Foreword
Compact mobile shelving systems are commonly found in healthcare, office,
library, and numerous other occupancies. The advantage to the use of these
systems is that aisles are eliminated, with the exception of one moveable aisle, in
most cases saving up to 50% of floor area.
The 2007 edition (current) of NFPA 13 does not provide specific design criteria
for compact storage arrangements. As a result, design engineers are required to
utilize protection criteria for either rack storage (storage depth 30 inches or
deeper) or shelf storage (storage depth less than 30 inches), depending on the
dimensions of the compact storage array. This is problematic because the fire
protection design challenges presented by compact mobile shelving units are
fundamentally different than those presented by stationary shelf and/or rack
storage.
In 2007, the Foundation conducted a review of available fire test data for these
systems, identified gaps, and made recommendations for future testing to
develop rational fire protection criteria for this storage system. This report
presents the result of that test program: a filing survey, bench scale and two
large scale fire tests were conducted.
The Research Foundation expresses gratitude to the report authors and to the
Project Technical Panel and Sponsors listed on following page.
The content, opinions and conclusions contained in this report are solely those of
the authors.
Compact Mobile Shelving System
Fire Testing Project
Technical Panel
Carl Anderson, City of Tacoma Fire Department
Doug Erickson, American Society for Healthcare Engineering
Ralph Gerdes, Architect
Gary Lougheed, National Research Council of Canada
Joe Noble, Noble Consultants
Rich Pehrson, Futrell Fire Consultants
John O’Neill, Protection Engineering Group
Jim Lake, NFPA staff liaison
Sponsor representatives
Chris Batterman, Spacesaver Corporation
Russ Fleming, National Fire Sprinkler Association
Scott Franson, The Viking Corporation
James Golinveaux, Tyco Fire and Building Products
Roland Huggins, American Fire Sprinkler Association
Ken Linder, SwissRe Insurance
Tom Multer, Reliable Automatic Sprinkler Company
Sponsors
American Fire Sprinkler Association
National Fire Sprinkler Association
Reliable Automatic Sprinkler Company
Spacesaver Corporation
SwissRe Insurance
Tyco Fire and Building Products
The Viking Corporation
Contributing Sponsor Representative
Terry Town, Stelterr Factory Direct Ltd.
COMPACT MOBILE SHELVING SYSTEM
FIRE TESTING PROJECT
THE FIRE PROTECTION RESEARCH FOUNDATION
ONE BATTERYMARCH PARK
QUINCY, MASSACHUSETTS
Prepared For:
Ms. Kathleen Almand
Executive Director
The Fire Protection Research Foundation
One Batterymarch Park
Quincy, Massachusetts 02169
SEC Project No.: 2007123-002
June 2, 2008
2008© Schirmer Engineering Corporation
This report is the property of Schirmer Engineering Corporation. Copies retained by the client shall be utilized
only for his use and occupancy of the project, not for the purpose of construction of any other projects.
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
i
June 2, 2008
SEC Project No.: 2007123-002
TABLE OF CONTENTS
EXECUTIVE SUMMARY ............................................................................................................. iii
INTRODUCTION / PROBLEM STATEMENT ............................................................................... 1
LITERATURE REVIEW ................................................................................................................ 2
EXPERIMENTAL TEST PLAN...................................................................................................... 5
BENCH SCALE TESTING...........................................................................................................7
FULL ARRAY TESTING .............................................................................................................9
FULL SCALE TEST 1 ..............................................................................................................12
FULL SCALE TEST 2 ..............................................................................................................13
EXPERIMENTAL RESULTS AND ANALYSIS ........................................................................... 15
BENCH SCALE TESTING.........................................................................................................15
FULL SCALE TEST 1 – OPEN ARRAY ......................................................................................17
FULL SCALE TEST 2 – CLOSED ARRAY WITH SOLID BARRIERS ................................................18
CONCLUSION AND RECOMMENDATIONS ............................................................................. 24
INDEX OF TABLES
TABLE 1. CEILING SPRINKLER PROTECTION PARAMETERS (FULL SCALE TESTING) .... 11
TABLE 2. BENCH SCALE RESULTS........................................................................................ 15
TABLE 3. SUMMARY OF SIGNIFICANT FULL SCALE TEST RESULTS .............................. 233
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
ii
June 2, 2008
SEC Project No.: 2007123-002
TABLE OF FIGURES
FIGURE 1. BENCH-SCALE TEST ARRANGEMENT (70% NON-UNIFORM LOADING) ......... 8
FIGURE 2A. FULL SCALE COMPACT MOBILE SHELVING UNIT (SECTION VIEW)............... 9
FIGURE 2B. FULL SCALE COMPACT MOBILE SHELVING UNIT (ISOMETRIC VIEW) ......... 10
FIGURE 3A. FULL SCALE TEST 1 IGNITION LOCATION (ELEVATION VIEW) ..................... 12
FIGURE 3B. FULL SCALE TEST 1 IGNITION LOCATION (PLAN VIEW) ................................ 13
FIGURE 4. FULL SCALE TEST 2 (CLOSED ARRAY WITH LONGITUDINAL BARRIERS) ... 14
FIGURE 5. LOCALIZED BURNING OBSERVED IN NON-UNIFORMLY LOADED SHELF.... 16
FIGURE 6. DEPENDENCE OF FIRE SEVERITY ON EQUIVALENCE RATIO ...................... 16
FIGURE 7. FIRE DAMAGE TO FLUE SPACE BEHIND IGNITION ......................................... 18
FIGURE 8. GAS TEMPERATURES WITHIN THE ARRAY - TEST 2 ...................................... 19
FIGURE 9. SPRINKLER SYSTEM FLOW – TEST 1............................................................... 22
FIGURE 10. STEEL BEAM TEMPERATURES AT CEILING – TEST 2 .................................... 23
APPENDIX
APPENDIX A – SUMMARY OF RELEVANT EXISTING EXPERIMENTAL DATA
(LITERATURE REVIEW)
APPENDIX B - FILING SURVEY RESULTS
APPENDIX C - UNDERWRITERS LABORATORIES TEST REPORT
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
iii
June 2, 2008
SEC Project No.: 2007123-002
EXECUTIVE SUMMARY
A series of fire tests, both bench and full scale, was conducted in an effort to solve the problem
of how to achieve fire control for paper files stored in compact mobile shelving units. The testing
focused on units which are subject to fluid use (non-archival) and protected by automatic
sprinklers positioned only at ceiling level (no in-rack protection). Recognizing the importance of
file packing density and configuration on fire severity, a series of 5 bench scale tests was
conducted preliminarily to the full scale series. A survey of user filing practices was conducted
to determine feasible bounds for storage configurations to be studied in the bench scale series
(Appendix B). The bench scale tests were then used to determine the sensitivity of the fire
severity to the combined effects of ventilation and thermal feedback from the shelving unit.
Results indicated that slumping files, which are the result of sparse file storage in any given
section of the array, increase the thermal feedback mechanism while also allowing for a smaller
fuel to air mass ratio (i.e. equivalence ratio). This creates a synergy between these
mechanisms and the rate of reaction is consequently optimized. For conservative design, this
phenomenon was exploited to the extent permitted by practical use of compact mobile shelving
units as determined in the filing survey. The result was the selection of a 70% packing density
distributed uniformly throughout the array in each full scale test (with the exception of the
ignition shelf).
Two full scale tests were completed, Test 1 with open array, and Test 2 with closed array. The
full scale testing component of this series afforded an opportunity to assess the effects of the
array configuration on the thermal feedback and air access issues highlighted by the bench
scale testing. In both tests, the high density storage configurations yielded relatively quick
smoke detection and slow fire growth, thereby potentially affording responding firefighters ample
time to produce fire control and suppression. However, the relatively slow growth of the fire in
both cases studied was not the result of the automatic sprinkler protection provided. Rather,
slow fire growth was generally due to the near balance between the energy generated by the
developing fuel-rich fire and the heat losses to the massive fuel structure surrounding the area
of burning. In this manner, the high density storage array proved capable of accomplishing the
same heat absorption objective of the automatic sprinkler system; however, this mechanism of
energy conservation is a delicate balance of several complex variables.
The implementation of solid barriers into the shelving unit acts to decrease the flow of oxygen
within the structure and also increases the dissipation of heat away from the reaction zone. By
minimizing the magnitude of energy generated by the fire while simultaneously maximizing heat
losses to other parts of the system, the experimental setup of Test 2 succeeded in providing fire
control for 81 minutes following ignition. During this time, no sprinklers were active. However,
the energy balance achieved in this test is the product of a specific type and configuration of
both the fuel and the shelving unit.
The reader should note that the findings of this project applies only to the storage of paper files
stored in the manner tested. Storage of other materials could result in a much faster fire
development.
A third full scale test should be conducted. This test should be a repeat of Test 1, however with
horizontal barriers installed as in Test 2. This scenario should be tested to confirm the ability of
the horizontal barriers to limit fire damage in a full range of likely shelf positions.
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
1
June 2, 2008
SEC Project No.: 2007123-002
INTRODUCTION / PROBLEM STATEMENT
The stated purpose of NFPA 13 is to provide a reasonable degree of protection for life and
property from fire through standardization of design, installation and testing requirements for
sprinkler systems, including private fire service mains, based on sound engineering principles,
test data and field experience. The referenced reasonable degree of protection alludes to the
fundamental design objective of automatic sprinkler systems, which is to provide either fire
suppression or fire control depending upon the specific application. Fire suppression is defined
by the standard as a degree of protection which sharply reduces the heat release rate of a fire
and prevents its re-growth by means of direct and sufficient application of water through the fire
plume to the burning fuel surface. In contrast, fire control is defined by the standard as a
degree of protection which limits the size of a fire by distribution of water so as to decrease the
heat release rate and pre-wet adjacent combustibles, while controlling ceiling gas temperatures
to avoid structural damage. The difference between these objectives is ultimately a matter of
efficiency. In applications where fire suppression by automatic sprinklers is made impractical by
parameters such as fuel type, distribution and shielding, the adequacy of a shift to control
objectives is explored.
The fundamental question of how fire control can be achieved is often approached in terms of
the necessary design parameters for automatic sprinkler system protection of the hazard. This
approach begins with questions as basic as the appropriate type of sprinkler for the application,
the necessary flow rate of water from the device (including other discharge characteristics), the
arrangement of sprinklers with respect to the hazard and the total number of sprinklers needed
to control or suppress a fire occurring within a given design area. This research into appropriate
strategies for achieving fire control in compact mobile shelving systems begins with these same
basic questions.
Compact mobile shelving systems present a unique challenge to automatic sprinkler protection
due to the variability of fuel shielding inherent to the design of the product. This study focuses
on a specific unit design which is commonly used in business occupancies where storage is
fluid and the overall fire hazard is assumed to be relatively low. In the past, challenges with fuel
shielding in rack storage arrangements have been solved by revising the relative geometry of
the fire / sprinkler interaction with the use of in-rack sprinklers. However, such an approach is
considered impractical for this product for multiple reasons. First, the mobility and fluidity of use
of the unit in a business application presents practical challenges for incorporating the
necessary network of supply piping to fixed sprinklers within the device. Second, and perhaps
more importantly, the product is already widely used in occupancies designed for Light Hazard
automatic sprinkler protection. Focusing on a solution incorporating the in-rack sprinkler
approach could therefore result in significant logistical obstacles including financial hardship for
existing users. Consequently, a means for achieving fire control with sprinklers positioned only
at the ceiling level is sought in this effort.
One of the most fundamental topics to be resolved in any area of fire research are the relevant
characteristics of the fuel being studied in terms of intrinsic (i.e. thermal parameters) and
extrinsic properties (i.e. geometric parameters). This issue is especially significant for
evaluation in compact mobile shelving systems, which may potentially be used to store fuel
loads with some level of variability in these characteristics. A survey of filing practices by
existing product users was conducted in an effort to capture realistic measurements of such
properties for justification of the experimental setup (Appendix B). During this survey, significant
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
2
June 2, 2008
SEC Project No.: 2007123-002
variability was noted in both the type and arrangement of fuel stored within compact mobile
shelving units. For example, although storage of paper files is a common use, there are
applications in the health care industry which differ markedly in the way the product is used.
Some locations included in the survey utilized compact mobile shelving units to store medicines,
garments and plastic-based medical supplies. This is a significant obstacle to achieving a
globally applicable solution for fire protection of compact mobile shelving units. The current
study is relevant only to protection of paper files.
This study addresses means to achieve fire control for paper files stored in compact mobile
shelving units, which are subject to fluid use (non-archival) and protected by automatic
sprinklers positioned only at ceiling level (no in-rack protection).
LITERATURE REVIEW
The 2007 edition (current) of NFPA 13 does not provide specific design criteria for compact
storage arrangements. As a result, designing engineers are required to utilize protection criteria
for either rack storage (storage depth 30 inches or deeper) or shelf storage (storage depth less
than 30 inches), depending on the dimensions of the compact storage array. This is
problematic because the fire protection design challenges presented by compact mobile
shelving units are fundamentally different than those presented by stationary shelf and/or rack
storage. In each of these applications, the configuration of the storage arrays obstructs the
natural flow field of water from a ceiling level automatic sprinkler to the burning fuel below. The
degree to which the water spray is obstructed is a function of the following problem specific
parameters:
1. Location of the burning fuel with respect to the location of the nearest sprinklers.
2. Dimensions of the storage array with respect to the dimensions of the surrounding
enclosure. This includes height of the array relative to the height of the room as well as
clearance between the top of the array and the sprinklers above.
3. Size of longitudinal flue spaces within the storage array.
4. Characteristics of the natural (unobstructed) water flow field including flow streamlines
and characteristic droplet momentum as determined by the design of the automatic
sprinkler system.
5. Any additional overhead obstructions specific to a given application but unrelated to
either the storage array or the sprinkler system.
The most significant difference between traditional rack or shelf storage and compact mobile
shelving is the internal mobility of the storage array. The ability to save space by manipulating
aisle widths adds a new unknown variable to the fire protection design problem. This new
unknown creates a set of geometric permutations for the obstruction of overhead sprinkler
sprays that is unique to the compact mobile shelving application. In addition, this mobility
generates complexity for protection schemes traditionally used in rack storage applications such
as in-rack sprinklers. In an effort to mitigate this issue, Section 2.2.4.3 of the Factory Mutual
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
3
June 2, 2008
SEC Project No.: 2007123-002
Global Data Sheet 8-9 “Storage of Class 1, 2, 3, 4 and Plastic Commodities” provides additional
guidance. This document calls for protection “…of rack storage in movable racks the same as
multiple-row rack storage. Supply in-rack sprinklers via flexible in-rack sprinkler system
connections, or other arrangements that provide sufficient water to the in-rack sprinklers.”
However, no specific design criteria are provided for compact storage with a depth of 30 inches
or less as FM Global considers the position of the array to be always closed, thus resulting in
depths which are 30 inches or greater.
