---
Performance of Residential 
Buildings (Flood and Wind), 
One-to Two-Family and 
3Chris Jones 
Larry Tanner 
David Johnston 
Dave Low 
Mark Matulik 
Tom Smith 
Wallace Wilson 
Multi-Family 

Assessing the structural and building envelope performance of 
residential buildings was one of the main goals of the MAT. 

3.1 Structural Performance 
Assessing the structural and building envelope performance of residential buildings was one of 
the main goals of the MAT (the other being the assessment of critical facility performance—see 
Chapter 4). Making these assessments required location-specific information, gathered prior 
to and during the MAT’s field investigations, and knowledge of the flood and wind loads and 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 

3-1 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

conditions to which the buildings were exposed during Hurricane Ike. In a few cases, additional 
data were gathered after field work was completed, but in most cases building performance 
judgments were based on information available to the MAT while in the field. Although the 
MAT believes its assessments of buildings described in this chapter are correct, statements made 
herein are not intended to represent final judgments as to the cause of damage to individual 
buildings—the MAT recognizes that further investigation by others may refine or alter judgments 
made by the MAT. Nevertheless, general damage patterns and trends observed by the 
MAT are valid and can be used as the basis for recommendations to improve residential design 
and construction. 

3.1.1 Foundation Performance 
Foundations in coastal areas must be able to perform several functions: 

n Elevate the building above the surge and wave crest level 
n Remain intact and functional despite scour and erosion effects 
n Provide a continuous load path from the elevated building to the ground, and resist all 
vertical and lateral loads transferred from the elevated building to the foundation 
n Resist flood loads—including storm surge, wave, and floodborne debris impacts—acting on 
the foundation and on any below-flood level obstructions that do not break away 

Failure to perform any of these functions can result in building damage or loss. The MAT observed 
foundations that performed well (Figure 3-1), and foundations that failed to satisfy one 
or more of the requirements listed above. 

Failures of the most common type of foundation observed by the MAT—the open (e.g., pile or 
column) foundation—were usually associated with one of two factors: insufficient embedment 
into the ground, or breakage of the piles or columns. 

Embedment Failures. Embedment failures occur where a foundation is not deep enough in the 
ground to resist wind and flood loads pushing on the structure; a leaning foundation or overturned 
building results (Figure 3-2). Scour and erosion can exacerbate this mode of failure by 
reducing embedment. 

Pile and Column Breakage. Pile and column breakage occur where the strength of the piles or columns 
is inadequate to resist the bending moments or shear forces caused by the flood and wind 
loads acting on the structure (Figures 3-3 and 3-4). Scour and erosion contribute to this mode 
of failure by increasing the un-braced pile/column length and by increasing the bending moments 
in the pile/column. 

The methods used to secure an elevated building to the top of the foundation can affect the 
overall foundation strength. Connections at the tops of the piles or columns that do not provide 
fixity (i.e., resistance to rotation) allow greater stresses to develop in the piles or columns than 
would develop with connections that rigidly tie the structural elements together. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

In most buildings the MAT evaluated, timber construction was used and the tops of the piles or 
columns were connected to the elevated buildings with bolted connections (Figure 3-5). This 
type of connection provides limited fixity; weakness in this type of connection can be overcome 
in some instances through the use of larger piles or columns and other design details that help 
to stiffen the foundation. 


Figure 3-1. 
Louisiana house 
sufficiently elevated on a 
foundation that withstood 
Ike flood loads 

Figure 3-2. 
A house on timber piles 
was pushed over by wind 
and flood loads and the 
load path failed at the 
connection between the 
floor beam and the piles. 
Embedment and elevation 
were also insufficient at 
this Bolivar Peninsula, TX, 
site. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-3.
Broken timber piles
(Galveston Island, TX)


Figure 3-4.
This concrete column
failed due to lateral loads.
Note limited overlap of
reinforcing steel at the
bottom of the column.



MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-5. 
Typical bolted connection 
between wood columns 
and wood beams 

3.1.1.1 Foundation Function 1: Elevate the Building 
Elevation is one of the most important keys to a successful coastal building. The MAT observed 
many residential buildings along the Gulf shoreline that were elevated above the effects of Ike’s 
storm surge and waves, and sustained no significant damage; on the other hand, nearby buildings 
that were at lower elevations were heavily damaged or destroyed (Figure 3-6). 

Figure 3-6. 
Unlike the older and 
lower house on the 
right, the Zone V house 
on the left was elevated 
approximately 5 feet 
above the 16-foot National 
Geodetic Vertical Datum 
(NGVD) BFE and sustained 
no structural damage due 
to flooding (Crystal Beach, 
TX) 


HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

It was difficult to obtain HWMs for flood levels 
on Bolivar Peninsula due to the magnitude of The MAT’s observations of wave damage 
the destruction there. However, the MAT was and analysis of building elevation data 
able to supplement high water mark data collect-indicate that the wave crest elevation on 
ed by government agencies (see Section 1.2.3) much of the Bolivar Peninsula reached 
with wave damage data in elevated houses that approximately 2 to 5 feet above the BFE. 
remained standing after Ike. For instance, houses 
such as the one shown in Figure 3-6 indicate that the wave crest elevation at that location was 
below the bottom of the elevated floor system (due to the fact that the floor system was intact). 
Damage such as that shown in Figure 3-7, where the shore-parallel floor joists were displaced 
landward, indicates the onset of wave damage to an elevated floor system. By carefully examining 
several such houses and by acquiring the corresponding lowest floor elevations from NFIP 
Elevation Certificates, the MAT concludes that the wave crest elevation reached approximately 
18 to 20 feet NGVD, an elevation approximately 2 to 4 feet higher than the BFEs at these particular 
houses. Although it is possible that the wave crest reached higher elevations relative to the 
BFE, it is unlikely based on the lack of wave damage at some houses that were approximately 5 
feet above the BFE (waves apparently passed beneath those elevated houses). 

Figure 3-7. 
The landward 
displacement of shore-
parallel floor joists 
indicates the onset of 
wave damage to an 
elevated floor system 
(Bolivar Peninsula, TX) 


The MAT also observed many bay shoreline and inland examples that demonstrate the importance 
of elevation. Houses situated at higher elevations, whether because of elevated foundations 
or because of being sited on high ground, sustained little or no damage, while adjacent houses 
with lower elevations were damaged or destroyed. In many cases, undamaged bay-front houses 
were elevated above the surge and wave elevation on pile foundations. Figure 3-8 shows a house 
elevated above the BFE and Ike wave effects that suffered no significant damage due to flooding. 
However, the nearby at-grade house shown in the inset was heavily damaged. On some 
bay-front shorelines (or inland areas) where storm wave heights were smaller and where erosion 
did not threaten a house, siting on natural high ground or fill provided the required elevation 
and support for the house (Figure 3-9). 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

The MAT observed many houses in more inland locations that were not elevated high enough 
to avoid Ike flooding, and were apparently subject to surge inundation, low-velocity storm surge 
flow, and, in some cases, minor wave action. These houses sustained varying degrees of flood 
damage depending on site-specific flood depths, flood loads, and construction details. Some 


Figure 3-8. 
This elevated house (Zone 
V, BFE = 15 feet) suffered 
no significant damage due 
to flooding. The nearby 
at-grade house in the 
background, shown in 
the inset, was heavily 
damaged (Baytown, TX). 

Figure 3-9.
Adjacent houses south of Baytown, TX. The house on the left (Zone X) was above the surge and wave runup level
and sustained no flood damage. The house on the right (Zone A, BFE = 13 feet) was at a lower elevation and was
largely destroyed by surge, waves, and floating debris.


HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

were inundated by several feet of flooding, as they had been during Hurricane Rita in 2005. The 
house shown in Figure 3-10 was not subject to wave action during Ike and suffered no apparent 
structural damage. However, flood damage to contents and finishings were likely severe. Other 
houses sustained significant structural damage due to storm surge flow (Figure 3-11). Some 
floated or were washed off their foundations (Figure 3-12). 

Figure 3-10. 
The Ike flood level reached 
approximately 3 feet 
above the floor slab (1 
foot above the 6-foot 
BFE) of this Zone A house 
(see inset), which was 
reported to have been 
similarly inundated during 
Hurricane Rita. The MAT 
was told that the house 
will be elevated (Lake 
Charles, LA). 
Figure 3-11. 
This house sustained 
significant structural 
damage due to storm 
surge and small waves 
above the 9-foot BFE in 
Zone A (Bridge City, TX). 
Note flood debris line on 
the roof. 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


The tallest residential foundations the MAT 
observed were at Fortified... for safer living® 
houses (see text box and Section 2.4) on Bolivar 
Peninsula. The houses are elevated with 
their lowest floor at approximately 27 feet 
NGVD (21 feet above the ground), 10 feet 
above the BFE (Figure 3-13). These foundations 
are reinforced, cast-in-place concrete 
columns connected to concrete slabs and 
drilled concrete shafts (extending 10 feet below 
grade). Ten of the 13 houses survived Ike, 
and three were destroyed. 

Figure 3-12. 
This house floated off 
its foundation due to 
insufficient elevation and 
inadequate connections 
between the foundation 
and the house (Golden 
Meadow, LA) 


WARNING 

Elevation alone is not adequate to ensure a 
building will perform well during a high wind 
and flood event. A building must be elevated 
on a well-designed and constructed foundation. 
Some of the tallest foundations the Ike 
MAT observed either failed or were in danger 
of failing. 

The houses had substantial timber decks connected to the columns at or just above the BFE, 
approximately mid-way between the ground and the elevated houses. Although not designed 
as breakaway decks, the decks broke away during Ike, probably a result of both wave and flood-
borne debris effects. The deck failures damaged some of the concrete columns where the decks 

were connected (Figure 3-14). 

The concrete columns left standing between 
the slabs and the (destroyed) elevated decks 
were observed to have a series of horizontal 
cracks in the columns (Figure 3-15). These 
cracks likely resulted from the columns bending 
in response to a combination of wind loads 
on the elevated houses, flood loads (waves, 
currents, debris) on the columns, and transfer 
of flood loads from the decks to the columns. 

FortiFied... For saFer living® 

The Fortified… for safer living® designation 
is from the Institute for Business and Home 
Safety. The “Fortified®” program specifies design 
and construction guidelines to increase 
a house’s resistance to natural disasters 
such as hurricanes. For more information: 

www.disastersafety.org/text.asp?id=fortified. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-13. 
Looking toward the Gulf, 
past Zone V houses on 
tall concrete column 
foundations (with the lowest 
floor 10 feet above the 17foot 
BFE). Four of the five 
tall houses shown in this 
photograph survived Ike (red 
circle indicates destroyed 
house). The red arrow points 
to exposed geotextile tube 
(under former dune). Note 
other destroyed houses (not 
on tall foundations) seaward 
of the highway (Bolivar 
Peninsula, TX). 


Figure 3-14. 
Ground-level view of 
elevated houses with inset 
showing typical column 
damage where the timber 
deck broke away (Bolivar 
Peninsula, TX). 
MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-15. 
Concrete column showing cracking that 
was likely caused by extreme column 
bending stresses due to lateral loads 
on the elevated house and foundation 
(Bolivar Peninsula, TX) 

3.1.1.2 Foundation Function 2: Resist Scour and Erosion 
Residential building performance in coastal areas often depends on the capability of the building 
foundation to accommodate a lowering of the ground elevation and loss of soil support. 
The lowering of the ground is often accompanied by high winds, storm surge, large waves, and 
debris propelled by wind or water, which further magnify any adverse effects of soil loss. 

For foundation design purposes, it is important to distinguish the nature and extent of soil loss 
expected around a building, since these can affect the stillwater flood depth and the magnitude 
of the flood conditions at the site (see erosion and scour text box on next page). 

The MAT observed significant levels of erosion and scour near buildings situated along the 
Gulf of Mexico. Erosion was widespread along the Gulf of Mexico shoreline of Follets Island, 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Galveston Island, and Bolivar Peninsula, TX, and portions of southwest Louisiana. Scour was 
particularly evident around building foundations on Bolivar Peninsula and at Holly Beach 
(Cameron Parish, LA). The MAT believes that erosion and scour were among the major contributors 
to structural failure of buildings close to the Gulf shoreline. Significant erosion and 
scour were not observed by the MAT along the bay shorelines, although there may have been 
some locations where such erosion and scour occurred. 

EROSION AND SCOUR 
Erosion is a lowering of the ground surface over a large area, usually brought on by a coastal storm or 
long-term shoreline recession. Erosion increases the unbraced length of vertical foundation elements 
and increases the stillwater depth at the building, allowing larger waves to reach the foundation. 
Scour is a localized loss of soil immediately around an object or obstruction. Scour also increases the 
unbraced length of vertical foundation elements, but does not act to increase the stillwater flood depth 
across which waves propagate (thus, scour can be ignored for wave height calculation purposes). 
Walls, columns, pilings, pile caps, footings, slabs, and other objects found under a coastal building can 
contribute to localized scour. 
Depending on the building location, soil characteristics, and flood conditions, a building may be subject 
to either coastal erosion or scour, or both. Refer to Hurricane Ike Recovery Advisory on Erosion, Scour 
and Foundation Design (Appendix D) for additional information. 
A preliminary review of pre-and post-Ike aerial photographs suggests that between 100 and 
200 feet of dunes and vegetation were lost during Ike along much of the Gulf shoreline (see 
Figure 3-16). 

