6- Improving the Efficiency of Natural Draft Cooling Towers -6

Page 1

www.elsevier.com/locate/enconman

Energy Conversion and Management 47 (2006) 1086–1100

Improving the efficiency of natural draft cooling towers

J. Smrekar, J. Oman *, B. Sirok

Faculty of Mechanical Engineering, Askerceva 6, SI-1000 Ljubljana, Slovenia

Received 24 January 2005; accepted 28 July 2005Available online 12 September 2005

Abstract

This study shows how the efficiency of a natural draft cooling tower can be improved by optimising the heattransfer along the cooling tower (CT) packing using a suitable water distribution across the plane area of thecooling tower. On the basis of cooling airmeasurements, it is possible to distribute the water in such away thatit approaches the optimal local water/air mass flow ratio and ensures the homogeneity of the heat transfer anda reduction of entropy generation, thus minimising the amount of exergy lost. The velocity and temperaturefields of the air flow were measured with the aid of a remote control mobile robot unit that was developed toenable measurements at an arbitrary point above the spray zone over the entire plane area of the coolingtower. The topological structures of themoist air velocity profiles and the temperature profiles above the sprayzone were used as input data for calculation of the local entropy generation in the tower. On the basis of themeasured boundary conditions, a numerical analysis of the influence of the water distribution across the cool-ing tower�s plane area on entropy generation and exergy destruction in the cooling tower was conducted.� 2005 Elsevier Ltd. All rights reserved.

Keywords: Cooling tower; Entropy generation; Exergy destruction; Cooling tower efficiency; Heat transfer

0196-8904/$ - see front matter � 2005 Elsevier Ltd. All rights reserved.doi:10.1016/j.enconman.2005.07.012

* Corresponding author. Tel.: +386 1 4771 303; fax: +386 1 2518 567.E-mail address: [email protected] (J. Oman).

Page 2

Nomenclature

A condition of airCT cooling towerh specific enthalpy, kJ/kgL height of CT packing, m_m mass flow rate, kg/s_S entropy flow rate, kW/K_E exergy flow rate, kW_Q heat flux, kWs specific entropy, kJ/kg Kt, T temperature, �C, KP power on generator, MWr water/air mass flow ratiow velocity of moist air in cooling tower, m/sv velocity of ambient air in vicinity of cooling tower, m/sW condition of waterx humidity ratio of moist air

Greek letterse effectivenessg efficiency of cooling towerq density of moist air, kg/m3

/ relative humidity

Subscripts0 condition of ambient1 inlet2 outletA, B, C, D measurement points of air condition in vicinity of cooling towera dry airgen generationi ith segment in cooling towerL localm wet bulb temperatureo outlet from cooling towerR heat rejectionopt optimalw waterz condition of air in vicinity of cooling tower

J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100 1087

Page 3

1088 J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100

1. Introduction

The cooling system is one of the most important parts of a power plant. Its task is to extractas little heat as possible from the thermodynamic cycle to the environment, thereby ensuringimproved efficiency of the power plant. Estimating the operating efficiency of a cooling systemusually involves the coefficient _QR=P [1]. A reduction of this coefficient means that more energyfrom the fuel was successfully transformed into useful work, which means that less heat was lostto the environment. Being able to reduce the coefficient _QR=P depends on the integral and differ-ential characteristics of the circulating water system in the plant. It is important to remember thatbecause we are dealing with large energetic flows in power plants, small improvements to the cool-ing system can mean large fuel savings and a reduction in the amount of pollution produced bythe exhaust gases.

In this study, the main element of the cooling system is a natural draft cooling tower, which, inphysical terms, is a relatively simple device. Although the basic parts of the tower can easily bedescribed [6] and understood, the processes of heat and mass transfer are very complex. In naturaldraft cooling towers, a process of counter flow heat transfer, which the water is cooled by air,takes place. Between the water and the air, a boundary layer is established, which is consideredto be saturated air at the same temperature as the water. In the heat transfer process more thantwo-thirds of the heat is transferred by evaporation, with the rest being transferred by convection.A mathematical description of the variables that are changing along the cooling tower (CT) pack-ing and in the spray zone is very complicated. It could be described by a system of ordinary dif-ferential equations [10,14]. However, an analysis of the heat and mass transfer in cooling towerscan also be described empirically with the NTU method, providing the construction characteris-tics of the tower are known [11–13,16,17].

