The Fukushima Syndrome, 1

This is the ninth article in the Where We Came From series.

On March 11, 2011, a wall of water made an incursion on the Pacific coast of Japan, travelling six miles inland to erase 15,000 people and the places they lived.

The tsunami also attacked the Fukushima Daiichi nuclear power plant and took out its generators, leaving the reactor for dead. What followed was a desperate attempt to shut down the reactor’s enraged metal heart in a battle that continues to this day. Fukushima is not Chernobyl but the environmental impact remains significant as radioactive material continues to leak from the site.

In my teens, I was fascinated with nuclear physics and still retain an unhealthy familiarity with the mechanics of nuclear fission. Yet I learnt nothing of reactor design and where it could go wrong. I wondered how such an incident could still happen in the wake of previous incidents at Windscale, Three Mile Island, and Chernobyl. What is it about a nuclear reactor that is susceptible to this scale of disaster? Why can’t you just switch the damn thing off?

And so, instead of being sensible and using the vast resources of the internet to answer these questions, I fired up a nuclear reactor simulation written by Chris Crawford in Atari Basic.

Origins

SCRAM: A Nuclear Power Plant Simulation (Atari, 1981) featured in many of the early Atari marketing brochures as Atari wanted to push the more serious side of their new Atari 400/800 home computer range. Atari was struggling to get away from their name being associated with games due to the explosive success of the Atari Video Computer System (later rebranded as the 2600), and their trademarked slogan in these early years was “Computers for people”.

SCRAM, though, is a game, complete with score that isn’t called score. The game looked complex, but not seductively so: it resembled work rather than play, and I never developed any interest in it.

Looking at it now, the tone of both manual and game is not to scare, but to inform and educate. It was an extension of a nuclear power plant simulator Crawford had been touring Californian high schools with prior to becoming a game developer.

SCRAM was developed in the wake of the 1979 Three Mile Island (TMI) accident, in which a core meltdown occurred, and the manual makes several explicit references to it including a full account of the accident in an appendix. The TMI accident was the result of both equipment failure and operators unable to interpret a disaster in-progress from reactor instrumentation. SCRAM reflects both these aspects in its simulation.

It doesn’t say anything like this:

Nature Hates Nuclear

SCRAM models a pressurised water reactor (PWR) – the most common reactor design in use – and the manual goes into great detail on how the pieces all fit together. However, Fukushima Daiichi employs a boiling water reactor (BWR) design. The difference between the two designs is not significant enough to kill my little experiment. Both models use water (called coolant in this context) to carry heat away from the reactor core. In a BWR, the coolant around the core converts to steam and powers a turbine directly while the PWR transfers the heat to a separate system with a turbine.

The game both simplifies and complicates the running of a reactor. While the controls and indicators offered the player are merely a handful, all the usual safeties and redundancies have been omitted. SCRAM emulates real-life by not having any sort of “damage control” readout. The player’s job is interpreting temperature and pressure indicators to read the reactor state, stabilising the reactor after component failures and sending workers to fix faults.

SCRAM pits the player against a vengeful nature that throttles the reactor with an endless sequence of earthquakes. Each earthquake damages precisely one component and you only have a finite amount of workers with which to handle repairs. It’s not about how you win because nature always wins; it’s about how long you don’t lose. Your job in the simulation is to keep the plant running long enough to hit a specific energy production target then put the reactor into cold shutdown. The game is a strange paradox. Crawford’s intent appears to be educational with a pro-nuclear stance but the game contradicts this entirely: reactors always fail.

I spent serious commuting time picking apart forty pages of SCRAM instruction, previously celebrated by Gamephemera (you can also download a PDF of the manual over there). But then I hit the section titled “Thermodynamics 1A”.

Thermal conductivity? Temperature gradients? This did not look good.

The Fukushima Syndrome concludes next week.