- Email: [email protected]
CO2-dispersion studies in an operation theatre under transient conditions
Energy and Buildings 40 (2008) 231–239 www.elsevier.com/locate/enbuild
CO2-dispersion studies in an operation theatre under transient conditions C.P. Karthikeyan, Anand A. Samuel * School of Mechanical and Building Sciences, Vellore Institute of Technology University Vellore 632014, Tamil Nadu, India Received 4 January 2007; received in revised form 30 January 2007; accepted 17 February 2007
Abstract Maintaining air quality and thermal comfort inside an operation theatre equipped with horizontal jet flow type air-conditioning units, has been a challenge to engineers. The objective of this study is to analyze the airflow pattern in such operation theatres and the influence of location of the airconditioners. The outcome of this study is expected to reduce the post-operation problems faced due to excess concentration of contaminants. Experimental studies were conducted in 10 different hospital operation theatres. Parameters such as air temperature and carbondioxide concentration were measured at discrete points chosen in the theatre. A 3D time-dependent numerical model was developed to simulate the airflow in terms of parameters such as velocity, temperature and CO2 distributions in an operation theatre under transient conditions. The Eulerian approach using the volume fraction of the mixture of air and CO2 was used to solve the numerical model. Finite volume approach was attempted in this work with PISO (pressure implicit with split operators) algorithm for the pressure correction equations. The simulated results were compared with the experimental results for validity. The locations of the air-conditioners were changed in the numerical model to analyze the airflow patterns and the contaminant distribution. # 2007 Elsevier B.V. All rights reserved. Keywords: Operation theatres; Airflow patterns; CO2-dispersion; CFD; PISO and transient conditions
1. Introduction Indoor air quality is a critical factor in the design of any hospital operation theatre. Studies of operating room air distribution indicate that the supply of air from the ceiling, with a downward flow movement to several ports located on opposite wall yielded probably, the most effective airflow pattern, maintaining the concentration of contaminants to a minimum acceptable level [1]. On account of ease of handling and economic considerations, window type air-conditioning units are used in many operation theatres. Window type units produce horizontal jet flow in contrast to the conventional downward flows. Simulation of airflow patterns for different locations of the air-conditioning units using computational fluid dynamics techniques help in analyzing air quality and thermal comfort. The variation in contaminant distribution, location of dead zones and air velocity in the operation theatre are the major factors that indicate the occurrence of post-operative wound infections.
Major contaminants in any hospital operation theatre are categorized as biological (microorganisms—fungi, bacteria and virus), chemical (waste anesthetic gases, CO2, etc.) and particulate matter [2]. Microorganisms are the major cause for ‘post-operative wound infections’. There are three main sources of contaminants: (a) the supply air to the theatre, (b) the surgical team members, and (c) the emissions of body vapors and organisms from the pathogens, these sources on mixing with the supply air gets deposited on the patient cause infections. The contaminants such as waste anesthetic gases and carbon dioxide, if not controlled properly can lead to severe complications. Most of the studies in the literature focus on air diffusion and solid contaminant behavior in indoor environments using CFD based integrated approach [3]. Studies on the distribution of CO2 in the operation theatres have not been reported in the literature. An attempt is made here to study the CO2 dispersion in an operation theatre using which the air quality is analyzed. 2. Indoor air quality in operation theatres
* Corresponding author. Tel.: +91 416 2243091/93; fax: +91 416 2243092. E-mail addresses: [email protected] (C.P. Karthikeyan), [email protected] (A.A. Samuel). 0378-7788/$ – see front matter # 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.enbuild.2007.02.023
CO2 concentration has been widely used as an indicator of indoor air quality. A Limit of 1000 ppm of CO2 is
232
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
Nomenclature Cm
constant used in turbulence (k–e) model, Cm = 0.09 CO2 carbon dioxide (ppm—parts per million) Hroom height of the room (m) K turbulent kinetic energy (m2 s2) l = 0.07L
QEAS S t0 TI Uinlet V vdr;p vm
l is the characteristic length at the inlet (m) quality of grid cells by Equi-Angle Skew technique source or sink non-dimensional time-step (s) turbulent intensity (%) the inlet velocity (m s1) velocity vector drift velocity of the primary phase (m s1) mass averaged velocity (m s1)
Greek letters ap volume fraction of the primary phase Gf exchange coefficient (laminar + turbulent) e turbulence dissipation rate (m2 s3) ueq characteristic angle corresponding to an equilateral cell of similar form umax maximum angle between the edges of the element in degrees umin minimum angle between the edges of the element in degrees f dependent variable r density (kg m3) rp density of the primary phase (kg m3) recommended to satisfy comfort criteria and CO2 concentration is also an useful indicator of insufficient outside air intake and ventilation problems [22]. In this study CO2-dispersion is used in predicting the regions of high contamination in the operation theatre. The patients in hospital isolation rooms constantly produce transmissible air-borne organisms by coughing, sneezing or talking, which if not under control, results in spreading air-borne infection. The air supply to the operation theatre should normally be free from such air-borne infectious sources. CO2 being a common gas and a basic indicator of air quality, is used for simulation studies, the numerical model developed can also be used for the simulation of any other waste anesthetic gas in theatres. Hypercapnia is a condition that is abnormally caused by shallow respiration or hypoventilation. It is an excessive amount of CO2 in the blood. Very high concentrations of atmospheric CO2 results in hypercapnia [7]. Studies have shown that there are numerous health effects associated with the exposure of waste anesthetic gases [4]. These include increased risk of spontaneous abortion in females exposed to anesthetic gases in hospitals with incidences of 1.5– 2 times greater than in unexposed females. The prevention of post-operative infection is dependent on several factors
including effective sterilization and disinfection procedures, good surgical technique, theatre design, bacterial contamination of theatre air, discipline—which include restricting the movement of staff near the operating table, appropriate use of prophylactic antibiotics, etc. [5]. It has been found that velocities as high as 0.6 m s1 are unsatisfactory because the exposed internal tissues of the patient are over-cooled and dried out by evaporation [6]. Parameters such as air temperature and humidity also play a vital role in surgeries. If the operation room temperature is higher during the eye surgeries, then the cornea dries out and laser beam removes more corneal tissue and an over-correction can result. The reverse occurs if the temperature in the operating theatre is too low. If the humidity is higher than the cornea is wetter and the laser removes less tissue with each pulse, increasing the chance of an under-correction [8]. These factors prove the need for the study and its significance. 2.1. Operation theatre indoor airflow and contaminant distribution simulation Airflow simulation using CFD techniques is an useful tool to analyze air movements in non-standard operating room situations [9]. Modern computational fluid dynamics models are used to predict the air velocity, temperature, turbulence level and contaminant distributions. The current work uses a transient model based on eulerian approach for simulation of indoor airflow and contaminant distribution. In 1997, Lo [16] has made a study addressing contamination control in an operating room. The study considered an operating room under isothermal condition and analyzed the distribution of contaminants. In a first assumption made in the study, the effect of significant thermal plumes in the room was ignored. In a second assumption it was considered that the particles in the room can be considered to follow Brownian motion. These assumptions were reconsidered by Memarzadeh and Manning [18] in their study conducted in reducing the risks in surgery. The basic assumption made in their study was that the squames were simulated as particles being released from several sources surrounding the occupant. These particles were then tracked for a certain period of time. The Brownian motion is strictly applicable to particles that are 1 mm or less in diameter. Bacteria and virus do conform to this criteria but bacteria are usually transported in operating rooms by squames [17], which are considerably bigger (in the range of 10 mm) and so do not necessarily follow Brownian motion. Numerous studies on the simulation of particle trajectories in rooms are reported in the literature but much little is focused on the of gas dispersion, hence this study. Studies conducted on airflow regimes in air-conditioned operating theatres suggest that the supply of air from the ceiling, with a downward flow movement to several exhaust ports located on opposite walls yielded, probably the most effective airflow patterns maintaining the contaminant concentrations to a minimum acceptable level [10–12]. In contrast this study focuses on the window type air-conditioning units, which are used on account of ease of handling and economic considerations. The window type models have a major
