The effect of the contaminant emission rate on the velocity field and contaminant distribution with the presence of an obstacle in a large space

. In the industrial field, the prediction of the contaminant gas distribution is very meaningful. However, when the leakage is high, not only the contaminant distribution will not follow the pattern of the original flowfield, but the contaminant buoyancy or negative buoyancy will affect the flowfield conversely. In this study, we focus on the effect of the contaminant emission rate on the velocity field and contaminant distribution with an obstacle in a large space by means of CFD simulation. Two leaking positions and five emission rates of the source have been taken into consideration. When the emission rate is high enough, the flowfield structure will be altered and new vortexes will appear. The contaminant dimensionless concentration distribution is totally different from the low-emission-rate conditions. The flammable region becomes significant, which leads to the potential risk of explosion.


Introduction
It is unrealistic to assume that the pollutant concentration is well-mixed due to the dimensions for large spaces. The significant spatial variations in contaminant concentration have been investigated in many studies in manufacturing plants and other laboratory experiments [1][2][3][4][5]. However, the effect of gas buoyancy is rarely taken into account, which will influence the flowfield and the concentration field.
Obstacles usually exist in realistic plants. The windward and leeward sides of the obstacles are usually featured with small velocities and vortexes due to the surrounding flow patterns [6][7]. When the leakages occur in such areas, it's easy to cause accumulation. Many studies have been done concerned with the concentration distribution with the existence of obstacles in the atmospheric field [6,[8][9] Buoyancy strength is related to the contaminant emission rate and the environmental velocity field [10]. When the emission rate is small with regard to the nearby velocity, the contaminant gas can be regarded as passive and transport with the airflow. When the emission rate is high to some extent, the effect of the density starts to appear.
In this study, we simulate the contaminant gas leaked at two sides of the obstacle with sulfur hexafluoride(SF6) as the contaminant gas in the same ventilation mode by CFD. The effect of the buoyancy on the flow field and concentration field are discussed. The critical range in which the gas is taking a noticeable effect on the flowfield and the concentration field.

Domain, computational grid, and boundary conditions
We investigate a 5m(X)*6m(Y)*6m(Z) space with a 1m*1m*1m cubic obstacle at the center of the floor. The air is supplied from the middle of the 5m-wide wall and exhausted from the bottom of the opposite wall. The air change rate of the space is 3h -1 . The air jet first impinges the opposite wall and then spreads and recirculates.
We designate the space between the obstacle and the wall of the outlet as A side and the other side as B side. As for the contaminant leakage location, one is 0.5m from the obstacle on A side and 0.5m above the floor, denoted by A(0, 4m, 0.5m), the other is 0.5m from the obstacle on B side and 0.5m above the floor, denoted by B(0, 2m, 0.5m).
We assume the contaminant source is a point source.
We built half the space, a 2.5m*6m*6m space, in ANSYS Geometry Modeler, as shown in Fig.1.

Computational setup
The commercial software ANSYS FLUENT is used for the steady RANS computations based on a control volume approach for solving flow and mass fraction equations. The Green-Gauss cellbased scheme is used for gradient discretization.
The advection terms are discretized using a second-order upwind scheme. The semi-implicit method for the pressure-linked equation (SIMPLE) algorithm is used for the pressurevelocity coupling.
The spatial distribution of airflow, temperature, and species in the zone is governed by the conservation laws of mass, momentum, and energy. The governing advection-diffusion equations of the fluid are all in the form of [11]: Where is the density of the fluid, ∅ is the scalar under discussion, is time, ⃑ ⃑ is the velocity vector, Γ ∅ is the diffusion coefficient, Where u, v, w are respectively the velocities in X, Y, and Z direction; T is the temperature; Cp is the specific heat; Yi is the species mass fraction; is the kinetic viscosity; is the temperature conduction coefficient; d is the mass transfer coefficient; P is the pressure; g is the gravitational acceleration; is the density; Q is the energy generated by the source; Qm is the contaminant strength generated by the source.
The local mass fraction of the species, , is calculated by solving a convection-diffusion equation for the species with the parameter in the last row of Table 1.
Since the study here is focused on the effect of the negative buoyancy on the flow field and the concentration field, we only use the standard kmodel as the turbulence model and standard wall function as the near wall treatment. The y+ values are kept in the range of 30-300.

Configurations
We use two contaminant source locations in this study as mentioned before. For each source location, there are five emission rates. All the configurations are listed in Table 2. (3) Where C is the concentration of SF6, kg/m 3 ; C0 is contaminant concentration when the room is fully mixed, kg/m 3 ; q is contaminant emission rate, kg/s; QV is room ventilation rate, m 3 /s.
We also need a criterion to judge the negative buoyancy strength, which is: Where θ is the dimensionless number to discriminate passive gas originally [10]; g is the                 In B5 condition, the shape of iso-surface for C * =2

Contaminant distribution
is similar to the vortex on the z=0.2m plane. The flammable region is turning noticeable from B4 to B5 ((from θ=7.331 to θ=9.950).
Comparing Fig.11 and Fig.12  With the increase of θ, the center of gravity for the zone of C * >2 becomes lower, the flammable region enlarges. With the increase of θ, the flammable region becomes noticeable, which means that the space is exposed to potential risk.