Witnessing Clean Agent System Acceptance Tests
... according to NFPA 2001 and 12A
This extract from the Retrotec Manual was designed to help AHJ witness door fan tests.
For an overview, get Retrotec’s “Environmentally Friendly” video. It discusses halon but
all the principals are the same. Special pricing is available to AHJ for software and
manuals. Contact us at [email protected] or call 360-738-9835 ext 308 for pricing.
Visit our Website at www.retrotec.com
Contents
About the Author.
Who discourages the discharge test ?
EPA
FSSA
NFPA 2001, Section 4-7
Who encourages the door fan test ?
What’s wrong with the discharge test?
Clean agent rooms must have tighter rooms than with halon. Less span between design and minimum extinguishing
concentrations. Doesn’t simulate fire event.
Why a door fan test is recommended.
Easily and inexpensively repeated. Requires more tightness-looks at leaks on all sides of the enclosure.
What is a door fan test?
A sophisticated diagnostic tool that pressurizes the enclosure to simulate a discharge test. Looks at many other variables.
How is the retention time calculated?
How to witness the door fan test.
Step by step guide. Easy to witness. How to section follows.
Small Room Testing...2500 cu.ft. or less
List of who accepts fan tests.
10’s of thousands performed worldwide.
3 K. Witnessing Tests - Page 1
About the Author
The author invites AHJ to contact him at [email protected] or call 360-738-9835 ext
308 to clear up any confusion created by the test or our explanations.
Mr. Genge participated in the first known door fan test of a halon protected enclosure, at
a military installation in Canada in 1988. Since then he has tested many hundreds of
gaseous agent protected enclosures.
Colin L. Genge was the first one to propose door fan testing for predicting retention
times. He wrote all of the Appendix B material and has been reposing for most of its
concurrent changes. He has trained the bulk of all Appendix B testers directly or through
trainers that he himself has trained.
Mr. Genge has been responsible for technical support to over 500 users of Retrotec
equipment in the field. He played a key role in the development of the most widely used
Door Fan Standard (CGSB CAN@ 149 10-M 85).
A partial list of clients includes:
- American Express
- Shearson Lehman Bros.
- Bethlehem Steel
- ABC Network Studies
- FAA (Federal Aviation Administration)
- Eli Lilley Co.
- GTE Telephones
- Canadian Pacific Railways
- Raytheon
- Anheuser Busch
- Glendale Federal
- U.S. Navy
- BC Hydro
- Los Alamos Labs
- Fletcher Challenge
- Canadian Airlines
Currently, Mr. Genge does door fan training courses;
familiarizes inspecting authorities with the procedure and how to ensure the test is
being correctly performed;
develops specialized test methodologies and
provides technical support to Retrotec’s door fan owners.
3 K. Witnessing Tests - Page 2
Who discourages the discharge test ?
The EPA has recommended that all discharge tests be eliminated.
The FSSA encourages all consulting engineers to specify and for all installers to perform door fan tests on every installation.
NFPA 2001 Section 4-7.2.2.10 states: “A discharge test is generally not recommended;”.
Who encourages the door fan test ?
IRI , FM, other insurers and 2001 state that the door fan test must be repeated annually if any extra holes have been made in
the enclosure.
quoted from
NFPA 2001
page 2001-37
“ 4-7.2.3* Review Enclosure Integrity. All total flooding systems shall have the enclosure
examined and tested to locate and than effectively seal any significant air leaks that could
result in a failure of the enclosure to hold the specified agent concentration level for the
specified holding period. The currently preferred method is using a blower door fan unit
and smoke pencil. If quantitative results are recorded, these could be useful for comparison
at future tests. ”
What’s wrong with the discharge test?
In the past the discharge test was often used as a quick way to test the enclosure. While the enclosure may have passed this
discharge test, the design may not have been adequate in a fire event where the hazard needs to be smoke tight on all sides.
The discharge test assumes static conditions that don’t occur in a fire event.
The discharge test only verified agent distribution in one location. Usually the most favorable. This may have led to
assuming that other approval steps could be overlooked.
The discharge test was never repeated. The room leakage would increase steadily, compromising the system from day one.
Now, clean agent rooms must be tighter than with halon. New agents all have less margin for error. Halon protected rooms
could lose over 50% and still maintain their extinguishing capabilities whereas the new agents can normally only lose about
20%.
Why a door fan test is recommended.
Quick. Can easily be repeated annually without disruption.
Inexpensive.
Locates room leakage problems such as faulty dampers.
Can identify intermittent problems that a one time discharge test would miss.
Shows overall hazard compartmentalization in all directions.
Conservative.
What is a door fan test?
The door fan itself merely measures the enclosure leakage area and the pressures that may exist across it. The computer on
board does the rest of the simulation and comes up with the prediction. The Retrotec software walks the user through all the
steps in a controlled way to ensure each step is done in accordance with NFPA. The few steps that need to be verified are
contained in the next section.
measuring room leakage areas
The Infiltrometer is temporarily installed in a doorway leading from the protected space (or test room) to a large open area or
outdoors. The fan speed is adjusted to obtain a pressure between the test room and the volume surrounding the test room. This
pressure (usually 10 to 15 Pa or 0.04" to 0.06" W.C.) is similar to the steady state pressure (column pressure) exerted by the agent
at floor level at the start of a typical 10 minute retention period.
3 K. Witnessing Tests - Page 3
The pressure created by the Infiltrometer causes air to move through leaks at a detectable rate. This makes it very easy to pinpoint
exactly where leakage occurs using a chemical smoke.
