an evaluation of service weather shelter areas at procter
TRANSCRIPT
1
AN EVALUATION OF SEVERE WEATHER SHELTER AREAS AT PROCTER & GAMBLE FACILITIES
IN GREATER CINCINNATI OHIO
EXECUTIVE DEVELOPMENT
BY: Ray Webber The Procter & Gamble Company Cincinnati, Ohio
An applied research project submitted to the National Fire Academy as part of the Executive Fire Officer Program
January 2003
2
ABSTRACT
This project evaluated severe weather shelter areas at Procter & Gamble facilities
in greater Cincinnati Ohio. The purpose of this project was to determine if current severe
weather shelter area selection and design was adequate to protect employees during
damaging winds. This research project used evaluative research to determine (a) the
frequency of damaging winds in greater Cincinnati (b) accepted methods of selecting
weather shelters in existing buildings (c) the preferred method of designing shelters in
new buildings (d) how shelters were being designed at Procter & Gamble facilities and
(e) what level of protection these shelter areas afforded building occupants.
The primary procedure used in this project was a review of literature pertaining to
the frequency and severity of wind events in southwest Ohio and the construction and
designation of severe weather shelter areas. A survey of severe weather shelters at
company facilities was conducted and is discussed in the results section of this project.
Personal observations and field surveys of severe weather shelter areas also appear in the
results section. Construction features of these shelters were evaluated against applicable
standards, test data, and observed performance during severe weather incidents.
This research concluded that the performance of Procter & Gamble shelter areas
under actual tornado conditions would vary greatly, and that there could be no assurance
of safety for occupants while in these shelters. The recommendations from this project
were (a) to immediately re-evaluate shelter areas using nationally recognized criteria (b)
to request funding for formal engineering surveys of shelter areas, (c) to require
engineered shelter areas in the design of new facilities in high risk geographic regions,
3
and (d) to include information regarding the limitations of existing shelters in employee
training.
4
TABLE OF CONTENTS
PAGE
Abstract 2
Table of contents 4
Introduction 6
Background and Significance 7
Literature Review 9
Procedures 23
Results 26
Discussion 31
Recommendations 33
Reference List 37
Appendix A Hamilton County tornadoes 1950 – 1995 39
Appendix B Warren County tornadoes 1950 – 1995 40
Appendix C Butler County tornadoes 1950 – 1995 41
Appendix D Tornado activity in the United States 42
Appendix E The Fujita Scale 43
Appendix F Wind zones in the United States 44
Appendix G Risk assessment worksheet 45
Appendix H Load determination flowchart for shelters 46
Appendix I HCRC shelter areas 47
Appendix J MVL shelter areas 49
Appendix K GO shelter areas 50
6
INTRODUCTION
The Procter & Gamble (P&G) Company constructs and operates its facilities in
accordance with applicable laws, codes and industry standards. Numerous organizations
within this company review current and proposed facility design to ensure risks to people,
profits, and property are managed appropriately. The problem is that designated severe
weather shelter areas at Cincinnati Ohio based P&G buildings are not specifically
designed to protect occupants from tornadoes or damaging winds.
The purpose of this research project was to determine if current severe weather
shelter area selection or design was adequate to protect occupants during tornadoes or
other damaging wind events at Cincinnati based P&G facilities.
Evaluative research methodology was used to answer the following questions: (a)
What is the frequency of weather incidents in greater Cincinnati Ohio that produce winds
capable of damaging commercial buildings, and potentially injuring building occupants?
(b) How should severe weather shelter areas in existing buildings be selected? (c) How
should severe weather shelter areas be designed or constructed in new buildings? (d)
How are severe weather shelter areas currently designed or selected at P&G facilities in
Cincinnati, Ohio? (e) What level of protection does current shelter design and or
selection afford occupants from severe weather in Cincinnati, Ohio based P&G facilities?
7
BACKGROUND AND SIGNIFICANCE
The P&G company’s world headquarters in Cincinnati Ohio. The main
administrative offices as well as five major research and development campuses are
situated in and around Cincinnati. These and other facilities in greater Cincinnati house
approximately fifteen thousand P&G employees. The major P&G sites evaluated in this
project are (a) The General Offices, (b) Ivorydale Technical Center, (c) Winton Hill
Business Center (d) Sharon Woods Technical Center, (e) Miami Valley Laboratories, (f)
The Health Care Research Center.
In the past, P&G facility management personnel designated severe weather shelter
areas in these buildings. Recently, the individual site fire chiefs and risk managers took
on this task. The frequency of designating these areas was low, because most building
renovations were cosmetic and involved little structural changes. Prior to this time, the
individual building managers designated these shelter areas. The building manager role
no longer exists. This role stems from a time when individual buildings at a given
facility operated independently of each other. Facility management practices in one
building were typically quite different from those in another building situated only
several yards away.
Shelter areas were never specifically designed to protect occupants from severe
weather. Recently, the risk managers and site fire protection leaders (also called “Site
Fire Chiefs”) became increasingly uncomfortable with designating these areas because
they recognized the limitations of their experience with this subject, and the potential for
injury or death. (C.R. Burgher, personal communication, May 10, 2002) The individuals
8
designating these shelter areas could make no realistic guarantees of safety during severe
weather incidents. They had no technical data or other information that enabled them to
qualify the performance of any given shelter area during a damaging wind event.
Building and fire codes gave little more than vague recommendations regarding the
establishment of shelter areas. Shelter areas were established based simply what
individuals thought was the least dangerous place to be during a tornado or wind storm.
Even more recent building construction and renovations at P&G did not include
areas specifically designed to protect occupants from severe weather, despite the
documented concerns of risk managers and fire protection leaders.(R.A. Webber,
personal communication, January 8, 2002) Shelter areas are simply not required to be
constructed or designated in anything other than schools in the State of Ohio (Ohio Fire
Code[OFC], 1998, p.7-3) A section of the P&G “Design Guidelines for Corporate
Buildings” standard addresses the need for severe weather shelters. That document states
that any area built specifically (but not necessarily exclusively) for the sole purpose of
providing an area of shelter would exceed legal requirements. The document also states
however, that it is “good practice to ensure that areas exist within a building that can
provide adequate shelter for employees” (Procter & Gamble[P&G], 2002 p. 1)
Persons advocating the evaluation and construction of severe weather shelters
have no basis in code requirements. However, they are bound by corporate standards to
provide areas that afford “adequate” shelter for employees. As will be shown later in this
research project, weather shelter areas can exhibit varying degrees of “dimensions of
quality” as defined in the Executive Development student manual (National Fire
Academy, [NFA], 1998, p.10-25) There is little more than an intuitive sense on the part
9
of risk managers on how well shelters would protect occupants during a tornado or other
damaging wind event.
This applied research project relates to the service quality / marketing module
taught in the National Fire Academy’s Executive Development course. The terminal
objective of this module states that participants “shall be able to evaluate the services
provided by their organization and develop strategies to improve organizational quality
and service standards” (NFA, 2002, p. 10-2) This applied research project also relates to
the United States Fire Administration operational objective “To promote within
communities a comprehensive, multi-hazard risk-reduction plan lead by the fire service
organization” (NFA 2002, p. II.-2) by assessing the shelter areas used to protect
employees during severe weather.
LITERATURE REVIEW
The literature review contained herein will serve as the foundation for this applied
research project. The focus of this review is the five research questions that need to be
addressed. First, what is the frequency of weather incidents in greater Cincinnati,Ohio
that produce winds capable of damaging commercial buildings and potentially injuring
building occupants? Second, how should severe weather shelter areas in existing
buildings be selected? Third, how should severe weather areas be designed and
constructed in new buildings? Fourth, how are severe weather shelter areas currently
designed or selected at P&G facilities? And finally, what level of protection does current
shelter design and or selection afford occupants from severe weather in Cincinnati Ohio
based P&G facilities?
