Ventilation Performance in

Operating Rooms: a Numerical

Assessment

CONG WANG

Doctoral Thesis
Stockholm, Sweden 2019
TRITA-ABE-DLT-1939 KTH School of ABE
SE-100 44 Stockholm
ISBN 978-91-7873-338-5 SWEDEN

Akademisk avhandling som med tillstånd av Kungl Tekniska högskolan framlägges till
offentlig granskning för avläggande av teknologie doktorsexamen i Byggvetenskap 2019-
11-27 klockan 10:00 i kollegiesalen, Kungl Tekniska högskolan, Brinellvägen 8,
Stockholm, Sweden.

© Cong Wang, October 2019

Abstract

Surgical site infections (SSIs) remain one of the most challenging postoperative complications
of healthcare and threaten the lives of millions of patients each year. Current evidence has
shown a positive relationship between the airborne concentration of bacteria-carrying
particles (BCPs) in the operating room (OR) and the rate of infections. The OR ventilation is
crucial for mitigating the dispersion of airborne bacterial contaminants and thus controlling
the risk of SSIs. A variety of ventilation schemes have been developed for OR use. Each has
pros and cons and may be better suited than another for operations under certain conditions.
The proper functioning of OR ventilation is also affected by external and internal disruptions.
By applying Computational Fluid Dynamics (CFD), the present study investigates the airflow
and contaminant distribution in ORs under different conditions.

The airflow distribution is of critical importance in removing or diluting airborne

contaminants. The conventional mixing ventilation is not able to reliably create an ultraclean
environment. The usage of mixing ventilation in infection-prone surgery should be limited,
especially when a large surgical team is involved. Laminar airflow (LAF) ventilation demands
a sufficient airflow rate to achieve desired performance. Temperature-controlled airflow
(TAF) ventilation represents an effective ventilation scheme that can serve as an energy-
efficient alternative to LAF.

Door openings have a detrimental impact on the microbiological cleanliness of the OR. The
temperature in the OR and adjacent space should be well controlled to minimize the inter-
zonal contaminant transfer. Temporarily reducing the OR exhaust flow during door operation
forms a directional airflow towards the adjacent space, which is found to be an effective
solution to ensure the isolation.

Surgical lamps serve as physical obstructions in the airflow path and significantly deteriorate
the performance of LAF ventilation. It is highly recommended to improve the shape and
design of the lamps in the LAF ventilation. TAF is found to be less sensitive to the presence
of surgical lamps in the airflow path. The buoyancy-driven airflow used by TAF is more
capable of circumventing obstacles than the inertia-driven flow used by LAF. Thermal plumes
developed from the surgical equipment in the OR have the potential to distort the buoyancy-
driven airflow in TAF.

The thesis conducts a comprehensive literature review of important topics in OR ventilation.

The present study enhances the understanding of the strengths and limitations of different
ventilation schemes and increases the knowledge of the design and usage of OR ventilation.

Keywords: operating room ventilation, bacteria-carrying particles, laminar airflow, mixing

ventilation, temperature-controlled airflow, computation fluid dynamics, Lagrangian particle
tracking

i
Sammanfattning

Postoperativa infektioner tillhör idag de mest utmanande komplikationerna inom sjukvården.

De utsätter årligen miljontals patienter för allvarlig hälsofara. Bevis finns för ett samband
mellan luftburen koncentration av bakteriebärande partiklar i operationssalen och
infektionsfrekvensen. Ventilationen i operationssalen är avgörande för att minimera
spridningen av luftburna bakteriella föroreningar och därmed kontrollera risken för
postoperativ infektion. För ventilation i operationsrum har olika luftföringsprinciper
utvecklats. Dessa har för- och nackdelar och vissa kan vara bättre lämpade än andra under
givna förhållanden. Ventilationens funktion påverkas också av yttre och inre störningar. Med
numeriska beräkningsmetoder (CFD) undersöks i denna avhandling luftflöden och
föroreningsspridning i operationsrum.

Såväl luftflödets fördelning i rummet som luftflödets volym är av avgörande betydelse för
eliminering och utspädning av luftburna föroreningar. Konventionell omblandande ventilation
har såväl teoretiska som praktiska begränsningar i detta sammanhang och dess användning
bör därför begränsas vid infektionsbenägen operation. Detta gäller särskilt vid stora
operationsteam. Ventilation med så kallat laminärt luftflöde (LAF) kräver tillräckligt luftflöde
för att uppnå önskat resultat. Ventilation med temperaturstyrt luftflöde (TAF) har befunnits
vara ett (energi)effektivt alternativ till LAF.

Dörröppningar till operationssalen har negativ inverkan på den mikrobiologiska renheten i

rummet. Temperaturen i operationsrummet och angränsande utrymme bör kontrolleras väl för
att minimera överföring av föroreningar mellan zonerna. Genom att tillfälligt reducera
frånluftflödet i operationsrummet under dörrdrift bildas ett riktat luftflöde mot intilliggande
utrymmen som effektiv visat sig säkerställa isoleringen.

Vissa kirurgiska lampor fungerar som fysiska hinder i luftflödesvägen och försämrar avsevärt
funktionen vid LAF-ventilation. Därför rekommenderas starkt att inte förbise lampans form
och placering. TAF-ventilation har visat sig vara mindre känslig för kirurgiska lampor i
luftflödesvägen. Det termiskt drivna luftflödet vid TAF syns mera kapabelt att klara hinder än
tröghetsdrivet flöde vid LAF. Värmeplymer som den kirurgiska utrustningen genererar i
operationsrummet kan påverka det termiskt drivna luftflödet vid TAF.

Avhandlingen innehåller en omfattande litteraturstudie inom området. Avhandlingen ökar

förståelsen för olika ventilationsprincipers funktion och kunskapen om hur ventilation skall
användas i sjukhusmiljöer.

Nyckelord: operationsrumsventilation, bakteriebärande partiklar, laminärt luftflöde,

blandningsventilation, temperaturstyrt luftflöde, beräkningsfluiddynamik, Lagrangian particle
tracking

ii
Preface

The major sponsor for the research was the Swedish Research Council for Sustainable
Development (FORMAS). The work presented in this thesis was carried out between March
2017 and September 2019 at the Division of Sustainable Buildings (Division of Fluid and
Climate Technology) in the Department of Civil and Architectural Engineering at KTH Royal
Institute of Technology, Stockholm, Sweden.

Firstly of all, I would like to express my deepest gratitude towards my supervisor, Professor
Sture Holmberg, for his ample guidance, strong encouragement and patient support during the
development of this work. He gave me the opportunity to carry out this research work and
always encouraged me to overcome difficulties in the pursuit of my research. He always has
my deepest appreciation for his constant confidence in me over the past three years.

I am also grateful to Professor Ivo Martinac for the support and encouragement he provided
for me during the licentiate studies. He gave me the opportunity to start my doctoral studies
and helped me develop myself on both an academic and a personal level.

I would also wish to thank my co-supervisor, Dr. Sasan Sadrizadeh, for his consistent support
and valuable input to my work. I thank him not only for the technical help but also for his
kindness and valuable advice.

Last, I would like to thank all my colleagues at the Division of Sustainable Buildings.
Without them, my life in the past two and half years would not have been so joyful and
fruitful.

Stockholm, October 2019

Cong Wang

iii
Publications

This doctoral thesis is based upon the following papers:

Paper I. Polak, J., Afshari, A., Sadeghian, P., Wang, C., & Sadrizadeh, S. (2019).
Improving the performance of heat valve ventilation system: a study on the
provided thermal environment. Building and Environment, 164, 106338.

Paper II. Wang, C., Holmberg, S., & Sadrizadeh, S. (2018). Numerical study of
temperature-controlled airflow in comparison with turbulent mixing and
laminar airflow for operating room ventilation. Building and Environment, 144,
45-56.

Paper III. Wang, C., Holmberg, S., & Sadrizadeh, S. (2019). Impact of door opening on
the risk of surgical site infections in an operating room with mixing ventilation.
Accepted by Indoor and Built Environment.

Paper IV. Wang, C., Sadrizadeh, S., & Holmberg, S. (2018). Influence of the shape of
surgical lamps on the airflow and particle distribution in operating rooms.
In Proceedings of Roomvent & Ventilation 2018: Excellent Indoor Climate and
High Performing Ventilation, Espoo, Finland.

Paper V. Sadeghian, P., Wang, C., Duwig, C., & Sadrizadeh, S. (2019). Impact of the
surgical lamp inclination and design on the operating room ventilation
performance: a numerical study. Submitted for journal publication.

Paper VI. Wang, C., Sadeghian, P., & Sadrizadeh, S. (2019). Effect of staff number on
the bacteria contamination in operating rooms with temperature-controlled
airflow ventilation and turbulent mixing ventilation. In Proceedings of Building
Simulation 2019: 16th Conference of IBPSA, Rome, Italy.

Paper VII. Wang, C., Sadrizadeh, S., & Holmberg, S. (2018). Numerical assessment of
the influence of heat loads on the performance of temperature-controlled
airflow in an operating room. In Proceedings of 39th AIVC Conference – 7th
TightVent Conference – 5th Venticool Conference: Smart Ventilation for
Buildings, Juan-les-Pins, France. Submitted to International Journal of
Ventilation by invitation from the Conference Scientific Committee.

v
Other publications by the author (not included in this thesis):

Paper VIII. Huang, Y., Ge, F., Wang, C., & Hu, Z. (2019). Numerical study on the heat
and mass transfer characteristics of the open-type cross-flow heat-source tower
at low ambient temperature. International Journal of Heat and Mass
Transfer, 145, 118756.

Paper IX. Lind, M.C., Sadrizadeh, S., Venås, B., Sadeghian, P, Wang, C., & Harsem,
T.T. (2019). Minimizing the airborne particle migration to the operating room
during door opening. In Proceedings of 10th International Conference on
Indoor Air Quality, Ventilation and Energy Conservation in Buildings, Bari,
Italy.

Paper X. Niu, C., Wang, C., & Sadrizadeh, S. (2019). Numerical simulations of
overhead island kitchen exhaust devices. In Proceedings of 11th International
Symposium on Heating, Ventilation and Air Conditioning, Harbin, China.

Paper XI. Wang, C., & Sadrizadeh, S. (2018). Numerical assessment of a novel
ventilation strategy for operating rooms in comparison with turbulent mixing
and laminar air flow. In Proceedings of the 15th Conference of the International
Society of Indoor Air Quality & Climate, Philadelphia, USA.

Paper XII. Alsved, M., Wang, C., Civilis, A., Sadrizadeh, S., Ekolind, P., Skredsvik, H.,
Höjerback, P., Jakobsson, J., & Löndahl, J. (2018). Experimental and
computational evaluation of airborne bacteria in hospital operating rooms with
high airflows. In Proceedings of the 5th Working & Indoor Aerosols
Conference, Cassino, Italy.

Paper XIII. Wang, C., Sadeghian, P., & Sadrizadeh, S. (2017). Application of open-source
CFD software to the indoor airflow simulation. In Proceedings of 38th AIVC
Conference – 6th TightVent Conference – 4th Venticool Conference: Ventilating
Healthy Low-Energy Buildings, Nottingham, UK.

Paper XIV. Kilkis, S., Wang, C., Björk, F., & Martinac, I. (2017). Cleaner energy
scenarios for building clusters in campus areas based on the Rational Exergy
Management Model. Journal of Cleaner Production, 155, 72-82.

Paper XV. Wang, C., Kilkis, S., Tjernström, J., Nyblom, J., & Martinac, I. (2017). Multi-
objective optimization and parametric analysis of energy system designs for the
Albano university campus in Stockholm. Procedia Engineering, 180, 621-630.

vi
Paper XVI. Wang, C., Martinac, I., & Magny, A. (2016). Multi-objective optimization of
energy system designs for the Albano university campus in Stockholm.
In Proceedings of YRSB16 – iiSBE Forum of Young Researchers in Sustainable
Building 2016, Prague, Czech Republic.

Paper XVII. Wang, C., Martinac, I., & Magny, A. (2015). Multi-objective robust
optimization of energy systems for a sustainable district in Stockholm.
In Proceedings of 14th Conference of International Building Performance
Simulation Association, Building Simulation 2015, Hyderabad, India.

Licentiate Wang, C. (2016). Optimal design of district energy systems: a multi-objective

Thesis approach. KTH Royal Institute of Technology.

vii
Nomenclature

Abbreviations

ACH air changes per hour

BCP bacteria-carrying particle

CFD computational fluid dynamics

CFU colony-forming units

COR adjacent corridor

DNS direct numerical simulation

DRW discrete random walk

FVM finite volume method

HEPA high-efficiency particulate air

LAF laminar airflow

LES large-eddy simulation

LPT Lagrangian particle tracking

OR operating room

PSI-C particle source in-cell

RANS Reynolds averaged Navier-Stokes

S. aureus Staphylococcus aureus

SIMPLE semi-implicit method for pressure linked equations

SSI surgical site infection

TAF temperature-controlled airflow

UDF unidirectional airflow

USD United States dollar

ix
Latin letters

𝑎 ,𝑎 ,𝑎 constants in Morsi and Alexander’s drag law (-)

𝑐 volume concentration of BCPs (CFU/m3)

𝐶 drag coefficient (-)

𝐶 coefficient in 𝑘 𝜀 models (-)

𝑑 particle diameter (m)

𝐹 additional forces per unit particle mass (m/s2)

𝐹 drag factor or inverse of relaxation time (1/s)

g gravitational acceleration (m/s2)

𝑘 turbulent kinetic energy (m/s2)

𝑙 eddy length scale (m)

𝑛 number of persons

𝑄 airflow rate (m3/s)

𝑞 BCPs source strength of a person (CFU/s)

𝑅𝑒 Reynolds number (-)

𝑅𝑒 particle Reynolds number (-)

𝑆 source term of transport equation

𝑡 time (s)

𝑡 eddy crossing time (s)

𝑡 eddy lifetime (s)

𝑡 eddy interaction time (s)

𝑇 Lagrangian integral time (s)

𝑢 instantaneous fluid velocity (m/s)

𝑢 mean fluid velocity (m/s)

𝑢 fluctuating fluid velocity (m/s)

𝑢 particle velocity (m/s)

𝑉 velocity (m/s)

x
𝑦 non-dimensional wall distance (-)

Greek letters

Γ diffusion coefficient of transported quantity

𝜖 turbulence dissipation rate (m2/s3)

𝜁 random variable following Gaussian distribution (-)

𝜇 fluid viscosity (kg/(mꞏs))

𝜇 turbulent viscosity (kg/(mꞏs))

𝜌 fluid density (kg/m3)

𝜌 particle density (kg/m3)

𝜏 particle relaxation time (s)

𝜙 transported quantity

xi
Contents

Abstract ...................................................................................................................................... i 

Preface ...................................................................................................................................... iii 

Publications ............................................................................................................................... v 

Nomenclature ........................................................................................................................... ix 

1  Introduction.................................................................................................................... 1 
1.1  Background ............................................................................................................. 1 
1.2  Motivation ............................................................................................................... 2 
1.3  Methods ................................................................................................................... 3 
1.4  Limitations............................................................................................................... 3 

2  Literature Review .......................................................................................................... 5 

2.1  Bacteria-Carrying Particles ..................................................................................... 5 
2.2  Air Quality Evaluation ............................................................................................ 6 
2.2.1  Particle Counting or Microbiological Sampling?....................................... 6 
2.2.2  Passive Air Sampling or Active Air Sampling? ......................................... 7 
2.3  Ventilation Schemes ................................................................................................ 8 
2.3.1  Mixing and Laminar Airflow ..................................................................... 9 
2.3.2  Temperature-Controlled Airflow ............................................................. 11 
2.4  Door Opening and Foot Traffic ............................................................................. 11 

3  Methodology ................................................................................................................. 15 

3.1  Physical Cases ....................................................................................................... 15 
3.1.1  Helsingborg General Hospital .................................................................. 15 
3.1.2  New Karolinska Solna University Hospital ............................................. 17 
3.2  Airflow Simulation ................................................................................................ 18 
3.2.1  Governing Equations ................................................................................ 18 
3.2.2  Turbulence Modelling .............................................................................. 18 
3.2.3  Boundary Conditions................................................................................ 19 

xiii
 3.2.4  Numerical Aspects ................................................................................... 20 
3.2.5  Mesh and Convergence ............................................................................ 20 
3.3  Door Motion Simulation ....................................................................................... 21 
3.4  Particle Transport Simulation ................................................................................ 23 
3.4.1  Eulerian and Lagrangian Formulation...................................................... 23 
3.4.2  Lagrangian Particle Tracking ................................................................... 23 
3.4.3  Discrete Random Walk ............................................................................ 25 
3.4.4  Particle Concentration .............................................................................. 26 
3.5  Model Validation ................................................................................................... 28 
3.5.1  Validation Study I – Heat Valve Ventilation ........................................... 28 
3.5.2  Validation Study II – Underfloor Air Distribution................................... 30 
3.5.3  Validation Study III – Ventilated Two-Zone Chamber ........................... 32 

4  Results and Discussions ............................................................................................... 35 

4.1  Ventilation Schemes .............................................................................................. 35 
4.1.1  Operating Airflow Rates .......................................................................... 35 
4.1.2  Identical Airflow Rates ............................................................................ 38 
4.1.3  Dilution Model ......................................................................................... 41 
4.2  Door Openings ...................................................................................................... 43 
4.2.1  Increased Contamination .......................................................................... 43 
4.2.2  Positive Pressure ...................................................................................... 49 
4.2.3  Directional Airflow .................................................................................. 51 
4.3  Internal Disruptions ............................................................................................... 54 
4.3.1  Surgical Lamps ......................................................................................... 54 
4.3.2  Surgical Staff ............................................................................................ 58 
4.3.3  Surgical Equipment .................................................................................. 59 

5  Conclusions and Future Work ................................................................................... 63 

5.1  Conclusions ........................................................................................................... 63 
5.2  Future Work .......................................................................................................... 64 

Bibliography ........................................................................................................................... 67 

xiv
 1.1. BACKGROUND

Chapter 1

Introduction

1.1 Background
Indoor air quality has a significant impact on the health and comfort of occupants. In office
and residential buildings, poor indoor air quality has been linked to Sick Building Syndrome
and degraded work performance. Over the past decades, indoor air quality has received
considerable attention from both the academic community and political institutions for
improving the occupants’ health and wellbeing. In the operating room (OR) environment, air
quality becomes even more critical and is found to be related to the rate of surgical site
infections (SSIs), which threaten the lives of millions of patients each year.

SSIs are “infections of the incision or organ or space that occur after surgery” [1]. Some
infections can be superficial involving the skin only, whereas some others can be more
serious and involve deeper tissues under the skin, organs, and implanted material [2,3]. SSIs
represent the most common healthcare-associated infections among patients [4], constituting
up to 19.6% of all such infections in Europe in 2011–2012 [5]. In the US, a prevalence survey
of healthcare-associated infections estimated 157,500 surgical site infections associated with
inpatient surgeries in 2011 [6]. In Sweden, nearly 65,000 patients suffer from healthcare-
associated infections each year, 23% of which are SSIs [7]. The National Healthcare Safety
Network in the US reported 16,147 primary incisional SSIs following 849,659 operative
procedures in 39 procedure categories from 2006 to 2008, with an overall SSI rate of 1.9%
[8]. The real rate of SSIs is likely to be underestimated as approximately 50% of SSIs become
evident after discharge from hospital [9]. SSIs remain one of the most challenging
postoperative complications of healthcare, which contribute substantially to the increased
mortality, prolonged hospitalization and repeat surgeries [4,5,10–14]. Patients who develop
SSIs have two to eleven times higher mortality rates than the ones who do not develop SSIs
[15]. The mortality rate associated with SSIs is about 3%, with 77% of deaths among patients
who develop SSIs being attributable directly to such infections [10,15]. Depending on the
type of surgery, SSIs can extend the length of hospital stay by 3–21 days [11]. In addition to
the negative impact on quality of life for the patient, SSIs also constitute a considerable
financial burden on healthcare providers [16–18]. Broex et al. [16] reported that a patient with

1
CHAPTER 1. INTRODUCTION

SSI approximately doubles the cost of a patient without SSI. It was estimated that SSIs, on
average, prolonged the hospitalization by 9.7 days and increased the cost by 20,842 USD per
admission [13].

SSIs are caused by bacteria that enter an open wound during surgery. Due to the catastrophic
impact of SSIs, there is a growing demand for interventions for the prevention of SSIs. The
prevention of SSIs includes timely administration of preoperative antibiotic prophylaxis,
meticulous operative technique, and maintenance of clean operating room environment [3,4].
The use of antibiotics has been an essential component for the prevention of SSIs over the
past decades. However, the reliance on antibiotic prophylaxis cannot continue due to the
increasing incidence of antibiotic-resistance. Bisht et al. [19] reported that nearly 70% of the
bacteria that cause hospital infections are resistant to at least one antibiotic agent. Lyon and
Skurry [20] found that more than 95% of Staphylococcus aureus were resistant to Penicillin,
which was the first antibiotic agent discovered by Sir Alexander Fleming in 1929. The
growing resistance to antibiotics has now been considered as a global threat to the human
race, ranking alongside terrorism [21]. Therefore, more efforts should be spent in limiting the
spread of airborne bacteria and maintaining clean surgical environments in ORs.

Strong evidence has been found to support the correlation between the airborne bacterial
contamination in ORs and the risk of infections. The extensive Medical Research Council
study of 8,000 total hip and knee replacement at 19 different hospitals established an
essentially strong linear relationship between the count of viable bacteria in the air and the
rate of deep infections following surgeries [22,23]. Nowadays it has been widely accepted
that infection-prone surgeries should not be taken in the environment with airborne bacteria
concentration higher than 10 colony-forming units per cubic meter (CFU/m3) [24,25].
Accordingly, many medical and government agencies have developed standards and
guidelines that regulate the use of ventilation and stipulate microbiological cleanliness of air
in ORs.