Good operating conditions of the cooling tower mean that there is a homogeneous heat transferacross the entire plane area of the tower. Homogeneity is shown by a uniform temperature field ofthe air above the spray zone across the plane area. Possible anomalies can be due to the construc-tion characteristics of the water distribution system, the impassability of the CT packing andnozzles, or the influence of the surrounding air on the inflowing air velocity to the cooling tower.Usually, these anomalies are not known a priori, but they can normally be determined with asuitable measurement system. In order to do this, a variety of diagnostic methods to determinethe effectiveness of the heat transfer [4,5] that take into account the cooling tower�s operationalirregularities were developed. The temperature and the velocity field of the air are closely con-nected as was explained in Ref. [2] with an analogy to the convective transfer of mass, heatand momentum.

An improved cooling tower performance is the result of an optimum mass flow rate of coolingwater with respect to the power plant�s operating conditions [9]. For this kind of operation, pumpswith a variable speed, which is unusual for today�s cooling systems with large water mass flowrates, are required.

Besides eliminating local anomalies in the temperature and velocity fields, it is also possible toimprove conditions with a proper distribution of water across the cooling tower�s plane area. It isthis distribution of water that is analysed here. The result is a reduction of entropy generation andan improvement in the cooling tower�s efficiency, which also results in an increased efficiency ofthe power plant. The analysis considers a constant total mass flow rate of cooling water, which

Page 4

J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100 1089

is optimally distributed according to the air flow conditions over which we have no influence withnatural draft cooling towers.

2. Experimental section

2.1. Description of the measurements and the measuring equipment used in the cooling tower

The basis of the presented measuring system is a mobile robot unit that moves gradually overthe entire measuring plane. The sensors are mounted on the robot in such a way that the systemfulfils all of today�s standards. A vane anemometer, designed for operation in 100% humidity, anda Pt-100 temperature sensor are used. The details of the equipment and its calibration are de-scribed elsewhere [7]. The duration of a measurement at a single point, depending on the chosenmagnitude of the observation scale (quasi-steady state condition), is 15–30 s. After the measure-ment is performed, the servo positioning system moves to a new pre-chosen measuring point.

The following integral parameters are measured simultaneously according to the DIN 1947standard [3]: the inlet and outlet temperatures of the cooling water, the cooling water flow rate,which is measured with an ultrasonic flow meter, and the output power of the thermo-energeticsystem. The air velocity and the temperature in the cooling tower�s measuring plane are measuredcontinuously with stationary sensors. As the velocity and temperature measurements are not per-formed simultaneously, the stationary velocity meter, together with the integral parameters, serveas correction elements for these measurements. Other parameters that cannot be measured by themobile unit (the parameters in and below the CT packing) were also acquired with a stationaryreference point, which is represented as a vertical segment in Fig. 2. Besides the mobile unitand the vertical segment, the measuring system also included devices for data acquisition andanalysis.

The position of the mobile unit was determined by measuring the radial distance and the anglefrom the reference point in the measurement plane. The radial distance and angle were then con-verted to x, y coordinates. The distance was measured using a linear cable extension transducerwith a 50 m range and a typical uncertainty of less than 0.1 m. The angle was measured by a de-vice tracking the position of the cable and rotating the cable extension transducer. An angularresolution of about 0.15� was achieved. The results of the measurements and their analysis arepresented in a report [15].

2.2. Measurements of the ambient air parameters and the exit air mapping resultsfrom the cooling tower

Using the results of the velocity and temperature measurements from the measuring plane ofthe cooling tower, three-dimensional topological structure diagrams of the velocity and tempera-ture data were obtained. These results are shown in Fig. 1 and provide a direct means of evalu-ating the extent to which the falling hot water droplets and the cooling air mix in the coolingtower. Because the air mappings completely characterize the water/air interface across the planearea of the cooling tower, problems with the water distribution and the fill system can be identifiedin the mappings and successfully solved.

Page 5

Fig. 1. Topological structure of the velocity and temperature exit air profile.

