Sie sind auf Seite 1von 11

Applied Thermal Engineering 105 (2016) 763773

Contents lists available at ScienceDirect

Applied Thermal Engineering


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

Two types of welded plate heat exchangers for efficient heat recovery in
industry
Olga P. Arsenyeva b,, Leonid L. Tovazhnyanskyy a, Petro O. Kapustenko a, Genadii L. Khavin b,
Anna P. Yuzbashyan a, Pavlo Yu. Arsenyev b
a
b

National Technical University Kharkiv Polytechnic Institute, 21 Frunze Str., 61002 Kharkiv, Ukraine
AO SPIVDRUZHNIST-T LLC, Krasnoznamenny per. 2, off. 19, 61002 Kharkiv, Ukraine

a r t i c l e

i n f o

Article history:
Received 18 December 2015
Revised 19 February 2016
Accepted 13 March 2016
Available online 21 March 2016
Keywords:
Plate heat exchangers
Heat transfer
Crude oil preheat train
Fouling mitigation

a b s t r a c t
The developments in design theory of welded Plate Heat Exchangers (PHEs), aiming to enhance the heat
recovery and efficiency of energy usage, are presented. The thermal and hydraulic performance of the
unit is estimated using two approaches: by proper selection of plate corrugation pattern and by adjusting
the numbers of passes for heat exchanging streams. The optimisation problem targeting the minimal heat
transfer area under the requirements of proper operating conditions is observed. The optimising variables
include the number of plates with different corrugation geometries in one pass. To estimate the value of
the objective function in a space of optimising variables the mathematical model of PHE is developed. The
possibilities of their application as heat exchangers in preheat train of crude oil distillation unit of the oil
refinery are analysed basing on obtained design parameters with the effect of flow movement arrangement in the unit and its influence on shear stress and fouling formation. The comparison of Plate-andShell and Compabloc types of welded PHE is discussed.
2016 Elsevier Ltd. All rights reserved.

1. Introduction
The separation of the oil products into fractions such as LPG,
naphtha, kerosene, diesel and atmospheric residue is carried out
in the crude oil distillation unit due to the different boiling point
temperatures. Oil refining is energy-intensive, requiring significant
amounts of heat energy. From 7% to 15% of the crude oil input is
used by the refinery processes, according to Szklo and Schaeffer
[1]. The amount of energy consumed by the distillation unit comes
up to 3545% of the total energy consumption by the refinery [2].
Usually this process is performed by the crude oil heating by other
products, available at the refinery, among which are light and
heavy distillates, atmospheric residue and others. These hot products are pumped in the preheat trains prior to the distillation fire
heater to increase the temperature of crude oil by heat recuperation. The bigger amount of recuperated heat can lead to considerable saving of energy required to be supplied in the fire heater and
corresponding reduction of green house gases emission. But the
increasing of heat recovery for the preheat train with traditional
shell-and-tube heat exchangers (HE) requires additional heat
transfer area in a number of locations. This has an adverse effect
Corresponding author. Tel.: +380 577202278; fax: +380 577202223.
E-mail address: o.p.arsenyeva@gmail.com (O.P. Arsenyeva).
http://dx.doi.org/10.1016/j.applthermaleng.2016.03.064
1359-4311/ 2016 Elsevier Ltd. All rights reserved.

on economics considerations, as the installation of additional heat


transfer area can be very expensive. One way to overcome this
problem is to enhance the performance of heat exchangers with
intensified heat transfer technique, as it is shown by Jiang et al.
[3] on example of shell-and-tube HEs enhanced with tube inserts.
Another important issue is to maintain the heat recovery on higher
level over the period of plant operation before cleaning heat
exchangers [4], that is also possible with fouling mitigation on
intensified heat transfer surfaces.
The heat transfer intensification is an intrinsic feature of modern Plate Heat Exchangers (PHEs). The design and operation of
PHEs are well described in the literature, e.g. [5]. PHE channels
are formed by specially corrugated plates produced by stamping
from thin metal sheets. It is observed that high heat transfer coefficients and low fouling tendencies are typical characteristics of
PHE channels of complex geometry because of high levels of turbulence, effects that are similar in principle to those observed in
enhanced tubes and tubes with inserts [6]. The advantages of PHEs
in many applications were demonstrated by plate-and-frame type
of PHEs, which were initially introduced in 1930th and gradually
undergone significant improvements in their construction and
design, especially in the last decades [7]. In many industrial applications plate-and-frame PHE has up to four times higher overall
heat transfer coefficients than traditional shell-and-tube HEs at

764

O.P. Arsenyeva et al. / Applied Thermal Engineering 105 (2016) 763773

the same conditions, require much less material for heat transfer
surface, much more compact and suitable for economically viable
solutions [8]. It is confirmed by a number of researchers, as e.g.
Hajabdollahi et al. [9] have found in their study case of water to
water HE that the comparison of the optimum results for plateand-frame PHE shown 13% improvement in the total cost compared with shell-and-tube heat exchanger at the same operating
conditions. To similar conclusion came Perevertaylenko et al. [10]
in their study of cost effective ways for amine absorption unit
design in CO2 post-combustion capture process. The use of PHEs
in Heat Exchanger Network (HEN) of Absorption Desorption Unit
allowed with the same or even 15% smaller purchased cost of heat
exchangers to save up to 13% more energy as compare to optimised
HEN with conventional shell-and-tube HEs.
Important feature of plate-and-frame PHE is its ability to be disassembled for mechanical cleaning of heat transfer surface, as the
channels between plates are sealed by elastomeric gaskets. On the
other hand, the presence of gaskets is limiting the range of plateand-frame PHE application by temperatures lower than maximum
180 C and pressures below 25 bar. In the construction of welded
PHE the gaskets between plates are eliminated, that allows to
widen significantly the range of its application on temperatures
and pressures. Such PHEs can be used in preheat trains of crude
oil distillation units with temperatures up to 350 C [12]. Nowadays there are a number of different by construction principles
types of welded PHEs produced by contemporary PHE manufacturers, as is discussed in a book by Klemes et al. [5]. In this paper are
considered two recently most widely used types of welded PHEs:
Plate-and-Block HE and Plate-and-Shell HE (PSHE).
The most known representative of Plate-and-Block type is Compabloc HE [11], which was originally developed and manufactured
in 1980s. According to the data, published by Andersson et al. [12]
for the moment of publication it was installed more than 750 Compabloc HEs in oil refineries around the world on different positions,
among which 200 units were installed for operating in the crude
oil preheat train. Comparing with the conventional shell-andtube units, Compabloc is more compact and requires less space
for installation. The value of surface area per unit volume for heat
transfer core can reach up to 200 m2/m3 and more, while for shelland-tube HEs this parameter is from 7 to 10 m2/m3. The complex
channel geometry formed by the corrugated plates provides the
different channel spacing, which varies from zero value in contact
points to the double height of the corrugation in the largest gap.
The Compabloc heat exchanger is based on the square shaped
plates welded in blocks arranged in different passes combination
for hot and cold streams movement, see Fig. 1. At the presented
picture each stream goes through four individual groups of plates
with cross flow between two heat exchanging streams. The flows
are directed into passes using the baffle plates extending across
the whole cross-section of the heat exchanger collector. The hot
stream enters the unit from the top, and cold from the bottom of
the unit creating overall counter flow arrangement with cross flow
in individual groups of plates. While cross flow is less effective for
heat transfer, the conditions for streams distribution across individual channels are much better, than in conventional plate-andframe PHE with pure counter flow. There are much smaller local
hydraulic resistances in channels inlet and outlet zones, where
the channel cross section area not changing so dramatically as in
distribution zones of counter flow PHE. The stream is entering
Compabloc PHE channel through the full its width, no additional
change of cross section area. In plate-and-frame PHE and PSHE
with parallel flow of streams (see Fig. 2) part of the channel width
is blocked to arrange the outlet (or inlet) of opposed stream. It creates additional hydraulic resistance at channel inlet and exit.
The Plate-and-Shell type HE was first commercially produced
by Vahterus Oy Company [13] in 1990s and now is manufactured

Fig. 1. The construction and operation principle of Compablock heat exchanger


(courtesy of OAO AlfaLaval Potok).

