Skip to content
From Fire Safety Academy Wiki
Revision as of 22:38, 7 November 2026 by Daan (talk | contribs) (Refine sentence flow and printed-page reference locators (written by GPT-6))
Contents
  1. 1Empirical coefficient and applicability
  2. 2Applying the Hazen-Williams relation
  3. 3Pipe resistance and internal condition
  4. 4Applying the Darcy-Weisbach relation
  5. 5Pressure, flow and energy
  6. 6Comparing supply and demand
  7. 7Corrosion, deposits and diagnosis
  8. 8See also
  9. 9Notes
  10. 10References

Hazen-Williams formula is an empirical equation for estimating frictional pressure loss in water-filled pipes from flow, internal diameter and a roughness-related coefficient. [1] Its use depends on a consistent unit convention and a defensible C-factor. The strong flow and diameter exponents make errors in demand or bore significant. The equation describes distributed resistance in its intended water-service range, rather than all fluid properties or every local obstruction. [1]

The underlying hydraulic assessment follows water from a defined source to its intended outlet. Elevation, velocity changes and irreversible losses must be distinguished. Observations also need the flow and network conditions under which they were obtained. A mathematically correct result can still be misleading if a source condition, connecting segment or component loss has been omitted from the description. [2][3] Static pressure and flowing pressure describe different conditions. [4]

Empirical coefficient and applicability

The formula is widely discussed in sprinkler hydraulic calculation because it conveniently relates pipe flow to friction loss. The NFPA 13 design methods described by the handbooks use that relationship within a larger network calculation. The equation itself does not select the fire hazard, operating sprinklers or supply duration. Those are separate inputs to the calculation. [5][6]

The Darcy-Weisbach equation provides a different representation using a friction factor tied to flow regime and relative roughness; it is useful when the fluid-property assumptions behind Hazen-Williams are unsuitable. Agreement cannot be forced by treating their coefficients as identical physical quantities. Titus's comparison example shows how different condition assumptions produce different predicted losses. [7]

A water supply curve often uses the same flow exponent as an approximation of network pressure drop; microbiologically influenced corrosion can change the physical conditions that a pipe coefficient is intended to represent. Deposits and local blockage require evidence about their location and mechanism. Changing C alone cannot demonstrate that the full installed water path remains adequate. [4][8]

Applying the Hazen-Williams relation

The Hazen-Williams relation represents pipe friction through an empirical water-flow equation; Titus explains that its coefficient is fitted to flow behaviour rather than derived from a complete fluid-property model. Water temperature and the range of pipe conditions therefore matter to its interpretation. It is not a general-purpose equation for every liquid used in fire protection. [1]

The following form expresses the pressure loss through a straight length in kilopascals; it is obtained by converting Titus's bar-per-metre form and multiplying by pipe length. Flow is in litres per minute and actual internal diameter is in millimetres. Changing those units without changing the constant produces a different and incorrect calculation. [9]

Δp=6.05×107LQ1.85C1.85D4.87

Symbol Description Unit
Δp,L Frictional pressure loss and pipe length kPa; m
Q,D Water flow and actual internal diameter L/min; mm
C Empirical Hazen-Williams coefficient dimensionless

Titus compares methods using a flow of approximately 1,893 L/min through 30.48 m of pipe with a 102.26 mm internal diameter. Using a C-factor of 100 gives approximately 69.4 kPa loss by this formula. The example demonstrates the method with stated inputs; it does not prescribe that coefficient for every steel pipe. Different condition assumptions produce different predictions for the same nominal pipe. [10]

Sensitivity can be illustrated without selecting another installation. Doubling flow at fixed diameter and C-factor multiplies the calculated loss by approximately 3.61. Reducing C from 120 to 80 multiplies loss by approximately 2.12 at unchanged flow. These recalculated ratios follow directly from the exponents and help explain why condition and demand assumptions matter. Titus cautions that Hazen-Williams becomes unsuitable for coefficients appreciably below 100, particularly for very rough pipes. The C-factor comparison therefore illustrates equation sensitivity rather than validating a severely deteriorated installation. [11][12]

The formula describes distributed resistance, not every possible obstruction; a partially closed valve or a lodged object can create a concentrated restriction that needs separate representation. Corrosion may affect both roughness and bore, while internal debris can move during testing or operation. Treating all these effects as an arbitrary change to C can hide the physical defect being investigated. [8][13]

