Catastrophe: Put a (stronger) lid on it


Natural catastrophe models have been transformed through scientific research, growth in global and localised observation systems and modeling technology enhancements. But despite this, uncertainty remains, say Claire Souch and Michael Kistler.

One of the key drivers of uncertainty is knowledge about the actual quality of building stock. Every event is a critical opportunity to learn about the reality of how buildings respond to storm-force winds and Hurricane Ike in 2008, along with the 2004 and 2005 storms, revealed unexpected flaws in parts of the building stock.

In the aftermath of the 2004-2005 storms, and again after Hurricane Ike, extensive investigations into the paid claims revealed significant uncertainty in the building quality of the risks exposed, as well as varying practices in policy claims behaviour. However, while RMS has collected more than $18bn-worth (£11.5bn) of claims data in-house, there are still geographical gaps in the data, as not all at-risk areas of the US have had their building stock tested in the last 20 years. Therefore, the question remains: what can we do to understand the vulnerability of buildings in areas that have not experienced hurricane losses in recent years?

Completing the picture of risk
The answer to the gaps lies in carefully applying lessons from a rich sampling of claims data along with deep expertise on construction quality to areas such as the north-east US, which experiences fewer landfalling storms. To calibrate the building stock for these areas, engineering judgement and data on building quality are needed so claims analysis learnings can be applied from one region to another. Prior to Hurricane Ike, RMS' view on the vulnerability of Texan properties to hurricane-force winds already accounted for large-scale regional variations.

The quality of building stock in areas with little loss experience was estimated by taking account of building regulations (current and historical), adjusted by available regional estimates of code enforcement as captured by the Building Code Effectiveness Grade (BCEG) score published by ISO, which assigns a grade from one to 10 to a given risk, where one indicates ideal enforcement practice, while a 10 indicates no recognisable enforcement. In spite of this regional accounting for construction quality, this type of localised building quality shortfalls, especially in roof performance, shows the ongoing necessity to capture far more rigorous quality assessment.

Close examination of claims data following Hurricane Ike revealed several patterns across the building stock in the landfall area; one of the most striking was the performance of the rooves. Roof damage drives overall building losses, as it is one of the weakest elements of a property and the first to be impacted by strong winds. Once the roof is gone, the wind and rain can then penetrate the interior causing more damage and eventually partial or full structural failure.

As hurricane winds blow over, a local area of comparatively lower pressure is formed above the roof through aerodynamic forces. The combination of lower pressure above the roof and higher pressures directly below causes the covering or the roof deck to detach, depending on how strongly it is anchored down. Notably, in the presence of unfavourable workmanship, premature roof failure due to uplift causes large losses in a structure - on the other hand, if the roof deck and covering remain intact, the structure escapes with significantly less damage.

In Harris County, Texas, windspeeds at or below 90mph caused widespread roof damage during Ike, and even complete roof failures, while the building codes specified windspeed tolerances at far higher peak gusts. In one case, a relatively new retail building with stipulated wind tolerances of over 110mph failed when the peak winds at roof level during Ike were estimated at 78mph. In sub-standard properties, other irregularities can contribute to premature failure, such as flaws in air conditioning unit fittings, vents, window flashing, and other external fixtures. Moreover, irregularly shaped buildings can cause hurricane winds to apply uneven twisting forces on the roof, which can also lead to a premature breach. While the cause of poor workmanship is not always immediately clear at a location, some patterns have emerged from our post-Ike analysis: cost-cutting measures by contractors during the building boom-bust cycles of the 1980s, lack of municipal enforcement or even hastily completed repairs after prior landfalling hurricanes (such as Hurricane Alicia in 1983).

RMS recently convened a panel of highly experienced roofing consultants with deep knowledge of regional construction quality gained through many years' worth of building inspection programmes, to quantify the relative workmanship of the roof systems by region in the US, especially in areas with little recent storm activity. This study not only highlighted regional differences in building quality but variations in workmanship, even at the contractor level. Moreover, patterns of enforcement and compliance have been closely analysed; the study revealed varying local regulatory practices ranging from robust building inspection and compliance programmes to contractors' self-inspection.

This building consultant expertise, combined with wind-load resistance engineering simulation tools, enables a more thorough assessment of building performance. It helps to create a robustly quantified perspective of structural vulnerability in areas with few landfalling storms, like the north-eastern US, therefore reducing vulnerability uncertainty and providing an enhanced view of the risk.

Claire Souch is vice president and Michael Kistler is senior manager of Nat Cat & Portfolio Solutions at RMS

  • LinkedIn  
  • Save this article
  • Print this page  

You need to sign in to use this feature. If you don’t have an Insurance Post account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an indvidual account here: