LINE SIZING

1 Line Sizing

According to a 1979 American survey, as much as 30% of the total cost of

a typical chemical process plant goes for piping, piping elements and valves. A
significant amount of operating cost (energy) is also used up in forcing flow
through its components.

A significant amount of the maintenance cost is also for the piping and
associated things. Proper sizing, optimal in some sense, is therefore very
necessary.

Piping must be sized before the plant is laid out. Layout must be complete (i.e.
equipment must be located, pipe racks established, layout of individual pipe runs
decided, etc.) for calculating realistic pressure drop and doing pipe sizing for
each pipe segment. This 'chicken and egg* scenario means that decisions
regarding pipe sizing and plant layout must be iterative in most cases. That is
normally not the practice except in few very large engineering organizations,
which can afford it. Having to carry out pipe sizing at a premature stage
invariably means that the recommended pipe size may not meet process
requirement or may not be the most economic, etc.

Normally a layout is assumed drawing on past practices and experience and

pipes are sized. No second iteration is carried out. Actual layout, which emerges
later, may be significantly different than what was assumed during sizing.

2 Types of Flow

The flow can be categorized on several basis as follows:

1. Single phase, Two phase, Multiphase

2. Horizontal, Inclined
3. Through straight run-pipes, through complex routings
4. Isothermal, non-isothermal
5. Incompressible, compressible
6. Laminar, Turbulent

Although the flow can be categorized on several basis the classification

based on number of phases involved is the most commonly used. When the
flowing medium has uniform physical properties across the flow cross-section, the
flow is a single-phase flow. Flow of pure single liquids, solutions of solids in liquids,
mixtures of completely miscible liquids, mixtures of gases and/or vapors come in
this category.

All other flows are multiphase flows. The two-phase flow would involve two
distinct phases such as liquid with its vapor, a liquid, with an incondensable gas,
etc. A liquid or gas/vapor stream with suspended solid particles is also a two-
phase flow. However, a two-phase flow would normally refer to two fluid phases.
When two immiscible liquids are involved with their vapor and/or another inert
gas, it is a three-phase flow and so on.

Energy required to sustain such flows in pipes/tubes is very important

information, which has to be generated through calculations of pressure drop that
the flow would cause in a conduit of given cross-section, and extent. This
information is then used in locating equipments, sizing pipes, deciding their
routes, rating pressure generating equipments, etc.

Temperature of the flowing medium affects physical properties such as

density and viscosity, which in turn have a bearing on the pressure drop. When
the temperature is constant over the pipe segment under consideration, or the
temperature change along the flow path is not significant enough so as to cause
appreciable change in the physical properties, it is treated as an isothermal flow.
When the temperature change is significant, it is non-isothermal flow. When the
density of the flowing medium is not strongly correlated with the pressure, the
medium is termed as incompressible and the flow as incompressible flow. Liquid
flow (single, two or multiphase) would come in this category naturally. However
when gases/vapors which are compressible (that is their density is a strong
function of pressure) are involved, but the pressure drop along the flow path is
not significant enough to affect the medium density, their flow may also be
treated as incompressible flow. Otherwise, the flow of gases/vapors is a
compressible flow.

In some flow situations, especially two and multiphase flows, the

inclination of the flow conduit from horizontal is of great significance. Also
whether the flow in the inclined conduit upward or downward is also an important
consideration. In the case of single-phase flow, the inclination is important in the
sense that affects the overall energy balance given for flow situation by the
famous Bernoulli's equation.

3 Bernoulli's Equation

In its original form, Bernoulli's equation is merely statement of conversation

of energy for flowing medium. Consider a segment of an inclined conduit of
variable cross-section as shown in Fig 8.1 and fluid flowing through it. The energy
of the fluid at any location may be expressed in terms of a vertical column of the
flowing fluid itself. The height at any point along the conduit is then seen as
comprising of three components, the pressure (P/p), velocity head (v 2/2g) and
elevation head (Z). Bernoulli's theorem states that the sum of these three
components is constant everywhere along the flow path. This is true if there are
no external inputs or withdrawals from the conduit Applied at the two points 1 and
2 of the inclined pipe shown (Fig.), the Bernoulli's equation can be written as
follows:

P1/p + V12/2g + Z = P2/p + V22/2g + Z2

Z2
Z1 1 2

Bernoulli’s Equation

When the pipe is horizontal (Z1= Z2) and the conduit cross-section is
uniform (V1=V2), the pressures at the two points, 1 and 2, should be equal, This is
not the case because the flow is confined by the pipe and there is a resistance to
flow caused by friction between the fluid and the wall, friction between different
layers of fluid flowing at non-optimal. Also, what is optimal today may not be
optimum over a long period (due to fouling, change in relative cost, change in
operating schedule which affects the utilization time of the pipeline, etc.)