1090 J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100

All measurements were performed at a single operating regime, with the power station�s blockoutput power of 350 MW and a cooling water volume flow rate of 180,000 m3/h. The followingambient air parameters were measured simultaneously: the ambient air velocity at four locations(vA, vB, vC, vD), the air temperature in the vicinity of the cooling tower (tz) and the air density inthe vicinity of the cooling tower (q).

In Table 1, the average values of the various parameters for the duration of the analysis arepresented. It is clear that the ambient parameters did not change significantly and, therefore,did not influence the measurement results inside the cooling tower.

Fig. 1 shows the velocity and temperature structures of the exit air acquired with the robotmobile unit. From the left diagram of Fig. 1, one can recognize the local non-homogeneities inthe velocity field, which can be caused by different factors, such as constructional properties(obstacles) and the local fill system impassability. The non-homogeneous temperature field pre-sented in the right diagram of Fig. 1, is closely connected with the non-homogeneous velocity field[8]. Good operation of a cooling tower is defined by homogeneous conditions for heat and masstransfer across the plane area of the tower. The result is a uniform temperature field of the exit airabove the spray zone across the plane area. In general, a non-optimal operating condition isreflected in a deviation of the actual operating point from the expected nominal value.

Table 1Ambient air parameters

Quadrant 1 2 3 4

vA (m/s) 1.8 2.2 2.1 2.3vB (m/s) 2.2 1.9 2.7 2.4vC (m/s) 1.6 2.2 1.8 1.9vD (m/s) 2.2 2 2.1 1.8tz (�C) 9.8 10.2 9.5 10q (kg/m3) 1.23 1.22 1.23 1.23

Page 6

J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100 1091

2.3. Measurements of the temperatures and mass flow rates on a vertical segment of the cooling tower

To determine the basic heat and mass transport properties of the cooling tower, measurementswere made of the aerodynamic and thermodynamic characteristics of the air and water flows in aselected reference vertical segment of the cooling tower, shown in Fig. 2. Measurements on thevertical segment served as a measurement correction element for the non-simultaneous measure-ments made with the robot mobile unit. The vertical segment was also used to determine the char-acteristic function of the analysed cooling tower [4,5] and for a study of the variation of the massflow rate of water on the effectiveness of the heat transfer. From this, it is possible to determine thelocal effectiveness of the cooling tower simply from the local air mapping measurements.

A reference vertical segment of the cooling tower was selected in a region of the cooling towerwhere the geometrical properties are homogeneous.

The vertical segment in Fig. 2 consists of the CT packing located at the bottom, a spray elementin the central part and a drift eliminator in the upper part of the segment. The following param-eters were measured for the experimental system: the inlet moist air temperature ta1, the outlettemperature of the saturated air ta2, the inlet water temperature tw1, the outlet water temperaturetw2, the mass flow rate of the water _mw and the velocity of the moist air wa.

The uncertainty of the measurements made with the Pt-100 thermometers was estimated to beless then 0.25 �C. The air velocity measurements were performed using a pre-calibrated vaneanemometer. The same type of vane anemometer was used for the measurements made withthe mobile robot unit. The sampling frequency was 1 Hz, and the total acquisition time was700 s. The inlet air humidity was determined on the basis of dry and wet bulb thermometers.

Fig. 2. The selected vertical segment of the cooling tower.

Page 7

1092 J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100

All the measurements in the cooling tower were performed according to the DIN 1947 standard[3].

3. The influence of the water/air mass flow ratio on the effectiveness of the heat transfer

The improvement in the heat transfer can be explained by Fig. 3, which shows the situation inthe CT packing, where the counter flow heat exchange from the water to the air is taking place.More than two-thirds [6] of the transferred heat is rejected from the water by evaporation with theremainder being transferred by convection.

Fig. 3 shows the increase in the air�s relative humidity along the CT packing as a result of theevaporative cooling of the water. Air enters at the bottom side of the CT packing and becomessaturated at some height; in the diagram, this happens at section P–P. Above this section, theair would be saturated if the temperature were constant. However, above this section, the processof evaporation is actually still occurring because of the increasing air temperature and, thus, thedecreasing relative humidity of the air. How big a portion is transferred by evaporation dependson the position of section P–P, which means the water/air mass flow ratio.