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
advantage of phenomenal energy saving compared to the centralized systems. The effect of the locations of the airconditioners on the airflow patterns and the contaminant distributions were analyzed in the present study. 3. Investigations in operation theatres Experiments were conducted in 10 hospital operation theatres during the time of surgery. Parameters such as CO2concentration of the natural air in the room, temperature, and air humidity were measured in the operating room over time. CO2 measuring instrument used in this study is calibrated for uncertainty and has an accuracy of <30 ppm CO2. These parameters were measured at three different planes (T-top plane of measurement, C-center plane of measurement and B-bottom plane of measurement), equidistance of 0.5 m between them are shown in Fig. 1. The center plane was assumed to be exactly at the height of the operating table (1.1 m high from the floor), one above and one below it. Four measuring points were located in each plane to the right, left, head and the leg portions of the patient placed on the table. The measuring points along the length of the operating table were at a distance of 1 m away on either side on the mid-lengths (1 m) of the operating table. The measuring points along the width were 0.5 m away on either side and lie on the mid-widths (0.75 m) of the operating table [21,23]. The operating table has a dimension of height length width as 1.1 m 2 m 1.5 m, respectively. Measurements were recorded for every 5 min. The surgery in most of the cases lasted for around 30 min. The dimensions of the operating room (L H W) chosen for the study was 8.8 m 3 m 4 m. Three air-conditioning units were installed in the theatre. The inlets of the air-conditioning units were 0.10 m 0.2 m in size, each located at the rear wall as shown in Fig. 1 with a centroid of (2.1, 0, 2.0), (0.4, 0, 2.0) and (2.9, 0, 2.0) for the left, center and right air-conditioning units, respectively. The window type models have the exhaust (E) and inlet (I) side by side as shown in Fig. 1. The measured average velocity of the jet coming out of the air-conditioning units is found to be 5 m s1 in all the three units. The location of the operation table is made to be exactly at the center of the operating room. The room model volume is descretized into hybrid cells of size 0.25 units each. The boundary conditions used in the analysis are the velocity, turbulent kinetic energy, k = 1.5(Uinlet TI) and turbulence dissipation rate, ***e =
Fig. 1. Room model with three measurement planes. Length along the x-axis, Height along the y-axis and width along the z-axis.
233
(Cm0.75k1.5)/l. The no-slip boundary condition was used for the walls. The wall temperatures were taken to be 300 K. The boundary conditions for the human models considered in the study were temperature of 300 K. The room air was initially patched with an initial CO2 concentration of 400 ppm compared to the room inlet air concentration of 480 ppm. 3.1. Modeling and simulation Airflow modeling based on computational fluid dynamics solves the fundamental conservation equation for mass, momentum and energy in the form of Navier–Stokes equation, the general form of which is: @ðrfÞ þ divðrVf G f grad fÞ ¼ Sf @t
(1)
The Navier–Stokes equations and the scalar equations for turbulence and CO2-dispersion, etc. are solved by superimposing a grid of cells that describe the physical room geometry. The room geometry was discretized into a nonuniform grid of hybrid cells (cells of tetrahedral, triangular and hexahedral cells). The geometry was divided into three zones; the center zone, which is the occupied zone is refined with smaller cells when compared to the outer two zones. Zoning is done to make the predictions of the contaminant distribution and airflow simulation more accurate in the experimental zone than the other non-occupied zones. Zoning also initiates faster convergence and reduces computational efforts. Grid dependency tests were performed to ensure that the results were appropriate and would not vary on increasing the grid density. Apart from grid dependency test, the quality of the grid cells is also checked using the equi-angle skew technique (Eq. (2)) umax ueq ueq umin QEAS ¼ max ; (2) 180 ueq ueq where QEAS = 0 describes an equilateral element, and QEAS = 1 describes a completely degenerated (poorly shaped) element. Experimental and numerical simulation studies have been reported by Kameel and Khalil [11,12] on the effect of turbulence, velocity and humidity on airflow patterns in conventional down flow type operation theatres. The current work is based on the finite volume approach to consider the discretistaion and the solution of the equations. PISO (pressure implicit with split operators) algorithm was used solve the pressure-correction equations [13,14] and k–e model was used to solve the turbulence viscosity [15]. Although the Gauss–Seidel scheme rapidly removes local (high-frequency) errors in the solution, global (low-frequency) errors are reduced at a rate inversely related to the grid size. Thus, for a large number of nodes, the solver ‘‘stalls’’ and the residual reduction rate becomes prohibitively low. Multigrid scheme used in the code accelerates the convergence of the solver by computing corrections on a series of coarse grid levels. The use of the multigrid scheme greatly reduces the number of iterations and the CPU time required to obtain a converged solution, particularly since the model contains a large number of control volumes.
234
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
3.2. PISO technique PISO (pressure implicit with split operators) was used for the transient calculations on the highly skewed meshes. PISO maintained a stable calculation with a time step size of 0.1 s and an under-relaxation factor of 1.0 combining both momentum and pressure. For steady-state indoor airflow simulation problems, PISO does not provide any noticeable advantage over SIMPLE or SIMPLEC with optimal under-relaxation factors. The PISO algorithm takes a little more CPU time per solver iteration, but it can dramatically decrease the number of iterations required for convergence, especially for transient case simulation. One of the limitations of the SIMPLE and SIMPLEC algorithms is that new velocities and corresponding fluxes do not satisfy the momentum balance after the pressurecorrection equation is solved. As a result, the calculation must be repeated until the balance is satisfied. To improve the efficiency of this calculation, the PISO algorithm is used. In PISO algorithm, momentum equations are first solved as a ‘predictor’ tentatively using the pressure field found in the preceding iteration or time step. The continuity equation, rewritten as the divergence of the momentum equations is then employed as the corrector. These results in a Poisson equation for the pressure using which pressure–velocity coupling can be achieved. Results showed that PISO is more efficient in timedependent calculations. The numerical calculation of the flow was based on an Eulerian approach, which used the concept of phasic volume
fraction. Different components such as air and CO2 are treated mathematically as interpenetrating continua. Volume fractions are assumed to be continuous functions of space and time. The mixture model uses a single-fluid approach. It allows the phases to be interpenetrating. The volume fractions aq andap for a control volume can therefore be equal to any value between 0 and 1, depending on the space occupied by phase ‘q’ and phase ‘p’. The room airflow simulation used in this study involves phases moving with a same velocity making the slip velocity to be zero. The basic equations used in the model involve an additional feature of volume fraction added to them. The velocity, density, viscosity, etc. are mass averaged for the mixture. The equation for the volume fraction of the secondary phases is given by [19]: @ðap rp Þ þ divðap rp vm Þ ¼ divðap rp vdr;p Þ @t
(3)
The time discretization was done using a first-order implicit method. The transient simulations were done by defining a nondimensional time step [20]. The non-dimensional time step (t0 ) was chosen based on inlet air velocity, height of the room and a physical time step. The formula is t0 ¼
U inlet t H room
(4)
At rest, the normal adult inhales between 0.10 and 0.12 l/s of air and of this only about some 5% are absorbed as oxygen by
Fig. 2. Room models considered for the study.
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
lungs. The exhaled breath contains between 3 and 4% of carbon dioxide, which amounts to about 0.0004 l/s. The accepted level for a maximum concentration of carbon dioxide within an occupied space is 5000 parts per million (or) 0.5% by volume, for an exposure of 8 h. These data were used as the boundary conditions for the CO2 exhalation in the human models.