4 Pipe Sizing Procedures

Pipe sizing is generally done using one of the following criteria:

 Velocity considerations

 Available pressure drop considerations

 Economic considerations

The degree of difficulty increases as one goes from (1) to (3). While
pressure drop calculation in an integral part of (2) and (3), it would need to be
calculated in case (1) also to quantify energy requirement, sizing pressure
providing equipment such as pumps/ compressors, etc. To be conversant with
pressure drop calculation procedures for variety of flow types that are
encountered is thus very important.
Different velocities and the small or big swirls created in the liquid due to flow
turbulence. Flow against these resistances causes generation of heat raising the
temperature of the fluid as it flows. This temperature rise is not enough to do any
work and this energy transformed into thermal energy is good as lost energy. This
expressed in pressure units or expressed in terms of an equivalent column of the
flowing fluid is called frictional pressure drop or head loss.

Incorporating this fact into the Bernoulli's equation yields the following form
which is generally used in calculating frictional pressure drop in flow:

P1/p + V12/2g + Z = P2/p + V22/2g + Z2 + P/p (8.2)

4.1 Single Phase Pressure Drop Calculations

Single-phase flow is classified as LAMINAR, TRANSIENT OR TURBULENT. The

deciding factor is the REYNOLD'S NUMBER defined as follows:

D
Re  (8.3)


It is ratio of inertia force to viscous force.

It is a Dimensionless number if the quantities are in consistent units. For

Reynolds’s number values up to 2000, the flow is termed laminar and for values
above 4000, it is a turbulent flow. The range 2000-4000 is termed as the
transition region. D in the definition of the Reynolds’s number is the actual
diameter if the flow cross-section is circular such as in commonly used pipes
However, for other cross-section (rectangular, square, annular, etc.), D is defined
in terms of the Hydraulic radius (Rw) as follows:

The HYDRAULIC RADIUS is defined as ratio of flow cross sectional area

to the wetted perimeter, for example, in case of a rectangular cross-section with
sides a and b, the flow cross-section is ab while the wetted perimeter is 2a+2b.
With D defined in this general sense in the definition of Reynolds’s number, the
limiting values of the number for laminar, transient and turbulent flows remain the
same as given earlier. The linear velocity used in the definition of Reynolds’s
number is obtained by dividing the volumetric flow rate by cross-sectional area
for flow.
Alternative but equivalent forms of definition of Reynolds’s number which
are commonly used are as follows:

DG
Re  (8.4)

Where G is the linear mass velocity of fluid.

The frictional pressure drop is calculated using Darcy’s equation as follows.

P f Dv 2
 (8.5)
 2 gD

fD is termed as the Darcy’s friction factor and is related to the Reynolds’s number
and pipe roughness. The applicable and widely used graphs are given in several
textbooks.

For turbulent region, the friction factor value should be read an appropriate
curve for a pipe of roughness  by calculating its ratio with pipe diameter (/D).
The log-log plot is difficult to read and the reading is error prone due to non-
linearity of scale. Several correlations are therefore proposed by various authors
so that the friction factor can be calculated from the Reynolds’s number. Some of
the famous correlations are given later.

In the case of implicit correlations, an iterative approach is necessary to

get the value of the friction factor for given value of Reynolds’s number. Newton-
Rhapson method may be used for getting the value in less iterations.

Farming's equation is also used in place of Darcy's equation as follows:

P 4 f Dv2
 (8.5)
 2 gD

Comparison should show that the Darcy's friction factor is obviously four
times the Farming's friction factor, f F. While using any friction factor vs Reynolds’s
number graph to read friction factor and then while using if in the formula to
calculate the pressure drop, care must be taken to choose the compatible graph
and compatible correlation. This is often a source of error.

