Heat Transfer Theory

Objectives
• Understanding the principles of Heat Transfer
– Liquid-to-liquid

– Steam-to-liquid

• Understanding how CASv5 works

– The calculations behind the results

– To evaluate the suggested designs from CASv5

• Confident in discussing heat transfer with our customers

Topics to cover
• Modes of heat transfer
• Flow principles
• The heat balance
• The heat transfer equation
• Thermal length concept
• Pressure drop
• Heat and friction losses
• Physical properties
• Shear stress
• Design safety factors
• Fouling
Modes of heat transfer
• Law of physics
– Heat = Energy

– If you take a hot spot

… and a cold spot

… the heat will always be transferred from the hot to the cold
 Three ways to transfer heat
• Radiation Reflected
– Electromagnetic waves
– When it reaches a body it has 3 options: Absorbed

Transmitted
• Conduction
– Molecular or atomic vibrations
– No material transport

• Convection
– Energy is transferred by the motion and
intermixing of small mass elements
– Natural convection caused by density difference
– Forced convection is man-made (ex., pump)
Example, a day at the beach

Radiation
Convection

Conduction
 Which of these three are
important in heat exchangers?

• Radiation? - Negligible
• Conduction? - Interesting!
• Convection? - Yes! The most effective way of heat transfer
Flow principles
• Two types of flow
– Laminar

Conduction

Flow profile Velocity profile

• Orderly flow throughout the fluid

• Parabolic flow profile
– Fluid at the wall moves slower
– Due to the friction from the wall surface
• Ex., viscous fluids or water at low velocity
• How is heat transferred through the above pipe?
 Flow principles
• Two types of flow Conduction
– Turbulent

Convection

Flow profile Velocity profile

• No orderly flow
• Random eddy motion mixes the fluid
• Always a laminar film closest to the wall
• Ex., water at higher velocity
• How is heat transferred in the middle and at the wall?
Flow principles
• Turbulent flow  Convection  Better heat transfer
• How does these parameters affect the laminar film?
– Velocity?
Higher velocity  Thinner laminar film
– Viscosity?
High viscosity  Thicker laminar film
Type of plate, which we will get back to...
 Heat transferred in a HE
The temperature profile at one point of the plate wall
Wall
T1, Bulk temperature on hot side

Hot side

Flow direction T3
T4
Flow direction
Cold side

T2, Bulk temperature on cold side

Heat transfer (Q) driven by

temperature difference

More turbulence  Thinner laminar film

 Less temperature drop  Better Heat Transfer
 The Heat Balance
• Liquid-to-liquid
Cold fluid out at T2 Out Definitions
Mass flowrate m2
Cold fluid in at T2 In Q = Heat load, W
Q (rate of heat transfer)
Mass flowrate m2 He
at Hot fluid in at T1 In
Tr m = Mass flow rate, kg/s
an
s fe Mass flowrate m1
r Cp = Specific heat, J/kg°C
(the energy needed
Hot fluid out at T1 Out to heat 1 kg of the fluid
Mass flowrate m1 with 1°C)

Heat released by the hot fluid: Q1=m1*Cp1*(T1 In-T1 Out)

Heat absorbed by the cold fluid: Q2=m2*Cp2*(T2 Out -T2 In)

Heat losses are negligible  Q1= Q 2

 The Heat Balance - calculation
• What is the cold fluid outlet temperature?
T2 Out= X
T2 In= 20°C
m2 = 120 kg/s
Q
Cp2 = 4,2 kJ/(kg °C) He
at T1 In = 80°C
Tr
an
s fe m1 = 100kg/s
r

T1 Out = 40°C
Cp1 = 4,0 kJ/(kg °C)

Heat Load: Q1 = m1*Cp1*(T1 In-T1 Out) = 100 kg/s * 4.0 kJ/(kg °C) * (80-40)°C
Q1 = 16 000 kJ/s = 16 000 kW
Q1= Q 2: 16 000 kW = Q2=m2*Cp2*(T2 Out -T2 In)
16 000 kW = 120 kg/s * 4.2 kJ/(kg °C) * (X-20)°C
X = 16 000 / (120 * 4.2) + 20 = 52°C
 The Heat Balance
• Steam-to-liquid
Everyone knows what steam is?
- The three three phases of water:
Solid (s) Liquid (l) Gas (g)
Ice, Snow Liquid water Steam = water vapour
Crystals, Molecules move Molecules leaves
all molecules apart the liquid and vaporise
tight together

Melting Heat absorbed by Heat of vaporisation /

energy the water (Cp) Heat of condensation
 The Heat Balance
• Steam described in diagrams
Enthalpy-Temperature Temperature-Pressure
Enthalpy Temperature
(°C)
Superheated
eated vapour
Saturated uperh
S
Saturated steam
120°C
Hvap
Liquid
id
Liqu
T
Cp*

Temperature 2 bar Pressure

Tboil (°C) (bar)

Hvap = Heat needed to vaporise 1 kg of a fluid (kJ/kg)

The same amount of energy is released during condensation
 The Heat Balance
• Steam-to-liquid
Cold fluid out at T2 Out
Mass flowrate m2
Cold fluid in at T2 In
Q
Mass flowrate m2 He
at Steam (g) at T1 In
Tr
an Mass flowrate m1
s fe
r Hvap (kJ/kg)

Condensate (l) out

Heat released by the steam: Q1=m1* Hvap

Heat absorbed by the cold fluid: Q2=m2*Cp2*(T2 Out -T2 In)

Heat losses are negligible  Q1= Q 2

 The Heat Balance - calculation
• How much steam is needed to heat the water?
Water out at 90°C

Water in at 60°C Q
Mass flowrate 50 kg/s He
at Steam (g) at 120°C
Cp2 = 4,2 kJ/(kg °C) Tr 2 bar (1 bar gauge)
an
s fe Mass flowrate X
r Hvap = 2200 kJ/kg

Condensate (l) out

Heat Load: Q2 = m2*Cp2*(T2 Out -T2 In) = 50 kg/s * 4.2 kJ/(kg °C) * (90-60)°C
Q2 = 6 300 kW
Q 2= Q 1: 6300 kW = Q1 = m1* Hvap
6300 kW = X kg/s * 2200 kJ/kg
X = 2.86 kg/s
 The Heat Balance
• So why is steam used?
– Less than 3 kg steam can heat 50 kg water from 90 to 60°C
– It is a high-energy carrier !!!

• Steam is the most common way in industry to

distribute energy for heating purposes

• Steam in CAS 1-phase programme

– Only pure and saturated water vapour (steam)
– Total condensation and no sub-cooling of condensate
– No inert or non-condensable present in the steam !!
– Inerts are for example Nitrogen and Oxygen
– If not the above  2-phase programme
The Heat Transfer equation
Q = k * A * LMTD
• Q = Heat Load, W (same as Q1 = Q2 before)
• k = k-value, overall heat transfer coefficient (OHTC), W/m²°C
– Higher k-value = More efficient heat transfer

– Described later on how this value is calculated

• A = heat transfer area (m²) - our goal is to minimise this !!!

• LMTD= Logarithmic mean temperature difference
– The driving force of heat transfer is temperature difference

– LMTD describes the temperature profile in the HE

– How many knows what the LMTD is and can calculate it?
 Q = k * A * LMTD
• LMTD= Logarithmic mean temperature difference
– Depend on counter-current or co-current flow

Counter-Current Flow Co-Current Flow

T1 in T1 in
1
T2 out 1 T1 out
T1 out 2
2 T2 out
T2 in
T2 in
Area Area

1   2
LMTD 
 1 
ln  

  2
 Q = k * A * LMTD - calculation
• What is the LMTD for the two cases below?
Counter-Current Flow Co-Current Flow
90°C 90 °C
1
40 °C 1 45°C
45°C 2
2 40 °C
20 °C
20 °C
Area Area

LMTD = (50-25) / ln(50/25) 1   2 LMTD = (70-5) / ln(70/5)

LMTD 
= 25 / ln 2 = 36.1°C  1  = 65 / ln 14 = 24.6°C
ln  

  2

Counter-current flow gives a higher LMTD

Q = k * A * LMTD
• LMTD and counter-current flow
– Counter-current flow always gives a higher LMTD for liq/liq.
• Remember the Q = k * A * LMTD

• For the same heat load (Q) and the same k-value...

• A higher LMTD…
• Allows a smaller Area  More competitive design !
– Counter-current flow allows temperature cross
• Hot side temperature outlet
is lower than cold side outlet 1

2
 Q = k * A * LMTD
• LMTD and co-current flow
• When do we need to use co-current flow?
Co-Current Flow Counter-Current Flow

1
1 Wall temperature
2
2

Area Area
– To control the wall temperature (more constant in co-current)
– For example,
• Heat sensitive media such as food stuff
• Avoid crystallisation below a certain temperature
• Avoid scaling (precipitation of CaCO3 above certain temperature)
 Q = k * A * LMTD
• LMTD with steam as heating media
Counter-Current Flow Co-Current Flow

1 2

2 1

Area Area

– Pure steam condense at constant temperature

– Same LMTD for co-current and counter-current
 Q = k * A * LMTD - calculation
• Calculate the LMTD and the Area for
co-current and counter-current flows
Fluid 1 Fluid 2
m [kg/s] 100 120
Cp [kJ/kgC] 4,0 4,2
Tin [C] 80 20
Tout [C] 50 ?
k-value [W/m2°C] 6690

Heat Load: Q1 = m1*Cp1*(T1 In -T1 Out) = 100 kg/s * 4.0 kJ/(kg °C) * (80-50)°C
Q1 = 12 000 kW
Q1= Q2: 12 000 kW = Q2 = m2*Cp2*(T2 Out -T2 In)
12 000 kW = 120 kg/s * 4.2 kJ/(kg °C) * (T2 Out -20)°C
T2 out = 12 000 / (120 * 4.2) + 20 = 43.8°C
 Q = k * A * LMTD - calculation
Counter-Current Flow Co-Current Flow
80°C 80 °C
1
43.8 °C 1 50°C
50°C 2
2 43.8 °C
20 °C
20 °C
Area Area

LMTD = (36.2-30) / ln(36.2/30) 1   2 LMTD = (60-6.2) / ln(60/6.2)

LMTD 
= 6.2 / ln 1.2067 = 33°C  1  = 53.8 / ln 9.677 = 23.7°C
ln  
Q = k * A * LMTD

  2 Q = k * A * LMTD
12000 kW = 6.69 kW/m2°C * A * 33°C 12000 = 6.69 * A * 23.7
A = 12000/6.69/33 = 54.3 m2 A = 12000/6.69/23.7 = 75.7 m2
 Q = k * A * LMTD
The k-value consists of 3 different heat transfer resistances
Wall Flow direction
T1, Bulk temperature on hot side
Cold side
Called Film heat transfer Wall conductivity, 
1-value coefficient on hot side
Wall thickness, 
T3 Resistance
Hot side T4 from the wall
Flow direction
Film heat transfer Called
coefficient on cold side 2-value

T2, Bulk temperature on cold side

Heat transfer (Q) driven by

temperature difference
 Q = k * A * LMTD
• The k-value formula
k = Overall heat transfer coefficient, W/m²°C
 = Film heat transfer coefficient, W/m²°C
 = Wall thickness, m
 = Wall conductivity, W/m°C

We get a high k-value when... A high k-value means…

The -values are high meaning: Less area is needed...

– High turbulence For the same heat load
– Thin laminar film
Q = k * A * LMTD
 Less heat transfer resistance
And the plate is thin with high conductivity
Q = k * A * LMTD
• How do we get the k-value?
 is known, it is the plate thickness

 from tables of metal conductivity

• How do we get the -values?