10
Regarding to the first question, there is a significant amount of literature available
that addresses the frequency and severity of weather incidents in southwest Ohio. Ohio is
situated on the eastern edge of what has commonly been called “tornado alley” (Ohio
Insurance Institute, 2001). The region in question includes the counties of Hamilton,
Warren and Butler, in southwestern Ohio. P&G operates major facilities in both
Hamilton and Warren counties. Miami Valley Laboratories is situated in Hamilton
county but is less than one mile from the Butler county line. From the period of 1950 –
1995, the National Oceanographic and Atmospheric Administration (NOAA) recorded
eleven tornadoes in Hamilton County (see Appendix A). These tornadoes ranged in
strength from F1 – F5. (National Climactic Data Center, [NCDC], n.d.). In the same
period, thirteen tornadoes were reported in both Warren and Butler counties (see
Appendixes B and C). Combined, these storms produced 683 injuries and 12 deaths
(NCDC, n.d.)
Southwestern Ohio is situated in a geographic zone that receives approximately
thirty to fifty tornadoes per 10,000 square miles each year. (GE Global Asset Protection
Services, 1998). This region includes states as far west as Colorado and as far south as
Texas. The average annual occurrences of tornadoes can vary greatly over relatively
short distances. According to this same data, the average annual incidence increases
sharply in south central Indiana, about 90 miles from greater Cincinnati. This area
records between 50 – 70 tornadoes / year / 10,000 square miles. Federal Emergency
Management Agency (FEMA) publication number 320, includes data from NOAA which
identifies the total number of tornadoes recorded per 1,000 square miles. Nearly all of
Ohio is in a region that has experienced between one and five total tornadoes per 1,000
11
square miles (Federal Emergency Management Agency, [FEMA], 1999). FEMA (2000)
summarized the occurrences of F3, F4 & F5 tornadoes between 1950 – 1998.
Southwestern Ohio, eastern Indiana, and parts of Kentucky have received more than 25
violent tornadoes per 3700 square miles(see Appendix D). When analyzed yet another
way, Cincinnati Ohio, along with Indianapolis Indiana, Lexington Kentucky, and
Owensville, Kentucky are all in a region which has received as many F3, F4 and F5
tornadoes as the more familiar areas of “Tornado Alley” in the plains states of Kansas,
Oklahoma, and Texas (the Fujita scale or “F” scale is presented in Appendix E).
Additional review of the NCDC’s database from 1993 to present, showed that in
Hamilton County, there were 52 thunderstorms or non tornadic wind events producing
winds of 50 knots (57.6 miles / hour) or greater. These events were responsible for two
deaths, 12 injuries and over $2.2 million in property damage. During this same time
period, there were two tornadoes, which combined were responsible for four deaths, 65
injuries and $82.2 million in property damage (NCDC, n.d.)
Similarly, from 1993 to present, Butler and Warren Counties combined have
recorded 3 tornadoes, and 93 wind events. These storms caused 4 deaths, 28 injuries and
over $9.7 million in damages (NCDC, n.d.)
Due in part to the statistical history of these severe winds, southwest Ohio is
situated in “wind zone IV”, the most extreme wind zone category in the United States
(see Appendix F for wind zones in the United States). Design wind speeds for a three
second gust in wind zone IV can be as high as 250 miles per hour (FEMA, 1999). When
determining the risk from a damaging wind event, in a given geographic area, both the
wind zone and the tornado activity are considered. Using the risk assessment worksheet
12
in FEMA 320, (see Appendix G) which is based solely on these two factors, we find that
the Cincinnati Ohio area is at a high risk for extreme winds. Shelter is the preferred
method of protection against damaging wind events (FEMA 1999).
An informal interview was conducted with Ed Ray, site meteorologist for the
Fernald Environmental Management Project (FEMP), to determine the frequency and
severity of non-tornadic winds in southwest Ohio. The FEMP is a United States
Department of Energy facility located in north west Hamilton County, about 20 miles
from the city of Cincinnati. Mr. Ray is tasked with short and long term weather
forecasting at this facility to guard against business interruption. His forecasts also serve
to protect workers and the surrounding community from severe weather and its effects on
the former nuclear facility. Mr. Ray indicated that straight line winds (non rotating
damaging winds) are not as closely tracked and recorded as tornadoes. These winds can
vary greatly in intensity and direction over a relatively small geographic area. Mr. Ray
pointed out however, that meteorological data of this type was tracked by the NCDC
through measurements are taken at the Greater Cincinnati / Northern Kentucky
International Airport.
The NCDC’s(2001) Meteorological Data for 2001 for the Cincinnati / Northern
Kentucky area was provided by Mr. Ray to the author. The information presented
indicates that the maximum five second wind gust recorded in 2001 was 61 miles per
hour, on July 8. The maximum two minute sustained wind speed was 48 miles per hour,
recorded on October 24. By comparison, an F0 tornado contains winds estimated to be
between 40 and 72 miles per hour (National Climactic Data Center [NCDC], 2001)
13
Wind speed alone however, is not an accurate predictor of tornado damage. In
1997 an F5 tornado struck the town of Jarrell, Texas. The estimated wind speed of this
storm was only 160 miles per hour. However, the storm moved only about four to five
miles per hour across the ground. Because this storm moved so slowly, exposing
buildings for a longer period of time, the damage caused was severe (Schultz & Metz,
2001. ¶ 26).
It is important to note here that the F scale, or Fujita scale does not directly
correspond to wind speed. Since the 1970s, the F scale has been used to categorize
tornado damage and imply tornado intensity. The F scale relies on damage to man made
structures. Thus, if a tornado occurs in an unpopulated area, it cannot be assigned an F
rating regardless of its size or intensity, because there was no damage to man made
structures (Thompson & Vescio, 1998).
The second research question asks how shelter areas should be selected in
existing buildings. The vast majority of this information comes from FEMA 361, Design
and Construction Guidance for Community shelters. Few other sources discuss in detail
how to select severe weather shelter areas. Appendix B of FEMA 361 contains a
number of site assessment checklists, one of which deals with identifying shelter areas in
existing buildings. FEMA tells us that the first step in selecting a shelter area is to
determine the maximum possible number of occupants at any given time who could use
the shelter. The next logical step, is then to look for available space to house those
occupants, while noting other potential dangers (FEMA 2000).
Once these potential shelters are identified, the screening criteria contained in
FEMA 361, is applied to those areas. This screening consists of detailed checklists and
14
questions regarding such items as structural composition, glazing, envelope protection
and non structural features. These items are assigned a weighted value and eventually
tallied so that a given shelter area receives a particular score. “The scores will identify
refuge areas that are candidates for retrofit designs as well as those that are poor
candidates because of excessive vulnerabilities” (FEMA, 2000, p. 2-10). Based on the
nature of these deficiencies, some areas may already provide substantial occupant
protection. Others however, may simply be unsuited to serve as shelters (FEMA, 2000).
The Metropolitan Emergency Managers Association (MEMA) of Minneapolis/
St. Paul Minnesota has also distributed guidelines for determining shelter areas in
buildings. In this guideline, the best and worst areas for shelters are delineated. Best
areas include: basements, rooms constructed of reinforced concrete, small interior rooms
with no windows, and any protected area away from doors and windows. Worst areas
include: auditoriums, rooms with large windows and doors, hallways exposed to the
direction of the tornado, and rooms beneath heavy roof mounted equipment
(Metropolitan Emergency Managers Association [MEMA], n.d.).
This guideline also includes a checklist similar to that included in appendix B of
FEMA 361. MEMA’s checklist also uses a weighted grading system to evaluate features
such as the shelter’s level above or below grade, its location within the building, glazing
surface area, wall construction, and ceiling span (MEMA, n.d). The checklist score and
shelter selection process rationale are explained in the following statement by MEMA
(n.d,):
15
Remember, you are simply trying to locate the safest area in your building. Even
if point totals are relatively low, the area with the highest total in your building
would still be much safer than being in a car, a mobile home or outside. (p.)
No widely used building code requires design standards for extreme wind events.
