Report M3 Rick Van Pruissen 20100603

7/22/2019 Report M3 Rick Van Pruissen 20100603

1/44

May 24th

201

Eindhoven University of Technology

Building Service

Master project

Supervisors

Prof. ir. P.G.M. Rutte

Dr. ir. M.G.L.C. Looman

Spaceco

olingusingawetcoolingtower

EVAPOR

ATIVECO

OLING

RickvanPruissen


2/44

PrefaceThis document is the report of my 3rd master project of the master programme Building Services of

the department Building Physics and Systems at Eindhoven University of Technology. This project is

based on an assignment from DWA installatie- en klimaatadvies [5]. This project could possibly be

an introduction to my graduation project, this possibility will be re-evaluated during or after the

project.

I would like to thank prof.ir. Paul Rutten and dr.ir. Marcel Loomans for their advice and patience

during this project. I would also like to thank Martijn Menschaar from GEA Polacel for his elaborate

introduction into the world of cooling towers.

Rick van Pruissen

May 2010


3/44

SummaryCooling is getting increasingly important in buildings. This rise in cooling demand is commonly met

with active cooling systems (chillers). Disadvantages of this method are: increased (fossil) energy

usage, larger environmental impact (materials, refrigerant and emissions) and higher costs

(investment and exploitation). In this report the application of a wet cooling tower is investigated for

an office building in the Netherlands to cover the total cooling load without the use of a chiller.

From the working principle and physics of the different types of cooling towers follows that the type

of cooling tower with the highest potential is a wet cooling tower operating in counterflow

(henceforth cooling tower). From this analysis also follows that the performance of a certain cooling

tower depends on the inlet temperature, the outlet temperature and the wet-bulb temperature.

These temperatures are the limiting factors for the coefficient of performance (COP) of a cooling

tower. The smaller the difference between WBT (wet-bulb temperature) and the inlet temperature,

the lower the performance of the cooling tower. The outlet temperature follows from the WBT, inlet

temperature and cooling tower design. For the application of a cooling tower the requirements for

the building are: a low latent cooling load, the presence of a high temperature cooling dissipation

system (concrete core conditioning or chilled ceilings) and high efficiency ventilation (e.g.

displacement ventilation). The supply temperature of a high temperature cooling system should be

18oC maximum. A difference of at least 2K should exist between cooling tower outlet temperature

and wet-bulb temperature making the maximum wet-bulb temperature 16oC.

Using data from a manufacturer the COP of an actual cooling tower was determined for a WBT of

16oC. This COP is approximately 35. Lower capacities at this WBT will increase the COP because the

fan speed can be lowered. Lower WBTs will also increase the COP.

An analysis of the Dutch climate proves that this temperature is exceeded for 546 hours per year on

average over 20 years. In 1994, a commonly used Dutch reference year, this number was 476 of

which 217 occurred during office hours. The duration of these excesses ranges from 1 hour to 3 days

in 1994. From a coarse investigation of thermal storage techniques follows that a sensible buffer to

bridge 1 day of excess would become very large (232.3m3) due to the small temperature difference

over the cooling system. A latent (Paraffin) buffer to bridge 1 day would be approximately 10.6m3.

A coarse evaluation of the economic applicability of cooling towers in the Netherlands shows that the

investment for the cooling tower investigated was 197 /kW. The closest alternative sources in this

capacity range (~100kW) are ventilation air and open air in combination with a heat pump, both with

an investment of approximately 550 /kW. Sensible buffering with water adds approximately 180

/kW to the total, which is still considerably lower than direct alternatives. A latent buffer adds 725

/kW to the total making the investment considerably larger than its direct alternatives.

Preceding findings imply that cooling towers are technically and economically feasible for application

in the Netherlands as means of direct cooling. When applied as base cooling, cooling towers can be

used constantly with a high COP (35 compared to typically 3-6 for a conventional chiller-dry-cooler).

When used for full load cooling, flexible comfort boundaries must be applied. A buffer could lower

the amount of exceeding hours. A buffer to avert the exceeding hours is technically (size) and

economically unviable.


4/44

Table of contents

1 INTRODUCTION........................................................................................................................................ 5

1.1 GENERAL .................................................................................................................................................. 5

1.2 GOAL ....................................................................................................................................................... 6

1.3 METHOD .................................................................................................................................................. 6

1.4 REPORT OUTLINE ........................................................................................................................................ 6

2 THE WORKING PRINCIPLE OF COOLING TOWERS ..................................................................................... 8

2.1 GENERAL .................................................................................................................................................. 8

2.1.1 Heat transfer methods ...................................................................................................................... 9

2.1.2 Airflow ........................................................... ................................................................... ............... 10

2.2 COOLING POTENTIAL ................................................................................................................................. 11

2.2.1 Dry cooler ........................................................................................................................................ 11

2.2.2 Evaporative coolers ......................................................................................................................... 12

2.2.3 Technical comparison ...................................................................................................................... 12

2.3 CONDITIONS FOR APPLICATION .................................................................................................................... 13

2.3.1 High temperature cooling (HTC) ................................................................. ..................................... 13

2.3.2 Ventilation ....................................................................................................................................... 13

2.4 PERFORMANCE OF A COOLING TOWER .......................................................................................................... 15

2.4.1 Coefficient of performance .............................................................. ................................................ 15

2.4.2 Calculation ....................................................................................................................................... 16

3 APPLICABILITY OF EVAPORATIVE COOLING ............................................................................................ 19

3.1 SPATIAL VARIATIONS ................................................................................................................................. 193.2 TEMPORAL VARIATIONS ............................................................................................................................. 20

3.3 APPLICABILITY IN THE NETHERLANDS ............................................................................................................ 20

3.3.1 Mean diurnal patterns..................................................................................................................... 20

3.3.2 Wet-bulb temperature .................................................................................................................... 22

3.3.3 Critical hours.................................................................................................................................... 23

3.4 BUFFERING.............................................................................................................................................. 24

3.4.1 Sensible water .............................................................................................................................. 25

3.4.2 Latent .............................................................................................................................................. 25

3.5 DISCUSSION............................................................................................................................................. 26

4 ECONOMIC APPLICABILITY ..................................................................................................................... 27

4.1.1 Investment ....................................................................................................................................... 27

4.1.2 Exploitation ..................................................................................................................................... 274.1.3 Conclusion ....................................................................................................................................... 28

5 CONCLUSION ......................................................................................................................................... 29

5.1 GENERAL ................................................................................................................................................ 29

5.2 RECOMMENDATIONS................................................................................................................................. 29

REFERENCES ................................................................................................................................................... 31

APPENDICES ................................................................................................................................................... 32

APPENDIX I:MOLLIER DIAGRAM ............................................................................................................................... 33

APPENDIX II:FREQUENCY OF OCCURRENCE OF WET-BULB TEMPERATURES ........................................................................ 35

APPENDIX III:TECHNICAL DATA GEAPOLACEL ............................................................................................................ 36APPENDIX IV:PART OF ASHRAE FUNDAMENTALS ABOUT THE HEAT BALANCE AND PERFORMANCE OF COOLING TOWERS.......... 38


5/44

1 Introduction1.1 GeneralCooling is getting increasingly important in buildings. In 2008, 71% of the office buildings in theNetherlands had a cooling system. [9] Multiple factors contribute to this trend [10]:

Changing building styles: modern architectural styles poorly adapted to the prevailingclimate often do not allow for efficient solar and thermal control e.g. fully glazed buildings.

Increasing internal loads: increasing use of computers and related equipment leads to highercooling loads as all supplied energy is converted into heat.

Increase of ambient temperature: in densely built urban environments the ambienttemperature rises above the temperature outside these areas. This phenomenon is called

heat island effect and causes a higher cooling load. The changing climate might also be

contributing this effect.

Increasing comfort requirements: as cooling is becoming more and more common therequirements of the building users also tend to be more demanding.

Improved building insulation: The improvements in insulation (construction and glazing) ofbuildings cause the buildings to be better able to retain heat inside. In a winter situation this

is an improvement because it reduces heat loss which means less heat has to be supplied to

the building. In a summer situation the heat needs to be transported from inside to outside

to cool the building, insulation impedes this process. However, insulation also impedes the

energy flow from outside in. Whether the insulation is increasing or decreasing cooling load

depends on the specific building characteristics.

A common way to meet this rise in cooling demand is to increase the capacity of the (active) cooling

system. Disadvantages of this method are:

Increased (fossil) energy usage. Larger environmental impact (materials, refrigerant and emissions). Higher costs (investment and exploitation).

To limit or avert the disadvantages of an active cooling system, passive cooling systems could be

used. In this report passive cooling system refers to a cooling system which uses considerably less

energy than conventional systems or none at all. The relatively small amount of energy that some

passive systems use could possibly be generated by means of sustainable energy sources i.e.

photovoltaic, wind, etc. Photovoltaic seems the obvious choice as the highest gain occurs roughly

when the highest cooling power is needed. A passive cooling system achieves this low energy usage

by making optimal use of natural heat sinks and optimising heat exchange. The main principles of

passive cooling are [11]:

Evaporative cooling: lowering the temperature of air by adding water to it. In this processsensible heat from the air is transformed into latent heat.

