3-00 Hydraulic Principles

FM Global Property Loss Prevention Data Sheets
3-0
September 2006 Page 1 of 59
HYDRAULICS OF FIRE PROTECTION SYSTEMS
Table of Contents
Page THEORY OF WATER FLOW IN PIPES ........................................................................................................ 4 Bernoullis Theorem and Applications .................................................................................................... 4 Hazen-Williams Formula ......................................................................................................................... 6 Pressure Loss at Fittings ........................................................................................................................ 7 Discharge from Nozzles ......................................................................................................................... 7 Discharge Coefficient .............................................................................................................................. 8 Theoretical Discharge ........................................................................................................................... 13 Nozzle K Factor .................................................................................................................................... 13 Combined Sprinkler Discharge and Pipe Flow .................................................................................... 14 TESTING WATER SUPPLIES ..................................................................................................................... 15 Purposes ............................................................................................................................................... 15 Hazards ................................................................................................................................................. 15 Flow and Pressure Measurements ....................................................................................................... 15 Test Planning ........................................................................................................................................ 17 Private Water Systems ......................................................................................................................... 18 Public Water Systems .......................................................................................................................... 19 Test Arrangement ................................................................................................................................. 21 Test Procedure ..................................................................................................................................... 22 Determining Hydraulic Gradients .......................................................................................................... 23 Special Water Test Techniques and Considerations ............................................................................ 26 PRESENTING WATER SUPPLY DATA ...................................................................................................... 28 Supply Curves (Fig. 16) ........................................................................................................................ 28 Friction-Loss Curve for a Single Pipeline (Fig. 17) .............................................................................. 28 Curve Combination - Methods and Meaning ....................................................................................... 30 Calculation of Yield from a Single Supply ............................................................................................ 31 Calculation of Yield from Supplies in Parallel (Fig. 21) ........................................................................ 35 Calculation of Yield from Supplies in Series (Fig. 22) ......................................................................... 36 HYDRAULICS OF SPRINKLER SYSTEMS ............................................................................................... 37 Sprinkler System Demand Specifications ............................................................................................ 37 Use of Velocity Pressure ...................................................................................................................... 39 Sprinkler System Flow and Pressure Adjustments .............................................................................. 46 Multilevel Systems ................................................................................................................................ 46 Looped and Gridded Systems .............................................................................................................. 47 Overlay Systems ................................................................................................................................... 57 EXISTING SYSTEMS - CONVERTING TO NEW SPECIFICATIONS ....................................................... 57
List of Figures
Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. 1. 2. 3. 4. 5. 6. 7. 8. 9. Relationships between various hydraulic factors in typical piping ..................................................... 5 Typical hydrant butts ........................................................................................................................... 8 Water flow devices ............................................................................................................................ 16 Typical pressure gauge locations ..................................................................................................... 16 Pitot tube with gauge and air chamber ............................................................................................ 16 Taking pitot readings at hose nozzle ................................................................................................. 17 Simplified, freehand line diagram of water supplies ......................................................................... 17 Single-source water supply ............................................................................................................... 18 Single source with underground loop ................................................................................................ 18
2006 Factory Mutual Insurance Company. All rights reserved. No part of this document may be reproduced, stored in a retrieval system, or transmitted, in whole or in part, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without written permission of Factory Mutual Insurance Company.
3-0
Page 2
Hydraulics of Fire Protection Systems

Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig.
Fig. Fig. Fig. Fig. Fig. Fig.
10. Two sources with underground loop ............................................................................................... 19 11. Typical public water system near a plant site ................................................................................. 20 12. Two dead-end pipes from a public water system near a facility site .............................................. 21 13. Single source with partial loop ........................................................................................................ 22 14. Relations between water pressure and elevation in pipe having uniform flow .............................. 23 15. Pipeline for which hydraulic gradient is computed (upper) and plot of profile and hydraulic gradient to the pipe line (lower) ...................................................................................... 25 16. Supply curves .................................................................................................................................. 29 17. Friction loss curve for a single pipeline .......................................................................................... 30 18. Combination of supply and friction loss curves to determine yield ................................................ 31 19. Typical loop in distribution system .................................................................................................. 32 20. Friction loss and supply curves in parallel pipe (loop) systems ..................................................... 33 21. Calculation of yield from supplies in parallel. Note: Curves (e), (g), and (h) do not show more than 150% of rated pump flow because that flow is the maximum guaranteed for all pumps and all installations. ............................................................................................................ 34 22. Calculation of yield from supplies in series .................................................................................... 35 23. Sprinkler layout used in waterflow calculation examples ................................................................ 38 24. Flow and pressure adjustments for specific array of operating sprinklers ..................................... 47 25. Demand lines for a multilevel sprinkler system .............................................................................. 48 26. Typical looped feed sprinkler system ............................................................................................. 49 27. Graphical determination of flow split in loop .................................................................................. 51
List of Tables
Table Table Table Table Hazen-Williams Pipe Coefficients For Underground Use ................................................................. 7 Hazen-Williams Pipe Coefficients For Sprinkler System Use .......................................................... 7 Typical Discharge Coefficients .......................................................................................................... 8 Theoretical Discharge Through Circular Orifices, U.S. Gallons Per Minute (l/min). Computed For Discharge Coefficient of 1.00 (Seldom Reached In Practice). Reduce the Given Discharge By Multiplying By A Coefficient Suited to the Particular Opening Used. See Table 3. .................... 9 5. Values of K For Various Discharge Orifices. .................................................................................. 14 6. Data For Example of Gradient Test. .............................................................................................. 24 7. Pump Discharge Pressures ............................................................................................................ 36 8. Sprinkler System Calculation For Example. ................................................................................... 41 8. Sprinkler System Calculation (contd.) ........................................................................................... 42 9. Velocity Pressure Factors ............................................................................................................... 43 10. Sprinkler System Calculation For Example .................................................................................. 44 10. Sprinkler System Calculation For Example (contd) ..................................................................... 45 11. Comparison Of Results For Different Calculation Methods .......................................................... 46 12. Sprinkler System Calculation For Example 1 ............................................................................... 52 12. Sprinkler System Calculation For Example 1. {contd) ................................................................ 53 13. Sprinkler System Calculation For Example 2 .............................................................................. 55 13. Sprinkler System Calculation For Example 2 (contd) ................................................................. 56 14. Comparison of Results From Examples 1 and 2 ......................................................................... 57 15. Conversion Factors and Formulas ............................................................................................... 57 1. 2. 3. 4.
Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table
2006 Factory Mutual Insurance Company. All rights reserved.

3-0
Page 3
Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig.
Fig. Fig. Fig. Fig. Fig. Fig.
9. Single source with underground loop ................................................................................................ 17 10. Two sources with underground loop ............................................................................................... 18 11. Typical public water system near a plant site ................................................................................. 19 12. Two dead-end pipes from a public water system near a facility site .............................................. 20 13. Single source with partial loop ........................................................................................................ 21 14. Relations between water pressure and elevation in pipe having uniform flow .............................. 22 15. Pipeline for which hydraulic gradient is computed (upper) and plot of profile and hydraulic gradient to the pipe line (lower) ...................................................................................... 24 C-1. 11212 in. diameter drain ............................................................................................................... 28 C-2. 2 in. diameter drain ....................................................................................................................... 29 C-3. 3 in. diameter drain. ...................................................................................................................... 30 16. Supply curves .................................................................................................................................. 31 17. Friction loss curve for a single pipeline .......................................................................................... 32 18. Combination of supply and friction loss curves to determine yield ................................................ 33 19. Typical loop in distribution system .................................................................................................. 34 20. Friction loss and supply curves in parallel pipe (loop) systems ..................................................... 35 21. Calculation of yield from supplies in parallel. Note: Curves (e), (g), and (h) do not show more than 150% of rated pump flow because that flow is the maximum guaranteed for all pumps and all installations. ............................................................................................................ 36 22. Calculation of yield from supplies in series .................................................................................... 37 23. Sprinkler layout used in waterflow calculation examples ................................................................ 40 24. Flow and pressure adjustments for specific array of operating sprinklers ..................................... 49 25. Demand lines for a multilevel sprinkler system .............................................................................. 50 26. Typical looped feed sprinkler system ............................................................................................. 51 27. Graphical determination of flow split in loop .................................................................................. 53
List of Tables
Table Table Table Table 1. 2. 3. 4. Hazen-Williams Pipe Coefficients For Underground Use ................................................................. 6 Hazen-Williams Pipe Coefficients For Sprinkler System Use .......................................................... 6 Typical Discharge Coefficients .......................................................................................................... 7 Theoretical Discharge Through Circular Orifices, U.S. Gallons Per Minute (l/min). Computed For Discharge Coefficient of 1.00 (Seldom Reached In Practice). Reduce the Given Discharge By Multiplying By A Coefficient Suited to the Particular Opening Used. See Table 3. ... 8 5. Values of K For Various Discharge Orifices. .................................................................................. 13 6. Data For Example of Gradient Test. .............................................................................................. 23 C-1. Equivalent Lengths of Fittings (ft) .............................................................................................. 27 7. Pump Discharge Pressures ............................................................................................................ 39 8. Sprinkler System Calculation For Example. ................................................................................... 43 8. Sprinkler System Calculation (contd.) ........................................................................................... 44 9. Velocity Pressure Factors ............................................................................................................... 45 10. Sprinkler System Calculation For Example .................................................................................. 46 10. Sprinkler System Calculation For Example (contd) ..................................................................... 47 11. Comparison Of Results For Different Calculation Methods .......................................................... 48 12. Sprinkler System Calculation For Example 1 ............................................................................... 54 12. Sprinkler System Calculation For Example 1. {contd) ................................................................ 55 13. Sprinkler System Calculation For Example 2 .............................................................................. 57 13. Sprinkler System Calculation For Example 2 (contd) ................................................................. 58 14. Comparison of Results From Examples 1 and 2 ......................................................................... 59 15. Conversion Factors and Formulas ............................................................................................... 59
Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table
3-0
Page 4

Automatic sprinklers with adequate water supplies are the best protection for most fire hazards. To obtain the most economical system for any location, fire protection systems must be designed individually. To design these systems, knowledge of water flow in pipes is needed. This data sheet covers basic theory of water flow in pipes and applications, including water supply testing and sprinkler system hydraulics. September 2006. Minor editorial changes were made for this revision. May 2006. Data Sheet 3-0 was updated to remove any reference to Data Sheet 2-76, which was made obsolete. THEORY OF WATER FLOW IN PIPES Bernoullis Theorem and Applications Problems of water flow in pipes are usually solved by procedures based on Bernoullis theorem, which states that in steady flow, without friction, the sum of velocity head, pressure head, and elevation head is constant for a particle throughout its course. This theorem can be expressed by the following equation: (v2/2g) + (p/w) + z = H (equation 1)
where v = velocity, ft/s (m/s) g = acceleration of gravity = 32.2 ft/s2 (9.81 m/s2) p = pressure, lb/ft2 (Pa) w = weight of water per unit volume = 62.4 lb/ft3 (9810 N/m3) z = elevation head (or potential head), distance above an assumed datum, ft (m) H = total head of water, ft (m) In equation 1 the terms v2/2g and p/w express velocity head and pressure head, respectively. Velocity head = hv = v2/2g or v = 2ghv Pressure head = hp = p/w or p = whp (equation 2) (equation 3)
Because total head (H) is constant, a change in velocity results in the conversion of velocity head to pressure head, or vice versa. For a pipeline flowing full between points a and b, Bernoullis theorem can be modified to include friction, as follows: hab (va2/2g) + (pa/w) + za = (vb2/2g) + (pb/w) + zb + (equation 4)
where hab is the total dynamic head lost between points a and b. Figure 1 illustrates the relationships between various factors in typical piping. Heads are indicated by heights to which water rises in the vertical tubing. Velocity magnitudes and directions are indicated by arrows. At each location, B, C, D, E, and F, pressure head, hpF, hpB, etc., is a measure of the potential energy of water in the pipe; velocity head, hvF, etc., is a measure of the kinetic energy of the water; and the sum, hpF + hvF, etc., total head, is a measure of the total energy of the water. Flow rate in a pipeline or discharge through an orifice can be expressed in terms of velocity and crosssectional area of the stream: Q = Av or v = Q/A where Q = flow rate, ft /s (m /s) A = cross-sectional stream area, ft2 (m2) v = average water velocity, ft/s (m/s) In Figure 1(a), QF = QB because no water flows out of the pipe (once stabilized in the vertical tubing). Therefore, from equation 5, AFvF = ABvB, and since the pipe does not change size, AF = AB, so that vF = vB. This means kinetic energy is the same at F and B, and hvF = hvB. Pressure head decreases linearly between F and B. The difference, hpF - hpB, is friction loss or the total dynamic head lost between F and B. (See equation 4.) The loss rate per unit length is (hpF - hpB)/L, where L is length between F and B. In Figure 1(b), water is flowing out of the pipe at B, and the fitting between D and E reduces cross-section area. Flow at F is the sum of flows at B and C, QF = QB + QC. Since QB > 0, QC must be less than QF, (ACvC < AFvF). Cross section areas at F and C are the same, so vC < vF. This means kinetic energy at C is less
3 3
(equation 5)

3-0
Page 5
Fig. 1. Relationships between various hydraulic factors in typical piping
than at F or hvC < hvF. Pressure head, hpC, is shown slightly less than hpF because of friction loss straight through the fitting. Flow out of the pipe at B has substantially increased velocity, because all potential energy is given up to friction or converted to kinetic energy. Thus, total head at F, hpF + hvF, equals hFB + hvB, friction loss between F and B plus velocity head at B. From experimental observation, friction loss between F and B for discharging water is approximately equal to velocity head at F, hFB hvF. Between C and D, QC = QD, vC = vD, hvC = hvD, and hpC drops slightly because of friction, to hpD.
3-0
Page 6

From D to E, QD = QE or AEvE = ADvD (from equation 5) but AE < AD, so vE must be greater than vD. Since velocity increases, kinetic energy does also, and hvE > hvD shows this. Since friction loss in reducers is greater than in pipes, the rate of drop from hpD to hpE is greater than from hpC to hpD. Formulas, such as Q = Av, depending on velocity or velocity head, are not suited to the solution of problems involving flow in closed pipes because simple equipment for direct measurement of velocity or velocity head in closed pipes is not available. Ordinary pressure gauges tapped into pipelines register pressure head only, and therefore give no indication of velocity head or rate of flow. Although suitable devices and procedures for direct pipeline gauging have been developed for studies of waterworks distribution systems, penstocks, etc., these methods are seldom employed in private fire protection practice. Hazen-Williams Formula For fire protection and waterworks, the most widely accepted pipeflow formula was developed by G. S. Williams and Allen Hazen. In conventional fire protection units, the Hazen-Williams formula is: p = cQ1.85/C1.85 d4.87 (equation 6)
where p = loss per unit length, psi/ft (b/m) (kPa/m) c = constant = 4.52 (6.06 105) (6.06 107) (Use 6.06 105 for p in bars and 6.06 107 for p in kPa.) Q = flow rate, gpm (l/min) C = Hazen-Williams pipe coefficient d = internal pipe diameter, in. (mm) Other formulas, developed by Chezy, Darcy, Weisbach, Fanning, Reynolds, and others, are sometimes used. Head is converted to pressure by the relation: 1 ft = 0.433 psi (1 m = 0.098bar) (9.8 kPa) Under this relation, terminology changes as follows: Total head corresponds to total pressure. Velocity head corresponds to velocity pressure. Elevation head corresponds to elevation pressure. Pressure head corresponds to normal pressure. Friction loss tables in Data Sheet 2-89 give values of p for varying flow rates, pipe coefficients, and pipe diameters. Factors for changing from one pipe coefficient to another are also given. Suggested values of C for various kinds of pipe used underground are given in Table 1. The rate of change in the value of C with age in unlined cast-iron pipe depends on corrosive activity of the water. Saturation Index is a commonly used measure of a specific water s corrosive quality and should be available from the water utility in the area. It establishes three categories: positive, zero, and negative, corresponding to mildly, moderately, and severly corrosive waters. Data Sheet 2-81 contains a more extensive discussion of the corrosive quality of water.