The objective of the literature review was to analyze existing fire test data on the subject matter
and evaluate the need for additional research. The following literature was identified as relevant
to the compact mobile shelving application:
ƒ
“Fire Tests in Mobile Storage Systems for Archival Storage” – Prepared for General
Services Administration by Factory Mutual Global – June 1978
ƒ
“Archives II Mobile High Density Shelving Fire Protection System” – Prepared for
HOK/Ellerbe Becket by Gage Babcock & Associates (GBA) – November 1989
ƒ
“Design Criteria for Sprinklered Mobile Shelving Units for the National Library of Canada”
(Research Report No. 200) – Prepared for Public Works Canada by National Research
Council of Canada – November 2005
ƒ
Block, J.A. “A Theoretical and Experimental Study of Non-propagating Free-Burning
Fires”, 13th Symposium (International) on Combustion. 1971. Pittsburgh, Pennsylvania:
The Combustion Institute.
ƒ
Heskestad, G. “Modeling of Enclosure Fires”. Journal of Fire and Flammability, 1975. v6
(3): p. 253 – 273.
ƒ
“Performance Testing of Light Hazard Active Fire Protection of Compact Shelving
Systems Containing Class III Commodity (Paper Files in Paper Folders)” – Prepared by
Southwest Research Institute – August 31, 2007
ƒ
“Compact Mobile Shelving System Fire Test Plan” - Prepared by Schirmer Engineering –
November 16, 2007
The table provided in Appendix A presents a summary of the data gathered from the first three
of these tests. The remaining two publications address more fundamental combustion theory.
The data shown in the referenced table illustrates a formidable obstacle to comparison of test
results within each individual series as well as between separate test series. This obstacle is
the consistent variation of multiple design parameters influential to fire growth and development
from test to test. The end result of such variations is a number of stand-alone test results, which
are insufficient in providing conclusive evidence of the dependency of fire growth and
development on any given independent design variable. For instance:
1. Fuel type, total loading and packing density were varied in each of the three Factory
Mutual (FM) tests. In addition to these variations, spacing within the shelving array was
manipulated in each test, leading to characteristically different ventilation of the design
fires. Therefore, although automatic sprinkler design criteria were held constant,
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
4
June 2, 2008
SEC Project No.: 2007123-002
variations in results can not be definitively attributed to any one of these influential
categories.
2. Fuel type, total loading and packing density were varied in the two tests conducted by
Gage Babcock & Associates (GBA).
In addition to these variations, further
manipulations included the size of longitudinal flue spaces, clearance between the
sprinklers and the top of the fuel bed, sprinkler design density, the number of shelving
modules and the shelf depth. Certain aspects of each influential category were held
constant between tests, but variations in results can not be definitively attributed to any
one of these influences on the burning rate.
3. Sprinkler type, discharge density and spacing were varied in each of the five tests
conducted by the National Research Council of Canada (NRCC). In addition to these
variations, additional manipulations included the size of longitudinal flue spaces.
Furthermore, mechanical failures of the shelving arrangement during early stages of
initial tests likely led to erroneous conclusions.
4. The array itself consisted of solid steel shelves, end panels and tops for both the FM and
GBA test series, while the NRCC tests were conducted with open shelving arrays (i.e. no
end panels, uprights or tops). This fundamental difference in the test array effected
many influential fire growth parameters such as obstruction of the sprinkler spray,
oxidizer access to the burning fuel and incident heat flux to adjacent fuels. This
difference alone is enough to prevent meaningful comparison of results between the
NRCC test series and its counterparts.
5. The aspect ratio of an individual rack or shelf may be defined as the ratio of length to
depth. This ratio is important for determining the burning characteristics of fuel within
the space. Shallow compartments provide minimal shielding of the burning fuel from a
sprinkler spray above. Additionally, well ventilated shallow compartments typically yield
flame extensions upward along the face of the unit. In contrast, deeper compartments
provide increased shielding and consequently a higher degree of heat transfer internally
within the module. Typically a depth of 30 inches is used as the maximum value for
defining a shallow compartment (shelf) and the minimum value for a deep compartment
(rack). Utilizing this definition, the data in Table A1 (Appendix A) illustrates that all of the
FM tests and one of the two GBA tests were rack storage fires. Conversely, the second
GBA test together with all of the NRCC tests were shelf storage fires.
6. Gas temperatures at the ceiling level were significantly different throughout the testing
review. One of the subtle but significant influences on this result is the ratio of the fuel to
sprinkler clearance versus the height of the storage array. As this ratio decreases, the
temperature measurement is taken much closer to the source of the fire plume. This
geometric comparison between the fuel and its surrounding enclosure illustrate a stark
contrast between each of the three testing efforts. Ceiling gas temperatures were
consistently significantly higher in the NRCC tests; however it is unclear whether this is
due to the decreased clearance between the burning fuel and the ceiling, the lack of
enclosed shelves or the performance of the automatic sprinkler system.
Perhaps the most meaningful conclusion to be reached from an analysis of the data in Table A1
may be made with respect to the influence of ventilation on the rate of fire growth. Tests 1 and
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
5
June 2, 2008
SEC Project No.: 2007123-002
3 of the FM series contain variations in both the fuel commodity and the ignition location;
however, despite these multiple variations, there is evidence of an expected trend in the data.
In Test 1, ignition occurs in a module whose front face is open to a wide aisle. This fire is
ultimately controlled by 3 sprinklers activating at 17, 17 and 19 minutes, respectively. In Test 3,
ignition occurs in module whose front face is open to a ½ inch wide longitudinal flue space. This
fire is ultimately controlled by 4 sprinklers activating at 48, 78, 121 and 125 minutes
respectively. The effect of changing the fuel commodity on fire growth is uncertain from these
results; however, it is expected that the fire open to the aisle would have better access to
ventilation thereby producing a more robust fire plume at a faster rate of development. Despite
the strength of this plume, the relative decrease in shielding of the burning fuel allows the
sprinkler spray to control the fire. It is important to also mention the presence of a catwalk in
Test 3, which significantly obstructed the sprinkler spray.
To date, researchers have correctly identified that the size of the flue space governs both the
degree of fire shielding from sprinklers and also the magnitude of internal heat transfer within
the module. However, a key function of these flue spaces that has been overlooked is the
access that they provide for oxidizer to reach the burning fuel. For enclosed shelving, the only
significant path of air access to the burning fuel is through the longitudinal flue. As the width of
the flue is decreased, the dominant influence on the burning rate shifts from the sprinkler spray
to a balance between the ratio of fuel to oxidizer and thermal feedback between adjacent
shelves.
Perhaps the most formative work on free-burning wood crib phenomena was the development
of a fundamental theory by a Harvard University researcher named James A. Block. Block was
one of the first to derive a relationship between the internal openness of the wood crib, which he
termed the porosity, and the burning rate. Block’s theory was essentially that the burning rate
within a densely packed crib is limited by airflow through the internal flue spaces. As the
porosity is increased, a maximum fuel-limited rate of burning is reached. The limitation in this
second case is not the size of the flue spaces, but rather the exposed surface area of the fuel.
Further increasing the porosity to an extreme magnitude results in unsustained combustion as
the internal radiation is not strong enough to cause the fire to “jump the aisles” between
adjacent sticks. It is important to draw an analogy to Block’s theory because compact shelving
modules offer the unique fire protection circumstance of the ability to control fuel porosity during
a fire event. This is important because fire requires heat, fuel, and oxidizer to sustain itself.
Porosity is a parameter within the crib / compact shelving module / unit of solid fuel, which aims
to express the interaction of these critical elements to produce a certain rate of burning. This
theory suggests potentially unique approaches to achieving fire control.
EXPERIMENTAL TEST PLAN
Due to the potential for a high level of fuel shielding, fire control is the more feasible design
objective for automatic sprinkler system protection of compact mobile shelving units. As
identified in the standard definition, one of the primary components of fire control is the limitation
of fire size and spread. This in turn leads to reduced heat exposure to the structural elements of
the surrounding enclosure.
Fire spread in compact mobile shelving units may occur as surface spread along the face of the
array, interior spread through the depth of the unit or remote spread via radiant heat transfer
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
6
June 2, 2008
SEC Project No.: 2007123-002
between fuel surfaces separated by aisle space. An independent series of full-scale fire tests
conducted by Southwest Research Institute illustrated the importance of the bulk density of fuel
within the array in governing the dominant mode of fire spread. These tests were conducted
with a fuel bulk density large enough to limit the impetus for interior fire spread to the relatively
slow process of heat conduction through a semi-infinite thermally thick solid. A more
challenging scenario for fire protection is found when the rate of reaction is increased. Such an
increase in the reaction rate may be achieved by manipulating a number of system parameters
contributing to the provision of heat, fuel and oxidizer to a single cell (control volume). The
control volume may be envisioned from a number of perspectives. In this case, meaningful
results may be achieved by focusing on control volumes associated with a single shelf, the
shelving array and the surrounding room.
According to the results of the filing survey, the bulk density of fuel per shelf in compact mobile
storage arrays is typically high. Such tightly packed file storage is conducive to a fuel rich
environment within the closed portion of the array. This differs from other common forms of
storage, such as rack storage, which even filled to capacity offers ample porosity and exposed
fuel surface area to support fire development and growth. A review of existing test data
underscores the significance of this fuel packing density variable as a determinant for fire
spread interior to the array. Certainly in real world applications, portions of compact mobile
shelving units could be less than “full” for many reasons, including the ability to provide future
storage capacity. Tests were conducted to establish the density of the commodity that provides
the reasonable most severe conditions for interior fire spread and severity. This density was
used consistently throughout the array in all subsequent tests as detailed further in the following
discussion.
The bulk density of the fuel may also be expressed in terms of its porosity. One of the most
prominent efforts to estimate the burning rate of a cellulosic fuel as a function of its porosity was
a combined theoretical and experimental effort with wood cribs. The results of this effort
determined that as the packing density of the wood cribs increases, the burning of the individual
sticks constituting the crib more closely resembles the burning of a single large block of wood.
In concept, this is not unlike the result observed in the initial Spacesaver test series where
individual files and papers were packed together in a way which minimized the porosity of the
block of fuel. According to the experimentally validated theory for wood cribs, the optimum
reaction rate is proportional to exposed surface area and fuel thickness with the constant of
proportionality determined by additional influences from material type (i.e. density and molecular
composition) and moisture content.1
The first step toward specifically applying this concept to papers packed inside file folders is to
determine a standard unit of fuel thickness that will be used throughout the array. It is
suggested for practical reasons that this standard thickness correspond to a single file folder
filled to capacity with tightly packed paper. Based on data collected from the filing survey, this
standard thickness was chosen to correspond to 110 sheets of paper per folder, which produces
an approximate 1 inch thickness with the folder lying flat on its face. The array was
consequently be loaded with folders filled uniformly to this capacity. In theory, the condition
providing the largest porosity of the bulk fuel would occur with uniform spacing between folders.
However, this condition is not realistic as each individual folder would need to be provided with
1
Scale Modeling of the Transient Behavior of Wood Crib Fines in Enclosures, Jonathan Perricone – Interflam
Conference, 2007
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
7
June 2, 2008
SEC Project No.: 2007123-002
its own upright support. The more realistic condition includes a limited number of such supports
per shelf (3 per shelf in the Spacesaver units) supporting leaning records. As the population of
folders per shelf increases, the spacing between folders will decrease. Fuel porosity was
maintained consistent across each shelving section subdivided by upright compacting supports.
Bench Scale Testing
In order to determine a feasible bulk fuel density which poses the most challenging fire
protection scenario, a preliminary series of experiments was conducted with the focus placed
upon the burning of fuel inside of a single shelf. The purpose of these preliminary tests was not
to derive specifics of fire dynamics with respect to the array, but rather to make relative
comparisons of fire severity at the reduced volume. For the arrays considered in this program,
a single shelf with a length of 36 inches, a depth of 24.5 inches and a height of 9.75 inches was
used. An error in fabrication resulted in a shelf height which differed from those used in the
actual array; however, this discrepancy was irrelevant given that the goal of the bench scale
testing was to assess relative fire behavior at the bench scale. This goal was achieved by
maintaining consistent shelf dimensions for all bench scale tests.
The full front and back faces of the compartment were open for each test. Based on the results
of the filing survey, it was assumed that the minimum feasible fuel loading would be 70% of the
total shelf space. A total of 5 free-burning single-shelf tests were therefore conducted within the
range of 70-100% shelf fuel loading. The first four of these tests were conducted with a uniform
fuel load across the length of the shelf corresponding to 100% (fully packed) in the first test,
70% in the second, 85% in the third, and 100% in the fourth. The fifth test utilized the same
number of files as Test 2, yet concentrated packing to one side of the shelf such that a small
section at the extreme end of the shelf was very lightly loaded while the remainder of the shelf
was fully packed. Cerawool refractory blanket insulation was provided on the top surface of the
shelf to simulate conduction losses to shelving above in an actual assembly, which would be
filled with records (Figure 1). Measurements (in addition to video recording) taken during these
tests include the following:
•
•
•
•
•
•
•
Fuel mass loss rate
Total heat release rate (tests conducted under hood)
Multiple gas temperatures at top center depth of shelf above fuel
Solid surface temperatures at all inside and outside surfaces of the shelf
Heat flux from one of the open faces of the compartment
Heat flux from the flame extension back toward location of shelf above
Flame height
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
8
June 2, 2008
SEC Project No.: 2007123-002
FIGURE 1. BENCH-SCALE TEST ARRANGEMENT (70% NON-UNIFORM LOADING)2
Ignition was consistently achieved with one-quarter of a standard igniter, which consisted of a
1.5 inch diameter by 3 inch long cellulosic fuel bundle soaked with 2 fluid ounces of gasoline
and wrapped in a polyethylene bag, which was fully inserted between folders at the ignition
location. This location was centered along the bottom front face of the shelf and fully inserted
for the first 4 tests and moved to the bottom center of the lightly loaded shelving section for the
final test.
These initial tests were intended to provide an estimate of the most severe fuel loading
condition on any given shelf within the array. It is important to note that these results will be
specific to the type of fuel used. The introduction of a mixed commodity into the array, such as
limited plastics, will change factors as fundamental as the stoichiometric proportions of the
reaction. As a result, data regarding the rate of reaction as a function of bulk fuel porosity may
be significantly altered.
The array itself does pose additional influences on the burning fuel in the form of thermal
feedback across and ventilation within individual flue spaces. However, these flue spaces are
2
Photo courtesy of UL
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
9
June 2, 2008
SEC Project No.: 2007123-002
limited to a minimum uniform width of 1 inch throughout the array. Therefore, additional tests to
determine the most severe flue spacing were deemed unnecessary. It was recognized prior to
this stage of testing that aspects of the full array would produce effects related to heat
dissipation and bulk array ventilation which are significant factors in the final solution. However,
the bench-scale testing was used simply as a tool for relative comparison of the general fire
severity.
