Download as pdf or txt
Download as pdf or txt
You are on page 1of 19

1

THERMALDESALINATIONPROCESS
A. MSFENGINEERING
Contents(ThisispartialcontentoftheoriginalsourceThermal
DesalinationProcessMSFEngineeringforM.A.Darwish)

1. OncethroughMSFSystem
a. SinglestageFlashSystem
b. TwoStagesFlashSystem
c. nStageFlashSystem

2. RecirculationMSFSystem

B. MSFModeling
Contents(ThisispartialcontentoftheoriginalsourceMSFModelingand
ComputerPackageforH.Eltony)

1. OncethroughMSFSystem
2. RecirculationMSFSystem

2
THERMAL DESALINATION PROCESS
MSF ENGINEERING
M.A. Darwish
Department of Mechanical and Industrial Engineering, Kuwait University, Kuwait
Contents (This is partial content of the original source)
1. Once-through MSF System
2. Recirculation MSF System

1. Once-through MSF system
1.1 Single-stage Flash System
The once-through MSF represents the limiting case of the ME system when the feed flow
rate is increased to the extent that the vapor is generated by flashing only and the boiling
heat transfer surface is entirely omitted. The flash vapor is condensed on the feed heater-
type condensers.
In a single-stage flash desalting system (Figure 1), the feed M
f
, after pre-treatment, is pre-
heated from T
cw
to T
f1
in the condenser tubes. It is then heated in the brine heater (called
heat input section, HIS) to temperature T
bo
by condensing the supply steam M
s
(main
energy source). The feed, leaving the brine heater at temperature T
bo
and pressure P
o
,
enters the first flashing chamber (or stage) kept at P
1
of saturation temperature T
b1
<T
bo
.
The feed becomes unstable and part of it is spontaneously vaporized (i.e. flashed in order
to reach stability by dropping its temperature to T
b1
). The vapor released by flashing
flows upward through demisters (to separate entrained liquid brine) to the condenser
tubes in the upper portion of the chamber. The vapor condenses there and forms the
product. The heat released by vapor condensation is gained (or recovered) by the heated
feed. The temperature of the flashed vapor (from brine at T
b1
) suffers some losses T
l
due
to non-equilibrium, boiling point elevation, and pressure drops through the demister and
condenser tubes. The following energy balances can be carried out by assuming the
following.
a) Constant and average specific heat, C
p
,for the feed M
f
, distillate M
d
, and brine
M
b
, streams a
b) Constant and average latent heat value, ,for condensing vapor and supply steam.
c) A completely insulated stage.


3

Energy Balances:
(i) For the condenser tubes
1
( )
f p f cw d
M C T T M = (1.1)
(ii) For flashing brine
1
( )
f p bo b d
M C T T M = (1.2)
(iii) For the brine heater
1
( )
s f p bo f
Q M M C T T = = (1.3)
where Q is the heat input rate (kJ s
-1
) and M
s
is the steam flow rate assumed to be
supplied at saturated vapor condition and leaves at saturated liquid condition.
For the whole desalting unit
4
1 1 f p cw S b p b d p V
M C T M M C T M C T + + (1.4)
Since
f b d
M M M = +
and
1 1 V b l
T T T =
then
1
( )
S F p b cw
M M C T T
Equations (1.1), (1.2), and (1.4) give
1 1 f cw bo b
T T T T =
1
1
1
d f cw
s b cw
M T T
M T T

