The Theory and Practice of Variable Orifice in Automatic Sprinkler Systems

Transcription

1 The Theory and Practice of Variable Orifice in Automatic Sprinkler Systems Abstract In the industry of Fire Protection a large number of Automatic Fire Sprinklers has been developed trying to cover all kinds of fire situations. For each case a different type of sprinkler is used, with different shape and size, in order to control (or suppress) the fire according to proven standards aiming to use enough water for the task but not too much in order to avoid water damage. At the event of fire, sprinklers start operating one after the other. The first sprinkler employs the maximum water available according to the pressure provided to the system. As more sprinklers start operating the pressure in each sprinkler drops but the total water flow increases until the system reaches the point of controlling the fire. It would be advantageous to employ larger sprinklers in order for the firsts to provide larger flows and control the fire sooner. On the other hand, should the number of operating sprinklers increase the total amount of water provided may be much more than needed resulting in a need for a larger water tanks and fire pumps and potentially causing extensive Water Damage. The Water Flow depends on the Pressure and on a Constant Factor K defined by the diameter and shape of the orifice. This article demonstrate how by developing a Variable Orifice it is possible to optimize the Water Flow of the sprinkler by maximizing the Flow of the first operating sprinklers and on the same time not exceeding the maximum flow determined by the standards. 1/21

2 The development of the Theory and Practice the Variable Orifice in this article is accompanied by the results of tests and by suggestions for farther development of this idea. 1 Introduction From the day the Automatic Sprinkler was invented, the orifice of the sprinkler has had a fixed value defined by the diameter and shape of the orifice. This characteristic creates some limitations to sprinkler performance, such as: 1.1 Limitation in the usage of sprinklers. Different applications (i.e. different occupancies) require different sprinklers. 1.2 In a process of fire development, more water throughput from the initial sprinklers opened would enable faster control of the fire, but flow is limited by the orifice size. 1.3 Each sprinkler exhibits optimal performance at a limited range of pressures. In order to improve sprinkler efficiency, and to broaden the range of operation, etc. it is suggested in this article to develop a sprinkler with a Variable Orifice. 2 Current situation 2.1 The basic characteristic of the water flow via the Fire Sprinkler is defined by the equation: Q=K*P ½ 2/21

3 Where: Q = The Flow [gpm] P = The Pressure [psi] K = A constant Coefficient [gpm/psi ½ ] This equation is the basis of the Hydraulic Calculation of the Sprinkler Systems. The K factor is specific to each sprinkler, for example: In case of Light or Ordinary Hazard a K factor of 5.6gpm/psi ½ may be sufficient, while for Extra or Special Hazard, a higher K factor will be required. 2.2 Poor choice of K factor may cause one of two results: Using a K factor that is too small will result in the need for high pressure in order to supply the required water flow. This may require a larger Fire Pump or in case of connection to the public water system, lake of sufficient Pressure A K factor that is too high will waste water because of the minimum starting pressure (7psi) needed, which may also require a bigger Water Tank or in case of connection to the public water system, insufficient pressure will lead to lack of sufficient water flow. 2.3 Recent tests showed that in Sprinkler System in Storage buildings it is very efficient to use sprinklers with a bigger K factor. [1. Thomas & Tomes]. 3 Theory of the Variable Orifice 3.1 In an average system, the standard demands calculation of 12 sprinklers, approximately, to control the fire at a safe area. When fire breaks up, the sprinklers start operating one after the other according to the development of the fire. The statistic shoes that in 27% of the cases only one sprinkler was 3/21

4 needed to control the fire, and in 65% of the cases up to five sprinklers were required. The number of sprinklers needed depends on the fire development rate and on the amount of water applied at the starting point of the fire. The Hydraulic Calculation takes into account the total water capacity needed to extinguish a fire in the worst case, and, accordingly, the total pressure needed to deliver it to the fire. At the beginning of the response, the first operating sprinkler will receive the entire pressure available in the system. This pressure will be reduced, gradually, during the development of the fire, as additional sprinklers start operating. 3.2 Example: A building with Ordinary Hazard Group II Occupancy (which may be considered as an average case), 150ft long, 125ft wide, 25ft high with a sprinkler system consisting of K=8gpm/psi ½ sprinklers, 20 1½ branch lines 10 apart with 10 sprinklers, each 12.5 apart, and a 3 cross main at the middle along the building, and a 3 feed main: The minimum flow needed for the first sprinkler (the furthermost) is: Q = 0.2gpm/ft 2 x 12.5ft x 10ft = 25gpm The results of the hydraulic calculation are: First sprinkler pressure: P 1 = 9.8psi. Total flow of the system at source: 318.6gpm. Maximum pressure needed at source: 46.6psi. When the first sprinkler start operating at the maximum above pressure, it will flow 45.3gpm at 30.1psi which is 81% more water! Later, the flow will be gradually reduced as more sprinklers start operating. This fact explains how one sprinkler may control the fire at its early stages. 4/21