Full Array Testing
These tests focused on the compact mobile shelving systems typically found in offices and other
Light Hazard Occupancies. The general dimensions of the compact mobile storage array was
held constant with a height of 8 feet, depth of 2 feet (24 inches), and width of 3 feet (36 inches)
per shelf. Two 12-inch deep fixed storage units were positioned each at opposite ends of the
array as this is the standard for the majority of the compact units. One aisle with a width of 36
inches was provided between the units. The total size of the array, including this aisle space,
was approximately 200 square feet (15 feet long by 13 feet 10 inches wide). Unlike the mobile
storage units, the fixed storage units were equipped with metal backing. Vertical sheet metal
barriers were also provided in the transverse direction between each of the 5 storage bays (i.e.
at 3 foot intervals). A total of eight shelf levels were located within the unit as shown in Figures
2A and 2B. All shelving was noncombustible.
FIGURE 2A. FULL SCALE COMPACT MOBILE SHELVING UNIT (SECTION VIEW)
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
10
June 2, 2008
SEC Project No.: 2007123-002
FIGURE 2B. FULL SCALE COMPACT MOBILE SHELVING UNIT (ISOMETRIC VIEW)
Commodity was letter sized paper inside file folders stored perpendicular to the aisle. These file
folders were 12 inches long, requiring a total of 24 inches of shelf depth. Although a 1 inch
space was provided between the 24 inch deep mobile shelf units, due to randomness in file
loading by lab personnel the paper folders overlapped in some areas of the flue space thereby
potentially obstructing vertical flow of air from below and water from above. However, this
randomness is assumed to mimic actual storage practices given that the loading personnel
were not given specific instructions to avoid such a condition. Based on the results of the bench
scale tests, it was decided that all shelves would be loaded to 70% capacity as uniformly as
possible. Recognizing that each shelf within the array was divided into 4 sections by vertical file
holders, direction was given to laboratory personnel to load uniformly with respect to each
section of shelving. It was assumed that the result would yield as uniform a fuel loading
condition as could be realistically expected for the array.
The laboratory chosen for conducting these tests was the large scale test facility at
Underwriter’s Laboratories (UL) located in Northbrook, Illinois (see Appendix C for complete
facility description). A small 10,000 square foot section at the center of this 14,400 square foot
room was used for these full-scale tests. The ceiling was lowered to the 10 foot elevation.
Smoke exhaust capacity was provided at the uppermost corners of the test facility at rates low
enough to minimize any potential impact to the test results. A total of 36 automatic sprinklers
were provided underneath this simulated ceiling in accordance with the requirements of NFPA
13 for Light Hazard Occupancy.
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
11
June 2, 2008
SEC Project No.: 2007123-002
TABLE 1. CEILING SPRINKLER PROTECTION PARAMETERS (FULL SCALE TESTING)
PARAMETER
System Type
DESCRIPTION
Wet pipe ceiling level with
looped piping serving 2.5 inch
diameter branch lines
Sprinkler Type
Temperature Rating
Nominal Flow Coefficient
Orifice Size
Sprinkler Spacing
Coverage Area per Sprinkler
Deflector to Ceiling Clearance
Storage to Ceiling Clearance
Nominal Flow per Sprinkler
Nominal Pressure at Sprinkler
Water Delivery Density
Closed head quick response
standard spray pendant
o
155 F
1/2
5.6 gpm/psi
1/2 inch
15 feet
225 square feet
8 inches
18 inches
22.5 gpm
16.1 psig
2
Constant 0.10 gpm/ft
Ignition was consistently achieved with one standard igniter, which consisted of a 3 inch
diameter by 6 inch long cellulosic bundle soaked with 8 fluid ounces of gasoline and wrapped in
a polyethylene bag, which was fully inserted between folders at the ignition location.
Measurements (in addition to the video recording) taken during these tests included the
following:
•
•
•
•
•
•
•
•
•
Gas temperatures at ceiling level at various locations above the array
Sprinkler temperatures
Gas and solid shelf surface temperatures within the array
Total incident heat flux to the target arrays
Commodity moisture content
Sprinkler activation times
Sprinkler system density
Building ventilation below the simulated ceiling
Activation of an ionization smoke detector centered at ceiling level above ignition
Heat flux targets were provided on all four sides for each test. The target arrays were
comprised of the same corrugated cardboard material which constitutes Class III commodity
boxes. These targets were approximately the same height as the mobile array and separated
from the main array by a distance of 3 feet. These targets were utilized in lieu of heat flux
gauges in an effort to provide a more comprehensive measurement of critical heat flux level (i.e.
ignition of the cardboard) due to the significantly larger viewing angle afforded by the larger
target.
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
12
June 2, 2008
SEC Project No.: 2007123-002
Full Scale Test 1
The first full scale test incorporated the standard shelving unit configuration depicted in Figures
2A and 2B with one 36 inch wide aisle located centrally within the array, as shown on the test
diagram in Figure 3. A nominal 1 inch flue space was maintained between each of the
remaining shelving units within the array with no longitudinal barriers except for the metal
backing along the back face of the fixed units at each end of the array. A vertical sheet metal
divider support (same metal gage as the rest of the unit) was installed at the center depth of
each shelving bay. This divider was approximately 4.75 inches tall and served to hold in place a
total of 3 vertical sheet metal dividers per shelving bay (Figure 2B).
The shelf where ignition was designed to occur was loaded to mimic the 70% non-uniform fuel
loading studied in bench-scale Test 5. This was done in recognition of the results of the bench
scale test series which demonstrated most significant fire growth and spread in this
configuration. The intent for the full scale test was to provide a robust ignition source consisting
of a fully involved shelf for a practical but conservative design scenario. All remaining shelves
within the array were loaded at a uniform 70% packing density. The location of the ignition shelf
was chosen at the bottom center of the array on the face fronting the open aisle and directly
centered between 4 ceiling level sprinklers. Within the ignition shelf, the standard igniter was
located in the same manner as the bench scale test with non-uniform fuel loading (Test 5).
FIGURE 3A. FULL SCALE TEST 1 IGNITION LOCATION (ELEVATION VIEW)3
3
Figure Courtesy of UL
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
13
June 2, 2008
SEC Project No.: 2007123-002
FIGURE 3B. FULL SCALE TEST 1 IGNITION LOCATION (PLAN VIEW)
A B C D E
F G H I J
Full Scale Test 2
The second full scale test consisted of a closed compact shelving array with a nominal 1 inch
flue space between adjacent shelving units. A single 36 inch wide aisle was provided between
the fixed shelving unit at the North end of the array and the closest moveable carriage. This
was done to utilize the fixed unit as a target array for the measurement of radiant heat flux. In
addition to this change, longitudinal vertical sheet metal barriers (same metal gage as the rest of
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
14
June 2, 2008
SEC Project No.: 2007123-002
the unit) running the full height of the shelving array were provided at the locations shown in
Figure 4. These barriers were incorporated with the intent of slowing fire spread within the
depth of the unit. All other significant experimental parameters remained consistent with those
utilized in the first full scale test of the series. This includes the location of the ignition shelf and
the fuel loading scheme within that shelf. Consequently, the ignition location for this test did not
front an open aisle but rather a nominal 1 inch wide flue space. This location was once again
centralized below the nearest 4 sprinklers at ceiling level.
FIGURE 4. FULL SCALE TEST 2 (CLOSED ARRAY WITH LONGITUDINAL BARRIERS)
A B C D E F G H I
Fire barriers
J
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
15
June 2, 2008
SEC Project No.: 2007123-002
EXPERIMENTAL RESULTS AND ANALYSIS
Bench Scale Testing
The results of the bench scale tests were intended to gage fire severity as a function of packing
density. The concept of fire severity may incorporate a number of design parameters
depending on the application. In this instance, the primary design concern is the rate of fire
spread from the ignition cell to all adjacent cells. Therefore, the severity of the fire is most
appropriately quantified in terms of the rate of heat release by the burning fuel. Table 2
provides important information regarding experimental inputs affecting this quantity as well as
the results in terms of peak values.
TABLE 2. BENCH SCALE RESULTS
Test ID
B-1
B-2
B-3
B-4
B-5
Packing Density
All sections 100%
All sections 70%
All sections 85%
All sections 100%
Left section = 100%
Left-center section = 100%
Right-center section = 57%
Right section = 29%
Moisture Content
5.6%
5.6%
5.6%
5.4%
Peak Heat Release
Rate (kW)
1.3
16.5
8.9
7.4
5.4%
38.4
The most striking conclusion to be drawn from these results is that they are evidence of the how
the efficiency of the combustion process is dependent on the packing density of the fuel within
the shelving arrangement. For cases in which the shelf was very densely packed, the process
of combustion was observed to be very similar to the burning of a thermally thick solid in which
conduction heat losses play an important role. For very sparsely populated shelving, a
fundamental change in fire dynamics occurs. Although folders still burn as thermally thick
solids, the more significant focus becomes the synergistic effect of increased airflow and
thermal feedback from the surrounding metal shelf boundaries. Both of these influences
increase in magnitude as slumping files present natural pockets of air within the fuel as well as
at the uppermost regions of the shelf. In this latter case, the burning is somewhat analogous to
a localized compartment fire (Figure 5).
Based on this analogy, the results may be presented in terms of the peak heat release rate (a
measure of fire severity) as a function of an estimated ratio of the mass of fuel versus the mass
of air stored within the boundaries of the shelf (Figure 6). For the non-uniformly loaded case,
the size of the control volume is reduced to focus on the concentrated burning observed almost
exclusively in the lightly loaded (29% packing density) quarter of the shelf. The non-linearity
observed in this plot in addition to a peak value well below ballpark stoichiometric values is
indicative of the thermal feedback to the burning fuel. This influence is most pronounced in the
case of the lightly loaded shelf with slumped files, which behaves more like a well-ventilated
compartment fire within that section of shelving. Ultimately, these results highlight the potentially
significant influence of even localized file arrangement on the anticipated fire severity.
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
16
June 2, 2008
SEC Project No.: 2007123-002
FIGURE 5. LOCALIZED BURNING OBSERVED IN NON-UNIFORMLY LOADED SHELF
FIGURE 6. DEPENDENCE OF FIRE SEVERITY ON EQUIVALENCE RATIO
Peak Heat Release Rate (kW)
45
40
35
30
25
20
15
10
5
0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
Fuel Mass / Air Mass
The results of the bench scale test series were utilized to determine the manner in which the full
scale compact shelving array would be loaded consistently for the full scale test series.
Consequently, a uniform packing density of 70% was chosen for each section of shelving
throughout the entire array with the exception of the shelf in which ignition was located. For
both full scale tests, this particular shelf was loaded non-uniformly in the same manner as test
B-5 which produced the most severe heat release rate. This was done to make the ignition
model robust yet with a relatively short duration of influence on burning in the larger array.
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
17
June 2, 2008
SEC Project No.: 2007123-002
Full Scale Test 1 – Open Array
The first full scale test incorporated the burning of an entire compact shelving array as
described in the experimental setup section of this report. Smoke detection was recorded 44
seconds following ignition. Although ignition occurred in a shelf that fronted an open aisle and
was located centrally at the bottom of the array, flame spread during the initial stages of fire
growth did not occur vertically along the open face. Instead, as the fire burned within the
ignition shelf, added thermal feedback from the fuel mass across the 1 inch flue space at the
back of the shelf increased the intensity of the initial fire. This feedback mechanism, with
sufficient oxidizer gaining access through the open aisle, quickly resulted in fire spread through
the depth of the array and vertically upward within the flue space behind ignition. The fact that
this flame spread occurred prior to spread up the vertical face of the open aisle is evidence of
the relative significance of thermal feedback as a mechanism to increase fire intensity. Figure 7
presents the observation of more significant damage in terms of steel deformation in the flue
space behind ignition than in the area facing the open aisle. This result is indicative of higher
heat absorption in this area of the array, due to thermal feedback across the flue.
At approximately 10 minutes after ignition, flames protruded from the top of the flue space
behind ignition and banked off the ceiling above soon thereafter. Eventually, despite the
operation of sprinklers overhead, which began at 36 minutes and 9 seconds after ignition, the
fire built such significant inertia within the flue space that auto-ignition of a large area of fuel
(several shelves in the central portion of the array) fronting the open aisle was observed over a
very short duration. Though additional sprinklers operated following this observation, radiant
heat transfer across the 3 foot wide aisle was sufficient to cause remote flame spread (aisle
jump) at 54 minutes after ignition. It is important to note that these newly ignited files were
exposed to water from operating sprinklers above the aisle prior to their ignition.
Within less than 50 minutes after the activation of the first sprinkler, a total of 12 additional
sprinklers activated. The sprinkler protection provided was unable to prevent fire spread across
the open aisle and to the ends of the array during this time frame. The maximum one-minute
average gas temperature at the ceiling above ignition was recorded as 1,460oF with a maximum
one-minute average steel beam temperature of 1,063oF. Although these results seem severe at
first appraisal, it is noteworthy that the period of fire growth preceding the well-developed stage
took place over a relatively benign duration of 36 minutes with smoke detection occurring prior
to 1 minute after ignition. There was no fire spread to the target arrays.
The Light Hazard sprinkler protection scheme provided for this configuration failed to provide
fire suppression or control. Smoke detection coupled with the characteristically slow growth of
the fire observed in this case suggests a relatively large response time window for fire
department personnel; however, the conclusion that such slow fire growth is characteristic of
compact mobile shelving is only as certain as the assumptions on which it is based. It should
be highlighted that this result is based largely on the assumption of a specific fuel loading
arrangement as a practical worst-case scenario. Results of the bench scale test series
indicated that fire severity is highly sensitive to this complex parameter.
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
18
June 2, 2008
SEC Project No.: 2007123-002
FIGURE 7. FIRE DAMAGE TO FLUE SPACE BEHIND IGNITION
Full Scale Test 2 – Closed Array with Solid Barriers
In an effort to exploit the potentially beneficial result of slow fire growth, the second full scale
test was designed to incorporate solid steel barriers to divide the compact array into a series of
smaller fuel areas. As an added benefit toward achieving this goal, these barriers theoretically
provided resistance to airflow throughout the depth of the array as well as the disruption of
thermal feedback across the flue spaces in which they were located (Figure 4). Based largely
on the observation of the dominant mode of initial fire growth during the first full scale test, a
closed array was utilized for the second experiment. All other significant test variables were
held constant with respect to the first full scale experiment.
The results of the second full scale test once again yielded early smoke detection. This time,
detection occurred at approximately 1 minute and 52 seconds after ignition with visible flaming
noted at the top level of storage within the flue space fronting ignition just over three minutes
afterward. During the following 76 minutes of slow fire development, perhaps the most
important observation was a noticeable flow of buoyant gas at the ceiling yet the absence of
flames. Observations of temperatures during the test and damage patterns at the conclusion of
this test indicated higher heat exposure surrounding the point of ignition than in the areas above
that point (Figure 8).