Fig. 2. The construction and operation principle of Plate-and-Shell PHE, after Freire
and Andrade [14].

by all major producers of PHEs. The PSHE consist of a number of


welded together round plates, as shown in Fig. 2. The schematic
drawing of heat transfer plate is presented in Fig. 3a. The plates
with prolonged form are also manufactured by some producers.
In this PHE the counter flow arrangement of streams can be
organised and also multiple passes on both streams. The possible
corrugations forms are presented in Fig. 3b. In Compablocks
produced by AlfaLaval the triangular shape of corrugations is used
(see Figs. 3b and 4).
In some applications PSHE can be less costly than Compabloc
HE and consume less material for production. Welded construction
of plate pack prevents any intermixing between channels (Freire
and Andrade [14]). PSHE allow using this type of heat exchangers
for high temperature and pressure. They can be used for general
cooling and heating duties, as condensers, evaporators, reboilers
and steam heaters and operate under the temperature beyond

O.P. Arsenyeva et al. / Applied Thermal Engineering 105 (2016) 763773

765

Fig. 3. The drawing of PSHE plate (a) and possible forms of corrugations on PHE plates (b).

Fig. 4. The possible organisation of passes inside PHE: (a) single pass arrangement; (b) multi pass arrangement.

400 C and the pressure up to 100 bar [5]. Many researchers investigated the application of PSHE for different duties. Nakaoka and
Uehara [15] carried out the research of PSHE used for ocean thermal energy conversion. In this work they provide the technique
for prediction of overall heat transfer coefficient and friction factor
for the water side of the heat exchanger. Freire and Andrade [14]
observe the possibility of PSHE application as steam generator in
naval nuclear reactors. The application of PSHE in oil refinery for
pre-heat train of crude oil is the efficient solution for energy saving
and needs proper design of the units and investigation of fouling
formation in this equipment for the observed duties.
The welded construction allows to extend considerably the PHE
application range on working temperature and pressure, while
limiting the access to heat transfer surface for mechanical cleaning
of fouling deposits. The cleaning is possible with rinsing by chemical solutions or with partial access to heat transfer surface from
the sides of PHE channels for both streams surfaces in Compabloc
PHE or only for one of them in PSHE. In welded PHEs made on the
principle of conventional plate-and-frame PHE the access to heat
transfer surface is not possible for both streams. Such type of
PHE is not considered in present study.
The efficient use of welded PHEs in different recuperation systems and heat exchanger networks (HENs) require reliable method
for their rating and sizing. The commercial, proprietary computer
software models available in the industry can be used for detailed
design, especially on a stage of ordering the equipment. But these
codes are not suitable for the grass root HEN optimisation and its
retrofit [3]. Hence compact but accurate enough and reliable
method is required for the welded PHE performance calculation,

which is suitable for HE simulation, especially on a stage of HEN


optimisation. The essentials of different heat exchangers design
are exhaustively described and analysed in literature, see e.g.
[16]. The features of PHE construction renders to their thermal
and hydraulic design procedure considerable differences compare
to design of shell-and-tube HEs [8]. The detailed analysis of different approaches to plate-and-frame PHE design is presented by
Arsenyeva et al. [17] and more recently in a book [5]. The features
of welded PHEs construction introduce some additional specifics in
their design procedure. PSHE with parallel flow of heat exchanging
streams can be modelled practically in the same way as plate-andframe PHE [18]. But the Compabloc type HE has the construction
feature of cross flow in one pass of streams movement with possibility to arrange overall counter flow in PHE as a whole unit. This
issue was not much addressed in the literature, e.g. Tamakloe
et al. [19] have used mixingmixing model of cross flow in their
modelling of Compabloc PHE. Compare to model of not mixing
streams or model with one stream not mixing another fully
mixed, mixingmixing model gives lowest prediction of heat
exchanger efficiency [16]. From the point of hydraulic modelling
the important difference with plate-and-frame PHE is the possibility to arrange much better entry and exit zones of the PHE channel
without additional blockage in places of gaskets support. These
zones can make considerable contribution to a total pressure loss
in PHE [20]. The correct accounting of all these features of welded
PHEs is important requirement for their accurate modelling and
design.
The implementation of enhanced heat transfer equipment into
industry requires the reliable operation of the installed equipment

766

O.P. Arsenyeva et al. / Applied Thermal Engineering 105 (2016) 763773

during the proper time between shutdowns for maintenance. As


the process goes on, the heat exchangers tend to the fouling formation on heat transfer surface due to chemical and physical properties of heat carrier streams. It affects the heat transfer and
hydraulic performance of the units. In refineries 20% of all energy
consumption is considered to be lost due to the fouling in heat
exchange equipment [21]. The grows of fouling deposit on the heat
transfer surface affects the heat transfer capacity of the heat
exchanger, that results in more fuel consumption. The application
of heat transfer equipment with lower level of fouling formation
can reduce the energy consumption in oil refinery applications.
The problem to correctly design the heat exchanger accounting
the fouling formation on its surface with possibility to predict the
level of deposit in time and provide the reliable maintenance
schedule for the heat transfer equipment cleaning is of big importance, as shown by Assis et al. [22]. Depending on the origin of the
crude oil, its composition varies. The main important factor for
fouling formation is the amount of asphaltenes, their tendency
for solubility and reactivity with the rise of temperatures. The
problem of prediction the fouling level in heat transfer equipment
is complicated because of the variability of crude oil types and its
compositions. Extensive experimental researches of fouling in heat
exchangers of preheat trains at crude oil distillation units have
established that under some conditions exchangers may not foul.
It is particularly interesting since if to find and maintain such
favourable conditions in heat exchanger then the fouling problem
is ultimately solved. To deal with this phenomenon Ebert and Panchal [23] introduced the concept of threshold fouling models for
crude oil fouling inside tubes of shell-and-tube heat exchanger,
expressing fouling rate as a difference between deposit formation
and suppression terms and introducing dimensional parameters
which are vary for different crude oils and can be adjusted. Polley
et al. [24] modified the form of threshold model introducing in
removal term the rate of convective mass transfer between surface
of fouling deposit and bulk of fluid. For tube side of shell-and-tube
HEs the threshold conditions predicted with this model were confirmed in industrial tests, see e.g. [25]. Later for complex channel
geometries Polley et al. [26] have proposed to use EbertPanchal
model with deposition term expressed through film heat transfer
coefficient. To extend fouling model to tubes with inserts and other
channels of complex geometries Yang and Crittenden [27] proposed another approach. They modified Yeap et al. [28] model by
incorporating surface shear stress into the removal term and
equivalent velocity creating the same shear stress in smooth tube
as in tube with heat transfer enhancement. In paper [29] the influence of equivalent velocity was expressed through surface shear
stress, that made model application for different intensified surfaces more convenient.
Expensive cost related to the maintenance shutdown of the
refinery impose that it should be planned carefully and heat transfer equipment in operation should stand the operating periods
between cleaning. It requires the corresponding design of the heat
exchangers, with the maximum allowable fouling formation
deposit not to be achieved between the shutdowns. It needs the
reliable estimation of shear stress at the heat transfer surface in
the channels and determining fluid properties.
The properties of crude oils are significantly varies depending
on its type and origin, being significantly dependent of operating
conditions, especially temperature [30]. At temperatures below
50 C at pipeline transportation conditions heavy crude oils exhibit
a non-Newtonian behaviour and require knowledge of their rheological properties [31]. As shown by Djemiat et al. [32], rheology is
important at a low shear rate, while it was also noted that the
Newtonian behaviour occurs at high values of the gradient of shear
rate. When designing HEs at higher temperatures and shear stresses crude oil and its products are regarded as Newtonian fluids,

which properties are significantly changing with the temperature.