1.2 Motivation
The history of surgical procedures can date back to prehistoric times [26]. By 400 BC, the
Greek physician Hippocrates had noted the effect of clean air on the prevention of contagious
diseases [27]. The need for a clean surgical environment was first acknowledged in late 19th
century, driven by the frequent infections in surgical wounds [28,29]. Nowadays operating
room ventilation plays an important role in improving microbiological air quality, maintaining
thermal comfort and optimizing energy usage. The ventilation airflow aims to effectively
remove or dilute the airborne bacterial contaminants from the surgical critical zone. Over the
past century, various ventilation schemes have been developed for OR use, such as turbulent
mixing airflow ventilation, laminar airflow ventilation, displacement ventilation, and hybrid
ventilation. These ventilation schemes differ in the primary airflow pattern created in the

2
 1.3. METHODS

ORs. It still remains debatable among medical professionals and ventilation engineers as to
the most appropriate ventilation for OR use. The pros and cons of each ventilation scheme
should be thoroughly analyzed and compared. The applicability of various ventilation
schemes may also vary with the working conditions. The presence of surgical personnel and
surgical facilities in the OR may interact with the ventilation airflow and deteriorate the
ventilation performance. Furthermore, door openings and foot traffic may alter the established
airflow pattern and causes potential failure of the ventilation. Therefore, it is of great
importance to investigate the airflow distribution and contaminant dispersion in ORs with
various ventilation schemes under different working conditions.

It is believed that critical parameters such as ventilation airflow rates, supply air temperatures,
and air distribution including the location of supply/exhaust vents should be correctly set to
provide a clean surgical environment. Thus, the investigation in the present study looks into
the ventilation schemes and their applicability under different working conditions – varying
ventilation rates, different number of surgical staff, different heat loads, and different design
of surgical lamps. This study aims to enhance the understanding of the strengths and
limitations of existing ventilation schemes and increase the knowledge of the design and
usage of OR ventilation.

1.3 Methods
Traditional measurements have run into problems of both technology and costs. In the context
of ORs, field measurements are further faced with logistic and ethical constraints. Therefore,
we develop and apply computer-based airflow and contamination simulation methods to
ensure reliable infection risk assessment in ORs. We use Computational Fluid Dynamics
(CFD) to resolve the Reynold Averaged Navier-Stokes (RANS) equations for the airflow and
Lagrangian particle tracking (LPT) to compute the particle trajectories for the contaminant
dispersion. The two-equation Realizable 𝑘 𝜀 model is used to model the turbulent flow. The
Discrete Random Walk (DRW) model is adopted to account for the turbulent dispersion of
particles. With numerical simulations, we obtain high-quality spatial and temporal distribution
of bacterial contaminants in ORs.

The ventilation performance of the ORs in Helsingborg General Hospital is investigated. To

extend the applicability of our research, we also adopt the OR designed for the New
Karolinska Solna University Hospital.

1.4 Limitations
The present study has been conducted by means of numerical simulations only. Field
measurements are not performed due to limited time and resources. Despite the fact that the

3
CHAPTER 1. INTRODUCTION

numerical models are well validated against laboratory experiment, there is a lack of adequate
information of the initial and boundary conditions. Furthermore, the airflow and contaminant
dispersion are subject to considerable uncertainties in ORs during surgery. Such uncertainties
cannot be accounted and quantified in numerical simulations. Therefore, field measurements
in real ORs under working conditions are necessary for future studies. The investigation of
ventilation performance in this study is limited to specific designs of different ventilation
schemes. Many other design variants exist for each ventilation scheme. Caution is needed
when interpreting the findings and extending the conclusions of the present study.

4
 2.1. BACTERIA-CARRYING PARTICLES

Chapter 2

Literature Review

2.1 Bacteria-Carrying Particles

SSIs are caused by bacteria entering incision sites during surgical procedures. Some bacterial
species are highly capable of causing infections. Staphylococcus aureus (S. aureus), for
instance, has generally been accepted as the most common cause of SSIs [30–33]. Numerous
examples of S. aureus causing wound infections have been reported in the literature [34–36].
Although bacteria have a wide diversity of shape and sizes, they are often attached to particles
that are much larger than themselves [37]. It has been reported that the majority of bacteria
dispersed in an OR are carried on desquamated skin scales, shed from the human body [38–
41]. The human skin has an average area of 1.75 m2 and a complete layer is desquamated on
average in less than 24 hours [42,43]. During moderate physical activity, a person
disseminates more than 107 skin scales every day, about 10% of which carry viable
microorganisms [42,44]. During natural walking movement, a person releases about 104 skin
scales per minute [45]. Woods et al. [40] estimated a typical generation rate of 6.0 µg/min of
7.5 µm mean diameter skin flakes for a clothed person. Current evidence in the literature
indicates that the primary source of bacteria-carrying particles (BCPs) in the OR is the
surgical team. Many studies have confirmed that the level of airborne BCPs is associated with
the number of people present in the OR [46–49].

The skin is an ecosystem of diverse and highly variable habitats colonized by a broad
diversity of microorganisms [50]. The dispersal of skin bacteria varies greatly from individual
to individual. Wide variations in both the number of shed skin fragments and the proportion
of these particles carrying bacteria have been noted [37,41,51], which together can contribute
to as high as 1000-fold difference in the number of shed BCPs between individuals [52].
Males tend to shed more skin bacteria than do females, with males being more heavily
colonized than females [41,42]. Another striking observation is that the dispersal of S. aureus
is more profuse among men than women [53,54]. People are found to shed more skin bacteria
into the air after a shower than before it [55], but the regular use of antibacterial soaps
containing hexachlorophene or trichlorocarbanilide has shown a statistically significant effect
on reducing the dispersal of skin bacteria [56,57]. A person may become a so called

5
CHAPTER 2. LITERATURE REVIEW

‘disperser’ of S. aureus if he or she acquires a skin disorder or a septic lesion [31]. The design
of surgical garments, facemasks, caps, etc., also has a significant effect on the amount of skin
bacteria transferred to the air [54,58–61]. Surgical clothing serves as filters and retains the
skin fragments shed from the personnel. Mitchell et al. [58] found that a considerable
reduction in the dispersal of skin bacteria was achieved with the non-woven fabric clothing
compared with the conventional clothing made of loose-weave cotton. Thus, there is growing
need for surgical clothing that is both comfortable to wear and capable of retaining as many
skin fragments as possible.

The skin fragments shed from humans have a wide size distribution. Mackintosh et al. [52]
measured the dimension of dispersed skin fragments and found that they were essentially
plate-like with a thickness of about 2–5 µm. The minimum projected diameter, defined as the
distance between two parallel bounding lines, extends below 5 µm and beyond 60 µm [52].
Noble [62] measured the size distribution of airborne particles in hospitals and reported that
the mean aerodynamic diameter of particles carrying S. aureus was 13.5 µm. Noble et al. [63]
further examined the size of airborne infected particle in offices, hospital wards and operating
rooms and concluded that micro-organisms related to human disease and carriage were
usually found on particles in the range 4–20 µm aerodynamic diameter. In their study, the
median aerodynamic diameter of particles carrying aerobic flora found in operating rooms
was 12.3 µm. Hansen et al. [64] found that the number of particles above 5 µm was closely
correlated to the bacteria count in 105 operating procedures. Memarzadeh and Manning [65]
pointed out that the bacteria-carrying particles shed from humans were in the order of 10 µm
in diameter.

2.2 Air Quality Evaluation

The microbiological quality of air in operating rooms is a key factor in controlling SSIs.
However, there are no established regulations for air monitoring in operating rooms. The
literature is divided on the best method to quantify air contamination in operating rooms.

2.2.1 Particle Counting or Microbiological Sampling?

Particle counting, as a technique derived from cleanroom technology, is commonly used for
industry. Particle counting measures the presence of particles of varying sizes and numbers.
With the use of ultraclean ventilation in operating rooms, there is a tendency to apply particle
counting for hospital use. Microbiological sampling evaluates the microbiological
contamination in the environment related to infections and presents the quantity of
microorganisms in colony-forming units (CFU). Monitoring operating rooms by
microbiological sampling is time consuming and results are delivered after a couple of days.
By contrast, particle counting can be done more frequently and deliver online results.

6
 2.2. AIR QUALITY EVALUATION

There is no international consensus on whether particle counting or microbiological sampling

is more appropriate to evaluate air quality in ORs. Several studies have analyzed the
relationship between particle counts and BCPs counts in the air but conflicting results are
reported. Whether particle counting could be predictive of microbiological contamination of
air in an OR remains controversial.

Seal and Clark [66] compared the count of airborne particles in eight size ranges with the
number of BCPs in both ultraclean and turbulently ventilated operating rooms. They found a
significant correlation between the number of particles sized 5–7 µm and the level of
microbiological contamination in ORs. In contrast, Jalovaara and Puranen [67] observed no
significant correlation between bacterial and particle counts during surgeries performed in a
conventionally ventilated OR. A review study by Dharan and Pittet [68] recommended the use
of particle counting as a routine procedure for air monitoring in operating rooms. They
suggested that the bacteria sampling should be restricted to the investigation of epidemics,
validation of protocols, and maintenance procedures of operating rooms. Landrin et al. [69]
observed an absence of correlation between particle counting and microbiological counting in
four empty ORs equipped with conventional ventilation via high-efficiency particulate air
(HEPA) filters. They suggested that particle counting should not replace microbiological
sampling as routine evaluation of microbial contamination in conventionally ventilated ORs.
Hansen et al. [64] measured particle and bacteria counts during 105 procedures under laminar
airflow conditions and found that counts of particles ≥5 µm correlated significantly with
bacteria counts. Stocks et al. [46] investigated whether the air particulate concentration could
predict the airborne BCPs concentration during 22 arthroplasty procedures in an OR with
mixing ventilation. They found that the number of large particles (≥10 µm) correlated with
the number of BCPs in air sampled within 40 cm from the surgical incision. Cristina et al.
[70] found no statistically significant correlation between the microbial loads and particle
counts for either particle size group (≥0.5 µm and ≥5 µm) during 95 arthroplasty procedures.
They concluded that microbiological monitoring remained the most suitable method to
evaluate the quality of air in ORs. A large multicenter observational study by Birgand et al.
[71] found a strong correlation between airborne particle counts and microbial contamination
during 60 operations in 13 ORs and thus concluded that particle counting could serve as a
good surrogate of airborne microbiological contamination in operating rooms.

2.2.2 Passive Air Sampling or Active Air Sampling?

The microbiological sampling of air in ORs can be performed by two principal methods:
active air sampling and passive air sampling. In active air sampling, the sampling instrument
(e.g. a slit sampler) actively pulls in a pre-defined volume of air through or over a collection
device such as a filter or an agar plate. The microorganisms collected by the collection device
will be incubated and the level of contamination is expressed as colony-forming units per unit
volume of air (i.e. CFU/m3). In contrast, passive air sampling requires the use of a settle plate
(e.g. a Petri dish with a culture medium) exposed to the air for a certain period of time. Viable

7
CHAPTER 2. LITERATURE REVIEW

microbiological particles that eventually sediment out of the air and settle onto the surface of
the plate over the time of exposure are subsequently incubated. The microorganisms on the
settle plate are counted and by relating both the plate area and exposure time the level of
contamination can be presented in the form of CFU/m2/h.

The passive air sampling is more efficient in sampling larger, faster-settling particles, as
larger particles are more likely to sediment than are smaller particles [72]. Considering the
size distribution of BCPs, the passive sampler needs to work continuously over a long time
period to collect a sufficient amount of microorganisms. In contrast, the active sampler
samples air over a short time period and thus can reflect the transient effect of individual
events on the air quality in ORs, such as door openings and foot traffics in ORs [73].
However, the passive air sampling is more convenient in practice than the active air sampling,
as the settle plate is easily available and can be used without additional devices.

The international standard (ISO19468-1) describes the available air sampling techniques but
leaves the choice of method open [74,75]. The best method to assess the microbiological
contamination of air remains debatable and studies in this field are controversial. Some
researchers recommend the use of passive air sampling, arguing that the surface count of
bacteria is clinically more relevant to the risk of SSIs than the air count as the airborne
contaminants that fall onto a critical surface is more harmful [72,76–78]. Napoli et al. [79]
suggested that both methods could be useful for general monitoring of air contamination but
the choice should be made depending on the specific information to be obtained. Friberg et al.
[78] argued that the airborne contamination would be unhelpful in assessing the surgical site
contamination and thus proposed a corresponding standard for surface contamination <350
CFU/m2/h as a complement to the accepted standard for airborne contamination <10 CFU/m3.
Studies have been conducted to compare the level of microbial contamination obtained
through the two sampling methods. Some studies have documented a significant correlation
between two methods [72,74,79], whereas some others have found an inconsistent correlation
[78] or an absence of correlation [80].

2.3 Ventilation Schemes

The need for a clean indoor environment for operations was first acknowledged in late 19th
century, driven by the frequent infections in wounds [28,29]. The provision of ventilation for
operating rooms has since become increasingly important. The purpose of OR ventilation is to
provide comfortable working conditions for the surgical personnel and more importantly to
protect the patient against exogenous bacterial infections [81]. Airborne BCPs can settle
directly on the open wound or on instruments which subsequently come into contact with the
wound. A well-designed ventilation system is thus crucial to minimize the risk of airborne
infections. By supplying sterile air to the operating room, the conditioned airflow should be
able to effectively remove or dilute the airborne BCPs from the surgical zone. The ventilation

8
 2.3. VENTILATION SCHEMES

system should also control such environmental factors as the temperature, humidity, and air
movement to maintain an appropriate level of thermal comfort. A variety of ventilation
schemes have been developed and introduced into medical operations over the past century.

2.3.1 Mixing and Laminar Airflow

Operating rooms use two main airflow distribution principles [82,83]. One principle is the
conventional turbulent mixing ventilation based on the principle of dilution. In an OR with
mixing ventilation, turbulent air streams are supplied into the room through the ceiling- or
wall-mounted diffusers and extracted near the floor, which creates highly mixed airflow
patterns throughout the entire OR. The incoming air is mixed with infected air in the OR and
thus dilutes the airborne bacterial contamination. The mixing ventilation aims to create a
uniform low concentration of contaminated air in the operating room. When a sufficient
amount of air is supplied, the contaminant concentration can be reduced to a level that is
associated with an acceptable risk of infections.

Modern ORs tend to adopt the other principle based on the displacement of contaminated air
– laminar airflow (LAF) ventilation, which directs a linear stream of low-turbulence air to the
entire operating zone at a high momentum. While not truly laminar, the entire body of air
flows at nearly identical speed and along parallel streamlines. This creates a unidirectional
airflow pattern with little turbulence in the surgical zone and avoids mixing with surrounding
air. Thus, LAF ventilation is also referred to as unidirectional airflow (UDF) ventilation. The
high speed of incoming air makes it possible to directly wash away the BCPs generated
within the surgical critical zone. To maintain sufficient momentum of the airflow and ensure
an adequate washing effect, the discharge velocity at the supply diffuser must be high enough.
Therefore, the airflow rate necessary for LAF to achieve expected performance can be several
times greater than those of conventional mixing ventilation [84–86].

The concept of LAF originated from the electronic and aerospace industry that demands a
particle-free environment. Generally, the air is introduced through a HEPA filter, which
removes 99.97% of particles greater than 0.3 µm in size. The combination of LAF and the
HEPA filter, defined as ultraclean ventilation, was first introduced to OR use in 1960s [87].
To date, LAF has predominately been used for infection-prone surgeries (e.g. orthopedics and
vascular implant surgeries) to minimize the microbiological contamination of wounds and
thus the risk of SSIs [88,89]. The performance of the LAF ventilation has been extensively
investigated and thoroughly compared with the mixing ventilation. The majority of literature,
including both measurement- and simulation-based studies, has suggested that LAF is
superior to mixing ventilation, with respect to minimizing the contamination of wounds with
airborne bacteria [64,65,71,90–92]. Memarzadeh and Manning [65] numerically evaluated
different ventilation schemes and concluded that laminar airflow, if appropriately designed,
was the best option for controlling the risk of BCPs deposition on the critical surgical site.
Some clinical evidence has also confirmed the reduced SSIs rates with LAF ventilation

9
CHAPTER 2. LITERATURE REVIEW

compared with conventional mixing ventilation [88,89,93]. In the landmark study of Lidwell
et al. [93], the protective effect of LAF was evaluated in a multi-center controlled trial of
8055 orthopaedic procedures in 1970s and the SSIs rate decreased from 1.5% in the control
(non-LAF) group to 0.6% in the LAF group. Bosanquet et al. (2013) undertook a
retrospective review of 170 vascular procedures performed over a one-year period and found
a statistically significant reduction in SSIs incidence with LAF.

However, some recent publications have questioned the protective effect of LAF in
preventing SSIs and even suggest that the use of LAF may have a negative effect on infection
rates [94–99]. Brandt et al. [94] performed active SSIs surveillance of 99,230 orthopaedic and
general surgical procedures in 63 surgical departments stratified according to the type of
ventilation: conventional mixing ventilation with HEPA filtration and HEPA-filtered LAF
ventilation. The authors noted a significantly higher risk for severe SSIs after hip prosthesis
using LAF ventilation, with an odds ratio of 1.63 as compared with the conventional mixing
ventilation. Hooper et al. [97] made a retrospective review of data from New Zealand Joint
Registry and documented a statistically significant increased deep infection rates after
prosthetic hip and knee surgery in LAF ventilated ORs compared with conventionally
ventilated ORs. A systematic review and meta-analysis by Bischoff et al. [99] suggests that
LAF ventilation confers no benefit compared with conventional mixing ventilation in
reducing the risk of SSIs in orthopaedic and abdominal surgery. Based on their findings and
cost-effectiveness analyses in the literature, the authors strongly recommend that LAF
ventilation should not be installed in new operating rooms.

The contradictory clinical evidence in the literature may be ascribed to the inappropriate
design and use of LAF, poor compliance with OR protocols, and OR traffic [100]. Diab-
Elschahawi et al. [85] evaluated the impact of LAF ceiling size and noted significantly
reduced bacteria contamination of the instrument tables in the larger sized LAF. Korne et al.
[101] assess the compliance with best practice in the placement of instruments tables and
other sterile devices within the LAF protection zone and found a compliance rate of less than
10%. Smith et al. [102] noted that increased OR traffic, measured by the number of door
openings, was associated with significantly increased bacterial contamination. Other studies
also show that the air supply velocity, the convective heat dissipation, and the obstructions in
the airflow paths such as medical lamps, as well as personnel movement in the OR, may
easily disrupt the unidirectional airflow and hence strongly influence the BCPs dispersion and
deposition [103–111]. Cao et al. [112] measured the air speed above the lying patient in two
real ORs equipped respectively with LAF and mixing ventilation and found that the airflow
distribution of LAF could be greatly affected by surgical facilities. They therefore considered
mixing ventilation a more robust method of delivering air to ORs. In a recent study, Jain and
Reed [113] reviewed the current evidence behind LAF and provided important
recommendations for the effective use of LAF in the operating room environment.

The installation and validation of LAF are more expensive than those of conventional mixing
ventilation [68,99]. LAF also requires continuous technical maintenance and results in

10
 2.4. DOOR OPENING AND FOOT TRAFFIC

considerably high energy use [64,100]. Evans [114] documented a cost of 60,000–90,000
USD for the installation of an exponential LAF system in a new OR. An Italian study reported
that LAF was 24% more expensive than the conventional mixing ventilation in terms of
installation costs and 34% more expensive in terms of operation costs [115]. The installation
of LAF in an existing OR can be even more prohibitive due to site and financial constraints.
Thus, a mobile laminar airflow unit is sometimes used in combination with turbulent mixing
ventilation to improve the air cleanliness at the surgical area and instrument tables [116–119].
Friberg et al. [119] evaluated the performance of a mobile LAF unit as an addition to the
conventional mixing ventilation and found that the use of a mobile LAF unit reduced the
bacterial contamination to the same level as the complete LAF ventilation.

2.3.2 Temperature-Controlled Airflow

An alternative ventilation scheme, named temperature-controlled airflow (TAF), has recently

been developed for OR use [24]. In contrast to the conventional mixing and LAF ventilation,
TAF supplies air at different temperature levels and controls the air movement based on the
temperature gradient. Specifically, cool air is discharged into the operating zone from ceiling-
mounted central diffusers to displace the contaminated air, whereas warm air is dispersed into
the periphery of the OR through surrounding diffusers to further dilute the airborne BCPs
concentration. The central supply air is kept about 1.5–3°C cooler than the ambient air in the
OR and thus the incoming air is accelerated by gravity due to density difference and falls
down to the surgical zone. The downward air streams are expected to overcome the
convective currents developed from the heat sources – that is, the surgical staff, surgical
equipment and lamps – and wash BCPs away from the patient and sterile instruments. The
downward air velocity, dictated by the temperature difference between the incoming air and
the ambient room air, should be sufficient for the clean air to reach the surgical site and
provide the adequate washing effect. Hence, TAF can be deemed as a combination of features
of conventional mixing ventilation that dilutes the microbiological contamination and
unidirectional airflow ventilation that directly removes the airborne BCPs.

The TAF ventilation system has so far been installed in several hospitals in Sweden,
Germany, Switzerland, the Netherlands and the United States. Based upon the information
from Swedish hospitals, the installation cost for the TAF system is higher than for
conventional mixing ventilation systems. However, our research works have shown that the
TAF ventilation is more capable of providing an ultraclean environment than the mixing
ventilation and is more energy-efficient than the LAF ventilation.

