CATASTROPHE MANAGEMENT SERVICES
BRIEFING
Managing Severe Thunderstorm Risk
Exposure Data Details for Underwriting and Catastrophe Risk Management
Executive Summary
In May 2013, the Midwest witnessed major tornadoes causing widespread property damage, ranging from the loss of a few roof
shingles to complete collapse. The ability of buildings to withstand tornado damage varies widely and depends on numerous
factors, such as the intensity of the tornado, location of the structure with respect to the tornado path, age of the building,
construction type, secondary building attributes (e.g., roof type, roof age, roof shape) and construction quality. In particular,
when there are low wind speeds, specific building attributes correlate significantly with the damage to buildings.
Based on historical data since 1980, on average 86.6% of tornadoes occurring annually in the United States are of an intensity
less than or equal to EF1 strength (wind speeds less than 110mph) and 3.6% are of an intensity greater than or equal to EF3
strength (wind speeds greater than 135mph).
Less intense and more frequent tornadoes cause significant loss to property insurance portfolios on an aggregated basis year
over year. As such, using information related to detailed building characteristics in underwriting, risk quantification and the risk
management process can help minimize loss potential to an insurance portfolio from most reoccurring but less intense tornado
events.
This briefing summarizes several building damage patterns and building characteristics – which can influence tornado damage
potential to buildings – as observed by Willis Re’s damage reconnaissance teams from the May 2013 Midwest tornadoes as well
as by other research organizations.
Introduction
Two of the many key findings from our May
2013 field damage surveys include Tornado
Damage Isolines (TDIs) and the variation in
degree of damage (damage profile) across the
tornado path. TDIs are contours of comparable
building damage levels within a tornado path
and they can help us envision the impact of
tornadoes from a loss or building damage
perspective. TDIs represent an average damage
ratio (DR) to a typical residential building (1-2
story wood frame building with brick veneer
construction). Many buildings within the 60-80% TDIs range are damaged beyond repair and need total replacement. Because of
this, it is reasonable to assume a 100% loss within the 60-80% and 80-100% TDI contours.
© Copyright 2014 Willis Limited / Willis Re Inc. All rights reserved: No part of this publication may be reproduced, disseminated, distributed, stored in a retrieval system,
transmitted or otherwise transferred in any form or by any means, whether electronic, mechanical, photocopying, recording, or otherwise, without the permission of Willis Limited /
Willis Re Inc. Some information contained in this document may be compiled from third party sources and we do not guarantee and are not responsible for the accuracy of such.
This document is for general information only and is not intended to be relied upon. Any action based on or in connection with anything contained herein should be taken only
after obtaining specific advice from independent professional advisors of your choice. The views expressed in this document are not necessarily those of Willis Limited / Willis Re
Inc., its parent companies, sister companies, subsidiaries or affiliates (hereinafter “Willis”). Willis is not responsible for the accuracy or completeness of the contents herein and
expressly disclaims any responsibility or liability for the reader's application of any of the contents herein to any analysis or other matter, or for any results or conclusions based
upon, arising from or in connection with the contents herein, nor do the contents herein guarantee, and should not be construed to guarantee, any particular result or outcome.
Willis accepts no responsibility for the content or quality of any third party websites to which we refer.
A rapid decrease in the degree of damage moving out from the centerline of the tornado was observed in Moore, OK and in
Texas. For example, in the Moore tornado (EF5 intensity), damage to residential buildings was observed at 80% at about 0.1
mile, 10% at 0.25 miles and 1% at 0.35 miles out from the centerline of the storm. This variation in building damage reflects the
intensity of winds at various points across the storm path. The intensity of wind gusts decreases moving away from the
innermost TDIs. Locations in the outer bands of TDIs experienced significantly less than EF5 strength winds. Buildings with
certain damage mitigation characteristics and newer construction performed better relative to buildings with no damage
mitigation characteristics and older construction within outer TDIs.
Building Damage Observations
1.
In general, newer construction performed significantly better
than older construction in areas within outer TDIs / exposed
to less intense winds. Within the innermost TDIs, almost no
variation in the performance of newer and older construction
was observed as these locations were subjected to high suction
forces.
2.
It was widely observed that at low wind intensity levels
(outermost TDIs), the performance of brick veneer was better
than wood, metal or vinyl cladding. Brick veneer also showed
relatively good resistance to small wind-borne debris.
