"Exposure Protection" and "Fire Resistive Rating" are two completely different terms that are often confused when it comes to the application of automatic fire sprinklers. NFPA standards such as 101,Life Safety Code; 80A,Protection of Buildings from Exterior Fire; and 13,Installation of Sprinkler Systems, all reference protection criteria for exposure protection. None of these, however, state that the closely spaced sprinklers required for exposure protection establish a fire rating as required by the building code.

Figure 1. The top of a typical K5.6 distribution pattern at various flows.
When sprinklers are referenced for exposure protection for glass, the comment is usually followed by the phrase: "Sprinklers shall be capable of completely wetting the entire surface of the glass." Figure 1 is an example of a standard spray sprinkler used to protect a glass surface. It shows that if the minimum spacing between standard spray sprinklers is 6 ft. (which means they will not wet each other during operation), they will not wet the entire surface of the glass when installed at the recommended 6-ft. spacing. Figure 1 also shows that at 3 ft. away (half the recommended spacing) from the sprinkler, the top of the discharge pattern will be approximately 12 in. below the sprinkler. How can a standard spray sprinkler wet the entire surface of the glass if the distribution pattern falls 12 in.?

Currently, there is also confusion between the term, "Listed Window Sprinklers," as advertised by some manufacturers. If you were to look in the Underwriters Laboratories (UL) Fire Protection Equipment Listing Directory, you will find a listing category titled: "Sprinklers, Outside (VOKR)." Under this heading, you will find an explanation stating, in part: "These sprinklers are non-automatic and are intended for protection of windows, walls and roofs against exposure fires." These nozzles (none in this category have operating elements) are intended for use with deluge systems to provide exposure protection and will not provide a fire resistive rating.

What are these nozzles, and why don't they provide a fire resistance rating? Let's first visit the purpose of these UL-listed sprinklers. Chapters 4-7, 7-9.6, and 8-7 of the 1999 edition of NFPA 13, as well as NFPA 80A, provide design criteria for protecting a building from a fire from a neighboring building. This is known as "exposure protection." In short, exposure protection requires that sprinklers be spaced closely together (usually 6 ft. apart), just outside the exterior wall of a protected building. Depending on the particular building code in effect, exposure protection is required at different floor levels when a building is within a given distance of a neighboring building or property line, and the wall does not meet minimum fire rating requirements. This chapter goes on to give minimum levels of sprinklers required, as well as minimum flow rates and corresponding orifice sizes to meet these flow rates. The flows from these outside sprinklers must then be added to the inside sprinkler hydraulic calculations. The minimum flows vary in such a manner that much smaller orifices than usual may be sufficient. Hence, the listing of orifices starts as small as 1/4". Because these sprinklers are outside the building, these systems typically utilize deluge valves and open sprinklers. Water in the piping for these sprinklers would be subject to freezing.

For quite some time, standard spray sprinklers with fusible elements have been used to protect glazing in walls that were required to be fire rated. These walls may have been required to maintain an occupancy separation by the building code. They may also have been required to carry a two-hour fire rating because they were part of an egress corridor. This is typically required to assure safe passage of occupants out of a building during a fire.

During the design phase of a building, the design team--possibly consisting of the architect, engineer, insurance underwriter, (hopefully) the local fire authority and others--would meet to discuss the requirements for the particular project. The aesthetic desire for the use of glazing to create an open, airy feel in the building would often result in conflict with required building and fire codes. This conflict would usually be resolved by concessions from the building and fire authorities. It seemed reasonable that if closely spaced sprinklers were located near the glass, this would create a kind of "water curtain," thereby protecting the glass from failure.

Figure 2. A comparison of standard spray sprinklers and a new model of window sprinkler.
There are a few inherent problems with this theory, however. Where is the ceiling in relation to the top of glass? Is it 6 in., 1 ft., or possibly 2 ft. above the glass? How far laterally from the glass will the sprinkler be located? What flow rate should be used for the sprinklers for the hydraulic calculations? Some of these questions were answered by referring to other sections of NFPA 13 that dealt with "water curtain" requirements around openings, or by looking at the aforementioned "exposure protection" requirements, or a combination of the two. Regardless of the best intentions and engineering judgment of the building and fire officials, questions and uncertainty remained. Would this sprinkler, with its particular spray pattern, located 9 in. off and 1 ft.-4 in. above the top of the glass, actually wet the entire surface of the glass? Unfortunately, the water distribution does not bend around the header or mullion, and the top portion of the glass would remain dry, as shown in Figure 2. At this point, the most that you have is a water curtain with no fire resistive rating. Does it matter if the sprinkler doesn't wet the entire surface of the glass and leaves dry spots? The answer to this is a resounding YES.

More questions remain. Will the sprinkler(s) respond quickly enough to prevent thermal shock once the water discharges onto the hot glazed surfaces? What data and testing was used to allow the use of these sprinklers in this particular application? Did the building code or fire code reference specific criteria and requirements for this installation? Who assumes liability, the installing contractor or approving authority?

Figure 3. The standard time-temperature curve specified in ASTM E-119.

A New Way of Thinking

To answer these questions, sprinkler manufacturers had to consider a new design for automatic sprinklers specifically for use with glazed windows. The new window sprinkler that resulted from these deliberations was designed to offer both exposure protection and fire resistance to meet multiple code requirements.

And because these sprinklers and their use, in conjunction with glazed partitions, went a step beyond the typical use of automatic fire sprinklers, it was also necessary to step beyond the traditional listing procedures and reports. Evaluations and corresponding reports were sought from the National Evaluation Service (NES), the ICBO Evaluation Service and the Building Materials Evaluation Commission (BMEC) of Canada. And a "Specific Application Sprinkler" listing was established at Underwriters Laboratory under the category "Window Sprinklers" to test the sprinklers and ensure that they do provide a fire resistive rating.

In 1995, the test program was established in conjunction with UL for the purposes of determining if the window sprinklers, when located within specific parameters, and with a specific discharge pattern, would maintain the integrity of single pane, heat-strengthened glass for two hours, when subjected to a window exposure fire test, as well as a 40 kW fire test at the base of the glazing.

The window exposure fire test utilized test furnaces which were calibrated to bring room temperature up to 1,832 degrees F (1,000 degrees C) over a period of two hours. This was done to maintain the standard time-temperature curve specified in ASTM E-119 (Figure 3) and CAN/ULC-S101-M89.

Figure 4. UL's 40 kW fire test for window sprinklers.
In addition to this test, a 40 kW fire test was performed using a pan of lit heptane at various distances off the face of the glass partition (Figure 4). This could simulate a waste paper basket fire, or fire of similar load starting at the base of the glass. By simulating this relatively small fire at the base of the glazing, a more challenging scenario was created. A fire of this size, concentrated in a small area of the glass may not be large enough to generate quick enough temperature increase in the room, but can create enough stress on the glazing surface to cause thermal shock of the glass when the sprinklers eventually operate.

These tests included glazing assemblies connected with vertical mullions as well as glazing assemblies that were connected by a silicone butt joint.

After successful completion of the tests and the establishment of subsequent installation parameters, these new types of window sprinklers were approved for use as an "equivalent method of construction," when acceptable to the Authority Having Jurisdiction, to provide an equivalency of a two-hour rated partition when using glazed assemblies. Thus, by stepping beyond the traditional use of automatic sprinklers, a new design was able to unite the concepts of exposure protection and fire resistance in these glazed window applications.