Sensor-based gas tracking: Data acquisition, linearization, and successive source tracking and separation

摘要

The propagation of information encoded in the fluctuations of chemical species of natural gas transported within natural gas distribution grids allows for a new gas tracking method. In contrast to state of the art gas and calorific value tracking based on numerical gas-flow simulations, the presented method is solely based on signal processing techniques. Due to a wider diversification of natural gas sources, gas and therefore calorific value tracking is of great interest for gas grid operators to provide fair invoicing of gas customers. The calorific value of substitute natural gas, e.g. produced by biogas or power-to-gas plants, injected concurrently into natural gas grids, typically deviates significantly over time in comparison to the prevailing natural gas in the grid. Sampling gas features like the chemical species contained in injected gases, or the standard density, by means of calibrated sensors, e.g. by process gas chromatography or infrared sensors, provides time-dependent signals that can be taken for calorific value extrapolation from source nodes to downstream nodes. To that end, we provide a general estimation of gas feature related transmission characteristics of natural gas distribution grids based on a straightforward one-dimensional diffusion model. Additionally, we present an accurate technique to estimate transit times and source fractions on gas grid exit nodes. We show with a field experiment, that gas tracking based on gas sensor signals is feasible and that our new method improves on state of the art gas tracking software based on computational fluid dynamics, already in use for gas customer invoicing.