2.4 Door Opening and Foot Traffic

ORs may have door openings in direct connection to uncontrolled areas such as corridors that
are more contaminated. Ljungqvist et al. [120] measured the airborne BCPs concentration in

11
CHAPTER 2. LITERATURE REVIEW

the corridor and found that the level of contamination could be as high as 180 CFU/m3. A
well-designed ventilation system should not only establish an appropriate airflow pattern to
efficiently dilute or remove BCPs that originate from inside the OR but also be able to prevent
the migration of contaminants from the surrounding hospital environment. To that end, ORs
are usually maintained at a positive pressure relative to adjacent spaces to prevent the entry of
contaminated air through gaps and cracks from areas outside the OR. The positive pressure
acts as a barrier against spaces with lower air cleanliness as contamination is transferred in the
direction of decreasing static pressure [121]. Pressure difference has been proved to be an
effective measure to realize static isolation in clean rooms [122] and hospital isolation rooms
[123]. Many national standards and guidelines also mandate positive pressure in ORs to keep
out airborne contaminants. However, opening the OR door can defeat the positive pressure
and allow air exchange between two spaces. Instantaneous airflow measurements in an empty
OR have shown that the door opening permits an inflow of air even under isothermal
conditions [124].

The mechanism driving the airflow through an open door is a combination of several factors:
the temperature difference, the pressure difference, ventilation airflow, human passage, and
the motion of the door itself. Some early studies attempted to derive analytical expressions to
quantify the air volume migration as a function of temperature difference [125–128]. More
recent research has focused on the effect of door opening and closing on the containment
failure in isolation rooms using scale models [121,129,130,130–133], full-scale mock-ups
[134–136], and CFD simulations [137–144]. The majority of such studies have concentrated
on hinged doors and evaluated the door swing pumping effect caused by the motion of a
hinged door [129–132,137,140]. While hinged doors are common in hospital isolation rooms,
modern ORs tend to use sliding doors. Several studies have compared sliding doors and
hinged doors and found that sliding doors induce smaller air volume exchange than hinged
doors [133,135,136,139]. However, further analyses showed that the gravity (or buoyancy)-
driven flow dominates the air exchange through the doorway [142] and the difference
between the two types of doors becomes insignificant when a temperature difference is
involved [139].

The frequency and duration of door openings depend on the type of surgery. Lynch et al.
[145] recorded 5–87 door openings per hour from 28 procedures, with one door cycle taking
approximately 20 seconds. Elliott et al. [146] observed that the average door openings ranged
from 33 per hour in general surgery to 54 per hour in cardiac surgery. For total joint
arthroplasty, several observational studies reported 0.62–0.85 mean door openings per minute
[102,145,147]. The majority of literature has found a strong negative impact of door openings
and corresponding foot traffic on the OR environment. Statistical studies based on
measurements during surgery have shown that door openings are associated with increased
environmental contamination in ORs [47,92,102,148–150]. For instance, Mathijssen et al.
[148] suggested that a single door opening contributed to a 5% increase in the odds of
microbial contamination ≥20 CFU/m3. However, Alsved et al. [24] found no significant

12
 2.4. DOOR OPENING AND FOOT TRAFFIC

correlation between the total number of door openings per surgery and the average BCPs
concentration at the wound. As explained by the authors themselves, this was due the low
number of door openings and low BCPs concentration in the corridor outside the OR. Several
studies also show that door openings and corresponding foot traffic are directly associated
with the rate of SSIs. Babkin et al. [151] found that keeping the door shut led to a decrease in
SSIs rates. Crolla et al. [152] noted a significant correlation between the development of SSIs
and higher number of door openings in a cohort study of 1537 colorectal procedures. These
studies confirm that controlling OR foot traffic is critical for the prevention of infections.

13
 3.1. PHYSICAL CASES

Chapter 3

Methodology

Experiments and CFD simulations are two approaches commonly adopted to investigate
indoor air movement, temperature, and contaminant distribution. Compared with the
experimental approach, CFD can provide rich information about the air distribution and offer
great possibilities to conduct parametric analyses. Such features have made CFD a powerful
tool in analyzing and designing indoor ventilation systems. When it comes to ventilation in
ORs, it is rather difficult to conduct high-quality measurements due to logistical and ethical
constraints. Many previous studies have utilized CFD to predict the ventilation airflow and
particle dispersion (e.g. [21,25,103,104,106,153,154]). In this thesis, only the CFD approach
is adopted to investigate the performance of OR ventilation schemes and relevant influential
factors. To ensure the accuracy and reliability of the simulation results, the numerical models
used in the CFD simulations are validated against well-documented benchmark measurements
available in the literature.

3.1 Physical Cases

3.1.1 Helsingborg General Hospital

The ORs in Helsingborg General Hospital (Helsingborgs lasarett) are selected as the physical
model for Papers II, IV, and VI. Helsingborg General Hospital is an emergency care hospital
that operates 24 hours a day, located in Helsingborg, Sweden. There exist three rectangular-
shaped ORs in the hospital, which measure L 6.4 m × W 6.3 m × H 3.0 m. They are identical
in size but use three distinct ventilation schemes: LAF, TAF and mixing ventilation. The
selected ORs are used by the Orthopaedic Surgery Department, which hosts approximately
2500 orthopaedic surgical procedures each year [24]. The physical configurations of the ORs
are presented in Figure 1.

As depicted in Figure 1a, the first OR is ventilated by a vertical LAF unit at an airflow rate of
12000 m3/h (i.e. 100 ACH). Clean air is supplied to the OR from a (2.78 m × 2.78 m) ceiling
diffuser, resulting in a supply velocity of 0.43 m/s. Partial walls enclosing the ceiling diffuser

15
CHAPTER 3. METHODOLOGY

Figure 1. Isometric view of the OR configurations in Helsingborg General Hospital.

are adopted to strengthen the unidirectional airflow. The partial walls terminate at 2.1 m
above the floor, which separate the clean zone and the periphery of the OR. Room air is
extracted mainly through eight (0.90 m × 0.38 m) top vents in immediate adjacency to the
partial walls and partly through three (0.6 m × 0.3 m) side vents on one sidewall.

The second OR is provided with TAF ventilation, as shown in Figure 1b. The cool air is
discharged into the surgical zone through eight semi-spherical diffusers evenly mounted in a
circular platform, whereas the warm air is dispersed into the periphery of the OR through
eight additional ceiling-mounted diffusers in a parallel arrangement. All diffusers are identical
in size and each one has a surface area of 0.16 m2. The TAF ventilation operates at an airflow
rate of 5600 m3/h (i.e. 46 ACH). The supply air is evenly distributed between the diffusers
and exhausted by four (0.56 m × 0.46 m) low-level vents, located on two opposite-facing
sidewalls near the corners of the OR.

Figure 1c presents the third OR with conventional turbulent mixing ventilation, which
operates at an airflow rate of 3200 m3/h (i.e. 26 ACH). The supply diffuser is located on the
top of one sidewall, over the full length of the wall with a width of 0.64 m. Outgoing air is
evacuated through six (0.6 m × 0.3 m) return grilles on the opposite sidewall: three at the
ceiling and three at the floor levels.

16
 3.1. PHYSICAL CASES

3.1.2 New Karolinska Solna University Hospital

The OR designed for the New Karolinska Solna University Hospital (NKS) is chosen as the
physical model in Papers III, V, and VII. NKS is the new university hospital built in Solna,
Sweden. The OR has an overall dimension of L 8.6 m × W 7.7 m × H 3.2 m. The OR is
separated from the adjacent corridor (COR) by a partition wall with a thickness of 0.2 m. The
corridor measures L 7.5 m × W 2.8 m × H 3.2 m. The partition wall was opened up to create a
doorway, which measures L 1.5 m × H 2.1 m. Figure 2 shows an isometric view of the OR
and corridor.

The OR is installed with a turbulent mixing ventilation system. As shown in Figure 2, clean
air is supplied to the OR through 24 diffusers in the ceiling at a total airflow rate of 2.0 m3/s.
Each diffuser has a dimension of 0.6 m × 0.6 m and contains 9 × 9 nozzles. The arrangement
of the nozzles directs the majority of the supply air to the periphery and rest to the center of
the OR. The supply opening that hosts all supply diffusers covers a projected area of 8.64 m2.
Air is extracted from the OR through four (0.90 m × 0.45 m) low-level exhaust vents.

The LAF and TAF ventilation had also been proposed in the design phase of the OR. The
LAF ventilation was described in Paper V and the TAF ventilation in Paper VII.

Figure 2. Isometric view of the OR configuration and the adjacent corridor in the New
Karolinska Solna University Hospital.

17
CHAPTER 3. METHODOLOGY

3.2 Airflow Simulation

3.2.1 Governing Equations

The airflow simulation model is built upon the fundamental laws of physics – the
conservation law of mass, momentum, and energy. A set of equations that govern the three-
dimensional behavior of airflow are derived from the conservation laws. These equations can
be written in the general form of a transport equation, which consists of a transient term, a
convection term, a diffusion term and a source term [155]:

() 
 (V )   ()  S (1)
t

where 𝜌 is the air density, 𝑉⃗ is the air velocity vector,  denotes the transported quantity, Γ
is the (effective) diffusion coefficient of  , and S is the source term. By assigning the right
values for  , Eq. 1 can represent any of the conservation equations. When   1 , Eq. 1
becomes the continuity equation, which describes the conversation of mass; when 
represents each of the three velocity components, Eq. 1 turns into the momentum equations,
which dictates the conservation of momentum; when the air enthalpy is assigned to  , Eq. 1
becomes the energy equation.

3.2.2 Turbulence Modelling

One of the major challenges in indoor airflow simulation is the turbulent flow. The complex
state of indoor airflow and heat transfer imposes strict requirements for appropriate turbulence
treatment. The Direct Numerical Simulation (DNS) or Large Eddy Simulation (LES) is
capable of resolving turbulence with sufficient accuracy. However, the required
computational resources are highly demanding. Due to the scale and complexity of indoor
airflow simulations, DNS or LES has still been beyond the scope of practical applications.
Due to its low computational cost and reasonable accuracy, the Reynolds Averaged Navier-
Stokes (RANS) simulation serves as a good alternative.

The RANS approach relies on the closure of the Reynolds stress tensor, which can either be
directly computed from transport equations for each individual component (Reynolds-stress
models) or modelled by the eddy-viscosity hypothesis – a relationship with the mean flow
through the turbulent viscosity. Among all eddy-viscosity based turbulence models, the two-
equation 𝑘 𝜀 models are the most commonly used ones for a broad range of engineering
applications and indoor airflow simulations [156]. The 𝑘 𝜀 models solve two transport
equations for turbulent kinetic energy 𝑘 and dissipation rate 𝜀 . The turbulent viscosity is
constructed through t  Ck2 /  , where C  is a critical coefficient. The two-equation 𝑘 𝜀

18
 3.2. AIRFLOW SIMULATION

models are computationally efficient and stable, compared with more sophisticated Reynold
stress models [157].

The present study adopts the Realizable 𝑘 𝜀 model, a relatively recent variant of 𝑘 𝜀
models developed by Shih et al. [158]. In the Realizable 𝑘 𝜀 model, the transport equation
for 𝜀 is derived from an exact equation for the transport of the mean-square vorticity
fluctuation. The derivation of the model satisfies certain mathematical constraints on the
Reynolds stresses and thus the model is consistent with the physics of turbulent flows.
Compared with other variants of 𝑘 𝜀 models, the Realizable 𝑘 𝜀 model is expected to
provide superior performance for flows involving boundary layers under strong adverse
pressure gradient. The Realizable 𝑘 𝜀 model has shown its capability of accurately
predicting indoor air flow and heat transfer in previous research works (e.g. [21,25,153,159])
and has been proven to be appropriate for complex indoor environment simulations [160,161].
In the Realizable 𝑘 𝜀 model, C  is not a constant like in the standard 𝑘 𝜀 model but
varies as a function of mean flow and turbulence properties. More detailed descriptions of the
model can be found in Shih et al [158].

3.2.3 Boundary Conditions

The conditions at the boundaries of the computational domain need to be specified in order to
solve the aforementioned equations. The correct setup of boundary conditions plays a crucial
role in achieving valid and reliable simulation results. The boundary conditions adopted in all
simulations in this study are given below.

The no-slip condition is used for velocity at solid surfaces including all OR walls and exposed
surfaces of staff and furnishings. Constant heat flux is specified at the heat-emitting surfaces
of internal heat sources including the surgical staff, medical equipment and the downward-
point face of surgical lamps. Other solid surfaces are all set as adiabatic walls (i.e. zero heat
flux). The thermal radiation is neglected and thus all sensible heat loads are taken as 100%
convective. Due to the large ventilation rate in ORs, this simplification proves to have little
impact on the simulation results and conclusions [106,162].

The ‘pressure-outlet’ condition is specified at air exhaust openings in all simulations except
the ones in Paper III where the ‘outlet-vent’ condition is adopted. The ‘velocity-inlet’
condition is used at the air supply openings. Depending on the type of supply diffusers, a
uniform or pre-simulated non-uniform velocity profile is imposed at the air inlets. The inlet
conditions for corresponding turbulent quantities are determined from the turbulent intensity
and hydraulic diameter at the supply openings.

A turbulent intensity ranging 5–20% is selected in the simulations based on the type of
ventilation. The supply air temperature at the inlets is taken as 20°C for LAF and mixing
ventilation. In the TAF ventilation, the central diffusers supply air at 18.5 ℃, whereas the

19
CHAPTER 3. METHODOLOGY

temperature at the surrounding diffusers is adapted so as to establish a 1.5–3°C temperature

gradient.

3.2.4 Numerical Aspects

The aforementioned conservation equations are solved by the finite volume method (FVM) –
the most common approach in CFD. The computational domain is subdivided into a finite
number of small control volumes (i.e. cells). The control volumes do not overlap and fill the
spatial domain completely. The solution variables are defined at the centroid of control
volumes and the conservation of relevant properties is satisfied in each control volume. By
integrating the conservation equations over each control volume, FVM converts the partial
differential equations into a system of algebraic equations.

In FVM, the volume integral of a differential term is transformed into the surface integral
using the divergence theorem, which is evaluated as fluxes at the faces of each control
volume. The face values of relevant variables are calculated from the centroid values of
neighboring control volumes by various discretization schemes. Therefore, the discretization
schemes are important in determining the accuracy of the numerical solutions. In the present
study, the ‘second-order upwind’ scheme is used to discretize all the convective terms and the
central-difference scheme is adopted for the discretization of the diffusion terms. A staggered
scheme named PRESTO! is used to compute pressure values at cell faces. The PRESTO!
scheme is recommended for flows with high swirl numbers and high-Rayleigh number natural
convection. For transient simulations, a first-order implicit formulation is employed for the
temporal integration.

The resulting algebraic equations are non-linear and coupled with each other. An iterative
process is necessary to simultaneously solve these equations. The widely used procedure in
CFD to solve the discretized flow equations is the Semi-Implicit Method for Pressure Linked
Equations (SIMPLE). Therefore, the SIMPLE algorithm is chosen in this work for both
steady and transient simulations.

3.2.5 Mesh and Convergence

For indoor environment simulations with contaminant dispersion, the mesh should be
sufficiently fine to accurately resolve the turbulent flow field and compute particle
trajectories. Mesh refinement is applied to the regions with high gradients of transported
quantities (e.g., air inlets, air outlets, and surgical critical zones) to capture the detailed flow
characteristics.

The meshes used in the present work are sufficiently fine to resolve the near-wall viscous
sublayer. Consequently, the log-law wall functions are no longer valid for such fine meshes.
The ‘Enhanced Wall Treatment’ is thus employed to treat the turbulent boundary layers in the

20
 3.3. DOOR MOTION SIMULATION

near-wall region. EWT is a hybrid near-wall formulation, which combines a two-layer model
with a blending function to implement a smooth transition between the viscosity-affected
region and the fully-turbulent region. The use of EWT alleviates the strict requirements for
gird resolutions in the near-wall region, as it is designed to work well for a diversified mesh
resolution ranging from low-Reynolds-number meshes to intermediate meshes and to wall-
function meshes.

Grid convergence analyses are always performed to ensure reasonable grid-independent

solutions. The mesh is continuously refined until no significant differences in airflow
properties and particle concentration can be observed. Numerical convergence is judged using
several criteria. The convergence criterion for the scaled residuals is set 1  108 for the
energy equation and 1 105 for all other equations. The overall mass and heat balance is also
checked and the net imbalances is less than 0.5% of the smallest flux through the domain
boundary upon completion of the simulation. In addition, to guarantee a stable steady
solution, values of solution variables are monitored at specified locations to ensure no
significant changes between iterations.

3.3 Door Motion Simulation

The simulation of transient flow due to the motion of a door or any object can be handled
directly and indirectly [163]. The direct simulation implements the real movement of the real
‘object’ inside the computational domain, which requires updating the volume grid at each
instant in the region affected by the motion. Some previous studies have adopted the dynamic
mesh technique (e.g., mesh deformation and remeshing) to simulate the door opening and
closing [138,143,164]. During runtime, a new mesh is generated at each time step by
deforming the old mesh or remeshing the deforming zone accordingly. The dynamic mesh
simulation demands substantial computational cost and is also face with mesh quality issues.
The mesh-to-mesh interpolation also introduces errors into the solution. The indirect
approach, by contrast, considers only the influence of the object movement on the flow,
which is mainly based on defining specific source terms in the governing equations [163]. By
applying a momentum source to the flow with assigned values in the region containing the
object at a specific instant, the fluid is forced to move following the motion of the object. This
indirect approach has also been adopted in previous research works to simulate the door
opening and closing [141,142,165]. The main disadvantage of the indirect approach is the
uncertainty and inaccuracy. Applying a momentum source is not able to fully stop the flow
through the door body and a small leakage occurs.

Paper III adopts a simplified direct approach to simulate the motion of a sliding door. In this
simplified approach, the door is modelled as an internal wall of zero-thickness. The boundary
mesh at this internal wall is divided into a number of rectangular sections, each corresponding
to a specific door position. During the door opening and closing process, the effect of the door

21
CHAPTER 3. METHODOLOGY

Figure 3. Modelling the motion of a sliding door – sequentially changing the boundary type
from the 1st to 50th sections.

motion is obtained by running a Fluent Scheme script that continually updates the type of the
boundary mesh for each door position. Despite having discrete door motions, this approach
can overcome the numerical difficulties in the dynamic mesh technique and avoid the
uncertainty in the indirect approach. A similar approach was employed by Carneiro et al.
[166] and achieved success in simulating the opening of both a sliding door and a hinged door
in cold rooms.

In Paper III, the complete cycle of door opening and closing takes 20 s. The door starts to
open at t=0 s and becomes fully open at t=5 s. The door remains fully open for the next 10 s
(i.e. up to t=15 s) and takes another 5 s to close (from t=15 s to t=20 s). The door moves at a
speed of 0.30 m/s. The model of the sliding door is shown in Figure 3. A total of 50 sections
are created to simulate the door motion; so each section corresponds to an opening width of
0.03 m for a total opening of 1.5 m. Thus, the type of the boundary mesh zones is updated
every 0.1 s during runtime in the simulation.

22
 3.4. PARTICLE TRANSPORT SIMULATION

3.4 Particle Transport Simulation

3.4.1 Eulerian and Lagrangian Formulation

The numerical simulation of particle transport carried by turbulent airflow requires the
modelling of the continuous phase – the airflow, the discrete phase – the particles, and the
interaction between them. The airflow is most commonly modelled using the Eulerian
formulation, as presented in the previous subsection. Approaches begin to differ in the
modelling of particle motions. There are generally two distinct strategies for modelling
particles transport: the Eulerian method and the Lagrangian method. The Eulerian approach
treats the particle phase as a continuum and solves the particle statistics from a transport
equation in the form of Eq. 1, established using particle and flow field properties. The particle
concentration is thus obtained at each point of the numerical domain. The Lagrangian
formulation, on the contrary, directly considers the discrete phase of particles and tracks the
pathways of each individual particle based on the equation of motion. By tracking a large
number of particles, the spatial and temporal development of particle dispersion can be
predicted from the statistics of particle trajectories. Thus, the Lagrangian method is often
referred to as Lagrangian particle tracking (LPT).

The Lagrangian formulation provides a deeper insight into particle dynamics than the
Eulerian formulation. Nevertheless, the Lagrangian formulation is rather computationally
expensive as a sufficiently large number of particle trajectories need to be computed to obtain
statistical convergence. In the Eulerian formulation, on the other hand, the particle
concentration becomes stable once the simulation is converged. Furthermore, additional post-
processing is required in the Lagrangian framework to obtain the statistical information such
as the particle concentration from an enormous number of computed particle trajectories.

The applications of the two approaches in indoor environments have been extensively
validated against experiment in the literature. Good agreements with measured data have been
observed for both approaches [167–171].

3.4.2 Lagrangian Particle Tracking

Both the Eulerian and Lagrangian approaches have been successfully applied in previous
studies to simulate BCPs transport in ORs (e.g., Ref. [103,104,106] using the Eulerian
approach; Ref. [21,25,153] using the Lagrangian approach). The present study adopts the
Lagrangian approach, which predicts particle trajectories by solving the following particle
motion equation
 
du p   g (  p   ) 
 FD (u  u p )  F (2)
dt p

23
CHAPTER 3. METHODOLOGY

where 𝑢 is the velocity, ρ is the density, and g is the gravitational acceleration; the subscript
𝑝 denotes particle related variables whereas unsubscribed symbols refer to airflow quantities.
Eq. 2 is a direct result from the Newton’s second law for a particle in spherical shape. The
location and velocity components of a particle can be computed by integrating Eq. 2 along the
particle trajectory.