= <

(1.5)
1
d
s
M
PR
M
= < (1.6)
1 1
( ) ( )
f
d p bo b p f cw
M
M C T T C T T

= =

(1.7)
A typical example of a single-stage MSF system where,
T
bo
=90C (top brine temperature, TBT),
T
cw
=30C (seawater temperature), and
T
b1
T
f1
=4.2C (temperature approach of the condenser)
would give
T
bo
T
cw
=60C, T
f1
=57.9C, T
b1
=62.1C, M
d
/M
s
= 0.869, and
M
f
/M
d
=19.9 ( is assumed equal to 2330 kJ kg
-1
).
5
The value of M
f
/M
d
(=19.9) is large compared to M
f
/M
d
(=2.5) for a single-effect or
ME system when the feed total dissolved solids (TDS) =42000 p.p.m. and brine
blowdown TDS =70 000 p.p.m.
The areas of heat transfer surface for both the brine heater and condenser heat exchangers
can be determined by equating the thermal load of each exchanger with the overall heat
transfer coefficient U, multiplied by the surface area multiplied by the effective
temperature (LMTD). For the brine heater (with subscript b for the parameters involved)
s b b b
M U A LMTD = (1.8)
where
1 2
1
2
ln
b b
b
b
b
T T
LMTD
T
T

=


(1.9)
1 1
2
( )
( )
b s f
b s bo
T T T
T T T
=
=

For the condenser (with subscript c)
d c c c
M U A LMTD = (1.10)
1 2
1
2
ln
c c
c
c
c
T T
LMTD
T
T

=


(1.11)
1 1
2 1 1
( )
( )
c V cw
c V f
T T T
T T T
=
=

Recalling M
d
/M
s
=(T
f1
- T
cw
)/(T
b1
- T
cw
) and T
V1
=T
b1
- T
l
, the total specific area
A
t
/M
d
=(A
b
+A
c
)/M
d
is expressed by
1 1 1
1 1 1 1 1
ln ln
( ) ( )
t b cw s f b l cw
d f cw b bo f s bo c f cw b l f
A T T T T T T T
M T T U T T T T U T T T T T

= +



(1.12)

6
The typical example mentioned earlier, with
T
s
=100C, T
bo
=90C,
T
b1
=62C, T
cw
=30C,
T
f1
=58C, T
l
=1.2C
and by assuming U
b
=3 kW m
-2
C
-1
and U
c
=2.5 kW m
-2
C
-1

has a specific total area A
t
/M
d
=119.6 m
2
kg
-1
s
-1
.
7
1.2 Two-stage Flash System
Now a two-stage flash desalting system can be considered, as shown in Figure 2. The
feed M
f
entering the second stage condenser tubes at T
cw
is successively pre-heated to T
f2

and to T
f1
in the second and first stage condensers, respectively, before it enters the brine
heater. The supply steam M
s
raises the temperature of the feed in the brine heater from T
f1

to T
bo
before it enters the first stage as flashing brine which is kept at pressure P
1
(of
saturation temperature T
b1
<T
bo
).

The vapor generated in the first stage D
1
(due to the pressure drop of the flashing brine) is
condensed on the outer surface of this stage condenser tubes and, thus, raises the
temperature of the feed circulating in this condenser from T
f2
to T
f1
. The unflashed brine
(B
1
=M
f
- D
1
) and the formed distillate D
1
proceed to the second stage kept at pressure P
2

of saturation temperature T
b2
<T
b1
. Again both flashing brine and distillate suffer a
pressure drop, partially evaporated by flashing and their temperatures drop to the second
stage saturation temperature. The vapor generated in the second stage is condensed on
these stage condenser tubes and heats the feed flowing in the condenser from T
cw
to T
f2
.
The unflashed brine (B
2
=M
f
- D
1
- D
2
) leaves the second stage as brine blowdown at T
b2
.
The term D
2
represents the vapor evaporated (flashed) from the flashing brine in the
second stage only. The mass and energy balances for the two-stage flash desalting system
give the following.
8
Brine heater
1
( )
s f p bo f
Q M M C T T = =
Flashing brine entering the first stage
1 1
( )
f p bo b
D M C T T =
First stage condenser tubes
1 1 2
1 2
( )
where,
f p f f f p c
c f f
D M C T T M C T
T T T
= =
=

Second stage flashing brine
2 1 1 2
( ) ( )
f p b b
D M D C T T =
or
1
2
( )
f p b
M D C T
D