By measuring the air flow rate and the pressure
created, the computer calculates the Equivalent
Leakage Area (ELA), or the total area of all the
cracks, gaps, and holes in the test room. The flow
pressure across the blower inlet is converted into
flow by computer or manually.
Airflow
Airflow
Airflow
Airflow
The measurement is done by first blowing air out
of the room (depressurization) and then into the
room (pressurization). The two
readings are averaged to reduce
errors due to:
Airflow
Airflow
Airflow
1.
HVAC operation-major problem can
Airflow
produce 30% errors
2.
Other static pressures-WIND occasionally can
errors
if the room
gauge is to the same pressure exerted by the agent
Theproduce
door fan
pressurizes
the enclosure
fluctuating. 5Pa of fluctuation will produce aon30%
the error.
floor after discharge. The flow needed to create this pressure is used to
3.
Inaccurate gauge zeroing- 5 to 10 % possiblecalculate the leakage area of the enclosure.
4.
One-way leaks-almost never a factor
3 K. Witnessing Tests - Page 4
How the retention time is calculated.
The agent mixes violently upon discharge resulting in a homogeneous mixture. Pressures created in the first few seconds of
discharge (referred to as dynamic discharge pressure) are ignored in the retention time prediction model because they are so short
and because large factors for loss are already allowed for in the concentration formulae.
A small positive pressure is created by the heavier-than-air agent pressing down upon the floor. Flow develops whenever a hole
has a pressure difference across it. The greater the pressure and the larger the hole, the greater the agent lost. A small negative
pressure will develop at the top of the room that will allow a similar volume of air to flow back into the room from leaks at the
higher elevations.
If air moving equipment in the room is shut off at discharge, the agent mixture will tend to stay separate from the air infiltrating
through the upper leaks.
The intersection between the pool of agent mixture and clean air above is referred to as the agent/air interface. This is called the
descending interface case. This interface drops, as agent is lost out of the room through leaks in the floor and lower wall area.
Air from outside the room generally replaces agent by infiltrating through leaks in the upper half of the room.
If air moving equipment is left on during the retention period, the infiltrating air will become mixed in with the agent. This is
called the continual mixing case. The concentration at the floor will decay at the same rate as the concentration near the ceiling.
In some cases, air flow into or out of the room is created by other causes (e.g. damper or duct leakage). This air flow produces a
static pressure which pushes the agent out faster. This static pressure is therefore usually eliminated.
The door fan measures total leakage areas and static pressures. Below ceiling leaks can then be measured separately using a flex
duct or plastic on the ceiling to neutralize ceiling leaks. All other variables such as room volume and heights are easily measured
on site.
The model predicts how many minutes it will take for the descending interface to reach the minimum protected height specified
by the Authority Having Jurisdiction Or, for the concentration to fall to the minimum percent acceptable for the continual mixing
case.
… how door fan test
relates to
a discharge test
The minimum protected height for the descending
interface case
The less common continual mixing case
Taping is ok on ceiling for second test to give more
accurate prediction
usually chosen at the
same place where
Is equivalent to
the highest acceptance probe would be
The minimum allowable concentration
Taping does not accurately reflect test condition.
3 K. Witnessing Tests - Page 5
How to witness the door fan test.
This section is a comprehensive check-list showing each and every point that must be inspected to complete an
approval according to NFPA 2001.
Technician training and experience- should have certificate from course attendance. We will be listing all
testers on our Website at www.retrotec.com after they have completed a test to show they understand the basics.
4-6.1 All persons who might be expected to inspect, test, maintain, or operate fire extinguishing systems
shall be thoroughly trained and kept thoroughly trained in the functions they are expected to perform.
Software conformance to NFPA 12A & 2001. Retrotec version HA5 is up to date for all but the inert gases that
need HA5.1. The newest version HA6 has very minor adjustments to density and has a few new agents added but
essentially gives the same result as HA5.1. The example in NFPA 12A can be run to see if the same result is
achieved but is very tedious. This would be an option for software the AHJ was unfamiliar with. In general, the
Retrotec software conforms to NFPA exactly.
Room pressure gauge calibration certificate (less than one year old)
This is required by the NFPA test although we have found that unless the gauge has been damaged or moves
unevenly that they are usually within 10%.
System calibration - must be completed every five years as stated on a dated 5 year certificate.
Field Calibration check procedure- can be requested to see if the equipment can measure the correct leakage
area. It doesn’t test the software.
Room set-up
There must be a complete flow path back to Infiltrometer otherwise all leaks may not be measured. All doors in
the zone must be open. The HVAC system and all dampers must be in position they’d be in at discharge.
Door Fan Equipment set-up
A doorway must be selected that opens into the largest most open space. Door Fan panel taping is OK as long as
the doorway is tighter than the panels. Gauge leveling and zeroing must be completed before any tubes are
hooked up. Range choice extremely critical for good results. The range on the printout must match the range
diagram shown in this section. Gauges must be tapped prior to each reading.
Common Needed Clarifications
Inert gases are all heavier than air and will according to 2001 run out of room leaks. They are not as heavy as
halocarbons and usually only run out at half the rate. All inert clean agents need relief areas according to their
manufacturers.
Covering ceiling level leaks to measure leaks in the lower part of the room is not the same as taping up leaks to
pass a discharge test. The leakage of the ceiling has already been measured in the previous test and now the lower
leaks can be measured separately to get a more accurate prediction.
All enclosures must be re-tested yearly if any doubt exists as to whether the room has had any more holes put in it
in the last year.
Rooms must be tested positively and negatively to eliminate bias due to duct leaks not because of positive
pressures after discharge.
Use with Retrotec Software Version HA6 or higher and Manual #6.