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
19
June 2, 2008
SEC Project No.: 2007123-002
FIGURE 8. GAS TEMPERATURES WITHIN THE ARRAY - TEST 2
The temperatures profiles depicted in Figure 8 were taken at the top center of each flue space
(8th level of shelving), with the exception of one measurement taken at the lowest level of
shelving in the flue adjacent to ignition. The information provided facilitates the following
significant observations about fire growth within the array which could not be made visually
during the test:
•
For the first 20 minutes of this test, the accumulation of significant amounts of heat was
limited in location to the flue spaces directly in front and directly behind the ignition shelf
on the lowest level (1st level E). During this time, fire growth within flue space C-D
lagged significantly behind growth in flue space E-F.
•
At approximately the 20 minute mark, the gas temperature at the top of flue space E-F
begins to decay toward a new quasi-steady value of 300oC. At the same time, growth
within flue space C-D increases. Observations of fire damage at the end of the test
reveal that the decay in temperature was not due to the complete consumption of fuel
within a certain area. Rather, it is the result of the high density storage burning initially at
the surface and eventually lacking sufficient oxygen to continue burning the mass of files
at the same rate in the same location. In other words, this is evidence of local ventilation
limitations within the array.
•
For the first 20 minutes of the test, the temperatures at the top of flue space C-D (8th
level) are significantly higher than those at the bottom (1st level). After a transition stage
of several minutes, the situation reverses for the remainder of the test. Higher heat
accumulation at the bottom of the array is evidence of flame quenching by heat
absorption within the array at the upper levels. This coincides with the observation of no
visual flaming above the array during this time period.
•
The solid barrier separating shelving bays F and G (Figure 4) assists in slowing the
transfer of heat to flue space G-H. During the hour following ignition, the gas
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
20
June 2, 2008
SEC Project No.: 2007123-002
temperatures within this area remain very near ambient. Following this period, towards
the end of the test, an approximate rate of gas temperature rise of 25 degrees per
minute is observed until a maximum of nearly 300oC is reached. This relatively slow rate
of rise suggests an absence of flaming combustion behind the barrier, which would be
expected to produce a more severe rate of temperature rise.
Together, these observations confirm the mechanism for slow fire growth within the unit. The
rate of the combustion process is determined by a delicate balance commonly expressed as the
conservation of energy. When the generation of heat by the chemical reaction supersedes the
magnitude of heat losses from the system, energy is reinvested in the system and the process
is self-sustaining4. In other words, fire needs sufficient heat in conjunction with the right mixture
of fuel and oxidizer to sustain itself. For ventilation limited fires, the generation of heat is limited
by the diffusion of oxygen into the reaction zone. This limited access to oxygen can slow the
rate of heat generated by the system and significantly impact the energy balance, resulting in
slower fire growth and spread. For the first full scale test, ample access was provided via close
proximity to the open aisle in the middle of the array. With a closed array including solid
barriers, the access of air to the fire area was limited thereby resulting in a less intense reaction.
Simultaneously, as flaming gases rose vertically within the unit, they lost such significant heat to
the surrounding unburned fuel that the flame was quenched at an elevation within the unit. The
result was the observation of gasified yet unburned fuel reactants rising to the ceiling.
Eventually, by the time the fire was able to spread laterally to the end of the array, additional
oxidizer offset the balance. This resulted in rapid flame spread through the premixed gases
located at the top of the array. This flame spread fed the fire from the top-down to produce a
well developed reaction throughout the array. It was at this point, nearly 81 minutes after
ignition, when four ceiling sprinklers activated within the span of 40 seconds. During this most
intense time-period of fire growth prior to the termination of the test, the maximum one-minute
average steel beam temperature was measured at 846oF with a maximum one-minute average
gas temperature of 1560oF above ignition.
The results of Test 2 demonstrated that fire control can be achieved for a certain period of time
without automatic sprinkler protection, provided that the energy balance is manipulated by
minimizing energy generation and maximizing heat losses to the array.
Methods of
accomplishing these objectives include limiting the access of air into the interior of the unit while
also introducing enough energy absorbing material (i.e. densely packed fuel and solid steel
barriers) to limit flame height. However, recognizing that the result is a delicate balance among
a number of influential parameters, it is prudent to investigate any practical disruptions of this
balance that could limit the range of applicability of the solution. Such influences include the
following:
4
•
Introduction of additional oxidizer into the unit. This could be accomplished by repeating
full scale Test 2 in the same open array configuration used for Test 1. Such a test would
isolate the energy absorbing contribution of the solid barriers to slow flame spread.
•
Manipulation of the fuel packing arrangement and density within the array. The results
of the bench scale test series demonstrated the sensitivity of fire growth to the specific
fuel arrangement. A minimum practical packing density has been assumed in this test
Quintiere, J. Fundamentals of Fire Phenomena. John Wiley & Sons, LTD. England, 2006.
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
21
June 2, 2008
SEC Project No.: 2007123-002
series based on limited filing survey information. A uniform arrangement of files has also
been assumed throughout the array.
•
Variation in the type fuel burning within the array (i.e. introduction of plastics) influences
the optimal balance of fuel and oxidizer.
•
Variation of the geometry of fuel burning within the array (i.e. introduction of shapes
other than standard file folders) influences thermal feedback within the compartment.
•
Fuel type and geometry also dictate the energy balance (energy generation and
absorption) within the array, which may potentially be used as a means for fire control.
The full scale tests were intended to simulate fire behavior for paper files stored in a specific
size compact mobile shelving unit subject to frequent use. This was done to mimic an expected
use of the product in an office setting. It is therefore appropriate to discuss performance
objectives that define what is meant by fire control. The following criteria are often used for
such an assessment:
•
•
•
The extent of visual fire damage laterally to the ends of the array
Sufficient radiant heat transfer to cause remote ignition of a target fuel load
The magnitude of heat to which the ceiling / structure is exposed. This may be
assessed in terms of gas temperatures at the ceiling.
Automatic sprinklers act to control or suppress a fire by applying water and consequently
removing heat from the reaction zone. Therefore, evidence of control may be crudely assessed
visually by inspection of the extent and magnitude of heat damage to an array of fuel. The
general idea behind this method of assessment is that undamaged portions of the array
represent areas where enough heat was absorbed by sprinkler water to prevent the
advancement of the flame front within close proximity. However, this straightforward method of
assessing sprinkler performance becomes potentially very inaccurate in situations where
additional heat absorbing influences are present. As evidenced by the results of full scale Test
2 in this series, the relatively high density storage contained within the compact mobile shelving
array acts as both a shield to sprinkler water from above as well as an alternative heat absorber
for achieving fire control. These characteristics make it difficult to assess sprinkler performance
by visual observation of heat damage at the end of a test. However, the recorded system flow
shows a consistent positive slope, indicating that the fire growth within the unit is essentially
unchecked by the rate of water application (Figure 9).
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
22
June 2, 2008
SEC Project No.: 2007123-002
FIGURE 9. SPRINKLER SYSTEM FLOW – TEST 1
Remote ignition of a target fuel load during a test is evidence that the source fire has grown
enough to pose a significant threat to its surroundings. This occurrence is typically defined as a
failure point for the attainment of fire control because it requires a significant buildup of heat
within the reaction zone, contrary to the action of heat absorption by sprinkler water. Remote
ignition of a target fuel load across a 3 foot wide aisle was observed during the first full scale
test in this series. However, the event occurred approximately 54 minutes after ignition (53
minutes after smoke detection). Assuming that this result is repeatable, it would provide the fire
department with significant response time prior to the onset of such high heat buildup within the
array. However, from the perspective of assessing sprinkler performance, it is important to note
that this aisle jump occurred following the activation of 5 sprinklers and contributed to the
subsequent activation of another 8 within the following 30 minutes. As a result, it is clear that
the sprinkler protection did little to control the developing fire. Concerning the second full scale
test, it is reasonable to deduce that the lack of target ignition during the second full scale test
was principally due to the very low view factor between the targets and the closed array.
Assessing the severity of heat exposure to the ceiling necessary includes both the magnitude of
the heat flux to the ceiling and the materials comprising the exposed construction. Traditional
performance criteria for steel beams at the ceiling level stipulate a maximum steel temperature
of 1000oF at any point for the duration of one minute and a gas temperature not to exceed
1000oF for the duration of 5 minutes at any point during the test. These standard criteria are
based on a conservative approximation of the heat exposure necessary to significantly reduce
the strength of a standalone steel beam (not part of an assembly)5. Such temperatures were
measured in the first full scale test with a potential for similar results prior to termination of the
second test. However, in both cases, these results were obtained quite late in the experiment
(Figure 10).
5
Lie, T.T. Structural Fire Protection, American Society of Civil Engineers, NY: 1992.
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
23
June 2, 2008
SEC Project No.: 2007123-002
FIGURE 10. STEEL BEAM TEMPERATURES AT CEILING – TEST 2
It is important to note here that in an office environment, variations on ceiling construction
materials, such as a suspended ceiling, is common. As a result, application of these standard
criteria for standalone steel beams is perhaps overly conservative. Acoustic tile, which is a
common material used in suspended ceiling construction, would be expected to sustain
significant damage under lower levels of heat exposure than those established by the steel
criteria. However, a suspended ceiling is not a structural element and could actually afford
some degree of thermal insulation to more critical building components. Such interactions were
not studied in detail during this test series.
Table 3 summarizes significant results with an emphasis on peak and average peak
temperatures. Note that both fires eventually exhibit similar maximums. The distinguishing
characteristic between the scenarios is the prolonged period of growth in Test 2.
TABLE 3. SUMMARY OF SIGNIFICANT FULL SCALE TEST RESULTS
PARAMETER
Peak Gas Temperature at Ceiling Above Ignition
Maximum 1 Minute Average Gas Temperatures at
Ceiling Above Ignition
Maximum 1 Minute Average Steel Beam Temperature
Above Ignition
Peak Steel Beam Temperature
Period of Fire Growth Preceding Sprinkler Operation
Duration of Test Following Last Sprinkler Operation
Total Number of Sprinklers Activated
Remote Fire Spread (Across Aisle or to Targets)
TEST 1 RESULT
(Open Array)
o
1551 F
o
1460 F
o
TEST 2 RESULT
(Closed array, barriers)
o
1585 F
o
1560 F
o
902 F
o
1063 F
846 F
o
928 F
36 min, 9 sec
4 min, 2 sec
13
Across 3 foot wide
aisle at 54 min
81 min, 11 sec
9 sec
4
None
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
24
June 2, 2008
SEC Project No.: 2007123-002
CONCLUSION AND RECOMMENDATIONS
Valuable information has been gained from the study of paper files burning in compact mobile
shelving units under a 10 foot ceiling with Light Hazard automatic sprinkler protection. The
potentially high degree of shielding inherent to the function of these storage units is problematic
given the limitations placed on the access of water discharged by ceiling level sprinklers. In
addition, significant complexity is introduced by the number of potential permutations of fuel type
and storage arrangement within any given unit. A survey of user filing practices was conducted
to estimate the range of potential fuel loads to be expected in paper file storage applications.
Bench scale testing was utilized to determine relative fire severity as a function of this packing
density. The results of this first phase of testing included the following:
•
Densely packed files burn similar to a solid slab of fuel. In the most extreme cases
(100% packed), burning is limited to flame spread across the outside surface of the fuel
mass. Very limited burning is possible in sporadic areas where fuel, oxygen and heat
are all in sufficient supply.
•
As the density of storage is decreased uniformly, the “slab” of fuel becomes porous,
thereby allowing for fuel, heat and oxidizer to combine in sufficient quantities to produce
flames within the shelving unit.
•
The shelving unit presents a mechanism for thermal feedback to the burning fuel. This
exchange of radiant energy intensifies the rate of reaction thereby increasing fire
severity. The magnitude of this influence is directly related to the exposed surface area
of the fuel with respect to the shelving. This parameter typically increases when files are
slumped as opposed to standing perfectly upright.
•
Slumping files are the result of sparse file storage in any given section of the array. Such
a configuration not only increases the thermal feedback mechanism, but also allows for
a smaller fuel to air mass ratio (i.e. equivalence ratio). This creates a synergy between
these mechanisms and the rate of reaction is consequently optimized. For conservative
design, this phenomenon was exploited to the extent permitted by practical use of
compact mobile shelving units as determined in the filing survey. The result was the
selection of a 70% packing densely distributed uniformly throughout the array in each full
scale test (with the exception of the ignition shelf).
•
The dependence of the peak heat release rate on the equivalence ratio is evidence that
the results of these tests with paper files should not be blindly applied to other fuel types
with different chemical properties (i.e. plastics).
The full scale testing component of this series afforded an opportunity to assess the effects of
the array configuration on the thermal feedback and air access issues highlighted by the bench
scale testing. In both tests, the high density storage configurations yielded relatively quick
smoke detection and slow fire growth, thereby potentially affording responding firefighters ample
time to produce fire control and suppression. However, the relatively slow growth of the fire in
both cases studied was not primarily the result of the automatic sprinkler protection provided.
Rather, slow fire growth was the produce of a near balance between the energy generated by
the developing fuel-rich fire and the heat losses to the massive fuel structure surrounding the
Fire Protection Research Foundation
Compact Mobile Shelving System
Fire Testing Project
25
June 2, 2008
SEC Project No.: 2007123-002
area of burning. In this manner, the high density storage array proved capable of accomplishing
the same heat absorption objective of the automatic sprinkler system; however, this mechanism
of energy conservation is a delicate balance of several complex variables. In addition the
following important conclusions have been reached:
•
The automatic sprinkler system provided cooling of both the ceiling and top of the
storage arrays in both tests. However, due to the shielding of the fire, sprinkler
discharge could not reach the majority of the commodity. It is unlikely an increase in
sprinkler system density would improve the situation due to this shielding.
•
The implementation of solid barriers into the shelving unit acted to decrease the flow of
oxygen within the structure and also increased the dissipation of heat away from the
reaction zone. By minimizing the magnitude of energy generated by the fire while
simultaneously maximizing heat losses to other parts of the system, the experimental
setup of Test 2 succeeded in providing fire control for 81 minutes following ignition.
During this time, no sprinklers were active. The energy balance achieved in this test is
the product of a specific type and configuration of both the fuel and the shelving unit.
•
A third full scale test should be conducted. This test should be a repeat of Test 1,
however with horizontal barriers installed as in Test 2. This scenario should be tested to
confirm the ability of the horizontal barriers to limit fire damage in a full range of likely
shelf positions.
Submitted by,
SCHIRMER ENGINEERING CORPORATION SCHIRMER ENGINEERING CORPORATION
____________________________________
Jonathan Perricone, P.E.
