Media Disinformation & Nuclear Power

Since the widespread and significant destruction caused in Japan by the magnitude 9.0 earthquake just over two weeks ago, a few things have stood out. One is a rather substantial amount of attention raised towards charity and rebuilding efforts, which of course is a Very Good Thing. Another is the attention raised towards the damaged Fukushima nuclear power plants. Unfortunately, this has once again raised that rather poor reputation that nuclear power has for lack of safety – which I believe, based on the statistics, to be inaccurate and grossly exaggerated.

I have commented on nuclear power in the past, and my research has led me to consider it a useful and perhaps imperative source of energy in the near future as a replacement for fossil fuels. A large portion of the media disagree. However, it seems to me that several of the more outspoken critics of nuclear power in the media haven’t done their research, and base their criticisms solely on the conspicuous failures of nuclear power, whether they apply to the scenario in Fukushima or not. As such, I’d just like to make a few brief comments.

Comparing the scenario at Fukushima to Chernobyl is not accurate or helpful. Chernobyl, as anybody who has heard about it knows, was the single-most devastating disaster in the history of nuclear power, and represented the nadir of Soviet engineering. The disaster led to thousands of deaths and intensely damaged the reputation of nuclear power. Yet, as I’ve noted in the past, comparing modern nuclear power plants, or even the elderly Generation II power plants which the reactors at Fukushima represent, to Chernobyl is dubious at best.

The fact is that the Chernobyl power plant was (and still is) composed of crude, unsafe RBMK reactors designed for the production of weapons-grade plutonium, with any power production strictly being a useful side effect. There was no sort of concrete shielding around the reactor core, unlike all designs produced after and most contemporary designs, and safety systems were distinctly sub-par. Unlike the Fukushima scenario, the Chernobyl disaster was created by gross human negligence rather than an extraordinary natural disaster. The two can’t be compared directly if you’re looking for any sort of credence.

There were more pressing, more important issues to be addressing. I don’t think it escaped everybody’s attention that several thousands lay dead or missing in the aftermath of the earthquake and tsunami, along with tens of thousands more displaced from their homes. Given that this was happening there and then, I think it was more important to report than a nuclear power plant for which experts believed that the reactor core shielding had stayed intact, and therefore, based on precedent, was unlikely to cause any widespread damage. Yet, because of the nuclear connection, we got a lot of information on the nuclear power plant and not enough on the rather more devastating effects of the natural disaster. Way to set out your priorities, reporters!

The lack of objective fact being reported didn’t help. It seemed to me that this was just one of those perfect opportunities for anti-nuclear campaigners to try to dismantle the development of nuclear power again, despite the fact that their campaigns in the past hadn’t helped things and perhaps had slowed down the development of newer, safer power plants than the Generation II reactors at Fukushima. Where objective fact was reported, it was limited and lacked detail. Hardly what I’d call a fair look at the situation.

Historical Computer Systems – The EDSAC

At the beginning of the computer age, there were various computer systems of differing construction and operational principles. Among these were the Z3, an electromechanical computer designed by the German civil engineer, Konrad Zuse, which incorporated data storage through the use of punched 35mm film, the electronic Atanasoff-Berry Computer designed by John Atanasoff, Howard Aiken’s Harvard Mark 1 and the famous ENIAC, designed by John Mauchly and J. Presper Eckert. While these machines varied in design and operation, they were all built with one specific task: to make the task of calculation quicker.

Before George Stibitz demonstrated the relay-based electromechanical Complex Number Calculator, the earliest electrical tabulator to exist, complex calculation was done using slow desk-mounted mechanical tabulators. With the simultaneous developments of Zuse, Atanasoff, Aiken and Britain’s Alan Turing, which came just in time to assist the war effort in the Second World War, the world had seen the nascent developments of a technology which would eventually revolutionise society. First, though, the computer had to become more than an elaborate calculator.

