The following organizations in the United States issue codes and standards for piping systems and components:
Parallel federal specifications also have been developed by government agencies and are used for many public works projects. Chapter IV of ASME Standard B31.9 lists applicable U.S. codes and standards for HVAC piping. In addition, it gives requirements for safe design and construction of piping systems for building heating and air conditioning. ASME Standard B31.5 gives similar requirements for refrigerant piping.
1.2 DESIGN CONSIDERATIONS
Pipes are conduits in which fluids [compressible (e.g., air, steam) and noncompressible (e.g., water)] flow in a system, in response to a pressure differential. Piping system designers should assess the following aspects:
-
Code requirements.
-
Load: the amount of energy or fluid to be moved through the pipe to where it is needed; determination of load is not covered in this chapter (see Chapters 16 to 18 for information on load calculations).
-
Working fluid and fluid properties in the pipe.
-
Pressure and temperature of the fluid.
-
External environment of the pipe: outdoor installations deal with temperature extremes, environmental contaminants, and ultraviolet radiation. Other environments could contain caustic chemicals. Soil can contain elements that can be corrosive to underground pipe systems.
-
Installation cost.
-
Pipe’s resistance to chemical attack from the fluid.
When designing a fluid flow system, two related but distinct concerns emerge: sizing the pipe and determining the flow/pressure relationship. The two are often confused because they can use the same equations and design tools. Nevertheless, they should be determined separately.
This chapter focuses on sizing the pipe during the design phase, and to this end presents design charts and tables for specific fluids in addition to the equations that describe fluid flow in pipes. Once a system has been sized, it should be analyzed with more detailed methods of calculation to determine the pump head, if applicable, required to achieve the desired flow. Computerized methods are well suited to handling the details of calculating losses around an extensive system.
Not discussed in detail in this chapter, but of potentially great importance, are physical and chemical considerations such as pipe and fitting design; materials; and joining methods appropriate for working pressures and temperatures encountered, as well as resistance to chemical attack by the fluid. For more information, see Eshbach (2009), Heald (2002), and Nayyar (1999).
For fluids not included in this chapter or for piping materials of different dimensions, manufacturers’ literature frequently supplies pressure drop charts. The Darcy-Weisbach equation, with the Moody chart or Colebrook equation, can be used as an alternative to pressure drop charts or tables.
Each HVAC system and, under some conditions, portions of a system require a study of the conditions of operation to determine suitable materials. For example, because the static pressure of water in a high-rise building is higher in the lower levels than in the upper levels, a heavier pipe or different materials may be required for different vertical zones.
Table 1 lists some typical systems and materials used for heating and air-conditioning metallic piping. The list is not all inclusive, because piping systems are constantly being developed. The pressure and temperature rating of each component selected must be considered; the lowest rating establishes the operating limits of the system.
Nonmetallic (Plastic) Pipe Systems
Nonmetallic pipe is used in HVAC and plumbing. Plastic is light, generally inexpensive, and corrosion resistant. Plastic also has a low “C” factor (i.e., its surface is very smooth), resulting in lower pumping power requirements and smaller pipe sizes. Plastic pipe’s disadvantages include rapid loss of strength at temperatures above ambient and a high coefficient of linear expansion. The modulus of elasticity of plastics is low, resulting in a short support span. Some jurisdictions do not allow certain plastics in buildings because of toxic products emitted during fires. Plenum-rated plastic and insulation may be used to achieve a plenum rating; check with the authority having jurisdiction (AHJ).
Table 2 lists nonmetallic materials used for service water and heating and air-conditioning piping. The pressure and temperature rating of each component selected must be considered; the lowest rating establishes the operating limits of the system.
Some piping systems are governed by separate codes or standards. Generally, any failure of the piping in these systems is dangerous to the public, so some local areas have adopted laws enforcing the codes, such as the following:
-
Boiler piping: ASME Standard B31.1 and the ASME Boiler and Pressure Vessel Code (Section I) specify piping inside code-required stop valves on boilers that operate above 15 psig with steam, or above 160 psig or 250°F with water. These codes require fabricators and contractors to be certified for such work. The field or shop work must also be inspected while it is in progress, by inspectors commissioned by the National Board of Boiler and Pressure Vessel Inspectors.
-
Refrigeration piping: ASHRAE Standard 15 and ASME Standard B31.5.
-
Plumbing systems: Local codes.
-
Sprinkler systems: NFPA Standard 13.
-
Fuel gas: NFPA Standard 54/ANSI Standard Z223.1.
Pressure drop caused by fluid friction in fully developed flows of all well-behaved (Newtonian) fluids is described by the Darcy-Weisbach equation:
where
| Δp |
= |
pressure drop, lbf/ft2 |
| f |
= |
friction factor, dimensionless (from Moody chart, Figure 13 in Chapter 3) |
| L |
= |
length of pipe, ft |
| D |
= |
internal diameter of pipe, ft |
| ρ |
= |
fluid density at mean temperature, lbm/ft3 |
| V |
= |
average velocity, fps |
| gc |
= |
units conversion factor, 32.2 ft · lbm/lbf · s2 |
This equation is often presented in head or specific energy form as
where
| Δh |
= |
head loss, ft |
| g |
= |
acceleration of gravity, ft/s2 |
In this form, the fluid’s density does not appear explicitly (although it is in the Reynolds number that influences f).
The friction factor f is a function of pipe roughness ε, inside diameter D, and parameter Re, the Reynolds number:
where
| Re |
= |
Reynolds number, dimensionless |
| ε |
= |
absolute roughness of pipe wall, ft |
| μ |
= |
dynamic viscosity of fluid, lbm/ft · s |
The friction factor is frequently presented on a Moody chart (Figure 13 in Chapter 3) giving f as a function of Re with ε/D as a parameter.
A useful fit of smooth and rough pipe data for the usual turbulent flow regime is the Colebrook equation:
Another form of Equation (4) appears in Chapter 21, but the two are equivalent. Equation (4) is useful in showing behavior at limiting cases: as ε/D approaches 0 (smooth limit), the 18.7/Re
term dominates; at high ε/D and Re (fully rough limit), the 2ε/D term dominates.
Equation (4) is implicit in f; that is, f appears on both sides, so a value for f is usually obtained iteratively.
A less widely used alternative to the Darcy-Weisbach formulation for calculating pressure drop is the Hazen-Williams equation, which is expressed as
or
where C = roughness factor.
Typical values of C are 150 for plastic pipe and copper tubing, 140 for new steel pipe, down to 100 and below for badly corroded or very rough pipe.
Valves and fittings cause pressure losses greater than those caused by the pipe alone. One formulation expresses losses as
where K = geometry- and size-dependent loss coefficient (Tables 3 to 6).
Example 1.
Determine the pressure drop for 60°F water flowing at 4 fps through a nominal 1 in., 90° threaded elbow.
Solution: Use Equation (7). From Table 3, the K for a 1 in., 90° threaded elbow is 1.5.
The loss coefficient for valves appears in another form as Cv, a dimensional coefficient expressing the flow through a valve at a specified pressure drop.
where
| Q |
= |
volumetric flow, gpm |
| Cv |
= |
valve coefficient, gpm at Δp = 1 psi |
| Δp |
= |
pressure drop, psi |
See the section on Control Valve Sizing in Chapter 47 of the 2020 ASHRAE Handbook—HVAC Systems and Equipment for more information on valve coefficients.
Example 2.
Determine the volumetric flow through a valve with Cv = 10 for an allowable pressure drop of 5 psi.
Solution: Use Equation (8).
Alternative formulations express fitting losses in terms of equivalent lengths of straight pipe (see Tables 8 and 27). Pressure loss data for fittings are also presented in Idelchik (1986).
Equation (7) and data in Tables 3 and 4 are based on the assumption that separated flow in the fitting causes the K factors to be independent of Reynolds number. In reality, the K factor for most pipe fittings varies with Reynolds number. Tests by Rahmeyer (1999a, 1999b, 2002a, 2002b) (ASHRAE research projects RP-968 and RP-1034) on 2 in. threaded and 4, 12, 16, 20, and 24 in. welded steel fittings demonstrate the variation and are shown in Tables 6 and 7. The studies also present K factors of diverting and mixing flows in tees, ranging from full through flow to full branch flow. They also examined the variation in K factors caused by variations in geometry among manufacturers and by surface defects in individual fittings.
Hegberg (1995) and Rahmeyer (1999a, 1999b) discuss the origins of some of the data shown in Tables 6 and 7. The Hydraulic Institute (1990) data appear to have come from Freeman (1941), work that was actually performed in 1895. The work of Giesecke (1926) and Giesecke and Badgett (1931, 1932a, 1932b) may not be representative of present-day fittings.
Further extending the work on determination of fitting K factors to PVC piping systems, Rahmeyer (2003a, 2003b) (ASHRAE research project RP-1193) found the data in Tables 8 and 9 giving K factors for Schedule 80 PVC 2, 4, 6, and 8 in. ells, reducers, expansions, and tees. The results of these tests are also presented in the cited papers in terms of equivalent lengths. In general, PVC fitting geometry varied much more from one manufacturer to another than steel fittings did.
Losses in Multiple Fittings
Typical fitting loss calculations are done as if each fitting is isolated and has no interaction with any other. Rahmeyer (2002c) (ASHRAE research project RP-1035) tested 2 in. threaded ells and 4 in. ells in two and three fitting assemblies of several geometries, at varying spacing. Figure 1 shows the geometries, and Figures 2 and 3 show the ratio of coupled K values to uncoupled K values (i.e., fitting losses for the assembly compared with the sum of losses from the same number of isolated fittings).
The most important conclusion is that the interaction between fittings always reduces the loss. Also, although geometry of the assembly has a definite effect, the effects are not the same for 2 in. threaded and 4 in. welded ells. Thus, the traditional practice of adding together losses from individual fittings gives a conservative (high-limit) estimate.
Calculating Pressure Losses
The most common engineering design flow loss calculation selects a pipe size for the desired total flow rate and available or allowable pressure drop.