Associate Engineer
_____________________________________
Garner A. Palenske, P.E.
Vice President/Regional Engineering Manager
JP:jm
S:\SNDGO\PROJECTS\JOB 2007\207123 NFPA Compact Storage Testing\rpjp.161.Fire Testing Project 6.2.08.doc
APPENDIX A
SUMMARY OF RELEVANT EXISTING EXPERIMENTAL DATA
(LITERATURE REVIEW)
TABLE A l . COMPACT SHELF STORAGE TEST DATA
fire controlled, but
in the transverse direction, tier 12
tapes
GBA-89-2
NRCC-05-1
Double the amount of
plastic computer tapes
Paper files in storage
boxes
vertical steel barriers at 42 inches
in the transverse direction, tier 12
inches high
Solid steel shelves, end panels
and tops. 36 inches in depth,
vertical steel barriers at 42 inches
in the transverse direction, tier 12
inches high.
Additional steel
vertical barrier, provided to
subdivide the shelf depth from 36
inches to 18 inches
Open shelves without end panels
or tops. 6 tiers, 10 inches tall, 32
inches wide, 30 inches deep, with
steel barrier to subdivide the shelf
depth to 15 inches.
Open shelves without end panels
or tops. 6 tiers, 10 inches tall, 32
inches wide, 30 inches deep, with
steel barrier to subdivide the shelf
depth to 15 inches.
Open shelves without end panels
or tops. 6 tiers, 10 inches tall, 32
inches wide, 30 inches deep, with
steel barrier to subdivide the shelf
depth to 15 inches.
Open shelves without end panels
or tops. 6 tiers, 10 inches tall, 32
inches wide, 30 inches deep, with
steel barrier to subdivide the shelf
depth to 15 inches.
Open shelves without end panels
or tops. 6 tiers, 10 inches tall, 32
inches wide, 30 inches deep, with
steel barrier to subdivide the shelf
depth to 15 inches.
-
36 in
7.833 ft
42 in
(45.5 + 48.75) in
-
l8 in
32 in
= 2.33
7.833 ft
= 2.13
-
l5 in
7.8 in
7.89
ft
= 0.48 + 0.52
throughout
= 0.48 -+ 0.52
1-2 inches
throughout
0.20
None
Onloff
None
= 0.08
minutes.
fire controlled
% inch, 165 Degree
F,QR, 1 0 f t x l O f t
spacing
Unit closed, no
aisle
2 sprinklers operated
Q 12 and 60 minutes.
fire controlled
fire controlled
Onloff
Unit closed, no
aisle
Uncontrolled, test
terminated
Uncontrolled, test
terminated
1100 OF sustained with 4
sprinklers operating
0.46
Pendant, QR
Unit closed, no
aisle
Uncontrolled, shelf
failure, test terminated
Uncontrolled, shelf
failure, test terminated
1650 OF sustained prior
to shelf failure and
termination of test
Unit closed, no
aisle
Fire extinguished
0.70
Sidewall, QR
See report for
sprinkler layout
Fire extinguished
1 inch
Ambient to 1650 OF and
back to ambient in 1
minute.
0.50, then
0.60 at 25
minutes
Sidewall, QR
See report for
sprinkler layout
Unit closed, no
aisle
Uncontrolled, then
extinguished when
higher density was
applied
Uncontrolled, then
extinguished when
higher density was
applied
Not reported.
1 inch
Unit closed, no
aisle
Delayed control
0.70
Sidewall, QR
See report for
sprinkler layout
Delayed control
1 inch
1300 OF and climbing
until increased sprinkler
density suppresses fire.
Waffle ceiling
assembly. Upright,
% inch, 165 Degree
F,QR, 1 0 f t x l O f t
spacing
700 OF
p
p
p
p
p
p
NRCC-05-2
NRCC-05-3
NRCC-05-4
NRCC-05-5
Paper files in storage
boxes
Paper files in storage
boxes
Paper files in storage
boxes
Paper files in storage
boxes
Test series ID = Agency - Year - Test Number
FM = Factory Mutual
GBA = Gage Babcock & Associates
NRCC = National Research Council of Canada
32 in
-=
l 5 in
2.13
7.8 in
7.89
ft
32 in
7.8 in
l5 in
7.89
32 in
7.8 in
l 5 in
7.89
32 in
7.8 in
-= 2.13
-- = 2.13
-
l5
= 2.13
7.89
ft
ft
ft
= 0.08
= 0.08
= 0.08
= 0.08
~
APPENDIX B
FILING SURVEY RESULTS
FILING SURVEY
In pursuit of determining a practical range of conditions for the bulk fuel density, a survey
of filing practices in compact mobile shelving units was conducted. This survey
consisted of visits to a total of three healthcare related sites local to the San Diego area.
Locations are referenced herein as Hospitals A, B and C, respectively. It is important to
note that the last of these visits (Hospital C) utilized compact mobile shelving units for
the purpose of storing medical supplies. A total of 2 units at this facility were surveyed.
The first of these units, manufactured by Metro, was 9 feet tall and 8 feet long, with each
module subdivided into 4 rows and 2 columns of metal wire rack shelving. The depth of
each of the 4 modules comprising the unit was 30 inches. It is estimated that stored
commodity types and proportions throughout the unit were as follows:
•
•
•
•
•
•
Rubber gloves ~ 10% total fuel load
Foam pillows, foam sandles ~ 30% total fuel load
Inflatable leg braces ~ 10% total fuel load
Electronic pressure monitoring kits ~ 10% total fuel load
Neck braces ~ 20% total fuel load
Cloth garments ~ 20% total fuel load
The second of these units, manufactured by Linco, was located in a separate room.
This unit was 6 feet tall and 6 feet long, with each module subdivided into 6 rows of
shelves spanning the entire length of the unit. The depth of each of the 5 modules
comprising the unit was 30 inches. It is estimated that stored commodity types and
proportions throughout the unit were as follows:
•
•
•
Stainless steel surgical equipment in plastic bags ~ 75% total fuel load
Rubber gloves ~ 15% total fuel load
Face masks, doctor’s surgical robes ~ 10% total fuel load
Although these findings are not particularly useful in characterizing the storage of file
folders in compact mobile shelving units manufactured by Spacesaver, they serve as an
important reminder that compact mobile shelving units may be utilized for a much wider
range of commodity storage applications.
The remaining two site visits focused exclusively on file folder storage in compact mobile
shelving units manufactured by Spacesaver. The unit located at the Hospital A storage
site was stored in a room with a 9.5 foot tall ceiling. The unit itself was 6 feet and 10
inches tall, leaving a clearance of 32 inches to the underside of the suspended ceiling
above. Pendent style automatic sprinkler protection was provided below the suspended
ceiling with a roughly estimated deflector to storage clearance of 24 inches. Commodity
storage was observed to be limited to manila file folders with plastic tab identification
labels. Folders were observed to have a typical thickness of approximately 1.5 inches.
Pictures are provided in Figure B1.
FIGURE 1. SPACESAVER STORAGE – HOSPITAL A
The bulk fuel density for each of the 28 individual shelves in a single module was
evaluated based on the number and thickness of file folders present and the total length
of available space remaining on the shelf with all files held to one side in the upright
position. Prior to this measurement, it was noted that no dividers were observed
throughout the entire storage unit. This measurement technique indicated the following
distribution of bulk densities for that particular module. This distribution was taken as
representative of the entire unit. The mode of this data set was 100% fully packed,
which was a value measured for 15 of the 28 shelves. The remainder of the unit was
packed at bulk fuel densities ranging from 50% - 95%. The 95% confidence interval for
the actual mean of this data yields a range of bulk densities from 83% - 94%.
The same survey procedure was undertaken at Hospital B. The compact mobile
shelving unit at this location, manufactured by Spacesaver, was stored with an estimated
clearance between the top of storage and the sprinkler deflector of approximately 24
inches. The unit contained a total of 10 modules subdivided into 4 rows and 6 columns.
The stored commodity consisted of paper files in manila folders with plastic identification
tabs as shown in Figure B2. The typical thickness of these folders was estimated as 1.5
inches. One issue of significance as evidenced by this photograph is the prevalence of
partial storage across the length of individual shelves. In other words, it was not
uncommon for an entire section of any given shelf to be completely void of any files.
Furthermore, for the section of such shelves which was loaded, this block was found to
be less than fully packed.
The bulk fuel density for each of the 24 individual shelves in a single module was
evaluated based on the number and thickness of file folders present and the total length
of available space remaining on the shelf with all files held to one side in the upright
position. Prior to this measurement, it was noted that dividers were present throughout
the storage unit. The mode of this data set was 80% fully packed, which was a value
measured for 10 of the 28 shelves. The remainder of the unit was packed at bulk fuel
densities ranging from 50% - 95%. The 95% confidence interval for the actual mean of
this data yields a range of bulk densities from 70% - 79%.
FIGURE B2. SPACESAVER STORAGE – HOSPITAL B
APPENDIX C
UNDERWRITERS LABORATORIES TEST REPORT
Fire Performance Evaluation of
Nominal K=5.6 Ceiling Sprinklers For Protecting
8-Ft. High Storage of
Class III Commodity in Compact Mobile Shelving
Under a 10-Ft. Ceiling
Technical Report
Underwriters Laboratories Inc.
Project 07CA64050, NC5756
for the
Fire Protection Research Foundation
April 3, 2008
Copyright © 2008, Underwriters Laboratories Inc.
NC5756
07CA64050
Issued: 04/03/2008
EXECUTIVE SUMMARY
A series of five (5) bench scale and two (2) large scale fire tests was conducted to develop data
regarding the level of protection provided by nominal K-5.6 (gpm/psi ½) pendent, quick response
ceiling sprinklers in protecting an arrangement of paper in standard manila paperboard folders
stored in moveable compact mobile shelving units to a nominal height of 8- ft.
Bench scale testing consisted of a single 36-in x 24-in x 9.75-in sheet metal shelf loaded with
varying densities of commodity. A total of five bench scale fire tests were conducted on a load
cell beneath a calorimeter where total heat release rate was measured over the course of each
test. Based on the results of these bench scale tests, the project sponsors determined that a 70%
loading density by volume would be used as the basis for both large scale fire tests.
The tests were conducted under a 10- ft high ceiling with 155°F pendent ceiling sprinklers
installed on 15 by 15-ft. spacing with the deflectors positioned 8- in. below the ceiling. The first
large scale test was conducted with the moveable compact mobile shelving units in the open
position, including a 3-ft. center aisle. The second large scale test was conducted with the
moveable compact mobile shelving units in the closed position, and included full- height
longitudinal sheet metal barriers in two of the four 24- in deep moveable shelving units.
For both large scale tests, the ceiling sprinkler system was supplied with water resulting in a
nominal sprinkler discharge density of 0.10 gpm/ft2 . The ignition location for each of the two
tests was located between four sprinklers.
The large scale fire tests utilized both moveable and fixed compact shelf units. The individual
shelving fixtures in the moving units of the systems were nominally 3- ft wide, 2- ft deep, and 8- ft
tall. Vertical transverse sheet metal barriers were provided between each of the five (5) storage
bays at 3- ft. intervals. There were eight (8) shelf levels positioned with a nominal vertical
spacing of 8, 18.5, 29, 39.5, 50, 60.5, 71, 81.5 and 95-in. from the floor. There were five (5)
shelf fixtures in each row with a total length of 15- ft.
For the first large-scale fire test, an open shelving arrangement was used consisting of six
compact shelf units. Five of the compact shelf units were moveable. Four of the moveable units
were approximately 24-in. deep and one moveable unit was approximately 12- in. deep. The
sixth compact shelf unit was fixed and approximately 12- in. deep. The compact shelf
configuration was open with a 36-in aisle space provided in the center of the array. No
longitudinal barriers were provided. The target arrangement consisted of cardboard fixed to
upright structures. The paper- faced cardboard was positioned across a 36- in. aisle space on the
north, south, east and west sides of the compact shelving array. The cardboard affixed to the
target assembly spanned 16- ft. long x 8- ft high.
The first large scale fire test was conducted for 90 minutes (90:00). A total of thirteen (13)
ceiling sprinklers operated during the 90- minute test period. The fire in Test 1 propagated to the
east, west and south ends of the main shelving array. However, the fire was contained within the
main shelving units and did not spread to the adjacent targets.
1
NC5756
07CA64050
Issued: 04/03/2008
For the second large-scale fire test, a closed shelving arrangement was used consisting of five
moveable compact shelf units in the main array and one fixed compact unit as a target. Four of
the moveable units were approximately 24- in. deep and one unit was approximately 12- in. deep.
Full height longitudinal sheet metal barriers were installed in two of the four 24- in deep
moveable units. The sixth compact shelf unit was a target which consisted of a 12- in. deep
compact shelving unit positioned across a 36-in aisle on the north side, filled to the same
capacity of the main shelving array. Cardboard fixed to upright structures were also placed
across a 36- in. aisle space on the south, east and west sides of the compact shelving array. The
cardboard target assemblies spanned 16- ft. long x 8- ft. high.
The second large scale fire test was conducted for 82 minutes (82:00). A total of four (4) ceiling
sprinklers operated during the 82- minute test period. The fire in Test 2 propagated to the east
and west ends of the array, but was generally limited to the area between the full- height
longitudinal sheet metal barriers. The fire was contained within the main shelving units and did
not spread to the adjacent target arrays.
A summary of the parameters and results is presented in Table E1.
2
NC5756
07CA64050
Issued: 04/03/2008
FIRE TEST DATE
Test Code
01/11/2008
01110802
01/18/2008
01180802
Compact Mobile Shelving
Compact Mobile Shelving
Parameters
Storage Type
Shelving Array Configuration
Open
Closed
Commodity Type
Paper in manila folders with 70%
loading density
Paper in manila folders with 70%
loading density
Nominal Storage Height (ft)
8
8
Nominal Ceiling Height (ft)
10
10
Deflector Clearance to Storage (in)
18
18
Aisle Width (in.)
36
36
Shelf Type
Solid Steel
Solid Steel
Longitudinal Barrier
None
Solid Steel
Transverse Barrier
Solid Steel
Solid Steel
Nominal Shelf Depth (in.)
