Residence time distributions and computational fluid dynamics to characterize dilution ventilation in a confined space model.

Abstract

Residence Time Distributions (RTDs) can be useful in evaluating air mixing for dilution ventilation in confined spaces (CSs). Computational Fluid Dynamics (CFD) can be used for simulation and visualization of air mixing and contaminant dispersion. The development of sophisticated computational methods and the commercial availability of user-friendly CFD software, which can be used on desktop computer workstations, has made CFD a much more viable tool for industrial hygiene (IH) applications today than it was a few years ago. Ventilation experiments were conducted using a tracer gas released in a cubical CS model with a single opening at the top. The experimental ventilation variables were ventilation mode (supply/exhaust), airflow rate, and ventilation duct height. Eight test cases were studied by injecting tracer gas as a pulse into the CS model, and deriving RTDs from exit stream tracer concentration versus time data. Two air mixing parameters based on RTD--Mixing Deficiency (MD) and Relative Mixing Time (RMT)--were defined and computed from physical experiments and CFD simulations. CFD simulations were performed using two commercially available CFD codes, in three steps: (1) steady-state solution of the flow field, (2) transient injection of tracer gas, and (3) transient decay of tracer gas. Comparisons were made between experimental and simulated RTDs, and between experimental and simulated air mixing parameters (MD and RMT). Supply ventilation was generally more effective than exhaust ventilation for contaminant (tracer) dilution, required less ventilation time, and produced good dilution regardless of ventilation duct height. Exhaust ventilation at low flowrate and low duct height produced good air mixing; poor mixing occurred with other combinations of flowrate and duct height. CFD simulations were more reliable in predicting RTDs and air mixing parameters for supply ventilation test cases than for exhaust ventilation test cases. Qualitative descriptions of steady-state flow fields from CFD simulations of exhaust ventilation test cases indicated poor air mixing, consistent with experimental observations. This study showed that commercially available CFD software can be used to simulate ventilation airflow and contaminant dilution in CSs. RTD-based air mixing parameters were used to describe air mixing in CSs, and RTDs were used to compare physical experiments and CFD simulations.