partner login

BMS News

Summary of IBHS Disaster Dynamics Academy – Evaluating Garage Doors as a Damage Amplifier

Most Americans have fallen in love with the automobile, which is why miles of roadways reach across the country and so many residential structures in the U.S. have garages to keep these assets safe and clean. However, garages can become major problems in high wind events. When the doors to these large open spaces fail, it can cause major structural damage.
This past week, I was fortunate enough to attend the Insurance Institute for Business & Home Safety’s (IBHS) fourth Disaster Dynamics Academy session – Inside the Eye: Defending Homes Against Hurricane Damage. In this BMS Insight, I will share my experiences and what I learned about garage doors being a damage amplifier.

A Meteorologist’s Dream
Throughout my career, I have read countless papers about hurricanes, helped price and understand various aspects of hurricane risk, examined gigabytes of hurricane wind and claims data, and helped insurance companies recover after the devastation these events can cause. However, growing up and living as far away from an ocean as one can get, I’ve had limited firsthand experience with hurricanes. In fact, my only experience was on September 4, 2010, while living in Halifax, Nova Scotia, Canada, when Category 1 Hurricane Earl raced across the area. Most people wisely took shelter, but I took the more unconventional approach as a meteorologist to experience the hurricane-force winds for myself (as safely as possible). On that day, I headed out to Lawrencetown Coastal Heritage Park beach to experience my first hurricane. It truly gave me a new appreciation for what a hurricane can do and for what the insurance clients I support are facing every year.

This is video taken of my Hurricane Earl experience in Nova Scotia on September 4, 2010.
Driving rain made it hard to get a clear image of me on Lawrencetown Coastal Heritage Park beach, NS, Canada – Category 1 Hurricane Earl

Because I have spent most of my life living in the Upper Midwest, I’ve experienced my share of strong wind events while supporting insurance clients that have had to endure the same damaging windstorms. In fact, I’ve had firsthand experience with some of the research that the IBHS is currently conducting. I recently had a pole barn built on a big open farm field. On June 11, 2017, a strong line of thunderstorms worked their way across the western metro area of Minneapolis. A downburst of wind from these storms blew in the barn door, also causing damage to part of the roof. Luckily, there was no structural damage to the newly constructed barn, but, as a result, our replacement door was upgraded to a reinforced W4 door with a wind rating of 110 mph to prevent another similar incident.

This is the standard garage door that was installed on our pole barn. It was no match for winds of 55 to 70 mph.

Garage Doors Are a Wind Amplifier
During the Disaster Dynamics Academy, the IBHS showcased to its members the steps that can be taken to limit storm damage to various residential and commercial structures. After several years of IBHS demonstrations showing how fortified and non-fortified structures handle hurricane-force winds in its giant wind tunnel, this week they tested garage doors. After all, the garage door is a wind damage amplifier. Garage doors have a large surface area that can be affected by hurricanes and other severe convective winds events, such as downbursts or tornadoes. They are an amplifier because a breach often results in larger structural damage to the roof and walls. However, there are many questions about how this damage occurs, and the IBHS wind tunnel is the perfect place to find answers.
The IBHS conducts scientific research to improve building material performance and construction standards. Recent testing has included asphalt shingle performance research and continuous load path testing. The IBHS has also done extensive field investigation into past hurricanes and strong tornadoes, such as the 2013 Moore tornado, to help understand the various aspects of wind damage. One of the key findings from the garage door field investigation is that roughly 60% of homes had structural roof damage if the garage door was damaged, while roof damage hardly ever occurred if the garage door remained intact. This supports the notion that the garage door is an amplifier of damage.

