MSF - Systems Optimized For High Fuel Efficiency: Authors: Fritz Alt Presenter: Fritz Alt
MSF - Systems Optimized For High Fuel Efficiency: Authors: Fritz Alt Presenter: Fritz Alt
Abstract
Since fuel efficiency and environmental impact became over the past years more important factors in
conceptual studies for future combined power & water generation projects, the MSF technology with
design and operation parameters as applied in the past may be no longer the preferred technology,
primarily because of the relatively high energy consumption.
Considering, that for conceptual studies and the realization of future combined power & water
generation plants also only reliable and well proven technologies may be considered, this paper will
describe suitable optimized MSF – cross tube evaporators with an increased range of performance ratios,
wherein only traditional, proven process parameters like top brine temperature, brine concentration,
flash chamber liquid loading, feed water treatment etc. and only proven, typically specified materials are
applied. Only tentatively higher number of flash stages in combination with simple changes to the flash
stage configurations are applied, leading beside higher performance ratios largely to improved operation
characteristics and eventually to higher plant availability and reduced life cycle cost.
The optimized MSF – cross tube evaporators are providing the possibility to increase the performance
ratio up to 16 kg / 2326 kJ (ratio of mass of distillate produced to thermal energy consumed) on a cost
competitive level. Furthermore, with one configuration option, cross tube evaporators can be build with
unit capacities significantly above 20 migd, while the space requirement may be at the same time
reduced.
With competitive capital costs for the optimized desalination plants and increased performance ratios,
the fuel efficiency of combined power & water generation plants may be improved significantly beyond
the linear change of evaporator performance ratios, since a large portion of turbine bypass steam used in
many cases may be significantly reduced or eliminated.
The International Desalination Association World Congress on Desalination and Water Reuse 2013 / Tianjin, China
REF: IDAWC/TIAN13-016
I. INTRODUCTION
The desalination market requires thermal desalination systems with performance ratios in the range of
12-16 kg/2326 kJ or beyond, in order to meet market requirements like improved fuel efficiency of
combined power & water generation plants. The foundation for the optimization of the evaporators
described in this paper is the use of well proven design and operation parameters as well as the use of
commonly preferred and specified evaporator materials.
This paper will describe the basic parameters applied to the optimization of the MSF-cross tube
evaporators and the range of achievable performance ratios. Further the direct thermal and electrical
energy consumption as well as the possible impact of fuel energy savings on the power plant side will be
compared.
For the optimization of the evaporators, two basic parameters are considered. Those are, the improved
use of heat transfer area and a more compact design of the flash stages. Goal of the optimization is to be
able to design and build high performance MSF units on a competitive capital cost level, while
maintaining or improving operation characteristics compared to the traditional system design.
The heat transfer area is a significant factor in the evaporator design optimization, since the heat
exchanger tubes are a major cost factor. Herein the available LMTD (log mean temperature difference)
is beside the heat transfer coefficient the main factor to be considered. The temperature difference of the
coolant (cooling water in the heat rejection section or re-circulating brine in the heat recovery section)
between tube bundle inlet and discharge of an evaporator flash stage changes with the number of flash
stages, while the vapor temperature relative to the coolant temperature will primarily depend on the
plant performance ratio, resulting in LMTD values as shown in Figure 1.
The LMTD will increase with increasing number of flash stages and constant performance ratio as
indicated with the vertical line. This means the required tube surface area would be reduced. On the
other side when increasing the number of flash stages, the performance ratio can be increased with
constant LMTD and constant heat transfer area, as indicated with the horizontal line in Figure 1.
Those benefits can be seen also in Figure 2, which shows the specific heat transfer area required per one
m3 of daily evaporator unit capacity for different evaporators calculated for common design conditions.
The curves MSF 18, MSF 21 and MSF 24 show the specific tube surface areas required for typical cross
tube evaporators with 18, 21 and 24 flash stages when they are designed for a range of performance ratio
from about 8 to 11 kg/2326 kJ. For those evaporators a specific tube surface area in the range of 2.0 –
3.3 m2 per m3 of daily design production capacity may be required.
4 18 STAGES
24 STAGES
C
30 STAGES
3
LMTD
36 STAGES
42 STAGES
2 48 STAGES
0
8 9 10 11 12 13 14 15 16 17 18
PERFORMANCE RATIO kg / 2326 kJ
Figure 1: LMTD over performance ratio with different number of flash stages in heat recovery section
18
16
14
kg / 2326 kJ
M SF 24
M SF 21
12
M SF 18
M SF A/ C
PR
10
6
2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0
SPECIFIC HEAT TRANSFER AREA m 2 / m 3 / day
The black curve shows the approximate or average specific tube surface area required for the optimized
evaporators, wherein the lower end of the curve may represent an evaporator configuration with 30 flash
stages and a performance ratio of approximately 10.5 kg/2326 kJ, while the upper end has been
calculated for 48 flash stages and a performance ratio of about 17.5 kg/2326 kg. When following in
Figure 2 for example a horizontal line at performance ratio 10 or 12 kg/2326 kJ, it can be seen how
much less tube surface area would be required when using the optimized evaporators designed with
tentative higher number of flash stages. Following any line in vertical direction, it can be seen how
much the performance ratio can be improved, with a fixed heat transfer area when using the optimized
evaporator configurations.
