Recently, I received a message from a colleague, David Monyae, who offered the critique that I have been speaking much about the fourth industrial revolution while SA is fraught with the problems of the second, such as security of electricity supply.
It is a misconception that one needs to be able to master technologies from earlier revolutions to understand the technologies of the fourth industrial revolution. Nonetheless, how is electricity generated and why are we struggling to supply it?
There are two main ways of generating electricity. The dominant way is to place an electric conductor, such as copper wire, close to a magnet and move it. This process makes an electric generator. Eskom generates electricity by moving a giant electric conductor located close to a giant magnet. The second way of generating electricity is to harvest it directly from the sun using devices such as solar panels.
If generating electricity is as simple as moving a giant conductor located close to a magnet, how come Eskom is struggling to achieve it? To understand this, we need to understand how such giant conductors close to magnets are moved.
There are many ways in which these conductors are moved, and one of these is to use coal to heat water, which generates steam to push the conductor. Another way is to use nuclear fission, where the atoms of radioactive materials such as uranium are separated, thereby releasing heat, which boils water, which becomes steam, which moves the conductor.
Once electricity is generated, it is transmitted from the power station to a location or a city, from where it is distributed to houses.
Eskom generates, transmits and distributes electricity. However, some municipalities, such as the City of Johannesburg, distribute electricity within the city.
The concept of municipalities generating electricity was so difficult that all of them abandoned it.
What remains are unused electricity generation plants like the one in Soweto overlooking the University of Johannesburg campus in the township. It is now used for bungee jumping.
So why is Eskom struggling to deliver electricity? The first problem is commercial. Eskom is unable to collect its revenue. It is reportedly owed more than R28bn by municipalities. Soweto residents reportedly owe it as much as R15bn. Furthermore, it is unable to optimally price coal, and carries a huge debt of R450bn. The other commercial problem is that the business model adopted after the construction of the Medupi and Kusile power stations was based on high tariffs, and the National Energy Regulator of SA has stopped this model.
Eskom requires a complicated treasury department that is skilled in pricing options and derivatives instruments and thereby is able to handle future costs of debt, coal and accounts payable.
The second problem is technical. The people who designed the Medupi and Kusile power stations may have been brilliant engineers, but they were not practising in SA. They designed the two power stations and worked with locally registered engineers to evaluate the designs.
Medupi and Kusile are the largest supercritical power stations in the world. Supercritical power stations involve a complicated design and have never been constructed at this scale. They have to operate at specific temperature and pressure, and, unlike ordinary coal power stations - where steam is used to move the conductor - in supercritical power stations it is a superfluid - neither liquid nor gas - that moves the conductor.
Why we pursued this difficult and untested technology on this scale needs to be understood.
The construction process followed was not based on the turnkey principle, where the project is constructed and handed over to the customer as a completed product, but on the order-to-build principle. The question why we did not have adequate professional engineers, engineering technologists, project managers and financial specialists to be able to negotiate a good construction deal, especially given that it was a first of its kind, needs to be answered.
The result is that the construction of these two power stations, which was expected to cost approximately R200bn, ended up escalating to R450bn. The technical problem became the commercial problem.
This naturally led to maintenance problems. If a company is short of money and the skills balance sheet is also weak, maintenance is compromised.
Three maintenance strategies are deployed in industry: run-to-failure, scheduled maintenance and predictive maintenance. Run-to-failure is where assets are used until they fail. The disadvantage of this is that it leads to unplanned power cuts. In scheduled maintenance, assets are replaced when they reach the end of their life cycle. In this strategy, assets that still work are replaced to prevent unexpected power cuts. When there is a scheduled time for power cuts, it is because the power supplier is performing scheduled maintenance. The maintenance strategy of the fourth industrial revolution is predictive maintenance.
The question we ought to ask is what strategy is Eskom using to maintain its assets?
President Cyril Ramaphosa announced in his state of the nation address that he will break up Eskom into three parts: generation, transmission and distribution. The idea of breaking up Eskom is not new. In the previous decade, the government established Electricity Distribution Industry Holdings, meant to consolidate all electricity distribution in one company. Ultimately, this was abandoned.
It is important that this time we pursue what is optimal for SA and create structures that are efficient from financial, technological and operational perspectives.
To run an electric company is a multiple-objective and complicated exercise - optimising the commercial objective acts against intensifying maintenance and technical capacity.
We need to be decisive and understand that to sort out Eskom we need to invest, and this will require financial sacrifice. Second, we need to technically capacitate Eskom from the board to the lowest levels. Third, we need to adopt a maintenance strategy and stick with it.
• Professor Marwala is the vice-chancellor and principal of the University of Johannesburg and author of 'Handbook of Machine Learning'. He writes in his personal capacity..