|
TEAM
|
A
LESSON IN TANK DESIGN FROM HURRICANE ANDREW ANDREW
STRUCK HOME In 1969 I was transferred from Pittsburgh, PA. to Miami. My first Florida home was in South Miami Heights an area west of Perrine, that was devastated by Hurricane Andrew. My first hurricane experience came in the summer of 1969 when Camille ripped through the Hattisburg, Mississippi area. My boss, an officer in the National Guard, was called into action to help with the cleanup effort. This experience taught me that a hurricane was not a force to be taken lightly. I saw the aftermath of hurricanes David, Elena and Hugo. TEAM Consultants performed inspections of tanks in Charleston, S.C. where several petroleum tanks had side walls blown in as a result of Hugo. These tanks were empty and obviously not designed for winds above 100 m.p.h. These images taught respect for a hurricane's fury. If Andrew had headed for Tampa, I would have prepared for its arrival. I would have felt comfortable in the fact that my concrete block house with a barrel tile roof and real plaster on the walls would have been strong enough to withstand any storm. I was not prepared to see the destruction that occurred in South Florida. On August 27, four days after the storm hit, I made a trip to south Dade County to make a assessment of several damaged storage tanks and check up on my two oldest children who still live there. My daughter and her husband had been evacuated from their home in a new subdivision east of Old Cutler road, south of Cutler Ridge. They went to my son’s home located about a mile north of Perrine and about two blocks west of U.S.#1. My son and his wife live in an older home (concrete block and real plaster wall construction, built in the late 50’s or early 60’s). My daughter and her husband lost almost everything in their home which is severely damaged. My son’s home had roof damage and several windows blew out, but the walls did not appear to have structural damage. To use the word awesome to describe the overall devastation is an understatement. A strip about 25 miles wide from Florida City to Cutler Ridge is virtually leveled. Most of the buildings are so severely damaged they are going to need to be rebuilt from the ground up. Most buildings will need to be inspected for structural damage. There are no trees left intact. Those trees that are still standing have no branches. I noticed several chain link fence rows where the fence posts were bent over about 10 degrees from horizontal. The fence had blown off the post and was wrapped around the tree stumps in adjacent fields. FROM THEORY TO REALITY When I worked in Pittsburgh, standard wind design of 100 m.p.h. was the normal consideration in my design reviews of tanks. In Florida and the Caribbean, however most tank designs involved wind speeds in excess of the 100 m.p.h. that is standard for both AWWA and API tank designs. Most tanks built in southeast coastal regions require high wind design considerations. Many consulting engineers required coastal tank designs to utilize a 150 m.p.h. wind velocity and be designed in accordance with the building codes. Until recently, South Florida designs were to be in accordance with the South Florida Building Code (SFBC) now superseded by the Standard Building Code (SBCCI). One requirement of the building code is for structures with pitched roofs to apply a “pressure coefficient” that results in an uplift force on the roof. Technically, cone roof tanks fall into this category, and design by building code requirements would require uplift consideration. In high wind conditions this uplift force can exceed the weight of the roof plate, requiring that the roof rafters, girders and columns be secured. Both AWWA D100 and API-650 address wind forces in tank design. Both standards increase wind force in proportion to the square of the wind speed. The standard design force is 30 PSF on a flat projected surface at 100 m.p.h., and would equate to 2.25 times that force or 67.5 PSF at 150 m.p.h. Neither standard makes reference to uplift forces on the roof. ROOF UPLIFT FORCES In 1972, a Florida Keys Aqueduct Authority specification for a 5 million gallon steel water storage tank required the design and construction to be in accordance with the South Florida Building Code with a design wind speed of 150 m.p.h. This tank (164' dia. +), was to be constructed on Stock Island, Key West. Until this time I was not aware of the implication of this requirement for uplift loads on the roof. Closer analysis revealed that the uplift load exceeded the roof plate weight therefore requiring anchoring the roof. Normal construction is to simply lay the plate on the roof support framing. The only attachment to the tank is the shell-to-roof weld. In this instance the design engineer insisted that the roof be secured to meet the uplift requirement. A wind tunnel test was suggested to show the effect of a hurricane wind force on a cone roof tank since no incident of a roof failure could be documented due to hurricane wind forces. This was not done and the result was an expensive hold-down system including concrete foundations under each interior column. ANDREW PERFORMED THE WIND TEST Hurricane Andrew has answered some of these questions. Two 200' dia. x 48' high flat bottom, cone roof, oil storage tanks took a direct hit from the storm. Both tanks were damaged from what appears to be uplift forces on the roof. The roofs did not blow off the tanks, but structural damage did occur. A brief damage assessment is as follows: 1. The roof-to-shell weld had broken in several areas around the perimeter. This weld did exactly as it is designed to do since it is a weak joint designed to fail in case of explosion, thus allowing the roof to fail before the sidewalls and prevent product loss. The damage appears similar to that caused by an explosion. 2. The perimeter top angle was subject to excessive compressive forces and failed in many areas. This caused buckling and flat spots in the upper tank shell. 3. The rafters went into compression, causing several to break loose from the shell connection also causing twisting of the support columns. These tanks had about 35 feet of product at the time of the storm. No product was lost. There was extensive damage to everything in the area and the proportional amount of damage to these tanks would be considered minimal CONCLUSION Had the roof been designed to resist these uplift forces by securing the entire roof support system, damage may have occurred to the tank bottom or sidewalls resulting in product loss. These tanks will require repairs, but I do not believe the initial increased cost of construction, nor the possibility of other tank damage resulting in product loss warrants designing a flat bottom, cone roof tank to resist uplift forces as occurred in Hurricane Andrew. TEAM Consultants inspected several other flat bottom cone roof tanks in the area. Three large, empty tanks in the Cutler Ridge area had sidewall damage but no evidence of roof damage due to uplift. These tanks were not designed for high wind and did not have wind girders. ELEVATED TANKS WITHSTAND ANDREW’S FURY Most pictures of south Florida show complete devastation. In many pictures the only structure still standing is the elevated water tank. Although every elevated tank we reviewed had some damage, I am aware of only one that collapsed. It appears that flying debris damaged a strut causing this structure to buckle. The City of Homestead has two 500,000 gallon Hydropillar style elevated tanks. These tanks, designed for a wind speed of 150 m.p.h., were constructed in 1977. TEAM Consultants had recently completed an inspection of these tanks for the city and found them to be in good structural condition. We are happy to report that both tanks still appear to be structurally sound. Each tank has a 12 ft. wide by 12 ft. high overhead truck door and a 3 ft. man door in the base. Both doors in both tanks were blown out during the storm. No other damage was noted. The Florida City elevated tank appears to be OK also, although a thorough inspection was not performed. The exterior ladders were blown off the tank. The tank exterior was certainly high pressure washed, and the letters on the tank bowl were partially washed away. We inspected several other elevated tanks in the area. One had windage rods broken, ladders and platforms damaged, level indicators blown away and roof manholes covers blown off. We strongly suggest that all tanks that were in the path of this storm be inspected for structural integrity. The aftermath of hurricane Andrew will remain with us for a long time to come. Many building codes are going to be re-addressed, however, the steel tank industry can stand proud of its record of building structures that have been able to withstand the greatest forces of nature. The industry is to be commended for the stringent and conservative standards established for design and construction of storage tanks both ground storage and elevated
|
Send mail to webmaster@tankteam.com
with questions or comments about this web site.
|