Because either formulation of fitting losses requires a known diameter, pipe size must be selected before calculating the detailed influence of fittings. A frequently used rule of thumb assumes that the design length of pipe is 50 to 100% longer than actual to account for fitting losses. After a pipe diameter has been selected on this basis, the influence of each fitting can be evaluated.
Metallic Pipe. Although stress calculations are seldom required, the factors involved should be understood. The main areas of concern are (1) internal pressure stress, (2) longitudinal stress caused by pressure and weight, and (3) stress from expansion and contraction.
ASME Standard B31 standards establish a basic allowable stress S equal to one-fourth of the minimum tensile strength of the material. This value is adjusted, as discussed in this section, because of the nature of certain stresses and manufacturing processes.
Hoop stress caused by internal pressure is the major stress on pipes. Because some forming methods form a seam that may be weaker than the base material, ASME Standard B31.9 specifies a joint efficiency factor E, multiplied by the basic allowable stress to establish a maximum allowable stress value in tension SE. (Table A-1 in ASME Standard B31.9 lists values of SE for commonly used pipe materials.) The joint efficiency factor can be significant; for example, seamless pipe has a joint efficiency factor of 1, so it can be used to the full allowable stress (one-quarter of the tensile strength). In contrast, butt-welded pipe has a joint efficiency factor of 0.60, so its maximum allowable stress must be derated (SE = 0.6S).
Equation (9) determines the minimum wall thickness for a given pressure. Equation (10) determines the maximum pressure allowed for a given wall thickness.
where
| SA |
= |
allowable stress range, psi |
| Sc |
= |
allowable cold stress at coolest temperature system will experience, psi |
| Sh |
= |
allowable hot stress at hottest temperature system will experience, psi |
Both equations incorporate an allowance factor A to compensate for manufacturing tolerances, material removed in threading or grooving, and corrosion. For the seamless, butt-welded, and electric resistance welded (ERW) pipe most commonly used in HVAC work, the standards apply a manufacturing tolerance of 12.5%. Working pressure for steel pipe (see Table 16) has been calculated using a manufacturing tolerance of 12.5%, standard allowance for depth of thread (where applicable), and a corrosion allowance of 0.065 in. for pipes 2 1/2 in. and larger and 0.025 in. for pipes 2 in. and smaller. Where corrosion is known to be greater or smaller, pressure rating can be recalculated using Equation (10). Higher pressure ratings than shown in Table 16 can be obtained (1) by using ERW or seamless pipe in lieu of continuous-weld (CW) pipe 4 in. and less, and seamless pipe in lieu of ERW pipe 5 in. and greater (because of higher joint efficiency factors); or (2) by using heavier-wall pipe.
Longitudinal stresses caused by pressure, weight, and other sustained forces are additive, and the sum of all such stresses must not exceed the basic allowable stress S at the highest temperature at which the system will operate. Longitudinal stress caused by pressure equals approximately one-half the hoop stress caused by internal pressure; thus, at least one-half the basic allowable stress is available for weight and other sustained forces. This factor is taken into account in Table 11.
Stresses caused by expansion and contraction are cyclical, and, because creep allows some stress relaxation, the ASME Standard B31 series allows designing to an allowable stress range SA as calculated by Equation (11). Table 15 lists allowable stress ranges for commonly used piping materials.
where
| SA |
= |
allowable stress range, psi |
| Sc |
= |
allowable cold stress at coolest temperature system will experience, psi |
| Sh |
= |
allowable hot stress at hottest temperature system will experience, psi |
Nonmetallic. Both thermoplastics and thermosets have an allowable stress derived from a hydrostatic design basis stress (HDBS). The HDBS is determined by a statistical analysis of both static and cyclic stress rupture test data as set forth in ASTM Standard D2837 for thermoplastics and ASTM Standard D2992 for glass-fiber-reinforced thermosetting resins.
The allowable stress, called the hydrostatic design stress (HDS), is obtained by multiplying the HDBS by a service factor. HDS values recommended by some manufacturers and those allowed by ASME Standard B31 are listed in Table 18.
The pressure design thickness for plastic pipe can be calculated using the code stress values and the formula in Equation (12):
where
| t |
= |
pressure design thickness, in. |
| p |
= |
internal design pressure, psig |
| D |
= |
pipe outside diameter, in. |
| S |
= |
hydrostatic design stress (HDS), psi |
The minimum required wall thickness can be found by adding an allowance for mechanical strength, threading, grooving, erosion, and corrosion to the calculated pressure design thickness.
Another method of rating pressure rating of piping used by manufacturers is the standard dimension ratio (SDR), which is the ratio of the pipe diameter to the wall thickness:
where
| D |
= |
pipe outside diameter, in. |
| s |
= |
pipe wall thickness, in. |
An SDR of 11 means that the outside diameter D of the pipe is 11 times the thickness of the wall s. A high SDR means that the pipe’s wall is thin compared to its diameter, and a low SDR means that the pipe’s wall is thick relative to pipe diameter. SDR is inversely correlated with pressure rating: high SDR indicates a low pressure rating, whereas low-SDR pipes have higher pressure ratings.
There are many formulations of the polymers used for piping materials, and different joining methods for each, so manufacturers’ recommendations should be followed. Most catalogs give pressure ratings for pipe and fittings at various temperatures up to the maximum the material will withstand.
A procedure for sizing piping systems is as follows:
-
Determine system type (open, closed, compressible, incompressible, pumped, gravity feed, domestic, etc.).
-
Determine type and properties of fluid to be conveyed in the pipe.
-
Determine temperatures used (high, low) and temperature differentials.
-
Identify system pressures encountered in the system (working, maximum, low, fill, and relief pressures).
-
Determine load at each device (e.g., heating or cooling requirements, fixture units for plumbing) to find flow.
-
Sketch main, risers, and branches, and indicate equipment to be served and each device’s flow rate.
-
Determine flow of supply pipe for each pipe segment run by summing the loads at the furthest device and running back to the source.
-
Determine flow of each return pipe by starting at the first device returning water and summing the loads back to the source (when applicable).
-
Determine equivalent length of pipe in the main lines, risers, branches, and returns. Because pipe sizes are not known, the exact equivalent length of various fittings cannot be determined. Add the equivalent lengths, starting at the beginning and proceeding along the mains, risers, branches, and returns (when applicable).
-
In domestic or gravity feed: calculate the approximate design value of the average pressure drop per 100 ft of equivalent length of pipe determined in step 9. In pumped system: calculate pressure drop or head H using the flow rate and pressure drop for pipe from Equations (2) or (6), the valves and fittings using head drop from Equation (7), and head from the devices from the manufacturer’s data.
where
| Δp |
= |
average pressure loss per 100 ft of equivalent length of pipe, psi |
| ps |
= |
pressure at the source, psig |
| pf |
= |
minimum pressure required to operate device, psig |
| pm |
= |
pressure drop through any meters, psi |
| H |
= |
height of highest fixture above source (if open system), ft |
| L |
= |
equivalent length determined in step 4, ft |
-
In domestic or gravity system: from the expected rate of flow (step 5) and Δp (step 10), select pipe sizes. In pumped system: select the pump using the flow rate and calculated H.
1.6 PIPE-SUPPORTING ELEMENTS
Pipe-supporting elements consist of (1) hangers, which support from above; (2) supports, which bear load from below; and (3) restraints, such as anchors and guides, that limit or direct movement as well as support loads. Pipe-supporting elements must withstand all static and dynamic conditions, including the following:
-
Weight of pipe, valves, fittings, insulation, and fluid contents, including test fluid if using heavier-than-normal media
-
Occasional loads such as ice, wind, and seismic forces or testing loads (e.g., hydrostatic loads on a steam pipe)
-
Forces imposed by thermal expansion and contraction of pipe bends and loops
-
Frictional, spring, and pressure thrust forces imposed by expansion joints in the system
-
Frictional forces of guides and supports
-
Other loads (e.g., water hammer, vibration, reactive force of relief valves)
-
Test load and force
In addition, pipe-supporting elements must be evaluated in terms of stress at the points of connection to the pipe and the building. Stress at the point of connection to the pipe is especially important for base elbow and trunnion supports, because this stress is usually the limiting parameter, not the strength of the structural member. Loads on anchors, cast-in-place inserts, and other attachments to concrete should not be more than one-fifth the ultimate strength of the attachment, as determined by manufacturers’ tests. All loads on the structure should be communicated to and coordinated with the structural engineer.
The ASME B31 standards establish criteria for the design of pipe-supporting elements, and the Manufacturers Standardization Society of the Valve and Fittings Industry (MSS) has established standards for the design, fabrication, selection, and installation of pipe hangers and supports based on these codes.
MSS Standard SP-69 and the catalogs of many manufacturers illustrate the various hangers and components and provide information on the types to use with different pipe systems. Table 10 lists maximum safe loads for threaded steel rods, and Tables 11 and 12 show suggested pipe support spacing for metal and PVC pipes, respectively.
Loads on most pipe-supporting elements are moderate and can be selected safely in accordance with manufacturers’ catalog data and the information presented in this section; however, some loads and forces can be very high, especially in multistory buildings and for large-diameter pipe, particularly where expansion joints are used at a high operating pressure. Consequently, a qualified engineer should design or review all anchors and pipe-supporting elements, especially for the following:
-
Steam systems operating above 15 psig
-
Hydronic systems operating above 160 psig or 250°F
-
Risers over 10 stories or 100 ft
-
Systems with expansion joints, especially for pipe diameters 3 in. and greater
Hanger Spacing and Pipe Wall Thickness
Table 11 suggests minimum pipe hanger spacing for use unless exceeded by the local authority having jurisdiction or engineering calculations. The primary factors determining pipe wall thickness are hoop stress caused by internal pressure, and longitudinal stresses caused by pressure, weight, and other sustained loads. Detailed stress calculations are seldom required for HVAC applications because standard pipe has ample thickness to sustain the pressure and longitudinal stress caused by weight (assuming hangers are spaced in accordance with Table 11).
