Energy and Buildings 41 (2009) 311–319

A scale model to evaluate water evaporation from indoor swimming pools

F. Asdrubali *

Department of Industrial Engineering, University of Perugia, Via G. Duranti 67, 06125 Perugia, Italy

A R T I C L E I N F O

Article history:

Received 11 February 2008

Received in revised form 28 August 2008

Accepted 5 October 2008

Keywords:

Evaporation

Heat loads

Swimming pools

Scale model

A B S T R A C T

The evaluation of water evaporation from indoor swimming pools is a topic of considerable practical

interest, since evaporation may cause the highest energy consumption of the pool plant. A purposely

designed experimental apparatus was used to measure the water evaporation rate from a pool scale

model inserted into a climatic chamber to control environmental conditions. The experimental data were

obtained varying various parameters such as water temperature, air temperature, relative humidity and

air velocity. The results were used to propose a prediction model for water evaporation which was

compared to other methods found in the literature, showing a good agreement.

� 2008 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Energy and Buildings

journal homepage: www.e lsev ier .com/ locate /enbui ld

1. Introduction

The study of water evaporation from indoor swimming poolsis a topic of major interest both for the design of HVAC—heating,ventilation and air-conditioning systems and of water heatingplants: in indoor swimming pools, in fact, the highest thermalload is often due to water evaporation, which thereforerepresents the main source of energy consumption of the entireplant.

A precise evaluation of the evaporation flow rate in these basinsis therefore very important, to the extent that it makes it possibleto correctly design the air conditioning system and to reduceenergy consumption. Furthermore, water evaporation addsmoisture to the building atmosphere and the resulting highhumidity may cause discomfort to the occupants and damage tomaterials by promoting rot and corrosion.

Water evaporation from free surfaces is a function of manyparameters, such as water temperature, air temperature andrelative humidity, air velocity, as well as number and kind ofactivity of the occupants. There are various studies in the literatureconcerning the evaluation of evaporation from indoor pools [1–4];most of them have considered data from real, occupied orunoccupied pools, but there is no evidence of recent studies doneon small-size scale models.

The main objective of this study is to develop data and relationsof general validity to calculate the water evaporation flow fromindoor swimming pools, starting from experimental investigations

* Tel.: +39 0 75 5853716; fax: +39 0 75 5853697.

E-mail address: fasdruba@unipg.it.

0378-7788/$ – see front matter � 2008 Elsevier B.V. All rights reserved.

doi:10.1016/j.enbuild.2008.10.001

in various conditions of temperature, relative humidity and airvelocity.

To this extent, an experimental apparatus was designed at theThermotechnical Labs of the University of Perugia to carry outinitial water evaporation measurements from a scale model of aswimming pool [5,6]; the apparatus, inserted into a climaticchamber, made it possible to control air temperature and relativehumidity but not water temperature. The preliminary resultsencouraged the design and construction of an improved apparatus,which also allowed precise control of water temperature, as in realindoor swimming pools.

Measurement results were used to implement a correlation topredict evaporation and, therefore, to estimate heat loads in indoorswimming pools under different service conditions.

2. Literature review

Many methods for evaluating evaporation from water basinshave been proposed over the years [7–9], although only a few arespecifically related to indoor swimming pools. Some of them werederived from experimental measurements in real pools, othersfrom energy balances of the pool or basin, others from theevaluation of the amount of condensate on the cooling coil of theair-conditioning unit, assuming that this is equal to the amount ofwater evaporated from the pool surface. Shah recently provided asummary of available methods [3,10,11] both for unoccupied andoccupied pools: the two cases have to be discussed separately,since for various reasons evaporation is higher in occupied pools,most notably because of the increase in contact area between airand water. Occupants, in fact, cause waves, ripples and mist,increasing with the number of occupants and their activity.

