Applied Thermal Engineering 40 (2012) 397e408

Contents lists available at SciVerse ScienceDirect

Applied Thermal Engineering

journal homepage: www.elsevier.com/locate/apthermeng

A numerical technique for Total Site sensitivity analysis

Peng Yen Liew a, Sharifah Radah Wan Alwi a, *, Petar Sabev Varbanov b, Zainuddin Abdul Manan a,
Jir Jaromr Klemes b
a

Process Systems Engineering Centre (PROSPECT), Faculty of Chemical Engineering, Universiti Teknologi Malaysia, 81310 UTM Johor Bahru, Johor, Malaysia
Centre for Process Integration and Intensication e CPI2, Research Institute of Chemical and Process Engineering, Faculty of Information Technology, University of Pannonia,
Egyetem u. 10, H-8200 Veszprm, Hungary
b

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 19 October 2011
Accepted 11 February 2012
Available online 22 February 2012

Total Site Heat Integration (TSHI) is a methodology for the integration of heat recovery among multiple
processes and/or plants interconnected by common utilities on a site. Until now, it has not been used to
analyze a sites overall sensitivity to plant maintenance shutdown and production changes. This feature
is vital for allowing engineers to assess the sensitivity of a whole site with respect to operational changes,
to determine the optimum utility generation system size, to assess the need for backup piping, to
estimate the amount of external utilities that must be bought and stored, and to assess the impact of
sensitivity changes on a cogeneration system. This study presents four new contributions: (1) Total Site
Sensitivity Table (TSST), a tool for exploring the effects of plant shutdown or production changes on heat
distribution and utility generation systems over a Total Site; (2) a new numerical tool for TSHI, the Total
Site Problem Table Algorithm (TS-PTA), which extends the well-established Problem Table Algorithm
(PTA) to Total Site analysis; (3) a simple new method for calculating multiple utility levels in both the PTA
and TS-PTA; and (4) the Total Site Utility Distribution (TSUD) table, which can be used to design a Total
Site utility distribution network. These key contributions are clearly highlighted via the application of the
numerical technique to two Case studies.
2012 Elsevier Ltd. All rights reserved.

Keywords:
Total site problem table algorithm (TS-PTA)
Total Site
Heat cascade
Numerical approach
Site minimum utility targets
Process integration

1. Introduction
Pinch Analysis is an established technology for reducing energy
consumption that has been widely applied in various industries for
more than 30 years. Dhole and Linnhoff [1], Raissi [2] and Klemes
et al. [3] extended traditional heat integration, which focuses on
direct heat transfer among process streams at a single site, to heat
integration for multiple sites. This is known as Total Site Heat
Integration (TSHI), sometimes called site-wide integration. Direct
heat transfer is not always suitable for inter-process heat recovery
due to the required high degree of operational exibility and the
long-distance piping needed, which makes it very costly [4]. TSHI
using indirect heat transfer utilising existing utility systems is
typically more cost effective because the existing plant piping
system can be used. TSHI heat integration is linked by a common
central or sectional utility system.
Dhole and Linnhoff [1] have introduced Site Sink and Source
Proles (SSSP), a graphical tool that can be used to evaluate fuel
consumption, cogeneration, emissions and cooling needs for
an integrated site. A simple exergy model was proposed for
* Corresponding author. Tel.: 60 07 5535533; fax: 60 07 5581463.
E-mail address: shasha@cheme.utm.my (S.R. Wan Alwi).
1359-4311/$ e see front matter 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.applthermaleng.2012.02.026

cogeneration capacity estimation based on SSSP, and the model was

further extended by Raissi [2] and Klemes et al. [3]. Based on SSSP,
Klemes et al. [3] developed the Total Site Prole (TSP) and the Site
Utility Grand Composite Curve, which can be used to evaluate Total
Site potential heat recovery. Subsequently, Marchal and Kalitventzeff [5] introduced a mathematical programming tool for
minimising Total Site energy costs. Their work also included an
integration of combined heat and power production using a steam
network. Matzuda et al. [6] have successfully studied the heat
recovery potential for a large steel plant using TSP analysis.
An advanced approach to these concepts, known as top-level
analysis, is one that allows for scoping, i.e., selecting site
processes to target for heat integration improvements [7]. The
utility system is rst optimised for the current steam and power
demands. This is followed by an assessment of the potential benet
of reducing steam demands at various levels by successively optimising the system in steps of steam demand reduction. This results
in a set of curves for steam marginal prices for the system under
consideration.
Perry et al. [8] extended the Total Site concept to a broader
spectrum of processes in addition to the industrial process.
A potential for the integration of renewable energy sources was
introduced to reduce the carbon footprint of a Locally Integrated