=
where T
b
=T
b1
T
b2
and D
2
is the vapor released from flashing brine. Vapor flashed
from distillate D
1
entering the second stage, D
1V
, is equal to
1 1 2 1
1
( )
p V V p b
V
DC T T DC T
D


= =
Accumulated vapor condensed on the second stage condenser tubes is equal to
1
2 1 2
p b f p c
V
DC T M C T
D D D


+ = + =
Second stage condenser
2 1 2
( ) ( )
f p f cw f p b b
M C T T M C T T =
Overall system (assuming no heat losses)
2
( )
s f p b cw
M M C T T =
The above equations give
9
1 2 bo f b cw
T T T T =
1 1 2 bo b f f b c
T T T T T T T = = =
1 2 2 b b f cw
T T T T =
1 1 2
1 2 1
( ) ( )
( )
bo b b b
d f p f p
T T T T
M D D M C M D C


= + = +
If (T
bo
T
b1
) is chosen equal to T
b1
T
b2
, then
2
2
bo b
T T T =
and
f p f p p
d f
M C T M C T C T
M M


= +




Define
y =C
p
T /
Then
M
d
=M
f
(1-y) y +M
f
y
M
d
=M
f
[1-(1-y)
2
]
2
1
1 (1 )
f
d
M
M y
=

(1.13)
Equations (3) and (13) give
2
1
1 (1 )
( )/
d
s p bo f
M y
M C T T

=

(1.14)


10
If the temperature approaches of the two-stage condensers (T
b1
T
f1
), and (T
b2
T
f2
) are
kept the same and equal to 4.2C and T
bo
=90C and T
cw
=30C, then
( ) ( )
2 2 bo cw bo b b cw
T T T T T T = +
60 =(T
bo
T
b2
) +(T
bo
T
f1
) =(T
bo
T
b2
)+(T
bo
T
b1
) +(T
b1
T
f1
)
60 =3 T +(T
b1
T
f1
)
This gives T =18C, T
f2
=48.6C, y =C
p
T/ =0.0335, M
f
/M
d
=15.16 and
M
d
/M
s
=1.604.
Thus, the addition of another stage improves the performance ratio M
d
/M
s
from 0.875 to
1.604, decreases M
f
/M
d
from 19.5 to 15.615, and, thus, reduces the energy rejected with
the blowdown brine (both brine blowdown flow and temperature are decreased).
The areas of heat transfer surfaces of the brine heater and stage condensers can be
determined as before. For the brine heater
d b b b
M U A LMTD =
1 2
1
2
ln
b b
b
b
b
T T
LMTD
T
T

=



1 1
2
( )
( )
b s f
b s bo
T T T
T T T
=
=

For the first stage condenser area A
1

( )
1 2 1 1
=
f p f f c C
M C T T U A LMTD
For the second stage condenser area A
2

( )
2 2 2
=
f p f cw c C
M C T T U A LMTD
11
1.3 Multistage (n-stage) systems



12
A flow sheet of a once-through MSF system is given in Figure 3a and 3b. The previous
analysis can be extended to include n number of stages. Assumptions:
1) insulated stages
2) equal temperature drop of the flashing brine, i.e.
T
bo
T
b1
=T
b1
T
b2
=T
b2
T
b3
= =T
b,n-1
T
b,n
=T
b
3) equal temperature rise in the feed seawater in the condenser of stage, i.e.
T
f1
T
f2
=T
f2
T
f3
=T
f3
T
f4
= =T
f,n
T
cw
=T
f
Accordingly,
nT
b
=T
bo
T
b,n
, and nT
f
=T
f1
T
cw

So, using the fact that T
b
=T
f
=T, yield:
T
bo
T
b,n
=T
f1
T
cw
.

