Thursday, April 20, 2017

Another failed prediction

As you know, I love to point out the mistakes made by those who make future predictions. Since I was little more than a teenager, I have been saving clippings from the press and scientific journals that make more or less reasonable forecasts about the evolution of science and technology. In an earlier article I have pointed out that such predictions are seldom met, even when made by people who are both scientists and visionary, famous science fiction authors such as Arthur C. Clarke or Isaac Asimov.
I just unearthed an article published by Alexander Kusko in the IEEE Spectrum magazine in April 1968, with the following title:
A prediction of power system development, 1968 to 2030
And the following subtitle:
By predicting the trend of future power system design some 60 years hence, we should be better equipped to solve some of the technical and sociological problems that the industry faces today.
The assumptions on which Kusko's predictions were based were the following:
  1. The population will triple. What did actually happen? The world’s population in 1968, according to UN data, was about 3.5 billion people. The world population in 2015 was 7.35 billion. According to UN estimates, the world population in 2030 will amount to between 8.2 and 8.8 billion people. Far from the 10.5 billion estimated by Kusko.
  2. Most of electrical energy will be generated by nuclear power plants. Magnificent forecast, as nuclear energy currently generates about 10% of the world’s total electricity consumption.
  3. The total electric energy consumption will double approximately every ten years. This is the usual failed prediction. As energy consumption had been doubling every ten years for some decades, to predict what would happen in the next 60 years, it was assumed that this trend would continue indefinitely.
  4. Increased government regulation of the electric power industry. That this is a successful or a failed prediction depends on the country.
  5. Governments will be increasingly interested on pollution, the end of nuclear fuels and the conservation of nature. This assumption can be considered correct, although it is so broadly defined, that it can always be considered successful.
  6. Computers will be increasingly used to operate and control power systems. True. The same has happened with systems of all kinds.
  7. Load distribution will be increasingly managed. This is also true.
  8. Cheaper cables will be invented to carry greater power capacities. This assumption can also be considered correct.
So far the starting assumptions in the paper. The following are Kusko’s predictions:
  1. By the year 2000, the global production of electricity will be 10 times larger than that of 1967. By 2030, it will be 80 times larger. In fact, in 2000, the world production of electricity was only 2.5 times larger than that in 1973. In 2012, it was just 3.7 times larger. It is hardly feasible that in eighteen more years it will be multiplied by 20, so as to become 80 times larger than in 1967. This prediction should be considered failed.
  2. Beginning in the year 2000, superconducting cables must be used to withstand the enormous energy load. Another prediction that has not been fulfilled.
  3. Nuclear plants will need to be built in densely populated areas to serve their consumption needs. Another clearly failed prediction.
  4. Generators based on gas turbines and similar energy sources will only serve to provide energy variations around the base level. Taking into account that these energy sources are now providing over 60% of the world’s electricity consumption, this prediction must also be considered unsuccessful.
Evolution of World Energy Consumption
(Wikipedia)
As I have said before in these posts, natural processes are never exponential, except apparently during some time: sooner or later, growth curves get near to the logistic curve. In the attached figure, taken from Wikipedia, it can be seen that the growth of energy consumption between 1968 and 2015 has been approximately linear, rather than exponential. It is true that between 1920 and 1973 there was an almost exponential growth, but the oil crisis of 1973 led to great campaigns for energy saving that managed to control that growth. It is curious that the end of the exponential growth in energy consumption advanced by twelve years the turning point in the growth of world population.
I am proposing here a golden rule for the analysis of economic, technological and social predictions:
Any prediction based on the future extension of current trends is probably incorrect.


Manuel Alfonseca

1 comment:

  1. most interesting...your golden rule makes a lot of sense, although few will heed it

    ReplyDelete