Notes about nuclear EMP
by Jerry Emanuelson
The main description of nuclear EMP on the Futurescience EMP page describes the mainstream scientific opinion about typical cases of nuclear EMP. On this page, I will go into some more detail and give some more detailed references.
There are several things that make it difficult to describe accurately what an EMP attack would be like:
In addition to the above uncertainties, the magnitude of all of the components of a nuclear EMP are dependent upon the strength and orientation of the Earth's magnetic field. In general, the strength of the EMP is roughly proportional to the strength of the geomagnetic field. The strength of the Earth's magnetic field is much more important than the orientation. The Earth's magnetic field varies according to location, and it also varies slowly over time at any particular location. Here are some examples of the variation of the Earth's magnetic field at various locations that might be of interest:
The magnetic field strengths are the total magnetic field strengths in nanoteslas. The angle shown in parenthesis is angle of the magnetic field downward direction. The influence of the downward tilt on the fast E1 pulse is quite complicated, but when there is less downward tilt, the intensity of the EMP tends to be more symmetrical around the detonation point. The fast E1 pulse will be greatest where the electrons traveling in a direction away from the detonation point penetrate the Earth's magnetic field at a right angle. This region is south of the detonation point in the northern hemisphere and north of the detonation point in the southern hemisphere. You can see from the table that the total magnetic field was weakest of any of the examples at the Starfish Prime detonation site, and also that the direction of the magnetic field was tilted toward the ground at a relatively shallow angle in comparison to many likely EMP targets.
From the best information that I can find, the total strength of the Earth's magnetic field at the detonation point of Soviet Test 184 in 1962 was very close to 46,900 nT. There are a lot of uncertainties about the exact detonation points of Tests 187 and 195, although they were in the same general area.
There is an English-language web site in Kyoto that has a lot of information about the Earth's magnetic field, including explanations of geomagnetic poles and a versatile and comprehensive online global magnetic field calculator. If you enter longitude and latitude into this calculator, be sure to observe NORTH latitude and EAST longitude. If you are using south or west coordinates, be sure to enter a minus sign in front of the number.
The United States NOAA now maintains its own geomagnetic field calculator. It is at:
There has been a lot of confusion among the general public because of conflicting statements about the weapon yield necessary to generate a significant EMP. One of the most comprehensive technical articles describing the magnitude of the E1 component generated by different types of weapons and weapon yields is A Calculational Model for High Altitude EMP by Louis W. Seiler, Jr (Report ADA009208). The graph showing peak electric field strength vs. various weapon yields and altitudes in the Wikipedia EMP article is derived from Figure 6 and Figure 8 from that Seiler report.
Those graphs as shown in the Wikipedia article, however, appear to be for a 30,000 nT (0.3 gauss) geomagnetic field. As Figure 9 from the original Seiler article shows, increasing the geomagnetic field to a higher strength (such as the 49,000 nT that exists over much of the United States) causes a proportional increase in the EMP field strength. Saturation effects are still encountered at a certain level, though, due to ionization of the upper atmosphere. The Seiler article can be quite confusing for non-technical readers because there are so many variables to consider (and the quality of the reproduction in the PDF from the Defense Technical Information Center is not very good). If it appears to be all too confusing, the message to take away from the article is that there are no simple answers to the question of EMP field strength vs. weapon size.
The most important consideration from the Wikipedia article graph (see the graph reproduced below) can be seen by looking at the lower blue thermonuclear/pre-ionization curve. That curve shows the E1 EMP from a typical thermonuclear weapon, which is triggered by a small fission explosion (and which, in a typical case, releases about .03 kilotons of prompt gamma rays that occurs a tiny fraction of a second before the main thermonuclear energy release). It shows that the smaller thermonuclear weapons (such as those that are preferred in today's arsenals) are much less efficient at producing E1 as compared to simpler single-stage fission weapons. This blue curve representing "typical" thermonuclear weapons are what would have been produced in 1962-era thermonuclear weapons which had no consideration for enhancing the EMP effect. Simple changes in weapon design could significantly alter the blue thermonuclear curve on this diagram. The advantage of single-stage fission weapons for producing E1 EMP disappears at multi-megaton yields (where single-stage fission weapons are impossible anyway).