3-0
Page 7
Table 1. Hazen-Williams Pipe Coefficients For Underground Use Kind of Pipe Cast-iron, unlined 10 years old 15 years old 20 years old 30 years old 50 years old Cast-iron, unlined, new Cast-iron, cement-lined Cast-iron, bitumastic-enamel-lined Cement-asbestos Approved plastic-lined steel Approved glass-fiber-reinforced plastic Approved PVC Mild 105 100 95 85 75 Water Corrosiveness Moderate 90 75 65 55 50 120 140 140 140 145* 160* 150* Severe 75 60 55 45 40
*If using the Hazen-Williams formula, use these coefficients. If using nominal pipe size and Table 1 from Data Sheet 2-89, use the following artificial C coefficients to compensate for differences in internal diameter between these pipes and Schedule 40 steel pipe:FM Approved plastic-lined steel 165; FM Approved glass-fiber-reinforced plastic 180; FM Approved PVC 165.
Fire protection pipe, normally without flow, is expected to deteriorate less rapidly than pipe subject to continuous or intermittent draft. Cement-lined, bitumastic-enamel-lined, or cement asbestos pipe is relatively smooth with little or no reduction in carrying capacity over a reasonable period of time. Unlined steel pipe exposed to water for various time periods has a wide range of C values. To allow for quick early deterioration under field conditions, therefore, use approximate C values from Table 2 for sprinkler systems.
Table 2. Hazen-Williams Pipe Coefficients For Sprinkler System Use C Value* System Type Hydraulic Design Pipe Schedule Steel 120 100 150 150 Copper** * Use indicated values unless measurements indicate other values should be used. ** Use indicated values with actual pipe size in Hazen-Williams formula. Use C = 130 with nominal pipe size to compensate for differences in internal diameter between copper tube and Schedule 40 steel pipe. Kind of Pipe Steel Copper** Wet 120 150 Dry 100 150
Pressure Loss at Fittings In typical public water or fire protection pipe, losses arising from changes in flow direction and changes in velocity are called loss due to fittings. Such losses are proportional to velocity head (v2/2g) and can be equated to losses in a length of straight pipe. Data Sheet 2-89 gives equivalent pipe lengths for various fittings. In underground pipe computations, fitting loss is generally a small portion of total loss and is not normally considered. In sprinkler system computations, fitting loss is normally considered. Experiments have shown that pressure loss due to fittings occurs primarily downstream from the fittings and involves cavitation and turbulence. Thus, tees have larger losses than short radius elbows, which have larger losses than long radius elbows, etc. Discharge from Nozzles From equations 2 and 5, it follows that Q = A2ghv (equation 7)
In a jet discharging from a nozzle, all the available head (velocity head plus pressure head) is converted to velocity head, which can be measured with a Pitot gauge. When velocity head and nozzle diameter are known, theoretical discharge can be computed from equation 7.
3-0
Page 8

Discharge Coefficient Actual discharge is less than computed discharge because velocity is not uniform over the cross section of the stream. Therefore, a correction factor, or discharge coefficient, is needed in order to use equation 7. Thus Q = cA2ghv Table 3 gives typical discharge coefficients.
Table 3. Typical Discharge Coefficients Type of Orifice Hydrant butt, smooth, well-rounded outlet* Hydrant butt, square outlet* Hydrant butt, inset outlet* Smooth Underwriter nozzles Deluge nozzles Open pipe, smooth and well-rounded, at least 10 diameters long Open pipe, burred opening or less than 10 diameters long
*See Fig 2.
(equation 8)
c 0.80 0.70 0.60 0.97 0.99 0.90 0.80
Discharge coefficients used in calculations of flow from hydrants depend upon the character of the hydrant
Fig. 2. Typical hydrant butts
outlet. Figure 2 shows three general types of hydrant outlets.

3-0
Page 9
Table 4. Theoretical Discharge Through Circular Orifices, U.S. Gallons Per Minute (l/min). Computed For Discharge Coefficient of 1.00 (Seldom Reached In Practice). Reduce the Given Discharge By Multiplying By A Coefficient Suited to the Particular Opening Used. See Table 3.
3-0
Page 10


3-0
Page 11
3-0
Page 12

*For pressure in bars, multiply by 0.01.

3-0
Page 13
Hydrants with orifice plates or valved butts are not in any of the above categories. Use hoses with nozzles to determine flow. In convenient fire protection units, equation 8 may be rewritten Q = acd2Pv (equation 9)
where Q = rate of flow, gpm (l/min) a = constant = 29.8 (0.666) (0.0666) (Use 0.666 for Pv in bars and 0.0666 for Pv in kPa.) c = discharge coefficient d = orifice diameter, in. (mm) Pv = velocity pressure or Pitot pressure, psi (bar) (kPa) Theoretical Discharge Table 4 gives theoretical discharges in gpm (l/min) for various orifice diameters and velocity (Pitot) pressures when c = 1.00. To obtain actual discharge, find theoretical discharge corresponding to known Pv and d, and multiply by appropriate coefficient from Table 3. Nozzle K Factor Equation 8 can also be written in the form Q = KPv where K = acd2 K is theoretically constant for a given orifice. If nozzle K is known and Pv is measured, discharge in gpm (l/min) is found from equation 10. Nominal values of K for various nozzles are given in Table 5. Sprinkler K is calculated from a variation of equation 10: Q = KPn (equation 11) (equation 10)
where Pn is measured upstream from the sprinkler in a pipe or reservoir. When Pn is measured in a reservoir, Pv is essentially zero, so we use Pn = Pt and write: Q = KPt (equation12)
3-0
Page 14

Table 5. Values of K For Various Discharge Orifices. Type of Orifice Sprinkler Sprinkler Sprinkler Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Nozzle Open pipe, smooth and well-rounded (c = 0.85) FM nozzle (c = 0.86) (c = 0.87) (c = 0.88) (c = 0.89) Hydrant butt (c = 0.80) Hydrant butt (c = 0.80) Hydrant butt (c = 0.80) Hydrant butt (c = 0.80) Hydrant butt (c = 0.80) Nominal Size in. (mm) 1732 ( 14) 1 2 ( 13) 38 ( 10) 12 ( 13) 78 ( 22) ( 25) 1 ( 27) 1116 ( 29) 118 ( 30) 1316 ( 32) 114 ( 33) 1516 ( 35) 138 ( 37) 1716 ( 38) 112 ( 40) 1916 ( 41) 158 ( 43) 11116 ( 44) 134 ( 46) 11316 ( 48) 178 ( 49) 11516 2 ( 51) 2 ( 51) ( 57) 214 214 ( 57) 214 ( 57) 214 ( 57) ( 60) 2 3 8 212 ( 64) 258 ( 67) 4 (102) 412 (114) K English 8.0 5.6 2.8 7.2 22.2 29.1 32.8 36.8 41.0 45.4 50.1 54.9 60.0 65.4 70.9 76.8 82.8 89.0 95.5 102 109 116 101 130 131 133 134 134 149 164 381 484 (Metric)* (115) ( 81) ( 41) ( 104) ( 320) ( 420) ( 470) ( 530) ( 590) ( 650) ( 720) ( 790) ( 860) ( 940) (1020) (1110) (1190) (1280) (1380) (1470) (1570) (1670) (1460) (1860) (1890) (1910) (1940) (1940) (2150) (2370) (5500) (6960)
* Use value shown when pressure is in bars and 110 value when pressure is in kPa.
Acceptable K values for FM Approved 12-in. (13-mm) sprinklers range between 5.3 (76) (7.6) and 5.8 (84) (8.4). For most purposes, use K = 5.6 (81 bar) (8.1 kPa). Combined Sprinkler Discharge and Pipe Flow Discharge from single sprinklers is calculated from equation 12. Flow in a pipe is available from equation 6, p = cQ1.85/C1.85d4.87. For a specific piece of pipe, c, C and d are constant, so we can solve for Q to get Q = Kp0.54 where K = Cd
2.63
(equation 13) /c
0.54
is also constant.
If several sprinklers in a constant elevation system are operating, calculations use both p0.5 and p0.54. A group of sprinklers with piping can thus be represented by Q = Kpr, where K is constant and 0.5 < r < 0.54. Pinpointing r within this range is impractical. When friction loss is relatively small, as with large pipe, short branch lines, or small groups of sprinklers, r is close to 0.5. Conversely, when friction loss is relatively large, r is close to 0.54. For a fixed number of operating sprinklers, total pressure throughout a sprinkler system at constant elevation is directly proportional to total pressure at any given point. If total pressure at one point is changed from Pt1 to Pt2, total pressure at every other point with the same elevation is multiplied by Pt2/Pt1. From Q = Kpr, we get Q1 = Kp1r and Q2 = Kp2r which vield Q2/Q1 = (p2/p1)r. Thus to change from p1 to p2, find Q2 from: Q2/Q1 = or
p2/p1
, for r = 0.5
(equation 14) (equation 15)
Q2/Q1 = p2/p1)0.54 , for r = 0.54

3-0
Page 15
Although technically this method applies only when the sprinklers are at exactly the same elevation, it is satisfactory in single-story sections of sprinkler systems unless the ceiling is steeply pitched. TESTING WATER SUPPLIES The performance of public and private water supplies for fire protection service is best determined from periodic flow tests. Such tests are particularly important for the design of new fire protection systems. Site selection and timing must be carefully chosen to reveal conditions that might exist at the time of the fire. Testing procedures are based on principles of water flow in pipes and discharge from orifices. Available fire flow must be determined even if known to be less than demand or even if water supply is less than reliable. Purposes There are several reasons for flow testing water supplies: 1. Fire flow available to a given area can be compared with demands specified in existing standards. 2. Fire pumps and drivers are subject to numerous malfunctions that can be discovered and corrected promptly. 3. Closed valves and other obstructions become apparent. 4. Water supply deterioration can be detected. 5. The ability of plant personnel to use the fire protection system can be assessed. 6. Weakened underground pipe sometimes ruptures under fire pump or fire department pressures. It is far better that this condition be discovered during a test than during a fire. Hazards Ensure qualified personnel direct flow tests because personal injury and property damage can result from improperly conducted tests. Some of the more common problems that arise are: 1. Improperly secured nozzles can work loose and injure personnel. 2. Excessive flows may draw vacuums in high buildings and in hilly country, possibly contaminating water supplies, damaging boilers, and interrupting industrial processes. Avoid reducing pressure in public mains below 20 psi (1.38 bar) (138 kPa). 3. Local flooding at low spots, such as truck docks, basements, pits, tunnels, etc., is possible. 4. Water damage can occur to storage if it is located outdoors or in low-lying buildings where water can flow through doorways. 5. High voltage electrical equipment can be shorted by solid hose streams, with danger to personnel. Note that spray from fire protection equipment avoids this problem. 6. Rapid valve operation results in water hammer, which can rupture piping and damage sprinkler equipment. Flow and Pressure Measurements Flow tests are usually made by discharging water through one or more hydrant or nozzle outlets. At the same time readings are taken with pressure gauges at sprinkler risers, nonflowing hydrants, or other direct connections to supply pipes (Figs. 3 and 4). Flow rate is determined by measuring the velocity pressure of each stream with a Pitot gauge (Fig. 5). The Pitot orifice is held firmly in the center of the jet. The knife edge is held directly against the nozzle end with the blade at right angles to the nozzle axis. Velocity pressure is read directly on the gauge in psi (bar) (kPa) (Fig. 6). The corresponding discharge in gpm (l/min) is then obtained from tables or equations 9 or 10. Note discharge coefficients and exact diameter for all nozzles and use reliable pressure gauges. Pitot readings under 10 psi (0.69bar) (69 kPa) are questionable. Avoid them whenever practical by reducing the size or number of orifices. Pressures observed at sprinkler risers, nonflowing hydrants, or other direct connections to the supply pipe are called static if there is no test flow. Pressures observed with test streams flowing are called residual. When testing, always read both static and residual pressures. Handle gauges carefully and calibrate them yearly
3-0
Page 16

Fig. 3. Water flow devices
Fig. 4. Typical pressure gauge locations
Fig. 5. Pitot tube with gauge and air chamber
or whenever they are damaged or otherwise suspect. Calibrate gauges at several pressures, since deviations can vary from one pressure to another, and affix records of deviations to gauges for ready reference during water tests.

3-0
Page 17
Fig. 6. Taking pitot readings at hose nozzle
Test Planning Flow tests must indicate the ability of water supplies to meet water demands in the demand area. Before testing, plan a course of action, indicating where to make flows, where to read pressures, and which valve operating sequence to use. Occasionally, physical conditions require changes in plans. Such changes are usually simple. For example, a given hydrant can be dangerous to use without hoses because of land slope or high voltage electrical equipment, while an adjacent hydrant is relatively safe. When accuracy is needed, plan to use hose streams rather than hydrant butts. Accuracy might be needed if costly recommendations are being considered. Check for local regulations that may affect planning. For example, if may be necessary to notify the public water service before flow testing private systems. For public water systems and large plants, condense water supply information onto a single small sheet of paper in diagram form (Fig. 7). Show all sources and connections as well as major loops and valves.
Fig. 7. Simplified, freehand line diagram of water supplies
3-0
Page 18

When planning flow tests and when testing, keep the following fundamentals in mind: 1. Make a flow equal to or greater than demand near the demand area. 2. Keep fire protection in service at all times. 3. Flow each major loop and each source or connection (except pressure tanks). 4. Reopen all valves and flow all sources together. Private Water Systems Water supplies range from simple to complex, and planning varies accordingly. Figure 8 illustrates a simple supply. Flow the hydrant and read static and residual pressures at the riser below the alarm check valve. Flow sufficient water to meet the combined sprinkler and hose stream demand.
Fig. 8. Single-source water supply
If there is a single source with an underground loop (Fig. 9), plan as follows: 1. Select hydrant and riser nearer the demand area and read static pressure at the riser. 2. Close valve A and flow hydrant (south loop), reading residual pressure at the riser. 3. Open valve A. 4. Close valve B and flow hydrant (north loop), reading residual pressure at the riser. 5. Open valve B and flow hydrant (both loops) with flow rate sufficient to prove that water supply satisfies water demand. Read residual pressure at the riser. 6. Read static pressure.
Fig. 9. Single source with underground loop
If there are two sources or two connections and an underground loop (Fig. 10), plan as follows: 1. Select hydrant and riser nearest the demand area. 2. Close valves B and C and read static pressure at the riser. 3. Flow hydrant (gravity tank, south loop), reading residual pressure at the riser. 4. Open valve B.

3-0
Page 19
5. Close valve A and flow hydrant (gravity tank, north loop), reading residual pressure at the riser. 6. Read static pressure at the riser and open valve C. 7. Close valve B and flow hydrant (public), reading static and residual pressures at the riser. 8. Open valves A and B and flow hydrant (both sources, both loops) with flow rate sufficient to prove that water supply satisfies water demand. Read residual pressure at the riser. 9. Read static pressure at the riser.
Fig. 10. Two sources with underground loop
Tests of more complex supplies are planned similarly, keeping in mind the fundamentals. Not only is it necessary to select a flow to prove the ability of the supply to meet the demand, but total flow must draw from all sources. A test that does not draw from all sources of a multi-source system, when feasible, lacks completeness. For example, typical gravity tank pressure is 50 psi (3.45 bar) (345 kPa) and typical public pressure is 80 psi (5.52 bar) (552 kPa). If the test result with all valves open is reported as 80 psi (5.52 bar) (552 kPa) static, 1000 gpm (3790 l/min), 51 psi (3.52 bar) (352 kPa) residual, the water supply is seriously misrepresented because there is no flow from the gravity tank. Gravity tanks can increase flow dramatically with small pressure losses, depending on relative locations of gravity tank, public connections, and flowing hydrant. Public Water Systems To test public water supplies at proposed facility sites, visit the local water service to diagram appropriate sections of the water map and to arrange for tests. Previous test results are sometimes on file and may be useful for planning tests. Plan a course of action before testing. Generally, flow from a single hydrant indicates whether or not the supply is adequate for the demand. In case of high demands, flow two or more hydrants. Select hydrants to give significant results. Ensure pressure readings and flows are as close to plant connection points as practical. Figure 11 shows a typical situation. 1. Example (Fig. 11) If all pipes are the same size, flow either hydrant, with pressure readings at the other, since most of the friction loss occurs in the part of pipe A that is out of the picture. 2. Example (Fig. 11) If pipes B, C, and D are small and A is large, test flow at either hydrant has little effect on pressure in A. With test flow at either hydrant, the nonflowing hydrant is about halfway from A to the flowing hydrant via pipe that is carrying part of the test flow. Therefore, pressure at the nonflowing hydrant reflects about half the change near the flowing hydrant. Such pressure readings may be acceptable, but it is preferable to flow the west hydrant and read pressures north of it. 3. Example (Fig. 11) If pipe C is very small, pressure at the nonflowing hydrant is much closer to pressure in pipe A than to pressure at the flowing hydrant. Read pressures at nearby hose bibs or flow the west hydrant and read pressures north of it. Make sure hose bibs are supplied from the pipe under test and that pressure is unaffected by domestic draft or pressure regulator during measurements.
3-0
Page 20

Fig. 11. Typical public water system near a plant site
4. Example (Fig. 11) If pipes C and D are small and A and B are large, test flow at west hydrant with pressure readings at east hydrant represents available supply at the west side of the facility site adequately. Test flow at east hydrant with pressure readings at a nearby hose bib represents available supply at the east side of the facility site. Make sure hose bibs are supplied from the pipe under test and that pressure is unaffected by domestic draft or pressure regulator during measurements. Test flow at the east hydrant with pressure readings at the west hydrant is not representative, because test flow must pass through small pipes C and D where pressure loss per unit length is relatively high. West hydrant pressures are thus much higher than pressures near the flowing hydrant. Similarly, high flows available at the west hydrant do not represent water supply at the east hydrant. 5. Example (Fig. 12) Another situation that occurs sufficiently often to merit special attention is illustrated in Figure 12. Since the two dead-end pipes are part of a single water system, flow at A reduces flow available at B and vice-versa. Therefore, to determine total flow available to the facility site, flow hydrants A and B simultaneously and read residual pressures as near as possible to both A and B. Then, in case the facility is supplied from only one pipe, flow A and B separately and read residual pressures upstream from each flowing hydrant.

3-0
Page 21
Fig. 12. Two dead-end pipes from a public water system near a facility site
Public water systems often require special treatment. Some items needing attention follow: 1. If systems are direct or intermittent pumping, determine facilities in operation during tests. 2. If there is elevated storage, test with pumps shut down, if possible. 3. If special fire pressure is available, test if possible, and determine (a) notification required, (b) time required to obtain service, and (c) facilities in operation. 4. Check for pressure regulating valves. 5. Check for variations in operating procedures from day to night or from summer to winter. Test Arrangement When using hose streams lash nozzles to substantial supports such as posts, trees, or deeply driven stakes. Do not hold nozzles by hand. When hoses are not available and waterflow from hydrants or pump headers must be redirected to avoid damage, use elbows and piping with adapters from hose thread to pipe thread. Ensure pipe downstream from elbows is at least ten diameters long to obtain smooth flow. If Pitot pressures are erratic or higher at orifice edges than at centers, pipes must be longer than ten diameters to smooth flows. Read static and residual pressures at locations where they respond to the normal pressure existing in pipes through which test water is flowing. Ensure there is no check valve or shut valve between the column of flowing water and the point where residual pressure is read. This precaution reduces the probability of reading trapped overnight high pressures or of finding no pressure drop when flowing from a typical public water supply. Use FM Global valve supervision procedures. Make sure tanks are full and that valves controlling tank fill lines are shut before starting test. Ensure nonflowing butts on the flowing hydrant are tightly capped. Ensure all valves closed during water tests are shut tightly. Rsidual pressure readings measure normal pressure where the column of still water starting in the gauge meets the column of flowing water, adjusted for elevation difference. This location should be near the demand area. Figure 13 shows how test flow and residual pressure can be taken near the demand area with meaningless results because the junction of the two water columns is much further away. Read residual pressure at B with test flow at C so the column of still water starting in the gauge meets the column of flowing water a short distance from the demand area. If residual pressure is read at A with test flow at C, the column of still water starting in the gauge meets the column of flowing water at D, indicating available water supply at the city connection and defeating the purpose of making a flow test near the point of highest demand. If both sprinkler systems are hydraulically designed, read residual pressures at both A and B to obtain the best water supply information for calculating both systems.
3-0
Page 22

Fig. 13. Single source with partial loop
Test Procedure The last step before testing is to make sure any central station, fire service, or proprietary supervisory services have been notified that a flow test is being conducted. Then proceed with the test as follows: 1. Close valves to isolate the desired source and underground loop. Use FM Global valve supervision procedures. 2. Read static pressure and bleed off any trapped high pressure. 3. Inspect area for possible damage from discharge. 4. If flowing an open hydrant butt, determine orifice diameter and coefficient. (See Table 3.) 5. Open hydrant valve wide or open header valves to create desired flow. 6. Make sure discharge is doing no damage. 7. Allow ample time for residual pressure to stabilize. It often takes several minutes for pressure to stop changing. 8. Read pressures at flowing orifices with Pitot gauge and note residual pressure. 9. Close hydrant or header valves slowly. 10. Read static pressure. 11. Convert Pitot readings to flows using Table 4 or equation 9 and plot, as illustrated in Figure 16. 12. Compare with anticipated results to determine if test setup is correct or if problems exist. 13. Correct problems and replan test sequence if necessary. 14. Proceed with test sequence and record results. When testing pumps, achieve desired flow by adding or removing nozzle tips to change orifice size, changing number of streams, or throttling valves. To alter flow from hydrants, change orifice size or number of streams. Do not throttle hydrant valves.

3-0
Page 23
For most tests, readings at two flow rates with all valves open are recommended to provide a check on results, and if public water is a supply, to help assess the effect of normal public (non-test) flow or automatic starting pumps. 15. Compare with previous tests or with anticipated results to ensure that no further testing is needed. Dismantle test setup. 16. Make sure all valves are locked or sealed open and follow FM Global valve supervision procedures. 17. Refill gravity and pump suction tanks and make sure fire pumps are left on automatic. 18. Complete FM Global valve supervision procedures. 19. Notify supervisory service that test is complete and ascertain that valve tamper signals and flow signals were transmitted properly. Determining Hydraulic Gradients If fire flow tests show that yield from a water supply is unaccountably low, identify the causes by determining the hydraulic gradient. Low yields can be caused by partly closed valves, excessively tuberculated pipes, or obstructions. Pipe size errors on plans can make yields appear abnormal. Hydraulic gradients isolate locations of abnormally high losses, which cause low yields. The hydraulic gradient is a profile of residual pressure. It assumes a uniform rate of flow and simultaneous residual pressure readings at various points along the pipe. In practice, moving one gauge progressively from test point to test point while test flow is maintained yields maximum consistency in pressure readings. If there are tees, crosses, bends, valves, or meters in pipe being tested, obtain the pressure loss in these devices from Data Sheet 2-89 and deduct it from the observed drop in pressure before calculating coefficients for given pipe. Relations between pressure and elevation in pipe having uniform flow are illustrated in Figure 14.
Fig. 14. Relations between water pressure and elevation in pipe having uniform flow
3-0
Page 24

Tests of private mains usually are made on much shorter runs of pipe than tests of public mains. In order to reduce the number of tests, choose mains that are typical of the age and probable conditon of the system. Where obstructions are suspected, investigate that portion of the yard system. Ensure heavy flow is induced through the test section to drop pressure from station to station as much as possible, reducing effects of fluctuating pressure or small inaccuracies in gauge readings. Gauge readings without test flow in fire protection mains supplied from gravity tanks or private reservoirs indicate true static pressures, provided there is no facility-use draft or significant leakage. If gauges are attached to hydrants connected to level water mains, static readings are approximately the same at each station. If the pipe profile is as shown in Figure 14, or if gauges are attached to sprinkler risers at various elevations, readings differ from station to station. Under these circumstances, static readings constitute a measure of gauge elevation at each station; high readings indicate a low spot, as at point A, and low readings a high point, as at B. Because of the flow normally prevailing in public mains, observed static pressures (that is, pressure readings without test flow) are actually residual pressures and trace combinations of elevation and normal gradient instead of elevation alone. Determine relative gauge elevations from topographic maps, municipal survey data, or (with caution) simply by observation. When elevation figures are not available, use static gauge readings. These readings are most reliable between midnight and 4 a.m., when normal draft is minimal. Observed static pressures are then more representative of true elevations than at other times. Example of Gradient Test Given: Gradient test data for pipeline shown in Figure 15.
Table 6. Data For Example of Gradient Test.
1 2 3 4 Static Pressure, psi (bar) (kPa) 81 (5.58)(558) 910 (277) 500 (152) 435 (133) 910 (277) 8 (203) 8 (203) 6 (152) 8 (203) 5 Residual Pressure, psi (bar) (kPa) 71 (4.9)(490) 6 7 Loss Station to Station, psi (bar) (kPa) 8 9 10 11 Gauge Elevation, psi (bar) (kPa) 12 Gradient Elevation, psi (bar) (kPa) Given Proposed 75 (5.17) (517) 65 (4.48) (448) 59.5 (4.1) (410)
Station A
Pipe Length, ft (m)
Diameter in.(mm)
Total Loss psi (bar) (kPa) 10 (0.69) ( 69)
Loss per ft (m) psi (mbar) (kPa)
fc
4 75 (0.28) (28) (5.17)(517) 10 (0.69)( 69) 0.011 ( 2.49)(0.249) 0.04 (9.09)(0.909) 0.046 (10.37)(1.037) 0.011 ( 2.49)(0.249) 1.0 100 10 65 (0.69) (69) (4.48)(448) 2 (0.14) (14) 0 5 (0.34)(34) 45 (3.1) (310)
75 55 20 (5.17)(517) (3.79)(379) (1.38)(138)
C D E
83 43 40 20 (5.72)(572) (2.96)(296) (2.76)(276) (1.38)(138) 85 25 60 20 (5.86)(586) (1.72)(172) (4.14)(414) (1.38)(138) 80 (5.52)(552) 10 70 (0.69)( 69) (4.83)(483) 10 (0.69)( 69)
3.6 1.0 1.0
50 100 100
25 54.7 (1.72)(172) (3.77)(377) 15 (1.03)(103) 44.7 (3.08) (308)
Note: Column numbers (1) to (12) relate to explanation in text below.
Required: 1. Calculate pipe coefficients. 2. Plot hydraulic gradient and pipe profile. 3. Recommend improvement. Procedure: 1. Tabulate gradient test data in Table 6. 2. Calculate unkowns and complete tabulation in Table 6. 3. Plot gradient and profile.

3-0
Page 25
Fig. 15. Pipeline for which hydraulic gradient is computed (upper) and plot of profile and hydraulic gradient to the pipe line (lower)
Explanation of Table 6: Columns 1 to 5, inclusive, contain data from Figure 15. Column 6. Static - residual pressure = total loss Column 7. Difference in total loss, station to station
3-0
Page 26

Column 8. Loss per foot (per meter): 10/910 = 0.011 psi (A to B) (0.69/277 = 0.00249 bar [0.249 kPa]) 20/500 = 0.04 psi (B to C) (1.38/152 = 0.00909 bar [0.909 kPa]) Column 9. fc = Column 8 Loss/ft (m) (from Data Sheet 2-89, Table 9)
For 8-in. (203-mm) pipe (0.011 psi/ft [2.49 mbar/m] [0.249 kPa/m]) 0.011 psi per ft/0.011 psi per ft = 1(A to B) (2.49 mbar per m/2.49 mbar per m = 1) 0.040 psi per ft/0.011 psi per ft = 3.6 (B to C) (9.09 mbar per m/2.49 mbar per m = 3.6) For 6-in. (152-mm) pipe (0.046 psi/ft [10.37 mbar/m] [1.037 kPa/m]) 0.046 psi per ft/0.046 psi per ft = 1(C to D) (10.37 mbar per m/10.37 mbar per m = 1) Column 10. Values of C from Data Sheet 2-89, Table 14, corresponding to fc Column 11. Assume zero elevation at station with the maximum static pressure, in this case Station D, where static = 85 psi (5.86 bar) (586 kPa). Elevations of other stations are found by subtracting respective static readings from 85 psi (5.86 bar) (586 kPa). Thus: A = 85 - 81 = 4 psi (5.86 - 5.59 = 0.28 bar [28 kPa]). B = 85 - 75 = 10 psi (5.86 - 5.17 = 0.69 bar [69kPa]) etc. Column 12. Gradient elevations are obtained by adding residual pressure to gauge elevations. Thus: A = 71 + 4 = 75 psi (4.90 + 0.27 = 5.17 bar [517 kPa]) B = 55 + 10 = 65 psi (3.79 + 0.69 = 4.48 bar [448 kPa]) etc. Differences in gradient elevation equal losses between corresponding stations. Recommendations: 1. Clean and line 8-in. (203-mm) pipe BC, restoring C to 100. 2. Replace 6-in. (152-mm) pipe CD with 8-in. (203-mm) cement-lined pipe. Show expected gradient with recommendations completed (Fig. 15). Loss B to C: 1 500 0.011 = 5.5 psi (1 152 0.00249 = 0.38 bar [38 kPa]) Gradient elevation at C = 65 - 5.5 = 59.5 psi (4.48 - 0.38 = 4.1 bar [410 kPa]) Loss C to D: 0.537 435 0.011 = 2.8 psi (0.537 133 0.00249 = 0.18 bar [18 kPa]) Gradient elevation at D = 59.5 - 2.8 = 56.7 psi (4.1 - 0.18 = 3.92 bar [392 kPa]) Gradient elevation at E = 56.7 - 10 = 46.7 psi (3.92 - 0.69 = 3.23 bar [323 kPa]) Special Water Test Techniques and Considerations Water testing is often less straightforward than has been indicated thus far. This section presents some problems that arise occasionally, and their possible solutions. Problem: There are no pressure gauge mounting fittings below check valves on sprinkler risers. Solution: Mount the gauge on the fitting above the check valve and open the 2-in. (51-mm) drain just enough to allow the indicated pressure to drop and stabilize. At flows low enough to avoid adding significantly to the test flow, essentially no pressure is lost in the check valve.

