**2. Materials and methods**

Definition and quantification of the contaminant sources under OT effective use conditions is a complex task. Existing databases refer to constant particle mass emission of a person as a function of his/her activity. Several studies analyze different sources and human activities causing high levels of indoor particle concentration, diffusion and re-suspension processes. Some authors [20] present studies on the influence of periodic bending movement of the surgeon on the air flow field and bacteria carrying particle distribution using a combined approach based on an Eulerian RANS model, modified drift-flux and moving mesh. In the study by Brohus et al. [17], bacteria attributed to the skin flake emission of staff and patient are simulated as a gaseous contaminant. The most widely used experimental methods allow particle size calculation for discrete intervals, so as to check whether microclimatic and IAQ conditions are in compliance with standards and specific rules. Some authors have developed a useful method for determining individual emission rates from ambient air contaminant concentrations caused by multiple indoor sources [21]. Other studies provide a sampling method integrated with statistical analysis. [22]. In a recent paper bacteria sources are inves‐ tigated [23] in a university classroom both occupied and vacant as concentration differences due to the two conditions. Besides the influence of the airflow pattern, it has been demonstrated that particle dynamic behaviour is strongly dependent on particle size and size-related forces [24]. Chih-Shan Li et al. [25] carried out a series of field tests by active and passive sampling of air and surface measuring bacterial and fungal concentrations without referring to particle size and diameter. Another paper [26] based on particle counting combined with statistical analysis, evaluates aerobic bacterial sedimentation and connected index of microbial air contamination. There, results refer to different particle diameters but not emission sources. Some authors [1] have carried out transient simulations using the Renormalization Group (RNG) k-ε turbulence model, assuming a released rate of bacteria-carrying airborne particu‐ lates from surgical staff of 100 BCP/min per person and for patients of 400 BCP/min per wound. In general, CFD simulation concerning aerosol particle transport and diffusion processes can be solved by three main different numerical approaches: the first method consists in solving the particle concentration field by using diffusion-transport equations based on "passive scalar transport" in which the vector of particle transport is the motion field of the fluid but gravi‐ tational effects and frictional forces on particle motion are disregarded; the second is the Eulerian-Eulerian method and the third is the Eulerian-Lagrangian one. These last two methods solve the airflow field based on the Eulerian approach, but the particle phases are treated differently: in the first, particle concentration is directly solved using gravitational effect in the transport term, in the second one an ordinary differential equation is solved for any particle path. A modified drift-flux model based on the second approach has been applied to modelling particle transport [27,28,20]. At present, there are not many studies in the literature concerning numerical modelling in which the contaminant concentration for the IAQ analysis has been performed starting from sources of particle emission with a distribution dependent on particle diameter [29,30]. However, Italian and international standards require, for environment classification, a particle count for diameter dimension. Actually, there is a lack of data concerning specific emission sources in terms of particles issued in time unit by people. A recent study provides the quantification of size-resolved particle concentrations in indoor air of a classroom under occupied and vacant conditions [29]. Another important

516 Current Air Quality Issues

#### **2.1. Experimental measurements and data analysis**

Experimental measurements were conducted in a real orthopaedic OT of the University Hospital of Parma (Italy) referring to [32,33] and during the different conditions as given: **at rest**, i.e. room with all services functioning and equipment installed and operable/operating, but without surgical/healthcare staff and the patient; **operational "correct use condition"** reproducing the effective use conditions of the OT, with all services and equipment function‐ ing, surgical and healthcare presence. Experimental data were collected after installation of all acquisition equipment in the room and during simulated hip surgery lasting 20-25 minutes and involving some PhD students. The surgical staff and healthcare assistants were in upright stationary positions and surrounding the patient lying on the operating table. The sliding door was closed; **operational "incorrect use condition"**, i.e. the same conditions described for operational "correct use", but considering the sliding door opening and closing and one healthcare assistant walking from outside through the door and up to the patient's head on the operating table, and then turning round and moving back to the corridor. The HVAC system was working during all tests, supplying a constant incoming airflow rate at controlled temperature. The measurements campaign was carried out over a two day period, from 8 a.m. to 2 p.m. Microclimate parameters were measured over time at discrete points PT01, PT02, PT03 that are shown in Figure 1. In particular: 1 hot wire anemometer, 1 air-temperature and humidity sensor and 1 globe thermometer were set below the central air inlet diffuser on the plane of the surgical table (PT01); 1 air temperature and humidity sensor and 1 differential pressure sensor near the sliding door on a tripod (PT02); 1 differential pressure sensor and 1 air temperature and humidity sensor were set near the air outlet diffuser on a utility pole (PT03). Differential air pressure, between monitored room and the adjoining ones, was evaluated in proximity to the sliding door, at 1 m with respect to the floor. Some telescopic tripod poles were used for instrument arrangement during tests. Instruments were connected to a radio master R-Log data logger system. Measurements were made every two seconds. In