A building designed and constructed in accordance with local codes does not mean it
will withstand a tornado (Federal Alliance for Safe Homes, 2002). Code requirements
for shelter selection in existing buildings are virtually nonexistent. The Ohio Basic
Building Code (OBBC) applies to P&G facilities in greater Cincinnati. The OBBC is
actually a state customized version of the Building Officials and Code Administrators
(BOCA) model code and contains no guidelines for establishing shelter areas in existing
buildings. Similarly, the Ohio Fire Code only recommends that shelter areas be selected
in accordance with certain widely accepted criteria such as avoiding glass, and keeping
shelters on the lowest level of a building (OFC, 1998). These recommendations are only
referenced in the section applicable to schools. There are no shelter guidelines or
requirements for commercial, residential, or other institutional occupancies.
The third question the literature review shall address is how should severe
weather areas be designed or constructed in new buildings. Again, a review of the Ohio
Fire Code and the OBBC finds no requirements or shelter guidelines for facilities other
than schools. The OBBC contains a section dedicated to wind load consideration in
building design. No provisions are made for tornadoes or other damaging winds. The
basic wind speed design for buildings in Ohio is 80 miles per hour (Ohio Basic Building
Code, 1998).
16
The most in-depth reference was FEMA 361, Design and construction Guidance
for Community Shelters. Since the early 1970s studies have been conducted to determine
design parameters for shelters intended to provide protection against tornadoes and other
wind events. The results of this research, culminated in the publication of FEMA 320,
Taking Shelter from the Storm, Building a Safe Room Inside your House. While FEMA
320 is specific to small shelters, FEMA 361 provides guidance for the construction of
larger shelters. These types of shelters would be found in schools, hospitals, commercial
buildings and other facilities occupied by large numbers of people. The construction of
these shelters are of concern. The “May 1999 Building Performance Assessment Team
investigation of the tornadoes in Oklahoma and Kansas made it clear that a severe wind
event can cause a large loss of life or large number of injuries in high-occupancy
buildings…”(FEMA 2000, p.2-1).
The performance criteria established in FEMA 361 “are intended to minimize the
probability of death and injury during a high wind event by providing shelter occupants
with near-absolute protection.”(FEMA 2000, p.2-2). FEMA also states that the
information provided in their manual is believed to be the best, and most current material
available regarding shelter design. FEMA recognizes that no other known building, fire
or life safety code or engineering standard provides guidance, detailed information, or
recommendations to provide near-absolute protection to building occupants during high
wind events (FEMA 2000).
FEMA (2000) emphasizes the importance of determining the loads that will act on
a shelter area. Building materials and construction methods are very important in the
design of a shelter area. If the materials fail wind or debris may enter the shelter causing
17
the death or injury of its occupants. Therefore, shelter designers should select materials
and construction methods that will withstand the various loads that act upon a shelter.
The recommended methodology for determining those loads (and subsequent design
features) can be summarized as follows:
I. Determine the basic loads on a shelter
II. Determine applicable codes and standards for normal construction
III. Select the design wind speed for the shelter
IV. Calculate the extreme wind loads
V. Calculate seismic loads
VI. Determine appropriate load combinations
VII. Design shelter
VIII. Check exposed walls and roof areas for missile impact resistance
IX. Finalize design
(see Appendix H for a more detail in determining shelter loads)
Much attention is given to impact resistance of differently constructed wall
sections in FEMA 361. In this manual FEMA describes and illustrates a number of wall
sections that passed the impact resistance test at Texas Tech University’s WERC. In
these tests, a fifteen pound wood 2 X 4 is fired out of an air activated cannon into the test
wall section at one hundred miles per hour (FEMA 1999). The construction methods of
those wall section that were not penetrated by this missile varied greatly according to
FEMA (2000). Wall sections passing the impact resistance test included: plywood and
masonry infill, plywood and metal, expanded metal and sheet metal with plywood,
18
reinforced concrete walls, composite walls of reinforced concrete and brick, and
reinforced concrete masonry units (CMUs).
Engineering a wall section to withstand specifically the impacts associated with
windborne missiles is essential. “Even a reinforced masonry wall will be penetrated
unless it has been designed and constructed to resist impact during extreme winds”
(FEMA 1999, p.3-5). FEMA (1999) goes on to say that wind borne missiles can severely
damage buildings as well as threaten the safety of the occupants. Furthermore, research
at Texas Tech University WERC indicates that walls ceilings and doors commonly used
in ordinary construction cannot withstand the impact of missiles propelled by extreme
winds.
Research question number five was answered in part by conducting an informal
interview with Rachelle Burnside in November 2002. Mrs. Burnside currently serves as
the site risk leader at P&G’s Winton Hill Business Center. She also leads the P&G site
fire chiefs in setting policy for emergency planning, fire protection and emergency
response for corporate facilities in Cincinnati. Prior to this role, Mrs Burnside served as
the utilities manager at the General Offices from 1991 – 1995, and a building manager at
several P&G campuses from 1986 - 1991.
Mrs Burnside indicated that site fire chiefs and risk managers only recently were
tasked with designating shelter areas. In the 1980s, building shelter areas would have
been designated by building managers. Not all of these individuals had backgrounds in
engineering, and most had only a basic knowledge of building sciences. Mrs. Burnside
stated that since no codes existed to require severe weather shelters, building managers
19
had little guidance when designating them. They could only chose areas they personally
felt would offer some level of protection against severe winds.
Mrs. Burnside also indicated that prior to the 1980s there was very little safety
review of any changes in any P&G buildings. Effective change management is relatively
new to the company. She stated that this is why there was never any policy regarding
selection, construction or even the location of shelter areas. Each facility and each
building manager were left to do what they thought was best for their occupancy base.
Mrs. Burnside stresses however, that cost was not the reason that shelter areas were not
considered more carefully in the past. It was simply lack of information about the
subject.
Research question number five is not directly answered through the literature
review. To determine what level of protection current shelter areas afford, the author will
compare construction features of P&G shelters to similar features of buildings that were
in the path of severe winds. Literature pertaining to the performance of these specific
construction methods during actual and simulated wins storms will be reviewed.
In September of 2001, a tornado struck the University of Maryland in College
Park. Two persons were killed and property damage amounted to nearly $15 million
(Schultz & Metz 2001, ¶ 1). Schultz and Metz note that educational facilities such as
those at theUniversity of Maryland, often use makeshift storm shelters such as interior
corridors, storage rooms, restrooms, and other interior spaces (2001, ¶4). They go on to
say that these makeshift shelter areas often receive significant damage during tornadoes.
Walls collapse, roofs are blown off, and debris can be found in the very corridors
specified as refuge areas ( Schultz & Metz 2001, ¶5). FEMA’s Building Performace
20
Assessment Team (BPAT) conducted surveys of numerous buildings, including
educational facilities, after the May 3 1999 tornadoes that struck parts of Oklahoma and
Kansas. FEMA’s findings were consistent with those of Schultz & Metz.
The BPAT’s findings show a preponderance of damages associated with the
failure of roofing systems on commercial buildings. Wind pressure breached the
envelopes of multiple buildings, lifting the roof from the walls and creating numerous
damage patterns (FEMA, 1999). An inspection of buildings in Moore, Oklahoma with
tilt-up pre cast walls and a steel bar joist roofing systems, showed this phenomena. The
connections between the roofing systems and the tilt-up walls of many of these buildings
failed. In many cases, once the roof was removed, the end walls of the structure also
failed, sometimes collapsing into the buildings (FEMA, 1999).
The problem associated with uplift pressures was manifested in other wall and
roof systems as well. Load bearing masonry with steel joists, steel frame with masonry
infill walls, masonry walls with pre-cast hollow core floors, and light steel frame
buildings were all systems that showed damage associated with uplift forces. The modes
of failure include the failure of wall to roof masonry bonds, ductile iron fastener failure,
and column floor anchors withdrawing from concrete foundations (FEMA, 1999).
Schultz and Metz (2001) state that many educational facilities use hallways
constructed of non reinforced concrete masonry units (CMUs) for shelter areas. These
CMU wall sections can collapse, and can even be penetrated by wind generated missiles.