Nomenclature

A surface of evaporation area (m2)

D diffusivity

E evaporation rate (kg/(m2 s); kg/(m2 h), lbs/h)

G mass flow rate (kg/s, g/h)

I water latent heat of evaporation (kJ/kg)

K mass transfer coefficient (referred to pressure) (kg/

(Pa m2 s))

L water evaporation heat (kJ kg�1)

m mass (kg)

N number of pool occupants (–)

P saturation pressure (Pa, Hg)

Re Reynolds number (–)

Sc Schmidt number (–)

Sh Sherwood number (–)

t time (s)

T temperature (8C)

V velocity (m/s)

Y latent heat of evaporation (kJ kg�1)

W specific humidity (kg of moisture/kg of dry air)

Subscriptsa air

ev evaporation

los loss

max pool area plus waves area

p pool

r room

un unoccupied

v ventilation

w water

Greek lettersD interval (–)

D deviation (–)

y viscosity

r air density (kg/m3)

F relative humidity (%)

F. Asdrubali / Energy and Buildings 41 (2009) 311–319312

Most of the empirical equations for unoccupied pools are offollowing type:

E ¼ gApðDPÞn (1)

where E is the water evaporation rate per unit area of the pool (kg/(m2 s)); g is a constant; Ap is the pool surface area (m2);DP = Pw � Pr is the difference between water and room saturationpressures (Pa); and n is a value ranging from 1 to 1.2.

The most widely published and used correlation for waterevaporation rates is the one proposed by Carrier in 1918 [12] andlater reported in the ASHRAE Application Handbook [13]:

E ¼ ð0:089þ 0:40782 VaÞAp DP=Y (2)

where Va is the velocity of air parallel to the water surface (m/s)and Y is the latent heat of evaporation of water (kJ/kg).

In Eq. (2), E, Ap and DP are measured respectively in kg/h, m2

and Pa.

The expression is based on laboratory experiments in which airwas blown above the water surface of a pool. Some authors suggestthat the formula overestimates evaporation for unoccupied poolsand recommend it for evaluating evaporation losses from occupiedswimming pools [14].

Smith et al. [15,16] conducted tests on occupied andunoccupied swimming pools and gave empirical formulas basedon these data; their equations are:

For unoccupied pools:

E ¼ ðC þ 0:35VaÞAp DP=Y (3)

where C is a coefficient which depends on barometric pressure(C = 72 at 5000 ft elevation and C = 69 at sea level).

For occupied pools:

E ¼ ð0:068þ 0:063FuÞAp DP=I (4)

where Fu is the pool utilization factor (Fu = Amax/ApN); Amax is thepool area Ap increased by waves area; I is the latent heat ofevaporation of water (kJ/kg).

A different model has been proposed by Hannsen and Mathisenin [1]. Their formula for unoccupied pools may be written as

E ¼ 3� 10�5V1=3ðe0:06Tw �Fa e0:06Ta Þ (5)

where V ¼ ½V2a þ ð0:12ð4ð1�FaÞ � ðTa � TwÞÞ0:5Þ

2�0:5

; Tw is thewater surface temperature (8C); Ta is air temperature (8C); and Fa

is air relative humidity (–).Shah [10] proposed a correlation based on the analogy between

heat and mass transfer for unoccupied pools, later modified toimprove accuracy:

E ¼ KAprwðrr � rwÞ1=3ðWw �WrÞ (6)

where r is the air density (kg/m3); rr is the room air density, whilerw is the saturated air density; W is the specific humidity (kg ofmoisture/kg of dry air); and K is a constant.

In Eq. (5), K = 40 if rr � rw < 0.02; K = 35 if rr � rw > 0.02.The correlation was evaluated against undisturbed water pool

test data from various sources, covering a wide range of watertemperatures (7.1–94.2 8C), air temperatures (6.1–34.6 8C) and airrelative humidities (28–98%).

Shah recommends Eq. (6) for indoor water pools withundisturbed surfaces and unforced airflow over those surfaces.

He also proposed an empirical correlation based on test datafrom various sources for occupied pools:

E ¼ Apð0:113� 0:0000175Ap=N þ 0:000059 DPÞ (7)

Eq. (7) is recommended for normal activity occupied pools (N,number of pool occupants less than 45), under the followingconditions: water temperature (25–30 8C), air temperature (26–31.7 8C) and air relative humidity (33–72%). Finally, Shah [11] alsoproposed another formula for pools with very intense activity suchas diving and water polo.

As stated previously, some of the correlations mentioned arederived from energy balances and others from experimentalmeasurements in real pools, but there is no evidence in theliterature of measurements carried out on scale models, apart fromsome studies quoted by Shah in [11], some of which date back tomore than 60 years ago and cover a range of air and watertemperatures that is too wide to be considered reliable. The aim ofthis paper is therefore to provide new experimental results forevaporation rates from water basins, thanks to a scale model andan apparatus which allows one to accurately control all the mainparameters influencing the phenomenon.