Table 1 the average values of experimental measures upon N samples (*EX PAV* =1 / *<sup>N</sup>* ∑ *N EX Pi* )

*i*=1

are shown for the different room conditions. Data dispersion with respect to the average values was quantified by means of the sample standard deviation σ(N-1), computed using Bessel's correction and the percentage deviation with respect to the average value is also shown in Table 1. For air active sampling, a particle counter (Climet CL 754) was used. The equipment can sample a volume of 75 l/min of air and gives the number of suspended particles, divided according to the diameter (≥0.3; ≥0.5, ≥1.0; ≥5.0 µm), allowing one to deduce particle concen‐ tration in the diameter ranges (0.3-0.5 µm); (0.5-1 µm); (1-5 µm). Experimental data of air temperature and relative humidity (RH) are globally consistent: standard deviations are narrow, much more for temperature (less than 1 °C) than for RH (less than 3 percentage points); normalized deviation does not exceed a 6% threshold in any cases. At point PT01 a detected dispersion of experimental data appears, that becomes very significant for air velocity measurements. The higher variation of the measured velocity in PT01 during operational conditions, due to medical staff movement, produces an increase in RH levels in the central zone (PT01). Temperature values, recorded during operational conditions, are slightly lower than values in at rest conditions, because the local ventilation system operates in the full working phase. Analysing particle concentration measurements, excluding PT1 at operational conditions for particles of 0.3 µm diameter and PT05 at rest conditions for particles of 5 µm diameter, particle distribution for each diameter is similar for the two OT use conditions at the different measurement points.

As expected, the highest particle concentration values are during operational conditions. Referring to the absolute value, the higher differences between the two OT use conditions were checked at PT05 for each particle diameter, while the lower differences were checked at PT01. Particle concentration measured at PT01 for the two use conditions of OT and for each diameter, respects the standard limit imposed by the ISO5 at rest classification [34,35] and Grade B EU-GMP clean-room classification (Annex1), in accordance with EN ISO 14644, for both at rest and operational conditions. Referring to the GMP at operational conditions, at PT01 the conditions imposed by Grade B are respected by all the various particles (for all different

Microclimate and Indoor Air Quality in an Operating Theatre under real use Conditions — An Experimental… http://dx.doi.org/10.5772/59671 519


**Table 1.** Microclimatic parameters measured at the discrete points for different conditions: average values (EXPav), sample standard deviation (σ(N-1)) and percentage error (ι).

diameters) at operational conditions, and the concentration of 0.5 µm diameter particles is within the standard limits for operational conditions, but in particular at PT01, PT03, PT04 and PT05, it is higher.

#### **2.2. Numerical modelling**

the operating table, and then turning round and moving back to the corridor. The HVAC system was working during all tests, supplying a constant incoming airflow rate at controlled temperature. The measurements campaign was carried out over a two day period, from 8 a.m. to 2 p.m. Microclimate parameters were measured over time at discrete points PT01, PT02, PT03 that are shown in Figure 1. In particular: 1 hot wire anemometer, 1 air-temperature and humidity sensor and 1 globe thermometer were set below the central air inlet diffuser on the plane of the surgical table (PT01); 1 air temperature and humidity sensor and 1 differential pressure sensor near the sliding door on a tripod (PT02); 1 differential pressure sensor and 1 air temperature and humidity sensor were set near the air outlet diffuser on a utility pole (PT03). Differential air pressure, between monitored room and the adjoining ones, was evaluated in proximity to the sliding door, at 1 m with respect to the floor. Some telescopic tripod poles were used for instrument arrangement during tests. Instruments were connected to a radio master R-Log data logger system. Measurements were made every two seconds. In

Table 1 the average values of experimental measures upon N samples (*EX PAV* =1 / *<sup>N</sup>* ∑

different measurement points.