3-0
Page 27
Problem: There are no hydrants, sprinkler risers, or hose bibs within 1000 ft (305 m) of the flowing hydrant. Solution: Mount the gauge on the second butt of the flowing hydrant. Indicated pressure differs from nearby residual pressures by friction loss in hydrant and supply pipe. Losses in hydrants vary considerably, but average from 2 psi (0.14 bar) (14 kPa) at 500 gpm (1890 l/min) to 5 psi (0.34 bar) (34 kPa) at 1000 gpm (3790 l/min).
3-0
Page 28

Problem: Pitot tube is unavailable. Solution: Mount the gauge on the second butt of the flowing hydrant. Indicated pressure equals velocity pressure at flowing butt plus friction loss in butt minus velocity pressure in hydrant barrel. Friction loss in butt approximately equals velocity pressure in hydrant barrel, so indicated pressure approximately equals Pitot pressure. If a hose and nozzle are used on the hydrant, subtract friction loss in the hose from indicated pressure to obtain the approximate Pitot reading. PRESENTING WATER SUPPLY DATA Investigation of water supplies for fire protection may be expedited by using graphs that provide means for the effective presentation of hydraulic conditions and indicate what changes or improvements are needed. There are two kinds of flow curves supply curves and loss curves - that appear as curved lines on linear cross-section paper. The same curves appear as straight lines on semi-exponential paper having linear subdivisions for pressure on the vertical axis and exponential (N1.85) subdivisions for flow on the horizontal axis. This semi-exponential (N1.85) paper fits the Hazen-Williams pressure-drop-flow relation, given in equation 6. Numerical values on the vertical and/or horizontal scales of N1.85 paper may be multiplied or divided by any constant as may be necessary to fit problems onto sheets or to magnify portions of curves to facilitate study. N1.85 paper is particularly useful for the following purposes: 1. Showing supply curves (Fig. 16) 2. Calculating friction-loss curve in a single pipeline (Fig. 17) 3. Combining friction-loss and supply curves to determine yield (Fig. 18) 4. Solving parallel-pipe (loop) systems (Fig. 19 and 20) 5. Calculating yield from combinations of supplies (Fig. 21 and 22) Supply Curves (Fig. 16) Supply curves show the rate of flow available at any residual pressure. They are drawn by plotting flow against residual pressure obtained from water test data or from data based on characteristics of specific supplies. Typical supply curves shown in Figure 16 are obtained by plotting the supply available at discharge outlets or at typical points in a public water system. When plotting a fire-flow curve with supply from a public water system, three or more test points, including static, are desirable to check on accuracy and to avoid errors that might be introduced by extrapolation beyond limits of the test data. In addition, extra pumps may come on line automatically with higher flows. If three test points are used, a curve instead of the usual straight line may appear because the water system, in addition to test flow, is also supplying a normal demand of unknown quantity. True static pressure may be higher than observed static pressure because of friction loss to the test location developed by normal flow. Under such circumstances, the curve drawn between points representing observed conditions will give a reasonably accurate indication of the available fire flow. Friction-Loss Curve for a Single Pipeline (Fig. 17) To plot a friction loss curve for a pipeline of uniform size throughout, determine the losses through the pipe at two different rates of flow; plot these points and connect them with a straight line. Because the loss is zero with no flow, it is customary to use the 0-0 coordinate as one of the two points. Select the other point near the top or right-hand edge of the graph. Example: Plot the friction loss curve for 1,120 ft (342 m) of 8-in. (203-mm) pipe having Hazen-Williams C of 90. Friction loss per foot (per meter) from equation 6 is p = cQ1.85/C1.85d4.87. Data Sheet 2-89 tabulates p and multipliers, fc = (C2/C1)1.85 (Table 14, DS 2-89), to convert from C1 to C2. Total friction loss is Pf = pfcL where L is pipe length. (equation 16)

3-0
Page 29
Fig. 16. Supply curves
1. Assume 1,500 gpm (5680 l/min) flow. 2. From Data Sheet 2-89, Tables 14 and 9, fc = 1.22 for converting C1 = 100 to C2 = 90 p = 0.024 psi/ft (5.4 mbar/m) (0.54 kPa/m) for 8-in. (203-mm) pipe, for assumed flow, and for C = 100. 3. From equation 16, Pf = 0.024 1.22 1120 = 32.8 psi Use 33 psi. (0.0054 1.22 342 = 2.25 bar [225 kPa]. Use 2.25 bar [225 kPa].) 4. Plot a point at 33 psi (2.25 bar) (225 kPa) and 1500 gpm (5680 l/min). Connect it to the 0-0 point with a straight line (Fig. 17). This is the required friction loss curve. It illustrates loss in the given pipe at any flow, and conversely, flow through the pipe at any loss.
3-0
Page 30

Fig. 17. Friction loss curve for a single pipeline
To plot the friction loss curve for a single pipeline made up of various lengths and diameters of pipe, proceed as follows: 1. Determine friction loss for each element of the pipeline as described above, using the same rate of flow for each element. 2. Add friction losses for individual elements. 3. Plot the sum of friction losses at the selected flow. 4. Draw a line connecting this point with the 0-0 point. This is the friction loss curve for the entire pipeline. Curve Combination - Methods and Meaning The following statements summarize relationships between operations with curves and actual situations: 1. Two curves can be combined vertically or horizontally. 2. Vertical combination is accomplished by selecting a flow and adding or subtracting pressures. 3. Horizontal combination is accomplished by selecting a pressure and adding or subtracting flows. 4. Vertical combination or flow selection implies a series (end-to-end) arrangement of pipes, sources, and elevations, since inflow equals outflow. a. Elevation is normally interpreted as fixed pressure (horizontal line) in series with the source (or in series with the pipe system) and is independent of flow. It is subtracted from the source curve if higher than the source, and added if lower. b. Pump curves are not straight lines. Generally, combined source curves are not straight lines. Pipe curves are straight lines. Elevation curves combined with pipe curves are straight lines. 5. Horizontal combination or pressure selection implies a parallel (side-by-side) arrangement of pipes, sources, and hoses, since pressure difference across the system is the same by any route. a. Hose demand is normally interpreted as a fixed flow (vertical line) in parallel with the source and independent of pressure. Hose demand is then subtracted from the source curve at several selected pressures. Combined source curve and hose demand is not a straight line.

3-0
Page 31
Fig. 18. Combination of supply and friction loss curves to determine yield
Calculation of Yield from a Single Supply Combination of friction loss and supply curves in pipe systems (Fig. 18). In Figure 18, supply and friction loss curves are combined vertically to indicate yield at a point some distance from the source, as at B in the example shown. A test made at point A on the public main indicates that flow of 950 gpm (3600 l/min) reduces static pressure of 100 psi (6.89 bar) (689 kPa) to residual pressure of 90 psi (6.21 bar) (621 kPa). Static pressure at Point B is 80 psi (5.52 bar) (552 kPa). The difference in static pressure is due to elevation. A is connected to B by 767 ft (234 m) of 6-in. (152-mm) pipe with a known coefficient of 110. Curve (a) is the supply curve of water available at point A, developed from test data as follows:
3-0
Page 32

Fig. 19. Typical loop in distribution system
1. Plot one point at 100 psi (6.89 bar) (689 kPa) and no flow. 2. Plot one point at 950 gpm (3600 l/min) and 90 psi (6.21 bar) (621 kPa). 3. Connect the two points with a straight line, curve (a). Curve (b) is the friction loss curve in the pipe between A and B, which is developed as for Figure 17. Flow of 1,500 gpm (5680 l/min) is assumed, for which friction loss from equation 16 using Data Sheet 2-89, Tables 9 and 14, is Pf = 0.098 0.838 767 = 63 psi (0.0222 0.838 234 = 4.35 bar [435 kPa]) Curve (c) is the supply curve at B before correction for elevation. It is obtained by subtracting friction loss curve (b) from supply curve (a) as follows: 1. At no flow, static pressure is 100 psi (6.89 bar) (689 kPa) and there is no friction loss. Plot a point at this pressure and flow. 2. At 1,500 gpm (5860 l/min), residual pressure is 77 psi (5.31 bar) (531 kPa) and friction loss is 63 psi (4.35 bar) (435 kPa), a difference of 14 psi (0.97 bar) (97 kPa). Plot a point at 1,500 gpm (5680 l/min) and 14 psi (0.97 bar) (97 kPa). 3. Connect these two points by a straight line, curve (c). Where, as in discharge from a centrifugal pump, supply curve (a) plots as a curve rather than a straight line, friction loss at a number of flow rates must be deducted from the supply curve to develop a true curve representation of yield at B. 4. Draw curve (d) parallel to, and 20 psi (1.38 bar) (138 kPa) below, curve (c). Curve (d) represents supply, or net yield, at B and is obtained by correcting supply curve (c) for elevation difference. Combination of friction loss and supply curves in parallel-pipe (loop) systems (Fig. 19 and 20) Solutions for parallel-pipe (loop) problems are based on two principles: 1. Pressure difference between any two points of a network is the same by any route through the network. 2. Flow toward any point equals flow away from the point. Example: In the example of Figures 19 and 20, it is not possible to make a flow test at the proposed facility site, but a test at point A is available. The problem involves the water supply to expect at point C when pipe B is installed. The test at point A indicates that flow of 1,400 gpm (5,300 l/min) reduces pressure from 70 psi

3-0
Page 33
Fig. 20. Friction loss and supply curves in parallel pipe (loop) systems
(4.83 bar) (483 kPa) static to 60 psi (4.14 bar) (414 kPa) residual. Static pressure at point C is 95 psi (6.55 bar) (655 kPa). The difference in static pressures is due to elevation difference. To obtain the water supply curve at point C: 1. Plot supply curve (d) (water available at point A). 2. Plot friction loss curves (a) and (b) for each side of the loop. This applies the second principle above. Flow into one end of a pipe equals flow out the other end, regardless of size changes or bends. a. Assume 1,500 gpm (5680 l/min) in each side. b. From Data Sheet 2-89, Table 3, PB = 0.059 psi/ft (13.3 mbar/m) (1.33 kPa/m). c. Since Table 3 in Data Sheet 2-89 is generated for C = 140, and proposed pipe has C = 140, fc = 1 and equation 16 yields: PfB = 0.059 1320 = 77.9 psi. (0.0133 403 = 5.36 bar [536 kPa]). d. From Data Sheet 2-89, Table 9, PC = 0.098 psi/ft (22.2 mbar/m) (2.22 kPa/m), also fc = 0.838 for converting C1 = 100 to C2 = 110.
3-0
Page 34

Fig. 21. Calculation of yield from supplies in parallel. Note: Curves (e), (g), and (h) do not show more than 150% of rated pump flow because that flow is the maximum guaranteed for all pumps and all installations.
e. Equation 16 yields: PfC = 0.098 0.838 1260 = 103.5 psi (0.0222 0.838 384 = 7.15 bar [715 kPa]) 3. Plot friction loss curve (c) for the entire loop by adding the flow available through each side at some convenient pressure. This applies both principles. Friction loss is the same across each side. Flow at A equals the sum of flows at B and D, and equals flow at C. a. To add curves, it is easiest to select the highest round-number pressure at which the sum of flows is on the graph paper. At 25 psi (1.72 bar) (172 kPa), 710 gpm (2690 l/min) is available through pipe C, and 810 gpm (3060 l/min) through pipe B. b. Add these quantities and plot a point at 25 psi (1.72 bar) (172 kPa) and 1520 gpm (5750 dm). c. Connect this point to the 0-0 coordinate by a straight line. This is the friction loss curve (c) for the entire loop.