Aquifer: a (seasonal) buffer of water between layers of clay in the earths crust. This long-term energy storage makes it possible to use the heat of the summer in the winter and the


6/44

cold from the winter in the summer. This type of passive cooling is particularly profitable if

the usable floor area of the building(s) is over 6.000 m2. [9]

Ventilation:o Diurnal ventilative cooling: directly using ambient air if its temperature is lower than

the indoor temperature. Often this is also a side-effect of the ventilation needed to

guarantee indoor air quality.

o Nocturnal ventilative cooling: the ambient temperature during the night is lowerthan during the day. Nocturnal ventilative cooling uses this low temperature to cool

down the building by ventilating it with this air, extracting heat from the building,

using the building as a buffer for daytime heat load.

Radiant cooling: using long wave radiant heat loss to the dark sky during the night. Soil cooling: a few meters below the surface the soil has a constant temperature (10-12oC)

which in the Netherlands is lower than the required indoor air temperature. This lower

temperature can be used as a heat sink for the heat extracted from the building.

1.2 GoalFollowing the general introduction it is interesting to know the (technical) feasibility of the

implementation of passive cooling in a building in the Netherlands. The main focus will be

evaporative cooling with cooling towers for office buildings.

The main research question of this project is: Is passive cooling with cooling towers applicable and,

technically and economically feasible in the Netherlands?.

Sub-questions, derived from the main research question are:

1. What is the working principle of cooling towers?2. What is the efficiency/potential of a cooling tower?

a. What are the limiting factors?b. How does potential change in time on daily and yearly basis?

3. What conditions need to be met for application in the Netherlands?4. What are the main points of consideration for application in the Netherlands?

1.3 MethodWith a literature study on the topics of passive cooling, cooling towers and evaporative cooling, the

working principle of cooling towers will be investigated. From this investigation the efficiency limiting

factors and conditions for application will arise. These factors and conditions combined with basic

physics will give an insight in the theoretical potential and applicability in the Netherlands.

Using manufacturers data combined with climate data, the practical applicability and feasibility of

cooling towers in the Netherlands will be determined.

1.4 Report outlineIn chapter 2, the working principle of cooling towers is explained. First (2.1) the general concept of a

cooling tower is described and the different variants are introduced and compared with each other.

Next (2.2), the cooling potential of a cooling tower is investigated and important variables are

identified. These variables, among others, define the conditions for application (2.3). In the next


7/44

paragraph (2.4) the actual performance is determined using manufacturers data for an exemplar

cooling tower.

Chapter 3 describes the actual applicability of cooling towers in the Netherlands. In the first

paragraph of this chapter (3.1) the influence of location on the applicability is explained. The next

paragraph (3.2) the factor time is investigated in relation to applicability. In 3.3, Dutch weatherdata is examined to determine the temporal patterns of performance and critical moments. In the

last paragraph (3.4), a coarse estimation is made on economic applicability in the Netherlands.

In chapter 4 a coarse investigation is made for the economic applicability of cooling towers for

building cooling. Investment costs are estimated and together with the findings about the

performance in chapter 3 a conclusion is drawn.

In the last chapter, chapter 5 the conclusions of this report are drawn. Also included in chapter 5 are

recommendations for further research which have arose during this research.


8/44

2 The working principle of cooling towersCooling towers are devices which remove heat from a medium and reject this heat to the

atmosphere. The most widely known type cooling tower is the hyperbolic cooling tower as they are

common near power plants, refineries and chemical plants. Using a similar principle, cooling towers

for HVAC purposes exist to reject the heat of the chiller (active) or to directly cool the cooling circuit

(passive). The latter type will be the main focus of this report.

2.1 GeneralIn the following paragraphs the working principle of cooling towers will be explained. In appendix IV

a part of ASHRAE fundamentals [6] on cooling towers is included which explains the heat balance and

performance of cooling towers. The general working principle of a cooling tower is as follows:

1. Relatively cool ambient air and hot cooling water enter the tower.2. (Water is evaporated to cool down the air)3. The air is heated by the heat from the cooling water.4. Relatively hot (and moist) air and cool cooling water leave the tower.

Two kinds of temperature are used in the following paragraphs, dry-bulb and wet-bulb temperature.

The dry-bulb temperature is the actual measured temperature. The wet-bulb temperature is the

temperature the air would reach if it would be moistened to its maximum vapour content (saturation

pressure/temperature). Figure 2 illustrates the relation between dry- and wet-bulb temperature in a

cut-out of the Mollier diagram. This diagram describes the relation between different properties of

air e.g. temperature, humidity, enthalpy etc. The complete diagram can be found in appendix I.

Figure 2: Mollier diagram with dry- and wetbulb temperature

[Source: http://w3.bwk.tue.nl/nl/unit_bps/]

Figure 1: Hyperbolic cooling towers (left) and a typical cooling tower for HVAC purposes (right)[Source: http://www.wordpress.com/]


9/44

Figure 3: Cross-section of an open

loop (top) and a closed loop

(bottom) wet cooling tower.[Altered, original source:

http://www.betterbricks.com/]

2.1.1 Heat transfer methodsThe heat transfer between cooling water and the ambient air can occur in three ways:

1. The air entering the tower can directly be used to cooldown the cooling circuit. The lowest temperature that can

be theoretically achieved is close to the dry-bulbtemperature of the ambient air. This type of cooler is

called a Dry Cooler. The cooling air is separated from the

cooling circuit by a heat exchanger, usually a bundle of

pipes with fins.

2. In an open loop wet cooling tower the cooling circuit iscooled to a temperature below the dry-bulb temperature,

theoretically to a temperature near the wet-bulb

temperature. To achieve this lower temperature water is

evaporated. With this evaporation the sensible heat of the

air is decreased and the latent heat increased. The coolant

(water) itself is partially evaporated by injecting it in the

flow of ambient air in the cooling tower. The amount of

extra cooling power which can be achieved with

evaporation depends on the humidity of the ambient air

i.e. humid air can absorb less moisture than dry air. Note

that cooling water and transport medium are the same in

this type of cooler.

3. A hybrid type of the two types above exists and is called aclosed loop wet cooling tower. Two sub-types exist:

a. Internal exchanger: The coolant flows through apipe bundle in the cooling tower which is covered

in water. This way the coolant circuit is closed but

evaporation does occur i.e. the benefits of both dry and wet cooling combined.

b. External heat exchanger: The coolant flows through a heat exchanger outside of thecooling tower which is cooled by the water coming from the cooling tower.

Of course due to the extra heat exchange step the performance of these types of coolers is a

bit lower than that of a wet cooling tower. Because of this extra step this type is also called

an indirect cooling tower.

All types use fill in the cooling tower to increase the heat exchanging

surface of the water with the airflow. It also increases the turbulence of

the air and distributes the water, both increasing performance. Most

modern cooling towers use plastic as fill material. The fill for a certain

cooling tower must be chosen in such a way that it provides the highest

exchanging surface (m2/m3) without clogging. This depends on the

quality (amount of contamination) of the water which is used.Figure 4: Plastic fill material[Source: http://www.spec-net.com/]


10/44

2.1.2 AirflowTo generate the required airflow, 3 principles exist:

In a Natural draft cooling tower the airflow is created by using buoyancy. The densitydifferences of a natural draft cooling tower occur because warm moist air has a lower density

than relatively dry and cool air. To create a sufficient pressure difference the height of thecooling tower needs to be high enough.

In a Mechanicaldraft cooling tower the airflow is created by a fan. Two subtypes exist:o Forced draft; in this type the air is forced through the cooling tower by a fan. This

method causes a relatively high entry velocity and a low exit velocity.

o Induced draft; in this type the air is pulled through the cooling tower by the fan. Thismethod causes a low entry velocity and a high exit velocity. The high exit velocity

reduces the risk of recirculation in comparison with a forced draft cooling tower.

A hybrid type of the two types above exists and is called Fan assisted natural draft. As thename suggests the natural airflow is assisted by a fan with this type.

Counterflow and crossflow

All these types can exist in counterflow, crossflow or intermediates. This describes the direction of

the airflow compared to the flow of water. In general the counterflow type can achieve better

efficiencies. The main advantages of crossflow types are a lower noise production and a smaller

installation height. This means that the crossflow type is easier to install out of sight, making it more

suitable for application for space cooling. In figure 5 these principles are illustrated.

The reason for the difference in performance between the two types is explained with basic heat

exchanger theory. The air entering the counterflow cooling tower (Twb1) has the lowest wet-bulb

temperature just after the intake and heats up along the height of the cooling tower to (Twb2). The

water entering the counterflow cooling tower (Tw1), is the hottest near the top of the cooling tower

and cools down along the height of the cooling tower to (Tw2). This means that the coolest air comes

in contact with the coolest water, thereby achieving the lowest possible output water temperature

(Tw2). In figure 6 on the next page this is illustrated in the left half.