518 Current Air Quality Issues

are shown for the different room conditions. Data dispersion with respect to the average values was quantified by means of the sample standard deviation σ(N-1), computed using Bessel's correction and the percentage deviation with respect to the average value is also shown in Table 1. For air active sampling, a particle counter (Climet CL 754) was used. The equipment can sample a volume of 75 l/min of air and gives the number of suspended particles, divided according to the diameter (≥0.3; ≥0.5, ≥1.0; ≥5.0 µm), allowing one to deduce particle concen‐ tration in the diameter ranges (0.3-0.5 µm); (0.5-1 µm); (1-5 µm). Experimental data of air temperature and relative humidity (RH) are globally consistent: standard deviations are narrow, much more for temperature (less than 1 °C) than for RH (less than 3 percentage points); normalized deviation does not exceed a 6% threshold in any cases. At point PT01 a detected dispersion of experimental data appears, that becomes very significant for air velocity measurements. The higher variation of the measured velocity in PT01 during operational conditions, due to medical staff movement, produces an increase in RH levels in the central zone (PT01). Temperature values, recorded during operational conditions, are slightly lower than values in at rest conditions, because the local ventilation system operates in the full working phase. Analysing particle concentration measurements, excluding PT1 at operational conditions for particles of 0.3 µm diameter and PT05 at rest conditions for particles of 5 µm diameter, particle distribution for each diameter is similar for the two OT use conditions at the

As expected, the highest particle concentration values are during operational conditions. Referring to the absolute value, the higher differences between the two OT use conditions were checked at PT05 for each particle diameter, while the lower differences were checked at PT01. Particle concentration measured at PT01 for the two use conditions of OT and for each diameter, respects the standard limit imposed by the ISO5 at rest classification [34,35] and Grade B EU-GMP clean-room classification (Annex1), in accordance with EN ISO 14644, for both at rest and operational conditions. Referring to the GMP at operational conditions, at PT01 the conditions imposed by Grade B are respected by all the various particles (for all different

*i*=1 *N EX Pi* )

> A solid model of the OT was made: the OT geometry is outlined by a rectangular-shaped room with smoothed corners, a 43 m2 base area and 120 m3 volume (Figure 1).

> A sliding door 1.4 m wide and 2.2 m high connects the room to an entry corridor. The locations of the staff members, surgical lamps and equipment for the operating table, defined in the model are shown in the same Figure 1. The surgical staff and patient bodies are outlined as rectangular solid boxes. The room is equipped with an operating bed and lighting system made up of three joined arms, each one holding three lamps. Two further solid boxes represent an operating trolley and an electro-medical device. One of the walls of the room has a window, facing an internal space. There are two rectangular supply ceiling diffusers each of 0.56 m2 surface, located in the central zone of the ceiling, that strengthen the unidirectional flow. Each diffuser provides a constant flow rate of 3969 m3 h-1 of fresh air, so that there are 66 total air changes per hour in the room. Although the real HVAC system is temperature-controlled by a remote ambient thermostat located inside the room, in the numerical models a constant inlet temperature value is set for the incoming fresh air. Two groups of 14 conical outgoing grilles (cross-section of 0.0128 m2 ) are arranged over two of the opposite four walls corresponding to the OT smoothed corners. Following the indications given in the standards [36,37], the global OT air-volume (TV) is divided into 3 zones, labelled from now on as Breathing Zone (BZ, pink

**Figure 1.** System description: in the map (upper portion of the figure) positions (labelled as PT01, PT02, PT03, PT04 and PT05) where instruments were located are shown; some photos taken during the experimental campaign are pro‐ vided. In the bottom portion, the geometry of the numerical model: solid elements used for simulating people, room furnishing (operating table, trolleys) and equipment (ceiling diffusers, lighting system) are shown as well as some indi‐ cations about physical constraints for the system.

in Figure 2), Occupied Zone (OZ, blue in Figure 2) and Peripheral Zone (PZ, corresponding to TV-OZ). This allowed the computation of air quality indexes for each zone. Numerical models were built-up to simulate airflows, climate and air quality conditions in the real orthopaedic OT used for experimental measurements.

Commercial software, that allows multi-physical analyses through solution of the related governing equations by a Finite Element approach [38], was used. Fluid-dynamics and thermal analysis were firstly solved under the assumption of Newtonian fluid and incompressible flow. A standard k-ε closure scheme [39] was used for solving velocity and pressure field by an eddy viscosity approach. Then, further transport-diffusion equations were considered for solving RH, CO2 concentration, mean age of air and particulate concentration for diameters 0.4, 0.75, 3 µm. These dimensions represent the average particle diameter referring to the particle diameter range that were experimentally measured ((0.3-0.5 µm); (0.5-1 µm); (1-5 µm)). The basic formulation of the PDE used for computations is:

Microclimate and Indoor Air Quality in an Operating Theatre under real use Conditions — An Experimental… http://dx.doi.org/10.5772/59671 521

