Mega-warehouse fires: We can’t throw the baby out with the sprinkler water
We have essentially eliminated most of the redundant (and time-tested) fire protection features that we need to fight fires in these structures
By Steve Lohr
Forty-nine years ago, I enrolled in one of my first college-level fire protection courses – Building Codes and Construction. As part of that course, we studied two large-loss fires involving the near destruction of buildings that were both unusually large in size and had been considered essentially fireproof, immune to such large losses as the result of fire.
The first fire occurred at a GM Transmission Plant in Livonia, Michigan, in 1953. Approximately 4,200 people evacuated the 1.5-million-square-foot flat-roof building. Six individuals were killed. The building was only 20% sprinklered, and fire walls were absent in this four-year-old building. Unprotected steel roof trusses collapsed in minutes, allowing melted roofing materials to drip and feed the larger fire.
Subsequent code changes included limits on combustible tar in built-up roofs, assignment of hazardous operations to separate buildings, increased sprinkler use in industrial occupancies (hallelujah!), fire coating of steel framing, the introduction of automatic fire doors and, ultimately, the development of NFPA 204: Guide for Smoke and Heat Venting.
The second fire occurred at the McCormick Place Convention Center in Chicago in 1967. One individual was killed. Constructed six years earlier, the building lacked sprinklers yet was considered to be fire resistant by industry standards at the time. The open-space building included a tremendous fuel package from a home furnishings trade show that was set to open to the public about one hour after the fire ignited.
Chicago firefighters arrived within six minutes after discovery of the fire and were unable to stop the blaze, which ultimately resulted in a total loss of this unique structure. It was later determined that high-ceiling sprinklers would have likely controlled the fire. The replacement building included 40,000 sprinkler heads (again, hallelujah!). Note: I do not mean hallelujah as sarcastic or tongue in cheek. I assure you there is no bigger supporter of sprinklers in our profession. However, we should not throw the baby out with the bath water or, in the case of mega-warehouses, the sprinkler water.
While we have made significant strides in NFPA code adoption, it is clear by the continuing stream of destructive large building fires and mega-warehouse fires that we have more work to do, particularly as these massive incidents pose complex risks to both firefighters and occupants. Although the risk is intended to be mitigated by sprinklers and other fire protection systems, something is clearly not working – at least not as designed in the laboratory.
The fire service needs to have a conversation about modern-day mega-warehouses.
The first task is to determine what constitutes a “mega-warehouse,” as it seems there is no clear and consistent definition to adequately classify these buildings. Further, the challenges posed by large warehouses and manufacturing facilities are not the same as mega-warehouses; it can be the difference between a 200,000-square-foot structure and a 2-million-square-foot structure. Oddly, they require the same fire flow per the standards. This should be unacceptable to the agencies responsible for delivering manual firefighting services.
A careful review of fire flow and fire flow duration in NFPA 1: Fire Code, NFPA 13: Standard for the Installation of Sprinkler Systems and other relevant codes applicable to these massive buildings should quickly raise fire department concerns about our ability to both access and conduct manual firefighting operations in these buildings that are often sized in acres, not square feet.
Coupled with the practical limitations of a hydraulically calculated firefighting water supply that by code receives a 75% credit just for having sprinklers, access to interior portions of 1.5 million square feet of high-density rack storage that is 10 feet apart, 40 feet high and up to 2,000 feet in length is of great concern to many fire chiefs. Further, consider standpipe connections fed by 1-inch pipes with 1½-inch hose connections that are likely not sufficient to supply larger manual streams necessary to combat fires that sprinklers do not control. The codes require a hose stream allowance of only 250 gpm, inadequate if manual firefighting becomes necessary. Most safe structural firefighting policies plan for a three-line attack. Typically these hose streams would combine for a 450-500-gpm fire flow. Wherever these buildings are present, fire departments should be raising questions about how they would manage significant fires.
Even when a fire can be accessed for manual firefighting, how many departments are resourced to deploy appropriately sized attack line(s) or master stream(s) in a timely manner hundreds of feet to the interior, or to the center portions of a rubber roof assembly (non-combustible by code) that is literally hundreds of feet from the building’s perimeter?
In plain language, the fire department in Anytown USA is challenged with a fuel package and roof assembly that, in one form or another, is equivalent to solidified gasoline. It is interesting to note that no matter the size of the structure, the traditional fire hazard assessments used to evaluate risks from fire in any structure remain the same:
- Potential ignition sources;
- Potential fuel sources and fuel packages;
- The arrangement of fuel packages;
- The presence of fire protection features, including fire department access and available water supply;
- Building height, area and compartmentation; and
- Occupancy classification and human factors.
Fully sprinklered buildings remain the best fire protection money can buy. The fight for sprinklers has spanned decades, and as an industry, the fire service has allowed improvements in fire sprinkler technologies to trump almost every other known strategy to reduce the spread of fires in these buildings.
All sprinkler protection is partially based upon pre-wetting of adjacent fuels. Traditionally, this was stuff stored in cardboard where pre-wetting was in fact preventative. I am not convinced that fuels that are primarily hydrocarbon-based, stored in boxes that are now wrapped in plastic, and sitting on skids that are also hydrocarbon based will be as effective as originally conceived. It is at least probable that pre-wetting will in fact run-off of the plastic wraps and simply drip to the floor. Further testing is required.
Consider the impact of Ni-Cad batteries that we are struggling with in electric vehicles, scooters, tools, lawn equipment, etc. – all of which require excessive amounts of water application for hours once they are in thermal runaway. Again, more testing is indicated.