Another friction factor is also defined by Churchill (which is half of

Farming's friction factor). The corresponding formula for pressure drop
calculation thus has a factor 8 in the numerator instead of 4 in Farming's
equation. So, one needs to be really very careful in handling this prevailing
multiple definition scenario. Generally, chemical engineering literature uses
Farming's friction factor and Process industry follows the Darcy's friction factor.

If one uses the f vs R e, plot, it is necessary to note whether it is for Fanning,

Darcy or Churchill friction factor. There is a simple way to do it which any
engineer should know. If you don't, ponder over it a little and you would get it.

Several simplified correlations are available to calculate friction factors from

Reynolds’s number under different conditions of flow. Some of the commonly
used ones are given below with reference to the Darcy's definition of friction
factor.
Suitable multiplying factors must be used to convert these correlations for
other friction factors.

LAMINAR REGION

f=64/R (8.6)

TURBULENT REGION

Rough commercial pipes, Re less than 50000:

f=56.8 x lO-10 Re-2 (8.7)

Smooth Pipe, Re less than 3400000

1.126 Re 
f  19.65 ln  1 / 2  (8.8)
 f 

Blazius equation f= 0.046 Re-0.2is also used for fully developed turbulent region.

Most f vs Re plots would mark transition between developing turbulent flows

by a broken line. Most flow situations in process in industry would fall in the
fully developed turbulent region and Blazius equation (especially the one with Re
with exponent -0.2) given above is widely used.

The roughness factor  is dependent on the pipe material and method of

fabrication and some representative values are given in the Table 1. Note the
wide variation in perceptions of the roughness by different authors. In most plots,
Moody's roughness values are used. Because of the variation in friction factor
definition and roughness values, it is advisable to stick to one plot with full
knowledge of the friction factor it pertains to and the roughness values it refers to.
Surface roughness of various pipes and tubes (SI)
Miller Moody Kutateladze
Tubing Drawn 0.0025 0.0015 0.0015-0.01
Glass 0.0025 - 0.0015-0.01
Steel New 0.025 0.046 0.06
Light Rust 0.25 - 0.2
Concrete Smooth 0.025 0.3 0.8
Rough 0.5 3.0 9.0
Cast Iron Uncoated 0.15 0.25 0.3
Coated 0.15 0.12 -
Galvanized Smooth finish 0.025 - -
Normal finish 0.15 0.15 -
 The frictional pressure drop calculated by any of the above methods
should be multiplied by the effective length of the pipe segment to get the net
frictional drop across the segment. This is then used in the Bernoulli's equation to
obtain the actual pressure drop between pipe origin and destination. The
effective length is the actual pipe length if the pipe line is straight and long
enough so that pressure drop due to extra turbulence created at the entrance
when fluid enters the pipe from an equipment or at the exit when the
pipe feeds into another equipment are relatively insignificant as
compared to overall frictional pressure drop. In case the pipe has fittings
such as elbows, tees, valves, expanders, reducers, etc., an hypothetical
straight pipe length of same diameter as the run pipe on which the fittings exits is
added in place of each of the fittings. The effective length is the sum of the
straight-run pipe length plus the total equivalent for all fittings. Entrance and
exit of fluid in and from the pipe segment also adds to turbulence and to
extra pressure drop. This effect is also incorporated by adding equivalent length
of these. The actual equivalent lengths for important fittings are given in real
terms (i.e. length of pipe to be added)

In another approach, equivalent lengths of fittings are mentioned in terms

of diameters of the pipe. This number should then be multiplied by the pipe size
to get the equivalent length of pipe to be added. The equivalent lengths for
valves and fittings in terms of diameters are reported in several books and are
not given here.

Using actual pipe length as per tables is a more accurate approach.

Above procedure is applicable to fluids, i.e., liquids and gases.