The physical phenomena in the CT packing can be explained using an example of the measuredvalues in the vertical segment, which are presented in Fig. 4. In the vertical segment, air enters incondition A1 with a relative humidity of 0.82 and a temperature of 10 �C, and saturated air leavesin condition A2 at 31 �C. The mass flow rate of the water is 4.7 kg/s with an inlet temperature of33.9 �C and an outlet temperature of 24.2 �C, as shown in the right diagram in Fig. 3 with pointsW1 and W2, respectively. The mass flow rate of the water was then reduced to 4.2 kg/s. The inlettemperature of the water and the air remained unchanged, while the outlet water temperature

Fig. 3. Condition of the air through the CT packing and the Mollier h–x diagram, which shows the condition of the airthrough a vertical segment for two different mass flow rates of water.

Page 8

Fig. 4. Measured temperatures and velocities of the air and water in a vertical segment for two different mass flow ratesof water.

J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100 1093

decreased to 21.9 �C. At the same time, the outlet temperature of the saturated air did not changesignificantly. This can be easily explained with the equations for energy and mass balance:

_ma dh ¼ _mw dhw þ _ma dxhw. ð1Þ

While the mass flow rate of the water was reduced, the mass flow rate of the air remained approx-imately the same, which means more cooling air per unit of water mass and, consequently, a loweroutlet water temperature. The cause is the higher proportion of air and, thus, a greater differencein the partial pressures of the water vapour between the boundary layer and the air flow along theCT packing.

Under normal operating conditions, the section P–P is at the middle height of the CT packing.This situation suits a process in the CT packing where the water flow is relatively high comparedto the air flow, leading to a high water/air mass flow ratio. In this case, the entering air quicklyapproaches saturation, and the rest of the CT packing does not cause such intensive evaporation,which increases the proportion of the convective heat transfer. This means that it has not reachedthe optimal moistening of the CT packing. The analysed example is presented by the full curvebetween points A1 and A2 in the Mollier h–x diagram in Fig. 3. We can conclude that this is aless desirable process of heat transfer in cooling towers.

In the second case, when the mass flow rate of water was decreased, and the air flow rate re-mained approximately the same, we obtained a better water/air mass flow ratio, the consequence

Page 9

1094 J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100

of which was a drop in the temperature of the outlet water. Thus, section P–P in Fig. 3 moved upthe CT packing. This situation suits a process where the water flow is relatively small in compar-ison to the air flow, which means a lower water/air mass flow ratio. Because of this smaller water/air mass flow ratio, the air is in a nearly saturated state in the upper part of the CT packing, whichmeans intensive water cooling by evaporation, although convective heat transfer is also present. Inthis way, the height of the packing is better used. In the Mollier h–x diagram in Fig. 3, this exam-ple suits the broken line between A1 and A2, which intersects the saturation line later (at a highersection of the CT packing height) than the full line. In this case, the water is cooled to 21.9 �C(state W 0

2), which is 2.3 �C less than in the normal operating regime, where the air is near satura-tion at a lower height in the CT packing. This water/air mass flow ratio is more suitable from theheat transfer point of view. The positions of the lines in the Mollier h–x diagram in Fig. 3 werecalculated from Refs. [11,17].

Because the task of the cooling tower is to cool the water as much as possible, the consequenceof an improved heat transfer effectiveness is a bigger drop in the temperature of the output water,which is shown in Figs. 4 and 5. The effectiveness of the heat transfer on the water side is calcu-lated using the following equation:

e ¼ hw1 � hw2hw1 � hwm

; ð2Þ

where hwm is the specific enthalpy of the water evaluated from the wet bulb temperature of theentering air, which is the minimum temperature to which the water can be cooled. The effective-ness of the heat transfer is improved by about 8% with a 0.51 kg/s lower mass flow rate of water,which is equivalent to 10.8%.

Fig. 5. Heat transfer effectiveness of vertical segment by decreasing the water mass flow rate.

Page 10

J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100 1095

3.1. Improved heat transfer by proper distribution of water across the plane areaof the cooling tower

The analysed cooling tower had the same type of nozzles across the plane area of the coolingtower. The water distribution system with flow channels is exposed to atmospheric pressure,which means that with the correct operation of the nozzles and channels, the water distributionacross the plane area would be uniform. On the basis of the air–velocity and air–temperature dia-grams in Fig. 1, we can determine the position and the level of entropy generation. At the edge ofthe analysed plane area, there is a relatively extensive region of high air velocities and low air tem-peratures. These represent the areas with a high entropy generation. The rounded areas in thevelocity and temperature fields, shown in Fig. 1, that deviate from the expected values are calledareas of anomalies and should, in any case, be additionally investigated [4].