24
24
Ignition Location Beneath Walkway Sprinklers
Between 4 Sprinklers
Between 4 Sprinklers
Ceiling Sprinkler Temperature Rating °F
155
155
Deflector to Ceiling (in)
8
8
Nominal Ceiling Sprinkler Discharge Coefficient K (gpm/psi ½)
5.6
5.6
Nominal Ceiling Sprinkler Discharge Pressure (psi)
16.1
16.1
2
Nominal Ceiling Sprinkler Discharge Density (gpm/ft )
0.10
0.10
Ceiling Sprinkler Spacing (ft x ft)
15 x 15
15 x 15
Length of Test (min:s)
90
82
First Ceiling Sprinkler Operation (min:s)
36:09
81:11
Last Ceiling Sprinkler Operation (min:s)
85:58
81:51
Number of Operated Ceiling Sprinklers
13
4
Peak Gas Temperature at Ceiling Above Ignition °F
1551
1585
Maximum 1 Minute Average Gas Temperature at Ceiling Above
Ignition °F
1460
1560
Peak Steel Temperature at Ceiling Above Ignition °F
1063
928
Maximum 1 Minute Average Steel Temperature Above Ignition °F
902
846
Fire Spread Across Aisle
Yes
No
Fire Spread To the Ends of the Array
Yes
Yes
Fire Spread to Target
No
No
Results
Table E1 – Summary of Test Parameters and Results
3
NC5756
07CA64050
Issued: 04/03/2008
NOTE
This Report was prepared as an account of a test sponsored by and under the direction of the Fire
Protection Research Foundation. In no event shall UL be responsible for whatever use or nonuse
is made of the information contained in this Report and in no event shall UL, its employees, or
its agents incur any obligation or liability for damages arising out of or in connection with the
use, or the inability to use, information contained in this Report.
4
NC5756
07CA64050
Issued: 04/03/2008
Table Of Contents
INTRODUCTION......................................................................................................................... 7
1
TEST FACILITY.................................................................................................................. 7
2
EQUIPMENT AND INSTRUMENTATION ..................................................................... 8
2.1
2.2
2.3
3
8
8
9
TEST ARRANGEMENT ................................................................................................... 11
3.1
3.2
3.2.1
3.2.2
3.3
3.4
3.4.1
3.4.2
4
BENCH SCALE INSTRUMENTATION
CEILING SPRINKLER SYSTEMS
LARGE -SCALE INSTRUMENTATION
BENCH SCALE TESTING AND RESULTS
LARGE -SCALE TESTING
Test 1
Test 2
TEST METHOD – LARGE SCALE TESTS 1 AND 2
TEST RESULTS
Test 1
Test 2
11
14
14
17
20
24
24
24
SUMMARY......................................................................................................................... 45
APPENDIX A ................................................................................................................................ 1
GRAPHICAL PRESENTATION OF
HEAT RELEASE RATE DATA FOR
BENCH-SCALE TESTING
1
1
1
APPENDIX B ................................................................................................................................ 1
GRAPHICAL PRESENTATION OF
TEMPERATURES AND W ATERFLOW - TEST 1
1
1
APPENDIX C ................................................................................................................................ 1
GRAPHICAL PRESENTATION OF
TEMPERATURES AND W ATERFLOW - TEST 2
1
1
List of Tables
TABLE 1
TABLE 2
TABLE 3
SPRINKLER W ATER SUPPLY PARAMETERS ........................................................................................................ 8
BENCH-SCALE TEST PARAMETERS AND RESULTS.......................................................................................... 11
SUMMARY OF TEST PARAMETERS AND RESULTS........................................................................................... 26
Table Of Figures
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5
FIGURE 6
FIGURE 7
FIGURE 8
FIGURE 9
FIGURE 10
TEST FACILITY ....................................................................................................................................................... 7
NOMINAL K=5.6 PENDENT SPRINKLER.............................................................................................................. 9
SPRINKLER AND INSTRUMENTATION LOCATIONS........................................................................................... 10
BENCH-SCALE TEST A RRANGEMENT – ELEVATION VIEW............................................................................ 12
BENCH-SCALE TEST A RRANGEMENT – PLAN VIEW....................................................................................... 12
BENCH-SCALE TEST A RRANGEMENT ............................................................................................................... 13
PLAN VIEW – TEST 1........................................................................................................................................... 15
PLAN VIEW DETAIL – TEST 1............................................................................................................................. 16
EAST / WEST ELEVATION VIEW – TEST 1........................................................................................................ 17
PLAN VIEW – TEST 2........................................................................................................................................... 18
5
NC5756
07CA64050
FIGURE 11
FIGURE 12
FIGURE 13
FIGURE 14
FIGURE 15
FIGURE 16
FIGURE 17
FIGURE 18
FIGURE 19
FIGURE 20
FIGURE 21
FIGURE 22
FIGURE 23
FIGURE 24
FIGURE 25
FIGURE 26
FIGURE 27
FIGURE 28
FIGURE 29
FIGURE 30
FIGURE 31
FIGURE 32
FIGURE 33
FIGURE 34
FIGURE 35
FIGURE 36
FIGURE 37
FIGURE 38
FIGURE 39
FIGURE 40
FIGURE 41
FIGURE 42
FIGURE 43
Issued: 04/03/2008
PLAN VIEW DETAIL – TEST 2............................................................................................................................. 19
EAST ELEVATION VIEW – TEST 2...................................................................................................................... 20
TEST SET -UP PHOTOS - TEST 1.......................................................................................................................... 21
TEST SET -UP PHOTOS - TEST 1.......................................................................................................................... 22
TEST SET -UP PHOTOS - TEST 2.......................................................................................................................... 23
EXTENT OF FIRE SPREAD – TEST 1 PLAN VIEW .............................................................................................. 27
SPRINKLER OPERATIONS – TEST 1.................................................................................................................... 28
RACK ORIENTATION FOR DAMAGE A SSESSSMENT VIEWS – TEST 1............................................................ 29
DAMAGE – TEST 1, UNIT A................................................................................................................................ 29
DAMAGE - TEST 1, UNIT B................................................................................................................................. 30
DAMAGE - TEST 1, UNIT C................................................................................................................................. 30
DAMAGE – TEST 1, UNIT D................................................................................................................................ 31
DAMAGE - TEST 1, UNIT E................................................................................................................................. 31
DAMAGE - TEST 1, UNIT F ................................................................................................................................. 32
DAMAGE - TEST 1, UNIT G................................................................................................................................. 32
DAMAGE - TEST 1, UNIT H................................................................................................................................. 33
DAMAGE - TEST 1, UNIT I .................................................................................................................................. 33
DAMAGE - TEST 1, UNIT J.................................................................................................................................. 34
POST TEST PHOTOS – TEST 1 ............................................................................................................................. 35
EXTENT OF FIRE SPREAD – TEST 2 PLAN VIEW .............................................................................................. 36
SPRINKLER OPERATIONS – TEST 2.................................................................................................................... 37
RACK ORIENTATION FOR DAMAGE A SSESSSMENT VIEWS – TEST 2............................................................ 38
DAMAGE – TEST 2, UNIT A................................................................................................................................ 38
DAMAGE - TEST 2, UNIT B................................................................................................................................. 39
DAMAGE - TEST 2, UNIT C................................................................................................................................. 39
DAMAGE – TEST 2, UNIT D................................................................................................................................ 40
DAMAGE - TEST 2, UNIT E................................................................................................................................. 40
DAMAGE - TEST 2, UNIT F ................................................................................................................................. 41
DAMAGE - TEST 2, UNIT G................................................................................................................................. 41
DAMAGE - TEST 2, UNIT H................................................................................................................................. 42
DAMAGE - TEST 2, UNIT I .................................................................................................................................. 42
DAMAGE - TEST 2, UNIT J.................................................................................................................................. 43
POST TEST PHOTOS – TEST 2 ............................................................................................................................. 44
6
NC5756
07CA64050
Issued: 04/03/2008
INTRODUCTION
This Test Report describes the Special Service Investigation conducted for the Fire Protection
Research Foundation to develop fire test data relative to the level of protection provided for a fire
scenario involving a compact mobile shelving arrangement of paper in manila folders.
1
TEST FACILITY
The fire tests were conducted in Underwriters Laboratories Inc.’s large-scale fire test facility
located in Northbrook, Illinois. The large-scale fire test building used for this investigation
includes four fire test areas that are used to develop data on the fire growth and fire suppression
characteristics of commodities, as well as the fire suppression characteristics of automatic water
sprinkler systems. A schematic of the test facility is shown in Figure 1.
ADD Test Facility
Large Scale
Fire Test Facility
Heat Release Calorimeter & RDD
PDPA Test Facility
Conditioning
Room
Warehouse
Figure 1
Test Facility
The fire tests were conducted in a 120 by 120 by 54- ft. high room fitted with an adjustable height
ceiling. The ceiling was positioned 10-ft above the floor. The test room was equipped with an
exhaust system including a regenerative thermal oxidizing (RTO) smoke abatement system.
Make-up air was provided through four inlet ducts positioned along the walls of the test facility.
The floor of the test facility was smooth, flat and surrounded with a grated drainage trench to
insure adequate drainage from the test area. The water runoff from the suppression system drain
was collected through a 180,000-gallon water treatment system.
7
NC5756
07CA64050
2
2.1
Issued: 04/03/2008
EQUIPMENT AND INSTRUMENTATION
Bench Scale Instrumentation
The instrumentation used in the bench-scale fire testing consisted of the following devices:
•
•
•
•
•
•
•
•
2.2
Three (3) 1/16- in. diameter, Type K inconel sheathed thermocouples attached to the shelving
exterior to obtain shelf temperatures
Three(3) 1/16- in. diameter, Type K inconel sheathed thermocouples positioned within the
shelving to obtain gas temperatures
Two (2) Radiometers
Moisture meter to measure the commodity moisture content of the fuel package
Weigh-scale in the appropriate range
Electronic data acquisition system at a two-second-scan rate to obtain the oxygen
consumption, temperature, and heat flux data
Video and still cameras to capture and record images of the fire test
Stopwatches and timing devices located within the data acquisition system were used to
monitor and record significant events during the fire tests
Ceiling Sprinkler Systems
Thirty-six (36) quick-response standard spray pendent style sprinklers having a 155° F
temperature rating and a nominal flow coefficient (K) of 5.6 gpm/psi1/2 were installed on 15-x
15-ft. spacing in a closed head, wet pipe, automatic sprinkler system at the ceiling level. The
sprinklers were supplied through a looped piping system consisting of 2 ½- in. diameter branch
lines. The sprinklers were positioned with the distance between the deflector and ceiling
measured to be 8- in.
The water supply to the ceiling sprinklers was controlled to achieve the parameters shown in
Table 1.
Close- up photos of the sprinklers used in this test are shown in Figure 2. A plan view of the
sprinkler and instrumentation layout is shown in Figure 3.
Table 1
Sprinkler Location
Ceiling
Sprinkler Water Supply Parameters
Nominal flow per
sprinkler (gpm)
22.5
8
Nominal pressure at
sprinkler (psig)
16.1
NC5756
07CA64050
Issued: 04/03/2008
Figure 2
2.3
Nominal K=5.6 Pendent Sprinkler
Large-Scale Instrumentation
The instrumentation used in the testing consisted of the following devices:
`
• 1/16 in. diameter, Type K inconel sheathed thermocouples located below the ceiling adjacent
to each sprinkler to record temperatures.
• 1/16 in. diameter, Type K inconel sheathed thermocouples located below the Elevated
Access Walkway adjacent to each sprinkler to record temperatures.
• 1/16 in. diameter, Type K inconel sheathed thermocouples located 6, 12, and 18-in. below
the ceiling above the ignition location.
• Five 1/16 in. diameter, Type K inconel sheathed thermocouples embedded in a 50.5-in. long
steel beam attached to the bottom of the hard tile ceiling directly above the fire.
• Bristol Babcock Model 250815B pressure transducers in a 0-100 psi range were used to
measure the water pressure in the sprinkler system.
• Fischer Porter model X3311SFD20 – 12 inch magnetic flowmeter in the 0-3200 gallons per
minute (gpm) range were used to measure the water flow rate.
• Stopwatches and timing devices located within the data acquisition system were used to
monitor and record significant events during the fire tests.
• Video and Infrared cameras were used to capture and record images of the fire test.
9
NC5756
07CA64050
Issued: 04/03/2008
Figure 3
Sprinkler and Instrumentation Locations
10
NC5756
07CA64050
3
Issued: 04/03/2008
TEST ARRANGEMENT
3.1
Bench Scale Testing and Results
Bench scale testing was conducted beneath a 500-kW rated product calorimeter in a test room
measuring 50- ft. x 15-ft. x 15-ft. (l x w x h) with a 12.75-ft. x 12.75-ft. square-shaped collection
hood located centrally in the room approximately 7- ft. above the floor. Collected combustion
products are exhausted by way of a 24- in. x 24- in. plenum into an 18- in. diameter exhaust duct
for data measurement.
Bench scale testing consisted of a single 36-in x 24-in x 9.75-in sheet metal shelf loaded with
varying densities of commodity. A total of six (6) vertical sheet metal dividers approximately 9in. long x 6- in. high were attached to the shelf to evenly space the folders. The shelf enclosure
was insulated on the top with a single thickness of Cerawool refractory blanket. The stored
commodity consisted of sheet paper in common manila paperboard folders. Stored commodity
loading was based on sheet count, folder count, and weight. Ignition was accomplished using a
one-quarter standard igniter constructed from a 1.5-in. diameter by 3- in. long cellulosic bundle
soaked with two fluid ounces of gasoline and wrapped in a polyethylene bag.
Fire tests with a loading density of 100% were conducted twice to confirm baseline weight and
folder count, and the loading arrangement ignition location was similar in both tests. Fire tests
with a loading density of 70% were conducted twice to gather additional heat release data. In the
first 70% loading test, 110 folders were evenly spaced back to back within the 24-in. deep shelf.
In the second 70% loading test, the 110 folders were not evenly spaced but were more loosely
loaded in the ignition area.
With the test commodity samples loaded in the shelf assembly and positioned under the products
collector with video and data collection operating, ignition was accomplished and the fire was
allowed to burn freely until the fuel was either consumed or no longer sus tained combustion.
The bench-scale test parameters and results are presented in Table 2.
Test Load Density
No.
(%)
Moisture Initial Weight Weight Loss
Test Duration
Folder Count
Content (%)
(lbs)
(lbs)
(min)
Peak HRR
(kW)
1
100
5.6
169.2
0.5
153
60
1.3
2
70
5.6
118.4
16.6
110
60
16.5
3
85
5.6
143.1
5
128
30
8.9
4
100
5.4
168.6
0.2
153
40
7.4
5
70
5.4
118.3
22.4
110
78
38.4
Note: All folders contained an average of 110 sheets of standard 8.5-in x 11- in plain white paper
Table 2
Bench-Scale Test Parameters and Results
11
NC5756
07CA64050
Issued: 04/03/2008
A diagram of the bench scale shelf mockup and photos of the fire test arrangement are shown in
Figures 4, 5 and 6.
Figure 4
Figure 5
Bench-Scale Test Arrangement – Elevation View
Bench-Scale Test Arrangement – Plan View
12
NC5756
07CA64050
Issued: 04/03/2008
Figure 6
Bench-Scale Test Arrangement
13
NC5756
07CA64050
3.2
Issued: 04/03/2008
Large-Scale Testing
The large scale fire tests utilized both moveable and fixed units. The moveable units were
assembled on a carriage rail system which was attached to the test room floor. The individual
shelving fixtures in the moveable units were nominally 3- ft wide, 2- ft deep, and 8-ft tall.