5 3.3 Another way of improving sprinkler performance is choosing a sprinkler with a bigger K factor. This option is efficient only when the source of water has high potential of Flow and Pressure. In the example above, if the K factor was increased to K=16.8gpm/psi ½ the result of the hydraulic calculation would have been: Total Flow of the system at source: 650.4gpm. Maximum Pressure needed at source: 113.3psi. This pressure may not be available or alternatively a larger water tank would be required. These results are due to the minimum 7psi pressure required for sprinkler function. 3.4 In order to overcome those limitations, it is suggested to replace the Constant Sprinkler Orifice with a Variable Orifice Sprinkler (VOS). (Possible configurations will be shown in the next section). This will change the constant factor to a variable factor, defined as V, with the same dimensions of gpm/psi ½. The factor V will be a function of the pressure: V = F(p). And the basic equation of the hydraulic calculation will get the form: Q=V*P ½ The exact formula is defined according to the following conditions: It should comply with the range of pressure as established by the NFPA, i.e. Pressure between 7psi to 175psi and the K factor between 5.6gpm/psi ½ to 28gpm/psi ½ and It should comply with the range of the K factor as established by the NFPA, i.e. between 5.6psi to 28psi It should respond to the pressure in a way that when the pressure decreases so will the orifice size. The V Equation may take the form of: V=a*P b 5/21

6 Inserting the above NFPA conditions into this equation will give the formula for V as follows: V=2.12*P 1/2 By inserting to the equation of Q the formula of V, the Q equation will become: Q=2.12*P Note: The final function will be derived from further research and the construction of the Variable Orifice. Or in Graphic Chart: V( P) P Figure The graph in Figure 1 represent the value of V(gpm/psi ½ ) as a function of the pressure P(psi). 3.5 One way to determine the efficiency of the Variable Orifice Sprinkler is to compare the performance of the VOS to that of standard sprinklers, i.e. the output of the flow Q(gpm) as a function of the Pressure Two sprinklers were taken for this comparison (see Figure 2): 6/21

7 The standard sprinkler with K=5.6gpm/psi ½ which produces Q=5.6*P ½, marked by a doted blew line The largest sprinkler available K=25.2gpm/psi ½ which produces Q=25.2*P ½ marked by a dashed green line Those two compared to the VOS which produces Q=2.12*P marked by a strait red line. 400 Q V ( P ) 300 Q 5.6 ( P ) 200 Q 25 ( P ) P Figure Figure 2 clearly shows better performance at high pressure compared to both sprinklers will and it still affective at a low pressure (compared to the ESFR which is known to loose it momentum). Farther discussion on this subject to be found in section 5 Theoretical Feasibility Study. 7/21

8 4 Construction of the Variable Orifice There are various ways to construct the Variable Orifice. Below are the major models [2. Mehr]: 4.1 Add-on device: A1 A2 Figure 3 This device is installed between the sprinkler (which has an extra large orifice) and the pipe. Description: Figure A1 shows the device in a stand-by position. The thick black lines represent an elastic sleeve diaphragm containing an appropriate fluid under pressure from a piston pushed by a spring (indicated at the right hand side), forming an inside orifice. When pressure starts building in the device, it flattens the diaphragm and enlarges the orifice. 8/21

9 4.1.2 Figure A2 shows the device with maximum pressure, which pushes the diaphragm all the way to the device walls, enabling the orifice to reach its maximum size The advantage of this configuration is that it is based on existing tested sprinkler designs, saving a lot of research time, but it may not be the most cost effective configuration. 4.2 Special sprinkler built with a Variable Orifice: B1 B2 Figure 4 This special sprinkler resembles a concealed or an institutional sprinkler where the deflector pops out. However, the sprinkler incorporates two changes: The rods of the deflector are held in place by springs and a cone is added to the inside of the deflector. Using this arrangement, as the pressure is increased, the cross section of the orifice is enlarged. 9/21