The design and selection of HEs for crude oil preheat train must
account the possibility of working with crude oils of different origin with which refinery is expected to work and corresponding
temperature correlations for crude oil and its products properties.
In the presented here case study the one possible crude oil is
analysed to obtain clear picture on comparison of two types of
welded PHEs at different process conditions, which can be also
used for analysis of applications in other industries. In a paper
are presented the main features of the welded PHEs modelling
and design with Equations for heat transfer and pressure loss calculation in PHE channels of different geometry. The cross flow local
arrangement of streams and overall counter flow in Compabloc is
compared to pure counter flow in PSHE with accounting for plates
construction differences and fouling formation on heat transfer
surfaces.
2. Thermo-hydraulic design of PHE
The cold and hot streams inside PHE can be distributed using
single pass construction, or multi passes (see Fig. 4). As a rule,
the multi-pass arrangement of streams can provide more wide
range of possible distribution of the allowed pressure drop inside
PHE, but also it needs special and more complex construction for
the separation and organising the movement of streams to the
proper channels. Thus the single pass construction is easier to
design and further maintenance and, as a rule, is cheaper, than
multi pass. But in some applications, especially where the considerable difference between flow rates on the hot and cold side exist,
the design with single pass arrangement of heat carriers movement is not the best solution.
For the PHE design the operating conditions, in which this unit
will operate, should be known and they are the input parameters:
 Heat carriers and their thermo-physical properties for the hot
and cold side;
 t11, t12 are the inlet and outlet temperatures of the hot stream,
C;
 t21, t22 are the inlet and outlet temperatures of the cold stream,
C;
 qm1, qm2 are the mass flow rates of the hot and cold streams,
respectively, kg/s;
 DP1, DP2 are the allowable pressure drops for the hot and cold
streams, respectively, Pa.
The problem is to find the heat exchanger with minimal heat
transfer surface area, which will suit the given operating
conditions.
2.1. The calculation of heat transfer and pressure drop in PHE channels
The overall heat transfer coefficient can be expressed as:

1
1
1 dw

Rfoul
U h1 h2 kw

where h1 and h2 are the film heat transfer coefficients for hot (distillation products) and cold (crude oil) sides, W/(m2 K), which can
be found from the values of the Nusselt numbers for the sides calculated by corresponding equations; dw is the thickness of the plate
wall, m; kw is the heat conductivity of plate material, W/(m K); Rfoul
is the total thermal resistance of the fouling deposits.
The design of heat exchanger consists of the modelling of
steady-state behaviour at first for the clean heat transfer surface
without the fouling deposit. The effectiveness-NTU method is
applied for heat transfer modelling. The basis of this method is

767

O.P. Arsenyeva et al. / Applied Thermal Engineering 105 (2016) 763773

Fig. 5. The possible organisation of passes inside Compabloc PHE.

the estimation of three parameters for all the channels in HE: the
heat transfer effectiveness e in the channel, the number of heat
transfer units (NTU) for the channel and the ratio between heat
capacity flow rates for the hot (1) and cold (2) sides. For application
considered in this paper crude oil is flowing at the cold (2) side of
HE. The type of channels for both sides of considered PHEs (see
Fig. 3) is the same and indexes specifying side in following Equations of this paragraph could be omitted for brevity.
For the calculation of pressure drop on the main corrugated
field of PHE channel the geometric parameters of plates corrugations (see Fig. 3) must be accounted, that is made when calculating
the friction factor in the channels for hot (index i = 1) and cold
(index i = 2) sides by following Eq. [33]:

"
#121
12
12 p2i
1
fi 8 

;
3
Rei
Ai Bi 2
2

1316



6:4
exp Pr30
1
c 
3 1  0:012  Re0:27

"

1:75

A1i 380=tgbi 

where p1i, p2i, p3i, p4i, p5i are the parameters defined by channel corrugation form.

p1i exp0:15705  bi ; p2i

pbc2i

2:63

p4i 0:061 0:69 tgbi



bi
; p3i exp p  180
 c12 ;
i

 1 1  ci  0:9  b0:01
;
i

bi
10

where ci = 2b/S is the corrugation doubled plate spacing b to corrugation pitch S ratio; bi is the corrugations inclination angle, degrees;
Rei = wideqi/li is the Reynolds number; de is the equivalent diameter of the channel, de = 2b, m; wi is the stream velocity in the channel, m/s; l is the dynamic viscosity of fluid, Pa s; qi is the density of
the fluid, kg/m3.
To account for the pressure losses in distribution zones at the
inlet and outlet of the inter-plate channel between commercial
plates, Arsenyeva et al. [20] introduced the coefficient of local
hydraulic resistance in those zones fDZ, assuming that it accounting
the pressure losses in both these inlet and outlet zones. To obtain
the pressure loss of the stream in PHE the pressure losses in channel should be multiplied by the number of passes and added by
pressure loss in ports and collectors:



qi  w2porti
q  w2i
LF q  w2i
 X i 1:3 
Dpi fi   i
fDZ  i
de
2
2
2

where li and lwi are the dynamic viscosities for the stream and wall
temperatures, Pa s; Nu = hide/ki is the Nusselt number; ki is the thermal conductivity of the fluid, W/(m K); hi is the film heat transfer
coefficient, W/(m2 K); Pri is the Prandtl number; fi is the friction factor accounting for total pressure losses in the channel, calculated by
Eq. (2); wi is the share of pressure loss due to friction on the wall in
total loss of pressure; FX is the coefficient of surface area enlargement due to corrugation. For the range of Pr numbers 0.7 < Pr < 15
the exponent c in Eq. (4) is taken c = 0.4. For higher Pr values
15 < Pr < 1000, which can be observed for colder crude oil at HEs
near its entrance to preheat train, the equation presented in paper
[34] is used:

p5i

p5i 1

 wi  fi =F X 3=7  Prci  li =lwi 0:14

The value of wi is estimated according to relation [33]:


16
37530  p1i
C7
A5 ; Bi
0:9
Rei
7p3i
0:27  105
Re

6
B
Ai 4p4i  ln @

6=7

Nui 0:065  Rei

where LF is effective plate length, m; w is velocity in PHE channels,


m/s; wport is velocity in PHE ports, m/s; X is the number of passes for
the considered stream; de is equivalent diameter of PHE channel, m;
q is the density of fluid, kg/m3.
The film heat transfer coefficients are estimated using for hot
(1) and cold (2) sides the following relation [33]:

wi Rei =A1i 0:15sinbi


wi 1 at Rei 6 A1i

at Rei > A1i ;