3-0
Page 35
Fig. 22. Calculation of yield from supplies in series
4. Plot supply curve (e) at point C, before elevation correction, by subtracting friction loss through the loop (curve (c)) from the available supply at point A (curve (d)). At 1,400 gpm (5300 l/min), 60 - 21 = 39 psi (4.14 - 1.45 = 2.69 bar [269 kPa].) At 0 gpm (0 l/min), 70 - 0 = 70 psi (4.83 - 0 = 4.83 bar [483 kPa].) 5. Plot supply curve (f) 25 psi (1.72 bar) (172 kPa) above curve (e). This corrects for elevation as indicated by static pressures. Curve (f) is the expected water supply (yield) at point C. Calculation of Yield from Supplies in Parallel (Fig. 21) In the example shown, public water tests at point A indicate that flow of 600 gpm (2270 l/min) reduces static pressure from 105 psi (7.24 bar) (724 kPa) to 75 psi (5.17 bar) (517 kPa) residual. The centrifugal fire pump at B is rated 750 gpm (2840 l/min) at 100 psi (6.89 bar) (6.89 kPa). The connection between A and B is 1,640 ft (500 m) of 8-in. (203-mm) cast iron pipe with C of 70. Points B and C are 35 ft (10.7 m) higher than point A.
3-0
Page 36

Supplies from fire pump and public water connection are additive at any point C between the two sources after proper allowances are made for elevation difference and friction loss between individual sources and point of demand. To plot water supply curve (yield) at point C from fire pump and public water connection in simultaneous operation: 1. Plot supply curve (a), public water available at point A, by the method given for Figure 16. 2. Plot friction loss curve (b), for the pipe between points A and C, by the method given for Figure 17. Using equation 16 Pf = 0.011 1.94 655 = 14 psi (0.0025 1.94 200 = 0.97 bar [97 kPa]) at 1,000 gpm (3790 l/min) 3. Plot supply curve (c) at point C from public water at point A before elevation correction by subtracting friction loss indicated by curve (b) from curve (a). This is vertical combination of a series arrangement. 4. Correct curve (c) for elevation difference by drawing curve (d) 15 psi (1.03 bar) (103 kPa) below curve (c) to represent net yield at point C from point A (vertical combination). 5. Plot pump characteristic curve (e). 6. Plot friction loss curve (f) for pipe between points B and C. Using equation 16 Pf = 0.011 1.94 985 = 21 psi (0.0025 1.94 300 = 1.45 bar [145 kPa]) at 1,000 gpm (3790 l/min). 7. Plot supply curve (g), showing yield at point C from the fire pump at point B, by subtracting friction loss indicated by curve (f) from curve (e) (vertical combination). 8. Add curves (d) and (g) to form curve (h) (horizontal combination of a parallel arrangement). This is the required water supply curve (yield) at point C from fire pump and public water connection in simultaneous operation. To illustrate: At 80 psi (5.52 bar) (552 kPa), curve (d) (300 gpm [1140 l/min]) + curve (g) (825 gpm [3120 l/min]) = 1,125 gpm (4260 l/min). Calculation of Yield from Supplies in Series (Fig. 22) In the example shown, a test of the public water connection at point A indicates that flow of 1,000 gpm (3790 l/min) reduces pressure from 34 psi (2.34 bar) (234 kPa) static to 30 psi (2.07 bar) (207 kPa) residual. A 1,000-gpm (3790-l/min), 60 psi (4.14 bar) (414 kPa) fire pump takes suction from the public main through 211 ft (65 m) of 8-in. (203-mm) pipe with C = 80 and discharges through 163 ft (50 m) of 6-in. (152-mm) pipe with C = 120 to point C. Points A, B, and C are at the same elevation. To plot the water supply curve (yield) at point C from the booster pump: 1. Plot water supply curve (a) at point A available from public water. 2. Plot friction loss curve (b) for the pump suction pipe between points A and B. Using equation 16, Pf = 0.024 1.52 211 = 7.7 psi (0.0054 1.52 65 = 0.53 bar [53 kPa]) at 1,500 gpm (5680 l/min) 3. Plot suction-supply curve (c) to the pump at point B by subtracting curve (b) from curve (a) (vertical combination). 4. Plot characteristic curve (d) of the pump at zero suction pressure. 5. Plot pump discharge curve (e) by adding curves (c) and (d) at several flows as given in Table 7 (vertical combination). Note this means booster pump suction pressure equals public residual pressure at all flow rates.
Table 7. Pump Discharge Pressures Net head of pump, psi (bar)(kPa) Curve (d) 72 Residual suction pressure, psi (bar) (kPa) Curve (c) 35 Pump discharge pressure, psi (bar) (kPa) Curve (e) 107
Flow, gpm (l/min) 0

3-0
Page 37
(0) 500 (1890) 750 (2840) 1000 (3790) 1250 (4730) 1500 (5680)
(4.96)(496) 69 (4.76)(476) 65 (4.48)(448) 60 (4.14)(414) 52 (3.59)(359) 39 (2.69)(269)
(2.41)(241) 33 (2.28)(228) 30 (2.07)(207) 26 (1.79)(179) 22 (1.52)(152) 17 (1.17)(117)
(7.38)(738) 102 (7.03)(703) 95 (6.55)(655) 86 (5.93)(593) 74 (5.1 )(510) 56 (3.86)(386)
6. Plot friction loss curve (f) for pipe between points B and C. Using equation 16 Pf = 0.046 0.72 163 = 5.4 psi (0.0104 0.72 50 = 0.37 bar [37 kPa]) at 1,000 gpm (3790 l/min). 7. Plot supply curve (g), showing yield at point C, by subtracting curve (f) from curve (e) (vertical combination). This is the yield available at point C from the booster pump. HYDRAULICS OF SPRINKLER SYSTEMS Pressures and discharges of sprinklers in simultaneous operation are not readily calculated by any exact formula. The most practical methods for calculating water flow in sprinkler piping are to use manual calculation methods for either tree-type of looped sprinkler systems as outlined in the data sheet, or a computer software program. Computations by the methods explained in this section agree closely with test work. Sprinkler calculations are based on hydraulic principles of discharge from circular orifices and water flow in pipes. Discharge depends on total pressure and friction loss. Figure 1 and equations 12 to 15, with associated text, illustrate relationships between flows, discharges, and pressures. While sprinkler system computations can be made precisely, final results can vary 5% or more from actual flows measured experimentally. The reasons are: 1. Pipe coefficients vary widely with size and age and from location to location. 2. FM Approved sprinkler K values vary 5%. 3. Water supplies can easily be different from when tested. In this data sheet, assumed values are nominal and indicated accuracy in examples is for illustration and comparison only. Sprinkler System Demand Specifications Demand specifications for sprinkler protection often are expressed as rate of water application (density) for a given area (area of demand). Sprinkler systems so designed must be able to discharge a specified density over the hydraulically most remote area (defined in Data Sheet 2-8N, Chapter 7) as well as over every other similar area covered by the systems. Typically, the remote area is calculated and the pattern of pipe sizes established therein is extended to the rest of the system.
3-0
Page 38

Calculation Techniques Sprinkler system calculations rely heavily on equations 12 and 13. Since square root is convenient, equation 12, Q = KPt is often used solely. However, N1.85 paper is also convenient, so entire sprinkler systems are often represented by equation 13, Q = KPt0.54. (Note that Q = Kp0.54 approximates p = (Q/K)1.85 closely.) For most sprinkler systems and demand areas, use Q = Kp0.54 or N1.85 paper, but for individual sprinklers or small groups of sprinklers, use Q = Kp0.5. Calculate branch line pressure and discharge with some or all sprinklers operating by assuming a pressure at the end sprinkler in the demand area and working back to the cross main. Pressure losses in sprinkler fittings need not be considered. Proceed as follows: 1. At the end sprinkler (point A1 of Figure 23), assume a total pressure and calculate discharge using equation 12 and sprinkler K from Table 5. Accuracy to 1 gpm at 0.1 psi is ample (approximately 4 l/min and 7 mbar [0.7 kPa]).
Fig. 23. Sprinkler layout used in waterflow calculation examples
2. At the second sprinkler, point A2, add friction loss in the end pipe to total pressure at the end sprinkler to obtain total pressure. 3. Calculate discharge from the second sprinkler, using total pressure at point A2, equation 12, and sprinkler K from Table 5. 4. Calculate discharges and pressures for third and subsequent sprinklers on the same branch line similarly. To obtain total pressure at the cross main for a single branch line, add the following: 1. Total pressure at sprinkler nearest to cross main 2. Friction loss in pipe between nearest sprinkler and cross main 3. Pressure loss in elbow (or tee) at top of nipple connecting branch line to cross main (obtained from Data Sheet 2-89) 4. Friction loss in nipple if over 6 in. (152 mm) long (ignore elevation) 5. Pressure loss in tee at bottom of nipple Using flow and pressure thus obtained, calculate K from equation 12 to represent the entire branch line.

3-0
Page 39
If the first cross main nipple supplies two branch lines, the top fitting is a tee instead of an elbow (Point 7 of Fig. 23), and nipple and branch line sizes may all be different. If two branch lines connected to a single tee are operating, calculate discharges (Qi and Qj) and cross main pressures (Pti and Ptj) from each as shown in the example. If Pti does not equal Ptj , they must be equalized because two different pressures at a single point are impossible. (This condition is equivalent to the requirement that pressure difference between any two points of a network must be the same by any route through the network.) To equalize, select the lower (say Ptj) and adjust it up to Pti. Using equation 14, adjust Qj to QjPti/Ptj . For single branch lines or rows, equation 14 is more realistic than equation 15. It is unnecessary to recalculate other flows and pressures. Using flow and pressure thus obtained, calculate K from Equation (12) to represent both branch lines. When several branch lines are operating and all are alike, K is the same for each. Thus at point 9, branch row K is the same as at point 8 and total pressure is the sum of friction loss in the end length of cross main and total pressure at point 8. Flow to branch row CD is then Pt (at point 9) times K. Calculate remaining rows similarly. Often branch rows alternate in length, e.g., 15 sprinklers, 16 sprinklers, 15 sprinklers, etc. Find one K for 15-sprinkler rows and another for 16-sprinkler rows and proceed as above using appropriate K for each row. When flow and pressure are obtained for any point supplying all operating sprinklers, add elevation pressure and friction loss in pipe and fittings back to a selected part of the supply system (riser, fire pump, etc.). The following is common nomenclature on calculation forms: Pt = Pn + Pv = total pressure Pn = Pt - Pv = normal (pipe, gauge, or net) pressure Pv = velocity pressure Pf = friction loss (Equation 16) Pe = elevation pressure Q = flow rate C = Hazen-Williams coefficient K = 5.6 (81 bar)(8.1 kPa) for 12-in. (13 mm) sprinklers (Table 5). Set up calculations in tabular form starting with no. 1 sprinkler on branch line A. The note column is used for preliminary calculations. Example (Fig. 23) Given: A wet pipe sprinkler system with indicated pipe sizes and sprinklers 8 ft (2.44 m) apart, on branch lines 10 ft (3.05 m) apart. Required: Find pressure needed at point 18 to maintain flow of 960 gpm (3630 l/min) with indicated 40 sprinklers operating. Result: 70.3 psi (4.85 bar) (485 kPa) at top of riser. (The computation is given in Table 8.) In reporting results of calculations, round off to the nearest 10 gpm and 1 psi. (If calculations are in metric units, equivalent values are about 40 l/min and 70 mbar or 7 kPa). Use of Velocity Pressure This section is included to acquaint the reader with a technique commonly used in the industry. There is no need to use velocity pressure in sprinkler system calculations. In order to compensate for different methods of calibrating sprinkler K, velocity pressure is sometimes deducted from total pressure to obtain normal pressure, and sprinkler discharge is calculated from equation 11. This technique provides more nearly accurate results when sprinkler K is determined for normal pressure in the pipe or fitting, or for total pressure in the sprinkler. No compensation is needed and velocity pressure should not be used when sprinkler K is determined for total pressure in the pipe as it is in Table 5. When actual internal pipe diameter is known, equation 17, a variation of equation 9, gives values of velocity pressure. Pv = Q2/cd4(17) where:
3-0
Page 40

Pv = velocity pressure, psi (bar) (kPa) Q = Flow, gpm (l/min) c = constant = 888 (0.443 bar) (0.00443 kPa) (Use 0.443 for Pv in bars and 0.00443 for Pv in kPa). d = actual internal pipe diameter, in. (mm). Table 9 gives values of cd4 for schedule 40 steel pipe. Use equation 17 and actual internal pipe diameter for all other pipe. Calculation Techniques With Velocity Pressure When velocity pressure is considered, it is done as follows: 1. At the end sprinkler (point A1 of Figure 23) assume a total pressure and calculate discharge using equation 12. Velocity pressure is not considered at the end sprinkler. 2. At the second sprinkler (point A2 of Figure 23) add friction loss in the end pipe to total pressure at the end sprinkler to obtain total pressure at Point A2.
Table 8. Sprinkler System Calculation For Example. Pipe Fittings and Devices
Spkr. or Nozzle Ident. & Location A1 Pressure Summary, psi (bar) (kPa) Pt 10.0 (0.69) (69) Pe Pf 0.9 (0.06) (.6)
Flow, gpm (l/min) q 18 ( 68)
Pipe Size in. (mm) 1 (25)
Friction Loss, psi/ft (mbar/m) (kPa/m) 0.107 (24) (2.4)
Q 18 ( 68)
Equiv. Pipe Length, ft (m) Lgth 8 (2.44) Fit Tot 8 (2.44)
FM Global Operating Standards
A2
q 18 ( 68)
1 (25)
0.386 (87) (8.7) 0.238 (54) (5.4) 0.205 (46) (4.6) 0.329 (74) (7.4) 0.146 (33) (3.3)
A3
Q 36 (136) q 21 ( 80)
114 (32)
Q 57 (216)
Lgth 8 (2.44) Fit Tot 8 (2.44) Lgth 8 (2.44) Fit Tot 8 (2.44)
Pt 10.9 (0.75) (75) Pe Pf 3.1 (0.21) (21) Pt 14.0 (0.96) (96) Pe Pf 1.9 (0.13) (13)
A4
q 22 ( 83)
112 (38)
A5
Q 79 (299) q 23 ( 87)
112 (38)
A6
Q 102(386) q 25 ( 95) T
2 (51)
Q 127 (481) q 127 (481)
B4
Q q 22 ( 83) T 0.205 (46) (4.6)
112 (38)
Q 79 (299) q 79 (299) Q
q 87 (329) Q 214 (810)
Lgth 8 (2.44) Fit Tot 8 (2.44) Lgth 8 (2.44) Fit Tot 8 (2.44) Lgth 4 (1.22) Fit 10 (3.05) Tot 14 (4.27) Lgth Fit Tot Lgth 4 (1.22) Fit 8 (2.44) Tot 12 (3.66) Lgth Fit Tot Lgth Fit Tot
Pt 15.9 (1.09) (109) Pe Pf 1.6 (0.11) ( 11) Pt 17.5 (1.20) (120) Pe Pf 2.6 (0.18) (18) Pt 20.1 (1.38) (138) Pe Pf 2.0 (0.14) (14) Pt 22.1 (1.52) (152) Pe Pf Pt 15.9 (1.10) (110) Pe Pf 2.5 (0.17) (17) Pt 18.4 (1.27) (127) Pe Pf Pt 22.1 (1.52) (152) Pe Pf
Normal Pressure, psi (bar) (kPa) Notes Pt Assume 10 psi (0.69 bar) (69 kPa) at sprinkler. Pv C = 120, K = 5.6 (81, 8.1) Pn (Use 81 for p in bars and 8.1 for p in kPa) Flow = 5.6 10= 17.7 gpm (81 0.69= 67.3 l/min) Pt Flow = 5.6 10.0= 18.5 gpm Pv (81 0.75 = 70.1 l/min) Pn Pt Flow = 5.6 14 = 21.0 gpm Pv (81 0.96 = 79.3 l/min) Pn Pt Flow = 5.6 15.9 = 22.3 gpm Pv (81 1.09 = 84.6 l/min) Pn Pt low = 5.6 17.5 = 23.4 gpm Pv (81 1.20 = 88.7 l/min) Pn Pt Flow = 5.6 20.1 = 25.1 gpm Pv (81 1.38 = 95.2 l/min) Pn Pt Contribution from 6-sprinkler branch line at point 7. Pv Pn Pt Flow and pressure at point B4 are the same as at Pv point A4 Pn Pt Contribution from 4-sprinkler branch line at point 7 Pv before adjustment Pn Pt Adjust pressure to 22.1 psi (1.52 bar) (152 kPa) Pv q = 79 22.1/18.4 = 86.6 gpm Pn (299 1.52/1.27 = 327 l/min)
Page 41
3-0
Table 8. Sprinkler System Calculation (contd.) Now carry flow down to cross main, compute branch line K, and finish calculation.
3-0
Page 42
Spkr. or Nozzle Ident. & Location 7
Flow, gpm (l/min) q
Pipe Fittings and Devices T
Friction Loss, psi/ft (mbar/m) (kPa/m) 0.384 (87) (8.7) 0.384 (87) (8.7) 0.622 (141)(14.1) 0.495 (112) (11.2)
Notes Riser nipple = 1 ft (0.305 m). Include length and ignore elevation. Include tee at bottom of nipple.
Q 214 (810) q
2 (51)
Q 214 (810)
Equiv. Pipe Length, ft (m) Lgth 1 (0.305) Fit 10 (3.05) Tot 11 (3.35) Lgth 10 (3.05) Fit Tot 10 (3.05)
Pressure Summary, psi (bar) (kPa) Pt 22.1 (1.52) (152) Pe Pf 4.2 (0.29) ( 29) Pt 26.3 (1.81) (181) Pe Pf 3.8 (0.26) (26)
Normal Pressure, psi (bar) (kPa) Pt Pv Pn Pt Pv Pn
K = 214/ 26.3 = 41.7 (810/ 1.81 = 602 [60.2]) (Use 602 for p in bars and 60.2 for p in kPa.)
q 229 (867)
212 (63)
10
Q 443 (1680) q 251 (950)
3 (76)
Q 694 (2630)
Lgth 10 (3.05) Fit Tot 10 (3.05) Lgth 10 (3.05) Fit Tot 10 (3.05)
Pt 30.1 (2.07) (207) Pe Pf 6.2 (0.43) (43) Pt 36.3 (2.50) (250) Pe Pf 5.0 (0.35) (35)
11
q 268 (1010)
3 (76)
12
Q 962 (3640) q
312 (89)
14
Q 962 (3640) q T
4 (102)
16
Q 962 (3640) q 2-E
6 (152)
18
Q 962 (3640) q
6 (152)
Q 962 (3640)
Lgth 10 (3.05) Fit Tot 10 (3.05) Lgth 20 (6.1) Fit Tot 20 (6.1) Lgth 15 (4.57) Fit 20 (6.1) Tot 35 (10.67) Lgth 60 (18.3) Fit 28 ( 8.53) Tot 88 (26.8) Lgth Fit Tot
0.906 Pt 41.3 (2.85) (285) (205) (20.5) Pe Pf 9.1 (0.63) (63) 0.446 Pt 50.4 (3.48) (348) (101) (10.1) Pe Pf 8.9 (0.61) (61) 0.241 Pt 59.3 (4.09) (409) (55) (5.5) Pe Pf 8.4 (0.58) (58) 0.033 Pt 67.7 (4.67) (467) ( 7) (0.7) Pe Pf 2.9 (0.20) (20) Pt 70.6 (4.87) (487) Pe Pf
Pt Flow = 41.7 30.1 = 229 gpm Pv (602 2.07 = 866 l/min) Pn Pt Flow = 41.7 36.3 = 251 gpm Pv (602 2.50 = 952 l/min) Pn Pt Flow = 41.7 41.3 = 268 gpm Pv (602 2.85 = 1016 l/min) Pn Pt Pv Pn Pt Pv Pn Pt Pv Pn Pt70.6(4.87)(487) Flow and total pressure at point 18, top of riser. Pv 0.7 (0.05) ( 5 ) Pn69.9(4.82)(482)4) Normal (gauge) pressure at Point 18.
Using Equation 15, adjust flow to 960 gpm (3630 l/min) and pressure to (960/962)1.85 70.6 = 70.3 psi ((3630/3640)1.85 4.87 = 4.85 bar [485 kPa])