Support spacings for PVC and CPVC pipe systems are influenced by operating temperatures. Table 12 recommends horizontal spacing based on pipe size, schedule, material (PVC or industrial-grade CPVC), and operating temperature. Hangers and supports should not be clamped tightly because the axial movement of the pipe would be restricted. The charts are based on continuous spans and uninsulated lines carrying liquids. They are not applicable where loads between supports are concentrated (e.g., for valves, flanges) or where there is a change in direction. Hangers/supports should be located adjacent to joints, branch connections, and changes in direction. Risers should be in installed independently of adjacent horizontal hangers/supports.
For cast iron pipe, maximum spacing should be 12 ft, with at least one hanger/support for each pipe section.
1.7 PIPE EXPANSION AND FLEXIBILITY
Temperature changes cause dimensional changes in all materials. Table 13 shows the coefficients of expansion for metallic piping materials commonly used in HVAC. For systems operating at high temperatures, such as steam and hot water, the rate of expansion is high, and significant movements can occur in short runs of piping. Even though rates of expansion may be low for systems operating in the range of 40 to 100°F, such as chilled and condenser water, they can cause large movements in long runs of piping, which are common in distribution systems and high-rise buildings. Therefore, in addition to design requirements for pressure, weight, and other loads, piping systems must accommodate thermal and other movements to prevent the following:
An unrestrained pipe operates at the lowest overall stress level. Anchors and restraints are needed to support pipe weight and to protect equipment connections. Anchor forces and bowing of pipe anchored at both ends are generally too large to be acceptable, so general practice is to never anchor a straight run of steel pipe at both ends. Piping must be allowed to expand or contract through thermal changes. Ample flexibility can be attained by designing pipe bends and loops or by including supplemental devices, such as expansion joints.
End reactions transmitted to rotating equipment, such as pumps or turbines, may deform the equipment case and cause bearing misalignment that may ultimately cause the component to fail. Consequently, manufacturers’ recommendations on allowable forces and movements that may be placed on their equipment should be followed.
Detailed stress analysis requires involved mathematical analysis and is generally performed by computer programs. However, such involved analysis is not typically required for most HVAC systems because the piping arrangements and temperature ranges at which they operate are usually simple to analyze. Expansion stresses discussed in this section relate only to aboveground pipe located in open air, or preinsulated pipe.
The guided cantilever beam method of evaluating L bends can be used to design L bends, Z bends, pipe loops, branch take-off connections, and some more complicated piping configurations. The guided cantilever equation [see Equation (17)] is generally conservative because it assumes that the pipe arrangement does not rotate. The anchor force results will be higher because of the lack of rotation, and rigorous analysis is recommended for complicated or expensive systems.
Equation (15) may be used to calculate the length of leg BC needed to accommodate thermal expansion or contraction of leg AB for a guided cantilever beam (Figure 4).
where
| L |
= |
length of leg BC required to accommodate thermal expansion of long leg AB, ft |
| Δ |
= |
thermal expansion or contraction of leg AB, in. |
| D |
= |
actual pipe outside diameter, in. |
| E |
= |
modulus of elasticity, psi |
| SA |
= |
allowable stress range, psi |
For the commonly used A53 Grade B seamless or ERW pipe, an allowable stress SA of 22,500 psi (see Table 15) can be used without overstressing the pipe. However, this can result in very high end reactions and anchor forces, especially with large-diameter pipe. Designing to a stress range SA of 15,000 psi and assuming E = 27.9 × 106 psi, Equation (15) reduces to Equation (16), which provides reasonably low end reactions without requiring too much extra pipe. In addition, Equation (16) may be used with A53 continuous (butt-) welded, seamless, and ERW pipe, and B88 drawn copper tubing.
The guided cantilever method of designing L bends assumes no restraints; therefore, care must be taken in supporting the pipe. For horizontal L bends, it is usually necessary to place a support near point B (see Figure 4), and any supports between points A and C must provide minimal resistance to piping movement; this is done by using slide plates or hanger rods of ample length, with hanger components selected to allow for swing no greater than 4°.
For L bends containing both vertical and horizontal legs, any supports on the horizontal leg must be spring hangers designed to support the full weight of pipe at normal operating temperature with a maximum load variation of 25%.
The force developed in an L bend that must be sustained by anchors or connected equipment is determined by the following equation:
where
| F | = | force, lb |
| Ec | = | modulus of elasticity, psi |
| I | = | moment of inertia, in4 |
| L | = | length of offset leg, ft |
| Δ | = | deflection of offset leg, in. |
Z bends, as shown in Figure 5, are very effective for accommodating pipe movements. A simple and conservative method of sizing Z bends is to design the offset leg to be 65% of the values used for an L bend in Equation (15):
where
| L |
= |
length of offset leg, ft |
| Δ |
= |
anchor-to-anchor expansion, in. |
| D |
= |
pipe outside diameter, in. |
The force developed in a Z bend can be calculated with acceptable accuracy as follows:
where
| C1 |
= |
4000 lb/in. |
| F |
= |
force, lb |
| D |
= |
pipe outside diameter, in. |
| L |
= |
length of offset leg, ft |
| Δ |
= |
anchor-to-anchor expansion, in. |
Pipe loops or U bends are commonly used in long runs of piping. A simple method of designing pipe loops is to calculate the anchor-to-anchor expansion and, using Equation (15), determine the length L necessary to accommodate this movement. The pipe loop dimensions can then be determined using W = L/5 and H = 2W.
Note that guides must be spaced no closer than twice the height of the loop, and piping between guides must be supported, as described in the section on L Bends, when the length of pipe between guides exceeds the maximum allowable hanger spacing for the size pipe.
Table 14 lists pipe loop dimensions for pipe sizes 1 to 24 in. and anchor-to-anchor expansion (contraction) of 2 to 12 in.
No simple method has been developed to calculate pipe loop force; however, it is generally low. A conservative estimate is 200 lb per inch diameter (e.g., a 2 in. pipe will develop 400 lb of force and a 12 in. pipe will develop 2400 lb of force). Additional analysis should be done for pipes greater than 12 in. in diameter, because other simplified methodologies predict higher anchor forces.
Expansion and Contraction Control of Other Materials
To design expansion and contraction loops and bends for other materials, consult the Copper Development Association (CDA 2010) for copper pipes, and Plastic Pipe and Fitting Association (PPFA 2009) for plastic pipes.
Cold springing or cold positioning of pipe consists of offsetting or springing the pipe in a direction opposite the expected movement. Cold springing is not recommended for most HVAC piping. Furthermore, cold springing does not allow designing a pipe bend or loop for twice the calculated movement. For example, if a particular L bend can accommodate 3 in. of movement from a neutral position, cold springing does not allow the L bend to accommodate 6 in. of movement.
Analyzing Existing Piping Configurations
Piping is best analyzed using a computer stress analysis program, which can provide all pertinent data, including stress, movements, and loads. Services can perform such analysis if programs are not available in house. However, many situations do not require such detailed analysis. A simple, satisfactory method for single and multiplane systems is to divide the system with real or imaginary anchors into a number of single-plane units, as shown in Figure 6, that can be evaluated as L and Z bends.
Stewart and Dona (1987) surveyed the literature relating to water flow rate limitations. Noise, erosion, and installation and operating costs all limit the maximum and minimum velocities in piping systems. If piping sizes are too small, noise levels, erosion levels, and pumping costs can be unfavorable. If piping sizes are too large, installation costs are excessive. Therefore, pipe sizes are chosen to minimize initial cost while avoiding the undesirable effects of high velocities. ASHRAE Standard 90.1 has been accepted by authorities having jurisdiction (AHJs) as a code and, as such, limits the flow for energy conservation. The table (Table 21) is reproduced with modification showing velocity limitations.
Various upper limits of water velocity and/or pressure drop in piping and piping systems are used. One recommendation places a velocity limit of 4 fps for 2 in. pipe and smaller, and a pressure drop limit of 4 ft of water/100 ft for piping over 2 in. Other guidelines are based on the type of service (Table 22) or annual operating hours (Table 23). These limitations are imposed either to control the levels of pipe and valve noise, erosion, and water hammer pressure or for economic reasons. Carrier (1960) recommends that the velocity not exceed 15 fps in any case.
Velocity-dependent noise in piping and piping systems results from any or all of four sources: turbulence, cavitation, release of entrained air, and water hammer. In investigations of flow-related noise, Ball and Webster (1976), Marseille (1965), and Rogers (1953, 1954, 1956) reported that velocities on the order of 10 to 17 fps lie within the range of allowable noise levels for residential and commercial buildings. The experiments showed considerable variation in noise levels obtained for a specified velocity. Generally, systems with longer pipe and with more numerous fittings and valves were noisier. In addition, sound measurements were taken under widely differing conditions; for example, some tests used plastic-covered pipe, whereas others did not. Thus, no detailed correlations relating sound level to flow velocity in generalized systems are available.
Noise generated by fluid flow in a pipe increases sharply if cavitation or release of entrained air occurs. Usually, the combination of high water velocity with a change in flow direction or a decrease in pipe cross section, causing a sudden pressure drop, is necessary to cause cavitation. Ball and Webster (1976) found that at their maximum velocity of 42 fps, cavitation did not occur in straight 3/8 and 1/2 in. pipe; using the apparatus with two elbows, cold-water velocities up to 21 fps caused no cavitation. Cavitation did occur in orifices of 1:8 area ratio (orifice flow area is one-eighth of pipe flow area) at 5 fps and in 1:4 area ratio orifices at 10 fps (Rogers 1954).
Some data are available for predicting hydrodynamic (liquid) noise generated by control valves. The International Society of Automation compiled prediction correlations in an effort to develop control valves for reduced noise levels (ISA 2007). The correlation to predict hydrodynamic noise from control valves is
where
| SL |
= |
sound level, dB |
| Cv |
= |
valve coefficient, gpm/(psi)0.5 |
| Q |
= |
flow rate, gpm |
| Δp |
= |
pressure drop across valve, psi |
| t |
= |
downstream pipe wall thickness, in. |
Air entrained in water usually has a higher partial pressure than the water. Even when flow rates are small enough to avoid cavitation, the release of entrained air may create noise. Every effort should be made to vent the piping system or otherwise remove entrained air.