IBHS Garage Door Testing
In the IBHS wind tunnel garage door testing, they were focused on determining the most appropriate methods and cost-effective practices to ensure that buildings are adequately constructed for the hazards they face. However, the questions around garage doors are complex. The internal pressure may be three times more than the external pressure, resulting in a large pressure difference between the inside and outside of the building. This can lead to major structural failure when the pressure equalizes. Damage that is commonly associated with a failed garage door is the collapse of an adjacent side wall. Some garage door failures may result in the loss of roof decking above the garage, or in extreme cases, the entire garage may collapse. If the garage door fails, it can also lead to a large amount of water entering the building, which can increase damage to contents from wind-driven rain.
The IBHS garage door testing looked at many different types of doors, including the double car door, single car door, twin single-car door, commercial roll-up doors, garage doors with and without automatic openers, and even insulated and non-insulated doors. There are many other variables that needed to be tested to help understand door failure, such as critical wind angles to the door and different wind speeds. The IBHS tested both wind rated doors and non-wind rated doors, as well as the types of hardware used to install the door. One could have a wind rated door, but it may come with normal hardware that could be prone to failure.

This is my raw iPhone video of the test and the resulting damage. The test showed that the door folded in, and the hardware holding the door also failed. Also notice some of the testing of the fabric roofing material letting lose.
This is up close picture of the door damage. Notice the door folded in and the hardware on the lower right also failed. Remember if you install a wind rated door the hardware needs to match.

The Test
I can’t comment on all of the initial testing the IBHS has been conducting, but I can share my experience with the Disaster Dynamics Academy garage door test session. The test involved a normal big box store double garage door without an automatic opener. The standard two stall garage area was equipped with various pressure sensors along the wall and door; after all, the pressure on the walls is critical to what causes the overall failure. As the test began, the IBHS 105 Vaneaxial fan array, each with a 350 horsepower motor, were started at a steady 50 mph and were slowly ramped up every few seconds. As the fans approached hurricane-force levels, you could hear the wind whip around inside the six-story, 21,000 square foot testing area as turbulent winds began to flex the double garage door. As expected, the door folded in and hardware failed at around 90 mph. If it were not for the bracing behind the door to support the array of sensors, the door would have likely folded in completely and been ripped off its rollers and pushed to the back of the two stall garage space. Maybe the most exciting part of the test was the fact that the fortified structure sustained very little damage to category 3 winds, which was expected, as this was a test for doors and not to see the results of building failure. If the IBHS did a building failure test with every door, the testing could take years to complete, which is why the fortified construction is so important. Given the personal experience with my barn, and then seeing first-hand how easily a garage door can fail in the IBHS test facility, I believe there needs to be more attention paid to this wind damage amplifier. Currently, however, there are very few insurance companies that underwrite for such wind variables. What’s even more concerning is that very few areas of the country have building codes requiring wind-rated doors. Hopefully, with the important work that the IBHS is doing, the most appropriate methods and cost-effective practices for garage door construction can be determined. In addition, building codes could be established and existing code standards updated, all from using science in one of the coolest facilities I have ever been in.

This is the IBHS test facility. Behind me are the 105 fans that can create winds as high as a Category 3 hurricane.

BMS Tropical Update July 11th

As mentioned in the last BMS Tropical Update, a storm is brewing in the Gulf of Mexico.  The National Hurricane Center has now named this system Tropical Storm Barry and expects it to potentially become a hurricane by Saturday and make landfall shortly thereafter, with possible impacts along the coasts of Texas and Louisiana, particularly in the city of New Orleans.  As many already know, New Orleans sits below sea level and has had its share of flooding problems in the past.  Since water is the new wind risk for the insurance industry, this BMS Insight will focus on this threat in a bit more detail, as any wind damage at this time is uncertain given the high forecast uncertainty around the overall intensity and track of the storm.

The forecast models are suggesting as much as 18 inches of rain for the region, which isn’t good, considering that the area is already inundated with rain. New Orleans has already flooded this week from rain, with many streets covered with a dangerous amount of water.  The Mississippi River is expected to rise to 19 or 20 feet by the weekend, which is near the height of the city’s levees.

By Saturday, it may have winds of 85 mph, making it a Category 1 hurricane on the Saffir-Simpson wind scale.  It should be noted, however, that there is a lot of uncertainty around the forecasted intensity of the storm. This intensity will depend on how much wind shear the storm encounters and how much time it spends over the warm waters of the Gulf of Mexico.  Currently, the worst-case scenario appears to be a Category 1 hurricane.  However, forecasts can change fast, as we experienced with Michael last year.  Most of the model guidance is suggesting a lopsided tropical storm with a landfall somewhere along the Louisiana coastline.