Figure 5: Schematic view of mechanical draft cooling towers in counter (left) and cross flow (right) [8]


11/44

In the right half ofFigure 6, the principle of crossflow is depicted. In contrast to counterflow, the out

flowing (Tw2) water does not come into contact with fresh air (Twb1) but with air which already has

been moistened in other parts of the cooling tower. For example in Figure 6, the air flowing through

part 18 of the fill has already passed part 16 and 17. This decreases the cooling potential of the airand thus the exit temperature. The effect of this is that there is a temperature gradient horizontally

as well as vertically, causing the mixed exit temperature of the water to be higher than that of a

comparable (volume and type of fill) counterflow cooling tower.

2.2 Cooling potentialThe cooling potential of a cooling tower depends on the climate in which it is used. In general the

cooling potential of evaporative (wet and hybrid) cooling towers is higher than the potential of a dry

cooler.

2.2.1 Dry coolerA dry cooler is an air to water heat exchanger. Heat flow from the water to the ambient air depends

on the total heat exchanging surface (A), the total heat transfer coefficient (U) and the temperature

difference between the water and the ambient air. A common formula to describe the heat transfer

in a heat exchanger is as follows:

= , , ( , , )ln , ,

( , , )

Both A and U are type specific i.e. depending on the model. Hence, the performance of a certain dry

cooler is directly related to the ambient dry bulb temperature: the cooler the ambient air, the larger

Figure 6: Schematic representation of the fill in a cross- and counterflow cooling tower [4](Twb = Wet-bulb temperature, Tw = water temperature)


12/44

the cooling potential. Because there always has to be a temperature difference to exchange heat, the

ambient temperature needs to be below the desired cooling water temperature. In turn, the cooling

water temperature needs to be below the desired temperature of supply air of the building to be

able to cool this air. These former statements imply that a dry cooler can only be used if the ambient

air has a dry bulb temperature several degrees lower than the temperature of set point of the supply

air.

2.2.2 Evaporative coolersAs mentioned before, evaporative coolers use the principle of evaporation to increase the cooling

capacity compared to a dry cooler. Because a hybrid cooler is basically a dry cooler in which water is

evaporated, the potential of a hybrid cooler is at least as high as that of that of a comparable (same

U and A) dry-cooler. According to Costelloe and Finn (2007), temperate and maritime European

regions have a considerable unrealised potential for evaporative cooling with high temperature

building cooling systems like cooling ceilings and displacement ventilation [12].

The cooling potential of an evaporative cooling tower (wet and hybrid) depends on the wet-bulbtemperature. The wet-bulb temperature depends on the relative humidity of the air i.e. the dryer the

air, the lower the wet-bulb temperature for a given dry-bulb temperature. This makes the potential

of an evaporative cooler directly related to the humidity of the ambient air. Because the temperature

that can be achieved using evaporation is lower than the dry-bulb temperature, an evaporative

cooling tower can be used at higher ambient temperatures compared to dry-coolers.

Although in theory both hybrid and wet cooling towers can achieve the same cooling water

temperature there is a difference in practice. Because in a wet cooling tower the actual water from

the cooling system is evaporated (partly), there is a very high contact area between the water and

the air. This makes it relatively easy for the air to reach its wet-bulb temperature. The hybrid coolingtower on the other hand has another heat exchange step over which a temperature difference must

exist. This means that the cooling water temperature will always be higher than the temperature of

the moistened air, the wet-bulb temperature. In practice this temperature difference is around 2 to 3

Kelvin depending on the type (size) of heat exchanger and cooling tower [7, 10].

2.2.3 Technical comparisonThe main weakness for both dry and evaporative coolers in the Netherlands is the fact that when the

building has a high cooling demand (in summer), the cooler has a relatively low cooling potential and

vice versa. This means that for space cooling, only in spring and autumn a dry cooler could be used

directly. In summer, when the cooling demand is high, the cooling potential of a dry cooler is too low.In the winter the cooling potential of a dry cooler is highest but the demand is (almost) zero. Of

course dry-coolers could be used directly for purposes with a high cooling demand not in the summer

like data centres. This type of application goes beyond the scope of this report. Because of the low

cooling potential in the summer, the dry cooler will not be investigated any further in this report.

Using the principle of evaporation, the potential of a cooling tower increases and might be sufficient

to be used for space conditioning. Because evaporative cooling is theoretically more efficient, the

focus will be on evaporative cooling in this report.


13/44

2.3 Conditions for applicationTraditionally office buildings are ventilated with an all-air ventilation system. In this setting all

ventilation air is centrally conditioned i.e. heated, cooled and (de)humidified. Typically these systems

require a cooling temperature between 8-12oC. These low temperatures make it possible to

dehumidify using the central chiller. For the application of evaporative cooling towers a few

conditions must be met:

A low latent cooling load; High temperature cooling; Low ventilation rates.

In modern office buildings these conditions can be met. With the high insulation standard, low air

leakage and high efficiency ventilation applied in modern office buildings, the cooling load is often

dominated by the internally generated sensible load [7]. This sensible load can often be met with

high temperature cooling devices. The latent load, if present, is met in the central air handling unit

and will be left out of consideration in this report.

2.3.1 High temperature cooling (HTC)The applicability of evaporative cooling depends on two temperatures: the achievable temperature

of the cooling water and the temperature of the cooling device at which enough heat is extracted

from the indoor air i.e. the required temperature. Both temperatures are of course time dependent

as is the climate. In the ideal case the cooling water temperature is always lower than the required

temperature. As described before, the cooling water temperature depends on the climate. The

required temperature depends on the type of cooling device used and the set point temperature of

the room which is being cooled.

High temperature cooling devices such as chilled ceilings or concrete core conditioning, can achieve

enough cooling power at smaller temperature differences between cooling water and room

temperature in comparison with traditional less efficient techniques. This is mainly achieved by using

a large heat exchanging surface. The usage of such high temperature cooling devices is essential for

applying evaporative cooling during periods of high cooling demand [7]. Several studies show that

chilled ceilings can maintain reasonable thermal comfort with a supply temperature maximum of

18oC [7, 13]. Faco et al state that even supply temperatures from 18-20oC are sufficient in some

cases [13]. This depends on the requirements for the indoor climate i.e. the comfort boundaries.

2.3.2

VentilationTraditionally, the ventilation air of a building serves two purposes: supply of clean air and transport

of energy (heating and cooling). In combined air-water systems these two goals are separated to a

great extend.

Air only systems

In a traditional building the heating and cooling energy is transported by the ventilation air. The

amount of transported energy is often normative for the ventilation rate i.e. the amount of air

needed to transport the energy is higher than the amount needed to maintain a reasonable indoor

air quality (IAQ). The main reason for this is that air is a poor energy transporting medium. The low

heat capacity (cp = 1.000 [J/kg.K]) and low density ( = 1 [kg/m

3

]) of air imply a volumetric heatcapacity of (cp*) 1.000 [J/m

3.K]. This low capacity causes the air temperature to change rapidly


14/44

when heat is added or removed. For performance and comfort reasons the air supply temperature is

bound to a certain bandwidth. This bandwidth combined with the low volumetric heat capacity of air

causes a higher ventilation rate than needed for indoor air quality.

Air-water systems

In air-water systems both water and air are supplied to condition the building spaces. This way,maintaining IAQ and energy transport are separated; the air is used to provide clean air and the

water is used to transport the energy. The preference of water above air for this purpose follows

from the volumetric heat capacity of water with a density of 1.000 [kg/m3] and a heat capacity of

4.200 [J/kg.K]: 4.200.000 [J/m3.K]. In comparison with air this capacity is 4.200 times higher.

With the use of high temperature cooling the ventilation rate can be lowered in many cases because

the air is not needed as a transport medium for cooling (and heating) purposes. This reduction in

ventilation airflow reduces the energy consumption for ventilation and the size of ventilation

ductwork. [14]

Ventilation efficiency

By increasing the ventilation efficiency the ventilation can be lowered, lowering the cooling load. An

example of high efficiency ventilation is displacement ventilation. This principle eliminates the mixing

caused by other ventilation strategies. In the ideal case of this principle the new air pushes the older

air out of the ventilated space and relatively warm air pollutants from occupants etc. rise up due to

buoyancy. This ensures that the air near the occupant is as fresh as possible and the extracted air has

a relatively high pollutant concentration compared to mixed ventilation, making the ventilation more

efficient.

