Case Studies & Lessons Learned

Session Chair:

• Armin Fricke, Chemstations Europe

Confirmed Talks:

Abstracts:

Machine Learning Meets Production

Nicole Büttner, Merantix

Research on machine learning topics is rapidly advancing in breadth and depth promising break-through changes on how companies operate. How to translate these cutting-edge technologies into real business value is only in its infancy. MX Labs brings cutting-edge academic research to life in applied machine learning projects across industries like healthcare, logistics, automotive and chemicals. As an example, for clients like Volkswagen, Bosch and TÜV, we have created powerful solutions for quality control in production processes, image recognition systems for autonomous driving as well as automated inventory.

This session will focus on:

• introducing examples of state-of-the-art machine learning technologies
• giving insights in how to leverage machine learning for generating real business value
• showcasing some current relevant industry applications

Using Power-to-Ammonia to Transform Fluctuating Wind Power into Base-Load

Marc Hölling, Hamburg University of Applied Sciences

The energy revolution in Europe is focused on the installation of renewable energies, which are mainly wind power and solar power. Due to their strong dependence on wind and sun, their production is highly fluctuating and not matching the consumption profiles of industry and households. With the on-going shut-down of base-load power plants, like e.g. coal fired power plants, the renewable energies have to be transformed into base-load by means of energy storages. In this case study, a 3 MW wind turbine is combined with a Power-to-Ammonia process as an energy storage. In times of high electricity production, hydrogen is produced by an electrolyser. Due to the low volumetric energy density of hydrogen, it is not directly used as the energy storage. Instead, it is mixed with nitrogen to react to ammonia in a modified of Haber-Bosch-process. The reaction parameters for this exothermic reaction are chosen to be about 155 bar and 420 °C. Ammonia is separated by condensation at the reactor outlet and, since it can be stored as a liquid, it shows a huge advantage in terms of energy density (approx. 3150 kWh/m³). In times of low electricity production, the energy storage is discharged, i.e. ammonia is fed to the same reactor but at a low pressure and a higher temperature. Thus, ammonia is split up into nitrogen and hydrogen, which can be used in a fuel cell. The heat for the reverse reaction is delivered by the combustion of a part of the hydrogen. The overall efficiency of this Power-to-Ammonia process is found to be about 30%. The results are compared to other storage technologies in terms of efficiencies, storage size, investment costs and finally electricity generation costs. For a long-term energy storage, which is capable of transforming a wind turbine into base-load, the Power-to-Ammonia process is looking very promising.

Status of and Perspectives for Power to Products

Alexander van der Made, Senior Principal Advisor, New Energy Technologies

Increasingly, our energy system will become electrified because costs of renewable energies like solar PV and wind continue to decrease. However, whereas the traditional fossil fuels like coal, oil, and natural gas offer 24/7 availability, wind and solar energy are only available intermittently – with fluctuations varying from seconds to seasons. I will explore the use of cheap renewable electricity to create carbon-based fuels and chemicals with a zero-carbon footprint fro water and CO2. Technically such “power to products” (PTP) or “liquid sunshine” schemes are feasible, yet economically they are far from being attractive. Also, often such schemes cannot lay claim to having truly a zero-carbon footprint. I will discuss what technical breakthroughs are required to make PTP schemes being zero-carbon and economically attractive.