If I have done anything in my research life that I can say has directly benefitted the whole planet, it would be the work I did as a research engineer at KU Leuven in Belgium. I came up with a method to optimise the layout of a windfarm without the need for a super-computer. In other words, I introduced a method that allows engineers who want to design a windfarm to locate the wind turbines in such a way that can significantly increase the energy extracted from the windfarm, without incurring any extra costs. Also, my simulations showed that the rectangular grid which is a typical layout in many older windfarma is the worst possible layout. The following image can give you an idea why.
To know the context, look at the above video. It is a visualisation of a typical simulation I used to perform. To simulate the flow of wind in a windfarm, one needs t to solve Navier-Stokes equations. I used Large Eddy Simulations (LES) method to approximate the solution of the Navier-Stokes equations and for a typical simulation of a wind farm with 1km x 3km x 4km dimensions for an hour of implementation time, I needed to use 160 cores of the Flemish Super-Computer for a few hours. If we want to optimise the layout of a windfarm, we need to run such a simulation many times. Which makes it computationally intractable. What I did, was to show that using an analytical model, if it is calibrated in a certain way, we can approximate the output of the LES simulations with an accuracy of more than 90%.
If you are interested to know more about the technical details, you can look at this open-access paper we published in Wind Energy Systems, a journal published by the European Academy of Wind Energy.