Early fission weapons typically released about 0.3 percent of the total energy yield as prompt gamma radiation. About ten times as much gamma radiation is generated at the interior of the weapon, but most was absorbed by the chemical explosive and the weapon casing. The 21 kiloton Nagasaki weapon, for example, may have released about .06 kilotons as prompt gamma radiation, but was probably actually less because of its extremely heavy construction. By using newer chemical high explosives to start the detonation -- and by using more modern casing materials that are more transparent to gamma radiation -- a larger amount of prompt gamma radiation can be made available for generating the E1 EMP. (There are probably many other ways of enhancing the prompt gamma radiation from a small nuclear weapon. One obvious way might be by using U-235, in the form of moderately enriched uranium, rather than U-238, or natural uranium, for the tamper.)
Some scientists have disagreed with the prevalent opinion on the magnitude of EMP effects. One notable scientist who has made the case in scientific papers for a lesser EMP effect is Dr. Mario Rabinowitz. See his technical paper on the effect of fast EMP, and his technical paper on the effect of (late time) MHD-EMP. Dr. Rabinowitz was addressing these two papers primarily at EMP effects on the power grid, and he was working with only unclassified data (and consequently some of the data in the first paper is incorrect). Because of the fact that Dr. Rabinowitz did not have access to classifed data at the time (much of which has since been de-classified), he appears to have greatly underestimated the coherence of the E1 pulse. The high intensity and brief duration of the pulse are due to the fact that each of the 1025 initial electrons ejected by the gamma radiation is producing an electromagnetic pulse nearly simultaneously. This causes each of the very small pulses from each initial collision to add together to form a very large amplitude pulse. Dr. Rabinowitz does make an excellent point in the latter paper about transformers being much less likely to fail as a result of EMP vs. a solar storm due to the duration of each event (since solar storms last far longer, giving the transformers more time to overheat to destruction). That is not a reason to continue with the high level of risks that is being taken with the North American power grid due to the use of very old transformers, with few replacements on hand, and with no manufacturing capability for these very large transformers on the North American continent until very recently. There is still a significant risk of failures during an EMP just due to the extreme age of the transformers.
Also, there is a near-certainty of a large-scale failure of most of the large transformers in the power grid in the event of a solar storm of the magnitude of the 1859 solar storm or a larger geomagnetic storm. The risk of large-scale failure due to solar storms is especially great in Canada, along both coasts of the United States, and along the northern part of the continental United States and all of Alaska. A large solar superstorm would also cause severe damage to power grids in northern Europe and Australia. The United States could find itself at the end of a very long line waiting for transformers after a large solar superstorm, putting most of the United States without electrical power for a period of 4 to 10 years. The construction of the first factory of the 21st century on the North American continent for building these large transformers was completed in 2012, and another is scheduled for completion in 2013. The mere existence of factories is insufficient, however, since the transformers must be built and be available for use before an electromagnetic disaster occurs.
In the paper by Dr. Rabinowitz on the fast (E1) component of nuclear EMP, he correctly points out that in an attack using multiple detonations, the gamma radiation from the different detonations will interfere with each other since the first detonation will ionize the upper atmosphere. Like the pre-ionization caused by the initial fission explosion in a multi-stage thermonuclear weapon, the first nuclear explosion in a multiple-nuclear weapon EMP attack will ionize the upper atmosphere, making the subsequent explosions relatively ineffective at producing the fast E1 component.
Because of this ionization problem, a possible course of action for producing a maximum effect of both the E1 and E3 components would be to first detonate a weapon optimized for producing gamma radiation (while sacrificing some explosive power in order to maximize the gamma output and resulting E1 pulse), then follow a few seconds later with a thermonuclear weapon with maximum energy output in order to produce a maximum level of geomagnetically induced currents (the E3 component) in order to destroy most of the critical transformers in the power grid.
