LETTERS TO THE EDITOR
Sprinkler system water supply analysis
I read with great disappointment Gary Keith’s and Donald Garner’s article on sprinkler system water supply analysis in the December 1991 issue. Although the article is on a timely and well-deserving subject, the authors presented the fire service with misleading and inaccurate information.
In the first sentence of the second paragraph, the authors stress the importance of calculating the water “needed” for the operation of a sprinkler system, including hose streams, as part of evaluating the adequacy of a sprinkler system. The primary content of the article is the authors’ attempt to give the readers a method of calculating the demand for any given sprinkler system. Unfortunately, the authors define demand as “being equal to density multipled by the demand area, expressed in gpm.” This is only one-half of the demand requirement! The authors failed to inform the readers that equally important to determining the flow (gpm) requirement for any given sprinkler system is determining the pressure (psi) required to deliver the flow. Nowhere in the article do the authors discuss the required pressures. After going through the calculation procedure for determining the flow portion of demand, the authors simply assume a 5S-psi pressure requirement. Where did this come from? Why is it not 85 psi?
Furthermore, the authors consider their demand calculation a “theoretical demand,” with which I wholeheartedly agree. From my experience, the actual sprinkler flow demand typically ranges from 15 to 90 percent greater than the theoretical demand calculation presented by the authors. For this reason, the theoretical demand method presented by the authors never should be used to properly analyze a sprinkler system and for all practical purposes is of little use to the fire service.
The actual sprinkler demand cannot be determined by the methods presented by the authors. The actual required demand of an existing sprinkler system, both flow and pressure, can only be determined by hydraulic calculations performed on the system by those knowledgeable in the performance of hydraulic calculations. Mathematically, the end result will always be greater than the theoretical demand the authors presented.
The authors provided very little useful information for the fire service to calculate sprinkler system demand for use in preplans and fireground operations. Perhaps the readers should have been provided with information on how to interpret and use the information found on the hydraulic data plates required to be attached to the base of sprinkler system risers. These plates (or signs) contain the sprinkler system demand, including required flow, pressure to deliver the flow, and the hose stream requirement. Where systems are not provided with the data plates, the fire service should work with owners and the original installers to determine the system requirements. Once this information has been obtained, it is then a simple matter to compare the sprinkler system demand with the available water supply.
I also would like to point out that the authors’ methods in plotting the water available with hose stream adjustment curv e is not consistent with the methods recognized by the NFPA and the sprinkler industry . I refer vour readers to NFPA 13 for a more accurate and accepted procedure. The hose stream allowance is required to be available at the sprinkler demand point. I know of no requirement or recognized standard that requires the hose stream allowance be available above the sprinkler demand as presented by the authors.
I^st, there is one very disturbing statement made by the authors. In the third from last paragraph, the authors state that if sprinkler demand were above the water supply with hose stream deduction curve, “the end result likely would be an uncontrolled fire, even though the building is fully sprinklered.” This is far from the truth. I would venture to guess that every fire service reader has been on a fire extinguished by a sprinkler system in which a hoseline was not even charged, substantiating that hose streams in many cases are never required. In addition, there are several inherent factors of safety built into a sprinkler system. It has been stated in numerous publications that more than 95 percent of all fires in fully sprinklered buildings are extinguished with fewer than five heads. Considering that a design area may contain three to six times as many heads, one could reason that onesixth to one-third of the design demand was needed to extinguish 95 percent of the fires. I realize that this is a simplistic approach and that the percentages may be mathematically incorrect, but they do present a point. There are many variables, both internal and external, to a sprinkler system that will determine if a system will operate satisfactorily wTien the system demand exceeds the supply. An unequivocal statement that an uncontrolled fire is likely to occur is simply not a fact. It is only an increase in probability.
J. Michael Thompson, PE
Gage-Babcock and Associates
Vienna, Virginia
Ciary Keith and Donald Garner respond: It is important to note that the purpose of our article simply was to show that these water supply/sprinkler demand comparisons and graphs exist. It has been our experience, after many years of responding to fire service questions, that firefighters most often want help in understanding the relationship between the available water supply and sprinkler system demand.