This loss occurred in areas with natural dunes and in areas where previously eroded dunes 
had been rebuilt and reinforced with geotextile tubes (see Figures 3-13 and 3-17). As of 2003, 
approximately 7.6 miles of geotextile tube dune reinforcement had been installed along the 
Texas shoreline, mostly along the Bolivar Peninsula and western Galveston Island shorelines 
(Gibeaut et al., 2003). Virtually all of these tubes were uncovered by Ike, and many were 
destroyed. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-16. 
Pre- and post-Ike aerial 
photographs of the east 
end of Galveston Island, 
TX, illustrating some of 
the most significant loss 
of dunes and vegetation 
during Ike 

SouRcE: uSGS, http:// 
coastal.er.usgs.gov/ 
hurricanes/ike/photocomparisons/
galveston.html 

The Ike MAT noted that the amount of scour around pile foundations was far greater than that 
observed during previous post-storm investigations, both in terms of frequency of occurrence 
and depth of scour. Most of the scour was observed at foundations with concrete slabs at ground 
level, but this is likely due to the prevalence of this type of construction; significant scour was 
also observed around some pile foundations before the slabs had been constructed. Significant 
scour (several feet deep, tens of feet in diameter) was observed after Ike at hundreds of the 
buildings that were still standing near the Gulf shoreline. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-17. 
Exposed geotextile tubes 
formerly covered by sand 
and dune vegetation. 
Note erosion behind the 
tubes and under Zone V 
(BFE = 18 feet) buildings 
(yellow arrow at left). 


Figures 3-18 and 3-19 show buildings at Holly Beach, LA, both of which sustained significant 
scour around foundations. Of the approximately 20 pile-elevated houses in existence at Holly 
Beach prior to Hurricane Ike, nearly half experienced 
significant foundation scour (virtually all 
of buildings at Holly Beach were destroyed by The amount of scour around pile founda-
Hurricane Rita in 2005, and the houses observed tions observed by the Ike MAT far exceeded 
by the Ike MAT had been built since 2005). what current design guidance predicts. 

Figure 3-18. 
Foundation scour 
observed at Holly 
Beach, LA 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-19.
Foundation scour
observed at Holly Beach,
LA (Zone V, ABFE = 16
feet)


Figure 3-20 shows a case of extreme foundation scour at a house on Bolivar Peninsula. The 
scour depression shown was reported by a local contractor to have been as much as 10 feet deep. 
The house was able to withstand the scour and the wind and flood loads acting on the structure, 
but lack of soil support allowed the bottoms of some of the piles supporting the deck on the 
right side of the house to be shifted toward the left. 

Figure 3-20. 
Foundation scour was 
reported to be 10 feet 
deep—note the bottoms 
of the piles on the right 
side of building that have 
been pushed toward 
the building (Bolivar 
Peninsula, TX; Zone V) 


HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

In some cases, pile foundations subject to erosion and/or scour were not embedded deeply 
enough to resist the loads and conditions that were present during Ike. Figure 3-21 shows such 
a case where a pile foundation shifted under an elevated house. Scour and erosion contributed 
to the failure. The presence of the attached, but broken, concrete slab could also have contributed 
to the foundation failure by reducing the lateral support formerly provided by the intact 
slab, and by causing eccentric loading of the piles (see Section 3.3.3). 

Figure 3-21. 
Failure of a timber pile 
foundation undermined 
by scour and erosion. 
Inset shows close-up 
of concrete slab failure 
and rotation of some 
of the foundation piles 
(Galveston Island, TX; 
Zone V). 


One other aspect of scour was noted by the MAT—linear scour features that result in the loss of 
soil around or under buildings when storm surge flow is channeled or directed across a building 
site. This process usually takes place where storm surge flow is constrained between large 
buildings or gaps in shore protection, or when storm surge return flow to the sea follows paths 
of least resistance, such as along canals and roads (Figure 3-22). Some of the many buildings lost 
during Ike were likely lost as a result of this process. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-22. 
Linear scour features 
tend to align with canals 
and roads as storm 
surge returns to the Gulf. 
Houses such as this one 
were fortunate not to 
be undermined and lost 
during Ike, as many homes 
undoubtedly were (Bolivar 
Peninsula, TX). 

3.1.1.3 Foundation Function 3: Provide a Continuous Load Path to the Ground 
Loads acting on a building follow many paths through the building and must eventually be resisted 
by the ground, or the building will fail. Loads accumulate as they are routed through key 
connections in a building (connections between members are usually the weak links in a load 
path). Load paths must be continuous, from the top of the building, through the foundation, 
and into the ground; failed or missed connections cause the loads to be rerouted through unintended 
load paths, potentially overloading those paths and resulting in structural failure. A 
graphic illustrating vertical and horizontal load paths from an elevated building to the foundation 
and into the ground is shown in Figure 3-23. 

Connections between structural members are often the weak point in a load path, and the MAT 
observed many load path failures at the floor system-pile connection. The MAT also observed 
instances where this connection was adequate to prevent structural failure during Ike. Figure 
3-24 shows an example of a wood-frame house elevated on concrete columns. The house survived 
with no structural damage even though the owner reported a flood level above the lowest 
floor. The attachment of the timber floor beams to the concrete columns provided load path 
continuity and prevented the house from floating or washing off its foundation (Figure 3-25). 
Although this house survived Hurricane Ike, this type of connection only provides limited resistance 
to lateral loads and applied moments—had the house experienced a higher surge or 
stronger winds, it may not have survived. The MAT estimated the 3-second gust wind speed (Exposure 
C) during Ike was approximately 85 mph at this site, but if wind speeds or lateral flood 
loads had been higher, this house could have sustained structural damage. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-23.
Example load path through
a pile foundation (note:
some building components
are not shown)


SouRcE: FEMA P-762, 

LocaL oFFiciaLS 
GuiDE FoR coaStaL 
conStRuction 

Figure 3-24. 
This Bridge City, TX, house 
sustained no structural 
damage, despite the fact 
that the owner reported 
that Ike flood levels rose 
above the lowest floor 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-25. 
A 5/8-inch diameter 
galvanized steel anchor 
bolt in red circle (with 
washer and nut, not 
visible in this photograph) 
provided connections 
between beam and 
column for the house 
shown in Figure 3-24. 
This does not appear to be 
an engineered connection. 

Some designs rely on connections between columns and beams to provide fixity (resistance to 
rotation), particularly in commercial or multi-family buildings of concrete construction. Figure 
3-26 shows one such example—reinforcing steel that will extend into a concrete beam (under 
construction) and connect columns and beams. The cast-in-place concrete connection will provide 
resistance to rotation. 

Figure 3-26.
Reinforcing steel
extending from the top of a
concrete column (building
under construction)
(Galveston Island, TX)



The MAT noted instances of other types of foundation load path failures, including those at the 
point where a column attached to a pile cap, slab, or grade beam. Deterioration of timber piles 
contributed to load path failures in some foundations (Figure 3-27). The deterioration could have 
been the result of inadequate preservative treatment or poor design/construction practice. In other 
cases, deterioration was observed that did not result in foundation failure during Ike; however, 
such a weakened foundation would be more susceptible to failure in the future (Figure 3-28). 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-27. 
Deterioration in the wood 
piling likely contributed 
to the foundation failure 
(Bolivar Peninsula, TX) 

Figure 3-28. 
Deterioration in wood 
piling. The foundation did 
not fail during Ike, but it 
was weakened and will be 
more susceptible to failure 
in the future (Galveston 
Island, TX). 


The MAT also noted cases where houses survived Ike, but must not have been exposed to high 
winds or large flood loads; otherwise the lack of load path continuity would have resulted 
in foundation failure. The house shown in Figure 3-29 is resting on top of precast concrete 
piers, stacked concrete masonry units (CMUs), and shallow footing pads—the necessary 
structural connections are missing. This design will not provide a continuous load path from 
the elevated house to the ground, and does not comply with minimum NFIP or building 
code requirements. This foundation will likely fail if it is subject to high winds and/or waves, 
velocity flow, or scour. Additional discussion of load paths is provided in Section 3.1.2.3. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-29. 
House resting (i.e., with 
no structural connection) 
on top of precast concrete 
piers, stacked CMUs, and 
shallow footing pads (New 
Iberia, LA) 

3.1.1.4 Foundation Function 4: Resist Flood Loads 
Flood loads acting on a coastal building can include: 

n Hydrostatic loads (pressure from standing or slowly moving water). Vertical hydrostatic 
forces are known as buoyant forces, and cause objects to float, including houses that 
are poorly attached to their foundations. Lateral hydrostatic forces will not harm pile or 
column (open) foundations, but can cause damage to foundation walls and enclosure walls 
that do not have the flood openings required to allow inside and outside water levels to 
equalize. 
n Hydrodynamic loads (forces caused by fast-moving water, the up-rush of broken 
waves, etc.). Storm surge flowing past or around a foundation or building will lead to 
hydrodynamic loads. 
n Wave loads (caused by waves breaking on or striking a building foundation). Wave loads 
are high magnitude, short duration loads that can cause rapid destruction of inadequately 
elevated or constructed buildings. Hundreds of waves can strike a building during an 
episode of hurricane flooding. 
n Floodborne debris impacts (parts of broken structures striking a building, or becoming 
lodged in a building foundation and transferring other flood loads to the foundation). 
Large numbers of buildings destroyed by flood forces contributed to large quantities of 
floodborne debris, and undoubtedly led to additional building failures during Ike. 

Flood damages to residential buildings observed by the MAT were consistent with the nature 
and magnitudes of the flood loads described above. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

n In locations where waves were small and 
flood velocities were low, there was little 
damage to houses elevated above the flood 
level on NFIP-compliant foundations. 
Houses constructed at grade, or not elevated 
high enough above the ground to escape 
the flooding, were inundated and sometimes 
dislodged from their foundations. 

n In locations where waves were larger, flow 
velocities were greater, and floodborne 
debris generation was significant. Houses 
not elevated high enough were severely 
damaged or destroyed. Houses elevated 
above the wave crest level were still 
subject to damage or destruction if their 
foundations could not withstand the flood 
loads and failed. 

The typical wave damage patterns described 
above are illustrated in Figure 3-30. Damage 
to properly designed and constructed elevated 
houses is generally minor until the waves reach 
the elevated floor system, at which point the 
damage increases dramatically with increasing 

Typical, low-rise residential buildings near 
the shoreline can be designed and constructed 
to resist wind loads, but must be 
elevated high enough on a pile or column 
(open) foundation to avoid flood loads. 

Wind pressures acting on walls of low-rise 
buildings are almost always less than 100 
pounds per square foot (psf), and these 
loads can be resisted easily by proper 
design and construction. However, fast-
moving storm surge and floodborne debris 
can exert pressures several times higher 
than wind pressures against a building 
wall. Wave pressures against walls can 
reach several hundred, or in extreme cases, 
thousands of psf. 

Even lateral flood loads acting on pile 
or column foundations can reach 1,000 
pounds or more against each pile or column. 
These loads can be resisted, but only 
by properly designed and constructed open 
foundations. 

water level and wave height. The importance of adding freeboard—elevating above the wave 
crest level—is apparent (see Section 3.1.3 and the Ike Recovery Advisory, Designing for Flood Levels 
above the BFE). 


Figure 3-30. Idealized depth-damage relationship for an elevated building subject to waves 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Wave effects and floodborne debris impacts 
were a major cause of building structural fail-An estimated 3,600 buildings, (approxiure 
during Hurricane Ike, both on lands near mately 61 percent of the pre-Ike buildings) 
the Gulf of Mexico and immediately adjacent to on Bolivar Peninsula were destroyed by 
many bay-front shorelines. Damage was more se-Hurricane Ike, and approximately 2,200 
vere and widespread along the Gulf shoreline, (37 percent) more were damaged (Halff 

Associates, 2008). Much of the Peninsula 

as would be expected, since the wave heights 

was inundated by an estimated 6 to 10 

were larger there. Also, the loss of many build-

feet of stillwater, and experienced wave ef


ings along the Gulf shoreline added greatly to fects above that level—meaning that Ike 
the debris stream available to strike and damage flood levels exceeded the BFE for virtual-
other buildings farther inland. ly all of the Peninsula. This would explain 

the widespread loss of elevated houses 
It is not always possible to separate damages on the Peninsula, and the survival of only 
caused by waves alone from that caused by flood-those houses elevated the highest, with 
borne debris, especially since the debris is carried deep foundations resistant to waves, de-
by the surge and waves. However, the direct and bris, storm surge, erosion, and scour. 
indirect effects of waves should be considered 
one of the two most damaging aspects of coastal 
flooding for coastal residential buildings (erosion and scour being the other). 

Figure 3-31 shows a comparison of pre- and post-Ike photographs for the Crystal Beach area of 
the Bolivar Peninsula. The Peninsula is the region where Hurricane Ike storm surge levels and 
wave heights appear to have reached maximums along the Gulf shoreline. Buildings along the 
Gulf shoreline of the Peninsula were likely subject to the greatest flood forces during Ike, and 
sustained the worst damage. Damage in this area has been compared to the Mississippi coast 
following Hurricane Katrina. 

Figures 3-32 and 3-33 show examples of houses affected by waves and the inland penetration of 
large debris fields. The combination of waves and debris led to the destruction of many houses 
on Bolivar Peninsula. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-31. 
Pre- and post-Ike aerial 
photographs of the Crystal 
Beach area of Bolivar 
Peninsula, TX 

SouRcE: uSGS1 


1 http://coastal.er.usgs.gov/hurricanes/ike/photo-comparisons/bolivar.html 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-32. 
Seaward side of Zone V 
house struck by waves. 
The deck, the elevated 
floor system, and the 
seaward walls were 
destroyed or heavily 
damaged (Bolivar 
Peninsula, TX). 

Figure 3-33. 
Roofs, walls, and other 
parts of destroyed houses 
washed landward to and 
inland of this location, 
approximately ½ mile 
from the Gulf shoreline of 
Bolivar Peninsula, TX 

Although flood levels and wave conditions were not as severe on Galveston Island as on the Bolivar 
Peninsula, many houses were also lost there, largely as a result of waves and erosion. Figure 
3-34 shows one example, approximately 3 miles west of the seawall, where two Gulf-front houses 
were lost. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-34.
Broken piles beneath
destroyed Gulf-front
houses, Galveston Island,
TX (west of the seawall)



Figures 3-35 and 3-36 show examples of Ike wave damage typical for Galveston Bay, where wave 
heights were less than those on the Gulf shoreline. These at-grade buildings were gutted or destroyed 
by storm surge, waves, and floodborne debris. In both cases, nearby buildings elevated 
on pile foundations survived, with damage only to breakaway walls and access stairs. 

