Applied Energy 88 (2011) 4358–4365

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Application of infrared thermography for the determination of the overall heat

transfer coefficient (U-Value) in building envelopes
Paris A. Fokaides a, Soteris A. Kalogirou b,⇑
a
University of Cyprus, Department of Civil and Environmental Engineering, Cyprus
b
Cyprus University of Technology, Department of Mechanical Engineering and Materials Science and Engineering, Cyprus

a r t i c l e i n f o a b s t r a c t

Article history: Infrared (IR) thermography constitutes a reliable measurement method for the determination of spatially
Received 9 January 2011 resolved surface temperature distributions. IR thermography may be used for several research problems,
Received in revised form 24 April 2011 applications, and measurement environments with a variety of physical arrangements. In this work the
Accepted 6 May 2011
results of the determination of the overall heat transfer coefficient (U-Value) with the use of IR thermog-
Available online 23 July 2011
raphy for building envelopes are presented. The obtained U-Values are validated by means of measure-
ments performed with the use of a thermohygrometer for two seasons (summer and winter), as well as
Keywords:
with the notional results provided by the relevant EN standard. Issues related to the applicability of the
Infrared (IR) thermography
Overall heat transfer coefficient (U-Value)
method due to the non-steady heat transfer phenomena observed at building shells are also discussed. A
Sensitivity analysis more precise validation of the proposed technique was also performed with the use of heat flux meters.
Thermal emissivity The percentage absolute deviation between the notional and the measured U-Values for IR thermography
Reflected temperature is found to be in an acceptable level, in the range of 10–20%. Finally, a sensitivity analysis is conducted in
order to define the most important parameters which may have a significant influence on the measure-
ment accuracy.
Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction authorities did not kept detailed building records at the time of
construction. Therefore, the question to answer is how reliable is
The use of infrared (IR) thermography has increased dramatically the implementation of IR thermography for the determination of
throughout the world over the past few years. This technique is the U-Value of building shells.
employed for the measurement of surface characteristics for a vari- The discussion on the applicability of the IR thermography for
ety of research investigations involving all possible heat transfer building diagnostics has been held since this technique was widely
phenomena. The method is especially important and useful commercialized at the early 90s. One of the most important ques-
because it gives spatially resolved surface temperature distribu- tions to answer is whether this technique may provide reliable
tions non-intrusively, even when large gradients of surface quantitative measurements, or if the extend of this application is
temperature are present [1]. IR thermography can be used in build- limited to qualitative results. Another issue which is also under
ing envelopes to detect heat losses, missing or damaged thermal discussion, especially in the case of building thermography is the
insulation in walls and roofs, thermal bridges, air leakages and non-steady character of the heat transfer in building envelopes
sources of moisture. IR thermography can also be employed in and the interpretation of an instantaneous thermogram under
building diagnostics for the determination of the thermo physical non-steady conditions.
properties of building envelopes. Currently, this application is The main areas for using IR thermography in building diagnos-
becoming more important, as the knowledge of the U-Value is a tics are presented in detail by Balaras and Argiriou [3]. In particu-
precondition for the classification of the energy performance of lar, representative examples of building envelopes, mechanical and
existing buildings [2]. This information may not be in some cases electrical systems inspection in audited office buildings are pre-
accessible or available, especially in the case of old buildings or sented to demonstrate common problems and data interpretation.
in member states of the European Union where the regulatory In this study, it is suggested that building IR thermography on
external building elements should be performed either at night
or during a cloudy day. This was found to be important in order
⇑ Corresponding author. Address: Cyprus University of Technology, Department
to avoid the problem of temperature increase which occurs as a re-
of Mechanical Engineering and Material Sciences and Engineering, P.O. Box 50329,
Limassol 3603, Cyprus. Tel.: +357 2500 2621.
sult of the incident solar radiation, and the impact from the ab-
E-mail addresses: [email protected] (P.A. Fokaides), [email protected] sorbed solar energy, which presents a time lag of a few hours.
(S.A. Kalogirou). Additionally in this work it is pointed out that measurements

