Pipe Fitting Losses

Pipework fittings such as bends, tees, reducers etc., cause pressure loss or
resistance in a heating system.
When making approximate calculations 10%, 15%, 20% or more may be added
to the pressure loss in straight pipe runs.
For accurate calculations the fitting loss should be determined separately for
each fitting, as outlined below.
The concept of equivalent length is used and is defined as the length of straight
pipe which would give a friction pressure loss equivalent to one velocity head.
The DArcy equation is;
H = 4 . f . l . v
2
/ 2 . g . d
Where; H = head loss due to friction in a straight pipe (m)
f = friction coefficient
l = length of pipe (m)
d = diameter (m)
v = velocity of fluid (m/s)
g = acceleration due to gravity (m/s
2
)
The DArcy equation can be rewritten for pressure instead of head.
Pressure drop (p
l
) = ( 4 . f . l ) / d x ( v
2
. r . g ) / 2 . g
Where; p
l
= Pressure loss in a pipe section (Pa)
f = friction coefficient for pipe
l = length of pipe (m)
d = diameter of pipe (m)
v = water velocity (m/s)
r = density of water (kg/m3)
g = acceleration due to gravity (m/s
2
)

To simplify the above equation we get;
Pressure drop (p
l
) = ( 4 . f . l ) / d x ( v
2
. r ) / 2
or; Pressure drop (p
l
) = ( 4 . f . l ) / d x ( . r . v
2
)
For the friction pressure loss to equal one velocity head;
Velocity pressure = ( . . v
2
)
Then 1.0 = ( 4 . f . l ) / d
The length (l) is now called equivalent length (l
e
) and by rearranging the
above formula we get;
1.0 x d = 4 . f . l
e

l
e
= d / 4 .f

Values of equivalent length are given in the FLOW of WATER in PIPES TABLE
for water at 75
o
C, see CIBSE guide C (2001) section 4 , Flow of Fluids in Pipes
and Ducts, Tables 4.9 to 4.33 for various types of pipes.
These values should be corrected for each particular type of fitting.
The correction factors of Velocity pressure loss factors are called (Zeta)
factors.
The resistance in a fitting is converted to equivalent straight lengths of pipe,
e.g. a bend may have a resistance equivalent to 1.2 metres of straight pipe.
The Total Equivalent Length of a Fitting = Equivalent Length x Pressure Loss
factor z(Zeta).
Total Equivalent Length of a Fitting (T.E.L.) (m) = (le) x (Zeta
factor).
See CIBSE guide C (2007) section 4.4.1 and section 4.10 for more details of
fittings zeta factors.
Examples Of Zeta Factors
The following are some examples of pressure loss ( zeta) factors for pipe
fittings:


Some Pipe Cross Sectional Areas
Area = .d
2
/4
Area (28mm) = .0.028
2
/ 4 = 0.0006158 m
2
.
Area (22mm) = .0.022
2
/ 4 = 0.0003801 m
2
.
Area (15mm) = .0.015
2
/ 4 = 0.0001767 m
2
.
Some Ratios of Pipe Areas
For 22mm x 15mm reduction
The ratio of area A
2
/A
1
= 0.0001767 / 0.0003801 = 0.465.
Therefore for reduction = 0.35 approximately from the above table.
For 28mm x 22mm reduction
The ratio of area A
2
/A
1
= 0.0003801 / 0.0006158 = 0.617.
Therefore for reduction = 0.25 approximately from the above table.


