Edited Version September 8, 1999
EIIP Classroom Online Presentation

"Are Chemical Plants at Risk For Hurricane Winds"

Dr. Marc Levitan
Acting Director - Hurricane Center
Associate Professor of Civil Engineering
Louisiana State University (LSU)

The original unedited transcript of the September 8, 1999 online Virtual Classroom presentation is available on the EIIP Virtual Forum (http://www.emforum.org). The following version of the transcript has been edited for easier reading and comprehension. Typos were corrected, date/time/names attributed by the software to each were deleted but content of discussions, questions, and responses are as stated by each participant. Answers from the presenter to questions by the audience are grouped beneath the appropriate question to facilitate meaning.


Amy Sebring: Welcome to the EIIP Virtual Classroom!

One quick note about any URL's that may be used in the session; they are live links and you can click on them and view the referenced site in your browser window. Subsequent "slides" may display behind your chat window, so you may need to bring the browser window forward. Background information for today's session may be found at <http://www.emforum.org/vclass/990908.htm>.

We will have a presentation for about thirty minutes, and then have audience Q&A for the last thirty minutes. We will review the instructions for Q&A as we are about to begin that portion.


We are pleased to welcome Dr. Marc Levitan, Acting Director of the Hurricane Center at Louisiana State University (LSU) and Associate Professor of Civil Engineering. Dr. Levitan has been actively engaged in wind engineering research, practice, and teaching for the past 15 years. His primary research focus is in the fields of wind loading on structures, wind damage assessment, wind damage mitigation, and hurricane sheltering and evacuation issues. A couple of photos of him may be found on "Dr. Marc's Wind Engineering Page" at <http://www.geol.lsu.edu/workshop/students/levitan/index.html>

Marc is going to tell us about his paper "Are Chemical Plants at Risk from Hurricane Winds?" A copy of the full paper may be downloaded from the Hurricane and Hazards Internet Conference at <http://hurricane.lsu.edu/internetconf.htm>.

Marc, thank you very much for joining us today.


Marc Levitan: Thank you, Amy. Glad to be here.

Many places on the US coastline are home to large concentrations of chemical/petrochemical/agrichemical plants. Areas such as Corpus Christi and Houston TX, Lake Charles, New Orleans, and Baton Rouge Louisiana, the Mississippi and New Jersey coasts, among others.

Are these plants at risk from hurricane winds?

These plants are filled with complex structures, many of which there are no wind loading standards available for. Figure wind1 shows a typical plant, with process structures on the left and a tank farm on the right (photo credit: unknown).

[Slide 1]

The two biggest potential impacts from hurricane wind damage are release of hazardous materials into the environment and the economic impact of repairing damaged structures, including loss of production while the repairs are underway.

Hurricane preparedness procedures that lessen one of the potential impacts often have a negative effect on the other potential impact. Consider the common case of a steel storage tank.

These are very thin shell structures. Empty, they can be easily damaged or destroyed by moderate hurricane wind speeds, like squeezing an empty soda can between two fingers. A full tank is much stronger, much like an unopened soda can.

Hurricane preparedness procedures often call for tanks to be filled. This significantly increases the wind resistance of the tank, thereby reducing the potential that a release will occur. However, in the event of damage, the entire contents of a full could potentially be released.

When questioned about the wind threat, a common response is 'Our facility is designed to withstand hurricane winds. The general attitude seems to be that expected wind damage would be minor, such as damage to storage trailers, sheet metal roofing and siding failures, minor debris damage, etc. The idea that major process structure, pipe rack, or other engineered structure could suffer significant wind damage or even collapse does not appear to be widely considered.


One reason for this relative lack of concern is the perception that industry-wide, the various structures found in the process industries do not fail in high winds, so there must not be much of a problem. With the exception of the past few years, the Atlantic basin has enjoyed nearly three decades of relative calm, experiencing less frequent and intense tropical storms and hurricanes than the long term average.

Institutional memory also plays a part in this sense of security. Most of the plants have experienced several minor hurricanes or tropical storms in the past 30 years and come through with fairly minor damage, which has served to build up confidence in the ability of the structures to withstand hurricanes.

Major hurricanes are defined as Category 3 and above. Very few major hurricanes have hit areas with chemical plants. Hurricane Andrew fortunately came ashore between New Orleans and Lafayette, skirting by the concentration of plants on the lower Mississippi River.

Hurricane Andrew did, however, take a large toll in offshore structures. Nearly 300 manned and unmanned platforms were damaged or destroyed due to the combined effects of wind and waves, including some newer structures. A refinery on the island of St. Croix suffered serious damage during the passage of Category 4 Hurricane Hugo in 1989. In addition to other damage, several petroleum storage tanks were damaged which led to a significant oil spill.

Hurricane Alicia in 1983 in the Houston area did some damage to area chemical and petrochemical plants. Thus, recent history demonstrates the potential for significant damage to chemical and petrochemical structures in Category 3 and stronger hurricanes.


A commonly held perception is that if an engineered structure is "designed to code" then it will withstand all but perhaps the most intense hurricane with very minimal damage.

Current design standards have significant limitations that are not always appreciated. They are primarily life safety codes. Factors such as environmental damage and economic impact are not considered. Additionally, codes and standards do not address some of the structure types most commonly found in chemical plants.

The fact that many of the structures found in the plants are not addressed by building codes and standards has led to wide variations in design procedures and estimates of wind loads, even if the wind speed is specified. Most firms involved in the design or operation of chemical and petrochemical facilities developed their own wind loading guidelines.

Wind loads on open frame structures are not addressed by codes. These are highly complex structures, supporting vessels, reactors, piping, and other equipment. An example of this type of structure is shown in Figure wind2.

[Slide 2]

Wind loads on a sample open frame structure were computed using 13 design guides from large petrochemical producers and engineering/construction firms. A wind speed of 96 mph was used in all for all calculations. The results (listed below in thousands of pounds) varied from 90,000 to 182,000, more than a factor of 2.

90 112 116 116 120 129 141 141 148 154 160 162 182

average = 136

The implications of the wide ranging variations are significant. If a plant specified that this structure be designed to ASCE 7 wind speeds, one firm would be designing for a wind force as low as 90 kips, while another would be designing for as much as 182 kips (1 kip=1000 pounds).

The best estimate of the actual wind load 147 kips. While many design guides yielded results in this ballpark or more conservative, five out of 13 estimated wind loads roughly 20-40% too low. Pipe rack structures are very common in plants, supporting the pipes that connect the various processes throughout the plant (see Figure wind3).

[Slide 3]

Results of a similar comparison study for pipe racks yielded much greater variations, as much as 400% from the lowest estimate to the highest.


There have been very few significant hurricanes in recent decades to strike chemical and petrochemical plants. Some wind damage has been seen to occur during Category 3 and larger storms, but this information is not widely known. The experience of relatively minor damage to plants during Category 1 storms in this period has perhaps lulled many into a false sense of security.

Limitations of the codes and standards under which they plants were designed are not fully appreciated. These factors have led to what appears to be an unfounded expectation that significant damage due to high winds is very unlikely.

Finally, this is occurring at a time of increasing wind risk, due to long term climactic trends of increasing Atlantic hurricane activity and the potential impact of global climate change.

Amy Sebring: Thank you, Marc. We will now take questions from the audience. If you have a question or comment, please indicate by inputting a question mark (?) to the chat screen. Then compose your question but hold it until you are recognized; then hit Enter or Send. First question, please?

[Audience Questions]


Amy Sebring: While folks are thinking, I have a question. I understand from a previous wind presentation that duration of the winds have almost an unknown impact on loading. Did your calculations factor in wind duration?

Marc Levitan: The calculations shown were for the peak gust. Some types of construction are sensitive to the duration of the load while other types are not. The structures in an industrial plant are typically steel construction (with some concrete). These structure types are not that susceptible to load duration. Wood construction, on the other hand, is much more sensitive to the duration of loading.


Russell Coile: Isn't the situation possibly even bleaker? During Hurricane Andrew, apparently every anemometer in the hurricane blew over, etc., including the Weather Service. No one, really knows the maximum gust speeds --- perhaps more than 150 mph.


Marc Levitan: Good question, Russell. In Florida, there is still a debate raging between meteorologists and wind engineers over what the surface wind speeds were. However, it appears that although in some isolated areas there were gusts in the 150 mph+ range. Most of the areas saw substantially less. Also, there was a problem with the South Florida code at the time, the specified wind speed was 120 mph, but that was for a sustained wind. They didn't have a factor to account for peak gusts. The current wind speed design map for south FL specifies peak gust speeds of 140-150 mph, depending on location.


Amy Sebring: Although you may not have been looking for it, did you find any kind of correlation between conservative design and hazard posed by the contents? Or are the same "specs" used regardless of contents?

Marc Levitan: I have not looked into that. Others in our university and a group at Tulane are now beginning to address those type of questions.


Terry Storer: Is there any factor to account for "dirty wind", those winds with all kinds of foreign objects being blown about? It would seem that what is being blown about could significantly effect structural integrity.

Marc Levitan: Another excellent question. For industrial/petrochemical type structures, the debris field would typically include metal siding and perhaps some other items from adjacent areas. Plans typically do a good job of picking up and tying down items which could present a significant debris potential. Also the nature of the steel structures is that they are not that susceptible to small debris, like residential or light commercial construction is.