0306-2619/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.apenergy.2011.05.014
 P.A. Fokaides, S.A. Kalogirou / Applied Energy 88 (2011) 4358–4365 4359

Nomenclature

Symbols Abbreviations
A area (m2) HFM heat flux meter
d thickness (m) INT integer part of test duration, according to ISO
h thermal convection coefficient (W/m2 K) 9869:1994
k thermal conductivity (W/mK) IR infrared
I_ power intensity (W)
Q_ heat flux (W) Indices
R thermal resistance (m2 K/W) ATM atmosphere
RH relative humidity of ambient air (%) CONV convective
T temperature (°C) EM emitted
x distance between thermographer and target (m) IN indoor
U overall heat transfer coefficient (U-Value) (W/m2 K) MEAS measured
OBJ object
OUT outdoor
Greek letters
RAD radiative
e thermal emissivity (–)
REF reflected
r Stefan–Bolzmann constant = 5.67  108 (W/m2 K4)
TOT total
s thermal transmittance (–)
W wall

should be carried at low wind speeds in order to minimise as much be converted to temperature by means of the Stefan–Bolzmann’s
as possible the influence of convective heat losses. law. In order to achieve a precise conversion of the thermal energy
Grinzato et al. [4] presented some application examples of to temperature, an understanding of the various types of heat
quantitative building IR thermography for diagnostic purposes. In transfer occurring between the target object surface and the ther-
this study four models were investigated in detail, each one pre- mal camera sensor is required.
senting a specific building construction deficiency. The general The energy which reaches the thermal sensor consists not only
outcome of this study is that defects under non-steady heat trans- from the energy emitted by the target itself, but also from the en-
fer conditions are more easy to be detected than in steady state. In ergy reflected by the surroundings and intercepted by the object’s
this work, the question of the thermal stimulus, i.e., the incident surface. These scalar quantities are attenuated by the atmosphere
solar radiation, is also raised since its power, depends on several in the measurement path as shown in Fig. 1. Thus, the thermal en-
parameters and may not be easily predictable. Therefore, by imple- ergy at the thermal sensor consists of three terms.
menting this technique the heat stimulus parameter may differ,
resulting in poor reproducibility of results. I_MEAS ¼ I_EM þ I_REF þ I_ATM ð1Þ
Avdelidis and Moropoulou [5] presented a review on emissivity All three parameters may be determined on the basis of Stefan–
measurement techniques and indicated the importance of surface Bolzmann’s law. As far as the target surface emitted energy is con-
emissivity values on building diagnostics. In this work, the emis- cerned, the parameter which needs to be identified is its thermal
sivity of selected building materials, such as plasters, marbles, emissivity (e). For the determination of the reflected energy, it is as-
and porous stones were determined at various temperatures and sumed that the reflective temperature is the same for all reflecting
given in tables. The measurements were performed in the mid- surfaces. This is to a great extent correct when performing indoor
and long-wavelength regions of the infrared spectrum, using dif- building IR thermography. In accordance to Kirchhoff’s law, it is also
ferent approaches. The discussion of the results was performed assumed that the emissivity of the surroundings is equal to one. Fi-
in terms of the approach employed, the effect of the wavelength nally for the atmosphere emission, the atmosphere transmittance
and temperature effects, as well as the materials surface state. At (s) is required. The total radiation power received by the object’s
this point however, it is important to note that not only the mate- surface is expressed in terms of the target object thermal intensity
rial as such but also its condition (age, roughness, exposure to envi- as follows [7]:
ronment) as well as its shape have an important influence on its
emissivity [6]. Therefore, it is questionable whether values taken 1 1e_ 1s_
I_OBJ ¼ I_TOT  IREF  IATM ð2Þ
from emissivity tables for building IR thermography purposes es e es
should be used, since this value must be determined for each Therefore, according to Eq. (2), the parameters which are required
measurement. for the detailed determination of the temperature field in a thermo-
In this paper the objective is to determine the U-Value of typical gram are the emissivity of the object, the reflective temperature,
building constructions in Cyprus using IR thermography and vali- and the transmittance of the atmosphere. The latter can be
date the results with the notional values obtained from relevant
EN standards and the use of thermohygrometer for two seasons,
summer and winter and with the use of heat flux meters. This will
prove whether the IR thermography is suitable for such
estimations.