The left hand side of Eq. 2 represents the inertial force (per unit mass) and the terms on the
right hand side are body forces that act on the particle. The first term on the right hand
represents the drag force (per unit mass), which is linked to the air-particle velocity difference
through the inverse of particle relaxation time 𝐹 – to be discussed below. The second term
refers to the gravitational or buoyant force due to the action of gravity. The third term is used
to incorporate additional forces exerted on the particle, which may, depending on the particle
properties and flow conditions, include the pressure gradient force, the thermophoretic force
due to temperature gradient, the Brownian force, the Saffman’s lift force due to shear, and
unsteady forces such as the virtual mass effect and Basset force caused by the acceleration of
particles with respect to the airflow.

According to the analysis by Zhao et al [172], the pressure gradient force, the Basset force,
and virtual mass effect are at least two orders of magnitude smaller than the drag force and
can be neglected without compromising the accuracy. The Brownian force and Saffman’s lift
force may become comparable to the drag force when sub-micron particles are considered in a
flow field. Chang and Hu [173] investigated the transport mechanisms of size-dependent
airborne particulate matters in the indoor environment. They concluded that the Brownian
force cannot be neglected for particle size smaller than 0.5 µm and the Saffman’s lift force is
important for a specific range of particle size between 2.5 and 5 µm.

FD is the inverse of particle relaxation time  , computed by

1 18 CD Re p
FD   (3)
  p d p2 24

where 𝜇 is molecular viscosity of the air, 𝜌 is the particle density, 𝑑 is the particle diameter,
𝐶 is the drag coefficient, and 𝑅𝑒 is the particle Reynolds number defined as

 
d p u  up
Re p  (4)


The drag coefficient 𝐶 in Eq. 3 is dependent on the particle Reynolds number. For a smooth
spherical particle, 𝐶 can be evaluated from Morsi and Alexander’s empirical drag law [174]

a2 a
CD  a1   32 (5)
Re p Re p

24
 3.4. PARTICLE TRANSPORT SIMULATION

where 𝑎 , 𝑎 , and 𝑎 are piecewise constants that apply over a broad range of 𝑅𝑒 .

The modelling of interaction between airflow and particles and between particles themselves
depends on the strength of coupling. The types of coupling can be determined by the volume
fraction of particles [175]. For indoor aerosol applications, the particle volume fraction is
sufficiently low and the impact of particles on the airflow field is negligible [170,176]. Thus,
a one-way coupling between airflow and particles is assumed in the present study, that is,
from airflow to particles but not vice versa. In steady-state simulations, the turbulent airflow
field is resolved in the first place and particles are injected into the flow domain afterwards. In
dynamic simulations, the particle tracking always coincides with the flow time of the transient
flow solver.

Particle trajectories terminate at the boundaries of the computational domain. Specifically,

particles are considered as leaving the domain when they reach the air exhaust openings (i.e.,
the ‘escape’ boundary condition). On the other hand, particles are assumed to be ‘trapped’
after impacting onto a solid surface, since they usually cannot accumulate enough rebound
energy to overcome adhesion [177]. Such a treatment is typical for indoor aerosol applications
and has been adopted by a majority of previous studies (e.g. [170,172,176,178,179]).

3.4.3 Discrete Random Walk

The dispersion of particles due to turbulence is predicted by the Discrete Random Walk
(DRW) model. The DRW model takes into account the effect of instantaneous turbulent
velocity fluctuations on the particle trajectories using stochastic tracking. Specifically, the
turbulent dispersion of particles is predicted by integrating Eq. 2 for each individual particle
along the particle path, using the instantaneous flow velocity expressed as

u  u  u' (6)

which is the sum of the mean airflow velocity 𝑢 and the fluctuating velocity 𝑢 . Assuming
isotropic turbulence, the component of the fluctuating velocity can be evaluated from the
turbulent kinetic energy

u'i   i u'i 2   i 2k / 3 (7)

where  is a random variable following Gaussian distribution with zero mean and unit
variance. The value of  is kept constant over a time interval of one eddy interaction, which is
the smaller of the eddy lifetime t e and the eddy crossing time t cross

25
CHAPTER 3. METHODOLOGY

tint  min(te , tcross )

k
te  2TL  2CL
 (8)
le
tcross   ln(1    )
 u  up

where t in t is the interaction time between particles and the fluid eddy, TL  CLk /  is the fluid
Lagrangian integral time with C L  0.15 for the 𝑘 𝜀 turbulence models,  is the particle

relaxation time, and le  C k

3/4 3/2
/  is the eddy length scale. When the interaction with the
current eddy is ended (i.e., t is reached), a new value of the fluctuating velocity will be
obtained by applying a new value of  .

With the DRW model, two particles with the same initial condition may have different
trajectories. When a sufficient number of particles are tracked, the stochastic tracking leads to
an equivalence of effective diffusion of particles in turbulent flows.

The isotropic turbulence is implied in the DRW model when obtaining the instantaneous flow
velocity. In the presence of walls, however, particle deposition is dictated by phenomena in
the boundary layer. Within the viscous sublayer, the flow is highly anisotropic and the normal
component of turbulence intensity is order of magnitude smaller than its counterparts parallel
to the wall (i.e., stream-wise and span-wise). The isotropic decomposition of the turbulent
kinetic energy may lead to over-prediction of particle deposition [180,181]. Thus, researchers
have developed different variants of the standard DRW model to account for the anisotropic
turbulence in the boundary layer. However, it is noted that the DRW model works well when
the near-wall grid is sufficiently fine, like those used in DNS simulations [182,183]. In this
study, an inflation of boundary layers is applied to the walls with sufficient mesh refinement
to resolve the near-wall viscous sublayer. The resulting non-dimensional wall distance of first
layer cells has been kept in the range 1 𝑦 5 for the majority of the wall surfaces. In
addition, Lai and Chen [171] found that super-micron particles are not sensitive to the near-
wall turbulence anisotropy. On the other hand, the unsmooth and rough wall surfaces in
operating rooms may lead to more particle deposition when the dispersed particles have a
diameter of 1-10 µm [184]. Considering the size of BCPs, the standard DRW model
employed in the present study is capable of providing sufficient accuracy in the prediction of
particle transport.

3.4.4 Particle Concentration

Both the surface BCPs contamination (in CFU/m2/h) and airborne (volume) BCPs
contamination (in CFU/m3), representing the results of passive and active air sampling in
measurements respectively, are reported in the present study. To mimic the passive sampling,

26
 3.4. PARTICLE TRANSPORT SIMULATION

the entire upper surface of the operating table and instrument tables are taken as the settle
plates. The number of particles that impact onto these surfaces is monitored and scaled,
assuming that the OR is exposed to airborne contaminants over a period of one hour. The
airborne contamination is evaluated as the average volume concentration in a pre-defined
volume centrally positioned above the operating table.

The volume concentration of particles can be computed from the particle trajectory
information through the particle source in-cell (PSI-C) scheme [176]. The following formula
calculates the particle number concentration


N
n  ti , j
i 1 i
Cp, j  (9)
Vj

where index i and j denote the i  th trajectory and the j  th cell respectively, C p is the
mean particle concentration in a cell, V is the volume of the computational cell, 𝑛 is the
number flow rate of each trajectory, and ti , j is the time required for a particle to traverse the
j  th cell along the i  th trajectory, that is, the particle residence time.

Due to the stochastic nature of the DRW model, an adequate number of trajectories need to be
computed to reach statistical convergence. As the number of trajectories increases, the
fluctuation between solutions diminishes. The appropriate number of trajectories is case
dependent. It is suggested that the statistical analysis should start from a sample size that is
comparable to the number of concentration cells [176]. In the present study, the number of
trajectories is continually increased until the particle concentration becomes statistically
stable.

Certain engineering standards and design guidelines for hospital ventilation recommend the
use of the dilution model to estimate the level of contamination at a given airflow rate or the
required airflow rate for a given level of contamination. Assuming perfect dilution with fully
mixed airflow, the airborne contamination can be estimated from the following analytical
formula:

n p  qs
c (10)
Q

where c is the airborne BCPs concentration (CFU/m3), n p is the number of persons (i.e.,
surgical staff), qs is the source strength of BCPs releasing from a person (CFU/s), and Q is
the airflow rate (m3/s). The dilution model implies that increasing the airflow rates will
proportionally decrease the airborne contamination.

27
CHAPTER 3. METHODOLOGY

3.5 Model Validation

In order to apply the numerical models to the investigation of ventilation performance and
particle dispersion in ORs, it is necessary to validate the models to confirm their accuracy and
efficacy. Ideally, experiments should be conducted and numerical models should be validated
against measured data before applying them to simulations. However, it is rather difficult to
conduct high-quality measurements in ORs due to logistic and ethical constraints. Therefore,
we validate the numerical models against well-documented benchmark experiments.

3.5.1 Validation Study I – Heat Valve Ventilation

The airflow model is firstly validated in Paper I, in which the performance of two air supply
configurations was evaluated in a heat valve ventilation system. The air temperature and
velocity is studied both experimentally and numerically. The experiment was conducted in a
full-scale chamber, which measured L 3.1 m × W 3.6 m × H 2.7 m. The chamber represented
a room in a low-energy residential building that consisted typically of a floor, a ceiling, three
internal walls and one external wall with a window. The chamber was contained in a
laboratory hall. The chamber walls were well insulated and the temperature inside the hall
was kept as close as possible to the temperature inside the chamber in order to minimize the
heat transfer through the chamber walls and mimic the adiabatic condition. A cooling panel
with a total area of 2.1 m2 was located on one chamber wall to represent the window. The
geometry of the chamber was given in Figure 4a. In Configuration I, air was supplied to the
chamber through a circular valve located below the window (X = 1.8 m; Y = 0.0 m; Z = 0.4
m). In Configuration II, air was supplied through three nozzles placed in the upper part of the

a) b)

Figure 4. Physical configuration of the experimental chamber: a) isometric view of the

chamber; b) front view of the measurement sections (dashed outlines), sampling lines (blue
lines) and sensors (red dots).

28
 3.5. MODEL VALIDATION

Figure 5. Comparison of simulated and measured temperature profiles at three sampling lines
(L1–L3) in Section C: a) Configuration I (air supplied through the circular valve); b)
Configuration II (air supplied through three nozzles).

wall opposite to the window (X = 1.8 m; Y = 3.1 m; Z = 2.5 m). In both configurations, air
was supplied at 51 °C and an airflow rate of 7 l/s (0.84 ACH). Room air was extracted
through the gap below the door leaf. The air velocity and temperature were measured at three
sampling lines (L1–L3) at each of three vertical sections (Section A, Section B and Section
C), as shown in Figure 4b.

The comparison between simulated and measured temperature profile is shown in Figure 5 for
Section C at sampling lines L1, L2, and L3. The simulated temperature profile well matches
the measured one under Configuration I. Despite the under-prediction of temperature in the
lower part of the room under Configuration II, the simulation shows good overall agreement
with the measurement. The relative error between simulated and measured temperature values
is on average 2.2% and 4.6% for Configuration I and II, respectively.

29
CHAPTER 3. METHODOLOGY

3.5.2 Validation Study II – Underfloor Air Distribution

Zhang and Chen [176] conducted an experimental study on indoor particle dispersion in a
full-scale environmental chamber with an underfloor air distribution system. The geometry of
the chamber is shown in Figure 6. The chamber measured L 4.8 m × W 4.2 m × H 2.4 m. Air
was supplied at a total flow rate of 0.0944 m3/s (i.e. 340 m3/h) from two openings located on
the floor and was exhausted from one outlet on the ceiling. Four heated manikins placed on
the floor and six lamps installed on the ceiling were used to represent heat sources in an
ordinary office room. Monodisperse spherical particles with an average diameter of 0.7 µm
and density of 912 kg/m3 were generated by a condensation aerosol generator from a non-
soluble liquid. Particles were released into the room from a point source located 0.3 m above
the floor (X=1.5 m; Y=2.1 m; Z=0.3 m). As shown in Figure 6, airflow velocity and
temperature were measured at seven poles (from V1 to V7, cyan lines), whereas particle
concentration was monitored at the exhaust outlet and at five height levels of another six
poles (from P1 to P6, green lines). The average surface temperature was also recorded for the
chamber walls. A more detailed description of the experiment and boundary conditions can be
found in Zhang and Chen [176].

Figure 7 compares the simulated and measured airflow velocity and temperature profiles at
V1–V7. The simulated velocity agrees well with the measured one. The simulation also
captures the temperature stratification in the chamber reasonably well, although it under-
predicts the temperature level in the lower part of the chamber. The discrepancy from the
measured temperature was also observed in other numerical studies [185,186], which might
be ascribed to the inadequate information of boundary conditions. In the experiment, the
temperature at the walls might not necessarily remain constant or uniform. As only the
surface averaged temperature was available in the literature, uniform temperature profiles
were used for the wall boundary conditions in the simulation. If well-defined boundary
conditions were provided, the airflow model would better capture the flow characteristics.

Figure 8 compares the simulated and measured particle concentration at P1–P7. The
measurement shows relatively homogeneous particle dispersion, whereas the LPT/DRW
model predicts more scattered concentration. Significant fluctuations can be observed at poles
P1–P2, and particles are highly concentrated at certain locations. Similar discrepancies were
also observed in other simulation-based studies [185,187]. Besides the insufficient
information of boundary conditions, this discrepancy could be caused by the known
deficiency of the standard DRW model, which shows a tendency for sub-micron particles to
concentrate in low-turbulence regions in strongly inhomogeneous diffusion-dominated flows
[188]. This adverse impact can be minimized as the particle size increases. Since the size of
BCPs range mainly 5–20 µm, such a phenomenon can be avoided in the simulations presented
in this work.

30
 3.5. MODEL VALIDATION

Figure 6. Sketch of the chamber configuration and measurement arrangements in Zhang and
Chen [176].

a)

b)

Figure 7. Comparison of simulated and measured velocity and temperature profiles at seven
poles: a) velocity; b) normalized temperature; (rectangular symbols: experimental data; dash
line: simulated data)

31
CHAPTER 3. METHODOLOGY

Z [m]

Z [m]

Z [m]

Z [m]

Z [m]

Z [m]
Figure 8. Comparison of simulated and measured particle concentration profiles at six poles
(rectangular symbols: experimental data; dash line: simulated data).

3.5.3 Validation Study III – Ventilated Two-Zone Chamber

The previous experiment of the underfloor air distribution validates the static airflow and
particle transport model in a non-isothermal indoor environment. The validation study in this
subsection aims to validate the unsteady particle tracking model in dynamic airflow
simulations.

Figure 9. Sketch of the ventilated two-zone chamber in the experiment of Lu at al [189].

32
 3.5. MODEL VALIDATION

a)
Particle concentration [ g/m3 ]

b)
Particle concentration [ g/m3 ]

Figure 10. Comparison of simulated and measured particle mass concentration in a) Zone 1
and b) Zone 2.

33
CHAPTER 3. METHODOLOGY

The particle dispersion experiment in a two-zone chamber conducted by Lu et al [189] is

selected to validate the unsteady particle tracking model. As shown in Figure 9, the two-zone
chamber had an overall dimension of L 5.0 m × W 3.0 m × H 2.4 m, with a thin partition in
the middle of the chamber separating two zones. The partition carried a W 0.70 × H 0.95 m
sliding door in the center line of the chamber, which could be either fully open or fully closed.
There was a high-level supply opening (1.0 m × 0.5 m) on one side of the chamber in Zone 1
and a low-level exhaust opening (1.0 m × 0.5 m) on the other side of the chamber in Zone 2.
The air change rate in the experiment was 9.216 ACH.

Initially, the sliding door was closed and the ventilation was switched off. Oil smoke particles
of size 0.5–5 µm and density 865 kg/m3 were injected into Zone 1 and mixed with the room
air. Once a uniform distribution of the smoke particles was achieved in Zone 1, the sliding
door was opened and the ventilation was switched on. Air entered the chamber through the
supply opening in Zone 1 and was extracted through the exhaust opening in Zone 2. The
particle mass concentration in each zone was measured for a total period of 26 minutes. The
initial particle mass concentration achieved in the experiment was 1×10-5 µg/m3. In the
simulation, it was assumed that the initial concentration was equally allocated to five particle
size groups (1 µm, 2 µm, 3 µm, 4 µm, and 5 µm). The time evolution of the simulated and
measured particle mass concentration in each zone is compared in Figure 10. It can be seen
that the simulated particle mass concentration agrees well with the measured one. The
agreement between simulated and measured particle concentration in both zones is generally
reasonable and acceptable. The simulation tends to slightly over-predict the particle mass
concentration in Zone 1, which was also noted in other simulation studies [103,173]. As
pointed out by Lu et al. [189], this can be explained by two unrealistic assumptions adopted in
the simulation: a) the particle mass is assumed to be equally distributed in each size group; b)
the particle rebounding is not considered in the simulation.

34
 4.1. VENTILATION SCHEMES

Chapter 4

Results and Discussions

This chapter consists of three sections.

 Section 4.1 is based upon Paper II, which investigates and compares the performance of
three distinct ventilation schemes at different airflow rates.
 Section 4.2 is based upon Paper III, which studies the impact of door opening on the
ventilation airflow and risk of infections.
 Section 4.3 is based upon Papers IV–VII, which focuses on the effect of internal
constellation on the disruption of the ventilation airflow.

4.1 Ventilation Schemes

The performance of three ventilation schemes with respect to airborne BCPs contamination
mitigation is evaluated and compared in Paper II. The three ventilation schemes are laminar
airflow (LAF), temperature-controlled airflow (TAF), and mixing airflow ventilation.

The three ORs in Helsingborg General Hospital, as presented in Section 3.1.1 are adopted as
the physical cases. The three ventilation schemes are designed to operate at different airflow
rates. LAF operates at the highest airflow rate (100 ACH), the mixing ventilation at the lowest
airflow rate (26 ACH), and TAF at a moderate airflow rate (46 ACH). Thus, the performance
of the three ventilation schemes is firstly assessed at their respective operating airflow rates.
The three ventilation schemes are further compared at two identical airflow rates.

4.1.1 Operating Airflow Rates

The simulated airflow patterns are exhibited in Figure 11, which plots the velocity vectors at
two vertical planes passing the center of the operating table. The presence of surgical lamps
disrupts the airflow path and creates a stagnant area underneath the lamps. In spite of this, a
nearly uniform airflow distribution dominates the critical zone in LAF ventilation, resulting in
parallel down-flow of clean air with minimum turbulence in the operating zone. Bacteria-

35
CHAPTER 4. RESULTS AND DISCUSSIONS

carrying particles can be directly removed by the airflow and an effective washing effect can
be achieved.

TAF is also capable of creating strong down-flow air streams in the operating zone. The cool
air falls down to the operating table and is gradually accelerated by gravity. The airflow
reaches its maximum speed approximately at the breathing level of the staff. Thus, TAF can
also be expected to effectively sweep away BCPs released from the surgical staff and prevent
them from approaching the patient. Compared with LAF and TAF, the mixing ventilation
showed small variances in the velocity distribution. The velocity vectors look more irregular
and no clear airflow patterns can be identified, indicating a mixed airflow field in the entire
OR.

The simulated BCPs contamination at the operating table and two instrument tables are

Figure 11. Velocity vector plot at the center-planes of the operating table: (a, b) LAF at 100
ACH; (c, d) TAF at 46 ACH; (e, f) Mixing at 26 ACH.

36
 4.1. VENTILATION SCHEMES

presented in Figure 12 and Figure 13. The three ventilation systems results in substantially
different levels of contamination. Both LAF and TAF achieve high air cleanliness at the
operating table, with volume BCPs concentration below 0.1 CFU/m3. The lowest
contamination is found with LAF, which uses the highest airflow rate. TAF operates at half
the airflow rate of LAF, but the resulted BCPs concentration in TAF is only slightly higher
than in LAF. The mixing ventilation uses the least airflow and results in the worst air
cleanliness. The airborne contamination at the operating table in the mixing ventilation is
about 30 CFU/m3 and surface contamination about 650 CFU/m2/h.

As expected, the mixing ventilation shows more uniform BCPs concentration between the
operating table and instrument tables, indicating a better-mixed airflow in the mixing
ventilation than in LAF and TAF. By contrast, both LAF and TAF show much higher BCPs
concentration at the instrument tables than at the operating table. The highest level of
contamination is found at instrument table 2 with LAF. This implies that the space above the
instrument tables is not well ventilated. The instrument tables are positioned behind the
surgical staff downstream the surgical zone, which is not considered good work practice.

Both the surface and volume BCPs concentration give relatively consistent results for
different case studies and different sampling locations. The numerical results are also
generally consistent with on-site measurements by Alsved et al. [24], except that the
simulation over-predicted the contamination level for LAF and TAF at the instrument tables.
As explained above, this could be attributed to the positioning of the instrument tables.

100

10
CFU/m3

0.1

0.01
LAF TAF Mixing
Operating table Instrument table 1 Instrument table 2

Figure 12. Simulated airborne contamination for three ventilation schemes at operating
airflow rates. Horizontal dashed line indicates the recommended limit for an ultraclean
environment.

37
CHAPTER 4. RESULTS AND DISCUSSIONS

1000
CFU/m2/h

100

10
LAF TAF Mixing
Operating table Instrument table 1 Instrument table 2

Figure 13. Simulated surface contamination for three ventilation schemes at operating airflow
rates. Horizontal dashed line indicates the recommended limit for an ultraclean environment.

4.1.2 Identical Airflow Rates

The three ventilation schemes are further evaluated at two identical airflow rates: 4800 m3/h
(40 ACH) and 7200 m3/h (60 ACH). The airflow patterns are presented in Figure 14 (at 40
ACH) and Figure 15 (at 60 ACH). The simulated airborne contamination and surface
contamination are presented in Figure 16 and Figure 17, respectively.

