Causality Analysis of the Green Ammonia Synthesis Process Using the Convergent Cross Mapping Algorithm

Abstract

Ammonia synthesis has been gradually altered from the gray process (using fossil fuels as hydrogen sources) to the green process (directly or indirectly using water electrolytic cells as hydrogen sources powered by renewable energy), with the motivation of sustainable development and carbon neutrality. The fluctuating nature of renewable energy and the location mismatch between power plants and the production complex make gray ammonia production roadmaps likely to fail or embrace the change. Establishing knowledge graphs in the form of causal relationship network diagrams will help enterprise decision-makers and engineers better understand the process and generate correct production operations and scheduling. In this study, a chemical engineeringinformed method is introduced to generate causal networks of multiple load conditions for green ammonia production. Expert knowledge of chemical engineering is embedded in the determination of the existence and corresponding time delay of the causal relationships of variable pairs. In an industrial case study, the skeletal knowledge graphs and evolution of the control mechanisms were identified in a comparison of the derived causal networks. Extended applications are expected with further integration of controlling theory and algorithms.

Time-Series Forecasting of Large-Scale Green Ammonia Production: An Intelligent Dynamic Modeling Case in the Inner Mongolia Area of China

“…The data sets generated by the UniSim software encompass 279 features, which include streams, controls, operations, and other variables pertinent to the green ammonia synthesis system. The model inputs are derived from these data sets, following a process of simple sieving features that do not exhibit time variance and selection based on domain knowledge 62 about ammonia synthesis. The input variables, accompanied by detailed explanations, are provided in the Supporting Information S8.…”

“…In our previous study, 62 we obtained causal relationships during load fluctuations of key control variables and controlled variables. These relationships facilitate the flexible production and control optimization of green ammonia synthesis.…”

Time-Series Forecasting of Large-Scale Green Ammonia Production: An Intelligent Dynamic Modeling Case in the Inner Mongolia Area of China

“…The import variables of Unisim dynamic model and their descriptions are listed in Table S10, which chosen by chemical knowledge about ammonia synthesis 6 .…”