Figure 3-35. 
Damage to at-grade house 
in Zone V likely caused by 
wave and surge along the 
northern Galveston Bay 
shoreline in Baytown, TX 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-36. 
Likely wave and debris 
damage to townhouse 
building along the 
western Galveston Bay 
in Seabrook, TX. The 
building was supported 
on shore-parallel masonry 
walls, and is landward of 
another building that was 
destroyed by Ike. 

3.1.2 Main Wind Force Resisting System 
According to ASCE 7-05, the MWFRS is an assemblage of structural elements that provide support 
and stability for the overall structure. The MWFRS can be thought of as the portion of a 
building’s structural frame that collects wind loads from the building envelope and transfers 
those loads to the ground via the building’s foundation. Elements of the building envelope 
that do not qualify as a part of the MWFRS are identified as C&C. While some may consider the 
foundation to be part of the MWFRS, the following discussion will focus on that portion of the 
structural system above the foundation. 

3.1.2.1 High Winds 
High winds can originate from a number of events—tornadoes, hurricanes, extra-tropical cyclones, 
and other coastal storms. The most current design wind speeds are given by the national 
load standard, ASCE 7-05. Figure 3-37, taken from ASCE 7-05, shows the geographic distribution 
of design wind speeds for Gulf of Mexico and Atlantic Ocean portions of the United States. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


3-28 MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS and Louisiana 
3 Performance of Residential Buildings (Flood and Wind), One- to Two -Fami ly and Multi-Fami ly 
Figure 3-37. ASCE wind speed map (ASCE 7-05) 
High winds are capable of imposing large lateral (horizontal) and uplift (vertical) forces on 
buildings (see Figure 3-38). Residential buildings can suffer extensive wind damage when they 
are improperly designed and constructed and when wind speeds exceed design levels. The damages 
shown in Figures 3-39, 3-40, and 3-41 exemplify poor design and construction, since Ike’s 
winds were less than design levels. 
The effects of high winds on a building will depend on several factors: 
n Maximum wind speeds, gustiness of the winds, wind directions, and duration of high winds 
n Height of building above ground 
n Exposure or shielding of the building (by topography, vegetation, or other buildings) 
relative to wind direction 
n Topographic effects (hills and escarpments) that create wind speedup 
n Strength of the structural frame, connections, and envelope (walls and roof) 
n Shape of building and building components

PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Wind Direction 
n Number, size, location, and resistance to damage of openings (e.g., windows, doors, and vents) 
n Presence and strength of shutters or opening protection 
n Type, quantity, and velocity of windborne debris 

Proper design and construction of residential structures, particularly those close to open water 
or near the coast, demand that every factor mentioned above be investigated and addressed 
carefully. Failure to do so may ultimately result in building damage or destruction by wind. 
Hurricane Ike winds removed the roof structure on the house shown in Figure 3-41. Hurricane 
straps could have been added, thereby greatly increasing the resistance to wind. Refer to IBHS 
2005 Standards for proper connection of roof structural elements. 

Figure 3-38. 
Code-defined MWFRS 
wind loads on an elevated 
residential structure 

SouRcE: FEMA 55 


Figure 3-39. 
Galveston, TX, West End 
Beach house with roof 
structure removed by 
Hurricane Ike. 
The cause of the failure 
is unknown, but Ike wind 
speeds in this area were 
below design speeds 
(Hurricane Ike estimated 
wind speed in this area: 93 
mph, Exposure C). 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-40. 
This West Bay, Galveston 
Island, TX, apartment 
experienced gable-end 
wind damage as a result 
of sheathing failure and 
poor connection of the 
brick veneer to the stud 
walls (Hurricane Ike 
estimated wind speed 
in this area: 90 mph, 
Exposure B) 

Figure 3-41. 
The roof structure was 
poorly connected to 
this house in Grand Isle 
(Jefferson Parish, LA) and 
was blown off by 50-mph 
Ike winds (Exposure B) 


3.1.2.2 Combination of Loads – MWFRS and C&C 
Some elements of low-rise buildings are considered to be part of the C&C or part of the MWFRS, 
depending upon the wind load being considered. Using the example of the exterior walls of a 
masonry building, the MWFRS provisions are used to determine the in-plane shear forces in the 
design of these masonry walls, and the C&C provisions are used to determine the out-of-plane 
design bending loads. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


HURRICANE IKE IN TEXAS and Louisiana MITIGATION ASSESSMENT TEAM REPORT 3-31 
Performance of Residential Buildings (Flood and Wind), One- to Two -Fami ly and Multi-Fami ly 3 
The pressure (positive/inward or negative/outward suction) exerted by wind flowing over and 
around a building varies with time and location on the building. The highest pressures occur 
over small areas for a very short time in the regions of a building where the wind flow separation 
is quite significant (such as at corners of roofs and walls, ridges, hips, and overhangs). This flow 
separation can cause small vortices to form that can cause much higher pressures in small 
localized areas. These flow separation regions generally occur along the edges of the roof and 
corners of exterior walls (see Figures 3-42 and 3-43). Therefore, the design wind pressures 
for the design of the C&C element can be nearly three times the pressure used to design the 
structural framing of the building. Proper assessment of the design wind pressures is critical to 
developing the design of a building’s structural frame and the selection of appropriate exterior 
cladding. 
Figure 3-42. 
Areas of roof covering 
loss (red arrows) indicate 
zones of higher wind 
pressure on a roof 
Figure 3-43. 
Galveston, TX, West End 
Beach house with roof 
damage in high pressure 
zones (Hurricane Ike 
estimated wind speed 
in this area: 95 mph, 
Exposure C) 

PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

In addition to these external pressures, openings and the natural porosity of the building components 
contribute to internal pressures. As seen in Figure 3-44, internal pressures introduced 
by building openings are additive to (or subtractive from) the external pressures. The magnitude 
of the internal pressures depends on whether the building is “enclosed,” “partially enclosed,” or 
“open” as defined by ASCE 7-05. In hurricane-prone regions as defined by ASCE 7-05, in order 
for a building to be considered “enclosed” for design purposes, glazing must either be impact-
resistant or protected with shutters or other devices that are impact-resistant. This requirement 
also applies to indoor glazing and skylights. Refer to Section 3.2.4 for the discussion on windows 
and shutters and their performance in Hurricane Ike. As previously stated, Hurricane Ike was 
not a wind design event and therefore the MAT did not observe any notable examples of building 
failures resulting from internal pressurization. 


Figure 3-44. 

Effect of wind on an enclosed building and a building with a wall opening producing a partially enclosed building 

by allowing internal pressurization of the structure 

3.1.2.3 Load Paths 
Figures 3-45 and 3-46 illustrate the load path concept for the elevated portion of a building. 
Wind loads collected and concentrated as shown in these figures must be passed through the 
foundation to the ground (see Figure 3-23). 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-45. Depiction of a building load path 

SouRcE: FEMA 489 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-46. 
Load path around 
openings and connection 
to foundation pile 

SouRcE: FEMA 499 


Figure 3-47 shows a house on western Galveston Island that collapsed during Ike. High water 
levels and waves acted on the foundation while winds (blowing from land toward the Gulf of 
Mexico) pushed the house seaward. The result was a foundation failure—the foundation could 
not provide the required load path continuity to the ground without breaking. 

Figure 3-47. 
Collapse of a West 
Galveston Island, TX, 
house due to foundation 
failure 
Broken Pilings 
MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-48 shows a house on Bolivar Peninsula that remained standing although severely damaged 
by surge, waves, and wind. The house survived because its MWFRS and foundation load 
paths remained intact. 

Figure 3-48. 
Though much of the 
cladding and structural 
sheathing was destroyed 
by Ike’s surge, the MWFRS 
of this Bolivar Peninsula, 
TX, beach house remained 
intact and connected 


Piling connections to floor beams of elevated structures were routinely observed by the MAT. 
However, unless the building was substantially damaged or under construction, most load path 
connections of wall and roof structural elements were covered by building finishes and not visible 
for inspection. Some beam-to-piling connections were found to be strong and robust as 
seen in Figures 3-49 and 3-50. Many others were weakly connected with nails, too few bolts, or 
columns weakened by deep mortises (Figures 3-51). 

Figure 3-49. 
Strong concrete column-
to-beam steel saddle 
connector 


HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-50. 
Strong interior wood 
column-to-beam 
connector with building 
shear wall connector 
Bolted Beam connection 
Building Shear Wall connector 
Floor Joist (Lollipop) 
connectors 
Figure 3-51. 
Poor beam connection to 
corner column 

New construction was frequently observed with robust construction such as sill plates bolted to 
slabs-on-grade, studs clipped to double top plates, and wall-to-roof construction (Figures 3-52, 
3-53, and 3-54). 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-52. 
Studs and sill plate 
connected in new house 
(sill bolts yellow arrows 
and clips red arrows). 
However, sill bolts are 
spaced too far apart (2 feet 
is the maximum spacing 
allowed) and did not have 
3-inch by 3-inch by 1/8inch 
washers per 2003 
IRC and TDI-adopted IBHS 
guidelines. Blue line shows 
3-foot spacing (Webster, 
TX). 

Figure 3-53. 
Studs clipped to double 
top plate; rafter-to-topplate 
connector has yet 

to be installed. Better 
framing practices could 

have avoided some of 
the problems shown in 
this photo. Ceiling joists 

are not well nailed to the 
rafter and may twist in 

the future. The builder 
did not take advantage of 
aligning wall framing and 
rafter framing to simplify 

connections for wind 

loads. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-54. 
Wall-to-roof strapping. 
Details for uncommon 
framing details should be 
specifically provided by 
the designer on building 
plans, including specifying 
the specific connection 
and application to ensure 
a continuous load path. 


Numerous new and older houses, however, were observed without proper hurricane connections 
or improperly installed connections (Figures 3-55 and 3-56). 

Figure 3-55. 
Toe-nailed connection 
of floor joists on band 
beam on house under 
construction. Floor joist 
should be installed using 
either galvanized metal 
joist hangers or ledger 
beams (LaPorte, TX). 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-56. 
Existing house shear wall 
connector incorrectly 
located (red arrow). 
Connector should be 
located on column line or 
the beam and beam-tocolumn 
connection should 
be designed to resist 
the uplift load, which 
is carried by a nailed 
connection in the absence 
of bolts (Sunset Crystal 
Beach, Bolivar Island, TX). 

3.1.3 Elevation and Freeboard 
The observations of the Ike MAT investigation clearly demonstrate the importance of elevating 
buildings above the flood level, including any effects of waves and floodborne debris. Elevating 
only to the BFE does not guarantee a house will remain free of flood damage during a specific 
hurricane or coastal flood event. As was stated in Section 2.1.1, FISs and FIRMs may not depict 
the true lateral and vertical extents of actual flooding during the base flood event (100-year 
flood event) for a variety of reasons. Nor will construction to the 100-year flood event shown on 
the maps offer protection against floods that exceed the true base flood. 

The key to successful coastal buildings is to construct them higher than the BFE by adding freeboard. 
The desired amount of freeboard will depend on a number of factors, but the age of 
the FIRM and the nature of the building being 
constructed are the most important factors. Old 

The MAT recommends any post-Ike re-

FIRMs tend to be less accurate than newer FIRMs 

construction and new construction in 

in showing the contemporary 1-percent-annual-Ike-flooded areas be carried out with a 
chance flood level. Critical facilities should be minimum of 3 feet of freeboard above the 
constructed with higher freeboard than typical BFEs shown on the Effective FIRMs at 

residential and commercial structures. the time of Ike (refer to Section 3.1.1.1 for 
additional information). Freeboard is necessary 
to compensate for out-of-date flood 
hazard maps and to provide an additional 
degree of flood protection not afforded by 
the Effective FIRMs. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

3.1.4 Siting Effects on Structural Performance 
While many people recognize that how buildings are constructed will affect flood damage to 
that building (e.g., building floor elevation and foundation design), they may not appreciate the 
importance of where buildings are constructed in determining flood damage. Post-hurricane inspections 
typically observe the greatest flood damage, loss of coastal buildings, and loss of roads 
and infrastructure in the area closest to the shoreline. This was also the case with Hurricane Ike. 

The greatest damage occurs in the area closest to the water since buildings and infrastructure 
situated there are subject to the most extreme flood forces and conditions during a hurricane, 
i.e., the highest waves and the greatest erosion (Figures 3-57 and 3-58). Buildings situated closest 
to the shoreline are also at greatest risk for the effects of long-term erosion, sea level rise, and 
other long-term changes affecting the shoreline (see Surfside Beach text box). 

Figure 3-57. 
Post-Ike photograph of 
West Galveston Island, 
TX, illustrating increased 
severity of flood damage 
near the shoreline 


Figure 3-58. 
Post-Ike photograph of 
Bolivar Peninsula, TX, 
illustrating increased 
severity of flood damage 
near the shoreline 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

SURFSIDE BEACH, TX


The closer a building is located to the shoreline, the more vulnerable it becomes. This is not only due to 
the increasing flood forces close to the shoreline, but also because a building’s foundation designed for 
a given location and set of conditions (ground elevation, stillwater flood depth, wave height, etc.) will find 
itself exposed to a different set of conditions (lower ground, higher wave height, etc.) as the shoreline 
erodes over time, and the building may not be able to withstand those new conditions. A classic case is 
Surfside Beach, TX, where long-term erosion had resulted in dozens of houses standing on the beach, 
seaward of the line of vegetation. Many of these houses were ordered removed by the State of Texas; 
some were removed, but others remained and litigation resulted. Hurricane Ike destroyed most of the 
houses standing on the beach (see photos below). 


The presence of reinforced dunes and revetments and seawalls can reduce damage slightly in 
areas close to the shoreline when those dunes and erosion control structures remain intact during 
a storm event. However, when they fail they offer little protection to upland buildings. Of 
the structures observed by the MAT, only the Galveston Seawall provided significant protection 
to buildings against wave attack and erosion. The recently completed Surfside Beach revetment 
appears to have survived Ike with minor damage, and undoubtedly offered some protection to 
upland buildings, but this revetment was not subject to the extreme forces that the Galveston 
seawall and shorelines farther east were. 

Although wave and erosion effects in the bays were not as severe as on the Gulf coast, buildings 
sited close to the bay shoreline were at increased risk to flood damage, relative to buildings farther 
from the bay shoreline. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

3.2 Envelope Damage 
The MAT observed building envelope damage as far west as the west end of Galveston Island, TX, 
and as far east as Terrebonne Parish, LA, a distance of approximately 150 miles. The MAT also 
observed building envelope damage as far inland as the north side of the City of Houston, approximately 
45 miles from the coast (see Figure 1-16). Sections 3.2.1 through 3.2.6 describe building 
envelope performance, including roof systems, non-load-bearing walls and wall coverings, doors, 
windows and shutters, soffits and roof ventilation, and exterior-mounted equipment. 

Blow-off of building envelope components frequently results in damage to adjacent buildings 
and vehicles, as well as the building itself. The most notable building envelope issues during 
Hurricane Ike, and the most common windborne building envelope debris, were roof coverings 
and vinyl siding. Figure 3-59 illustrates the magnitude of building envelope debris that occurred 
in some areas. 

As expected, the building envelope on older houses did not perform as well as on new houses. Specifically, 
houses constructed prior to 1985 in Texas and prior to the adoption of the IRC in 2005 in 
Louisiana exhibited the poorest envelope performance. Post-1985 Texas home construction in the 
counties affected by Hurricane Ike were governed by the Texas Windstorm Program (refer to Section 
2.3), and all post-2005 houses in Louisiana were governed by the newly adopted IRC. 

The extent and magnitude of envelope damage observed by the MAT was greater than would 
be anticipated given that the estimated actual wind speeds of Hurricane Ike were less than the 
design speeds given by ASCE 7-05 and IRC 2006. The poor performance of the newer houses is 
therefore related to the lack of contractor knowledge of proper hurricane construction, material 
installations not conforming to manufacturer’s requirements for hurricane zones, and poor 
code enforcement. 

Figure 3-59. 
A substantial amount 
of siding (the white 
lines scattered around 
the ground), along with 
roofing materials, blew 
off these West Galveston, 
TX, houses (Hurricane Ike 
estimated wind speed in 
this area: 
90 mph) 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

3.2.1 Roof Systems 
Historically, damage to roof coverings is the leading 
cause of building performance problems LIQUID-APPLIED ROOF 
during hurricanes. In the rains accompanying a COVERING 
hurricane, rainwater entering a building through 

The MAT observed one residence that

damaged roofs can cause major damage to the 

had a liquid-applied roof covering over a 

interior finishes and contents. Unless quick ac-concrete deck. FEMA investigations after 
tion is taken to dry a building, mold bloom can Hurricane Marilyn (1995) in the u.S. Virgin 


quickly occur in the hot, humid southern cli-Islands found that this type of roof covering 
mate. Drying of buildings was hampered after has excellent wind performance.
Hurricane Ike by the lack of electrical power to 
run fans and dehumidifiers. 


The MAT observed a variety of roof coverings, including asphalt shingles, metal panels, metal 
tiles, and tile. In the areas observed by the MAT, roof covering damage was common, and quite 
variable as shown in Figure 3-60. This type of variability is consistent with what was observed by 
the Hurricane Charley, Ivan, and Katrina MATs (see FEMA 488, Hurricane Charley in Florida: Ob-
servations, Recommendations, and Technical Guidance [April 2005a]; FEMA 489, Hurricane Ivan
in Alabama and Florida: Observations, Recommendations, and Technical Guidance [August 2005e]; 
and FEMA 549, Hurricane Katrina in the Gulf Coast: Building Performance Observations, Recommen-
dations, and Technical Guidance [July 2006b]). 


Figure 3-60. 
Some of the roofs on 
these Jamaica Beach, 
TX, houses had no roof 
covering damage, while 
one had moderate 
damage (blue arrow) 
and one had extensive 
damage, including loss of 
underlayment (red arrow) 
(Hurricane Ike estimated 
wind speed in this area: 
90 mph) 


HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

At several residences, a large amount of roof covering was blown away, as shown in Figures 3-60, 
3-66, 3-67, and 3-69. However, more commonly, roof covering damage was limited to a small 
area such as at corners, eaves, rakes, or ridges. In the case of asphalt shingled roofs, sometimes a 
few shingles in the field of the roof were blown away. Had Hurricane Ike’s winds been closer to 
current design wind speeds, the roof covering damage would likely have been greater. The following 
subsections present asphalt shingle, metal panel, and tile roof observations. 

3.2.1.1 Asphalt Shingles 
ASPHALT SHINGLE CLASS 

Most of the residences observed by the MAT had 

RATINGS 

asphalt shingle roof coverings. There were two 
notable observations, as discussed below: 1) use Testing and labeling is prescribed in ASTM 
of shingles that had been tested and labeled in D 7158.* The following classes of shingles 
accordance with relatively new criteria, and 2) are specified in this standard: 
the use of roof tape at deck sheathing joints. Class D: Suitable for use up to 90 mph 

New Shingle Labels. Asphalt shingles are now Class G: Suitable for use up to 120 mph 
available with Class D, G, or H labels (see text Class H: Suitable up to 150 mph 
box). At the time of Hurricane Katrina (2005), 

Class F: Shingles with this classification 

only a limited number of shingles were available 

are tested in accordance with the old test 

with the new ratings. However, several products method prescribed in ASTM D 3161, a test 
are now available with the new classifications.2 
method widely recognized antiquated 

as 

for evaluating the wind resistance of self-
Figure 3-61 shows a shingle bundle wrapper at a sealing shingles 
house under construction at the inset in Figure 

* Wind speeds cited are design wind speeds 
3-61. The shingle bundle wrappers indicate the in IBc/IRc/AScE 7 (based on Exposure 
shingles meet Class H (i.e., suitable for up to 150 c, and a maximum mean roof height of 60 
mph). The IRC/ASCE 7 design wind speed for feet). 
this locaton is 120 mph, hence use of a Class G 
shingle would have been sufficient. This is the 
first and only house observed by a MAT wherein it was known that shingles meeting one of the 
new Class ratings was installed. There was no apparent wind damage to this house. 

The MAT observed several other newly installed roofs, but was unable to determine if the shingles 
met any of the new classifications. Even if the shingles did meet Class G or H, failure could 
have initiated along the rake, eave, or hip/ridge unless there was special securement (such as 
that shown in Technical Fact Sheet 20, Asphalt Shingle Roofing for High-Wind Regions, in FEMA 
499), as described below. 

The newly constructed house shown in Figure 3-62 is on the same block as the one shown in Figure 
3-61. Bleeder strips were installed along the rake; however, as discussed in Section 5.4.1.3 of 
the Katrina MAT report (FEMA 549), unless the shingles are hand-tabbed as described in Technical 
Fact Sheet 20, bleeders do not provide reliable securement. 

2 See the following TDI Web site for product listings: http://www.tdi.state.tx.us/wind/documents/ashglcnf08ibcircrev031009b.pdf 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-61.
View of a shingle bundle
wrapper at the Webster,
TX, house shown in the
inset. This shingle has a
Class H rating (red arrow)
(Hurricane Ike estimated
wind speed in this area:
104 mph).


Figure 3-62.
Shingle damage at a house
near the one shown in
Figure 3-61 (Webster, TX)


HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-63 shows a house under construction on Bolivar Peninsula that lost shingles along the 
eave (it is also shown on the front cover of this report). Failure along eaves commonly occurs because 
of incorrect application of the starter course and lack of hand-tabbing (as recommended in 
Technical Fact Sheet 20). For further discussion of eave issues, see Section 5.4.1.2 in FEMA 549. 

Figure 3-63. 
Loss of shingles along 
the eave in Bolivar 
Peninsula, TX (Hurricane 
Ike estimated wind speed 
in this area: 110 mph) 


Figure 3-64 shows a house that was reportedly constructed in 2005 on Bolivar Peninsula. It lost 
shingles and underlayment at a corner area (red circle at the inset) and shingles in the field of 
the roof near the exhaust fan (blue arrow). Loss of shingles was likely due to lack of hand-tabbing. 
These shingles reportedly met Class F. 

Figure 3-64. 
Loss of shingles and 
underlayment in a corner 
area, and loss of shingles 
from the field of this roof 
in Bolivar Peninsula, TX 
(Hurricane Ike estimated wind 
speed in this area: 110 mph) 
MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-65 shows a house under construction on Bolivar Peninsula. It lost shingles along the 
hip. Also, at areas along the exposed hip, either the underlayment did not completely lap over 
the hip line, or if it did, portions of the underlayment blew away. Water could leak into the 
building in the vicinity of the two red arrows. Unless hip and ridge shingles are hand-tabbed, as 
recommended in Technical Fact Sheet 20, they are very susceptible to blow-off (for further discussion, 
see Section 5.4.1.1 in FEMA 549). 

Taping of Sheathing Joints. Figure 3-66 shows some relatively new Fortified...for safer living® houses 
in the Audubon Village area of Bolivar Peninsula (refer to Section 3.1.1.1 text box for more information 
on Fortified...for safer living® homes). 

As shown in Figure 3-66, some of the roof coverings had no apparent damage, but the shingles 
and underlayment were blown off of one roof (red arrow). Also, a portion of the roof overhang 
blew off of one of the houses (blue arrow). When the MAT observed blow-off of roof framing 
and/or sheathing, it typically occurred on older buildings, rather than new construction. 

Figure 3-65. 
Loss of hip shingles and portions of the 
underlayment in Bolivar Peninsula, TX 
(Hurricane Ike estimated wind speed in 
this area: 110 mph) 


HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-66. 
Roof covering and roof 
structure damage at 

Fortified... for safer living® 

houses in the Audubon 
Village area on Bolivar 
Peninsula, TX (Hurricane 
Ike estimated wind speed 
in this area: 110 mph) 


The Fortified...for safer living® requirements include special provisions pertaining to attachment 
of roof underlayment in order to make them more wind-resistant in the event the shingles are 
blown off. The MAT was unable to determine whether or not the failed underlayments complied 
with the Fortified...for safer living® requirements. However, according to IBHS investigators 
deployed after Hurricane Ike, two layers of #15 felt were installed. (Use of two layers of #15 is 
one of the underlayment options in the current Fortified...for safer living®.) The underlayment 
was attached with plastic capped-head nails, spaced at 6 inches on center along the laps and 12 
inches on center in the field (this spacing is consistent with the original Fortified...for safer living® 
spacing guidance). This underlayment and attachment spacing is consistent with underlayment 
Option 2 in Technical Fact Sheet 19, Roof Underlayment for Asphalt Shingle Roofs, in FEMA 499. 

The Fortified...for safer living® requirements also 
include a requirement to tape the sheathing The IBHS is preparing a report on Audubon 

joints with a minimum 4-inch-wide modified Village. This report is expected to be available 
on the IBHS Web site by the end of 

bitumen roof tape. The tape is intended to 

2009. Refer to: www.ibhs.org 

provide an additional line of defense against 
water infiltration in the event the shingles and 
underlayment blow off. The use of roof tape was recommended in the 2000 edition of FEMA 
55 and it is recommended in Technical Fact Sheet 19 in FEMA 499. 

Several of the Fortified...for safer living® houses that lost underlayment had taped joints, including 
the one shown in Figure 3-67. However, as shown in Figure 3-68, the taping was not effective. 
Observations by IBHS investigators revealed application problems with the tape. Staples were 
used to attach the tape because bonding problems were experienced during application. Apparently 
the applicator did not realize the tape was intended to prevent water from leaking through 
the sheathing joints. With the tape in an un-bonded and wrinkled condition, it was incapable 
of fulfilling its intended purpose. Self-adhering modified bitumen roof tape normally bonds 
quite well to sheathing. Bonding problems are commonly attributed to dust on the sheathing, 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

wet sheathing, a surfacing on the sheathing that interfered with the bonding, or using inappropriate 
tape. According to IBHS, problems with bonding self-adhering modified bitumen to 
oriented strand board (OSB) had been previously experienced at a demonstration project. In 
evaluating that demonstration project, IBHS discovered that although the OSB manufacturer 
had recommended application of a primer before installation of the self-adhering modified bitumen 
because of the presence of a wax on the OSB, a primer had not been installed. 

According to IBHS, the shingles at the Fortified...for safer living® houses at Audubon Village met 
Class H (i.e., suitable for use up to 150 mph). 

Figure 3-67. 
This Fortified...for safer 
living® house had taped 
sheathing joints (red 
arrow) 


Figure 3-68. 
This tape did not provide 
a watertight seal. Note 
the wrinkles (which allow 
water migration) and the 
staples (blue arrow) 
that were used to attach 
the tape (Audubon 
Village, TX). 

cREDIT: IBHS 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

3.2.1.2 Metal Panels 
Several metal panel roofs performed ex-

Metal panels were the second most common ceptionally well during Hurricane charley 
type of residential roof covering observed by (2004), even though they were exposed to 
the MAT. However, there were substantially few-very high winds. For further discussion, see 
er metal roofs than asphalt shingle roofs. Their FEMA 488. 
performance was quite varied, as illustrated by 

For guidance on metal roofs, see Hurricane 

a new housing area near the west end of the 

Ike Recovery Advisory, Metal Roof Systems 

Galveston seawall. All of the houses in that area 

in High-Wind Regions (Appendix D).

