Analogies between SARS-CoV-2 infection dynamics and batch chemical reactor behavior. (14th December 2020)

• Record Type: Journal Article
• Authors: Manenti, F.; Galeazzi, A.; Bisotti, F.; Prifti, K.; Dell'Angelo, A.; Di Pretoro, A.; Ariatti, C.
• Is Part Of: Chemical engineering science. Volume 227(2020)
• Publication Date: 2020-12-14
• Subjects: SARS-CoV-2 -- Infection dynamics -- Batch chemical reactor -- Predictive model -- Non-linear regression -- Pandemic mathematical model; Chemical engineering -- Periodicals; Génie chimique -- Périodiques; Chemical engineering; Periodicals; Electronic journals; 660
• DOI: 10.1016/j.ces.2020.115918
• Languages: English
• ISSNs: 0009-2509

Abstract

Highlights: The batch reactor dynamics is used to predict SARS-CoV-2 spreading. The reactor model can phenomenologically explain the virus spreading. The model predicts the peak (day and entity) and the infection extinction. Initial Value Problem for ODE has been solved with an unknown initial condition. Algorithm robustness and convergence have been tested. Abstract: The pandemic infection of SARS-CoV-2 presents analogies with the behavior of chemical reactors. Susceptible population (A), active infected population (B), recovered cases (C) and deaths (D) can be assumed to be molecules of chemical compounds and their dynamics seem well aligned with those of composition and conversions in chemical syntheses. Thanks to these analogies, it is possible to generate pandemic predictive models based on chemical and physical considerations and regress their kinetic parameters, either globally or locally, to predict the peak time, entity and end of the infection with certain reliability. These predictions can strongly support the emergency plans decision making process. The model predictions have been validated with data from Chinese provinces that already underwent complete infection dynamics. For all the other countries, the evolution is re-regressed and re-predicted every day, updating a pandemic prediction database on Politecnico di Milano's webpage based on the real-time available data.