I fully recognize that Early Suppression Fast Response sprinklers are a very effective weapon, delivering much higher flows and densities than traditional sprinklers. However, they are not a magic bullet when any of the design assumptions are not met.
If we assume that some of our height, area and fire flow requirements for many occupancies, including warehouses, are simply wrong in the codes, I challenge you to carefully consider two broader questions:
- When is “large” beyond the practical ability of fire departments to function?
- Have we traded away too many other known strategies for fire protection to be effective in these buildings when manual firefighting becomes a necessity?
Traditional features of fire protection like realistic fuel package assessments, draft curtains, compartmentation, adequate separation of known hazards, fire-resistant separations designed to limit horizontal fire and smoke spread, the impact of hundreds of trailers full of stuff limiting access to water supply and access into the structure and, most importantly, height, area, and fire flow requirements in the codes combine to pose unacceptable risks.
When a fire originates outside of the sprinklered space, the fire pump or any given riser is out-of-service, or any other calamity that is part of maintenance of effort during the life cycle of a building, we have essentially eliminated most of the redundant features of fire protection that are time-tested.
Once the fire department arrives and deploys large-caliber streams, we further accelerate the depletion of the available water supply that most systems will not adequately sustain for more than two hours or promptly replenish. When this occurs, we have now entered a losing battle that most departments will not overcome on their best day. Much of the fire service is operating with a traditional pre-connect mindset that assumes that if we are fast, close, hot and wet, all will end well. We are seeing evidence with fires in these super-sized buildings that this simply is not true.
In the jurisdiction where I currently serve as fire chief, there are seven mega-warehouses within 12 square miles. The smallest is 300,000 square feet, the largest 1.5 million square feet. Our plans-review team members are evaluating a proposal to connect two that will result in a facility that’s larger than 2 million square feet.
Venturing 30 miles north or south, there are two dozen or more nearly identical buildings. This is a perplexing and compounding risk. Like many agencies, we would be ill-equipped to face a fire that exceeds 12-20 activated sprinkler heads.
If fire departments acknowledge that once fires grow to the 12-20 head threshold and/or we cannot access large areas of the building and/or we exceed the available water supply for any reason, we must apply our risk management models and accept that we are in the middle of a financial transaction between the owner/occupant and their insurers who specialize in highly protected risks (HPRs). The fire department is well advised to proceed with caution. The parties invested in these warehouses have a different view than the fire service regarding acceptable loss. We have been trained and conditioned to save as much as possible, regardless of the cost. This is an antiquated model that needs reevaluation.
Preparing for warehouse fires
Our department has implemented some standard and non-standard initiatives to prepare for responses to large warehouse fires:
- In-building radio bidirectional amplifiers to guarantee radio coverage;
- FDCs located immediately adjacent to man doors to prevent blocking by trucks or trailers;
- Requirement that private yard hydrants are located at the nose of trailers rather than the rear of trailers staged around the outside perimeter of the property;
- Addition of a 2½-inch hose connection at each of the sprinkler risers inside the man doors, which is in addition to the minimum code requirement;
- Addition of portable master streams, multiple standpipe packs and leader lines on every pumper to enhance rapid deployment of large-caliber streams from the man door connections on the risers;
- Two new LDH FDCs on the private fire protection loop near the tank/pump house in addition to the required 2½-inch FDCs;
- Placement of a municipal hydrant off the private loop just outside of the property line to supplement the private water supply when high-demand flows are needed;
- Pumpers equipped with dual supply beds to facilitate large-capacity water supply;
- Risk management training for all personnel to enhance go/no-go decision-making;
- Implementation of a construction and fire protection features training program unique to these large buildings; and
- Development of an SOG that is significantly different than our standard safe structural firefighting SOG. At every rank, I have found the practical horizontal reach of an elevated stream without the impact of wind to be about 100 feet (+ or- 10%). Consider your ability to deploy one or more elevated streams when any aerial is automatically set back from the building’s edge by a 53-foot-long trailer, a 50-foot-tall building that is 2,000 feet long and 600 feet wide.
The fundamental principles of fire protection that have traditionally provided some redundancy to sprinkler protection should be entered back into the equation. Height, area and practical fire flows, adequate separation, total fuel available, arrangement of fuel packages, isolation of known hazards, enforcement of the human factors necessary to prevent ignition, and maintenance of fire safety features are essential in these structures if we are going to meet our mandate to protect the building and, more importantly, the building’s occupants.
It is past time to collaboratively work with the best in our industry to fairly evaluate the risks presented by mega-warehouses, then make reasonable recommendations to ensure that we have not created a fire protection perfect storm. This must include input from the insurance industry, which willingly underwrites these structures and their contents. If the answer is that the number of significant losses is less than what the industry is willing to accept, then so be it.
Until then, fire departments beware. These structures simply do not fit into the traditional pre-connect firefighting model, and we should not risk the community’s fire protection resources – that includes our members and equipment – for a transaction that is almost surely a loser.
About the Author
Steve Lohr serves as the fire chief in Hagerstown, Maryland. In 2014, he retired as chief of the Montgomery County (Maryland) Fire-Rescue Department after a 30-year career serving in every rank. Lohr is a graduate of the University of Maryland Fire Protection Engineering program, the National Fire Academy’s EFO program, and the Executive Leader’s program at the Naval Post Graduate School. In addition, he has served as an adjunct instructor in the Montgomery College Fire Protection curriculum and lectured on topics of fire service interest. Contact Lohr via email.