A $280 MILLION “STATE OF THE ART” SPRINKLER DISASTER
BY FRANCIS L. BRANNIGAN, SFPE
An item abstracted from the National Fire Protection Association report “Warehouse Fire–New Orleans” describes a veritable fire protection “Pearl Harbor.”1 Every department protecting a huge warehouse, even with “state-of-the-art”2 sprinkler protection, should get a copy and study it with the warehouse management.
The basic facts are the following:
The warehouse was built in 1991. The undivided area was 930,020 square feet. The walls were tilt-slab concrete.
The ceiling was 72 feet above the floor.
The racks in the high-rack area with in-rack sprinklers were 63 feet high.
Other racks–some of which had no in-rack sprinklers–were 21 to 30 feet high.
There were 30 overhead sprinkler systems and 17 in-rack sprinkler systems. All were properly alarmed. The water supply was adequate.
The incendiary fire was started in combustible material, such as wicker baskets, in a 21-foot-high stack that had no in-rack sprinklers. The sprinklers were 50 feet above the stacks; they operated but did not control the fire.
Employees fought the fire, apparently briefly, because the fire department re-ceived a phone call from the warehouse at the same time of the first sprinkler water flow. Eleven minutes after the first sprinkler water flow, the first New Orleans (LA) Fire Department unit on the scene reported fire through the roof. The fire went to five alarms and was declared to be extinguished about five hours later.
It took a five-alarm assignment to suppress the fire in the racks without rack sprinklers.
All the sprinkler systems were shut down after the fire had been extinguished, to limit water damage.
Apparently without checking the electrical system, the facility personnel restored power to conveyors to get back into production. Arcing ignited combustibles. Personnel turned the sprinkler valves back on, but the fire could not be controlled, and the entire warehouse area was totally destroyed in a fire that lasted six days.
Some NFPA Comments
The failure of the sprinklers 50 feet above the 21-foot-high storage is very significant and had not been anticipated in the code. Storage 25 feet high or higher requires a space of not more than 10 feet above the stacks.
Procedures for shutting down and monitoring sprinkler systems are not specifically addressed in current documents; neither is the question of which systems should be shut down.
The code requires in-rack sprinklers for the portable racks as they are installed in the warehouse. However, “it is nearly impossible and very impractical” to install such sprinklers.
There are no recommended procedures for restoring utilities.
My Comments
In 1967, a huge exhibition hall in Chicago, the original McCormack Place, was totally destroyed by a fire in the contents–a tremendous accumulation of fuel, the Home Show exhibits. As a result, in the first edition of Building Construction for the Fire Service (1971), I discussed the problem of a heavy exhibit fire load in an unsprinklered arena. I pointed out that if the building were sprinklered, at least 20 sprinklers would operate on any fire. I reasoned that this would justify setting up heavy-caliber stream devices in advance to cover all areas and to be operated if a fire were not immediately extinguished. A far-out idea–I do not know that any fire department took it seriously, but manufacturers developed the nozzles, which can oscillate to cover the assigned area and the cross-zoned visual fire detectors. The system has been installed in several arenas. A “water cannon” operated as designed when fireworks were set off in the San Antonio Alamodome. The management had ignored instructions on how to handle such a situation. Why not install such a system in warehouses where in-rack sprinklers are not practical?
Fire that cannot be extinguished by the sprinklers menaces the entire warehouse. In a Pennsylvania high-rack grocery warehouse, scrap cardboard boxes were loaded into steel railroad cars within the warehouse, as pictured on page 609 of BCFS3. A fire in a railroad car would trip sprinklers all over the warehouse without affecting the fire. Such side-stream inconsequential operations should be located in a low-value building outside the warehouse. I wrote to the president of the company and was recently informed that this no longer occurs due to the hazard it presents and that, in fact, the car wells (pits with tracks low enough so the car floor meets the warehouse floor) are being eliminated.
The fact that the sprinklers high above the fire were ineffective should raise the question of whether sprinklers high above the fuel in such cases as the wooden structure in a hotel atrium (BCFS3, page 468) are reliable.
SPRINKLER RELIABILITY
My interest in fire protection started almost 70 years ago, when I used to hang around the neighborhood fire station in New York City. I was fascinated by the bell alarm system for dispatching companies. The system was very detailed. Any unit could be called to any box location in the greater city. A huge force of more than 30 companies could be called to a location with a single bell signal. The ordinary signal was the location of a fire alarm box. For instance, ******* ********** **** indicated Fire Alarm box 7 0 4 located at 7th Ave and 33rd Street. An assignment card indicated the units that were to respond.
