The source of the boiler and mill problems at Medupi and Kusile

They didn't listen to the experts

One of the first lessons in engineering is to learn from the mistakes of others, and that is why it is so important to always speak to experienced engineers that are about to retire or those who have already retired.

After commenting on national television on Eskom’s performance and asking a few friends in the industry around as to what is going on, I was alerted to a 2019 article that was written by Mr. Alex J. Ham who worked for Eskom for 30 years, where he held the position of Chief Engineer of Power Station Design. He retired in 1994, but clearly he is concerned about the way that the current power stations are being operated.

Eskom’s Unplanned Capacity Loss Factor is 16 GW on average with a 4 GW standard deviation (ideally both numbers should be close to zero), meaning that the current management are out of their league in terms of maintenance and operation.

As it turns out Kusile and Medupi’s 800 MW nameplate capacity doesn’t mean much, because they were not designed for South Africa’s coal quality. Eskom was trying to cut costs in the design and now it is biting them.

• The boiler internal volume and dimensions are undersized for the slow-burning coals used.

• The flue gas velocities are too high, causing excessive erosion of the burners, as well as of the furnace tubing.

• The change during the early contract stage from the tube mills originally specified, to vertical mills, was a major error. The change was proposed by the boiler contractor with a price reduction, which Eskom accepted. The pulverised coal fineness required cannot be achieved, and the mill wear will be excessive.

As is usual with experienced engineers, they often come with the solution immediately after having identified the problem. The reason being that they draw from past experiences and their own costly mistakes.

To my mind, the possible available corrective actions and way forward are:

• It would not be realistic to attempt to beneficiate the coal to improve its reactivity. This would only reduce the ash content, but the cost could not be justified.

• The source of the problem is the coal pulveriser mills not being able to produce the more stringent fineness required. This results in coal/char particles burning higher up in the furnace, causing excessive reheater temperatures, etc.

• One could attempt to increase the rotational speed of the vertical spindle mills by at least 15%. This would require gearbox modifications. It would also result in increased mill internal wear, but could improve the fineness of the pulverised coal produced.

• In my opinion, the only real option for a solution is to change the vertical mills to tube mills of similar to those used at Lethabo, Kendal, etc. This should be done to the uncompleted units at Kusile first. This is a major construction change, but in my view, possible and worthwhile, and would change the combustion behaviour in the entire boiler. There would be two years to complete the change on the last units at Kusile.

• It is most likely that output of the boilers will have to be de-rated, as the furnaces are undersized, and the existing coal mills are incapable of achieving the required fineness at their full output rate.

What is the root cause for this failure in engineering? The designers deviated from the specifications that local engineers specified.

All of this would not have occurred had the boiler contractor utilised the boiler plant specification which we at Eskom helped evolve over many years, and applied successfully at the Lethabo, Kendal and Matimba power stations. These stations achieved availabilities of over 80% for several years.

The cost of this error is that Kusile and Medupi will have to be downrated from 800MW to 720MW and that the mills will have to be upgraded with more load shedding ahead.

South Africa has 1 stage of additional load shedding because of a bad design.

Why didn’t Eskom listen to experts like Mr. Ham?