Email
Newsletters
R&I ONE®
(weekly)
The best articles from around the web and R&I, handpicked by R&I editors.
WORKERSCOMP FORUM
(weekly)
Workers' Comp news and insights as well as columns and features from R&I.
RISK SCENARIOS
(monthly)
Update on new scenarios as well as upcoming Risk Scenarios Live! events.

Reducing Cat Risks

Wind Turbines Slow Down Hurricane Winds

Hurricane winds are dissipated by offshore wind farms.
By: | March 6, 2014 • 3 min read
Topics: Catastrophe | Energy
03062014WindTurbines

Off the New York coastline would be a perfect place for an array of wind turbines, according to a Stanford professor. It would not only offer clean energy to the Big Apple but it would protect it the next time a Superstorm Sandy comes calling.

“If you have a large enough array of wind turbines, you can prevent the wind speeds [of a hurricane] from ever getting up to the destructive wind speeds,” said Mark Jacobson, a professor of civil and environmental engineering at Stanford University.

Computer models demonstrated that offshore wind turbines reduce peak wind speeds in hurricanes by up to 92 mph and decrease storm surge by up to 79 percent, said Jacobson, who worked on the study with University of Delaware researchers Cristina Archer and Willett Kempton.

//

“The additional benefits are there is zero cost unlike seawalls, which would cost about $30 billion,” he said, noting that the wind turbines “generate electricity so they pay for themselves.”

The researchers studied three hurricanes, Sandy and Isaac, which struck New York and New Orleans, respectively, in 2012; and Katrina, which slammed into New Orleans in 2005. Generally, 70 percent of damage is caused by storm surge, with wind causing the remaining 30 percent, he said.

That’s why onshore wind farms would not be as effective, he said. While they would reduce the wind speed, they wouldn’t impact storm surge.

In 2013, one of the “most inactive” Atlantic hurricane seasons on record, insured losses totaled $920 million, according to Guy Carpenter, which relied on information from the Mexican Association of Insurance Institutions. The most noteworthy events were Hurricane Ingrid in the Atlantic and Tropical Storm Manuel in the Pacific, which displaced thousands as they caused excessive rainfall, flooding and mudslides.

Advertisement




According to the Insurance Information Institute, Katrina was the costliest hurricane in insurance history, at $48.7 billion, followed by Andrew in 1992 at $25.6 billion and Sandy at $18.8 billion. Economic losses, of course, were much higher.

Wind turbines, which can withstand speeds of up to 112 mph, dissipate the hurricane winds from the outside-in, according to Jacobson’s study. First, they slow down the outer rotation winds, which feeds back to decrease wave height. That reduces the movement of air toward the center of the hurricane, and increases the central pressure, which in turn slows the winds of the entire hurricane and dissipates it faster.

The benefit would occur whether the turbines were immediately upstream of a city, or along an expanse of coastline. It could take anywhere from tens of thousands to hundreds of thousands of wind turbines off the coast to offer sufficient hurricane protection.

At present, there are no wind farms off the U.S. coastline, although 18 have been proposed for off the East Coast. Proposals have also been made for off the West Coast and the Great Lakes. There are 25 operational wind farms off the coast of Europe.

“Overall,” Jacobson and his colleagues concluded in the study, “we find here that large arrays of electricity-generating offshore wind turbines may diminish hurricane risk cost-effectively while reducing air pollution and global warming, and providing local or regionally sourced energy supply.”

Anne Freedman is managing editor of Risk & Insurance. She can be reached at afreedman@lrp.com.
Share this article:

Black Swan

Toxic Tornado

When a nuclear reactor melts down due to a powerful tornado, deadly contamination rains down on a metropolitan area.
By: | August 4, 2014 • 9 min read
082014_02_cs_tornado_craft_shop

It is a warm, humid spring day in Dallas/Fort Worth when strong thunderstorms begin to develop alongside a high-altitude weather system that includes strong winds and convective energy coming from the Rocky Mountains.

By mid-afternoon, the atmosphere reaches a tipping point. A massive supercell thunderstorm along the weather front produces large, damaging hail and what is later designated as an EF5 tornado, with winds in excess of 200 mph.