2. Methodology

IR thermography is based on the measurement of the radiant

thermal energy distribution (heat) which is emitted from an ob-
ject’s surface, according to the Planck’s law. This radiation may Fig. 1. Thermal radiation components between target and thermal camera.
4360 P.A. Fokaides, S.A. Kalogirou / Applied Energy 88 (2011) 4358–4365

determined from the knowledge of the atmosphere’s temperature the target to measure. Therefore, both were fixed on the target
and relative humidity, and the distance between the target and for some time prior to the measurement.
the sensor. Therefore, as part of the thermography process, it is nec- The calculation of the U-Value by means of IR building thermog-
essary to record also the ambient temperature as well as the rela- raphy was accomplished based on the assumption that the heat
tive humidity of the atmosphere with the use of additional transfer from the wall to the thermal camera sensor resulted due
instruments. The determination of the emissivity and the reflective to thermal radiation as well as thermal convection, given by
temperature is achieved by means of in situ calibration, which is de-
scribed in the following paragraph. Q_ RAD ¼ 4erAT 3W ðT W  T REF Þ ð3Þ
The measurement procedure which was followed in this work is
described in detail in [8]. The building element to be examined was Q_ CONV ¼ hIN AðT W  T IN Þ ð4Þ
an exposed one; however the measurement was performed from
The U-Value was determined according to [11]
the inside of the building. The building element to be measured
needs to be free of any objects and was not conditioned directly Q_ ¼ UAðT IN  T OUT Þ ð5Þ
by air blowing devices such as split units or exposed to sun for
the past few hours. In this context, it was also important to achieve The U-Value calculation took into consideration the individual heat
a quasi-steady heat transfer state during the measurement for at transfer coefficients of both the radiant and the convective heat
least 3–4 h prior to measurement, and to verify that a minimum stream. Values for the thermal convection coefficient (h) were taken
temperature difference exists between the inside and the sur- from (11). Hence, in terms of this study, the U-Value was calculated
rounding area of the building of 10 °C. An important step for each by the following relationship.
thermogram was the measurement of the reflected ambient tem- ½4erT 3W ðT W  T REF Þ þ hIN ðT W  T IN Þ
perature and the target’s thermal emissivity (e). The reflective U MEAS ¼ ð6Þ
ðT IN  T OUT Þ
ambient temperature was determined by means of a crumpled
piece of aluminium foil fixed on the surface, in the field of view
of the thermal image (Fig. 2). Based on standards [9,10], the re-
flected ambient temperature is the average temperature of the foil 3. Infrastructure and investigated sample
target, if the IR camera emissivity at this point is set to equal 1.0.
For the emissivity, a piece of black tape (e = 0.95) was fixed on For the needs of the present study a FLIR T360 thermal camera
the target. During the post processing of the image, it was possible was employed [12]. The field of view of this camera is 25  19°/0.4
to determine the exact temperature of the target in the area of the (m) and the IR resolution is 320  240 pixels. The camera performs
known emissivity. In both cases it is of importance for the alumin- measurements in the long-wavelength infrared range of the elec-
ium foil, and for the black tape to be in thermal equilibrium with tromagnetic spectrum (7.5–13 lm). The specifications of the ther-
mal camera are given in Table 1. The post processing of the results
was performed by means of the FLIR Build IR ver. 2.0 Software.
Additionally, two thermohygrometers were employed. One was
used for the determination of the atmosphere condition at the
measurement area (ambient temperature and relative humidity).
The second thermohygrometer (TESTO 635, accuracy ±0.1 °C) was
used for the validation of the surface temperature measurement,
as captured by the IR camera. Both instruments were equipped
with a data logger, thus continuous time series of measurements
were recorded. The U-Value of the building element was also ob-
tained with the use of the TESTO 635 instrument. The calculation
was based on a thermal balance which relates heat transfer due
to thermal conduction and convection on both sides of the wall
(Eq. (7)). Hence, a detailed comparison between the results pro-
vided by IR building thermography and those calculated by the
application of the thermohygrometer can be obtained.
T W  T IN
U MEAS ðW=m2 KÞ ¼ hIN ð7Þ
T OUT  T W
Additionally, heat flux meters were used to validate the results. For
this purpose, the International Thermal Instrument (ITI) Company,
Model No. GHT-1C(801) heat flux meter was used. The flux meters
used are able to record both the heat flux and surface temperature.
The measurements were recorded with the help of the Pico TC-08

