Influence of air velocity on indoor environment quality in unidirectional flow operating theatres: A study based on Computational Fluid Dynamics

. It is necessary to characterise air-conditioning airflow in omanuscriprder to optimize hospital Indoor Environment Quality in high-performance operating theatres, and also reduce the risk of nosocomial infection due to pathogen contamination. The aim of this article is to study the prevalence of optimal healthy conditions from controlled air flow quality in hospital facilities, and to minimize energy consumption. To this purpose, the indoor air movement was modelled by Computational Fluid Dynamics technology. The optimal results showed that it is necessary to drive ultra-clean air ranging between 0.25 m/s and 0.40 m/s, values which are adequate to perform efficient sweeping and cleaning of the air near the patient, maintaining unidirectional air flow permanently as the air passes through the surgical field. These speeds must be taken into account as calculation parameters in new hospital facility projects, and as control parameters for the existing operating theatres.


Introduction
Healthcare-associated infections are among the most common and harmful causes of unintended harm to health care. European Commission studies confirm that it is possible to reduce these adverse effects by 20% per year, as infection control methods are not optimized. Approximately, 37,000 deaths annually are caused by nosocomial infections. That is why the European Parliament is developing a Legislative Project against hospital contagion [1].
In Spain, prevalence of patients with nosocomial infection, contracted during his or her stay in hospital, was 7.74% in 2017, about 25% of which was caused after underwent a surgical intervention [2].
Air-conditioning of healthcare buildings is one of the most complex installations that Hospital Engineering has to deal with [3]. Hospital areas are very different each other, both technical requirements of facilities and type of services they must provide, specially respect of Indoor Environment Quality (IEQ) and thermal comfort. Furthermore, characteristics of supply air and its movement along the room allow cleaning of airborne pathogens. Air-conditioning systems are therefore an effective way to control infection in hospitals caused by airborne bioparticles.
IEQ has great importance in operating theatres, as there are sources of indoor infection that can contaminate the room and so reduce IEQ level: medical personnel, patients, equipment, instruments, etc. Moreover, movement of personnel and instruments causes spread of airborne pathogens [4,5]. Air-conditioning is the key tool for achieving adequate IEQ, which must keep the number of airborne bioparticles present below reference values as they are the vehicle for spreading disease [6].
Intensity of infection [7] is proportional to concentration of infectious airborne bioparticles and the time of exposure to them. It is therefore crucial that number of air changes per hour is sufficient to ensure a low concentration of bioparticles in the environment and to minimize air age in the room.
García Sanz-Calcedo and Monzón González (2014) analysed the economic impact of applying measures to guarantee environmental biosafety on health building construction projects. They checked that measures minimized impact on workers' and patients' health in refurbishing of health centres in Extremadura (Spain) [8].
Operating theatres are part of hospital's clean rooms. Heating, ventilating and air-conditioning (HVAC) system for operating rooms should protect the room against crosscontamination [9]. Overpressure conditions provided to the room will prevent it. An overpressure value of 20 Pa will be enough, allowing Ultra Clean Ventilation (UCV) system to work optimally avoiding the entry of pathogens from adjacent areas. Lydon et al. (2014) observed the effect of opening and closing doors in two pressure scenarios (0 Pa and 20 Pa) with respect to adjacent spaces [10]. The results concluded that the UCV system operates efficiently with positive pressure (20 Pa) but fails when there is no pressure difference between the operating room and surrounding areas.
Tinker and Roberts (1998) applied Computational Fluid Dynamics (CFD) technology to characterize the airconditioning system of a hospital operating room in the United Kingdom [11]. These authors concluded that a thermal plume is generated by thermal radiation from surgical lights. This is only noticeable when airflow passes through them below the critical velocity value. Critical velocity value will be 0.10 m/s, since with a velocity greater than 0.3 m/s, the formation of such natural convection currents is not detected.
Ventilation of an operating theatre is performed by imposing a number of air changes per hour or, what is the same, a fresh from outside air flow per volume unit. However, this parameter does not ensure that ventilation will achieve indoor air quality throughout the room, as design of the ventilation system may result in by-pass between supply and exhaust, backwater areas, etc.
In addition, number of real air renovations executed should be used to evaluate performance of a ventilation system. This parameter quantifies the air residence time at all points of the room to be homogeneous [12]. Equalizing number of renovations carried out to those planned and reducing average air age in the room improves the efficiency of the HVAC system. Therefore, ventilation will be adequate when it is possible to move the driven air evenly throughout the room, avoiding problems of bypass or backwash. Moreover, minimizing air fresh flow to promote energy efficiency. In order to achieve adequate ventilation with minimum fresh air flow it is necessary to achieve maximum ventilation efficiency.
Quantifying the efficiency of air conditioning is no easy process as it depends on the flow pattern, room geometry, layout of the drive and extraction, etc. This is why the CFD is used. Analysis using CFD technology allows an in-depth study of the complex airflow generated in operating theatres. That is, their movement pattern and fluid property distributions (air age, temperature, pollutant dispersion, etc.). They are thus a virtual trainer for the optimisation of ventilation systems [13].
The aim of this study is to carry out an analysis of IEQ achieved in a high-performance operating theatre for different air supply velocities from central air diffuser. Results will make it possible to calculate an appropriate velocity air supply to design new HVAC installations in hospital operating rooms, valid also as a control parameter of existing.