The most recent tornado of this size as designated by the National Weather Service was on May 20, 2013, when an EF5 struck Moore, Okla., killing 24 people, flattening neighborhoods and schools, and injuring more than 350 people.

This Texas tornado is much, much worse.

Video: An EF5 tornado in May 2013 flattened much of Moore, Okla.

Moving in the usual southwest to northeast direction, it creates a damage path about 1 mile wide and nearly 200 miles long, and directly strikes the Comanche Peak Nuclear Power Plant in Glen Rose, Texas, about 40 miles west of Fort Worth and 60 miles west of Dallas.

The power plant’s reactor was built to withstand winds up to 300 mph, but it can’t withstand what happens after the tornado throws around multiple gas-filled tanker trucks, which explode and kill numerous workers.

Matthew Nielsen, director of Americas product management at RMS, created the model for our Comanche Peak Nuclear Power Plant black swan scenario.

Matthew Nielsen, director of Americas product management at RMS, created the model for our Comanche Peak Nuclear Power Plant black swan scenario.

Debris fills the air as the powerful winds destroy much of the plant’s emergency equipment, making it impossible to maintain proper conditions and temperature within the reactor. The remaining power plant workers feverishly try to manually shut down the nuclear reactor before it melts down. They can’t.

When the reactor’s heat exceeds the ability of the plant’s processes to cool it down, radioactive gases begin to snake their way into the funnel stacks. The radioactive contamination is carried by the ferocious winds directly toward Dallas/Fort Worth.

Communication fails as area power lines go down, so it is difficult to warn the 7 million residents of the Metroplex, as Dallas/Fort Worth is known. Residents know the tornado has been sighted and try to prepare, but they don’t know that deadly airborne toxins are being carried toward them.

The Damage

About 10,000 homes and 700 commercial structures in the direct path of the tornado are completely destroyed and another 35,000 suffer damage, according to a model built by RMS. Roofs are ripped off apartment houses and multi-family dwellings. Vehicles are tossed around like toys, and with the storm striking at rush hour, workers on the roads are exposed to flying debris and high winds.

Advertisement




Even with residents sheltering in basements and safe rooms, fatalities reach into the 500-700 range — putting this event in line to be the deadliest tornado in U.S. history, after the Tri-State tornado of 1925, which killed 695 people in Missouri, Illinois and Indiana.

But it is the unseen radioactive contamination that ultimately makes the deadliest mark on the area.

Immediate fatalities from radiation poisoning number about two dozen, but as the contaminated rainfall seeps into the ground soil and water supply, the long-term health of the residents — and their descendants — is jeopardized. So, too, are the cattle and other agricultural products of Texas, which leads the nation in the number of ranches and farms it holds.

Chernobyl and Fukushima are the only events of a similar nature, even though the United States has seen its own recent near misses.

The radioactivity causes large swaths of area to be cordoned off, making it difficult to repair transmission and power lines as well as homes and businesses.

“The hard truth is that many businesses will close and many people will move from the area,” said Todd Macumber, president of international risk services, Hub International.

Chernobyl and Fukushima are the only events of a similar nature, even though the United States has seen its own recent near misses.

In 2011, a tornado knocked out power to the Browns Ferry Nuclear Power Plant near Huntsville, Ala., requiring the shut down of its three reactors. The plant fired up backup diesel generators until power was restored. The storm also disabled the plant’s sirens, which are needed to warn nearby residents in a crisis.

That same year, a tornado barely missed damaging 2.5 million pounds of radioactive waste at the Surrey Power Station in southeastern Virginia, although it touched down in the plant’s electrical switchyard and disabled power to the cooling pumps. The operators needed to activate backup diesel generators to run the two reactors until power was restored.

Twenty-eight years after the radioactive disaster at Chernobyl in 1986, some parts of the Ukraine remain a toxic wasteland. And in Japan, an initial evacuation area of about 2 miles surrounding the Fukushima Daiichi Nuclear Power Plant was soon widened to about 12.5 miles.

About 300 tons of  highly radioactive water has leaked from storage tanks at the Fukushima Daiichi Nuclear Power Station.

About 300 tons of highly radioactive water has leaked from storage tanks at the Fukushima Daiichi Nuclear Power Station.