Follow the numbers on the Retrotec form or on computer screen.
3 K. Witnessing Tests - Page 6
form #
Screen #
Questions on forms
or computer screen
what to look for
usually inspect tags to get weight to agent. Increasing agent
weight decreases the retention time most often unless the agent is
continually mixed after discharge where more agent greatly
increases retention time.
Within 1000 ft. is OK. Used to re-calculate the design
concentration but does not affect retention time.
1.1
16
Actual Agent Weight in
Cylinder(s):
1.2
14
Elevation Above Sea Level
(within 1000 ft. __ corrects
initial conc.):
1.3
17
Normal Room Temperature
(corrects initial conc.):
Within 10F is OK. Used to re-calculate the design concentration
but does not affect retention time.
1.6
21
Net Agent Protected Volume:
1.7
24
Calculated Initial
Concentration (from Retrotec
software):
Used to re-calculate the design concentration. Must be remeasured. Retention time increase linearly with volume i.e. 10%
volume increase gives 10% retention time increase.
The Retrotec will re-calculate the design concentration based on
all the above 4 inputs. This result must be acceptable.
1.9
30
Minimum Required Retention
Time (e.g. 10 minutes):
1.10
25
Maximum Agent Height from
Slab:
1.11
32
Will Major Air Circulating Fans
Shut Down On Discharge?
 YES 
NO
 YES if air handlers are shut off, the agent will form a
descending interface. Go to 1.14
If NO, what is Minimum
Specified Final Concentration
(e.g. 5%)?
%
What is the minimum extinguishing concentration the agent can
fall to by the end of the retention period? For FM200 this could
be 5.5%, 31% for INERGEN etc.
1.11
29
Often 10 minutes but could be longer for remote sites. Must be
long enough to allow for trained personnel to arrive. See section
on small rooms. 10 minutes could be required for deep seated
fires.
must be re-measured from floor slab to highest combustible.
 NO if air handlers will keep running, the agent will stay
mixed over the retention time. Go to 1.11
Skip to 3.2
Height measured from the slab that the agent can fall after the
retention period. The higher the height, the longer the retention
time. This is a very critical value. e.g. in a 10’ room changing
this height from 7.5’ to 8’ will about halve the retention time.
Typically minimum protected heights near the ceiling can be
nearly impossible to pass and continual mixing may have to be
considered.
 Whole Room: is always done first unless its a subfloor only
test.
 BCLA Flex Duct and  BCLA Poly Under Ceiling:
measures leaks below the ceiling only so any leak going through
the ceiling can be taped off since it has already been measured at
the Whole Room stage.
The pressure that would be across the enclosure at discharge
must be predicted. If a pressure existed on the day of the test but
the building HVAC would be shut down at actual discharge,
then use zero. If the HVAC is designed to stay on, then use the
measured pressure. If the subfloor A/C will stay on at discharge,
measure the pressure between the subfloor and outside the room.
will only affect the result by about 1% per 20F of temperature
difference
1.14
28
Minimum Protected Height is:
 75% of
Max. Height
OR

Highest Combustible
3.2
56
Type of test:  Whole Room
 BCLA Flex Duct  BCLA
Poly Under Ceiling
3.7
33
Record the static pressure just
prior to agent discharge.
3.9
57
Temperature in zone and
return path space at time of
fan test (within +10oF):
61
is used to correct the readings for room leakage area and should
Static pressure during the
not change the result unless more than 3Pa.
door fan test
3 K. Witnessing Tests - Page 7
3.10
3.12
door fan test.
4.5
65
Room pressure
4.5
68
Range
4.5
74
4.5
77
Flow Pressure(s)
Depressurization forcing air
out of zone
Pressurization Test forcing
air into the zone
4.7
79
Measured Whole Room ELA:
4.7
80
Max. Allowable ELA to Pass:
4.7
79
Assumed BCLA (1/2 Total
ELA):
5.12
79
Measured BCLA:
5.12
81
Max. Allowable BCLA to Pass:
5.12
82
Interface Height or % at Spec.
Time:
The computer prompts the operator to take the room up to a
range of pressures; often 10 to 13 Pa. The room pressure is
increased by the door fan blower. The room pressure is read
from the left hand gauge on the console or upper gauge on the
gauge clip. The pressure reading must be verified within 1 Pa
This is critical. Ensure that the door fan blower is set up
according to the range sheet that follows. The correct range
must be entered. Each range will change the result by over 50%.
Must be read at the same time as the flow pressure from the
The test must normally be done both ways to compensate for
stray static pressures in the room usually caused by duct leaks.
If the door ways can be tested with the smoke puffer to see that
there is no smoke movement prior to the door fan test, then the
requirement to test both ways could be waived. When doing the
plastic on the ceiling test, it is usually not possible to test both
ways. To get a fairly accurate estimate of the true leakage when
testing only one way, follow the Retrotec software procedure for
an estimate starting on screen 77.
The total Equivalent Leakage Area (ELA) of all room leaks;
floor, walls and ceiling
The total ELA that will just allow the enclosure to pass the
acceptance criteria
The Below Ceiling Leakage Area (BCLA) is the total of all wall
and floor leaks. The first prediction is base on assuming that
half the (ELA) leaks are located in the worst place for leakage;
the floor.
If the worst case assumption causes the enclosure to fail, the next
step is to make an actual measurement or is in some cases
estimate ( see small room section ) of the actual BCLA to get a
longer and more accurate prediction of retention time. Taping
of ceiling registers and leaks is OK because the above ceiling
leaks have already been measured in the previous test. Putting
plastic on the ceiling is an acceptable way to measure below
ceiling leaks in small rooms.