Methodology
CFD technology was employed to calculate air velocity profile in two interesting areas: at patient's level and over surgeon's head.
Range of simulated air velocities is found in reference literature as recommended for this application [4]. UNE 100713:2005 also defines a recommended range of values. An operating theatre with typical dimensions and characteristics of this type of theatre was modelled. The arrangement of extraction grilles is also the usual one in operating theatres of this type, since the same extraction duct is used for the grilles in the lower part of the wall and for the grilles in the upper part.
Firstly, control volume of the system, i.e. the computational domain, was defined. Geometry of the room has the typical dimensions of European operating theatres. The modelled geometry (Fig. 1) presents six boundary planes, whose names are: floor, ceiling and wall i (i = 1-4), were defined with the condition "wall" and applied the physical characteristics to be affected by the  Spatial discretization of the control volume, i.e. the meshing of the geometry, was established. The meshing generates a series of small volumes called cells, inside of which they house the nodes, on which the calculations will be made. Distribution of cells, their shape and size are determined from the mesh consistency study described in the results section. Five cases were raised that present enormous differences in the characteristics of the mesh. Table 1 shows these characteristics. It is checked that the values of the results are similar and therefore the consistency of the mesh is validated. The number of cells and the characteristics of Case 03 was used to carry out our studies. This is because it provides a consistent fluid field solution with the lowest possible computational cost.
IEQ was assessed in terms of air velocity profiles generated at the head height of medical personnel present and over the patient. For this purpose, a stationary calculation scheme has been proposed which will provide stabilized value for an infinite time of the velocity magnitude.
In the computational domain, conservation laws are established as equations of fluid field government. The equation of conservation of mass applied to steady state study is given by Equation (1): The equation of conservation of momentum is given in Equation (2). For its definition, it has been taken into account variation of the fluid density according to the approximation proposed by Boussinesq (term β). According to the Boussinesq Model, the consequences of this variation in density are simplified and will only affect the conservation equation of momentum.
The energy conservation equation taking into account that there are no internal sources of heat generation is given by Equation (3): The Reynolds-Averaged Navier-Stokes (RANS) approach is used to resolve the fluid field. In the case of a stationary study, continuity is verified through Equation (4), since the partial time derivative is simplified.
Considering the viscous and incomprehensible fluid, the momentum equations can be written in their compact form as in Equation (5). They can be written in Cartesian tensor form as follows: where ρ is the density and µ is the viscosity of the fluid, t is the time, xi will represent each axis of the coordinates system, ui are the components of the velocity vector and Si is the source term. Standard k-ε model was employed to model turbulence in fluid field and Boussinesq hypothesis was used. This hypothesis uses Reynolds stress tensor with the mean velocity gradients indicated in Equation (6): where µt represents turbulent viscosity, k is the turbulent kinetic energy and δij is Kronecker Delta. This implies solving two more equations. It concerns turbulent kinetic energy equation, k, and turbulent kinetic ratio equation, ε.
The method of two standard equations k-ε [14] was chosen as turbulent closing model. As a complement, it has been combined with the Standard Wall Functions which implement the Universal Wall Law. These are suitable when the fluid has negligible pressure gradients, as occurred evaluating HVAC systems applied to operating rooms.
An air velocity supplied through the large central diffuser were defined as boundary condition with negative Y-direction. Drive velocity magnitude varied from 0.10 m/s to 0.40 m/s, at 0.05 m/s intervals. Thus, there was seven case studies. Each of the velocity magnitude values will generate a different flow rate value. A forced extraction will then be applied to each of the 10 grids in the room. The value of the extract flow rate is calculated as a proportion of the total impulse flow rate. The values for the impulsion and extraction conditions of each study are shown in Table 2. Finally, the apertures of the doors are defined as free outlets at atmospheric pressure and the rest of the surfaces as "wall" (ux = uy = uz = 0). The fluid characteristics are considered constant except for the density, which is considered a variation according to the Boussinesq approximation, to take into account the buoyancy term in the momentum equation. For each simulation, the flow governing equations are solved, considering the initial and boundary conditions imposed, using the algorithm proposed by the Finite Volume Method. The ANSYS Fluent software configured with Pressure-based solution technology is used to execute it. The Coupled method is employed to solve the finite system of algebraic equations. The way it solves velocity and diffusion gradients will approximate by the Least Square Cell Based order which is suitable for unstructured meshes. The pressure is solved by the Pressure Staggering Option (PRESTO). Finally, a second upstream approximation, 2nd Order Upwind, is applied to solve the momentum equation, the turbulent kinetic energy equation and the turbulent kinetic energy dissipation rate equation.