One day I counted off a strange signal … 3-704-13. My mentor explained, “That`s a special building box. It`s an alarm from a building near 704 box. It can originate from a private manual fire alarm box in the building–from an automatic fire alarm system or a water flow on a sprinkler system.”3 He showed me the card and said, “See, it`s a sprinkler alarm. Engine 1 and Ladder 24 will respond. It`s most unlikely we`ll hear any more from it. The sprinklers will have the fire out or at least controlled, so two companies can handle it.”4
Help had to be summoned once in a while. Garments are cut on a wide cutting table. Sprinklers were not required under the tables at the time, where a load of scraps accumulated. This could provide a nasty smoky worker. I recall one “all hands worker” in a W. 37th Street high-rise garment loft.5 Only the employees of the fire area evacuated the building. Everybody else kept working; they were paid on piece work. The clerk of one company, a buff, left to watch the fire from the street and was docked two hours pay. He was the senior lieutenant in the fire department when he retired.
Insurance engineers had the same absolute confidence in sprinklers. During Word War II, the availability of lift trucks made rack storage a reality. At Norfolk, Virginia, the Navy had huge wooden sprinklered warehouses with expensive war reserve stock (70 million 1947 dollars of radar equipment in one fire area) stacked in racks, 16 feet high. I set up prefire plans for heavy-caliber streams to confine the fire to the area of origin.
Navy fire protection engineers with insurance company experience but no fire experience scoffed, “There is no need to plan for a major fire; the building is sprinklered.” They never answered the question, “How will a half-inch spray of water work its way down through this stack to a fire in the bottom of the stack?”6 The same attitude in varying degrees persists to this day in the fire service.
It is reinforced by another myth: “The insurance company makes them keep it all up to requirements.”7 A common point of view, erroneous in my opinion, is that the sprinklers are just a matter between the owner and the insurance company.
The answers usually given to the question “What action is taken by the fire department when it is notified that a sprinkler system is out of service?”8 are indicative of serious complacency. This complacency should be countered by the following considerations:
The record of huge losses in warehouses insured by highly protected risk insurers and presumably protected by “state of the art” sprinkler protection (BCFS3, pages 605-612).
Major fire losses are not a private matter. They can seriously affect the economic health of the community.
Some “sprinkler ” installations have been completely fraudulent in that while sprinkler heads were visible, there was no water supply.
Although articles, such as that written by Richard Shulte,9 lift the curtain on slack practices that may be common in construction, we somehow do not expect the contractor to use them because human lives are at stake. Shulte tells of an early work experience in which he discovered that a large duct would completely mask the sprinkler discharge. He was told, “Forget it unless the fire department or insurer objects.”
In a number of cities, fire departments have absorbed the function of plan review from the Building Department.
Many significant reductions in static fire protection and life safety provisions are permitted if dynamic fire protection in the form of sprinklers is installed. The Life Safety CodeT has adopted to some degree the concept I have been preaching and enforced where I had the authority, for 50 years …. “Where sprinklers are installed for life safety, very special precautions, up to and including closing the building, are to be taken if the sprinkler protection is impaired.”
Professor Vincent Brannigan, J.D., a thorn in the side of those who believe “performance codes” are the golden elixir for the construction industry, has pointed out that the substitution of “performance codes” for “prescriptive codes” means that the public authorities will have “cradle to the grave” responsibilities to see that the provisions of the performance code remain in operation and will perform as designed.
All these things point to the necessity for developing fire officers with a more sophisticated understanding of sprinkler system design and maintenance than is generally present at this time.
One fire chief excellently expressed the function of a sprinkler system as “my first-due engine company.” The municipality spends a sizable sum to keep an engine company in service. Therefore, the modest cost of educating some personnel so they can properly assure the management (which often is totally ignorant of the sprinkler system and its function) will maintain the sprinkler system properly so that it will act adequately when needed is more than justified.
Some years ago the entire sprinkler system in Macy`s department store in New York City was totally shut down and drained to deal with an intractable leak. I don`t know whether the fire department was notified, but if it was, the only action taken was to note on the blackboard, “Macy`s sprinklers out of service.” The sprinklers are the basic life safety factor for the thousands of customers and employees in the huge building. When the system was put back, a valve was left closed. A fire occurred in the unprotected area, and a firefighter died. I understand that procedures have been improved.
