Dataset

This repository contains CO2 measurements and contaminant models. The CO2 measurements are the experiments conducted in a train carriage of the British Rail Class 802. These experiments were conducted with six passengers seated near the centre or middle of the carriage. The CO2 of these passengers was measured over time by multiple sensors located along the length of the carriage. The location of the CO2 sensors was changed depending on the passenger seating locations. During the start of an experimental run, the ventilation was turned off, in order to build up CO2 concentrations. Afterwards, the ventilation was turned back on, resulting in a strong decrease in the CO2 concentrations. We used these measurements to compare with the 1D-model and to estimate the turbulent diffusion coefficient and the ventilation rate. The CO2 dataset (11082020_CO2_H802.xlsx) shows the CO2 concentration (in ppm) as a function of the time. In addition, an overview is included of when the experimental runs started and ended (including when the ventilation was turned on and off). The fully-mixed and 1D-model are included in this repository as MATLAB scripts. The main script is named 'carriage_models.m' and the required model functions can be found in the folder 'model_functions'. The main script provides an overview of the required parameters. In addition, the script calculates the remaining parameters (e.g. flow rates). The 1D-model is solved numerically by iterating in time while the fully-mixed model is solved analytically. Note that this script only illustrates the working of the contaminant models. The infection risks can be estimated from these models by using a source of infectious quanta as the contaminant source and by calculating the infection risk based on the exposure using the modified Wells-Riley equation. This project was partially funded by the TRACK EPSRC project (reference EP/V032658/1).

Model is in the form of MATLAB files, data is in an excel spreadsheet. Please cite the associated paper if using data or models.