The total BCLA that will just allow the enclosure to pass the
acceptance criteria
The height the interface will fall to during the specified retention
time (usually 10 minutes).
The % the concentration will fall to during the specified
retention time (usually 10 minutes).
More detail on how variables affect retention time and to what degree. The following table shows how retention time is affected.
Assumed all other variables are held constant.
input
effect on retention time
magnitude
room volume
doubles and retention time doubles
linear
room leakage
doubles and retention time halves
inverse linear
agent weight
quadruples and retention time halves
inverse sq. root
minimum
protected increasing from 75 to 85% of room height will halve retention time
inverse sq. root
height
agent height
doubles and retention time more than doubles assuming minimum somewhat linear
protected height stays same.
room test pressure
testing at 13 instead of 10Pa may increase apparent leakage area by not much
5%
Static @ Fan Test
should not affect leakage area measurement but may
none
Elevation
only affects concentration, loss based on weight of agent
none
Discharge Temp.
only affects concentration, loss based on weight of agent
none
Setup conditions and how they effect retention time.
3 K. Witnessing Tests - Page 8
inadequate return path
static pressures
equipment calibration
usually not more than a few percent. Check the pressure across
each wall.
small differences for small pressures. May require qualified
engineering judgment above 3Pa.
seldom more than 10% of result unless gauges stick. Can be
checked with field calibration check.
3 K. Witnessing Tests - Page 9
Flow Ranges... must appear exactly as shown
Range 18F - for standard Range 18R - for
Range 9 - for slightly tighter
depressurization test
pressurization test
rooms
Range 5
Range 1.3
Range 3
Range 1.2
Range 1.4
Range 1.1
Range 0.1
Note: 18 refers to 18" diameter inlet, F to forward blowing (i.e. away from the operator), R to reverse (or towards the
operator).
Clear flow pressure tube must be in port labeled 18, Plate Off.
9 and all other whole numbers refer to the number of 4" holes open, including the middle hole. All decimals refer to the
number of 2" holes open, e.g. ".3" of 1.3 refers to 3 x 2" holes open plus of course the 1 x 4" hole in the middle (motor
cooling).
Clear flow pressure tube must be in port labeled 9, Plate On.
3 K. Witnessing Tests - Page 10
hp or DOS Printout
============================================================
RETROTEC DISCHARGE SIMULATOR VER. HA6.1.1
RETENTION TIME PREDICTION MODEL
-------1
============================================================
Location:
Room Name:
Testing Company:
Technician:
Date:
Test #:
YOUR BEST CLIENT
SAMPLE
YOUR CO. NAME
YOUR NAME
DATE
TEST 2 10am
Units:
BCLA Flex Duct Test.
-------2
IMPERIAL Units (with Pascals) -------3
Gas Being Modeled:
Lbs./Kgs. of Agent In Cylinder(s):
Net Room Volume (ft3):
Room Height (ft):
Minimum Protected Height (ft):
Minimum Retention Time (min.):
Initial Gas Concentration (%):
Static Pressure @ Discharge:
1301
650.00
25340.00
9.00
7.00
10.00
6.27
-1.00
-------4
-------5
-------6
-------7
-------8
-------9
-------10
-------11
============================================================
Whole Room Leakage (ELA in2):
Hole in Ceiling (in2):
Hole below Ceiling (BCLA in2):
950.4 -------12
859.12 -------13
91.41 -------14
This Room PASSES the Test as the Predicted Retention Time is---15
23.5 minutes for the agent/air interface to drop below the
minimum protected height .
Witnessed By:
ABC INSURANCE -------16
X ____________________________________
Conforms To 1992 NFPA 12A and 1996 NFPA 2001 Acceptance Procedure. -----17
-------18
Interface Height @ 10 Minutes:
8.12 -------20
===========================================================
PAGE 1 OF 2
Licensed To: Your Company Appears Here
Registration #:100
3 K. Witnessing Tests - Page 11
============================================================
FAN TEST READINGS & DATA
============================================================
Location:
Room Name:
Test #:
YOUR BEST CLIENT
SAMPLE
TEST 2 10am
Temperature IN: 70
Temperature OUT: 70
-------21
Static Pressure @ Fan Test:
0
-------22
23-------- DEPRESSURE
Operator and Gauges Location
Room Pressure Gauge Reading
Corrected Room Delta P(pa)
Blower Range Config Used
Flow Pressure Gauge Reading
Corrected Flow Pressure
Calculated Air Flow (cfm)
Temp. Corrected Air Flow
Leakage Area (in2)
Average Leakage Area (in2)
IN
11.0
-11.0
1.4
175.0
175.0
339.4
339.4
95.1
91.4
PRESSURE -------24
IN
10.0
10.0
3.0
52.0
42.0
298.2
298.2
87.7
-------34
-------25
-------26
-------27
-------28
-------29
-------30
-------31
-------32
-------33
============================================================
RM=
1.520
AT=
0.374
C3=
4.072
GD=
6.283
AR= 261.565
PAGE 2 OF 2
--35
--37
--40
--43
--45
PC=
8.6
ALL=
0.036
C4=
0.000
K1=
2.2062
T= 1409.314
--36
--38
--41
--46
PA=
FA=
CF=
K2=
EL=
10.0
0.096 --39
0.982 --42
0.0050
500.00 --47 TD= 70.0--44
Licensed To: RETROTEC INC
Registration #:100
3 K. Witnessing Tests - Page 12
Interpreting the Printout
Refer to the accompanying sample printout.
1.
2.
18. Any deviations to the Procedure will be printed out here
(See next section).
Should be version HA6 or higher.