398

P.Y. Liew et al. / Applied Thermal Engineering 40 (2012) 397e408

Energy Sector (LIES). In a LIES, heat sources and sinks can be derived
from small-scale industrial plants, large building complexes (such as
hotels and hospitals), ofces and residential areas.
One of the major challenges in implementing the Total Site
concept involving renewable energy is the variation of energy
supply and demand with time and location. Therefore, Varbanov
and Klemes [9] suggested performing Total Site targeting in a set of
time slices to maximise heat recovery within each time slice. Varbanov and Klemes [10] then further extended the concept to heat
storage, heat waste minimisation and carbon footprint reduction. A
Total Site heat cascade is also introduced in this work to illustrate
these concepts.
Bandyopadhyay et al. [11] proposed a modication of the Site
Grand Composite Curve (SGCC) that incorporates assisted heat
transfer. This type of heat transfer takes into account the nonmonotonic parts or pockets of the process GCC, which were not
considered by Dhole and Linnhoff [1], although this may not be
practical for many integrated sites. The results of their study show
that the modied SGCC tends to increase heat recovery potential,
particularly those within each process, a feature which is not
considered in the TSP. However, the economy of this design has to
be explored, as an increased integration using nonstandard steam
mains can be costly.
Kapil et al. [12] proposed the recovery and upgrading of lowgrade heat from processes. The work has proposed a new methodology for estimating the cogeneration potential for a site utility
system via bottom-up and top-down procedures. Ghannadzadeh
et al. [13] presented Iterative Bottom-to-Top Model (IBTM) as a new
shaftwork targeting model to estimate the cogeneration potential
for site utility systems prior to the detailed design.
Fodor et al. [14] further developed a TSHI targeting method to
allow for a variation of the minimum temperature difference
(DTmin) among Total Site processes. Previous works by Dhole and
Linnhoff [1] and Klemes et al. [3] assumed a uniform DTmin on
a Total Site. Fodor et al. [14] and Varbanov et al. [15] proposed the
use of a utility and process-specic DTmin between utility and
process streams, which is more realistic in practical applications.
The Total Site methodology and the concepts developed by
Dhole and Linnhoff [1] and used in recent studies are based on
a graphical method, with the typical advantages and disadvantages
of such approaches. Numerical methodologies that provide similar
benets such as the Problem Table Algorithm (PTA) for heat pinch
and the Water Cascade Analysis (WCA) for water pinch are therefore desirable.
The PTA is a numerical tool for intra-process heat integration
proposed by Linnhoff and Flower [16]. This tool is the equivalent to
the use of Composite Curves (CCs) and Grand Composite Curve
(GCCs) in the graphical method and supports a more precise
graphical construction by providing exact values for the crucial
points. The algorithm was extended to multiple utility targeting by
Costa and Queiroz [17]. The PTA was also recently extended to the
Unied Targeting Algorithm (UTA) by Shenoy [18]. The UTA is
a powerful tool for obtaining the maximum resource recovery for
Process Integration problems including heat and mass exchange,
water, hydrogen, carbon emissions and material reuse networks.
However, the method proposed by Costa and Queiroz [17] involves
rather complex calculations, whereas the UTA cannot be used for
TSHI problems. To make the PTA a more powerful tool, simpler
method for multiple utility targeting would be benecial. Additionally, the PTA can also be extended to TSHI.
In the current study, a new numerical tool for targeting TSHI is
proposed, known as the Total Site Problem Table Algorithm
(TS-PTA). This numerical tool is an alternative to the graphical TSHI
approach and is suitable for both the uniform and non-uniform
DTmin methods proposed by Dhole and Linnhoff [1], Klemes et al.

[3], Fodor et al. [14] and Varbanov et al. [15]. Although graphical
approaches are advantageous in terms of providing valuable visual
insights, they are difcult to construct, especially for large problems, and may yield some inaccuracies inherent in the graphical
nature of the method. The Problem Table Algorithm (PTA), which is
a numerical tool introduced by Linnhoff and Flower [16] as an
alternative to the Composite Curves, has been among the preferred
analytical tools used to compensate for the limitations of the
graphical approaches. In this work, the PTA method is extended to
include TSHI analysis.
The previous works cited have generally not deeply studied the
exibility of integrated plants. A numerical tool therefore offers
a good opportunity to evaluate the sensitivity of each plant in TS
integration. The Total Site Sensitivity Table (TSST) can be used as
a tool to explore site-wide sensitivities to various operational
changes and variations. A typical case is when one site process must
be closed down for regular maintenance or due to an accident.
Using the TSST, the effect of a plant shutdown can be assessed, and
suitable measures can be taken during the design and operational
stages to ensure other site utility supplies are not disrupted.
2. Methodology
A summary of the procedure involving the four methodologies
is described in the following.
2.1. Tool 1: Total Site Problem Table Algorithm (MU-PTA)
The initial steps follow the same procedure as the PTA for
individual process. First, the shifted temperatures for the process
streams in each individual process are calculated as described in
Smith [19], Kemp [20] and Klemes et al. [21]. The PTA is constructed
as described by Linnhoff and Flower [16] and Smith [19]. The
multiple utility cascade procedure for each individual plant is as
follows:
a. Above the pinch region:
i. Subtract half of the minimum temperature difference
within each process, DTmin,pp/2, from the shifted temperature to return it to a normal temperature, and then add the
minimum temperature difference between the utility and
process stream (DTmin,up) [14,15].
ii. Cascade the heat available in each temperature interval
from the highest temperature to the pinch temperature.
When a negative value results, an external heat enthalpy
representing the utility is added immediately to the
temperature interval during cascading.
iii. The amount of each utility type required can be determined
by summing the external heat enthalpies from below each
utility temperature to the next utility temperature.
b. Below the pinch region:
i. DTmin,pp/2 is added, and DTmin,up is subtracted, to the shifted temperatures [14,15].
ii. The heat available in each temperature interval is cascaded
from the lowest temperature to the pinch temperature, and
the external cooling utility required is immediately added
to the temperature interval when there is positive value in
the cascade.
iii. The amount of each utility type to be generated is obtained
by adding the external cooling utility above each utility
temperature but below the next-highest temperature
utility.
This tool also could be used for single process Heat Integration
which has different temperature shifting at the beginning.

P.Y. Liew et al. / Applied Thermal Engineering 40 (2012) 397e408

Specically, at above pinch, DTmin,pp/2 is added to the shifted

temperature in step a(i). At below pinch point, DTmin,pp/2 is subtracted in step b(i).
2.2. Tool 2: Total Site Problem Table Algorithm (TS-PTA)
The TS-PTA is a continuation of the PTA table with an extension
of four columns. This table represents an algebraic version of the
Site Composite Curve (SCC) in a graphical TSHI analysis. The
procedure is described as follows:
a. The heat sinks above the pinch region in each process are
added as net heat sinks according to utility type. Similarly, the
heat sources below the pinch region are added to become net
heat sources according to utility level.
b. The net heating requirement at each utility level is formulated
by deducting the net heat source from the net heat sink.
c. The net heating requirements are then cascaded from the top
to the bottom.
d. Analogous to the PTA, the most negative value of the previous
cascade is then used to initiate a new cascade, after rst
changing it to a positive value.
e. Similar steps to construct a PTA involving multiple utilities are
performed as follows:
i. Above the Total Site Pinch, the net heat requirement is
cascaded from the top to the bottom. An external heating
utility is added into the system when there is a negative
value to balance the heat decit at different utility levels.
ii. Below the Total Site Pinch, the neat heat requirement is
cascaded from the lowest temperature cooling utility to the
pinch. A negative external cooling utility is added when
a positive value occurs in the cascade.

2.3. Tool 3: Total Site Utility Distribution (TSUD) table

To visualise the site distribution network, a TSUD table can be
constructed as follows:
a. The table lists the heat sources and sinks of each site according
to utility. The external heat requirement calculated in the TSPTA is also recorded.
b. Arrows are used to indicate possible utility exchanges from one
site to another or from a utility plant to a site.

2.4. Tool 4: Total Site Sensitivity Table (TSST)

The Total Site Sensitivity Table (TSST) is a practical tool for
analysing the effects of variations in Total Site operating conditions
on heat distribution and utility generation. The TSST is constructed
as below:
a. The TS-PTA is used to determine the utilities necessary for
different operating conditions, e.g., when one of the plants is

Table 1
Stream data for Plant A of Case study 1 with DTmin 20  C; modied example from
Canmet ENERGY [22].

399

Table 2
Stream data for Plant B of Case study 1 with DTmin 10  C; modied example from
Kemp [20].
Stream

Ts ( C)

Tt ( C)

DH(MW)

mCp (kW/ C)

Ts ( C)

Tt ( C)

B1
B2
B3
B4
B5

200
240
200
30
50

50
100
119
200
250

0.450
0.210
1.863
0.680
0.400

3.0
1.5
23.0
4.0
2.0

195
235
195
35
55

45
95
114
205
255

Hot
Hot
Hot
Cold
Cold

shutdown. The ndings are recorded in the table based on the

different types of utilities.
b. Variations of normal operation with various operating conditions are calculated by subtracting the utility requirements in
normal operations from the utility requirements under
different operating conditions according to utility type.
A more detailed explanation of all the tools using different steps
is described below.
3. Demonstration Case study
The four tools are used to demonstrate their application to Total
Site sensitivity analysis.
3.1. Step 1: construct the Problem Table Algorithm (PTA) to
determine QHmin, QCmin and the pinch temperature for each
individual plant
The temperature of cold streams (Tc) and the temperature of hot
streams (Th) in an individual plant are converted to shifted cold
stream temperatures (Tc) and shifted hot stream temperatures
(Th). Tc is shifted by adding half of the minimum temperature
difference between processes, DTmin,pp, whereas Th is shifted by
subtracting half of DTmin,pp. Assuming a DTmin,pp of 20  C for plant A
and a DTmin,pp of 10  C for Plant B, Tables 1 and 2 show the shifted
temperatures of all streams in Plants A and B of Case Study
1. Table 3 shows the utility temperature levels available at the
plants, which are used in the next step. The minimum utility/
process temperature difference, DTmin,up, is 10  C.
PTA are performed for both Plant A and Plant B. The completed
PTAs for Plants A and B are shown in Tables 4a and 4b, respectively.
As shown in Table 4, plant A requires 2250 kW of hot utility and
400 kW of cold utility with a shifted pinch temperature of 60  C.
Plant B requires 100 kW of hot utility and 1543 kW of cold utility
with a shifted pinch temperature of 195  C. Figs. 1 and 2 are the
GCCs for plants A and B. The results from these GCCs are similar to
the results obtained from the PTA given in Tables 4a and 4b.
3.2. Step 2: construct a Multiple Utility Problem Table Algorithm
(MU-PTA) for each individual plant to obtain targets for multiple
utility levels as heat sources and sinks for TSHI
MU-PTA are constructed to target the amounts of various utility
levels selected as potential sinks and sources for use in Total Site

Table 3
Site utility data for Case study 1.

Stream

Ts ( C)

Tt ( C)

DH(MW)

mCp (kW/ C)

Ts ( C)

Tt ( C)

Utility

Temperature ( C)

A1
A2
A3
A4

200
150
50
50

100
60
120
220

2.00
3.60
4.90
2.55

20
40
70
15

190
140
60
60

90
50
130
230

High-pressure steam (HPS)

Medium pressure steam (MPS)
Low-pressure steam (LPS)
Cooling water (CW)

270
179.93
133.59
15e20

Hot
Hot
Cold
Cold

400

P.Y. Liew et al. / Applied Thermal Engineering 40 (2012) 397e408

Table 4a
Single utility cascade table for Plant A of Case study 1.

mCp (kW/ C)
T ( C)

( C)

20

40

70

15

(kW/ C)

(kW)

230
40

-15

45

-25

-45

50

integration. The multiple utility cascade methodology is an extension of the PTA with an additional 4 columns. The multiple utility
cascade calculations are similar to GCCs and can be used identify
pockets and target the exact amounts of utilities needed within
a given utility temperature interval. Note that multiple utility
cascades must be performed based on the pinch regions for each
plant that were determined in Step 1.
3.2.1. Multiple utility cascades in the region above the pinch of each
individual plant
All shifted temperatures (T0 ) in the region above the pinch
(column 1, Table 5) from Table 4 PTA are reduced by DTmin,pp/2 to
return them to normal temperatures and then the minimum
temperature difference between the utility and the process
(DTmin,up) is added, as shown in column 2, Table 5; the resulting
temperature is labeled T00. The utility temperatures listed in Table 3
were also added into Table 5 to make it easier to determine the
utility distribution at a later stage.
Heat is again cascaded starting from the highest temperature
segment to the pinch temperature, as shown in column 7, Tables 5a
and 5b. Note that there are no changes in the calculations of sum
mCp and sum DH for each temperature level. This cascade is
known as a multiple utility heat cascade; it differs from the
previous heat cascade in the PTA (column 6, Tables 4a and 4b)
because it is performed interval-by-interval. If a negative value is
encountered while cascading one of the temperatures, external
utilities are immediately added at that point (the amount of
external utility added is listed in column 8) equal to the negative
value. The cascade then becomes zero at that temperature, e.g., at

40

-600

1650

-350

1900

100

2350

-900

1350

-2250

-1850

400

-1350

60
10

2250

-1000

90
30

450

130
40

Initial
Heat
Cascade

250

140
10

7
Single
Utility
Heat
Cascade

-600

190
50

400

a shifted temperature of 190  C, the cascade initially gives a value

of 600 kW at column 7. Therefore, 600 kW of external utility is
added at this interval, as listed at column 8. The cascade now
becomes zero here, as shown in column 7, Table 5. The cascade is
then continued, and the procedure is repeated.
Once the multiple utility heat cascades are completed, the
amounts of each type of utility consumed in the process are
obtained by adding the utility consumed below the utility
temperature (from column 8, Table 5) to before the next utility
temperature. For example, Table 5a shows that 600 kW of highpressure steam (HPS) at a temperature of 270  C is consumed in
plant A between 270  C and 179.93  C. Thus, 1650 kW of lowpressure steam (LPS) is used between 133.59 and 60  C for plant
A. The same procedure is repeated for plant B to yield a requirement of 100 kW of high-pressure steam.
3.2.2. Multiple utility cascades for the region below the pinch of
each individual plant
A similar methodology is used for multiple utility cascading
below the pinch temperature. All temperatures available below the
pinch are shifted by adding DTmin,pp/2 and then subtracting the
minimum temperature difference between the utility and process,
DTmin,up (see the region below the pinch in column 2, Table 5) to
obtain the temperatures in the utility temperature scale. Utility
temperatures are then added to the temperature list, as in column
2, Table 5.
However, multiple utilities are instead cascaded starting from
the bottom temperature to the pinch temperature, and any positive
heat value encountered while cascading must be zeroed out by

P.Y. Liew et al. / Applied Thermal Engineering 40 (2012) 397e408

401

Table 4b
Single utility cascade table for Plant B of Case study 1.

(kW)

Initial
Cascade

Single
Utility
Cascade

100

-40

60

-55

45

-100

1641.5

1741.5

1613

1713

1493

1593

1483

1583

1443

1543

mCp (kW/ C)
T' ( C)

( C)

1.5

23

(kW/ C)

255
20

-2

-40

235
30

-0.5

-15

205
10

-4.5

-45

195
81

21.5

1741.5

114
19

-1.5

-28.5

95
40

-3

-120

55
10

-1

-10

45
10
35

Fig. 1. Grand Composite Curve for Plant A of Case study 1 [15].

-4

-40

Fig. 2. Grand Composite Curve for Plant B of Case study 1 [15].

402

P.Y. Liew et al. / Applied Thermal Engineering 40 (2012) 397e408

Table 5a
PTA with multiple utility heat cascades for Plant A of Case study 1.

T
( C)

T
( C)

mCp (kW/ C)
( C)

20

40

70

15

(kW/ C)

(kW)

270
0

-15

190

Heat Sink/
Source
600

MPS

LPS

1650

600

50.35
50.35

199.65

140

250
6.41

45

288.45

133.59

538.45
3.59

45

161.55

130

700
40

-25

-1000

90

300
0

30

-45

-1350

1350

60

60

60

Pinch

10
50

HPS

39.93

90

Utility
Consumed/
Generated
(kW)

-600

179.93

130

Multiple
Utility
Heat
Cascade

10.07

140

230
40

190

0
40

230

40

400

50

-400
0

35
15

generating utilities (see the lower part of column 7 and 8 in Table 5).
For the region below the pinch, the negative values encountered
during multiple utility cascading represent pockets in the GCC.
The amount of utility that can be generated can be determined
by adding the amounts of excess heat from above the utility
temperature to the next utility temperature level. For example,
plant A can generate 400 kW of CW using process heat between 50
and 10  C. For plant B, 216.50 kW of medium pressure steam (MPS)
at 190 to 179.93  C and 996.31 kW of LPS between 179.93 and
133.59  C can be generated, whereas 330.19 kW of CW is consumed.
The proposed method differs from the one developed by Costa
and Queiroz [17]. The method in this study was developed through
a detailed observation of multiple utility targeting in the GCC. In
addition, the method proposed herein is a direct continuation of
the PTA, in which the multiple utility cascade actually uses most of
the information from the PTA. The method proposed by Costa and
Queiroz [17] includes an interpolation step for nding the upper
and lower temperature boundaries after utility targeting. However,

0
0

CW

400

the proposed methodology targets utilities according to temperature intervals, with the utility temperatures becoming temperature
boundaries, to distinguish the amounts of each utility type. The
calculations involved in this proposed method are also simpler than
those of the previously proposed method.
3.3. Step 3: construct the Total Site Problem Table Algorithm
(TS-PTA) to determine the amounts of utilities that can be
exchanged among processes
This part is an extension of the PTA to represent the Site CC in
TSHI. The utilities available from each plant are arranged from
highest to lowest temperature. The utilities generated below the
pinch temperature for all sites, as determined in Step 3, are added
together to represent the net heat source (see column 3, Table 6).
The utilities consumed above the pinch temperature for all sites, as
determined in Step 2, are added together to represent the net heat
sink (see column 4, Table 6). Fig. 3 shows the TSP and the Site

P.Y. Liew et al. / Applied Thermal Engineering 40 (2012) 397e408

403

Table 5b
PTA with multiple utility heat cascade for Plant B of Case study 1.

T
( C)

Multiple
Utility
Utility Consumed/
Heat
Generated
Cascade
(kW)

mCp (kW/ C)

T
( C)

( C)

1.5

23

(kW/ C)

(kW)

270

0
10

-2

-40

40

240

0
30

-0.5

-15

15

210

205

100

0
20

235

HPS

260

255

Heat Sink/
Source

0
10

-4.5

-45

45

200

190

Pinch

195

10.07

21.5 216.51

-216.50

179.93

0
46.34

21.5 996.31
0

24.59

21.5 528.69

-1.5

-28.5

330.19

-170
40

-3

-120

50

-50
10

-1

-10

40

-40
10

-4

-40

30

35

CW

90

45

996.31

-198.5
19

55

LPS
-330.19

109

95

216.50

-996.31

133.59

114

MPS

0
15

15

Table 6
Total site Problem Table algorithm (TS-PTA) for Case study 1.
1

Utility

Utility Temp. ( C)

Net heat
source (kW)

Net heat
sink (kW)

Net heat
requirement (kW)

Initial heat
cascade

Final single heat

cascade

Multiple utility
heat cascade

External utility
requirement (kW)

HPS

270

MPS
LPS

179.93
133.59

0
216.50
996.31

700
0
1650

1137.19

700

437.19

483.50

653.69

216.50

700

700

216.50

653.69

437.19
1137.19

CW

15e20

730.19

0 (Pinch)

730.19

L730.19

730.19
407

404

P.Y. Liew et al. / Applied Thermal Engineering 40 (2012) 397e408

Fig. 3. TSP and SCC for Case study 1 [15].

Composite Curve for Case study 1, as proposed by Varbanov et al.

[15]. The net heat sink and the net heat source from Fig. 3 are the
same as in the TS-PTA (columns 3 and 4, Table 6). The net heat sink
is subtracted from the net heat source to obtain the net heat
requirement (column 5, Table 6). The locations with negative
amounts of net heat indicate heat decits, whereas the locations
with positive values indicate heat surpluses. The Second Law of
Thermodynamics species that heat can only be transferred from
a higher temperature to a lower temperature. Therefore, the heat
surplus at higher temperature utilities can be transferred to utilities
with lower temperatures that have heat decits. For example, the
217 kW of MPS in Case study 1 can be transferred to LPS, which has
heat decit of 654 kW, instead of disposing of this excess heat with
an external cooling utility. As a result, the net heat requirement is
cascaded from top to bottom, starting with an initial value of zero.
The most negative value in the initial heat cascade (column 6,
Table 6) is then used to determine the amount of external heating

utility needed for the system by making it positive and cascading

column 5 again (see column 7, Table 6). This gives a value of
1137.19 kW of external heating needed. The value at the bottom of
the cascade represents the total cooling utility needed by the
system, which is 730.19 kW. The location at which the value
becomes zero is the Total Site Pinch Point, which is between the LPS
and CW temperatures.
Similarly to Step 2, the utilities in Table 6 can be separated into
two parts, i.e., the regions above and below the Total Site Pinch
region. Multiple utility cascades above the Total Site Pinch point use
the same method as in Step 2(a) (see column 8 and 9, Table 6). The
net heat requirement (column 5, Table 6) is cascaded (column 8 and
9) from the top to the pinch point by assuming that there is no heat
supplied at a temperature above the HPS. The same amount of
external heating utility is added when there is a negative value in
the cascade, e.g., 700 kW of HPS and 437.19 kW of LPS are needed in
Case study 1 as heating utilities. Step 2(b) is similar for the region
below the Total Site Pinch, as shown below the pinch in columns 8
and 9 of Table 6. Multiple utilities are cascaded (columns 8 and 9)
from the bottom to the pinch point, and cooling utility is added
when there is a positive value in the cascade until it reaches zero.
Note that cooling utilities below the Total Site Pinch are represented by negative numbers. For Case study 1, 730.19 kW of
external cooling water (CW) is required to dispose of the excess
heat.
The effect of multiple utility cascading above the Total Site Pinch
in Table 6 is clearly evident in Fig. 3. The heat sources at MP and LP
temperatures are provided to the heat sink at LP. A heating
requirement is necessary for LP instead of MP, which is less
economical. Fig. 3 clearly shows that the heat requirement of
437.19 kW also can be fullled by using Hot Water (HW) at a range
between 50 and 60  C.

3.4. Step 4: construct a Total Site Utility Distribution (TSUD) table

to visualise the utility ow in the sites
The SCC does not adequately display the utility distribution
when there are several processes involved on the integrated site.
The amounts of utility distribution for each site from on-site utility
systems can be visualised using the TSUD table (Table 7). All the

Table 7
Total site Utility Distribution (TSUD) table for Case study 1.

P.Y. Liew et al. / Applied Thermal Engineering 40 (2012) 397e408

Table 8
Stream data for Plant A [8] with DTmin,pp 20  C.
Stream

Ts ( C)

Tt ( C)

DH (kW)

mCp (kW/ C)

Ts ( C)

Tt ( C)

A1
A2
A3
A4
A5

170
150
25
70
30

80
55
100
100
65

5000
6477
1500
1050
5250

55.5556
68.1818
20.0000
35.0000
150.0000

160
140
35
80
40

70
45
110
110
75

Hot
Hot
Cold
Cold
Cold

405

heat sources and heat sinks in the various plants are listed separately according to utility type, as shown in columns 3 and 4. The
external utilities calculated from Step 4 are also listed in Table 7.
Arrows within the table show that heat sources can be transferred
to heat sinks for the same type of utility. If there are extra heat
sources, heat can be transferred to the lower utility levels.
4. Application of the TS-PTA to TS sensitivity with changes
and variations

Table 9
Stream data for Plant B [8] with DTmin,pp 10  C.
Stream

Ts ( C)

Tt ( C)

DH (kW)

mCp (kW/ C)

Ts ( C)

Tt ( C)

B1
B2
B3
B4
B5
B6

200
20
100
150
60
75

80
100
120
40
110
150

10,000
4000
10,000
8443
1000
7000

83.3333
50.0000
500.0000
76.7575
20.0000
93.3333

195
25
105
145
65
80

75
105
125
35
115
155

Hot
Cold
Cold
Hot
Cold
Cold

Table 10
Stream data for Plant C [8] with DTmin,pp 20  C.
Stream

Ts ( C)

Tt ( C)

DH (kW)

mCp (kW/ C)

Ts ( C)

Tt ( C)

C1 Hot
C2 Hot
C3 Cold
C4 Cold
C5 Cold
C6 Cold
C7 Cold
C8 Cold
C9 Cold
C10 Cold
C11 Cold

85
80
25
55
33
25
30
25
30
18
21

40
40
55
85
60
60
121
28
100
25
121

23.85
96.40
17.70
77.40
6.48
77.00
12.74
151.68
59.50
100.80
5.00

0.5300
2.4100
0.5900
2.5800
0.2400
2.2000
0.1400
50.5600
0.8500
14.4000
0.0500

75
70
35
65
43
35
40
35
40
28
31

30
30
65
95
70
70
131
38
110
35
131

Table 11
Stream data for Plant D [8] with DTmin,pp 10  C.
Stream

Ts ( C)

Tt ( C)

DH (kW)

mCp (kW/ C)

Ts ( C)

Tt ( C)

D1 Cold
D2 Cold

15
15

60
80

6000
5000

133.3333
76.9232

20
20

65
85

Table 12
Site utility temperatures.
Utility

Temperature ( C)

High-pressure steam (HPS)

Steam (ST)
Hot water (HW)
Cooling water (CW)

170
125
50e60
20

As mentioned previously, the TS-PTA can be benecial for analysing the sensitivity of the TSHI to plant shutdowns due to maintenance or upsets and to design mitigation strategies. This is
illustrated using Case study 2 from Perry et al. [8]. Here, there are
four sites considered in Locally Integrated Energy Sectors (LIES):
two industrial process plants, a hospital complex and a combined
residential and ofce complex. The stream data for the four plants
are listed in Table 8e11 . Plants A and C are assumed to have the
same DTmin,pp of 20  C, whereas Plants B and D both have a DTmin,pp
of 10C. Table 12 shows the types of utilities serving the area, with
a DTmin,up of 10  C.
Steps 1 to 4 were performed for the processes in Case study 2.
The nal TS-PTA values for the standard operation of the plants
comprising the TS are listed in Table 13. Due to its numerical nature,
it is very convenient to manipulate data in the TS-PTA to obtain new
values for various cases. For example, to consider a plant shutdown,
we omit the contributions from the shutdown plant from the heat
sinks and sources in columns 3 and 4 of Table 13. The new external
utility requirements are then obtained. Table 14 summarises the
external utility variations when one of the plants is shutdown. We
refer to Table 14 as the proposed Total Site Sensitivity Table (TSST),
which can be used to gain many insights into utility system design.
The variance in Table 14 is calculated by subtracting the amounts of
external utilities during plant shutdowns from the values needed
during normal operation. A positive variance above the Total Site
Pinch indicates that the central utility has a heat surplus that is not
used in any sinks. The utility systems have the following options:
(i) Fewer utilities can be generated, if permitted by the turn down
ratio.
(ii) The heat surplus can be disposed of using an external cooling
utility, which would incur a penalty cost.
(iii) The heat surplus can be sold to other plants.
(iv) For HP or MP steam, if a plant has a combined heat and power
system (CHP) with a double-stage extraction turbine, the heat
surplus can be used to generate extra electricity for the plant.
(v) The heat surplus can be cascaded downwards to locations with
negative variances provided they are still located in the same
TS-PTA pinch region.
A positive variance below the Total Site Pinch represents surplus
cooling utility produced by the utility plant, and it can be cascaded

Table 13
Total Site Problem Table algorithm (TS-PTA) during normal operation.
Utility

Utility Temp.( C)

HP

170

Net heat
source (kW)

Net heat
sink (kW)

Net heat
requirement (kW)

Initial heat
cascade

Final heat
cascade

Multiple utility
cascade

11,937.90

3.04

11,934.86

4805.18

6484.70

ST
HW

125
50e60

3.04
2967.17
1495.71

7769.31
7606.92

3.04

3.04

4802.14

4802.14

6111.22

6111.22
10,916.40

CW

10

35.15

Amount of
utility needed

L227.27

35.15
10,881.25

35.15

406

P.Y. Liew et al. / Applied Thermal Engineering 40 (2012) 397e408

Table 14
Total Site Sensitivity Table (TSST).
Utility

Total Site external utility requirement, kW

Normal
operation

Plant A
shutdown

HP

3.04

3.04

ST

4802.14

7769.31

HW

6111.22
Pinch
35.15

6821.32
Pinch
35.15

CW

Variance from
normal operation

Plant B
shutdown

Variance from
normal operation

Plant C
shutdown

3.04

2967.17

810.77

3991.37

4665.19

136.95

710.08

6896.82
Pinch
0

758.60

5863.17
Pinch
35.15

248.06

35.15

Variance from
normal operation
3.04

Plant D
shutdown

Variance from
normal operation

3.04
1161.15
Pinch
1247.65

0
3640.99
4863.57

35.15

STEP 1: Perform Problem Table Algorithm (PTA) for all individual process

STEP 2: Construct multiple utility cascade for each individual process

Above pinch temp. (heat sink)

Below pinch temp. (heat source)

Shift all the temperatures by deduct Tmin,pp/2 and add

with Tmin,up

Shift all the temperatures by add Tmin,pp/2 and deduct

with Tmin,up

Cascade the heat available from the highest

temperature towards pinch temperature, external utility
added when there is negative value in the cascade

Cascade the heat available from the lowest

temperature towards pinch temperature, external utility
added when there is positive value in the cascade

Sum the external heat enthalpy below the utility

temperature until before the next utility temperature

Sum the external heat enthalpy above the utility

temperature until before the next utility temperature

STEP 3: Perform Total Site Problem Table Algorithm (TS-PTA)

Formulate Net heat sink and Net heat source by adding heat sink from above pinch region at each
processes and heat source from below pinch region according to utility level
Calculate Net heat requirement by deducting net heat source with net heat sink

Cascade the net heat requirement from top to bottom by taking the most negative value in the
previous cascade as hot utility provided
Above Total Site Pinch

Below Total Site Pinch

Cascade the heat available from the highest

temperature towards pinch temperature, external utility
added when there is negative value in the cascade

Cascade the heat available from the lowest

temperature towards pinch temperature, external utility
added when there is positive value in the cascade

STEP 4: Construct Total Site Utility Distribution

(TSUD) Table

Application: Construct Total Site Sensitivity Table

(TSST)

Record all the heat sinks and sources of different

processes according to types of utility. Record also site
utility requirement as calculated in STEP 4

Record all the utility requirement calculated in STEP 4

Represent the heat flows from one process to another

or from utility to a process

Calculate the variance of normal operation with

situation when one of the plant shutdown

Fig. 4. Summary of the proposed methodology.

Omit the heat sink and source from

process shutdown.

Cascade the net heat requirement from top to bottom by assuming no hot utility provided

P.Y. Liew et al. / Applied Thermal Engineering 40 (2012) 397e408

upwards to serve as an extra heat source at higher temperature. A

negative variance indicates the central utility has a heat decit, and
more external utility must be generated. Based on this decit, the
designers can determine the maximum size of utility system that
must be built.
Based on Case study 2, the following conclusions can be drawn:
(i) HP - If Plant C is shutdown, there will be excess HP. Because all
the above pinch variance for utilities below HP (ST and HW)
are also positive, this heat cannot be cascaded downwards. The
HP must be diverted for electricity generation if a CHP system
is available, disposed of using cooling utilities or sold to other
plants. For other plant shutdowns, there is no effect on HP
consumption.
(ii) ST - For ST, more ST is needed if Plant A is shutdown, and
excess ST is generated if Plants B, C and D are shutdown. The
boiler generating ST should have a maximum design capacity
that can reach 7770 kW and the boiler could be turn down to
810 kW because if Plant B is shutdown, part of the surplus ST
available can be cascaded downwards to satisfy the HW
demand, a negative variance).
(iii) HW - For HW, more HW is needed if Plant A and B are shutdown
and excess HW is generated if Plant C and D are shutdown.
Hence, the boiler/heater generating HW should have
a maximum design capacity that can reach 6897 kW and a turn
down of not more than 4863 kW. If the turn down is more than
4863 kW, extra cooling utilities will be needed or the extra HW
can be sold to other plants. The HW utility requirement if Plant
B is shutdown can be obtained from the surplus ST available.
(iv) CW - If Plant B is shutdown, there will be 35 kW of extra
cooling water capacity available. This extra cooling water can
be used to remove the surplus heat from ST, or the cooling
tower can be shutdown.

5. Methodology summary
Fig. 4 presents a summary of the overall procedure for the four
useful tools proposed in this study: the Problem Table Algorithm
with multiple utility targets, the Total Site Problem Table Algorithm
(TS-PTA), the Total Site Utility Distribution (TSUD) table, and the
Total Site Sensitivity Table (TSST).
6. Conclusions
In the following, we present a summary of the contributions of
this work:
1) A new method was developed for calculating multiple utility
levels in the PTA that is simpler than that presented by Costa
and Queiroz [17]. This work introduced the use of multiple
utility cascades to determine multiple utility levels for individual PTAs and TS-PTAs. This tool enables the multiple utility
targeting for individual processes to be done effectively using
the numerical approach which produces more accurate results.
2) The TS-PTA was introduced for TSHI. We further demonstrated
that the TS-PTA yields more accurate results for TSHI analysis
when compared with a graphical approach, which is prone to
inaccuracies. The tool saves time and effort in determining
amounts of heat interchange among plants compared with
graphically constructed CCs, GCCs, TSPs and SCCs. This tool
could be explored further for the variable supply and demand
Total Site problem as proposed by Varbanov and Klemes [9].
Also, TS-PTA could be used for continuous and batch processes
that may not be conveniently solved using graphical tools.

407

3) The Total Site Utility Distribution (TSUD) table can be benecial

for the design of a Total Site utility distribution network. This
tool can be used to visualise and design the heat transfer
network in the system, between utility streams and process
streams.
4) The Total Site Sensitivity Table (TSST) is introduced to analyse
Total Site sensitivity. A typical example is, TSST can be use for
analysing the variation in a plants utility requirements when
one of the integrated site plants is shutdown for reasons such
as scheduled maintenance (e.g., for repairing faulty parts or
clearing unwanted material in the reactor), periodic shutdowns
(e.g., summer district heating shutdowns in the northern
hemisphere), operability problems or unpredicted accidents.
TSST results can also be used for utility design and production
planning.
The present research can be extended for the optimisation of
cogeneration potential. A prior study on assisted heat transfer [11]
can also be integrated into the TS-PTA. These developments should
be especially useful in increasing the applicability of the TS-PTA.
Heat storage in Total Site system also could be explored through
the mathematical tool proposed.

Acknowledgements
The authors would like to thank the Universiti Teknologi
Malaysia for providing nancial support through the UTM International Education Experience Fund and the nancial support from
the Hungarian project TMOP-4.2.2/B-10/1-2010-0025 and to the
University of Pannonia in Hungary for supporting the collaboration.