Roof systems over these hallways are typically of lightweight construction, and are
susceptible to collapse or uplift forces as described in FEMA 342. Once the roof of a
shelter area is compromised, shelter occupants are at risk to flying debris and storm
21
generated missiles (FEMA 1999). A survey of Kelly Elementary School in Moore
Oklahoma, illustrated this point. Kelly Elementary designated shelters in windowed
corridors that consisted of brick to a height of about seven feet. Glass extended from the
top of the masonry to the top of the wall. These hallways were completely destroyed by
the storm. Glass, concrete blocks, large sections of walls and other debris filled the very
area in which occupants would have been seeking shelter. “Obviously, had these
corridors been used for shelter during the impact of a violent tornado, numerous injuries
or deaths would have occurred” (FEMA 1999, p. 6-25).
Reinforced concrete however, fares better against wind pressure and storm
generated missiles. The BPAT survey of Westmoore High School in Moore Oklahoma
revealed that a 12 inch thick, 35 foot tall wall remained essentially undamaged. This
exterior wall was constructed of reinforced concrete, and was within 100 yards of the
tornado vortex. This particular tornado was classified as “violent” because it displayed
F4 and F5 damage characteristics (FEMA 1999).
Other areas of Westmoore High School were used successfully as shelter areas.
The building’s central core as well as a reinforced concrete stairwell sheltered occupants
during this tornado (FEMA 1999). There was no information regarding any injuries to
these shelter occupants.
Other common shelter areas such as storage areas and restrooms have limitations.
The walls and roofs of storage rooms may have the same structural problems as corridors
shown by FEMAs 1999 BPAT surveys in Oklahoma. Materials may fill storage rooms,
and moving those materials would waste valuable time when occupants are seeking
shelter. Restrooms, like storage areas and hallways, have walls that may not have been
22
designed to withstand a tornado’s winds. Doors, door frames and associated hardware
are typically not capable of withstanding wind pressure or missile impact associated with
tornadoes. (Schultz & Metz, 2001, ¶ 10).
Schultz & Metz (2001) and FEMA (1999) illustrate the vulnerabilities of
numerous construction methods and materials especially when they exist as features of
shelter areas. FEMA (1999, p.7-5) states
“Although some areas typically offer a relatively greater level of protection than
others, when people take refuge in a portion of a building that was not specifically
designed and built as a tornado shelter, they are at significant risk of being injured
or killed if a tornado of any intensity directly strikes the building or passes
nearby.”
In summary, the literature reviewed indicates tornadoes and thunderstorms
producing damaging winds are not uncommon in southwest Ohio. These events are well
documented and have claimed lives, injured people and caused millions of dollars in
damage in the greater Cincinnati area alone. While these events are somewhat less
frequent and severe than those in tornado alley, greater Cincinnati is certainly at risk for
tornadoes.
Definitive guidelines, while not required by code, exist to evaluate, and retrofit
shelter areas in existing buildings, and provide occupants with an increased level of
protection from severe weather. Similarly, several sources of technical information
provide very specific guidance for constructing shelters in new buildings. However,
building codes fail to require all but the most fundamental shelter guidelines.
Construction features common to P&G buildings have undergone testing and actual
23
tornadoes in other parts of the United Sates and performance of these materials and
methods vary. Failure is most often caused because these structural systems and
components are subjected to loads and impacts outside their design limitations. The
sources consulted indicate that shelter areas should be specifically engineered to protect
occupants from a given worst credible wind event.
PROCEDURES
The desired outcome of this applied research project was to determine if the
severe weather shelters at P&G facilities in Cincinnati provided adequate protection for
building occupants. Evaluative research was used to compare existing methods of
selecting and constructing shelter areas against recognized standards and guidelines. This
same method of research was also used to identify the risk to P&G facilities from
regional severe weather patterns and frequencies.
Research for this paper began with a brief literature review at the National Fire
Academy (NFA) Learning Resource Center (LRC) in July of 2002. The author then
submitted the applied research project proposal to the assigned evaluator in August of
2002. The evaluator provided general feedback to the author that indicated the
fundamental components of the applied research project were sound. A more
comprehensive literature review was then undertaken.
This literature review consisted of on-line sources, journal articles, conference
proceedings and papers, codes, engineering and construction standards. Internal P&G
current best approaches (CBAs), guidelines, procedures, and personal communications
were referenced to determine the overall policy for the selection of severe weather shelter
areas.
24
In order to answer the fourth research question; how are shelter areas currently
designed or constructed at P&G facilities in Cincinnati, Ohio, the author used a
combination of a survey, an interview and personal observations. A survey was
conducted to gather general information about severe weather shelter areas at the
different P&G facilities. The purpose of the survey was to determine who selected the
shelter areas, if the shelter areas were designed to protect against wind events, and if it
was felt these areas offered adequate protection to building occupants. The survey was
sent to the P&G site fire chiefs, the persons currently tasked with fire protection and life
safety at each facility. Five surveys were sent to each of the major P&G facilities in
greater Cincinnati. The author completed a survey for his own facility, the Winton Hill
business Center. Only two surveys were returned so the author contacted the remaining
three site fire chiefs via telephone, and recorded their responses to the survey questions.
An interview was conducted with Rachelle Burnside, the site risk leader for
Winton Hill Business center. Mrs. Burnside was chosen because of her background in
facility and risk management at P&G facilities. She provided information as to how
shelter areas historically were selected and how the current method of selecting shelters
came about.
The author then traveled to two other Procter & Gamble facilities in Cincinnati to
view shelter areas first hand. The General Offices and the Health Care Research Center
were selected because of their unique construction features and the large number of
persons occupying each site. The site fire chiefs from these facilities escorted the author
to answer any questions and clarify information gathered while reviewing the shelter
areas. Observations regarding shelter locations, size and construction features were
25
recorded, and sent back to the fire chiefs to check for accuracy. The author, being
familiar with his own facility, was able to recall the majority of shelter information from
memory. Several brief tours were taken to clarify ambiguous information on blueprints
and electronic drawings.
An informal interview was conducted with Ed Ray, site meteorologist at the
Fernald Environmental Management Project, in late November of 2002. Mr Ray was
interviewed to determine the frequency of non-tornadic winds in southwest Ohio. This
was undertaken because the author recognized that wind events other than tornadoes
might be capable of producing damage and endangering building and shelter occupants.
Although Mr. Ray was not able to provide much of this information first hand, he was
able to refer the author to a reliable source; the National Climactic Data Center.
The fifth research question pertaining to the quality of shelter areas at P&G
facilities, essentially draws upon answers from the first four research questions. The
author examines damage from previous wind events in buildings similar to those found at
P&G facilities. Construction features of existing shelter areas are also compared to
recognized shelter construction and selection standards.
There were limitations to this research. First, the author was not able to view
every single shelter area at every single site, due to time constraints. There are six major
P&G facilities in the Cincinnati area, most of which are quite large and complex. Winton
Hill Business Center (WHBC) alone comprises over one million square feet in 12
different buildings. Many P&G buildings are multiple stories with very similar floor
plans. In these buildings, shelter areas tend to be in the same location on every floor. For
the sites visited, the author surveyed only one of any identical shelter areas. This
26
research project is not a comprehensive evaluation of every single shelter area at every
P&G site in greater Cincinnati.
Another limitation of this research is that the author is neither a professional
engineer nor an architect. So, the review of P&G shelter construction features is in
general terms. This review is given from the author’s basic understanding of building
construction.
Finally, there was a general lack of information regarding shelter design,
construction or selection from any source other than FEMA. Much of the material
included in the literature review references FEMA 361, or 320. The only definitive
guidance for shelter design or selection came from FEMA.
The terms used in this research project are in plain language and the use of
technical jargon is avoided wherever possible. Any ambiguous terms can easily be
defined by consulting FEMA 361 or FEMA 320.
RESULTS
Using a literature review, four of the five research questions were answered. First, what
is the frequency of weather incidents in greater Cincinnati that produce winds capable of
damaging commercial buildings and potentially injuring building occupants? In general
terms, the maximum climatological probability of a violent (defined as F4 or greater)
tornado within 25 miles of any point in the United States during the course of a year is
about 5%(Concannon, Brooks, & Doswell 2000, pp. 2-3). While these violent storms
only represent about 2% of the total number of tornadoes, they are responsible for 67% of
all tornado related deaths. (Concannon et al., p.2).