Now, three years after three of Fukushima’s six reactors melted down, the area is still unlivable and 40 miles away, diagnoses in children of thyroid cancer, which is caused by radiation poisoning, are skyrocketing, according to some reports.

Nearly 16,000 people died in the 2011 earthquake and tsunami that struck Japan, causing the meltdown. About 160,000 people were evacuated, 130,000 buildings were destroyed and $210 billion in damage was sustained.

The Texas scenario has a lot of variables, said Matthew Nielsen, director of Americas product management at RMS, who created the model for our Comanche Peak Nuclear Power Plant black swan scenario.

The likelihood of a tornado, with thunderstorms and hail, causing massive structural damage is about 1 in 200 years, he said. Such an event would result in at least $20 billion in insured losses and uninsured losses of about the same amount.

But a tornado following the exact path as this scenario — striking the power plant and heading into the Dallas/Fort Worth Metroplex — has a much, much smaller chance — about 1 in 10,000 years.

“Given the fact that tornadoes are very rare, it isn’t something that I think people should be screaming and running around frantically about,” Nielsen said. “But it’s certainly something that could happen.”

As for losses due to the radiation? “There’s not a lot of historical data points that we can confidently say that that portion would be x or y billion,” he said.

The Recovery

Any rebuilding will be delayed by the threat posed by radioactive contamination, which may spread over a large area via the thunderstorms and storm water runoff.

From an insurance perspective, all personal and commercial lines of insurance have a nuclear energy hazard exclusion. American Nuclear Insurers (ANI) provides third-party liability insurance for all power reactors in the United States.

“We are responsible for the insurance coverage protecting the operators from claims alleging bodily injury or property damage offsite from [radioactive] materials,” said Michael Cass, vice president and general counsel at ANI, a joint underwriting association with 20 insurance company members.

082014_02_cs_tornado_nuclear350pxNEWThe ANI was created under the Price-Anderson Act of 1957 and provides a primary policy limit of $375 million for claims due to offsite consequences from the release of radioactive materials from the 100 operating nuclear power plants in the United States. It also covers some plants that are shut down or in the process of being decommissioned, he said.

The ANI also covers costs related to emergency response and evacuation, including food, clothing and shelter, he said.

The joint underwriting association also administers an additional excess layer of about $13.2 billion, the costs of which would be borne by the power plant operators, and would be apportioned equally among them.

For any claims above $13.6 billion (which includes both the primary and excess layers), the Price-Anderson Act requires the U.S. Congress to “take steps to come up with a scheme to provide full compensation to the public and to continue claims payments,” Cass said.

“They could assess or tax the energy industry in some fashion or form. It doesn’t say that specifically, but that is what is alluded to.”

None of the insurance companies that are ANI members would be adversely affected if such a black swan event were to occur, he said.

Advertisement




“There would be a loss reserve recorded on their balance sheets, per participation in our pool, but we do have funds set aside for these catastrophic events where we wouldn’t be requiring any additional funds,” Cass said.

Damage to the power plant itself would be covered by Nuclear Electric Insurance Ltd., which insures electric utilities and energy companies in the United States. Current limits are $1.5 billion per site on the primary program, and up to $1.5 billion per site in its excess program.

Allan Koenig, vice president, corporate communications at Energy Future Holdings, which operates Comanche Peak, said the plant is robustly protected. It has two independent systems that can provide off-site power as well as backup diesel generators, to allow the units to be safety shut down in the event of natural catastrophes.

He also noted the plant has safety shields for fuel storage casks, a 45-inch-thick steel-reinforced concrete containment building wall, and fire protection redundancies.

As for the affected businesses and homeowners, they may be left in a swirling vortex of coverage confusion. The situation would have the flavor of what happened after Superstorm Sandy, when coverage often depended on whether damage was caused by flooding or wind surge.

The question for Texas insureds would be whether the damage was caused by the tornado or by the radioactivity.

“It’s an incredibly complex question and a complex issue that is really only solvable and resolvable if and when the incident occurs,” said John Butler, vice president of the environmental practice at Hub International.

“What it boils down to is the chicken and the egg scenario,” he said. “What came first? Either event has the ability on its own to create a total loss.”

Resilience and redundancy should be the key takeaways from this, said Peter Boynton, founding co-director of the Kostas Research Institute for Homeland Security at Northeastern University in suburban Boston.

