NG Production Lecture 11
NG Production Lecture 11
• Lecture 11
1
Transportation
The transmission of natural gas to the consumer may be divided
into three distinct pipeline units:
1. The gathering system
2. The main truck line
3. Distribution Lines
This part focuses on design and operation of natural gas
pipelines in onshore and offshore fields
Pipeline Design
Many factors must be considered in the design of long-
distance gas pipelines. These include:
1. The nature and volume of the gas to be transmitted.
2. The length of the line.
3. The type of terrain to be crossed. 2
4. Maximum elevation of the route.
Sizing Pipelines
The capacity of gas transmission of a pipeline is
controlled mainly by its size.
Complex equations have been developed for sizing
natural gas pipelines in various flow conditions.
The Weymouth equation, the Panhandle equation, and
the Modified-Panhandle equation are used for relating
the volume transmitted through a gas pipeline to the
various factors involved, thus deciding the optimum
pressure and pipe dimensions.
From these equations ,pipe diameter and wall thickness
for a desired rate of gas throughout can be calculated. 3
Definition of friction factor
Flow of natural gas in pipelines always results in some mechanical
energy being converted into heat.
In single-phase gas flow in pipe, the internal energy losses due to
v i s c o s i t y e f f e c t s and l o s s e s d u e t o t h e r o u g h n e s s o f t h e
inner wall of the pipeline(skin friction).
Under turbulent flow conditions that always exist in natural gas
transmission pipelines, the energy losses must be determined
experimentally and then correlated as some function of the flow
variables.
The lost work is usually calculated using a friction factor by
dimensional analysis.
𝛆
𝐅𝐫𝐢𝐜𝐭𝐢𝐨𝐧 𝐟𝐚𝐜𝐭𝐨𝐫 = 𝐟 = 𝒇(𝐍𝐑𝐞 , )
𝐃
𝜺 4
Where: 𝑵𝑹𝒆 =Reynolds number , =Relative roughness
𝑫
Definition of friction factor
The first law of thermal dynamics (conservation of energy) is the
theoretical basis of most fluid flow equations
Under steady-state fluid and heat flow conditions, the
conservation of energy gives the expression for the pressure
gradient due to viscous shear or frictional losses.
𝒅𝒑 𝒅𝒍𝒘
( )𝒇 = 𝝆 (11-1)
𝒅𝑳 𝒅𝑳
Where:
𝐝𝐏
( )𝐟 = 𝐩𝐫𝐞𝐬𝐬𝐮𝐫𝐞 𝐠𝐫𝐚𝐝𝐢𝐞𝐧𝐭 𝐝𝐮𝐞 𝐭𝐨 𝐟𝐫𝐢𝐜𝐭𝐢𝐨𝐧 𝐥𝐨𝐬𝐬𝐞𝐬
𝐝𝐋
𝐈𝐛𝐟
𝐩= 𝐩𝐫𝐞𝐬𝐬𝐮𝐫𝐞, 𝟐
𝐟𝐭
𝐈𝐛𝐦
𝛒 = 𝐟𝐥𝐮𝐢𝐝 𝐝𝐞𝐧𝐬𝐢𝐭𝐲,
𝐟𝐭 𝟑
𝒇𝒕−𝑰𝒃𝒇
𝐥𝐰 = mechanical energy (loss of work) converted to heat,
𝑰𝒃𝒎 5
𝑳 =pipe length , 𝒇𝒕
The equation that relates lost work per unit length of pipe and
the flow
𝒅𝒍𝒘 𝒇𝒖𝟐
= (11-2)
𝒅𝑳 𝟐𝒈𝒄 𝑫
Where:
𝐟𝐭
𝐮 = 𝐟𝐥𝐨𝐰 𝐯𝐞𝐥𝐨𝐜𝐢𝐭𝐲,
𝐬𝐞𝐜
𝐈𝐛𝐦 .𝐟𝐭
𝐠 𝐜 = 𝐠𝐫𝐚𝐯𝐢𝐭𝐚𝐭𝐢𝐨𝐧𝐚𝐥 𝐜𝐨𝐧𝐯𝐞𝐫𝐬𝐢𝐨𝐧 𝐟𝐚𝐜𝐭𝐨𝐫 = 𝟑𝟐. 𝟏𝟕𝟒
𝐈𝐛𝐟 .𝐬𝐞𝐜 𝟐
𝐃 = 𝐩𝐢𝐩𝐞 𝐝𝐢𝐚𝐦𝐞𝐭𝐞𝐫, 𝐟𝐭
𝐟 = 𝐌𝐨𝐨𝐝𝐲 𝐟𝐫𝐢𝐜𝐭𝐢𝐨𝐧 𝐟𝐚𝐜𝐭𝐨𝐫
By substituting equation (11-2) into equation (11-1) gives:
𝒅𝑷 𝒅𝒍𝒘 𝝆𝒇𝒖𝟐
( )𝒇 = 𝝆 = (11-3)
𝒅𝑳 𝒅𝑳 𝟐𝒈𝒄 𝑫
A similar equation, using the Fanning friction factor, 𝒇′ is
𝒅𝑷 𝟐𝝆𝒇′ 𝒖𝟐 𝝆𝒇𝒖𝟐 𝟐𝝆𝒇′ 𝒖𝟐
( )𝒇 = , = (11-4) 6
𝒅𝑳 𝒈𝒄 𝑫 𝟐𝒈𝒄 𝑫 𝒈𝒄 𝑫
Therefore, 𝒇 = 𝟒𝒇′ or 𝒇𝑴𝒐𝒐𝒅𝒚 = 𝟒𝒇𝑭𝒂𝒏𝒏𝒊𝒏𝒈
Reynolds Number
The Reynolds number (𝑵𝑹𝒆 ) is defined as the ratio of fluid momentum force to
viscous shear force.
𝐦𝐨𝐦𝐞𝐧𝐭𝐮𝐦 𝐟𝐨𝐫𝐜𝐞
𝐍𝐑𝐞 =
𝐯𝐢𝐬𝐜𝐨𝐮𝐬 𝐟𝐨𝐫𝐜𝐞
𝐝𝐮 𝐝𝐮
𝐅𝐦𝐨𝐦𝐞𝐧𝐭𝐮𝐦 = 𝛒𝐮 = 𝐅𝐥𝐮𝐱 ∗
𝐝𝐥 𝐝𝐥
Example: Water jet striking a surface.
𝐝𝐮
𝐅𝐯𝐢𝐬𝐜𝐨𝐮𝐬 = 𝛍
𝐝𝐥
Example: Honey flowing slowly through a pipe.
The Reynolds number can be expressed as a dimensionless group defined as
𝑫𝒖𝝆
𝑵𝑹𝒆 =
𝝁
Where:
𝐃 = 𝐩𝐢𝐩𝐞 𝐈𝐃 , 𝐟𝐭
𝐟𝐭
𝐮 = 𝐟𝐥𝐮𝐢𝐝 𝐯𝐞𝐥𝐨𝐜𝐢𝐭𝐲,
𝐬𝐞𝐜
𝐈𝐛
𝛒= 𝐟𝐥𝐮𝐢𝐝 𝐝𝐞𝐧𝐬𝐢𝐭𝐲, 𝐦𝟑
𝐟𝐭 7
𝐈𝐛
𝛍= 𝐟𝐥𝐮𝐢𝐝 𝐯𝐢𝐬𝐜𝐨𝐬𝐢𝐭𝐲, 𝐦
𝐟𝐭.𝐬𝐞𝐜
The Reynolds number can be used as a parameter to distinguish between laminar
and turbulent fluid flow.
The change from laminar to turbulent flow is usually assumed to occur at a
Reynolds number of 2,100 for flow in a circular pipe.
If U.S. field units is used:
𝐟𝐭 𝐈𝐛𝐦
𝐃 𝐢𝐧 𝐟𝐭 , 𝐮 𝐢𝐧 , 𝛒 𝐢𝐧 and 𝛍 in 𝐜𝐞𝐧𝐭𝐢𝐩𝐨𝐢𝐬𝐞 (𝐜𝐏)
𝐬𝐞𝐜 𝐟𝐭 𝟑
The Reynolds number equation becomes
𝑫𝒖𝝆
𝑵𝑹𝒆 = 𝟏. 𝟒𝟖𝟖 (11-6)
𝝁
If a gas of specific gravity 𝜸𝒈 and viscosity 𝝁 (𝒄𝑷) is flowing in a pipe with an inner
diameter 𝑫 (𝒊𝒏) at flow rate 𝒒 (𝑴𝒄𝒇𝒅) measured at base conditions of 𝑻𝒃 (°R) and
𝑷𝒃 (𝒑𝒔𝒊𝒂), the Reynolds number can be expressed as:
𝟕𝟏𝟏𝑷𝒃 𝒒𝜸𝒈
𝑵𝑹𝒆 = (11.7)
𝑻𝒃 𝑫𝝁
As 𝑻𝒃 is 𝟓𝟐𝟎 °R and 𝑷𝒃 varies only from 𝟏𝟒. 𝟒 𝒑𝒔𝒊𝒂 to 𝟏𝟓. 𝟎𝟐𝟓 𝒑𝒔𝒊𝒂 in the United States, the
𝟕𝟏𝟏𝑷𝒃
value varies between 𝟏𝟗. 𝟔𝟗 and 𝟐𝟎. 𝟓𝟒 8
𝑻𝒃
For all practical purposes, the Reynolds number for natural gas flow
problems may be expressed as
𝟐𝟎 𝒒𝜸𝒈
𝑵𝑹𝒆 = (11.8)
𝝁𝑫
Where:
𝒒 =gas flow rate at 𝟔𝟎𝒐 𝑭and 𝟏𝟒. 𝟕𝟑 𝒑𝒔𝒊𝒂 , 𝑴𝒄𝒇𝒅
𝜸𝒈 =gas specific gravity (air =1)
𝝁 =gas viscosity (𝒄𝑷) at in - situ temperature and pressure
𝑫 =pipe diameter in inch
9
Relative Roughness
The frictional losses of fluid energy and pressure depend on the roughness of the
inside wall of a pipe.
Wall roughness is a function of:
1. Pipe material.
2. Method of manufacture.
3. The environment to which it has been exposed.
From a microscopic sense, wall roughness is not uniform, and thus the distance
from the peaks to valleys on the wall surface will vary greatly.
The absolute roughness of a pipe is a measure of the average height of surface
irregularities (roughness) on the inner surface of the pipe.
It’s usually denoted by ε (epsilon) and expressed in millimeters (mm) or feet
10
11
The effect of roughness is not due to its absolute dimensions, but to its dimensions
relative to the inside diameter of pipe.
Relative roughness, 𝒆𝑫 , is defined as the ratio of the absolute roughness to
the pipe internal diameter:
𝜺
𝒆𝑫 =
𝑫
Where 𝜺 and 𝑫 have the same unit.
The absolute roughness is not a directly measurable property for a pipe,
which makes the selection of value of pipe wall roughness difficult.
The way to evaluate the absolute roughness is to compare the pressure
gradients obtained from the pipe of interest with a pipe that is sand-
roughened.
If a pipe's inner wall is coated with sand grains of known size 𝜺, that pipe is
considered sand roughened, and it helps standardize the study of friction and
pressure drop in turbulent flow.
If measured pressure gradients are available, the friction factor and
Reynolds number can be calculated and an effective 𝒆𝑫 obtained from the
Moody diagram 12
If no information is available on roughness, a value of ε = 0.0006 inches is
recommended for tubing and line pipes.
13
Equations for Friction Factor
Fluid flow ranges in nature between two extremes: laminar or streamline flow
and turbulent flow. Within this range are four distinct regions
Friction factor,𝒇 can be obtained from Moody friction factor chart.
It is a log-log graph of (𝐥𝐨𝐠 𝐟) versus (𝐥𝐨𝐠 𝐍𝐑𝐞 ).
In a Moody diagram, there are four distinct regions:
1. Laminar
2. Critical
3. Transition
4. Turbulent
Due to the characteristics of the complex nature of the curves, the
equation for the friction factor in terms of the Reynolds number and
relative roughness varies for each of the four regions.
Figure 11–4 is a Moody friction factor chart covering the full range of flow 14
conditions
15
Laminar Single-Phase Flow
The “𝒇” factor for laminar flow can be determined analytically.
The Hagen-Poiseuille equation for laminar flow is:
𝒅𝒑 𝟑𝟐𝝁𝒖
( )𝒇 = (11.10)
𝒅𝑳 𝒈𝒄 𝑫𝟐
Equating the frictional pressure gradients given by Equation (11.3) and
𝒅𝑷 𝒅𝒍𝒘 𝝆𝒇𝒖𝟐
Equation (11.10) gives , ( )𝒇 = 𝝆 = (11-3)
𝒅𝑳 𝒅𝑳 𝟐𝒈𝒄 𝑫
𝝆𝒇𝒖𝟐 𝟑𝟐𝝁𝒖
= (11-11)
𝟐𝒈𝒄 𝑫 𝒈𝒄 𝑫𝟐
𝟔𝟒𝝁 𝟔𝟒
𝐟 = 𝐃𝐚𝐫𝐜𝐲 𝐟𝐫𝐢𝐜𝐭𝐢𝐨𝐧 𝐟𝐚𝐜𝐭𝐨𝐫 = = (11-12)
𝑫𝝆𝒖 𝑵𝑹𝒆
16
Turbulent Single-Phase Flow
Studies of turbulent flow have shown that the velocity profile and pressure
gradient are very sensitive to the characteristics of the pipe wall, that is, the
roughness of the wall.
Although a number of empirical correlations for friction factors are available, only
the most accurate ones are presented.
For smooth wall pipes in the turbulent flow region, Drew, Koo, and McAdams
(1930) presented the most commonly used correlation
𝟎.𝟓
𝒇 = 𝟎. 𝟎𝟎𝟓𝟔 + 𝟎.𝟑𝟐 (11-13)
𝑵𝑹𝒆
which is valid over a wide range of Reynolds numbers
𝟑 ∗ 𝟏𝟎𝟑 < 𝑵𝑹𝒆 < 𝟑 ∗ 𝟏𝟎𝟔
For rough wall pipes in the turbulent flow region, the effect of wall roughness on
friction factor depends on the relative roughness and Reynolds number.
Nikuradse (1933) friction factor correlation is still the best one available for fully
developed turbulent flow in rough pipes:
𝟏 17
= 𝟏. 𝟕𝟒 − 𝟐 log(𝟐𝒆𝑫 ) (11-14)
𝒇
This equation is valid for large values of the Reynolds number where the
𝟏
effect of relative roughness is dominant. ( 𝒇 = 𝟏. 𝟕𝟒 − 𝟐 log(𝟐𝒆𝑫 ) )
Guo and Ghalambor (2002) showed that the Nikuradse friction factor
correlation is also valid for gas flows with solid particles and liquid mist.
The correlation that is used as the basis for modern friction factor
charts was proposed by Colebrook (1938):
𝟏 𝟏𝟖.𝟕
= 𝟏. 𝟕𝟒 − 𝟐 log(𝟐𝒆𝑫 + ) (11-15)
𝒇 𝑵𝑹𝒆 𝒇
This equation is applicable to smooth pipes and to flow in transition
and fully rough zones of turbulent flow
Equation (11.15) is not explicit in 𝒇. However, values of 𝒇 can be obtained by an
numerical procedure such as Newton-Raphson iteration.
An explicit correlation for friction factor was presented by Jain (1976):
𝟏 𝟐𝟏.𝟐𝟓
= 𝟏. 𝟏𝟒 − 𝟐 log(𝒆𝑫 + 𝟎.𝟗 ) (11-16) 18
𝒇 𝑵𝑹𝒆
19
Theoretical Pipeline Equations
There are several equations in the petroleum industry for calculating
the flow of gases in pipelines
The most common pipeline flow equation is:
The Weymouth equation, which is generally preferred for smaller-
diameter lines (D ≤ 15 in).
The Panhandle equation and the Modified Panhandle equation are
usually better for larger-sized transmission lines.
Based on the first law of thermodynamics, the total pressure gradient
is made up of three distinct components
𝒅𝒑 𝒈 𝒇𝝆𝒖𝟐 𝝆𝒖𝒅𝒖
= 𝝆 sin 𝜽 + + (11-17)
𝒅𝑳 𝒈𝒄 𝟐𝒈𝒄 𝑫 𝒈𝒄 𝒅𝑳
𝒈
𝝆 𝒔𝒊𝒏 𝜽 =the component due to elevation or potential energy change
𝒈𝒄
𝒇𝝆𝒖𝟐
=the component due to frictional losses
𝟐𝒈𝒄 𝑫
20
𝝆𝒖𝒅𝒖
=the component due to acceleration or kinetic energy change
𝒈𝒄 𝒅𝑳
The elevation component is pipe angle dependent. It is zero for
horizontal flow.
The friction loss component applies to any type of flow (laminar or
turbulent) at any pipe angle and causes a pressure drop in the
direction of flow.
The acceleration component causes a pressure drop in the direction
of velocity increase in any flow condition in which velocity changes
occurs. It is zero for constant-area, incompressible flow.