2.2. The e-NTU design of PHE


In multi-pass PHE the collectors are blocked in certain places to
organise flow of hot and cold streams through consecutive groups
of channels. The examples of distribution of streams inside single
pass and multi-pass PHE are shown in Fig. 4. The temperature distribution in the provided passes can be different and both countercurrent and co-current flows may occur in different groups of
channels with parallel flow of streams like in PSHE. In Compablock
type PHE in separate groups of channels exists the cross flow of
streams, with overall counter flow in PHE. The example of such
arrangement is shown in Fig. 5. The approach of multi-pass PHE
design considers the unit with different blocks as a system of
single-pass packs of plates. The individual packs are calculated as
separate units with proper temperature.
The operation conditions for all channels in such pack are the
same. The flow arrangement on Fig. 4b demonstrates four passes
for the hot stream (X1 = 4) and two for the cold stream (X2 = 2).
The heat transfer area of the block is determined by Fb = F/(X1X2),
where F is the total heat transfer area of the unit, m2. The change
of hot stream temperature in each block is Dti, i = 1, 2, . . ., 8. The
total number of blocks is nb = X1X2. According to the e-NTU
method the number of heat transfer units (NTU) in single block,
assuming that flow heat capacity of hot stream smaller than that
of cold stream, can be expressed as:

NTUb U b  F b  X 2 =G1  cp1

where Ub is the overall heat transfer coefficient in this block, W/


(m2 K); G1 is the mass flow rate of hot stream, kg/s; cp1 is specific
heat of hot stream, J/(kg K).

768

O.P. Arsenyeva et al. / Applied Thermal Engineering 105 (2016) 763773

The effectiveness eb for the block inside heat exchanger according to [16] is as follows.
For counter current flow:

1  expNTUb  Rb  NTUb
;
1  Rb  expNTUb  Rb  NTUb
NTUb

1 NTUb

eb
eb

for Rb 1
8

where Rb = G1cp1X1/(G2cp2X2) is the relation between flow heat


capacities of streams going through block; G2 and cp2 are mass flow
rate [kg/s] and specific heat [J/(kg K)] of cold stream.
For co-current flow:

1  expNTUb  Rb  NTUb
;
1 Rb
0:51  expNTUb 

eb
eb

for Rb 1
9

In case of cross flow in plate heat exchanger Tovazhnyansky


[35] experimentally confirmed that the Equation for one fluid
unmixed and another mixed can be applied approximately in following form:

eb 1  exp 


1  expNTUb  Rb 
;
Rb

10

The temperature change of the hot and cold stream in block i


can be calculated through the temperature difference between
streams entering the block Dtb inlet :

Dt1i ebi Dt b

inlet

and Dt 2i Dt 1i  Rb

11

Arsenyeva et al. [17] have shown that for any number of passes
in the unit the following system of equations in matrix form can be
obtained:

ZDt i  ebi D

12

where [Dti] is the vector-column of temperature drops in blocks;


[eiD] is the vector-column of the right parts of the equations in
the system; D is the temperature difference of streams entering
HE, C; [Z] is the matrix of system coefficients (i is the row number;
j is the column number), which elements are:

ebi Rb

n
  h i

o

3
1  sign j  int i1
1 X 1 0:5 1 ; if j > i
Xi
7
7
if i j 7;
5
i


o
X 2  j 0:5 1 ;
1  sign inti1
if j < i
X2

6 2X1
6
zij 6 1;
4 n
ebi

2X 2

13
The friction factors and heat transfer in blocks are found
depending on the used corrugation geometry of the plates according to Eqs. (2)(6). The numerical solution of this type of linear
algebraic Equations system (12) easily performed on a PC numerically, after which the outlet temperature of hot (1) stream in the
PHE can be calculated as:

t calc
12 t 11 

X1
X
i1

!
X2
1 X

dti1X 2 j ;
X 1 j1

14

The outlet temperature of the cold (2) stream is determined by


heat balance of PHE. Eqs. (1)(14) with relations determining necessary thermal and physical properties of streams represent the
mathematical model of PHE with known geometrical parameters
of plates and their corrugations. Its solution represents rating
of PHE. The algorithm of PHE design is presented in paper [17].
For welded PHEs the application in one PHE of plates with different
corrugation angle is not considered here.
The influence of multi pass principle on the efficiency of heat
transfer between streams in PHE can be estimated by analysis of
e-NTU relations in separate blocks of plates with different flow

arrangements. The most efficient way to exchange heat is counter


current flow arrangement which e-NTU relation expressed by Eq.
(8). It can be achieved in PSHE, as also in conventional plate-andframe and brazed PHEs. For the equal numbers of passes for both
streams this advantage is also preserved, especially for a big numbers of plates in one pass, when the influence of the channels with
co-current flow at the border between passes can be neglected.
When the numbers of passes are not equal, some blocks of plates
inevitably have co-current flow arrangement which e-NTU relation
is expressed by Eq. (9) by which the heat transfer effectiveness significantly drops up to 45% even at small values of NTU like
NTU = 0.2. It leads to significant decrease of overall efficiency for
the whole PHE with parallel streams flow and unequal number
of passes. could be
For blocks of plates with cross flow arrangement (like in Compabloc type PHE) the decrease of effectiveness compare to counter
flow is much lower, especially at NTUb smaller than 1. At
NTUb = 0.5 it is only 2.4% and at NTUb = 1 it is 6.3%, decreasing
with NTU value down to 22% at NTUb = 4. The combination of
the blocks with low NTUb in PHE with overall counter flow
arrangement of cross flow blocks can give the solution for total
NTU required by the application conditions with very small loss
of efficiency compare to pure counter current flow. It renders the
significant advantage for PHE with cross flow inside passes in case
of not equal passes numbers, compare to PHE with parallel flow of
streams in which some number of channels is working in cocurrent flow with very low effectiveness. The presented in this
paper case study enables to discuss this difference in more detail.
2.3. Time related behaviour of welded PHEs due to fouling
The models of heat transfer in plate heat exchangers includes
the parameter of thermal resistance of fouling deposit on heat
transfer area, what requires the modelling of dynamic conditions
inside heat exchanger. The reliable dynamic model of heat exchanger performance including an appropriate fouling model may
result in better accuracy in predicting fouling behaviour, as can
be concluded by analysis of recent developments in fouling
research covered in the book edited by Coletti and Hewitt [36].
In crude oil preheat trains the fouling on heat transfer surface
has a very complex mechanism, caused by both organic and inorganic causes, which depend on composition of the used crude oil
and also on operating conditions of heat exchanger. According to
threshold fouling models [23] the fouling accumulation rate is
the result of the difference between the fouling deposition rate
ud and fouling removal rate ur. When ud is equal to ur, it causes
the condition, when the fouling deposit thickness is stable and
dont grow. Before it happen the removal mechanism is stronger
than deposition. The adhesion of fouling to the heat transfer surface can only start after certain threshold conditions, such as
those of surface temperature and surface shear stress, have been
met.
The modification of threshold model for channels of PHE and
the dependence of the fouling deposition on wall shear stress was
discussed by Arsenyeva et al. [29]. It considers the fouling removal
rate for threshold condition ur is proportional to shear stress at the
wall:

ur C  sw ;

15

where C is a proportionality coefficient, 1/(Pa s).


The wall shear stress in the channels of PHE is determined as:

sw fs  w  q  w2 =8;

16

where the friction factor f is calculated according to Eq. (2); the


share of friction losses w has been estimated using Eq. (6).