In cases the temperature varies across the pipe segment, the physical
properties vary. Also if the fluid is gas/vapor, its volumetric flow rate may vary due
to pressure changes arising out of temperature change as well as due to
pressure drop. To account for these effects, it may be a good practice to divide
the whole line into segments over each of which, the temperature changes is not
so significant as to change the properties drastically. The properties are suitably
updated to incorporate temperature and pressure changes as one traverse these
hypothetical segments. Calculation over all the segments thus gives the total
pressure drop.
Resistances of Elbows, Tees and Bends
(Resistance in equivalent pipe length, ft)
0
Nominal 90 Elbows 900 Bends Tee
Pipe R = 1D R =1.5 D R = 5D R = 10 Flow Flow
Size, in D through through
Branch straight
1.5 4.5 3 2.5 4 8 3
2 5.25 3.5 3 5 11 3.5
2.5 6 4 3.5 6 13 4
3 7.5 5 4 7.5 16 5
 4 10.5 7 5.5 10 20 7
6 15 10 8.5 15 30 10
8 21 14 11 20 40 14
10 24 16 14 25 50 16
12 32 21 16 30 60 21
14 33 22 19 33 65 22
16 39 26 21 38 75 26
18 44 29 24 42 86 29
20 48 32 27 50 100 32
24 57 38 32 60 120 38

Change in pressure across the pipe may be of importance in case of

compressible fluids. It may be ignored if it is less than 10% of the total fluid
pressure. However, if it is more than this engineering tolerance, above approach
of segmenting the pipeline may be adopted. A good practice would be to
calculate pressure drop over the pipe run assuming fluid properties at inlet or
average temperature/pressure conditions to begin with. If the pressure drop so
calculated is within 10% or less of the actual pressure levels at which the fluid is
flowing, one may ignore the effect of temperature/pressure change. If the
pressure drop exceeds 10% of flow pressure, the above approach of segmenting
may be restored to.

5 Two Phase Flow

Pressure drop in the case of a two phase flow is dependent on the flow
regime. For two phase flow conditions, 7 regimes are possible as distributive flow
(bubble or mist), intermittent flow (plug or slug), segregated flow (stratified or
annular) and wavy flow. Flow regime identification is done by following Baker's
procedure.

The Baker parameter values are then used to identify the flow regime from the
plot. Slug flow must be avoided in process piping applications.

The pressure drop calculation then proceeds as per several correlations offered
by several researchers. Only two commonly used ones discussed here.

5.1 Lockhart Martinelli Method

Assuming that that only the liquid flows in the pipe line, calculate the
pressure drop that it would cause over unit length, (P)L. Similarly, considering
that only vapor/gas flows in the pipe, calculate the pressure drop per unit length,
(P) V. Single phase correlations are to be used in getting these two pressure
drops. Lockhart Martinelli Modulus, X, is then defined as follows;

X 2  (P ) L /( P )V (8.9)
 For this value of modulus, a multiplier Y L or YV is then read from the plot
and it is appropriately used in one of the following relations to get the two phase
pressure drop, (P)LV per unit length. Multiplying this with the effective length
(after including equivalent lengths of the fittings) of the pipe, one gets the total
two-phase frictional drop.

(P)LV = YL(P)L (8.10)

5.2 Baker's method

Depending on the regime identified earlier, an appropriate correlation or

plot is used to get Baker's modulus,  and it is multiplied with pressure drop with
only gas flowing to get the two phase pressure drop.

These correlations were derived by the respective authors by extensive

experimentation on air-water flow, but mostly on smaller diameter pipes. There
applicability for larger dimension industrial pipes is suspect. However, these
remain the most used correlations. Better approaches to two phase flow pressure
drop estimation are available but are seldom used.

In two phase flow calculations, confidence levels are low. Also, it is not
safe to overdesign here as the flow regime may change and one may get an
undesirable flow regime such as slug flow. Extreme precaution is therefore
necessary at engineering stage in designing pipes for two phase flow and must
be ready to handle problems that may surface at the commissioning stage.

The Baker map is applicable only if the flow line is horizontal. Inclination
has a great effect on flow pattern and the flow regime may change for same
vapor and liquid flows in same size pipe line if the inclinations are different. Also,
in inclined pipes, it matters whether the flow is upward or downward. Extensive
work has been reported on these aspects but industrial practices ignore this fact

6 Jacketed Piping

In a Chemical Process Industry transportation of material, especially in

fluid form, poses a variety of problems. The problems are more when the fluid is
viscous and has to be maintained at higher temperatures than the ambient
throughout transport. Variations in ambient temperature from winter to summer
also affect the flow characteristics of the liquid. The fluid inside the pipe can also
undergo phase changes and the viscosity can change to adversely affect the fluid
flow pattern. If it is only the atmospheric changes that create problems, then the
temperature variation in the pipe is kept within the acceptable range by heat
tracing. If the fluid has to be kept at a certain temperature all throughout the
process of transportation, then the pipelines need jacketing. When a pipe of
higher diameter is put over the service pipe, and when heating and cooling
medium passes as required, through the annular space created between the
two pipes, then it is termed as a jacketed pipe. The inner pipe is called the core pipe
and the outer pipe is called the jacket.

6.1 Mechanical design of jacketed piping

There are two numbers of pipes involved in jacketed piping with different
design considerations. The core pipe is subjected to internal pressure when there
is fluid flow through the same. The pipe is subjected to external pressure when
there is fluid flow through the jacket. It could be that these pressures get balanced
and the pipe gets relaxed. But for mechanical design, the worst condition is to
be considered. For any cylindrical surface subjected to external pressure, design
is more complicated than for those subjected to internal pressure.

Code ASME B31.3 under clause 304.1.3 specifies that to determine the wall
thickness for straight pipe under external pressure, the procedure outlined in the
BPV Code Section VIII Division l.UG-28 throuuh UG-30 shall be followed.
The combination of core and jacket pipes shall be selected based on:

1. The properties of the heating /cooling medium.

2. The flow required to maintain the temperature.
3. The criticality of the service.
4. The differential expansion of the core and jacket when the material of
construction of core and jacket are different.

The jacketed pipe poses problems, in design, fabrication and erection,

different from that of the non-jacketed piping. This article is intended to highlight
the problems of mechanical design of jacketed piping.

BPV Code Section VIII Division 1,UG-2S tlirough UG-30 shall be followed.
When standard piping details are prepared the design length L cannot be exactly
predicted. Hence the length to diameter (L/D) ratio of 50 is considered as a
standard practice. Since the pipes are manufactured with standard thicknesses,
the selection has to be done from the available thickness ranges. The thickness
considered for calculation should be after allowing for the mill tolerance and the
corrosion allowance.

Size Combination of Core and Jacket Pipe:

i) Straight Pipe:

The size combination of the Core and jacket pipes are determined by the
annular space necessary to obtain the required flow to maintain the heat transfer.
In the absence of any specific process data, the following combinations are most
widely used:

Table 8.3 Core – Jacket for Pipe

Size of Core Pipe 15 20 25 40 50 65 80 100 150
NB
(mm)
Size of Jacket Pipe 40 40 50 65 80 100 100 150 200
NB
(mm)

ii) Elbows:

The bending radius of elbows/bends are so selected that the core pipe
and the jacket pipe maintain the same centerlines even at the change of direction.
The ideal combinations of the core and jacket bends shall be as follows: -

Table 8.4 Core – Jacket for Bend

Core pipe Bends Radius Jacket Pipe Bend Radius Remarks
NB (mm) (mm) NB (mm) (mm)
15 60 4D 40 57 LSD NOTE 1
20 60 3D 40 SI LSD NOTE 1
25 15 3D 50 76 LSD NOTE 1
40 57 LSD 65 62 L0D NOTE 2
50 16 LSD 30 76 ID NOTE 2
65 95 LSD 100 102 ID NOTE 2
80 114 LSD 100 102 ID NOTE 2
100 152 LSD 150 152 ID NOTE 2
150 229 LSD 200 203 ID NOTE 2

NOTE: i) Use 1.5D(LR) std. elbow for jacket.

ii)Use 1.5D(LR) std. elbow for core and 1D(SR) std elbow for jacket.

6.3 Thickness of Jacket Pipe:

Jacket pipes have to be designed for internal pressure, which the jacket
fluid exerts. The formula given in clause 319.4.1 of the code shall be followed
with the standard procedure.
Types of Jacketing: Depending upon the criticality of the requirement, the
jacketing of the piping system can be done:

i) Only on straight pipe keeping all bends and flange welds exposed.
ii) On straight pipes and elbows but keeping the flange size same as that of
the core pipe
iii) On straight pipes and elbows with flange size that of the jacket pipe.
(Reducing flanges)

This type is used where jacketing requirements are critical. The slip on type
flanges are modified to get a seating of the jacket pipe to achieve a proper
welding joint.