27
An interview with meteorologist Ed Ray of the Fernald Environmental
Management Project yielded little direct information regarding the frequency of these
events. However, Mr. Ray was helpful in suggesting sources, such as the NCDC, which
proved to be valuable to this research. This information appears in the following
paragraphs.
According to the NCDC (2001) from 1950 – 1995 there were 37 tornadoes in
Hamilton Butler and Warren counties. In the past 10 years, these same counties
combined have recorded 5 tornadoes, and 145 events producing winds of 50 knots (57.6
miles per hour) or greater (NCDC, n.d.). On April 9, 1999, an F4 tornado struck the town
of Montgomery, in northeastern Hamilton County. This violent storm was responsible
for four deaths, 65 injuries, and over $82 million in property damage. The path of this
storm was only about five miles from P&G’s Sharon Woods Technical Center, on Reed
Hartman Highway.
How should severe weather shelters in existing buildings be selected? Although
there is little guidance in building and life safety codes, methods have been established to
select shelter areas in existing buildings. If a shelter must be specified in a building that
has no specifically engineered shelters, then any potential shelter areas must be evaluated
using a given set of criteria. Both FEMA 361 (2000) and MEMA (n.d.) provide a
weighted checklist to evaluate the performance characteristics of a potential weather
shelter. The MEMA checklist however, appears to be a simplified version of the
checklist provided in FEMA 361, appendix B. Each proposed shelter area receives a
numerical score which can be compared to other shelter areas. The highest point value of
any area evaluated would indicate the best location within a building for shelter (MEMA,
28
n.d.). Both FEMA (2000) and MEMA (n.d.) state however that some buildings may not
have any areas that are suitable as severe weather shelters, and that these checklists
should not serve as a substitute for a detailed engineering analysis.
How should severe weather areas be designed or constructed in new buildings?
The author found that, there is as little guidance in code requirements for new buildings
as there is for existing buildings. FEMA 361 (2000) states that the decision to design and
construct a shelter is based on a risk assessment process that includes, the nature,
probability, and severity of the event; in this case a tornado or damaging winds. Once an
architect, engineer or future building occupant decides that a shelter is the best way of
protecting occupants, detailed design and construction guidance can be found in several
FEMA publications.
FEMA 361, FEMA 320, and FEMA’s National Performance Criteria for
Tornado Shelters all give varying levels of detail regarding the construction of severe
weather shelters. These standards are developed by professional engineers and are based
on performance data from building materials and methods tested at Texas Tech’s WERC.
When constructed to these standards, shelters will provide occupants with near absolute
protection (FEMA, 2000).
How are severe weather shelter areas currently designed or selected at
P&G facilities in Cincinnati, Ohio? This question was answered in part using the survey
found in appendix B of this research project. According to survey respondents, all of the
facilities for which they were responsible had severe weather shelters. Five of the sites
indicated that the site fire chief was responsible for selecting the most appropriate shelter
in their facility. Four surveys indicated that more than one person has designated or had
29
designated severe shelter areas. One respondent indicated that a source defined as
“other” was exclusively responsible for this selection. None of the shelter areas at the
facilities surveyed were specifically designed to shelter occupants during tornados or
other severe weather. Question number four of the survey asks why occupants are
directed to go to certain areas if those areas not designed to provide shelter during severe
weather. The specifics of the responses varied, but all respondents indicated that
designated shelter areas represent the best place to send occupants during severe weather.
Five out o the six respondents indicated that the local authority having jurisdiction (AHJ)
did not require shelter areas in their facilities nor did they assist in establishing them.
Two site fire chiefs indicated that the local fire department assisted them in designating
shelter areas at their facility. Finally, five of the six site fire chiefs indicated that their
severe weather shelter areas provided adequate protection against severe weather.
The interview with Rachelle Burnside revealed that P&G never had a consistent
method to designate shelter areas. For many years, facility managers were left to make
the decision about shelter areas themselves. Mrs. Burnside indicated that the lack of code
requirements was a very significant factor in why engineered shelters do not exist at P&G
facilities. Risk assessments and safety reviews of facility designs are relatively recent
initiatives. According to Mrs. Burnside, building managers would have rarely researched
such issues as shelter selection criteria if they were not required by code. Facility safety
was only one of many responsibilities for a building manager.
P&G promulgates design standards and policies for its corporate owned buildings.
One of these standards; Severe Weather Shelters does provide some guidance for the
establishment of shelter areas. It should be noted that at the time this research project
30
was submitted, this standard was still in draft format. P&G states that severe weather
shelters should be established and identified in all corporate facilities (Procter & Gamble
[P&G], 2002). This standard does not require that shelter areas be specifically designed
to protect occupants against severe weather. It does however, offer six criteria for
tornado shelters. Shelter areas should be: large enough to accommodate building
occupants, below grade whenever possible, free of windows or glass, away from bearing
walls, away from rooms with long spans, and distant from tall structures such as towers
or chimneys (P&G, 2002). If implemented, this standard would be considered mandatory
for any new construction or significant renovation of P&G facilities.
The field observations of different severe weather shelters by the author showed a
wide array of construction techniques and materials. Shelters ranged from interior rooms
of drywall on metal studs, to restrooms and stairwells, to otherwise ordinary elevator
lobbies. Other shelters were large, below grade areas of reinforced concrete. Detailed
descriptions of shelter areas at four different P&G facilities can be found in appendix I -
M of this research project.
The final research question asks the following. What level of protection does
current shelter design and or selection afford occupants from severe weather in Cincinnati
Ohio based P&G facilities. To answer this question, the author compared construction
features common to both P&G facilities and to structures that had received tornado and
storm damage in other parts of the United States. Construction methods and materials
used in P&G shelter areas showed a wide range of performance when subjected to actual
severe weather events and tests at Texas Tech’s WERC. Reinforced concrete, which can
be found in some shelters at P&G’s General Offices, and the Health Care Research
31
Center, showed good windborne missile resistance, as well as resistance to wind pressure.
On the other hand, materials such as CMUs, commonly found in numerous P&G shelter
areas show poor resistance to the missile impact test performed at the WERC (FEMA,
2000). Hallways, stair towers, restrooms and utility cores at P&G are all used as shelter
areas. They have different methods of construction which exhibit different levels of
effectiveness against debris impact and wind pressure.
Referencing the information presented, the author concludes that there is at least
a reasonable threat of tornados in greater Cincinnati. Concannon and Doswell (2000)
indicate that there is no long term change in the threat from significant tornados.
Therefore, these events should be expected in the future. They are inevitible.
Shelter areas should be engineered specifically to provide protection for building
occupants in new as well as existing buildings. If the threat of severe weather is so high
that shelter areas are pre designated, they should provide occupants with a reasonable
expectation of safety. “Engineered shelters not only provide the best protection against
loss of life for individuals subject to a tornado, but also furnish the only protection
reliably capable of providing survival” (FEMA, 1999, p. 6-1).
Given the wide range of shelter construction methods and components, P&G
shelters would exhibit varying levels of performance if subjected to a tornado. Shelter
areas may adequately protect some occupants, while other areas may subject occupants to
risk of injury or death.
DISCUSSION
The literature review confirms that tornados and other damaging wind events pose
more than a remote threat to P&G facilities in southwestern Ohio. FEMA (1999) places
32
southwest Ohio in wind zone IV, the most extreme wind zone in the United States.
Furthermore, tornadoes and other wind events have caused multiple deaths and injuries
and caused millions of dollars in property damage in greater Cincinnati alone (NCDC,
n.d). There is more than enough evidence to suggest wind events should not be
discounted as potential sources of injury and death at P&G facilities in greater Cincinnati.