If the ambient differentiations and operational irregularities of the water distribution system,the CT packing and the nozzles are not taken into account, we can focus on improving the oper-ation of the cooling tower. This kind of improvement involves determining the optimal water/airmass flow ratio on a local basis across the plane area of the cooling tower. From the left diagramof Fig. 1, it is clear that there are three relatively distinctive areas with different air velocities. It isalso clear that because of the specific hyperbolic shape of the cooling tower, the air velocity de-creases from the edge of the tower to the interior. Because of the uniform inlet water temperaturesand uniform mass flow rates of water across the plane area and because of the different air massflow rates, the outlet water temperatures are different, which results in increased entropy genera-tion in the cooling tower. With a constant water/air mass flow ratio, the same amount of air perunit of water is ensured, which results in the same outlet water temperatures across the plane areaof the cooling tower and smaller entropy generation. To achieve a constant water/air mass flowratio there has to be a variable mass flow rate of water that is adapted to the mass flow rate of theair across the plane area. Because of a lower total entropy generation, a lower outlet water tem-perature is achieved and, thus, a better overall cooling tower efficiency is ensured. A suitable dis-tribution of water relative to the air flow can be achieved by regulation of the water distributionsystem or by different sizes of nozzles across the plane area of the cooling tower.

3.2. Determination of the optimal water/air mass flow ratio on a local basis for a cooling tower

For a cooling tower, it is possible to determine the optimal water/air mass flow ratio on a localbasis, which, for fixed operational and ambient conditions, achieves the best heat transfer effec-tiveness. The optimal ratio can be calculated from measurements of the local air velocities, towhich the total amount of water has to be adapted so that the water/air mass flow ratio willbe minimised. At the same time, a more constant water/air ratio across the cooling tower�s planearea is achieved by a suitable water distribution, which ensures the homogeneity of the heat trans-fer and gives the best operating conditions.

The sum of the local measured mass flow rates of air is equivalent to the total mass flow rate ofair through the cooling tower:

X

i

_ma;i ¼ _ma. ð3Þ

Page 11

1096 J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100

The portion of the mass flow rate of the air on a local basis is

ki ¼_ma;i

_ma

. ð4Þ

The total amount of cooling water has to be distributed across the plane area of the cooling towerrelative to the mass flow rate of the air. Thus, the local mass flow rate of water has to be propor-tional to the local mass flow rate of air:

_mw;i ¼ ki � _mw; ð5Þ

where _mw is the known total mass flow rate of water. From the local air flow and a suitable waterflow, calculated by Eq. (5), the optimal water/air mass flow ratio is obtained, which is constantalong the entire plane area of the cooling tower:

ropt;i ¼_mw;i

_ma;i¼ const. ð6Þ

Because of the changing ambient conditions, the power plant load and the resulting conditions inthe cooling tower, it is reasonable to calculate the average value of the optimal water/air massflow ratio for a given cooling tower. For the analysed example, this ratio is equal to 1.54 for a180,000 m3/h flow rate of cooling water.

4. Entropy generation and exergy destruction with the optimal water/air mass flow ratio

Fig. 6 shows two diagrams, the first of which represents the water/air mass flow ratio acquiredfrom measured values, while the second diagram represents the entropy generation across theplane area of the cooling tower, calculated with the equation:

_Sgen;i ¼X

out

_m � s�X

in

_m � s. ð7Þ

In Fig. 6, a significant non-uniform distribution of the water/air mass flow ratio can be seen,which has an unfortunate influence on the homogeneity of the heat transfer and the overall effi-ciency of the cooling tower. The scheme in Fig. 3 can be used to explain the reduction in entropygeneration shown in Fig. 7, where we distributed the same amount of water with respect to themass flow rate of air. Fig. 6 shows that the water/air mass flow ratio is very low on the edgeof the cooling tower, which means that we have large amounts of air compared to the amountsof water. It could happen that in some regions, the air does not become saturated along the fillsystem, which represents unused potential for evaporation and, thus, less transferred heat. Inthe interior of the cooling tower, we can see the opposite state, shown in Fig. 6, where thewater/air mass flow ratio is too high. In this case, the condition of the air approaches saturationin the lower part of the CT packing, while the upper part of the CT packing is not so effectivelyused for the process of evaporation. Both examples show non-optimal usage, depending on the airflow distribution across the plane area. Consequently, there are large non-homogeneities, whichcontribute to the generation of entropy.