7 Jumper Pipes
To maintain the continuation of fluid flow in the jacket, jumper pipes are
provided. The location of jumper pipes on the horizontal jacketed pipe is decided
based on the type of fluid in the jacket pipe. There can be a single jumper or two
jumpers and these can be placed in one of the following patterns:

When the heating fluid is in vapor form and no condensation is expected,

arrangement as per Pattern I can be used. The arrangement in Pattern II makes
the vapor phase as well as the liquid phase continuous and is ideal when steam is
used as a heating medium in the jacket. When the jacketing fluid is a liquid under
sufficient pressure, arrangement as per Pattern III or Pattern IV can be used.
The arrangement for connection as per Pattern IV is difficult to fabricate, as the
hole on the jacket pipe has to be cut to profile. Pattern V shows the arrangement
in a vertical pipeline. In all types of arrangements it should be ensured that the
jumper joins the jacket pipe at minimum distance from the breakout flange in
order to avoid cold / hot spots due to stagnancy. The jumper should also be
provided with dismantling arrangement, either flange joints or unions, flange joint
being preferred.

Spacers:

In order to keep the core pipe concentric with the jacket pipe, supports are
provided at definite intervals. These are done by welding flat or bar to the core
pipe called spacers. These spacers will be stitch welded to the core pipe. Flats
are preferred as the restrictions in the jacket flow are minimum in this case. The
arrangement shall be as shown in the sketch below:

NOTES: *
Material of spacers shall as that of the core pipe.
Spacers near to pipe bend should be located at least 1000 mm away from the
centerline of bend.
Jacket Barriers:

It is absolutely essential that the proper flow of the fluid in the jacket is
maintained for proper heat transfer. Whenever there is stagnancy or inadequate
flow in the jacket, the hot spot or the cold spot gets formed which affects the
process fluid flow in the core pipe. To avoid this, the fluid in the jacket may have to
be directed properly. This is established by the provision of jacket barriers inside
the jacket. These are mainly used where there are branch -offs.

8 Plastic Piping

The latest entry in to the materials list for manufacturing pipe is plastic.
Plastic has emerged as a reliable, safe and cost effective alternative material.
There is a broad range of plastic compounds being developed today.

For piping systems, two categories are most effective: fluoroplastics and
thermoplastics. Fluoroplastics are found in materials like PTFE, PVDF, ECTFE,
CTFE, PFA, and FEP. As a group, fluoroplastics perform extremely well in
aggressive chemical services at temperatures from -328 0F to +5000F.
Thermoplastics are those that require melting during the manufacturing process.
These plastics can be welded or injection molded into shapes for machining into
piping system components.

For some piping systems, it is now inconceivable not to use plastics. Pipes
made from plastic are replacing traditional, expensive materials like glass or
ceramic-lined pipe. Some plastics such as UHMW PE, PVDF, CTFE, and nylon
have such excellent wear resistant that they prove in Taber Abrasion Tests to be
five to ten times better in this regard than 304 Stainless Steel.

8.1 Joining Plastic Pipe

Plastic pipe can be joined by one of the following methods: threading,

solvent cement, or fusion. Threading plastic is not a viable option because it is
expensive. Heavy wall thicknesses are required, and leaks from high pressures
and expansion and contraction are difficult to control. Joints made with solvent
cement have been proven more reliable. Though, once hardened, cemented
joints cannot be disassembled. They offer good resistance to abrasive chemical
and high-pressure commodities and are available in a large selection of fittings
without the need of threads. Heat fusion must be performed on plastic
compounds that are resistant to chemical solvents. Pipe can be either be butt-
joined or socket-joined. Heat fusion can be used with thinner wall thickness and
are pressure resistant beyond the burst pressure of the pipe.
 Though fabrication with plastic pipe may sound simple, caution must be
exercised when using plastic pipe. The effectiveness of a particular grade of
plastic must be tested before it is chosen for a particular service. Four important
variables must be evaluated: chemical resistance, pressure limitations,
temperature limitations, and stress. The various molecular components of plastic
make them susceptible top chemical reactions with certain compounds.
Hazardous mixtures must be avoided. Pressure and temperature limitations must
be established for obvious reasons. Excessive stresses in the form of restricted
expansion and contraction, and frequent or sudden changes in internal pressure
and temperature must be avoided.