Results
This section shows the results for mesh consistency study and for air velocity profiles study over medical personnel and patients present in the room.
For mesh consistency study, five cases with different mesh characteristics were solved. Number of elements, density and shape and size of the cells were varied for mesh consistency study. Carrying out simulation were obtained results shown in Table 3 for each case studies. These results are represented graphically in Fig. 2. A high correlation is shown between the values resulting from the proposed studies. For different values of the number of cells, a similar result is obtained for the variable of interest: the pressurization of the room. This means that the mesh provides a consistent solution.   3 shows the air velocity profile generated by each of the different study supply velocities. This profile is established on a horizontal line located at a height of 1.75 m and which crosses the room through the median plane in the direction of the Z-axis. Represents the air velocity profile to which the medical personnel present in the operating room will be exposed. Fig. 4 shows the air velocity profile generated by each of the air velocity magnitudes proposed for the study. In this case, they are those established on the horizontal line situated at a height of 1.20 m and which crosses the room through the median plane in the direction of the X-axis. These are the different profiles to which the patient will be exposed according to the velocity of impulsion chosen in the diffuser.

Discussion
Results show that CFD technology is a suitable tool for designing operating theatres, as it ensures proper ventilation quality and minimizes the risk of nosocomial infection by optimizing energy and environmental costs. Air must be driven at a speed that favours unidirectionality to facilitate high levels of cleanliness and low levels of turbulence. These characteristics minimize the presence of pathogens in the surgical field. A diffuser surface large enough to cover the entire surgical field is also necessary.
To ensure unidirectionality, the key parameter is the ultra-clean supply air velocity. Factors that can interrupt the unidirectionality of the flow are: low supply air velocity and thermal contour generated around equipment or people due to thermal radiation. These thermal contour will induce a flow of air called natural convection.
In healthcare applications, medical requirements take precedence over the comfort of people in terms of the drive air velocity [15]. In this way, a high drive air velocity can be annoying for comfort of people, but is necessary for cleaning the surgical field.
It was also observed that driving ultra-clean air with a velocity less than 0.10 m/s will cause it to reach the head of medical personnel with values below the critical value of 0.10 m/s. There are also additional values of drive