Figure 7: Rough principle of displacement ventilation[Adapted, original source: http://www.aivc.org/]


15/44

2.4 Performance of a cooling towerTo be able to compare a cooling tower with other cooling techniques, the performance must be

known for different circumstances. In this paragraph the coefficient of performance (COP) will be

calculated for different wet-bulb temperatures. For these calculations, data from a cooling tower

from GEA Polacel will be used. The cooling tower used is of the open type. The technical data for this

cooling tower has been obtained from the GEA PolaSelections website [1]. Because a counterflow

cooling tower has the highest efficiency, this type will be investigated.

2.4.1 Coefficient of performanceIn general sense, the energetic efficiency of any device can be determined by dividing the amount of

useful energy delivered, by the energy used to generate it. The amount of useful energy of a cooling

tower is its cooling power expressed in watt (W). This cooling power can be calculated with the

following formula:

.

=

(

) [W]

With: Pc.tower: Cooling power of the cooling tower [W]

v: Flow of water [m3/s]

water: Density of water [kg/m3]

cwater: Specific heat capacity of water [J/kg.K]

Tin: Temperature of the water entering the cooling tower [K]

Tout: Temperature of the water leaving the cooling tower [K]

The energy used to generate this cooling power can be divided in two parts: the energy used to

power fan and the energy used to power the water pump. The main contributor to the total is the

energy to power the fan. To save energy, the fan speed is reduced when the maximum cooling power

is not needed. Traditionally this was realised with a two-speed motor. Modern cooling towers use a

frequency controller to reduce the speed which is needed to match the capacity to the heat load. The

lowest achievable speed is about 40% of the maximum speed of the motor. In figure 8 the relation

between power and fan speed is shown for a single-speed, two-speed and variable speed (frequency

controlled) motor. The hatched parts indicate the power savings. Note that due to the parabolic

curve, for example at 40% of the maximum fan speed the fan uses only 10% of the energy.

The rated power of the motor is type specific and can be found in the technical data of the assessed

cooling tower. The second contributor to the energy usage of a cooling tower is the energy to power

Figure 8: Fan speed vs. fan power [6]


16/44

the water pump. This pump is used to keep the fill wet by circulating the water through the cooling

tower. The power needed for the water pump will be calculated with the following formula:

= ( +( )+ ) [W]With: v: Flow of water [m3/s]

pnozzle: Pressure difference over the nozzles of the cooling tower [Pa]

g: Gravitational constant [m/s2]

h: Height from inlet to the base [m]

Ppipe: Pressure difference over the pipe transporting the water [Pa]

from the base to the inlet

pump: Total efficiency of the water pump [-]

The capacity of the water pump cannot be controlled like the fan because the water distribution

nozzles are designed for a certain pressure. Lowering the water flow/pressure will cause the fill to be

inefficiently wetted and it can cause channelling i.e. the water is not distributed over the fill but the

water flows in channels through the fill seriously lowering performance of the cooling tower.

One contributor to the total has been left out of consideration: the energy used to keep the water in

the reservoir at the bottom ice-free. This contributor is ignored because the heating element in the

water reservoir is only used in the winter to start-up the cooling tower. Once it is running is can keep

itself ice-free. In addition to this, the cooling tower will probably not be used for space-cooling in the

winter with exception of spaces with an exceptionally high internal heat load e.g. server rooms and

theatres.

The three factors mentioned before combine into the formula which is used to calculate the overall

COP of a cooling tower:

= = ( ) + [-]With: COP: Coefficient of performance of the cooling tower [-]

Pfan: Power input of the fan [W]

Ppump: Power input of the water pump [W]

This formula will be used in the following paragraphs to calculate the performance of a cooling tower

for different wet-bulb temperatures.

2.4.2 CalculationAs an example for a counter flow cooling tower the model CMC4-DL-90-PS3/3 of GEA Polacel will beused. According to the technical data, this model can deliver a cooling power of 81 kW with a water

temperature entering at 20oC and exiting at 18

oC at a wet-bulb temperature of 16

oC. This model has

an extra efficient electromotor for a 2% extra investment compared to the standard electromotor on

this cooling tower. It is assumed that this extra investment will pay off.

The following table shows the performance of the cooling tower described before, at different web-

bulb temperatures with a constant inlet temperature of 20oC and a constant air velocity in the

cooling tower.


17/44

Table 1: Performance of the cooling tower at hand with a constant inlet temperature, constant outlet temperature,

constant air velocity and descending wet-bulb temperature [1]

Wet bulb

temp (C)

Inlet

temp.

Outlet

temp (C)

Cooling

range (C)

Approach

(C)

Cooling

cap. (kW)

Air vel.

(m/s)

Water

load

(m3/m2h)

0 20 11.4 8.6 11.4 323.3 2.5 11.3

1 20 11.7 8.3 10.7 310.9 2.5 11.3

2 20 12 8.0 10 298.2 2.5 11.3

3 20 12.4 7.6 9.4 285.2 2.5 11.3

4 20 12.7 7.3 8.7 271.9 2.5 11.3

5 20 13.1 6.9 8.1 258.3 2.5 11.3

6 20 13.5 6.5 7.5 244.3 2.5 11.3

7 20 13.9 6.1 6.9 230 2.5 11.3

8 20 14.3 5.7 6.3 215.3 2.5 11.3

9 20 14.7 5.3 5.7 200.3 2.5 11.3

10 20 15.1 4.9 5.1 184.7 2.5 11.311 20 15.5 4.5 4.5 168.7 2.5 11.3

12 20 15.9 4.1 3.9 152.2 2.5 11.3

13 20 16.4 3.6 3.4 135.2 2.5 11.3

14 20 16.9 3.1 2.9 117.7 2.5 11.3

15 20 17.3 2.7 2.3 99.6 2.5 11.3

16 20 17.8 2.2 1.8 81.0 2.5 11.3

The data shows that at design wet-bulb temperature the cooling capacity is 81 kW. With descending

wet-bulb temperature, this capacity increases. Because the air velocity in the cooling tower is kept

constant, the power for the fan is also constant at 1.6 kW [1]. For the calculation of the pump power,

the height (h) is estimated to be 2.5 meters and the pump efficiency 65%. According to the

documentation the water flow is 32 m3/h and the nozzle pressure must be between 14.3 and 48 kPa,

a value of 30 kPa is used for the calculation.

Using the formula stated in the former paragraph, the power of the pump is estimated:

= +( )+ =(

32

3600)(30 000+(10009.812.5)+2300)

0.65= 777 [W]

Using the equation from the former paragraph the total efficiency of the cooling tower can now be

determined for a wet-bulb temperature of 16

o

C:

= = ( ) + = 81.01.6 + 0.777 = 34.1 [-]With this formula, the COP of the cooling tower can be defined per wet-bulb temperature with a

constant air flow. Because the cooling power of the cooling tower increases with a lower WBT and

the input energy remains the same, the COP of the cooling tower will also increase with a decreasing

WBT. In the following table the calculated efficiency of the cooling tower at hand is given for a wet-

bulb temperature between -2 and 16oC.


18/44

Figure 9: Air velocity to meet design parameters with

changing WBT [1]

Table 2: Efficiency of the cooling tower at full fan power

WBT

[oC]

-2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

COP 146.5 141.3 136.0 130.8 125.5 120.0 114.4 108.7 102.8 96.8 90.6 84.3 77.7 71.0 64.0 56.9 49.5 41.9 34.1

As mentioned before, the capacity of the

cooling tower is lowered by reducing the fan

speed when full power is not needed. The

following graph from the manufacturers

data shows the fan speed for different wet-

bulb temperatures with the same conditions

as before i.e. 81 kW, cooling from 20 to 18oC.

With the graphs in figure 8 and figure 9 the

fan power can be obtained when the fan

speed is reduced. In the following table the

fan speed, fan power and COP per WBT for

81 kW cooling power at design temperatures.

Table 3: Performance of the cooling tower at hand with constant cooling capacity, constant outlet temperature and

variable fan power

WBT

[oC]

Cooling power

[kW]

Air velocity

[m/s]

Fan power

[% of max.]

Fan power

[W]

C.O.P.

[-]

0 80.9 0.5 4% 160*

86.4

1 80.9 0.5 4% 160* 86.4

2 80.9 0.5 4% 160* 86.4

3 80.9 0.5 4% 160

*

86.44 80.9 0.6 5% 160* 86.4

5 80.9 0.6 5% 160* 86.4

6 80.9 0.6 5% 160*

86.4

7 80.9 0.7 6% 160* 86.4

8 80.9 0.7 6% 160* 86.4

9 80.9 0.8 7% 160*

86.4

10 80.9 0.9 9% 160* 86.4

11 80.9 1.0 11% 176 85.0

12 80.9 1.1 13% 208 82.2

13 80.9 1.3 20% 320 73.814 80.9 1.5 28% 448 66.1

15 80.9 1.8 45% 720 54.1

16 80.9 2.4 100% 1600 34.1*

Actual fan power is 10% because of the 40% speed minimum of a frequency controlled fan

As stated before, the fan power cannot be lower than 10% of the maximum power. For this reason

the fan power of the cases with a WBT lower than 11oC is 10%/160W. In these cases the tower can

be switched off for an amount of time. To prevent damage to the cooling tower the heat capacity

must be designed in such a way that the cooling tower is not switched on and off too often. The

cooling tower could also be operated longer than needed, but care has to be taken to prevent

condensation on the heat dissipation system (HTC).