Equation (11.17) applies for any fluid
in steady-state,
one-dimensional flow for which ρ, f, and u can be defined
𝒅𝒑 𝒈 𝒇𝝆𝒖𝟐 𝝆𝒖𝒅𝒖
= 𝝆 sin 𝜽 + + (11-17)
𝒅𝑳 𝒈𝒄 𝟐𝒈𝒄 𝑫 𝒈𝒄 𝒅𝑳
21
Consider the followings:
Steady-state flow of dry gas
𝝆𝒖𝒅𝒖
Constant-diameter, =𝟎
𝒈𝒄 𝒅𝑳
𝒈
Horizontal pipeline, 𝝆 sin 𝜽 = 𝟎.
𝒈𝒄
The mechanical energy equation, Equation (11.3), becomes
𝒅𝑷 𝒅𝒍 𝝆𝒇𝒖𝟐
( )𝒇 = 𝝆 𝒘 = (11-3)
𝒅𝑳 𝒅𝑳 𝟐𝒈𝒄 𝑫
𝒇𝒕−𝑰𝒃𝒇
𝐥𝐰 = mechanical energy (loss of work) converted to heat,
𝑰𝒃𝒎
𝒑(𝑴𝑾)𝒂
𝝆=
𝒛𝑹𝑻
𝒅𝑷 𝒅𝒍𝒘 𝝆𝒇𝒖𝟐 𝒑(𝑴𝑾)𝒂 𝒇𝒖𝟐
( )𝒇 = 𝝆 = = (11-18)
𝒅𝑳 𝒅𝑳 𝟐𝒈𝒄 𝑫 𝒛𝑹𝑻 𝟐𝒈𝒄 𝑫
22
Where,
(𝑴𝑾)𝒂 =apparent molecular weight
This equation is the base for development of many pipeline equations
Integrating Equation (11.18) gives
(𝑴𝑾)𝒂 𝒇𝒖𝟐 𝒑
𝒅𝒑 = 𝒅𝑳 (11-19)
𝑹 𝟐𝒈𝒄 𝑫 𝒛𝑻
If temperature is assumed to be constant at average value in a pipeline, 𝑻 , and gas
deviation factor, 𝒛 , is evaluated at average temperature and average pressure, 𝒑 ,
Equation (11.19) can be evaluated over a distance 𝑳 between upstream pressure, 𝒑𝟏 ,
and downstream pressure, 𝒑𝟐 :
𝟐𝟓𝜸𝒈 𝒒𝟐 𝑻𝒛𝒇𝑳
𝒑𝟏 𝟐 − 𝒑𝟐 𝟐 = (11-20)
𝑫𝟓
Where:
𝛄𝐠 = 𝐠𝐚𝐬 𝐠𝐫𝐚𝐯𝐢𝐭𝐲(𝐚𝐢𝐫 = 𝟏)
𝐪 = gas flow rate ,𝐌𝐌𝐬𝐜𝐟𝐝(𝐚𝐭 𝟏𝟒. 𝟕 𝐩𝐬𝐢𝐚, 𝟔𝟎𝐨 𝐅
𝐓 =average temperature,𝐨𝐑
𝐳 = gas deviation factor at 𝐓 and 𝐩
𝐩𝟏 +𝐩𝟐
𝐩=
𝟐
𝐋 = 𝐩𝐢𝐩𝐞 𝐥𝐞𝐧𝐠𝐭𝐡, 𝐟𝐭 23
D = pipe internal diameter, in.
f = Moody friction factor
Equation (11.20) may be written in terms of flow rate measured at arbitrary
base conditions (𝑻𝒃 and 𝒑𝒃 ):
𝟐𝟓𝜸𝒈 𝒒𝟐 𝑻𝒛𝒇𝑳
𝒑𝟏 𝟐 − 𝒑𝟐 𝟐 = (11-20)
𝑫𝟓
𝑪𝑻𝒃 (𝒑𝟏 𝟐 −𝒑𝟐 𝟐 )𝑫𝟓
𝒒= (11-21)
𝒑𝒃 𝜸𝒈 𝑻𝒛𝒇𝑳
where 𝑪 is a constant with a numerical value that depends on the
units used in the pipeline equation.
If 𝑳 is in miles and 𝒒 is in 𝒔𝒄𝒇𝒅, 𝑪 = 𝟕𝟕. 𝟓𝟒.
The use of Equation (11.21) involves an iterative procedure
The gas deviation factor ,𝒛, depends on pressure,𝒑, and the friction factor, 𝒇,
depends on flow rate
𝐳 = 𝐟(𝐩),
𝐟 = 𝐟(𝐟𝐥𝐨𝐰 𝐫𝐚𝐭𝐞)
This problem prompted several investigators to develop pipeline flow 24
equations that are noniterative or explicit