This line indicates the type of test conducted.
options are:
BCLA Flex Duct Test,
BCLA Test Using Plastic,
Subfloor Only Test,
Analysis of Manually Entered ELA/BCLA's.
Other
19. The maximum size of ELA permitted to result in a
retention time equal to item 9. If the test was a BCLA test,
and it failed, this value would be the maximum allowable
BCLA.
3.
Indicates which unit of measure were used. Note that
Pascals are used for pressure measurement even in the
imperial version. 249 pa = 1" H2O.
20. Where the descending interface would be at the time
specified in item 9. If mechanical mixing takes place (0 to
Screen 27), the average concentration throughout the room
at the specified time will be displayed.
4.
Agent used in the analysis.
21. The temperatures entered at Screens 57 and 58.
5.
Agent weight entered at Screen 16.
22. The static pressure entered at Screen 61.
6.
Net enclosure volume calculated from the inputs
Screens 20/21 and 22.
7.
Room height as entered at Screen 25/26.
8.
Minimum protected height entered at Screen 28. If
mechanical mixing will take place (0 entered at Screen 27)
the Minimum Specified concentration will be displayed
here (usually about 75% of the initial concentration).
9.
to
23. This column contains the Depressurization fan test data.
24. This column contains the Pressurization fan test data.
25. Where the gauges were located during each set of
readings.
Minimum retention time in minutes entered at Screen 30.
10. Calculated initial concentration displayed at Screen 24.
26. The room pressure gauge reading entered (from the 0-60
pa gauge).
27. The actual difference in room pressure obtained, after
subtracting the static pressure.
28. The fan calibration range (18,9,5,3,1.4,1.3,1.2,1.1,0.1,36
or 99)
11. Static pressure entered at Screen 33/34.
12. The Whole Room Leakage comes from the Whole Room
Test and includes all ceiling, wall and floor leaks. If this is
a BCLA test (line 2) then this value comes from a
previous Whole Room Test.
13. The difference between item 12 and item
14.
Note that the model always simplifies
the location of
room leakage to be one
hole in the ceiling and one in the
floor, even though a visual inspection may uncover
no
leakage in either location.
29. The flow pressure read off the 0-250 pa gauge. It equals 0
if CFM were entered directly by using range 99, or if
multiple blowers were used.
30. If the fan was blowing towards the operator (e.g. when
pressurizing the room with the pressure gauges inside the
room), the flow pressure value item 29 is reduced
internally by the room pressure gauge reading in item 26.
When 2 fans have been used (range 36), this is a
composite value of the 2 flow pressures entered.
31. This is the non-corrected air flow in CFM.
14. The BCLA (Below Ceiling Leakage Area).
32. The temperature corrected air flow.
15. If the retention time is greater than the value
for
Minimum Retention Time at item
9, this paragraph
reads "PASS". This room either needs more sealing or a
more accurate assessment of the BCLA using a
Ceiling Leakage Neutralization test.
33. These are the ELA's measured in each direction. These
values would be BCLAs if this were a Below Ceiling test.
16. The Authority Having Jurisdiction (AHJ)
can sign
here to indicate that he or she witnessed the test.
17. Statement of conformance to the 12A
Procedure.
34. The average of the two leakage areas in item 33. If the
test was only conducted in one direction, the value here is
the result of the correction factor applied per Section
10.5.
Items 35 to 47 are printed out so that an inspecting
authority can check the calculations manually using the
3 K. Witnessing Tests - Page 13
NFPA Procedure. The acronyms used are those used in
12A.
35. RM = mixture density in kilograms per cubic meter.
36. The first PC value (8.6 pa) is the calculated column
pressure, the second value (10.000 pa) is what was used
for the minimum test pressure as 10 pa is always used as
the minimum.
37. AT = ELA * 0.61 in square meters.
38. ALL = BCLA * 0.61 in square meters.
39. FA = Lower Leak Fraction. Equals .5 in worst case.
Typically is actually found to be less than .25 when ceiling
leakage is neutralized in an enclosure having a suspended
ceiling.
40/41. C3 and C4 are constants for equation
simplification.
42. CF is the altitude correction factor which is applied when
determining the concentration.
43. The gas density in kgs/m3.
44. The normal room temperature entered at Screen 17.
45. AR is the effective floor area in square meters.
46. T is the retention time in seconds.
47. The elevation above sea level entered at Screen 14.
3 K. Witnessing Tests - Page 14
Small Room Testing...2500 cu.ft. or less
Some compromises can be made is small rooms for two reasons:
They are not likely to have a large fire that would threaten the rest of the building.
Once trained personnel arrive and open the door, the enclosure integrity (and some of the agent) is lost anyway so
shorter retention times can be considered.
1) Selection of a retention time.
Select an appropriate retention time for the specific enclosure. NFPA 2001 states “... the design concentration ... shall be
maintained for a sufficient period of time to allow effective emergency action by trained personnel”. Yet most
specifications state it must be 10 minutes. Why? I would suggest the following guidelines for rooms that do not have the
possibility of a deep seated fire. Determine how long it would take for personnel to arrive then use this as the specified
retention time. As soon as they enter a small room, the integrity is compromised anyway.
My suggestion is for room volumes(cu. ft.) of:
That, the minimum retention time (min.) be :
provided that it is reasonable to expect that
“trained personnel could show up in that time.
The maximum leakage area allowed
if the time was 10 minutes is
If the retention time was reduced as above
the leakage areas could be a more reasonable
2500
8
1250
6
625
4
350
3
50
25
12.5
7 (sq.in.) respectively
62
42
32
23 (sq.in.) respectively.