Skip to content
Skip to main navigation
Skip to first column
Skip to second column
MYCHEME




search...

search com_search 97

Home Calculation Methods Pressure Drops in Pipe Fittings
Calculating Pressure Drops in Pipe Fittings
Article Index
Calculating Pressure
Drops in Pipe Fittings
Page 2
All Pages
Page 1 of 2
Introduction
Piping on a process plant
does more than run in a straight line. Piping consists of straight lengths of pipe punctuated by any
number of fittings - including bends, valves and T-pieces. Line losses from pipe work fittings cannot be
discounted.
Pipe fittings impose a pressure drop as they:
Change the fluid flow direction
Change the size of the cross-sectional flow path, causing the fluid to either accelerate or de-accelerate.
Present an obstruction in the flow path.
Often, pipe fitting pressure losses make up a sizable chunk of the total system pressure drop. Any estimate of the
hydraulics of a pipe line must consider the impact of the pipe work fittings.
Other articles of interest:
Pressure drop in pipe lines
Compressible flow pressure drops
Standard pipe sizes
Calculation Method
In another article, we introduced the Darcy Equation for calculating pressure drop in straight lengths of pipe:

The equivalent equation for pipe fittings is

Where:
P
Fittings
- Pressure drop across fitting, Pa
- Fluid Density, kg/m
3

u - Fluid velocity in the pipe, m/s
K - Resistance Coefficient (Dimensionless)
The Resistance Coefficient, K, is sometimes referred as the number of velocity heads and is specific to each
type of fitting. Values of the resistance coefficient for various fitting types can be found in the literature, but typical
values are listed below.

Resistance
Cofficient, K
45
o
Elbows 0.3
90
o
Elbows (Standard Radius) 0.6-0.8
30
o
Mitre Bends 0.2
45
o
Mitre Bends 0.375
60
o
Mitre Bends 0.625
90
o
Square Elbows 1.2
180
o
Return Bend (Close Pattern) 1.25
T-Piece (Side Connection) 1.2-1.8
T-Piece (Flow through) 0.5
Globe Valves (Fully Open) 1.2-6.0
Gate Valves (Fully Open) 0.15
Gate Valves (
3
/
4
Open) 1
Gate Valves (
1
/
2
Open) 4
Gate Valves (
1
/
4
Open) 16
Plug Valves (Fully Open) 0.45
3-Way Valve (Straight Thru') 0.75
3-Way Valve (Side Connection) 2.25
Ball Valves 0.075
Butterfly Valves (less than 10" 250mm NB) 1.125
Butterfly Valves (10" 250mm NB - 14" 350mm
NB) 0.875
Butterfly Valves (Greater than 14" 350mm NB) 0.625
Check Valves (Swing Type) 2.5
Check Valves (Lift Type) 15
Pipe Exits 1
Pipe Entrances 0.78
Some textbooks use Equivalent Length instead of Resistance Coefficient to give a measure of resistance to fluid
flow in fittings (see box below).
Calculating Pressure Drops in Pipe Fittings - Page 2
Article Index
Calculating Pressure
Drops in Pipe Fittings
Page 2
All Pages
Page 2 of 2
Limitations with this Calculation Method
As with many engineering calculations, this
approach is only approximate and should be used with care. It is based on two important assumptions.
Firstly, it assumes that the Resistance Coefficient is constant for any given fitting type regardless of size. For
example, it assumes that a fully open gate valve has Resistance Coefficient of 0.15, regardless whether it is a
inch or 24 inch diameter. This would only be true if the valves were geometrically similar. In reality, this will not be
the case. Secondly, the Resistance Coefficients given above are for fully developed flow i.e. fully turbulent.
However, for many applications, this method produces acceptable results. For gases and low viscosity liquids
(e.g. water), pipelines are designed to produce fully turbulent flow. In addition, pipe sizes are selected before the
final pipe routing and actual number of pipe fittings is known. These uncertainties mean that this calculation
method should give results within the margin of error.
However, this calculation method breaks down at low Reynolds Numbers (espicially in laminar and transitional
flow). This will be an issue with high viscosity liquids where flow is typically at low Re numbers
There are refinements to this method which take into account these approximations. These are know as the 2K
and 3 K methods. However for most situations, the simplified approach outlined above is acceptable.





Pipe Friction Loss Calculations
Flow of fluid through a pipe is resisted by viscous shear stresses within the fluid and
the turbulence that occurs along the internal pipe wall, which is dependent on the roughness of the
pipe material.