• Thanks to three smart men:
– Mr Nusselt
– Mr Prantl
– Mr Reynolds

• They found three different dimensionless numbers that

describes the heat transfer
Q = k * A * LMTD
• The -value
– Nusselt number
 = film heat transfer coefficient (W/m²°C)
  Dh
Nu   = thermal conductivity of fluid (W/m°C)
 Dh = hydraulic mean diameter (m)

– Describes the heat transfer

– Contains the -value which we need to calculate the k-value
– The hydraulic mean diameter is know
• Depend of the pressing depth of the plate heat exchanger
• Dh for PHE = 2 * the pressing depth
  is also known as we know which fluids we deal with
Q = k * A * LMTD
• The -value
– Prantl number
 = thermal conductivity, W/m°C
Cp  
Pr   = dynamic viscosity, Pas
 Cp = specific heat, J/kg°C

– Describes the fluids

– All parameters are known as we know which the fluids are:
• Thermal conductivity - how easily the fluid conducts heat
• Viscosity - how thick the fluid is (affect the laminar film)
• Specific heat - how much heat is absorbs/released (same as in Q)
Q = k * A * LMTD
• The -value
– Reynolds number

G  Dh Dh = hydraulic mean diameter, m

Re  
 = dynamic viscosity, Pas
G = mass flow rate in each channel, kg/m²s

– Describes the flow conditions

– Dh and  - as before
– G is the mass flow rate in each channel which is calculated by:
• The total flow rate of the fluid
• Divided by the number of channels in the plate heat exchanger
Q = k * A * LMTD
• The -value
– These three numbers are related to each other as follows

y
Nu  Pr  B  Re n

– y is a function of the Prantl number

– B is a function of geometry and Reynolds
– n is a function of geometry and Reynolds
– y, B and n is established through laboratory test on each plate

Higher Re  Higher Nu  High -value  Higher k-value

Q = k * A * LMTD
• Example of Nu, Pr, Re relationship for M15M
Three different channels for each plate
M15-M • High theta (high heat transfer)
• Medium theta (medium)
• Low theta (lower heat transfer)

For the same Re-number

the PHE has a higher Nu Pr
than a Shell & Tube
 High -value
 Higher k-value
Thermal length
• Describes how difficult a duty is thermally
• Two names for the same thing:
– Number of Transfer Units (NTU)
– Theta,  (mainly used in Alfa Laval)

• We use the “Theta” concept in several ways:

– Thermal duty (high / low theta duties)
– Unit (high / low theta PHE models)
– Plates (high / low theta plates)
– Channels (high / medium / low theta channels)
Thermal length
• There is no good or bad Theta
– Just different thetas on the duties
– And different PHE models and plate to choose
– So we fit the duty for the customers duty

• Typically,
– Low theta means  < 1
– High theta means  > 1
Thermal length - duty
• Theta is calculated for the hot and cold side

– “How many times the LMTD that the fluid is cooled/heated”

T1 in
1 Lower 
1   2
T2 out Hig
LMTD 
her T1 out  1 

2
ln  
 2 
T2 in
Area
Thermal length - units
• What characterises a high theta unit (PHE model)?
– Long plate  Longer time for the fluid to be cooled/heated
– Low pressing depth  Less fluid / plate to be cooled/heated
– Both high and low theta units are needed to find good designs
Thermal length - units
• Example of PHEs in high and low theta
Port size Low theta  1 High theta  1
30 mm M3
60 mm M6M M6
100 mm M10M M10B
150 mm M15M M15B
200 mm AK20/M20M A20B
250 mm MX25
300 mm MA30M M30
450 mm A45

M10-B and M10-M are the same size (length and width)
• B means 2.5 mm pressing depth  High 
•M means 4.0 mm pressing depth  Low 

Units in yellow are On Request =

available but not with standard price and delivery
 Thermal length - units
• Also possible to make multi-pass design
– For very high theta duties
– If there is no plate that fits in single pass
– Choose best available unit and make it multi-pass
• Example, 2 pass hot side / 2 pass cold side