The flashing brine enters the first stage of the flow rate M
f
and leaves the first, second,
third, and n stages at flow rates of :
M
f
(1-y), M
f
(1-y)
2
, M
f
(1-y)
3
and M
f
(1-y)
n
, respectively,
where y =C
p
T/ is the fraction of vapor released from the flashing brine. The flashing
brine leaving the last stage n is the brine blowdown B, i.e. B
n
=M
b
=M
f
(1-y)
n
.
The distillate generated from the flashing brine in the first, second, third, , and n stages
is
y M
f
, y(1-y) M
f
, y(1-y)
2
M
f
, , and y(1-y)
n-1
M
f
, respectively.
Accumulated distillate M
d
in all stages is M
d
=M
f
M
b
and M
d
=M
f
- M
f
(1-y)
n


1
1 (1 )
f
n
d
M
M y
=

(1.15)
A simpler approximate relationship between M
f
and M
d
can be obtained by considering
that the heat gain by all condensers is equal to the heat loss by the average flashing brine
due to a temperature drop from T
bo
-T
b,n
, i.e.
(M
f
M
d
/2)C
p
(T
bo
-T
b,n
) M
d
,
13
where (M
f
M
d
/2) is the average flow rate of the flashing brine between entering at flow
rate M
f
and leaving at flow rate (M
f
- M
d
). This gives

,
1
0.5 0.5 0.5
( )
f
d p bo b n p
M
M C T T n T C n y

= + = + = +



The heat gain in the brine heater Q is equal to Q =M
f
C
p
(T
bo
T
f1
) =M
s
.
A heat balance for the whole system gives
M
s
. =M
f
C
p
(T
bo
T
f1
) =M
f
C
p
(T
b,n
T
cw
).
The performance ratio D/S is then equal to

1
[1 (1 ) ]
( )
n
f
d d
s f s p bo f
M
M M y
PR
M M M C T T

= = =

(1.16)
It is noticed that the terminal temperature difference (TTD) is T
bo
T
f1
, which is equal to
T
bo
T
f1
=(T
bo
T
b1
) +(T
b1
T
f1
) =T +T
b1
T
f1
and, consequently,
,
1 1
[1 (1 ) ]
( )
n
d
bo b n s
p b f
M y
PR
T T M
C T T
n

= =

+



where T =(T
bo
T
b,n
)/n, and (T
b1
T
f1
) is the temperature approach in the first-stage
condenser. It is clear that T decreases with an increase in n and (T
b1
T
f1
) depends on
the condenser heat transfer surface and temperature losses. Thus, M
d
/M
s
depends on n and
the condenser's heat transfer surface area. It is also noticed that the temperature range
(TR) is equal to

, ,
1 1 1
1 1
( ) ( )
( ) ( )
( 1) ( )
bo cw bo b n b n cw
bo f b f
b f
TR T T T T T T
n T T T n T T T T
TR n T T T
= =
= + = + +
= + +


The condenser and brine heater heat transfer surface areas can be determined as outlined
before.
14
For the brine heater
f p s b b b
M C TTD M U A LMTD = =
For the stage i (see the temperature distribution in Figure 3b)

, , ,
, ,
,
( )
ln
( )
ln
( )
f p c i c i c i
f p c i c i
d d bi fi l
bi fi l
f p c i
d d c bi fi l
M C T U A LMTD
M C T U A
T
M M
T T T T
T T T
M C A
T
M M U T T T
=


=
+




=





Since temperature approach are the same for all stages, then A
c1
= A
c2
= = A
c,n
=A
c
.
Therefore,

, c i
c
d d
A
n A
sA
M M
= =


The effects of the design parameters such as the number of stages n and top brine
temperature (TBT) T
bo
and operating parameters such as seawater temperature T
cw
and
temperature losses T
l
on the performance ratio M
d
/M
s
, feed to distillate ratio M
f
/M
d
, and
heat transfer surface areas are shown in Figures 4 and 5. The use of once-through MSF
(compared to recirculation MSF which is discussed later) is very limited. Of 68 MSF
plants reviewed by Birkett and Newton (1981), only four were once-though plants. The
main reasons for not using once-through systems extensively compared to the
recirculation systems are as follows.
(a) The system uses a high feed to distillate ratio M
f
/M
d
(for T
bo
=90C, T
cw
=30C,
n =27, and T
b1
-T
f1
=4C, then M
f
/M
d
=10.2). This means that a large quantity of
feed M
f
(10.2 M
d
) is pre-treated (i.e. costly pre-treatment and a large quantity of
brine M
b
(9.2M
d
) is to be blown down at a temperature T
b,n
of 6C above the
seawater temperature (i.e. high energy loss).