“If we can retain a percentage of the critical function of whatever system we are talking about, the difference between 0 percent and 30 percent when the bad thing happens is huge.” — Peter Boynton, founding co-director of the Kostas Research Institute for Homeland Security, Northeastern University

Instead of viewing catastrophic events from an emergency management perspective, where the discussion revolves around what was — or was not — managed well, it’s better to look at the way design can lead to “continuity of function,” he said.

When Boynton was head of emergency management for the state of Connecticut, he managed the statewide response in 2011 to Hurricane Irene, which knocked out 70 percent of the state’s electric grid, leaving residents unable to access many gas stations, ATMs and grocery stores.

If the state had designed a “resiliency approach” prior to the event, it could have built in a pre-determined amount of redundancy into the system so that, say, an additional 20 percent or 30 percent of the grid remained viable.

Advertisement




“If we can retain a percentage of the critical function of whatever system we are talking about, the difference between 0 percent and 30 percent when the bad thing happens is huge,” Boynton said.

In the Texas scenario, if the crisis planning included a redundancy for warning nearby residents even when the power and communication lines failed — such as by using satellites to create a minimal level of continuity — the amount of death and destruction could have been lessened.

“Otherwise, we really are setting ourselves up for an impossible discussion,” he said. “You can’t just pick up these pieces at the moment of crisis. You have to understand how system design can play a role.”

Analyzing such a black swan scenario is a useful exercise, said Justin VanOpdorp, manager, quantitative analysis, at Lockton.

“Can this actually happen? Yes. Will it? Maybe not,” he said. “I think what it does is, it helps to think through it just to be prepared for those situations when they do arise.”

R8-14p32-36_02Tornado2.indd

____________________________________________________________________

Additional 2014 black swan stories:

Bigger Than the Big One

When the 8.5 magnitude earthquake hits, sea water will devastate much of Los Angeles and San Francisco, and a million destroyed homes will create a failed mortgage and public sector revenue tsunami.

Sub-Zero Sucker Punch

A double dose of ice storms batter the Eastern seaboard, plunging 50 million people and three million businesses into a polar vortex of darkness and desperation.

Anne Freedman is managing editor of Risk & Insurance. She can be reached at afreedman@lrp.com.
Share this article:

Sponsored: Rising Medical Solutions

Beware of Medical Hyper-Inflation!

Workers’ comp medical costs are spiking in hidden pockets across the country.
By: | August 4, 2014 • 5 min read
SponsoredContent_Rising

Historically, medical inflation rates nationwide have been fairly consistent. However, data is now showing that medical inflation is not a “one size fits all” phenomenon, with hyperinflation spikes occurring in some locations…but not others.

This geographical conundrum means hyperinflation can occur as narrowly as two hospitals having dramatically different charges on the same street in Anytown, USA. So, uncovering these anomalies is akin to finding the proverbial needle in a haystack.

“In recent years, workers’ compensation saw claim frequency decline, while severity rates went up. This basically means that increased job safety has offset increased medical costs,” explained Jason Beans, CEO of Rising Medical Solutions, a national medical cost management firm. “So, whenever a client’s average cost-per-claim went up, it was almost always caused by catastrophic, outlier-type claims.”

But beginning in 2013 and extending into 2014, Beans said, things changed. “I’ve never seen anything like it in my 20-plus years in this industry.”

SponsoredContent_Rising“Our analytics made it very clear that small pockets around the country are experiencing what could only be described as medical cost hyperinflation. The big spikes in some clients’ claim costs were driven by a broader rise in medical costs, rather than catastrophic claims or severity issues.”
– Jason Beans, CEO, Rising Medical Solutions

Data dive uncovers surprising findings

On a national level, most experts describe medical costs increasing at a moderate annual rate. But, as often is the case, sometimes a macro perspective glosses over a very different situation at a more micro level.

“Our analytics made it very clear that small pockets around the country are experiencing what could only be described as medical cost hyperinflation,” explained Beans. “The big spikes in some clients’ claim costs were driven by a broader rise in medical costs, rather than catastrophic claims or severity issues.”