(around a dozen) had metal panel roofs. Three 
of the houses experienced panel blow off. Two 
of these failures are shown in Figures 3-69 and 3-70. Fortunately, as shown in the figures, the 
underlayment did not blow away, so it provided leakage protection. The panels shown in Figure 
3-69 have snap-lock seams. One side of the seam was attached with concealed fasteners. 
The seam unlatched, but lack of roof access prevented MAT investigation of the cause of the 
unlatching. 

Figure 3-69. 
These snap-lock seam 
panels were attached with 
fasteners through one side 
of the seam (Galveston 
Island, TX; Hurricane Ike 
estimated wind speed in 
this area: 110 mph) 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

The panels shown in Figure 3-70 were attached with concealed clips, which unlatched from the 
panels. The first row of clips (just above the red line) was several inches from the end of panels; 
this first row should have been within a few inches from the eave. 


Figure 3-70. 
These architectural metal 
panels unlatched from 
the concealed clips. The 
red line shows location 
of the first row of clips 
(Galveston Island, TX; 
Hurricane Ike estimated 
wind speed in this area: 
110 mph). 

3.2.1.3 Tile 
The MAT observed very few tile roofs. As with asphalt shingles and metal panels, the performance 
was quite varied. Figures 3-71 and 3-72 show two houses along the coast of Galveston 
Island. The roof shown in Figure 3-71 was observed from the air and ground. No tile damage 
(including hips) was observed. Figure 3-72 shows damage at hips, the eave, and the field (which 
was likely caused by windblown eave and/or hip tiles. 

For further information on tile roof performance, see the MAT reports for Hurricane Charley 
and Hurricane Ivan (FEMA 488 and 489, respectively), wherein a large number of tile roofs 
were observed. For guidance on design and installation of tile, see Technical Fact Sheet 21, Tile 
Roofing for High-Wind Areas, in FEMA 499. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 

3-51 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-71. 
This tile roof on Galveston 
Island, TX, did not 
experience any wind 
damage (Hurricane Ike 
estimated wind speed in 
this area: 106 mph) 

Figure 3-72. 
This tile roof on Galveston 
Island, TX, experienced 
hip, eave, and field 
damage (Hurricane Ike 
estimated wind speed in 
this area: 106 mph) 


3.2.2 Non-Load-Bearing Walls and Wall Coverings 
This section covers exterior wall coverings (also known as cladding or siding) including brick veneer 
(Section 3.2.2.1), vinyl siding (Section 3.2.2.2), fiber-cement siding (Section 3.2.2.3), and 
wood and hardboard siding (Section 3.2.2.4). In the area visited by the MAT, the most common 
exterior wall coverings were fiber-cement lap siding; vinyl siding; and panels of wood, hardboard, 
or fiber cement. Although not a prevalent residential cladding, Exterior Insulation Finish System 
(EIFS) was observed in a few locations. Because most of the houses surveyed were elevated, brick 
was predominantly observed on a few commercial or institutional buildings, and on one multifamily 
residential complex. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

In Louisiana, the MAT observed a variety of siding and cladding failures, despite the fact that 
wind speeds were less than current code-specified values. The damage observed was mostly, 
but not always, on older buildings, which (presumably) had been designed and constructed 
without full consideration of wind resistance (Figures 3-73 and 3-74). 

Figure 3-73.
Loss of siding due to
winds, Chauvin, LA
(Terrebonne Parish;
Hurricane Ike estimated
wind speed in this area:
50 to 60 mph, Exposure B)



Figure 3-74. 
Loss of siding at Holly 
Beach, LA, home (Cameron 
Parish; Hurricane Ike 
estimated wind speed 
in this area 80 mph, 
Exposure C) 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


3-54 MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS and Louisiana 
3 Performance of Residential Buildings (Flood and Wind), One- to Two -Fami ly and Multi-Fami ly 
Gable end walls are frequently covered with a non-structural sheathing, such as foam plastic or 
thin fiberboard and gypsum sheets. Because there is no interior wall covering, the sheathing 
and cladding assembly is exposed to the full force of the wind pressure differential between the 
attic and outside. Where this pressure is negative (that is, the side of the house is downwind or 
parallel to the wind direction), substantial suction pressure is exerted against the sheathing, 
which can transfer the load to the cladding and thereby produce cladding failure. The MAT 
observed several cases where both sheathing and cladding over the gable end were blown out 
(Figures 3-75 and 3-76). 
Figure 3-75. 
Complete loss of thin 
gypsum sheathing and 
brick veneer from gable 
end (West Bay, Galveston 
Island, TX; Hurricane Ike 
estimated wind speed 
in this area: 90 mph, 
Exposure B) 
Figure 3-76. 
Loss of fiberboard 
sheathing and fiber 
cement siding from gable 
end wall of an apartment 
complex (West Bay, 
Galveston Island, TX; 
Hurricane Ike estimated 
wind speed in this area: 90 
mph, Exposure B) 
Brick Veneer 
Loss of Gypsum Sheathing

HURRICANE IKE IN TEXAS and Louisiana MITIGATION ASSESSMENT TEAM REPORT 3-55 
Performance of Residential Buildings (Flood and Wind), One- to Two -Fami ly and Multi-Fami ly 3 
3.2.2.1 Brick Veneer 
Numerous brick veneer failures were observed at one Galveston apartment complex. Figure 
3-77 shows failed brick veneer at one complex. The brick ties were randomly spaced with the 
horizontal spacing ranging from 32 inches to 16 inches on-center and the vertical spacing ranging 
from 48 inches to 24 inches on-center. Many of the corrugated ties were rusted and broken, 
were not embedded in the masonry, or had minimal embedment. Figure 3-78 illustrates common 
problems with brick veneer installations and Figure 3-79 illustrates proper methods of 
installation. 
Figure 3-77. 
Collapsed brick veneer wall with brick ties 
at varied spacing, some of which were only 
partially imbedded or not embedded (West 
Bay, Galveston Island, TX; Hurricane Ike 
estimated wind speed in this area: 90 mph, 
Exposure B) 
Brick Tie Not Embedded to 
Masonry 
Gypsum Sheathing Substrate

3-56 MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS and Louisiana 
3 Performance of Residential Buildings (Flood and Wind), One- to Two -Fami ly and Multi-Fami ly 
The Brick Industry Association’s (BIA’s) Technical Notes 28 Anchored Brick Veneer, Wood Frame 
Construction (2002) specifies a maximum tie spacing of 24 inches in each direction for 16-inch 
stud spacing for buildings in standard 90-mph wind zones. Table 3-1 indicates the required tie 
spacing for high wind zones. Though Galveston experienced less than design wind speeds, the 
proximity of the adjacent complex shown in Figure 3-77 may have produced increased wind 
pressures, thereby producing the catastrophic failure of the poorly anchored brick veneer. However, 
the installed tie spacing was not suitable for this back bay Galveston location with a design 
wind speed of 120 mph. 
FEMA Hurricane Ike Recovery Advisory, Attachment of Brick Veneer in High-Wind Regions (Appendix 
D), provides recommended practices for brick veneer attachment. The advisory is based on 
observations from Hurricanes Ivan, Katrina, and Ike. 
Figure 3-79. Proper installation and embedment of corrugated brick ties 
Figure 3-78. Misalignment of the tie reduces the embedment and promotes veneer failure

PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Table 3-1. Brick Veneer Tie Spacing 

Wind Speed (mph) 
(3-Second Peak Gust) Wind Pressure (psf) 
Maximum Vertical Spacing for Ties (inches) 
16-inch stud spacing 24-inch stud spacing 
90 -19.5 24a,b 16a 
100 -24.1 24a,b 16a 
110 -29.1 20½ b 13½ 
120 -34.7 17 NAc 
130 -40.7 15 NAc 
140 -47.2 13 NAc 
150 -54.2 11 NAc 
Notes: 

1. The tie spacing is based on wind loads derived from Method 1 of AScE 7-05, for the corner area of buildings up to 30 feet 
high, located in Exposure B with an importance factor (I) of 1.0 and no topographic influence. For other heights, exposures, or 
importance factors, engineered designs are recommended. 
2. Spacing is for 2 ½-inch long 8d common (0.131-inch diameter) ring-shank fasteners embedded 2 inches into framing. Fastener 
strength is for wall framing with a Specific Gravity G=0.55 with moisture contents less than 19 percent and the following 
adjustment factors, ct=0.8; and c, cM, ceg, and ctn=1.0. Factored withdrawal strength W'=65.6#. 
D

3. The brick veneer tie spacing table is based on fastener loads only and does not take into account the adequacy of wall framing, 
sheathing, and other building elements to resist wind pressures and control deflections from a high-wind event. Prior to repairing 
damaged brick veneer, the adequacy of wall framing, wall sheathing, and connections should be verified by an engineer. 
a Maximum spacing allowed by the American concrete Institute (AcI) 530-08. 

b In locales that have adopted the 2006 IBc/IRc, the maximum vertical spacing allowed by AcI 530-05 is 18 inches. 

c 24-inch stud spacing exceeds the maximum horizontal tie spacing of AcI 530-08 prescribed for wind speeds over 110 mph. 

3.2.2.2 Vinyl Siding 
Vinyl siding was the most frequently used exterior cladding and was found in all the areas 
observed by the MAT, on both newer and older buildings. Vinyl siding was observed to be commonly 
used to re-cover older wood cladding (Figure 3-80). Panel widths observed were typically 
between 8 and 12 inches, with double-four (two 4-inch) faces, double-five, and triple 3 ½-inch 
profiles being the most common. Siding was most commonly installed over plywood or OSB 
sheathing, and usually, with a water-resistant barrier (house wrap) over the sheathing. Where the 
siding was covering older wood plank or panel siding, a layer of foam sheathing was frequently 
applied. The foam sheathing, typically ½-inch to 1-inch extruded polystyrene sheets, served as 
both additional thermal insulation and flat substrate against which to place the siding. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-80. 
Typical vinyl siding failure. Vinyl was installed 
over older wood cladding (red arrows) (Sea 
Isle, TX; Hurricane Ike estimated wind speed 
in this area: 95+ mph, Exposure C). 


Vinyl siding failure was frequently initiated at the building corners and along the bottom edges 
of elevated houses. The higher wind corner pressures produced unlatching along the bottom 
strip that resulted in the unzipping of the entire wall (Figure 3-81). When vinyl siding was blown 
off, the water-resistant barrier (either asphalt-saturated felt or housewrap) was often blown away. 
Though not witnessed by the MAT, this loss of the siding and underlayment could have allowed 
wind-driven rainwater to enter the wall cavity and the house, thereby causing water damage 
to interior finishes and contents. Vinyl siding and soffits that become windborne debris can 
potentially break unprotected glazing. 

The most important factors influencing whether vinyl siding will remain on the wall during a 
high wind event are: (1) selection of siding appropriate for the basic wind speed at the location, 
and (2) the use of proper application techniques and installation details. The latter category 
includes use of proper accessories such as starter strips, receivers and utility trim; nail selection 
and placement; and locking of successive panel courses to each other. 

Detachment of vinyl siding attributed to application deficiencies is frequently seen after high 
wind events (e.g., excessive spacing between fasteners and improper nail head size of the fasteners). 
In other cases, while proper fastening may have been used, the type of vinyl siding used 
may not have been appropriate for use in high wind locations. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-81. Improper installation led to extensive loss of siding up the house wall. The bottom lock of the lowest 
course of siding was cut off, and utility trim substituted for the correct starter strip. The poorly retained bottom 
edge pulled out under wind pressure, leading to extensive loss of siding up the house wall (Tiki Island, TX; 
Hurricane Ike estimated wind speed in this area: 88 mph, Exposure B). 

Siding that is intended for locations with a basic wind speed greater the 110 mph usually has a 
double-layer nail hem (Figure 3-82). This double layer strengthens the vinyl at the point where 
the nail attaches so the siding better resists tearing or pull-through of the nail head. Conventional 
vinyl siding has a single-layer nail hem. Most of the siding that was removed from the 
wall (and therefore exposed for inspection by the MAT) had a single nail hem and was thus 
not likely to have been rated for high wind locations. Although it is possible that the siding that 
stayed on the wall (and therefore wasn’t inspected) was predominantly high-wind rated, it seems 
likely that a significant percentage of the siding installed in the high wind zones of this area of 
the Texas coast is not intended for that application. This conclusion would appear reasonable, 
since winds produced by Ike varied from maximum 3-second gusts of 90 mph on the west end 
of Galveston to 110 mph on the east end of Bolivar Peninsula, and the ASCE 7-05 assigned wind 
speeds for these locations is 130 mph. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 

3-59 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-82. 
Vinyl siding rated for high 
wind has a double-layer 
nail hem 


As with any building system, even high-wind rated siding needs to be properly installed in order 
to function as designed. The MAT observed several common installation methods that tended 
to allow siding to be blown from the building by Hurricane Ike, including: 

1. Starter strip attachment along the first (lowest) course of siding 
Starter strips consist of a nail hem and locking profile that matches the shape of the lock on the 
lower edge of the siding panel (called the buttlock). The starter strip is fastened to the lowest 
part of the wall and the first course of siding is locked into it. If this lock is not strong, wind can 
get under the first course and detach it from the starter strip. The loose piece of siding will place 
stress on the lock of the course above, as well as its own nail hem, leading to successive loss of 
courses up the wall. In order to protect against this, the starter strip should be designed for use 
with the particular profile (shape) of siding being used, and the siding should be firmly locked 
into the starter strip. 

Proper use of the starter strip is particularly important with elevated structures, where the wind 
passes at high velocity underneath the structure as well as against the walls. On Galveston Island, 
Bolivar Peninsula, and Tiki Island, where elevated houses were predominant, a large percentage 
of siding loss originated at the lowest course and led to loss of the courses above. The MAT 
saw numerous instances where a “generic” starter strip (having just a bulge, rather than a lock 
shaped to match the siding) was used. In other cases, J-channels, which do not lock into the 
panel at all, or field-fabricated substitutes for starter strips were used. Elevated structures with 
poorly implemented starter strips were most vulnerable to siding loss starting at the lowest edge 
of the elevated wall (Figure 3-83). 

