TECHNICAL MEMO

To: Manager, Process Technologies

From: [insert your name]
Date: [insert submission date]
Re: Pipe Networks

1. Introduction
1.1 Parallel pipes connection

For many application of fluid flow, the pipes - as the fluid flowing medium, have to be connected in
either series or parallel connection.

When two or more pipes are linked in such a way that the flow from one pipe branches or divides into
two or more distinct pipes before reuniting into a single pipe, the pipes are said to be in parallel
[ CITATION Red \l 1033 ]. Along the branch pipes, the total discharge Q is divided into components Q 1
and Q2, resulting in –

Q=Q1 +Q2
(1)

1.2 Hardy Cross Method

For pipe networks with loops, Hardy cross method is generally employed to determine the flow rates
through each pipes. The Hardy Cross technique is an iterative approach for calculating flow in pipe
network systems with known inputs and outputs but unknown flow within the network. This method
iteratively solves for flows in a pipe network using continuity of flow and continuity of potential. In the
case of pipe flow, conservation of flow means that at each pipe junction, the flow in equals the flow out.
The total directional head loss along any loop in the system is zero, which is known as conservation of
potential [CITATION Wik212 \l 1033 ].

If f is the friction factor, L is the length of pipe, d is its diameter and g is acceleration due to gravity, the
relationship between head loss and flow rate is provided by Equation 2 (Darcy-Weisbach Equation).

h f =r Q2 (2)

32 fL
where , r=
g π 2d5

The Hardy cross method of balancing heads starts with an estimated assumption (Q 0) that ensures flow
continuity at each junction, then balances the flows until potential continuity is reached over each loop
in the pipe network. The iteration is continued until the change in flow calculated by Equation 3 is close
to zero.
 −∑ r Q20
∆ Q= (3)
∑ 2r Q0
2. Problem definition
The problem involves evaluating the pipe network depicted in Figure 1, which spans two loops (I and II),
as well as the flow rate of water entering and exiting from four corners of these loops (A, C, D, and F), all
of which are made of similar construction materials. The temperature and pressure are assumed to
remain constant throughout the pipe network, the entire pipe assembly is at the same level, and the
pipe flow may be called steady state. The task is to analyze the flow rate in each pipe in order to fulfill
the external flows entering and exiting the network using the pipeline data provided in Table 1.

Figure 1. Given pipe network

Table 1. Pipeline data

Line Length (ft.) Internal diameter (inches)

1 2000 18

2 900 24

3 2800 12

4 1100 24

5 2200 12

6 750 18

7 2600 24
3. Assumptions
Following assumptions were made during the evaluation of given pipe networks:

i. The friction factor for each pipe are assumed to be 0.05.

ii. The flow is assumed to be steady.
iii. Constant temperature and pressure are considered throughout the pipe network.
iv. All the pipes are considered to be in the same level.
v. The change in flow obtained from Equation 3 is just an assumption due to higher order terms
ignored in the Taylor expansion.

4. Calculation
We assume the friction factor to be 0.05. So, ‘r’ on each pipe can be calculated as shown below,

32 f L1 32 ×0.05 × 2000
r 1= 2 5
= =1.33
gπ d 18 5
1 2
32.174 × π ×( )
12

32 f L2 32 × 0.05× 900
r 2= 2 5
= =0.14
gπ d 24 5
2 2
32.174 × π ×( )
12

32 f L3 32 ×0.05 × 2800
r 3= 2 5
= =14.11
gπ d 12 5
3 2
32.174 × π ×( )
12

32 f L 4 32 ×0.05 ×1100
r4 = 2 5
= =0.17
gπ d 24 5
4 2
32.174 × π ×( )
12

32 f L5 32 ×0.05 × 2200
r 5= 2 5
= =11.08
gπ d 12 5
5 2
32.174 × π ×( )
12

32 f L6 32× 0.05 ×750

r6 = 2 5
= =0.5
gπ d 18 5
6 2
32.174 × π ×( )
12
 32 f L7 32× 0.05 ×2600
r7 = 2 5
= =0.41
gπ d 24 5
7 2
32.174 × π ×
12( )
Let the initial flow assumption (in ft3/s) around the pipe networks be,

