See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/245145360

Thermal conductivity analysis of a briquette with additive hazelnut shells

Article  in  Building and Environment · July 2005

DOI: 10.1016/j.buildenv.2004.09.008

CITATIONS READS

4 314

1 author:

Burhan Çuhadaroğlu
Karadeniz Technical University
20 PUBLICATIONS   189 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

5th International Anatolian Energy Symposium View project

All content following this page was uploaded by Burhan Çuhadaroğlu on 20 October 2017.

The user has requested enhancement of the downloaded file.

 ARTICLE IN PRESS

Building and Environment 40 (2005) 942–948

www.elsevier.com/locate/buildenv

Thermal conductivity analysis of a briquette with additive

hazelnut shells
- uhadaroğlu
B. C
Department of Mechanical Engineering, Karadeniz Technical University, Trabzon 61080, Turkey

Abstract

Briquette is a building wall component, which is widely used in developing countries. This study, supported by the State Planning
Organization of Turkey, aims to improve the thermal insulation characteristic of the wall briquette by adding hazelnut shells to its
mortar. For this purpose, briquette specimens, which consist of pumice, sand, cement and various ratios of hazelnut shell, have been
prepared and the thermal conductivity of the specimens has been measured in conjunction with parameters such as temperature,
humidity, etc. The results obtained show that the thermal conductivity of the specimens decreases with increasing ratio of addition
of hazelnut shells under all circumstances.
r 2004 Elsevier Ltd. All rights reserved.

Keywords: Briquette; Thermal conductivity; Hazelnut shells

1. Introduction standards during the last decade concentrate on the

whole building. This implies that a building with all
Heating energy consumption in residential buildings its components will have an upper limit on heating
is an important part of total energy consumption in energy use. The revised Turkish thermal insulation
developing countries. According to the data provided by standard (TS 825) [3], which was published in 1998, aims
the International Energy Agency (IEA) [1] as seen in to restrict heat loss and fuel consumption in buildings.
Table 1, the energy consumption by other consumers As seen in other developing countries, the common
tends to decrease over the years in non-OECD building technique in rural areas of Turkey depends on
countries. This fact shows that energy efficiency in wall masonry using briquette. Briquette is made of
residential buildings should be considered as a major mortar of pumice, sand (or gravel), cement and water.
parameter for improving heating energy conservation in Detailed information on briquette preparation and
developing countries. Because of the lack of technolo- production was given by Özdeniz [4]. Common com-
gical innovation on energy conversion in developing plaints about walls made of briquettes are their low
countries, insulation techniques play an important role thermal insulation, high moisture absorption, high
in the conservation of heating energy. condensation risk and high driving rain penetration.
Some efforts have been made to reduce heating energy Traditionally, the walls of buildings are site assembled
consumption in both OECD and non-OECD countries. from basic components with separate materials serving
A study on the common standards for heating energy as structural members, weather shield and thermal
restriction in buildings reveals that [2] while the insulation. In a number of resource-poor developing
standards restricted only the U-values of building countries, buildings are constructed of masonry material
components until recent years, the revised and improved without thermal insulation. The aim of this study is to
combine the functions of structural and thermal
Fax: +90 462 325 5526. insulation on a composed briquette with additive
- uhadaroğlu).
E-mail address: burhan@ktu.edu.tr (B. C hazelnut shells.

0360-1323/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.buildenv.2004.09.008
 ARTICLE IN PRESS
- uhadaroğlu / Building and Environment 40 (2005) 942–948
B. C 943

Table 1
Energy consumption in OECD and non-OECD countries by sector [1]

Sector OECD Non-OECD

1973 2000 1973 2000

Mtoe % Mtoe % Mtoe % Mtoe %

Industry sector 1075.99 37.9 1086.39 30.1 480.31 35.9 1099.66 33.4
Transport sector 721.78 25.5 1219.38 33.8 234.80 17.6 569.36 17.3
Other sectorsa 939.37 33.1 1188.87 32.9 580.05 43.4 1557.24 47.3
Nonenergyuseb 99.20 3.5 116.97 3.2 42.26 3.1 67.33 2.0
Total energy consumption 2836.34 100 3611.61 100 1337.42 100 3293.59 100
a
Compromises agriculture, commercial and public service, residential and non-specified.
b
Covers white spirit, paraffin waxes, lubricants, bitumen and other products.