Erosion in piping systems is caused by water bubbles, sand, or other solid matter impinging on the inner surface of the pipe. Generally, at velocities lower than 10 fps, erosion is not significant as long as there is no cavitation. When solid matter is entrained in the fluid at high velocities, erosion occurs rapidly, especially in bends. Thus, high velocities should not be used in systems where sand or other solids are present or where slurries are transported.
With age, the internal surfaces of pipes become increasingly rough. This reduces the available flow with a fixed pressure supply. However, designing with excessive age allowances may result in oversized piping. Age-related decreases in capacity depend on type of water, type of pipe material, temperature of water, and type of system (open or closed) and include
-
Sliming (biological growth or deposited soil on the pipe walls): occurs mainly in unchlorinated, raw water systems.
-
Caking of calcareous salts: occurs in hard water (i.e., water bearing calcium salts) and increases with water temperature.
-
Corrosion (incrustations of ferrous and ferric hydroxide on the pipe walls): occurs in metal pipe in soft water. Because oxygen is necessary for corrosion to take place, significantly more corrosion takes place in open systems.
Allowances for expected decreases in capacity are sometimes treated as a specific amount (percentage). Dawson and Bowman (1933) added an allowance of 15% friction loss to new pipe (equivalent to an 8% decrease in capacity). The HDR Design Guide (1981) increased the friction loss by 15 to 20% for closed piping systems and 75 to 90% for open systems. Carrier (1960) indicates a factor of approximately 1.75 between friction factors for closed and open systems.
Obrecht and Pourbaix (1967) differentiated between the corrosive potential of different metals in potable water systems and concluded that iron is the most severely attacked, then galvanized steel, lead, copper, and finally copper alloys (e.g., brass). Freeman (1941) and Hunter (1941) showed the same trend. After four years of cold- and hot-water use, copper pipe had a capacity loss of 25 to 65%. Aged ferrous pipe has a capacity loss of 40 to 80%. Smith (1983) recommended increasing the design discharge by 1.55 for uncoated cast iron, 1.08 for iron and steel, and 1.06 for cement or concrete.
The Plastic Pipe Institute (1971) found that corrosion is not a problem in plastic pipe; the capacity of plastic pipe in Europe and the United States remains essentially the same after 30 years in use.
Extensive age-related flow data are available for use with the Hazen-Williams empirical equation. Difficulties arise in its application, however, because the original Hazen-Williams roughness coefficients are valid only for the specific pipe diameters, water velocities, and water viscosities used in the original experiments. Thus, when the Cs are extended to different diameters, velocities, and/or water viscosities, errors of up to about 50% in pipe capacity can occur (Sanks 1978; Williams and Hazen 1933).
When any moving fluid (not just water) is abruptly stopped, as when a valve closes suddenly, large pressures can develop. Although detailed analysis requires knowledge of the elastic properties of the pipe and the flow-time history, the limiting case of rigid pipe and instantaneous closure is simple to calculate. Under these conditions,
where
| Δph |
= |
pressure rise caused by water hammer, lbf/ft2 |
| ρ |
= |
fluid density, lbm/ft3 |
| cs |
= |
velocity of sound in fluid, fps |
| V |
= |
fluid flow velocity, fps |
The cs for water is 4720 fps, although the pipe’s elasticity reduces the effective value.
Example 3.
What is the maximum pressure rise if water flowing at 10 fps is stopped instantaneously?
Solution: Δph = 62.4 × 4720 × 10/32.2 = 91,468 lb/ft2 = 635 psi
Sizing service water piping differs from sizing process lines in that design flows in service water piping are determined by the probability of simultaneous operation of multiple individual loads such as water closets, urinals, lavatories, sinks, and showers. The full-flow characteristics of each load device are readily obtained from manufacturers; however, service water piping sized to handle all load devices simultaneously would be seriously oversized. Thus, a major issue in sizing service water piping is to determine the diversity of the loads.
The procedure shown in this chapter uses the work of R.B. Hunter for estimating diversity (Hunter 1940, 1941). The present-day plumbing designer is usually constrained by building or plumbing codes, which specify the individual and collective loads to be used for pipe sizing. Frequently used codes (including the ICC International Plumbing Code and the PHCC National Standard Plumbing Code) contain procedures quite similar to those shown here. The designer must be aware of the applicable code for the location being considered.
Federal mandates are forcing plumbing fixture manufacturers to reduce design flows to many types of fixtures, but these may not yet be included in locally adopted codes. Also, the designer must be aware of special considerations; for example, toilet usage at sports arenas will probably have much less diversity than codes allow and thus may require larger supply piping than the minimum specified by codes.
Table 24 gives the rate of flow desirable for many common fixtures and the average pressure necessary to give this rate of flow. Pressure varies with fixture design.
In estimating load, the rate of flow is frequently computed in fixture units that are relative indicators of flow. Table 25 gives the demand weights in terms of fixture units for different plumbing fixtures under several conditions of service, and Figure 9 gives the estimated demand corresponding to any total number of fixture units. Figures 10 and 11 provide more accurate estimates at the lower end of the scale.
The estimated demand load for fixtures used intermittently on any supply pipe can be obtained by multiplying the number of each kind of fixture supplied through that pipe by its weight from Table 25, adding the products, and then referring to the appropriate curve of Figure 9, 10, or 11 to find the demand corresponding to the total fixture units. In using this method, note that the demand for fixture or supply outlets other than those listed in the table of fixture units is not yet included in the estimate. The demands for outlets (e.g., hose connections and air-conditioning apparatus) that are likely to impose continuous demand during heavy use of the weighted fixtures should be estimated separately and added to demand for fixtures used intermittently to estimate total demand.
The Hunter curves in Figures 9, 10, and 11 are based on use patterns in residential buildings and can be erroneous for other usages such as sports arenas. Williams (1976) discusses the Hunter assumptions and presents an analysis using alternative assumptions.
So far, the information presented shows the design rate of flow to be determined in any particular section of piping. The next step is to determine the size of piping. As water flows through a pipe, the pressure continually decreases along the pipe because of loss of energy from friction. The problem is then to ascertain the minimum pressure in the street main and the minimum pressure required to operate the topmost fixture. (A pressure of 15 psig may be ample for most flush valves, but manufacturers’ requirements should be consulted. Some fixtures require a pressure up to 25 psig. A minimum of 8 psig should be allowed for other fixtures.) The pressure differential overcomes pressure losses in the distributing system and the difference in elevation between the water main and the highest fixture.
The pressure loss (in psi) resulting from the difference in elevation between the street main and the highest fixture can be obtained by multiplying the difference in elevation in feet by the conversion factor 0.434.
Pressure losses in the distributing system consist of pressure losses in the piping itself, plus the pressure losses in the pipe fittings, valves, and the water meter, if any. Approximate design pressure losses and flow limits for disk-type meters for various rates of flow are given in Figure 12. Water authorities in many localities require compound meters for greater accuracy with varying flow; consult the local utility. Design data for compound meters differ from the data in Figure 12. Manufacturers give data on exact pressure losses and capacities.
Figure 13 shows the variation of pressure loss with rate of flow for various faucets and cocks. The water demand for hose bibbs or other large-demand fixtures taken off the building main frequently results in inadequate water supply to the upper floor of a building. This condition can be prevented by sizing the distribution system so that pressure drops from the street main to all fixtures are the same. An ample building main (not less than 1 in. where possible) should be maintained until all branches to hose bibbs have been connected. Where street main pressure is excessive and a pressure-reducing valve is used to prevent water hammer or excessive pressure at fixtures, hose bibbs should be connected ahead of the reducing valve.
The principles involved in sizing upfeed and downfeed systems are the same. In the downfeed system, however, the difference in elevation between the overhead supply mains and the fixtures provides the pressure required to overcome pipe friction. Because friction pressure loss and height pressure loss are not additive, as in an upfeed system, smaller pipes may be used with a downfeed system.
The maximum safe water velocity in a thermoplastic piping system under most operating conditions is typically 5 fps; however, higher velocities can be used in cases where the operating characteristics of valves and pumps are known so that sudden changes in flow velocity can be controlled. The total pressure in the system at any time (operating pressure plus surge of water hammer) should not exceed 150% of the pressure rating of the system.
Procedure for Sizing Cold-Water Systems
The recommended procedure for sizing piping systems is as follows:
-
Sketch the main lines, risers, and branches, and indicate the fixtures to be served. Indicate the rate of flow of each fixture.
-
Using Table 25, compute the demand weights of the fixtures in fixture units.
-
Determine the total demand in fixture units and, using Figure 9, 10, or 11, find the expected demand.
-
Determine the equivalent length of pipe in the main lines, risers, and branches. Because the sizes of the pipes are not known, the exact equivalent length of various fittings cannot be determined. Add the equivalent lengths, starting at the street main and proceeding along the service line, main line of the building, and up the riser to the top fixture of the group served.
-
Determine the average minimum pressure in the street main and the minimum pressure required for operation of the topmost fixture, which should be 8 to 25 psi.
-
Calculate the approximate design value of the average pressure drop per 100 ft of equivalent length of pipe determined in step 4 and using Equation (1).
where
| Δp |
= |
average pressure loss per 100 ft of equivalent length of pipe, psi |
| ps |
= |
pressure in street main, psig |
| pf |
= |
minimum pressure required to operate topmost fixture, psig |
| pm |
= |
pressure drop through water meter, psi |
| H |
= |
height of highest fixture above street main, ft |
| L |
= |
equivalent length determined in step 4, ft |
-
If the system is downfeed supply from a gravity tank, height of water in the tank, converted to psi by multiplying by 0.434, replaces the street main pressure, and the term 0.434H is added instead of subtracted in calculating Δp. In this case, H is the vertical distance of the fixture below the bottom of the tank. The pressure conversion factor 0.434 is determined by the weight of water occupying a 1 ft3 volume, or 62.4/144 = 0.434 psi per foot of water.
-
From the expected rate of flow determined in step 3 and the value of Δp calculated in step 6, choose the sizes of pipe from Figure 14, 15, or 16.
Example 4.