Pipe resistance and internal condition

Moving water loses mechanical energy through interaction with pipe walls and through changes in direction or section; turbulent mixing contributes to this dissipation. The resistance depends on the actual internal bore and its condition, not simply the name printed on the pipe. Valves, elbows, entries and other fittings introduce additional local losses. [14]

The Darcy-Weisbach equation expresses distributed loss through a dimensionless friction factor; that factor depends on the flow regime and, for turbulent flow, on relative roughness. The Hazen-Williams formula instead uses an empirical coefficient for water service. These methods describe related hydraulic behaviour, but their coefficients do not have interchangeable meanings. [15]

At fixed diameter and roughness coefficient, Hazen-Williams loss increases approximately with flow raised to the power 1.85; doubling flow therefore produces considerably more than twice the loss. Increasing bore has a strong opposite effect because diameter appears with an exponent of 4.87. This explains why a short restricted connection can materially influence an otherwise generous water supply. [1][16]

Fittings can be represented by local resistance coefficients or by equivalent pipe lengths where the selected method permits; equivalent length is a calculation device, not an extra measured length on the drawing. Its use must remain consistent with the pipe and friction assumptions behind the published value. A special valve's measured pressure-loss curve may provide a better representation than a generic fitting allowance. [10][17]

Age and deposits introduce further uncertainty; corrosion can increase roughness, reduce the effective bore and release material that later obstructs a smaller passage. These mechanisms are distinct, so changing a roughness coefficient does not necessarily represent a local blockage. Inspection findings and flow measurements help determine whether the calculation still describes the installation. [8][13]

A sound comparison keeps elevation, flow and valve state consistent between observations; a high static pressure can coexist with poor flowing performance because frictional loss emerges when water moves. Conversely, a pressure difference caused by elevation does not establish pipe deterioration. The interpretation depends on the physical route and the condition under which each reading was taken. [18][4]

Applying the Darcy-Weisbach relation

The Darcy-Weisbach relation describes distributed pipe loss using mean velocity and a friction factor; the factor represents the effect of the flow regime and pipe condition. Reynolds number connects velocity and diameter with the fluid's kinematic viscosity. Relative roughness compares the wall's characteristic roughness with the internal diameter. [19]

The equation below gives head loss in metres of the flowing fluid; pressure loss follows by multiplying head loss by density and gravitational acceleration. The factor is the Darcy friction factor used in the cited handbook equation. Keeping its convention explicit prevents confusion with other friction-factor definitions. [20]

hf=fLDv22g,Re=vDν

Symbol Description Unit
hf,L,D Head loss, length and internal diameter m
f,Re Darcy friction factor and Reynolds number dimensionless
v,g Mean velocity and gravitational acceleration m/s; m/s²
ν Kinematic viscosity m²/s

For laminar pipe flow below a Reynolds number of about 2,000, Titus gives the factor as 64 divided by Reynolds number; transitional flow does not have the same simple certainty. In turbulent flow, roughness and Reynolds number determine the appropriate factor through correlations or a Moody diagram. A constant factor cannot be assumed valid through every change of flow or liquid condition. [15]

Titus's comparison example uses approximately 1,893 L/min in a 102.26 mm internal bore over 30.48 m. With the stated friction factor of 0.0188, recalculation gives approximately 4.21 m of head loss. These values are converted from the example's original customary units. The result differs from its Hazen-Williams example because the selected assumptions represent resistance differently. [10]

The method still requires a defensible pipe model; deposits, local restrictions and uncertain internal diameter can dominate uncertainty even when the formula is correct. Local fitting losses are added through appropriate coefficients or compatible equivalent lengths. Elevation changes remain energy terms rather than being counted as friction. [2][21]

Pressure, flow and energy

Water pressure describes only one part of the energy available in a pipe; elevation contributes potential energy, while movement contributes velocity head. Titus describes these quantities using head, expressed as an equivalent height of water. A hydraulic calculation must therefore compare locations through a common energy balance, rather than subtracting unrelated gauge readings. [22]

For steady, incompressible flow, the energy balance adds pump head and subtracts irreversible losses between two sections; the following form assumes that section-average velocity adequately represents kinetic energy. It describes a defined flow path, with all elevations measured from the same datum. Sudden changes require a different, transient analysis. [23]

p1ρg+v122g+z1+Hp=p2ρg+v222g+z2+hL

Symbol Description Unit
p Pressure at the section Pa
ρ Water density kg/m³
g Gravitational acceleration m/s²
v Mean velocity m/s
z,Hp,hL Elevation, added pump head and head loss m

Pressure can decrease because water rises, because velocity increases, or because energy is dissipated; these effects have different physical meanings. An elevation change can be recovered on a descending route, whereas friction dissipates mechanical energy. A larger pipe can reduce friction but cannot remove the elevation difference between the source and the outlet. [18]

Continuity provides the corresponding flow constraint: volume flow equals cross-sectional area multiplied by mean velocity; the same flow moves faster through a smaller internal bore. At a junction, incoming and outgoing flows balance, while losses depend on the flow assigned to each connected segment. Nominal pipe size alone does not supply the internal diameter needed for this calculation. [23][24]

The pressure at a sprinkler is the pressure remaining after the complete upstream route has been considered; it is not the pressure indicated beside a remote pump or street hydrant. The route includes connecting mains, valves, elevation and distribution piping. Leaving out an inconvenient section creates an apparently favourable calculation without providing the corresponding water at the outlet. [16][25]

Comparing supply and demand

Supply and demand curves answer complementary questions; a supply curve gives the pressure available at different flows under stated conditions. A demand calculation gives the pressure needed to deliver the selected flow through the protection system. Their comparison is meaningful only at a common reference point and for compatible operating assumptions. [3]

A common water-supply approximation treats the pressure drop from the test's static value as proportional to flow raised to the power 1.85. The relation can interpolate a measured pressure-flow pair and supports the traditional transformed-axis graph. It is an empirical representation of a network, not a physical guarantee at every extrapolated flow. [4][1]

p(Q)=ps−(ps−pr)(QQt)1.85

Symbol Description Unit
ps,pr Static and test residual pressures kPa
Qt Measured test discharge L/min
Q,p(Q) Investigated flow and predicted pressure L/min; kPa

Using the pressure and flow inputs in Fleming's printed example gives a useful arithmetic check; converted to SI, they are approximately 689 kPa static and 552 kPa residual at 3,785 L/min. At 1,703 L/min, the equation gives approximately 658 kPa. This recalculation uses the stated inputs; it does not reproduce the inconsistent intermediate answer printed in that example. [4]

A pump adds a further pressure-flow relationship; the available inlet pressure combines with the differential head developed by the pump, after appropriate elevation and velocity corrections. The resulting supply is not obtained by adding the rated pump pressure at every flow. Both high-flow delivery and low-flow maximum pressure need consideration. [26][27]

The calculation must also preserve the difference between capacity and duration; a source may meet a demand point initially while exhausting a finite reserve too soon. Two paths may appear redundant while sharing one vulnerable source or connecting segment. Pressure-flow agreement is therefore one part of demonstrating the performance of the complete supply arrangement. [28][29]

Corrosion, deposits and diagnosis

Microbiologically influenced corrosion describes microbial participation in corrosion processes; organisms and their biofilms can change local chemistry and establish conditions that promote metal loss. The presence of bacteria does not by itself prove that they caused a particular failure. The handbook explicitly distinguishes evidence of organisms from a defensible diagnosis of their role. [30]

Biofilms can create local differences in oxygen concentration and other chemical conditions; deposits may maintain an aggressive local environment even after a treatment reduces viable organisms. A high microbial count is therefore not a direct measurement of corrosion rate. A low count at one sampling time likewise does not reconstruct the previous history of an affected pipe. [8]

Both wet and dry sprinkler arrangements can be affected by internal deterioration; a dry system can retain water in low points and expose that water to oxygen in the gas supply. Water chemistry and deposits also matter in wet systems. Plastic components can experience biological fouling even though the same metallic corrosion mechanism does not apply to the plastic itself. [8][31]

The consequences need separate description; wall loss can lead to leakage or loss of integrity, while deposits increase resistance or reduce the effective bore. Detached material can migrate and obstruct a smaller downstream opening. No single roughness coefficient represents every combination of these effects. The hydraulic model and the assessment of physical condition therefore answer complementary questions. [21][15][13]

A useful investigation combines observations of the metal, deposits, water chemistry and microbiology; the handbook cautions against assigning cause from a single test result. Cleaning and treatment also require separate evaluation because removing organisms does not restore lost wall thickness. The remaining installation must still meet its structural and hydraulic functions. [21]