Hazelnut is a popular product of the Black Sea region specimens were allowed to dry and harden for a week in
of Turkey, and 570 000 ton was produced in 2001, which open air and were then dried at 105
5  C in a drier.
corresponds to approximately 70% of the world’s total
production [5]. Hazelnut shells are waste by products 2.2. Measurement apparatus
which emerge from the processing of inner hazelnut and
are completely consumed in stoves for the heating in The thermal conductivity of a material is a basic
traditional residences in the urban areas of the Black Sea parameter of heat transfer. In order to measure the
region in Turkey. Because of the high moisture content thermal conductivity of a material, the methods of
and the low efficiency of the stoves, hazelnut shells guarded hot plate [8] and heat flowmeter [9] are used. To
consumed as fuel deteriorate outdoor air quality by maintain the flow of the heat produced in an electric
augmenting the particulate matter and other air pollu- resistance from specimens, the guarded hot plate
tion emissions [6,7]. An unknown characteristic of method is an absolute method and the heat does need
hazelnut shells is their low thermal conductivity of not be measured by another method. In the heat
about 0.1 W/mK. The average thermal conductivity of a flowmeter method which was used in the present study,
briquette 0.8 W/mK, which depends on parameters such the heat conducted from specimens was set to cooling
as density, moisture, temperature, etc., can be lowered water and this heat was measured by applying the first
by addition of hazelnut shells to the mortar. In the law of thermodynamics.
present study, the thermal conductivity values of the The apparatus consists of a self-clamping specimen
briquette specimens prepared by addition of hazelnut stack assembly with electrically heated source, calori-
shells have been measured for various additive ratios. meter base, dewar vessel enclosure to ensure negligible
loss of heat and a constant head cooling water supply
tank (Fig. 2). A multipoint thermocouple switch is
2. Experimental procedures mounted on the steel cabinet base and two mercury and
glass thermometers are provided for water inlet and
2.1. Preparation of test specimens outlet temperature readings. Four NiCr/NiAl thermo-
couples are fitted and connections are provided for a
Five types of briquette mortar were prepared for the suitable potentiometer instrument to give accurate
tests as seen in Table 2. The pumice used is brought temperature readings. The specimen is lapped on a
from the eastern border of Turkey. River sand, which comparison material in the dewar vessel and the
has a lower salt ratio than sea sand, was used. Both thermocouples are inserted in the holes to read
pumice and hazelnut shells were sifted in order to temperatures.
prepare mortar appropriate to the sizes of specimens. In
order to improve the experimental accuracy, two 2.3. Measurements conducted
equivalent steel moulds were used to compress the
mortar so that the 10 specimens had been tested totally. After the specimens were dried at 105
5  C in a
While the mortars were being compressed in the moulds, drier, they were cooled and weighed. Their sizes were
particular care was taken to apply different force in measured and the densities were determined by dividing
order to get specimens that had 2 different densities. The the weight by the overall volume including the cores and
perspective view of a mould and a specimen of 25 mm pores. In order to observe the relation between the
diameter and 38 mm length is shown in Fig. 1. The moisture content and the thermal conductivity of the
 ARTICLE IN PRESS
944 - uhadaroğlu / Building and Environment 40 (2005) 942–948
B. C

Table 2
Briquette specimens tested and their contents

Contents Specimens

Br00.0 Br02.3 Br04.5 Br08.7 Br12.5

Pumice (1–2 mm) 32 g 32 g 32 g 32 g 32 g

%38.1 %37.2 %36.4 %34.8 %33.3
Sand 32 g 32 g 32 g 32 g 32 g
%38.1 %37.2 %36.4 %34.8 %33.3
Cement 20 g 20 g 20 g 20 g 20 g
%23.8 %23.3 %22.7 %21.7 %20.9
Hazelnut shell (2–4 mm) — 2g 4g 8g 12 g
— %2.3 %4.5 %8.7 %12.5

different moisture contents as

lave;T 1 W 1 þ lave;T 2 W 2
lave;W ¼ ; (3)
W1 þ W2
where W 1 and W 2 are two different moisture contents
of a specimen. The thermal conductivity of any speci-
men at average density is also predicted by the thermal
conductivity values of two specimens which have
different densities as
lave;T 1 r1 þ lave;T 2 r2
lave;r ¼ ; (4)
r 1 þ r2
where r1 and r2 are the densities of two different
specimens that have the same hazelnut shells additive
ratio.
Fig. 1. Perspective view of the mould and a specimen.
2.4. Regression analysis

