Boiling and Condensation

Patricio F. Mendez
University of Alberta
ChE 314 Heat Transfer

Contents
1 Phase Change: Boiling 2
1.1 Boiling modes . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Saturated pool boiling . . . . . . . . . . . . . . . . . . . . . . 2
1.3 The boiling curve of pool boiling . . . . . . . . . . . . . . . . 3
1.3.1 Surface flux during nucleate boiling . . . . . . . . . . . 5
1.4 Critical and minimum heat flux for nucleate pool boiling . . . 6
1.5 Total condensation rate . . . . . . . . . . . . . . . . . . . . . . 7
1.6 Parametric effects on pool boiling . . . . . . . . . . . . . . . . 8
1.6.1 Effects of subcooling . . . . . . . . . . . . . . . . . . . 8
1.6.2 Effects of surface roughness . . . . . . . . . . . . . . . 8
1.6.3 Effects of surface arrangements . . . . . . . . . . . . . 8
1.7 Example of calculation (example 10.1 in the textbook) . . . . 9
1.8 Forced convection boiling . . . . . . . . . . . . . . . . . . . . . 11

2 Phase Change: Condensation 13

2.1 Correlations for film condensation . . . . . . . . . . . . . . . . 14
2.2 Film condensation on radial surfaces . . . . . . . . . . . . . . 16
2.3 Condensation in horizontal tubes . . . . . . . . . . . . . . . . 17

1
1 Phase Change: Boiling
1.1 Boiling modes
Boiling occurs when the temperature of the surface Ts exceeds the saturation
temperature Tsat / Heat is transferred from the solid surface to the liquid, and
the resulting heat flx can be calculated by Newton’s law of cooling:

qs00 = h (Ts − Tsat ) = h∆Te Eq. (10.3)

where ∆Te = Ts − Tsat is termed the excess temperature.

Note: Tsat is not T∞ ! Tsat corresponds to the liquid pressure.
According to the motion of the fluid, boiling can be divided as:
• Pool boiling: equivalent of free convection
• Forced convection cooling: equivalent of forced convection
According to the fluid temperature, boiling can be categorized as:
• Saturated: T∞ > Tsat
• Subcooled: T∞ < Tsat

1.2 Saturated pool boiling

In saturated pool boiling, which is a extensively studied mode, the tem-
perature through most of the fluid remains slightly above Tsat with a sharp
increase close to the solid surface. Bubbles generated at the interface between
the liquid and solid rise to the liquid-vapor interface.

2
Figure 1: Temperature distribution in saturated pool boiling with a liquid-
vapor interface

1.3 The boiling curve of pool boiling

Understanding of underlying physical mechanisms can be obtained by exam-
ining the boiling curve. Shiro Nukiyama in the 1970s was the first to identify
different regimes of pool boiling by the experimental apparatus shown as be-
low, termed as power-controlled heating, where the excess temperature ∆Te
is the dependent variable and the heat flux is the independent variable. Both
heating curve and the cooling curve are showed as below:

Figure 2: Nukiyama’s power-controlled heating apparatus and the obtained

boiling curve for saturated water at atmospheric pressure.

3
Surface heat flux depends on both the convection coefficient h and the excess
temperature, ∆Te = Ts − Tsat . Different boiling regimes can be delineated
according to the value of ∆Te and they are: free convection, nucleate (begin
with isolated bubbles and then develops into jets and columns), transition
regime and film:

Figure 3: Typical boiling curve for water at 1 atm: surface heat flux qs00 as a
function of excess temperature, ∆Te = Ts − Tsat

Classification and important features of each regime:

• Free convection: ∆Te < 5 ◦ C where free convection correlations can be
applied
• Nucleate boiling:
5 ◦ C < ∆Te < 30 ◦ C where heat transferred from the surface to liquid

4
 5 ◦ C < ∆Te < 15 ◦ C generation of bubbles and bubbles stir liquid
15 ◦ C < ∆Te < 30 ◦ C jets interfere with liquid motion at surface
00
Critical heat flux qmax occurs at ∆Te ≈ 30 ◦ C, which reaches the max-
imum heat transfer in nucleate boiling and any more heat introduced
will increase the surface temperature drastically.
• Transition: 30 ◦ C < ∆Te < 120 ◦ C
It is unstable and switches between the regimes of nucleate and film
and does not have technological use at the present.
• Leidenfrost point ∆Te ≈ 120 ◦ C
Heat flux reaches a minimum and drops run floating on the vapor. It is
the minimum heat transfer by film and below this value, boiling mode
switches to nucleate.
• Film boiling ∆Te > 120 ◦ C
During this regime, a stable vapor blanket forms between the surface
and the liquid. Vapor has very low thermal conductivity and radiation
is relevant.