10 4.2.2 Figure B1 shows the device at a stand-by position and Figure B2 at maximum pressure with the orifice at its maximum size. The configuration of the cone will determine the exact function of the V factor Development of this type of sprinkler may require extensive research, more than needed for a regular sprinkler, and as such its development is not anticipated to be cost effective. 4.3 Finally, a Variable Orifice Sprinkler could be built by adding a special low cost insert to an existing Large Orifice sprinkler: C C1 C2 Figure 5 The insert enables conversion of a Large Orifice Sprinkler like ESFR to a multi purpose one with the ability of controlling fire in various occupancies. 10/21

11 4.3.1 Figure C shows a Large Orifice Sprinkler with the insert in the stand-by position Figure C1 shows the insert in the closed position from side, top and axonometric views Figure C2 shows similar views of the insert in the open position. 4.4 In principal, all configurations must not on/off position, and shall have a minimum opening to ensure proper operation. 5 Theoretical Feasibility Study This paragraph compares a model of typical sprinkler system installed with various K factor sprinklers versus the same system installed with Variable Orifice sprinklers, in order to establish the optimal performance and the limitations of each type of sprinkler. 5.1 The model: Cover a space of 150 x 125 x Made of schedule 10 steel pipes as follows: 1½ Branches, long with 10 sprinklers installed 12.5 apart. A 3 Cross main, along the building at the middle. A 3 riser, connecting the cross main with the water source, including an alarm check valve The model was calculated with four different types of sprinklers: K=8gpm/psi ½, K=11.2gpm/psi ½, K=16.8gpm/psi ½ and sprinkler with a variable V factor calculations were made to enable comparison between four sets of available pressures at source: the minimum required to comply with the standard, a low Pressure of 60psi, a medium Pressure of 100psi, and high pressure of 175psi. 11/21

12 5.1.5 For each system with each type of sprinkler, up to 12 sprinklers were calculated, starting for from the first sprinkler and adding one additional sprinkler unit for each calculation. 5.2 The first set of calculations, shown in Table 1 determines the minimum pressure needed to comply with the standard requirements For each calculation, the Flow (in gpm) is given. The number in brackets represents the minimum pressure (in psi) at source. Number of P = min Sprinklers K=8.0 K=11.2 K=16.8 V (21.8) 29.6 (19.5) 44.4 (21.4) 25.0 (22.5) (24.4) 60.0 ( (29.4) 50.5 (25.1) (25.2) (29.4) (32.2) 80.6 (26.0) (28.4) (35.8) (44.0) (30.0) (30.1) (31.5) (49.9) (31.2) (32.2) (32.9) (52.0) (32.2) (33.0) (35.8) (60.3) (34.5) (34.9) (38.5) (66.3) (36.4) (39.7) (46.6) (91.5) (42.7) (42.4) (50.9) (103.2) (45.9) (44.8) (54.2) (110.2) (48.1) (46.6) (56.3) (113.3) (49.6) Table Conclusions: The results for V are similar to those of K= The flow values are higher as the K factor increases, but this result may not be applicable because of the high pressure needed at source. 5.3 The second set of values in Table 2 illustrates the flow delivered for each case where the pressure at source is 60psi. 12/21

13 5.3.1 The results from Table 1 are shown in Table 2 for comparison. Number of P = min P = 60 Sprinklers K=8.0 K=11.2 K=16.8 V K=8.0 K=11.2 K=16.8 V (21.8) 29.6 (19.5) 44.4 (21.4) 25.0 (22.5) (24.4) 60.0 ( (29.4) 50.5 (25.1) (25.2) (29.4) (32.2) 80.6 (26.0) (28.4) (35.8) (44.0) (30.0) (30.1) (31.5) (49.9) (31.2) (32.2) (32.9) (52.0) (32.2) (33.0) (35.8) (60.3) (34.5) N/A (34.9) (38.5) (66.3) (36.4) N/A (39.7) (46.6) (91.5) (42.7) N/A (42.4) (50.9) (103.2) (45.9) N/A (44.8) (54.2) (110.2) (48.1) N/A (46.6) (56.3) (113.3) (49.6) N/A Table Conclusions: At the worst case, when all 12 sprinklers operates, the flow in the V type system almost converges to the flow in the K=8 type, which means less waste of water and minimal size of Water Tank The advantage of systems with higher value of K factor is the delivery of high flow at the beginning of the fire, i.e. fewer sprinklers may be needed to control the fire. On the other hand, at the worst case a higher degree of water damage may be expected Sprinklers with high value of K factor will not be allowed in this case, as it will not reach the minimum pressure of 7psi required by the standard. 5.4 Table 3 shows the results of the calculation for a medium pressure of 100psi and a high pressure of 175psi, respectively. 13/21