O.P. Arsenyeva et al. / Applied Thermal Engineering 105 (2016) 763773

2.4. The model validation

The rate of fouling thermal resistance Rf at time t is:

dRf t
u d  C  sw :
dt

17

The expression for the deposition term through wall shear


stress [29]:

ud

Am  Pcu  T 2=3
 q2=3  l4=3
s
1 Bm  P cu  2  sw  q4=3  l1=3  T s2=3  expE=R  T s

18

where

Pcu

1
11:75

2  sw

9:38
 q0:75

0:25

1
1:75

769

19

In this equations Ts is the surface temperature, K; q is the fluid


density, kg/m3; l is the fluid dynamic viscosity, Pa s; R is the universal gas constant equal to 8.314 J/(mol K). In the crude oil study
reported by Yang and Crittenden [27], the parameter values were
found to be E = 52,100 J/mol, Am = 7.93  1010 kg2/3 K1/3 m5/3
(kW)1 s1/3 h1
and
Bm = 1.8  105 m13/3 kg2/3 s8/3 K2/3.
5
2
1
1 1
C = 1.60  10 m K (kW) Pa h . The threshold fouling conditions can be determined at condition when fouling rate calculated
by Eq. (17) is equal to zero. With this assumption, accounting for
Eqs. (18) and (19), the nonlinear algebraic Equation is obtained
determining the link between the temperature of the heat transfer
surface and shear stress corresponding to threshold conditions. The
results of its solution are presented in Fig. 6 by curve, above which
the fouling will start to develop. At conditions with shear stress
and temperature below this curve the fouling is not starting and
heat transfer surface remains clean during all operation period.
At conditions above threshold, as it is reported by Tamakloe
et al. [19] for the processes of fouling deposit in Compabloc heat
exchangers and by Klemes et al. [5] for PHEs, the higher shear
stress values, the less fouling deposit layer formed for the same
period of time. Thus, under the modelling of the PHE and its design,
it is needed to estimate the shear stress value. After using it to
determine the time limits for the allowable operating conditions
of the unit, before the fouling deposit thermal resistance will rise
to critical value. The critical value can be evaluated by the estimation of some percent of losses of heat transfer efficiency comparing
to the clean condition. As the shear stress depends of the geometry
of the channels, it is possible to find the geometric parameters of
the plates corrugations in PHE with the optimal value of shear
stress.

Fig. 6. The curve of threshold fouling conditions for temperature of fouling surface
Ts and shear stress sw. The points are corresponding to the conditions in heat
exchangers considered in a case study: squares 16 are for Compablock; dots 16
are for PSHE.

The validity of the correlations for calculation of friction factor


(Eq. (2)) and heat transfer coefficients (Eqs. (4)(6)) was confirmed
by comparison with experimental data for the main corrugated
field of the channels formed by plates with different geometrical
parameters of corrugations published by a number of researchers.
These equations account for the main parameters that characterise
the geometrical form of the channel, but they are not sensitive to
the scale of the channel in the range of channel spacing from
1.12 to 10 mm [33]. The range of geometrical parameters for plates
corrugations [33] is: the corrugations angle b from 14 to 65; corrugations aspect ratio c from 0.5 to 1.5; the surface area enlargement factor Fx from 1.14 to 1.5. This applies for Reynolds
numbers from 100 to 25,000 and Prandtl numbers from 0.7 up to
1000 [34]. It was concluded that the mean-squire error of correlation for experimental data on friction factor is 9% [37] and on heat
transfer is 6.5% [33]. These ranges of corrugations geometrical
parameters, as well as Reynolds and Prandtl numbers are including
their possible variations when calculating PHEs for process conditions in the research considered in this paper.
The validation of the thermal and hydraulic design for clean
Compabloc PHE and it with fixed fouling thermal resistance Rfoul
in Eq. (1) is performed by comparison of calculated heat transfer
surface area with the data for the same process conditions of 24
different quotations by Alfa-Laval. The discrepancies were not
exceeding 10%. This design method was also used in developed
software and calculation of welded PHEs with square plates of
similar to Compabloc type produced by Pavlogradchemmash factory in Ukraine [5], confirmed by testing in industry, as described
in [35].
The design of PSHE is made by the same principle as for design
of frame-and-plate HEs which validity was checked with a number
of industrial applications [17]. The method of calculating total
pressure loss and heat transfer in PHE channels by dividing it on
the main corrugated field and distribution zones at the inlet and
outlet of the channel is discussed and its accuracy shown in [20].
It proved useful for modelling of PHEs in HENs to estimate
enhanced heat transfer area targets [38].
To validate the model of fouling behaviour with a time the
results of calculations were compared with predictions of fouling
in Compablock HE presented by Tamakloe et al. [19]. The calculations are made for the same flow rates and temperatures 320 C of
hot residue inlet and 235 C cold crude oil inlet and 280 C crude
oil outlet temperature in clean HE. The physical properties of crude
oil and residue are calculated by equations presented in paper [19].
The wall shear stress by data in compared paper is equal to 86 Pa.
But in that case there is no distinction between pressure loss due to
friction forces at the wall and due to drag forces due to change of
flow direction and separation. In our calculations this shear stress
is multiplied on the share of pressure loss due to friction on the
wall in total loss of pressure w calculated by Eq. (6), the accuracy
of which was confirmed also by CFD modelling [39]. Resulting
shear stress is equal to 64 Pa. The calculated time dependence of
crude oil outlet temperature is presented in Fig. 7 by dashed curve.
The discrepancy with solid line, representing the data of Tamakloe
et al. [19] is not more than 2 C at time about 2 years. Counting that
in paper [19] another fouling model is used, and that the error in
fouling experiments reaches up to 50% (see [19,27]), it is fairly reasonable accuracy. The fouling model can be used to estimate the
differences in welded PHEs time behaviour at the same process
conditions.
The proposed approach is applied, when designing PHEs for
crude oil preheat train. In a case study the two types of welded
PHEs are observed for the application in these duties accounting
for the fouling formation.

770

O.P. Arsenyeva et al. / Applied Thermal Engineering 105 (2016) 763773


Table 2
Physical properties of heat exchanging streams.

Fig. 7. The development in time of outlet crude oil temperature for example in
paper [19]: 1 according to [19]; 2 calculated in this study.

3. Case study

Heat carrier

t, C

Cp, kJ/(kg K)

k, W/(m K)

l  103, Pa s

Crude Oil
Crude Oil
Product 1
Product 2
Residue 2
Product 3
Product 4

250
140
170
160
320
270
200

3.081
2.480
2.9
2.644
3.28
3.15
2.81

0.0605
0.1118
0.0714
0.09727
0.064
0.0565
0.0854

0.2898
0.7728
0.04443
0.1107
0.226
0.0457
0.068

even four times. If to compare the flow rates for the cold and hot
sides from Table 1, the position 5 has the bigger one for the hot
side, and others for the cold side. The temperature of heat carriers
vary from 81.4 C to 346 C. It enables to make comparison of HEs
in a wide range of process conditions, that makes it interesting also
for other applications in industry.

3.1. Process conditions

3.2. The design of Compabloc PHEs

The required process conditions in heat exchangers are taken on


example of oil preheat train at a refinery currently operating with
shell-and-tube HEs. There complex flowsheet with streams splitting including totally 23 positions for HEs of which six mostly representative and suspected for fouling are taken for this case study.
The required flow rates and temperature programs for these HEs
are presented in Table 1. The design is performed for pressure drop
1 bar at each side of HE. The refinery can operate with four different sorts of crude oils and for comprehensive HEs design the operation with all of them should be analysed. In a case study only
calculations for one sort of light crude oil are presented. The
required for design properties of this crude oil are approximated
by following relations for its density q2, kg/m3; dynamic viscosity
l2, Pa s; heat conductivity k2, W/(m K); specific heat capacity cp2, J/
(kg K).

