Community Research and Development Information Service - CORDIS

H2020

HPC4E — Result In Brief

Project ID: 689772
Funded under: H2020-EU.2.1.1.
Country: Spain
Domain: Energy

Harnessing the power of exascale high performance computing to maximise energy efficiency

Energy demands are rising and the hunt is on for effective alternatives to fossil fuel sources – but new computational approaches are needed to understand their potential.
Harnessing the power of exascale high performance computing to maximise energy efficiency
Energy needs worldwide will increase yearly until 2020 and far beyond. A report by the International Energy Agency (IEA) estimates that the global energy demand is set to grow by 37 % by 2040. Intensive numerical simulations and prototyping are needed to weigh up the real value of new energy sources and improve their throughput. The potential benefit of exascale high performance computing (HPC) and data intensive algorithms in the energy industry is widely recognised.

The EU-funded HPC4E project applied the new exascale HPC techniques to energy industry simulations. “The project customised the techniques, going beyond what is currently the state-of-the-art, to run exascale HPC simulations for different energy sources,” explains project coordinator Dr José Cela.

“Our aim was to ensure the software used in energy industry is ready to exploit exascale computers efficiently. This implies modifications in the software both at the algorithm and the parallel programming level.”

Demand for novel computing solutions grows in line with the increase in data

The Oil and Gas (O&G) industry has been one of the most active buyers of HPC technology in recent decades. Seismic data processing is an extremely demanding task in terms of computational demand. Its essential task is transforming seismic data records (i.e. “sound” tracks recording the response of the Earth to external impulses) into maps of the subsurface.

“The ability to conduct 3D surveys has meant the amount of recorded data has exploded,” says Dr Cela. While the move from 2D to 3D brings benefits, such as reducing the uncertainty in exploration, it also generates far greater quantities of data that need processing. “High-frequency data results in a considerable overhead in terms of compute time, roughly increasing by a factor x16 each time we double such frequency.”

The HPC4E project has developed the 3D acoustic and elastic Full WaveFrom Inversion software to generate 3D maps of a terrain’s physical properties. This alleviates some of the uncertainty in the search of hydrocarbons.

“This software has been programmed to run efficiently in exascale computers, reducing the execution time of these problems to the time window acceptable for the industry. The HPC4E project has created geophysics software to be used in oil and gas exploration that sets the standard for the industry. This software enables companies to measure how efficient the codes run on exascale computers are and improve them if needed.”

The competitiveness of wind energy can also be improved with an accurate wind resource assessment, wind turbine and farm layout design and short-term micro-scale wind simulations to forecast daily power production.

The HPC4E project adopted a new approach to couple computational fluid dynamics models at wind farm level with meteorological data at regional level. Harnessing the power of high performance computing the project created ALYA, a software package that demonstrates perfect scalability for more than 100 000 processors. All these improvements produce a reduction of the error around 10 %.

“We have also worked on the needs of the biogas industry,” says Dr Cela. “By developing complete model for the combustion of biogas, we’ve established the conditions necessary to guarantee safe combustion,” he adds. The project tabulated the chemistry of biogas composed of different elements and applied their findings to the simulation of industrial combustors.

“The use of super computers is highly efficient when running software to simulate different combustion phenomena,” Dr Cela explains.

Keywords

HPC4E, energy, computing, oil and gas, biomass, wind energy, turbines, supercomputing, computational fluid dynamics, exascale high performance computing
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top