14 5.4.1 As more water is delivered with fewer operating sprinklers, it is expected that the number of sprinklers required by the standard will decrease. Theoretically the minimum flow to control the fire for this case is 1500ft 2 x 0.2gpm/ ft 2 = 300gpm (as in example 3.2 above). In Table 3, for each set of calculations, the first flow accedes 300gpm is marked in blue. Number of P = 100 P = 175 Sprinklers K=8.0 K=11.2 K=16.8 V K=8.0 K=11.2 K=16.8 V N/A N/A N/A Table Conclusions from the theoretical feasibility study: Variable Orifice Sprinkler (VOS) does not overflow the area of operation when maximum sprinklers are required; as more sprinklers open, the orifice decreases In the process of sprinkler operation, the first sprinklers deliver significantly more water which may eliminate the need for more sprinklers to control the fire and to use the advantage of Early Suppression in all Hazard Occupancies In the case of medium and high pressure at the source, three or two sprinklers (respectively) will exceed the minimum 300gpm needed to control the fire. This conclusion is, of course, theoretical. Further research is needed to determine 14/21

15 the exact number of sprinklers required as a function of the pressure at source, and to determine the optimum pressure at source The above theoretical feasibility study does not take in account the large droplet effect, which may improve the performance of the VOS specifically for regular hazard occupancies versus ordinary sprinklers, as will be demonstrated below. 6 Physical Feasibility Study A partial feasibility study was done by a group of students at the Worcester Polytechnic Institute (WPI) under the supervision of Professor Jonathan R. Barnett [3. Barnett] to compare the performance of the VOS to that of the ESFR. 6.1 The way to construct the VOS is to take or Construct a sprinkler with a very large orifice and to add a mechanism that will reduce the size of the orifice as the pressure declines. As described in paragraph 4 above, there are 3 basic ways to accomplish it: To build a new type of sprinkler, to resemble the one shown in Figure 3 only for testing of the principal: (All Figures and Tables in this section are from WPI) The drawback of this solution is the needs of full research including design of the deflector To take a sprinkler with a large orifice and to add the mechanism between the sprinkler and the piping, as shown in Figure 2. Although it has the advantage of flexibility as it does not requires building a special sprinkler, this is not a cost effective solution The solution shown in Figure 4 seems to be the most flexible and cost effective. 6.2 The team at WPI chose to build a sprinkler similar to that shown in Figure 3, because of the time limit assigned to the project and the lack of tools to produce 15/21

16 the inserted element. But as the goal of their project was to investigate the feasibility of the principal of the Variable Orifice by comparing between the performances of the VOS and the ESFR at high and low pressure and develop a new sprinkler, this arrangement serves the goal. 6.3 The first parameter determined by WPI for the specific sprinkler (Figure 6) is the Variable K Factor. This was done by two separate procedures: Figure The first procedure included recording pressure by use of a pressure gauge and determining flow by use of a stop watch and graduated cylinder An additional test was performed to determine the K-factor range of the variable orifice sprinkler, except more precise measures were used. This time, a pressure transducer was used to average pressure over time and a scale was used to determine the volume of water dispersed. The test was conducted to provide more a more accurate means of determining K-factors. The two figures below show the K factor function for the deflection of the sprinkler deflector which is a direct function of the pressure. 16/21

17 The correspondence between these two figures to the function of the VOS, shown in Figure 1, is outstanding. K-factor vs. Deflection K-factor (gpm/psi^(1/2)) Deflection (in) Figure 7: K-factor vs. Deflection - Procedure 1 K-Factor vs. Deflection K-Factor (gpm/psi^(1/2)) Deflection (in) Figure 8: K-factor vs. Deflection - Procedure 2 17/21