Assume a minimum street main pressure of 55 psig; a height of topmost fixture (a urinal with flush valve) above street main of 50 ft; an equivalent pipe length from water main to highest fixture of 100 ft; a total load on the system of 50 fixture units; and that the water closets are flush valve operated. Find the required size of supply main.
Solution: Use Equation (1):
| ps | = | Street main pressure (given) = 55 psig |
| H | = | 50 ft (given) |
| Pf | = | 15 psig from Table 24 |
| Flow | = | 51 gpm from Figure 11 |
For a trial run, use 1 1/2 in.; then Pm= 6.5 psig from Figure 12 at 51 gpm. The pressure drop available for overcoming friction in pipes and fittings is 55 – 0.434 × 50 – 15 – 6.5 = 12 psi.
At this point, estimate the equivalent pipe length of the fittings on the direct line from the street main to the highest fixture. The exact equivalent length of the various fittings cannot be determined because the pipe sizes of the building main, riser, and branch leading to the highest fixture are not yet known, but a first approximation is necessary to tentatively select pipe sizes. If the computed pipe sizes differ from those used in determining the equivalent length of pipe fittings, a recalculation using the computed pipe sizes for the fittings will be necessary. It is common practice for the first trial to assume that the total equivalent length of the pipe fittings is 50% of the total length of pipe. In this example, 100 ft × 50% = 50 ft.
The permissible pressure loss per 100 ft of equivalent pipe is 12 × 100/(100 + 50) = 8 psi or 18 ft/100 ft. A 1 1/2 in. building main is adequate.
The sizing of the branches of the building main, the risers, and the fixture branches follows these principles. For example, assume that one of the branches of the building main carries the cold-water supply for three water closets, two bathtubs, and three lavatories. Using the permissible pressure loss of 8 psi per 100 ft, the size of branch (determined from Table 25 and Figures 14 and 11) is found to be 1 1/2 in. Items included in the computation of pipe size are as follows:
Table 26 is a guide to minimum pipe sizing where flush valves are used.
Velocities exceeding 10 fps cause undesirable noise in the piping system. This usually governs the size of larger pipes in the system, whereas in small pipe sizes, the friction loss usually governs the selection because the velocity is low compared to friction loss. Velocity is the governing factor in downfeed systems, where friction loss is usually neglected. Velocity in branches leading to pump suctions should not exceed 5 fps.
If the street pressure is too low to adequately supply upper-floor fixtures, the pressure must be increased. Constant- or variable-speed booster pumps, alone or in conjunction with gravity supply tanks, or hydropneumatic systems may be used.
Flow control valves for individual fixtures under varying pressure conditions automatically adjust flow at the fixture to a predetermined quantity. These valves allow the designer to (1) limit flow at the individual outlet to the minimum suitable for the purpose, (2) hold total demand for the system more closely to the required minimum, and (3) design the piping system as accurately as is practicable for the requirements.
The Darcy-Weisbach equation with friction factors from the Moody chart or Colebrook equation (or, alternatively, the Hazen-Williams equation) is fundamental to calculating pressure drop in hot- and chilled-water piping; however, charts calculated from these equations (such as Figures 14, 15, and 16) provide easy determination of pressure drops for specific fluids and pipe standards. In addition, tables of pressure drops can be found in Crane Co. (1976) and Hydraulic Institute (1990).
The Reynolds numbers represented on the charts in Figures 14, 15, and 16 are all in the turbulent flow regime. For smaller pipes and/or lower velocities, the Reynolds number may fall into the laminar regime, in which the Colebrook friction factors are no longer valid.
Most tables and charts for water are calculated for properties at 60°F. Using these for hot water introduces some error, although the answers are conservative (i.e., cold-water calculations overstate the pressure drop for hot water). Using 60°F water charts for 200°F water should not result in errors in Δp exceeding 20%.
Range of Usage of Pressure Drop Charts
General Design Range. The general range of pipe friction loss used for design of hydronic systems is between 1 and 4 ft of water per 100 ft of pipe. A value of 2.5 ft/100 ft represents the mean to which most systems are designed. Wider ranges may be used in specific designs if certain precautions are taken.
Piping Noise. Closed-loop hydronic system piping is generally sized below certain arbitrary upper limits, such as a velocity limit of 4 fps for 2 in. pipe and under, and a pressure drop limit of 4 ft per 100 ft for piping over 2 in. in diameter. Velocities in excess of 4 fps can be used in piping of larger size. This limitation is generally accepted, although it is based on relatively inconclusive experience with noise in piping. Water velocity noise is not caused by water but by free air, sharp pressure drops, turbulence, or a combination of these, that cause cavitation or flashing of water into steam. Therefore, higher velocities may be used if proper precautions are taken to eliminate air and turbulence.
Air in hydronic systems is usually undesirable because it causes flow noise, allows oxygen to react with piping materials, and sometimes even prevents flow in parts of a system. Air may enter a system at an air/water interface in an open system or in an expansion tank in a closed system, or it may be brought in dissolved in makeup water. Most hydronic systems use air separation devices to remove air. The solubility of air in water increases with pressure and decreases with temperature; thus, separation of air from water is best achieved at the point of lowest pressure and/or highest temperature in a system. For more information, see Chapter 13 of the 2020 ASHRAE Handbook—HVAC Systems and Equipment.
In the absence of venting, air can be entrained in the water and carried to separation units at flow velocities of 1.5 to 2 fps or more in pipe 2 in. and under. Minimum velocities of 2 fps are therefore recommended. For pipe sizes 2 in. and over, minimum velocities corresponding to a head loss of 0.75 ft/100 ft are normally used. Maintaining minimum velocities is particularly important in the upper floors of high-rise buildings where the air tends to come out of solution because of reduced pressures. Higher velocities should be used in downcomer return mains feeding into air separation units located in the basement.
Example 5.
Determine the iron pipe size for a circuit requiring 20 gpm flow.
Solution: Enter Figure 4 at 20 gpm, read up to pipe size within normal design range (1 to 4 ft/100 ft), and select 1 1/2 in. Velocity is 3.1 fps, which is between 2 and 4. Pressure loss is 2.9 ft/100 ft.
Valve and Fitting Pressure Drop
Valves and fittings can be listed in elbow equivalents, with an elbow being equivalent to a length of straight pipe. Table 27 lists equivalent lengths of 90° elbows; Table 28 lists elbow equivalents for valves and fittings for iron and copper.
Example 6.
Determine equivalent feet length of pipe for a 4 in. open gate valve at a flow velocity of approximately 4 fps.
Solution: From Table 27, at 4 fps, each elbow is equivalent to 10.6 ft of 4 in. pipe. From Table 28, the gate valve is equivalent to 0.5 elbows. The actual equivalent pipe length (added to measured circuit length for pressure drop determination) will be 10.6 × 0.5, or 5.3 equivalent feet of 4 in. pipe.
Tee Fitting Pressure Drop. Pressure drop through pipe tees varies with flow through the branch. Figure 17 shows pressure drops for nominal 1 in. tees of equal inlet and outlet sizes and for the flow patterns shown. Idelchik (1986) also presents data for threaded tees.
Different investigators present tee loss data in different forms, and it is sometimes difficult to reconcile results from several sources. As an estimate of the upper limit to tee losses, a pressure or head loss coefficient of 1.0 may be assumed for entering and leaving flows (i.e.,
).
Example 7.
Determine the pressure or head losses for a 1 in. (all openings) threaded pipe tee flowing 25% to the side branch, 75% through. The entering flow is 10 gpm (3.71 fps).
Solution: From Figure 17, bottom curve, the number of equivalent elbows for the through-flow is 0.15 elbows; the through-flow is 7.5 gpm (2.78 fps); and the head loss or pressure drop is based on the exit flow rate. Table 27 gives the equivalent length of a 1 in. elbow at 3 fps as 2.7 ft. Using Figure 14, the head loss is 4 ft/100 ft for 1 in. pipe and 7.5 gpm flow.
From Figure 17, top curve, the number of equivalent elbows for the branch flow of 25% is 13 elbows; the branch flow is 2.5 gpm (0.93 fps); and the head loss or pressure drop is based on the exit flow rate. Table 27 gives the equivalent of a 1 in. elbow at 1 fps as 2.2 ft. Using Figure 14, the head loss is 0.55 ft/100 ft for 1 in. pipe and 2.5 gpm flow.
Pressure losses in steam piping for flows of dry or nearly dry steam are governed by Equations (2) to (8) in the section on Design Equations. This section incorporates these principles with other information specific to steam systems.
Required pipe sizes for a given load in steam heating depend on the following factors:
-
The initial pressure and the total pressure drop that can be allowed between the source of supply and the end of the return system
-
The maximum velocity of steam allowable for quiet and dependable operation of the system, taking into consideration the direction of condensate flow
-
The equivalent length of the run from the boiler or source of steam supply to the farthest heating unit
Initial Pressure and Pressure Drop. Table 29 lists pressure drops commonly used with corresponding initial steam pressures for sizing steam piping.
Several factors, such as initial pressure and pressure required at the end of the line, should be considered, but it is most important that (1) the total pressure drop does not exceed the initial gage pressure of the system (in practice, it should never exceed one-half the initial gage pressure); (2) pressure drop is not great enough to cause excessive velocities; (3) a constant initial pressure is maintained, except on systems specially designed for varying initial pressures (e.g., subatmospheric pressure), that normally operate under controlled partial vacuums; and (4) for gravity return systems, pressure drop to heating units does not exceed the water column available for removing condensate (i.e., height above the boiler water line of the lowest point on the steam main, on the heating units, or on the dry return).
Maximum Velocity. For quiet operation, steam velocity should be 8000 to 12,000 fpm, with a maximum of 15,000 fpm. The lower the velocity, the quieter the system. When condensate must flow against the steam, even in limited quantity, the steam’s velocity must not exceed limits above which the disturbance between the steam and the counterflowing water may (1) produce objectionable sound, such as water hammer, or (2) result in the retention of water in certain parts of the system until the steam flow is reduced sufficiently to allow water to pass. These limits are a function of (1) pipe size; (2) pitch of the pipe if it runs horizontally; (3) quantity of condensate flowing against the steam; and (4) freedom of the piping from water pockets that, under certain conditions, act as a restriction in pipe size. Table 30 lists maximum capacities for various size steam lines.