The analysis showed that we should distribute water from the cooling tower�s interior, where wehave more than enough water relative to the air flow, to the cooling tower�s exterior. We have

Page 12

Fig. 7. Entropy generation from an optimal water/air mass flow ratio across the plane area of the cooling tower.

Fig. 6. Measured relations between water/air mass flow ratio and entropy generation across the plane area of thecooling tower.

J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100 1097

obtained the best possible results for the given operating conditions when we have the optimaldistribution of water across the plane area of the cooling tower. Consequently, we had less entro-py generation on a local basis, shown in Fig. 7. Table 2 shows a comparison of the results for anumerical calculation of the optimal water distribution and the uniform water distribution.

Page 13

Table 2Comparison of the integral parameters from a uniform water distribution across the plane area of the cooling towerwith the optimal water distribution

Uniform water distribution Optimal water distribution

_Sgen (kW/K) 101.4 60.6_Elost (kW) 28,696 17,150tw,o (�C) 24 22.6g (%) 38.9 44.4

1098 J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100

The calculations showed that the results of the water distribution according to air flow are sig-nificantly better. The sum of the local entropy generations gives the total entropy generation ofthe cooling tower, which is 101.4 kW/K. In this case, the measured outlet water temperature fromthe tower is 24.0 �C, and the cooling tower�s efficiency, from Eq. (2), is 38.9%, where the values arerelated to the inlet and outlet water temperatures of the cooling tower, respectively. From thetotal entropy generation of the cooling tower, the lost exergy can be calculated using the Gouy–Stodola theorem:

_Elost ¼ T 0 � _Sgen. ð8Þ

In the case of the measured values, the exergy lost is 28,696 kW.

Fig. 7 shows numerical results when the water/air mass flow ratio is constant across the entireplane area of the cooling tower, and for the given boundary conditions, it is 1.54. In a numericalcalculation of the water and air outlet parameters, as a result of the water distribution, the sameboundary conditions as with the measured values were used, i.e., the same inlet water tempera-ture, the same inlet air conditions and the same air velocity profile across the cooling tower planearea. With the last assumption, we included the anomalies in the air flow that are a consequence ofthe constructional irregularities. This assumption makes it possible to see the difference in thequality of the operation of the cooling tower before and after the water distribution with the sameconstructional conditions. The numerical analyses for the cooling tower were based on algorithmsfrom Ref. [13]. The thermodynamic properties of the water and the moist air were calculated fromRefs. [18,19].

Fig. 7 shows better results in comparison with Fig. 6. A suitable water distribution in terms ofthe air flow conditions results in a homogeneous heat transfer across the plane area of the coolingtower, which gives a lower and more uniform entropy generation. In spite of this, a two times lar-ger entropy generation can be seen on the edge of the cooling tower, which is a consequence of thelarger mass flow rates of water and air.

The entropy generation and lost exergy in this case are 60.6 kW/K and 17,150 kW, respectively,which is 40.2% less than in Fig. 6, where there is a uniform water distribution. As a result of thelower entropy generation, we obtained a lower output water temperature from the cooling tower,which, in the case of the numerical analysis, is 22.6 �C, i.e., 1.4 �C less than with a uniform waterdistribution. The consequence of the water temperature drop is a 5.5% higher cooling tower effi-ciency of 44.4%.

Because of the unchanged total amount of water, this is the result of an improved heat transfer,a lower outlet water temperature from the cooling tower and, thus, also the outlet water temper-ature from the condenser, which dictates the pressure in the condenser.

Page 14

J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100 1099

5. Conclusion

With the aid of a robotized method, measurements of the temperature and velocity fields in acooling tower were performed for the given power plant parameters, cooling tower constructionalcharacteristics and ambient air velocity conditions in the vicinity of the cooling tower. The lasttwo parameters influence the homogeneity of the heat transfer, from which we can see the anom-alies in the cooling tower�s operation. Homogeneity in the heat transfer could not only be achievedwith fault free construction characteristics but also with a proper distribution of water across theplane area of the cooling tower.