The OR with LAF is normally ventilated at 100 ACH. At reduced airflow rates, the
performance of LAF has drastically worsened. This indicates a change of the unidirectional
airflow pattern and a failure of the washing effect. Compared with Figure 11a–b, Figure 14a–
b and Figure 15a–b clearly show that at a low supply velocity the downward airflow does not
have enough momentum to overcome the upward convective currents. Therefore, local
thermal plumes are developed. The interaction between the downward supply air and upward
buoyant flow forms a strong recirculation area above the operating table. This may entrain a
great amount of BCPs from the surgical staff into the wound site. The lower airflow rate
results in a disordered airflow pattern that partly resembles turbulent mixing airflow. Hence,
with a reduced air supply velocity, LAF is unable to preserve the unidirectional airflow
pattern and provide an adequate washing effect [190]. These observations are in line with the
findings in previous research works [105,190], although the critical airflow rate (or supply air
velocity) that triggers such a change in the airflow pattern is case-dependent.

38
 4.1. VENTILATION SCHEMES

Figure 14. Velocity vector plot at the center-planes of the operating table at 40 ACH: (a, b)
LAF; (c, d) TAF; (e, f) Mixing.

In contrast to the LAF ventilation, the variation in ventilation rates does not have a significant
impact on the performance of the TAF ventilation. The airflow pattern retains most of the
important characteristics and does not show marked differences at alternative airflow rates.
Even at 40 ACH, the down-flow air streams are still sufficient to provide the necessary
washing effect. As a result, TAF preserves an ultraclean environment in the operating zone at
both 40 and 60 ACH. At the instrument tables, the BCPs concentration slightly increases at a
lower airflow rate. These findings confirm that TAF taking advantage of the buoyant effect
utilizes the airflow more efficiently than LAF relying on high momentum.

Increasing the airflow rate of the mixing ventilation from 26 ACH to 40 ACH improves the
air cleanliness in the operating room. When the airflow rate is further increased to 60 ACH, a
remarkable improvement in the cleanliness of air is observed at the operating table and
instrument table 1. This is ascribed to the changes in the airflow patterns. With a high supply

39
CHAPTER 4. RESULTS AND DISCUSSIONS

Figure 15. Velocity vector plot at the center-planes of the operating table at 60 ACH: (a, b)
LAF; (c, d) TAF; (e, f) Mixing.

air velocity, as shown in Figure 15e–f, the air streams supplied from the diffuser impinge
directly on the operating table and sweep off infectious particles from the operating zone. The
remarkable difference in the contamination at the two instrument tables indicates a non-
uniform airflow distribution. This is caused by the asymmetric positioning of the exhaust
openings. Horizontal displacement of airflow takes place from +X to –X direction and
transports infectious particles to the space above instrument table 2. Previous studies have
shown that the arrangement of the exhaust openings has a significant influence on the
particulate contaminant dispersion in cleanrooms [191,192].

40
 4.1. VENTILATION SCHEMES

100 100

10 10
CFU/m3

CFU/m3
1 1

0.1 0.1
LAF TAF Mixing LAF TAF Mixing
Operating table Instrument table 1 Instrument table 2 Operating table Instrument table 1 Instrument table 2

Figure 16. Simulated airborne contamination at 40 ACH (left) and 60 ACH (right). Horizontal
dashed line indicates the recommended limit for an ultraclean environment.

10000 10000

1000 1000
CFU/m2/h

CFU/m2/h

100 100

10 10

1 1
LAF TAF Mixing LAF TAF Mixing
Operating table Instrument table 1 Instrument table 2 Operating table Instrument table 1 Instrument table 2

Figure 17. Simulated surface contamination at 40 ACH (left) and 60 ACH (right). Horizontal
dashed line indicates the recommended limit for an ultraclean environment.

4.1.3 Dilution Model

The dilution model implies that increasing the airflow rates will proportionally decrease the
airborne contamination. However, using Eq. 10 in Section 3.4.4 may lead to substantial over-
or under-prediction of the airborne BCPs concentration at specific locations [154,163,193],
since perfect mixing is never fully achieved under real use conditions.

For comparison purposes, Figure 18 presents the simulated airborne BCPs concentration
together with the estimated values from Eq. 10 on a logarithmic scale. Although higher
airflow rates are generally associated with better air cleanliness, it is clear from Figure 18 that
the airborne BCPs concentration does not increase (decrease) linearly with lower (higher)
airflow rates. With regard to the LAF ventilation, increasing the ventilation rate from 40 ACH
to 60 ACH does not show marked improvement in air cleanliness and even slightly increases
the BCPs concentration at instrument table 1. Similar phenomena were also observed by
Woods et al. [40], in which the particle concentration in the vicinity of the surgical wound
was lower at 12 ACH than at 18.8 ACH when using an overhead ceiling diffuser air

41
CHAPTER 4. RESULTS AND DISCUSSIONS

CFU / m3

CFU / m3

CFU / m3
Figure 18. Comparison of the simulated airborne contamination with the dilution model.
Horizontal dashed line indicates the recommended limit for an ultraclean environment.

distribution system. This is because the reduced down-flow air velocity allowed the upward
convective current developed from the surgical team to lift particles out of the surgical field.

Increasing the ventilation rate from 40 ACH to 60 ACH steadily mitigates the airborne
contamination at the instrument tables in the TAF-ventilated OR. However, the minimum
airborne BCPs concentration at the operating table is achieved at the operating airflow rate,
i.e. 46 ACH. This finding can be ascribed to the fact that a high airflow rate adds turbulence
and provokes the cross-contamination. In addition, a higher airflow rate leads to a lower
ambient temperature and thus a smaller temperature gradient, which reduces the driving
forces acting on the incoming cool air. As a consequence, the washing effect provided by the
downward airflow at 60 ACH is not as strong as the one at 46 ACH, which is confirmed by
the smallest falling speed at 60 ACH above the operating table. Hence, both the airflow rate
and supply temperature need to be carefully optimized and controlled under real use
conditions to maximize the performance of the TAF system.

The results suggest that increasing the airflow rate alone will not always guarantee sufficient
control of particulate concentration at specific locations within an OR, as also confirmed by
experimental studies [194–196]. The airflow distribution plays a critical role in the removal
and dilution of particulate contamination. Hence, specific analyses of the airflow distribution
are needed to determine the likely pathways of BCPs within ORs.

42
 4.2. DOOR OPENINGS

4.2 Door Openings

The impact of door openings on the airborne BCPs contamination in the OR is studied in
Paper III. The OR in the New Karolinska Solna University Hospital is adopted as the physical
case. The OR has mixing ventilation and is separated from an access corridor by a sliding
door. Two thermal conditions are considered in the study: the summer condition and the
winter condition. The average temperature in the OR is about 3 °C lower (higher) than in the
corridor under the summer (winter) condition.

First, the cross-door airflow and BCPs spread caused by the sliding door opening under a
positive pressure of 5 Pa is studied. This is taken as a reference case (Case A) for subsequent
studies. Three different levels of positive pressure are then applied and the effect on the
mitigation of inter-zonal BCPs transfer is evaluated. Finally, we examine whether temporarily
reducing the OR exhaust airflow rate during door operation can improve the isolation
conditions.

4.2.1 Increased Contamination

Figure 19 shows the temporal variation of pressure difference between the OR and the
corridor during the door opening-closing cycle. The pressure difference drops sharply as the
door opens and is even reversed when door is fully open under the summer condition. The
positive pressure is not restored until the door is just about to be fully closed. The slight
difference in the pressure variation between the summer and winter condition can be ascribed
to the thermal pressure due to the distinct thermal conditions. This finding is consistent with
previous studies [164,197].

5
summer
winter
4

0
0 5 10 15 20
Time [s]

Figure 19. Temporal variation of the pressure difference between the OR and corridor.

43
CHAPTER 4. RESULTS AND DISCUSSIONS

Figure 20. Temperature contour plot at the center-plane of the sliding door for Case A: (a) t=0
s; (b) t=5 s; (c) t=15 s; (d) t=20 s; (left: summer condition; right: winter condition).

Figures 20–22 respectively present the temperature field, the velocity field, and the airborne
BCPs concentration at the plane passing the center line of the door at four time instants (t=0,
5, 15, and 20 s). When the door is closed, most of the supply air is directed by the diffusers to
the periphery of the OR at the ceiling level and then falls down to the floor along the
sidewalls. The incoming air further travels to the surgical zone and gets heated by the surgical
personnel and equipment, forming upward convective currents. When the door is open, the
OR air and corridor air encounter each other at the door interface and the interaction between
the warm and cold air causes natural convection due to the temperature difference.
Consequently, the corridor air flows into the OR through the upper part of the doorway under
the summer condition, whereas the corridor air penetrates into the OR at the floor level under
the winter condition. The upward warm air from the corridor interacts with the downward
supply cold air in the OR, forming a vortex near the doorway under the summer condition
(see Figure 21c left). Under the winter condition, by contrast, a neutral point can be roughly

44
 4.2. DOOR OPENINGS

Figure 21. Velocity vector plot at the center-plane of the sliding door for Case A-5: (a) t=0 s;
(b) t=5 s; (c) t=15 s; (d) t=20 s; (left: summer condition; right: winter condition).

identified slightly above the center of the door (see Figure 21c right). As seen from Figure 22,
BCPs are accumulated mainly in the middle and upper part of the OR under the summer
condition, whereas under the winter condition BCPs are confined at the floor level. It is
noteworthy that, under the winter condition, the BCPs travel to the surgical zone at the foot
level of the surgical staff and then get lifted to the operating table by human convective
boundary layer flow, which constitutes a potential risk to patient safety (see Figure 21d right).

Figure 23 shows the overall contamination in the OR within a five-minute period from the
onset of door operation. The single door cycle increases the overall airborne BCPs
contamination in the OR by approximately 2.1 CFU/m3. The airborne contamination starts to
decay when the door is closed. It takes nearly five minutes to reduce the overall OR
contamination to 5% of its initial level. The decay curves of BCPs concentration do not
exactly follow an exponential course as they are not completely mixed with the room air.

45
CHAPTER 4. RESULTS AND DISCUSSIONS

Figure 22. Airborne BCPs concentration at the center-plane of the sliding door for Case A-5:
(a) t=0 s; (b) t=5 s; (c) t=15 s; (d) t=20 s; (left: summer condition; right: winter condition).

Despite the similar initial contamination level, the two decay curves differ from each other.
This can be ascribed to the different airflow patterns induced near the doorway between the
summer and winter conditions. Under the summer condition, the contaminants accumulate in
the middle and upper space of the OR, whereas the contaminants are confined in the lower
space of the OR under the winter condition. Therefore, the airborne contaminants are
evacuated more quickly under the winter condition than under the summer condition, as the
penetrated BCPs are closer to the low-level exhaust vents under the winter condition.

The penetrated BCPs continue to spread in the OR after the door is fully closed. Certain BCPs
might approach the surgical site and cause infections. Therefore, it is necessary to evaluate the
local contamination of surgical site, which is closely linked to the rate of infections. Figure 24
presents the development of the airborne BCPs concentration at the surgical site over the five-
minute period. Although the overall OR contamination is similar under the two thermal

46
 4.2. DOOR OPENINGS

conditions, the resulting local contamination is drastically different. As pointed out above, the
BCPs travel to the surgical critical zone at the floor level under the winter condition and are
then raised to the level of surgical wound by the human thermal plumes. Correspondingly, the
airborne BCPs concentration reaches the highest of 12.5 CFU/m3 approximately 20 seconds
CFU/m3

Figure 23. Temporal development of the overall OR contamination over a 5-minute period
from the onset of door operation in Case A-5.
CFU/m3

Figure 24. Temporal development of the local contamination at the surgical site over a 5-
minute period from the onset of door operation in Case A-5.

47
CHAPTER 4. RESULTS AND DISCUSSIONS

a)

2.5
A-5 (summer)
A-10 (summer)
2.0 A-15 (summer)
A-5 (winter)
A-10 (winter)
A-15 (winter)
1.5

1.0

0.5

0
0 5 10 15 20
Time [s]

b)

450 summer winter

Number of CFU

400

350

300
5 Pa 10 Pa 15 Pa

Figure 25. OR contamination caused by door opening in Case A-5 (5 Pa), Case A-10 (10 Pa)
and Case A-15 (15 Pa): (a) temporal development of overall OR contamination; (b)
maximum number of BCPs in the OR.

after the door is fully closed. Such a contamination level poses a threat to patient safety in
infection-prone surgery. Under the summer condition, by contrast, the BCPs can barely
spread into the surgical critical zone and the consequent contamination at the surgical site

48
 4.2. DOOR OPENINGS

stays below 1 CFU/m3 over the entire period, which does not constitute a significant risk
factor for SSIs occurrence. As the overall level of OR contamination is directly associated
with the volume of air exchange across the doorway, the simulation results suggest that
similar air exchange volumes can lead to completely different levels of local contamination at
the surgical site. This finding implies that it is not sufficient to focus only on the air volume
migration for a valid and reliable assessment of the infection risks caused by the door
opening.

4.2.2 Positive Pressure

The positive pressure in the OR is utilized to prevent the entry of contaminants through the
gaps between the OR and adjacent spaces. The positive pressure proves to be useful when all
openings are closed, as the contamination is transferred in the direction of decreasing static
pressure. However, the pressure drop effectively becomes zero when the door is opened, as
indicated in the preceding subsection. The effect of increasing positive pressure in the OR is
investigated in this subsection. The positive pressure in the OR relative to the adjacent
corridor is increased from 5 Pa (Case A-5) to 10 Pa (Case A-10) and eventually to 15 Pa
(Case A-15). The incurred airborne BCPs contamination in the OR is evaluated and
compared.
OR exhaust flow [m3 /s]

Figure 26. Temporal development of the OR exhaust flow rate in Case B (10% reduction),
Case C (20% reduction), and Case D (30% reduction).

49
CHAPTER 4. RESULTS AND DISCUSSIONS

Figure 27. Temperature contour plot at the center-plane of the sliding door at t=15 s for a)
Case A; b) Case B; c) Case C; d) Case D; (left: summer condition; right: winter condition).

Figure 25a shows the temporal development of the overall contamination during door
operation and Figure 25b compares the maximum number of CFU in the OR. Increasing the
positive pressure slightly reduces the intrusion of airborne contaminants under both summer
and winter conditions. However, the effect of pressure difference is marginal, diminishing as
the positive pressure increases. Under the summer condition, for instance, the maximum
number of CFU in the OR is reduced by only 10% when the positive pressure is doubled and
by 18% when the positive pressure is further increased to 15 Pa. It is obvious that the positive
pressure levels necessary to minimize or eliminate the contaminant migration would be
considerably high. Constantly maintaining such a high positive pressure will result in elevated
noise and unnecessary operating cost. Thus, increasing the pressure difference between the
OR and the adjacent space is not an efficient way to prevent contaminant intrusion caused by
door opening. Despite the important role in static state, the pressure difference must be
combined with other measures to enhance the dynamic isolation.

50
 4.2. DOOR OPENINGS

Figure 28. Velocity vector plot at the center-plane of the sliding door at t=15 s for a) Case A;
b) Case B; c) Case C; d) Case D; (left: summer condition; right: winter condition).

4.2.3 Directional Airflow

In order to prevent the infectious air from entering the OR, the airflow extracted from the OR
is reduced while the door is opening. During the opening phase of the door, the loss of
pressure in the OR is monitored and the OR exhaust flow rate is decreased accordingly. After
the outgoing airflow becomes stabilized, the resulting exhaust flow rate is reduced by 10%,
20% and 30% in Cases B, C, and D, respectively. During the closing process of the door, as
the pressure is gradually restored, air extraction is increased to balance the mass flow and
avoid pressure overshoot in the OR. Figure 26 presents the temporal development of the OR
exhaust flow in Cases B, C, and D.

By reducing the exhaust flow rate, a certain amount of air is forced to flow out of the OR
through the opening area. Figures 27−29 compare the temperature, velocity, and airborne

51
CHAPTER 4. RESULTS AND DISCUSSIONS

Figure 29. Airborne BCPs concentration at the center-plane of the sliding door at t=15 s for a)
Case A; b) Case B; c) Case C; d) Case D; (left: summer condition; right: winter condition).

BCPs concentration at the center-plane of the door at t=15 s between Cases A, B, C, and D.
As the net outflow through the door opening increases, the pattern of cross-doorway airflow is
changed towards forming a one-way directional airflow (see Figure 28). Correspondingly, as
seen from Figures 27−29, the air exchange and contaminant intrusion are constantly
suppressed from Case A to Case D. These figures suggest that the outflow of air can
overcome the buoyant flow and mitigate the transfer of BCPs. Thus, a barrier effect is
expected to be established as sufficient air flows out of the OR through the door opening.

The airborne contamination caused by the door opening evolves over time. In order to
compare the barrier effect in different cases, the mean and maximum contamination levels
over the five-minute period in the four cases are plotted against the reduction rates of the OR
exhaust flow in Figure 30 (overall OR contamination) and Figure 31 (local contamination at
the surgical site). As the exhaust flow rate is continually decreased from Case A to Case D,

52
 4.2. DOOR OPENINGS

a) b)

1 2.5
summer summer
winter winter
0.8 2

0.6 1.5

0.4 1

0.2 0.5

0 0
0% 10% 20% 30% 0% 10% 20% 30%
Flow Rate Reduction Flow Rate Reduction

Figure 30. Comparison of the OR overall contamination at different reduction rates of OR

exhaust flow: a) mean contamination over a 5-minute period from the onset of door operation;
b) maximum contamination over a 5-minute period from the onset of door operation.

a) b)

1.2
summer
winter

0.8
CFU/m 3

0.4

0
0% 10% 20% 30%
Flow Rate Reduction

Figure 31. Comparison of the local contamination at the surgical site at different reduction
rates of OR exhaust flow: a) mean contamination over a 5-minute period from the onset of
door operation; b) maximum contamination over a 5-minute period from the onset of door
operation.

a declining trend in both overall and local contamination can be observed. In Case B, for
instance, a 10% decrease in the exhaust flow rate leads to a more than 40% reduction in
overall OR contamination. The maximum contamination at the surgical site is reduced from
12.5 CFU/m3 to 2.5 CFU/m3 from Case A to Case C under the winter condition, which
significantly lowers the risk of infections. In Case D with a 30% decrease in the exhaust rate,
the BCPs migration is considerably suppressed, resulting in nearly zero overall and local
contamination. The simulation results show that decreasing the airflow extraction during door
operation can be an effective solution to ensure isolation and thus minimize the risk of SSIs.

53
CHAPTER 4. RESULTS AND DISCUSSIONS

4.3 Internal Disruptions

OR ventilation aims to establish a certain airflow pattern inside the room. The door opening
can disrupt the airflow inside the OR, as discussed in Section 4.2. The ventilation airflow can
also be affected by internal disruptions caused by surgical lamps, surgical equipment and
surgical personnel. This section analyzes and discusses the impact of internal disruptions on
the ventilation airflow and BCPs dispersion.

4.3.1 Surgical Lamps

Surgical lamps are considered as one of the major disturbances of airflow in an OR that
impairs the proper functioning of ventilation, especially in the LAF ventilation. Chow et al.
[96] indicated that the medical lamp positioning in a LAF system could have a serious impact
on the airborne dispersion of BCPs and thus increased the risk of infections. Zoon et al. [108]
conducted experimental and simulation studies and found that different lamp shapes resulted
in considerably different airflow pattern in the OR. The impact of the shape and design of
surgical lamps on the ventilation airflow and BCPs concentration is investigated in Paper IV
and Paper V respectively.

Shape of lamps

Two distinct lamp shapes are investigated in Paper IV: a cylinder-like closed-shape lamp and
an open-shape lamp. The open-shape lamp is composed of five individual spots connected to
a central hub. The two types of lamps have the same outer diameter of 0.6 m, corresponding
to a projected area of 0.28 m2 for the closed-shape lamp and 0.08 m2 for the open-shape lamp.
The influence of the lamps on the airflow is evaluated under two ventilation schemes: LAF
and TAF. The ORs in Helsingborg General Hospital are adopted as the physical cases.

The combination of two shapes of lamps and two ventilation schemes results in four
simulation cases. The surface BCPs contamination is presented in Table 1. Two ventilation
schemes in combination with two types of surgical lamps result in dramatically different
contamination levels. The closed-shape lamp in LAF results in more than 100 CFU/m2/h
surface contamination. When using an open-shape lamp in LAF, there are nearly zero BCPs
depositing onto the operating table. Replacing the closed-shape lamp by the open-shape lamp
in TAF reduces the surface contamination from about 10 CFU/m2/h to 3 CFU/m2/h.

Table 1. Simulated surface BCPs contamination (CFU/m2/h) at the operating table.

Closed-Shape Lamp Open-Shape Lamp

LAF 104 < 0.1

TAF 9.6 3.1

54
 4.3. INTERNAL DISRUPTIONS

Figure 32. Velocity vector plot under surgical lamps for a) LAF with closed-shape lamp; b)
LAF with open-shape lamp; c) TAF with closed-shape lamp; d) TAF with open-shape lamp.

The significant differences between cases can be explained by the airflow patterns. As shown
in Figure 32, the closed-shape lamp positioned below the supply diffuser obstructs the flow
path and hinders the formation of a unidirectional airflow. Instead, a stagnant zone is formed
under the lamp with remarkably reduced air speed. As seen in Figure 32a, the unidirectional
airflow is immediately separated at the upper edge of the lamp and two vortices are
developed, which creates a recirculation region under the lamp. The washing effect is thus
severely diminished and the recirculating air in this stagnant zone may entrain a significant
amount of BCPs from the source of contaminants, i.e. the surgical staff. When using the open-
shape lamp, the obstruction to the airflow is rather limited and LAF can maintain a relatively
unidirectional flow pattern. Therefore, the microbiological cleanliness of air at the surgical
site is remarkably improved by using the open-shape lamp. The findings that surgical lamps
obstruct the airflow and negatively influence airborne contamination in the LAF ventilation
are in line with previous studies [107,108].

TAF shows good performance with both closed-shape and open-shape lamps. The presented
airflow patterns and surface BCPs contamination indicate that TAF is less sensitive to the
shape of objects in the airflow path. Regardless of the lamps used, TAF is capable of
providing a clean zone over the entire surgical site. The airflow passes around the surgical
lamp, which implies that the buoyancy-driven flow utilized by TAF might be more effective
in circumventing obstacles than the inertia-driven flow used by LAF. Although a small
stagnant zone is also observed under the closed-shape lamp in TAF, it is confined and
surrounded by the clean supply air, which precludes the entrainment of particles from the
surgical staff. It is worth noting that TAF operates at half the flow rate of LAF, but achieves
comparable level of CFU concentration when using the open-shape lamp and TAF gives even
better performance than LAF when using the closed-shape lamp.

The drastically different airflow pattern under the lamp between LAF and TAF can be
explained by the effect of the so-called ‘aiding thermal buoyancy’ on the downward flow past

55
CHAPTER 4. RESULTS AND DISCUSSIONS

a bluff body [198]. Sharma et al. [199] showed that the flow separation could start at very low
Reynolds numbers for isothermal flow around a square cylinder. The Reynolds number of the
flow past the closed-shape lamp in LAF is around 15,000 and therefore the wake region is
dominated by the vortices caused by the flow separation. When the buoyancy is added, as is
the case in TAF, the presence of temperature difference between the supply air and ambient
room air induces a downward force acting on the supply air. Such a force is superimposed to
the airflow momentum and aids the downward movement of the supply air. Some
fundamental research works in fluid mechanics and heat transfer have studied the
characteristics of flow around a heated/cooled bluff obstacle [200–204]. Experiments and
numerical simulations have been conducted to investigate the influence of aiding thermal
buoyancy on the recirculation region behind the obstacle. It has been shown that increasing
the strength of aiding thermal buoyancy gradually diminishes the flow separation [199,204].
The flow separation can be completely suppressed at some critical value of the buoyancy
parameter [202,204]. In those studies, the Richardson number is used to quantify the
buoyancy effect at a given Reynolds number. However, in the context of TAF in ORs with
surgical equipment and personnel, the critical value of the buoyancy parameter for which the
flow separation around the surgical lamps completely vanishes is case-dependent and the
exact quantification is difficult. In spite of this, the buoyancy-induced airflow utilized by TAF
has shown its superiority to momentum-induced airflow used by LAF in circumventing
physical obstruction in the airflow paths.

Design of lamps

An innovative design of the surgical lamp – fan-mounted lamp, is assessed and compared
with a conventional closed-shape lamp in Paper V. The performance of the two designs of
lamps is evaluated under two ventilation schemes: LAF and mixing ventilation. The OR in the
New Karolinska Solna University Hospital is adopted as the physical case.

The two lamps have an identical overall dimension. Both have a hexagonal cross-section with
an outer diameter of 0.7 m and thickness of 0.15 m. A 0.2 m diameter fan is installed inside
the fan-mounted lamp, which creates a constant pressure jump of 0.6 Pa.

The simulated BCPs contamination is reported in Table 2. The fan-mounted lamp resolves the
problem that the conventional lamp has caused in the LAF ventilation. When using the
closed-shape lamp, LAF performs even worse than the mixing ventilation, resulting in
airborne BCPs concentration 63 CFU/m3. Such a high level of BCPs concentration poses a
serious threat to patient safety in infection-prone surgery. By contrast, zero contamination is
achieved in LAF ventilation with the fan-mounted lamp. It is noteworthy that using the fan-
mounted lamp also reduces the contamination level in the mixing ventilation.

The effect of the lamp design can be further illustrated by Figure 33, which presents the
velocity contours at two vertical planes passing the center of the operating table. The closed-
shape lamp blocked the unidirectional airflow and results in a stagnant area under the lamp

56
 4.3. INTERNAL DISRUPTIONS

Figure 33. Velocity contour plot at the center-planes of the operating table.

(see Figure 33a). Thus, the washing effect expected in LAF ventilation is severely weakened.
The stagnant area is eliminated by the fan-mounted lamp that blows air to the operating table
(see Figure 33b). In the mixing ventilation, the air blown from the fan-mounted lamp can also
help wash the BCPs away from the surgical site. Therefore, the fan-mounted lamp improves
the performance of both LAF and mixing ventilation. Quantitatively, applying the fan-
mounted lamp decreases the contamination level by nearly 100% in the LAF ventilation for
both surface and airborne contamination. In the mixing ventilation, the surface contamination
decreases by 26% and the airborne contamination by 41% when using the fan-mounted lamp.
Hence, the fan-mounted lamp proves to be more effective in the LAF ventilation than in the
mixing ventilation.

Table 2. Simulated airborne and surface BCPs contamination for different lamp designs.

Surface Contamination Airborne Contamination

(CFU/m2/h) (CFU/m3)

LAF Mixing LAF Mixing

Closed-shape lamp 338 249 63 44

Fan-mounted lamp <1 183 <1 26

57
CHAPTER 4. RESULTS AND DISCUSSIONS

4.3.2 Surgical Staff

The surgical staff has a two-fold effect on the contaminant distribution in an OR. First, the
human bodies in the OR serve as an obstruction and heat source that alter the airflow. Second,
the surgical staff contributes to the source of BCPs. Paper VI investigates the effect of
surgical staff number on the BCPs contamination in two ORs respectively ventilated by
turbulent mixing airflow and TAF. Four cases with 3, 5, 7, and 9 surgical staff members
surrounding the operating table are analyzed. The ORs in Helsingborg General Hospital are
adopted as the physical cases.

The simulated airborne and surface contamination at the operating table is plotted against
staff number in Figure 34. The simulation results are also compared with the analytical
estimation by the dilution model (Eq.10 in Section 3.4.4).

The contamination level rises steadily as more surgical staff members are present in the OR
with mixing ventilation. The surface contamination increases linearly, whereas the growth
rate of the airborne contamination shows slight fluctuations between cases. This can be
ascribed to the fact that the presence of persons alters the airflow and affects local airborne
BCPs concentration.

The TAF ventilation maintains a sufficiently low level of contamination regardless of the
number of individuals in the OR. The airborne contamination fluctuates slightly between 0.3
and 1.0 CFU/m3 at the operating table and surface contamination grows very slowly as the
staff number increases. The highest level of surface contamination is found to be 60
CFU/m2/h in the case of 9 persons, whereas the maximum airborne contamination is achieved
in the case of 7 persons. This inconsistency between airborne and surface contamination is

a) b)
CFU / m3

Figure 34. Simulated BCPs contamination at the operating table under the conditions of
different number of surgical staff: a) airborne contamination; b) surface contamination;
(horizontal dash line indicates the recommended limit for ultraclean environment).

58
 4.3. INTERNAL DISRUPTIONS

also attributable to the local airflow pattern and the positioning of the persons.

The dilution model over-predicts the airborne contamination in all cases for both mixing and
TAF ventilation, which reflects that the airflow is not fully mixed even in the mixing
ventilation. For the mixing ventilation, the simulated airborne BCPs concentration tends to
converge to the analytical estimation as more persons are present in the OR. For the TAF
ventilation, however, the airborne contamination remains very low and the departure from the
dilution model is expanded as the staff number increases. This implies that TAF uses airflow
in a more efficient way than the mixing ventilation based on the dilution principle.

4.3.3 Surgical Equipment

Previous results have shown that TAF has advantages in overcoming physical obstructions in
the airflow path, such as the surgical lamps. These findings suggest that TAF utilizing the
buoyancy effect can serve as a promising alternative to LAF that relies on the inertia force.
However, the upward convective airflow developed from the heat sources (e.g., the surgical
equipment) interacts with the downward supply airflow and may deteriorate the performance
of TAF. Surgical equipment develops strong thermal plumes, which may prevent the clean air
from being delivered to the surgical site. The influence of the heat released from surgical
equipment on the performance of TAF is thus investigated in Paper VII.

Twelve pieces of equipment are considered, representing the most commonly used medical
equipment during operations. Four simulation cases are performed with heat loads gradually
increased by 25% of the rated power of the equipment. The heat load starts from 4.4 kW in
Case 1 and increase to 7.2 kW in Case 4. Details of the power of the equipment can be found
in Table 2 in Paper VII.

25

20

15
CFU/m3

10

0
Case 1 Case 2 Case 3 Case 4

Operating table Instrument table 1 Instrument table 2

Figure 35. Simulated airborne BCPs contamination at the operating table and two instrument
tables in Case 1, Case 2, Case 3, and Case 4.

59
CHAPTER 4. RESULTS AND DISCUSSIONS

Figure 36. Temperature contour plot at the center-plane of the operating table.

60
 4.3. INTERNAL DISRUPTIONS

Figure 35 presents the simulated airborne contamination at the operating table and two
instrument tables. At the operating table, Cases 3–4 result in significantly higher BCPs
concentration than Case 1–2. Specifically, an abrupt increase in airborne BCPs concentration
is observed when the heat load is increased from Case 2 to Case 3. This indicates fundamental
changes in the airflow pattern, as confirmed by the temperature contour plot in Figure 36. The
washing effect is severely weakened and the dilution effect becomes predominated in Cases
3–4. Consequently, furthering increasing the heat load from Case 3 to Case 4 only slightly
increases the airborne BCPs concentration at the operating table. From Case 1 to Case 2, the
airborne BCPs concentration slightly decreases as the heat load increases. This can be
ascribed to the fact that a heavier heat load results in a higher ambient temperature and thus a
higher temperature gradient, which strengthens the driving force acting on the supply air.
Thus, the downward airflow in Case 2 provides a stronger washing effect and thus better air
cleanliness than in Case 1.

The two instrument tables are placed outside the protection zone. The dilution principle
replaces the washing effect, that is, the airflow dilutes the airborne contamination rather than
directly washes away infectious particles. The BCPs concentration at the instrument tables
shows a slight declining trend as the heat loads increases. This is due to the fact that higher
heat dissipation leads to stronger upward thermal plumes, which have more chance to lift
particles out of the range of instrument tables.

The findings suggest that TAF can function properly under moderately high heat loads and
provide good protection for patients. When the heat load becomes extremely heavy, it is
difficult for TAF to establish the desired airflow pattern and TAF fails to provide an efficient
washing effect. Despite the distorted airflow pattern, however, TAF is still able to work in
such a way that it resembles the mixing ventilation and dilutes the airborne contamination.

61
 5.1. CONCLUSIONS

Chapter 5

Conclusions and Future Work

5.1 Conclusions
Operating room ventilation is crucial for mitigating the dispersion of airborne bacterial
contaminants and controlling the risk of SSIs. Various ventilation schemes have been
developed for ORs. Each has pros and cons and may be better suited than another for surgical
procedures under certain conditions. Furthermore, the proper functioning of OR ventilation is
also affected by external and internal disruptions. By applying CFD simulations, the present
study investigates the airflow distribution and ventilation effectiveness with respect to
reducing the BCPs contamination in ORs under different working conditions. This thesis aims
to enhance the understanding of the strengths and limitations of various ventilation schemes
and increase the knowledge of the design and usage of OR ventilation. Important findings
from this research work are summarized below.

A sufficient airflow rate is needed for LAF to function properly and achieve desire
performance. At reduced airflow rates, the unidirectional airflow is difficult to be established
and the performance of the LAF ventilation can be drastically degraded. The conventional
mixing ventilation is not able to reliably create an ultraclean environment. The contamination
level steadily rises in the mixing ventilation as the number of staff increases. The usage of
mixing ventilation in infection-prone surgery should be avoided. The TAF ventilation uses the
airflow in a more efficient way than the mixing ventilation and this advantage becomes
increasingly important as additional surgical staff is added in the OR. The TAF ventilation
maintains a sufficiently low level of contamination regardless of the number of persons in the
OR. TAF also represents an effective ventilation scheme that can serve as an energy-efficient
alternative to LAF. In the examined case, the TAF ventilation operates at half the airflow rate
of LAF but achieves comparable performance.

Increasing the airflow rate does not always result in better control of contamination in ORs.
Clean air is not only a result of high airflow rates. The airflow distribution is of critical
importance in removing or diluting airborne contaminants. The dilution model fails to take
into account the effect of airflow pattern on the ventilation effectiveness. Therefore, caution is

63
CHAPTER 5. CONCLUSIONS AND FUTURE WORK

needed when using the dilution model to estimate the airflow rate or contamination level in
the OR. A specific analysis of the airflow is highly recommended for each design of OR
ventilation.

Door openings have a detrimental impact on the microbiological cleanliness of the OR. The
temperature in the OR and adjacent space should be well controlled in order to minimize the
inter-zonal contaminant transfer, as the temperature difference is one of the major driving
forces that determine the two-way airflow across the doorway. The conventional approach
that quantifies only the air volume migration is not sufficient for a reliable assessment of the
impact of door openings on the risk of SSIs. Depending on the airflow distribution in the OR,
the resulting local contamination at the surgical wound can be dramatically different even if
the air exchange volumes are similar. Therefore, the dispersion of penetrated contaminants
within the OR must be considered when evaluating the impact of door openings. Temporarily
reducing the OR exhaust flow during door operation forms a directional airflow towards the
adjacent space, which is found to be an effective solution to ensure isolation and minimize the
contaminant transfer. In the examined case, a 20–30% reduction in the OR exhaust flow
significantly suppresses the air volume migration and decreases the airborne contamination to
a sufficiently low level.

Surgical lamps serve as physical obstructions in the airflow path and constitute a major factor
that deteriorates the performance of LAF ventilation. The lamps distort the unidirectional
airflow and create stagnant regions above the operating table. It is highly recommended to
improve the shape and design of the lamps in the LAF ventilation, as the optimized shape and
design can minimize the negative impact of the lamps on the ventilation airflow. TAF is
found to be less sensitive to the presence of surgical lamps in the airflow path. The buoyancy-
driven airflow used by TAF is more capable of circumventing obstacles than the inertia-
driven flow used by LAF. Therefore, the shape and design of the lamps play a less important
role in the TAF ventilation than in the LAF ventilation. Thermal plumes developed from the
surgical equipment in the OR have the potential to distort the buoyancy-driven airflow in
TAF. Simulation results show that TAF is able to tolerate moderately heavy heat loads and
provide good protection for the patient.

5.2 Future Work

While the present study has presented a detailed assessment of ventilation performance in
ORs, several aspects can be further investigated to get a complete overview of OR ventilation.
It is necessary to refine the understanding of the airflow and contaminant dispersion in ORs,
which contributes to the more effective design and more appropriate usage of OR ventilation.

Although dynamic simulations have been performed to investigate the impact of door
openings in this study, the effect of human activities has been neglected. The surgical staff
has been assumed to be stationary and motionless, which is not the case during real operations

64
 5.2. FUTURE WORK

that are generally completed within a few hours. Apart from the continuous small-scale
movements taking place around the patient related to the operation itself, the surgeons
perform numerous complex manipulative movements such as bending the body and reaching
out arms as well as other more vigorous activities. The scrub nurse may walk from the OR
periphery to the operating table and the circulating nurse walks around the perimeter of the
OR. The impact of such human activities on the airflow and contamination distribution should
be studied in future work.

The assessment of OR ventilation in the present study has been patient-oriented, which
focuses on the protection of the patient against infections caused by bacteria shed from the
surgical staff. However, the OR ventilation should also be able to provide a comfortable
working environment for the surgical staff and protect them from being exposed to waste
anesthetic gases and infectious agents released from the patient wound. Thermal comfort
directly affects the work performance of the surgical staff and is essential in realizing a less
error-prone situation. Therefore, the thermal comfort and protection of the surgical team will
be evaluated in future studies.

From the methodological point of view, both the RANS simulation and the LPT/DRW model
adopted in this study are subject to known deficiencies. Future research works should develop
and apply advanced numerical models to the investigation of airflow and contaminant
dispersion in ORs. As one of the major limitations of the present study is the lack of direct
field measurements, future studies will focus on performing measurements in real ORs during
activity if time and resources permit.

65
 BIBLIOGRAPHY

Bibliography

[1] S.I. Berríos-Torres, C.A. Umscheid, D.W. Bratzler, B. Leas, E.C. Stone, R.R. Kelz,
C.E. Reinke, S. Morgan, J.S. Solomkin, J.E. Mazuski, E.P. Dellinger, K.M.F. Itani,
E.F. Berbari, J. Segreti, J. Parvizi, J. Blanchard, G. Allen, J.A.J.W. Kluytmans, R.
Donlan, W.P. Schecter, Centers for Disease Control and Prevention Guideline for the
Prevention of Surgical Site Infection, 2017, JAMA Surgery. 152 (2017) 784–791.
[2] A.J. Mangram, T.C. Horan, M.L. Pearson, L.C. Silver, W.R. Jarvis, Guideline for
Prevention of Surgical Site Infection, 1999, American Journal of Infection Control. 27
(1999) 97–134.
[3] R.A. Weinstein, M.J.M. Bonten, Laminar airflow and surgical site infections: the
evidence is blowing in the wind, The Lancet Infectious Diseases. 17 (2017) 472–473.
[4] D.J. Anderson, K. Podgorny, S.I. Berríos-Torres, D.W. Bratzler, E.P. Dellinger, L.
Greene, A.-C. Nyquist, L. Saiman, D.S. Yokoe, L.L. Maragakis, K.S. Kaye, Strategies
to Prevent Surgical Site Infections in Acute Care Hospitals: 2014 Update, Infection
Control & Hospital Epidemiology. 35 (2014) 605–627.
[5] J.M. Badia, A.L. Casey, N. Petrosillo, P.M. Hudson, S.A. Mitchell, C. Crosby, Impact
of surgical site infection on healthcare costs and patient outcomes: a systematic review
in six European countries, Journal of Hospital Infection. 96 (2017) 1–15.
[6] S.S. Magill, J.R. Edwards, W. Bamberg, Z.G. Beldavs, G. Dumyati, M.A. Kainer, R.
Lynfield, M. Maloney, L. McAllister-Hollod, J. Nadle, S.M. Ray, D.L. Thompson, L.E.
Wilson, S.K. Fridkin, Multistate Point-Prevalence Survey of Health Care–Associated
Infections, The New England Journal of Medicine. 370 (2014) 1198–1208.
[7] Sveriges Kommuner och Landsting, Vårdrelaterade infektioner - KUNSKAP,
KONSEKVENSER, KOSTNADER, 2017.
[8] Y. Mu, J.R. Edwards, T.C. Horan, S.I. Berrios-Torres, S.K. Fridkin, Improving Risk-
Adjusted Measures of Surgical Site Infection for the National Healthcare Safely
Network, Infection Control & Hospital Epidemiology. 32 (2011) 970–986.
[9] L. Mckibben, T. Horan, J. Tokars, G. Fowler, D. Cardo, M. Pearson, P. Brennan,
Guidance on Public Reporting of Healthcare-Associated Infections: Recommendations
of the Healthcare Infection Control Practices Advisory Committee, American Journal
of Infection Control. 33 (2005) 217–226.
[10] S.S. Awad, Adherence to Surgical Care Improvement Project Measures and Post-
Operative Surgical Site Infections, Surgical Infections. 13 (2012) 234–237.
[11] R. Coello, A. Charlett, J. Wilson, V. Ward, A. Pearson, P. Borriello, Adverse impact of
surgical site infections in English hospitals, Journal of Hospital Infection. 60 (2005)
93–103.
[12] K.B. Kirkland, J.P. Briggs, S.L. Trivette, W.E. Wilkinson, D.J. Sexton, The Impact of
Surgical-Site Infections in the 1990s: Attributable Mortality, Excess Length of
Hospitalization, And Extra Costs, Infection Control & Hospital Epidemiology. 20
(1999) 725–730.

67
BIBLIOGRAPHY

[13] G. de Lissovoy, K. Fraeman, V. Hutchins, D. Murphy, D. Song, B.B. Vaughn, Surgical

site infection: Incidence and impact on hospital utilization and treatment costs,
American Journal of Infection Control. 37 (2009) 387–397.
[14] S.S. Magill, W. Hellinger, J. Cohen, R. Kay, C. Bailey, B. Boland, D. Carey, J. de
Guzman, K. Dominguez, J. Edwards, L. Goraczewski, T. Horan, M. Miller, M. Phelps,
R. Saltford, J. Seibert, B. Smith, P. Starling, B. Viergutz, K. Walsh, M. Rathore, N.
Guzman, S. Fridkin, Prevalence of Healthcare-Associated Infections in Acute Care
Hospitals in Jacksonville, Florida, Infection Control & Hospital Epidemiology. 33
(2012) 283–291.
[15] D.J. Anderson, K.S. Kaye, D. Classen, K.M. Arias, K. Podgorny, H. Burstin, D.P.
Calfee, S.E. Coffin, E.R. Dubberke, V. Fraser, D.N. Gerding, F.A. Griffin, P. Gross, M.
Klompas, E. Lo, J. Marschall, L.A. Mermel, L. Nicolle, D.A. Pegues, T.M. Perl, S.
Saint, C.D. Salgado, R.A. Weinstein, R. Wise, D.S. Yokoe, Strategies to Prevent
Surgical Site Infections in Acute Care Hospitals, Infection Control & Hospital
Epidemiology. 29 (2008) S51–S61.
[16] E.C.J. Broex, A.D.I. van Asselt, C.A. Bruggeman, F.H. van Tiel, Surgical site
infections: how high are the costs?, Journal of Hospital Infection. 72 (2009) 193–201.
[17] J. Shepard, W. Ward, A. Milstone, T. Carlson, J. Frederick, E. Hadhazy, T. Perl,
Financial Impact of Surgical Site Infections on Hospitals: The Hospital Management
Perspective, JAMA Surgery. 148 (2013) 907–914.
[18] J.A. Urban, Cost Analysis of Surgical Site Infections, Surgical Infections. 7 (2006)
s19–s22.
[19] R. Bisht, A. Katiyar, R. Singh, P. Mittal, ANTIBIOTIC RESISTANCE –A GLOBAL
ISSUE OF CONCERN, Asian Journal of Pharmaceutical and Clinical Research. 2
(2009) 6.
[20] B.R. Lyont, R. Skurray, Antimicrobial Resistance of Staphylococcus aureus: Genetic
Basis, MICROBIOL. REV. 51 (1987) 47.
[21] S. Sadrizadeh, A. Afshari, T. Karimipanah, U. Håkansson, P.V. Nielsen, Numerical
simulation of the impact of surgeon posture on airborne particle distribution in a
turbulent mixing operating theatre, Building and Environment. 110 (2016) 140–147.
[22] O.M. Lidwell, E.J.L. Lowbury, W. Whyte, R. Blowers, S.J. Stanley, D. Lowe, Airborne
contamination of wounds in joint replacement operations: the relationship to sepsis
rates, Journal of Hospital Infection. 4 (1983) 111–131.
[23] O.M. Lidwell, E.J. Lowbury, W. Whyte, R. Blowers, S.J. Stanley, D. Lowe, Infection
and sepsis after operations for total hip or knee-joint replacement: influence of
ultraclean air, prophylactic antibiotics and other factors., The Journal of Hygiene. 93
(1984) 505–529.
[24] M. Alsved, A. Civilis, P. Ekolind, A. Tammelin, A.E. Andersson, J. Jakobsson, T.
Svensson, M. Ramstorp, S. Sadrizadeh, P.-A. Larsson, M. Bohgard, T. Šantl-Temkiv, J.
Löndahl, Temperature-controlled airflow ventilation in operating rooms compared with
laminar airflow and turbulent mixed airflow, Journal of Hospital Infection. 98 (2018)
181–190.
[25] S. Sadrizadeh, A. Tammelin, P. Ekolind, S. Holmberg, Influence of staff number and
internal constellation on surgical site infection in an operating room, Particuology. 13
(2014) 42–51.
[26] H. Ellis, S. Abdalla, S. Abdalla, A History of Surgery, 3rd edition, CRC Press, 2018.
[27] G.B. Phillips, R.S. Runkle, Biomedical Applications of Laminar Airflow, 1st ed., CRC
Press, 1973.
[28] E. Abel, P.-E. Nilsson, L. Ekberg, P. Fahlén, L. Jagemar, R. Clark, O. Fanger, K.
Fitzner, L. Gunnarsen, P.V. Nielsen, J. Stoops, A. Trüschel, P. Wargocki, Achieving

68
 BIBLIOGRAPHY

the desired indoor climate - Energy efficiency aspects of system design,

Studentlitteratur, 2003.
[29] M. Essex-Lopresti, Operating theatre design, The Lancet. 353 (1999) 1007–1010.
[30] S.C. Mellinghoff, J.J. Vehreschild, B.J. Liss, O.A. Cornely, Epidemiology of Surgical
Site Infections With Staphylococcus aureus in Europe: Protocol for a Retrospective,
Multicenter Study, JMIR Research Protocols. 7 (2018) e63.
[31] A. Hambraeus, Aerobiology in the operating room—a review, Journal of Hospital
Infection. 11 (1988) 68–76.
[32] D.J. Anderson, K.S. Kaye, Staphylococcal Surgical Site Infections, Infectious Disease
Clinics of North America. 23 (2009) 53–72.
[33] D.J. Anderson, Surgical Site Infections, Infectious Disease Clinics of North America.
25 (2011) 135–153.
[34] G.A.J. Ayliffe, B.J. Collins, Wound infections acquired from a disperser of an unusual
strain of Staphylococcus aureus, Journal of Clinical Pathology. 20 (1967) 195–198.
[35] R.B. Kundsin, Documentation of airborne infection during surgery, Annals of the New
York Academy of Sciences. 353 (1980) 255–261.
[36] E.I. Tanner, J. Bullin, C.H. Bullin, D.R. Gamble, An outbreak of post-operative sepsis
due to a staphylococcal disperser, Journal of Hygiene. 85 (1980) 219–225.
[37] R.R. Davies, W.C. Noble, Dispersal of bacteria on desquamated skin, Lancet (London,
England). 2 (1962) 1295–1297.
[38] J. Dankert, J.B. Zijlstra, H. Lubberding, A garment for use in the operating theatre: the
effect upon bacterial shedding, Journal of Hygiene. 82 (1979) 7–14.
[39] C. Noguchi, H. Koseki, H. Horiuchi, A. Yonekura, M. Tomita, T. Higuchi, S.
Sunagawa, M. Osaki, Factors contributing to airborne particle dispersal in the operating
room, BMC Surgery. 17 (2017).
[40] J.E. Woods, D.T. Braymen, R.W. Rasmussen, G.L. Reynolds, G.M. Montag,
Ventilation Requirements in Hospital Operating Rooms – Part I: Control of Airborne
Particles, ASHRAE Transactions. 92 (1986) 396–426.
[41] W.C. NOBLE, J.D.F. HABBEMA, R. VAN FURTH, I. SMITH, C. DE RAAY,
Quantitative studies on the dispersal of skin bacteria into the air, Journal of Medical
Microbiology,. 9 (1976) 53–61.
[42] W.C. Noble, Dispersal of skin microorganisms, British Journal of Dermatology. 93
(1975) 477–485.
[43] L.H. Jansen, M.T. Hojyo‐Tomoko, A.M. Kligman, Improved fluorescence staining
technique for estimating turnover of the human stratum corneum, British Journal of
Dermatology. 90 (1974) 9–12.
[44] W.C. Noble, R.R. Davies, Studies on the dispersal of staphylococci, Journal of Clinical
Pathology. 18 (1965) 16–19.
[45] G.W. Sciple, D.K. Riemensnider, C.A.J. Schleyer, Recovery of Microorganisms Shed
by Humans into a Sterilized Environment, APPL. MICROBIOL. 15 (1967) 5.
[46] G.W. Stocks, S.D. Self, B. Thompson, X.A. Adame, D.P. O’Connor, Predicting
bacterial populations based on airborne particulates: A study performed in nonlaminar
flow operating rooms during joint arthroplasty surgery, American Journal of Infection
Control. 38 (2010) 199–204.
[47] A.E. Andersson, I. Bergh, J. Karlsson, B.I. Eriksson, K. Nilsson, Traffic flow in the
operating room: An explorative and descriptive study on air quality during orthopedic
trauma implant surgery, American Journal of Infection Control. 40 (2012) 750–755.
[48] G.W. Stocks, D.P. O’Connor, S.D. Self, G.A. Marcek, B.L. Thompson, Directed Air
Flow to Reduce Airborne Particulate and Bacterial Contamination in the Surgical Field
During Total Hip Arthroplasty, The Journal of Arthroplasty. 26 (2011) 771–776.

69
BIBLIOGRAPHY

[49] R.O. Darouiche, D.M. Green, M.A. Harrington, B.L. Ehni, P. Kougias, C.F. Bechara,
D.P. O’Connor, Association of Airborne Microorganisms in the Operating Room With
Implant Infections: A Randomized Controlled Trial, Infection Control & Hospital
Epidemiology. 38 (2017) 3–10.
[50] E.A. Grice, J.A. Segre, The skin microbiome, Nature Reviews Microbiology. 9 (2011)
244–253.
[51] O.M. Lidwell, C.A. Mackintosh, A.G. Towers, The evaluation of fabrics in relation to
their use as protective garments in nursing and surgery. II. Dispersal of skin organisms
in a test chamber., The Journal of Hygiene. 81 (1978) 453–469.
[52] C.A. Mackintosh, O.M. Lidwell, A.G. Towers, R.R. Marples, The dimensions of skin
fragments dispersed into the air during activity., The Journal of Hygiene. 81 (1978)
471–479.
[53] J. Hill, A. Howell, R. Blowers, EFFECT OF CLOTHING ON DISPERSAL OF
STAPHYLOCOCCUS AUREUS BY MALES AND FEMALES, The Lancet. 304
(1974) 1131–1133.
[54] N.J. Mitchell, D.R. Gamble, CLOTHING DESIGN FOR OPERATING-ROOM
PERSONNEL, The Lancet. 304 (1974) 1133–1136.
[55] R. Speers, H. Bernard, F. O’grady, R.A. Shooter, INCREASED DISPERSAL OF
SKIN BACTERIA INTO THE AIR AFTER SHOWER-BATHS, Lancet (London,
England). 1 (1965) 478–480.
[56] R. Speers, F.W. O’Grady, R.A. Shooter, H.R. Bernard, W.R. Cole, INCREASED
DISPERSAL OF SKIN BACTERIA INTO THE AIR AFTER SHOWER BATHS:
THE EFFECT OF HEXACHLOROPHENE, The Lancet. 287 (1966) 1298–1299.
[57] P.E. Wilson, A Comparison of Methods for Assessing the Value of Antibacterial
Soaps, Journal of Applied Bacteriology. 33 (1970) 574–581.
[58] N.J. Mitchell, D.S. Evans, A. Kerr, Reduction of skin bacteria in theatre air with
comfortable, non-woven disposable clothing for operating-theatre staff., BMJ. 1 (1978)
696–698.
[59] A. Tammelin, B. Ljungqvist, B. Reinmüller, Single-use surgical clothing system for
reduction of airborne bacteria in the operating room, The Journal of Hospital Infection.
84 (2013) 245–247.
[60] C.M. Doig, The effect of clothing on the dissemination of bacteria in operating theatres,
British Journal of Surgery. 59 (1972) 878–881.
[61] R. Blowers, M. Mccluskey, DESIGN OF OPERATING-ROOM DRESS FOR
SURGEONS, The Lancet. 286 (1965) 681–683.
[62] W.C. Noble, The size distribution of airborne particles carrying Clostridium welchii,
The Journal of Pathology and Bacteriology. 81 (1961) 523–526.
[63] W.C. Noble, O.M. Lidwell, D. Kingston, The size distribution of airborne particles
carrying micro-organisms, The Journal of Hygiene. 61 (1963) 385–391.
[64] D. Hansen, C. Krabs, D. Benner, A. Brauksiepe, W. Popp, Laminar air flow provides
high air quality in the operating field even during real operating conditions, but
personal protection seems to be necessary in operations with tissue combustion,
International Journal of Hygiene and Environmental Health. 208 (2005) 455–460.
[65] F. Memarzadeh, A.P. Manning, Comparison of operating room ventilation systems in
the protection of the surgical site, ASHRAE Transactions; Atlanta. 108 (2002) 3–15.
[66] D.V. Seal, R.P. Clark, Electronic particle counting for evaluating the quality of air in
operating theatres: a potential basis for standards?, The Journal of Applied
Bacteriology. 68 (1990) 225–230.

70
 BIBLIOGRAPHY

[67] P. Jalovaara, J. Puranen, Air bacterial and particle counts in total hip replacement
operations using non-woven and cotton gowns and drapes, Journal of Hospital
Infection. 14 (1989) 333–338.
[68] S. Dharan, D. Pittet, Environmental controls in operating theatres, Journal of Hospital
Infection. 51 (2002) 79–84.
[69] A. Landrin, A. Bissery, G. Kac, Monitoring air sampling in operating theatres: can
particle counting replace microbiological sampling?, The Journal of Hospital Infection.
61 (2005) 27–29.
[70] M.L. Cristina, A.M. Spagnolo, M. Sartini, D. Panatto, R. Gasparini, P. Orlando, G.
Ottria, F. Perdelli, Can Particulate Air Sampling Predict Microbial Load in Operating
Theatres for Arthroplasty?, PLOS ONE. 7 (2012) e52809.
[71] G. Birgand, G. Toupet, S. Rukly, G. Antoniotti, M.-N. Deschamps, D. Lepelletier, C.
Pornet, J.B. Stern, Y.-M. Vandamme, N. van der Mee-Marquet, J.-F. Timsit, J.-C.
Lucet, Air contamination for predicting wound contamination in clean surgery: A large
multicenter study, American Journal of Infection Control. 43 (2015) 516–521.
[72] B. Friberg, S. Friberg, L.G. Burman, Correlation between surface and air counts of
particles carrying aerobic bacteria in operating rooms with turbulent ventilation: an
experimental study, Journal of Hospital Infection. 42 (1999) 61–68.
[73] P. Hoffman, H. Humphreys, Air sampling: settle plates or slit samplers?, Journal of
Hospital Infection. 49 (2001) 299–300.
[74] C. Napoli, V. Marcotrigiano, M.T. Montagna, Air sampling procedures to evaluate
microbial contamination: a comparison between active and passive methods in
operating theatres, BMC Public Health. 12 (2012) 594.
[75] ISO:, Cleanrooms and associated controlled environments – Biocontamination control
– Part 1: General principles and methods, 2003.
[76] W. Whyte, A. Hambraeus, G. Laurell, J. Hoborn, The relative importance of the routes
and sources of wound contamination during general surgery. II. Airborne, Journal of
Hospital Infection. 22 (1992) 41–54.
[77] G.J. Taylor, G.C. Bannister, Infection and interposition between ultraclean air source
and wound, The Journal of Bone and Joint Surgery. British Volume. 75 (1993) 503–
504.
[78] B. Friberg, S. Friberg, L.G. Burman, Inconsistent correlation between aerobic bacterial
surface and air counts in operating rooms with ultra clean laminar air flows: proposal of
a new bacteriological standard for surface contamination, Journal of Hospital Infection.
42 (1999) 287–293.
[79] C. Napoli, S. Tafuri, L. Montenegro, M. Cassano, A. Notarnicola, S. Lattarulo, M.T.
Montagna, B. Moretti, Air sampling methods to evaluate microbial contamination in
operating theatres: results of a comparative study in an orthopaedics department,
Journal of Hospital Infection. 80 (2012) 128–132.
[80] W.J. Sayer, N.M. MacKnight, H.W. Wilson, Hospital Airborne Bacteria as Estimated
by the Andersen Sampler versus the Gravity Settling Culture Plate, American Journal
of Clinical Pathology. 58 (1972) 558–566.
[81] T.T. Chow, X.Y. Yang, Ventilation performance in operating theatres against airborne
infection: review of research activities and practical guidance, Journal of Hospital
Infection. 56 (2004) 85–92.
[82] P.N. Hoffman, J. Williams, A. Stacey, A.M. Bennett, G.L. Ridgway, C. Dobson, I.
Fraser, H. Humphreys, Microbiological commissioning and monitoring of operating
theatre suites, Journal of Hospital Infection. 52 (2002) 1–28.

71
BIBLIOGRAPHY

[83] S. Iudicello, A. Fadda, A Road Map to a Comprehensive Regulation on Ventilation

Technology for Operating Rooms, Infection Control and Hospital Epidemiology. 34
(2013) 858–860.
[84] S. Fischer, M. Thieves, T. Hirsch, K.-D. Fischer, H. Hubert, S. Bepler, H.-M. Seipp,
Reduction of Airborne Bacterial Burden in the OR by Installation of Unidirectional
Displacement Airflow (UDF) Systems, Medical Science Monitor : International
Medical Journal of Experimental and Clinical Research. 21 (2015) 2367–2374.
[85] M. Diab-Elschahawi, J. Berger, A. Blacky, O. Kimberger, R. Oguz, R. Kuelpmann, A.
Kramer, O. Assadian, Impact of different-sized laminar air flow versus no laminar air
flow on bacterial counts in the operating room during orthopedic surgery, American
Journal of Infection Control. 39 (2011) e25–e29.
[86] O.M. Lidwell, I.D.G. Richards, S. Polakoff, Comparison of Three Ventilating Systems
in an Operating Room, The Journal of Hygiene. 65 (1967) 193–205.
[87] Friberg Barbro, Ultraclean Laminar Airflow ORs, AORN Journal. 67 (2012) 841–851.
[88] D. Bosanquet, C. Jones, N. Gill, P. Jarvis, M. Lewis, Laminar flow reduces cases of
surgical site infections in vascular patients, Annals of The Royal College of Surgeons
of England. 95 (2013) 15–19.
[89] R.G. Kakwani, D. Yohannan, K.H.A. Wahab, The effect of laminar air-flow on the
results of Austin-Moore hemiarthroplasty, Injury. 38 (2007) 820–823.
[90] A. Erichsen Andersson, M. Petzold, I. Bergh, J. Karlsson, B.I. Eriksson, K. Nilsson,
Comparison between mixed and laminar airflow systems in operating rooms and the
influence of human factors: Experiences from a Swedish orthopedic center, American
Journal of Infection Control. 42 (2014) 665–669.
[91] T. Hirsch, H. Hubert, S. Fischer, A. Lahmer, M. Lehnhardt, H.-U. Steinau, L.
Steinstraesser, H.-M. Seipp, Bacterial burden in the operating room: Impact of airflow
systems, American Journal of Infection Control. 40 (2012) e228–e232.
[92] M. Rezapoor, A. Alvand, E. Jacek, T. Paziuk, M.G. Maltenfort, J. Parvizi, Operating
Room Traffic Increases Aerosolized Particles and Compromises the Air Quality: A
Simulated Study, The Journal of Arthroplasty. 33 (2018) 851–855.
[93] O.M. Lidwell, E.J. Lowbury, W. Whyte, R. Blowers, S.J. Stanley, D. Lowe, Effect of
ultraclean air in operating rooms on deep sepsis in the joint after total hip or knee
replacement: a randomised study., British Medical Journal (Clinical Research Ed.). 285
(1982) 10–14.
[94] C. Brandt, U. Hott, D. Sohr, F. Daschner, P. Gastmeier, H. Rüden, Operating room
ventilation with laminar airflow shows no protective effect on the surgical site infection
rate in orthopedic and abdominal surgery, Annals of Surgery. 248 (2008) 695–700.
[95] P.A. Lipsett, Do We Really Need Laminar Flow Ventilation in the Operating Room to
Prevent Surgical Site Infections?, Annals of Surgery. 248 (2008) 701–703.
[96] T.E. Salassa, M.F. Swiontkowski, Surgical Attire and the Operating Room: Role in
Infection Prevention, The Journal of Bone and Joint Surgery – American Volume. 96
(2014) 1485–1492.
[97] G.J. Hooper, A.G. Rothwell, C. Frampton, M.C. Wyatt, Does the use of laminar flow
and space suits reduce early deep infection after total hip and knee replacement?: THE
TEN-YEAR RESULTS OF THE NEW ZEALAND JOINT REGISTRY, Bone & Joint
Journal. 93-B (2011) 85–90.
[98] P. Gastmeier, A.-C. Breier, C. Brandt, Influence of laminar airflow on prosthetic joint
infections: a systematic review, Journal of Hospital Infection. 81 (2012) 73–78.
[99] P. Bischoff, N.Z. Kubilay, B. Allegranzi, M. Egger, P. Gastmeier, Effect of laminar
airflow ventilation on surgical site infections: a systematic review and meta-analysis,
The Lancet Infectious Diseases. 17 (2017) 553–561.