F. Asdrubali / Energy and Buildings 41 (2009) 311–319 313

3. Experimental apparatus

Various apparatuses have been built at the ThermotechnicalLabs of the University of Perugia for the experimental determina-tion of water evaporation rates. Preliminary analyses were carriedout by means of a scale model of a swimming pool inserted into aclimatic chamber for air temperature and relative humiditycontrol. Evaporation was determined using such apparatus,varying air temperature, humidity and air speed and measuringthe resulting water temperature and evaporated water [5,6].

The experimental campaign described in this paper was carriedout thanks to an improved apparatus, which includes a water heatregulation system. In indoor swimming pools water temperature isin fact controlled by a specific plant and is not a function of airtemperature.

The experimental apparatus (Fig. 1) for the tests is made up of:

(1) E

nvironmental test chamber: Mazzali Climatest Model climaticchamber, volume for testing 300 l, dimensions 700 mm� 660 mm � 680 mm, temperature range �40/+150 8C, uni-formity�0.3 8C, relative humidity range from 15% to 98% (�3%);the chamber is equipped with two centrifugal fans to guaranteeuniformity and with heating elements and a refrigeratingmachine to control air temperature.

(2) P

recision balance: Scaltec model Bel Engineering Ultra Mark4000, digital screen, range from 0 to 4 kg, accuracy �0.01 g;operating temperature conditions between 0 and 40 8C, relativehumidity between 20% and 85%, used to value evaporated mass.

(3) E

xternal aluminum container: dimensions are 300 mm� 300 mm � 95 mm, internally insulated with polyurethanefoam 50 mm thick, so that heat exchange occurs mainlythrough the free water surface.

(4) In

ternal aluminium container: dimensions 250 mm � 150 mm� 70 mm, container capacity 2.5 l.

(5) T

emperature regulator (Gefran 2000). (6) T emperature probe: PT100, Class A, Din 43760, precision�0.04 8C, immersed in the water and connected to regulator (5);for measuring water temperature.

(7) T

hermal resistor used to heat water inside the container. (8) W ooden box used to regulate air velocity inside the climatic

chamber.

(9) A ir velocity probe: hot wire type, model BSV 101, with

compensation for operating temperature conditions between0 and 40 8C and relative humidity between 10% and 95%; theinstrument precision is �0.01 m/s in the range 0–l m/s; an

Fig. 1. Experimental apparatus for water e

externally controllable mechanical stick allows one to move theprobe appropriately over the container surface.

(10) F

luke 45-01 mode for five figures and double fluorescentdisplay, for signal acquisition of air velocity.

The container (4) can store up to 2.5 l of water and it is putinside another insulated container. In this way, the first containeris thermally insulated and at the same time it can be easilyextracted. A wooden structure (8) with various openings isinserted into the climatic chamber, making it possible to regulateair flow velocity in proximity of the free water surface and to varyits rate by moving the openings.

A thermoregulation system guarantees a temperature oscilla-tion of water of about �0.1 8C from the fixed one. The use of anelectrical impulse system to regulate water temperature allows adrastic reduction of temperature oscillations caused by thermalinertia of heating elements. Measured water temperature oscillationsand air temperatures during a typical test are shown in Fig. 2.

During the preliminary stage, air velocities were measured todetermine the perfect positioning of the wooden structureopenings in order to guarantee different air velocities over thefree water surface. Velocity is measured at about 10 mm fromwater surface in five different positions.

4. Measurement methodology

At the beginning of each test, container (4) was filled with about2500 g of distilled water and placed inside the insulated box andinto the climatic chamber. Water temperature, air relativehumidity, air velocity and air temperature values were then set,according to the combinations given in Table 1, typical of indoorswimming pools. In particular, water temperature ranged from 20to 30 8C, and air temperature from 22 to 32 8C: in order toguarantee thermal comfort conditions for pool occupants, espe-cially when they get out of the pool, air temperature has to be 28higher than water temperature. The relative humidity valuesinvestigated ranged from 50% to 70%; as concerns air velocity, wasnot only the value of 0.05 m/s suggested by ASHRAE [17] and UNI[18] standards for indoor swimming pools considered, but also, forthe sake of completeness, the higher values of 0.08 and 0.17 m/s.

Once the stationary conditions in the chamber were reached,water temperature and evaporated water mass were measuredevery 10 min.