Like most other buildings in the United States, P&G facilities do not feature
engineered severe weather shelters, nor does the company require them in new
construction. Based on the responses from the surveys, it appears as if there may never
have been a formal evaluation of any shelter area at a P&G facility in greater Cincinnati.
FEMA provides in depth guidance for the construction and designation of shelter areas in
both new and existing buildings. It is the author’s conclusion that FEMA guidelines
represent the benchmark for shelter design and selection. Despite the quality of the
FEMA guidelines, they are not codified. Building codes simply do not require design
standards for extreme high wind events (Federal Alliance for Safe Homes, 2001).
Even though no codes require the establishment of shelter areas, P&G site fire
chiefs and risk leaders designate existing building areas as shelters. Site building
managers began designating shelter areas years ago when they deduced that some parts of
the building should be avoided during damaging winds. Schultz and Metz (2001)
indicate that commonly used shelter areas, similar to those found at P&G facilities may
not provide sufficient occupant protection when subjected to strong winds. Similarly
FEMA (1999) showed that some building elements similar to P&G shelter areas
performed poorly when subjected to severe winds. However, some P&G shelter areas are
composed of materials and employ methods that fared rather well at the WERC facility.
33
These materials also performed well in actual tornadoes such as the May 1999 tornadoes
in Oklahoma and Kansas.
There exists a wide variety of construction methods and materials in P&G shelter
areas. These materials and methods exhibit an even wider range of performance against
windborne missiles and wind pressure (FEMA, 1999). The author must conclude
therefore, that there can be no reasonable expectation of a consistent level of occupant
protection should a tornado of any size affect a P&G facility. Furthermore, it is the
author’s opinion that there exists a significant risk of death or multiple severe injuries for
many P&G shelter occupants should they be subjected to a severe or devastating tornado.
RECOMMENDATIONS
This author has been directly involved with the establishment of severe weather
shelter areas at P&G’s Winton Hill Business Center since 1998. Unfortunately, this
author designated many of these areas based on the only known references at the time;
vague recommendations in codes and corporate procedures. It is the recommendation of
this author that a project be established to begin to incorporate engineered weather
shelters in P&G buildings in southwest Ohio. This project should include: The
immediate re-evaluation of all existing shelter areas, funding for formal architectural or
engineering analysis of shelter areas, and accurately conveying the level of risk to shelter
occupants through employee training.
All shelter areas at P&G facilities should be re-evaluated using the criteria shelter
assessment worksheet in FEMA 361 ? . Existing shelter areas, as well as new potential
shelters should be evaluated. This may represent an opportunity to consider some areas
that may in fact be more suited to use as a shelter than thse already established. The
34
FEMA worksheet assigns a “score” to each potential shelter area that would allow site
fire chiefs or risk leaders to chose the best potential shelter areas evaluated. This
recommendation represents, in the author’s opinion, the bare minimum intervention
necessary to move toward a consistent level of occupant protection. The proposed shelter
assessment can add benefit by serving as a baseline of information for emergency
planning purposes. Selecting shelter areas based on this survey shows that P&G is using
a nationally recognized standard to protect building occupants, rather than a best guess by
a site fire chief or risk manager.
Next, funding should be provided for formal engineering assessments of proposed
and existing shelter areas. These assessments should also include recommendations as to
how to best fortify existing shelter areas so they provide a level of protection
commensurate with nationally recognized standards. Obviously, cost will be an
important factor in the recommendations of these assessments. Therefore, these
recommendations should represent options for providing near absolute occupant
protection, using the most cost effective means possible.
After these engineering assessments are completed, funding should be provided to
implement these formal recommendations and actually add or fortify shelter areas.
Documentation of this process would be extremely important. The P&G change
management program must recognize any shelter areas so that future building
renovations do not affect critical shelter features.
P&G must acknowledge the need for shelter areas in new and renovated corporate
facilities that are at risk to tornadoes and damaging winds. The P&G global design
guidelines for corporate buildings must address the risk of severe wind events by
35
geographic region, and then set specific shelter criteria based on that risk level. At the
very least, this document should indicate that FEMA 361 should be used to determine if
sheltering represents the best option, and if so, the standards to which shelters must be
designed. Based on the risk of severe weather, if sheltering is the best method to protect
occupants, then their inclusion in building design must not be optinal.
As soon as possible, building occupants must be made aware of the potential
varying performance of severe weather shelters. It is quite possible that many P&G
employees believe current shelter areas provide protection against any anticipated wind
event, which they do not. P&G must accurately convey to these employees that safety
cannot be guaranteed in these shelter areas during wind events. This recommendation
can actually help establish support for the proposed survey and subsequent fortification of
shelter areas. If building occupants understand that shelter areas may not provide the
level of protection they expected, they may convey their concerns to their management.
It is the author’s experience that managers and directors can have a significant influence
on the design of new and renovated P&G facilities.
The most logical place for the recommendations of this applied research project to
start is with the city wide site fire chief’s matrix. This group consists of P&G personnel
responsible for fire protection, life safety, emergency response, and emergency planning
at the major P&G sites in greater Cincinnati. The fire chief’s matrix is the group best
able to form a strategy that would result in the successful implementation of these
recommendations. It is anticipated that if any of these recommendations are
implemented, they will be preceded by a cost benefit analysis.
36
Future readers who may wish to replicate this study, or apply it to their own
organization would benefit from several recommendations. First, readers must be aware
that the vast majority of information regarding shelter construction and designation can
be found through FEMA. This agency seems to have compiled the most comprehensive
body of knowledge regarding shelter construction. Future researches must further
understand that this study found FEMA to be the sole credible source that provides
specific, detailed shelter design and construction information.
Second, a more in-depth analysis of shelter area performance would have only
been possible with a structural engineering or architectural expert. Readers wishing to
conduct future examinations of shelter areas at their facility would be well advised to
retain the services of such an individual. A registered professional engineer or architect
may be able to help a future reader make sense of the many complex engineering
standards and how they apply to common building practices. These individuals may also
be able to give authoritative opinions regarding the performance of existing shelter areas.
Those attempting similar research must consider that surveys of shelter areas can
be quite time consuming. This is especially true if the person conducting the survey does
not possess an in depth knowledge of architecture or structural engineering.
Lastly, one important element of this issue not addressed in this project is cost.
Future readers may want to consider cost implications of retrofitting, or designing in new
shelter areas to their facilities. A cost analysis associated with different shelter options
would be a very beneficial document.
37
REFERENCES
Concannon, P.R., Brooks, H. E., & Doswell, C.A. III, (2000, January), Climatological
risk of strong and violent tornadoes in the United States. Paper presented at the
Second Conference on Environmental Applications of The American
Meteorological Society. Long Beach, CA.
Federal Alliance For Safe Homes. (n.d.). Blueprint for safety. Retrieved November 11,
2002, from http://www.blueprintforsafety.org.
Federal Emergency Management Agency. (1999) Midwest tornadoes of May 3, 1999.
Building Performance Assessment Team Report. (FEMA publication no. 342).
Washington, DC.
Federal Emergency Management Agency. (1999). Taking shelter from the storm:
building a safe room in your house. (FEMA publication no. 320). Washington,
DC.
Federal Emergency Management Agency. (2000). Design and construction guidance for
community shelters. (FEMA publication no. 361). Washington, DC.
GE Global Asset Protection Services (1998). Windstorms. In GE GAP guidelines. (pp.
1-3). Hartford, CT.
Metropolitan Emergency Managers Association. (n.d.).Determining severe weather
shelters in buildings. Retrieved November 29, 2002, from
http://www.metroemergencymanagers.org.
National Climactic Data Center (2001). Local climatological data. Annual summary
with comparative data. Covington / Cincinnati. Available from
http://www.ncdc.noaa.gov.
38
National Fire Academy. (1998). Executive development. Student manual. Emmitsburg,
MD.
National Climactic Data Center. (n.d.). Storm events for Ohio. Retrieved December 1,
2002 from http://www.ncdc.gov.
Ohio Basic Building Code. Basic Wind Speed, section 1611.3. (1995).