19/44

3 Applicability of evaporative coolingThe applicability of evaporative cooling depends on several factors. In this chapter these factors will

be discussed and from this discussion the applicability in the Netherlands will be determined.

3.1 Spatial variationsBecause the cooling potential of cooling towers depends mainly on the climate, the applicability of

cooling towers depends on its geographical position.

A common way to measure applicability of evaporative cooling towers is to calculate the amount of

degree hours of wet-bulb temperature depression (WBTD). The WBTD is the difference between dry-

and wet-bulb temperature. As discussed before

this difference is the cooling potential for

evaporative cooling. The WBTD degree hours

are calculated by integrating the wet-bulb

temperature depression over time per hour. In

figure 10 a graphical representation of the

degree hours of WBTD are given for a summer

period in Europe.

Although this is a common way to measure the

potential, it should be used with care. For

instance, the figure shows that the central part

of Spain has a high potential for evaporative

cooling. This is true but the temperatures are also high, making the cooling power less suitable for

space cooling. This process is illustrated in a cut-out of the Mollier diagram in figure 11. In this figure

it is shown that the resulting wet-bulb temperature is high although the wet-bulb temperature

depression is also high (11K). Depending on comfort criteria additional cooling might still be required.

Note that the actual cooling temperature is higher than the indicated 24oC (see paragraph 2.2.1).

Figure 11: Mollier diagram illustrating relation between WBTD and applicability[Source: http://w3.bwk.tue.nl/nl/unit_bps/]

Figure 10: Degree hours of wet-bulb temperature

depression over the summer period in Europe [3]

The Netherlands


20/44

3.2 Temporal variationsThe potential for evaporative cooling does not only vary with place but also in time. This potential is

affected by numerous factors e.g. wind direction, humidity of the surrounding area, solar irradiation,

radiation of the soil (net radiation), etc. The latter has a particular large effect on the temperature

variation. The daily minimum and maximum wet-bulb temperatures typically coincide with the

minimum and maximum dry-bulb temperatures. In general the daily range of the wet-bulb

temperature is about one-third of that of the dry-bulb temperature. [10] This is a very coarse

estimate and of course this difference is constantly fluctuating with factors described earlier.

To determine the applicability, these fluctuations

must be known because the maximum indoor

temperature may not be surpassed too often,

depending on the comfort criteria. The lowest

wet-bulb temperatures occur just before sunrise

and the highest in the afternoon. [10]

The seasonal variations also have an effect on the

potential for evaporative cooling. As mentioned

before the seasonal cooling demand for space

conditioning of a typical building varies inversely

with the evaporative cooling potential during the

year. This means that the peak cooling demand

occurs when the evaporative cooling potential is

the lowest in the seasonal pattern and vice versa.

Figure 12 shows the expected cooling water

temperatures of an indirect cooling tower for

Dublin and Milan with an assumed difference

between WBT and outlet temperature of 3K

during the day and in the months that a cooling demand could be expected, based on TMY (typical

meteorological year) data. [7] These two cities were chosen as representatives for a Northern and

Southern city. The results for all north-west European, temperate climates are expected to be

somewhere between the results in this figure. [7]

3.3 Applicability in the NetherlandsIn the former paragraphs it is shown that the cooling potential for evaporative cooling towers

depends on temporal and spatial variations. To be able to analyse the application of evaporative

cooling with cooling towers, it must be known what the temporal variation in cooling availability is

for a certain location. For this research, the applicability in the Netherlands will be investigated. As

reference data for the Netherlands, data from the weather station in de Bilt will be used.

3.3.1 Mean diurnal patternsThe graph on the next page illustrates the mean diurnal wet-bulb temperature in the period 1981-

2000 in de Bilt in the summer.

Figure 12: Possible cooling water temperatures in

Dublin (top) and Milan (bottom) from May to October

with an assumed difference of 3K between cooling

water and wet-bulb temperature derived from TRY-data


21/44

Figure 13: Diurnal patterns of the mean wet-bulb temperature in the period 1981 2000 in de Bilt from May to October


22/44

In line with the former paragraphs, the graph in figure 13 shows that July and August are the months

in which the highest mean wet-bulb temperatures occur. This graph also shows that on average the

peak of the wet-bulb temperature occurs in the morning. Especially in July and August the maximum

of the mean wet-bulb temperature occurs early in the morning between 07:00 and 09:00 AM. This

means that at constant supply temperature the potential of the cooling tower starts increasing from

this point in time onwards.

This phenomenon is further illustrated in the next graph in which the mean DBT, WBT and the

difference between these two are depicted for August.

The maximum value of the difference between mean wet- and dry-bulb temperature (DBT-WBT),

indicating the potential of the cooling tower (WBTD), occurs around 02.00PM. This point in time

coincides with the moment at which the dry-bulb temperature, a (very) coarse indicator for the

cooling load, is highest. This graph also agrees roughly with the statement from (Santamouris, 2007)

that the range of the WBT is about one third of the range of the DBT. The graph disagrees with the

statement of (Santamouris, 2007) that the daily minimum and maximum wet-bulb temperatures

typically coincide with respectively the minimum and maximum dry-bulb temperatures.

3.3.2 Wet-bulb temperatureIn the table below the frequency of occurrence of different wet-bulb temperatures in de Bilt is

given, averaged over 20 years (1981-2000). The complete table can be found in appendix II.

Table 4: Occurrence of wet-bulb temperatures in 'de Bilt' averaged over 20 years (1981-2000)[15]

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC TOT TOT CUMM CUMM

(hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (%) (hrs) (%)

24 - 25 0 0 0% 0 0%

23 - 24 1 0 0 2 0% 2 0%

22 - 23 0 2 1 0 3 0% 6 0%

21 - 22 0 1 5 3 2 11 0% 16 0%

20 - 21 0 5 11 6 2 25 0% 41 0%

Figure 14: Diurnal patterns of the mean WBT, DBT and difference between these two for August in the period 1981-2000 in de Bilt from May to October


23/44

19 - 20 1 9 18 14 5 47 1% 88 1%

18 - 19 0 1 17 30 25 12 0 85 1% 173 2%

17 - 18 0 4 24 44 45 20 4 141 2% 313 4%

16 - 17 1 11 36 64 78 34 9 232 3% 546 6%

15 - 16 2 16 53 84 98 61 19 0 333 4% 878 10%

14 - 15 0 3 23 68 102 115 72 31 1 415 5% 1293 15%

From this table follows that on average the wet-bulb temperature exceeds 18 oC (the desired cooling

water temperature for HTC) for 173 hours per year. This means that it is not possible to rely

completely on evaporative cooling directly without the use of buffering and/or flexible comfort

boundaries.

As mentioned before, to be able to exchange heat a temperature difference must exist between

cooling water and wet-bulb temperature. In the potential calculation in paragraph 2.4 a difference of

2K was used as example. If a temperature difference of 2K over the cooling tower is taken into

account, the desired wet-bulb temperature would be 16oC for HTC and this temperature would beexceeded 546 hours per year with the most of these hours occurring in August (78).

It also follows from the table that throughout an average year the wet-bulb temperature is always

lower than 24oC. July and August are the months in which the highest wet-bulb temperatures occur.

This observation corresponds with the findings ofCostelloe and Finn [7] for a Northern city.

3.3.3 Critical hoursIn the former paragraph the mean temperatures per hour were examined. This is useful information

to get an insight in the average cooling potential and profiles. To be able to assess the applicability of

cooling towers for full passive cooling of an office building it is also useful to know if, when and forhow long the cooling power is insufficient.

In the Netherlands the climate data of 1994 is commonly used as a reference year for cooling power.

[16] In the following graph the amount of hours is displayed the threshold of 16oC for the WBT is

exceeded. The total amount of hours the WBT exceeded 16oC was 476 in 1994, 217 of these excesses

occurred during office hours (08:00 18:00).

Figure 15: Frequency in hours at which the WBT exceeds 16oC during office-hours


24/44

Figure 17: Excess of the threshold of 16oC between July 30

thand

August 1st

1994

As mentioned earlier not only the amount of hours the threshold of 16oC is exceeded is interesting

but also the degree to which this occurs consecutively i.e. if the threshold is surpassed for several

consecutive hours the thermal comfort could be jeopardised. The graph above (figure 16) illustrates

the moments at which the 16oC threshold

was exceeded during 1994. From this

graph follows that the threshold of 16oC is

exceeded several times and the duration

of these excesses range from 1 hour to 51

hours. The excess of 51 consecutive hours,

the longest that year, occurred during July

30th and August 1st. In figure 17 this period

is displayed in more detail. It is clearlyvisible that the threshold is exceeded for

almost 3 consecutive days. Note that it is

not certain that the full power is needed during periods of excess i.e. a WBT slightly higher than 16 oC

could still be sufficient if the demand is lower than the maximum capacity of the cooling tower at

16oC.