Russell Coile: Do you Civil Engineers have all the appropriate wind tunnel facilities you need, or do you need both BIG and armor-plated, so to speak tunnels for possible damage?

Marc Levitan: Russell, the type of wind tunnel testing needed right now is very fundamental. We (the wind engineering community) have studied wind loads on houses and typical commercial construction for the past 30 years. We understand the aerodynamics of these structures fairly well. No one has looked into the wind loads on industrial structures in a systematic way yet, just studied individual, specific buildings

for industrial clients. Here at LSU, we have had some basic testing on open frame and pipe rack structures under way for the past 3 years, and are now building a larger wind tunnel to handle more detailed, larger models.


Erwin Prater: I was wondering, Marc, if you've had a chance to look at the effects of hurricane Camille (1969, I believe) in Baton Rouge/New Orleans. I was wondering how the chemical plants faired during that storm.

Marc Levitan: Erwin, I don't think Camille caused much damage in N.O/B.R. to plants, or if so, I haven't heard about it.

Erwin Prater: I wasn't sure. That was well before my time in Louisiana. Amy had a question about conservative design and the type of chemicals stored. The Army stores nerve agents in underground bunkers, partially for this reason.


Terry Storer: Does the age of the structure provide any change in the threat profile?

Marc Levitan: It can. Plants are typically very caustic environments. Corrosion and other degradation of structural integrity are very likely if the structures are not adequately maintained. Also, in areas such as Louisiana, we have had much coastal land loss plants built 50 years ago were perhaps 30 miles from the coast, but are in some cases much closer than that now. Since the winds decrease rapidly with passage overland, this land loss in effect increases the wind speed for the same storm at the same location.


Amy Sebring: Are there specific failure points, welds, bolts, etc.?

Marc Levitan: Yes. Generally (true for almost any type of failure due to any cause) failures occur at connections or due to buckling.


Russell Coile: Back to wind tunnels - Isn't NASA always talking about "technology transfer", partnerships with industry & universities, blah, blah - and the US Air Force, also? Can you use their slow speed facilities?

Marc Levitan: Yes. We hope to use some NASA facilities when we need to test larger scale models but we aren't at that point yet.


Amy Sebring: I would like to ask: Our wind tunnels at LSU are capable of wind speeds of approximately 120 mph. What recommendation would you make to a local emergency manager or LEPC member? What questions should they be asking?

Marc Levitan: To work closely with their local plant managers. To find out the types of chemicals used in order to be prepared and to ask some questions about what design criteria were used for the structures.


Amy Sebring: Finally, you have alluded to your ongoing research on this topic; is there a specific direction you are taking?

Marc Levitan: We have several projects ongoing in our wind tunnels, but we also hope to expand our research into computational fluid mechanics approaches, and would also like to gather data on past failures and do field analysis of future wind-induced failures. REQUEST-if anyone has experience or knowledge of wind damage in a plant, please let me know so we can investigate and use this 'full scale data' to verify the data collected in wind tunnel experiments.

Final Question:

Amy Sebring: Marc, would you like to put up your email address?

Marc Levitan: [email protected]

Amy Sebring: Thank you very much Marc, and thank you audience. We will have a text transcript posted later today, and a reformatted version early next week. You can access these via the Transcripts link under Quick Picks on our home page.

Marc Levitan: I would also like to mention the home page address of our LSU Hurricane Center. This site http://www.hurricane.lsu.edu contains links to all of the hurricane-related research being conducted at LSU. It is still being developed, but it already contains useful and useable information, such as the Hurricanes and Chemical Hazards Internet conference, sponsored by EIIP.

Amy Sebring: Thanks, the link is also on the background page for today.

Our time is about up, but before we adjourn, Ava will give us a heads up on our upcoming events. Ava?

Avagene Moore: Thank you, Amy. And thank you, Marc, for your presentation. Excellent discussion, Marc, and good questions from the audience.

Next Tuesday, 9/14, 12 Noon EDT, our Round Table discussion is hosted by the FEMA Community & Family Preparedness program with Kellye Junchaya serving as the moderator. The topic will be CERT Programs with Sam Isenberg as our guest.

On Wednesday, 9/15, 12 Noon EDT, we will have a Panel Discussion in the Virtual Forum. The panelists will be folks involved in the Gulf States Shelter Study. Should be very interesting. Make plans to join us for both sessions next week. That's all for now, Amy.

Amy Sebring: Thank you, Ava. We will adjourn the session but you are invited to join us back in the Lobby room for a few more minutes of open discussion and to thank our guest.