This resistance is termed pipe friction and is usually measured in feet or metres head of the fluid,
which is why it is also refered to as the head loss due to pipe friction.
Head Loss in a Pipe
A large amount of research has been carried out over many years to establish various
formulae that can calculate head loss in a pipe. Most of this work has been developed based
on experimental data.
Overall head loss in a pipe is affected by a number of factors which include the viscosity of
the fluid, the size of the internal pipe diameter, the internal roughness of the inner surface of
the pipe, the change in elevation between the ends of the pipe and the length of the pipe along
which the fluid travels.
Valves and fittings on a pipe also contribute to the overall head loss that occurs, however
these must be calculated separately to the pipe wall friction loss, using a method of modeling
pipe fitting losses with k factors.
Darcy Weisbach Formula
The Darcy formula or the Darcy-Weisbach equation as it tends to be referred to, is now
accepted as the most accurate pipe friction loss formula, and although more difficult to
calculate and use than other friction loss formula, with the introduction of computers, it has
now become the standard equation for hydraulic engineers.
Weisbach first proposed the relationship that we now know as the Darcy-Weisbach equation
or the Darcy-Weisbach formula, for calculating friction loss in a pipe.
Darcy-Weisbach equation:

hf = f (L/D) x (v2/2g)

where:
hf = head loss (m)
f = friction factor
L = length of pipe work (m)
d = inner diameter of pipe work (m)
v = velocity of fluid (m/s)
g = acceleration due to gravity (m/s)

or:

hf = head loss (ft)
f = friction factor
L = length of pipe work (ft)
d = inner diameter of pipe work (ft)
v = velocity of fluid (ft/s)
g = acceleration due to gravity (ft/s)
The establishment of the friction factors was however still unresolved, and indeed was an
issue that needed further work to develop a solution such as that produced by the Colebrook-
White formula and the data presented in the Moody chart.
The Moody Chart
The Moody Chart finally provided a method of finding an accurate friction factor and this
encouraged use of the Darcy-Weisbach equation, which quickly became the method of choice
for hydraulic engineers.
The introduction of the personnel computer from the 1980's onwards reduced the time
required to calculate the friction factor and pipe head loss. This itself has widened the use of
the Darcy-Weisbach formula to the point that most other equations are no longer used.
Hazen-Williams Formula
Before the advent of personal computers the Hazen-Williams formula was extremely popular
with piping engineers because of its relatively simple calculation properties.
However the Hazen-Williams results rely upon the value of the friction factor, C hw, which is
used in the formula, and the C value can vary significantly, from around 80 up to 130 and
higher, depending on the pipe material, pipe size and the fluid velocity.
Also the Hazen-Williams equation only really gives good results when the fluid is Water and
can produce large inaccuracies when this is not the case.
The imperial form of the Hazen-Williams formula is:

hf = 0.002083 x L x (100/C)^1.85 x (gpm^1.85 / d^4.8655)

where:
hf = head loss in feet of water
L = length of pipe in feet
C = friction coefficient
gpm = gallons per minute (USA gallons not imperial gallons)
d = inside diameter of the pipe in inches
The empirical nature of the friction factor C hw means that the Hazen-Williams formula is
not suitable for accurate prediction of head loss. The friction loss results are only valid for
fluids with a kinematic viscosity of 1.13 centistokes, where the velocity of flow is less than
10 feet per sec, and where the pipe diameter has a size greater than 2 inches.
Notes: Water at 60 F (15.5 C) has a kinematic viscosity of 1.13 centistokes.

Common Friction Factor Values of C hw used for design purposes are:

Asbestos Cement 140
Brass tube 130
Cast-Iron tube 100
Concrete tube 110
Copper tube 130
Corrugated steel tube 60
Galvanized tubing 120
Glass tube 130
Lead piping 130
Plastic pipe 140
PVC pipe 150
General smooth pipes 140
Steel pipe 120
Steel riveted pipes 100
Tar coated cast iron tube 100
Tin tubing130
Wood Stave 110
These C hw values provide some allowance for changes to the roughness of internal pipe
surface, due to pitting of the pipe wall during long periods of use and the build up of other
deposits.