Table 1
Technical specifications of thermal camera.

Field of view 25°  19°/0.4

(m)
Thermal sensitivity 50 mK
Spectral range 7.5–13 lm
IR resolution 320  240 pixels
Object temperature 20 °C to 650 °C
range
Fig. 2. Arrangement for determination of reflected temperature according to [6]: Accuracy ±2 °C
digital image and thermogram.
 P.A. Fokaides, S.A. Kalogirou / Applied Energy 88 (2011) 4358–4365 4361

Thermocouple Data Logger. The measurement campaign was car- As far as the roof is concerned, four typical cases were
ried in Cyprus and concerns five dwellings in two seasons (summer investigated:
and winter), and took place during August 2009 and February 2010
respectively. Furthermore, the measurement campaign carried out (1) Flat, non-insulated roof made of reinforced concrete (15 cm)
to validate proposed methodology with the use of heat flux meters and screed (10 cm) (ID2 and ID3).
took place in March 2011. The investigated buildings resulted from (2) Flat, insulated roof including 5 cm extruded polystyrene
the statistical processing of the existing building stock [13] and are (ID5).
considered to be representative of the construction solution tech- (3) Pitched non-insulated roof (ID1).
niques used in Cyprus. In order to achieve steady state conditions, (4) Pitched insulated roof including 10 cm of stone wool (ID4).
the indoor spaces of the dwellings to be investigated were condi-
tioned, e.g. either cooled or heated for at least 3 h prior to the per- All constructions also include a damp proof layer (asphalt
formance of the measurements. In the same context, in an effort to based, 0.5 cm).
avoid the effects of thermal inertia during the measurement, the The five glazing test cases include single clear glazing (ID2),
building elements to be measured were not exposed to the sun double clear glazing (ID1 and ID5) and double glazing consisting
for a few hours prior and during the measurement. The surface tem- of external laminated glass and internal low-emissivity glass (out-
perature of each building element was measured every 20 min for a er surface of inner glass) (ID3 and ID4).
total duration of 3 h and the U-Value resulted from the average cal-
culation of the 10 measurements for each building element.
The notional U-Values were obtained according to EN ISO 4. Results and discussion
6946:1997 [11] for masonry and roofs and EN 673:1997 [14] for
glazing and are given in Table 2. A variety of constructions were The results of the estimated U-Values for the heating and cool-
examined as follows: ing period obtained by applying the IR thermography and the ther-
Regarding masonry, the most common building solution in Cy- mohygrometer are given in Table 3. Specific conclusions regarding
prus is the 20 cm perforated brick, with an internal and an external the deviation of the measured U-Values from the theoretical ex-
layer of 2.5 cm plaster, a section of which is shown in Fig. 3; three pected values may be drawn from Fig. 4. This figure presents the
such cases were examined (ID2, ID4 and ID5). Dwellings with insu- percentage absolute deviation between the notional and the mea-
lated masonry (5 cm extruded polystyrene; ID3) as well as stone sured U-Values for IR thermography and thermohygrometer. As it
masonry (ID1) were also included in the sample. can be seen, the percentage absolute deviation was found to be
at an acceptable level, in the range of 10–20%. It is obvious that
the building element which gives the greater deviation in all cases
Table 2 tested by both techniques is the roof, followed by glazing. The most
Investigated buildings.
possible explanation is the fact that during daylight it is almost
ID (–) Zonea (–) Area (m2) Construction U-Value (notional) (W/m2 K) impossible to avoid solar exposure of roof, therefore effects of ther-
year (–) mal inertia seems to prevail. By comparing the results obtained by
Masonry Roof Glazing
the thermohygrometer and the IR thermography with the notional
1 1 80 1920 2.1 1.52 2.7 values, the latter appears to be less accurate. It should be noted
2 1 180 1971 1.39 3.33 5.8
however, that there are uncertainties in all methods as explained
3 2 150 1989 0.42 3.33 1.6
4 2 180 2004 1.39 0.56 1.6 below. Additionally, the IR thermography results were derived
5 3 98 2008 1.39 0.55 2.7 from a total sample of 10 measurements, whereas the results of
a the thermohygrometer were achieved by means of continuous
Zone 1: seaside, Zone 2: inland, Zone 3: semi-mountainous.
time series of measurements. Particularly, the time step of the

