**5. Numerical results**

diameter (di

528 Current Air Quality Issues

values.

and operational (oper) conditions, computed as follows:

while the error bars are plotted by assuming the following:

where values suggested in the previous literature are also reported.

) represents the difference between average data (Av) acquired for the at rest (rest)

<sup>=</sup> {} {} - *ii i dd d oper rest EXP Av EXP Av EXP* (3)

max{} {} min <sup>+</sup> <sup>=</sup> - *ii i dd d oper rest Err EXP EXP* (4)

min{} {} max - <sup>=</sup> - *ii i dd d oper rest Err EXP EXP* (5)

Numerical values plotted, refer to the "final" particle emission rate per average diameter applied to the exposed person surfaces carried out from application of the computing proce‐ dure shown in Figure 5. Values of emission rate for diameter range are provided in Table 5,

**Figure 6.** Particles concentration per diameter in different locations: comparison between experimental and numerical

Despite an iterative application of our proposed "guess and check" procedure, experimental/ numerical difference in particle concentration at point PT01 remained quite high: the numerical model overestimates particle contents in the air at this location. Otherwise, a good agreement can be pointed out from comparison of particle concentration at the other different locations. Obtained results of particle emission rate per diameter and per person (Table 5), provide lower values with respect to those proposed by Quian et al. However, this is an expected result,

#### **5.1. Microclimate and ventilation assessment**

This chapter section is devoted to numerical results analysis and discussion. Figure 7-a-c provides the air velocity fields in a horizontal section (z=1.5) for the different room conditions studied. Transient simulation results (for the "incorrect use conditions", Figure 7c) refer to step 3, when the door is completely closed and the medical assistant walks through the room, until he reaches the top of the operating table. In particular, Figure 7c refers to the first 13 seconds, when the moving person has reached the table midpoint. Distribution of velocity magnitude in the operational zone is significantly modified from the "at rest" towards both the "opera‐ tional conditions". In Figure 8a-c velocity profiles along x (y=8; z=1.5), y (x=3; z=1.5) and z (x=3; y=8) direction are given for the different room conditions. The "incorrect use conditions" still refer to the time instant t=13 s.

**Figure 7.** Air velocity field [m/s] in a horizontal plane (z=1.5) for "at rest" (a), "correct operational use" (b) and "incor‐ rect operational use" (c).

In these diagrams zero values represent the "imprint" of a fixed object/person or moving person standing in that location for the considered instant. Velocity profiles underline the modifications of the air flow patterns due to the operational conditions. In particular, air flow in the surgical zone is strongly modified by medical staff presence: important differences, produced by this effect, are shown with the gap between dashed lines and continuous/dasheddotted lines in Figs. 8a-8c.

**Figure 8.** Velocity profiles for "at rest" (dashed line), "operational correct use (continuous line) and "operational incor‐ rect use" (dotted-dashed line) along x (a) (y=8; z=1.5), y (b) (x=3; z=1.5) and z (c) (x=3; y=8).

The effect due to the person movements on the airflow patterns in the surgical zone, appears to be less important. In Figs.8b,8c continuous and dashed-dotted lines are almost overlapped, while in Figure 8a they are remarkably distant only for x<2.5. Velocity profiles show variable trends and high curve slopes in each case: this is really very important when the efficacy of "unidirectional" or "laminar" airflow is discussed for similar applications. Now the influence of OT use conditions on indoor thermal field variations, is discussed. Figure 9 shows the air temperature distribution by means of contour plots and horizontal slice (z=1.5), obtained for "incorrect use conditions", step 3, time 13 s. Thermal "imprint" of a walking person is clear. On the left side of Figure 9, thermal profiles obtained along the x axis (y=8; z=1.5, see the line sketch in the coloured map) for the different room conditions, is provided. Temperature variation is evident in the "correct operational use" compared with the empty room charac‐ terized by a predominant isotherm profile. The additional thermal load, due to the walking person, produces a further local temperature increase. In the operating zone thermal levels remains very close to the design value despite the different use conditions. The mean air temperature value, computed all over the OT, is within the limits suggested by the Italian and International standards for the correct use conditions (23.5 °C) and slightly outside the limits for the incorrect use conditions (24.4 °C). Taking into account results obtained on the air RH, it can be detected that vapour production, due to persons presence, determines an air moisture content that is not well balanced by the incoming air at the considered hygrometric conditions. Figure10 shows (left side) RH distribution in a horizontal slice (z=1.5) for "correct operational conditions".

In these diagrams zero values represent the "imprint" of a fixed object/person or moving person standing in that location for the considered instant. Velocity profiles underline the modifications of the air flow patterns due to the operational conditions. In particular, air flow in the surgical zone is strongly modified by medical staff presence: important differences, produced by this effect, are shown with the gap between dashed lines and continuous/dashed-

**Figure 8.** Velocity profiles for "at rest" (dashed line), "operational correct use (continuous line) and "operational incor‐

The effect due to the person movements on the airflow patterns in the surgical zone, appears to be less important. In Figs.8b,8c continuous and dashed-dotted lines are almost overlapped, while in Figure 8a they are remarkably distant only for x<2.5. Velocity profiles show variable trends and high curve slopes in each case: this is really very important when the efficacy of "unidirectional" or "laminar" airflow is discussed for similar applications. Now the influence of OT use conditions on indoor thermal field variations, is discussed. Figure 9 shows the air temperature distribution by means of contour plots and horizontal slice (z=1.5), obtained for "incorrect use conditions", step 3, time 13 s. Thermal "imprint" of a walking person is clear. On the left side of Figure 9, thermal profiles obtained along the x axis (y=8; z=1.5, see the line sketch in the coloured map) for the different room conditions, is provided. Temperature variation is evident in the "correct operational use" compared with the empty room charac‐

rect use" (dotted-dashed line) along x (a) (y=8; z=1.5), y (b) (x=3; z=1.5) and z (c) (x=3; y=8).

dotted lines in Figs. 8a-8c.

530 Current Air Quality Issues

**Figure 9.** On the left side: temperature [°C] map plotted on contours and in a horizontal slice (z=1.5) for the "incorrect operational use" (step 3, time 13 s). On the right side: temperature profiles along the x-direction (y=8; z=1.5, see the line sketched on the slice) for "at rest" (dashed line), "operational correct use" (continuous line) and "operational incorrect use" (dotted-dashed line) conditions.

**Figure 10.** On the left side: RH (%) field in a horizontal plane (z=1.5) for the "correct operational use". On the right side: RH profile along the y-direction (x=1.1; z=1.5) for "operational correct use" (continuous line) conditions.

The mean value computed all over the room in this condition is 67.8%, which is exceeding any maximum threshold suggested by all the standards. The RH profile lying on a horizontal line along the y axis (x=1.1; z=1.5, sketched in the horizontal slice) is also reported in the right portion of Figure10. An increasing level of the moisture content along the y-direction can be noticed: RH at the back of the operating zone presents higher level. This could be due to the lack of ventilation all over the room that can determine stagnation zones. The mean age of air (τ) was also evaluated, using the steady airflow achieved for "at rest" and "correct operational conditions" as transport field for τ computations. The lower τ value corresponds to the higher air washing effect of the ventilation system for the considered zone. Results are plotted in Figure 11, where τ distribution is reported for both the analyzed conditions. In the same, the τ profiles along a horizontal line lying on the x direction (y=8; z=1.5, sketched on the horizontal slices) are provided. The τ values are not critical in both analyzed conditions. A much more uniform distribution is found for the empty room, and in this condition the medical staff act as a "constraint" for the local airflow, allowing a slightly lower air age value. It should be noticed that a lower air age does not directly mean better air quality. We were finally interested in assessing the effect of the sliding door opening/closing during the simulated "incorrect operational conditions". Figure12 provides the air velocity field on a horizontal plane (z=1.5) for step 2, at time 7 seconds: the medical assistant is walking in the room and the sliding door is shutting behind his back. In the same figure, as an enlargement, air velocity vector distri‐ bution is shown for the zone close to the sliding door. The important velocity field variation, in the zone of the sliding door, and also the one due to the medical assistant's movement through the room, is evident. As a consequence, an important rate of air outflows from the room. Then a total amount of 16.3 m3 of air outflows toward the corridor during steps 5-6-7 (door opening/person crossing/door closing) was estimated.

**Figure 11.** On the left side: mean age of air [s] in a horizontal slice (z=1.5) for the "at rest" (a) and "correct operational use" (b). On the right side: mean age of air profiles along the x-direction (y=8; z=1.5, see the line sketched on the slices) for "at rest" (continuous line) and "operational correct use" (dashed line) conditions.

Figure13 provides a representation of the airflow rate (continuous black line) and the total volume of air (grey-filled surface) outgoing the OT during the door opening. Indeed, the effect of the door opening on the average pressure level inside the OT was also investigated. Figure13 also shows the average pressure trend as a function of time during step 1 (door opening): a significant pressure decrease can be seen from the plotted data. Starting from its initial value, i.e. 32.6 Pa, the OT average overpressure with respect to the corridor level becomes very low, i.e. 1.2 Pa. The overpressure variation due to an "unforeseen event" occurring during an "incorrect use condition" ofthe OT, can determine a temporary non compliance of the pressure scheme with limits suggested by all the considered standards giving specific indications for it.

The mean value computed all over the room in this condition is 67.8%, which is exceeding any maximum threshold suggested by all the standards. The RH profile lying on a horizontal line along the y axis (x=1.1; z=1.5, sketched in the horizontal slice) is also reported in the right portion of Figure10. An increasing level of the moisture content along the y-direction can be noticed: RH at the back of the operating zone presents higher level. This could be due to the lack of ventilation all over the room that can determine stagnation zones. The mean age of air (τ) was also evaluated, using the steady airflow achieved for "at rest" and "correct operational conditions" as transport field for τ computations. The lower τ value corresponds to the higher air washing effect of the ventilation system for the considered zone. Results are plotted in Figure 11, where τ distribution is reported for both the analyzed conditions. In the same, the τ profiles along a horizontal line lying on the x direction (y=8; z=1.5, sketched on the horizontal slices) are provided. The τ values are not critical in both analyzed conditions. A much more uniform distribution is found for the empty room, and in this condition the medical staff act as a "constraint" for the local airflow, allowing a slightly lower air age value. It should be noticed that a lower air age does not directly mean better air quality. We were finally interested in assessing the effect of the sliding door opening/closing during the simulated "incorrect operational conditions". Figure12 provides the air velocity field on a horizontal plane (z=1.5) for step 2, at time 7 seconds: the medical assistant is walking in the room and the sliding door is shutting behind his back. In the same figure, as an enlargement, air velocity vector distri‐ bution is shown for the zone close to the sliding door. The important velocity field variation, in the zone of the sliding door, and also the one due to the medical assistant's movement through the room, is evident. As a consequence, an important rate of air outflows from the

**Figure 11.** On the left side: mean age of air [s] in a horizontal slice (z=1.5) for the "at rest" (a) and "correct operational use" (b). On the right side: mean age of air profiles along the x-direction (y=8; z=1.5, see the line sketched on the slices)

Figure13 provides a representation of the airflow rate (continuous black line) and the total volume of air (grey-filled surface) outgoing the OT during the door opening. Indeed, the effect of the door opening on the average pressure level inside the OT was also investigated. Figure13

of air outflows toward the corridor during steps 5-6-7

room. Then a total amount of 16.3 m3

532 Current Air Quality Issues

(door opening/person crossing/door closing) was estimated.

for "at rest" (continuous line) and "operational correct use" (dashed line) conditions.

**Figure 12.** Air velocity field [m/s] in a horizontal slice (z=1.5) for step 2 and time 7 s (left side) and an enlargement with velocity vectors representation, in the proximity of crossing zone of the door towards the corridor (right side) during "incorrect operational use" conditions.

**Figure 13.** Time evolution of average OT pressure (dashed black line), airflow rate (continuous black line) and total volume of air out-coming the OT (grey-filled surface) during the sliding door opening (Step 5 of the "incorrect opera‐ tional use" conditions).

#### **5.2. IAQ indexes evaluation**

Referring to the air freshness concept, the calculated CO2 concentration, was found to be considerably below the critical limit of 3000 parts per million (ppm) in the global room volume and 1000 ppm in the operating table zone. CO2 concentration is high only in the breathing zone where staff are exhaling. The driving effect due to the velocity field prevails over its spatial distribution. The simulation results obtained for CO2 concentration and its distribution in the BZ and OZ, but also in the PZ and in the total volume of the OT, are in agreement with those provided by recent studies [15, 53]. Anyway, results showed that there is a significant increase in the CO2 concentration from the fundamental zone BZ, to the OZ but progressively to the PZ and TV. Moreover, some IAQ indexes were computed using simulation results and discussed. The proposed indexes are usually applied for IAQ assessment and a quantitative evaluation of ventilation system performance with regard to contaminant removal and infection risk control [54, 55, 56]. Once the distribution of the dependent variables inside the OT, i.e. air velocity, CO2 and particle concentration, were evaluated by simulations, IAQ indexes were calculated. Some of them were expressed in the form of a continuous distribution (local indexes), others were referred to the average values of dependent variables, in the BZ, OZ, PZ and TV. The first investigated IAQ parameter was the mean age of air (τ) [41, 42, 43]. The air age concept is generally defined as the average time for air to travel from a supply inlet area to any location in a forced ventilated room [57, 58, 59, 8]. The mean age of air was calculated as a dependent variable, as explained in the modeling section. It provides a measure of air freshness, so its lower values are more favourable. The τ parameter trend was controlled during a transient simulation of 1800 seconds, starting from an initial state corresponding to the steady state condition discussed in the above section. Figure 14 shows a spatial distribution of τ in a horizontal and in a vertical slice of the room, once the transient time to achieve the steady state was expired. Results show that lifetime of air located in the central portion of the OT is much lower than that concerning the peripheral portion of the room (Figure 14).

**Figure 14.** Mean age of air in horizontal (z=1) and vertical (x=3.0) slices.

To quantify this result, the average value of τ was computed in the OT different zones (i.e. BZ, OZ, PZ, TV), and called as τZj, where Zj means the generic j-zone (Table 6). Generally, these

values are quite low. Comparison of these values with the theoretical residence time of air inside the OT (defined as the ratio between the total volume of the room (VTV, m3 ) and the mass flow rate of incoming ventilating air (Vvent, m3 /s)), shows that the ratio is always higher than 1.


**Table 6.** Values of mean age of air and Air Change Efficiency in the different zones.

**5.2. IAQ indexes evaluation**

534 Current Air Quality Issues

**Figure 14.** Mean age of air in horizontal (z=1) and vertical (x=3.0) slices.

To quantify this result, the average value of τ was computed in the OT different zones (i.e. BZ, OZ, PZ, TV), and called as τZj, where Zj means the generic j-zone (Table 6). Generally, these

Referring to the air freshness concept, the calculated CO2 concentration, was found to be considerably below the critical limit of 3000 parts per million (ppm) in the global room volume and 1000 ppm in the operating table zone. CO2 concentration is high only in the breathing zone where staff are exhaling. The driving effect due to the velocity field prevails over its spatial distribution. The simulation results obtained for CO2 concentration and its distribution in the BZ and OZ, but also in the PZ and in the total volume of the OT, are in agreement with those provided by recent studies [15, 53]. Anyway, results showed that there is a significant increase in the CO2 concentration from the fundamental zone BZ, to the OZ but progressively to the PZ and TV. Moreover, some IAQ indexes were computed using simulation results and discussed. The proposed indexes are usually applied for IAQ assessment and a quantitative evaluation of ventilation system performance with regard to contaminant removal and infection risk control [54, 55, 56]. Once the distribution of the dependent variables inside the OT, i.e. air velocity, CO2 and particle concentration, were evaluated by simulations, IAQ indexes were calculated. Some of them were expressed in the form of a continuous distribution (local indexes), others were referred to the average values of dependent variables, in the BZ, OZ, PZ and TV. The first investigated IAQ parameter was the mean age of air (τ) [41, 42, 43]. The air age concept is generally defined as the average time for air to travel from a supply inlet area to any location in a forced ventilated room [57, 58, 59, 8]. The mean age of air was calculated as a dependent variable, as explained in the modeling section. It provides a measure of air freshness, so its lower values are more favourable. The τ parameter trend was controlled during a transient simulation of 1800 seconds, starting from an initial state corresponding to the steady state condition discussed in the above section. Figure 14 shows a spatial distribution of τ in a horizontal and in a vertical slice of the room, once the transient time to achieve the steady state was expired. Results show that lifetime of air located in the central portion of the OT is much lower than that concerning the peripheral portion of the room (Figure 14).

Indeed, this comparison consists in computing the following Air Change Efficiency (ACE) index:

$$ACE = \frac{V\_{TV} \int \dot{V}\_{\text{vent}}}{\tau\_{Z\_{\text{j}}}} \cdot 100\tag{6}$$

whose values in the different zones are given in Table 6. The ACE index measures how effectively ventilation systems replace the air in a room with fresh air. In the BZ the average lifetime of air is more than 2 times lower than the theoretical residence time (about 53.7 seconds), that can be deduced analytically. The Local Air Change Efficiency (LACE) is expressed by the following expression:

$$LACE = \frac{V\_{TV} \int \dot{V}\_{\text{vent}}}{\pi} \cdot 100\tag{7}$$

The LACE index characterizes the conditions at a specific point (defined as the ratio between the minimum replacement time, as previously defined, and the local mean age of the air). It is possible to observe (Figure15) that the zone corresponding to the operating table is more favourable from this point of view, reaching LACE values of up to 500-600%.

Knowing the concentration field computed for CO2 and particles, the Ventilation Effectiveness (VE) index was also assessed. The VE index measures how quickly a contaminant is removed from an air volume by quantifying the efficiency with which the internal pollutant is diluted or removed. It depends on the airflow patterns, and is expressed as follows:

$$VE = \frac{C\_E - C\_S}{C\_{Z\_f} - C\_S} \tag{8}$$

where CE is the mean value of contaminant concentration (i.e. CO2 and particles) calculated at the air-recovery grilles (Exhaust), CS is the contaminant concentration at the air inlet diffusers (Supply) and CZj is the mean value of the contaminant concentration in a specific OT zone (i.e. BZ, OZ, PZ, TV).

**Figure 15.** LACE index in horizontal (z=1.2) and vertical slices (x=3.0).

Similarly, the Contaminant Removal Effectiveness (CRE) index, expressed by the ratio between the concentration of contaminants at the exhaust point and the mean value of contaminant concentration within a specific zone was evaluated:

$$CRE = \frac{C\_E}{C\_{Z\_{j}}} \tag{9}$$

Figure 16 shows the VE and CRE indexes, computed using CO2 and particle concentrations. Because the particle concentration value was assumed to be zero at the inlet air diffusers (CS=0), VE and CRE expressions correspond to each other. Due to the very low effect of the settling velocity on particle concentration distribution, a very low quantitative difference was found in computing the CRE (or VE) index as a function of the different particle diameter ranges studied. Therefore, we referred to a single CRE index representative for any particle diameter range studied. Values of the computed global indexes are shown in Figures 17 (CO2) and 18 (particles). The CRE distribution at a specific point, that is known as Local Contaminant Removal Effectiveness (LCRE) was also calculated by:

$$LCRE = \frac{C\_E}{C} \tag{10}$$

23

Local index distributions are shown in Figure 17 for CO2 and in Figure 18 for the particles concentration. Comparison between the LCRE indexes, computed for CO2 and particles, highlights the combined effect of the two sources (nose for CO2 and body surface for particles) which provides very different trends. The mass transport effect is particularly evident in both cases.

3 **6. Conclusions Figure 16.** VE (a) and CRE (b) computed for CO2 concentration in the different zones and VE (or CRE) computed for particles concentration in the different zones (c).

4 An experimental and numerical investigation on the airflow patterns and thermal field in a 5 real OT is presented in this chapter. Different scenarios were considered, then measured and 6 simulated, representative of "at rest" and "operational/effective use" conditions of the OT.

9 quantitative microclimatic parameters (air velocity, temperature, RH and pressure) the

1 **Fig. 16.** VE (a) and CRE (b) computed for CO2 concentration in the different zones and VE

2 (or CRE) computed for particles concentration in the different zones (c).

#### 7 Numerical models were successfully validated against experimental data. In this chapter 8 crucial results of our investigation, obtained through comparison and discussion of **6. Conclusions**

Running Title

where CE is the mean value of contaminant concentration (i.e. CO2 and particles) calculated at the air-recovery grilles (Exhaust), CS is the contaminant concentration at the air inlet diffusers (Supply) and CZj is the mean value of the contaminant concentration in a specific OT zone (i.e.

Similarly, the Contaminant Removal Effectiveness (CRE) index, expressed by the ratio between the concentration of contaminants at the exhaust point and the mean value of contaminant

<sup>=</sup> *CE CRE*

<sup>=</sup> *CE LCRE*

*C*

*CZ j*

Figure 16 shows the VE and CRE indexes, computed using CO2 and particle concentrations. Because the particle concentration value was assumed to be zero at the inlet air diffusers (CS=0), VE and CRE expressions correspond to each other. Due to the very low effect of the settling velocity on particle concentration distribution, a very low quantitative difference was found in computing the CRE (or VE) index as a function of the different particle diameter ranges studied. Therefore, we referred to a single CRE index representative for any particle diameter range studied. Values of the computed global indexes are shown in Figures 17 (CO2) and 18 (particles). The CRE distribution at a specific point, that is known as Local Contaminant

(9)

(10)

BZ, OZ, PZ, TV).

536 Current Air Quality Issues

**Figure 15.** LACE index in horizontal (z=1.2) and vertical slices (x=3.0).

concentration within a specific zone was evaluated:

Removal Effectiveness (LCRE) was also calculated by:

10 influence of unforeseen movements of medical staff, sensible and latent heat, but also CO2 11 and particles, released by persons in the ambient and sliding door opening/closing on the An experimental and numerical investigation on the airflow patterns and thermal field in a real OT is presented in this chapter. Different scenarios were considered, then measured and simulated, representative of "at rest" and "operational/effective use" conditions of the OT. Numerical models were successfully validated against experimental data. In this chapter crucial results of our investigation, obtained through comparison and discussion of quantita‐ tive microclimatic parameters (air velocity, temperature, RH and pressure) the influence of unforeseen movements of medical staff, sensible and latent heat, but also CO2 and particles, released by persons in the ambient and sliding door opening/closing on the OT climate and IAQ, are provided. It can be noticed that, variations in use (resulting in different internal sensible and latent heat loads, moving objects and room confinement) can play an important role in terms of microclimatic system performance against standard suggestions, even those lacking compliance with standard limits are found to be mainly local or temporary.

**Figure 17.** LCRE for CO2 in a horizontal slice (z=1.6).

**Figure 18.** LCRE for particles concentration in a horizontal slice (z=1.4).

From the case study presented here, the efficiency of the AHU-HVAC system for providing the right indoor microclimatic conditions in compliance with standard thresholds is seen not only related to a good plant design, but also to medical staff and assistant behaviour and room use. Measurements of particle concentration, with and without persons, are used in combination with iteratively computed numerical results in order to assess particle emission rate by the occupants for given particle dimensions. Particle concentration fields are also numerically solved for several particle diameters, by using an Euler approach based on the Cunningham formulation of settling velocity. Numerical results are successfully checked against the experimental evidence. CO2 concentration levels numerically computed in the OT agree well with data reported in the literature for similar applications. Some consolidated indexes, adopted for monitoring the IAQ, are computed for gaseous contaminant (CO2) and particle concentra‐ tion, both in terms of spatial distribution and overall values referring to specific OT zones. Results obtained, by applying our proposed methodology for estimating the particle emis‐ sion rate, highlight a good agreement with the small number of contributions in the literature concerning particle emission for different diameters. Our study contributes to better under‐ standing the additional environmental "load" induced by the medical staff in an OT, based on an innovative strategy proposed and applied to quantify the particle emission rate released by occupants, for given ranges of particle diameters. Our integrated experimental and numerical approach, is in accordance with some recent surgical infection control guidelines, that high‐ light the importance of surgical staff behaviour control in order to decrease air contamination and wound colonisation. These recommendations include restricting the movements and the number of persons in the OT, but they are often general and based on expert advice. An integrated approach, such as that proposed here, based on CFD application and periodic experimental monitoring campaigns of OT, can contribute to providing valuable suggestions for medical staff information and education concerning the analysed topics, and in general can hopefully stimulate careful considerations on specific procedures for OT design and use.