This last row is more in keeping with how tight rooms can be made as they get smaller. For example each room regardless of
size must have a door and door usually leak about 5 to 20 sq.in. depending on how well they are weather-stripped. To try to
achieve a total of 7 sq.in. in this room is not really practical.
2) Test the room to get the
most accurate prediction of retention time.
When we were in the habit of discharge testing small rooms with Halon 1301, it was common for small rooms to fail that
test. Now that we commonly test small rooms using the door fan and Appendix B, these rooms similarly fail this test.
The results are similar between the door fan test and the discharge test except in one instance - if the upper half of the
room leaks a lot more than the lower half, then, the Appendix B test will be overly conservative and may fail a room that a
discharge test would pass.
Reason; On the first test, the total leakage of the whole room is measured. The assumption is then made that half the
leaks are in the floor and the other half in the ceiling. This usually gives a very conservative (shortest retention time) result
because most of the leaks are usually above the ceiling where the agent will not leak out. If the ceiling leaks can be measured
or calculated the actual hole in the floor can be used to make the retention prediction. Where most of the leaks are in the
ceiling this prediction will be much longer.
There are several solutions to this measurement problem:
• perform double fan ceiling neutralization test on rooms with a T-bar ceiling. This test will separate the below
ceiling leaks from the above by allowing one fan to depressurize the above ceiling space while the other
depressurizes the below ceiling space. The fan speeds are altered until smoke at the T-bar ceiling neither rises nor
falls indicating a balanced condition.
• cover the ceiling with plastic and re-test to measure below ceiling leaks alone. Use this test where the above ceiling
space cannot be pressurized or where two fans and flex duct are not available. This is time consuming in all but the
smallest rooms
• do an audit of the leaks and determine from this what the lower leaks are compared to the above ceiling leaks.
Section 4-7.2.3 of NFPA 2001 in the 1996 version, allows for the use of a smoke pencil and blower door fan unit
only and section B-1.2.2.5 allows for the technical judgment in assessing upper Vs lower leaks. This is the basis for
the following spreadsheet method of quantifying this ratio. Use this table as an example (from BCLA%.XLS): an
example follows
3 K. Witnessing Tests - Page 15
Spreadsheet for calculating BCLA based on a visual leak audit
est. upper leaks
sq. in. est. % full speed corrected est. leak
scaled down to
wall to ceiling joint
122
50%
61
49.5
upper slab penetrations
60
100%
60
48.7
upward duct openings
140
25%
35
28.4
upper open conduits
10
10%
1
0.8
total upper leaks
157
127.4
est. lower leaks
wall to lower slab joint
lower slab penetrations
lower open conduits
downward duct openings
door leaks
window leaks
stub wall leaks
total lower leaks
25
5
0
0
10
0
50
10%
100%
100%
100%
50%
100%
75%
2.5
5
0
0
5
0
37.5
50
est. total lower and upper
actual lower and upper ELA
factor
The calculated BCLA =
2.0
4.1
0.0
0.0
4.1
0.0
30.4
40.6
207
168
1.23
40.6 sq. in.
which is
3 K. Witnessing Tests - Page 16
24% of the ELA
List of who accepts fan tests.
Version 2.0, September 1990
This partial list shows some of the many who have accepted door fan testing in lieu of discharge testing.
Fire Marshals
Brown University
University of Mass.
State of VT
City of Danvers
City of Wilmington
City of Canton
City of Addison
North Carolina Fire Marshal
Savannah, GA Fire Marshal
Denver Fire Prev. Bureau
Boulder Fire Dept.
Ft. Collins Fire Dept.
Westminster Fire Dept.
Castlewood Fire Dept.
Colo. Springs Fire Dept.
Arvada Fire Dept.
Thornton Fire Dept.
Anchorage Fire Marshal
Fire Sub Code Official
Fire Sub Code Official
Fire Sub Code Official
Fire Sub Code Official
NJ
Fire Sub Code Official
Fire Sub Code Official
Coral Springs Fire Dept.
Pompano Beach Fire Dept.
FL
Cincinnati Fire Dept.
Omaha Fire Div.
Lincoln Fire Prevention
Nebraska State Fire Marshal
Phoenix Dept. of Fire Prev.
City of San Diego Fire Dept.
North Chicago Fire Dept.
Waukegan Fire Dept.
Highland Park Fire Dept.
Countryside Fire Protection
Wauconda Fire Dept.
Lake County Fire Insp.
Ogden City Fire Dept.
Orem City Fire Dept.
Provo City Fire Dept.
City of Phoenix Fire Dept.
Memphis Fire Dept.
Ventura County Fire Dept.
Long Beach City Fire Dept.
Newport Beach Fire Dept.
CA
Torrance Fire Dept.
Los Angeles City Fire Dept.
Providence, RI
Boston, MA
Montpilier, VT
Danvers, MA
Wilmington, MA
Canton, MA
Addison, TX
Raleigh, NC
Savannah, GA
Denver, CO
Boulder, CO
Ft. Collins, CO
Westminster, CO
Englewood, CO
Colo. Springs, CO
Arvada, CO
Thornton, CO
Anchorage, AK
Woodcliff Lake, NJ
Newark, DE
Edison T.W.P., NJ
South Brunswick,
Somerset, NJ
Morris T.W.P., NJ
Coral Springs, FL
Pompano
Beach,
Cincinnati, OH
Omaha, NE
Lincoln, NE
Nebraska
Phoenix, AZ
San Diego, CA
North Chicago, IL
Waukegan, IL
Highland Park, IL
Dist. Mundelein, IL
Wauconda, IL
Lake County, IL
Ogden, UT
Orem, UT
Provo, UT
Phoenix, AZ
Memphis, TN
Camarillo, CA
Long Beach, CA
Newport Beach,
Torrance, CA
Los Angeles, CA
Skokie Fire Dept.