Table 1 Horizontal plane profiles
4. CO2-dispersion in the operating room
1
At the floor (Y = 1.5) At the mid-height (Y = 0)
The dimensions of the actual operation theatre in terms of length, breadth and width are 8.8 m 3 m 4 m. The dispersion pattern of CO2 in the operation theatre was analyzed by considering six different combinations of positions of operation table and air-conditioners. The room models considered are shown in Fig. 2. The differences in combinations are in (i) the location of the air-conditioning units, (ii) the location of the operation table and (iii) the position of the theatre staff in the operation theatre. In the first three cases (Case 1, Case 2 and Case 3 in Fig. 2) the operating table was positioned perpendicular to the rear and front wall. In Case 1, the air-conditioning units were placed perpendicular to the rear wall with the centroid as mentioned previously. The staffs were positioned three on either side of the table facing the side walls. The patient was placed on the table as shown in all the cases. In Case 2, the air-conditioning units were shifted towards the floor by a distance of 0.5 m without changing the positions of the operating table and the theatre staff. In Case 3, the airconditioning units were shifted towards the ceiling by a distance of 0.5 m from the position in Case 1, without changing the positions of the operating table and the theatre staff. The other three options considered are shown in Case 4, Case 5 and Case 6. In Case 4, the operation theatre staffs were made to face the front and rear wall, three on either side of the table, placed parallel to the front and rear walls. The airconditioning units were placed similar to Case 1. In Case 5, the air-conditioning units were moved towards the floor by a distance of 0.5 m maintaining the position of the table and the staffs as in Case 4. In Case 6, the air-conditioning units were moved towards the ceiling by a distance of 0.5 m maintaining the position of the table and the staffs as in Case 3.
2
Case study no.
Area weighted average values for horizontal planes at the floor level (Y = 1.5 at XZ plane) and at the mid-height (Y = 0 at XZ plane) of the rooms (Ref. Fig. 1) Velocity (m/s)
Turbulence intensity (%)
CO2 ratio
0.232 0.422
3.73 12.7
0.474 0.488
At the floor (Y = 1.5) At the mid-height (Y = 0)
0.376 0.222
5.9 7.1
0.452 0.478
3
At the floor (Y = 1.5) At the mid-height (Y = 0)
0.247 0.200
3.3 7.7
0.469 0.475
4
At the floor (Y = 1.5) At the mid-height (Y = 0)
0.242 0.444
4.9 12.6
0.494 0.498
5
At the floor (Y = 1.5) At the mid-height (Y = 0)
0.376 0.252
5.8 7.8
0.452 0.482
6
At the floor (Y = 1.5) At the mid-height (Y = 0)
0.297 0.240
3.8 7.9
0.471 0.478
air-conditioning units near the ceiling, table perpendicular to the front and rear walls and three theatre staffs on either side of the table facing the side walls showed a better dispersion of CO2 in the occupied zone (Ref. Tables 1 and 2) The distribution of CO2 in the two planes (vertical and horizontal) is shown for some of the cases in Figs. 3–9. A comparison was made between the cases, with the theatre staff and the table parallel and perpendicular to the rear wall. The case with table perpendicular to the front and rear wall was more efficient in maintaining a low profile (average value of CO2 ratio = 0.475) of CO2 as well as the turbulence intensity (average value of Table 2 Vertical plane profiles Case study no.
Area weighted average values for the vertical planes (YZ-planes) along the A/C inlets (Ref. Fig. 1) Velocity (m/s)
Turbulence Intensity (%)
CO2 ratio
1
Left A/C (X = 2.096) Mid A/C (X = 0.4) Right A/C (X = 2.896)
0.440 0.458 0.332
12.2 11.4 11.2
0.481 0.495 0.422
2
Left A/C (X = 2.096) Mid A/C (X = 0.4) Right A/C (X = 2.896)
0.412 0.417 0.374
12.3 9.4 11
0.459 0.504 0.466
3
Left A/C (X = 2.096) Mid A/C (X = 0.4) Right A/C (X = 2.896)
0.37 0.445 0.374
11.2 11.8 11.1
0.472 0.493 0.453
4
Left A/C (X = 2.096) Mid A/C (X = 0.4) Right A/C (X = 2.896)
0.454 0.468 0.389
12.3 11.9 11.6
0.498 0.498 0.482
5
Left A/C (X = 2.096) Mid A/C (X = 0.4) Right A/C (X = 2.896)
0.442 0.432 0.396
12.5 10.4 11.3
0.461 0.518 0.469
6
Left A/C (X = 2.096) Mid A/C (X = 0.4) Right A/C (X = 2.896)
0.39 0.495 0.374
11.2 11.8 11.1
0.478 0.498 0.458
4.1. Results and discussion These models discussed in Section 4 were used in the simulation studies. After simulating the airflow patterns for the required time, planes were created to analyze the outputs obtained. Vertical planes were created at the planes corresponding to the inlets (left air-conditioner, X = 2.096 m, center air-conditioner, X = 0.4 m and right air-conditioner, X = 2.896 m) of the air-conditioners (I) (Ref. Fig. 1) and horizontal planes were created at the planes (T, C and B) as shown in Fig. 1, in the mid-height (Y = 0 m) and at the floor level (Y = 1.5 m) in all the room models. The results of the planes T, C, B were used in the comparison with the experimental/measured results. The other two planes were used in estimating the area weighted average values of velocity, CO2 ratio and turbulence intensity. The room model with
235
236
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
Fig. 6. Carbon dioxide distribution in the horizontal center plane (Y = 0) for the room model with air conditioning units located near the floor of the operating room as in CASE2. Fig. 3. Carbon dioxide distribution in the center plane (X = 0.4) for the room model with air conditioning units located near the floor, Case 2.
Fig. 7. Carbon dioxide distribution in the horizontal center plane (Y = 0) for the room model with air conditioning units located near the ceiling of the operating room as in CASE3.
Fig. 4. Carbon dioxide distribution in the center plane (X = 0.4) for the room model with air conditioning units located near the ceiling, CASE3.
Fig. 5. Carbon dioxide distribution in the center plane (X = 0.4) for the room model with air conditioning units located at the center as in the operating room, CASE1.
turbulence intensity = 7.7%) in the occupied zone. The airconditioning units placed at the center and near the floor (Cases 1, 2, 4 and 5) produced a high level of turbulence (average turbulence intensity value = 12.5%) around the operating table enhancing the CO2 dispersion and the distribution of other contaminants.
Fig. 8. Carbon dioxide distribution in the horizontal center plane (Y = 0) for the room model with air conditioning units located at the center of the wall in the operating room as in CASE1.
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
Fig. 9. Carbon dioxide distribution in the horizontal center plane (Y = 0) for the room model with air conditioning units located near the ceiling of the operating room as in CASE6.
Experimental studies were conducted in 10 different hospital operation theatres equipped with window-type air-conditioning units as mentioned earlier. The variation of CO2 ratio with the time ratio is shown for 10 different hospital operation theatres in Figs. 10 and 11. The CO2 ratio is defined as the ratio of actual carbon dioxide in the operation theatre at a particular instant to the threshold limit value (TLV) of carbon dioxide for indoor environments. The time ratio is defined as the ratio of time lapsed from the start of surgery to the total time of surgery. If the curves reach a value of 1.0 in the y-axis the particular case can lead to post-operative problems. In the case studies conducted in all the 10 hospital operation theatres four hospital operation theatres tend to reach a value of 1.0 in the y-axis (Figs. 10 and 11). These hospital operation theatres are highly prone to post-operative wound infection problems in addition to poor air quality and thermal comfort.