1.3.1 Surface flux during nucleate boiling

The surface heat flux during nucleate pool boiling regime can be calculated
as follows and it is valid for shapes much bigger than the size of a bubble:
 1/2  3
g (ρl − ρv ) cp,l ∆Te
qs00 = µl hf g Eq. (10.5)
σ Cs,f hf g Prl n

where the subscripts l and v denote the saturated liquid and vapor states,
respectively and σ(N/m) is the surface tension. hf g is the latent heat asso-
ciated with liquid-vapor transformation, which is denoted as if g in our class.
Experimentally determined values of the coefficient Cs,f and the exponent of
the Prandtl number are presented as follows:

5
Figure 4: Values of Cs,f and exponent n for various surface-fluid combinations

All properties associated with Eq. 10.5 are evaluated at Tsat .

1.4 Critical and minimum heat flux for nucleate pool

boiling
The critical (maximum) heat flux for nucleate pool boiling can be expressed
as follows and it is valid for shapes much bigger than the size of a bubble:
 1/4
00 σg(ρl − ρv )
qmax = Chf g ρv Eq. (10.6)
ρ2v

6
Eq. (10.6) is independent of the surface material and weakly dependent on
the heated surface geometry.
• For large horizontal cylinders, spheres, finite shapes,
C = π/24 ≈ 0.131 (the Zuber constant).
• For large horizontal plates, C = 0.149.
All properties in Eq. (10.6) are evaluated at Tsat . Gravity term g 1/4 is
confirmed for rotating machinery and microgravity.
The minimum heat flux for nucleate pool boiling from a large horizontal
plate:
 1/4
00 σg(ρl − ρv )
qmin = 0.09ρv hf g Eq. (10.7)
(ρl + ρv )2

where the properties are evaluated at the Tsat . Eq. (10.7) is accurate to ap-
proximately 50% for most fluids at moderate pressures but is not as accurate
for higher pressures.

1.5 Total condensation rate

According to Newton’s cooling law q = Ah̄L (Tsat − Ts ), the total condensa-
tion rate may then be determined as:

q hL A(Tsat − Ts )
ṁ = 0
= Eq. (10.34)
hf g h0f g

where h0f g is a modified latent heat used to account for the inclusion of
thermal advection effects:

h0f g = hf g + 0.68cp,l (Tsat − Ts )

Eq. (10.34) is generally applicable to any surface geometry, however, calcu-

lation of h̄L may vary from different geometry and flow conditions.

7
1.6 Parametric effects on pool boiling
1.6.1 Effects of subcooling
The liquid is termed as subcooled if it is maintained at a temperature less
than the saturation temperature Tsat . Temperature difference is defined as:

∆Tsub ≡ Tsat − Tl

For nucleate boiling, the influence of subcooling is considered to be negligible,

however, the critical and minimum heat fluxes are known to increase linearly
with ∆Tsub .

1.6.2 Effects of surface roughness

Surface roughness (by machining, grooving, scoring, or sandblasting) has
negligible effects on the critical and minimum heat fluxes. However, increased
surface roughness result in a larger increase in heat flux in the nucleate boiling
regime because a roughened surface can provide more and larger sites for
bubble growth.

Figure 5: Enlarged profile of a roughened surface with more and larger nu-
cleation sites

1.6.3 Effects of surface arrangements

Special surface arrangements can provide stable augmentation (enhance-
ment) of the nucleate boiling. There are two types of enhancement surfaces:

8
 1. coatings of very porous materials formed by sintering, brazing or flame
spraying, electrolytic deposition, or foaming
2. mechanically machined or formed double-reentrant cavities which en-
sure continuous vapor trapping

Figure 6: Typical structured enhancement surfaces for augmentation of nu-

cleate boiling: (a) sintered metallic coating and (b) mechanically formed
double-reentrant cavity

1.7 Example of calculation (example 10.1 in the text-

book)
The bottom of a copper pan, 0.3 m in diameter, is maintained at 118◦ C by
an electric heater. Estimate the power required to boil water in this pan.
What is the evaporation rate? Estimate the critical heat flux. Properties:
for saturated water, liquid (100◦ C ): ρl = 957.9 mkg3 , cp,l = 4.217 Kg.K
kJ
, µl =
−6 N.s kJ −3 N
µf = 279 × 10 m2 , Prl = Prf = 1.76, hf g = 2257 Kg , σ = 58.9 × 10 m . For
saturated water, vapor (100◦ C ):ρv = 0.5956 mkg3 .