Vinyl siding installers should be advised to use starter strips that are specifically designed for 
the brand and model or profile of the siding that will be used and generic starter strips should 
be avoided. Installers should consult the manufacturer’s instructions to identify the starter 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

strip to be used. Installers should also test the fit of the starter strip to the siding to make sure 
it locks securely before installing. On elevated structures, the starter strip should not extend 
below the lowest edge of the vertical wall or the exposed edge may catch the wind blowing 
under the house. 

Figure 3-83. 
Use of a generic starter strip contributed to loss of siding 
on this house. The shape of the starter strip did not properly 
match the shape of the siding, and the relatively weak strip 
bent up at the end and released the siding lock (blue inset). 
The edge of the starter strip also extended slightly below the 
edge of the building, which further contributed to the failure 
(red inset) (Tiki Island, TX; Hurricane Ike estimated wind 
speed in this area: 88 mph, Exposure B). 
2. Locking of mid-wall siding courses 
Siding loss frequently begins midway up the wall rather than at the bottom course. Many of 
these instances are the result of failure to fully and securely lock the buttlock of the siding into 
the locking shape of the siding course below. This can happen when the siding is pulled up too 
tightly before being nailed, thereby placing it under tension when the siding is not fully pushed 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 

3-61 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

into the lock (Figure 3-84), or when the siding is allowed to sag before nailing. These modes of 
failures frequently occur when installers try to align the horizontal course lines on one wall with 
those of an adjacent wall by installing several courses loosely. 

Each course of siding should be installed by pushing the buttlock firmly upward into the lock 
of the course below until it snaps into place and goes no further. The siding should be held in 
the lock by pushing up from the bottom while the first several fasteners are placed to hold it in 
position. Siding should never be pulled up from the nail hem. When properly installed, siding 
should be able to slide back and forth without undue force; neither tight fasteners at the nail 
hem nor friction in the buttlock should prevent the siding from sliding. Installers should properly 
locate the starting points for siding on adjacent walls and check alignment of horizontal 
lines every course or two to avoid needing to make adjustments further up the wall. 

Figure 3-84. 
Loosely locked panel led 
to the siding failure of 
this Tiki Island, TX, house. 
The buttlock should be 
fully inserted into lock of 
panel below (Hurricane 
Ike estimated wind speed 
in this area: 90 mph, 
Exposure B). 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

3. Using utility trim at windows and other locations where the top edge of siding must be 
removed 
When a course of siding intersects the bottom of a window or other large opening, a section 
of the top portion of the panel must be removed to fit around the window. With the nail hem 
removed, special techniques must be used to stabilize and secure the cut edge of siding. An 
accessory called utility trim must be installed beneath the window. The cut edge of the siding 
panel is notched with a snap lock punch. The edge of the siding is inserted into the utility trim, 
which grabs and holds the punched notches (Figure 3-85). A furring strip may need to be used 
underneath the utility trim to place it at the right level to match the angle of the siding. An 
overlap between adjacent siding panels should never be located directly beneath a window or 
similar opening (Figure 3-86). The same technique must be used to finish the top course of siding 
where the nail hem is cut off to match the location of the eave line. 

Figure 3-85. 
Use of utility trim under window to securely attach 
cut and notched siding section 

SouRcE: VSI INSTALLATIoN MANuAL 

Figure 3-86. 
Siding partially detached 
by wind as a result of 
improper placement 
of joint directly under 
window. Factory-notched 
end is not held by utility 
trim (Tiki Island, TX; 
Hurricane Ike estimated 
wind speed in this area: 
88 mph). 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Although most cases of vinyl siding loss can be traced to improper installation techniques or use 
of incorrect products and accessories, there is room for improvement in product testing and 
documentation. It is recommended that the vinyl siding industry reevaluate the test standards 
used for validating the strength of the siding material and its installation. ASTM D 3679, Standard 
Specification for Rigid Poly (Vinyl Chloride) (PVC) Siding, specifies a 1.5 product safety factor. 
Given the MAT observations, this safety factor appears to be too low. ASTM D 5206, Standard Test 
Method for Windload Resistance of Rigid Poly (Vinyl Chloride) (PVC) Siding, tests the product installations 
using a static load. Considering the flexible nature of vinyl siding and the dynamic nature 
of wind loading, a dynamic test appears to be prudent for vinyl siding. Manufacturers should 
provide clearer and more explicit information in the product literature (including Web sites) 
and installation instructions on high-wind applications, including explicit information on: 

n Windload ratings for specific products and profiles, and any limitations or conditions 
needed to achieve the rated performance 
n Specific accessories (e.g., starter strips, trim pieces) needed to provide the rated 
performance 
n Any applicable fastener specifications, spacing frequency, and installation details needed 
for high-wind applications 

3.2.2.3 Fiber Cement Siding 
The MAT observed fiber cement siding on many residential structures, primarily as a reside 
cladding (Figure 3-87).3 The observations included lap (plank) siding of varying exposures, perforated 
soffit material, and siding panels and sheathing material below elevated structures. 

Figure 3-87. 
Fiber cement plank siding, 
installed as a reside over 
the original plywood 
siding, was torn from 
this West Bay, Galveston 
Island, TX, house 
(Hurricane Ike estimated 
wind speed in this area: 
100 mph, Exposure C) 


3 Reside cladding relates to the installation of a cladding material over an original cladding, usually sandwiched between foam 
board insulation and house wrap. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Lap siding damage varied from the loss of a few planks to entire walls (Figure 3-88). In most cases, 
the siding had been blind nailed at each stud (Figure 3-89), which is standard for non-high 
wind zones. Published ratings and ICC Evaluation Reports for the application of fiber cement 
lap siding in high wind zones require that the siding be face nailed through both layers of siding 
at the lap joint, shown in Figure 3-90. The spacing of the nails (16-inch or 24-inch) and permitted 
material exposure is dependent upon the thickness and width of the siding boards and wind 
zone. 

Figure 3-88.
Fiber cement lap siding
was blown off this West
Bay, Galveston Island,
TX, house (Hurricane Ike
estimated wind speed
in this area: 100 mph,
Exposure B)



Figure 3-89. 
Damaged fiber cement 
plank siding. Note that 
blind nailing alone (red 
arrows) is recommended 
only for 90-mph or less 
installation. Higher wind 
zone installations should 
include both blind and 
face nailing (Hurricane 
Ike estimated wind speed 
in this area: 93 mph, 
Exposure B). 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-90. Standard wind zone installation High wind zone installation 

Another area of vulnerability for fiber cement siding is the exposure of the underside of the 
first course of lap siding, or the bottom edge of panel siding. In setting the first (lowest) course 
of lap siding, a shim is used to place the board at the proper angle. If the shim is not flush with 
the bottom of the board, a lip is formed that can catch wind pressure, and force this board up. 
The first board acts as a lever under the second, and loss of siding progresses up the wall. This is 
a particular issue with elevated structures, where the wind accelerates under the building. The 
MAT observed numerous cases where a projecting lip of the first course on an elevated structure 
led to significant loss of siding (Figure 3-91). If the bottom edge of the panel extends below the 
lowest edge of the elevated structure, or there is a gap between the panel and the lowest structural 
member, wind pressure can catch the edge and pry the panel off. 

Shims under lap siding should be placed flush with the bottom of the first course, and panel 
siding should be fastened tightly to the substrate so that no gap is created at the lowest edge. 
Consideration should be given to placing a trim piece below the lower edge of the siding to 
fully close off the edge. Neither lap siding nor panel siding should extend below the lowest 
structural member of an elevated building, where it would be exposed to the full force of the 
wind (Figure 3-92). 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-92. 
Loss of fiber cement panels due to lower edge exposure. Inset shows 
lower edge exposed to wind (Bolivar Peninsula, TX; Hurricane Ike 
estimated wind speed in this area: 110 mph, Exposure C). 
Figure 3-91. 
Shim placement (red 
arrow) allowed the lower 
edge (red circles) of siding 
to be exposed, resulting 
in loss of siding at several 
locations around this 
elevated structure on 
Bolivar Peninsula, TX. Note 
the blind nailing shown 
by blue arrows (Hurricane 
Ike estimated wind speed 
in this area: 110 mph, 
Exposure C). 
Edge of panel extends below 
structural member; exposed to wind. 
Shim 
HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 

3-67 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

3.2.2.4 Wood and Hardboard Siding 
Most of the older houses on Galveston Island, Tiki Island, and Bolivar Peninsula that were originally 
constructed with plywood or hardboard siding had been re-sided with either vinyl or fiber 
cement siding. The performance of the remaining plywood and hardboard siding was basically a 
function of maintenance. The clapboard-sided house shown in Figure 3-93 was well maintained 
and performed well, though the second floor failure was produced when a non-breakaway wall 
was destroyed by surge. Failure of the plywood siding shown in Figure 3-94 appeared to be the 
result of decayed plywood removed by the wind pressures. 

Figure 3-93. 
Clapboard-sided house 
with siding that performed 
well; damage resulted 
from failure of a non-
breakaway wall (Tiki 
Island, TX; Hurricane Ike 
estimated wind speed 
in this area: 103 mph, 
Exposure C). 

Figure 3-94.
Decayed plywood
siding removed by wind
pressures (Tiki Island, TX;
Hurricane Ike estimated
wind speed in this area:
103 mph, Exposure C)



MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

3.2.3 Doors 
Failure of an exterior door has two important consequences. First, the failure can cause a rapid 
increase in internal pressure, which may lead to exterior wall, roof, interior partition, ceiling, 
or structural failure. Second, wind can drive rainwater through the opening, causing damage to 
interior contents and finishes, and leading to the development of mold. The essential elements 
of good high-wind door performance include product testing to ensure sufficient factored 
strength to resist design wind loads (both static and cyclic loading); suitable anchoring of the 
door frame to the building; proper flashing, sealants, tracks, and drainage to minimize water intrusion 
into wall cavities or into occupied space; and, for glazed openings, the use of laminated 
glass or shutters to protect against windborne debris damage, as discussed in Section 3.2.4. 

Many door failures observed by the MAT were the result of flood loads, which doors are not designed 
for. Personnel door failures in slab-on-grade houses and houses elevated below the BFE 
were commonly seen, along with catastrophic failure of the entire house. Garages with garage 
doors are frequently installed below elevated homes, and are designed to fail due to flood loads 
in conjunction with breakaway walls. 

3.2.4 Windows and Shutters 
Most building codes incorporate the wind provisions from ASCE 7-05 and require that buildings 
within the most hazardous portion of the hurricane-prone region, called the windborne debris 
region, be equipped with shutters or impact-resistant glazing and designed as enclosed structures. 
The 2003 IRC allows a residence without either protection to be designed as a partially 
enclosed structure (as if the windows and doors are broken out). Designing a partially enclosed 
structure typically requires upgrading structural components and connections. In Texas, the 
TDI requires opening protection for both Seaward and Inland I zones (refer to Figure 3-69 for 
wind zone locations). Few impact-resistant glazed window units were observed by the MAT, with 
homeowners and builders generally opting to use shutters to provide debris impact protection. 
However, the MAT observed four new houses being constructed on the east beach of Galveston 
Island that were installing impact-resistant glazing (Figure 3-95). 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-95.
Impact-resistant door and window glazing in new East Galveston, TX, house. Inset shows manufacturer’s label
indicating glazing is impact resistant.


The MAT observed that glazing at most houses was protected by some form of shutter. The shutter 
types varied from simple plywood to roll-down shutters. Figures 3-96 to 3-101 show a variety 
of shutters seen by the MAT. 

Figure 3-96. 
Clear Lake, TX, house 
with plywood shutters 
installed on the accessible 
first floor and roll-down 
shutters installed on the 
less accessible second 
floor (Hurricane Ike 
estimated wind speed 
in this area: 90 mph, 
Exposure B) 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-97.
Tiki Island, TX, house
with adjustable shutters
(Hurricane Ike estimated
wind speed in this area:
103 mph, Exposure C)


Figure 3-98.
Texas City, TX, house with
corrugated clear plastic
shutters (Hurricane Ike
estimated wind speed
in this area: 88 mph,
Exposure B)


HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-99.
Traditional wood swinging
shutters on Tiki Island, TX
(Hurricane Ike estimated
wind speed in this area:
103 mph, Exposure B)


Figure 3-100. Corrugated 
metal shutters on house 
in West Galveston, TX 
(Hurricane Ike estimated 
wind speed in this area: 
95 mph, Exposure C) 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-101. 
Snapped-on vinyl canvas 
window covers (red 
arrows) in West Bay, 
Galveston Island, TX 
(Hurricane Ike estimated 
wind speed in this area: 
90 mph, Exposure B) 

Since Ike’s winds were below design wind speeds in both Texas and Louisiana, no failures or debris 
impacts were observed. The 2006 IRC/IBC and TDI require that all shutters be attached to 
the building structure and not to the window frame, siding, or veneer (Figure 3-102); they require 
that all shutters be tested to ASTM Standards E 1886 and E 1996. The MAT observed plywood 
shutters mounted directly to the wall cladding or window frame as seen in Figure 3-103. Further 
information regarding shutters can be obtained from Technical Fact Sheet 26, Shutter Alternatives, 
in FEMA 499. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-102. Common methods for plywood shutter attachment to wood-frame and masonry walls 

SouRcE: FEMA 499 TEcHNIcAL FAcT SHEET 26 

Figure 3-103. 
Plywood shutters installed 
into the wall cladding (red 
circles) in Clear Lake, TX 
(Hurricane Ike estimated 
wind speed in this area: 
90 mph, Exposure B) 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

3.2.5 Soffit and Roof Ventilation 
Hurricane winds can drive large amounts of water through attic ventilation openings. The accumulating 
water soaks insulation, which can lead to mold growth and, in some cases, to the 
collapse of ceilings. Attic ventilation can be provided by a number of devices, most of which 
have been observed to allow water intrusion under certain conditions and some of which have 
been observed to blow away. These devices include: 

n Soffit vents 
n Ridge vents 
n Gable end vents 
n Off-ridge vents (not observed by Ike MAT) 
n Gable rake vents (not observed by Ike MAT) 
n Mechanical vents – wind-powered turbines or electric-powered fans (not observed by Ike MAT) 

3.2.5.1 Soffits 
The opening created where a roof extends beyond the plane of the wall below (called eaves 
on the downslope side of a roof and a rake for the end of a gable roof) is normally closed off 
with a soffit. Soffits typically have small openings, slots, or perforations to provide ventilation 
to the attic, this ventilation is particularly important in the hot, humid climate of coastal Texas 
and Louisiana. Soffit venting allows air to enter the attic space, circulate through the attic, and 
be exhausted through passive vents (ridge vents, gable-end vents) or mechanical vents (either 
wind-powered turbines or electric-powered fans). The soffits along the roof eave and rake are 
the primary line of defense against entry of wind-driven rain into attics. Rain driven into attics 
can cause significant damage as water soaks through ceiling materials and into the interior of 
the building. 