3-0
Page 43
Table 9. Velocity Pressure Factors Nominal Pipe Size, in. (mm) 12 ( 13) 34 ( 19) 1 ( 25) ( 32) 114 ( 38) 112 2 ( 51) ( 64) 212 3 ( 76) ( 89) 312 4 (102) 5 (127) 6 (152) 8 (203) 133 410 1,080 3,230 5,980 16,200 33,100 78,800 141,000 234,000 577,000 1,200,000 3,640,000 cd4 for pressure in psi (bar) (kPa) ( 27,000) ( 270) ( 83,000) ( 830) ( 218,000) ( 2,180) ( 654,000) ( 6,540) ( 1,210,000) ( 12,100) ( 3,290,000) ( 32,900) ( 6,700,000) ( 67,000) ( 16,000,000) ( 160,000) ( 28,600,000) ( 286,000) ( 47,300,000) ( 473,000) (117,000,000) (1,170,000) (244,000,000) (2,440,000) (765,000,000) (7,650,000)
3. To find velocity pressure at the second sprinkler, assume a discharge and add it to the end sprinkler discharge. Using this total, compute velocity pressure between second and third sprinkers from Equation (17), and subtract it from total pressure at the second sprinkler to obtain trial normal pressure 4. Using trial normal pressure and equation 11, calculate flow. This calculated flow is probably correct. If assumed flow and calculated flow are widely different, repeat the procedure in Step 3, assuming the flow just obtained. 5. Calculate discharges and pressures for third and subsequent sprinklers on the same branch line similarly. 6. Calculate branch line Ks and cross main flows and pressures as in the example without velocity pressure. There is no need to include velocity pressure in cross main calculations since branch line K is calculated from total pressure in the cross main more easily than from normal pressure in the cross main or total pressure in the nipple. Nevertheless, cross main velocity pressure is sometimes used in the industry. The following procedure illustrates one technique for cross main calculations with velocity pressure: 1. Obtain total pressure at the bottom of the nipple connecting branch line to cross main as indicated for the example without velocity pressure. Note that the tee at the bottom of the nipple is excluded. 2. Using flow and pressure thus obtained, calculate K from equation 12 to represent the entire branch line (or branch row if sprinklers on both sides of the cross main are operating). This K is calculated for total pressure at the nipple exit of the cross main tee rather than for total pressure at the upstream entrance of the cross main tee. 3. Add total pressure at the bottom of the nipple, friction loss in the tee at the bottom of the nipple, and friction loss in the end length of cross main. 4. Obtain velocity pressure and normal pressure at point 9 as for the second sprinkler in the description above. 5. Using K from step 2, Pn from step 4, and equation 11, calculate flow for row CD. 6. Calculate discharges and pressures for third and subsequent branch rows similarly. Example (Fig. 23) Given: A wet pipe sprinkler system with indicated pipe sizes and sprinklers 8 ft (2.44 m) apart on branch lines 10 ft (3.05 m) apart. Required: Find pressure needed at point 18 to maintain flow of 960 gpm (3630 l/min) with indicated 40 sprinklers operating. Result: 72.8 psi (5.01 bar) (501 kPa) at top of riser. (The computation is given in Table 10.) Since velocity pressure is used for compensation when sprinkler K is determined from normal rather than total pressure in the pipe, this example uses sprinkler K = 5.7 (82 for p in bars and 8.2 for p in kPa), rather than sprinkler K = 5.6 (81 for p in bars and 8.1 for p in kPa). Results of the two calculations are given in Table 11 and are closely comparable.
Table 10. Sprinkler System Calculation For Example
3-0
Page 44
Spkr. or Nozzle Ident. & Location Pipe Fittings and Devices Equiv. Pipe Length, ft (m) Pressure Summary, psi (bar) (kPa) Notes Assume Pt = 10.0 psi (0.69 bar) (69 kPa). C = 120, sprinkler K = 5.7 (82) (8.2) (Use 82 for p in bars and 8.2 for p in kPa.) Normal Pressure, psi (bar) (kPa) 0.107 (24, 2.4) 0.386 (87) (8.7) Friction Loss, psi/ft (mbar/m) (kPa/m)
Flow, gpm (l/min)
Pipe Size in. (mm)
A1
q 18 ( 68)
1 (25)
Q 18 ( 68)
Lgth 8 (2.44) Fit Tot 8 (2.44)
Pt 10.0 (0.69) (69) Pe Pf 0.9 (0.06) (6)
Pt Pv Pn
A2
q 18 ( 68)
1 (25)
Q 36 (136)
Lgth 8 (2.44) Fit Tot 8 (2.44)
Pt 10.9 (0.75) (75) Pe Pf 3.1 (0.21) (21)
Pt 10.9 (0.75) (75) Pv 1.2 (0.08) (8) Pn 9.7 (0.67) (67)
Assume 18 gpm (68 l/min). 18+18=36 gpm (68+68=136 l/min). Pv=1.2 psi (0.083 bar) (8.3 kPa) Pn=10.9-1.2=9.7 psi (0.75-0.08 = 0.67 bar [67 kPa]). Flow = 5.7 9.7 = 17.8 gpm. (82 .67 = 67 l/min).
A3 0.238 (54) (5.4) Lgth 8 (2.44) Fit Tot 8 (2.44) Pt 14.0 (0.96) (96) Pe Pf 1.9 (0.13) (13) Pt 14.0 (0.96) (96) Pv 1.0 (0.07) (7) Pn 13.0 (0.90) (90)
q 21 ( 80)
114 (32)
Q 57 (216)
Assume 20 gpm (76 l/min). 36+20=56 gpm (136+76=212 l/min). Pv = 1.0 psi (0.069 bar) (6.9 kPa). Pn=14.0-1.0=13.0 psi (0.96-0.07=0.89 b, [89 kPa]). Flow = 5.7 13.0 = 20.6 gpm* (82 .90 = 78 l/min).
A4 0.205 (46) (4.6)
q 22 ( 83)
112 (38) Lgth 8 (2.44) Fit Tot 8 (2.44) Pt 15.9 (1.09) (109) Pe Pf 1.6 (0.11) (11) Pt 15.9 (1.09) (109) Pv 1.2 (0.08) (8) Pn 14.7 (1.01) (101)
Q 79 (299)
Assume 25 gpm (95 l/min). 57+25=82 gpm (216+95=311 l/min). Pv = 1.2 psi (0.083 bar) (8.3 kPa). Pn = 15.9-1.2=14.7 psi (1.09-0.08=1.01 bar [101 kPa]). Flow = 5.7 14.7 = 21.9 gpm* (82 1.01 = 82 l/min).
A5 0.329 (74) (7.4) Lgth 8 (2.44) Fit Tot 8 (2.44) Pt 17.5 (1.20) (120) Pe Pf 2.6 (0.18) (18)
q 23 ( 87)
112 (38)
Q 102 (386)
Pt 17.5 (1.20) (120) Pv 1.8 (0.12) (12) Pn 15.7 (1.08) (108)
Assume 25 gpm (95 l/min). 79+25=104 gpm (299+95=394 l/min). Pv = 1.8 psi (0.12 bar) (12 kPa). Pn = 17.5-1.8=15.7 psi (1.20-0.12=1.08 bar [108 kPa]). Flow = 5.7 15.7 = 22.6 gpm* (82 1.08 = 85 l/min).
A6 0.146 (33) (3.3) Pt 20.1 (1.38) (138) Pe Pf 2.0 (0.14) (14)
q 25 ( 95)
2 (51)
2006 Factory Mutual Insurance Company. All rights reserved. Lgth 4 (1.22) Fit 10 (3.05) Tot 14 (4.27) Pt20.1 (1.38) (138) Pv 1.1 (0.07) (7) Pn 19.0 (1.31) (131) Lgth Pt 22.1 (1.52) (152) Pt
Q 127 (481)
Assume 28 gpm (106 l/min). 102+28=130 gpm (386+106=492 l/min). Pv = 1.1 psi (0.076 bar) (7.6 kPa). Pn = 20.1-1.1=19.0 psi (1.38-0.07=1.31 bar [131 kPa]). Flow = 5.7 19.0 = 24.8 gpm* (82 1.31 = 94 l/min). Contribution from 6-sprinkler branch line at point 7.
q 127 (481)
* Recomputation using this as the assumed value yields agreement between assumed and computed values.
Table 10. Sprinkler System Calculation For Example (contd)

Pipe Fittings and Devices Equiv. Pipe Length, ft (m) Notes Lgth 4 (1.22) Fit 8 (2.44) Tot 12 (3.66) Lgth Pt Pt 18.4 (1.27) (127) 0.205 (46) (4.6) Pressure Summary, psi (bar) (kPa) Normal Pressure, psi (bar) (kPa) T Friction Loss, psi/ft (mbar/m) (kPa/m)
Spkr. or Nozzle Ident. & Location
Flow, gpm (l/min)
Pipe Size in. (mm)
B4
q 22 ( 83)
112 (38)
Q 79 (299)
Pt 15.9 (1.09) (109) Pe Pt 15.9 (1.09) (109) Pv Flow and pressure at point B4 are the same as at point Pf 2.5 (0.17) (17) 1.2 (0.08) (8) A4. Pn 14.7 (1.01) (101) Contribution from 4-sprinkler branch line at point 7 before adjustment.
q 79 (299)
Adjust 18.4 psi (1.27 bar) (127 kPa) up to 22.1 psi (1.52 bar) (152 kPa), and using equation 14, adjust 79 gpm (299 l/min) up to 79 22.1/18.4 = 86.6 gpm (299 1.52/1.27 = 327 l/min). Combined flow for row b is 127 + 87 = 214 gpm (481 + 329 = 810 l/min). Now carry flow down to cross main, compute branch line K, and finish calculation. 0.384 (87) (8.7) 0.384 (87) (8.7) 0.606 (137) (13.7)
q 214 (810)
2 (51)
Q T
Lgth 1 (0.305) Fit Tot 1 (0.305)
Pt 22.1 (1.52) (152) Pe Pt Pf 0.4 (0.03) (3) Pv Pn Pt 22.5 (1.55) (155) Pe Pt Pf 7.7 (0.53) (53) Pv Pn K = 214/22.5 = 45.1 (810/ 1.55 = 651 [65.1]). (Use 651 for p in bars and 65.1 for p in kPa)
q 214 (810)
2 (51)
Q Lgth 10 (3.05) Fit Tot 10 (3.05) Lgth 10 (3.05) Fit Tot 10 (3.05) 0.484 (109) (10.9)
Lgth 10 (3.05) Fit 10 (3.05) Tot 20 (6.1)
q 223 (844)
2 12 (63)
Q 437 (1654)
Pt 30.2 (2.08) (208) Pe Pt 30.2 (2.08) (208) Pv Assume 216 gpm (818 l/min) Pf 6.1 (0.42) (42) 5.7 (0.39) (39) Pn 24.5 Total = 430 gpm (1630 l/min) (1.69) (169) Flow = 45.1 24.5 = 223 gpm (651 1.69 = 846 l/min) Pt 36.3 (2.50) (250) Pe Pt 36.3 (2.5) (250) Pv Pf 4.8 (0.33) (33) 5.7 (.39) (39) Pn 30.6 (2.11) (211) Assume 233 gpm (882 l/min) Total = 670 gpm (2540 l/min) Flow = 45.1 30.6 = 249 gpm (651 2.11 = 946 l/min)
10
q 249 (943)
3 (76)
Q 686 (2600) Lgth 10 (3.05) Fit Tot 10 (3.05) 0.856 (194) (19.4)
11
q 247 (935)
3 (76)
Q 933 (3530) 0.421 (95) (9.5) 0.228 (52) (5.2)
Pt 41.1 (2.83) (283) Pe Pt 41.1 (2.83) (283) Pv Assume 244 gpm (924 l/min) Total = 930 gpm (3520 l/min) Pf 8.6 (0.59) (59) 11.1 (0.76) (76) Pn 30.0 (2.07) (207) Flow = 45.1 30 = 247 gpm (651 2.07 = 937 l/min) Pt 49.7 (3.42) (342) Pe Pt Pf 8.4 (0.58) (58) Pv Pn Pt 58.1 (4.00) (400) Pe Pt Pf 8.0 (0.55) (55) Pv Pn
12 Lgth 20 (6.1) Fit Tot 20 (6.1) T
3 12 (89)
Q 933 (3530)
14
4 (102)
2006 Factory Mutual Insurance Company. All rights reserved. Lgth 15 (4.57) Fit 20 (6.10) Tot 35 (10.67) 2-E Lgth 60 (18.3) Fit 28(8.53) Tot 88 (26.8) Lgth 0.031 (7) (0.7) Pt 66.1 (4.55) (455) Pe Pt Pv Pn Pf 2.7 (0.19) (19) Pt 68.8 (4.74) (474) Pt 68.8 (4.74) (474) Pv 0.8 (0.06) (6) Pn 68.0 (4.88) (488)
Q 933 (3530)
16
6 (152)
Q 933 (3530)
18
6 (152)
Flow and total pressure at point 18. top of riser. Normal (gauge) pressure at point 18.
933 (3530)
Using Equation 15, adjust flow to 960 gpm (3630 l/min) and pressure to (960/933)1.85 68.8 = 72.5 psi ((3630/3530)1.85 4.74 = 5.0 bar [500 kPa]).
Page 45
3-0
3-0
Page 46