3.4 BufferingThe former paragraph illustrates the need for a cooling buffer, flexible comfort boundaries or a

combination of both. Apart from the need for a buffer because it is needed lower the exceeding

hours, a buffer has other advantages:

Increase in performance; because the WBT is generally lower during the night, the overallperformance of the cooling tower will increase when the buffer is loaded during the night.

Size reduction; because of the increase in performance and the fact that the cooling towerhas more time to cool down the water (18:00-08:00 compared to 08:00-18:00) the size of the

cooling tower could be designed smaller, lowering the investment for the cooling tower.

Night tariff; in the Netherlands (and many other countries) the electricity is cheaper duringthe night because the overall demand is lower. Loading the buffer during the night could be

cheaper than using the cooling tower during the day, depending on the storage efficiency

(the ratio between energy output and input).

Figure 16: Excess of the threshold in time during 1994


25/44

In this paragraph the main principles of thermal energy storage will be investigated. These principles

can be divided into sensible and latent storage. [17] To get an idea of the storage volume this will be

calculated for one suitable material per principle. Storage efficiency is not taken into account in this

rough estimation.

The buffer size will be calculated for the duration of 1 workday (08:00-18:00; 10 hours) and anequivalent energy storage of 2/3 of the maximum power of 81kW continuously.

With these assumptions the total energy to be stored for 1 day is equal to:

= 2 3 3600 = 2 3 81 10 3600 = 1.94 1063.4.1 Sensible waterIn sensible heat storage, thermal energy is stored by raising (or lowering) the temperature of a

certain buffer material. Water appears to be the most suitable material for sensible heat storage for

this purpose because it is de medium which is being used by the cooling tower and no additional heat

exchange would have to take place. Apart from this, water has a high specific heat capacity and is

inexpensive compared to its alternatives e.g. rock, concrete, oils, organic liquids, etc. [17] If the

building is made of concrete this could also be used as a buffer, combining building structure and

building services in one element.

To account for heat-loss during storage, the temperature of the stored water has to be lower than

the required temperature for cooling. The dimensions of the water-buffer will be calculated with a

storage temperature of 18oC, it is assumed that the actual temperature will be lower because of the

higher performance during the night, accounting for the heat-loss.

The specific heat capacity of water at 18oC is approximately 4.183 kJ/kg.K and its density is

approximately 998.3kg/m3

at this temperature. These figures combined with a temperature

difference of 2K over the cooling system result in a volumetric heat capacity of the water buffer of

8,351.8 kJ/m3. The total volume of the buffer with this material would be (1.94*106 / 8,351.8)

232.3 m3.

3.4.2 LatentLatent heat storage is based on the fact that a material needs a relatively large amount of energy to

undergo a phase-change. This change occurs at a the phase transition temperature, depending on

the material. Above and below this temperature the material usually behaves as sensible heat

storage.

Many materials exist which are suitable for latent heat storage at different temperatures ranging

from 10oC to 150oC. These materials can be divided in organic materials (paraffins and non-paraffins),

inorganic materials (salt-hydrates and metallics) and eutectics. As example a paraffin will be used

with 16 carbon atoms. This paraffin has a melting point of 16.7oC, a latent heat of fusion of 237.1

kJ/kg and a density of 773 kg/m3. The volumetric heat capacity follows from these values and is

1.83*105kJ/m3. The total volume of the buffer with this material would be 10.6 m3.


26/44

3.5 DiscussionLooking at the wet-bulb temperature depression (WBTD) the Netherlands has a medium (350-450

hours of WBT depression) potential for evaporative cooling compared to other European countries

50-900 hours of WBT depression).

The wet-bulb temperature profiles derived from the climate data in de Bilt between 1981 and 2000

correspond with the findings ofCostelloe and Finn [7] for a Northern European city: the highest wet-

bulb temperatures occur in July and August, the range of the WBT is roughly one-third of the range

of the DBT, and the maximum of the averaged WBT is 14oC. This last finding implies that on average

the WBT is sufficiently low to provide in the total cooling demand which requires a maximum

temperature of 16oC to be able to extract the heat from the building.

Despite of the former statement there are moments during the year when the cooling temperature is

not low enough i.e. greater than 16oC. It is assumed that these moments coincide with the moments

on which the maximum (design) capacity is needed, causing overheating. On average, between 1981

and 2000, the maximum temperature is exceeded for 546 hours. In 1994, a commonly used

reference year for cooling capacity, this number was 476 of which 217 were during office hours

(08:00-18:00). A further assessment of the climate data of 1994 has revealed that the threshold of

16oC is exceeded several times during that year and the duration of these excesses range from 1 hour

to 51 hours. This last statement illustrates the need for a cooling buffer, flexible comfort boundaries

or a combination of both when applying cooling towers in the Netherlands.

A buffer to overcome the full period of 3 days excess is undesirable, mainly for economic reasons

(see 4.1.1). This means that the solution to the excesses will always be flexible comfort boundaries,

possibly combined with a cooling buffer. When a buffer would be used to bridge one day, its size

would depend on the used principle: sensible storage volume would be approximately 232.3 m3 and

latent storage volume would be 10.6 m3. These volumes are likely to be even greater because of a

limited storage efficiency. Because the size of the buffer is inversely proportional to the temperature

difference, the size could be lowered if a lower buffer temperature could be achieved. Due to the

generally lower WBT during the night, it is possible that a smaller buffer is sufficient for 1 day.


27/44

Table 5: Specific price of cooling sources for a heat pump [2]

Source Specific price[/kWth]

100 kW 500 kW

Ventilation air 350 320

Aquifer (heating and cooling) 900 400Groundwater 850 350

Surface water 400 400

Soil collector 500 500

Open air 350 320

4 Economic applicabilityIn the former paragraphs the technical applicability of cooling towers has been discussed. Because

cooling towers are technically at least applicable in the Netherlands for base-cooling of office

buildings the economic side of this should also be reviewed. This determines the chances for cooling

towers to be applied in practice. The following review is by no means intended as a full economic

analysis but as an indication of the position of cooling towers in relation to other cooling principles.

The costs associated with a cooling tower are the investment and exploitation costs.

4.1.1 InvestmentAs example of the investment for a cooling tower, the tower from the counterflow example ( 2.4.2)

will be used to calculate the cost per kW capacity. This cooling tower costs 16.000 and has a

capacity of 81 kW at given

conditions (20oC inlet, 18

oC outlet

and 16oC WBT). The specific price

of a cooling tower in this range

follows from these figures:

approximately 197 /kW.

Additional investments would be

needed to make the cooling

system completely sustainable e.g. for photovoltaic electricity generation. In table 5 the specific

prices are given for alternative cooling sources in combination with a heat pump.

The price of a heat pump is not included in the price and would be approximately 200 /kW [2]. Heat

pumps also need high temperature cooling and low temperature heating dissipation systems to run

efficiently. This makes the comparison independent of the dissipation systems e.g. concrete core

conditioning and chilled ceilings.

A sensible water buffer costs approximately 63.41 /m3 (surface mounted, steel). [18] The total price

for the sensible buffer as calculated in paragraph 3.4.1 would be 180 /kW installed cooling power (1

euro = 1.25 dollar). A latent (paraffin) buffer system costs approximately 106.13-112.82 /kWh. [19]

The total price for the latent buffer as calculated in paragraph 3.4.2 would be between 706 and 750

euro per kW installed cooling power.

4.1.2 ExploitationThe exploitation costs consist of the energy cost, the cost of the water which is used and the cost for

maintenance. The energy cost is directly related to the performance of the cooling tower when

comparing to alternatives. This performance, a COP of approximately 35, is considerably better than

that of a heat pump with a cooling COP of 3-6. When a buffer is used which can be loaded during the

night, this could be beneficial for the total energy cost because of the lower night tariff. This would

also increase the overall investment.

The water costs increase the exploitation cost of a cooling tower compared to other cooling

techniques. The counter flow cooling tower used as example in 2.4.2 uses approximately 0.1 m3/h

on average due to evaporation loss. With a price of 0.73 per m [www.brabantwater.nl, d.d.

28/03/10] and an assumed amount of operation hours of 2.000 the total price for suppletion water

for evaporation on yearly basis is 147. Additional water is needed to keep the water in the cooling

tower clean as the concentration of minerals and contamination rises due to the evaporation losses.


28/44

The cooling tower needs to be serviced once a year. During maintenance the cooling tower is cleaned

to maintain performance and prevent corrosion, algae and chalk deposition. Also the oil needs to be

replaced in the fan reducer and the fan is checked for imbalance. The actual cost of this yearly

maintenance is not known. A conventional system with a dry-cooler/condenser also needs to be

serviced every year [6].