This conclusion is supported by several key data patterns:

  • Geographic dependency: While many payers operate at the national level, only relatively small, geographically clustered claims showed steep cost increases.
  • Median cost per claim: The median cost per claim, not just the average, increased greatly within these geographic clusters.
  • Hospital associated care: Some clusters saw a large increase in the rates and/or the number of services provided by hospital systems, including their broad array of affiliate locations.
  • Provider rates: Other clusters saw the same hospital/non-hospital based treatment ratios as prior years, but there was a material rate increase for all provider types across the board.
  • Utilization increases: Some clusters also experienced a larger number of services being performed per claim.

One of the most severe examples of hyperinflation came from a large Florida metropolitan area which experienced a combined 47 percent workers’ compensation healthcare inflation rate. Not only was there a dramatic increase in the charge per hospital bill, but utilization was also way up and there was a shift to more services being performed in a costlier hospital system setting.

“The growth of costs in this Florida market stood in stark contrast to neighboring areas where most of our clients’ claim costs were coming down or at least had flat-lined,” Beans said.

An Arizona metropolitan area, on the other hand, experienced a different root cause for their hyperinflation. Regardless of provider type, rates have significantly increased over the past year. For example, one hospital system showed dramatic increases in both charge master rates and utilization. “Even with aggressive discounting, the projected customer impact in 2014 will be an increase of $773,850 from this provider alone,” said Beans.

ACA: Unintended consequences?

So what is going on? According to Beans, a potential driver of these cost spikes could be unintended consequences of the Affordable Care Act (ACA).

First, the ACA may be a contributing factor in recent provider consolidation. While healthcare industry consolidation is not new, the ACA can prompt increased merger and acquisition efforts as hospitals seek to improve financials and healthcare delivery by forming Accountable Care Organizations (ACO). ACOs, the theory goes, can take better advantage of value-based fee arrangements in existing and new markets.

“As hospital systems grow by acquisition, more patients are being brought under hospital pricing structures – which are significantly more expensive than similar services at smaller facilities such as independent ambulatory surgery centers and doctors’ offices,” Beans said.

Unfortunately, there is little evidence that post-consolidation healthcare systems have become more efficient, only more expensive. For example, a recent PwC study reported that hospital IT infrastructure consolidation alone is projected to add 2 percent to hospital costs in 2015.

Another potential ACA consequence is group health insurers may have less incentive to keep medical costs down. An ACA provision requires that 85% of premium in the large group market must be spent on medical care and provider incentive programs, leaving 15% of premium to be allocated towards administration, sales and subsequent profits. “Fifteen percent of $5000 in medical charges is a lot less than 15% of $10,000,” said Beans. “This really limits a group health carrier’s incentive to lower medical costs.”

How do increased group health rates relate to workers’ comp? In some markets, a group health carrier may use its group health rates for their work comp network so any rate increase impacts both business types.

In the end, medical inflation is inconsistent at best, with varying levels driven by differing factors in different locations – a true “needle in the haystack” challenge.

What to do?

Managing these emerging cost threats, whether you have the capabilities internally or utilize a partner, means having the tools to pinpoint hyperinflation and make adjustments. Beans said potential solutions for payers include:

  • Using data analytics: Data availability is at an all-time high. Utilizing analytical tools to spot problem areas is critical for executing cost saving strategies quickly.
  • Moving services out of hospital systems: Programs that direct care away from the hospital setting can substantially reduce costs. For example, Rising’s surgical care program utilizes ambulatory service centers to provide predictable, bundled case rates to payers.
  • Negotiating with providers: Working directly with providers to negotiate bill reductions and prompt payment arrangements is effective in some markets.
  • Underwriting with a micro-focus: For carriers, it is vital that underwriters identify where these pockets of hyperinflation are so they can adjust rates to keep pace with inflation.

“This trend needs to be closely watched,” Beans said. “In the meantime, we will continue to use data to help payers of medical services be smarter shoppers.”

Contact Rising Medical Solutions: info@risingms.com | www.risingms.com

This article was produced by the R&I Brand Studio, a unit of the advertising department of Risk & Insurance, in collaboration with Rising Medical Solutions. The editorial staff of Risk & Insurance had no role in its preparation.


Rising Medical Solutions provides medical cost containment, care management and financial management services to the workers’ compensation, auto, liability and group health markets.
Share this article: