Progress in the Application of Ultraviolet
Germicidal Irradiation for Air Disinfection
35th Meeting: American Society for Photobiology
Providence, RI, 16 June, 2010
Edward A. Nardell, MD
Associate Professor, Harvard Medical School
Harvard School of Public Health
Brigham & Women’s Hospital
Division of Global Health Equity
(Partners In Health)
enardell@pih.org
Ultraviolet Germicidal Irradiation
(UVGI) – key concepts:

254 nm UV easily produced


Low skin/eye penetration


by mercury vapor lamps
because it is so reactive – absorbed by outer
layers – nucleic acids protected
Microbes are vulnerable

because of their tiny size – nucleic acids are
exposed
Photokeratitis
Relative effectiveness
UVGI Occupational ExposureSkin
Limit
cancer
Bactericidal
Skin erythem
Wave length (nm)
UVGI is old but still
underdeveloped technology

Applications of Germicidal, Erythemal
and Infrared Energy, 1946

Matthew Luckiesh, DSc, DE, GE Laboratory


24 other books on light
Ultraviolet Irradiation: its properties,
production, measurement, and
applications, 1922
Why upper room UVGI?
The challenge of airborne infections



What are airborne infections?
Why are they so hard to control with
ventilation alone?
What are the advantages of UVGI?



Why UV in ducts
Why upper room UV?
Where should it be applied?
Current upper room UVGI usage



Used in hospital waiting
rooms, clinics,
emergency rooms,
isolation and procedure
rooms
Not considered a
substitute for negative
pressure isolation or for
at least 6 ACH
ventilation
Used in shelters, jails,
prisons, homeless
shelters
Current UVGI fixtures (louvered)
Narrow beam
UV in ventilation ducts or portable room
air disinfection devices
Microbial Inactivation
Microbial inactivation characterized by
survival curves, based on target theory.

Number of survivors decreases by a
constant fractional amount for a given
increment of exposure.

Holds for a single “hit”, but multiple “hits” leads
to survival curves with a shoulder – commonly
seen experimentally.
Limitation common to all air
disinfection strategies


Ventilation
HEPA filtration


Limited by number of
air turnovers
UV air disinfection


In ducts – same
limitation
Upper room –
disinfect large
volume at once
120
100
80
p
60
40
20
0
0
2000
4000
6000
Ventilation, CFM
8000
Upper Room UV Air Disinfection
uses the entire room as a duct
UV-C
.
.
7 ft.
Disinfected air
displaced
Warm contaminated
air rises
Paddle fans assure good air mixing
UVGI Investigations
Type:
Source:
Sample:
Utilitiy:
1. Bench-scale
BCG
aerosol
Mech. A.S.
Culture
UV dose
humidity
2. Room-scale
BCG
aerosol
Mech. A.S.
Culture
Fixtures,
Vent/CFD
3. Hospital ward
Human
M. tb
Guinea
Pig/RFLP
4. Epi
Human
M. tb
Skin Test
Hospital
room
UV efficacy
Real world
UV efficacy
Bench-scale studies
r
i
c
q
UV light
a
m np
Air flow
b
h
c
j
o
k
l
g
f
e
d
n
Determination of UV susceptibility of
various airborne organisms

Z value
ln N0/Nuv
Z = µWatt x sec x cm-2
Z is the slope of the plot of the natural logarithm of colony
count against UV dose
Mtb at 50% humidity =
33 (23-42) Erdman
48 (44-55) 199RB
M. bovis BCG
37 (33-39)
Serratia marcescens
214 (183-245)
Humidity protects the organism
80
70
60
50
% Survival
22-33%
49-62%
85-91%
40
30
20
10
0
400
800
UV dose (uw.sec/cm2)
Determination of UV susceptibility of
various airborne organisms

Z value
ln N0/Nuv
Z = µWatt x sec x cm-2
Z is the slope of the plot of the natural logarithm of colony
count against UV dose
Mtb at 50% humidity =
33 (23-42) Erdman
48 (44-55) 199RB
M. bovis BCG
37 (33-39)
Serratia marcescens
214 (183-245)
Permutt analysis - 1
Z = 0.0041 = Eq ACH/UV dose = 15 Eq ACH/1 µW/cm2
1 µW/cm2
= 15 ACH
Air Changes per Hour, ACH

