Experimental Study on the Transport and Distribution of Carbon Monoxide in Indoor Spaces

Introduction
Increases in the world population and changes in the pattern of urban life together with the countless sources of air pollutants in semi-closed spaces have changed the air space of our cities into a hostile environment. Therefore, the problem of air pollution is not only limited to the outdoor environment but also became a real problem in indoor spaces. In modern cities, people spend 70-90 percent of their time in indoor spaces and their activities are limited more to these environments. Carbon monoxide is one of the most dangerous gaseous pollutants in the buildings with both residential and industrial operations that are frequently reported to cause diseases and death.
Carbon Monoxide (CO) is a colorless, odorless, and tasteless flammable gas that is slightly positively buoyant compare to air. Carbon monoxide is produced from the partial oxidation or burning of fossil fuel (coals, oil, and natural gas) or any other carbon-containing compounds. CO can build up indoor spaces and poison people and whoever (animals and plants) breathe it. Poisoning due to Carbon monoxide is a common type of fatal air poisoning reported around the world. The most common symptoms of CO poisoning are headache, dizziness, weakness, upset stomach, vomiting, chest pain, and confusion. Many researchers such as Sykes and Walker (2016) and Levy (2015) have investigated the health effects and the risk of CO poisoning on resident safety and peace. Various concentration of CO is found in fumes produced by burning fuel in workshop buildings and Industrial Units. It is ranged from 2 to 5 percent (20,000-50,000 ppm) and can cause long-term health problems to workers who breathe it. Various health organizations have established the Carbon Monoxide (CO) concentration limits. ASHRAE sets outdoor maximum levels at 35 ppm (1 hour averaging) and 9 ppm (8 hour averaging), while the World Health Organization limits CO concentrations based on exposure time ranging from 90 ppm (15-minute exposure) to 10 ppm (8-hour exposure). The ASHRAE indoor's maximum concentration limits for CO in industrial units is 50 ppm (8 hour averaging), which is usually hard to meet in place works and workshops. Harmful health effects of carbon monoxide make it essential to be able to predict the behavior of flow in indoor spaces for the successful implementation of any mitigation measures. Natural or forced ventilation is the simplest way to reach acceptable indoor air quality standards. Numerical and experimental models are common tools to predict the behavior of airflow in indoor and outdoor spaces. Experimental works also by providing a benchmark for the calibration of the numerical works and by developing the empirical equations for the prediction of flow patterns are commonly used to define airflow in indoor spaces. Modeling of airflow through the simulations of flow in the water tank has rarely been investigated. Chen et al, (2010) used a water tank and LIF technique to simulate contaminant distribution inside an airliner cabin using a one-tenth scale water model. The same approach has been followed in the current study. A small scale model of a workshop has made to simulated airflow distribution and transport. To correctly represent the flow, a small-scale model should be designed based on similarity analysis, in which the relevant dimensionless flow parameters are identical between small- and full-scale. The similarity of the forces is reached by the equality of Froude or Grashof and Prandtl numbers.

Materials and Methods
For the flow of carbon monoxide, regarding the density difference between CO plume and air the forces of inertia and buoyancy (density difference) are dominated. Due to the high turbulence of the flow, the viscous force is negligible compared to the two aforementioned forces. Therefore, the Froude number that consists of the important properties of the flow can be considered to set the similarities. On the other hand, regarding low velocities compare to the speed of sound (Mach number <1) the compressibility of flow can be ignored. So, the density difference and inertial forces are the only important forces and flow can be assumed incompressible carefully.
In this study, a simple regime of CO plume from a point source in a workshop building has been investigated. The building is a 3 mm perspex box that is 30 cm long, 30 cm wide and 35 cm tall with the sloping roof at both sides. The flow of contamination is a CO plume on the floor with a diameter of 1 cm. The discharge speed is 10 cm/s and the CO concentration is 20,000 ppm, so the Froude number is 1.44 and the Reynolds number is equal to 1124. A roof window, 1 cm wide and 10 cm long, is considered for the outflow.
The experimental works are performed at the Environmental Fluid Mechanic Laboratory of Babol Noshirvani University of Technology, Iran, using the Three-Dimensional Laser-Induced Fluorescence (3DLIF) system that especially developed for this purpose. The scaled experimental test facility placed upside-down fully submerged in the water tank. Conducting the experiment in a water tank besides index-matching is simpler to quantify flow mixing and dilution using the illumination of fluorescent.

Discussion of Results
In this study, the behavior of a plume of carbon monoxide in the indoor space of a stationary environment was investigated where no wind or forced ventilation implemented. CO is less dense than the ambient air and the flow moves upward due to its initial buoyancy. The flow reaches the steady-state condition after some seconds from the beginning. The concentration of contamination decreases due to flow entrainment and mixing while plume moves upward. The dimensions of buildings and outflow roof window determine the time requires to reach steady-state and it was 50 seconds for this experiment. So the experiment was recorded after this time when it became time-independent. This time would be exactly equal for the same size building in either water or air. However, the time is proportional to T_P/T_M =√(L_P/L_M ) between model and prototype. For the flow speed also the same relation i.e. V_P/V_M =√(L_P/(L_M )) , is established.
The mixing and dilution of the contamination are proportional to flow speed and its initial buoyancy. Due to the dominant forces, the flow goes upward to eventually reaching to the ceiling, then moves outward from the roof widow up to reach the steady-state condition. In the uniform steady flow of CO smoke, a stratification forms in the building in which the concentration gradually decreases from the floor (20,000 ppm) up to the ceiling (2,600).
As a result of this study, self-similarity was observed for the profile of concentration and plotted at different locations from the source ( ). The 2D configuration of flow, changes in flow width and centerline concentration are also plotted.

Conclusions
In this study, utilizing the LIF system, the spatial and temporal changes of CO concentration from a plume of contamination in a workshop have been investigated. Concentration variation along the centerline is plotted along with the changes in flow width and pollutant distribution in the building. For the aforementioned dimension with a point source at the floor and a roof window at the ceiling, it observed that the flow reaches the steady-state condition after about 50 seconds in which as stable stratification forms in the building. The CO concentration gradually increases from zero to maximum from the floor up to the ceiling. The pattern of changes depends to flow initial fluxes and the roof window's dimensions. The effect of walls on entrainment restriction and the ceiling on flow re-entrainment were observed and plotted as the self-similar profiles. The concentration of CO found in the range of danger at the height of human respiration in this building. It shows that natural ventilation can not decrease the high concentration of CO in this plume. Forced or mechanical ventilation is required.