The modelling of fouling in Compabloc heat exchanger is


described by Tamakloe et al. [19]. It was pointed out, that design
value for the shear stress in shell-and-tube heat exchangers is
maximally equal to 20 Pa, but in practice due to the pressure drop,
they operate typically with 1015 Pa. In Compablock PHEs the
shear stress can reach 100 Pa and more. That work is concentrated
on analysis of shear stress and its influence on fouling in heat
exchanger of different predetermined passes arrangement, which
can be considered as rating of HE with known construction parameters. The design of PHE for the required process conditions was
not considered.
The approach for Compablock design takes into account the
multi-pass arrangement of the separate blocks inside the heat
exchanger, with mathematical model described above in Section 2
and presented by Eqs. (1)(19). The flow between the corrugated
plates of plate heat exchanger has the complex structure, which
is determined by the geometry of the plates. The wall shear stress
is determined according to Eq. (16). The parameters of plates and
inter plate channels used in our calculations for some Compablock
PHEs are presented in Table 3. The design was made for parameters
of streams and their properties presented in Tables 1 and 2. The
allowable pressure drop for both streams at all heat exchangers
was taken equal to 100 kPa, the material of plates is stainless steel
AISI 316. The results of calculations are presented in Table 4.
The obtained results for shear stress in channels showed that
three heat exchangers on positions 1, 2 and 3 are situated at the
safe area, fairly below the threshold fouling conditions. They can
work clean, without fouling deposit in all time of operation. For
heat exchangers at positions 4, 5 and 6 the results are situated
above the curve for threshold fouling. For this HEs the estimation
of the performance deterioration due to fouling can be made using

q2 783  1  0:0011  T  293:15


l2 q2  106  fexpexp19:46  3:3  ln T  0:8g

20
21

k2 0:168  0:000466  T  293:15

22

cp2 1825 5:46  T  293:15

23

The properties of the oil distillation products are approximated


according to refinery data at temperatures corresponding to HE
positions. Some representative data of oil and distillation products
properties are presented in Table 2.
The process conditions show the significant difference in flow
rates for cold and hot streams, which in some positions reaches

Table 1
Process conditions.
HE #

Side

Heat carrier

Flowrate, t/h

tin, C

tout, C

Heat load
of HE, kW

Cold
Hot

Crude oil
Product 1

107.339
37

150
177.4

157
155.6

637.3

Cold
Hot

Crude oil
Product 2

107.339
39.2

81.4
228.3

117.5
145.4

2457

Cold
Hot

Crude oil
Residue 1

107.339
26.327

161.2
275

178.5
211

1312

Cold
Hot

Crude Oil
Product 3

82.422
102.426

224.6
289.9

250.5
270.2

1753

Cold
Hot

Crude Oil
Product 4

120.755
102.426

224.6
289.9

247.5
264.4

2259

Cold
Hot

Crude Oil
Residue 2

82.422
24.372

269.3
346

284
300

840.5

Table 3
The parameters of plates used in the design of PHEs (approximated according authors
data).
Parameter

Plate length, mm
Plate width, mm
Angle b,
Pitch S, mm
Spacing h, mm
Thickness d, mm
Plate area, m2
ZDZ

Type of Compablock PHEs


CP15

CPL30

CPK40

141
141
45
13
4
1
0.023
9

300
300
45
15
4.95
1
0.112
9

368
368
45
15
5
1
0.168
9

771

O.P. Arsenyeva et al. / Applied Thermal Engineering 105 (2016) 763773


Table 4
The results of Compabloc HE design.
HE no.

Heat transfer
area, m2

Velocity, m/s

Shear stress,
Pa

Hot

Cold

Hot

Cold

1
2
3
4
5
6

26.57
14.78
9.18
26.57
33.13
20.01

0.966
0.764
1.256
0.742
0.890
0.578

1.19
1.294
1.502
1.323
1.236
1.492

30.6
21.5
70.5
19.4
23.7
11.5

60.7
79.9
93.4
59.8
53.0
69.9

Passes,
Cold/Hot

4/8
3/4
2/6
5/2
4/3
4/5

Eqs. (17)(19) and data on shear stress from thermalhydraulic


design for clean HE.
The calculated development in time of the crude oil temperature at the outlet of HE6 is presented in Fig. 8 by dashed curve
(2). The 40% drop in thermal performance (to outlet temperature
278 C) will occur in 190 days. For HE4 (dashed curve in Fig. 9)
the 40% drop in thermal performance (to outlet temperature
240 C) will happen in about 730 or 2 years, that is the good performance in such conditions. Similar situation is for HE5. The predicted performance of all other HEs (1, 2 and 3), operating at
lower temperatures but with higher enough shear stress, is not
deteriorating with time.
3.3. The design of Plate-and-Shell HEs
The possibility to use PSHE heat exchanger in crude oil preheat
train is analysed. These heat exchangers have welded plates pack
placed in shell, what provides high pressure ratings (Fig. 2). It
has alternating channels for hot and cold media, and can operate
with counter-current or co-current flows with single or multipass arrangement of heat carriers. The plate pack has a welded
construction and consists of the rounded corrugated plates assembled together in one unit. The general form of the circle plate is
presented in Fig. 3a. Generally the plates with the same corrugation are welded together to form the channels for heat carriers
movement, presented in Fig. 3b. The different flowrates for oil
and product affects the flow velocities on hot and cold sides. As
it was discussed in Section 3.2, in PHE with parallel flow the
single-pass arrangement with counter flow is preferable compare
to multi-pass arrangement. Besides, the multi pass arrangement
on hot side of HE with single pass for crude oil is leading to different temperature programs at the passes. The part of the crude oil
stream flowing through channels close to hot stream inlet will be
overheated compare to required average temperature at crude oil

Fig. 9. The calculated outlet temperature change with time for HE at position #4:
(1) PSHE; (2) Compabloc HE.

outlet and more subjected to fouling. In a present study only single


pass arrangement for both streams is considered.
Initially the optimal geometrical parameters of plate corrugations were selected with objective function as the total heat transfer area of all HEs in preheat train in clean conditions with the use
of the methodology described in [20]. The constraints were the
maximal heat transfer area of one plate (1 m2) and the number
of plates in one HE (300). As the result the corrugation inclination
angle is taken equal to 65, the corrugation height equal to 2 mm
and corrugation pitch 6.5 mm. The results of design of single pass
counter flow PSHE with these parameters of plate corrugation are
presented in Table 5. The calculations were made for material of
plate AISI 316 with its thickness equal to 1 mm. The coefficient
of local hydraulic resistance in distribution zones of the plate
was taken fDZ = 38, as for plate-and-frame heat exchanger analysed
in paper [20]. The allowable pressure drop was taken the same as
for design of Compabloc PHE equal to 1 bar.
For the considered PSHEs the threshold fouling conditions are
checked by comparison the positions of corresponding temperatureshear stress points on a graph in Fig. 6 with threshold fouling
curve. The calculated time development of outlet crude oil temperature for PSHEs on positions suspected to fouling (HE6 and HE4) is
presented on graphs in Figs. 8 and 9.
3.4. The comparison of Compabloc and PSHE types of welded PHEs
As it can be judged from results in Tables 4 and 5, the required
heat transfer areas of PSHE in most cases are somewhat bigger
than that for Compabloc PHE. Only at positions 4 and 5, where
hot and cold streams flow rates are close each other on magnitude,
PSHE has about 7% smaller heat transfer area. It can be explained
by two major points.
(1) Drawback of not using unsymmetrical passes arrangement
in PSHE with significantly different flow rates of streams at
the same allowable pressure drops. To overcome this, the

Table 5
The results of PSHE design.
HE no.

