Validation of CFD Simulations of Fume Hood Performance In Laboratory Environments

Daniel Jun Chung Hii, School of Design & Environment, National University of Singapore
Yueyang He, Building Department, National University of Singapore

Fume hood is regarded as one of the most complex structure that affects the flow field in laboratories. According to literatures, the effectiveness of fume hood in terms of removing pollutants is affected by many factors, including room air change rates, positions of diffusers, positions of pollutants, numbers of occupancies, etc. Although the fume hood behaviors are complicated, accurate prediction of their performance is critical for the sake of minimizing energy consumptions and simultaneously maintaining the acceptable exposure risk. For the case of National University of Singapore (NUS), the newly-built chemical laboratories are commonly equipped with up to 8 fume hoods at a typical space of just around 120 m2. Due to this compact laboratory design, the flow field is very much impacted by the fume hoods than those observed in the conventional laboratory environments. To evaluate these flow behaviors and explore the optimal ventilation strategies for the university laboratories, the computational fluid dynamics (CFD) simulation tool is chosen considering its powerful visualization and flexible prediction techniques.

This study presents a validation of CFD simulations for solving the flow behaviors near fume hood in laboratories. An open source CFD simulation tool, OpenFOAM version 1712, is used in this validation running on the buoyantSimpleFoam solver. As mentioned above, fume hoods have the most complex structure in university laboratories, and hence a close depiction modeling of them is required to ensure the simulations of contaminants and ventilation flows are accurate later. In order to capture the real geometric features, especially the tiny holes and gaps of the porous baffle and foil, that affect the flow directly, on-site measurements were done on a 6' Labconco Xstream fume hood, in the campus laboratory. The basic benchmark test is to compare the simulated data with two series of measured data in ASHRAE 110-2016 face velocity tests, which were prepared by AccuTec and Labconco, respectively.

The initial simulation results show a maximum margin of error of 0.02 m/s from the measured data. Once the confidence of capturing the flow of the fume hood can be achieved, the next step is to replicate this simulation method for an entire laboratory, consists up to 8 fume hoods. At the laboratory scale, the real geometries of the tiny holes and gaps that affects the flow field needs to be simplified in order to reduce the computational cost. This will be done using porous media cutting down the computational cell count as well as the simulation time. The limitation of this study is the lack of actual room configurations with exhaust and supply locations information used in both ASHRAE 110 tests.

Learning Objectives

  • how to capture the flow of fume hoods accurately in CFD simulations;
  • how to model a fume hood for CFD simulations;
  • how to measure wind speed at the fume hood; and
  • how to compare the measured and simulated wind speed data of the fume hood.


Daniel Hii is a graduate of School of Design & Environment, National University of Singapore, with the research interest in harnessing environmental simulation and 3D visualization to solve issues related to the urban and built environment. These pertinent issues involved heat and air pollutants and thermal comfort. All these issues become increasingly critical as we face a future with more ageing population and increasingly hotter climate. He also did extensive indoor and outdoor site measureme

Yueyang He is a doctoral researcher at Department of Building, National University of Singapore (NUS). His research interests cover both indoor and outdoor computational fluid dynamics (CFD) simulations. Currently, he is involved in a research project, which attempts to explore the optimum lab ventilation rate for energy efficiency and chemical contaminant exposure risk in NUS.


Note: I2SL did not edit or revise abstract or biography text. Abstracts and biographies are displayed as submitted by the author(s).