Robert H. Gray, author of The Elusive Wow: Searching for Extraterrestrial Intelligence, has searched for radio signals from other worlds using the Very Large Array and other radio telescopes. You’ll find numerous links to his work in the archives here. In today’s essay, Gray takes a look at a classic benchmark for assessing the energy use of civilizations, introducing his own take on Earth’s position in the hierarchy and how these calculations affect the ongoing SETI effort. His article on the extended Kardashev scale appeared in The Astronomical Journal Photograph by Sharon Hoogstraten.

by Robert H. Gray

Human civilization has come an amazingly long way in a short time. Not long ago, our major source of energy was muscle power, often doing hard work, while today much more energy is available from fuels, fission, hydro, solar, and other sources without breaking a sweat. How far can civilization go?

It’s probably impossible to say how far civilizations can go in areas like art or government, because such things can’t be measured or forecast, but energy use is measurable and has trended upward for centuries.

The astrophysicist Nikolai Kardashev outlined a scheme for classifying civilizations according to the amount of energy they command, in order to assess the type of civilization needed to transmit information between stars. He defined Type I as commanding the energy available to humanity in 1964 when he was writing, Type II could harness the energy of a star like our Sun, and Type III would possess the energy of all of the stars in a galaxy like our Milky Way.

Harnessing the energy of stars might sound like science fiction, but solar panels are already turning sunlight into electricity at a modest scale, on the ground and in space. Gerald O’Neill and others have envisioned orbiting space settlements soaking up sunshine, and Freeman Dyson envisioned something like a sphere or swarm of objects capturing all or much of a star’s light.

Carl Sagan suggested using Arabic numerals instead of Kardashev’s Roman numerals, to allow decimal subdivisions, and he suggested more uniform power levels. He re-defined Type 1 as 1016 watts—very roughly the Sun’s power falling on the Earth—and he rounded off Type 2 and 3 levels to 1026 and 1036 watts respectively, so planetary, stellar, and galactic categories increase in steps of 1010 or ten billion. A simple formula converts power values into decimal Types (the common logarithm of the power in megawatts, divided by ten). In the recent year 2015, human power consumption was 1.7×1013 watts, or Type 0.72—we’re short of being a Type 1.0 planetary civilization by a factor of roughly 600. In 1800 we were Type 0.58, and in 1900 we were Type 0.61.

The 2015 total power consumption works out to an average of 2,300 watts per person, which is 23 times the 100 watts human metabolism at rest, but it’s not many times more than the 500-1,000 watts a human can produce working hard. Maybe we haven’t gone all that far, yet.

I recently extended the scale. Type 0 is 106 watts or one megawatt, which is in the realm of biology rather than astronomy—the muscle power of a few frisky blue whales or several thousand humans. That seems like a sensible zero point, because a civilization commanding so little power would not have enough to transmit signals to other stars. Type 4 is 1046 watts, roughly the power of all of the stars in the observable Universe.

One use for the scale is to help envision the future of our civilization, at least from the perspective of energy. If power consumption increases at a modest one percent annual rate, we will reach planetary Type 1 in roughly 600 years and stellar Type 2 in 3,000 years—roughly as far in the future as the Renaissance and ancient Greece are in the past. That simplistic growth rate would put us at galactic scale Type 3 in 5,000 years which is almost certainly wrong, because some parts of our galaxy are tens of thousands of light years away and we would need to travel faster than light to get there.

There are, of course, many limits to growth—population, land, food, materials, energy, and so on. But humans have a history of working around such limits, for example producing more food with mechanization of agriculture, more living space with high rise buildings, and more energy from various sources. It’s hard to know if our civilization will ever go much beyond our current scale, but finding evidence of other civilizations might give us some insight.

Another use for the scale is to help envision extraterrestrial civilizations that might be transmitting interstellar signals, or whose large-scale energy use we might detect in other ways.

If we envision ET broadcasting in all directions all of the time, they would need something like 1015 watts or 100,000 big power plants to generate a signal that our searches could detect from one thousand light years away using the 100-meter Green Bank Telescope. That means we need to assume at least a Type 0.9 nearly planetary-scale civilization—and considerably higher if they do anything more than broadcast—a civilization hundreds or thousands of times more advanced than ours. That seems awfully optimistic, although worth looking for. If we envision civilizations soaking up much of a star’s light with structures like Dyson spheres or swarms, then unintentional technosignatures like waste heat re-radiated in the infrared spectrum conceivably could be detected. Some infrared observations have been analyzed along those lines, for example by Jason Wright and associates at Penn State.

If, on the other hand, we envision ET transmitting toward one star at a time using a big antenna like the 500 meter FAST in China, then we need to assume only something like 108 watts or one-tenth of one big power plant, although the signal would be detectable only when the antenna’s needle beam is pointed at a target star. To catch intermittent signals like that, we will probably need receiver systems that monitor large areas of sky for long periods of time—ideally, all-sky and full-time—and we can’t do that yet at the microwave frequencies where many people think ET might transmit. A modest prototype microwave receiver system called Argus has been monitoring much of the sky over Ohio State University in Columbus for a decade with very low sensitivity, and an optical system called PANOSETI (Panoramic SETI) is planned by Shelly Wright of UCSD and Paul Horowitz of Harvard to potentially detect lasers illuminating us.

Detecting some signature of technology near another star would be a historic event, because it would prove that intelligence exists elsewhere. But the U.S. government has not funded any searches for signals since Sen. Richard Bryan (D-NV) killed NASA’s program in 1993, even though thousands of planets have been discovered around other stars.

Both Kardashev and Sagan thought civilizations could be characterized by the amount of information they possess, as well as by energy. An information scale much like the energy scale can be made using 106 bits or one megabit as a zero point—roughly the information content of one book. Sagan thought that 1014 or 1015 bits might characterize human civilization in 1973 when he was writing on the topic, which would be Type 0.8 or 0.9 using the power formula (he used the letters A, B, C… for 106, 107, 108… bits, but letters don’t allow decimal subdivisions). More recent estimates of humanity’s information store range from 1018 to 1025 bits or Types 1.2 to 1.5, depending on whether only text is counted, or video and computer storage are included.

Nobody knows what information interstellar signals might contain. Signals could encode entire libraries of text, images, videos, and more, with imagery bypassing some translation problems. What might motivate sending information between stars is an open question; trade is one possible answer. Each world would have its own unique history, physical environment, and biology to trade—and conceivably art and other cultural stuff as well. Kardashev thought that the information to characterize a civilization could be transmitted across the Galaxy in one day given sufficient power.

Whether any interstellar signals exist is unknown, and the question of how far civilization can go is critical in deciding what sort of signals to look for. If we think that civilizations can’t go hundreds or thousands of times further than our energy resources, then searches for broadcasts in all directions all of the time like many in progress might not succeed. But civilizations of roughly our level have plenty of power to signal by pointing a big antenna or telescope our way, although they might not revisit us very often, so we might need to find ways to listen to more of the sky more of the time.

Additional Resources

N. S. Kardashev, Transmission of Information by Extraterrestrial Civilizations, SvA 8, 217 (1964).

C. Sagan, The Cosmic Connection: An Extraterrestrial Perspective, Doubleday, New York (1973).

V. Smil, Energy Transitions: Global and National Perspectives, 2nd edition, Praeger (2017).

R. H. Gray, The Extended Kardashev Scale, AJ 159, 228-232 (2020).

R. H. Gray, Intermittent Signals and Planetary Days in SETI, IJAsB 19, 299-307 (2020).