Table 3
Notional and measured U-Values with the use of IR thermography and thermohy-
grometer in winter and summer.

Dwelling Building Notional Estimated U-Values

element U-Values
EN ISO IR thermography Thermohygrometer
6946:1997
and EN 673
Winter Summer Winter Summer
ID 1 Wall 2.1 2.35 2.32 2.28 2.05
Roof 1.52 1.71 1.74 1.67 1.73
Glazing 2.7 3.03 3.01 2.99 2.95
ID 2 Wall 1.39 1.56 1.58 1.48 1.51
Roof 3.33 3.74 3.75 3.63 3.68
Glazing 5.8 6.67 6.83 6.69 6.73
ID 3 Wall 0.42 0.47 0.52 0.39 0.43
Roof 3.33 3.53 3.57 3.51 3.53
Glazing 1.6 1.94 1.89 1.89 1.83
ID 4 Wall 1.39 1.39 1.38 1.29 1.31
Roof 0.56 0.89 0.93 0.87 0.87
Glazing 1.6 1.93 1.98 1.87 1.96
ID 5 Wall 1.39 1.52 1.66 1.54 1.63
Roof 0.55 0.59 0.53 0.57 0.57
Glazing 2.7 2.94 2.99 2.84 2.93
Fig. 3. Masonry U-Value according to [11].
4362 P.A. Fokaides, S.A. Kalogirou / Applied Energy 88 (2011) 4358–4365