In non-high wind regions, a soffit is typically attached with fasteners to the roof structure only 
on one side—on the house side or to the underside of the fascia—if at all. In such installations, 
the channel formed by a bend in the fascia cover receives and supports the end of the soffit. In 
high-wind zones, most soffit manufacturers indicate the soffit should be attached at both ends 
and at intermediate points so that there is no span greater than 12 inches. 

The primary materials observed for roof soffits in the surveyed area were vinyl, aluminum, fiber 
cement, and plywood. In general, fiber cement and plywood soffits remained connected to the 
house (Figure 3-104), while vinyl and aluminum soffits were more likely to have blown off. 

By far the most frequently observed form of failure was loss of the aluminum fascia cover from 
the fascia board (the vertical board used to close off the end of eave spaces or form the outer 
edge of the rake), as shown in Figures 3-105 and 3-106. The fascia cover normally covers the 
ends of vinyl and aluminum soffits. Aluminum fascia covers are typically nailed every few feet 
along the length with color matched trim nails. The IRC currently has no guidelines for the installation 
of fascia covers. Vinyl fascia covers are also available. They are typically installed using 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

utility trim along the upper side of the fascia board. The continuous nature of the attachment 
may provide better wind resistance than the aluminum covers. The MAT did not observe any 
vinyl fascia covers. 

Figure 3-104. 
Fiber cement soffit 
remained connected; soffit 
vent slots shown with 
red arrows. Fiber cement 
plank siding was damaged 
(Tiki Island, TX; Hurricane 
Ike estimated wind speed 
in this area: 88 mph, 
Exposure B). 


The frequent loss of fascia covers is a significant concern. In most instances where the fascia cover 
was observed by the MAT to be fully or partially removed, the soffit itself remained in place 
or lost only a few sections, as further shown in Figures 3-105 and 3-106. The loss of the fascia 
cover can increase the risk of loss of the soffit. Even where the soffit remains, rain can be driven 
directly past the exposed soffit ends. The MAT did not have access to the interior of houses to 
determine whether interior moisture damage was a frequent result of fascia cover loss, but such 
damage would be expected. 

The frequency of fascia cover failure suggests that design and installation of this component 
needs to be better addressed in construction standards for buildings in high wind locations. 
The fact that most soffits stayed in place despite loss of the fascia cover suggests that most soffit 
installations were performed properly or the design was sufficiently robust to resist winds that 
occurred during Ike. However, loss of soffits exposed by fascia cover removal would likely have 
been much greater had winds approached design speeds. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-105. 
Loss of aluminum fascia 
cover (red arrows) 
exposed ends of vinyl 
soffit (blue arrows) to 
direct entry of wind-
driven rain (Tiki Island, TX; 
Hurricane Ike estimated 
wind speed in this area: 88 
mph) 

Figure 3-106. 
Loss of fascia cover (red 
arrow) led to loss of soffit 
(blue arrow), exposing the 
attic to wind-driven rain 
(San Luis, TX; Hurricane 
Ike estimated wind speed 
in this area: 93 mph, 
Exposure C) 

3.2.5.2 Ridge Vents 
The exhaust portion of the attic ventilation system includes ridge vents, gable end vents, off 
ridge vents, and mechanical vents. The MAT only observed damage produced by ridge vents 
and gable end vents. To accommodate the ridge venting system, roof decking is cut or left short 
of the gable ridge beam. Buildings can be retrofitted for ridge vents by cutting the gable slot in 
the existing deck. The ridge vent is normally the last part of the roof cover to be installed. The 
ridge vent should be a tested assembly with a baffle in front of the vent tube that provides passageway 
for hot attic gases to escape. The baffle is intended to trip any flow of wind and water 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

blowing up the surface of the roof and deflect it over the top of the roof ridge. The ridge vent 
should be installed with stainless steel screws, not roofing nails, into the roof structure. The 
MAT team was unable to climb onto residential roofs, but it was reported by the homeowner 
that the damage to a second floor ceiling shown in inset of Figure 3-107 was the result of a leaking 
ridge vent. 

Figure 3-107. 
The roof ridge vent (red 
arrows) on this Bolivar 
Peninsula, TX, home 
leaked, and it is presumed 
that the water was shed 
down the underside of 
the roof decking and/ 
or structure, thereby 
producing ceiling damage 
along the wall of this 
second story room 
(Hurricane Ike estimated 
wind speed in this area: 
110 mph, Exposure C) 


3.2.5.3 Gable End Vents 
Virtually all gable end vents (Figure 3-108) will leak when the wall they are mounted on faces 
into the wind-driven rain. The pressure developed between the outside surface of the wall and 
the inside of the attic are sufficient to drive water uphill for a number of inches and, if there is 
much wind flow through the vent, water carried by the wind will be blown considerable distances 
into the attic. Remedial measures include installing shutters, preferably on the outside of the 
house (Figure 3-109). The gable end vent shown in Figure 3-110 was not attached to the building 
structure and was blown off the apartment building. 

Refer to FEMA Hurricane Ike Recovery Advisory, Minimizing Water Intrusion Through Roof Vents 
in High-Wind Regions (Appendix D), for further discussion of off-ridge vents, gable-end rake 
vents, and mechanical vents. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-108. Gable end vent (red arrow) Figure 3-109. Shuttered gable end vent (red arrow) 

Figure 3-110. 
Gable end vent blew off this Galveston, TX, 
Back Bay apartment building (Hurricane Ike 
estimated wind speed in this area: 90 mph, 
Exposure B) 
3.2.6 Exterior-Mounted Equipment 
Residential condensing units should be elevated in floodprone areas. Condensers at many residences 
observed by the MAT were supported on cantilevered platforms as shown in Figure 
3-111. Cantilevered platforms are preferred because they are less susceptible to damage from 
floodborne debris impacts than are pile or knee-brace supported platforms. Outside floodprone 
areas, condensers are normally mounted at grade or on rooftops. In all cases, the units should 
be permanently anchored to prevent them from being moved (Figure 3-112). 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 

3-79 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-111. 
Typical cantilevered 
condenser (Jamaica 
Beach, TX; Hurricane Ike 
estimated wind speed 
in this area: 80 mph, 
Exposure B) 

Figure 3-112.
Improperly secured
condensing unit was
knocked from its platform
(Kahala Beach, TX;
Hurricane Ike estimated
wind speed in this area:
80 mph, Exposure B)



Maintenance should be considered in the design and installation of elevated supports. Figure 
3-113 shows a unit that is closely caged, making maintenance difficult. If units are caged, the 
railings should either be removable or the platform made sufficiently large to allow service to 
the unit. Further information regarding equipment protection can be obtained from Technical 
Fact Sheet 29, Protecting Utilities, in FEMA 499. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-113. 
Elevated condenser is tightly 
enclosed, making service 
access difficult (Bermuda 
Beach, TX; Hurricane Ike 
estimated wind speed in this 
area: 95 mph, Exposure C) 

3.3 Other Damage 
3.3.1 Breakaway Walls 
The Ike MAT found that solid breakaway walls performed as expected in the vast majority of cases. 
The walls broke free without causing significant or structural damage to elevated buildings. In 
some cases, failure of the breakaway walls led to propagation of damage to the building exterior 
above the lowest floor (Figure 3-114). In other cases, attachment of utilities to breakaway walls either 
prevented their successful breakaway, or contributed to utility damage (Figure 3-115). 


Figure 3-114. 
Propagation of damage 
above lowest floor when 
breakaway walls broke free 
(Seabrook, TX) 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-115. 
Attachment of utilities to 
breakaway wall may have 
prevented the wall from 
breaking away, thereby 
resulting in additional 
damage to the structure 
(Galveston Island, TX) 


The MAT did not document specific cases where breakaway wall panels from one building led 
to identifiable damage to adjacent buildings. However, the ubiquitous presence of breakaway 
walls beneath elevated buildings would undoubtedly increase the quantity of floodborne debris 
during a severe flood event, and could potentially contribute to damage at adjacent structures. 

The MAT observed some breakaway walls in excess of 11 feet high (Figure 3-116). While FEMA 
promotes elevating houses above the BFE (i.e., adding freeboard), one of the unintended consequences 
appears to be an increased size of floodborne debris elements due to the presence 
of these taller breakaway walls. 

The MAT observed that louvered panels remained intact longer than solid breakaway walls under 
the same flood conditions. As a result, houses with louvered panels had less flood-related 
damage (and repair cost) and contributed less floodborne debris. Figure 3-117 shows louvered 
panels that allowed Ike floodwaters to pass into and out of the below-BFE enclosure without 
damage to the louvered panels. These louvers were installed on the same building shown in 
Figure 3-116, where the solid breakaway wall panel was displaced by floodwaters trapped inside 
the enclosure. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-116.
This 11-foot high breakaway wall panel was pushed
out by floodwaters trapped inside the enclosure
(Galveston Island, TX)



Figure 3-117. 
Louvered panels allowed 
Ike floodwaters to pass 
into and out of the below-
BFE enclosure without 
damage to the panels. 
The building shown here 
is the same as in Figure 
3-116, where a solid 
breakaway wall panel 
broke away. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Numerous building owners in one community 
(Tiki Island, TX) were observed to be replac-Based on these observations, the Ike MAT 
ing solid breakaway walls lost during Ike with recommends the use of louvered panels 
louvered panels (Figure 3-118). This action will rather than solid breakaway walls below the 

BFE. See the Hurricane Ike Recovery Advi


reduce future flood damages and can result in 

sory, Enclosures and Breakaway Walls, in 

lower flood insurance premiums. Zone V flood 

Appendix D for more details on this topic. 

insurance premiums are much less for a building 
with a below-BFE enclosure formed by louvers 
than for a building with an enclosure formed by breakaway walls. A building with an enclosure 
formed by louvers is classified the same as if it had insect screening or open lattice, i.e., as “free 
of obstructions,” while a solid breakaway wall enclosure results in a “with obstruction” rating for 
the building. 

Figure 3-118. 
Solid breakaway walls 
lost during Ike are being 
replaced with louvered 
panels (Tiki Island, TX) 


3.3.2 Sheathing on the Underside of Elevated Buildings 
Sheathing is typically installed on the underside of the lowest-floor joists on elevated buildings. 
Besides protecting batt insulation that is placed between joists, sheathing can also protect electrical 
and plumbing lines from floodborne debris. A variety of sheathing materials are used, 
most often sheets of plywood, hardboard, or fiber cement panels. The sheathing is sometimes 
covered with vinyl soffit material, or left uncovered and painted. 

In locations where the water level or waves reached the elevation of the building, sheathing and 
any covering was frequently found to be partially or completely removed (Figures 3-119 and 
3-120). This was particularly true of the thinner panel types, such as .-inch fiberboard. Other 
forms of damage, such as gouges from floodborne debris, were observed on the underside of 
panels. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-119. 
Plywood sheathing 
removed by storm surge 
(Jamaica Beach, TX, house 
on West Bay) 
Plywood Sheathing 
Figure 3-120. 
Fiber cement board 
sheathing (red arrow) 
was removed from the 
underside of this house, 
which was elevated 
approximately 12 feet 
above ground level (Bolivar 
Peninsula, TX) 

Several examples of vinyl soffit attached directly to floor joists, without sheathing, were observed 
by the MAT (Figure 3-121). 

Where floodwaters did not reach the underside of the building, damage due to wind accelerating 
underneath the building was often observed. In these cases, vinyl soffit was often blown 
off. In some cases, but not all, the sheathing above the soffit was also removed. The vinyl soffit 
covering on the Tiki Island house shown in Figure 3-122 was probably blown off by wind action 
rather than storm surge. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-121. 
Tiki Island house with 
vinyl soffit applied without 
sheathing removed by 
storm surge (netting was 
used to contain insulation 
in joist space) 

Vinyl Soffit 
Figure 3-122. 
Vinyl soffit covering 
over plywood sheathing 
partially removed (Tiki 
Island, West Galveston 
Bay; Estimated Hurricane 
Ike wind speed: 103 mph, 
Exposure C) 

For further information on the performance of sheathing on the underside of elevated buildings, 
see FEMA 489. 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

3.3.3 Parking Slabs and Grade Beams 
Many of the houses supported on pile foundations that the MAT visited had concrete slabs constructed 
at grade. These slabs were typically used as parking slabs beneath the elevated houses 
(Figure 3-123). Some of the slabs were thin (less than 4 inches thick); others were much thicker. 
Some had thickened edges and interior sections that acted as grade beams, presumably to 
stiffen the foundation. Virtually all slabs were reinforced with welded wire mesh and/or steel 
reinforcing bars. 