During all measurements, barometric pressure remainedwithin the range of 99.5–100.50 kPa.

vaporation measurements.

Table 2K value (10�8) as a function of water temperature and air humidity for V = 0.05 m/s.

Va = 0.05 m/s

20 21 22 23 24 25 26 27 23 23 30

50% 3.35 3.33 3.40 3.35 3.31 3.33 3.34 3.30 3.24 3.20 3.23

60% 2.94 3.16 3.33 3.48 3.60 3.72 373 3.59 3.74 3.76 3.65

70% 3.06 3.17 3.54 3.85 385 3.65 3.57 3.61 3.52 3.53 3.44

Fig. 2. Oscillations of air (Ta) and water (Tw) temperatures during a typical test.

Table 1Water evaporation measurements: investigated environmental conditions.

Tw (8C) 20 21 22 23 24 25 26 27 28 29 30

Ta (8C) 22 23 24 25 26 27 28 29 30 31 32

F (%) 50 50 50 50 50 50 50 50 50 50 50

60 60 60 60 60 60 60 60 60 60 60

70 70 70 70 70 70 70 70 70 70 70

Va (m/s) 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05

0.08 0.08 0.08 0.08 008 0.08 0.08 0.08 0.08 0.08 0.08

0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17

F. Asdrubali / Energy and Buildings 41 (2009) 311–319314

5. Results

All measurements were repeated at least two times and theresults averaged; evaporated water mass was then calculated as afunction of water temperature, for the different values of airrelative humidity and velocity (see, for example Fig. 3, for airvelocity = 0.05 m/s).

The results show that as expected evaporated mass flowincreases when water temperature increases and relative humiditydecreases.

The measured data allow the calculation of the coefficient ofmass transport for every condition investigated. In a system inwhich the transport mechanism is convective, mass flow can bedescribed by introducing a mass transport coefficient which from aphysical point of view, has the same significance as the convectioncoefficient in problems concerning heat transfer.

In particular, for problems which concern evaporation from abasin where air flow hits a water mass at a certain speed, masstransport evaporation is given by [19]

Gw ¼ kA½PwðTwÞ �FPwðTaÞ� (8)

Fig. 3. Evaporated water flow rate (g/h), as a function of water temp

where Gw is the water mass flow rate (kg/s); A is the evaporationsurface area; K is the mass transport coefficient; and Pw is the watersaturation pressure.

In Eq. (8) the concentration degree is expressed in terms ofpressure difference between saturated vapour in balance withbasin water and vapour which is in the air flow hitting theevaporating surface. From Eq. (8) we can calculate K,

K ¼ Dm

Dt

1

A½PwðTwÞ � fPwðTaÞ�(9)

K values, calculated using Eq. (9) thanks to the measured values ofevaporated water flow rate Dm/Dt and using the typical airvelocity in indoor pools of 0.05 m/s, are given in Table 2.

K depends on a geometrical parameter as shown by thedimensionless relation for laminar flow along a flat plate [19]:

Shx ¼ CðRexÞ1=2ðScÞ1=3 for Re<5� 105 (10)

where the Sherwood number (Sh) has the same or similarfunctional dependence on the Reynolds (Re) and Schmidt (Sc)numbers as the Nusselt number has on the Reynolds and Prandtlnumbers.

The Sherwood number is defined as

Shx ¼ Kx=D (11)

where x is the distance of free water surface from friction surfaceand D is diffusivity.

Sh assumes the significance of the ratio between total masstransfer and diffusive mass transfer.

The Schmidt number is defined as

Sc ¼ n=D (12)

where n is the viscosity.Sc assumes the significance of the ratio between momentum

diffusivity and mass diffusivity.Finally, the Reynolds number is defined as

Rex ¼ V x=n (13)

erature, for different values of relative humidity (Va = 0.05 m/s).

Fig. 4. Predicted water evaporation flow rate per unit area, as a function of water temperature, for different values of air relative humidity (Va = 0.05 m/s).

F. Asdrubali / Energy and Buildings 41 (2009) 311–319 315

Re assumes the significance of the ratio between inertial forces andviscous forces.

Hence, for the same values of D, V and n:

K ¼ mx1=2�1 (14)

so the existing relationship between the parameters of the realscale and the scale model of the swimming pool is

K

K¼ x

x

� �12�1

(15)

where K is the mass transfer coefficient relative to the realswimming pools; x is the real size of the swimming pool; x=x ¼ f isthe scale factor.