The basic concept applied for the optimization of the MSF cross tube evaporator configuration may be
easily understood when comparing the sizes and configuration of a single flash stage for example for a
20 migd and a 10 migd evaporator as illustrated in Figure 3. Considering the same brine liquid loading
in the flash stages, the 20 migd evaporator shell would have substantially double the width compared to
the 10 migd evaporator, which means the tube bundle will have twice the tube length. This would mean
roughly a linear change of the evaporator size. However, at the same time the coolant flow rates through
the tube bundles have to be doubled, resulting substantially in double the tube bundle cross section. As
such the volume of each single tube bundle is increasing substantially by a factor four. Consequently
beside the linear change of evaporator width also the height and length have to be increase. As a result,
beside the linear increase of required tube material, the consumption of the evaporator shell material is
increasing more than linear with the change of unit capacity, so that there is no benefit of “economy of
scale” when increasing the evaporator capacity.
.
Since the smaller flash stage for the 10 migd evaporator would function substantially like the larger flash
stage for the 20 migd evaporator, a 20 migd evaporator could be build more compact if a single flash
stage is replaced by two smaller flash stages of a 10 migd evaporator, without compromising in the
traditional reliable operation.
A typical cross section of a conventional 20 migd evaporator flash stage is shown in Figure 4 on the left
while two examples of possible configurations of flash stages for a 10 migd evaporator capacity are
shown for comparison next to it.
3
1 2
Figure 4: example with reduced flash stage height for a 10 migd evaporator
The basic concept of brine and vapor flow remains in the described configurations for the 10 migd
evaporator substantially the same as for the 20 migd evaporator, so that the general operation
characteristic of the conventional cross tube evaporator would be maintained.
A typical evaporator of smaller capacity using one of the possible tube bundle arrangements is shown in
Figure 10. The flash stage cross section on the right shows a uniform vapor release from the flashing
brine and a tentative drifting of vapor toward the cold end of the tube bundle, as it is the case for any
MSF cross tube evaporator configuration.
For an evaporator with 20 migd capacity this compact flash stage configuration can be used in two ways.
The first option would be the configuration type MSF-A [1, 2] with basically two 10 migd evaporators
side by side in one common evaporator shell as shown in Figure 6.
Figure 6: 20 migd evaporator formed by two 10 migd evaporators housed in one common shell
In this case basically the brine flow, vapor flow and condensation in each half of the 20 migd evaporator
would work identical to the flow characteristic in a single 10 migd evaporator. While half of the total
coolant flow required would pass in serial flow through the upper tube bundles, the other half would
pass in opposite flow direction through the lower tube bundles.
Examples of evaporators with optimized flash stage configurations are shown in Figure 8 in comparison
to a conventional cross tube evaporator. For comparison, each evaporator is designed for the same unit
capacity of 20 migd. The conventional cross tube evaporator shown on top comprises a total of 22 flash
stages which could reach with a specific tube surface area of 3.0 m 2 per m3 / day evaporator capacity a
performance ratio of about 10 kg / 2326 kJ. The evaporator shell would have a length of about 106 m
(excluding deaerator) and a width of about 24 m.
The evaporator below is of the MSF-A type comprising for comparison 44 flash stages with the flash
stage configuration having two tube bundles arranged on top of each other as shown in Figure 4, 5 and
The 3rd evaporator would be of the type MSF-C with 22 flash stages on each level with a flash stage
configuration applying the same flash stage length as for the conventional evaporator and a tube bundle
height cut in half. It would have with the same tube surface area the same performance ratio as the
conventional evaporator. With the same length, half the width and a total height (shell bottom plate to
roof plate) of 6.6 m, it would have about 20% less shell weight compared to the conventional
evaporator.
The 4th evaporator shown on the bottom of Figure 8 would be also of the MSF-C type but with 44 flash
stages on each level, with the tube bundle configuration as the type MSF-A evaporator shown in second
position. With the same tube surface area, it would reach the same performance ratio of 12.5 kg / 2326
kJ as the MSF-A type evaporator. With half the width and a total height (shell bottom plate to roof plate)
of 8.8 m, it would have about 12% less shell weight compared to the conventional evaporator and the
MSF-A type evaporator and would require about 50% of the foot print.