Cold out Cold in

Hot in Hot out
Thermal length - plates & channels
• We have two plate corrugations (L and H)
• These form three different channels (L, M and H)

L: Low theta H: High theta

L + L = L channels L + H = M channels H + H = H channels

• We choose between L, M and H channels

• Tailor-make it for the specific duty
 Thermal length - plates & channels
Low turbulence Medium turbulence High turbulence
& pressure drop & pressure drop & pressure drop

L + L = L channels L + H = M channels H + H = H channels

Advantages Benefits
• Efficient heat transfer • Increased heat recovery
• High wall shear stress • Low fouling
• Variable thermal length • Optimal design
• Strong construction • Insensitive to vibration
Thermal length - plates & channels
• Be careful with designs with only H-theta plates
 “Snow-ball effect” when adding a large margin
Margin

Lower channel flow rate

More Area Lower pressure drop

Lower Lower
k-value turbulence

• To avoid this
– Change to plate with higher theta (ex. M10M M10B)
– Apply small margins (10%) when only H-theta plates
Thermal length
• Why do we have different theta units & plates?
• To find good designs for almost all duties !
– Depending on the theta of the duty
– Depending on the allowable Pressure Drop

• Rule of thumb,
– The higher the allowable Pressure Drop
– The higher Theta can be achieved in the thermal duty
– But it requires more Area (higher price)
Pressure drop
• Pressure drop equation
G2  4 f  L 
P     n 
2   Dh 

P = Pressure drop (Pa) between inlet and outlet of PHE

G = Mass flow rate per channel (kg/m² s)
 = Density of the fluid (kg/m3)
Dh = Hydraulic mean diameter (m) (2*pressing depth for PHE)
L = Length of the plate (m)
n = End effects, the pressure drop generated at the inlet and
outlet of the plate
f = Friction factor which is a function of Reynolds number
Pressure drop
• Example, friction factor vs Re number on M15-M
The three different channels gives
different pressure drop - L, M, H
Pressure drop
• Pressure drop is the price paid for high heat transfer
– Usually customer specifies how much pressure drop is allowed
– Normal P in a PHE is about 100 kPa (1 bar)
– Always try to use as much as possible of the allowed P
– Higher P
 Higher turbulence
 Thinner laminar film
 Higher heat transfer
– Higher P gives a higher pumping cost (electricity)
– A balance between less Area and higher electricity cost
 Pressure drop
• Do we only have pressure drop in the channels?
• No, also in the ports and connections
• Rule of thumb: Pports+conn < 30% of Total P

Pressure
Pports+conn = Pinlet+ Poutlet
Pin
PInlet
If >30% of PTotal

P first P last  Maldistribution problem

PTotal channel channel • Fluid in 1st channel rushes
through to quickly

Poutlet • To little fluid goes in the

Pout last channel

Length of port  PHE does not perform!

 Are losses negligible?
• Let’s go back to this example and consider the heat loss
T2 Out= 52°C
T2 In= 20°C
m2 = 120 kg/s
Q
Cp2 = 4,2 kJ/(kg °C) He
at T1 In = 80°C
Tr
an
s fe m1 = 100kg/s
r

T1 Out = 40°C
Cp1 = 4,0 kJ/(kg °C)

Q was calculated to 16 000 kW

LMTD is about 24ºC and a typical k-value is 6000 W/m2 ºC
 Heat transfer area of about 110 m2
 Typically a M20M unit with about 130 plates
Are losses negligible?
– Size of about 1.8 m high * 0.7 m wide and 0.6 mm long
 External surface of about 5.5 m2
– Air surrounds the heat exchanger
 Natural convection heats the air
 Typically gives a k-value of 10 W/m2 ºC

– Assume air has a temperature of 10ºC (outdoor)

– The unit can maximum be 80 ºC on the surface
 Temperature difference (T) maximum 70 ºC