A statistical analysis technique was applied to the

specimens, some moisture from the specimens was parameters measured. The regression analysis procedure
allowed to be absorbed and the thermal conductivity can be used to summarize data as well as to study
measurements were done with this moisture content at relationships between variables. A general multiple
three different average temperatures. This procedure regression equation can be expressed as
was repeated with a different moisture content for each
Y i ¼ b0 þ b1 X 1i þ b2 X 2i þ    þ bp X pi þ ei : (5)
specimen. All the measurements and predictions on
thermal conductivity of the specimens are given in The notation X pi indicates the value of the pth
Tables 3–7. The thermal conductivity at average independent variable for case i. The b terms
temperature is predicted by are unknown parameters and the ei terms are indepen-
Pn dent random variables that are normally distributed
li T i
lave;T ¼ Pi¼1
n ; (1) with mean 0 and constant variance of s2 : The
i¼1 T i multiple regression model assumes that there is a normal
where l is a measured thermal conductivity of the distribution of the dependent variable for every
specimen at temperature T and the counter i denotes combination of the values of the independent variables
each datum while n denotes the number of total data. in the model [10]. The thermal conductivity is considered
Similarly, any average parameter f such as temperature, a dependent variable while temperature, moisture
moisture content and density is obtained by content and density are considered as independent
Pn variables.
li fi
fave ¼ Pi¼1n : (2) A commonly used measure of the goodness of fit
i¼1 li of a linear model is R2 ; sometimes called the coeffi-
The thermal conductivity at average moisture content is cient of determination. It is defined as the propo-
predicted by the thermal conductivity values at two rtion of the variation in the dependent variable
 ARTICLE IN PRESS
- uhadaroğlu / Building and Environment 40 (2005) 942–948
B. C 945

1 7

2 1 Electric resistance heater

2 Briquette specimen
6 4
3 Comparison material
3 8 4 Cooling water inlet
thermometer
5 Cooling water outlet
5 thermometer
6 Thermocouples
7 Constant head cooling
13
water supply tank
9 8 Water supply from tap
15
9 Overflow to drain
10 Cooling water inlet to
10 apparatus
11
11 Water flow rate valve
12 Scaled water container
13 Current scale
14 Current control switch
15 Specimen loosen lever
16 Multipoint thermocouple
switch
18 12 17 On/off switch
16 17 14
18 Digital potentiometer

Fig. 2. Schematic view of thermal conductivity measurement apparatus.

Table 3
The measurements and the predictions for the specimen Br00.0

Density Moisture Temperature Thermal Thermal Thermal Thermal

rðkg=m3 Þ content T(1C) conductivity conductivity at conductivity at conductivity at
W(%) measured average average average
(W/mK) temperature moisture content density
(W/mK) (W/mK) (W/mK)

31.4 0.717 0.911

0.43 37.1 0.860 T ave ¼ 37:6  C
1351 42.1 1.101 0.898

29.4 0.659 0.888 W ave ¼ %0:49

0.55 36.4 0.892 T ave ¼ 37:4  C
43.4 1.040 0.934

32.9 0.951 0.954 rave ¼ 1361 kg=m2

0.43 36.6 0.998 T ave ¼ 36:8  C
41.1 0.918 0.969

1370 27.1 0.854 0.984 W ave ¼ %0:43

0.43 38.4 1.049 T ave ¼ 38:2  C
47.3 1.005

and expressed as line, R2 is 0. The significance level of the b terms is

Pn statistically tested using the F statistic expressed as
2 ðY^ i  Ȳ Þ2
R ¼ 1  Pi¼1 ; (6)
n 2 R2 ðn  p  1Þ
i¼1 ðY i  Ȳ Þ F¼ ; (7)
qð1  R2 Þ
where Y^ i is the value of Y predicted by the regression
line, Y i is the value of Y observed and Ȳ is the mean where n is the number of cases in the equation and q is the
value of Y i ’s. If all the observations fall on the regression number of variables entered at this step [10]. The small
 ARTICLE IN PRESS
946 - uhadaroğlu / Building and Environment 40 (2005) 942–948
B. C