Fort Worth Fire Dept.
Burbank Fire Dept.
Imperial Valley Fire Dept.
CA
Albuquerque Fire Dept.
Ithaca Fire Dept.
Fort Worth Fire Dept.
Canadian Fire Commissioners Office
Aurora Fire Inspector
Valparaiso Fire Dept.
Canton Fire Dept.
Danvers Fire Dept.
Concord Fire Dept.
Plymouth Fire Dept.
Hyannis Fire Dept.
Lawrence Fire Dept.
City of Milwaukee
City of Wauwatosa, Fire Inspector
Youngstown Fire Dept.
Pittsburgh Fire Dept.
City of Lakehead Fire Dept.
City of Troy
City of Royal Oak
City of Pontiac
City of Auborn Hills
City of Billings
Clark County Fire Dept.
City of Las Vegas Fire Dept.
New York City Fire Dept.
Skokie, IL
Fort Worth, TX
Burbank, CA
Imperial
Valley,
Albuquerque, NM
Ithaca, NY
Fort Worth, TX
Nationwide
Aurora, IL
Valparaiso, IN
Canton, MA
Danvers, MA
Concord, NH
Plymouth, NH
Hyannis, MA
Lawrence, MA
Milwaukee, WI
Wauwatosa, WI
Youngstown, OH
Pittsburgh, PA
Lakehead, FL
Troy, MI
Royal Oak, MI
Pontiac, MI
Auborn Hills, MI
Billings, MT
Las Vegas, NV
Las Vegas, NV
New York, NY
Insurance Groups and Engineers
Industrial Risk Insurers
Factory Mutual Engineering
Kemper Insurance
Chubb Insurance
American Risk Management
Alexander & Alexander
N.C. Dept, of Insurance
CUH2A
Comp U Site
Mead Loss
Reed Stenhouse
Sedgewick James
Rolf Jensen & Assoc.
Cohos Evamy
Nationwide
Nationwide
Nationwide
Nationwide
Worldwide
Boston, MA
Raleigh, NC
Princeton, NJ
Montvale, NJ
Dayton, OH
London/Toronto, ON
Nationwide
Chicago, IL
Edmonton, ALTA
Government Halon End Users
EG & G (Rocky Flats Site)
Bureau of Reclamation
3 K. Witnessing Tests - Page 17
Golden, CO
Denver, CO
EG & G Florida Inc (NASA)
Center
EG & G Idaho Inc
Dept. of Energy
U.S. Postal Service
Federal Aeronautics Admin.
Los Alamos National Labs
Telesat Canada
Illinois Dept. Nuclear Safety
State of VT
Omaha Corp. of Engineers
MSE
Kennedy
Space
Idaho Falls, ID
Idaho Falls, ID
Palatine, IL
Various sites
Los Alamos, NM
Ottawa, ON
Peoria, IL
Montpelier, VT
Omaha, NEB
Butte, MT
Military Departments
US Army Corps of Engineers
Corps of Engineers
Griffiss AFB
Kirtland Airforce Base
U.S. Navy
Can. Dept. Of National Defense
CT National Guard
US Air Force
Phoenix, AZ
Whiteman AFB
Rome, NY
Albuquerque, NM
San Diego, CA
Canada Nationwide
Hartford, CT
Knobnoster, MO
Private Sector Halon End Users
G.T.E. Sprint
Brown University
University of Massachussetts
Digital Equipment Corp.
Citgo
Indell Davis
Conoco
Cherry Creek School District #5
Memorial Hospital
Pioneer Tele-Technology
ATC
O'Neal Steel
Excel Corp.
USX (Steel Mill)
Public Service Co. of Colorado
CIGNA Insurance
MCI
Adolph Coors Company
Roche Biomedical
Rockwell International
Burroughs Wellcome
Telecom USA
Centel of NC
Contel of VA
Cellular One
ARMCO Inc
Ashland Oil
South Central Bell
Contel Cellular
Cincinnati Gas & Electric
Procter & Gamble
Cargill Inc
Union Pacific Railroad
Nebraska Meth. Hospital
Nationwide
Providence, RI
Boston, MA
Nationwide
Tulsa, OK
Tulsa, OK
Ponca City, OK
Englewood, CO
Co. Spgs., CO
Springfield, MO
Springfield, MO
Birmingham, AL
Dodge City, KS
Birmingham, AL
Denver, CO
Thornton, CO
Nationwide
Golden, CO
Burlington, NC
Laurinburg, NC
Greenville, NC
Raleigh, NC
Hickory, NC
Mechanicsville, VA
Florida
Middletown, OH
Ashland, KY
Atlanta, GA
Atlanta, GA
Cincinnati, OH
Cincinnati, OH
Dayton, OH
Omaha, NE
Omaha, NE
General Motors
Robin Hood Mills
Ontario Hydro
Rogers Cable Television
P.P.G. Duplate
Royal Bank of Canada
B.F.Goodrich
Mack Trucks
Metro Toronto Police
ESSO/EXXON
American Cyanamid
Hoechst Celanese
Rohr Industries
McDonnell Douglas
College of Osteopathic Medicine
UOP Research Center
Kemper Group
Milwaukee Sanitary Dist.