Figure 3-123. 
Typical concrete parking 
slab beneath a pile-
supported house (Bolivar 
Peninsula, TX) 


Many of the slabs failed once they were undermined. Where the piles were embedded deep into 
the ground, the slab was either undermined (and sometimes settled) as shown in Figure 3-124, 
or the slab collapsed without visible damage to the foundation (Figure 3-125). 


Figure 3-124. 
Undermining of concrete 
slab that settled 
but remained intact 
(Galveston Island, TX) 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-125. 
Pile-founded house 
with a slab thickened to 
create grade beams. The 
unthickened portion of 
the slab collapsed when 
undermined (Galveston 
Island, TX). 


Where the piles were thought to be less well embedded, failure of the slab could have caused 
the pile foundation to rotate or rack (Figure 3-126). The MAT believes this was more common 
with older houses, and was likely a result of portions of the slab causing eccentric loads on the 
piles and the transfer of flood forces from the slab to the foundation. 

The MAT observed instances where the weight of the slab likely contributed to foundation failure 
and building settlement, illustrated in Figures 3-127 and 3-128. Figure 3-127 shows a Holly 
Beach, LA, house under construction at the time of Ike. The piles and elevated floor beams had 
been placed, and a thick slab had been cast; when Ike undermined part of the slab, it cracked 
and settled, pulling some of the piles and beams downward. Figure 3-128 shows a house at Surf-
side Beach, TX, that was subject to considerable scour and erosion—when the slab settled and 
collapsed, it could have pulled part of the house lower as it went. Pile embedment appears to 
have been the larger issue at the houses shown in Figures 3-127 and 3-128, and insufficient embedment 
likely allowed the slabs to induce or worsen building settlement. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-126. 
Slab failure probably 
contributed to foundation 
damage (West Galveston 
Island, TX) 

PHoTo couRTESy oF 
STuART ADAMS, LSu 
HuRRIcANE cENTER 

Figure 3-127. 
Slab undermining and 
settlement during Ike 
probably pulled piles 
downward. Inset shows 
that the dropped piles also 
caused the floor beam to 
deflect (Holly Beach, LA). 
HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 

3-89 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-128. 
The weight of this slab, 
undermined due to scour 
and erosion, could have 
contributed to settlement 
and racking of this 
elevated house (Surfside 
Beach, TX) 


The MAT observed several houses on Galveston Island where parking slabs were constructed 
in 4-foot square sections and unreinforced. This method of construction is consistent with that 
recommended in FEMA 55, Third Edition. Where these slabs were observed, their failure did 
not appear to adversely affect foundations or elevated buildings (Figure 3-129). Section III of 
the Galveston County Dune Protection and Beach Access Plan (2006) requires use of unreinforced 
fibercrete or concrete slab sections (maximum 4-inch thickness) within 200 feet of the vegetation 
line in eroding areas. 

Figure 3-129. 
Thin, unreinforced parking 
slab sections separated 
when undermined and 
collapsed in place, with 
no apparent adverse 
impact to the foundation 
(Galveston Island, TX) 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

3.3.4 Mold and Contamination 
Hurricanes introduce various forms of contaminants and pollution into floodwaters and flooded 
buildings. Hurricanes also lead to the post-event growth of mold in wind- and flood-damaged 
buildings. Figure 3-130 illustrates one of many examples of mold and mildew growth observed 
by the Ike MAT. Guidance on cleanup and restoration of flooded buildings can be found in the 
Hurricane Katrina Recovery Advisory 2, Initial Restoration for Flooded Buildings (July 2006d), and 
Katrina Recovery Advisory 4, The ABCs of Returning to Flooded Buildings (July 2006e). 

Figure 3-130. 
Mildew and mold forming 
on wall sheathing 
following flooding (Golden 
Meadow, LA) 


3.3.5 Other Issues and Problems 
The MAT observed other construction deficiencies and community enforcement problems. 
While the details of these particular deficiencies are not known, their existence indicates potential 
compliance issues that should be monitored and addressed in communities visited by 
the MAT. Figure 3-131 shows a case where floor beams and joists were improperly notched to 
allow for plumbing installation. This practice can weaken structural members and should only 
be done at the direction of a structural engineer. Figure 3-132 shows a case where flood vents 
did not penetrate through the entire enclosure wall—if this installation was complete when observed 
by the MAT, this practice is a clear violation of NFIP flood opening requirements. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-131.
Floor joists and beams
were notched to allow
for plumbing (Sulphur,
Calcasieu Parish, LA)



Figure 3-132. Flood vent openings (red circles) that do not extend through the walls (Hackberry, Cameron Parish, 
LA) 

3.4 Manufactured Housing 
The MAT visited several communities in south Louisiana and east Texas where large numbers 
of manufactured homes were damaged by some combination of storm surge, waves, floodborne 
debris, and wind. In some locations in southwest Louisiana, manufactured housing installed 
after Hurricane Rita was not elevated to or above the BFE. This may have occurred in existing 
manufactured housing parks where an NFIP exception allows some homes to be elevated 3 feet 
above grade, even where this is lower than the BFE, or it may have occurred through incorrect 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

application of the 3-foot exception. Whether this practice was allowed by the NFIP exception or 
not, the result was the same—large numbers of manufactured homes installed below the BFE 
after Hurricane Rita were heavily damaged or destroyed by Hurricane Ike. 

3.4.1 Texas 
In San Leon, TX, the MAT observed a manufactured home that was knocked off its foundation 
and destroyed. The home was located in a Zone AE (BFE = 11 feet) approximately 150 feet landward 
of a rip-rapped shoreline. High water marks in the area indicated water levels were over 12 
feet NGVD, and 5 feet or more above grade. Coastal A Zone conditions (wave heights between 
1½ and 3 feet) likely existed there during Ike. 

The home, shown in Figure 3-133, was placed on short, unreinforced and un-mortared “dry 
stack” masonry piers placed on pre-cast concrete pads (at 8-foot centers [+/-]), and was secured 
with ground anchors spaced at 4-foot centers (+/-) with metal stabilizer plates. 


Figure 3-133. Destroyed manufactured home (San Leon, TX) 

SouRcE: GooGLE MAPS FoR INSET SHoWING LocATIoN 

Evidence suggests that the home was displaced from its piers by moving floodwaters or waves. 
Scour, undermining the concrete pads beneath the piers, may have contributed. Ground anchor 
failures were not noted, but the straps connecting the home to the anchors had torn away 
from the house’s anchorage points (Figure 3-134). HUD’s 2007 Manufactured Home Construction 
and Safety Standard (MHCSS), 24 CFR Parts 3280 and 3285, place this site in a Wind Zone II. The 
MHCSS requires Wind Zone II homes to be secured and anchored to their steel frames and to 
wall ties. No wall ties were observed. This suggests that the home was either non-compliant or 
was installed before the HUD standards went into effect. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-134.
Scour depressions existed
around the masonry piers,
pads, and ground anchor
stabilizer plates (San Leon,
TX)



In Oak Island, TX, some manufactured homes were elevated on timber piles. The elevation 
prevented foundation failure, but some of the homes were still damaged by inundation 
(Figure 3-135). 

Figure 3-135. 
Manufactured home in 
Oak Island, TX. The house 
foundation did not fail, 
but the elevation was 
insufficient to prevent 
damage from inundation. 
MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

3.4.2 Louisiana 
The MAT observed that Zone A manufactured homes elevated at or above the BFE/ABFE on 
reinforced concrete or reinforced masonry piers with proper anchoring performed well. The 
best performance of foundations in Zone V was found to be timber piles embedded sufficiently 
to withstand erosion and scour effects. Zone V homes on piers resting on concrete pads often 
failed due to flood and erosion/scour effects. 

3.4.2.1 Cameron and Vermilion Parishes 
Many of the manufactured homes that were present in Cameron and Vermilion Parishes in 
2005 are no longer in place. Those structures were destroyed by Hurricane Rita and in many 
instances, had not been replaced. Many of those that had been replaced after Rita and not elevated 
to or above the BFE/ABFE were damaged by Ike. 

The manufactured homes shown in Figure 3-136 are located immediately east of the Cameron 
Parish offices along LA Hwy 82 in South Cameron. They are currently located within Zone A, 
but will be classified as Zone V when the pending new flood maps are adopted. The homes were 
not properly anchored and were forced off their foundation piers by the storm surge. 

Figure 3-136. 
Two manufactured homes 
in Cameron, LA. Homes 
were displaced off 
foundations and siding 
peeled due to inundation 
and storm surge of 
approximately 4 feet 
above ground. 


3.4.2.2 Jefferson Parish 
The manufactured home shown in Figures 3-137 and 3-138 is located in Zone A (BFE = 10 feet, 
ABFE = 12 feet) on Grand Isle, Jefferson Parish. Ike floodwaters were approximately 6 feet 
deep and did not reach the home, which was elevated in compliance with NFIP requirements. 
Support framing was in place, and strapping secured the walls and the steel chassis frame to 
the foundation. While effective during Ike, the strapping was installed using non-conventional 
methods. Its ability to resist a design wind event could not be determined. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

The home experienced some wind damage (vinyl siding and portions of the roof covering were 
dislodged) despite the fact that the Ike wind speeds and wind pressures were far below the HUD 
and ASCE 7-05 design wind speeds and pressures. Section 305 of the MHCSS, 24 CFR Part 3280, 
requires that siding be designed to resist wind loads for Exposure C specified in ANSI/ASCE 
7-88, or wind pressures specified the HUD Standard table titled Table of Design Wind Pressures. 
The MHCSS places Jefferson Parish in HUD Wind Zone III, and the Table of Design Wind 
Pressures requires exterior coverings within 3 feet of corners to resist +/- 58 psf, and exterior 
coverings in other areas to resist +/- 46 psf. The ASCE 7-05 wind pressures (for a 150 mph basic 
wind speed) are +49/-65.7 psf at the corners of a building and +49/-53.1 psf in other areas. 

Figure 3-137. 
This elevated 
manufactured home in 
Grand Isle of Jefferson 
Parish, LA, had siding and 
roof damage, but did not 
move from its foundation 


Figure 3-138. 
The framing and anchoring 
system of the house 
shown in Figure 3-137. 
Strapping secured the 
home’s walls and frames 
to its foundation. While the 
strapping held the home to 
its foundation during Ike, 
it could not be determined 
if the strapping would 
resist a design wind event 
(Estimated wind speed 
during Ike: less than 60 
mph, 3-second gust). 


MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


3.4.2.3 Lafourche Parish 
The Zone A home shown in Figure 3-139 in Lafourche Parish was elevated, but not above the 
BFE. It suffered flood damage from about 3 to 4 feet of water above the floor. Interviews with 
nearby residents indicated the floodwaters reached the eaves on the house with the green roof 
to the right. Flood velocities were not sufficient to shift the manufactured home off of its foundation 
and the floodwaters rose slowly enough to allow leakage into the home, thereby preventing 
the home from becoming buoyant and floating off its foundation. 

Figure 3-139. 
Zone A manufactured 
home in the Golden 
Meadows section of 
Lafourche Parish, LA, 
sustained 3 to 4 feet of 
flooding above the floor, 
but did not shift or float 
off of its foundation. The 
red arrow indicates the 
flood level reported by 
neighbors. 

3.4.2.4 Manufactured Home Anchoring and Support Systems 
Manufactured homes in SFHAs must be placed on foundation systems that will resist flotation, 
collapse, and lateral movement (Figure 3-140). The 2005 edition of the NFPA 225, Model 
Manufactured Home Installation Standard, contains performance requirements for flood-resistant 
manufactured home installations. The 2008 edition, issued in January 2009, also contains 
prescriptive flood-resistant installations. Other flood-resistant foundation solutions will be contained 
in the revised FEMA 85, scheduled to be completed in 2009. 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 

Figure 3-140. 
Prescriptive Flood-
Resistant Foundation 
Design 


3.5 Mitigation Projects 
The MAT typically looks at funded mitigation projects to determine if the projects were successful. 
The MAT visited 27 residential mitigation projects in Louisiana and 10 in Texas. Thirty-four 
of the projects visited were elevation projects, and three were acquisition projects. All of the 
projects received funds through the Hazard Mitigation Grant Program (HMGP) or through 
Increased Cost of Compliance payments via NFIP flood insurance policies. There were no structures 
visible at the three acquisition project sites, and the land had been cleared and restored. 

Three of the 34 elevation projects had not been undertaken at the time of the MAT visit. The remaining 
31 elevation projects had been completed and were successful as far as preventing Ike 
flood damage—none of the elevated buildings appeared to have been flooded during Ike, even 
though many of the building sites were inundated. Most of the buildings had been elevated on 
masonry piers, tall masonry columns, or timber piles. 

While most of the elevation projects appeared to have been constructed in accordance with 
applicable codes and standards, some load path deficiencies (Figure 3-141) were noted that indicate 
possible project design and/or compliance problems that should be investigated. Some 
of the elevated buildings sustained wind damage to the building envelope during Ike (Figure 
3-142); this is likely a result of older homes being elevated, as opposed to a problem with the 
elevation project itself. 

MITIGATION ASSESSMENT TEAM REPORT HURRICANE IKE IN TEXAS ANd LoUISIANA 


PERfORMANcE Of RESIdENTIAl BuIldINGS (flOOd ANd WINd), ONE- TO TWO-fAMIly ANd MulTI-fAMIly 


Figure 3-141. 
Zone A house elevated 
with Increased Cost of 
Compliance funds on 
masonry piers (Iberia 
Parish, LA). There was 
no evidence of pier 
reinforcement, mortar 
between masonry blocks, 
or ties between the piers 
and the elevated home. 


Figure 3-142. 
House elevated with 
Increased Cost of 
Compliance funds (Kemah, 
TX). Inset shows evidence 
of wind damage to roof 
(Hurricane Ike estimated 
wind speed in this area: 
90 mph, Exposure B). 

HURRICANE IKE IN TEXAS ANd LoUISIANA MITIGATION ASSESSMENT TEAM REPORT