4.1.3 ConclusionFrom table 5 follows that the investment for the cooling tower of 81 kW is considerably lower than

alternatives in the same power range. When including a sensible buffer the investment will still be

lower than direct alternatives (~380 /kW). The application of a latent buffer makes the total (~925

/kW) investment higher than direct alternatives. As mentioned earlier in this report (1.1), an

aquifer system is only economical for larger buildings (>6,000m2). This trend is also visible in table 5

as the specific price drops from 900 to 400 /kWth,source when going from 100kW to 500kW cooling

power. This price is still considerably larger than the price for a cooling tower but it can also be used

for heating, making the comparison less obvious.

The cost of the water used to compensate evaporation losses in the cooling tower is very small in

comparison with the investment. The cost for maintenance is comparable to that of a conventional

cooling system. The main advantage of a cooling tower is its performance/COP. This performance

makes the running costs of the system significantly lower compared with its alternatives.

From this coarse analysis follows that a cooling tower for passive cooling is at least economically

feasible in the lower range of cooling power. A complete analysis of actual investment and running

costs compared to its alternatives should be made to justify application.


29/44

5 Conclusion5.1 GeneralThe application of passive cooling with a cooling tower (without the use of a chiller) is technically

feasible in the Netherlands. To be able to extract enough heat from a building the cooling water

temperature should be 18oC at most. A difference of at least 2K between outlet temperature and

wet-bulb temperature should exist to exchange the heat with the ambient air in the cooling tower.

This makes the required wet-bulb temperature 16oC. The technical feasibility follows from the fact

that the average wet-bulb temperature in the Netherlands is lower than this limit during the cooling

season. This corresponds to the findings ofCostelloe and Finn 2002. [7]

When the minimum WBT of 16oC is not exceeded a cooling tower is a very efficient means of cooling

with a COP of 35. As cooling method for base cooling, a cooling tower in the Netherlands could be

used continuously. As cooling method to meet the full cooling load in an office building, flexible

comfort boundaries should be applied to bridge the hours the availability of sufficiently cool water is

too low. The indoor temperature will rise in this case. In the commonly used reference year 1994 this

occurred for 217 hours during office hours. A buffer could be used to lower the amount of hours the

outlet temperature is too high. A point of concern is the size of a buffer for 1 day: approximately 232

m3

for a sensible buffer and 10.6 m3

for a latent buffer.

The application of passive cooling with a cooling tower (without the use of a chiller and buffer) is also

economically feasible in the Netherlands, at least at cooling capacities of around 100 kW. This follows

from a coarse evaluation of the investment and exploitation costs in comparison with alternative

cooling sources with a heat pump in this capacity range. Including a sensible (water) buffer does not

change this statement. A latent (paraffin) buffer increases the investment per kW by 375%,

depending on the comfort criteria this investment could still be feasible if a sensible buffer is not

possible due to its size.

5.2 RecommendationsThis report is the result of an analysis of the applicability of cooling towers in the Netherlands. During

this research a few assumptions have been made and aspects have been left out of consideration to

restrict the extent of the research. Following are a few topics which could be investigated for further

research:

In this report data from a manufacturer has been used to calculate the COP of a certaincooling tower at different WBTs. To be certain about the performance of cooling towers the

results should be compared to actual measurements and possibly compared to other

manufacturers i.e. check the validity of the manufacturers data.

This report focuses solely on the application of cooling. When designing a new building in theDutch climate, heating and cooling should be treated at the same time to generate an

optimum design. It would be interesting to know what the effect would be on the

applicability of cooling towers in the Netherlands.

Because the difference between WBT and inlet temperature determine the performance of acooling tower, the use of a cooling tower could be extra beneficial when applied in places

with a high cooling load in the winter too e.g. ICT rooms, data centres and theatres.


30/44

Cooling towers produce noise when in operation. This could pose problems in urban areas,especially during the night. Many solutions exist to limit the amount of noise a cooling tower

makes. It would be interesting to know to what extend these measures work and if the

resulting sound pressure is sufficiently low.

Cooling towers have been in the news in a negative way because of the risk of Legionnairesdisease. Because of the low temperatures when applying cooling towers for direct building

cooling the risk is relatively low. According to the manufacturer the risk of Legionnaires

disease can be averted with proper maintenance.

From the literature study in this report follows that a wet cooling tower is more efficientthan, or at least as efficient as a dry-cooler. This raises the question why dry-coolers are used

most commonly. This might be because of deficient general knowledge (maintenance cost,

occupied space, performance, etc.) on the topic. Bad publicity of poorly maintained cooling

towers could also have had an effect on this.


31/44

References1. GEA Polacel. GEA PolaSelections. [cited 2010; Available from:

http://www.polacel.nl/polaselections.

2. SenterNovem, Cijfers en tabellen, in Kompas, energiebewust wonen en werken. 2007.

3. Molina, S.Y.E.E.J.L., Roof Cooling Techniques: A design handbook. 2006: Earthscan.

4. GEA Polacel, Koeltorenboekje, in 'Koeltorens'.5. Notitie: Afstuderen bij DWA installatie- energieadvies. 2009, DWA installatie- en

energieadvies.

6. ASHRAE: American Society of Heating Refrigerating and Air-Conditioning Engineers, 2008

ASHRAE handbook heating, ventilating, and air-conditioning systems and equipment. 2008,

ASHRAE: Atlanta, Ga. p. 1 v. (various pagings).

7. Costelloe, B. and B. Finn, Indirect evaporative cooling potential in air-water systems in

temperate climates. Energy and Buildings, 2002. 35: p. 573-591.

8. Jin, G.-Y., et al., A simplified modeling of mechanical cooling tower for control and

optimization of HVAC systems. Energy Conversion and Management, 2007. 48(2): p. 355-365.

9. SenterNovem, Energiedata 2008. 2008, Ministerie van Volkshuisvesting, Ruimtelijke

Ordening en Milieubeheer.10. Advances in Passive Cooling. Earthscan, ed. M. Santamouris. 2007: Earthscan.

11. Givoni, B., Indoor temperature reduction by passive cooling systems. Solar Energy, 2009. In

Press, Corrected Proof.

12. Costelloe, B. and D. Finn, Thermal effectiveness characteristics of low approach indirect

cooling systems in buildings. Energy and Buildings, 2007. 39.

13. Faco, J. and A.C. Oliveira, Thermal behaviour of closed wet cooling towers for use with

chilled ceilings. Applied Thermal Engineering, 2000. 20(13): p. 1225-1236.

14. Klimaatplafonds / Koelconvectoren. ISSO Publicatie 48. 1998.

15. KNMI klimaatdata en -advies. [cited 2010 January 6, 2010]; Available from:

http://www.knmi.nl/klimatologie.

16. Hensen, J., Referentiejaar of beter een langere referentieperiode?TVVL Magazine, 2005.17. Sharma, A., et al., Review on thermal energy storage with phase change materials and

applications. Renewable and Sustainable Energy Reviews, 2009. 13(2): p. 318-345.

18. Crowther & Partnes, L. and I. George Reeves Associates, Stratified Chilled-Water storage

Design Guide, in EM-4852. 1988, Electric Power Research Institute: California.

19. He, B. and F. Setterwall, Technical grade paraffin waxes as phase change materials for cool

thermal storage and cool storage systems capital cost estimation. Energy Conversion and

Management, 2002. 43(13): p. 1709-1723.
http://www.polacel.nl/polaselectionshttp://www.polacel.nl/polaselectionshttp://www.knmi.nl/klimatologiehttp://www.knmi.nl/klimatologiehttp://www.knmi.nl/klimatologiehttp://www.polacel.nl/polaselections


32/44

Appendices

I: Mollier diagram

II: Occurrence of wet-bulb temperatures in de Bilt averaged over 20 years (1981-2000)

III: Technical data GEA Polacel - CMC4-DL-90-PS3/3

IV: Part of ASHRAE fundamentals about the heat balance and performance of cooling towers


33/44

Appendix I: Mollier diagram


34/44


35/44

Appendix II: Frequency of occurrence of wet-bulb temperatures

Measured in de Bilt averaged over 20 years (1981-2000)[15]

