I am confident that you have read the paper by Per Bak, titled How Nature Works: The Science of Self-Organized Criticality, which was published by Copernicus in 1996. Or not, perhaps.

But it is not too complicated. He determined that, if you made a pile of sand by steadily dropping sand grains on a surface, the pile would build up in a sort of conical shape and then, all at once, the pile would avalanche to a less tall but more widely spread pile.

I bet that has your attention. Okay, the point is that, when the pile is just on the point of avalanche, this is known as the point of ‘criticality’. But Bak found that the point of criticality seemed not to depend on how many grains of sand were used in the pile – sometimes the pile got to a certain height before it collapsed, sometimes it got higher. The number of sand grains, beyond a certain point, did not matter. What was important was that the pile always rose up in a cone, heading to a point where it would, without fail, given enough grains of sand, avalanche to a flatter pile.

The theory was thus born that avalanches, though regular, were dependent on the very complex interaction of the sand grains beneath, and at a certain point the whole pile would slump if just one grain of sand was added . . . or not, all depending on the degree of interaction of the all grains beneath. The matter is that the pile will collapse and, the higher it becomes, the bigger the collapse. We just do not know when.

Let us think then about Eskom and that each thing that has been done to Eskom by the present government is a few grains of sand being poured onto a pile. We can start with the now famous giving away of Eskom surplus generation capacity to the aluminium smelters in Richards Bay and Maputo, Mozambique. This was a government decision, not an Eskom one. Then, when Eskom wanted to build more power stations, request for a budget to build new power stations was denied in 1998.

Then the CEO, Allen Morgan, was replaced in 2000 by Thulani Gcabashe, a strange choice by government. Till then, the Eskom top dog had always been an engineer; Gcabashe holds a BA degree from the University of Botswana. He seemed not to understand how to supply coal to power stations effectively. There had been a good arrangement where collieries supplied Eskom on long-term contracts. Eskom received good, cheap coal.

In 2002, Gcabashe stopped this successful supply for political reasons. Eskom began to use small suppliers in the spot market, moving the coal by road. Suppliers were chosen by race, sex and size rather than by reliability of coal supply. The financial managers reduced the coal stockpiles to save money. The next CEO, Jacob Maroga, was an engineer but was hamstrung by government. By this time, Medupi power station construction started – if it had been built on time, it would have been a great help – but having to use the ruling party as a subcontractor and the many delays owing to political interference have made any benefit from Medupi too little, too late.

Brain Dames took over from Maroga as CEO. He had impeccable credentials but was given the responsibility of fixing every thing that government inter- ference had messed up, not the least of which was forcing Eskom to buy inter- mittent power from wind turbine operators at a cost double that for which Eskom could sell the power.

Now Eskom has a Minister of Public Enterprises, Lynne Brown, who arguably does not know much about electricity supply and generation. She has appointed as Eskom CEO Tshediso Matona, who, from his CV, knows less.

What is certain is that 15% of Eskom power is supplied to mines. The owners certainly will not stay in business if the Eskom supply is expensive and unreliable. They will think about generating their own power. The grains of sand are building up and forming piles. I think that Eskom, like many highly interactive systems, is reaching self- organised criticality. And when it all falls in, it will fall big.