237
The CO2 dispersion profile in Fig. 3 (Case 2) shows regions of turbulent dispersion above the operation table. This may result in microorganisms and other contaminants being transported into the patient’s body during surgery. Area weighted average values of the CO2 concentrations, velocity and turbulence intensity were estimated in vertical and horizontal planes created in all the models. The vertical planes (i.e. in the Y–Z plane) and the horizontal planes (X–Z plane) were created as mentioned previously. Comparing the dispersion level in the vertical planes, Fig. 5 (Case 1) is better than Fig. 4. (Case 3). But comparing the dispersion levels in the horizontal planes, Case 3 is better in terms of turbulent CO2 dispersion than Case 1 and Case 2. In Fig. 9 the CO2 dispersion profiles (horizontal plane) is shown for the Case 6 where the turbulent dispersion is more when compared to Cases 1–3. Placing the length of the table parallel to the jet direction seems to be a better case than placing it perpendicular. The dispersion region almost covers the entire staff of the operation theatre with the patient. Case 3 seems to be a better option than all the others. There are other reasons for recommending the case of air-conditioning units located near the ceiling with the length of the operation table parallel to jet direction (Case 3). The reasons are highlighted below (i) The swirling/recirculation motion which pops-up the dust and microorganisms from the floor is minimum (turbulence intensity = 3.34%, where as for the case of airconditioning units located near the floor it is 5.88%). (ii) The direct impact of the jet causes over-cooling and drying of the exposed internal tissues of the patient and on the operation theatre staff at velocities as high as 0.6 m s1 which is minimized in the case of air-conditioning units located near the ceiling (0.2 m s1). (iii) The positioning of the air-conditioning units one beside the other along a straight line as shown in all the room models also help in maintaining a uniform distribution of air above the operation table. (iv) Velocity contours show the presence of dead zones in the occupied region, which are minimum in the case of airconditioning units located near the ceiling. These are important in determining the thermal comfort and contaminant levels.
Fig. 10. CO2 ratio vs. time ratio in operation theatres.
Fig. 11. CO2 ratio vs. time ratio in operation theatres.
Fig. 12. Comparison of experimental results with simulated results for the center plane.
238
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
distribution when compared to the other cases. It reached a maximum concentration of 490 ppm (CO2 ratio = 0.490) after a period of 30 min with the room initial concentration as low as 350 ppm (CO2 ratio = 0.350) 5. Conclusions
Fig. 13. Comparison of experimental results with simulated results for the bottom plane.
With respect to the position of the air-conditioning units, closer the location of the air-conditioning units to the floor, much higher is the contamination accumulation near the table in the operation theatre. This is based on the fact that the CO2 dispersion profiles can be used as a basic indicator of indoor air quality. CO2 concentration can also be an indirect indicator of the occurrence of post-operative wound infections. With respect to the position of operating table, the table may be positioned in such a way that it is free from the main stream jet by a distance of 0.5 m minimum on all sides. This result in reduced turbulence around the table, in situations where the above condition cannot be fulfilled, placing the air-conditioning unit near the ceiling is recommended. The theatre staffs should also be away from the main stream jet, placing the air-conditioning units above 2 m height will be helpful. It was also observed that the location of the operation theatre in terms of height from the ground level, the presence of trees outside the theatre nearer to the air-conditioning units, number of persons in the operation theatre and the equipments present in the operation theatre have influence on the CO2 dispersion in the room.
Fig. 14. Comparison of experimental results with simulated results for the top plane.
References
The experimental values of CO2 dispersion are compared with the simulated results as shown in Figs. 12–14. The experimental results are found to be higher than the simulated values. The factors that contribute to the variation are (a) local effects such as a pollutant source located near or anywhere else in the operating room or the movement of personnel in the room, (b) buoyancy effects due to the dense or hot gases that result due to the presence of heat sources such as operating lamps, etc. and (c) short-circuiting between the ventilation systems supply and exhaust. The coanda effect also plays a role in CO2 dispersion. Since the air-conditioning units (for the case under consideration) are facing the width of the operation theatre which is comparatively smaller than the length of the operation theatre. The coanda effect tends to make the jet adhere to the wall along the ceiling and more towards the opposite wall. Only the cooling effects will be felt by the operation theatre staff. In the study conducted in 10 hospitals it was noted that only two hospitals had their theatre located at the ground floor. Six hospitals had their theatre located in the first floor and two hospitals had their theatre in the second floor. Among the two hospitals which had their theatre located in the second floor one was surrounded by trees and the number of theatre staff was considerably low (six including the patient and the person measuring the airflow parameters) when compared to the other cases. This particular case seemed to be better in terms of CO2
[1] R. Kameel, E.E. Khalil, Energy efficient thermal comfort Vs air quality in air-conditioned health care applications, Building Simulation (2003) 621– 626. [2] Draft Guideline for environmental infection control in health care facilities, 2001, Centre for disease control and prevention—Health care Infection Control Practices Advisory Committee (HICPAC). [3] G. Einberg, Air diffusion and solid contaminant behavior in room ventilation—a CFD based integrated approach, Ph.D. Thesis, Department of Energy Technology, The Royal Institute of Technology, Sweden, April 2005. [4] U. Moscato, T. Esposito, G. Vanini, Nitrous oxide pollution: a geostatical method to assess spatial distribution of anesthetic gases and hospital staff exposure, Proceedings of Healthy Buildings 1 (2000) 487–492. [5] G. Mehta, et al., Microbiological surveillance of operation theatres, 2000. http://www.orthoteers.org/content/content.aspx?Article=372. [6] W. Lu, A. Howarth, N. Adam, S. Riffat, Modeling and measurement of airflow and aerosol particle distribution in a ventilated two-zone chamber, Building and Environment 31 (1996) 417–423. [7] L. Martin, The four most important equations in clinical practice, the clinical ultimate reference. www.globalrph.com. [8] K. Singh, Small Incision Cataract Surgery, Alpha Science Int’l Ltd., 2003. [9] Q. Chen, Computational fluid dynamics for HVAC: successes and failures, ASHRAE Transactions 103 (1) (1997) 178–187. [10] J.F. Pfost, A re-evaluation of laminar airflow in hospital operating rooms, ASHRAE Transactions 87 (2) (1981) 729–739. [11] R. Kameel, E.E. Khalil, Experimental and numerical investigations of airflow regimes in air-conditioned operation theatres, Building Simulation (2003). [12] R. Kameel, E.E. Khalil, Experimental and numerical investigations of heat transfer in air conditioned operation theatres, Building Simulation (2003). [13] B.E. Launder, D.B. Spalding, The numerical computation of turbulent flows, Method in Applied Mechanical Engineering 87 (2) (1974) 729–739.
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239 [14] Patankar, Revised algorithm for pressure velocity variation, 1980. [15] P.V. Nielsen, The selection of Turbulence models for prediction of room air flow, ASHRAE Transactions: Symposia (SF-98-10-1) (1999) 1119– 1128. [16] L.-M. Lo, Numerical studies of airflow movement and contaminant transport in hospital operating rooms, MSc Thesis, University of Minnesota, 1997. [17] F. Memarzadeh, A. Manning, Comparison of operating room ventilation systems in the protection of the surgical site, ASHRAE Transactions 108 (22) (2002) 3–15.
239
[18] F. Memarzadeh, A. Manning, Reducing risks of surgery, ASHRAE Journal (2003) 28–33. [19] FLUENT’S User Guide Version 6.1, Fluent Incorporated, Lebanon, NH, 2001. [20] Z.F. Tian, J.Y. Tu, G.H. Yeih, Large eddy simulation indoor airflow and contaminant concentration, Building Simulation (2003). [21] ANSI/ASHRAE 113-1990, Method of testing for room air diffusion. [22] ANSI/ASHRAE 62-1989, Acceptable ventilation for indoor air quality. [23] ANSI/ASHRAE 55-1992, Thermal environmental conditions for human occupancy.