9
Figure 7: Schematic of water boiling in a copper pan of prescribed surface
temperature

Solution:
According to Figure 1.4, for polished surfaces: Cs.f = 0.0128 and n = 1.
Substituting into Eq. (10.5), the boiling heat flux is:
 1/2  3
g (ρl − ρv ) cp,l ∆Te
qs00 = µl hf g Eq. (10.5)
σ Cs,f hf g Prl n
kg 1/2 4.217 kJ (118 ◦ C − 100 ◦ C) 3
" # " #
m
N.s kJ 9.8 2 (957.9 − 0.5956) 3 Kg.K
= 279 × 10−6 2 × 2257 s
N
m
kJ
m Kg 58.9 × 10−3 m 0.0128 × 2257 Kg × 1.76
W
= 8.36 × 105
m2

Hence, the boiling heat exchange rate is:

πD2 00 π (0.30m)2 W
q= Aqs00 = qs = × 8.36 × 105 2 = 5.91 × 104 W
4 4 m

Under steady-state conditions all heat addition to the pan will result in water
evaporation from the pan:

qs = ṁhf g

10
where ṁ is the rate at which water evaporates from the free surface to the
room. Hence, the evaporation rate can be computed as:

q 5.91 × 104 W kg kg
ṁ = == kJ
= 0.0262 = 94
hf g 2257 Kg s h

The critical heat flux for nucleate pool boiling can be estimated using Eq.
(10.6) and for large horizontal plates, C = 0.149:
 1/4
00 σg(ρl − ρν )
qmax = Chf g ρv
ρ2ν
" #1/4
kJ kg 58.9 × 10−3 m
N
× 9.8 sm2 × (957.9 − 0.5956) mkg3
= 0.149 × 2257 × 0.5956 3 ×
2
Kg m 0.5956 mkg3
MW
= 1.26 2
m

1.8 Forced convection boiling

Fluid flow In pool boiling is due primarily to the buoyancy-driven motion of
bubbles originating from the heated surface. For forced convection boiling,
flow is due to a directed motion of the fluid, together with the buoyancy ef-
fects. Conditions depend strongly on geometry, which may be either external
flow over heated plates and cylinders or internal flow. Internal, forced con-
vection boiling is commonly referred to as two-phase flow and is characterized
by rapid changes from liquid to vapor in the direction of flow (from section
10.5 in the textbook). In two-phase flow, boiling is accompanied with the
formation of bubbles at the inner surface of a heated tube and the existence
of a variety of two-phase flow patterns.
Consider the internal flow inside a vertical tube with a constant surface heat
flux qs00 as shown below with the variation on the convection coefficient h as
a function of the spatial variable x, increasing father down the tube:

11
Figure 8: Flow regimes for forced convection boiling in a tube.

12
Regimes of two-phase flow:
• liquid forced convection: liquid phase only
• subcooled flow boiling
Bubbles form adjacent to the heated wall and subcooled liquid flows
near the center of the tube.
• saturated flow boiling: bubbly (bubble flow and start of saturated flow),
slug (with bubble coalescence), annular(slow liquid on surface and fast
vapor in core), and mist (small drops carried by high vapor speed)
• vapor forced convection: vapor phase only and all drops have evapo-
rated

2 Phase Change: Condensation

When the temperature of a vapor is reduced to below its saturation temper-
ature, condensation occurs. There are three typical modes of condensation:
• Surface condensation: vapor condenses on a cool surface and may occur
in one of two ways, depending on the condition of the surface:
1. Film condensation: a liquid film covers the entire condensing sur-
face and with gravity, the film flows continuously from the surface.
It typically occurs at clean surfaces or fouled treated surfaces.
2. Dropwise condensation: typically occurs on at treated surfaces
(e.g. Teflon) and more than 90% of the surface is covered by
drops. Dropwise condensation has much higher heat transfer rate
than that of film (≈ 10 times) but is difficult to maintain.
• Homogeneous condensation: vapor condenses as droplets suspended in
a gas phase to form a mist
• Direct contact condensation: vapor condenses in contact with a cold
liquid

13
Figure 9: Modes of condensation: (a) Film (b) Dropwise condensation on
a surface (c) Homogeneous condensation or fog formation resulting from in-
creased pressure due to expansion (d) Direct contact condensation