**Figure 2.** OT zones description: BZ (pink) and OZ (blue).

in Figure 2), Occupied Zone (OZ, blue in Figure 2) and Peripheral Zone (PZ, corresponding to TV-OZ). This allowed the computation of air quality indexes for each zone. Numerical models were built-up to simulate airflows, climate and air quality conditions in the real orthopaedic

**Figure 1.** System description: in the map (upper portion of the figure) positions (labelled as PT01, PT02, PT03, PT04 and PT05) where instruments were located are shown; some photos taken during the experimental campaign are pro‐ vided. In the bottom portion, the geometry of the numerical model: solid elements used for simulating people, room furnishing (operating table, trolleys) and equipment (ceiling diffusers, lighting system) are shown as well as some indi‐

Commercial software, that allows multi-physical analyses through solution of the related governing equations by a Finite Element approach [38], was used. Fluid-dynamics and thermal analysis were firstly solved under the assumption of Newtonian fluid and incompressible flow. A standard k-ε closure scheme [39] was used for solving velocity and pressure field by an eddy viscosity approach. Then, further transport-diffusion equations were considered for solving RH, CO2 concentration, mean age of air and particulate concentration for diameters 0.4, 0.75, 3 µm. These dimensions represent the average particle diameter referring to the particle diameter range that were experimentally measured ((0.3-0.5 µm); (0.5-1 µm); (1-5 µm)). The

OT used for experimental measurements.

cations about physical constraints for the system.

520 Current Air Quality Issues

basic formulation of the PDE used for computations is:

$$\frac{\partial \left(\rho \phi\right)}{\partial t} + \nabla \cdot \left(\rho \phi \mathbf{U}\right) = \nabla \cdot \left(\Gamma \nabla \phi\right) + \Lambda \tag{1}$$

where *ρ* is the fluid density and **U**(u, v, w) is the air velocity vector. Table 2 shows the analytical formulation of different terms in eq. (1), depending on the specific physics referred to. From a physical point of view, source term F corresponds to the buoyancy force, QS is the metabolic sensible heat and QL the latent one. For comparing room ventilation performance in different conditions, the mean age of air (τ) was computed: it quantifies the average lifetime of air at a particular location of the room [41,42,43] once a steady state is achieved in terms of airflow patterns. In solving the particulate concentration, because we considered diameters higher than 0.01 µm, the Brownian diffusivity was disregarded with respect to turbulent diffusivity, that was assumed to be equal to the air kinematic turbulent viscosity [44]. Adopting an Euler approach, the following settling velocity was added to the transport vector vertical component:

$$\mathbf{w}\_{S} = \frac{\mathbf{C}\_{c} \rho\_{p} d\_{p}^{2}}{18\mu} \cdot \mathbf{g} \tag{2}$$

with Cc the Cunningham coefficient, ρp and dp particulate density and diameter, respectively. Source terms and physical properties values used in governing equations are listed in Table 3.


**Table 2.** Analytical formulation of terms in eq. (1), depending on the specific physics referred to.


**Table 3.** Physical properties of materials and source term values.

Movements of the sliding door and of one healthcare assistant were numerically simulated to study the incorrect operational use conditions. In transient analyses, once the sliding door is open, one assistant is expected to walk from the corridor space inside the room, moving along the surgical table up to the patient's head, then turning around and leaving the room. The mean velocity values applied for a person walking and door opening/closing are respectively 0.7 m s-1 s and 0.28 m s-1. The procedure adopted for simulating the "moving objects" inside the room was presented in detail in recent studies by the authors [18,19]. It is mainly based on the definition of specific source terms in the governing equations, assuming assigned values in the portions of the computational domains where the solid objects are located at a chosen time. Solid objects were not explicitly designed in the geometrical model because their movement is driven by assumed values 0 or 1 in specific logical functions during time. Solid object time-displacements and paths considered during transient analyses are shown in Figure 3. The black line represents displacement of the "moving object" sliding door (x direction into the model reference system), while the red line indicates movement belonging to the "moving object" healthcare assistant (y direction). The grey dashed lines schematically separate the different steps characterizing the simulated dynamics, that are highlighted by progressive numbers from 1 to 7, as shown in the grey horizontal bars in the upper part of Figure 3. Assuming the initial instant with the door closed and medical assistant standing still in front of it, in the corridor, 7 consecutive steps were simulated: step1 (0-5 seconds): sliding door opens, and medical assistant is standing still, waiting for the door to open; step 2 (5-11 seconds): the door is completely open and medical assistant is walking through it. After 1 second, the door starts to close, while the person is walking in the room; step 3 (11-15 seconds): the door is closed and medical assistant walks through the room until he reaches the operating table top; step 4 (15-27 seconds): medical assistant stops for 3 seconds, then moves in the opposite direction towards the sliding door and finally stops inside the room in front of the door waiting for it to open; step 5 (27-32 seconds): medical assistant is standing still until the door is completely open; step 6 (32-33 seconds): the door is completely open and the person is walking through it and then stopping in the corridor; step 7 (33-38 seconds): sliding door closes.