CO2-dispersion studies in an operation theatre under transient conditions C.P. Karthikeyan, Anand A. Samuel * School of Mechanical and Building Sciences, Vellore Institute of Technology University Vellore 632014, Tamil Nadu, India Received 4 January 2007; received in revised form 30 January 2007; accepted 17 February 2007
Abstract Maintaining air quality and thermal comfort inside an operation theatre equipped with horizontal jet flow type air-conditioning units, has been a challenge to engineers. The objective of this study is to analyze the airflow pattern in such operation theatres and the influence of location of the airconditioners. The outcome of this study is expected to reduce the post-operation problems faced due to excess concentration of contaminants. Experimental studies were conducted in 10 different hospital operation theatres. Parameters such as air temperature and carbondioxide concentration were measured at discrete points chosen in the theatre. A 3D time-dependent numerical model was developed to simulate the airflow in terms of parameters such as velocity, temperature and CO2 distributions in an operation theatre under transient conditions. The Eulerian approach using the volume fraction of the mixture of air and CO2 was used to solve the numerical model. Finite volume approach was attempted in this work with PISO (pressure implicit with split operators) algorithm for the pressure correction equations. The simulated results were compared with the experimental results for validity. The locations of the air-conditioners were changed in the numerical model to analyze the airflow patterns and the contaminant distribution. # 2007 Elsevier B.V. All rights reserved. Keywords: Operation theatres; Airflow patterns; CO2-dispersion; CFD; PISO and transient conditions
1. Introduction Indoor air quality is a critical factor in the design of any hospital operation theatre. Studies of operating room air distribution indicate that the supply of air from the ceiling, with a downward flow movement to several ports located on opposite wall yielded probably, the most effective airflow pattern, maintaining the concentration of contaminants to a minimum acceptable level [1]. On account of ease of handling and economic considerations, window type air-conditioning units are used in many operation theatres. Window type units produce horizontal jet flow in contrast to the conventional downward flows. Simulation of airflow patterns for different locations of the air-conditioning units using computational fluid dynamics techniques help in analyzing air quality and thermal comfort. The variation in contaminant distribution, location of dead zones and air velocity in the operation theatre are the major factors that indicate the occurrence of post-operative wound infections.
Major contaminants in any hospital operation theatre are categorized as biological (microorganisms—fungi, bacteria and virus), chemical (waste anesthetic gases, CO2, etc.) and particulate matter [2]. Microorganisms are the major cause for ‘post-operative wound infections’. There are three main sources of contaminants: (a) the supply air to the theatre, (b) the surgical team members, and (c) the emissions of body vapors and organisms from the pathogens, these sources on mixing with the supply air gets deposited on the patient cause infections. The contaminants such as waste anesthetic gases and carbon dioxide, if not controlled properly can lead to severe complications. Most of the studies in the literature focus on air diffusion and solid contaminant behavior in indoor environments using CFD based integrated approach [3]. Studies on the distribution of CO2 in the operation theatres have not been reported in the literature. An attempt is made here to study the CO2 dispersion in an operation theatre using which the air quality is analyzed. 2. Indoor air quality in operation theatres
* Corresponding author. Tel.: +91 416 2243091/93; fax: +91 416 2243092. E-mail addresses: [email protected] (C.P. Karthikeyan), [email protected] (A.A. Samuel). 0378-7788/$ – see front matter # 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.enbuild.2007.02.023
CO2 concentration has been widely used as an indicator of indoor air quality. A Limit of 1000 ppm of CO2 is
232
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
Nomenclature Cm
constant used in turbulence (k–e) model, Cm = 0.09 CO2 carbon dioxide (ppm—parts per million) Hroom height of the room (m) K turbulent kinetic energy (m2 s2) l = 0.07L
QEAS S t0 TI Uinlet V vdr;p vm
l is the characteristic length at the inlet (m) quality of grid cells by Equi-Angle Skew technique source or sink non-dimensional time-step (s) turbulent intensity (%) the inlet velocity (m s1) velocity vector drift velocity of the primary phase (m s1) mass averaged velocity (m s1)
Greek letters ap volume fraction of the primary phase Gf exchange coefficient (laminar + turbulent) e turbulence dissipation rate (m2 s3) ueq characteristic angle corresponding to an equilateral cell of similar form umax maximum angle between the edges of the element in degrees umin minimum angle between the edges of the element in degrees f dependent variable r density (kg m3) rp density of the primary phase (kg m3) recommended to satisfy comfort criteria and CO2 concentration is also an useful indicator of insufficient outside air intake and ventilation problems [22]. In this study CO2-dispersion is used in predicting the regions of high contamination in the operation theatre. The patients in hospital isolation rooms constantly produce transmissible air-borne organisms by coughing, sneezing or talking, which if not under control, results in spreading air-borne infection. The air supply to the operation theatre should normally be free from such air-borne infectious sources. CO2 being a common gas and a basic indicator of air quality, is used for simulation studies, the numerical model developed can also be used for the simulation of any other waste anesthetic gas in theatres. Hypercapnia is a condition that is abnormally caused by shallow respiration or hypoventilation. It is an excessive amount of CO2 in the blood. Very high concentrations of atmospheric CO2 results in hypercapnia [7]. Studies have shown that there are numerous health effects associated with the exposure of waste anesthetic gases [4]. These include increased risk of spontaneous abortion in females exposed to anesthetic gases in hospitals with incidences of 1.5– 2 times greater than in unexposed females. The prevention of post-operative infection is dependent on several factors
including effective sterilization and disinfection procedures, good surgical technique, theatre design, bacterial contamination of theatre air, discipline—which include restricting the movement of staff near the operating table, appropriate use of prophylactic antibiotics, etc. [5]. It has been found that velocities as high as 0.6 m s1 are unsatisfactory because the exposed internal tissues of the patient are over-cooled and dried out by evaporation [6]. Parameters such as air temperature and humidity also play a vital role in surgeries. If the operation room temperature is higher during the eye surgeries, then the cornea dries out and laser beam removes more corneal tissue and an over-correction can result. The reverse occurs if the temperature in the operating theatre is too low. If the humidity is higher than the cornea is wetter and the laser removes less tissue with each pulse, increasing the chance of an under-correction [8]. These factors prove the need for the study and its significance. 2.1. Operation theatre indoor airflow and contaminant distribution simulation Airflow simulation using CFD techniques is an useful tool to analyze air movements in non-standard operating room situations [9]. Modern computational fluid dynamics models are used to predict the air velocity, temperature, turbulence level and contaminant distributions. The current work uses a transient model based on eulerian approach for simulation of indoor airflow and contaminant distribution. In 1997, Lo [16] has made a study addressing contamination control in an operating room. The study considered an operating room under isothermal condition and analyzed the distribution of contaminants. In a first assumption made in the study, the effect of significant thermal plumes in the room was ignored. In a second assumption it was considered that the particles in the room can be considered to follow Brownian motion. These assumptions were reconsidered by Memarzadeh and Manning [18] in their study conducted in reducing the risks in surgery. The basic assumption made in their study was that the squames were simulated as particles being released from several sources surrounding the occupant. These particles were then tracked for a certain period of time. The Brownian motion is strictly applicable to particles that are 1 mm or less in diameter. Bacteria and virus do conform to this criteria but bacteria are usually transported in operating rooms by squames [17], which are considerably bigger (in the range of 10 mm) and so do not necessarily follow Brownian motion. Numerous studies on the simulation of particle trajectories in rooms are reported in the literature but much little is focused on the of gas dispersion, hence this study. Studies conducted on airflow regimes in air-conditioned operating theatres suggest that the supply of air from the ceiling, with a downward flow movement to several exhaust ports located on opposite walls yielded, probably the most effective airflow patterns maintaining the contaminant concentrations to a minimum acceptable level [10–12]. In contrast this study focuses on the window type air-conditioning units, which are used on account of ease of handling and economic considerations. The window type models have a major