2.1 Correlations for film condensation

In film condensation on the vertical surface as shown in Figure 10.a, transition
criterion of turbulent flow conditions can be expressed in terms of a Reynolds
number defined as:
4Γ
Reδ ≡ Eq. (10.35)
µl

As the condense mass flow rate is given by ṁ = ρl um bδ, the Reynolds number
can be expressed as:
4ṁ 4ρl um δ
Reδ = = Eq. (10.36)
µl b µl

14
where um is the average velocity in the film and δ is the thickness of the film.
The Reynolds number is the indicator of flow conditions:
• for Reδ .30, the film is laminar and wave free.
• for Reδ ≈1800, the transition from laminar to turbulent flow is com-
plete.

Figure 10: Film condensation on a vertical plate: (a) condensate rate for
plate of width b and (b) flow regimes

q hL A(Tsat −Ts ) 4ṁ

Combining Eq. (10.34) ṁ = h0f g
= h0f g
and Eq. (10.36) Reδ = µl b
=
4ρl um δ
µl
together, the average Nusselt number can be provided:

1/3
h̄L (νl 2 /g)
Reδ = 4P = 4P NuL
kl
where the dimensionless parameter P is:

kl L (Tsat − Ts )
P =
µl h0f g (νl 2 /g)1/3

15
In terms of P, average Nusselt numbers have the following expressions:
1/3
h̄L (νl 2 /g)
NuL = = 0.943P −1/4 P . 15.8 Eq. (10.43)
kl
1/3
h̄L (νl 2 /g) 1
NuL = = (0.68P + 0.89)0.82 15.8 . P . 2530 Eq. (10.44)
kl P
1/3
h̄L (νl 2 /g) 1 h i4/3
NuL = = (0.024P − 53) Prl 1/2 + 89 P & 2530, Prl > 1 Eq. (10.45)
kl P
In Eqs. (10.43) to (10.45), properties for liquid are evaluated at the film
temperature Tfilm = (Tsat + Ts )/2 and the value for vapor is evaluated at Tsat .

2.2 Film condensation on radial surfaces

The average Nusselt number for laminar film condensation on outer surface
of a sphere or a horizontal tube as presented in Figure 1.11 has the form:
1/4
ρl g (ρl − ρv ) h0f g D3

h̄D D
NuD = =C Eq. (10.46)
kl µl kl (Tsat − Ts )

where the coefficient C is:

• for the sphere C = 0.826
• for the tube C = 0.729
In Eq. (10.46), properties for liquid are evaluated at the film temperature
Tfilm = (Tsat + Ts )/2 and the value for vapor is evaluated at Tsat .

16
 Figure 11: Film condensation on a single horizontal tube and a sphere

2.3 Condensation in horizontal tubes

Flow conditions within the tubes depend strongly on the velocity of the
vapor, the mass fraction of vapor, and the properties of the fluid. For low
vapor velocities, Rev,i is defined as:
 
ρv um,v D
Rev,i = < 35000 Eq. (10.50)
µv i

where the subscript i refers to the inlet of the tube.

Average value of Nusselt number can be calculated using Eq. (10.46) with
C = 0.555 and h0f g = hf g + 0.375cp,l (Tsat − Ts ):
1/4
ρl g (ρl − ρv ) h0f g D3

h̄D D
NuD = =C Eq. (10.46)
kl µl kl (Tsat − Ts )

17
 Figure 12: Cross section of condensate flow for low vapor velocities

At high vapor velocities, the two-phase flow turns into turbulent and annular
with the vapor occupying the core of the annulus. The thickness of the outer
condensate layer increases in the flow of direction.

Figure 13: Longitudinal section of condensate flow for large vapor velocities

18
An empirical correlation for a local Nusselt number has the form of:
 
hD 0.8 0.4 2.22
NuD = = 0.023ReD,l Prl 1 + 0.89 Eq. (10.51a)
kl Xtt

where ReD,l = 4ṁ(1 − X)/(πDµl ), and X = ṁv /ṁ is the mass fraction of
vapor in the fluid. Xtt is the Martinelli parameter to capture the existence
of turbulent flows in both the liquid and vapor phases:
 0.9  0.5  0.1
1−X ρv µl
Xtt = Eq. (10.51b)
X ρl µv

When there is no vapor, equation reduces to internal flow equation.

Acknowledgement
This set of notes is based on class notes from previous years by Prof. Mendez,
and was first typed and illustrated by PhD student Ying Wang as part of
her Fraser and Shirley Russell Teaching Fellowship.

19