Vertical transverse sheet metal barriers were provided between each of the five (5) storage bays
at 3-ft. intervals. There were eight (8) shelf levels positioned with a nominal vertical spacing of
8, 18.5, 29, 39.5, 50, 60.5, 71, 81.5 and 95-in. from the floor. There were five (5) shelf fixtures
in each row with a total length of 15- ft.
Based on the results of the bench scale tests, the project sponsors determined that a loading
density of 70% would be used for both large scale tests. This loading density included 56 folders
per 12- in. x 36- in. loading bay, with the folders evenly spaced by partial vertical dividers. There
was a nominal 2- in. clear space between the top of the folders and the bottom of the next shelf
above the folders.
3.2.1
Test 1
For the first large-scale fire test, an open shelving arrangement was used consisting of six
compact shelf units. Five of the compact shelf units were moveable. Four of the moveable units
were approximately 24-in. deep and one moveable unit was approximately 12- in. deep. The
sixth compact shelf unit was fixed and approximately 12- in. deep. The compact shelf
configuration was open with a 36- in aisle space provided in the center of the array. A nominal 1in. flue space was maintained between each shelving unit. No longitudinal barriers were
provided. The target arrangement consisted of cardboard fixed to upright structures. The paperfaced cardboard was positioned across a 36- in. aisle space on the north, south, east and west
sides of the compact shelving array. The cardboard affixed to the target assembly spanned 16-ft.
long x 8- ft high.
The individual shelving unit bays were nominally 36-in wide, 10.5- in high and 24- in deep. At
the center of each 24- in deep bay, a 4.75- in tall sheet metal divider was located as a backing
device and vertical divider support. A total of six (6) vertical sheet metal dividers approximately
9-in. long x 6- in. high were attached to each 36- in wide, 10.5- in high and 24-in deep shelf bay to
evenly space the folders. A total of three (3) vertical sheet metal dividers approximately 9-in.
long x 6- in. high were attached to each 36- in wide, 10.5- in high and 12- in deep shelf bay.
The arrangement is shown in Figures 7,8 and 9. Photos of the arrangement are shown in Figures
13 and 14.
14
NC5756
07CA64050
Issued: 04/03/2008
Figure 7
Plan View – Test 1
15
NC5756
07CA64050
Issued: 04/03/2008
Figure 8
Plan View Detail – Test 1
16
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01110802
Main Array from East
Project 07CA64050
File NC5756
Figure 9
3.2.2
East / West Elevation View – Test 1
Test 2
For the second large-scale fire test, a closed shelving arrangement was used consisting of five
moveable compact shelf units in the main array and one fixed compact unit as a target. Four of
the moveable units were approximately 24- in. deep and one unit was approximately 12- in. deep.
Full height longitudinal sheet metal barriers were installed in two of the four 24- in deep
moveable units. The sixth compact shelf unit was a target which consisted of a 12- in. deep
compact shelving unit positioned across a 36-in aisle on the north side, filled to the same
capacity of the main shelving array. A no minal 1-in. flue space was maintained between each
shelving unit. Cardboard fixed to upright structures was also placed across a 36- in. aisle space
on the south, east and west sides of the compact shelving array. The cardboard target assemblies
spanned 16- ft. long x 8- ft. high.
The individual shelving unit bays were nominally 36-in wide, 10.5- in high and 24- in deep. At
the center of each 24- in deep bay, a 4.75- in tall sheet metal divider was located as a backing
device and vertical divider support. A total of six (6) vertical sheet metal dividers approximately
9-in. long x 6- in. high were attached to each 36- in wide, 10.5- in high and 24-in deep shelf bay to
evenly space the folders. A total of three (3) vertical sheet metal dividers approximately 9-in.
long x 6- in. high were attached to each 36- in wide, 10.5- in high and 12- in deep shelf bay.
The arrangement is shown in Figures 10, 11 and 12. Photos of the arrangement are shown in
Figure 15.
17
NC5756
07CA64050
Issued: 04/03/2008
Figure 10
Plan View – Test 2
18
NC5756
07CA64050
Issued: 04/03/2008
Figure 11
Plan View Detail – Test 2
19
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01180802
Main Array from East
Project 07CA64050
File NC5756
Figure 12
3.3
East Elevation View – Test 2
Test Method – Large Scale Tests 1 and 2
The nominal 8-ft. high storage array was positioned under a 10-ft. ceiling. The distance from the
top of the commodity to the sprinkler deflector was 18- in.
Thirty-six upright, nominal K=5.6 ceiling sprinklers were installed on a 15 by 15- ft. spacing for
this test. The moveable compact mobile shelving array was assembled between four sprinklers
such that the ignition location for each of the two tests, at the base of storage, was located
between four sprinklers.
The ceiling sprinkler system was supplied with water resulting in a nominal sprinkler discharge
density of 0.10 gpm/ft 2 .
Ignition was accomplished using one standard igniter constructed from a 3- in. diameter by 6-in.
long cellulosic bundle soaked with 8 oz. of gasoline and wrapped in a polyethylene bag. The
igniter was located on the base of the south side shelf (Rack E in Figures 18 and 32), within a
storage folder. The ignition location was centered between four (4) ceiling sprinklers.
20
NC5756
07CA64050
Issued: 04/03/2008
Shelf Setup
Carriage Arrangement
Carriage Arrangement
Shelf Detail
Shelf Detail
Figure 13
Loading Detail
Test Set-Up Photos - Test 1
21
NC5756
07CA64050
Issued: 04/03/2008
Ignitor
Figure 14
Final Test Arrangement
Test Set-Up Photos - Test 1
22
NC5756
07CA64050
Issued: 04/03/2008
Divider Detail
Barrier Detail
Shelf Detail
Shelf / Carriage Detail
Loading Detail
Final Test Arrangement
Figure 15
Test Set-Up Photos - Test 2
23
NC5756
07CA64050
3.4
3.4.1
Issued: 04/03/2008
Test Results
Test 1
The test was conducted for 90 minutes (90:00). A total of eight (13) ceiling sprinklers operated
during the 90- minute test period. The first sprinkler activated at thirty-six minutes nine seconds
(36:09) after ignition and the last sprinkler operated at eighty-five minutes fifty-eight seconds
(85:58) after ignition.
A combination photoelectric / ionization type smoke detector was located at the ceiling level
directly above ignition. The detector activated at 0 minutes 44 seconds (00:44) after ignition.
Visible flaming was noted at the top level of storage at approximately ten minutes (10:00) after
ignition. Flame travel was noted across the ignition aisle at approximately fifty-our minutes
(54:00) after ignition.
The fire in Test 1 propagated to the east, west and south ends of the main shelving array. The fire
was contained within the main shelving units and did not spread to the adjacent target arrays.
The maximum one- minute average steel beam temperature measured above ignition was 902 ºF
and the maximum one-minute average gas temperature measured above ignition was 1460 ºF.
A plan view of the extent of fire is presented in Figure 16 and sprinkler operation times in Figure
17. The test assembly orientation is shown in Figure 18. Elevation views of the extent of fire
spread and damage are shown in Figures 19 through 28. Post-test photos are presented in Figure
29. Tabulated test results are presented in Table 3. Temperature data vs. time charts are
presented in Appendix B.
3.4.2
Test 2
The second large scale fire test was conducted for 82 minutes (82:00). A total of four (4) ceiling
sprinklers operated during the 82- minute test period. The first sprinkler activated at eighty-one
minutes eleven seconds (81:11) after ignition and the last sprinkler operated at eighty-one
minutes fifty-one seconds (81:51) after ignition.
A combination photoelectric / ionization type smoke detector was located at the ceiling level
directly above ignition. The detector activated at 1 minute 52 seconds (01:52) after ignition.
Visible flaming was noted at the top level of storage at approximately five minutes thirty seconds
(05:30) after ignition.
The fire in Test 2 propagated to the east and west ends of the array, but was generally limited to
the area between the full- height longitudinal sheet metal barriers. The fire was contained within
the main shelving units and did not spread to the adjacent target arrays. The maximum oneminute average steel beam temperature measured above ignition was 846 ºF and the maximum
one- minute average gas temperature measured above ignition was 1560 ºF.
24
NC5756
07CA64050
Issued: 04/03/2008
A plan view of the extent of fire is presented in Figure 30 and sprinkler operation times in Figure
31. The test assembly orientation is shown in Figure 32. Elevation views of the extent of fire
spread and damage are shown in Figures 33 through 42. Post-test photos are presented in Figure
43. Tabulated test results are presented in Table 3. Temperature data vs. time charts are
presented in Appendix C.
25
NC5756
07CA64050
Issued: 04/03/2008
FIRE TEST DATE
Test Code
01/11/2008
01110802
01/18/2008
01180802
Compact Mobile Shelving
Compact Mobile Shelving
Parameters
Storage Type
Shelving Array Configuration
Open
Closed
Commodity Type
Paper in manila folders with 70%
loading density
Paper in manila folders with 70%
loading density
Nominal Storage Height (ft)
8
8
Nominal Ceiling Height (ft)
10
10
Deflector Clearance to Storage (in)
18
18
Aisle Width (in.)
36
36
Shelf Type
Solid Steel
Solid Steel
Longitudinal Barrier
None
Solid Steel
Transverse Barrier
Solid Steel
Solid Steel
Nominal Shelf Depth (in.)
24
24
Ignition Location Beneath Walkway Sprinklers
Between 4 Sprinklers
Between 4 Sprinklers
Ceiling Sprinkler Temperature Rating °F
155
155
Deflector to Ceiling (in)
8
8
Nominal Ceiling Sprinkler Discharge Coefficient K (gpm/psi ½)
5.6
5.6
Nominal Ceiling Sprinkler Discharge Pressure (psi)
16.1
16.1
2
Nominal Ceiling Sprinkler Discharge Density (gpm/ft )
0.10
0.10
Ceiling Sprinkler Spacing (ft x ft)
15 x 15
15 x 15
Length of Test (min:s)
90
82
First Ceiling Sprinkler Operation (min:s)
36:09
81:11
Last Ceiling Sprinkler Operation (min:s)
85:58
81:51
Number of Operated Ceiling Sprinklers
13
4
Peak Gas Temperature at Ceiling Above Ignition °F
1551
1585
Maximum 1 Minute Average Gas Temperature at Ceiling Above
Ignition °F
1460
1560
Peak Steel Temperature at Ceiling Above Ignition °F
1063
928
Maximum 1 Minute Average Steel Temperature Above Ignition °F
902
846
Fire Spread Across Aisle
Yes
No
Fire Spread To the Ends of the Array
Yes
Yes
Fire Spread to Target
No
No
Results
Table 3
Summary of Test Parameters and Results
26
NC5756
07CA64050
Issued: 04/03/2008
Figure 16
Extent of Fire Spread – Test 1 Plan View
27
NC5756
07CA64050
Issued: 04/03/2008
Figure 17
Sprinkler Operations – Test 1
28
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01110802
Damage Assessment Orientation
Project 07CA64050
File NC5756
Figure 18
Rack Orientation for Damage Assesssment Views – Test 1
TEST Code: 01110802
Damage Assessment - Unit A
Project 07CA64050
File NC5756
Figure 19
Damage – Test 1, Unit A
29
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01110802
Damage Assessment - Unit B
Project 07CA64050
File NC5756
Figure 20
Damage - Test 1, Unit B
TEST Code: 01110802
Damage Assessment - Unit C
Project 07CA64050
File NC5756
Figure 21
Damage - Test 1, Unit C
30
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01110802
Damage Assessment - Unit D
Project 07CA64050
File NC5756
Figure 22
Damage – Test 1, Unit D
TEST Code: 01110802
Damage Assessment - Unit E
Ignition Aisle
Project 07CA64050
File NC5756
Figure 23
Damage - Test 1, Unit E
31
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01110802
Damage Assessment - Unit F
Project 07CA64050
File NC5756
Figure 24
Damage - Test 1, Unit F
TEST Code: 01110802
Damage Assessment - Unit G
Project 07CA64050
File NC5756
Figure 25
Damage - Test 1, Unit G
32
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01110802
Damage Assessment - Unit H
Project 07CA64050
File NC5756
Figure 26
Damage - Test 1, Unit H
TEST Code: 01110802
Damage Assessment - Unit I
Project 07CA64050
File NC5756
Figure 27
Damage - Test 1, Unit I
33
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01110802
Damage Assessment - Unit J
Project 07CA64050
File NC5756
Figure 28
Damage - Test 1, Unit J
34
NC5756
07CA64050
Issued: 04/03/2008
Ignition Aisle
Aisle Opposite Ignition
Aisle “A”
Aisle “J”
Overhead from NE
Figure 29
Overhead from NE
Post Test Photos – Test 1
35
NC5756
07CA64050
Issued: 04/03/2008
Figure 30
Extent of Fire Spread – Test 2 Plan View
36
NC5756
07CA64050
Issued: 04/03/2008
Figure 31
Sprinkler Operations – Test 2
37
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01180802
Damage Assessment Orientation
Project 07CA64050
File NC5756
Figure 32
Rack Orientation for Damage Assesssment Views – Test 2
TEST Code: 01180802
Damage Assessment - Unit A
Project 07CA64050
File NC5756
Figure 33
Damage – Test 2, Unit A
38
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01180802
Damage Assessment - Unit B
Project 07CA64050
File NC5756
Figure 34
Damage - Test 2, Unit B
TEST Code: 01180802
Damage Assessment - Unit C
Project 07CA64050
File NC5756
Figure 35
Damage - Test 2, Unit C
39
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01180802
Damage Assessment - Unit D
Project 07CA64050
File NC5756
Figure 36
Damage – Test 2, Unit D
TEST Code: 01180802
Damage Assessment - Unit E
Ignition Aisle
Project 07CA64050
File NC5756
Figure 37
Damage - Test 2, Unit E
40
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01180802
Damage Assessment - Unit F
Project 07CA64050
File NC5756
Figure 38
Damage - Test 2, Unit F
TEST Code: 01180802
Damage Assessment - Unit G
Project 07CA64050
File NC5756
Figure 39
Damage - Test 2, Unit G
41
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01180802
Damage Assessment - Unit H
Project 07CA64050
File NC5756
Figure 40
Damage - Test 2, Unit H
TEST Code: 01180802
Damage Assessment - Unit I
Project 07CA64050
File NC5756
Figure 41
Damage - Test 2, Unit I
42
NC5756
07CA64050
Issued: 04/03/2008
TEST Code: 01180802
Damage Assessment - Unit J
Project 07CA64050
File NC5756
Figure 42
Damage - Test 2, Unit J
43
NC5756
07CA64050
Issued: 04/03/2008
Ignition Aisle
Aisle Opposite Ignition
Aisle A*
Aisle H*
*Note: Vertical barriers installed between B/C and F/G
Aisle C
Figure 43
Aisle D
Post Test Photos – Test 2
44
NC5756
07CA64050
4
Issued: 04/03/2008
SUMMARY
A series of five (5) bench scale and two (2) large scale fire tests was conducted to develop data
regarding the level of protection provided by nominal K-5.6 (gpm/ psi ½) pendent, quick response
ceiling sprinklers in protecting an arrangement of paper in standard manila paperboard folders
stored in moveable compact mobile shelving units to a nominal height of 8- ft.