18 6.4 The second parameter was a comparison of Droplet size in ESFR vs. VOS at High and Low Pressure. Sprinkler Droplet Size Description Type High Pressure Low Pressure The ESFR sprinkler had The ESFR barely had drops ESFR large drops but there was at this pressure. It was also a fair amount of mist that was created by the sprinkler as well. closer to a continuous drip of water at the center of the spray pattern with a small canopy at which the droplets were large, but with no thrust. VOS The VOS had smaller drops than the ESFR sprinkler, but the same amount of mist was created. Table 4 The VOS had large drops that were also mixed with a little mist. There was quite obviously a better developed spray pattern, but the water in the center of the spray pattern was noticeably reduced. The performance of the VOS deflector was not optimized and it can be refined to give equal or better performance than that of the ESFR. 6.5 The third parameter measured is a comparison of spray pattern of ESFR vs. VOS at high and low pressure Table 5 shows that the spray pattern radius of the VOS is larger than of the ESPR at high and at low pressure. Spray Pattern Radius [m] (ft) Sprinkler Type High Pressure Low Pressure ESFR 4.04 (13.3) 1.30 (4.25) VOS 4.36 (14.3) 2.19 (7.17) Table 5 18/21

19 6.5.2 The photos in Figure 8 illustrate the comparison between the spray patterns of the ESFR vs. the VOS and demonstrate clearly the wide range of operation of the VOS regarding the pressure range. Sprinkler Spray Pattern Type High Pressure Low Pressure ESFR VOS Figure Conclusions of the Physical Feasibility Study: The VOS performs similarly to the ESFR at medium and high pressure and better at a low pressure At very low pressure, the VOS behaves similarly to a standard sprinkler but with the advantages of larger drop size. 19/21

20 7 Final Conclusions and suggestions for Additional Research 7.1 The Variable Orifice Sprinkler can operate within a wide range of pressures. 7.2 The Variable Orifice Sprinkler demonstrates a high degree of flexibility, and can operate at all kinds of hazard occupancies from light hazard to extra hazard and special hazard, including storage etc. 7.3 As the first sprinklers deliver much more water, it is expected that fewer sprinklers will be needed to control or even to suppress the fire, as a result, less water might be needed and less water damage is expected. 7.4 Recommendations for additional research: Determination of the optimal function of V in the basic equation (Q=V*P ½ ) so that the first sprinklers will deliver the maximum possible water, while all sprinklers calculated will deliver a total amount of water, as required by the standard As the insert type device (Figure 5) seems to be the most cost effective, the insert should be designed to best comply with the optimal function of V, as determined above (section 7.4.1) A full scale research to re-determine the minimum number of sprinklers needed for each occupancy hazard should be performed Investigation to determine if the VOS fulfils the criteria of Quick Response Early Suppression (QRES) sprinklers as defined by the standard [4. Budnick & Fleming] which may improve the Fire Protection Level with much less water. 20/21

21 ACKNOWLEDGMENTS The author would like to acknowledge the excellent work done by Professor Jonathan Barnet and the students Kuhn Brian, Plati Joseph, and Purtell Andrew in testing the Variable Orifice at the WPI Worcester Polytechnic Institute. The author would also like to thank Mr. Russell Fleming Executive Vice President of the NFSA National Fire Sprinkler Association and Mr. Arthur Cote Executive Vice President & Chief Engineer of the NFPA National Fire Protection Association for reviewing this article and their helpful remarks and suggestions. REFERENCES [1] Thomas Bill & Toes Bill, The Effect of Large Orifice Sprinklers on High Challenge Fires, NFPA - Automatic Sprinkler Systems Handbook 2002 Supplement 2. [2] Mehr R. Ralph, Automatic Fire Sprinkler having a Variable Orifice US Patent Pen. No. 10/624, [3] Barnett R. Jonathan Professor, WPI Worcester Polytechnic Institute & Kuhn Brian, Plati Joseph, Purtell Andrew, Sprinkler Head Design A Major Qualifying Project Report [4] Edward K. Budnick & Russell P. Fleming, Developing an Early Suppression Design Protection for Quick-Response Sprinklers NFPA - Automatic Sprinkler Systems Handbook 1996 Supplement 4. 21/21