Fig. 8. The calculated outlet temperature change with time for HE at position #6:
(1) PSHE; (2) Compabloc HE.

1
2
3
4
5
6

Heat transfer area, m2

41.37
20.07
27.60
24.82
31.62
24.92

Velocity, m/s

Shear stress, Pa

Hot

Cold

Hot

Cold

0.270
0.420
0.229
1.210
0.930
0.289

0.808
0.922
0.866
0.871
1.005
0.916

7.01
14.13
8.47
87.01
55.67
12.52

120.98
150.96
135.91
89.83
115.37
97.88

772

O.P. Arsenyeva et al. / Applied Thermal Engineering 105 (2016) 763773

construction of plate with different cross section of arranged


channels on both sides can be used, as is proposed in paper
[40] for condensation process. Such option is not considered
in present study. When flow rates of streams are similar on
magnitude, the required heat transfer area of clean PSHE can
be smaller than that of Compabloc. It is mainly due to smaller hydraulic diameter.
(2) The additional hydraulic resistance at Compabloc in flow
distribution zones at inlet and outlet of the channels is smaller than in PSHE. As it is shown in Fig. 1, the stream is entering Compabloc channel through the full its width, no
additional change of cross section area required. In PSHE
with parallel flow of streams (see Fig. 2) part of the channel
width is blocked to arrange at the same side of HE the outlet
(or inlet) of opposed stream. It creates additional hydraulic
resistance at channel inlet and exit.
The advantages of PSHE construction is the possibility to use
pure counter current flow of streams, the channels with smaller
hydraulic diameter and with optimal corrugations angle. In Compabloc HE the change of corrugation angle at one side leads to
the opposite change of corrugation angle at another side, so the
angle of 45 seem as the best solution. However the good flow distribution and possibility to arrange multipass arrangement with
overall counter flow of streams makes this type of heat exchanger
relatively easy adopted to required process conditions.
The analysis of data in Tables 4 and 5 reveals that at the same
process conditions the wall shear stress in PSHE is considerably
higher than in Compabloc PHE. Considered PSHE has about two
times smaller plates spacing and hydraulic diameter of channels
than Compabloc. The friction factors in channels with corrugation
angle 65 is also much higher than at 45 corrugation angle. To satisfy the same allowable pressure drop such PSHE channel should
have much smaller length that allow to do it in one pass with
increase of the wall shear stress, as is explained with more details
in paper [20]. The higher level of shear stress can significantly mitigate the fouling formation. As it is seen at graph in Fig. 6, the most
points corresponding to temperatureshear stress conditions in
PSHE are situated below the fouling threshold curve. This indicates
that the fouling in PSHE at positions 15 even will not start or will
be very low. The point for the position 6 is also much closer to
threshold curve than that for Compabloc at the same position.
The calculated fouling time behaviour of PSHE shown in Fig. 8 indicates that the fouling tendency is about two times lower than for
Compabloc HE on this position. The outlet temperature of the
crude oil is dropping to 278 C in about 1 year compare to half a
year in Compabloc HE. There is even greater difference for HE4
position, as shown in Fig. 9. The drop in crude oil outlet temperature in PSHE is about 1 C in a 5 years, when Compabloc require
cleaning in a two years with 40% drop of heat load. As it is discussed in Section 1, the properties of different crude oils can vary
considerably, as well as their fouling propensities [41]. So the calculations of HEs must be made with accounting for possibility to
use different crude oils. The presented comparison can be regarded
as qualitative one that allows to conclude about the benefits of
using PSHE in crude oil preheat trains, where they can be competitive with Compabloc type PHEs.

4. Conclusions
The applications of compact heat exchangers for crude oil
preheat train duties increases heat transfer efficiency and energy
saving potential of industrial sites. The construction of welded
PHEs allows significant increase of their application range on
temperature (up to 400 C) and pressure (to 40 bar and more).

For the crude oil preheat train the welded PHEs of Compabloc type
and Plate-and-Shell type can be designed and introduced instead of
Shell-and-Tube HEs, that allows to decrease the heat transfer surface areas and mitigate fouling in heat transfer equipment.
The proposed mathematical models of welded PHEs enable to
design these HEs at different process conditions with accounting
for the influence of plates corrugations geometry on HE thermal
and hydraulic performance. The obtained results of the design
show that the application of PSHE units can decrease the fouling
formation on the heat transfer surface, as it has single-pass
arrangement of flow movement and higher wall shear stress level.
The approach of designing PSHE with variable cross section area for
cold and hot sides can significantly improve their heat and hydraulic performance for the processes with significant differences in
flow rates of cold and hot media. The correct calculation of PSHE
and optimisation of plates geometry requires the use of mathematical model and correlations to predict the heat transfer and friction
factors, which are presented.
The correct design of welded PHEs in crude oil preheat train
requires to be made with accounting for physical properties and
fouling propensity of different crude oils expected for the use in
specific refinery and their temperature dependence. It is also
requires the reliable data of fouling monitoring in PHEs for adequate estimation of empirical constants in fouling models.
Acknowledgements
The support of EC Project EFENIS (contract No ENER-FP7296003) is sincerely acknowledged.
References
[1] A. Szklo, R. Schaeffer, Fuel specification, energy consumption and CO2 emission
in oil refineries, Energy 32 (7) (2007) 10751092.
[2] M.A. Waheed, A.O. Oni, Performance improvement of a crude oil distillation
unit, Appl. Therm. Eng. 75 (2015) 315324.
[3] N. Jiang, J.D. Shelley, R. Smith, New models for conventional and heat
exchangers enhanced with tube inserts for heat exchanger network retrofit,
Appl. Therm. Eng. 70 (1) (2014) 944956.
[4] M. Pan, I. Bulatov, R. Smith, Improving heat recovery in retrofitting heat
exchanger networks with heat transfer intensification, pressure drop
constraint and fouling mitigation, Appl. Energy 161 (2016) 611626.
[5] J.J. Klemes, O. Arsenyeva, P. Kapustenko, L. Tovazhnyanskyy, Compact Heat
Exchangers for Energy Transfer Intensification: Low Grade Heat and Fouling
Mitigation, CRC Press, Boca Raton, USA, 2015.
[6] B.D. Crittenden, M. Yang, L. Dong, R. Hanson, J. Jones, K. Kundu, J. Harris, O.
Klochok, O. Arsenyeva, P. Kapustenko, Crystallization fouling with enhanced
heat transfer surfaces, Heat Transfer Eng. 36 (78) (2015) 741749.
[7] M.M. Abu-Khader, Plate heat exchangers: recent advances, Renew. Sustain.
Energy Rev. 16 (4) (2012) 18831891.
[8] L. Wang, B. Sunden, R.M. Manglik, PHEs. Design, Applications and Performance,
WIT Press, Southhampton, UK, 2007.
[9] H. Hajabdollahi, M. Naderi, S. Adimi, A comparative study on the shell and tube
and gasket-plate heat exchangers: the economic viewpoint, Appl. Therm. Eng.
92 (2016) 271282.
[10] O.Y. Perevertaylenko, A.O. Gariev, T. Damartzis, L.L. Tovazhnyanskyy, P.O.
Kapustenko, O.P. Arsenyeva, Searches of cost effective ways for amine
absorption unit design in CO2 post-combustion capture process, Energy 90
(2015) 105112.
[11] AlfaLaval. Welded plate-and-block heat exchangers. (2016) Available online
at:
<http://www.alfalaval.com/
products/heat-transfer/plate-heatexchangers/Welded-plate-and-block-heat-exchangers/>
(Last
visited
31.01.2016).
[12] E. Andersson, J. Quah, G.T. Polley, Experience in application of Compabloc heat
exchangers in refinery pre-heat trains, in: H. Muller-steinhagen, M.R.
Malayeri, A.P. Watkinson (Eds.), Proceedings of the International Conference
on Heat Exchanger Fouling and Cleaning VIII-2009, Schladming, Austria, 2009,
pp. 3943.
[13] Vahterus Oy. Plate & Shell Heat Exchanger, 2016. Available online at:
<www.vahterus.com/en/node/528> (Last visited 31.01.2016).
[14] L.O. Freire, D.A.d. Andrade, On applicability of plate and shell heat exchangers
for steam generation in naval PWR, Nucl. Eng. Des. 280 (2014) 619627.
[15] T. Nakaoka, H. Uehara, Performance test of a shell-and-plate-type condenser
for OTEC, Exp. Thermal Fluid Sci. 1 (1988) 275281.
[16] R.K. Shah, D.P. Sekulic, Fundamentals of Heat Exchanger Design, John Wiley &
Sons, New Jersey, USA, 2003.