(b) Large quantities of gas are released with flashing. This increases the danger of
corrosion, decreases the condensation heat transfer coefficient due to non-
condensing gases blanketing around the condenser tubes, and requires large
ejector capacity with more steam (or energy) consumption.
15
On the other hand, the once-through system is more simple than a recirculation MSF
system. The brine concentration in the first stage at high temperature is lower than the
corresponding one in the recirculation MSF system and this permits higher temperature in
the once-through system and decreases the scaling potentiality.
16

17
2. Recirculation MSF system
A recirculation MSF system of n stages, consists of three sections.
(a) The brine heater or the HIS.
(b) The heat recovery section of (n
r
=n
st
-n
j
) stages.
(c) The heat rejection section of n
j
stages.
Both recovery and rejection stages are trains of chambers similar to those of once-
through MSF systems with condenser tube banks arranged in long-tube (LT) or cross-
tube (CT) arrangements, as shown in Figure 7. In long-tube arrangements, the flow of the
recirculation stream in the heat recovery section (or cooling water in the heat rejection
section) inside the condenser tubes is parallel but in opposite direction (counterflow) to
the flashing brine flow at the bottom of the chambers. In cross-tube arrangements the
flow in the condenser tubes is in the traverse direction (or perpendicular) to the flashing
brine flow. The incoming seawater, at flow rate M
c
and temperature t
c
is pumped into the
condenser tubes of the last heat rejection stage and pre-heated successively in the heat
rejection stages. The cooling water stream leaving the rejection section is divided into
two streams: the first (M
c
- F) is rejected back to the sea and the second, the feed F, is
chemically treated before joining part of the flashing brine leaving the last stage to form
the recirculation stream. The recirculation stream R is pumped to the last recovery stage
(n-j) condenser tubes and is successively heated by passing through the recovery stage
heaters (i.e. condensers) and leaves the top (first) stage at temperature t
1
.
The recirculation flow then enters the HIS (i.e. brine heater) where it is heated to the TBT
(T
o
) by condensing the steam supply to the HIS as the main thermal energy input. The
temperature increase of the recirculation flow in the brine heater (T
bo
- T
f1
) is called the
terminal temperature difference (TTD). The recirculation flow leaving the HIS (e.g. at
pressure P
o
and temperature T
bo
) enters the first stage (kept at pressure P
1
of the
saturation temperature T
b1
<T
bo
) as flashing brine. Due to the pressure drop from P
o
to P
1
,
part of the flashing brine is spontaneously evaporated in order to drop its temperature to
T
b1
to stabilize the flow thermodynamically. The released vapor has saturation
temperature T
v1
=(T
n
-BPE) where is the boiling point elevation (BP is a function of
brine concentration and temperature). This vapor flows upwards to the condenser tubes
through demisters (to eliminate entrained brine droplets) and suffers temperature drops
due to pressure losses across the demisters and condenser tubes. The vapor is condensed
at the outer surface of the condenser tubes to form the product of the first stage and is
collected as product below the condenser at temperature T
b1
-T
l
where T
l
is the
summation of all temperature losses in the stage. The flashing brine proceeds to the
second stage kept at P
2
(<P
1
) of saturation temperature T
b2
(<T
b1
) and part of it is
vaporized spontaneously (flashed) to drop its temperature to T
b2
, and so on to the last
stage where the flashing brine leaves at temperature T
bn
. In addition, the product
accumulated in the first stage D
1
at T
b1
-T
l
is cascaded to the second stage and suffers a
pressure drop (and, consequently, temperature drop by vaporizing part of this product).
The vapors generated in the second stage from the flashing brine (of mass flow rate R -
D
1
) and from the incoming product (of mass flow rate D
1
) condense on the outer tubes of
18
this stage condenser. The accumulated distillate in the second stage is cascaded to the
third stage and so on to the last stage where it leaves at temperature T
n
-T
l
. The flashing
brine leaves the last stage at temperature T
bn
, highest concentration X
bn
, and flow rate M
b