Contrary to popular opinion, it is not all that difficult to move from the ability to make simple fission weapons to producing powerful thermonuclear weapons. Production of any nuclear weapon requires a large industrial capacity for making the basic fissionable material (usually uranium-235 or plutonium). From that point, further enhancement requires mainly technical knowledge and computing power. The average personal computer now contains more than the entire world's computing power at the time of the detonation of the first thermonuclear weapons. It is likely that thermonuclear weapons are now possessed by more than just the five nations that have publicly announced that they possess stockpiles of nuclear weapons. (Nearly all experts in the field believe that Israel possesses thermonuclear weapons in the megaton class.) The first nuclear-armed countries required several years to go from basic nuclear weapons to the ability to produce powerful thermonuclear weapons, but they had neither the computing power nor the technical knowledge that is widespread today. In the United States and the old Soviet Union, they had to make several tries before they hit upon the radiation implosion on a dry-fuel design common to all second-generation thermonuclear weapons. In both the United States and the Soviet Union, the first test of this design produced a much bigger explosion than expected, and resulted in human fatalities in each case.
A map of the location of large transformers in the United States that are most susceptible to geomagnetic current destruction was presented at a 2008 NASA workshop. A smaller version of that same map (taken from an Oak Ridge National Laboratories report) is shown above. This map shows the transformers that would likely be affected in a 4800 nT/minute solar storm or E3 event at a latitude of 50 degrees. The outlined areas are the areas of probable system collapse, and involve a population in excess of 130 million people.
That same workshop showed this photo of a large transformer destroyed by a solar storm and the actual damage done by geomagnetic currents during the March, 1989 solar storm. The NASA article implied that the transformer is from Quebec, but was actually a 500 KV transformer at the Salem nuclear plant in New Jersey that was destroyed in the same event. The closeup of the damage as seen on the upper right closeup in the NASA/PSE&G image shows the secondary winding that is normally capable of carrying up to 3000 amperes of alternating current. Transformers are easily upset by the DC-like currents induced by solar storms and the E3 component of nuclear EMP. The 3000 ampere secondary winding was destroyed by geomagnetic currents that were only about ten percent of the normal maximum AC operating current.
It is important to note that, in addition to components of the AC electrical grid being much more sensitive to DC overload, most circuit breakers and other kinds of overload protection have a much different response to continuous AC than to either DC or brief transients.
According to this NRC bulletin, another solar storm on September 19, 1989 damaged another transformer at the Salem Unit 2 power plant. Additional details about the transformer damage during that geomagnetic storm are available at a page on the Solar Terrestrial Dispatch site.
In the case of the transformer at the Salem nuclear plant that burned up in 1989, a replacement just happened to be available from a delayed construction project in another part of the country, so the replacement was installed in about six weeks. Without this bit of good luck, the power plant would have been out of commission for about a year.
A detailed summary of the March, 1989 geomagnetic disturbance effects on the electrical grid in North America is available at the North American Electric Reliability Corporation web site.
Another excellent publication by the North American Electric Reliability Corporation on future risks, published in 2010, is their report on High-Impact Low Frequency risks to the North American electric power grid.
For a report on the South Africa solar storm of November, 2003 that damaged several major transformers in the South African power grid, see Transformer failures in regions incorrectly considered to have low GIC-risk by C. T Gaunt and G. Coetzee published by The Electrical Systems Planning Research Laboratory in Brazil. It is important to note that the geomagnetic latitude of South Africa (in the southern hemisphere) is about the same as Florida and southern California (in the northern hemisphere).
The major insurance underwriters, Lloyd's of London, has recently published an important report on the suspectibility of the North American power grid to solar storms. The report, Solar Storm Risk to the North American Power Grid is available for a free PDF download from the Lloyds web site.
The worst effects of the solar-storm-like E3 component of an actual nuclear test occurred during Soviet Test 184 on October 22, 1962, which induced currents in a long underground power line that started a power plant fire in Kazakhstan. Few details are known about this power plant fire, but the power plant fire probably began when abnormal currents burned up a transformer. In Test 184, the E3 pulse began rising immediately after the detonation, but did not reach its peak until 20 seconds after the detonation. The E3 pulse then decayed over the next minute or so. The E3 component of the EMP that caused the failure of the underground power cable, and the power plant fire, was 1300 nT/min (nanotelsas per minute) in the region of the fire during the first 20 seconds after the detonation. For comparison, the solar storm that shut down the entire power grid of Quebec on March 13, 1989 had a magnitude of 480 nT/min, and caused the Quebec power grid to go from normal operation to complete collapse in 92 seconds. A solar storm on May 14-15 in 1921 produced a disturbance of 4800 nT/min over a relatively small part of our planet.