JAN FEB MRT APR MEI JUN JUL AUG SEP OKT NOV DEC TOT TOT CUMM CUM

(hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (hrs) (%) (hrs) (%

24 - 25 0 0 0% 0 0

23 - 24 1 0 0 2 0% 2 0

22 - 23 0 2 1 0 3 0% 6 0

21 - 22 0 1 5 3 2 11 0% 16 0

20 - 21 0 5 11 6 2 25 0% 41 0

19 - 20 1 9 18 14 5 47 1% 88 1

18 - 19 0 1 17 30 25 12 0 85 1% 173 2

17 - 18 0 4 24 44 45 20 4 141 2% 313 4

16 - 17 1 11 36 64 78 34 9 232 3% 546 6

15 - 16 2 16 53 84 98 61 19 0 333 4% 878 10

14 - 15 0 3 23 68 102 115 72 31 1 415 5% 1293 15

13 - 14 1 6 39 92 107 116 89 45 3 0 499 6% 1793 20

12 - 13 2 13 57 101 103 94 96 58 8 1 533 6% 2326 27

11 - 12 1 1 4 24 75 90 76 63 89 67 18 4 511 6% 2837 32

10 - 11 1 3 9 42 94 83 46 41 77 68 32 11 509 6% 3346 38

9 - 10 9 8 18 52 110 59 26 21 51 74 35 18 482 5% 3827 44

8 - 9 18 17 31 64 95 42 13 12 43 80 55 29 499 6% 4326 49

7 - 8 29 30 44 72 72 20 5 6 26 74 58 38 473 5% 4799 55

6 - 7 43 46 64 81 57 10 4 3 18 63 71 51 509 6% 5308 61

5 - 6 58 50 84 82 35 6 2 2 11 51 82 58 521 6% 5829 66

4 - 5 61 57 86 76 21 3 0 0 6 34 69 63 476 5% 6306 72

3 - 4 74 66 77 67 16 2 0 4 25 69 66 465 5% 6771 77

2 - 3 72 61 69 51 9 1 2 18 61 59 403 5% 7174 82

1 - 2 67 60 64 38 5 0 0 12 47 59 353 4% 7527 86

0 - 1 70 61 58 21 1 0 0 6 33 59 311 4% 7837 89

-1 - 0 55 56 45 14 1 0 2 28 56 258 3% 8095 92

-2 - -1 39 51 33 6 1 1 21 38 190 2% 8284 95

-3 - -2 28 39 21 2 0 0 14 30 136 2% 8420 96

-4 - -3 22 23 15 1 0 7 27 95 1% 8515 97

-5 - -4 20 16 9 0 4 20 68 1% 8583 98

-6 - -5 19 10 5 0 3 14 51 1% 8634 98

-7 - -6 15 8 3 1 13 39 0% 8673 99

-8 - -7 14 6 2 0 10 32 0% 8704 99

-9 - -8 11 4 1 0 8 24 0% 8728 10

-10 - -9 7 2 0 5 14 0% 8742 10

-11 - -10 5 1 0 4 11 0% 8753 10

-12 - -11 3 1 0 1 5 0% 8758 10

-13 - -12 2 1 0 0 3 0% 8761 10

-14 - -13 2 0 0 0 3 0% 8764 10

-15 - -14 1 0 1 0% 8765 10

-16 - -15 1 0 1 0% 8766 10

-17 - -16 0 0 0% 8766 10-18 - -17 0 0 0% 8767 10


36/44

Appendix III: Technical data GEA Polacel

CMC4-DL-90-PS3/3 (Counterflow cooling tower)


37/44

Datum : 21-1-2010 11:27:33

Ons projectnummer : 1000

Uw referentie : Test5 | TU Eindhoven

Pagina : 1 / 2

TECHNISCHE SPECIFICATIE PER KOELTOREN

Type : CMC4-DL-90-PS3/3

Fabrikaat : GEA Polacel Cooling Towers B.V.

Soort toren : Tegenstroom

Aantal cellen : 1

Kleur : Grijs met blauwe waterbak

Ontwerpgegevens

Koelcapaciteit : 75.0 kW

Waterhoeveelheid : 32.0 m3/h

Temperatuur in : 20.0 C

Temperatuur uit : 18.0 C

Natte bol temperatuur : 16.0 C

Max. toelaatbare watertemperatuur : 55.0 C

Hogere toelaatbare temperaturen zijn in overleg mogelijk.

Technische Specificatie

Hoogte koelpakket : 900.0 mm.

Benodigde sproeierdruk : 14.3 kPa, max. 49 kPa

Luchthoeveelheid per ventilator : 7.0 m3/s

Externe statische druk : 0.0 Pa

Aantal ventilatoren : 1Aantal motoren : 1

Genstalleerd vermogen per motor : 2.0 kW

Opgenomen vermogen per motor : 1.6 kW

Toerental elektromotor : 1500.0 min-1

Toerental ventilator : 430.0 min-1

Soort aandrijving : Indirect d.m.v. motorreductor

Aansluitspanning elektromotor : 400V-3Ph-50Hz

Beschermings- Isolatieklasse motor : IP 65 F

Vermogen verwarmingselement (optie) : 2.0 kW

Aansluitspanning verwarmingselement : 230V-1Ph-50Hz

Afmetingen en gewichten

Lengte : 1760.0 mm.

Breedte : 1760.0 mm.

Hoogte : 3410.0 mm.

Leeggewicht koeltoren : 700.0 kg

Bedrijfsgewicht koeltoren : 1410.0 kg


38/44

Datum : 21-1-2010 11:27:33

Ons projectnummer : 1000

Uw referentie : Test5 | TU Eindhoven

Pagina : 2 / 2

Waterverbruik

Gemiddeld verdampingsverlies : 0.1 m3/h

De maximale verdamping, welke de suppletie bepaalt, kan 25% hoger zijn.

Spuiverlies : Afhankelijk van de waterkwaliteit

Geluidsvermogen koeltoren

Het totale geluidsvermogen (PWL) van de koeltoren wordt bepaald door sommatie van de verschillendedeelbronnen: ventilatoruitblaas en luchtinlaat. Het geluid aan de luchtinlaat is een gevolg van het in dewateropvangbak vallende koelwater. Dit geluid kan gedempt worden met door Polacel ontwikkelde drijvendempers. De opgave is conform DIN 45635 Sub 46, klasse 3. De geluidsdruk op afstand kan wordenberekend met behulp van de tabel Conversie geluidsvermogen naar geluidsdruk tegenstroom.PWL - uitblaas ventilator : 85.7 dB(A)

PWL - luchtinlaat zonder dempers : 87.3 dB(A)

PWL - koeltoren totaal zonder dempers : 89.6 dB(A)

De toepassing van drijvende dempers reduceert het geluid aan de luchtinlaat.

PWL - luchtinlaat met dempers : 76.8 dB(A)

PWL - koeltoren totaal met dempers : 86.3 dB(A)

Materiaal specificatie

Ventilatorbladen/ type : Aluminium / 05-06-24L/B3TR Stock NL

Ventilatorframe : Thermisch verzinkt staal

Koelpakket/ type : Pp / 12 mm.Ondersteuning koelpakket : RVS 304/ glasvezelversterkt polyester

Druppelvanger/ type : Standaard PVC

Ondersteuning druppelvanger : RVS 304

Frame koeltoren : RVS 304

Omkasting koeltoren : Glasvezelversterkt polyester

Wateropvangbak : Glasvezelversterkt polyester

Ondersteuning wateropvangbak : Thermisch verzinkt staal

Luchtinlaatjaloezien : Glasvezelversterkt polyester

Sproeisysteem/ type sproeier : PPVogelrooster : Geen

Toren uitgerust met:

Warmwateraansluiting (F1)100

Koudwateraansluiting (F2) DN150

Aftap en overloop

Suppletie-afsluiter met vlotter (max. 250 kPa / 36,25 PSI)


39/44

Appendix IV: Part of ASHRAE fundamentals about the heat balance and

performance of cooling towers


40/44

Cooling Towers 39.1

general operating characteristics described. (See specific manufac-turers data for maximum accuracy when planning for test or criticaltemperature needs.)

A cooling tower selected for a specified design condition willoperate at other temperature levels when the ambient temperature isoff-design or when heat load or flow rate varies from the design con-dition. When flow rate is held constant, range falls as heat load falls,causing temperature levels to fall to a closer approach. Hot- and cold-water temperatures fall when the ambient wet bulb falls at constant

heat load, range, and flow rate. As water loading to a particular towerfalls at constant ambient wet bulb and range, the tower cools thewater to a lower temperature level or closer approach to the wet bulb.

COOLING TOWER THERMAL PERFORMANCE

Three basic alternatives are available to a purchaser/designerseeking assurance that a cooling tower will perform as specified:(1) certification of performance by an independent third party suchas the Cooling Technology Institute (CTI), (2) an acceptance test

performed at the site after the unit is installed, or (3) a performancebond. Codes and standards that pertain to performance certificationand field testing of cooling towers are listed inChapter 51.

Certification. The thermal performance of many commerciallyavailable cooling tower lines, both open- and closed-circuit, is cer-

tified by CTI in accordance with theirStandardSTD-201, whichapplies to mechanical-draft, open- and closed-circuit water coolingtowers. It is based on entering wet-bulb temperature and certifiestower performance when operating in an open, unrestricted environ-ment. Independent performance certification eliminates the need forfield acceptance tests and performance bonds.

Field Acceptance Test. As an alternative to certification, towerperformance can be verified after installation by conducting a fieldac

Report M3 Rick Van Pruissen 20100603

Documents

Transcript of Report M3 Rick Van Pruissen 20100603