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
advantage of phenomenal energy saving compared to the centralized systems. The effect of the locations of the airconditioners on the airflow patterns and the contaminant distributions were analyzed in the present study. 3. Investigations in operation theatres Experiments were conducted in 10 hospital operation theatres during the time of surgery. Parameters such as CO2concentration of the natural air in the room, temperature, and air humidity were measured in the operating room over time. CO2 measuring instrument used in this study is calibrated for uncertainty and has an accuracy of <30 ppm CO2. These parameters were measured at three different planes (T-top plane of measurement, C-center plane of measurement and B-bottom plane of measurement), equidistance of 0.5 m between them are shown in Fig. 1. The center plane was assumed to be exactly at the height of the operating table (1.1 m high from the floor), one above and one below it. Four measuring points were located in each plane to the right, left, head and the leg portions of the patient placed on the table. The measuring points along the length of the operating table were at a distance of 1 m away on either side on the mid-lengths (1 m) of the operating table. The measuring points along the width were 0.5 m away on either side and lie on the mid-widths (0.75 m) of the operating table [21,23]. The operating table has a dimension of height length width as 1.1 m 2 m 1.5 m, respectively. Measurements were recorded for every 5 min. The surgery in most of the cases lasted for around 30 min. The dimensions of the operating room (L H W) chosen for the study was 8.8 m 3 m 4 m. Three air-conditioning units were installed in the theatre. The inlets of the air-conditioning units were 0.10 m 0.2 m in size, each located at the rear wall as shown in Fig. 1 with a centroid of (2.1, 0, 2.0), (0.4, 0, 2.0) and (2.9, 0, 2.0) for the left, center and right air-conditioning units, respectively. The window type models have the exhaust (E) and inlet (I) side by side as shown in Fig. 1. The measured average velocity of the jet coming out of the air-conditioning units is found to be 5 m s1 in all the three units. The location of the operation table is made to be exactly at the center of the operating room. The room model volume is descretized into hybrid cells of size 0.25 units each. The boundary conditions used in the analysis are the velocity, turbulent kinetic energy, k = 1.5(Uinlet TI) and turbulence dissipation rate, ***e =
Fig. 1. Room model with three measurement planes. Length along the x-axis, Height along the y-axis and width along the z-axis.
233
(Cm0.75k1.5)/l. The no-slip boundary condition was used for the walls. The wall temperatures were taken to be 300 K. The boundary conditions for the human models considered in the study were temperature of 300 K. The room air was initially patched with an initial CO2 concentration of 400 ppm compared to the room inlet air concentration of 480 ppm. 3.1. Modeling and simulation Airflow modeling based on computational fluid dynamics solves the fundamental conservation equation for mass, momentum and energy in the form of Navier–Stokes equation, the general form of which is: @ðrfÞ þ divðrVf G f grad fÞ ¼ Sf @t
(1)
The Navier–Stokes equations and the scalar equations for turbulence and CO2-dispersion, etc. are solved by superimposing a grid of cells that describe the physical room geometry. The room geometry was discretized into a nonuniform grid of hybrid cells (cells of tetrahedral, triangular and hexahedral cells). The geometry was divided into three zones; the center zone, which is the occupied zone is refined with smaller cells when compared to the outer two zones. Zoning is done to make the predictions of the contaminant distribution and airflow simulation more accurate in the experimental zone than the other non-occupied zones. Zoning also initiates faster convergence and reduces computational efforts. Grid dependency tests were performed to ensure that the results were appropriate and would not vary on increasing the grid density. Apart from grid dependency test, the quality of the grid cells is also checked using the equi-angle skew technique (Eq. (2)) umax ueq ueq umin QEAS ¼ max ; (2) 180 ueq ueq where QEAS = 0 describes an equilateral element, and QEAS = 1 describes a completely degenerated (poorly shaped) element. Experimental and numerical simulation studies have been reported by Kameel and Khalil [11,12] on the effect of turbulence, velocity and humidity on airflow patterns in conventional down flow type operation theatres. The current work is based on the finite volume approach to consider the discretistaion and the solution of the equations. PISO (pressure implicit with split operators) algorithm was used solve the pressure-correction equations [13,14] and k–e model was used to solve the turbulence viscosity [15]. Although the Gauss–Seidel scheme rapidly removes local (high-frequency) errors in the solution, global (low-frequency) errors are reduced at a rate inversely related to the grid size. Thus, for a large number of nodes, the solver ‘‘stalls’’ and the residual reduction rate becomes prohibitively low. Multigrid scheme used in the code accelerates the convergence of the solver by computing corrections on a series of coarse grid levels. The use of the multigrid scheme greatly reduces the number of iterations and the CPU time required to obtain a converged solution, particularly since the model contains a large number of control volumes.
234
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
3.2. PISO technique PISO (pressure implicit with split operators) was used for the transient calculations on the highly skewed meshes. PISO maintained a stable calculation with a time step size of 0.1 s and an under-relaxation factor of 1.0 combining both momentum and pressure. For steady-state indoor airflow simulation problems, PISO does not provide any noticeable advantage over SIMPLE or SIMPLEC with optimal under-relaxation factors. The PISO algorithm takes a little more CPU time per solver iteration, but it can dramatically decrease the number of iterations required for convergence, especially for transient case simulation. One of the limitations of the SIMPLE and SIMPLEC algorithms is that new velocities and corresponding fluxes do not satisfy the momentum balance after the pressurecorrection equation is solved. As a result, the calculation must be repeated until the balance is satisfied. To improve the efficiency of this calculation, the PISO algorithm is used. In PISO algorithm, momentum equations are first solved as a ‘predictor’ tentatively using the pressure field found in the preceding iteration or time step. The continuity equation, rewritten as the divergence of the momentum equations is then employed as the corrector. These results in a Poisson equation for the pressure using which pressure–velocity coupling can be achieved. Results showed that PISO is more efficient in timedependent calculations. The numerical calculation of the flow was based on an Eulerian approach, which used the concept of phasic volume
fraction. Different components such as air and CO2 are treated mathematically as interpenetrating continua. Volume fractions are assumed to be continuous functions of space and time. The mixture model uses a single-fluid approach. It allows the phases to be interpenetrating. The volume fractions aq andap for a control volume can therefore be equal to any value between 0 and 1, depending on the space occupied by phase ‘q’ and phase ‘p’. The room airflow simulation used in this study involves phases moving with a same velocity making the slip velocity to be zero. The basic equations used in the model involve an additional feature of volume fraction added to them. The velocity, density, viscosity, etc. are mass averaged for the mixture. The equation for the volume fraction of the secondary phases is given by [19]: @ðap rp Þ þ divðap rp vm Þ ¼ divðap rp vdr;p Þ @t
(3)
The time discretization was done using a first-order implicit method. The transient simulations were done by defining a nondimensional time step [20]. The non-dimensional time step (t0 ) was chosen based on inlet air velocity, height of the room and a physical time step. The formula is t0 ¼
U inlet t H room
(4)
At rest, the normal adult inhales between 0.10 and 0.12 l/s of air and of this only about some 5% are absorbed as oxygen by
Fig. 2. Room models considered for the study.
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
lungs. The exhaled breath contains between 3 and 4% of carbon dioxide, which amounts to about 0.0004 l/s. The accepted level for a maximum concentration of carbon dioxide within an occupied space is 5000 parts per million (or) 0.5% by volume, for an exposure of 8 h. These data were used as the boundary conditions for the CO2 exhalation in the human models.