Hence

K ¼ 1ffiffiffiffif

p � K (16)

The scale factor between a real swimming pool and the model istherefore around 10. Therefore, to obtain the real scale value of K

from Eq. (16), the scale model value has to be multiplied by 3.16. Inthe measurements carried out on the scale model, the Re valuesremained lower than 5 � 103.

Eq. (10) can be used for Re lower than 5 � 105 and under thehypothesis that mass transfer of a laminar flow along a flat plate isequivalent to the that of air slowly lapping a water surface.

Experimental data confirm, as foreseen in theory [20], that thevalue of K remains constant in various tests during the same cycleof measurements in which the geometry of the system and airvelocity are unvaried; for example K is equal to the average value of3.4 � 10�8 kg/(m2 Pa s) for V = 0.05 m/s and of 4.2 � 10�8 kg/(m2 Pa s) for V = 0.08 m/s.

Fig. 5. Predicted water evaporation flow rate per unit area, as a function of wat

6. A model for predicting evaporation flow rate

When K value is known, it is possible to predict the waterevaporation flow rate in a basin as a function of water temperature,air temperature and humidity, using the equation:

G ¼ KðPwðTwÞ �FPwðTaÞÞ (17)

Figs. 4–6 show the predicted evaporation mass flow per unitarea using Eq. (17) as a function of water temperature for thedifferent values of relative humidity and, respectively, the threeinvestigated values of air velocity and the corresponding values ofK obtained for the scale model.

As previously stated, the experimental facility used in thisresearch is inserted into a climatic chamber for air relativehumidity and temperature control.

The proposed method has been compared only with predictionmethods found in the literature for unoccupied pools in similarenvironmental conditions and for three typical air velocity values(0.05; 0.08; 0.17 m/s). The Shah [10] model for unoccupied poolswas considered, since it is the most recent and accurate work; inaddition, the Hannsen–Mathisen (H&M) model [1] (formulated inforced ventilation conditions) and Smith model [15,16] forunoccupied pools were also included in the comparison.

For the comparison, the evaporation rate values evaluated usingthe real scale models were reduced according to the scale modelfactor as described in Section 5.

Predicted values in this work can be compared mainly with theones predicted by other models using two types of deviation data:

Average deviation ¼P

dn

(18)

Mean deviation ¼P

AbsðdÞn

(19)

er temperature, for different values of air relative humidity (Va = 0.08 m/s).

Fig. 6. Predicted water evaporation flow rate per unit area, as a function of water temperature, for different values of air relative humidity (Va = 0.17 m/s).

F. Asdrubali / Energy and Buildings 41 (2009) 311–319316

where deviation d is defined as

d ¼ data predicted by present work� data predicted by model

data predicted by model

(20)

n is the number of measured points.Figs. 7–15 show a comparison between the proposed model and

the models found in the literature, for relative humidity values of50%, 60% and 70% and for different water temperature values.

Figs. 10–12 in particular show that the proposed modelproduces values very similar to those of the Shah and Hannsen–Mathisen (H&M) models for an average velocity of 0.08 m/s(average d = 12% and 3.5%).

On the contrary, the values estimated using the proposed modelare lower than the ones estimated with the other models (Figs. 7–9) for air velocity of 0.05 m/s (typical of indoor swimming pools)(average d = �20% and mean d = 19% compared to Shah and H&M

Fig. 8. Comparison between predicted water evaporation rates per unit area: pr

Fig. 7. Comparison between predicted water evaporation rates per unit area: pr

models) and higher (Figs. 13–15) for air velocity of 0.17 m/s(average d = +27%; +17% and mean d = 27%; 17% compared to Shahand H&M models).

It can be concluded that the proposed method, based onthe experimental data found for the scale model, is in goodagreement with Shah and H&M prediction formulas forunoccupied pools. The Smith model always overestimates waterevaporation.

7. Errors evaluation

Error committed in the calculation of the mass transportcoefficient depends on errors in measurement of the variousparameters that regulate the evaporative phenomenon, and inparticular: water temperature, air temperature and relativehumidity, mass of evaporated water and time. The new experi-mental apparatus and the new measurement procedure originatethe absolute errors reported in Table 3.

oposed model and models found in the literature (F = 60%, Va = 0.05 m/s).

oposed model and models found in the literature (F = 50%, Va = 0.05 m/s).