Heat loss < k * A * T = 10 W/m2 ºC * 5.5 m2 * 70 ºC = 3850 W

% of Heat Transferred = 3.85 kW / 16 000 kW = 0.02%
Yes, that is negligible!
 Are losses negligible?
• Energy lost by friction in the PHE is given by:
Qfric = heat loss by frictional dissipation, W
m m = mass flowrate, kg/s
Q fric   P  = density, kg/m3
 P = pressure drop, Pa

– Let’s go back to the example

– Flow rate on Side 1 is 100 kg/s and Side 2 is 120 kg/s
– Assume a density of 1000 kg/m3 (water)
– Assume 100 kPa pressure drop (typical for PHE)
 Qfric side 1 = 100/1000*100 000 = 10 kW
 Qfric side 2 = 120/1000*100 000 = 12 kW
% of Heat Transferred = 22 kW / 16 000 kW = 0.14%
Yes, that is negligible (but considered in CAS)
Physical properties
• Physical properties of the fluids are very important
• Which physical properties are needed for each fluid?
– Density ()
– Specific heat (Cp)
– Thermal conductivity ()
– Viscosity ()

• For , Cp and :
– Changes only +/-10% within normal temperature program
– One value for each media is OK
Physical properties
• For viscosity ()
– Can change +/-500% within a normal temperature program
– Preferably 3 values for each media:
• One at or below the cold inlet temperature
• One at or above the hot inlet temperature
• One in between these temperatures

• Where do we find physical properties data?

– In CAS
– Ask the customer
– Tables and Handbooks
– Contact Market Segment for estimates
Physical properties
• Newtonian, non-Newtonian fluids and viscosity
– Newtonian fluids
• Most common

• Same viscosity at different velocities
• Viscosity only varies with temperature Newtonian
No
n-
Ne
wt
– Non-Newtonian fluids on
i an

• Viscosity varies depending on velocity Velocity

(and temperature)
• Example, ketchup, yoghurt and paint
• Contact Market Segment for design support
Shear stress
• What is shear stress?
– The force of the flow on the heat exchanger wall
– A measure of the turbulence in the heat exchanger
– Also called the Tao-value ()

• A high shear stress ensures

– High turbulence  High heat transfer
– High force on the wall  Reduced fouling

• Shear stress is calculated in CAS

 Shear stress
• The -value formula
w = shear stress at the wall, N/m² (Pa)
f    V 2 P  Dh f = friction factor, -
w   V = flow velocity
2 2 L L = channel length, m
Dh = hydraulic diameter (m)
(2 * pressing depth for PHE)
– Higher pressure drop  Higher shear stress
– A longer plate requires a higher P to achieve the same shear stress
as in a shorter plate (P/L constant)
– When going from one-pass to two-passes you need to double the P
to achieve the same shear stress (P/L constant)
– A plate with a larger pressing depth gives higher shear stress
providing that it has the same P as a plate with less pressing depth
 Shear stress
• Shear stress () as a function of velocity
The three different channels gives
For the same velocity different shear stress - L, M, H
the PHE gives a higher
shear stress than the
Shell & Tube
Shear stress versus fouling rate
Fouling resistance growth rate 45

40

35
( 10E-8 m2K/Wh)

30

25

20

15

10

0
0

0
10

30

90
50

70

15

17
11

13

19
Shear stess ( pa )

• Rule of thumb: Try to keep the shear stress >50 Pa

Design safety factors
• Why safety factors when designing?
– Variation in flow rates and physical properties
– Errors in predicting the k-value and pressure drop
– Allows fouling (dirt) on the plate and it still does the job

• Two ways of including safety factors

– Fouling factor Defined for Shell & Tubes
– K-value margin Defined for Plate Heat Exchangers

Many customers are used to S&Ts

You need to explain the difference !
Design safety factors
• Fouling factors (for S&T)
– The k-value formula with fouling factor (Rf)

1 1
  Rf
k Service k Clean
– Rf introduced as an “extra” heat transfer resistance