In this study, we have analysed the water distribution across the plane area of the coolingtower. We have adjusted the amount of water to suit the air flow conditions, which cannot beinfluenced with natural draft cooling towers. We have found that it is possible to determine theoptimal water/air mass flow ratio for a cooling tower, which has to be as small and as constantas possible across the entire plane area of the cooling tower. In this way, the optimal moistening ofthe CT packing is ensured, which results in a more effective heat transfer. With a constant water/air mass flow ratio, a constant local water outlet temperature is obtained, which decreases the en-tropy generation and the exergy lost from the cooling tower. The result is a lower outlet watertemperature from the cooling tower and, thus, from the condenser, which results in greaterefficiency of the power plant.

References

[1] El-Wakil MM. Power plant technology. New York, St. Louis, San Francisco: McGraw-Hill Book Company;1985 [Chapter 7].

[2] Novak M, Sirok B, Philpott DR, Bullen PR. Dynamic analysis of velocity and temperature fluctuations of coolingtower exit air mapping. In: Proc thermodynamics, heat and mass transfer of refrigeration machines and heatpumps, Seminaire EUROTHERM, No. 59. Nancy, France, 1998. p. 325–31.

[3] Standard DIN 1947. Thermal acceptance tests on wet cooling towers (VDI cooling tower code). 1989.[4] Sirok B, Blagojevic B, Novak M, Hocevar M, Jere F. Energy and mass transfer phenomena in natural draft cooling

towers. Heat Transfer Engrg 2003;24(3):1–10.[5] Hampe E. Cooling towers. Berlin: VEB Verlag fur Bauweses; 1975.[6] Stanford W. Cooling towers-principles and practice. 2nd ed. Carter Industrial Products; 1972.[7] Sirok B, Rus T, Novak M, Vrabic G. Thermal performance pre-upgrade exit air mapping and engineering analysis

of cooling tower performance. In; 60th Annual American Power Conference, Chicago, USA, 1998. p. 282–6.[8] Sirok B, Oman J, Hocevar M, Blagojevic B, Kustrin I. Transient phenomena in cooling tower induced by load

dynamics of power plant. In: 7th International Research/Expert Conference. Barcelona, Spain, 2003.[9] Bosnjakovic F, Knoche KF. Pinch analysis for cooling towers. Energy Convers Manage 1998;39:1745–52.[10] Fisenko SP, Petruchik AI, Solodukhin AD. Evaporative cooling of water in a natural draft cooling tower. Int J

Heat Mass Transfer 2002;45:4683–94.[11] Khan JR, Zubair SM. An improved design and rating analyses of counter flow wet cooling towers. ASME Trans

2001;123(August):770–8.[12] El-Dessouky HTA, Al-Haddad A. A modified analysis of counter flow wet cooling towers. J Heat Transfer

1997;119(August):617–26.[13] VDI-Warmeatlas. Algorithms for calculating heat transfer [Berechnung von Ruckkuhlwerken (Mj)], 9.

Auflage. Berlin: Springer; 2002.[14] Ibrahim GA, Nabhan NBW, Anabtawi MZ. An investigation into a falling film type cooling tower. Int J Refrig

1995;18(8):557–64.

Page 15

1100 J. Smrekar et al. / Energy Conversion and Management 47 (2006) 1086–1100

[15] Hocevar M, Sirok B, Jere F, Novak M. Nuclear Power Station Doel-3, Cooling Tower Measurement Report, No.2742. Ljubljana June 2003.

[16] Simpson WM, Sherwood TK. Performance of small mechanical draft cooling towers. Refrig Engrg 1946;52(6):525–43. 574–6.

[17] Khan JR, Yaqub M, Zubair SM. Performance characteristics of counter flow wet cooling towers. Energy ConversManage 2003;44:2073–91.

[18] Wagner W, Kruse A. Properties of water and steam, the industrial standard IAPWS-IF97 for the thermodynamicproperties and supplementary equations for other properties. Berlin: Springer; 1998.

[19] Nelson HF, Sauer Jr HJ. Thermodynamic properties of moist air: �40 to 400 �C. J Thermophys Heat Transfer2004;18(1):135–42.