predefined thermal resistances, according to the direction of the

heat flow, are considered. This is also an assumption which may im-
port an error factor, since the state of the boundary layer on the
buildings’ shell surface is not in all cases at the same energy level.
Finally regarding heat conduction, the conductivity of the materials
was retrieved mainly from ISO 10456:2007 [15]. This standard
however, does not take into consideration the fact that the thermal
conductivity is a function of temperature and may deviate accord-
ing to the season. Additionally, there are many cases where the
determination of the thermal conductivity of a building material
is not feasible. Perforated bricks thermal conductivity for example
is taken equal to 0.4 (W/mK), whereas according to TN 1745:2002
[16], the thermal conductivity of a perforated brick strongly de-
pends on its geometry. Furthermore for external and internal plas-
Fig. 4. Percentage absolute deviation between notional and measured U-Values for ter, it is assumed that the thermal conductivity of the material is
IR thermography (IR) and thermohygrometer (TH). equal to 1 (W/mK), whereas it is well known that plaster is a mix-
ture of asbestos- and cement-based mortars. Hence the thermal
thermohygrometer recording was 1 s and the total duration of the conductivity is strongly dependent on the mixtures’ content, which
measurement was 3 h, resulting in 10,800 measurements per may differ, even in the same building shell.
building element. Therefore, much more data were evaluated in As already explained in Section 2, the calculation procedure of
the case of the thermohygrometer possibly resulting in a more pre- the U-Value is based on the assumption that heat transfer phenom-
cise result. In most cases a small increase of the U-Value is ob- ena between the two sides of the building shell occur under steady
served during the summer. This is not surprising due to the fact state conditions. However during the day, the ambient air temper-
that the thermal conductivity is a function of temperature, and ature follows a sinusoidal pattern. Therefore, the achievement of
thus under elevated temperature conditions, it is expected to be steady state conditions may be satisfied only during some time
increased. periods of the day according to the season; either early in the
It is also highly possible that part of the deviation between the morning, or late in the afternoon and during the night. Addition-
measured and the notional results to be due to the fact that the ally, as stated earlier, in order to have a satisfying resolution anal-
theoretical calculation of the U-Value deviates from the real condi- ysis of the calculated U-Value, a temperature difference of at least
tion. Based on EN ISO 6946:1997 [11], U-Value for masonry and 10 °C between indoor and outdoor is necessary. All these precondi-
roof is calculated from tions and restrictions which affect the application of IR thermogra-
phy are understood better by examining the time variation of
1 temperature and the resulting U-Values. In Fig. 5, the indoor, out-
U¼ Pn di  ð8Þ
RW;IN þ RW;OUT þ i¼1 ki
door and on wall surface temperature (masonry – ID2, winter sea-
son), as well as the calculated U-Value, which results from these
In the case of suspended ceilings, an additional thermal resistance measurements, are presented. It is obvious that the necessary pre-
for still air is added in the denominator. This calculation takes into condition for the achievement of correct results is a quasi-steady
account only the heat transfer due to convection and conduction. state condition, which is achieved early in the morning and late
However in the case of perforated bricks, heat transfer due to ther- at night. During these intervals, the condition of a temperature dif-
mal radiation within the bricks’ gaps may also occur; this is also the ference between indoor and outdoor of at least 10 °C is also satis-
case for suspended ceilings. The contribution of this heat flux fied. Due to thermal inertia of the building envelope, during
becomes more significant in the case of elevated temperature midday the calculated U-Values are found to be more than two
differences between indoor and outdoor. As far as the heat convec- times higher than the expected value, due to thermal inertia of
tion is concerned, for the calculation of the notional U-Values, the building envelope. In this context it can also be concluded that

Fig. 5. Measurement of masonry U-Value over 24 h – ID 5 (thermohygrometer).

 P.A. Fokaides, S.A. Kalogirou / Applied Energy 88 (2011) 4358–4365 4363

the same target, externally exposed, experiences a much wider

range of temperatures, which may exceed 20 °C. In case of the out-
door measurement many objects with unknown thermal status re-
flect on the target, importing several unknown parameters in the
reflective part of the thermogram. This is not the case for an indoor
measurement, since the thermographer is able to determine the
thermal condition of most objects reflecting on the target, and even
to control the reflection index of the thermogram. These results
support the decision of the authors to perform the U-Value deter-
mination based on indoor measurements, since as the sensitivity
analysis has shown the reflected temperature is very important
for the performance of reliable measurements.

5. Validation of the proposed methodology

Fig. 6. Percentage temperature measurement sensitivity to thermal emissivity (e)
and ambient reflected temperature.
According to the European Standard EN15603 [17] the in situ
measurement to be applied in order to determine the U-Value of
the building envelope is the one described in ISO 9869:1994
there will be some times of the year during spring and autumn,
[18]. In terms of this method, the heat flow rate through a building
where the determination of the U-Value will not provide reliable
element is determined with the use of a heat flux meter or a calo-
results, mainly due to decreased indoor and outdoor temperature
rimeter, together with the temperatures on both sides of the ele-
differences.
ment under steady state conditions. Since steady state conditions
In order to determine which parameters in the U-Value relation
are never encountered on a site in practice, the heat flow rate
(Eq. (6)) are of significance for the achievement of precise results, a
and temperatures should be imposed over a sufficiently long
sensitivity analysis is also implemented. The values which are var-
period of time to give a good estimate of the steady state.
ied in terms of this analysis, are those referred in Section 2 as nec-
essary for the determination of the surface temperature through IR
thermography; namely, the reflective ambient temperature, the
thermal emissivity of the surface to measure, the ambient temper-
ature and relative humidity, and the distance between the ther-
mographer and the target. In equation form:

T MEAS ¼ f ðT REF ; e; T IN ; RH;xÞ ð9Þ

The influence of the last two parameters is found to be negligible for

the final result, since only a minor part of the recorded energy is
attenuated by the atmosphere. The factors which have a significant
impact on the measured temperature are the thermal emissivity
and the reflective ambient temperature. The results of the sensitiv-
ity analysis are presented in Fig. 6. It should be noted at this point
that the correct ambient reflective temperature of the measured
target was 22.6 °C and its thermal emissivity was equal to 0.93.
As can be seen from this figure, a deviation of 1 °C in the determi-
nation of the reflective temperature may cause an error of up to Fig. 7. Percentage U-Value calculation sensitivity to thermal emissivity (e) and
10% in the calculation of the surface temperature. This error how- ambient reflected temperature.

ever, is significantly increased when calculating the U-Value, as

shown in Fig. 7. Particularly a deviation of 10% in the surface tem-
perature is converted into a deviation of about 100% in the determi-
nation of the U-Value. Thus, the accuracy of the reflected
temperature plays a very important role in the final estimation of
the U-Value. These results seem to be in agreement with the uncer-
tainty analysis included in [8], proving that the most sensitive var-
iable at the quantification of a thermogram is the difference
between the reflected apparent temperature and the wall tempera-
ture, as well as the assumed emissivity of the building surface to
measure.
Fig. 8 presents the calculation of the reflected temperature for
an indoor and outdoor measurement of the same target (masonry
– ID4, summer season). The left axis of the diagram shows the tem-
perature scale for the outdoor temperature (ambient, surface and
ambient reflected), and the right axis is used for the same temper-
atures for the indoor estimation. From this diagram it can be con-
cluded that during a 24 h cycle, the reflected temperature of an Fig. 8. Ambient reflected, ambient and wall temperature – outdoor (primary axis –
indoor target remains quasi-constant and it is very close to the left) and ambient reflected, ambient and wall temperature – indoor (secondary axis
ambient room temperature; the ambient reflected temperature of – right).
4364 P.A. Fokaides, S.A. Kalogirou / Applied Energy 88 (2011) 4358–4365

Fig. 9. Measurement of masonry U-Value over 168 h and over 24 h – ID 3 (heat flux meter).