Friction Factor Calculations
The Darcy-Weisbach equation, for calculating the friction loss in a pipe, uses a dimensionless value
known as the friction factor (also known as the Darcy-Weisbach friction factor or the Moody friction
factor) and it is four times larger than the Fanning friction factor.
Friction Factor Chart / Moody Chart
The friction factor or Moody chart is the plot of the relative roughness (e/D) of a pipe against
the Reynold's number. The blue lines plot the friction factor for flow in the wholly turbulent
region of the chart, while the straight black line plots the friction factor for flow in the wholly
laminar region of the chart.

In 1944, LF Moody plotted the data from the Colebrook equation and the resulting chart
became known as The Moody Chart or sometimes the Friction Factor Chart. It was this chart
which first enabled the user to obtain a reasonably accurate friction factor for turbulent flow
conditions, based on the Reynolds number and the Relative Roughness of the pipe.
Friction Factor for Laminar Flow
The friction factor for laminar flow is calculated by dividing 64 by the Reynold's number.

Friction factor (for laminar flow) = 64 / Re
Critical Flow Condition
When flow occurs between the Laminar and Turbulent flow conditions (Re 2300 to Re 4000)
the flow condition is known as critical and is difficult to predict. Here the flow is neither
wholly laminar nor wholly turbulent. It is a combination of the two flow conditions.
Friction Factor for Turbulent Flow
The friction factor for turbulent flow is calculated using the Colebrook-White equation:
Colebrook-White Equation

Due to the implicit formation of the Colebrook-White equation, calculation of the friction
factor requires an iterative solution via numerical methods.
The friction factor is then used in the Darcy-Weisbach formula to calculate the fluid frictional
loss in a pipe.
Reynold's Numbers
Fluid flow in a pipe encounters frictional resistance due to the internal roughness (e)
of the pipe wall, which can create local eddy currents within the fluid. Calculation of the Reynold's
Number helps to determine if the flow in the pipe is Laminar Flow or Turbulent Flow.

Pipes that have a smooth wall such as glass, copper, brass and polyethylene cause less fritional
resistance and hence they produce a smaller frictional loss than those pipes with a greater internal
roughness, such as concrete, cast iron and steel.

The velocity profile of fluid flow in a pipe shows that the fluid at the centre of the stream moves
more quickly than the fluid flow towards the edge of the stream. Therefore friction occurs between
layers within the fluid.

Fluids with a high viscosity flow more slowly and generally not produce eddy currents, thus the
internal roughness of the pipe has little or no effect on the frictional resistance to flow in the pipe.
This condition is known as laminar flow.
Reynold's Number Calculation
The Reynolds number (Re) of a flowing fluid is calculated by multiplying the fluid velocity
by the internal pipe diameter (to obtain the inertia force of the fluid) and then dividing the
result by the kinematic viscosity (viscous force per unit length).

Kinematic viscosity = dynamic viscosity/fluid density

Reynolds number = (Fluid velocity x Internal pipe diameter) / Kinematic viscosity
Laminar Flow in a Pipe
Laminar flow occurs when the calculated Reynolds number is less than 2300 and in this case
the resistance to flow is independent of the pipe wall roughness.
Turbulent Flow in a Pipe
Turbulent flow occurs when the Reynolds number calculation exceeds 4000.
When Eddy currents occur within the flow, the ratio of the pipe's internal roughness to the
internal diameter of the pipe needs to be considered to calculate the friction factor, which in
turn is used to calculate the friction loss that occurs.
For pipes with a small diameter, the internal roughness can have a major influence on the
friction factor. For pipes with a large diameter the overall effect of the eddy currents are less
significant.
You can use this link to view information on the internal roughness for various pipe materials.
The relative roughness of the pipe and the Reynold's number can be used to plot the friction
factor chart.