1-value Layer of fouling

/
Rf

2-value
Design safety factors
• K-value margin (for PHEs)
– Defined as a % margin between kClean and kService

kclean  k service
M arg in  100 
k service

– K-value margin is based on experience

– Depend on application
– Normally 5-10%
– Too much margin  More area  Lower flow per channel 
Less turbulence  Fouling !
Design safety factors
• Typical S&T in water-water duty
– A normal Rf for S&T: 1,0  10-4 m2ºC/W

– A normal kClean for S&T: 2000 W/ m2ºC

1 1 1
  Rf   10  4  kService= 1667 W/ m2ºC
k Service k Clean 2000

– What does this correspond to in K-value margin?

kclean  k service 2000  1667
M arg in  100   100   20%
k service 1667

– Normal margin for S&T

• Works with lower turbulence  More fouling
• Area not so easy to access for cleaning
Design safety factors
• What if we apply the Rf on a PHE in water-water duty?
– Rf for S&T on PHE: 1,0  10-4 m2ºC/W

– A normal kClean for PHE: 6000 W/ m2ºC

1 1 1
  Rf   10  4  kService= 3750 W/ m2ºC
k Service k Clean 6000

– What does this correspond to in K-value margin?

kclean  k service 6000  3750
M arg in  100   100   60%
k service 3750

– Much too high margin

• Too many plate  Less turbulence  Fouling !
• Maybe not competitive?
Fouling
• What is fouling?
• Something that:
– Reduces the heat transfer
– Increases the pressure drop
– Something that destroys the plate material
– Something that leads to maldistribution
• 5 types of fouling
– Major debris
– Biological growth
– Scaling
– Sedimentation
– Burn-on
Fouling - major debris
• What is major debris?
– Large objects and particles that get stuck in the HE
– Example, rocks, branches, coca-cola cans, fish

• To avoid major debris clogging the HE use

– Strainers (large mesh)
– Filters (fine mesh)

• Why filter?
– Maintenance cost savings
– Trouble -free operation
– Increased operation time for heat exchangers
– Less fouling - less corrosion
– Pressure drop stability
Fouling - major debris
• When do you need a filter?
– Sea water
– River water
– Channel water
– Cooling tower water
– Process liquids

• Alfa Laval has three different products:

– Back-flushing equipment
– Alfa Port filter
– ALF filter
Backflushing
• Flow direction is reversed
• Flushes the debris out of the port and back to the source
Normal flow
Reversed flow
Alfa Laval Port Filter
• A simple solution without separate filter
• Cylindrical perforated tube
• Protects at the inlet port
• Available on most M-serie PHEs
• Mounted and dismantled from pressure plate
• Requires
– Hole in pressure plate
– Lining
– Inspection cover

Remove inspection cover Pull out and empty

 Alfa Laval Filters
Normal operation
• Advantages
– Easy to install - Saves space
– No extra pump capacity
– No disturbance of operation
during flushing Flushing
– Low pressure drop
– High reliability
– Low flushing pressure
– Easy and quick service
Back flushing
– Good corrosion resistance
 Alfa Laval Filters
Sizes Filter body
• ALF 10, 15, 20, 30, 40, 50, 60 • Rubber lined carbon steel
• Connections from 10 to 600 mm (ALF20-60)
• Flow rates up to 5000 m3/h • Glass fibre reinforced plastic
(ALF10-40)
• AISI 316 (On Request)
• Avesta 254 SMO (On Request)
Control system Internal wetted parts
• PLC digital circuit controller • AISI 316
• Pneumatic or electrical actuators • Avesta 254 SMO
• Flexible system with 6 programs • Titanium
• Correct pre-set times
Examples, ALF installations

5 x ALF60 - R
Fouling - biological growth
• What is biological growth?
– Micro organisms that grow on the heat transfer surface
– Example, algae, bacteria

• To avoid major debris clogging the HE use

– Chock poisoning
Environmental problems
– Chlorinating
– Cleaning In Place (CIP)
– CIP is described in a moment
Fouling - scaling
• What is scaling?
– Many fluids contains dissolved salts
– When the fluid is heated or cooled the salts solubility changes
– The salt precipitates on the heat transfer surface
• Normal solubility
– Increased solubility at higher temperature
– Most common Solubility
Sugar

