The cash-strapped Eskom needs to find ways of generating electricity more economically if it wants to survive. The brilliant dissertation of a master’s student of the NWU not only suggests a solution to this problem, but investigates ways to optimise the use of natural resources for power generation.

 

Jean-Pierre van Niekerk’s dissertation earned him the coveted Theodor Stewart Medal in the master’s category of the students’ competition at the annual conference of the Operations Research Society of South Africa (ORSSA) in Cape Town in September.

 

Burning the midnight oil pays off

 

For this year’s competition, Jean-Pierre’s supervisor, Prof Fanie Terblanche of the School of Industrial Engineering, nominated his dissertation, which Jean-Pierre completed in 2018.

 

“The medal is a reward for many hours of hard work and it means a lot for my career,” says Jean-Pierre, who works as a data scientist for Standard Bank. “I believe it is only by the grace of God that I was able to successfully complete my studies.”

 

Jean-Pierre is one of the first master’s students to have graduated from the NWU’s School of Industrial Engineering — a feat he shares with his twin brother, Thomas.

 

The answer lies in mathematical modelling

 

Jean Pierre’s dissertation on ways to minimise the cost of power generation considered the multitude of operational aspects that power utilities have to take into account. These include ageing infrastructure, stringent emissions legislation and operational limitations.

 

“Apart from the financial objective, the optimisation problem is also concerned with meeting the forecasted load demand to prevent grid instabilities.”

 

Jean-Pierre says given the magnitude and complexity of the problem, his dissertation proposes a mixed integer linear programming formulation to solve the non-linear Unit Commitment and Environmental Economic Load Dispatch (UCEELD) challenge that power utilities face.

 

“I found that the proposed mathematical model and solution approach have the ability to solve a realistically sized problem within a reasonable time,” he says.

 

In his dissertation, Jean-Pierre included other important aspects of the power generation problem, such as thermal generation, water consumption, outage constraints and demand uncertainty.