A more precise validation of the proposed technique was also achievement of precise results is implemented by means of a sen-
performed with the use of heat flux meters. For this purpose the sitivity analysis. It is found that the most sensitive variable at the
overall heat transfer coefficient (U-Value) of the wall of ID3 was quantification of a thermogram is the reflected apparent tempera-
also experimentally determined. The measured data was analyzed ture as well as the assumed emissivity of the building surface to be
by means of the average method, as described in [18]. This method measured. Also the calculation of the reflected temperature for an
assumes that the conductance or transmittance can be obtained by indoor and outdoor measurement of the same target is also deter-
dividing the mean density of heat flow rate by the mean tempera- mined, showing that the reflected temperature of the same target,
ture difference, the average being taken over a long enough period externally exposed, experiences a much wider range of
of time. The analysis was carried out over a period of 168 h, since temperatures.
the examined building element had a high specific heat per unit Finally, it should be noted that due to the uncertainties of all
area. Also it was verified that the U-Value obtained at the end of methods examined, the smaller deviation shown in some cases be-
the test did not deviate by more than 5% from the value obtained tween the notional U-Values and those estimated with the thermo-
either 24 h before or during a INT(2  h/3) time period of h hours, hygrometer does not necessarily mean that IR thermography is less
where INT was the integer part of the tests duration. accurate. On the contrary the authors believe that IR thermography
In Fig. 9 the U-Value, determined by the heat flux meters, is pre- results may be more correct due to the fact that the method con-
sented. According to the results shown in this figure, the deviation siders also radiation effects such as radiative temperature and
between the measurements over 1 week and the measurements of the emittance of the actual surfaces, which are ignored by the
the last day of the test, present a deviation which is less than 5%. other methods. Additionally, apart from the cost involved in pur-
The mean U-Value as obtained through the week measurements chasing an IR camera, the duration of the measurement procedure
equals 0.5 W/m2 K, whereas the corresponding result for the is much shorter, which is an important advantage.
144–168 h period (last day of the week) is 0.51 W/m2 K at a much
lower variability of the recorded values. These results are also very
Acknowledgement
close to the values provided the IR-thermography and close to the
values obtained by the notional methodology, as well as with the
This work was financed by a research grant from the Research
thermohygrometer, confirming the suitability of IR-thermography
Promotion Foundation of Cyprus under contract FTIUJHIF/FRNI/
for the implementation of the U-Value measurement.
0308 (DIT)/02 supported by EU Structural Funds. The authors
would also like to thank the Energy Service of the Ministry of Com-
6. Conclusions merce, Industry and Tourism of Cyprus for providing the instru-
mentation for this study.
In terms of this work the applicability of the infrared (IR) ther-
mography for the determination of the overall heat transfer coeffi-
cient (U-Value) of building envelopes is evaluated. The References
measurements were conducted in Cyprus in five dwellings within
[1] Kowalewski TA, Ligrani P, Dreizler A, Schulz C, Fey U, Egami Y. Temperature
two seasons, during August 2009 and February 2010. The percent- and heat flux. Springer handbook of experimental fluid mechanics. Berlin,
age absolute deviation between the notional and the measured U- Heidelberg: Springer-Verlag; 2007.
Values for IR thermography are found to be at an acceptable level, [2] Directive 2002/91/EC of the European Parliament and of the Council of 16
December 2002 on the energy performance of buildings.
in the range of 10–20%. The building element which revealed the [3] Balaras CA, Argiriou AA. Infrared thermography for building diagnostics.
greater deviation in all cases tested is the roof, followed by glazing, Energy Build 2002;34:171–83.
most probably due to thermal inertia effects which could not be [4] Grinzato E, Vavilov V, Kauppinen T. Quantitative infrared thermography in
buildings. Energy Build 1998;29:1–9.
controlled. In order to provide possible explanation for the devia-
[5] Avdelidis NP, Moropoulou A. Emissivity considerations in building
tion observed between the notional and the measured U-Values, thermography. Energy Build 2003;35:663–7.
a detailed discussion of possible errors by the implementation of [6] Training notes. Thermographer level 1. Infrared Training Center; 2009.
the relevant EN standard for the determination of U-Value is pre- [7] Schulz A. Infrared thermography as applied to film cooling of gas turbine
components. Meas Sci Technol 2000;11:948–56.
sented. The identification of the parameters in the calculation for- [8] User manual FLIR reporter building, programme version 1.1, publ. no. T559081,
mula of the U-Value which have a significant importance for the revision a348; 2009.
 P.A. Fokaides, S.A. Kalogirou / Applied Energy 88 (2011) 4358–4365 4365

[9] ASTM E1862 – 97(2002)e1. Standard test methods for measuring and [14] EN 673:1997. Glass in building – determination of thermal transmittance (U
compensating for reflected temperature using infrared imaging radiometers. value) – calculation method.
[10] ISO 18434-1:2008 (E). Condition monitoring and diagnostics of machines – [15] ISO 10456:2007. Building materials and products – hygrothermal properties –
thermography – part 1: general procedures. tabulated design values and procedures for determining declared and design
[11] EN ISO 6946:1997. Building components and building elements – thermal thermal values.
resistance and thermal transmittance – calculation method. [16] TN 1745:2002. Masonry and masonry products. Methods for determining
[12] http://www.flir.com. [Retrieved 18.11.09]. design thermal values.
[13] Panayiotou GP, Kalogirou SA, Florides GA, Maxoulis CN, Papadopoulos AM, [17] EN 15603:2008. Energy performance of buildings – overall energy use and
Neophytou M, et al. The characteristics and the energy behaviour of the definition of energy ratings.
residential building stock of Cyprus in view of directive 2002/91/EC. J Energy [18] ISO 9869:1994. Thermal insulation – building elements – in-situ measurement
Build 2010. doi:10.1016/j.enbuild.2010.06.018. of thermal resistance and thermal transmittance.