3.
Long span pre-engineered light metal buildings that are
typically used for warehouses, small manufacturing facilities
and churches to provide column free spaces tend to be more
Percentage of housing units by year
structure built
United States
Alabama
Colorado
Indiana
Michigan
Minnesota
Missouri
Oklahoma
Texas
Damage to trailer parks and mobile homes in general was very
high. Mobile homes with no tie-down mechanism or proper
anchors to the foundation failed significantly and were pushed
off their foundations.
5.
Engineered buildings, such as certain steel and concrete frame
buildings, performed better and withstood tornado winds
without structural collapse.
6.
Significant damage to exterior glazing and EIFS (Exterior
Insulation Finishing System) was observed in engineered high
Source: Willis Re
rise buildings. Significant damage to unreinforced masonry
(URM) infill walls was also observed in certain commercial and
industrial buildings.
7.
Metal roofing performed worse and reinforced concrete roofs,
composite concrete and steel roofs performed better than all
other roofing types.
8.
Loss of roof decks was observed widely and roof decks
separated from rafters or roof trusses due to the failure of
nails and / or inadequate nailing schedules between the
decking and rafter or truss to resist uplift forces.
9.
The performance of hip roof buildings was better than those
with gable roof in low intensity wind exposed locations. Gable
end wall failure was observed in many residential buildings
causing additional damage to the roof deck.
Page 2 of 4
30%
20%
20%
37%
39%
33%
32%
26%
17%
1960 to
1979
27%
30%
30%
26%
28%
26%
28%
33%
28%
1980 to
1999
28%
34%
32%
25%
23%
26%
27%
27%
33%
Source: U.S. Census Bureau, American Community Survey
damageable than wood frame construction.
4.
Pre
1960
Source: Willis Re
2000
or later
15%
16%
19%
13%
10%
14%
14%
14%
22%
10. The failure of roof truss-to-wall connections made walls
more vulnerable to collapse.
11. Significant damage to roof and wall attached equipment,
architectural elements like roof top HVAC units, entry
facades, carports, canopies, awnings, chimneys and others
was widely observed. Failure of these elements also caused
additional damage to structural elements to which these were
attached.
12. Significant damage to window and door glazing in
residential buildings caused increased internal
pressurization. Buildings within outer TDIs that were fitted
with insulated glazing units (i.e., double-paned windows)
performed better as the outer pane was sacrificed but the
Source: Willis Re
inner pane remained intact.
13. Many overhead garage doors failed and damage to older
double garage doors was significant.
14. In high intensity wind exposed locations (inner TDIs), 1 and 2
story residential buildings were lifted off of their foundations
due to either poor or no anchorage causing building
collapse in some cases.
15. Building damage from tree fall and flying debris was widely
observed. Large wind-borne debris like parts of roof trusses
and tree branches blew into nearby homes, causing significant
breach to the building envelopes.
Tornado Hazard
Source: Willis Re
A tornado generates a large vertical suction field producing
strong vertical uplift force; it also creates straight-line winds. Winds from EF0-EF2 intensity tornadoes are comparable to wind
hazard intensities in hurricane-prone regions. Tornadoes produce large amounts of wind-borne debris causing danger to life and
property. Tornado-generated wind-borne debris includes roof top gravel, pieces of tree limbs, appliances, furniture, heating,
ventilation and air-conditioning units, propane
tanks, trees and roof trusses among others.
Average annual number (percentage) of tornadoes by
Based on historical records from 1980-2013, the
United States has experienced 1,121 tornadoes per
year on average. As the table at right indicates, of
these, 971 were EF-0 and EF1, 110 were EF2, 32
were EF3 and eight were EF4 and EF5 intensity. In
other words, 86.6% of the tornadoes had wind gusts
estimated at 110 mph or lower and only 3.6% had
wind gusts of more than 135 mph. Clearly, the
frequency of less intense tornadoes is very high and
the frequency of catastrophic high intensity
tornadoes is low. On an aggregated annual basis,
less intense, more frequent tornadoes cause
significant loss to Property insurance portfolios.