Fig. 11. Comparison between predicted water evaporation rates per unit area: proposed model and models found in the literature (F = 60%, Va = 0.08 m/s).

Fig. 12. Comparison between predicted water evaporation rates per unit area: proposed model and models found in the literature (F = 70%, Va = 0.08 m/s).

Fig. 9. Comparison between predicted water evaporation rates per unit area: proposed model and models found in the literature (F = 70%, Va = 0.05 m/s).

Fig. 10. Comparison between predicted water evaporation rates per unit area: proposed model and models found in the literature (F = 50%, Va = 0.08 m/s).

F. Asdrubali / Energy and Buildings 41 (2009) 311–319 317

Relative error on K can be calculated applying the theory ofpropagation of errors at Eq. (8). Under the experimental conditionsexamined, such error varies from a minimum of 1% to a maximumof 11.6%, with a mean value of 7%, therefore lower than the onefound for the previous experimental apparatus (11%) [5].

8. Energy consumption due to water evaporation in pools

As stated previously, in indoor swimming pools the highestthermal load in winter is due to water evaporation. With the aim oflinking the proposed evaporation model to energy impacts, a real

Fig. 15. Comparison between predicted water evaporation rates per unit area: proposed model and models found in the literature (F = 70%, Va = 0.17 m/s).

Fig. 13. Comparison between predicted water evaporation rates per unit area: proposed model and models found in the literature (F = 50%, Va = 0.17 m/s).

Fig. 14. Comparison between predicted water evaporation rates per unit area: proposed model and models found in the literature (F = 60%, Va = 0.17 m/s).

F. Asdrubali / Energy and Buildings 41 (2009) 311–319318

case was considered and energy consumptions due to all thedifferent contributions were evaluated. An indoor swimmingfacility with two pools – which is quite a typical situation – wasconsidered: the biggest pool is 316.60 m2 (25.1 m � 12.6 m) whilethe one for children is 57.96 m2 (12.6 m � 4.6 m). The pools areinside a building whose surface is 792 m2 and whose volume is5100 m3 and are located in Perugia, central Italy (latitude:43870000N, longitude 128230000, altitude 490 m above sea level).

Table 3Absolute mean errors for the main parameters.

Parameter Absolute error

Water temperature �0.1 8CAir temperature �0.2 8CRelative humidity �3%

Evaporated flow rate �0.01 g

Time interval �15 s

The environmental conditions (internal and external air)assumed for the calculations are reported in Table 4, along withthe heat transfer coefficients of the various building components.

The building heat load due to water evaporation was calculatedby multiplying water latent heat (2500 kJ/kg) by the evaporatedflow rate G given by Eq. (17), using the values of coefficient K

calculated in this work. The result is 60.5 kW.A second contribution to energy consumption in the building is

due ventilation; a certain air flow rate has to be taken from theoutside, heated and inserted into the building to keep internalrelative humidity constant and equal to 50%. The higher is waterevaporation and the higher is ventilation flow rate and thecorresponding heat load. A value of 87 kW resulted from thecalculations.

Finally, the heat load due to dispersions through the buildingcomponents was evaluated in winter conditions, according toItalian law on energy efficiency in buildings [21], obtaining a valueof 56 kW.

Table 4Environmental conditions and building thermal properties used in the calculations.

Va (m/s) Ta (8C) Tw (8C) Wa (g/kg) F (%)

Internal conditions 0.08 29 27 12.6 50

External conditions – �2 – 2.6 80

Roof (W/(m2 K)) External wall (W/(m2 K)) Internal wall (W/(m2 K)) Soil (W/(m2 K)) Glass (W/(m2 K))

Heat transfer coefficients 0.84 0.84 1.38 0.57 3.5

Fig. 16. Different contributions to the total heat load in an indoor swimming pool.

F. Asdrubali / Energy and Buildings 41 (2009) 311–319 319

The overall heat load is the sum of the three contributions and isequal to 203.5 kW (Fig. 16): it may be noticed that the heat loaddue to water evaporation and the one due to ventilation(depending also from water evaporation) are respectively 30%and 42% of the overall heat load and are much bigger than the heatdispersed through the building components. It is thereforeextremely important to use an accurate model to evaluate waterevaporation from indoor pools: an error of about 10% in theevaluation of coefficient K in Eq. (17) may result in an error of about7% in the evaluation of the overall heat load.

