Journal of the Korean Society of Marine Engineering, Vol. 37, No. 7, pp.

784~791, 2013 (ISSN 1226-9549)

http://dx.doi.org/10.5916/jkosme.2013.37.7.784

Layout design optimization of pipe system in ship engine room for space
efficiency

Dong-Myung Lee1 · Soo-Young Kim2 · Byung-Young Moon3 · Gyung-Ju Kang†

(Received November 6, 2013； Revised November 14, 2013；Accepted November 21, 2013)

Abstract: Recent advanced IT made layout design fast and accurate by using algorithms. Layout design should
be determined by considering the position of equipment with satisfying various space constraints and its com-
ponent works with optimum performance. Especially, engine room layout design is performed with mother
ship data, theoretical optimal solution, design requirements and several design constraints in initial design
stage. Piping design is affected by position of equipment seriously. Piping design depends on experience of
designer. And also piping designer should consider correlation of equipment and efficiency of space. In this
study, space evaluation method has been used to evaluate efficiency of space. And also this study suggested
object function for optimal piping route, Average Reservation Index(ARI), Estimated Piping Productivity(EPP)
and with modified space evaluation method. In this study, optimum pipe routing system has been developed
to reflect automated piping route with space efficiency and experience of piping designer. Engine room is ap-
plied to the design of the piping in order to confirm validity of the developed system.
Keywords: Layout Design, Optimum Design, Engine Room, Average Reservation Index(ARI), Estimated
Piping Productivity(EPP)

1. Introduction after evaluation under designer’s knowledge and

Initial ship design of engine room is usually per- experience.

formed based on preceding design data such as refer- To perform the layout design more effectively, re-

ence ship data, theoretical optimum solutions, require- searches about room arrangement in ship have been car-

ment items and design constraints. And the designed ried out [1][2]. Hills et al. [3] studied effective design

data are used for the next design stage. Moreover, de- method for section layout or initial structural layout.

signers modify the layout under consideration of equip- Especially, Penteado et al. [4] conducted optimum lay-

ment performance analysis and the evaluation of equip- out of equipment of chemical process plant considering

ment operators. Finally, an optimum layout is selected safety. Many researchers have focused on the opti-
mization of room layout and plant equipment layout.
†Corresponding Author: School of Mechanical Engineering, Pusan National University, 2, Busandaehak-ro 63
beon-gil, Geumjeong-gu, Busan 609-735, Korea, E-mail: [email protected], Tel: 051-510-3442
1 Department of Naval Architecture & Ocean Engineering, Pusan National University, E-mail: [email protected],
Tel: 051-510-2754
2 Department of Naval Architecture & Ocean Engineering, Pusan National University, E-mail: [email protected],
Tel: 051-510-2346
3 Department of Naval Architecture Engineering, Kunsan National University, E-mail: [email protected], Tel:
063-469-1851

This paper is extended and updated from the short version that appeared in the Proceedings of the International sympo-
sium on Marine Engineering and Technology (ISMT 2013), held at BEXCO, Busan, Korea on October 23-25, 2013.
 Layout design optimization of pipe system in ship engine room for space efficiency

However, in company with equipment layout, pipe ing between unit equipment is considered from the ini-
layout is also important. Because pipe arrangement is tial stage of arrangement. Figure 1 illustrates algorithm
unconditionally affected from equipment layout, pipe for flowchart to evaluate the space efficiency accord-
path could not be determined before equipment ing to equipment arrangement.
arrangement. Also, even though equipment layout is
2.2 Fuzzy theory
confirmed, it is often modified due to high expenses of
Fuzzy theory is introduced by Prof. Zadeh at UC
pipe [5]. Thus, the optimization method of pipe path
Berkley in 1965 [6]. Fuzzy theory expresses un-
has difficulty in performing the design effectively in
certainty of phenomenon in itself. The theory expresses
short time using computer. If there is the design sys-
the real approximate and inexact state more effectively
tem satisfying the various design requirements, the ef-
than traditional logic system. Because the evaluation of
ficiency of pipe design could be increased. Pipe path
equipment arrangement contains many uncertain terms
design highly depends on the experiences of designer.
such large/small, far/near, often/sometimes, and emer-
In the pipe design, considering the correlation among
gency/normal, Fuzzy theory can be effectively applied
equipment and room efficiency is difficult. It is neces-
for the evaluation. Previous set theory based on boo-
sary to develop the optimum design system satisfying
lean logic explain that a specific object is belong to
the design requirements.
set A (1, true) or not to set A (0, false). On the other
Therefore, in this paper pipe design system is pro-
hand, from fuzzy set theory based on Fuzzy theory,
posed to optimize the pipe path for engine room
specific set A can be expressed by the number from
0(fully excluded) to 1 (perfectly included). For exam-
2. Theory for optimum pipe path design ple, for a set A of tall people, Crips set and Fuzzy set
are compared as shown in Figure 2.
2.1 Optimum arrangement evaluation methodology
Fuzzy set A attached to universal set U can be de-
Generally, machinery equipment is evaluated and
fined as follow:
produced under the established standards such as
standardization, departmentalization and ergonomics.
    ∈ (1)
The standards for evaluation of equipment arrangement
system consist of the following objection functions: the
where,   is membership function of x in A.
maximization of convenience for operations and main-
tenances; the minimization of equipment size; Estimate Membership function assigns each element of X num-
Pipe Productivity (EPP); the maximization of ber from 0 to 1.
Estimated Link Practicability (ELP); Reservation Index; The simplest membership functions are composed by
evaluation of cost function. Under the assigned algo- using lines. The simplest is triangular membership
rithm, initial layout design is calculated, and finally function which is composed of three points of triangle.
optimum layout is selected to minimize the cost Membership function of trapezoidal has flat shape at
function. For the design optimization of equipment ar- top, which is cut shape of truncated triangle curve.
range, easy, productivity, size of equipment, interfer- These line functions are easy for calculation. Besides,
ence, and locations of elements are considered. there are Gaussian membership function and general-
Designed equipment’s from experienced designers are ized bell membership function as shown in Figure 3.
compared and evaluated for optimum arrangement al- The evaluation items of descriptive language can be
ternatives under the considerations. Besides, pipe duct- included in the computer program, by expressing

Journal of the Korean Societ of Marine Engineering, Vol. 37, No. 7, 2013. 11 785
 Dong-Myung Lee · Soo-Young Kim · Byung-Young Moon · Gyung-Ju Kang

vague terms using Fuzzy function and organizing from

2.3 Shortest path search algorithm
Fuzzy rules. Moreover, from the Fuzzy inferences, de-
Shortest path means the shortest path among the
signers can quantify the evaluation standards as well
paths existing between two stationary points on graph.
as can evaluate arrangement alternatives fast and easy.
At direction graph the initial peak and final peak on
path are called source and destination, respectively.
Shortest path problem is divided according to the se-
lection of source and destination. Edsger Dijkstra [7]
proposed Dijkstra algorithm which search shortest path
from source to every destination. The shortest path
length obtained by conducting algorithm is stored into
array. If all the stationary points are included in the
subset of stationary points, the subset of shortest path
has shortest path. Disjkstra algorithm investigates the
shortest path by maintaining the set which is com-
posed of start peak and shortest path already known.
Firstly, only start peak is included in the set ex-
pressed array. Secondly, it searches the shortest path
Figure 1: Layout design for evaluation algorithm by selecting point which is shortest point from sta-
tionary point to among points which in not included
in set of stationary points, also by including only
shortest points into the set. The shortest points from
start stationary point to arbitrary stationary point are
stored into set using array. When all the points are in-
cluded in the set, the set have points having the short-
est points from start stationary point to all points
through these procedures [8].
Figure 2: A Comparison of Fuzzy Set and Crisp Set Moreover, Hart et al. [9] developed A* search algo-
rithm which combines heuristic method and formal
method, in order to estimate of the best route from
start point to destination node. A* algorithm guaran-
tees searching shortest path on graph and it performs
under minimum computation. From the research of
Hart et al. [9], admissibility and optimality properties
of A* algorithm are verified.
Therefore, because A* algorithm guarantees short
time and optimization, in this study A* algorithm is
used for the optimum system considering room effi-
ciency as well as pipe path optimization.

Figure 3: Membership Functions 3. Optimum pipe path search system

Journal of the Korean Societ of Marine Engineering, Vol. 37, No. 7, 2013. 11 786
 Layout design optimization of pipe system in ship engine room for space efficiency

Optimum pipe path system is developed using Fuzzy

theory and path search algorithm. Figure 4 illustrates
system organization for optimization of pipe path.

Figure 5: Fuzzy set for distance.

Figure 4: Flow chart of equipment arrangement evalua-

tion system

3.1 Fuzzy modeling for pipe path optimization Figure 6: Fuzzy set for Reservation Index
Eight Fuzzy rules are determined in order to esti-
mate reservation index. Based on the rules, Fuzzy in- posed and A* algorithm is used to calculate optimum
ference system contains input and output. Distance, in- pipe path. The applied A* algorithm guarantees the
put Fuzzy function as shown in Figure 5 consists of searching shortest path with minimum calculation using
linear Fuzzy function. Reservation index, output is heuristic information. Using heuristic information, the
composed of Fuzzy function as shown in Figure 6. modified cost accelerates searching to the direction of
Fuzzy rules for inference of Reservation Index are de- the objective node, instead of exploring all direction.
termined as shown in Table 1. Goodness of fit con- Consequently, it reduces time consuming for the search
cerning input can be obtained from the intersection of of small node and edge, and the exploring velocity
two propositions at sound area. Next, inference result increases.
of each rule can be found by applying these good nes- In this study, cost function for the generation of op-
ses of fit to consequent Fuzzy set, reservation index. timum pipe path can be expressed as follows:
Then, final inference result can be determined by accu-
mulating each rule. In fact, the final inference value Z               (1)
is obtained by using center of gravity method [10].
where, ARI denotes Average Reservation Index; X, Y,

3.2 Pipe path searching algorithm Z represent manhattan distance; EPP stands for
Estimated Pipe Productivity. Upon it, optimum pipe
In this study, pipe path searching system is pro

Journal of the Korean Societ of Marine Engineering, Vol. 37, No. 7, 2013. 11 787
 Dong-Myung Lee · Soo-Young Kim · Byung-Young Moon · Gyung-Ju Kang

paths are calculated using Reservation Index of equip-

ment as well as using heuristic information of distance
between nodes. In this study, the objective function for
optimum pipe path search is defined as cost function
and it can be formulate as follow:

  min             (3)

(a) (b)
3.3 Optimization application for pipe path
In this study, to validate the proposed optimization
pipe path system, the pipe path generation of engine
room is conducted. Table 2 demonstrates the lists of
equipment used for pipe path generation as well as
size and influence distances. Equipment shapes are dis-
played in Figure 7.
(c) (d)
Table 1: Fuzzy rules for inference of Reservation
Index
IF distance is FAR,
Rule 1:
THEN Reservation Index is LOW
IF distance is MIDDLE,
Rule 2:
THEN Reservation Index is MEDIUM
IF distance is NEAR,
Rule 3:
THEN Reservation Index is HIGH (e) (f)
Figure 7: Modeling for unit equipment: (a) pump;
(b) Generate; (c) Tank type 1; (d) Tank type 2; (e)
Table 2: List of Equipment and Influence Distance
2nddeckgenerator;(f)Engine
RMX RMY RMZ Influent
Equipment Distance
(mm) (mm) (mm) (mm) 4. Results and Discussion
Deck 1 21,000 33,000 500 1,000 Optimum pipe path generation of engine room is
Deck 2 21,000 33,000 500 1,000 conducted in order to validate the effectiveness of al-
Pump 1,000 750 2,600 100 gorithm proposed in this research. In order to define
Tank the space, cell size and number of cells are defined as
2,000 2,000 2,000 100
Type 1
shown in Table 3. Figure 8 illustrates the defined de-
Tank 2,000 2,000 2,000 100 sign space.
Type 2
Generator 1,000 1,000 2,000 100 Figure 9-10 show the design from the experiences
nd
2 Deck of designers. Figure 11 illustrates the reservation index.
3,100 2,100 2,000 1,500
Generator
Optimized designs using pipe design system are shown
Engine 5,900 3,650 2,900 2,000
in Figure 12-13. And corresponding reservation index

Journal of the Korean Societ of Marine Engineering, Vol. 37, No. 7, 2013. 11 788
 Layout design optimization of pipe system in ship engine room for space efficiency