tornado intensity (between 1980-2013)
United States
Alabama
Colorado
Indiana
Michigan
Minnesota
Missouri
Oklahoma
Texas
EF0 & EF1 EF2 (111 to EF3 (136 to EF4 & EF5
165 mph) (>165 mph)
(≤ 110 mph) 135 mph)
971 (87%)
32 (80%)
43 (96%)
18 (79%)
14 (83%)
31 (87%)
33 (85%)
49 (83%)
131 (89%)
110 (10%)
6 (14%)
2 (4%)
3 (15%)
2 (13%)
3 (9%)
4 (10%)
7 (12%)
12 (8%)
32 (3%)
2 (4%)
0 (1%)
1 (5%)
1 (4%)
1 (3%)
2 (4%)
2 (4%)
3 (2%)
8 (1%)
1 (1%)
0 (0%)
0 (1%)
0 (0%)
0 (1%)
0 (1%)
1 (1%)
1 (0%)
Numbers rounded to the nearest digits, so percentage numbers may not add up to 100%
Source: Storm Prediction Center, NOAA
Page 3 of 4
What You Can Do
Often, the collection of additional data related to building details requires additional resources and special inspections by
certified professionals. However, insurance companies can collect certain building characteristics, by visual inspections and via
simple questionnaires that policyholders can complete. A few examples of these easy-to-collect building characteristics that can
have significant impact on loss estimates include construction type, year built, height, roof shape, roofing type, roof age and wall
cladding type.
The performance of buildings during tornado events depends mainly on the location of the structure with regard to the tornado
path, construction / structure type, age of the building, construction quality and secondary building characteristics (e.g., roof
type, roof age, roof shape, connection types between structural components). Varying wind speed intensity has a bearing on the
impact of various building characteristics on the overall damage. Certainly, the impact of building mitigation factors on building
damage potential can be significant at wind gusts less than or equal to 135mph (EF2 intensity).
Underwriting guidelines should use information related to detailed building characteristics to help minimize annual loss
potential to an insurance portfolio from most reoccurring low intensity tornado events. As well, building details can be useful for
identifying poor performing risks and superior risks, establishing rating criteria.
As with underwriting, catastrophe loss modeling / loss estimates can also benefit from increased accuracy. Major commercially
available catastrophe risk models for Severe Thunderstorm (a.k.a. tornado and hail) include functionality that allows them to
take account of various building details in model loss estimates. The models include separate sets of secondary building
characteristics for the tornado and hail sub-perils within the severe thunderstorm peril.
In general, the more information we have at hand, the better equipped we are to understand and make decisions. Given what we
know about aggregate property loss based on tornado wind intensity and building characteristics, we highly recommend that
insurance companies begin to put a data improvement plan in place. Our Catastrophe Modeling Services department can provide
guidance on streamlining the process. Please contact us to see how we can help.
References
1.
2.
3.
4.
5.
6.
http://www.spc.noaa.gov/
http://www.census.gov/
Enhanced Fujita Scale, http://www.depts.ttu.edu/nwi/Pubs/FScale/EFScale.pdf
FEMA 2012, Spring 2011 Tornadoes: April 25-28 and May 22, Mitigation Assessment Team Report, FEMA P-908
Technical Investigation of the May 22, 2011, Tornado in Joplin, Missouri, Final Report, National Institute of Standards
and Technology (NIST), March 2014
May 2013 Tornadoes in Texas and Oklahoma – Willis Re’s post-event field damage survey report, June 2013
Contact us
Prasad Gunturi
Senior Vice President
Willis Re Inc.
7760 France Ave., Suite 450
Minneapolis, MN 55435
Phone: +1 952 841 6638
Email: [email protected]
The contents herein are provided for informational purposes only and do not constitute and should not be construed as professional advice. Any and all examples
used herein are for illustrative purposes only, are purely hypothetical in nature, and offered merely to describe concepts or ideas. They are not offered as solutions to
produce specific results and are not to be relied upon. The reader is cautioned to consult independent professional advisors of his/her choice and formulate
independent conclusions and opinions regarding the subject matter discussed herein. Willis is not responsible for the accuracy or completeness of the contents herein
and expressly disclaims any responsibility or liability for the reader's application of any of the contents herein to any analysis or other matter, nor do the contents
herein guarantee, and should not be construed to guarantee, any particular result or outcome.
Page 4 of 4