Ohio Fire Code. Tornado Shelters in Schools, section FM-708
Ohio Insurance Institute. (2001). Ohio insurance facts. Retrieved December1, 2002
from http://www.ohioinsurance.org/factbook2001.
Procter & Gamble. (2002). Design guidelines for corporate buildings. Severe weather
shelters. Cincinnati, OH.
Schultz, C., Metz, J. (2001, October 1). Emergency management: shelter from the storm.
[Electronic version]. American School & University.
Thompson, R.L., Vescio, M.D. (1998, September) The destruction potential index – A
method for comparing tornado days. Paper prepared for the 19th Conference on
Severe Local Storms. Minneapolis, MN.
Wind resistant construction program may change noncoastal construction. (1998, July /
August). The Building Official and Code Administrator. 28-31.
43
APPENDIX E
THE FUJITA SCALE
SCALE WIND ESTIMATE *** (MPH) TYPICAL DAMAGE
F0 < 73 Light damage. Some damage to chimneys; branches broken off trees; shallow-rooted trees pushed over; sign boards damaged.
F1 73-112 Moderate damage. Peels surface off roofs; mobile homes pushed off foundations or overturned; moving autos blown off roads.
F2 113-157
Considerable damage. Roofs torn off frame houses; mobile homes demolished; boxcars overturned; large trees snapped or uprooted; light-object missiles generated; cars lifted off ground.
F3 158-206
Severe damage. Roofs and some walls torn off well-constructed houses; trains overturned; most trees in forest uprooted; heavy cars lifted off the ground and thrown.
F4 207-260
Devastating damage. Well-constructed houses leveled; structures with weak foundations blown away some distance; cars thrown and large missiles generated.
F5 261-318
Incredible damage. Strong frame houses leveled off foundations and swept away; automobile-sized missiles fly through the air in excess of 100 meters (109 yds); trees debarked; incredible phenomena will occur.
47
APPENDIX I
Procter & Gamble
Workplace Services HCRC Severe Weather Shelter Profile
Facility description: The Health Care Research Center (HCRC) is a major Procter & Gamble research and development center. It is located north of Cincinnati, off interstate 71 in the City of Mason Ohio. The major building access, the “Central Functions Spine” runs north-south along Mason-Montgomery Road. The Central Functions Spine serves as the connection to all other portions of the facility. The development Wing (DW), Process Research Lab (PRL) and Central Utilities Plant (CUP) are located to the east of the Central Functions Spine. The Discovery Wing (DS), and South Building (SB) are located to the west. Some of the distinct portions of this building include; office areas, laboratories, pilot scale operations, public assembly, and administrative and mechanical areas. The facility comprises approximately ??? square feet under roof. Construction and occupancy were completed in phases during the mid to late 1990s During normal business hours, 2500 or more persons may be present at the facility. Shelter plan: During severe weather emergencies, employees are to go to the lowest level of the facility to designated shelter areas. They are alerted to a severe weather emergency by the site security center. They remain in designated shelter areas until the weather event passes and it is safe for them to exit. Shelter descriptions: In the DS wing, the shelter is the first level mechanical area. This area is underground on the south side and on ground level on the north and west sides. The shelter area boundaries consist of the exterior walls that are reinforced poured concrete. The ceiling is the underside of the upper floor. It consists of poured concrete on a metal deck. There are several shelter areas in the SB wing. The lowest level of this wing is the ground floor. Shelters include a service hallway behind the facility cafeteria and a mechanical area. The service hallway walls are interior concrete block partition walls on one side and reinforced poured concrete on the other. The ceiling is poured concrete on a metal deck. The mechanical area walls are also made of concrete block. The ceiling is the underside of the upper floor which is poured concrete on a metal deck. There are no windows in the shelter area, however numerous windows exist in adjoining rooms and spaces. The 1st floor ground level serves as the severe weather shelter for the DV wing. This area contains the mechanical equipment for various building services and utilities. The exterior walls are above ground and made of reinforced poured concrete. The ceiling is the underside of the floor above and is constructed of poured concrete on a metal deck. There are no exterior windows in this area.
48
The CF spine shelter can be found on the east side of the 1st floor. The first floor is also the ground floor in this area. The exterior walls are above ground and made of reinforced poured concrete. The ceiling is the underside of the floor above and is constructed of poured concrete on a metal deck. There are no exterior windows in this area. This is a long, relatively wide open corridor on the east side with a mechanical area on the west. The shelter areas for both the PRL and CUP wings are in the basement. The reinforced concrete poured walls here are all below grade. The ceiling of the shelter area is the underside of the first floor, and is poured concrete on a metal deck. These spaces serve primarily as mechanical and storage areas.
49
APPENDIX J
MVL SEVERE WEATHER SHELTER AREAS
MVL has the capability to house about 950 people; currently there are about 650. There are a total of 10 severe weather shelter areas on site. They are located in the following areas: 1. Bldg. #8 0S – Hallway ~ 480 sq.ft. ~ 90 people 2. Bldg. #8 00N – Hallway ~1500 sq.ft. ~ 140 people 3. Bldg. #8 Stairway ~ 700 sq.ft. ~ 40 people 4. Bldg. #12 0F – Hallway ~ 520 sq.ft. ~ 95 people 5. Bldg. #11 0S – Hallway ~ 2200 sq.ft. ~ 240 people 6. Bldg. #11 Lower Aud. ~ 1500 sq.ft. ~ 190 people 7. Bldg. #38 Upper Aud. ~ 1200 sq.ft. ~ 50 people * (see note)* 8. Bldg. #22 1M – Hallway ~ 600 sq.ft. ~ 60 people 9. Bldg. #9 Main Conf. Room ~ 150sq.ft ~ 20 people 10. Bldg. #28 Internal Core ~ 75 sq.ft. ~ 4 people * Shelter area #7 is only used for the people that are already in that room when the severe weather evacuation alarms are sounded. This auditorium is directly next to the main lobby (atrium), which is all glass. After discussing this situation with Colerain Township Fire Department, it was decided that the people in the auditorium would be in less danger sheltering in place as oppose to leaving that area which would have them walking through the glass atrium. All the shelter areas have a supply cabinet located within, which include a bullhorn, clipboards, pens, paper, Evacuation Assembly Area check off sheets, an orange vest for the Evacuation Assembly Area Leader, and a flashlight. The shelter areas are generally located in the interior of the buildings on or preferably below the ground floor. It was a joint effort between CTFD, our AHJ, the Site Risk Manager and the Site Fire Chief to identify the locations for the severe weather shelter areas. With being located on or below the ground floor, the floors are poured concrete, the ceilings are constructed from steel I-beams with a metal deck then at least 4 inches of poured concrete, and the walls are either poured foundation walls or block, either cement or glazed tile. Another consideration, besides the type of construction and being on or below the ground floor, in determining if an area would be suitable for a shelter area is looking at the potential hazards such as: natural gas, flammable/combustible chemicals, high voltage electricity (MCC’s), and heavy objects that could fall over trapping or causing injuries.
50
APPENDIX K
Facility and severe weather shelter description Procter & Gamble General Offices
Cincinnati Ohio
Facility description: The Procter & Gamble general offices (GO) located in Cincinnati, Ohio serve as the company’s world headquarters and base of operations. The facility is comprised of five interconnected buildings bounded by Sentinal street to the east, seventh street to the north, Sycamore street to the west, and Fifth street to the south. The buildings in this complex are referred to as the Central Building, the Sycamore Building, the Terrace, Tower East, and Tower North. An atrium connects Tower East and Tower North from floors one through five. The GO is owned and managed by the Procter & Gamble Company. The company also leases office space in the Polk and Chemed buildings which are situated at Fifth and Pike streets and Fifth and Sycamore streets respectively. These buildings are physically removed from the GO and are occupied by Procter & Gamble and other businesses as well. Shelter plan: During severe weather emergencies, employees go to the closest severe weather shelter area. They are alerted to a severe weather emergency via the site security center. They remain in these areas until the event passes and it is safe for them to exit. Shelter descriptions:
Central Building
The Central Building is an eleven story office complex that houses approximately 1600 to 1700 persons. The floor layout varies from floor to floor corresponding with different remodeling projects. Most floors are of the open office design, containing non structural, partial height office partitions. Glass windows are on all four sides of the structure on all floors. Designated shelter areas are in the elevator lobbies, stairwells, restrooms, the basement, and ceiling high offices and conference rooms. The upper floor elevator lobbies are open to the hallway. The ground floor elevator lobby is open to the main building lobby. The main building lobby walls are primarily constructed of glass. The two stairwells used as shelter areas run from the basement to the to floor. The walls are constructed partially of reinforced concrete of un-determined thickness, and partially of concrete masonry units. Each floor has a self closing, latching 1 hour rated fire door. Restrooms in this building also serve as shelter areas. The wall construction is undetermined but is believed to be concrete masonry units covered with dry wall. Restroom doors do not latch. Several conference rooms as designated as shelter areas. All conference rooms that are designated as shelter areas have walls that run from floor to ceiling. Generally, these walls are constructed of drywall on metal studs. The entire basement of this building is a designated shelter area and is completely below grade.
51
Sycamore Building
The Sycamore Building is an 11 story office building housing approximately seven hundred persons. As with the Central Building, the safe areas here consist of the elevator lobbies, rest rooms, one stairwell and ceiling high offices and conference rooms. The elevator lobbies in the Sycamore building are constructed of concrete and are on the interior of the building. There are no windows within sight of the lobbies. Restrooms in this building also serve as shelter areas. The wall construction is undetermined but is believed to be concrete masonry units covered with dry wall. Restroom doors do not latch. The northwest stairwell of this building is also a designated shelter area. The stairwell is of undetermined construction but believed to be reinforced concrete. Stairwell doors are self closing, latching one hour rated fire doors. Offices and conference rooms with floor to ceiling walls are used as severe weather shelter areas. The walls of these rooms are of undetermined construction but believed to be drywall on metal studs. The entire first floor exterior walls are glass. These windows are coated with a kevlar film for security concerns, but were not specifically designed to afford protection against severe weather. Terrace The terrace is the basement of the GO parking garage located on Seventh street. The entire terrace is windowless and below grade and is entirely constructed of poured reinforced concrete. The entire terrace is a designated severe weather shelter area. Tower The tower building is actually two office towers connected by an atrium. Tower North (TN) and Tower East (TE) are both seventeen stories and are connected on floors one through six by the atrium. The combined population of the towers is 1500 – 1700 persons. Exterior walls of the tower building are almost all glass. The footprint of both towers and the atrium comprise the tower lower level in the basement. The lower level is all below grade and is a severe weather shelter area in its entirety. The lower level also contains several large training rooms which when filled to capacity simultaneously, could hold approximately one thousand persons. The grade level of the tower building contains the auditorium. The auditorium seats approximately five hundred people and is a severe weather shelter. The walls are of undetermined construction but are believed to be poured concrete. Doors to the auditorium are heavy decorative wood and do not latch. There are no exterior windows to the auditorium. The ceiling of this space is of undertemined construction but is believed to be pre-cast concrete. The ceiling is the lower level for the floor of the cafeteria which is situated directly above it. Other shelter areas in the tower buildings include restrooms, which are on every floor. The restroom walls are drywall on metal studs and the doors are not latching. There are no exterior windows to the restrooms.
52
Elevator lobbies on each floor are also shelter areas. The lobby wall construction was not determined but is believed to be either pre-cast or reinforced concrete. The lobbies are open on two sides to the open office arrangement of the tower buildings. Small conference rooms are distributed throughout on all of the floors. These conference rooms have floor to ceiling drywall walls on metal studs with mostly non latching doors. Many of these conference rooms in the tower building are designated shelter areas.
53
APPENDIX L
Facility and severe weather shelter description Procter & Gamble Winton Hill Business Center
Cincinnati, Ohio
Facility Description: The Winton Hill Business Center is a major Procter & Gamble research and development center located within the city limits of Cincinnati Ohio. The buildings are situated on both sides of Center Hill Avenue, essentially splitting the site into halves. The two halves of the facility are connected by two pedestrian tunnels that run under Center Hill Ave. Several service drives run off of Center Hill Avenue and provide access to the site. This 12 building campus style facility consists of offices, laboratories, and pilot scale manufacturing processes. Most buildings contain combinations of offices and labs and were constructed between 1959 and 1979. The Food and Beverage, Family Care, and Fabric and Home Care global business units (GBUs) all have research and development operations at this facility. Shelter plan: During severe weather emergencies employees are alerted by the site-wide alarm system and instructed to go to the closest severe weather shelter. They remain in these areas until the severe weather event passes and it is safe for them to exit. Shelter descriptions: International Building This two story building with a basement is of ordinary construction. The walls are of brick on either poured concrete or concrete masonry units (CMUs) and the roof is a metal deck on open web bar joists. Interior partition walls are of CMUs, poured concrete or drywall on metal studs. The entire basement of this building is a shelter area. The basement consists of mechanical areas, small offices and laboratories, and a maintenance shop. The entire basement is below grade and there are no exterior windows. Foods Building This four story building with a basement is of ordinary construction. The walls are of brick on either poured concrete or concrete masonry units (CMUs) and the roof is a metal deck on open web bar joists. Interior partition walls are of CMUs, poured concrete or drywall on metal studs. Shelters exist throughout in mechanical cores restrooms and stairwells. Service Building This is a two story building of ordinary construction that houses the boiler plant. The walls are of brick on either poured concrete or concrete masonry units (CMUs) and the roof is a metal deck on open web bar joists. Interior partition walls are of CMUs, poured concrete or drywall on metal studs. The only shelter exists in the lower level hallway, outside the maintenance shop. Administration Building This is a three story building of ordinary construction that houses the boiler plant. The walls are of brick on either poured concrete or concrete masonry units (CMUs) and the roof is a metal deck on open web bar joists. Interior partition walls are of CMUs, poured concrete or drywall on metal studs. This building now features extensive open office space. A cafeteria is situated on the
54
bottom level and an auditorium on the first floor. Shelters exist in the stairwells, restrooms and the lower level hallway. Fem Care Building This is a three story building of ordinary construction that houses the boiler plant. The walls are of brick on either poured concrete or concrete masonry units (CMUs) and the roof is a metal deck on open web bar joists. Interior partition walls are of CMUs, poured concrete or drywall on metal studs. Shelters exist in the lower level hallway as well as in the restrooms and a few selected interior rooms. Mechanical cores also serve as shelter areas. Family Care West Building This is a four story office building. Large glass and metal panels make up the exterior skin of the building. These are held in place with structural steel framing members. Interior walls are of drywall on metal studs. There is very little masonry in this building. Shelter areas are in the three stairwells (which are concrete block) and several restrooms. Family Care East Building This building is partially vacant, and undergoing renovation. It is of similar construction to the foods and international buildings. Prototype Development Facility (PDF) This is a lightweight steel single story building. The building consists of one main wide open space used for prototype manufacturing. Walls are of drywall on metal studs. The shelter is the loading dock area and the restrooms just off the dock. Engineering Building
55
APPENDIX M
SAMPLE SURVEY
Severe weather shelters in Cincinnati based P&G facilities Survey
1) Does the facility or facilities for which you are responsible have designated severe weather shelter areas?
Yes No
2) Who selected the existing shelter areas in the buildings for which you are responsible?
Site Fire Chief Site Risk Manager Architect / Engineer Other Unknown
3) Are these shelter areas specifically designed to provide protection against severe
weather including tornadoes?
Yes No If yes, were they designed to a specific standard for severe weather shelters, and if so, what standard was used?
4) If these shelter areas were not specifically designed to provide protection against severe
weather, how were they selected? In other words, why are persons directed to go here
verses other portions of a building?
5) Did the local authority having jurisdiction (ie, fire or building department) require shelter areas to be established in existing structures? If so, did they assist in their selection?
6) Do you feel the shelter areas at your facility would provide building occupants with adequate protection during a severe weather event (ie, a reasonable expectation of avoiding death or serious injury)?