The Life Safety CodeT provides for valuable concessions in exit travel distance when the building is sprinklered. When the sprinklers are not in operating condition, the building is illegal because of inadequate exits. As I understand it, the CodeT requires action when the shutdown exceeds four hours. This would appear to permit the shutdown of the sprinkler system in a shopping mall for four hours on Christmas Eve, without any special precautions.
The list of disasters that occurred in such a “window of opportunity” is led by the Ringling Brothers Hartford Circus tragedy in which 188 died because the main exit was blocked for “only 10 minutes” by the connected cages that brought the tigers and lions into the arena.
When Atomic Energy Commission “bean counters” demanded to know why the management of MIT`s metallurgical project (which we had housed in an old dump of a factory) was paying sprinkler contractors overtime, the reply was, “Because of Brannigan`s rule. If we did the work in the daytime, we must send everybody home; the overtime is cost effective.”
ERRONEOUS TEXT
If you have a copy of my videotape “Why Masonry Walls Fall,” look at the last page of the text. If you find the words “Rich Tower,” instead of “Ritz Tower,” write to me at 2041 Daylily Road; Port Republic, MD 20676-2646 for a correct version.
RECENT ARTICLES CITED FOR SPECIAL
ATTENTION
“Strip Warehouses, A New Firefighting Challenge,” Bill Gustin (Fire Engineering, May 1997, page 48). This article is packed with practical information on dangerous conditions. Use for instruction. Company officers should be on the lookout for these and similar conditions.
I might supplement the article by pointing out that tilt-up concrete buildings such as those shown on page 49 require the roof for stability. If the roof is failing, the walls are unstable–clear the collapse zone. Generally there is a neoprene gasket between units. This can burn out and pass-fire from a vehicle parked outside or a lean-to shed or vice versa. The concrete wall is not a dependable fire barrier.
The column in the upper left picture appears to be unprotected steel. The failure of an unprotected steel column supporting a concrete floor killed four firefighters in Brackenridge, Pennsylvania.
“Plantation, Florida, Towne Mall Fire,” David Casey (Fire Engineering, April 1997, page 65). The combustible metal deck roof continues to take its toll. Industrial Risk Insurers (IRI) paid a tremendous loss on a carpet manufacturing plant in Georgia in early 1996. The key to the loss of the plant was a combustible metal deck roof. I have a sort of proprietary interest in this problem, because I discovered the problem in 1946 and proved the solution in 1948. Nobody was listening until FIA (IRI`s predecessor) suffered a $51 million loss at the General Motors transmission plant in Livonia, Michigan, in 1953.
Tests conducted at Factory Mutual showed that a 64-square-foot hole was required to successfully vent a structure 20 feet wide and 100 feet long in which only the combustible metal deck roof was burning; there were no contents. Because the high heat is seriously degrading the steel trusses (which in a standard fire test failed in seven minutes), it is foolhardy to attempt manual ventilation. Six Dallas, Texas, firefighters were plunged into the fire when attempting to ventilate a metal deck roof fire. Fortunately, they were on the edge of the roof and were recovered rapidly.
A combustible metal deck roof collapsed just as a midwestern fire officer was about to step on it. I am convinced that fumes from the combustible metal deck roof in the cabaret room–instead of the “insulation on electric wiring” blamed in the lawsuits–killed all but two of the 187 people who died in the Beverly Hills Supper Club fire in 1977 (pages 302-9 in BCFS3).
I NEED HELP … I would like to build a case file in three subject areas, with your assistance:
Near misses. More detailed stories of how disaster was avoided through preplanning and adequate command action and where injuries–but not fatalities–occurred. It is saddening beyond measure to have to use firefighter fatalities to hammer home points. Ben Franklin told us, “Experience keeps a dear school, but fools will learn in no other.”
Recently in our home area of Calvert County, Maryland, firefighters were battling a fast-spreading fire in a large, expensive home (loss estimate $750,000 for the house and contents). Typically, the house had truss floors and “lovely” knotty pine ceilings. They were making a desperate effort to save the house, but it was “built to burn.” Incident Commander Assistant Chief Wayne Hardesty of the Prince Frederick (MD) Fire Department, mindful of the truss floors, withdrew all personnel very shortly before the structure collapsed.