We agree that we did not emphasize the pressure requirement of sprinkler systems. The pressures shown are relative as a comparison within the examples presented. From a fire service perspective, the importance of adequate flow is often understated, especially when discussing supplementing the system through the fire department Siamese connection.
While we agree with Mr. Thompson that there are many examples of sprinklered fires being extinguished without the use of hoselines, we also are aware of numerous cases where the water supply for sprinklers was depleted by the improper use of hoselines. Too many times the sprinkler system is designed using all of the available water supply or, worse yet, without knowing exactly what the water supply can provide. We believe that the total design must include hoseline demand.
Again, our objective was to remind the fire service of the overlapping relationship among the occupancy of the building, the sprinkler system design, and the available water supply. Prefire planning must include more than accepting the philosophy of “don’t worry about that building, it’s sprinklered.”
Glenn Corbett adds: Mr. Thompson points out that, in his experience, “actual” sprinkler water demands range from 15 to 90 percent greater than calculated “theoretical” demands. My experience has indicated that actual demands are typically on the order of 10 to 20 percent greater than theoretical demands. I cannot recall having reviewed a system in which the actual demand was 90 percent over the theoretical demand.I would venture to say that sprinkler system designers who design systems with an actual water demand 90 percent over the theoretical demand will not be designing sprinkler systems for very long.
Failure to reap the economic benefits of hydraulically designing a sprinkler system with its resultant “optimization” of pipe sizing does not make good business sense in today’s highly competitive sprinkler-contracting/designing industry.
It is ironic that Thompson takes exception to Mssrs. Keith and Garner’s warning concerning sprinkler demands “above” the water supplycurve. Thompson champions the issue of ensuring that actual system demands be used when evaluating water supplies, yet diminishes the same concept when he writes of “inherent factors of safety” and that most fires are extinguished with but a few heads. He implies that these factors of safety (which are not specified in any design standard that 1 am aware of) guard against an uncontrolled fire when the system’s water supply demand lies above the water supply curve. I wouldn’t want to bet on that happening.
Fire flow testing
1 would like to comment on what I believe is a potentially significant omission in Glenn P. Corbett’s sidebar “Fire Flow Testing” (December 1991). He says that when establishing a water supply curve for sprinkler system design to “select two hydrants closest to the building and designate one as the pressure hydrant and the other as the flow hydrant.” However, Corbett did not mention that the test results are only truly valid at the test hydrant or that it is important to identify which hydrant should be used for test and which for flow.
It is important that the flow hydrants) be downstream from the test hydrant. In such a configuration, the water flowing from the flow hydrant must pass the test hydrant. This ensures that the flow at the flow hydrant is also available at the test hydrant at the pressure recorded at the test hydrant. Flowing the upstream hydrant would result in a higher pressure at a downstream test hydrant and would require additional friction loss calculations to determine the true residual pressure at that point.
On a dead-end line,, it is obvious which hydrant is downstream. However. in a gridded system one should consider the hydrant closest to the smaller main to be downstream, since more water usually will flow from the larger main. In a gridded system where both mains are the same diameter, age, C factor, etc., it probably would make little difference which is the flow or test hydrant.
This may seem like a trivial omission; however, the theme of the parent article, “Sprinkler System Water Supply Analysis, by Gary Keith and Donald Garner, is that “a sprinkler system is only as good as its water supply.”
It seems to me that since the parent article admits a five to 10 percent error in water system tests due to gauge error and parallax, we must guard against compounding that error. Mistakes like reversing flow and test hydrants could increase that error and make the water supply appear stronger than it really is.
David S. Miller
Fire Protection Engineer
Boulder (CO) lire Department
Glenn Corbett responds: I appreciate Mr. Miller’s comments concerning my sidebar on fire flow testing. Although the sidebar was intended as a firefighter’s procedural on how to actually conduct the test and graph the results, he raises the important consideration of which hydrant to select as the “test” hydrant. An entire article could be written about the hydrant selection process. However, I’d like to raise a few additional points for consideration.
As Miller points out, the “test” hydrant is the more important hydrant to consider when reviewing a new hydraulically designed sprinkler system or when evaluating an existing sprinkler system. The test hydrant is the “known” water supply reference node, with the actual residual pressure available in the water main shown on the cap gauge (corrected for the elevation difference between the main and the cap gauge).