Equivalent Length of Run. All tables for the flow of steam in pipes based on pressure drop must allow for pipe friction, as well as for the resistance of fittings and valves. These resistances are generally stated in terms of straight pipe; that is, a certain fitting produces a drop in pressure equivalent to the stated length of straight run of the same size of pipe. Table 31 gives the length of straight pipe usually allowed for the more common types of fittings and valves. In all pipe sizing tables in this chapter, length of run refers to the equivalent length of run as distinguished from the actual length of pipe. A common sizing method is to assume the length of run and to check this assumption after pipes are sized. For this purpose, length of run is usually assumed to be double the actual length of pipe.
Example 8.
Using Table 31, determine the equivalent length in feet of pipe for the run shown.
Figure 18 is the basic chart for determining the flow rate and velocity of steam in Schedule 40 pipe for various values of pressure drop per 100 ft, based on 0 psig saturated steam. Figures 19A through 19D present charts for sizing steam piping for systems of 30, 50, 100, and 150 psig at various pressure drops. These charts are based on the Moody friction factor, which considers the Reynolds number and the roughness of the internal pipe surfaces; they contain the same information as the basic chart (Figure 18) but in a more convenient form.
Using the multiplier chart (Figure 20), Figure 18 can be used at all saturation pressures between 0 and 200 psig (see Example 10).
3.4 LOW-PRESSURE STEAM PIPING
Values in Table 32 (taken from Figure 18) provide a more rapid means of selecting pipe sizes for the various pressure drops listed and for systems operated at 3.5 and 12 psig. The flow rates shown for 3.5 psig can be used for saturated pressures from 1 to 6 psig, and those shown for 12 psig can be used for saturated pressures from 8 to 16 psig with an error not exceeding 8%.
Both Figure 18 and Table 32 can be used where the flow of condensate does not inhibit the flow of steam. Columns B and C of Table 33 are used in cases where steam and condensate flow in opposite directions, as in risers or runouts that are not dripped. Columns D, E, and F are for one-pipe systems and include risers, radiator valves and vertical connections, and radiator and riser runout sizes, all of which are based on the critical velocity of the steam to allow counterflow of condensate without noise.
Return piping can be sized by Table 34, using pipe capacities for wet, dry, and vacuum return lines for several values of pressure drop per 100 ft of equivalent length.
Example 9.
What pressure drop should be used for the steam piping of a system if the measured length of the longest run is 500 ft, and the initial pressure must not exceed 2 psig?
Solution: It is assumed, if the measured length of the longest run is 500 ft, that when the allowance for fittings is added, the equivalent length of run does not exceed 1000 ft. Then, with the pressure drop not over one-half of the initial pressure, the drop could be 1 psi or less. With a pressure drop of 1 psi and a length of run of 1000 ft, the drop per 100 ft would be 0.1 psi; if the total drop were 0.5 psi, the drop per 100 ft would be 0.05 psi. In both cases, the pipe could be sized for a desired capacity according to Figure 18.
On completion of the sizing, the drop could be checked by taking the longest line and actually calculating the equivalent length of run from the pipe sizes determined. If the calculated drop is less than that assumed, the pipe size is adequate; if it is more, an unusual number of fittings is probably involved, and either the lines must be straightened, or the next larger pipe size must be tried.
High-Pressure Steam Piping
Many heating systems for large industrial buildings use high-pressure steam (15 to 150 psig). These systems usually have unit heaters or large built-up fan units with blast heating coils. Temperatures are controlled by a modulating or throttling thermostatic valve or by face or bypass dampers controlled by the room air temperature, fan inlet, or fan outlet.
Use of Basic and Velocity Multiplier Charts
Example 10.
Given a flow rate of 6700 lb/h, an initial steam pressure of 100 psig, and a pressure drop of 11 psi/100 ft, find the size of Schedule 40 pipe required and the velocity of steam in the pipe.
Solution: The following steps are shown by the broken line on Figures 18 and 20.
1. Enter Figure 18 at a flow rate of 6700 lb/h, and move vertically to the horizontal line at 100 psig
2. Follow along inclined multiplier line (upward and to the left) to horizontal 0 psig line. The equivalent mass flow at 0 psig is about 2500 lb/h.
3. Follow the 2500 lb/h line vertically until it intersects the horizontal line at 11 psi per 100 ft pressure drop. Nominal pipe size is 2 1/2 in. The equivalent steam velocity at 0 psig is about 32,700 fpm.
4. To find the steam velocity at 100 psig, locate the value of 32,700 fpm on the ordinate of the velocity multiplier chart (Figure 20) at 0 psig.
5. Move along the inclined multiplier line (downward and to the right) until it intersects the vertical 100 psig pressure line. The velocity as read from the right (or left) scale is about 13,000 fpm.
Note: Steps 1 through 5 would be rearranged or reversed if different data were given.
3.5 STEAM CONDENSATE SYSTEMS
The majority of steam systems used in heating applications are two-pipe systems (steam pipe and condensate pipe). This discussion is limited to sizing the condensate lines in two-pipe systems.
When steam is used for heating a liquid to 215°F or less (e.g., in domestic water heat exchangers, domestic heating water converters, or air-heating coils), the devices are usually provided with a steam control valve. As the control valve throttles, the absolute pressure in the load device decreases, removing all pressure motivation for flow in the condensate return system. To ensure the flow of steam condensate from the load device through the trap and into the return system, it is necessary to provide a vacuum breaker on the device ahead of the trap. This ensures a minimum pressure at the trap inlet of atmospheric pressure plus whatever liquid leg the designer has provided. Then, to ensure flow through the trap, it is necessary to design the condensate system so that it will never have a pressure above atmospheric in the condensate return line.
Vented (Open) Return Systems. To achieve this pressure requirement, the condensate return line is usually vented to the atmosphere (1) near the point of entrance of the flow streams from the load traps, (2) in proximity to all connections from drip traps, and (3) at transfer pumps or feedwater receivers.
The dry return lines in a vented return system have flowing liquid in the bottom of the line and gas or vapor in the top (Figure 21A). The liquid is the condensate, and the gas may be steam, air, or a mixture of the two. The flow phenomenon for these dry return systems is open channel flow, which is best described by the Manning equation:
where
| Q |
= |
volumetric flow rate, cfs |
| A |
= |
cross-sectional area of conduit, ft2 |
| r |
= |
hydraulic radius of conduit, ft |
| n |
= |
coefficient of roughness (usually 0.012) |
| S |
= |
slope of conduit, ft/ft |
Table 35 is a solution to Equation (22) that shows pipe size capacities for steel pipes with various pitches. Recommended practice is to size vertical lines by the maximum pitch shown, although they would actually have a capacity far in excess of that shown. As pitch increases, hydraulic jump that could fill the pipe and other transient effects that could cause water hammer should be avoided. Flow values in Table 35 are calculated for Schedule 40 steel pipe, with a factor of safety of 3.0, and can be used for copper pipes of the same nominal pipe size.
The flow characteristics of wet return lines (Figure 21B) are best described by the Darcy-Weisbach equation [Equation (1)]. The motivation for flow is the fluid head difference between the entering section of the flooded line and the leaving section. It is common practice, in addition to providing for the fluid head differential, to slope the return in the direction of flow to a collection point such as a dirt leg to clear the line of sediment or solids. Table 36 is a solution to Equation (1) that shows pipe size capacity for steel pipes with various available fluid heads. Table 36 can also be used for copper tubing of equal nominal pipe size.
Nonvented (Closed) Return Systems. For systems with a continual steam pressure difference between the point where the condensate enters the line and the point where it leaves (Figure 21C), Table 34 or Table 35, as applicable, can be used for sizing the condensate lines. Although these tables express condensate capacity without slope, common practice is to slope the lines in the direction of flow to a collection point (similar to wet returns) to clear the lines of sediment or solids.
When saturated condensate at pressures above the return system pressure enters the return (condensate) mains, some of the liquid flashes to steam. This occurs typically at drip traps into a vented return system or at load traps leaving process load devices that are not valve controlled and typically have no subcooling. If the return main is vented, the vent lines relieve any excessive pressure and prevent a backpressure phenomenon that could restrict flow through traps from valved loads; the pipe sizing would be as described for vented dry returns. If the return line is not vented, flash steam causes a pressure rise at that point and the piping could be sized as described for closed returns, and in accordance with Table 34 or Table 37, as applicable.
Passage of fluid through the steam trap is a throttling or constant-enthalpy process. The resulting fluid on the downstream side of the trap can be a mixture of saturated liquid and vapor. Thus, in nonvented returns, it is important to understand the fluid’s condition when it enters the return line from the trap.
The condition of the condensate downstream of the trap can be expressed by the quality x, defined as
where
| mv |
= |
mass of saturated vapor in condensate |
| ml |
= |
mass of saturated liquid in condensate |
Likewise, the volume fraction Vc of the vapor in the condensate is expressed as
where
| Vv |
= |
volume of saturated vapor in condensate |
| Vl |
= |
volume of saturated liquid in condensate |
The quality and the volume fraction of the condensate downstream of the trap can also be estimated from Equations (25) and (26), respectively.
where
| h1 |
= |
enthalpy of liquid condensate entering trap evaluated at supply pressure for saturated condensate or at saturation pressure corresponding to temperature of subcooled liquid condensate |
| hf2 |
= |
enthalpy of saturated liquid at return or downstream pressure of trap |
| hg2 |
= |
enthalpy of saturated vapor at return or downstream pressure of trap |
| vf2 |
= |
specific volume of saturated liquid at return or downstream pressure of trap |
| vg2 |
= |
specific volume of saturated vapor at return or downstream pressure of trap. |
Table 38 presents some values for quality and volume fraction for typical supply and return pressures in heating and ventilating systems. Note that the percent of vapor on a mass basis x is small, although the percent of vapor on a volume basis Vc is very large. This indicates that the return pipe cross section is predominantly occupied by vapor. Figure 22 is a working chart to determine the quality of condensate entering the return line from the trap for various combinations of supply and return pressures. If the liquid is subcooled entering the trap, the saturation pressure corresponding to the liquid temperature should be used for the supply or upstream pressure. Typical pressures in the return line are given in Table 39.