In the years following the Second World War, work by Turing, Eckert and Mauchly, along with John von Neumann, the Hungarian-born polymath, led to the development of the stored-program computer using what became known as the von Neumann architecture. The release of von Neumann’s paper, First Draft of a Report on the EDVAC, led to the development of a number of stored-program computers around the world. The first of these was the experimental SSEM, or Small-Scale Experimental Machine, a prototype computer developed in Manchester and designed to test the Williams tube, an early form of computer memory. This, in turn, influenced the development of the EDSAC at the University of Cambridge.

The EDSAC, or Electronic Delay Storage Automatic Calculator, was the first practical electronic stored-program computer. Completed in May 1949, the EDSAC took input from punched tape and printed output through a teleprinter, both methods which would find use long after the development of the EDSAC. Output could also be displayed on a series of CRT monitors, an interesting capacity which would play a role in a historic piece of software for the machine.

A picture of the room-filling EDSAC at the University of Cambridge.

In the years preceding and during the operation of the EDSAC, a number of computer scientists and mathematicians were devising applications for computers which would go beyond the military and academic mathematics tasks that the computers had been performing to that point. Alan Turing was conceiving of artificial intelligence, others were considering mathematics in fields such as chemistry and biology, and a few people had even considered experimental computer games. One of these was A. S. Douglas, a mathematician at Cambridge who, as part of his doctoral thesis on human-computer interactions, designed one of the first computer games.

OXO, or Noughts and Crosses, to give it its full title, was, as the name suggests, a simulation of tic-tac-toe designed to test the EDSAC’s capacity for doing things other than routine mathematics. Being a game with simple rules, the game wasn’t very sophisticated by today’s standards, but it worked – something notable when discussing a computer made less than a decade after the very first electronic computer.

The EDSAC design led to other applications as well. J. Lyons & Co. Ltd., a now-defunct restaurant, food manufacturing and hotel company in Britain, whose products included the popular Ready Brek cereal, had invested in the EDSAC project. From the EDSAC design came LEO I, one of the first commercial computers ever produced, and the first computer to be used for business applications. LEO I (standing for Lyons Electronic Office) efficiently ran through inventory and payroll jobs, first for J. Lyons & Co. themselves, then later for Ford UK. A LEO I computer was also used by the Met Office before the acquisition of their own computer in 1959.

The EDSAC as originally designed ran with a clock speed of 0.5MHz, and had 512 17-bit words in its mercury delay line memory. This was later expanded to 1024 words – just over 2kB of RAM, in modern parlance. All of this operated at approximately 700 operations per second, moderately quick but slightly compromised at the time, and beatable by even the least sophisticated modern microcontroller. Given that the EDSAC required 12kW of electrical power to operate, it demonstrates just how far we’ve come in the sixty-one years since the EDSAC ran its first application.

The EDSAC’s historical value is unquestionable for being the first practical von Neumann-compliant computer to be built. However, there are other reasons for the EDSAC to be historically interesting. The first are its applications outside of pure and applied mathematics, such as the derived LEO I’s business applications and OXO. The second is the existence of a fully-featured simulator for the EDSAC, made in 1996 by the computer scientist and historian, Martin Campbell-Kelly. While it doesn’t give the experience of a heated room full of vacuum tubes and paper tape readers, it’s still an interesting insight into the programming techniques of a first-generation computer.

Reports from the start of the year also suggest that a replica of the EDSAC is to be built at Bletchley Park (home of the Colossus code-breaking computers during the Second World War). As there are few components remaining from computers of the first generation, and only a single complete one in existence – Australia’s CSIRAC, also known as the first computer to play digital music – this will prove to be one of the few chances to see a computer with first-generation technology in action. (EDIT 10/09/2013: As pointed out in the comments by Robert Dowell, the WITCH, a British computer also from the first generation, was painstakingly restored to working condition by The National History of Computing at Bletchley Park, and was made operational in 2012.)

The EDSAC was later superseded by the EDSAC 2, another vacuum tube computer which served until the mid-1960s. During its time, it was responsible for accelerating several mathematical fields, as well as forming a bedrock for the British computer industry. LEO Computers Ltd., formed to sell LEO business computers, eventually amalgamated with English Electric, which led to the formation of ICL, one of Britain’s historically most successful computer manufacturers. The von Neumann architecture underpinning the EDSAC would go on to underpin almost every computer which followed.