**Table 2.** Analytical formulation of terms in eq. (1), depending on the specific physics referred to.

Movements of the sliding door and of one healthcare assistant were numerically simulated to study the incorrect operational use conditions. In transient analyses, once the sliding door is open, one assistant is expected to walk from the corridor space inside the room, moving along the surgical table up to the patient's head, then turning around and leaving the room. The mean velocity values applied for a person walking and door opening/closing are respectively 0.7 m s-1 s and 0.28 m s-1. The procedure adopted for simulating the "moving objects" inside the room was presented in detail in recent studies by the authors [18,19]. It is mainly based on the definition of specific source terms in the governing equations, assuming assigned values in the portions of the computational domains where the solid objects are located at a chosen time. Solid objects were not explicitly designed in the geometrical model because their movement is driven by assumed values 0 or 1 in specific logical functions during time. Solid

**Table 3.** Physical properties of materials and source term values.

522 Current Air Quality Issues

**Figure 3.** Path on the room map (right side) and time-displacement (left side) of the "moving objects": the black line identifies the transient position of the sliding door (moving along the x-axis in the chosen system of coordinates), the red one refers to the healthcare assistant (moving along the y-direction). The grey dashed lines separate the different steps characterizing the simulated dynamics (numbers in the grey horizontal bars).

The computational domain, representing the corridor, was considered only for simulations involving the sliding door opening/closing (incorrect use conditions). In this case, open boundary conditions were considered at the corridor transversal section. The following firsttype boundary conditions for supplied air at the ceiling diffusers were assumed: velocity magnitude (1.97 m/s), turbulence intensity (5%), air temperature (18 °C), RH (60%), CO2 concentration (350 ppm) and mean age of air (0 s). For modelling persons breathing, a RMS value, corresponding to the sinusoidal trend for inhaled/expired air, and for CO2 emission rate into the room were assumed. Adopted values in simulations were 2.0 m3 /h (medical staff) and 0.3 m3 /h (patient) for breathed airflow and 0.080 m3 /h (medical staff) and 0.012 m3 /h (patient) for CO2 emission rate. Outflow conditions were considered for all dependent variables at the recovery grids. At each solid/fluid interface, logarithmic wall functions were applied in the near wall airflow, that was considered parallel to the wall and being in a wall offset equals one hundred viscous units. Turbulent production was established equal to dissipation at walls. For the remaining dependent variables impermeable/insulation conditions at walls were imposed. The occupants contribution for particle emission rate was applied as a boundary flux [particle/(m2 s)] at the occupants/surrounding air interfaces. The procedure adopted to assess the particulate flux depending on particle dimension is explained in the next section of the present chapter. In thermal analysis, a convective thermal flux was applied to the walls, considering a heat exchange coefficient of 7.7 (W/(m2 K)) and a constant temperature of 20 °C in the adjacent rooms. Insulation conditions were applied to the solid/fluid interfaces for all other dependent variables that were solved. For the "at rest" conditions, the lamps were considered the only internal sources of sensible heat. Otherwise, in operational "correct use" conditions, internal heat sources were related to medical staff and patient presence, as sensible and latent heat loads. Other boundary conditions were not considered to change from "correct use" to "incorrect use" operational conditions: heat and vapour source values were un‐ changed, except for the additional load due to the walking healthcare assistant, whose location was variable in accordance with the moving object position during time. The governing equations together with their boundary conditions were spatially discretized on non-struc‐ tured grids, made of second order tetrahedral elements. Steady state solutions of discrete equations were carried-out by applying an iterative dumped Newton-Raphson scheme [45] based on the discretized PDE linearization by a first-order Taylor expansion. Algebraic systems of equations coming from differential operator discretization were solved by a PARDISO package, a direct solver particularly efficient to solve unsymmetrical sparse matrixes by a LU decomposition method. The convergence criterion was set to 1E-5. Time integration of governing equations for transient simulations was performed applying an Implicit Differen‐ tial-Algebraic (IDA) solver [46], which uses variable-order and variable-step-size Backward Differentiation Formulas (BDF). Because the time-matching scheme is implicit, a nonlinear system of equations was solved at each time step. All computations were carried out on a workstation with two 64-bit 6-core/12-thread processors speeding up to 2.3 GHz of frequency and handling 128 GB of RAM.