Table 11. Comparison Of Results For Different Calculation Methods Calculation Method Velocity pressure excluded and sprinkler K based on total pressure Velocity pressure included and sprinkler K based on normal pressure Flow 960 gpm (3630 l/min) 960 gpm (3630 l/min) Pressure 70.3 psi (4.85 bar) (485 kPa) 72.8 psi (5.01 bar) (501 kPa)
Sprinkler System Flow and Pressure Adjustments When operating sprinklers and piping are at one level, adjust flow and pressure to existing supplies or other restrictions for a fixed array of operating sprinklers by calculating pressure to top-of-riser and by plotting on N1.85 paper or by computing K to describe the entire operating sprinkler system downstream from the riser. Calculate adjustments as follows: Using equation 13 and calculation results from the example with velocity pressure excluded, K = 962/ (70.6)0.54 = 97 (3640/ (4.87)0.54 = 1550, 3640/(487)0.54 = 129) (Use 1550 if p is in bars and 129 if p is in kPa) If 7 psi (0.48 bar) (48 kPa) at the end sprinkler is desired, the top-of-riser requirement is: (7 psi/10 psi) (70.6 psi) = 49.4 psi ((0.48 b/0.69 bar) (4.87 bar) = 3.39 bar [339 kPa]) and from equation 13, Q = Kp0.54 = 97 (49.6)0.54 = 799 gpm (1550 (3.42)0.54 = 3010 l/min = 129 (342)0.54) Similarly, for a 1000 gpm (3790 l/min) flow limitation pressure required from equation 13 is p = (Q/K)1.85 = (1000/97)1.85 = 74.9 psi ((3790/1550)1.85 = 5.2 b, (3790/129)1.85 = 520 kPa), and pressure at the end sprinkler is: 10 psi (76.3 psi/70.6 psi) = 10.8 psi (0.69 b (5.29 b/4.87 bar) = 0.75 bar [75 kPa]). Adjust flow and pressure graphically as follows (Fig. 24): 1. Plot point A at 960 gpm (3630 l/min) and 71 psi (4.9 bar) (490 kPa). 2. Draw curve (a) through this point and the 0-0 point. 3. To reduce end sprinkler pressure from 10 psi (0.69 bar) (69 kPa) to 7 psi (0.48 bar) (48 kPa), multiply top-of-riser pressure by the ratio of end sprinkler pressures to get (7/10) 71 psi = 49.7 psi ((7/10) 4.9 b = 3.43 bar [343 kPa]) 4. Plot point B on Curve (a) at 50 psi (3.45 bar) (345 kPa) and read flow as 790 gpm (2990 l/min). 5. Even more easily, for a 1000 gpm (3790 l/min) limitation, plot point C on curve (a) at 1000 gpm (3790 l/min) and read pressure as 77 psi (5.3 bar) (530 kPa). Multilevel Systems For a system in which sprinklers operate simultaneously at several levels, calculate flow, pressure, and K for every level at the junction with the riser. Then, treating the riser as a cross main, work down from the top toward the point of supply, taking elevations into account at the junctions. To convert to another flow-pressure requirement, calculate new flow and pressure for the top level using previously obtained K, add elevation pressure (unchanged) down to next operating level, recalculate flow and pressure at that level using previously obtained K, etc. This method must be used for calculations since elevation pressure is independent

3-0
Page 47
Fig. 24. Flow and pressure adjustments for specific array of operating sprinklers
of flow and other pressures are not. To calculate a system K, and to attempt to change pressure requirements with it, would change elevation pressures equivalent to shrinking or expanding the system vertically. Graphical solutions using N1.85 paper are more practical. Given: Sprinkler arrays represented by curves (a), (b) and (c) in Figure 25. The arrays are at three levels in an open multi-story building. Problem: To adjust flow and pressure obtained from initial calculations for different conditions. Solution: Add curve (b) to curve (a) horizontally to obtain curve (d). Then add curve (c) to curve (d) horizontally to obtain curve (e). Discussion: Curve (e) (broken line) represents the entire sprinkler system. If the pressure isnt high enough to supply the top sprinklers, the first or second portion of Curve (e) gives proper flow to bottom levels. Looped and Gridded Systems Loops and grids to reduce friction losses or pipe sizes are becoming more common as computer programs are developed to reduce the time needed to process necessary repetitive procedures for balancing flows and pressures. Simple loops can be calculated by hand. It is impractical to calculate more involved systems by hand. Calculations are based on two principles stated earlier: 1. Pressure difference between any two points of a network is the same by any route through the network. 2. Flow toward a point equals flow away from the point.
3-0
Page 48

Fig. 25. Demand lines for a multilevel sprinkler system
Figure 26 shows a typical looped sprinkler system. For a given area of demand, part of the water is supplied around each side of the loop. If half the supply comes from each side, the point so supplied is the most hydraulically remote. This point is often (but not always) centered in the most hydraulically remote area of demand. Sometimes an area of demand completely separate from the remote cross main point generates the greatest water supply requirement. (See following examples.) Also, if two areas of demand are found such that one generates a higher flow requirement and the other a higher pressure requirement, water supplies must meet the demand for both areas. Loop calculations are easiest if all pipes are the same size. Different diameter pipes can be converted to one size, on paper, by using the Table of Equivalent Lengths in Data Sheet 2-89. For example, in Figure 26 there are 222 ft (68.7 m) of 6-in. (152-mm) pipe, plus an elbow, and 268 ft (81.7 m) of 4-in. (102-mm) pipe, plus 3 elbows. From Data Sheet 2-89, 100 ft (30.5 m) of 6-in. (152-mm) pipe is equivalent to 13.6 ft (4.15 m) of 4-in. (102-mm) pipe. Thus the equivalent length of the loop in 4 in. (102 mm) pipe is 268 + 30 + (222 + 14) (0.136) = 330 ft (81.7 + 9.1 + (67.7 + 4.3) (0.136) = 100.6 m). The most remote point is at A, 165 ft (50.3 m) from the supply connection. Calculate demand for an area centered at Point A.

3-0
Page 49
Fig. 26. Typical looped feed sprinkler system
Example 1 (Fig. 26) Given: A wet pipe system illustrated in Figure 26. Required: To find flow and pressure at point B needed to maintain 0.30 gpm/ft2 (12 mm/min) density over 2000 ft2 (186 m2) centered at point A. Result: 610 gpm (2310 l/min) at 46.4 psi (3.2 bar) (320 kPa) at top of riser. (The computation is given in Table 12.) Point A is the hydraulically most remote point in the loop. The 2000 ft2 (186 m2) centered at point A contains rows at points 1 through 5. Row 3 is partly supplied from each half of the loop.
3-0
Page 50

Flow divides in two parts according to the following ratios derived from equation 6: Q1/Q2 = (L2/L1)0.54 and Q1/Q = 1/([L1/L2]0.54 + 1) where Q1 flows through length L1 Q2 flows through length L2 and point c is 162 ft (49.4 m) from B on the left side and 168 ft (51.2 m) through equivalent 4 in. (102 mm) pipe on the right side. Q = Q1 + Q2 is the total flow. Thus, flow on the shorter side is Q1 = Q/([162/168]0.54 + 1) = 0.505 Q (Q/([49.4/51.2]0.54 + 1) = 0.505 Q) and the longer side is Q2 = Q - Q1 = 0.495 Q This flow split can also be obtained from N1.85 paper, Figure 27, as follows: 1. Calculate two lengths of pipe as above and call the longer pipe L2 and the shorter pipe L1. 2. Select a convenient flow line near the right side of the graph paper and call that flow Q1. 3. Plot L1 and L2, as though they were pressures, on flow line Q1. Adjust the scale so L2 is high on the graph for better definition. 4. Draw a straight line, curve (a), between point L2 and the 0-0 point. 5. Draw a horizontal line, curve (b), through point L1. 6. The flow at which curve (a) intersects curve (b) is Q2. 7. Flow splits in the ratio Q1/Q2 with the smaller flow in L2 and the larger flow in L1. Calculate row 3 starting at 28.7 psi (1.98 bar) (198 kPa) total pressure at the end sprinkler. Exclude velocity pressure. Total flow and pressure at B are 610 gpm (2310 l/min) at 46.4 psi (3.2 bar) (320 kPa). Flow from each sprinkler is at least 30 gpm (113.5 l/min), so density is 30/100 = 0.3 gpm/ft2 (113.5/9.29 = 12.1 mm/min) over the area of demand. In this example, pressure at the end sprinkler is chosen to yield desired density. Flow from the second sprinkler is so close to flow from the end sprinkler that the difference is insignificant. Friction in the cross main is negligible within the area of demand, demonstrating that pipe sizes can be chosen to minimize friction and resulting flow and pressure requirements. (equation 18) (equation 19)

3-0
Page 51
Fig. 27. Graphical determination of flow split in loop
Table 12. Sprinkler System Calculation For Example 1 Pipe Fittings and Devices Notes C = 120 K = 5.6 (81)(8.1) (Use 81 if P is in bars and 8.1 if P is in kPa.) q = 5.6 31.5 = 31.4 gpm (81 2.17= 119.3 l/min)
3-0
Page 52
Spkr. or Nozzle Ident. & Location a
Normal Pressure, psi (bar) (kPa) Pt Pv Pn Pt Pv Pn
Spkr. or Nozzle Ident. & Location c
Friction Loss, psi/ft Flow, gpm Pipe Size Equiv. Pipe (mbar/m) Pressure Summary, (l/min) in. (mm) Length, ft (m) (kPa/m) psi (bar) (kPa) q 30 (114) 1 (25) Lgth 10 (305) 0.276 Pt 28.7(1.98)(198) Fit (62.4) (6.24) Pe Q 30 (114) Tot 10 (3.05) Pf 2.8 (0.19) (19) q 31 (117) 1 (25) Cr Lgth 5 (1.52) 1.025 Pt 31.5 (2.17)(217) Fit 5 (1.52) (232) (23.2) Pe Q 61 (231) Tot 10 (3.05) Pf 10.2 (0.71)(71) Lgth Pt 41.7 (2.88)(288) Flow to the other two sprinklers in Row 3 is also 61 gpm (231 l/min.) From Point c, 50.5% of 122 gpm (462 l/min), or 62 gpm (235 l/min), flows from the short side. Continue calculating the short side back to Point B. Pipe Friction Fittings Loss, psi/ft Flow, gpm Pipe Size and Equiv. Pipe (mbar/m) Pressure Summary, (l/min) in. (mm) Devices Length, ft (m) (kPa/m) psi (bar) (kPa) q 62 (235) 4 (102) Lgth 10 (3.05) 0.002 Pt 41.7 (2.88)(288) Fit (0.4) (0.04) Pe Pf 0 Q 62 (235) Tot 10 (3.05) q 122 (462) 4 (102) Lgth 10 (3.05) 0.011 Pt 41.7 (2.88)(288) Fit (2.5) (0.25) Pe Pf 0.1(0.01)(1) Q 184 (697) Tot 10 (3.05) Normal Pressure, psi (bar) (kPa) Pt Pv Pn Pt Pv Pn E&T
K = 122/ 41.7= 18.9 (462/ 2.88= 272 [27.2]) (Use 272 for p in bars and 27.2 for p in kPa.)
q 122 (462) Q 306 (1159)
4 (102)
q 306 (1159)
q 60 (227) Q 60 (227)
4 (102)
q 122 (462) Q 182 (689)
4 (102)
Lgth 132 (40.23) 0.029 (6.6) Pt 41.8(2.89)(289) Pt q = 18.9 41.8= 122 gpm. Fit 30 (9.14) Tot (0.66) Pe Pv (272 2.89= 462 l/min.) 162 (49.37) Pf 4.7 (0.32)(32) Pn Short side flow at B Lgth Pt 46.5 (3.21)(321) From Point c, 60 gpm (227 l/min) flows from the long side. Calculate the long side back to B. Pt Lgth 10 (3.05) 0.001 (0.2) Pt 41.7 (2.88)(288) Pe Pv Fit (0.02) Pf 0 (0) Pn Tot 10 (3.05) Lgth 10 (3.05) 0.011 (2.5) Pt 41.7 (2.88)(288) Pt Fit (0.25) Pf 0.1 (0.01)(1) Pv Pn Tot 10 (3.05)
Table 12. Sprinkler System Calculation For Example 1. {contd)
Spkr. or Nozzle Ident. & Location g Notes
Flow, gpm (l/min) q 122 (462)
Pipe Fittings and Devices 2E
Friction Loss, psi/ft (mbar/m) (kPa/m) 0.029 (6.6) (0.66)
Q 304 (1151) q
6 (152)
Q 304 (1151) q 304 (1151)
q 306 (1159)
Equiv. Pipe Pressure Summary, Normal Pressure, Length, ft (m) psi (bar) (kPa) psi (bar) (kPa) Lgth 96 (29.26) Pt 41.8 (2.89)(289) Pt Fit 20 (6.10) Pv Pe Tot 116 (35.36) Pf 3.4 (0.23)(23) Pn Pt E&T Lgth 222 (67.67) 0.004 Pt 45.2 (3.12)(312) Pv Fit 44 (13.41) (0.9) (0.09) Pe Pf 1.1 (0.07)(7) Pn Tot 266 (81.08) Lgth Pt 46.3 (3.19)(319) Pt Long side flow at B Now combine flows from the two sides and average pressures since difference is small. Pt Short side flow at B Pt 46.5 (3.21)(321)
Q 610 (2310)
Page 53
3-0
3-0
Page 54