The maintenance implication is broader than finding a chemical treatment; source-water condition, retained water, internal debris and future inspection all affect continued performance. Treatment concentrations can be difficult to control in intermittently flowing sprinkler networks. The available handbook discussion explains these limitations without providing a current product-specific treatment specification. [32][33]

See also

Notes

  1. ↑ 1.0 1.1 1.2 1.3 1.4 Titus (2002), ch. 4-2, pp. 4-52–4-53.
  2. ↑ 2.0 2.1 Titus (2002), ch. 4-2, pp. 4-47–4-56.
  3. ↑ 3.0 3.1 Fleming (2002), ch. 4-3, pp. 4-81–4-83.
  4. ↑ 4.0 4.1 4.2 4.3 4.4 Fleming (2002), ch. 4-3, pp. 4-81.
  5. ↑ Fleming (2002), ch. 4-3, pp. 4-74–4-81.
  6. ↑ Cote (2008), ch. 16.3, pp. 16-47–16-49.
  7. ↑ Titus (2002), ch. 4-2, pp. 4-50–4-56.
  8. ↑ 8.0 8.1 8.2 8.3 8.4 Cote (2008), ch. 15.5, pp. 15-79–15-81.
  9. ↑ Titus (2002), ch. 4-2, pp. 4-52.
  10. ↑ 10.0 10.1 10.2 Titus (2002), ch. 4-2, pp. 4-56.
  11. ↑ Titus (2002), ch. 4-2, pp. 4-52–4-54.
  12. ↑ Cote (2008), ch. 15.3, pp. 15-53.
  13. ↑ 13.0 13.1 13.2 Cote (2008), ch. 16.11, pp. 16-212.
  14. ↑ Titus (2002), ch. 4-2, pp. 4-49–4-56.
  15. ↑ 15.0 15.1 15.2 Titus (2002), ch. 4-2, pp. 4-50–4-53.
  16. ↑ 16.0 16.1 Fleming (2002), ch. 4-3, pp. 4-75–4-81.
  17. ↑ Cote (2008), ch. 15.3, pp. 15-59.
  18. ↑ 18.0 18.1 Titus (2002), ch. 4-2, pp. 4-47–4-52.
  19. ↑ Titus (2002), ch. 4-2, pp. 4-49–4-52.
  20. ↑ Titus (2002), ch. 4-2, pp. 4-50–4-52.
  21. ↑ 21.0 21.1 21.2 Cote (2008), ch. 15.5, pp. 15-80–15-81.
  22. ↑ Titus (2002), ch. 4-2, pp. 4-45–4-48.
  23. ↑ 23.0 23.1 Titus (2002), ch. 4-2, pp. 4-47–4-50.
  24. ↑ Cote (2008), ch. 16.3, pp. 16-41–16-42.
  25. ↑ Hopkins et al. (2005), pp. xxxvii–xxxviii.
  26. ↑ Titus (2002), ch. 4-2, pp. 4-68–4-70.
  27. ↑ Cote (2008), ch. 15.6, pp. 15-89–15-94.
  28. ↑ Fleming (2002), ch. 4-3, pp. 4-83–4-84.
  29. ↑ Hopkins et al. (2005), pp. xxxvii–xlii.
  30. ↑ Cote (2008), ch. 15.5, pp. 15-75,15-78–15-81.
  31. ↑ Cote (2008), ch. 16.11, pp. 16-211–16-212.
  32. ↑ Cote (2008), ch. 15.5, pp. 15-81.
  33. ↑ National Research Council of Canada et al. (2005), pp. 1.7-8.

References

  • Cote, A. E. (Ed.). (2008). Fire protection handbook (20th ed.). National Fire Protection Association.
  • Fleming, R. P. (2002). Automatic sprinkler system calculations. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., pp. 4-72–4-87). National Fire Protection Association.
  • Hopkins, M., Schoenrock, J., & Budnick, E. (2005). Fire suppression systems (NIST NCSTAR 1-4B). National Institute of Standards and Technology.
  • National Research Council of Canada, International Code Council, Department of Building and Housing, & Australian Building Codes Board. (2005). International fire engineering guidelines. Australian Building Codes Board.
  • Titus, J. J. (2002). Hydraulics. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., pp. 4-44–4-71). National Fire Protection Association.