When a volume equivalent to the
volume of the room enters and is
exhausted



1 ACH well-mixed air removes 63% of air
contaminants
2 ACH well-mixed air removes 84% of air
contaminants
Any air disinfection method that is 63%
effective produces 1 Equivalent ACH
Permutt analysis - 2
Z = 0.0041 = Eq ACH/UV dose = 15 Eq ACH/1 µW/cm2
1 µW/cm2
= 15 ACH or 15 AC/60 min =
1 Eq AC/4 min = conc. reduced
by 63%
Permutt analysis - 3
Z = 0.0041 = Eq ACH/UV dose = 15 Eq ACH/1 µW/cm2
1 µW/cm2
20%
Instantaneous mixing
1 Eq ACH = 20 min (5 x 4 min)
Permutt analysis - 4
Z = 0.0041 = Eq ACH/UV dose = 15 Eq ACH/1 µW/cm2
32 µW/cm2 avg.= 480 Eq AC
1 Eq ACH/7.2 sec.
Instantaneous mixing
1 Eq ACH = 36 sec
or 100 Eq ACH
20%
Permutt analysis - 5
Z = 0.0041 = Eq ACH/UV dose = 15 Eq ACH/1 µW/cm2
100% air disinfection
25 ACH between upper
and lower room
= 25 Eq ACH
20%
Permutt analysis - 6
Z = 0.0041 = Eq ACH/UV dose = 15 Eq ACH/1 µW/cm2
32 µW/cm2 avg.= 480 Eq AC
1 Eq ACH/7.2 sec.
25 ACH mixing
20 Eq ACH in the lower rm
Permutt analysis - 7
Z = 0.0041 = Eq ACH/UV dose = 15 Eq ACH/1 µW/cm2
32 µW/cm2 avg.= 480 Eq AC
1 Eq ACH/7.2 sec.
100 ACH mixing
50 Eq ACH in the lower rm
Room-scale studies:
Riley-Middlebrook, 1976 - aerosolized BCG



A single 17 W UV lamp
added the equivalent of 10
air changes to an
unventilated room – air
mixing by radiator only
Silent, safe, no drafts, low
energy use, low
maintenance
Established current dosage
guideline 30 W fixture per
200 sq ft area.

Am Rev Resp Dis, 1976,
113:413-18.
Airborne Infection
- exposure chamber: HSPH
Aerosol generator: HSPH
Interior, Large exposure chamber, HSPH
Anderson air sampling, HSPH
Summary: efficacy of upper room UVGI
Microorganism
Riley (1976)
BCG
Miller (1999)
Mycobacterium
parafortuitum
0.65-2.1
DW
15-35
50-90
90
Yes
0
6
Yes
Particle size (m m)
0.5-3
0.2% BSA
Suspending medium
n/a
Temperature (°C) n/a
RH (%)
25
20, 40
Room size (m3)
61
Mechanical ventilation
No
ACH
2
2-4
Mixing fan
Yes
(during aerosolization)
UV output (W)
17
46
99(28)
UVoutput/Room size 0.28
0.75
1.1
(W/m3)
UV fixture type
C1
C1&W
CN&C2
UV Effectiveness (%) 83
88, 89
98
UV effect (ACH)
10
18-19, 33
Ko (2000)
BCG
1.1-4.7
10% FCS
4-26
41-69
46
Yes
6-8
6
No
99(28)
1.1
36(10)
0.78
59(15)
1.3
CN&C2
95
C2
52  19
C2&W
64  10
6-16, 19
9.8  6.4 11.7  7.1
Room Studies – Miller et al, 2002
University of Colorado at Boulder, NIOSH contract #200-97-2602


Full scale room studies – 87 m2 test chamber
5 fixtures, tot. 216 W producing



3 test organisms:


avg. 42 µW/cm2 in the irradiated upper zone,
only 0.08 µW/cm2 at eye level
B. subtilis, M. parafortuitum, and M. bovis.
2 types of experiments:


constant generation – gives effectiveness
decay – gives equivalent room air changes
Miller et al, 2002 (cont.)