Bench scale testing consisted of a single 36-in x 24-in x 9.75-in sheet metal shelf loaded with
varying densities of commodity. A total of five bench scale fire tests were conducted on a load
cell beneath a calorimeter where total heat release rate was measured over the course of each
test. Based on the results of these bench scale tests, the project sponsors determined that a 70%
loading density by volume would be used as the basis for both large scale fire tests.
The tests were conducted under a 10- ft high ceiling with 155°F pendent ceiling sprinklers
installed on 15 by 15-ft. spacing with the deflectors positioned 8- in. below the ceiling. The first
large scale test was conducted with the moveable compact mobile shelving units in the open
position, including a 3-ft. center aisle. The second large scale test was conducted with the
moveable compact mobile shelving units in the closed position, and included full- height
longitudinal sheet metal barriers in two of the four 24- in deep moveable shelving units.
For both large scale tests, the ceiling sprinkler system was supplied with water resulting in a
nominal sprinkler discharge density of 0.10 gpm/ft2 . The ignition location for each of the two
tests was located between four sprinklers.
The large scale fire tests utilized both moveable and fixed compact shelf units. The individual
shelving fixtures in the moving units of the systems were nominally 3- ft wide, 2- ft deep, and 8- ft
tall. Vertical transverse sheet metal barriers were provided between each of the five (5) storage
bays at 3- ft. intervals. There were eight (8) shelf levels positioned with a nominal vertical
spacing of 8, 18.5, 29, 39.5, 50, 60.5, 71, 81.5 and 95-in. from the floor. There were five (5)
shelf fixtures in each row with a total length of 15- ft.
For the first large-scale fire test, an open shelving arrangement was used consisting of six
compact shelf units. Five of the compact shelf units were moveable. Four of the moveable units
were approximately 24-in. deep and one moveable unit was approximately 12- in. deep. The
sixth compact shelf unit was fixed and approximately 12- in. deep. The compact shelf
configuration was open with a 36-in aisle space provided in the center of the array. No
longitudinal barriers were provided. The target arrangement consisted of cardboard fixed to
upright structures. The paper- faced cardboard was positioned across a 36- in. aisle space on the
north, south, east and west sides of the compact shelving array. The cardboard affixed to the
target assembly spanned 16- ft. long x 8- ft high.
The first large scale fire test was conducted for 90 minutes (90:00). A total of thirteen (13)
ceiling sprinklers operated during the 90-minute test period. The first sprinkler activated at
thirty-six minutes nine seconds (36:09) after ignition and the last sprinkler operated at eightyfive minutes fifty-eight seconds (85:58) after ignition.
45
NC5756
07CA64050
Issued: 04/03/2008
The fire in Test 1 propagated to the east, west and south ends of the main shelving array.
However, the fire was contained within the main shelving units and did not spread to the
adjacent targets. The maximum one-minute average steel beam temperature measured above
ignition was 902 ºF and the maximum one- minute average gas temperature measured above
ignition was 1460 ºF.
For the second large-scale fire test, a closed shelving arrangement was used consisting of five
moveable compact shelf units in the main array and one fixed compact unit as a target. Four of
the moveable units were approximately 24- in. deep and one unit was approximately 12- in. deep.
Full height longitudinal sheet metal barriers were installed in two of the four 24- in deep
moveable units. The sixth compact shelf unit was a target which consisted of a 12- in. deep
compact shelving unit positioned across a 36-in aisle on the north side, filled to the same
capacity of the main shelving array. Cardboard fixed to upright structures were also placed
across a 36- in. aisle space on the south, east and west sides of the compact shelving array. The
cardboard target assemblies spanned 16- ft. long x 8- ft. high.
The second large scale fire test was conducted for 82 minutes (82:00). A total of four (4) ceiling
sprinklers operated during the 82- minute test period. The first sprinkler activated at eighty-one
minutes eleven seconds (81:11) after ignition and the last sprinkler operated at eighty-one
minutes fifty-one seconds (81:51) after ignition.
The fire in Test 2 propagated to the east and west ends of the array, but was generally limited to
the area between the full- height longitudinal sheet metal barriers. The fire was contained within
the main shelving units and did not spread to the adjacent target arrays. The maximum oneminute average steel beam temperature measured above ignition was 846 ºF and the maximum
one- minute average gas temperature measured above ignition was 1560 ºF.
A summary of the parameters and results is presented in Table 3.
Report by:
Reviewed by:
Christopher Gates
Project Engineer
3019INBK
Tel: 847-664-2618
Fax: 847-313-2618
Michael McCormick
Staff Engineering Associate
3019INBK
Tel: 847-664-3303
Fax: 847-313-3303
46
NC5756
07CA64050
Appendix A – Graphs
APPENDIX A
Graphical Presentation of
Heat Release Rate Data for
Bench-Scale Testing
A-1
Issued 03/31/2008
NC5756
07CA64050
Appendix A – Graphs
Issued 03/31/2008
Fire Protection Research Foundation
HRR Comparison
Bench Scale Loading Density Data
40
35
30
Loading Density
HRR (kW)
25
70% Test 1
70% Test 2
20
85% Test 1
100% Test 1
15
100% Test 2
10
5
0
0
10
20
30
40
50
60
Time (min)
Fire Protection Research Foundation
HRR Comparison
Bench Scale Loading Density Data
40
35
30
HRR (kW)
25
Loading Density
70% - Test 1
20
70% - Test 2
15
10
5
0
0
10
20
30
40
Time (m)
A-2
50
60
70
80
NC5756
07CA64050
Appendix B – Graphs
Issued 03/31/2008
APPENDIX B
Graphical Presentation of
Temperatures and Waterflow - Test 1
B-1
NC5756
07CA64050
Appendix B – Graphs
Issued 03/31/2008
Fire Protection Research Foundation
Compact Mobile Shelving - Test 1
Sprinklers 1-6
90
80
70
60
(Deg C)
Spk.001
Spk.002
Spk.003
Spk.004
Spk.005
50
40
Spk.006
30
20
10
0
0
10
20
30
40
50
60
70
80
90
Time (Min)
Fire Protection Research Foundation
Compact Mobile Shelving - Test 1
Sprinklers 7-12
90
80
70
60
(Deg C)
Spk.007
Spk.008
Spk.009
Spk.010
Spk.011
50
40
Spk.012
30
20
10
0
0
10
20
30
40
50
Time (Min)
B-2
60
70
80
90
NC5756
07CA64050
Appendix B – Graphs
Issued 03/31/2008
Fire Protection Research Foundation
Compact Mobile Shelving - Test 1
Sprinklers 13-18
90
80
70
60
(Deg C)
Spk.013
Spk.014
Spk.015
Spk.016
Spk.017
50
40
Spk.018
30
20
10
0
0
10
20
30
40
50
60
70
80
90
Time (Min)
Fire Protection Research Foundation
Compact Mobile Shelving - Test 1
Sprinklers 19-24
90
80
70
60
(Deg C)
Spk.019
Spk.020
Spk.021
Spk.022
Spk.023
50
40
Spk.024
30
20
10
0
0
10
20
30
40
50
Time (Min)
B-3
60
70
80
90
NC5756
07CA64050
Appendix B – Graphs
Issued 03/31/2008
Fire Protection Research Foundation
Compact Mobile Shelving - Test 1
Sprinklers 24-30
90
80
70
60
(Deg C)
Spk.025
Spk.026
Spk.027
Spk.028
Spk.029
50
40
Spk.030
30
20
10
0
0
10
20
30
40
50
60
70
80
90
Time (Min)
Fire Protection Research Foundation
Compact Mobile Shelving - Test 1
Sprinklers 31-36
90
80
70
60
(Deg C)
Spk.031
Spk.032
Spk.033
Spk.034
Spk.035
50
40
Spk.036
30
20
10
0
0
10
20
30
40
50
Time (Min)
B-4
60
70
80
90
NC5756
07CA64050
Appendix B – Graphs
Issued 03/31/2008
Fire Protection Research Foundation
Compact Mobile Shelving - Test 1
Sprinkler System Flow
600
500
GPM
400
Flow
300
200
100
0
30
40
50
60
70
80
90
Time (Min)
Fire Protection Research Foundation
Compact Mobile Shelving - Test 1
Ceiling Steel Temperature above Ignition
Note: TC #1 malfunctioned
700
600
500
400
(Deg C)
AMB02
AMB03
AMB04
AMB05
300
200
100
0
0
10
20
30
40
50
Time (Min)
B-5
60
70
80
90
NC5756
07CA64050
Appendix B – Graphs
Issued 03/31/2008
Fire Protection Research Foundation
Compact Mobile Shelving - Test 1
Air Temperature above Ignition
900
800
700
(Deg C)
600
6-in. TC
500
12-in. TC
18-in. TC
400
300
200
100
0
0
10
20
30
40
50
60
70
80
90
Time (Min)
Fire Protection Research Foundation
Compact Mobile Shelving - Test 1
Ignition Aisle Shelf Steel Temperature
900
800
700
Steel temp 1-bay west of Ignition
600
(Deg C)
Steel temp above Ignition
500
TC1
TC3
400
300
200
100
0
0
10
20
30
40
50
Time (Min)
B-6
60
70
80
90
NC5756
07CA64050
Appendix B – Graphs
Issued 03/31/2008
Fire Protection Research Foundation
Compact Mobile Shelving - Test 1
Ignition Aisle Air Temperature
900
Air above Ignition - 8th Level
800
Air above Ignition
700
600
(Deg C)
TC2
500
TC4
Shelf 2 above
Ignition
TC5
400
TC6
TC8
300
200
1 Bay East of Ignition - 8th
Level
100
Air temp opposite Ignition (across
aisle)
0
0
10
20
30
40
50
60
70
80
90
Time (Min)
Fire Protection Research Foundation
Compact Mobile Shelving - Test 1
South Aisle Air Temperature
800
700
600
8th Level Air Temp (C)
(Deg C)
500
400
TC9
300
200
100
0
0
10
20
30
40
50
Time (Min)
B-7
60
70
80
90
NC5756
07CA64050
Appendix C – Graphs
Issued 03/31/2008
APPENDIX C
Graphical Presentation of
Temperatures and Waterflow - Test 2
C-1
NC5756
07CA64050
Appendix C – Graphs
Issued 03/31/2008
Fire Protection Research Foundation
Compact Mobile Shelving - Test 2
Sprinklers 1-6
90
80
70
60
(Deg C)
Spk.001
Spk.002
Spk.003
Spk.004
Spk.005
50
40
Spk.006
30
20
10
0
0
10
20
30
40
50
60
70
80
Time (Min)
Fire Protection Research Foundation
Compact Mobile Shelving - Test 2
Sprinklers 7-12
90
80
70
60
(Deg C)
Spk.007
Spk.008
Spk.009
Spk.010
Spk.011
50
40
Spk.012
30
20
10
0
0
10
20
30
40
Time (Min)
C-2
50
60
70
80
NC5756
07CA64050
Appendix C – Graphs
Issued 03/31/2008
Fire Protection Research Foundation
Compact Mobile Shelving - Test 2
Sprinklers 13-18
90
80
70
60
(Deg C)
Spk.013
Spk.014
Spk.015
Spk.016
Spk.017
50
40
Spk.018
30
20
10
0
0
10
20
30
40
50
60
70
80
Time (Min)
Fire Protection Research Foundation
Compact Mobile Shelving - Test 2
Sprinklers 19-24
90
80
70
60
(Deg C)
Spk.019
Spk.020
Spk.021
Spk.022
Spk.023
50
40
Spk.024
30
20
10
0
0
10
20
30
40
Time (Min)
C-3
50
60
70
80
NC5756
07CA64050
Appendix C – Graphs
Issued 03/31/2008
Fire Protection Research Foundation
Compact Mobile Shelving - Test 2
Sprinklers 25-30
90
80
70
60
(Deg C)
Spk.025
Spk.026
Spk.027
Spk.028
Spk.029
50
40
Spk.030
30
20
10
0
0
10
20
30
40
50
60
70
80
Time (Min)
Fire Protection Research Foundation
Compact Mobile Shelving - Test 2
Sprinklers 31-36
90
80
70
60
(Deg C)
Spk.031
Spk.032
Spk.033
Spk.034
Spk.035
50
40
Spk.036
30
20
10
0
0
10
20
30
40
Time (Min)
C-4
50
60
70
80
NC5756
07CA64050
Appendix C – Graphs
Issued 03/31/2008
Fire Protection Research Foundation
Compact Mobile Shelving - Test 2
Sprinkler System Flow
250
240
230
220
GPM
210
200
6" Flow
190
180
170
160
150
0
10
20
30
40
50
60
70
80
Time (Min)
Fire Protection Research Foundation
Compact Mobile Shelving - Test 2
Ceiling Steel Temperature above Ignition
Note: TC #1 malfunctioned
600
500
(Deg C)
400
AMB02
AMB03
AMB04
AMB05
300
200
100
0
0
10
20
30
40
Time (Min)
C-5
50
60
70
80
NC5756
07CA64050
Appendix C – Graphs
Issued 03/31/2008
Fire Protection Research Foundation
Compact Mobile Shelving - Test 2
Air Above Ignition
900
800
700
(Deg C)
600
500
6-in. TC
12-in. TC
18-in. TC
400
300
200
100
0
0
10
20
30
40
50
60
70
80
Time (Min)
Fire Protection Research Foundation
Compact Mobile Shelving - Test 2
Storage Array Steel Temperatures
NOTE: TC2 destroyed - 8th Level steel (B);
TC6 destroyed - 1st Level steel (F)
1000
Steel above Ignition
900
800
700
(Deg C)
600
TC3
1st Level steel (G)
TC5
500
TC9
400
300
200
8th Level steel (C)
100
0
0
10
20
30
40
Time (Min)
C-6
50
60
70
80
NC5756
07CA64050
Appendix C – Graphs
Issued 03/31/2008
Fire Protection Research Foundation
Compact Mobile Shelving - Test 2
Storage Array Air Temperatures
800
1st Level (C-D)
700
600
8th Level (E-F)
8th Level (C-D)
TC1
(Deg C)
500
TC4
TC7
400
TC8
TC10
300
200
8th Level (G-H)
100
8th Level (A-B)
0
0
10
20
30
40
Time (Min)
C-7
50
60
70
80