As per the graphic shown in Figure 2 and the examples of evaporator configurations described, it
becomes obvious that for individual cases it is a question of optimization between number of flash
stages and installed heat transfer area for a specified performance ratio, so that the performance ratio of
12.5 kg / 2326 kJ achieved in the described examples, could be also reached with lower number of flash
stages when the heat transfer area is adjusted accordingly. On the other side the tube bundles can be
designed with larger tube surface area, which allows also the increase of the performance ratio toward
15-16 kg / 2326 kJ.
When looking isolated at a desalination unit, the performance ratio may be an adequate measurement for
the thermal energy consumption. However, when it comes to the fuel efficiency of a combined power &
water generation plant and the fuel cost allocation, the load variations of power and water generation
become an important factor. Ideally a thermal desalination plant may receive the required LP-steam
from the discharge of a back pressure turbine. In practice, a certain amount of LP – steam may have to
be provided over a turbine bypass, wherein the amount of bypass steam may in most cased increase
during the winter season, when the ratio of power / water generation is dropping.
To get an idea how the allocated fuel consumption may be influenced by the desalination plant
performance ratio, three conditions are compared, the supply of 100% LP steam from the back pressure
turbine discharge, the supply of 100% LP steam over a turbine bypass and any intermediate situation.
For the comparison the following parameters are assumed. The HP-steam supplied from the boiler plant
has an enthalpy of 3,560 kJ/kg. The average boiler plant efficiency is assumed to be 88%. The
60.0
50.0
3
barrels / 1000 m
40.0
bypass 0%
bypass 20%
30.0
bypass 50%
20.0 bypass 100%
10.0
0.0
8 9 10 11 12 13 14 15 16 17 18
MSF PERFORMANCE kg / 2326 kJ
The lower green curve shows the fuel consumption for the ideal case, when the LP-steam is supplied
only from the back pressure turbine without any by-pass steam, while the upper red curve shows the fuel
consumption when 100% of the LP-steam consumed in the desalination unit is provided over a turbine
by-pass. An example for consideration in current conceptual studies may be when comparing a
desalination plant with performance ratio 10 kg/2326kJ and a portion of 50% LP-steam delivery over a
turbine by pass, versus the use of a desalination plant with performance ratio 12 kg/2326 kJ. With the
higher performance ratio, the consumption of steam supply over the by-pass would drop from 50% to
16.7%, reducing the overall fuel consumption for 1,000 m3 of distillate production from 31.4 barrels to
19.3 barrels equivalent to a saving of 38.4%. Another example may be when considering to replace
sometime in future an older desalination plant with performance ratio 8 kg/2326 kJ which uses also 50%
by-pass steam supply, with a new plant with performance ratio 16 kg/2326 kJ. In this case no by-pass
steam is required and the fuel consumption for 1,000 m3 of distillate production is dropping from 39.2
barrel to about 12 barrels equivalent to a saving of almost 70%.
Even so the described cases are based only on some rough assumptions, a comparison with more
detailed analysis for a specific case, including daily and annual cycles of power / water generation ratios
and changes of related boiler and turbine efficiency may provide tentatively similar results.
The approximate electric power consumption of the optimized desalination plants is shown in Figure 10
in comparison to the power consumption of desalination plants with conventional cross tube
evaporators. While the blue curve is calculated under consideration of in-tube velocities of about 2.1
m/s, the green curve is calculated with a reduced design velocity of 1.8 m/s. While most plants built in
the past, have been designed with the higher velocity range, simply to minimize capital cost, the
consideration of electric energy cost may become more important in future.
When comparing the examples of evaporators described in section 2.2 and shown in Figure 8, it can be
assumed that a desalination plant with optimized evaporators with 44 flash stages and performance ratio
12.5 kg/2326 kJ would have about the same power consumption as a plant with conventional cross tube
evaporators with 22 flash stages and performance ratio 10 kg/2326 kJ, when both are designed with the
same in-tube velocity.
8
7
3
electric energy kWh / m
5 MSF conventional
4 MSF A/C-lv
3 MSF A/C-hv
2
1
0
8 9 10 11 12 13 14 15 16 17 18
performance ratio kg / 2326 kJ
Designing an optimized evaporator with the lower in-tube velocity in the range of 1.8 m/s, an evaporator
with performance ratio 14-15 kg/2326 kJ would require about the same power consumption as a
desalination plant with conventional evaporators and performance ratio 10 kg/2326 kJ.
IV. CONCLUSION
• The MSF-cross tube evaporators can be built more compact without compromising on the
traditional operational characteristics of flash stages.
• The more compact design allows building evaporators with increased performance ratios on a
cost competitive level.
• Optimized evaporators with performance ratios in the range of 12-15 kg/2326 kJ may be build
with an electric power consumption below 4 kWh per m 3 of distillate production.
• With increased performance ratios, combined power & water generation plants can be
configured such that a significant lower amount of turbine by-pass steam would be used
compared to power & water generation plants with low performance desalination plants, so that
the fuel efficiency for the combined power & water generation could increase significantly.
V. REFERENCES