It must be recognized by all who perform hydraulic calculations, NFPA 13 does not require a separate or distinct “cushion” to be provided on the sprinkler system’s calculations. (Some people will use the term safety factor, hydraulic buffer or safety margin. Whatever term is used, this refers to a pressure margin below the water supply curve. The remainder of this document will use “cushion” because a safety factor to me implies it should be present.) In fact, NFPA 13 does allow one to design to the water supply curve. In fire protection, the Law of Supply and Demand can be stated as “The supply must be equal to or greater than the demand” or vice versa “The demand must be equal to or less than the supply”.
Recently this universal law has come under question and in the last NFPA 13 code cycle, there was a submittal requiring one to determine and apply a cushion to the water supply. This change did not make it into the 2016 edition of the standard; however, it appears that this is becoming a concern in the industry. The issue is so critical to 13 that the Fire Protection Research Foundation recently published a report entitled “Quantification of Water Flow Data Adjustments for Sprinkler System Design.” (http://www.nfpa.org/research/fire-protection-research-foundation/projects-reports-and-proceedings/suppression/other-sprinkler-protection/quantification-of-water-flow-data-adjustments-for-sprinkler-system-design)
The stated goal was “to identify the variables in water supplies that affect hydrant flow tests, which are used in the design of water based fire protection systems. The results of this research are intended to be used to establish recommendations for adjustment to water supply data to be used for the design of fire protection systems. Adjustments would be necessary to ensure that data used for design of fire protection systems represents the actual water supply system conditions during peak demand, accounting for parameters such as tank level during testing and normal system operations.”
A pretty ambitious goal! In reviewing the goal further, it seems like the conclusion is built in. The goal is to establish recommendations for adjustment to water supply data. Should not the initial goal of such a study be to determine if there is a need for adjustments? I keep going back to Chet Schirmer’s question as chair of the 13 committee whenever a change was proposed—“Where is America burning?” (For those of you not familiar with Chet Schirmer, he was a previous chair of the 13 committee and one of the individuals who developed the initial criteria for rack storage.) As they say in Missouri, the Show Me state, show the industry where the lack of adjustments to the water supply is causing system failure. Whatever happened to the adjustments to the water supply that I was told had to be done for any design? It was hammered into me during my early life as a designer—Perform the flow test then adjust for seasonal low pressure, then adjust for time of day low pressure. If we do all this, then why is there a need for additional cushion? From the discussions during the writing of NFPA 13, it appears this is not being done. One is just taking the water flow test at any time of the day or year and using it in the calculations. If the fire occurs on that day, at that time, I guess we are covered.
The other item hammered into me is there are natural safety factors built into hydraulics that one should recognize. These factors include:
• What is the likelihood that the fire will occur when we are experiencing the seasonal and time of day low pressure?
• What is the likelihood that the fire will occur at the hydraulically most remote part of the building?
• What is the likelihood that the fire will grow 1.2 times bigger in the direction of the branch lines? Or better yet, how does the fire know which way the branch lines are running?
• What, in actuality, is the hose stream allowance?
• When the first sprinkler discharges, what density is it putting out? The initial sprinkler doesn’t discharge at a rate of 0.1 gpm per square foot.
• What is the impact of velocity pressure on the system demand? Velocity pressure will normally lower the demand but is not frequently considered.
• How many fires actually reach 1,500 square feet? In 1970, NFPA Fire Journal published an article entitled Automatic Sprinkler Performance Tables which showed that 7 or less sprinklers on a wet pipe sprinkler system operated in 85.2% of fires. (Seven was used as the value because this represents the minimum number of sprinklers in a light hazard occupancy design.)
• How accurate is the Hazen Williams formula in calculating pressure drops? The Hazen Williams formula was developed for low flows in large pipes. It is conservative when applied to sprinklers. There have been numerous studies presented to the 13 committee arguing for the use of Darcy-Weisbach formula.
With all the safety factors built in, I would pose the question again—is an adjustment necessary? I would also amend the question to state “Is an adjustment necessary when the pressure is already adjusted for seasonal and time lows?”
As outlined in the Research Foundation report and as any contractor can tell you, many municipalities require a cushion to be provided. The supply curve shall be reduced by 5 psi or the supply curve shall be reduced by 10%. I came across one where the statement was made that the initial sprinkler discharge pressure shall be increased by 10%. This compounded itself every step of the way. US Department of Energy 1066 Standard required 10 psi safety margin from the water supply curve. FM Global limits velocity pressure to 20 fps which effectively builds in a cushion.
If you are in a jurisdiction and uncomfortable with a design to the curve, then by all means, mandate a cushion. But please make the required cushion apply to all projects and write it as an amendment to the standard so that the contractor is aware of it and can design accordingly. This appears to be what the NFPA sprinkler committee is trying to decide. Determine if a cushion should be required and then come up with a method so that one can design an appropriate system keeping in mind that the “purpose of this standard shall be to provide a reasonable degree of protection for life and property…” If we didn’t care about costs, we would mandate a private tank and fire pump for all systems.
The last item to discuss can be summarized with one line—“I measured with a micrometer and cut with a chain saw.” To those of you who have done flow tests on a cold day in January, you know that the accuracy of the test is inversely proportionate to the temperature of the air. The colder it gets, the quicker I can read the gauge. The gauge that is flickering as I gently locate the pitot tube by hand, ½ the diameter into the stream coming out of the hydrant, ½ the diameter away from the hydrant. That flickering gauge sets the water supply and is the criteria that determines the entire design of the sprinkler system. Maybe this is the reason for cushion—people are finally recognizing that there is an inherent inaccuracy in the flow test itself that one needs to compensate for.
Next month I plan on talking about some serious deterioration in water supplies that I have observed and go back to NFPA 25 once again.
As always, I welcome your comments: firstname.lastname@example.org
Jerry Schultz, P.E.