Gravity one-pipe air vent systems in which steam and condensate flow in the same pipe, frequently in opposite directions, are considered obsolete and are no longer being installed. Chapter 33 of the 1993 ASHRAE Handbook—Fundamentals or earlier ASHRAE Handbook volumes include descriptions of and design information for one-pipe systems.
Piping for gas appliances should be of adequate size and installed so that it provides a supply of gas sufficient to meet the maximum demand without undue loss of pressure between the point of supply (the meter) and the appliance. The size of gas pipe required depends on (1) maximum gas consumption to be provided, (2) length of pipe and number of fittings, (3) allowable pressure loss from the outlet of the meter to the appliance, and (4) specific gravity of the gas.
Gas consumption in ft3/h is obtained by dividing the Btu input rate at which the appliance is operated by the average heating value of the gas in Btu/ft3. Insufficient gas flow from excessive pressure losses in gas supply lines can cause inefficient operation of gas-fired appliances and sometimes create hazardous operations. Gas-fired appliances are normally equipped with a data plate giving information on maximum gas flow requirements or Btu input as well as inlet gas pressure requirements. The local gas utility can give the gas pressure available at the utility’s gas meter. Using this information, the required size of gas piping can be calculated for satisfactory operation of the appliance(s).
Table 40 gives pipe capacities for gas flow for up to 200 ft of pipe based on a specific gravity of 0.60. Capacities for pressures less than 1.5 psig may also be determined by the following equation from NFPA/IAS National Fuel Gas Code (NFPA Standard 54/ANSI Standard Z223.1):
where
| Q |
= |
flow rate at 60°F and 30 in. Hg, cfh |
| d |
= |
inside diameter of pipe, in. |
| Δp |
= |
pressure drop, in. of water |
| C |
= |
factor for viscosity, density, and temperature |
|
= |
0.00354(t + 460)s0.848μ0.152 |
| t |
= |
temperature, °F |
| s |
= |
ratio of density of gas to density of air at 60°F and 30 in. Hg |
| μ |
= |
viscosity of gas, centipoise (0.012 for natural gas, 0.008 for propane) |
| L |
= |
pipe length, ft |
Gas service in buildings is generally delivered in the low-pressure range of 7 in. of water. The maximum pressure drop allowable in piping systems at this pressure is generally 0.5 in. of water but is subject to regulation by local building, plumbing, and gas appliance codes [see also the NFPA/IAS National Fuel Gas Code (NFPA Standard 54/ANSI Standard Z223.1)].
Where large quantities of gas are required or where long lengths of pipe are used (e.g., in industrial buildings), low-pressure limitations result in large pipe sizes. Local codes may allow (and local gas companies may deliver) gas at higher pressures (e.g., 2, 5, or 10 psig). Under these conditions, an allowable pressure drop of 10% of the initial pressure is used, and pipe sizes can be reduced significantly. Gas pressure regulators at the appliance must be specified to accommodate higher inlet pressures. NFPA/IAS (2012) provides information on pipe sizing for various inlet pressures and pressure drops at higher pressures. More complete information on gas piping can be found in the Gas Engineers’ Handbook (1970).
The pipe used to convey fuel oil to oil-fired appliances must be large enough to maintain low pump suction pressure and, in the case of circulating loop systems, to prevent overpressure at the burner oil pump inlet. Pipe materials must be compatible with the fuel and must be carefully assembled to eliminate all leaks. Leaks in suction lines can cause pumping problems that result in unreliable burner operation. Leaks in pressurized lines create fire hazards. Cast-iron or aluminum fittings and pipe are unacceptable. Pipe joint compounds must be selected carefully.
Oil pump suction lines should be sized so that at maximum suction line flow conditions, the maximum vacuum will not exceed 10 in. Hg for distillate grade fuels and 15 in. Hg for residual oils. Oil supply lines to burner oil pumps should not be pressurized by circulating loop systems or aboveground oil storage tanks to more than 5 psi, or pump shaft seals may fail. A typical oil circulating loop system is shown in Figure 23.
In assembling long fuel pipe lines, be careful to avoid air pockets. On overhead circulating loops, the line should vent air at all high points. Oil supply loops for one or more burners should be the continuous circulation type, with excess fuel returned to the storage tank. Dead-ended pressurized loops can be used, but air or vapor venting is more problematic.
Where valves are used, select ball or gate valves. Globe valves are not recommended because of their high pressure drop characteristics.
Oil lines should be tested after installation, particularly if they are buried, enclosed, or otherwise inaccessible. Failure to perform this test is a frequent cause of later operating difficulties. A suction line can be hydrostatically tested at 1.5 times its maximum operating pressure or at a vacuum of not less than 20 in. Hg. Pressure or vacuum tests should continue for at least 60 min. If there is no noticeable drop in the initial test pressure, the lines can be considered tight.
Tables 41 and 42 give recommended pipe sizes for handling No. 5 and No. 6 oils (residual grades) and No. 1 and No. 2 oils (distillate grades), respectively. Storage tanks and piping and pumping facilities for delivering the oil from the tank to the burner are important considerations in the design of an industrial oil-burning system. The construction and location of the tank and oil piping are usually subject to local regulations and National Fire Protection Association (NFPA) Standards 30 and 31.
ASHRAE members can access ASHRAE Journal articles and ASHRAE research project final reports at technologyportal.ashrae.org. Articles and reports are also available for purchase by nonmembers in the online ASHRAE Bookstore at www.ashrae.org/bookstore.
ASHRAE. 2013. Safety standard for refrigeration systems. ANSI/ASHRAE Standard 15-2013.
ASHRAE. 2013. Energy standard for buildings except low-ride residential buildings. ANSI/ASHRAE/IES Standard 90.1-2013.
ASME. 2013. Pipe threads, general purpose, inch. Standard B1.20.1-2013. American Society of Mechanical Engineers, New York.
ASME. 2006. Pipe threads, 60 deg. general purpose (metric). Standard B1.20.2M-2006. American Society of Mechanical Engineers, New York.
ASME. 2015. Gray iron pipe flanges and flanged fittings: Classes 25, 125, and 250. Standard B16.1-2015. American Society of Mechanical Engineers, New York.
ASME. 2011. Malleable iron threaded fittings: Classes 150 and 300. Standard B16.3-2011. American Society of Mechanical Engineers, New York.
ASME. 2011. Gray iron threaded fittings: Classes 125 and 250. Standard B16.4-2011. American Society of Mechanical Engineers, New York.
ASME. 2009. Pipe flanges and flanged fittings: NPS 1/2 through NPS 24 metric/inch standard. Standard B16.5-2009. American Society of Mechanical Engineers, New York.
ASME. 2012. Factory made wrought buttwelding fittings. Standard B16.9-2012. American Society of Mechanical Engineers, New York.
ASME. 2009. Forged fittings, socket-welding and threaded. Standard B16.11-2009. American Society of Mechanical Engineers, New York.
ASME. 2009. Cast iron threaded drainage fittings. Standard B16.12-2009. American Society of Mechanical Engineers, New York.
ASME. 2013. Cast copper alloy threaded fittings: Classes 125 and 250. Standard B16.15-2013. American Society of Mechanical Engineers, New York.
ASME. 2012. Cast copper alloy solder joint pressure fittings. Standard B16.18-2012. American Society of Mechanical Engineers, New York.
ASME. 2013. Wrought copper and copper alloy solder-joint pressure fittings. Standard B16.22-2013. American Society of Mechanical Engineers, New York.
ASME. 2011. Cast copper alloy solder joint drainage fittings: DWV. Standard B16.23-2011. American Society of Mechanical Engineers, New York.
ASME. 2011. Cast copper alloy pipe flanges and flanged fittings: Classes 150, 300, 600, 900, 1500, and 2500. Standard B16.24-2011. American Society of Mechanical Engineers, New York.
ASME. 2011. Cast copper alloy fittings for flared copper tubes. Standard B16.26-2011. American Society of Mechanical Engineers, New York.
ASME. 2012. Wrought copper and wrought copper alloy solder-joint drainage fittings—DWV. Standard B16.29-2012. American Society of Mechanical Engineers, New York.
ASME. 2011. Ductile iron pipe flanges and flanged fittings: Classes 150 and 300. Standard B16.42-2011. American Society of Mechanical Engineers, New York.
ASME. 2016. Power piping. Standard B31.1-2016. American Society of Mechanical Engineers, New York.
ASME. 2016. Refrigeration piping and heat transfer components. Standard B31.5-2016. American Society of Mechanical Engineers, New York.
ASME. 2014. Building services piping. Standard B31.9-2014. American Society of Mechanical Engineers, New York.
ASME. 2015. Welded and seamless wrought steel pipe. Standard B36.10M-2015. American Society of Mechanical Engineers, New York.
ASME. 2015. Qualification standard for welding and brazing procedures, welders, brazers, and welding and brazing operators. Boiler and Pressure Vessel Code, Section IX. American Society of Mechanical Engineers, New York.