equal to
M
b
=M
r
M
d
(2.1)
where D is the accumulated distillate product from all stages. Part of this outgoing brine,
Bl, is rejected to the sea while the balance (B - Bl) forms a portion of the recirculation
stream. As mentioned earlier, the other part of the recirculation stream is F which is taken
from cooling seawater leaving the rejection section. This means that
(M
b
B
l
) +M
f
=M
r

B
l
=M
f
M
d
(2.2)

3.1 Recirculation arrangements
In recirculation MSF system, the feed M
f
(or make-up seawater) forms the portion of the
recirculation flow by one of the following arrangements.
(a) The first arrangement. The treated feed is added to the flashing brine in the last
rejection stage before blowdown takes place. This arrangement eliminates the need for a
separate brine blowdown pump. The recirculation pump forces the recirculation from the
last rejection stage to the condenser of last recovery stage and rejects the blowdown brine
from its discharge side. The saving attained by not using a blowdown pump is
encountered by waste in chemicals since the feed is dosed before blowdown and by waste
in the pumping energy since the head of the recirculation pump is higher than (four or
five times) the head required by a blowdown pump.
(b) The second arrangement. The blowdown is rejected before mixing with the feed. A
special pump is needed in this arrangement to reject the blowdown from the last rejection
stage (operated and vacuum condition) to the environment with discharging pressure little
higher than the atmospheric pressure.
For both arrangements, the mass and concentration balances give
M
b
=M
r
M
d
(2.3)
X
r
M
r
=X
b
(M
r
M
d
)
X
r
=X
b
(1- M
d
/ M
r
) (2.4)

19
where X
r
and X
b
are the salt concentrations of the recirculation flow M
r
at its entry to the
first stage and the brine flow in the last stage when the feed is not added to the last stage.
The feed to distillate ratio M
f
/M
d
can be determined for both arrangements in terms of the
salt concentration of the feed X
f
(depending on locality) and maximum brine
concentration X
b
(a design parameter usually taken as less than 70 000 p.p.m.) as follows.
In the first arrangement, where the feed F is mixed with the leaving brine in the last stage
before brine blowdown Bl is rejected, a salt balance gives X
b
(M
r
-M
d
) +X
f
M
f
=(B
l
+
M
r
)X
r
.
The use of Eqs (2.3) and (2.4) gives

M
f
/M
d
={[(M
r
/M
d
) 1] / [(M
r
/M
d
) (1-(X
f
/X
b
) -1} (2.5)

The ratio M
r
/M
d
is to be determined in the next section. Similarly, for the second
arrangement, a salt concentration balance at the mixing point of the dosed feed F (at X
f
)
and the recirculated brine M
r
-M
f
(at X
b
) gives X
b
(M
r
-M
f
)+X
f
M
f
=X
r
M
r
. The use of Eqs
(2.3) and (2.4) gives

M
f
/ M
d
=1 / (1-(X
f
/X
b
) (2.6)

For a typical case of T
bo
=90C, T
cw
=30C, =2330 kJ /kg, Cp=4.2 kJ /kg
o
C, X
f
=42000
p.p.m., and X
b
=70000 p.p.m., there are values of M
r
/M
d
=9.2 for both arrangements,
M
f
/M
d
=3.06 for the first arrangement, and M
f
/M
d
=2.5 for the second arrangement. This
shows that M
f
/M
d
in the first arrangement is higher (i.e. more water to be treated at
additional cost) than in the second arrangement.

You might also like