Table 4
The measurements and the predictions for the specimen Br02.3

Density Moisture Temperature Thermal Thermal Thermal Thermal

rðkg=m3 Þ content T(1C) conductivity conductivity at conductivity at conductivity at
W(%) measured average average average
(W/mK) temperature moisture content density
(W/mK) (W/mK) (W/mK)

31.3 0.490 0.692

0.49 36.8 0.684 T ave ¼ 37:3  C
41.1 0.854 0.835
1292
31.3 0.681 0.935 W ave ¼ %0:61
0.70 41.2 0.976 T ave ¼ 41:6  C
48.6 1.064 0.814
rave ¼ 1298 kg=m3
32.9 0.611 0.731
0.45 35.6 0.654 T ave ¼ 37:6  C
42.2 0.891 0.793
1305
31.1 0.518 0.842 W ave ¼ %0:51
0.57 41.2 0.833 T ave ¼ 41:4  C
46.6 1.066

Table 5
The measurements and the predictions for the specimen Br04.5

Density Moisture Temperature Thermal Thermal Thermal Thermal

rðkg=m3 Þ content T(1C) conductivity conductivity at conductivity at conductivity at
W(%) measured average average average
(W/mK) temperature moisture content density
(W/mK) (W/mK) (W/mK)

32.8 0.442 0.696

0.18 35.9 0.719 T ave ¼ 38:5  C
43.5 0.868 0.900
1191
31.9 0.804 0.939 W ave ¼ %0:62
0.94 35.7 0.756 T ave ¼ 37:9  C
43.3 1.190 0.747
rave ¼ 1221 kg=m3

31.8 0.462 0.427

0.51 37.8 0.475 T ave ¼ 36:7  C
41.7 0.356 0.602
1265
31.7 0.415 0.744 W ave ¼ %0:59
0.63 37.4 0.951 T ave ¼ 38:8  C
44.2 0.805

observed significance level of the F associated with the computing all possible regression equations, several
slope for independent variables supports the hypothesis other methods do not require as much computation and
that thermal conductivity and independent variables such are more frequently used. Among these procedures are
as temperature and density are linearly related. forward selection, backward elimination and stepwise
A variety of regression models can be constructed regression. None of these variable selection procedures
from the same set of variables. For instance, it can be is ‘‘best’’ in any absolute sense; they merely identify
built up by seven different equations from three subsets of variables that, for the sample, are good
independent variables: three with only one independent predictors of the dependent variable [10].
variable, three with two independent variables, and one In this study, the stepwise selection of independent
with all three. Although there are procedures for variables used on account of this selection is really a
 ARTICLE IN PRESS
- uhadaroğlu / Building and Environment 40 (2005) 942–948
B. C 947

Table 6
The measurements and the predictions for the specimen Br08.7

Density Moisture Temperature Thermal Thermal Thermal Thermal

rðkg=m3 Þ content T(1C) conductivity conductivity at conductivity at conductivity at
W(%) measured average average average
(W/mK) temperature moisture content density
(W/mK) (W/mK) (W/mK)

31.8 0.367 0.575

0.76 35.8 0.841 T ave ¼ 35:8  C
38.9 0.499 0.772
1260
32.3 0.628 0.887 W ave ¼ %1:09
1.30 36.0 0.805 T ave ¼ 37:9  C
42.2 1.154 0.777
rave ¼ 1292 kg=m3
32.7 0.447 0.746
0.69 37.7 0.901 T ave ¼ 38:4  C
42.3 0.838 0.782
1321
31.5 0.568 0.806 W ave ¼ %0:87
1.04 37.4 0.705 T ave ¼ 38:1  C
42.1 1.075

Table 7
The measurements and the predictions for the specimen Br12.5

Density Moisture Temperature Thermal Thermal Thermal Thermal

rðkg=m3 Þ content T(1C) conductivity conductivity at conductivity at conductivity at
W(%) measured average average average
(W/mK) temperature moisture content density
(W/mK) (W/mK) (W/mK)