Wind speed is just one of the insured loss concerns from the storm.  Historically, flooding has not been a big concern for the insurance industry, but the take-up rates of flood insurance are higher in Louisiana, including New Orleans.  The reinsurance markets are taking on some of the risk from the National Flood Insurance Program, so this flood risk needs to be watched.  Also, coastal flooding tends to be the deadliest aspect of a hurricane.

The main concern here is that the storm surge combined with heavy rainfall could reach 3 to 5 feet, which would flood many of the low-lying areas of Louisiana.  This storm is expected to generate a lot of rain, with some areas potentially seeing more than a foot, which can contribute to inland flooding along the banks of rivers.

This is the current National Weather Service rainfall forecast over the next 7 days. This rainfall forecast will depend highly on the track of Barry
Current National Weather Service Forecast river level in New Orleans. This forecast will change with the track and intensity of Barry

The biggest threat from the storm may be in New Orleans, where the Mississippi River is already at 16 feet due to the very wet spring and summer upstream.  Some states along the river have seen their wettest spring in recorded history.  All of this water has worked its way down the Mississippi River, which has been high for well over a month now.  The National Weather Service is now predicting the river to crest at 19 feet on Saturday.  That is one foot below the top of the river levees. This could change given the track of the forecast. This morning the forecasted river height was 20 feet, so the forecast is changing for the better since the newest forecast is now one foot below the top of some of the levees.

So, why is this a concern?  During both Hurricane Katrina in 2005 and Hurricane Isaac in 2012, storm surge pushed well upriver. In those cases, however, the river was only at a level of 3 feet or less.  During Katrina, a Category 3 hurricane with top winds of 125 mph, the river swelled to at least 15.25 feet at the Carrollton Gage in New Orleans.  The gage stopped operating when the water reached that level.  Several barges were deposited on top of the levees in Plaquemines Parish during the storm by surge water.  During Isaac, a Category 1 hurricane with top winds of 80 mph, the river rose to 9.5 feet at the Carrollton Gage.  While the current river level in New Orleans is 16 feet, levees and floodwalls protect the city up to water heights of about 20 feet.

How high is a 20-foot-high river stage for New Orleans?  It is higher than 99.9% of the city’s land surface, higher than every human being except those in multi-story buildings and about 30 feet higher than the lowest neighborhoods.

The red dots show where levees in New Orleans are under 20 feet.  This is where over-topping of levees is most likely if the forecast of 19 feet comes to fruition on Saturday. The red dots on the north side of the Mississippi River flooded during Katrina, including the Lower 9th Ward.
These are the areas of New Orleans that flooded during Katrina.

In summary, the precursor for a bad flooding event is set for New Orleans and remains for much of the hurricane season.  Many areas of Louisiana are flat, and flooding is a big concern with this upcoming storm.  There is uncertainty at this time around the potential strength and track of the storm, which also increases the level of uncertainty around the storm surge that may be expected.  However, the rainfall forecasts have been fairly consistent over the last several days, with most of the heaviest rain expected to fall along with the storm’s forecasted path.

BMS Tropical Update: July 9th

There is an old proverb used to describe the Atlantic hurricane season: “June – too soon; July – stand by; August – come they must; September – remember; October – all over.” In fact, we suggested in our May BMS Tropical Insight that the next best opportunity for tropical storms was going to be in early July. It appears that June was indeed too soon for tropical threats to the U.S. coastline. However, with the start of July, that is expected to change with development appearing likely in the northern Gulf of Mexico later this week and into the weekend.

The NHC is giving an 80% chance of a Tropical depression forming in Gulf of Mexico later this week.


A precursor vorticity center (areas of spin) is currently heading south through Georgia and will emerge into the northeast Gulf of Mexico by Wednesday. This circulation will take advantage of warm water in the area to build thunderstorms and may slowly intensify into a tropical system.
Dry air and dust from a few major outbreaks of Saharan dust have battered most of the Atlantic basin for the last several weeks. This combined with strong wind shear has prevented tropical development across the Atlantic basin since Subtropical Storm Andrea briefly roamed the waters of the west-central Atlantic in late May.