If the United States W49 warhead used for the Starfish Prime test had been used in Soviet Test 184, the E3 component would have been more than 5000 nT/min in the region of the power plant fire in Kazakhstan. According to recent studies, a disturbance in the present-day United States of 4800 nT/min would likely damage about 365 large transformers in the U.S. power grid, and would leave about 40 percent of the U.S. population without electrical power for as long as 4 to 10 years due to the loss of large transformers in the power grid.
In general, the abnormal geomagnetically induced current is about 1 ampere for each nanotesla per minute (nT/min) of the geomagnetic disturbance, although this is very dependent upon a large number of variables such as the length and orientation of the electrical conductors.
For a comprehensive recent report on the effects of geomagnetic storms and the EMP E3 component, see Severe Space Weather Events -- Understanding Societal and Economic Impacts by the National Research Council of the United States National Academies.
Another good source of information on this same subject can be found in this online University of Minnesota seminar by John Kappenman on the electric power grid vulnerability to geomagnetic storms. Although this seminar is directed toward engineering students, there is much in the seminar that a non-technical person can understand.
In U.S. Congressional hearings on EMP, Congressman Roscoe Bartlett has often pointed out that everyone pays good money to insure their homes against fire, although the chance of an individual house being destroyed by fire is one event every 300 to 500 years. There is wide agreement by nearly everyone familiar with solar storms that the probability of a severe solar storm that would simultaneously destroy the most of the power grids of the United States, Canada, northern Europe and Australia (by destroying most of the large power transformers) is much greater than the probability of an individual house burning down. It would take years (or decades) to restore the power grids after such an event. The probability of a nuclear EMP is much more controversial, but most people put it at much higher than the probability of one's house burning down.   In any case, nothing has been done to prevent the long-term loss of the power grid; even though much more is being spent on fire insurance than what it would cost to put plans and parts in place to quickly restore the power grid after a solar superstorm or an EMP event.
Also, see the Oak Ridge National Laboratory report #6665 on Electric Utility Industry Experience with Geomagnetic Storms, which points out that MHD-EMP, also known as the E3 pulse, may have severe effects on smaller electric distribution systems that are not affected by solar storms.
Regarding the troubling state of large transformers in the United States power grid, this quotation from Dr. Lowell Wood, during Senate hearings (S Hrg. 109-30) on March 8, 2005 is especially important:
The situation described above by Dr. Wood does not only apply to a large nuclear EMP attack. A geomagnetic superstorm, such as the 1859 Carrington solar storm, would also produce a similar disaster. The effects of a geomagnetic superstorm, however, would likely occur in many countries simultaneously, and would disable at least some of the overseas transformer manufacturing capability. In the years since Dr. Wood made the above statement, the delivery time for a large power grid transformer under "normal failure" conditions has risen to 3 years. Fortunately, as stated earlier, new factories for large transformers are finally starting to be re-built in the United States. Two large transformer plants are operational, but they cannot continue to operate without electricity.
For a good source of simpler explanations of the effects of solar storms on the power grid (which also applies, at least to a certain extent, to nuclear EMP), see these excellent University Corporation for Atmospheric Research tutorials.
Another good general introduction to solar storms and the power grid is this page at the solarstorms.org web site. Also see this American Geophysical Union publication about geomagnetic storms and the power grid.
For a excellent summary by the United States Nuclear Regulatory Commission on their response to concerns about the effects of severe solar storm on the U.S. power grid, and possible severe damage that could result at nuclear reactor sites, see the December 18, 2012 NRC Notice of Proposed Rulemaking.
For a later report contradicting Dr. Rabinowitz's scientific paper on the magnitude of fast EMP, and showing that it would be much higher over the United States, see this brief report by Conrad L. Longmire. The peak amplitude of the Starfish Prime EMP in Honolulu was only about 5,600 volts/meter. Longmire shows that the Starfish detonation could have produced EMP field strengths in excess of 30,000 volts/meter if the warhead had been detonated over the United States.