Nomenclature
Ts
Tt
T0
T00
CC
GCC
CW
HP
HW
LIES
LPS
mCp
PTA
Qcmin
Qhmin
SCC
SGCC
SSSP
TS
TSP
TSHI
TSST
TSUD
TS-PTA
UTA
DH
DTmin,pp

DTmin,up

Initial Supply Temperature ( C)

Final Target Temperature ( C)
Shifted Temperature ( C)
Double-shifted Temperature ( C)
Composite Curve
Grand Composite Curve
Cooling Water
High-Pressure Steam
Hot Water
Locally Integrated Energy Sector
Low-Pressure Steam
Heat Capacity Flowrate (kW/ C)
Problem Table Algorithm
Minimum Cooling Requirement (kW)
Minimum Heating Requirement (kW)
Site Composite Curve
Site Grand Composite Curve
Site SinkeSource Prole
Total Site
Total Site Prole
Total Site Heat Integration
Total Site Sensitivity Table
Total Site Utility Distribution
Total Site Problem Table Algorithm
Unied Targeting Algorithm
Stream Heat Load (kW)
Minimum Temperature Difference Between Process
Stream ( C)
Minimum Temperature Difference Between Utility And
Process Streams ( C)

408

P.Y. Liew et al. / Applied Thermal Engineering 40 (2012) 397e408

References
[1] V.R. Dhole, B. Linnhoff, Total site targets for fuel, co-generation, emission and
cooling, Comput. Chem. Eng. 17 (1993) S101eS109.
[2] K. Raissi, Total site integration. PhD Thesis, UMIST, Manchester, UK, 1994.
[3] J. Klemes, V.R. Dhole, K. Raissi, S.J. Perry, L. Puigjaner, Targeting and design
methodology for reduction of fuel, power and CO2 on total site, Appl. Therm.
Eng. 7 (1997) 993e1003.
[4] S. Ahmad, D.C.W. Hui, Heat recovery between areas of integrity, Comput.
Chem. Eng. 15 (12) (1991) 809e832.
[5] F. Marchal, B. Kalitventzeff, Energy integration of industrial sites: tools,
methodology and application, Appl. Therm. Eng. 18 (1998) 921e933.
[6] K. Matsuda, S. Tanaka, M. Endou, T. Iiyoshi, et al., Energy saving study on
a large steel plant by total site based pinch technology, Appl. Therm. Eng.
(2012). doi:10.1016/j.applthermaleng.2011.11.043.
[7] P.S. Varbanov, S. Doyle, R. Smith, Modelling and optimization of utiltiy
systems, Chem. Eng. Res. Des 82 (A5) (2004) 561e578.
[8] S. Perry, J. Klemes, I. Bulatov, Integrating waste and renewable energy to
reduce the carbon footprint of locally integrated energy sectors, Energy 33
(2008) 1489e1497.
[9] P.S. Varbanov, J.J. Klemes, Total site integrating renewables with extended
heat transfer and recovery, Heat Transfer Eng. 31 (9) (2010) 733e741.
[10] P.S. Varbanov, J.J. Klemes, Integration and management of renewables into
total slice with variable supply and demand, Comput. Chem. Eng. 35 (9)
(2011) 1815e1826.
[11] S. Bandyopadhyay, J. Varghese, V. Bansal, Targeting for cogeneration potential
through total site integration, Appl. Therm. Eng. 30 (2010) 6e14.

[12] A. Kapil, I. Bulatov, R. Smith, J.K. Kim, Site-wide low-grade heat recovery with
a new cogeneration targeting method, Chem. Eng. Res. Des (2012).
doi:10.1016/j.cherd.2011.09.001.
[13] A. Ghannadzadeh, S. Perry, R. Smith, Cogeneration targeting for site utility
systems, Appl. Therm. Eng. (2012). doi:10.1016/j.applthermaleng.2011.10.006.
[14] Z. Fodor, P. Varbanov, J. Klemes, Total site targeting accounting for individual
process heat transfer characteristics, Chem. Eng. Trans. 21 (2010) 49e54.
[15] P.S. Varbanov, Z. Fodor, J.J. Klemes, Total site targeting with process specic
DTmin, Energy (2012). doi:10.1016/j.energy.2011.12.025.
[16] B. Linnhoff, J.R. Flower, Synthesis of heat exchanger networks, AIChE J. 24
(1978) 2 parts. Part I: systematic generation of energy optimal network, 633642. Part II: evolutionary generation of networks with various criteria of
optimality, 642-654.
[17] A.L.H. Costa, E.M. Queiroz, An extension of the problem table algorithm for
multiple utilities targeting, Energ. Convers. Manage. 50 (2009) 1124e1128.
[18] U.V. Shenoy, Unied targeting algorithm for diverse process integration
problems of resource conservation networks, Chem. Eng. Res. Des 89 (12)
(2011) 2686e2705.
[19] R. Smith, Chemical Process: Design and Integration, John Wiley & Sons
Chichester, UK, 2005.
[20] I. Kemp, Pinch analysis and process integration, in: B. Linnhoff,
D.W. Townsend, D. Boland, G.F. Hewitt, B.E.A. Thomas, A.R. Guy, R.H. Marsland
(Eds.), A User Guide on Process Integration for Efcient Use of Energy, second
ed.. IChemE, Rugby, UK, 1994 Elsevier, Amsterdam, The Netherlands, 2007.
[21] J. Klemes, F. Friedler, I. Bulatov, P. Varbanov, Sustainable in Process Industry:
Integration and Optimization, McGraw Hill, New York, US, 2010.
[22] Canmet ENERGY, Pinch Analysis: For the Efcient Use of Energy, Water and
Hydrogen, Natural Resource Canada, Varennes, 2003.