When flow occurs between the Laminar and Turbulent flow conditions (Re 2300 to Re 4000)
the flow condition is known as critical and is difficult to predict. Here the flow is neither
wholly laminar nor wholly turbulent. It is a combination of the two flow conditions.
The Colebrook-White equation is used to calculate the friction factor for turbulent flow.
The friction factor is then used in the Darcy-Weisbach formula to calculate the fluid frictional
loss in a pipe.
Pipe Fittings Loss Calculations with K
Factors
Pipe fittings, valves and bends usually have some associated K factor or local loss coefficient, which
allows the calculation of the pressure loss through the fitting for a particular fluid flowing at a
specified velocity. Manufacturers of pipe work fittings and valves often publish a fitting's associated
'K' factor.
Pipe Fitting Loss Formula

Fluid head loss through a fitting can be calculated by the following equation:

h = K x v / 2g

where
h = pressure loss in terms of fluid head, i.e. fluid head loss
K = manufacturer's published 'K' factor for the fitting
v = velocity of fluid
g = acceleration due to gravity
Where the length of the pipe is relatively long, the effect of the fitting losses are usually
considered as minor losses, and are often ignored during initial analysis of the pipe system.
If the piping design contains a partially open valve then the effect and head loss through the
valve should always be included since the valve head loss may turn out to be significant.
Pipe Fittings and K factors database
Our Pipe Flow Expert software has a database that contains the K factors for many different
types of valves and fittings. It also has special wizard helpers that can calculate the K factor
for special types of fittings such as:
gradual enlargements
gradual contractions
sudden enlargements
sudden contractions
rounded entrances
long pipe bends
Addition information about losses through pipe fittings is published in 'Flow of Fluids
through valves, fittings and pipe' - Crane Technical Paper No. 410.





Chapter 11 : Applications of Viscous Flows Through Pipes

Lecture 37 :



Losses In Pipe Fittings
An additional loss of head takes place in the course of flow through pipe fittings like
valves, couplings and so on. In-general, more restricted the passage is, greater is
the loss of head.
For turbulent flow, the losses are proportional to the square of the average flow
velocity and are usually expressed by , where V is the average velocity of
flow. The value of K depends on the exact shape of the flow passages. Typical
values of K are
Approximate Loss Coefficients, K for Commercial Pipe Fittings .
Type and position of fittings Values of K


Globe valve,wide open 10
Gate valve, wide open 0.2
Gate valve, three-quarters open 1.15
Gate valve, half open 5.6
Gate valve, quarter open 24
Pump foot valve 1.5
90elbow(threaded) 0.9
45elbow(threaded) 0.4
Side outlet of T junction 1.8


Since the eddies generated by fittings persist for some distance downstream, the
total loss of head caused by two fittings close together is not necessarily the same
as the sum of the losses which,each alone would cause.
These losses are sometimes expressed in terms of an equivalent length of an
unobstructed straight pipe in which an equal loss would occur for the same average
flow velocity. That is

(37.5)
where, represents the equivalent length which is usually expressed in terms of
the pipe diameter as given by Eq. (37.5). Thus depends upon the friction
factor f , and therefore on the Reynolds number and roughness of the pipe.
.





Chapter 11 : Applications of Viscous Flows Through Pipes

Lecture 37 :


Power Transmission By A Pipeline
In certain occasions, hydraulic power is transmitted by conveying fluid
through a pipeline. For example, water from a reservoir at a high altitude is
often conveyed by a pipeline to an impulse hydraulic turbine in an
hydroelectric power station. The hydrostatic head of water is thus
transmitted by a pipeline. Let us analyse the efficiency of power transmission
under this situation.