For a collection of notes regarding the EMP detonations over Kazakhstan in 1962, see the minutes of a meeting with General Vladimir Loborev and other Russian officials who witnessed the tests. These notes were made from a meeting hosted by the U.S. Lawrence Livermore National Laboratories. (Reports of the Kazakhstan events can be confusing to some people because of name changes of some cities after the collapse of the Soviet Union and differences between the Russian and Kazakh languages.) Loborev gave further details of the Kazakhstan EMP events in a formal paper presented at a European Electromagnetics Conference in France in 1994. Also, see the page on this web site on the 1962 Soviet EMP tests for many additional details.
Additional important information about nuclear EMP is available from the United States EMP Commission. The 2004 Report is available here at the Global Security web site. A transcript of the session that introduced that report is here.
The most important report from the EMP Commission is the 2008 Critical National Infrastructures Report, which I have mentioned elsewhere.
EMP Commission reports often warn of the vulnerability of the circuits called SCADAs that control almost every important part of our infrastructures. An interesting presentation can be found here about the vulnerabilities of these circuits from a company that specializes in improving SCADA security. SCADAs are highly vulnerable to cyber-attack, as well as EMP.
For many years, the most important EMP simulator in the United States was a large structure called Trestle. A collection of notes relating to this device and EMP simulation and characteristics in general is now archived at the University of New Mexico related Summa Foundation site. One previously-mentioned 12-page brief report from this collection is particularly interesting because it shows why an EMP over the northern United States would be much stronger than the EMP from the Starfish Prime detonation.
The SUMMA Foundation at the University of New Mexico now has a 44-minute documentary movie online about the world's largest EMP simulator called TRESTLE: Landmark of the Cold War. : Dr. Carl E. Baum, who conceived the TRESTLE simulator, died of complications of a stroke on December 2, 2010.
A short 2007 article that summarizes the EMP problem and discusses the separate components of the EMP multi-pulse was published in this issue of Today's Engineer.
An interesting article on EMP, written by the late Conrad Longmire, the physicist who first figured out in detail the mechanism of high-altitude nuclear EMP, was published in 2004, but has only become available on the web in 2010. It is:
Longmire, Conrad L., "Fifty Odd Years of EMP", NBC Report, Fall/Winter, 2004. pages 47-51. U.S. Army Nuclear and Chemical Agency.
A U.S. Department of Energy report for Congress on the electrical power grid transformer situation and possibilities for mobile backups of somewhat smaller transformers is contained in a 2006 report called the MTS Report to Congress. As usual, lots of interesting reports, but almost no action.
The investigations of the EMP threat have come up with many solutions to convert a potential continent-wide megadisaster into a mere inconvenience.
The following quote is from Dr. Peter Vincent Pry of the United States EMP Commission staff at the same March 8, 2005 Senate hearing as mention in the above Lowell Wood quotation. Dr. Pry said:
In 2009, I wrote an article for Wikipedia about the 1962 series of United States high-altitude nuclear tests known as Operation Fishbowl. Since future changes to that article are now out of my control, I have posted my own version of the Operation Fishbowl article on this website.
As stated in the main article on EMP, in September, 2010, Oak Ridge National Laboratory published a series of reports for the Federal Energy Regulatory Commission, the Department of Energy and the Department of Homeland Security on the effects of electromagnetic disturbances on the United States electric power grid. The reports were written by the Metatech Corporation, and they provide an updated and comprehensive view of how electromagnetic disturbances such as nuclear EMP are likely to affect the United States electrical power grid. Many people will only be interested in the Executive Summary. Some of the other reports are hundreds of pages long.
An excellent overall source of scientific information about nuclear EMP continues to be:
which is a comprehensive source of material on the subject, maintained by Nigel B. Cook. It is an especially good source for links to technical reports and of eyewitness accounts of the 1958 and 1962 nuclear tests in space.
On other ways that the power grid can be destroyed other than by electromagnetic events, see:
Go to the main Futurescience EMP Page.
Go to the Sitemap of EMP Pages on this Web Site
Jerry Emanuelson's email address is firstname.lastname@example.org