Results (50% humidity):
 Constant generation (effectiveness):




B. subtilis
M. parafortuitum
M. bovis
Decay (release & mix):

M. parafortuitum
46 – 80% @ 216 W
83 – 98% @ 216 W
96 – 97% @ 216 W
16 ± 1.2 ACH @ 216 W
6.1 ± 0.8 ACH @ 108 W
19% increase @ 509 W
Z value = 1.2 ± 0.15 x10-3 cm2 u.sec-1sec-1
 Winter conditions (warm air from ceiling): 25% decrease
 Photoreactivation
 none detected at 40% humidity
 inconclusive at 100% humidity
Upper room UVGI effect on measles
in day schools, (Wells, Am J Hygiene, 35:97-121, 1942)
TUSS – TB UV Shelter Study

Prospective, double-blinded, placebo-controlled trial
of upper room UVGI in homeless shelters


5 cities:






Phil Brickner, MD, PI
NYC
Birmingham
New Orleans
Houston - two sites
Indeterminate results due to low infection rates and
poor retention of homeless subjects
Lots of data obtained on engineering and safety

Safe, reliable, low-maintenance
Experimental Hospital Ward
Riley (Wells) experimental TB ward
Baltimore, 1958 - 1962
The Airborne Infections Research (AIR)
Facility
Witbank, Mpumalanga Province, SA
Ventilation ducts in patient rooms
Paddle Fans Assure Good Air Mixing
UVGI Fixture Specifications –
a test of UVGI, not specific fixtures
• Develop specifications
• Advertise bid opportunity in the US and
S. Africa
• Review all specifications submitted
• Solicit fixtures for laboratory
measurements
• Receive and install fixtures
A IR , E xp e rim e n ta l P la n
G uinea P ig A ir S am pling
G uinea P ig
TB RFLP
B
A
O dd days
E ven days
3 patient room s
U V G I or
oth er
intervention
P lus com m on areas
Intervention on/off on alternative day s
P t. T B
RFLP
UV1
Intervention
TST-1
0
TST-2
0
TST-3
0
TST-4
0
TOTAL
0
UV2
TST-1
3
TST-2
12
TST3
0
TOTAL*
15
*p<0.0005
Control
1
3
5
0
9
17
30
1
48
Hazard of becoming infected
Results
Combined hazard ratio 4.9 (CI.95: 2.8, 8.6)
Critical parameters determining
upper-room UVGI effectiveness
•
Good UV fixture design
– Flood the upper room with
enough 254 nm
– Low levels of UV in the
occupied space
•
Good air mixing between
the lower and upper room
– A slow paddle fan is essential
•
Good maintenance of
fixtures.
Not all fixtures being sold are
well-designed for their application
New Fogarty Application
under review – would start Sept, 2010
• $400K for 1 year, multi-institutional, multidisciplinary
• Collaborators:
– MIT D-lab, Penn State, U. Colorado, Boulder
• 2 domestic, 1 international fellows
• Engineers, architects, public health scientists, MDs
• Proposal: “Sustainable Air Disinfection Technology
Innovations for Resource Limited Settings”
– one year of intensive study and hands-on research further
developing through innovation and optimization the longneglected technology of upper room germicidal ultraviolet air
disinfection for use in poor settings.
Upper room UV light for the prevention of
airborne
tuberculosis transmission
R Escombe1,2, R Ramirez3, RH
Gilman2,4,6, M Navincopa5, E Ticona5, P
Sheen6, C Noakes7, B Mitchell8, D
Moore1,2, JS Friedland1, C Evans1,2,4,6.
1. Wellcome Centre for Clinical Tropical Medicine & Department of Infectious Diseases & Immunity, Imperial College
London, Hammersmith Hospital Campus, UK.
2. Asociación Benéfica PRISMA, Lima, Perú.
3. Facultad de Medicina Veterinaria, Universidad Nacional Mayor San Marcos.
4. Johns Hopkins University, Baltimore, USA.
5. Hospital Nacional Dos de Mayo, Lima, Perú.
6. Universidad Peruana Cayetano Heredia, Lima, Perú.
7. School of Civil Engineering, University of Leeds, UK
8. Agricultural Research Service, USA.
HOSPITAL
NACIONAL
“DOS DE MAYO”
Kaplan-Meier survival based on PPD conversions
1.0
UV group
.8
Cumulative Survival
Ionizer group
.6
No intervention
group
.4
.2
0.0
0
100
200
300
400
500
600
Days of study
UV vs. No intervention:
Ionizers vs. No intervention:
Log rank 46
Log rank 20
p<0.0001
p<0.0001
UVGI reduced TB: 72%
Ionisers reduced TB: 58%
UV killing and relative humidity
Relative humidity in Lima 50-95%
Upper room UVGI safety