Since branch lines on the right side in Figure 26 contain three sprinklers instead of two, it might be more difficult to meet density and area requirements there than at the most remote point of the cross main. Check this as follows: Example 2 (Fig. 26) Given: The wet pipe system illustrated in Figure 26. Required: To find flow and pressure at point B to mainain 0.30 gpm/ft2 (12 mm/min) density over 2000 ft2 (186 m2) on the right side. Result: 646 gpm (2440 l/min) at 51 psi (3.52 bar) (352 kPa) at top of riser. (The computation is given in Table 13.) The most remote 2000 ft2 (186 m2) on the right side includes all sprinklers on rows 6, 7, and 8 and two sprinklers on row 9 as shown. From point m, equivalent length in 4-in. (102-mm) pipe for the left loop is 299 ft (91.1 m) and for the right loops is 31 ft (9.4 m). From Equation (19), flow on the short side is Q1 = Q/([31/1299]0.54 + 1) = 0.773 Q (Q/([9.4/91.1]0.54 + 1) = 0.773 Q) and on the long side is Q2 = Q - Q1 = 0.227 Q Of the total of 20 sprinklers, 77.3%, or 15.5, are supplied around the short side. This means that all sprinkers on rows 7, 8, and 9 plus one-and-a half on row 6 are supplied around the short side. Therefore the balance point is at k rather than at m. Refiguring equivalent length from k, the left loop is 298 ft (90.8 m) and the right loop is 32 ft (9.8 m). Flow on the short side is Q1 = Q/([32/298]0.54 + 1) = 0.769 Q (Q/([9.9/90.8]0.54 + 1) = 0.769 Q and on the long side is Q2 = Q - Q1 = .231 Q Of the total of 20 sprinklers, 23.1% or 4.6 are supplied around the long side. Calculate row 6 starting at 28.7 psi (1.98 bar) (198 kPa) on the end sprinkler. Exclude velocity pressure.
Table 13. Sprinkler System Calculation For Example 2 Pipe Fittings and Devices
Spkr. or Nozzle Ident. & Location h
Notes K = 5.6 (81) (8.1) (Use 81 if p is in bars and 8.1 if p is in kPa.) q = 5.6 31.5 = 31.4 gpm (81 2.17 = 119.3 l/min) q = 5.6 41.7 = 36.2 gpm (81 2.88 = 137.5 l/min)
Normal Pressure, psi (bar) (kPa) Pt Pv Pn Pt Pv Pn Pt Pv Pn
Pt Pv Pn
Essentially no friction loss in 6-in. (152-mm) pipe
Pt Pv Pn Pt Pv Pn Pt Pv Pn Pt Pt Pv Pn Pt Pt Pn
Friction loss is too small to record at accuracy being used
Friction Loss, psi/ft Flow, gpm Pipe Size Equiv. Pipe (mbar/m) Pressure Summary, (l/min) in. (mm) Length, ft (m) (kPa/m) psi (bar) (kPa) q 30 (114) 1 (25) Lgth 10 (3.05) 0.276 Pt 28.7 (1.98)(198) Fit (62.4) (6.24) Pe Q 30 (114) Tot 10 (3.05) Pf 2.8 (0.19)(19) i q 31 (117) 1 (25) Lgth 10 (3.05) 1.025 Pt 31.5 (2.17)(217) Fit (232) (23.2) Pe Q 61 (231) Tot 10 (3.05) Pf 10.2 (0.71)(71) Cr Lgth 5 (1.52) 0.636 Pt 41.7 (2.88)(288) j q 36 (136) 1 14 (32) Fit 6 (1.83) (144) (14.4) Pe Pf 7.0 (0.48)(48) Pt Tot 11 (3.35) Q 97 (367) 48.7 (3.36)(336) Flow to the other three sprinklers in Row 6 is also 97 gpm (367 l/min.) Of the flow to Row 6, 77% (4.6/6) or 149 gpm (549 l/min) is from the long side. Calculate the long side back to Point B. k q 149 (564) 4 (102) 3E&T Lgth 268 (81.69) 0.007 Pt 48.7 (3.36)(336) Fit 50 (15.24) (1.8)(0.18) Pe Pf 2. (0.18) (18) Q Tot 318 (96.93) B Pt 51.9 (3.54)(354) Now calculate the short side back to Point B. k q 45 (170) 6 Lgth 10 (3.05) 0 Pt 48.7 (3.36)(336) Pe (152) Fit Pf 0 Q Tot 10 (3.05) m q 194 (735) 6 Lgth 10 (3.05) 0.003 Pt 48.7 (3.36)(336) (152) Fit (0.68)(0.068) Pe Q 239 (905) Tot 10 (3.05) Pf 0 n q 194 (735) 6 Lgth 10 (3.05) 0.007 Pt 48.7 (3.36)(336) (152) Fit (1.6)(0.16) Pe Pf 0 1 (0.02)(2) Q 433 (1639) Tot 10 (3.05) t q Lgth Pt 48.8 (3.38)(338) Now include two sprinklers on row 9 and continue calculation back to point B. s q 61 (231) 1 Lgth 10 (3.05) 1.025 Pt 31.5 (2.17)(217) (25) Fit (232) (23.2) Pe Q Tot 10 (3.05) Pf 10.2 (0.7)(70) Cr Lgth 5 (1.52) .27 Pt 41.7 (2.87)(287) q 1 1 4 (32) Fit 6 (1.83) (61) (6.1) Pt Tot 11 (3.35) Pf 3.0 (0.21)(21) Q 61 (231) Sprinklers r and s are same as h and i
Page 55
3-0
Table 13. Sprinkler System Calculation For Example 2 (contd)
3-0
Page 56
Spkr. or Nozzle Ident. & Location t Equiv. Pipe Length, ft (m) Lgth Fit Tot q = 61 48.8/44.7 = 63.7 gpm (231 3.38/3.09 = 241 l/min) Combine flows at Point t. Pressure Summary, psi (bar) (kPa) Pt 44.7 (3.08)(308) Pe Pf Normal Pressure, psi (bar) (kPa) Pt Pv Pn
Flow, gpm (l/min) q
Pipe Size in. (mm)
Pipe Fittings and Devices
Friction Loss, psi/ft (mbar/m) (kPa/m)
Notes Balance flow and pressure to 48.8 psi (3.38 bar) (338 kPa) using Equation 14.
q 64 (242)
6 (152)
E&T
.01 (2.3) (0.23)
Q 497 (1881) q 149 (564)
Disregard small pressure difference at point B and combine flows.
Q 646 (2445)
Lgth 192 (58.52) Fit 44 (13.41) Tot 236 (71.93) Lgth Fit Tot
Pt 48.8 (3.36)(336) Pe Pf 2.4 (0.17)(17) Pt 51.2 (3.53) (353) Pe Pf
Pt Pv Pn Pt Pv Pn


3-0
Page 57
Total flow and pressure at B for examples 1 and 2 are given in Table 14.
Table 14. Comparison of Results From Examples 1 and 2 Example 1 Example 2 Flow 610 gpm (2310 l/min) 646 gpm (2445 l/min) Pressure 46.5 psi (3.21 bar)(321 kPa) 51 psi (3.52 bar) (352 kPa)
The design area of example 2 generates a greater water demand, hence is hydraulically more remote than the design area of example 1. Overlay Systems A second sprinkler system installed under an existing system is called an overlay system. Such arrangements are sometimes used when sprinkler demand increases greatly. If the two systems intersect at a single point, calculation of combined yield is quite simple. Calculate each separately to the intersection point, balance pressures, adjust flows, and calculate the remainder as for a single system. N1.85 paper can then be used to describe the combined system. When overlay systems intersect more than once (e.g., once overhead and once at base of riser), calculation of combined yield requires techniques used for looped and gridded systems. Once flow and pressure have been calculated back to the riser, N1.85 paper can be used to describe the combined system. EXISTING SYSTEMS - CONVERTING TO NEW SPECIFICATIONS When occupancy in existing buildings changes, new sprinkler system demands are generally specified. Sometimes new demands are obviously within sprinkler system capabilities. Other times new demands obviously require repiping and / or new water supplies. Often neither solution is obvious, and recalculation is needed to determine whether or not improvements are necessary. Typically, tree-type or loop-type sprinkler systems can be manually calculated, whereas other types of sprinkler systems require a computer software program for recalculation.
Table 15. Conversion Factors and Formulas VOLUME 1 ft3 = 1,728 in.3 = 7.48 U.S. gal = (0.0283 m3) 1 U.S. gal = 231 in.3 = 0.833 Imperial gal = (3.785 l) I Imperial gal = 1.2 U.S. gal = (4.542 l) 1 acre ft = 325,700 gal = (1233 m3) DISCHARGE 1 ft3/s = 448.3 gpm = 0.646 mgd = (1697 l/min) 1 million gal/day = 694.4 gpm = (2628 l/min) 1 gpm = 3.785 l/min) WEIGHT 1 gal water = 8.33 lb = (3.78 kg) 1 ft3 water = 62.4 lb = (28.3 kg) 1 m3 water = 2200 lb = (1000 kg) POWER 1 hp = 550 ft-lbf/s = 33,000 ft-lbf/min = 0.746 kW = (746 J) 1 kw = 1000 watts = 1.34 hp = (1000 J/s) PRESSURES 1 atmosphere = 14.7 psi = 29.9 in. mercury = 33.9 ft water = (1.014 bar = 101.4 kPa) 1 ft water = 0.881 in. mercury = 0.433 psi = (0.0299 bar = 2.99 kPa) 1 in. mercury = 1.135 ft water = 0.491 psi = (0.0339 bar = 3.39 kPa) 1 psi = 2.31 ft water = 27.7 in. water = 2.04 in. mercury = (.0689 bar = 6.89 kPa)
3-0
Page 58

DISTANCE 1 in. = 0.0833 ft = (25.4 mm) 1 ft = 12 in. = (0.3048 m) 1 mile = 5280 ft = (1.61 km) AREA 1 in.2 = 0.00694 ft2 = (645.16 mm2) 1 ft2 = 144 in.2 = (0.0929 m2) 1 acre = 43,560 ft2 = 0.00156 miles2 = (4047 m2) 1 mile2 = 640 acres = (2.59 km2) DISCHARGE DENSITY 1 gpm/ft2 = (40.7 mm/min). FRICTION LOSS 1 psi/ft = (0.226 bar/m = 22.6 kPa/m). K FACTORS Even though K factors are expressed without dimensions, they depend on units used to express flow and pressure. Orifices, K = Q P K Q P 1 gpm psi 14.4 l/min bar 1.44 l/min kPa Sprinkler systems, (K = Q/p0.54). K 1 16.1 1.33 HEAT 1 Btu = (1055 J) 1 Btu/lb = (2326 J/kg) TEMPERATURES C = 59 (F 32) K = C +273.17 F = 1.8 (C) +32 HYDRAULIC FORMULAS (Bernoullis equation) 1. v2/2g + p/w + z = H 2. v2/2g = hv (velocity head) 3. p/w = hp (pressure head) 4. va2/2g + pa/w + za = vb2/2g + pb/w + zb + hab or hab = (va2 - vb2)/2g + (pa - pb)/w + (za - zb) (Bernoullis equation modified for friction loss between points a and b) 5. Q = Av (flow rate) (Hazen-Williams formula) 6. p = cQ1.85/C1.85d4.87 7. Q = A 2ghv (orifice flow rate without discharge coefficient) (orifice flow rate) 8. Q = cA 2ghv 9. Q = acd2 Pv (orifice flow rate) (orifice flow rate) 10. Q = K Pv (orifice flow rate in testing apparatus) 11. Q = K Pn 12. Q = K Pt (orifice flow rate in testing apparatus when velocity is nearly zero) (pipe flow rate) 13. Q = Kp0.54 Q gpm l/min l/min P psi bar kPa

3-0
Page 59
14. Q2/Q1 = 15. 16. 17. 18. 19.
p2/p1
(proportionality of flow and pressure) (changing flow and pressure at a point in a piping system) (friction loss in pipe) (velocity pressure in pipe) (flow division in loop) (flow division in loop)
Q2/Q1 = (p2/p1)0.54 Pf = pfcL Pv = Q2/cd4 Q1/Q2 = (L2/L1)0.54 Q1/Q = 1/([L1/L2]0.54 + 1)
FIRE PUMPS Water horsepower = Qp/1710, Q in gpm and p is psi = (0.261 Qp, Q in l/min and p in bars) = (26.1 Qp, Q in l/min and p in kPa). Efficiency = water hp/input hp. Typical peak efficiency is 80 at 125 of rated flow.
There is no directly applicable NFPA standard. However, NFPA 13, Chapter 7, contains material on hyraulically designed sprinkler systems. NFPA standards do not take into consideration unusually corrosive water conditions when designing sprinkler systems.

3-00 Hydraulic Principles

Uploaded by

Copyright:

Available Formats

3-00 Hydraulic Principles

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

3-00 Hydraulic Principles

Uploaded by

Copyright:

Available Formats

FM Global Property Loss Prevention Data Sheets

HYDRAULICS OF FIRE PROTECTION SYSTEMS

Hydraulics of Fire Protection Systems

Fig. Fig. Fig. Fig. Fig. Fig.

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

Fig. Fig. Fig. Fig. Fig. Fig.

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

Fig. 1. Relationships between various hydraulic factors in typical piping

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

c 0.80 0.70 0.60 0.97 0.99 0.90 0.80

Fig. 2. Typical hydrant butts

outlet. Figure 2 shows three general types of hydrant outlets.

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

*For pressure in bars, multiply by 0.01.

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

(equation 14) (equation 15)

Q2/Q1 = p2/p1)0.54 , for r = 0.54

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

Hydraulics of Fire Protection Systems

Fig. 3. Water flow devices

Fig. 4. Typical pressure gauge locations

Fig. 5. Pitot tube with gauge and air chamber

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

Fig. 6. Taking pitot readings at hose nozzle

Fig. 7. Simplified, freehand line diagram of water supplies

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

Fig. 8. Single-source water supply

Fig. 9. Single source with underground loop

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

Fig. 10. Two sources with underground loop

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

Fig. 11. Typical public water system near a plant site

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems

2006 Factory Mutual Insurance Company. All rights reserved.

Hydraulics of Fire Protection Systems