Detonation depths are expressed as:
Meters × (power in kilotons)^1/3
Out of 140 Pacific tests, 121 (86%) produced no emissions.
In 8.4% of cases (remaining tests), a chimney formed due to collapse of the cavity created by the explosion.
Explanation: When detonating in unprepared volumes without aiming for secrecy, blast walls become vitrified. For a one-kiloton charge, a chamber forms around the bomb container, reaching only a few dozen cubic meters in size. This volume fills with extremely hot gas at high pressure; some escapes into surrounding material based on its porosity. Cooling then causes outward expansion, resulting in upward propagation of collapse. This creates characteristic depressions above explosion epicenters (e.g., Nevada desert). In alluvial terrain, this poses no problem. In French experiments, collapse occurred in 8.4% of cases, accompanied by fissuring that could release radioactive tritium, strontium, and cesium into karstic rocks and seawater. In four other tests (3%), tritium was detected in karstic layers despite no surface leakage.
Discussion:
Above 10 kilotons...
....... (not relevant to us)
Less than one kiloton:
Complete attenuation is achievable for explosions up to this power level provided cavities with the following diameters are used:
25 meters in salt
20 meters in hard rock
Many such cavities already exist or have been constructed worldwide. Excavation poses no major technical hurdles; costs remain modest compared to nuclear program development.
Containment remains a challenge; best results are achieved with salt. With hard rock, good containment can be attained if proper site selection, rock type, and depth are chosen. Pre-testing via chemical explosions helps assess cavity tightness. This was done during development of covert nuclear techniques in Nevada. Optimal test sites minimize seismographic detection likelihood (areas with relatively high seismic activity help mask signals in background noise).
Fully decoupled explosions under one kiloton produce seismic magnitudes below 2.6 m b.
Note: Blasting activities at Gardanne, at 1000-meter depth, generate magnitude 3 signals. Thus, covert nuclear explosions would yield similar signatures. In summary, for all Northern Hemisphere nations, events of such low magnitude would escape detection by the international monitoring system (CTBT).
Charges exceeding 10 kT
...not relevant to us
Disguise, deception, denial.
Conducting tests in active mining regions allows attribution of signals to normal mining operations via plausible justifications. If tests occur in underground storage cavities or salt mines, existing surface infrastructure naturally conceals operations. Fluid-filled cavities allow pumping without suspicion. Post-test, weak signals may be attributed to geophysical exploration, cavity collapse, or natural seismicity (see Leith & Simpson, 1986).
Similarly, underground nuclear tests in hard rock mines can be disguised as mining activity itself. Most nations capable of nuclear proliferation engage in routine mining activities (except Pakistan).
Explosions equivalent to 1000 tons of TNT linked to mining are rare; however, deep rock fracturing in certain regions can generate signals surpassing magnitude 5, potentially mimicking 10+ kiloton nuclear explosions.
Distinguishing between normal mining activity (typically magnitude 3–4 signals) and covert, attenuated nuclear explosions (also magnitude 3) becomes extremely difficult.
In granite, for charges between 1–10 kT, magnitudes range from 2 to 3.5 depending on tectonic environment; detection by Northern Hemisphere monitoring systems remains uncertain. Moreover, signals may remain unidentifiable (...)
General conclusion:
(...)
Under 10 kilotons, seismic signals from underground nuclear tests in appropriately sized spherical cavities would fall between magnitude 3 and 3.5. For one-kiloton charges, signals would reach magnitude 3...
Our commentary:
Based on this report’s conclusions, sophisticated attenuation techniques aren’t necessary for a one-kiloton nuclear charge (small fission bombs can serve as energy sources for currently developing electromagnetic weapons). It suffices to detonate within a sufficiently large spherical cavity: 20 meters in hard rock, 25 meters in salt. Such excavation poses no technical obstacles. Expansion of hot plasma gases efficiently reduces shock impact on surrounding material; resulting seismic signals reach merely magnitude 3, compatible with normal mining activity (e.g., tunnel blasting). In short: available seismographic data wouldn’t distinguish covert nuclear tests from regular mining operations.
Other attenuation techniques may also be combined. Since primary energy for electromagnetic weapons derives from “small A-bombs,” much of this energy converts to electromagnetic radiation; this system is inherently “stealthy,” as emitted radiation remains undetectable at sufficient depths.
Ultimately, France could have conducted countless underground nuclear tests at typical depths of 1000 meters, aimed at developing electromagnetic weapons powered by low-yield A-bombs, all largely undetected.
Back to News Back to Guide Back to Home Page
Number of views since September 29, 2003: