Microsoft Word - 211.docx


 CHEMICAL ENGINEERING TRANSACTIONS  
 

VOL. 61, 2017 

A publication of 

 

The Italian Association 
of Chemical Engineering 
Online at www.aidic.it/cet 

Guest Editors: Petar S Varbanov, Rongxin Su, Hon Loong Lam, Xia Liu, Jiří J Klemeš 
Copyright © 2017, AIDIC Servizi S.r.l. 

ISBN 978-88-95608-51-8; ISSN 2283-9216 

A Superstructure Approach for the Design of 

Heating Utility System 

Diban Pitchaimuthu*, Jui-Yuan Lee, Mahmoud El-Halwagi, Dominic C. Y. Foo 

Centre of Excellence for Green Technologies/Department of Chemical and Environmental Engineering, University of 

Nottingham Malaysia, Broga Road, 43500 Semenyih, Selangor, Malaysia. 

done_87@yahoo.com 

Heavy crude oil produced from reservoir normally requires heating in order to facilitate proper oil water 

separation at the separator. Providing heat helps to reduce the viscosity of the crude thus reducing the residence 

time required at the separator and its size. In conventional heating utility system, the heating medium is supplied 

to its users in parallel design where all the users receive the heating medium at the supply temperature. This 

design however may lead to overdesign of the heating utility system. In this paper, a novel superstructure 

approach is proposed to determine the optimum network design of the heating medium system with minimum 

total annualised cost. The novelty of this approach is that it allows determination of the global optimum solution 

for the system while taking into consideration of all possible network configuration. The approach caters for 

capital and operating costs trade-off for the heat exchanger network (HEN) and waste heat recovery unit 

(WHRU) in the heating utility system. An industrial case study is used to elucidate the newly proposed technique. 

1. Introduction 

The oil price has suffered a big drop in recent years from its all-time high price of USD 146/bbl in 2006 

(Macrotrends, 2017).  As a consequence, oil production companies have to be more prudent either in their new 

exploration or operation of their existing facility in this uncertain market condition. A few potential improvements 

in the design of the existing oil and gas platform were studied by Nyugen et al. (2016a) in their research paper. 

They summarised that modification such as limiting anti-surge recirculation at the gas compression, installation 

of multi-level (low and high pressure) production manifold, and waste heat recovery as having good energy 

saving potential. Other energy efficiency improvements such as heat integration by direct heat exchange 

between the hot and cold process streams was found not viable due to operation reasons (Nyugen et al., 2016b). 

The installation of organic Rankine cycle - ORC (Pierobon et al., 2014) or steam Rankine cycle - SRC (Nyugen 

et al., 2014) to recover waste heat is an attractive option in the near future. Another system that has potential 

for improvement but has been overlooked is the heating utility system. This is reported by de Oilivera and Van 

Hombeeck (1997) that petroleum heating and separation steps as the most inefficient exergy user in oil and gas 

platform. The heating utility system is normally used on offshore platform when heavy crude oil is produced from 

the reservoir. Heating the heavy crude helps to reduce its viscosity thus cutting down the residence time required 

at the separator and consequently its size. Common heating media used in oil and gas platform are oil and 

water due to their availability while ethylene glycol is sometimes used at offshore location with freezing 

temperature. This paper discusses a systematic approach based on superstructure method in designing the 

optimum heating medium system.  

Heating utility system like other central utility systems is a unique HIWN problem, as a single source (heating 

medium) is used to provide heating requirement to all its users. This is different from the conventional heat 

exchange network (HEN) whereby heat is integrated between process streams and when there exist insufficient 

or excess heat, steam or cooling water are used to meet the demand of the overall network. The steam and 

cooling water is assumed able to be obtained from a source at particular price. However, this is far from true. In 

a typical chemical plant, the heating or cooling medium is generated from a utility system. Recent superstructure 

works avoids the consideration of such system. Due to this, their optimisation model cannot be applied directly 

for design of heating medium system. There are a few handful research works carried out on the design of 

                               
 
 

 

 
   

                                                  
DOI: 10.3303/CET1761314

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Please cite this article as: Pitchaimuthu D., Lee J.-Y., El-Halwagi M., Foo D.C.Y., 2017, A superstructure approach for the design of heating 
utility system, Chemical Engineering Transactions, 61, 1897-1902  DOI:10.3303/CET1761314  

1897



central utility systems. Some graphical approaches were proposed to target the minimum flowrate for cooling 

water network (Kim and Smith, 2001) and chilled water network (Foo et al., 2014), while Ataei et al. (2014) 

proposed a similar approach for hot oil system which closely resembles heating utility system. It is to be 

highlighted that their works did not find the optimum overall cost of the system rather only optimum heating 

medium flowrate. On the other hand, Bade and Bandyopadhyay (2014) proposed an iterative approach to 

determine the optimum total overall cost solution for the hot oil system. In their approach, they varied the 

minimum approach temperature (ΔTmin) to calculate the minimum flowrate and the respective area required for 

the network. The downside of their proposed method is that it is a time-consuming process, as detailed network 

for different ΔTmin values needs to be determined, before the total cost can be calculated. Besides, their 

approach does not guarantee global optimum. The proposed new superstructure model that is presented 

guarantees global optimum for the utility system which is not considered in conventional HEN model.   

2. Heating utility system problem statement 

For a given set of heating medium user (heat exchangers) m ϵ M, each with a fixed heat load Qm to be provided 

by the heating medium (water, hot oil or glycol), and with specified maximum inlet (Tin, m) and outlet (Tout, m) 

temperatures. The heating medium (with return temperature THMR) is reheated using WHRU using exhaust gas 

from gas turbine generator and gas turbine compressor at specified maximum inlet (Tin, g) and outlet (Tout, g) 

temperatures. The reheated heating medium at circulation temperature (THM) is re-circulated among the heating 

medium users to provide the necessary heating. The main objective is to synthesise an optimum heating 

medium network, that have the lowest total overall cost while meeting the heat load requirements of all heating 

medium users. 

 

From

Nitrogen 

system

Waste heat 

recovery unit

Heating 

medium 

make mp 

Heating 

medium pump

H
e

a
ti
n

g
 m

e
d

iu
m

 s
u

p
p

ly
 h

e
a

d
e

r

H
e

a
ti
n

g
  

m
e

d
iu

m
 r

e
tu

rn
 h

e
a

d
e

r

Heating medium 

dump cooler

Sea water 

supply

Sea water 

return

TC

LC

NNF=Normally 

No Flow NNF

NNF

NNF

NNF

HP Flare 

Knock-Out 

Drum Heater

Flue gas

To ATM

Closed Drain 

Drum Heater

Heavy Oil 

Heater

Light Oil 

Heater

Crude Oil 

Heater

Diesel Flushing/

Pigging Heater

Test 

Separator 

Heater

 

Figure 1: PFD of heating utility system for an offshore platform  

3. Superstructure model for HEN  

The general superstructure model for heating utility system is shown in Figure 2. The model is a generic 

representation of a single heating medium user in the system. For every user, the heating medium at supply 

temperature (FHM) is allowed to mix with the used heating medium from its own or other users (Fm’,m) at the 

mixing point C. After heat exchange with the process fluid, the heating medium flowrate can be split at the 

splitting point, S either to WHRU for reheating (Fw,m) or reused by the users (Fm,m). 

The mass balance at point C and S and the energy balance at point C is given by Eqs(1-3). 

Fm= ∑ Fm',m
mϵM

+FHM.m                       ∀m ϵ M                                                                                                       (1) 

1898



 

Figure 2: Generic superstructure model for heating utility system 

Fm= ∑ Fm,m'
mϵM

+Fw.m                          ∀m ϵ M                                                                                                     (2) 

FmTin,m= ∑ Fm',m Tout,m
mϵM

+FHM.m THM                      ∀m ϵ M                                                                                (3) 

where Tin,m and Tout,m are the inlet and outlet temperature of the heating medium passing through heat exchanger 

of m user, Fm is the flowrate of the heating medium into the heat exchanger of m user and THM is heating medium 

supply temperature. The energy balance at each unit heat exchanger is calculated using Eq.(4).  