Q AB=6 ,Q AF =8 , Q FE =3.198 ,Q BC =0.8 , Q BE =5.2, Q ED=8.398 , Q CD =5.85

Table 2. Iteration I

Loop I Loop II

Pipe Pipe r Q0 hf = 2rQ0 Pipe Pipe r Q0 hf = 2rQ0

No. rQ02 No. rQ02

1 AB 1.33 6 47.88 15.96 5 BC 11.08 0.8 7.09 17.73

2 BE 0.14 5.2 3.786 1.456 2 BE 0.14 5.2 -3.786 1.456

3 FE 14.11 3.198 -144.31 90.25 7 ED 0.41 8.398 -28.92 6.89

4 AF 0.17 8 -10.88 2.72 6 DC 0.5 5.85 -17.11 5.85

Total -103.52 110.39 Total -42.73 31.926

∆ Q=0.937 ∆ Q=1.338

Since ∆ Q is positive, we need to add the change in flow for clockwise fluid flow and subtract from the
anti-clockwise fluid flow. So,

Q AB=6+0.937=6.937 QBC =0.8+ 1.338=2.138

QBE =5.2+0.937=6.137 QBE=6.137−1.338=4.799

Q FE=3.198−0.937=2.261 Q ED=8.398−1.338=7.06

Q AF =8−0.937=7.063 Q DC =5.85−1.338=4.512

Table 3. Iteration II

Loop I Loop II

Pipe Pipe r Q0 hf = 2rQ0 Pipe Pipe r Q0 hf = 2rQ0

 No. rQ02 No. rQ02

1 AB 1.33 6.937 64 18.45 5 BC 11.08 2.138 50.64 47.38

2 BE 0.14 4.799 3.22 1.34 2 BE 0.14 4.799 -3.22 1.34

3 FE 14.11 2.261 -72.13 63.8 7 ED 0.41 7.06 -20.43 5.79

4 AF 0.17 7.063 -8.48 2.4 6 DC 0.5 4.512 -10.18 4.512

Total -13.39 85.99 Total 16.81 59.022

∆ Q=0.15 ∆ Q=−0.285

For positive∆ Q , we need to add the change in flow for clockwise fluid flow and subtract from the anti-
clockwise fluid flow and for negative ∆ Q , we need to do the opposite. So,

Q AB=6.937+0.15=7.08 7 Q BC =2.138−0.285=1.853

Q BE =4.799+ 0.15=4.949Q BE=4.949+ 0.285=5.234

Q FE=2.261−0.15=2.11 1 QED =7.06+0.285=7.345

Q AF =7.063−0.15=6.913 Q DC =4.512+ 0.285=4.797

Table 4. Iteration III

Loop I Loop II

Pipe Pipe r Q0 hf = 2rQ0 Pipe Pipe r Q0 hf = 2rQ0

No. rQ02 No. rQ02

1 AB 1.33 7.087 66.8 18.85 5 BC 11.08 1.853 38.04 41.06

2 BE 0.14 5.234 3.83 1.46 2 BE 0.14 5.234 -3.83 1.46

3 FE 14.11 2.111 -62.88 59.57 7 ED 0.41 7.345 -22.12 6.02

4 AF 0.17 6.913 -8.12 2.35 6 DC 0.5 4.797 -11.50 4.797

Total -0.37 82.23 Total 0.59 53.337

∆ Q=0.0045 ∆ Q=−0.01

The change in flow is nearly equal to zero in both the loop. Hence, we can stop the iteration after
making the changes using the flow just obtained.
 Q AB=7.087+ 0.0045=7.0915 Q BC =1.853−0. 01=1.84 3

QBE =5.234+0.0045=5.2385 QBE=5.2385+ 0.01=5.2485

Q FE=2.111−0.0045=2.1065Q ED =7.345+ 0.01=7.35 5

Q AF =6.913−0.0045=6.9085 QDC =4.797+0.01=4.80 7

5. Conclusion
The pipe networks as shown in Figure 1 was solved using Hardy cross method. This method is an
iterative method which tends to reduce the head loss in each loop to zero while maintaining the law of
continuity. The flow values in each pipe are obtained after the 3 rd iteration with ∆ Q value close to zero.

The flow through pipes AB, AF, FE, BE, ED, BC and DC are 7.0915 ft 3/s, 6.9085 ft3/s, 2.1065 ft3/s, 5.2485
ft3/s, 7.355 ft3/s, 1.843 ft3/s and 4.807 ft3/s respectively.

6. References
2021. Hardy Cross method. March 17. Accessed October 10, 2021.
https://en.wikipedia.org/w/index.php?title=Hardy_Cross_method&oldid=1012679355.

Reddy, R. n.d. Flow through Pipes in Series and Parallel.

https://www.engineeringenotes.com/fluids/pipes/flow-through-pipes-in-series-and-parallel-
difference-diameters-equations-and-solved-problems/47568#:~:text=Pipes%20are%20said
%20to%20be,series%20is%20the%20same%20throughout.