With the Madden-Julian Oscillation (MJO) currently in a positive phase that supports upward motion over Central America, the absence of strong wind shear and warmer than normal waters across the Gulf of Mexico, it should not take much to organize a tropical named storm which, if it develops later this week, will be named Barry. For the last several days, the forecasted steering winds have suggested that any named storm would drift toward the central and western portions of the Gulf of Mexico this weekend. This would give it more time over open water and could potentially create a stronger storm. With this track the risk to the insurance industry grows due to the large amount of exposure in the Houston areas, the risk to offshore assets increases due to a number of petroleum rigs and refineries along the central and western Gulf Coast. Regardless of how much development we see, the forecast models are suggesting multiple days of showers and thunderstorms that could potentially bring flooding to the southern Plains, lower Mississippi Valley and/or the southern Appalachians starting this weekend and into next week, depending on the track and timing of the storm. Some models take the tropical storm as far west as Houston which has no issue flooding from even the smallest of weather systems. Currently, some of the guidance calls for 15″ of rain over the next 7 days for the area.

The overall model guidance from the ECMWF and GFS models have been back and forth over the last few days and the landfall location as suggested by the above GFS ensembles is anywhere from east Texas to Louisiana. Keep in mind this entire area will see a lot of rain.
Source: https://www.tropicaltidbits.com


The last time a named tropical system made landfall in the U.S. during the month of July was Tropical Storm Emily of 2017. Emily formed in the eastern Gulf of Mexico and moved into the central Florida Peninsula on the last day of July, so it’s not unusual to see the forecast of such an event happening this week. As I mentioned in our May BMS Insight, it appears that this season’s tropical activity will be more likely to develop closer to the U.S. coastline instead of out in the Main Development Region (MDR) of the Atlantic Ocean. The bigger question is just how much activity we’ll see this season as we move toward its climatological peak at the beginning of September.


Forecasters continue to call for five to eight hurricanes to develop this season. July still does not look like a very active period for named storm development, as large scale sinking air will continue over much of the Atlantic basin, along with on and off bursts of Saharan dust that may keep activity to a minimum. While El Niño conditions may suppress the total number of tropical storms and hurricanes in the Atlantic basin this season, the data is suggesting that El Niño is weakening and may become a non-factor as we head into September. With the warmest ocean temperatures right along the U.S. coastline, the insurance industry should continue to watch for tropical waves coming off Africa which could potentially strengthen into named storms when they get closer to the U.S. coastline versus the MDR in a more typical season.

Above are the different phases of the MJO and the influence on precipitation. Notice the red circled areas above. Tropical cyclones making landfall along the Gulf Coast during Phase 2 of the MJO during meteorological summer causes a statistically-significant positive precipitation anomaly to show up in the Deep South in the 1979-2008. Research here by Zhou et al.
Notice how the red circled areas match the forecasted areas for heavy rainfall over the next 5-7 days along the Gulf Coast.

We are also now in the period where the insurance industry can begin to watch the subseasonal forecast. I have mentioned before that one of the best ways to predict this is to use the MJO, which is a pulse of upward motion that travels the tropics but also creates areas of sinking stable air in other parts of the tropics. This symbiotic pattern of wetter and drier areas moves east as a unit and typically completes a full cycle in 45 to 60 days. When the MJO (or similar phenomena) promotes upward motion over the Atlantic, hurricane development and rapid strengthening becomes much more common. According to research performed by Phil Klotzbach of Colorado State University, hurricane damage in the U.S. during convectively active phases of the MJO since the early 1900s is nearly three times greater than during suppressed Atlantic phases.


Therefore, I believe the next round of tropical storms may begin during the second week of August. Based on the current state of African waves and favorable conditions for development along the U.S. coastline, we may also be seeing an active late September and early October. Until then, keep watching the skies off the Gulf Coast and along the southeastern U.S. for the development of named storms that might not coincide with the MJO phases.