Fig. 37.3 Transmission of hydraulic power by a pipeline to a turbine
The potential head of water in the reservoir = H ( the difference in the water level in
the reservoir and the turbine center)
The head available at the pipe exit (or at the turbine entry)
Where is the loss of head in the pipeline due to friction.
Assuming that the friction coefficient and other loss coefficients are constant,
we can write

Where Q is the volume flow rate and R is the hydraulic resistance of the pipeline.
Therefore, the power available P at the exit of the pipeline becomes

For P to be maximum, for a given head H, dP/dQ should be zero. This gives
(37.6)
or,

is always negative which shows that P has only a maximum value (not
a minimum) with Q.
From Eq. (37.6), we can say that maximum power is obtained when one
third of the head available at the source (reservoir) is lost due to friction in
the flow.
The efficiency of power transmission is defined as

(37.7)
1. The efficiency equals to unity for the trivial case of Q = 0.
2. For flow to commence and hence is a monotonically decreasing
function of Q from a maximum value of unity to zero.
3. The zero value of corresponds to the situation given by
when the head H available at the
reservoir is totally lost to overcome friction in the flow through the
pipe.
The efficiency of transmission at the condition of maximum power delivered
is obtained by substituting RQ
2
from Eq. (37.6) in Eq. (37.7) as


Therefore the maximum power transmission efficiency through a pipeline is 67%.


Exercise Problems - Chapter 11
1. Calculate the force F required on the piston to discharge of water through a syringe
(see Fig. 37.4), taking into account the frictional loss in the syringe needle only. Assume fully
developed laminar flow in the syringe needle. Take the dynamic viscosity of water .

Figure 37.4
2. A hydrocarbon oil (viscosity 0.025 pa-s and density 900 kg/m
3
) is transported using a 0.6 m
diameter, 10 km long pipe. The maximum allowable pressure drop across the pipe length is 1 MPa.
Due to a maintenance schedule on this pipeline, it is required to use a 0.4 m diameter, 10 km long
pipe to pump the oil at the same volumetric flow rate as in the earlier case. Estimate the pressure
drop for the 0.4 m diameter pipe. Assume both pipes to be hydrodynamically smooth and in the range
of operating conditions, the Fanning friction factor is given by:

3. Two reservoirs 1 and 2 are connected as shown in the Fig 37.5 through a turbine T. Given the
friction factor relation

for the connecting pipes, the turbine characteristics of water [ Q in m
3
/s] and an
ideal draft tube at the discharge end, find (a) the volume flow rate between the two reservoirs
and (b) the power developed by the turbine. Note:




Use an initial guess for power developed by the turbine as 1 MW. Show only two iterations . Also H
is head available at the turbine.

figure 37.5
Recap
In this course you have learnt the following

The Fannings friction coefficient for a flow through a closed duct,
in terms of wall shear stress, and C
f
= ( )(D
h
/L)P*/(1/2) V
2
in terms of piezometric
pressure drop Darcys friction factor is defined as f = 4C
f


Loss of head in a pipe flow is expressed in terms of Darcys friction factor as h
f
=
f(L/D)(V
2
/2g)

Friction factor in case of laminar fully developed flow is found by N-S equation and is
given by f = 64/Re. Friction factor for turbulent flow depends both on Re and the
roughness at pipe surface.

Flows, in practice, takes place through several pipes together either in series or parallel
or in combination of both of them. The relationship between the head causing the flow
H and flow rate Q can be expressed as H= RQ
2
, where R is the flow resistance in
the hydraulic path.

The loss of head due to friction over a length L of a pipe. Where the entire flow is
drained off uniformly from the side tappings, becomes 1/3 of that in a pipe of same
length and diameter, but without side tappings.

An additional head loss over that due to pipe friction takes place in a flow through pipe
bends and pipe fittings like valves, couplings and so on.

The hydraulic power can be transmitted by a pipeline. For a maximum power
transmission, the head due to friction in the flow equals to one third of the head at
source to be transmitted. The maximum power transmitted efficiency is 67%.