Membrane roofs. Single-ply membrane roofs present severe hazards during and after installation. City of New York (NY) Fire Department Battalion Chief Frank Montagna told me of a membrane roof fire that spread like a prairie fire. They were fortunate to be on the windward side.
Chief Charlie Dickenson of Pittsburgh, Pennsylvania, wrote me about a roofing fire involving containers (resembling 20-pound propane tanks) of “Insta StikTM,” a water-based adhesive pressurized by nitrogen. About 50 cylinders were propelled through the air when the nitrogen was heated by the fire. [Pressurized water extinguishers have no relief valves and have resulted in BLEVEs (boiling-liquid, expanding-vapor explosions) causing at least one fatality, if memory serves. I would never permit them to be hanging where a high fire load is present, the code notwithstanding.]
Unsprinklered high-rises. Unsprinklered high-rise office buildings are all across the country. Many have huge open areas filled with combustible computers, as did the trading floor of the First Interstate Bank in Los Angeles. Opponents of sprinklers cite low death rates in fires in such buildings and, in the usual obfuscation tactic, point the finger at dwellings. The problem in high-rises is the potential for mass casualties. However, what might be more effective in getting action would be the loss of an entire building. Such a disaster was narrowly averted in the Bankers Trust Fire in New York, where doubled company teams of the most experienced high-rise firefighters could not advance lines against the fire. The only saving grace was that the fire, by sheer chance, was just barely at the edge of the reach of exterior streams from tower ladders.
If enough stories are assembled, perhaps the complacent politicians would tell their golf-playing real estate buddies that, as much as they hate to do it, they cannot avoid passing a retrofit sprinkler law. n
Endnotes
1. Comeau, Ed, “Warehouse Fire–New Orleans,” National Fire Protection Association, March 21, 1996. Contact: NFPA Fire Investigations: (617)-984-7473, fax: (617) 770-0700.
2. I have often defined “state of the art” as meaning, “We don`t know if this will work–we`re trying it out on you.”
3. New York did not permit any private connections to its fire alarm system. All such alarms were handled by private fire alarm companies, which passed the alarm to the fire alarm office.
4. This excellent experience is borne out by “Sprinkler Experience in High-Rise Buildings” (SFPE Technology Report 79-1), Bob Powers, SFPE, then superintendent of the New York Board of Fire Underwriters. On the other hand, an alarm from one of the old thermostatic automatic alarms was very often the first signal of a greater-alarm fire.
5. In New York, a “loft building” rents out manufacturing space to tenants. After the Triangle Shirtwaist Fire, the Garment District moved en masse to midtown Manhattan into high-rise sprinklered loft buildings. The ordinary construction buildings left behind had low-grade occupancies and many serious fires. The area was called “Hells Hundred Acres.” It since has been “yuppiefied” into expensive living spaces and is called “SOHO” (south of Houston Street).
6. We did not face the prospect of lawsuits for negligence by “robbing the sprinkler system of water,” but we did face an exhaustive investigation into the circumstances of any major fire. Our engine companies were sternly warned to hook up to domestic hydrants only, because the high-pressure hydrants were on the mains feeding the sprinklers, and we would be charged by fire protection engineers with “robbing the sprinklers of water, thus causing the huge loss” instead of identifying the true cause–inadequate sprinkler protection.
7. For a frank discussion of insurance/fire department relations, see Building Construction for the Fire Service, Third Edition (BCFS3), pages 618-619, Francis L. Brannigan, the National Fire Protection Association.
8. See “Fire Department and Sprinklers” in Chapter 13 of BCFS3.
9. Shulte, R. G. “Quality Control in the Sprinkler Industry,” Plumbing Engineer, March 1996, 14.
 |
 |
 |
(Left) The fire totally destroyed the main portion of the warehouse. (Center) The 63-foot-high bay racks in the middle portion of the building had in-rack sprinklers. (Right) The portable racks, where the fire was started, were typically 21 feet high. There were no in-rack sprinklers. The distance between the top of the racks and the ceiling sprinklers was about 51 feet. (Photos © National Fire Protection Association, 1996. All rights reserved. Published with permission.)
FRANCIS L. BRANNIGAN, SFPE, recipient of Fire Engineering`s First Lifetime Achievement Award, has devoted more that half of his 56-year career to the safety of firefighters in building fires. He is well known for his lectures and videotapes and as the author of Building Construction for the Fire Service, Third Edition, published by the National Fire Protection Association. Brannigan is an editorial advisory board member of Fire Engineering. He may be reached at (301) 855-1982.