Table 1 Horizontal plane profiles
4. CO2-dispersion in the operating room
1
At the floor (Y = 1.5) At the mid-height (Y = 0)
The dimensions of the actual operation theatre in terms of length, breadth and width are 8.8 m 3 m 4 m. The dispersion pattern of CO2 in the operation theatre was analyzed by considering six different combinations of positions of operation table and air-conditioners. The room models considered are shown in Fig. 2. The differences in combinations are in (i) the location of the air-conditioning units, (ii) the location of the operation table and (iii) the position of the theatre staff in the operation theatre. In the first three cases (Case 1, Case 2 and Case 3 in Fig. 2) the operating table was positioned perpendicular to the rear and front wall. In Case 1, the air-conditioning units were placed perpendicular to the rear wall with the centroid as mentioned previously. The staffs were positioned three on either side of the table facing the side walls. The patient was placed on the table as shown in all the cases. In Case 2, the air-conditioning units were shifted towards the floor by a distance of 0.5 m without changing the positions of the operating table and the theatre staff. In Case 3, the airconditioning units were shifted towards the ceiling by a distance of 0.5 m from the position in Case 1, without changing the positions of the operating table and the theatre staff. The other three options considered are shown in Case 4, Case 5 and Case 6. In Case 4, the operation theatre staffs were made to face the front and rear wall, three on either side of the table, placed parallel to the front and rear walls. The airconditioning units were placed similar to Case 1. In Case 5, the air-conditioning units were moved towards the floor by a distance of 0.5 m maintaining the position of the table and the staffs as in Case 4. In Case 6, the air-conditioning units were moved towards the ceiling by a distance of 0.5 m maintaining the position of the table and the staffs as in Case 3.
2
Case study no.
Area weighted average values for horizontal planes at the floor level (Y = 1.5 at XZ plane) and at the mid-height (Y = 0 at XZ plane) of the rooms (Ref. Fig. 1) Velocity (m/s)
Turbulence intensity (%)
CO2 ratio
0.232 0.422
3.73 12.7
0.474 0.488
At the floor (Y = 1.5) At the mid-height (Y = 0)
0.376 0.222
5.9 7.1
0.452 0.478
3
At the floor (Y = 1.5) At the mid-height (Y = 0)
0.247 0.200
3.3 7.7
0.469 0.475
4
At the floor (Y = 1.5) At the mid-height (Y = 0)
0.242 0.444
4.9 12.6
0.494 0.498
5
At the floor (Y = 1.5) At the mid-height (Y = 0)
0.376 0.252
5.8 7.8
0.452 0.482
6
At the floor (Y = 1.5) At the mid-height (Y = 0)
0.297 0.240
3.8 7.9
0.471 0.478
air-conditioning units near the ceiling, table perpendicular to the front and rear walls and three theatre staffs on either side of the table facing the side walls showed a better dispersion of CO2 in the occupied zone (Ref. Tables 1 and 2) The distribution of CO2 in the two planes (vertical and horizontal) is shown for some of the cases in Figs. 3–9. A comparison was made between the cases, with the theatre staff and the table parallel and perpendicular to the rear wall. The case with table perpendicular to the front and rear wall was more efficient in maintaining a low profile (average value of CO2 ratio = 0.475) of CO2 as well as the turbulence intensity (average value of Table 2 Vertical plane profiles Case study no.
Area weighted average values for the vertical planes (YZ-planes) along the A/C inlets (Ref. Fig. 1) Velocity (m/s)
Turbulence Intensity (%)
CO2 ratio
1
Left A/C (X = 2.096) Mid A/C (X = 0.4) Right A/C (X = 2.896)
0.440 0.458 0.332
12.2 11.4 11.2
0.481 0.495 0.422
2
Left A/C (X = 2.096) Mid A/C (X = 0.4) Right A/C (X = 2.896)
0.412 0.417 0.374
12.3 9.4 11
0.459 0.504 0.466
3
Left A/C (X = 2.096) Mid A/C (X = 0.4) Right A/C (X = 2.896)
0.37 0.445 0.374
11.2 11.8 11.1
0.472 0.493 0.453
4
Left A/C (X = 2.096) Mid A/C (X = 0.4) Right A/C (X = 2.896)
0.454 0.468 0.389
12.3 11.9 11.6
0.498 0.498 0.482
5
Left A/C (X = 2.096) Mid A/C (X = 0.4) Right A/C (X = 2.896)
0.442 0.432 0.396
12.5 10.4 11.3
0.461 0.518 0.469
6
Left A/C (X = 2.096) Mid A/C (X = 0.4) Right A/C (X = 2.896)
0.39 0.495 0.374
11.2 11.8 11.1
0.478 0.498 0.458
4.1. Results and discussion These models discussed in Section 4 were used in the simulation studies. After simulating the airflow patterns for the required time, planes were created to analyze the outputs obtained. Vertical planes were created at the planes corresponding to the inlets (left air-conditioner, X = 2.096 m, center air-conditioner, X = 0.4 m and right air-conditioner, X = 2.896 m) of the air-conditioners (I) (Ref. Fig. 1) and horizontal planes were created at the planes (T, C and B) as shown in Fig. 1, in the mid-height (Y = 0 m) and at the floor level (Y = 1.5 m) in all the room models. The results of the planes T, C, B were used in the comparison with the experimental/measured results. The other two planes were used in estimating the area weighted average values of velocity, CO2 ratio and turbulence intensity. The room model with
235
236
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
Fig. 6. Carbon dioxide distribution in the horizontal center plane (Y = 0) for the room model with air conditioning units located near the floor of the operating room as in CASE2. Fig. 3. Carbon dioxide distribution in the center plane (X = 0.4) for the room model with air conditioning units located near the floor, Case 2.
Fig. 7. Carbon dioxide distribution in the horizontal center plane (Y = 0) for the room model with air conditioning units located near the ceiling of the operating room as in CASE3.
Fig. 4. Carbon dioxide distribution in the center plane (X = 0.4) for the room model with air conditioning units located near the ceiling, CASE3.
Fig. 5. Carbon dioxide distribution in the center plane (X = 0.4) for the room model with air conditioning units located at the center as in the operating room, CASE1.
turbulence intensity = 7.7%) in the occupied zone. The airconditioning units placed at the center and near the floor (Cases 1, 2, 4 and 5) produced a high level of turbulence (average turbulence intensity value = 12.5%) around the operating table enhancing the CO2 dispersion and the distribution of other contaminants.
Fig. 8. Carbon dioxide distribution in the horizontal center plane (Y = 0) for the room model with air conditioning units located at the center of the wall in the operating room as in CASE1.
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
Fig. 9. Carbon dioxide distribution in the horizontal center plane (Y = 0) for the room model with air conditioning units located near the ceiling of the operating room as in CASE6.
Experimental studies were conducted in 10 different hospital operation theatres equipped with window-type air-conditioning units as mentioned earlier. The variation of CO2 ratio with the time ratio is shown for 10 different hospital operation theatres in Figs. 10 and 11. The CO2 ratio is defined as the ratio of actual carbon dioxide in the operation theatre at a particular instant to the threshold limit value (TLV) of carbon dioxide for indoor environments. The time ratio is defined as the ratio of time lapsed from the start of surgery to the total time of surgery. If the curves reach a value of 1.0 in the y-axis the particular case can lead to post-operative problems. In the case studies conducted in all the 10 hospital operation theatres four hospital operation theatres tend to reach a value of 1.0 in the y-axis (Figs. 10 and 11). These hospital operation theatres are highly prone to post-operative wound infection problems in addition to poor air quality and thermal comfort.