ASTM. 2012. Standard specification for pipe, steel, black and hot-dipped, zinc-coated, welded, and seamless. Standard A53. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2015. Standard specification for seamless carbon steel pipe for high-temperature service. Standard A106. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2014. Standard specification for seamless copper water tube. Standard B88. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2016. Standard specification for seamless copper tube for air conditioning and refrigeration field service. Standard B280. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2011. Standard specification for rigid poly(vinyl chloride) (PVC) compounds and chlorinated poly(vinyl chloride) (CPVC) compounds. Standard D1784. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2015. Standard specification for poly(vinyl chloride) (PVC) plastic pipe, schedules 40, 80, and 120. Standard D1785. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2015. Standard test method for determining dimensions of thermoplastic pipe and fittings. Standard D2122. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2012. Standard specification for polyethylene (PE) plastic pipe (SIDR-PR) based on controlled inside diameter. Standard D2239. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2015. Standard specification for threaded poly(vinyl chloride) (PVC) plastic pipe fittings, schedule 80. Standard D2464. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2015. Standard specification for poly(vinyl chloride) (PVC) plastic pipe fittings, schedule 40. Standard D2466. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2015. Standard specification for poly(vinyl chloride) (PVC) plastic pipe fittings, schedule 80. Standard D2467. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2012. Standard specification for solvent cements for poly(vinyl chloride) (PVC) plastic piping systems. Standard D2564. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2014. Standard specification for acrylonitrile-butadiene-styrene (ABS) schedule 40 plastic drain, waste, and vent pipe and fittings. Standard D2661. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2014. Standard specification for poly(vinyl chloride) (PVC) plastic drain, waste, and vent pipe and fittings. Standard D2665. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2013. Standard test method for obtaining hydrostatic design basis for thermoplastic pipe materials or pressure design basis for thermoplastic pipe products. Standard D2837-13e1. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2012. Standard practice for obtaining hydrostatic or pressure design basis for “fiberglass” (glass-fiber-reinforced thermosetting-resin) pipe and fittings. Standard D2992. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2014. Standard specification for polyethylene plastics pipe and fittings materials. Standard D3350. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2016. Standard classification system and basis for specifications for rigid acrylonitrile-butadiene-styrene (ABS) materials for pipe and fittings. Standard D3965. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2015. Standard specification for threaded chlorinated poly(vinyl chloride) (CPVC) plastic pipe fittings, schedule 80. Standard F437. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2015. Standard specification for socket-type chlorinated poly(vinyl chloride) (CPVC) plastic pipe fittings, schedule 40. Standard F438. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2013. Standard specification for chlorinated poly(vinyl chloride) (CPVC) plastic pipe fittings, schedule 80. Standard F439. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2015. Standard specification for crosslinked polyethylene (PEX) tubing. Standard F876. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2011. Standard specification for crosslinked polyethylene (PEX) hot- and cold-water distribution systems. Standard F877. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2014. Standard specification for solvent cements for chlorinated poly(vinyl chloride) (CPVC) plastic pipe and fittings. Standard F493. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2015. Standard test method for evaluating the oxidative resistance of crosslinked polyethylene (PEX) pipe, tubing and systems to hot chlorinated water. Standard F2023. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2015. Standard specification for pressure-rated polypropylene (PP) piping systems. Standard F2389. American Society for Testing and Materials, West Conshohocken, PA.
ASTM. 2011. Standard specification for manufacture and joining of polyethylene (PE) gas pressure pipe with a peelable polypropylene (PP) outer layer. Standard F2830. American Society for Testing and Materials, West Conshohocken, PA.
AWWA. 2014. Thickness design of ductile-iron pipe. Standard C150/A21.50. American Water Works Association, Denver.
Ball, E.F., and C.J.D. Webster. 1976. Some measurements of water-flow noise in copper and ABS pipes with various flow velocities. The Building Services Engineer 44(2):33.
Carrier. 1960. Piping design. In System design manual. Carrier Air Conditioning Company, Syracuse, NY.
CDA. 2010. The copper tube handbook. Copper Development Association, New York.
Crane Co. 1976. Flow of fluids through valves, fittings and pipe. Technical Paper 410. Crane Company, New York.
Crane Co. 1988. Flow of fluids through valves, fittings and pipe. Technical Paper 410. Crane Company, New York.
Dawson, F.M., and J.S. Bowman. 1933. Interior water supply piping for residential buildings. University of Wisconsin Experiment Station Bulletin 77.
Ding, C., L. Carlson, C. Ellis, and O. Mohseni. 2005. Pressure loss coefficients in 6, 8, and 10 inch steel pipe fittings. ASHRAE Research Project TRP-1116, Final Report. University of Minnesota, Saint Anthony Falls Laboratory.
Eshbach, O.W. 2009. Eshbach’s handbook of engineering fundamentals, 5th ed. M. Kutz, ed. John Wiley & Sons, New York.
Freeman, J.R. 1941. Experiments upon the flow of water in pipes. American Society of Mechanical Engineers, New York.
Gas Engineers’ Handbook. 1970. Industrial Press, New York.
Giesecke, F.E. 1926. Friction of water elbows. ASHVE Transactions 32:303.
Giesecke, F.E., and W.H. Badgett. 1931. Friction heads in one-inch standard cast-iron tees. ASHVE Transactions 37:395.
Giesecke, F.E., and W.H. Badgett. 1932a. Loss of head in copper pipe and fittings. ASHVE Transactions 38:529.
Giesecke, F.E., and W.H. Badgett. 1932b. Supplementary friction heads in one-inch cast-iron tees. ASHVE Transactions 38:111.
Grinnell Company. 1951. Piping design and engineering. Grinnell Company, Cranston, RI.
HDR design guide. 1981. Hennington, Durham and Richardson, Omaha, NE.
Heald, C.C. 2002. Cameron hydraulic data, 19th ed. Flowserve Corporation, Irving, TX.
Hegberg, R.A. 1995. Where did the k-factors for pressure loss in fittings come from? ASHRAE Transactions 101(1):1264-78. Paper CH-95-20-3.
Hunter, R.B. 1940. Methods of estimating loads in plumbing systems. NBS Report BMS 65. National Institute of Standards and Technology, Gaithersburg, MD.
Hunter, R.B. 1941. Water distributing systems for buildings. NBS Report BMS 79. National Institute of Standards and Technology, Gaithersburg, MD.
Hydraulic Institute. 1990. Engineering data book. Hydraulic Institute, Parsippany, NJ.
ICC. 2012. International plumbing code®. International Code Council, Washington, D.C.
Idelchik, I.E. 1986. Handbook of hydraulic resistance. Hemisphere Publishing, New York.
ISA. 2007. Flow equations for sizing control valves. Standard 75.01.01-07. International Society of Automation, Research Triangle Park, NC.
Laschober, R.R., G.Y. Anderson, and D.G. Barbee. 1966. Counterflow of steam and condensate in slightly pitched pipes. ASHRAE Transactions 72(1):157.
Marseille, B. 1965. Noise transmission in piping. Heating and Ventilating Engineering (June):674.
MSS. 2009. Pipe hangers and supports—Materials, design, manufacture, selection, application, and installation. ANSI/MSS Standard SP-58. Manufacturers Standardization Society of the Valve and Fittings Industry, Vienna, VA.
MSS. 2003. Pipe hangers and supports—Selection and application. Standard SP-69. Manufacturers Standardization Society of the Valve and Fittings Industry, Vienna, VA.
Nayyar, M. 1999. Piping handbook. McGraw-Hill, New York.
NFPA. 2010. Installation of sprinkler systems. Standard 13. National Fire Protection Association, Quincy, MA.
NFPA/AGA. 2012. National fuel gas code. ANSI/NFPA Standard 54. National Fire Protection Association, Quincy, MA. ANSI/AGA Standard Z223.1-2002. American Gas Association, Arlington, VA.
NSF/ANSI. 2016. Plastics piping system components and related materials. ANSI/NSF Standard 14-2016. NSF International, Ann Arbor, MI.
NSF/ANSI. 2016. Drinking water system components—Health effects. ANSI/NSF Standard 61. NSF International, Ann Arbor, MI.
Obrecht, M.F., and M. Pourbaix. 1967. Corrosion of metals in potable water systems. AWWA 59:977. American Water Works Association, Denver, CO.
PHCC. 2012. National standard plumbing code. Plumbing-Heating-Cooling Contractors Association, Falls Church, VA.
Plastic Pipe Institute. 1971. Water flow characteristics of thermoplastic pipe. Plastic Pipe Institute, New York.
PPFA. 2009. PVC piping systems for commercial and industrial applications design guide. Plastic Pipe and Fitting Association, Glenn Ellyn, IL
Rahmeyer, W.J. 1999a. Pressure loss coefficients of threaded and forged weld pipe fittings for ells, reducing ells, and pipe reducers (RP-968). ASHRAE Transactions 105(2):334-354. Paper 4308.
Rahmeyer, W.J. 1999b. Pressure loss coefficients of pipe fittings for threaded and forged weld pipe tees (RP-968). ASHRAE Transactions 105(2):355-385. Paper 4309.
Rahmeyer, W.J. 2002a. Pressure loss data for large pipe ells, reducers, and expansions. ASHRAE Transactions 108(1):360-375. Paper 4533.
Rahmeyer, W.J. 2002b. Pressure loss data for large pipe tees. ASHRAE Transactions 108(1):376-389. Paper 4534.
Rahmeyer, W.J. 2002c. Pressure loss coefficients for close-coupled pipe ells. ASHRAE Transactions 108(1):390-406. Paper 4535.
Rahmeyer, W.J. 2003a. Pressure loss data for PVC pipe elbows, reducers, and expansions (RP-1193). ASHRAE Transactions 109(2):230-251. Paper 4653.
Rahmeyer, W.J. 2003b. Pressure loss data for PVC pipe tees (RP-1193). ASHRAE Transactions 109(2):252-271. Paper 4654.
Rogers, W.L. 1953. Experimental approaches to the study of noise and noise transmission in piping systems. ASHVE Transactions 59:347-360.
Rogers, W.L. 1954. Sound-pressure levels and frequencies produced by flow of water through pipe and fittings. ASHRAE Transactions 60:411-430.
Rogers, W.L. 1956. Noise production and damping in water piping. ASHAE Transactions 62:39.
Sanks, R.L. 1978. Water treatment plant design for the practicing engineer. Ann Arbor Science, Ann Arbor, MI.
Smith, T. 1983. Reducing corrosion in heating plants with special reference to design considerations. Anti-Corrosion Methods and Materials 30 (October):4.
Stewart, W.E., and C.L. Dona. 1987. Water flow rate limitations (RP-450). ASHRAE Transactions 93(2):811-825. Paper 3106.
Williams, G.J. 1976. The Hunter curves revisited. Heating/Piping/Air Conditioning (November):67.
Williams, G.S., and A. Hazen. 1933. Hydraulic tables. John Wiley & Sons, New York.