University of Wisconsin
Reliance Electric
Oshawa, ON
Rexdale, ON
Provincewide, ON
Toronto, ON
Oshawa, ON
Toronto, ON
Niagara Falls, ON
Oakville, ON
Toronto, ON
Sarnia, ON
North America
North America
San Diego, CA
San Diego, CA
Downers Grove, IL
Des Plaines, IL
Long Grove, IL
Milwaukee, WI
Madison, WI
Milwaukee, WI
End Users (cont'd)
Allen Bradley
Sears Roebuck
N. Chicago Refiners
Cimlinc
United Airlines
Machinery Systems Inc.
Fel Pro
FSC Paper
Telerate
Raytheon
Anhueser-Busch
Glendale Federal Bank
Merabank
Centel
Rockwell International
East Mesa Geothermal
Fletcher Challenge Canada
Canadian Airlines
ABC Studios
Canadian Pacific Railways
Eli Lily & Co.
Bethlehem Steel
Amdahl Computers
Niagara Mohawk Power
Hartwick College
Baseball Hall of Fame
Fonorola
Goodyear Corp.
Motorola Corp.
Cornell University
Alcan Aluminum Ltd.
Metropolitan Life
Rochester Institute of Tech.
Lawyers Publishing Coop
York County Resource Recovery
Riverside Hospital
3 K. Witnessing Tests - Page 18
Milwaukee, WI
Skokie, IL
N. Chicago, IL
Itasca, IL
Chicago, IL
Schaumberg, IL
Skokie, IL
Alsip, IL
Chicago, IL
Los Angeles, CA
Los Angeles, CA
Burbank, CA
Phoenix, AZ
Nationwide
Los Angeles, CA
East Mesa, CA
Crofton, BC
Vancouver, BC
New York, NY
Nationwide
Indianapolis, IN
Chesterton, IN
Sunnyvale, CA
New York State
Oneota, NY
Cooperstown, NY
Albany, NY
Niagara Falls, NY
Elma, NY
Ithaca, NY
Montreal, Quebec
Utica, NY
Rochester, NY
Rochester, NY
York, PA
Ottawa, Ontario
Walt Disney Inc.
Greenwich Hospital
Pratt and Whitney Canada
Hydro Quebec
Cantel Cellular
Bell Cellular
Recruit USA
Vermont Yankee Nuclear
Global Communications
Barnett Bank
Schneider Freight
Steelcase
Chateau Laurier Hotel
ABC Television
Northern Telecom
Rohm and Haas
Atwood Corp.
Christian Science Publ.
Wang Labs
Boston Gas
Central Maine Power
ADVO
Marshalls Dept. Stores
Motorola
Reuters News Service
National Semi Conductor
Addison Hospital
Boston Five Bank
Grossman's Dept. Stores
Univ. of Connecticut
Hills Dept. Stores
Uniroyal
Orlando, FL
Greenwich, CT
Halifax, NS
Montreal, Quebec
Canada
Canada
Jersey City, NJ
Vernon, VT
Toronto, Ontario
Jacksonville, FL
Green Bay, WI
Grand Rapids, MI
Ottawa, Ontario
New York, NY
Worldwide
Philadelphia, PA
Grand Rapids, MI
Boston, MA
Peabody, MA
Boston, MA
Augusta, ME
Hartford, CT
Andover, MA
Cambridge, MA
Smithtown, NY
Portland, ME
Gloucester, MA
Quincy, MA
Braintree, MA
Waterbury, CT
Canton, MA
Southboro, CT
Wilmington Water Dept.
Centel of VA
General Public Utilities
Clean Harbors
Andover Insurance
Microcom
Esqn Beckford
Mitre
Gibco
Marsulex
Big Y Stores
Southern NE Telephone
Putnam
Appex Telephone
SWIFT
Baker Shoes
U.S.X. Fire Prot. Division
Quantum Chemicals
Steinbrenner & Assoc.
Irving Tissue Mill
Raytheon Co.
IBM Corp.
International Paper
CH2M Hill
M.M.S.D
Diamond Star Motors
PPG
Conrail
The Hilman Co.
Bell Atlantic Mobile Systemss
End Users (cont'd)
Armco
Ohio Water Service
Ramada Inns
Chrysler Corp.
General Motors
Nissan Motor
Palonar Medical Ctr.
Rohr Industries
General Dynamics
Lockheed
Fluor-Daniel
Soltex-Polymer
Granite School Dist.
Utah County Geneva Steel
Salt Lake City Corp.
G.D. Searle Co.
Nova Gas
Alberta Govt. Telephone
Chrysler
Chubb EDP
Butler, PA
Boardman, OH
Phoenix, AZ
Detroit, WI
Detroit, WI
Nashville, TN
Escondido, CA
Chulla Vista, CA
San Diego, CA
Houston, TX
Houston, TX
Salt Lake City,UT
Vineyard, UT
Salt Lake City, UT
Skokie, IL
Calgary, ALTA
Edmonton, ALTA
Brampton, ON
Springfield,N
3 K. Witnessing Tests - Page 19
Wilmington, MA
Charlottesville, VA
Reading, PA
Quincy, MA
Andover, MA
Danbury, CT
Simsbury, CT
Norwood, MA
Grand Isle, NY
Norwalk, CT
Springfield, MA
New Haven, CT
Quincy, MA
Walthaven, MA
New York, NY
Milton, MA
Gary, IN
Morris, IL
Edmonton, ALTA
Saint John, NB
Bedford, MA
Essex Jct, VT
Jay, Maine
Milwaukee, WI
Milwaukee, WI
Normal, Ill
Pittsburgh, PA
Pittsburgh, PA
Pittsburgh, PA
Pittsburgh, PA
C:\Inetpub\wwwroot\RetroWeb3\Documents\Fire Witness.doc
Page 20 of 20
Friday, February 01, 2002