9. Conclusions

Water evaporation from indoor swimming pools frequentlycauses the highest energy consumption of the pool plant, so anaccurate evaluation of the thermal loads due to water evaporationis a topic of considerable practical interest.

A large experimental campaign has been carried out to measurewater evaporation flow rate from basins of assigned temperature.Measurements were carried out by means of an originalconception apparatus based on a pool scale model inserted intoa climatic chamber; experimental data, available in the range of airand water temperatures and air relative humidity typical of indoorswimming pools, were correlated and the mass transportcoefficient K, as well as the model scale factor, were calculatedfor the environmental conditions considered.

A new prediction model has therefore been proposed toevaluate evaporation flow rate from indoor swimming pools: themodel has been compared with some of the most widely knownmodels found in the literature, and in particular the Shah, Hannsenand Mathisen and Smith models, derived from measurements inreal pools, and a good agreement was found.

References

[1] S.O. Hanssen, H.M. Mathisen, Evaporation from swimming pools, in: Proceedingsof Roomvent ‘90, Oslo, 1990.

[2] K. Biasin, W. Kumme, Die Wasserverdunstung in einem Inneschwimmbad, Elec-trowaerme International 32 (1974) A115–A129.

[3] M.M. Shah, Prediction of evaporation from occupied indoor swimming pools,Energy and Buildings 35 (2003) 707–713.

[4] R.K. Linsley, M.K. Kohler, J.L.H. Paulhus, Hydrology for Engineers, Mc Graw HillInternational Book Company, 1982.

[5] F. Cotana, Experimental determination of the water vapour mass transferunder various environmental test conditions, Heat and Technology 18 (1)(2000) 69–77.

[6] F. Asdrubali, G. Baldinelli, A. Presciutti, Misure di portata di evaporazione dasuperfici a pelo libero di temperatura assegnata, in: Atti del XXI CongressoNazionale dell’Unione Italiana Termofluidodinamica, Genova, giugno, 2004 (inItalian).

[7] P.G. Whitehead, M. Robinson, Experimental basin studies an international andhistorical perspective, in: Proceedings of the Conference on Methods of Hydro-logic Basin Comparison, Oxford, UK, 1992.

[8] B.L. Sill, Free and forced convective effects on evaporation, Journal of HydraulicEngineering ASCE 109 (9) (1983) 1216–1231.

[9] P.J. Ryan, D.R.F. Harleman, K.D. Stolzenbach, Surface heat loss of cooling ponds,Water Resources Research 10 (5) (1974) 930–938.

[10] M.M. Shah, Calculating evaporation from swimming pools, Heating/Piping/AirConditioning Engineering 21–26 (2004).

[11] M.M. Shah, Rate of evaporation from undisturbed water pools to quiet air:evaluation of available correlations, International Journal of HVAC&R Research8 (2002) 125–131.

[12] W.H. Carrier, The temperature of evaporation, ASHVE Transactions 24 (1918)25–50.

[13] ASHRAE, ASHRAE Handbook HVAC Applications, Atlanta, USA, 1999.[14] R. Tang, Y. Etzion, Comparative studies on the water evaporation rate from a

wetted surface and that of a free water surface, International Journal of Buildingand Environment 39 (2004) 77–86.

[15] C.C. Smith, C.O.G. Lof, R.W. Jones, Energy requirements and potential savings forheated indoor swimming pools, ASHRAE Transactions (1993) 864–874.

[16] C.C. Smith, C.O.G. Lof, R.W. Jones, Rates of evaporation from swimming pools inactive use, ASHRAE Transactions 104 (1A) (1999) 514–523.

[17] ASHRAE, Ventilation for Acceptable Indoor Air Quality, ASHRAE Standard 62-89,American Society of Heating Refrigerating and Air Conditioning Engineers,Atlanta, USA, 1989.

[18] UNI 10637, Requisiti degli impianti di circolazione, trattamento, disinfezione equalita dell’acqua di piscina, Giugno, Milano, 1997 (in Italian).

[19] R.E. Treybal, Mass-transfer, McGraw-Hill International Editions, 1981.[20] F.P. Incropera, D.P. De Witt, Fundamentals of Heat and Mass Transfer, John Wiley

& Sons, New York, 1985.[21] F. Asdrubali, M. Bonaut, M. Battisti, M. Venegas:, Comparative study of energy

regulations for buildings in Italy and Spain, Energy and Buildings 40 (2008)1805–1825.