The San Antonio (TX) Fire Department Engineering Services Office requires that all hydraulic calculations terminate at the test hydrant. In addition, the designer’s site plan must show: ( 1) the type, size, and location of all intervening/surrounding public water mains between the test hydrant and the “private” (on-site) fire main; (2) all valves, including pressure-reducing and isolation valves (with normally closed or normally open designations) in the public water mains and private fire mains; (3) the elevations of the flow hydrant, test hydrant, and the base of sprinkler riser; (4) designation of all mains as either circulating or dead-end; (5) and time, date, location, and readings of flow test.
With this information in hand, the fire department plan reviewer has a good “picture” of the water supply surrounding the project. Specifically, the following types of conditions and problems usually become readily apparent:
- An elevation difference between the test hydrant and the sprinkler riser has been omitted in the hydraulic calculations. This is a very common error, leading to major sprinkler system design flaws.
- There is no direct connection between the public main at the test hydrant and the public main that has been “tapped” to serve the sprinkler system. In an extreme example sever-
- al years ago, I found that a flow test had been conducted on one water company’s main while the sprinkler system actually was supplied by another water company’s mains!
- A sprinkler system that is affected by a pressure-reducing valve (PRV) in the public main. A flow test can be performed “upstream” of a FRV while the sprinkler system tap is “downstream.” Significant errors in the hydraulic calculations will result if this is not spotted by the sprinkler system designer.
Assessing rescue risk
Rescue easily can be considered the most dangerous fire department operation. Although placing firefighters at great risk to save property is no longer acceptable (thankfully), incidents that require the rescue of human life add significantly to the risk analysis procedure during size-up. When we have more to gain (salvageable life), we tend to risk our safety more.
Generally, the fire service trains under standard operating procedures. Progressive training presents situations that deviate from the norm, causing rescuers to use problem-solving skills—firefighters learn to adapt to different situations. Each new rescue presents new challenges to team members.
Risks in rescues in which I have participated or witnessed were evaluated and often overcome by ingenuity, courage, and trust. Company officers or team leaders make life and death decisions based on limited information and time. The extent to which the officer will deviate from the SOP depends on many intangibles, including the probability of success (rescue vs. recovery) and his or her knowledge of equipment capabilities and the rescue team’s training and experience.
The use of firefighters as anchor points (“FDNY: Two Rescues in the Sky” by Patrick Brown, July 1991) became a highly controversial topic in the rope rescue community. You probably won’t find an SOP that uses people as anchors, but in this case a decision was made by the officer based on the conditions present and his knowledge of his team’s training and equipment capabilities. The bottom line was two successful rescues.
The use of a crane in rescue work (“Rescue ’90: The ‘Big One’ ” by Tom Carr, December 1990), as demonstrated by Captain Mike Brown of the Virginia Beach Fire Department, stirred up other technical rescue specialists. This, again, was not meant to show an SOP but rather an alternative when other life-saving means will not work. If a department’s equipment, training, and experience preclude the use of such alternatives, they simply should be crossed off the list of options. I submit (from personal knowledge gained through training with him) that if Captain Brown chose to use this alternative, the risk-gain factor was carefully weighed and there was no doubt that his team could successfully complete the task.
Plain and simply put, rescues involve risk; they always have and always will. We must place firefighter safety high on the decision list when determining whether or not to attempt rescues. Remember, however, that without complete knowledge of a rescue incident, intelligent and meaningful criticism cannot be made. We will, however, never rid ourselves of the speculations of “Monday-morning quarterbacks.”
With this in mind, some thoughts came to mind while reading the May 1992 issue of Fire Engineering, which highlighted the topic of rescue. Congratulations to the department members who displayed the skills and ingenuity necessary to complete the rescues in the manner that they deemed necessary. Since I have no firsthand knowledge of the situations, SOPs, or equipment and experience of the fire departments involved, the following is not offered as criticism but as an exchange of ideas concerning basic rescue practices.
- Rescue equipment. If your community relies on your fire department to provide technical rescue operations, make proper equipment a priority. (It takes a lot of convincing!) In rescues involving collapsed structures, high angles, confined spaces, etc., engine and truck company equipment generally is inadequate.