Table 3: Cell size and number of cells of design is shown in Figure 14. Average reservation index ac-
space cording to each design method is compared in Table 4.
　 Cell Size (mm) Number of Cells The average reservation index before pipe generation
X 1,000 20 shows low value of 23.04. The average reservation in-
Y 1,500 20
dex of design based on designer’s experience and of
Z 1,000 9

Figure 11: Reservation Index of design

Figure 8: Definition of design space

Figure 12: Pipe Arrangement of Deck 1

Figure 9: Pipe Arrangement of Deck 1

Figure 13: Pipe Arrangement of Deck 2

Figure 10: Pipe Arrangement of Deck 2

Journal of the Korean Societ of Marine Engineering, Vol. 37, No. 7, 2013. 11 789
 Dong-Myung Lee · Soo-Young Kim · Byung-Young Moon · Gyung-Ju Kang

Table 4: Comparison of Average Reservation Index have effect on EPP, so that the pipe condition is the
for Design Alternatives same. However, from the ARI the efficiency of the
Average room is improved.
Design Alternatives Reservation Index
Design Space 23.04
5. Conclusions
Designer 51.41 In this paper, Fuzzy theory and A* search algorithm
Optimized was applied for optimization of pipe route of engine
51.36
Design Result
room in the design stage of ship. From the research
the following conclusions were made:
1) A* algorithm is used for the search of optimum
path of pipe for engine room, and the analysis results
showed short calculation time and optimization of
path.
2) The efficiency of A* algorithm was maximized
by joining the traditional evaluation factors such as
ARI and EPP.
3) The proposed methodology can be effectively
used for the pipe design in engine room.

Acknowledgments
This research was financially supported by the
Figure 14: Reservation Index of Optimized design
Ministry of Education, Science Technology (MEST)
and National Research Foundation of Korea (NRF)
through the Human Resource Training Project for
Regional Innovation

References
[1] J. P. Hope, “The process of naval ship general
arrangement design and analysis”, Naval
Engineers Journal, vol. 93, no. 4, pp. 29-38,
1981.
Figure 15: After applying the layout optimization
[2] A. Cort and W. Hills, “Space layout design us-
(Changed the Pipe)
ing computer assisted methods”, Naval Engineers
Journal, vol. 99, no. 3, pp. 249-260, 1987.
optimized design are 51.41 and 51.36, respectively.
[3] W. Hills and M. Wels, “An efficient compart-
Figure 15 illustrates the differences of design based on
mentation method for use in preliminary ship de-
designer’s experience and optimized design.
sign”, Proceedings of ICCAS'89, pp. 141-153,
From the Figure 15, it is observed that the total
1989.
length of the pipe and number of bent part are the
[4] F. D. Penteado and A. R. Ciric, “An MINLP
same before and after the optimization. Therefore, it is
approach for safe process plant layout”, Industrial
clear that the design from the proposed system not

Journal of the Korean Societ of Marine Engineering, Vol. 37, No. 7, 2013. 11 790
 Layout design optimization of pipe system in ship engine room for space efficiency

& Engineering Chemistry Research, vol. 35, no.

4, pp. 1354-1361, 1996.
[5] J.-K. Park, Development of Evaluation System
for Machinery Arrangement in Engine Room,
M.S. Thesis, Department of Naval Architecture,
Pusan National University, Korea, 2002 (in
Korean).
[6] L. A. Zadeh, “Fuzzy sets”, Information and
Control, vol. 8, no. 3, pp. 338-353, 1965.
[7] E. W. Dijkstra, “A note on two problems in
connection with graphs”, Numeriche Mathematik,
vol. 1, no. 1, pp. 269-271, 1959.
[8] W.-B. Ku, Data Structure Realized Using C
Language, Seoul, Korea: Insolmedia Publisher,
2001 (in Korean).
[9] P. E. Hart, N. J. Nilsson, and B. Raphael, “A
formal basis for the heuristic determination of
minimum cost paths in graphs”, Systems Science
and Cybernetics, IEEE Transactions on , vol. 4,
no. 2, pp. 100-107, 1968.
[10] S.-K. Oh, Fuzzy Model and Control System by
C Programming, Seoul, Korea: Naeha Publisher,
2002 (in Korean).

Journal of the Korean Societ of Marine Engineering, Vol. 37, No. 7, 2013. 11 791