72
 BIBLIOGRAPHY

[100] S.M. McHugh, A.D.K. Hill, H. Humphreys, Laminar airflow and the prevention of
surgical site infection. More harm than good?, The Surgeon. 13 (2015) 52–58.
[101] D.F. de Korne, J.D.H. van Wijngaarden, J. van Rooij, L.S.G.L. Wauben, U.F.
Hiddema, N.S. Klazinga, Safety by design: effects of operating room floor marking on
the position of surgical devices to promote clean air flow compliance and minimise
infection risks, BMJ Qual Saf. 21 (2012) 746–752.
[102] E.B. Smith, I.J. Raphael, M.G. Maltenfort, S. Honsawek, K. Dolan, E.A. Younkins,
The Effect of Laminar Air Flow and Door Openings on Operating Room
Contamination, The Journal of Arthroplasty. 28 (2013) 1482–1485.
[103] T.-T. Chow, J. Wang, Dynamic simulation on impact of surgeon bending movement on
bacteria-carrying particles distribution in operating theatre, Building and Environment.
57 (2012) 68–80.
[104] T.-T. Chow, X.-Y. Yang, Performance of ventilation system in a non-standard
operating room, Building and Environment. 38 (2003) 1401–1411.
[105] T.T. Chow, X.Y. Yang, Ventilation performance in the operating theatre against
airborne infection: numerical study on an ultra-clean system, Journal of Hospital
Infection. 59 (2005) 138–147.
[106] T.T. Chow, Z. Lin, W. Bai, The Integrated Effect of Medical Lamp Position and
Diffuser Discharge Velocity on Ultra-clean Ventilation Performance in an Operating
Theatre, Indoor and Built Environment. 15 (2006) 315–331.
[107] W. Whyte, B.H. Shaw, The effect of obstructions and thermals in laminar-flow
systems, The Journal of Hygiene. 72 (1974) 415–423.
[108] W.A.C. Zoon, M.G.M. van der Heijden, M.G.L.C. Loomans, J.L.M. Hensen, On the
applicability of the laminar flow index when selecting surgical lighting, Building and
Environment. 45 (2010) 1976–1983.
[109] A. Aganovic, G. Cao, L.-I. Stenstad, J.G. Skogås, Impact of surgical lights on the
velocity distribution and airborne contamination level in an operating room with
laminar airflow system, Building and Environment. 126 (2017) 42–53.
[110] A. Aganovic, G. Cao, L.-I. Stenstad, J.G. Skogås, An experimental study on the effects
of positioning medical equipment on contaminant exposure of a patient in an operating
room with unidirectional downflow, Building and Environment. 165 (2019) 106096.
[111] G. Cao, M.C.A. Storås, A. Aganovic, L.-I. Stenstad, J.G. Skogås, Do surgeons and
surgical facilities disturb the clean air distribution close to a surgical patient in an
orthopedic operating room with laminar airflow?, American Journal of Infection
Control. 46 (2018) 1115–1122.
[112] G. Cao, A.M. Nilssen, Z. Cheng, L.-I. Stenstad, A. Radtke, J.G. Skogås, Laminar
airflow and mixing ventilation: Which is better for operating room airflow distribution
near an orthopedic surgical patient?, American Journal of Infection Control. 47 (2019)
737–743.
[113] S. Jain, M. Reed, Laminar Air Flow Handling Systems in the Operating Room,
Surgical Infections. 20 (2019) 151–158.
[114] R.P. Evans, Current Concepts for Clean Air and Total Joint Arthroplasty: Laminar
Airflow and Ultraviolet Radiation: A Systematic Review, Clinical Orthopaedics and
Related Research®. 469 (2011) 945–953.
[115] P. Cacciari, R. Giannoni, E. Marcelli, L. Cercenelli, Cost evaluation of a ventilation
system for operating theatre: an ultraclean design versus a conventional one [in Italian],
Annali Di Igiene: Medicina Preventiva E Di Comunita. 16 (2004) 803–809.
[116] C. Pasquarella, G.E. Sansebastiano, S. Ferretti, E. Saccani, M. Fanti, U. Moscato, G.
Giannetti, S. Fornia, P. Cortellini, P. Vitali, C. Signorelli, A mobile laminar airflow

73
BIBLIOGRAPHY

unit to reduce air bacterial contamination at surgical area in a conventionally ventilated

operating theatre, Journal of Hospital Infection. 66 (2007) 313–319.
[117] K.-G. Nilsson, R. Lundholm, S. Friberg, Assessment of horizontal laminar air flow
instrument table for additional ultraclean space during surgery, Journal of Hospital
Infection. 76 (2010) 243–246.
[118] D. Sossai, G. Dagnino, F. Sanguineti, F. Franchin, Mobile laminar air flow screen for
additional operating room ventilation: reduction of intraoperative bacterial
contamination during total knee arthroplasty, Journal of Orthopaedics and
Traumatology. 12 (2011) 207–211.
[119] B. Friberg, M. Lindgren, C. Karlsson, A. Bergström, S. Friberg, Mobile
zoned/exponential LAF screen: a new concept in ultra-clean air technology for
additional operating room ventilation, Journal of Hospital Infection. 50 (2002) 286–
292.
[120] B. Ljungqvist, B. Reinmüller, J. Gustén, L. Gustén, J. Nordenadler, Contamination
Risks due to Door Openings in Operating Rooms, European Journal of Parenteral &
Pharmaceutical Sciences. 14 (2009) 97–101.
[121] S.D. Keimig, M.L. Galvin, E.B. Sansone, The Influence of the Direction of Door
Swing and Door Velocity on the Transfer of Airborne Contaminants in a Scale Model,
Applied Industrial Hygiene. 2 (1987) 103–107.
[122] B. Ljungqvist, B. Reinmüller, Clean Room Design: Minimizing Contamination
Through Proper Design, 1st ed., Routledge, 1996.
[123] M. Hyttinen, A. Rautio, P. Pasanen, T. Reponen, G.S. Earnest, A. Streifel, P.
Kalliokoski, Airborne Infection Isolation Rooms – A Review of Experimental Studies,
Indoor and Built Environment. 20 (2011) 584–594.
[124] J.M. Villafruela, J.F. San José, F. Castro, A. Zarzuelo, Airflow patterns through a
sliding door during opening and foot traffic in operating rooms, Building and
Environment. 109 (2016) 190–198.
[125] B. Shaw, W. Whyte, Air movement through doorways - the influence of temperature
and its control by forced airflow, Journal of Building Services Engineering. 42 (1974)
210–218.
[126] D.E. Kiel, D.J. Wilson, Combining Door Swing Pumping with Density Driven Flow,
ASHRAE Transactions. 95 (1989) 590–599.
[127] O.M. Lidwell, Air exchange through doorways. The effect of temperature difference,
turbulence and ventilation flow, Journal of Hygiene. 79 (1977) 141–154.
[128] D.J. Wilson, D.E. Kiel, Gravity driven counterflow through an open door in a sealed
room, Building and Environment. 25 (1990) 379–388.
[129] L. Fontana, A. Quintino, Experimental analysis of the transport of airborne
contaminants between adjacent rooms at different pressure due to the door opening,
Building and Environment. 81 (2014) 81–91.
[130] A. Hathway, I. Papakonstantis, A. Bruce-Konuah, W. Brevis, Experimental and
Modelling Investigations of Air Exchange and Infection Transfer due to Hinged-Door
Motion in Office and Hospital Settings, International Journal of Ventilation. 14 (2015)
127–140.
[131] I.G. Papakonstantis, E.A. Hathway, W. Brevis, An experimental study of the flow
induced by the motion of a hinged door separating two rooms, Building and
Environment. 131 (2018) 220–230.
[132] J.W. Tang, I. Eames, Y. Li, Y.A. Taha, P. Wilson, G. Bellingan, K.N. Ward, J. Breuer,
Door-opening motion can potentially lead to a transient breakdown in negative-
pressure isolation conditions: the importance of vorticity and buoyancy airflows,
Journal of Hospital Infection. 61 (2005) 283–286.

74
 BIBLIOGRAPHY

[133] J.W. Tang, A. Nicolle, J. Pantelic, C.A. Klettner, R. Su, P. Kalliomaki, P. Saarinen, H.
Koskela, K. Reijula, P. Mustakallio, D.K.W. Cheong, C. Sekhar, K.W. Tham, Different
Types of Door-Opening Motions as Contributing Factors to Containment Failures in
Hospital Isolation Rooms, PLOS ONE. 8 (2013) e66663.
[134] J. Hendiger, M. Chludzińska, P. Ziętek, Influence of the Pressure Difference and Door
Swing on Heavy Contaminants Migration between Rooms, PLOS ONE. 11 (2016)
e0155159.
[135] P. Kalliomäki, P. Saarinen, J.W. Tang, H. Koskela, Airflow Patterns through Single
Hinged and Sliding Doors in Hospital Isolation Rooms, International Journal of
Ventilation. 14 (2015) 111–126.
[136] P. Kalliomäki, P. Saarinen, J.W. Tang, H. Koskela, Airflow patterns through single
hinged and sliding doors in hospital isolation rooms – Effect of ventilation, flow
differential and passage, Building and Environment. 107 (2016) 154–168.
[137] L. Chang, X. Zhang, Y. Cai, Experimental determination of air inleakage to pressurized
main control room caused by personnel entering, Building and Environment. 99 (2016)
142–148.
[138] L. Chang, S. Tu, W. Ye, X. Zhang, Dynamic simulation of contaminant inleakage
produced by human walking into control room, International Journal of Heat and Mass
Transfer. 113 (2017) 1179–1188.
[139] S. Lee, B. Park, T. Kurabuchi, Numerical evaluation of influence of door opening on
interzonal air exchange, Building and Environment. 102 (2016) 230–242.
[140] E.S. Mousavi, K.R. Grosskopf, Airflow patterns due to door motion and pressurization
in hospital isolation rooms, Science and Technology for the Built Environment. 22
(2016) 379–384.
[141] P.E. Saarinen, P. Kalliomäki, J.W. Tang, H. Koskela, Large Eddy Simulation of Air
Escape through a Hospital Isolation Room Single Hinged Doorway—Validation by
Using Tracer Gases and Simulated Smoke Videos, PLOS ONE. 10 (2015) e0130667.
[142] P. Saarinen, P. Kalliomäki, H. Koskela, J.W. Tang, Large-eddy simulation of the
containment failure in isolation rooms with a sliding door—An experimental and
modelling study, Building Simulation. 11 (2018) 585–596.
[143] Y.-C. Shih, C.-C. Chiu, O. Wang, Dynamic airflow simulation within an isolation
room, Building and Environment. 42 (2007) 3194–3209.
[144] J. Hang, Y. Li, W.H. Ching, J. Wei, R. Jin, L. Liu, X. Xie, Potential airborne
transmission between two isolation cubicles through a shared anteroom, Building and
Environment. 89 (2015) 264–278.
[145] R.J. Lynch, M.J. Englesbe, L. Sturm, A. Bitar, K. Budhiraj, S. Kolla, Y. Polyachenko,
M.G. Duck, D.A. Campbell, Measurement of Foot Traffic in the Operating Room:
Implications for Infection Control, American Journal of Medical Quality. 24 (2009)
45–52.
[146] S. Elliott, S. Parker, J. Mills, L. Meeusen, T. Frana, M. Anderson, A. Storsveen, A.
White, STOP: Can We Minimize OR Traffic?, AORN Journal. 102 (2015) 409.e1-
409.e7.
[147] P. Panahi, M. Stroh, D.S. Casper, J. Parvizi, M.S. Austin, Operating Room Traffic is a
Major Concern During Total Joint Arthroplasty, Clinical Orthopaedics and Related
Research®. 470 (2012) 2690–2694.
[148] N.M.C. Mathijssen, G. Hannink, P.D.J. Sturm, P. Pilot, R.M. Bloem, P. Buma, P.L.C.
Petit, B.W. Schreurs, The Effect of Door Openings on Numbers of Colony Forming
Units in the Operating Room during Hip Revision Surgery, Surgical Infections. 17
(2016) 535–540.

75
BIBLIOGRAPHY

[149] P. Perez, J. Holloway, L. Ehrenfeld, S. Cohen, L. Cunningham, G.B. Miley, B.L.

Hollenbeck, Door openings in the operating room are associated with increased
environmental contamination, American Journal of Infection Control. 46 (2018) 954–
956.
[150] M.T. Stauning, A. Bediako-Bowan, L.P. Andersen, J.A. Opintan, A.-K. Labi, J.A.L.
Kurtzhals, S. Bjerrum, Traffic flow and microbial air contamination in operating rooms
at a major teaching hospital in Ghana, Journal of Hospital Infection. 99 (2018) 263–
270.
[151] Y. Babkin, D. Raveh, M. Lifschitz, M. Itzchaki, Y. Wiener-Well, P. Kopuit, Z. Jerassy,
A.M. Yinnon, Incidence and risk factors for surgical infection after total knee
replacement, Scandinavian Journal of Infectious Diseases. 39 (2007) 890–895.
[152] R.M.P.H. Crolla, L. van der Laan, E.J. Veen, Y. Hendriks, C. van Schendel, J.
Kluytmans, Reduction of Surgical Site Infections after Implementation of a Bundle of
Care, PLoS ONE. 7 (2012) e44599.
[153] F. Romano, L. Marocco, J. Gustén, C.M. Joppolo, Numerical and experimental analysis
of airborne particles control in an operating theater, Building and Environment. 89
(2015) 369–379.
[154] S. Sadrizadeh, S. Holmberg, A. Tammelin, A numerical investigation of vertical and
horizontal laminar airflow ventilation in an operating room, Building and Environment.
82 (2014) 517–525.
[155] F. Moukalled, L. Mangani, M. Darwish, The Finite Volume Method in Computational
Fluid Dynamics - An Advanced Introduction with OpenFOAM® and Matlab, 2015.
[156] Q. Chen, COMPARISON OF DIFFERENT k-ε MODELS FOR INDOOR AIR FLOW
COMPUTATIONS, Numerical Heat Transfer, Part B: Fundamentals. 28 (1995) 353–
369.
[157] Q. Chen, Prediction of room air motion by Reynolds-stress models, Building and
Environment. 31 (1996) 233–244.
[158] T.-H. Shih, W.W. Liou, A. Shabbir, Z. Yang, J. Zhu, A new k-ϵ eddy viscosity model
for high reynolds number turbulent flows, Computers & Fluids. 24 (1995) 227–238.
[159] S. Sadrizadeh, S. Holmberg, A. Tammelin, A numerical investigation of vertical and
horizontal laminar airflow ventilation in an operating room, Building and Environment.
82 (2014) 517–525.
[160] J. Srebric, V. Vukovic, G. He, X. Yang, CFD boundary conditions for contaminant
dispersion, heat transfer and airflow simulations around human occupants in indoor
environments, Building and Environment. 43 (2008) 294–303.
[161] Z. Zhang, W. Zhang, Z.J. Zhai, Q.Y. Chen, Evaluation of Various Turbulence Models
in Predicting Airflow and Turbulence in Enclosed Environments by CFD: Part 2—
Comparison with Experimental Data from Literature, HVAC&R Research. 13 (2007)
871–886.
[162] S. Sadrizadeh, S. Holmberg, How safe is it to neglect thermal radiation in indoor
environment modeling with high ventilation rates?, in: Proceedings of the 36th AIVC-
5th TightVent-3rd Venticool Conference, 2015: pp. 23–24.
[163] H. Brohus, K.D. Balling, D. Jeppesen, Influence of movements on contaminant
transport in an operating room, Indoor Air. 16 (2006) 356–372.
[164] B. Zhou, L. Ding, F. Li, K. Xue, P.V. Nielsen, Y. Xu, Influence of opening and closing
process of sliding door on interface airflow characteristic in operating room, Building
and Environment. 144 (2018) 459–473.
[165] C. Balocco, A Simplified Numerical Model for Simulating Sliding Door and Surgical
Staff Movement in an Operating Theatre, in: 2012: p. 7.

76
 BIBLIOGRAPHY

[166] R. Carneiro, P.D. Gaspar, P.D. Silva, 3D and transient numerical modelling of door
opening and closing processes and its influence on thermal performance of cold rooms,
Applied Thermal Engineering. 113 (2017) 585–600.
[167] D. Hryb, M. Cardozo, S. Ferro, M. Goldschmit, Particle transport in turbulent flow
using both Lagrangian and Eulerian formulations, International Communications in
Heat and Mass Transfer. 36 (2009) 451–457.
[168] J.S. Shirolkar, C.F.M. Coimbra, M. Queiroz McQuay, Fundamental aspects of
modeling turbulent particle dispersion in dilute flows, Progress in Energy and
Combustion Science. 22 (1996) 363–399.
[169] E. Loth, Numerical approaches for motion of dispersed particles, droplets and bubbles,
Progress in Energy and Combustion Science. 26 (2000) 161–223.
[170] Z. Zhang, Q. Chen, Comparison of the Eulerian and Lagrangian methods for predicting
particle transport in enclosed spaces, Atmospheric Environment. 41 (2007) 5236–5248.
[171] A.C.K. Lai, F.Z. Chen, Comparison of a new Eulerian model with a modified
Lagrangian approach for particle distribution and deposition indoors, Atmospheric
Environment. 41 (2007) 5249–5256.
[172] B. Zhao, Y. Zhang, X. Li, X. Yang, D. Huang, Comparison of indoor aerosol particle
concentration and deposition in different ventilated rooms by numerical method,
Building and Environment. 39 (2004) 1–8.
[173] T.-J. Chang, T.-S. Hu, Transport mechanisms of airborne particulate matters in
partitioned indoor environment, Building and Environment. 43 (2008) 886–895.
[174] S.A. Morsi, A.J. Alexander, An investigation of particle trajectories in two-phase flow
systems, Journal of Fluid Mechanics. 55 (1972) 193–208.
[175] S. Elghobashi, On predicting particle-laden turbulent flows, Applied Scientific
Research. 52 (1994) 309–329.
[176] Z. Zhang, Q. Chen, Experimental measurements and numerical simulations of particle
transport and distribution in ventilated rooms, Atmospheric Environment. 40 (2006)
3396–3408.
[177] W.C. Hinds, Aerosol Technology: Properties, Behavior, and Measurement of Airborne
Particles, John Wiley & Sons, 2012.
[178] A. Makhoul, K. Ghali, N. Ghaddar, W. Chakroun, Investigation of particle transport in
offices equipped with ceiling-mounted personalized ventilators, Building and
Environment. 63 (2013) 97–107.
[179] B. Zhao, C. Chen, X. Yang, A.C.K. Lai, Comparison of Three Approaches to Model
Particle Penetration Coefficient through a Single Straight Crack in a Building
Envelope, Aerosol Science and Technology. 44 (2010) 405–416.
[180] E.A. Matida, K. Nishino, K. Torii, Statistical simulation of particle deposition on the
wall from turbulent dispersed pipe flow, International Journal of Heat and Fluid Flow.
21 (2000) 389–402.
[181] A. Li, G. Ahmadi, Dispersion and Deposition of Spherical Particles from Point Sources
in a Turbulent Channel Flow, Aerosol Science and Technology. 16 (1992) 209–226.
[182] J.B. McLaughlin, Aerosol particle deposition in numerically simulated channel flow,
Physics of Fluids A: Fluid Dynamics. 1 (1989) 1211–1224.
[183] C. Narayanan, D. Lakehal, L. Botto, A. Soldati, Mechanisms of particle deposition in a
fully developed turbulent open channel flow, Physics of Fluids. 15 (2003) 763–775.
[184] B. Zhao, J. Wu, Particle deposition in indoor environments: Analysis of influencing
factors, Journal of Hazardous Materials. 147 (2007) 439–448.
[185] M. Wang, C.-H. Lin, Q. Chen, Advanced turbulence models for predicting particle
transport in enclosed environments, Building and Environment. 47 (2012) 40–49.

77
BIBLIOGRAPHY

[186] C. Chen, W. Liu, F. Li, C.-H. Lin, J. Liu, J. Pei, Q. Chen, A hybrid model for
investigating transient particle transport in enclosed environments, Building and
Environment. 62 (2013) 45–54.
[187] N.P. Gao, J.L. Niu, Modeling particle dispersion and deposition in indoor
environments, Atmospheric Environment. 41 (2007) 3862–3876.
[188] J.M. MacInnes, F.V. Bracco, Stochastic particle dispersion modeling and the tracer‐
particle limit, Physics of Fluids A: Fluid Dynamics. 4 (1992) 2809–2824.
[189] W. Lu, A.T. Howarth, N. Adam, S.B. Riffat, Modelling and measurement of airflow
and aerosol particle distribution in a ventilated two-zone chamber, Building and
Environment. 31 (1996) 417–423.
[190] T.T. Chow, Z. Lin, W. Bai, The Integrated Effect of Medical Lamp Position and
Diffuser Discharge Velocity on Ultra-clean Ventilation Performance in an Operating
Theatre, Indoor and Built Environment. 15 (2006) 315–331.
[191] J. Eslami, A. Abbassi, M.H. Saidi, M. Bahrami, Effect of supply/exhaust diffuser
configurations on the contaminant distribution in ultra clean environments: Eulerian
and Lagrangian approaches, Energy and Buildings. 127 (2016) 648–657.
[192] S. Sadrizadeh, S. Holmberg, Optimal positioning of exhaust grills in an operating room
based on recovery test, in: Proceedings of 35th AIVC-4th Tightvent & 2nd Venticool
Conference, Poznań, Poland, 2014.
[193] E. Bjorn, P. V Nielsen, Dispersal of exhaled air and personal exposure in displacement
ventilated rooms, Indoor Air. 12 (2002) 147–164.
[194] W.B. Faulkner, F. Memarzadeh, G. Riskowski, K. Hamilton, C.-Z. Chang, J.-R. Chang,
Particulate concentrations within a reduced-scale room operated at various air exchange
rates, Building and Environment. 65 (2013) 71–80.
[195] W.B. Faulkner, F. Memarzadeh, G. Riskowski, A. Kalbasi, A. Ching-Zu Chang,
Effects of air exchange rate, particle size and injection place on particle concentrations
within a reduced-scale room, Building and Environment. 92 (2015) 246–255.
[196] E.S. Mousavi, K.R. Grosskopf, Ventilation Rates and Airflow Pathways in Patient
Rooms: A Case Study of Bioaerosol Containment and Removal, Annals of
Occupational Hygiene. 59 (2015) 1190–1199.
[197] S. Sadrizadeh, J. Pantelic, M. Sherman, J. Clark, O. Abouali, Airborne particle
dispersion to an operating room environment during sliding and hinged door opening,
Journal of Infection and Public Health. 11 (2018) 631–635.
[198] H. Laidoudi, M. Bouzit, The Effects of Aiding and Opposing Thermal Buoyancy on
Downward Flow around a Confined Circular Cylinder, Periodica Polytechnica
Mechanical Engineering. 62 (2018) 42–50.
[199] N. Sharma, Mixed convection flow and heat transfer across a square cylinder under the
influence of aiding buoyancy at low Reynolds numbers, International Journal of Heat
and Mass Transfer. 55 (2012) 2601–2614.
[200] H. Hu, M.M. Koochesfahani, Thermal effects on the wake of a heated circular cylinder
operating in mixed convection regime, Journal of Fluid Mechanics. 685 (2011) 235–
270.
[201] S.K. Singh, P.K. Panigrahi, K. Muralidhar, Effect of buoyancy on the wakes of circular
and square cylinders: a schlieren-interferometric study, Experiments in Fluids. 43
(2007) 101–123.
[202] D. Chatterjee, Dual role of thermal buoyancy in controlling boundary layer separation
around bluff obstacles, International Communications in Heat and Mass Transfer. 56
(2014) 152–158.

78
 BIBLIOGRAPHY

[203] N. Michaux-Leblond, M. Bélorgey, Near-wake behavior of a heated circular cylinder:

Viscosity-buoyancy duality, Experimental Thermal and Fluid Science. 15 (1997) 91–
100.
[204] D. Chatterjee, B. Mondal, Control of flow separation around bluff obstacles by
superimposed thermal buoyancy, International Journal of Heat and Mass Transfer. 72
(2014) 128–138.

79