237
The CO2 dispersion profile in Fig. 3 (Case 2) shows regions of turbulent dispersion above the operation table. This may result in microorganisms and other contaminants being transported into the patient’s body during surgery. Area weighted average values of the CO2 concentrations, velocity and turbulence intensity were estimated in vertical and horizontal planes created in all the models. The vertical planes (i.e. in the Y–Z plane) and the horizontal planes (X–Z plane) were created as mentioned previously. Comparing the dispersion level in the vertical planes, Fig. 5 (Case 1) is better than Fig. 4. (Case 3). But comparing the dispersion levels in the horizontal planes, Case 3 is better in terms of turbulent CO2 dispersion than Case 1 and Case 2. In Fig. 9 the CO2 dispersion profiles (horizontal plane) is shown for the Case 6 where the turbulent dispersion is more when compared to Cases 1–3. Placing the length of the table parallel to the jet direction seems to be a better case than placing it perpendicular. The dispersion region almost covers the entire staff of the operation theatre with the patient. Case 3 seems to be a better option than all the others. There are other reasons for recommending the case of air-conditioning units located near the ceiling with the length of the operation table parallel to jet direction (Case 3). The reasons are highlighted below (i) The swirling/recirculation motion which pops-up the dust and microorganisms from the floor is minimum (turbulence intensity = 3.34%, where as for the case of airconditioning units located near the floor it is 5.88%). (ii) The direct impact of the jet causes over-cooling and drying of the exposed internal tissues of the patient and on the operation theatre staff at velocities as high as 0.6 m s1 which is minimized in the case of air-conditioning units located near the ceiling (0.2 m s1). (iii) The positioning of the air-conditioning units one beside the other along a straight line as shown in all the room models also help in maintaining a uniform distribution of air above the operation table. (iv) Velocity contours show the presence of dead zones in the occupied region, which are minimum in the case of airconditioning units located near the ceiling. These are important in determining the thermal comfort and contaminant levels.
Fig. 10. CO2 ratio vs. time ratio in operation theatres.
Fig. 11. CO2 ratio vs. time ratio in operation theatres.
Fig. 12. Comparison of experimental results with simulated results for the center plane.
238
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239
distribution when compared to the other cases. It reached a maximum concentration of 490 ppm (CO2 ratio = 0.490) after a period of 30 min with the room initial concentration as low as 350 ppm (CO2 ratio = 0.350) 5. Conclusions
Fig. 13. Comparison of experimental results with simulated results for the bottom plane.
With respect to the position of the air-conditioning units, closer the location of the air-conditioning units to the floor, much higher is the contamination accumulation near the table in the operation theatre. This is based on the fact that the CO2 dispersion profiles can be used as a basic indicator of indoor air quality. CO2 concentration can also be an indirect indicator of the occurrence of post-operative wound infections. With respect to the position of operating table, the table may be positioned in such a way that it is free from the main stream jet by a distance of 0.5 m minimum on all sides. This result in reduced turbulence around the table, in situations where the above condition cannot be fulfilled, placing the air-conditioning unit near the ceiling is recommended. The theatre staffs should also be away from the main stream jet, placing the air-conditioning units above 2 m height will be helpful. It was also observed that the location of the operation theatre in terms of height from the ground level, the presence of trees outside the theatre nearer to the air-conditioning units, number of persons in the operation theatre and the equipments present in the operation theatre have influence on the CO2 dispersion in the room.
Fig. 14. Comparison of experimental results with simulated results for the top plane.
References
The experimental values of CO2 dispersion are compared with the simulated results as shown in Figs. 12–14. The experimental results are found to be higher than the simulated values. The factors that contribute to the variation are (a) local effects such as a pollutant source located near or anywhere else in the operating room or the movement of personnel in the room, (b) buoyancy effects due to the dense or hot gases that result due to the presence of heat sources such as operating lamps, etc. and (c) short-circuiting between the ventilation systems supply and exhaust. The coanda effect also plays a role in CO2 dispersion. Since the air-conditioning units (for the case under consideration) are facing the width of the operation theatre which is comparatively smaller than the length of the operation theatre. The coanda effect tends to make the jet adhere to the wall along the ceiling and more towards the opposite wall. Only the cooling effects will be felt by the operation theatre staff. In the study conducted in 10 hospitals it was noted that only two hospitals had their theatre located at the ground floor. Six hospitals had their theatre located in the first floor and two hospitals had their theatre in the second floor. Among the two hospitals which had their theatre located in the second floor one was surrounded by trees and the number of theatre staff was considerably low (six including the patient and the person measuring the airflow parameters) when compared to the other cases. This particular case seemed to be better in terms of CO2
[1] R. Kameel, E.E. Khalil, Energy efficient thermal comfort Vs air quality in air-conditioned health care applications, Building Simulation (2003) 621– 626. [2] Draft Guideline for environmental infection control in health care facilities, 2001, Centre for disease control and prevention—Health care Infection Control Practices Advisory Committee (HICPAC). [3] G. Einberg, Air diffusion and solid contaminant behavior in room ventilation—a CFD based integrated approach, Ph.D. Thesis, Department of Energy Technology, The Royal Institute of Technology, Sweden, April 2005. [4] U. Moscato, T. Esposito, G. Vanini, Nitrous oxide pollution: a geostatical method to assess spatial distribution of anesthetic gases and hospital staff exposure, Proceedings of Healthy Buildings 1 (2000) 487–492. [5] G. Mehta, et al., Microbiological surveillance of operation theatres, 2000. http://www.orthoteers.org/content/content.aspx?Article=372. [6] W. Lu, A. Howarth, N. Adam, S. Riffat, Modeling and measurement of airflow and aerosol particle distribution in a ventilated two-zone chamber, Building and Environment 31 (1996) 417–423. [7] L. Martin, The four most important equations in clinical practice, the clinical ultimate reference. www.globalrph.com. [8] K. Singh, Small Incision Cataract Surgery, Alpha Science Int’l Ltd., 2003. [9] Q. Chen, Computational fluid dynamics for HVAC: successes and failures, ASHRAE Transactions 103 (1) (1997) 178–187. [10] J.F. Pfost, A re-evaluation of laminar airflow in hospital operating rooms, ASHRAE Transactions 87 (2) (1981) 729–739. [11] R. Kameel, E.E. Khalil, Experimental and numerical investigations of airflow regimes in air-conditioned operation theatres, Building Simulation (2003). [12] R. Kameel, E.E. Khalil, Experimental and numerical investigations of heat transfer in air conditioned operation theatres, Building Simulation (2003). [13] B.E. Launder, D.B. Spalding, The numerical computation of turbulent flows, Method in Applied Mechanical Engineering 87 (2) (1974) 729–739.
C.P. Karthikeyan, A.A. Samuel / Energy and Buildings 40 (2008) 231–239 [14] Patankar, Revised algorithm for pressure velocity variation, 1980. [15] P.V. Nielsen, The selection of Turbulence models for prediction of room air flow, ASHRAE Transactions: Symposia (SF-98-10-1) (1999) 1119– 1128. [16] L.-M. Lo, Numerical studies of airflow movement and contaminant transport in hospital operating rooms, MSc Thesis, University of Minnesota, 1997. [17] F. Memarzadeh, A. Manning, Comparison of operating room ventilation systems in the protection of the surgical site, ASHRAE Transactions 108 (22) (2002) 3–15.
239
[18] F. Memarzadeh, A. Manning, Reducing risks of surgery, ASHRAE Journal (2003) 28–33. [19] FLUENT’S User Guide Version 6.1, Fluent Incorporated, Lebanon, NH, 2001. [20] Z.F. Tian, J.Y. Tu, G.H. Yeih, Large eddy simulation indoor airflow and contaminant concentration, Building Simulation (2003). [21] ANSI/ASHRAE 113-1990, Method of testing for room air diffusion. [22] ANSI/ASHRAE 62-1989, Acceptable ventilation for indoor air quality. [23] ANSI/ASHRAE 55-1992, Thermal environmental conditions for human occupancy.