- Atmospheric monitors. The use of human senses to detect toxic, hazardous, or oxygen-deficient air has proven fatal in many cases. Properly calibrated multifunction monitors should be available for confined-space rescue operations. Ventilation of the area and space may become extremely important.
- Supplied air breathing apparatus. Without them, the size of the space that you should enter is more limited. Removing your SCBA and pushing or pulling it also has been fatal to rescue personnel. If it doesn’t kill you, it may limit your ability to aid a victim in a tight spot to the extent of being almost nonfunctional. Entering a space that has not been atmospherically tested and without any breathing apparatus places rescue personnel at very high risk.
I’ll close by commenting on two points made by FDNY Captain Ray Downey:
- “Proper equipment should be seen as a means to satisfy standards intended to increase rescuer safety.” Work hard on getting your firefighters the proper equipment to do the job expected of them.
- “Remember, however, that training, experience, and ingenuity are the most vital ingredients to the successful conclusion of a rescue incident.” Amen to that.
Ron Zawlocki
Captain
Pontiac (MI) Fire Department
Size-up works for trusses
I would like to comment on the letter to the editor “Preplans work for trusses” (June 1992).
You can’t always plan an interior operation in a burning building with a truss roof. Captain Mike Maciaga asked for 10 minutes to fight the fire from the interior, and his team extinguished the blaze. He was successful. But he was also lucky. In Brooklyn, New York, on June 2, 1992, a fire occurred in a garage containing burning tires, and the roof and rear wall collapsed within 10 minutes after arrival of the first unit.
The companies and chiefs each had preplanned a fire in this dangerous vacant truss-roof building, and still two firefighters were buried beneath the brick wall pushed out by the collapsing timber-truss roof. The firefighters were seriously injured but will survive.
What Maciaga did was “gutsy.” The interior attack of the first initial attack hoseline extinguishes most fires and is the basic service the fire department provides for a community. But he was lucky.
A standard operating procedure for all truss (timber or lightweight) roof building fires should be:
Make an exterior size-up.
- If the fire is beyond the control of an initial attack hoseline, then set up an exterior attack.
- If no smoke or only minor smoke is showing, then enter and make an interior size-up of fire (see #2).
- Make an interior size-up:
- If it is a fire involving the contents of the building, then use standard operating procedure —extinguish with an interior attack hoseline.
- If it is a fire involving the truss structure itself (any type), then evacuate the building and use an exterior attack on the fire and protect exposures.
Vincent Dunn
Deputy Chief Division 3
City of New York Fire Department
A painful lesson
I recently read the story “Pennsylvania Collapse Kills Four” by John Stankiewicz (June 1992). It brought back a lot of memories. I was chief of Citizen’s Volunteer Hose Company #2 of Harrison Township (PA) (Station 20) at that time. Our company is a sister of Hilltop Hose Company #3 (Station 30). Our involvement at this tragedy was as a supply company, as we laid 1,000 feet of four-inch line. We originally were called to cover Brackenridge, but we soon were pressed into service. This tragedy brought our firefighters closer together.
I cannot estimate the attendance at the funeral homes where the four bodies were. I was in charge of three of the funeral processions. The first funeral was held in Harmar Township, and more than 400 firefighters stood in line, and about 60 companies were at the cemetery.
Two more funerals were held that same day in Harrison Township. We estimate about 500 firefighters and ambulance personnel and about 80 companies were in line. The last procession was held the following day in Harrison Township. This was a full fireman’s funeral procession. The Honor Guards from Pittsburgh and New Kensington (PA) led the procession. About 1,000 firefighters and citizens and more than 120 units were along the procession route.
The following Sunday a memorial service was held at Highlands High School in Harrison Township. It was attended by the governor, local dignitaries, and Chief Dennison of Pittsburgh; Allegheny County Fire Marshall John Klaus was the main speaker. It had been estimated that about 1,500 to 2,000 people would attend this memorial service—attendance was more than 4,000, including many local families —from grandparents to young grandchildren. Firefighters from all over the United States attended. The fire stations in the area opened their doors and provided lunches for attendees.
Congratulations on a very fine, indepth article on this tragedy. Our young firefighters had their eyes opened to the fact that death also rides the engines.
Jay Jordan
Assistant Chief
Citizen’s Volunteer Hose Company
Natrona Heights, Pennsylvania