TLV for UVGI (UV-C, 254 nm)

6.0 mJ/cm2 for 8 hr. exposure



Continuous staring for 8 hrs. at intensity of >
0.2 µW/cm2 will cause photokeratitis and skin
erythema
Will not cause lens cataracts or skin cancer
because of limited penetration
Compare to 240 µW/cm2 for 2 hrs peak
sunbathing outside.
Upper Air UVGI
10  >100 mW/cm2
ACGIH: 6.0 mJ/cm2 for
8-hour period
= 0.2 mW/cm2 for 8-hour
continuous eye-level
exposure
% Skin Penetration of UV
Depth,
mm
0
Skin
Layer
Surface
10
St. corneum 42
50
80
20
18
25
64
15
50
30
254 nm 297 nm 365 nm
UV-C UV-B UV-A
100 % 100 % 100 %
Viable layer 5
50
0.4
6
31
70
0.03
2
19
After Bruls, Photochem Photobiol,
1984
UVGI – poor design and installation
MDR TB Hospital – South Africa
Monitoring Human Exposures to Upper-Room Germicidal
Ultraviolet Irradiation
Melvin W. First, Robert A. Weker, Shojiro Yasui, and Edward A. Nardell
Journal of Occupational and Environmental Hygiene, 2005; 2: 285–292
Personal UVGI monitoring study
3 UV detectors:
Biofilm
Chemical
photometer
Electronic UV
meter
Prof. Melvin First
Study site: Hospital Room - UV Lamp
Old style fixtures
- intensity 10 x 0.2 µW/cm2
Unventilated bldg.
Drug-resistant TB
patients
No worker infections
Pt. B - always in bed
Patient B
11/30/01
1.50E-01
Peak
uW/cm2
Mean
uW/cm2
1.49E-01
5.69E-02
Irradiance (uW /cm 2)
1.20E-01
9.00E-02
6.00E-02
3.00E-02
0.00E+00
0
2000
4000
6000
8000
10000
Tim e (sec)
12000
14000
16000
18000
Summary - UV exposure
Peak
Mean
µW/cm2
µW/cm2
Pt. A
0.124
0.0454
Ratio Duration/ Exposure 8-hr exp. % TLV
P/M 8-hr
mJ/cm2
mJ/cm2 (6 mJ/cm2)
0.37 0.83
1.1
1.3
22
Pt. B
0.149
0.0569
0.38
0.54
0.9
1.6
27
Pt. C
0.190
0.0815
0.43
0.83
2.0
2.2
37
Pt. D
0.767
0.0356
0.05
0.87
0.9
1.0
17
Pt. D
0.847
0.0323
0.04
1.0
0.9
0.9
15
Pt. E
0.313
0.0282
0.09
0.75
0.6
0.8
13
R.N.
0.633
0.0141
0.02
0.60
0.2
0.4
7
Where are we going?

UVGI needs:




Better, more efficient fixture designs
Wider availability of expertise and UV
technology
CAD to plan UVGI installations
Greater understanding of human 254 nm
exposure
New Guidelines


NIOSH. Environmental Control of
Tuberculosis: Basic Upper-Room
Ultraviolet Germicidal Irradiation
Guidelines for Healthcare Settings . 2009
Recommendations:


30 – 50 µW/cm2 average irradiance
6.3 W total UVGI wattage per cubic meter
room volume treated
Conclusions:





UVGI is an old technology with new applications
Scientifically sound
Safe for room occupants
Clinical trials are difficult
Applications



Poor countries – TB, influenza
Rich countries – influenza, bioterrorism, future airborne
threats
Research: air sampling, personal UV monitoring, and
improved fixture designs underway.

CAD program under development