Qm=Fm(Tin,m-Tout,m)                       ∀m ϵ M                                                                                                        (4) 

where Qm is the total enthalpy required by the individual user. For determination of heating medium return 
temperature, THMR Eqs(5-6) are used. 

Fw= ∑ Fw,m       
mϵM

                                                                                                                                                       (5) 

FwTHMR= ∑ Fw,m Tout,m
mϵM

                                                                                                                                       (6) 

The limit for the inlet and outlet temperature of the heating medium entering and leaving the heat exchanger are 

given by Eqs(7-10). 

Tin,m≤ THM                                ∀m ϵ M                                                                                                                      (7) 

Tin,m≥ Tin,m Limit                          ∀m ϵ M                                                                                                              (8) 

Tout,m≥ Tout,m Limit                       ∀m ϵ M                                                                                                                    (9) 

Tout,m <THM                               ∀m ϵ M                                                                                                              (10) 

In order to reduce the search scope when computing the solution, flow limit for each connection is introduced 

using Eq.(11). The flow limit is derived based on American Petroleum Institute (API) 14E (1991), which 

recommends that the flow velocity should not exceed maximum velocity of 4.572 m/s to minimise flashing ahead 

of control valve. While for the minimum velocity, it is recommended to limit the velocity to 0.9144 m/s to minimise 

deposition of sand and other solid. Using this information together with the assumption that the biggest and 

smallest pipe size that are used are 6” and ½”, the upper and lower flow limit for the connection is calculated to 

be 432.891 and 3.1408 kW/°C. 

3.1408I ≤ Fm',m ≤ 432.89I                    ∀m ϵ M      I[0,1]                                                                                           (11) 

For HEN area calculation, the log mean temperature difference is determined using Eq.(12). 

Tlog,m=
(Tin, m-Tout,m Limit)-(Tout, m-Tin,m Limit)

LN[(Tin, m-Tout,m Limit)-(Tout, m-Tin,m Limit)]
           ∀m ϵ M                                                                      (12) 

The area required for the HEN is calculated using Eq(13). 

AHEN= ∑
Qm

Tlog,mhm
mϵM

                                                                                                                                           (13) 

where hm is the overall heat transfer coefficient for user m. 

 

1899



4. Sizing and costing of other equipment in the heating utility system 

Besides the HEN, heating utility system of the offshore platform consists of two other major units i.e. WHRU 

and pump.  Design variables of these units, e.g. area of WHRU and power consumption of the pump are to be 

considered before the cost optimum heating utility system design can be determined. For the WHRU, its log 

mean temperature (TWHRU) is calculated based on Eq(14) that follows, i.e., 

TWHRU=
(Tin, g-THM)-(Tout, g-THMR)

LN[(Tin, g-THM)-(Tout, g-THMR)]
                                                                                                            (14) 

where Tin, g is the gas turbine generator and/or gas turbine compressor exhaust gas temperature to the WHRU 

while Tout, g is the exhaust gas temperature leaving WHRU after heat transfer with the heating medium.  

Next, the area of WHRU (AWHRU) is calculated using Eq(15), 

AWHRU=
Q

UeTWHRU
                                                                                                                                              (15) 

where Q is the total heat load required to be transferred from the exhaust gas to the heating medium and Ue is 

the enhanced overall heat transfer coefficient. For brevity, the steps involved in calculating these enhanced 

overall heat transfer is excluded from this paper, however readers can refer to the detailed steps found in Nitsche 

and Gbadamosi (2016). 

The shaft power required by the pump (Ps) to circulate the heating medium in the close loop system is calculated 

using Eq(16), 

Ps=
1.1Fmingd

1×10
3
CPHMη

                                                                                                                                              (16) 

where g is the gravitational force, d is the net differential head required to transfer the heating medium to its 

required users, CPHM is the specific heat capacity of heating medium and η is the pump efficiency. A design 

margin of 10 % is included in the heating medium flowrate for pump design purpose. The capital cost of the 
equipment is annualised based on the lifespan of the offshore platform which is typically between a period of 

20-30 y using Eq.(17). 

AF=
i(1+i)

n

(1+i)
n
-1

                                                                                                                                                      (17) 

where i and n represent the interest rate and number of years respectively. The annualised capital cost of HEN, 

WHRU and the pump are calculated based on the correlations in Eqs(18-20) (Smith, 2005), 

CHEN=N AF [a+b (
AHEN

N
⁄ )

c

]                                                                                                                             (18) 

where N = number of units/shells, whichever is appropriate, a = installation cost, b = area cost, c  = constant. 

CWHRU=CBAF (
AWHRU

DB
)

M

                                                                                                                                   (19) 

where CWHRU = WHRU cost with capacity AWHRU, CB  = known base cost of equipment with capacity DB, M = 

constant depending on equipment type   

CP=CBAF (
PS

DB
)

M

                                                                                                                                                 (20) 

where CP = pump cost with capacity Ps, CB  = known base cost of equipment with capacity DB, M = constant 

depending on equipment type 

Note that the capital cost correlation for WHRU is based on that for air cooled heat exchanger. The operating 

cost of the heating medium system (CU) based on the power consumption of the circulation pump is determined 

using Eq.(21). 

CU=PSWG                                                                                                                                                         (21) 

where Ps is the pump power, W is the duration of usage and G is the electricity tariff. 

The optimisation equation is set to minimise the total annualised cost (TAC) of the heating medium system, 

given by Eq. 22.  The model will also determine the optimum network configuration for the system. To elucidate 

the proposed superstructure model, an industrial case study is presented in the following section. 

1900



TAC=CHEN+CWHRU+CP+CU                                                                                                                              (22) 

5. Case study 

In this section, an industrial case study is used to demonstrate the use of the proposed superstructure approach 

to design a cost-optimum heating utility system.  An offshore floating production storage offloading (FPSO) 

platform is to be installed at Angola (Africa). Its PFD is similar to that in Figure 1, with its limiting data for the 

cold streams shown in Table 1. The flue gas inlet (Tin, g) and outlet temperature (Tout, g) are 433 °C and 130 °C. 

While the heating medium temperature (THM) is set at 130 °C. Data for the design of WHRU and pump are as 

follow: overall heat transfer, Ue (326.5 kWm-2°C-1), height, d (51.8m), efficiency, η (60%), heating medium 

average specific heat capacity, CPHM (4.326 kWm-2°C-1).  

Table 1: Limiting cold streams data for industrial case study 

Unit, m 
Inlet 

Temperature, 
Tin, m (°C) 

Outlet 
Temperature, 

Tout, m (°C) 

Cold streams overall heat 
transfer coefficient, 

hm (kWm-2°C-1) 

Load, 
Qm (kW) 

Limiting heat capacity 
flowrate, 

Fm(kW°C-1) 

1 30 65 2 3,630 103.71 

2 39 65 1 9,174 352.85 

3 57 90 1.5 6,689 202.70 

Total    19,493 659.26 

Table 2: Data for costing calculation. *Data from Smith (2005). **Data adjusted based on CE index of October 

2014 (Chemical Engineering, 2015). 

Parameters Value Parameters Value 

i 10 % WHRU*, M 0.89 

n 25 y Pump*, CB $ 9.2594x104** 

a* 260,800** Pump*, DB 4 kW 

b* 3260** Pump*, M 0.55 

c* 1 Duration, W 8,760 h 

WHRU*, CB $ 1.0171x106** Electrical tariff, G 0.12 $/h 

WHRU*, DB 200 m2   

 

 

Figure 3: Final configuration of heating utility network for case study (heat capacity flowrate in kW/°C, 

temperature in °C, given in parenthesis). 

Solving the objective function in Eq(22) subject to the constraints in Eqs(1-21) yields the minimum total 

annualised cost of the heating medium as $ 632,679. The optimum HEN design obtained from the superstructure 

solution is shown in Figure 3. The design encompasses a series design of the HEN. Table 3 shows the 

comparison between the result obtained from the superstructure approach against the conventional design (see 

Figure 1), where the heating medium is fed to all heat exchangers at 130 °C and returns at 90 °C.  

 

Unit 3

6689 kW

Unit 2

9174 kW

Unit 1

3630 kW

Hot Oil Supply

355.00 (130)

355.00 (85.32)

355.00 (75.09)

Hot Oil Return
Waste Heat 

Recovery Unit

355.00 (111.16)

1901



The conventional design TAC is determined as $ 664,880, thus an annual saving of $ 32,201 (corresponds to 

4.8 % reduction) can be achieved through the series design shown in Figure 3. 

Table 3: Comparison of cost between optimum design from superstructure approach and conventional design. 

Elements This work Conventional design 

Annualised capital cost 

HEN ($/y) 295,053 259,974 

WHRU ($/y) 205,596 212,642 

Pump ($/y) 51,680 66,206 

Total annualised capital cost  ($/y) 552,329 538,822 

Total Utility Cost ($/y) 80,350 126,059 

TAC ($/y)  632,679 664,880 

6. Conclusions 

This work proposes a novel superstructure approach for the design of heating utility system, which is not covered 

in conventional HEN design.  The proposed method allows the lowest total annualised cost of the system to be 

determined by weighing the best trade-off between capital and operating cost of the system. The main 

advantage of this approach as compared to other techniques that has been published for central utility design 

is that it allows global optimum solution to be obtained. Besides, the proposed superstructure approach provides 

the solution for optimum HEN which produces the minimum TAC. An industrial case study was used for 

demonstration; it proves that the total annualised cost of the system is reduced by 4.8 % by changing the existing 

HEN from parallel to series design. 

References 

API recommended Practice 14E (RP 14 E), 1991. Recommended Practice for Design and Installation of 

Offshore Production Platform Piping Systems (5th Ed.), Washington, USA. 

Ataei A., Tahouni N., Seyedi S.M.H., Hashemian S.M., Yoo C.K., Panjeshahi M.H., 2014, A novel approach to 

hot oil system design for energy conservation, Applied Thermal Engineering, 66, 423-434.  

Bade M.H., Bandyopadhyay S., 2014, Thermal integration of heat transfer fluid systems, Asia-Pac. J. Chem. 

Eng, 9, 1-15. 

Chemical Engineering, 2015, Economic Indicators, <www.isr.umd.edu/~adomaiti/ chbe446/literature/ 

ChECostIndexJan2015.pdf>, accessed 5/05/2017. 

de Oliveira Jr. S., Van Hombeeck M., 1997, Exergy analysis of petroleum separation processes in offshore 

platforms, Energy Convers Manage, 38(15–17),1577–1584. 

Foo D. C. Y., Ng D. K. S., Leong M. K. Y., Chew I. M. L., Subramaniam M., Aziz R., Lee J. Y., 2014, Targeting 

and design of chilled water network, Applied Energy, 134, 589-599. 

Kim J-K., Smith R., 2001, Cooling water system design, Chem.Eng Science, 56, 3641-3658. 

Macro trends, 2017. Crude Oil Price History Chart. Macrotrends, n.d., <www.macrotrends.net/1369/crude-oil-

price-history-chart>, accessed 17/02/2017. 

Nitsche M., Gbadamosi R.O., 2016. Heat exchanger design guide; A practical guide for planning, selecting and 

designing shell and tube heat exchangers, Elsevier Inc, Oxford, UK, ISBN: 978-0-128-03822-2.  

Nguyen T.V., Voldsund M., Breuhaus P., Elmegaard B., 2016a, Energy efficiency measures for offshore oil and 

gas platforms, Energy, 117, 325-340. 

Nguyen T.V., Fullop T.G., Breuhaus P., Elmegaard B., 2016b, Life performance of oil and gas platforms: Site 

integration and thermodynamic evaluation, Energy, 73, 282-301. 

Nguyen T.V., Tock L., Maréchal F., Elmegaard B., 2014, Oil and gas platforms with steam bottoming cycles: 

retrofit integration and thermo-environomic evaluation, Applied Energy, 131, 222–237. 

Pierobon L., Benato A., Scolari E., Haglind F., Stoppato A., 2014, Waste heat recovery technologies for offshore 

platforms, Appl Energy,136:228–241. 

Smith R., 2005. Chemical process design and integration, John Wiley & Sons Ltd: West Sussex, UK. ISBN: 

978-1-119-09441-8. 

1902