– Precipitates when cooled down

– Be careful when cooling
– Example, sugar in water
Temperature
 Fouling - scaling
• Reversed solubility Solubility
Sugar
– Reduced solubility at higher temperature
– Precipitates when heated up
– Be careful when heating
CaCO3

Temperature
• Common problem
– Cooling water with CaCO3 and Ca(PO4)2
– Avoid cooling water outlet temperatures above 45-50ºC
– Design with high shear stress ()
– Recommend water treatment
– Regularly apply CIP
Cleaning In Place
• What is Cleaning In Place?
– A chemical agent is circulated in
the HE to dissolve fouling
– Important parameters
• Concentration of the chemical
• Temperature of the chemical
• Time of circulation
• Mechanical action (turbulence)

• Why use cleaning In Place?

– Removes fouling without opening
– Increases lifetime for heat exchangers
– Minimises downtime
– Cost effective
 Alfa Laval CIP units

• CIP-20 (liter) • CIP-75/200/400 (liter) • CIP-1500/3000 liter)

• Small pump • 1.1 kW pump • 4 kW pump
• Plastic • 30 m3/h at 50 kPa head • 80 m3/h at 50 kPa head
• 6 to 12 kW heaters • 4 x 12 kW heaters
• 11/2” connection • 80 mm conn. on pressure side
• Stainless steel • 100 mm conn. on suction side
• Stainless steel
Examples, Alfa Laval CIP liquids
Agent Solution based on Removes

Alfa Phos Phosphoric acid Scale (CaCO3), rust (Fe2O3)

Alfa Caus Sodium hydroxide Fat, grease, paint, organic

and biological deposits

Alfa Neutra Sodium hydroxide Neutralize acid cleaners

Alfa Pass Potassium hydroxide Passivate Carbon Steel

surfaces (ex., pipes)
 Cleaning In Place - example
• M30-FM with 300 plates
Manual cleaning Cleaning In Place
Time 50 hours 6 hours
Labour cost (600 SEK/h) 30 kSEK 3.6 kSEK
Chemicals (150L Alfa Phos) n/a 4 kSEK
Investment (CIP1500) n/a 190 kSEK
Pay-back 190/(30-7.6) = 8.5 times
Risks Damaged gaskets
Leaking unit
Down time
Lost production
Advantages No investment Short pay-back
Planned stops
Chemicals can be
reused
 Fouling - sedimentation
• What is sedimentation?
– Fine particles that settles on the heat transfer surface
– Reduces the k-value and the HE does not perform
– Difficult to remove in filter
– Hard to dissolve with chemicals (CIP)

• How to avoid it?

– Design with high turbulence / shear stress
• Utilise pressure drop
• Use H-theta plates
– Back-flushing can be an option
 Fouling - burn on
• What is burn on?
– Breakdown or polymerisation of molecules that stick to the plate
– Example, when you boil milk on the stove it burns easily
– Common in food and organic applications
– Caused by too high temperatures

• How to avoid it?

– Design with high turbulence / shear stress
• Utilise pressure drop
• Use H-theta plates
– Check what wall temperature can be allowed
– Apply co-current flow to reduce wall temperature
Fouling in PHE vs S&T
• PHE is considered to foul less than a S&T
– High turbulence  High shear stress  Less fouling
– Low wall temperatures due to efficient heat transfer
 Less risk of scaling and crystallisation
– Material is selected to avoid corrosion
(S&T have corrosion allowance)
– No zones of low velocity

Baffles
Fouling in PHE vs S&T
• Examples, Heat Transfer Research Institute (HTRI)
– HTRI study of typical fouling in cooling tower water

PHE S&T
Flow velocity (m/s) 0.45 1.8 m/s
Shear stress (Pa) ca 60 ca 15
Result: Fouling in PHE was 50-70% lower

– PHE  Higher turbulence at a lower velocity  Less fouling