O.P. Arsenyeva et al. / Applied Thermal Engineering 105 (2016) 763773


[17] O.P. Arsenyeva, L.L. Tovazhnyansky, P.O. Kapustenko, G.L. Khavin, Optimal
design of plate-and-frame heat exchangers for efficient heat recovery in
process industries, Energy 36 (8) (2011) 45884598.
[18] O. Arsenyeva, L. Tovazhnyansky, P. Kapustenko, G. Khavin, Mathematical
modelling and optimal design of plate-and-frame heat exchangers, Chem. Eng.
Trans. 18 (2009) 791796.
[19] E.K. Tamakloe, G.T. Polley, M. Picn-Nez, Design of Compabloc exchangers
to mitigate refinery fouling, Appl. Therm. Eng. 60 (2013) 441448.
[20] O. Arsenyeva, P. Kapustenko, L. Tovazhnyanskyy, G. Khavin, The influence of
plate corrugations geometry on plate heat exchanger performance in specified
process conditions, Energy 57 (2013) 201207.
[21] J. Aminian, S. Shahhosseini, Neuro-based formulation to predict fouling
threshold in crude preheaters, Int. Commun. Heat Mass Transfer 36 (2009)
525531.
[22] B.C.G. Assis, J.C. Lemos, E.M. Queiroz, F.L.P. Pessoa, F.S. Liporace, S.G. Oliveira, A.
L.H. Costa, Optimal allocation of cleanings in heat exchanger networks, Appl.
Therm. Eng. 58 (2013) 605614.
[23] W.A. Ebert, C.B. Panchal, Analysis of Exxon crude slip stream coking data, in: C.
B. Panchal et al. (Eds.), Fouling Mitigation of Industrial Heat-Exchange
Equipment, Begell House, New York, USA, 1997, pp. 451460.
[24] G.T. Polley, D.I. Wilson, B.L. Yeap, S.J. Pugh, Evaluation of laboratory crude oil
threshold fouling data for application to refinery preheat trains, Appl. Therm.
Eng. 22 (2002) 777788.
[25] M.R. Mozdianfard, E. Behranvand, Fouling at post desalter and preflash drum
heat exchangers of CDU preheat train, Appl. Therm. Eng. 89 (2015) 783794.
[26] G.T. Polley, E. Tamakloe, M. Picn Nuez, Models for Chemical Reaction
Fouling, in: 11th AIChE Spring Meeting, vol. Abstract No. 94c, Chicago, IL, USA,
2011.
[27] M. Yang, B. Crittenden, Fouling thresholds in bare tubes and tubes fitted with
inserts, Appl. Energy 89 (2012) 6773.
[28] B.L. Yeap, D.I. Wilson, G.T. Polley, S.J. Pugh, Mitigation of crude oil refinery heat
exchanger fouling through retrofits based on thermo-hydraulic fouling
models, TransIChemE 82 (2004) 5371.
[29] O.P. Arsenyeva, B. Crittenden, M. Yang, P.O. Kapustenko, Accounting for the
thermal resistance of cooling water fouling in plate heat exchangers, Appl.
Therm. Eng. 61 (2013) 5359.

773

[30] J.G. Speight, The Chemistry and Technology of Petroleum, CRC Press, Boca
Raton, USA, 2014.
[31] S.W. Hasan, M.T. Ghannam, N. Esmail, Heavy crude oil viscosity reduction and
rheology for pipeline transportation, Fuel 89 (5) (2010) 10951100.
[32] D.E. Djemiat, A. Safri, A. Benmounah, B. Safi, Rheological behavior of an
Algerian crude oil containing sodium dodecyl benzene sulfonate (SDBS) as a
surfactant: flow test and study in dynamic mode, J. Petrol. Sci. Eng. 133 (2015)
184191.
[33] O.P. Arsenyeva, L.L. Tovazhnyanskyy, P.O. Kapustenko, O.V. Demirskiy, Heat
transfer and friction factor in criss-cross flow channels of plate-and-frame heat
exchangers, Theor. Found. Chem. Eng. 46 (6) (2012) 634641.
[34] O.P. Arsenyeva, L.L. Tovazhnyanskyy, P.O. Kapustenko, O.V. Demirskiy,
Generalised semi-empirical correlation for heat transfer in channels of plate
heat exchanger, Appl. Therm. Eng. 70 (2) (2014) 12081215.
[35] L.L. Tovazhnyansky. The principles of heat transfer processes intensification,
development and optimization of new types of plate heat exchangers for
chemical industry. (Principy intensifikacii processov teploobmena. Razrabotka
I optimizaciya novyh tipov plastinchatyh teploobmennikov dlya himicheskih
proizvodstv.) The DSc thesis, NTU KhPI, Kharkiv, Ukraine, 1988 (in Russian).
[36] F. Coletti, G.F. Hewitt (Eds.), Crude Oil Fouling: Deposit Characterization,
Measurement, and Modeling, Elsevier Inc., London, UK, 2015.
[37] O. Arsenyeva, L. Tovazhnyansky, P. Kapustenko, G. Khavin, The generalized
correlation for friction factor in crisscross flow channels of plate heat
exchangers, Chem. Eng. Trans. 25 (2011) 399404.
[38] O.P. Arsenyeva, R. Smith, I. Bulatov, L. Tovazhnyanskyy, O. Kapustenko, G.
Khavin, Estimation of enhanced heat transfer area targets in process
industries, Computer Aided Chem. Eng. 32 (2013) 355360.
[39] I.A. Stogiannis, S.V. Paras, O.P. Arsenyeva, P.O. Kapustenko, CFD modelling of
hydrodynamics and heat transfer in channels of a PHE, Chem. Eng. Trans. 35
(2013) 12851290.
[40] L. Tovazhnyansky, P. Kapustenko, O. Perevertaylenko, G. Khavin, O. Arsenyeva,
Investigation of the new corrugation pattern for low pressure plate
condensers, Appl. Therm. Eng. 31 (2011) 21462152.
[41] T.C. Ho, A study of crude oil fouling propensity, Int. J. Heat Mass Transf. 95
(2016) 6268.