31.8 0.396 0.582

0.45 36.3 0.453 T ave ¼ 37:7  C
41.2 0.838 0.475
1062
31.5 0.542 0.446 W ave ¼ %0:97
1.64 37.4 0.467 T ave ¼ 36:6  C
43.2 0.359 0.666
rave ¼ 1085 kg=m3
31.8 0.534 0.756
0.49 37.0 0.766 T ave ¼ 37:5  C
41.0 0.999 0.850
1098
31.4 0.787 0.875 W ave ¼ %1:23
1.87 36.3 0.801 T ave ¼ 36:9  C
41.8 1.006

combination of backward and forward procedures and independent variables, i.e. the dependent variable is
is probably the most commonly used method. positively related to each independent variable. The
values of the coefficient of determination show that the
relations between the thermal conductivity and the
3. Results and discussion independent variables are strong in general ð%R2 X50Þ
except in the case of specimen Br12.5. It is also seen in
The results obtained from regression analysis are Table 8 that there is no relation between the thermal
presented in Table 8. As can be seen, the thermal conductivity and the independent variables for
conductivity increases with the increasing values of the the specimen of Br04.5. The significance levels of the
 ARTICLE IN PRESS
948 - uhadaroğlu / Building and Environment 40 (2005) 942–948
B. C

Table 8
Results of the linear multiple regression

Specimen l ¼ b0 þ b1 T þ b2 W þ b3 r

b0 b1 b2 b3 R2 ð%Þ F Significance
level of F

Br 00.0 0.327 1:61 102 — — 52.1 10.892 0.008

Br 02.3 0.579 2:97 102 0.396 — 94.6 78.447 0.000
Br 04.5 — — — — — — —
Br 08.7 0.952 4:60 102 — — 59.2 14.485 0.003
Br 12.5 8.528 — — 8:51 103 48.2 9.316 0.012
2
All specimens 1.033 2:53 10 — 6:77 104 40.4 19.347 0.000

1.1 4. Conclusion
1
Improving the thermal insulation characteristic of the
0.9
wall briquette by adding the hazelnut shells means
λ/λBr00.0

0.8 increasing the heat energy conservation in residential

0.7 buildings. The results of this paper give useful tips for
the energy conservation policies of developing countries
0.6
that have a shortage of energy sources. Even though the
0.5 results pertain to hazelnut shells, many other kind of
0 2 4 6 8 10 12 14 shells such as wallnut shell and pistachio nut shell may
Hazelnut shell additive ratio (%) also be used in briquette mortar as an additive in order
to reduce the thermal conductivity of the wall briquette.
Fig. 3. Variation of dimensionless thermal conductivity with hazelnut
It should also be stated that some strength character-
shell additive ratio.
istics of the briquette must be tested under different
circumstances.

F statistics for the relations reveal that the thermal

conductivity and the independent variables appearing in References
regression equations are linearly related.
[1] http://www.iea.org/statist/keyworld2002/keyworld2002.pdf.
The regression analysis with experimental data for all [2] Dilmac- S- A. Comparison of new Turkish thermal insulation
specimens shows that the thermal conductivity of the standard (TS 825), ISI 9164, EN 832 and German regulation.
specimens is related to the temperature and density in Energy and Buildings 2003;35(2):161–74.
weak relation ð%R2 o50Þ: Despite this weakness, the [3] TS 825 Thermal insulation rules in buildings. Ankara, Turkey:
significance level of the F supports the hypothesis that Turkish Standard Institution; 1998.
[4] Özdeniz MB. Hygrothermal performance of a new briquette
thermal conductivity and temperature/density are line- design. Building and Environment 1996;31(1):59–65.
arly related. [5] http://www.fiskobirlik.org.tr/duretim.xls.
As seen in Fig. 3, the thermal conductivity of the [6] C- uhadaroğlu B, Demirci E. Influence of some meteorological
specimens decreases with increasing hazelnut shells factors on air pollution in Trabzon city. Energy and Buildings
additive ratio. The regression analysis gives a linear 1997;25(3):179–84.
[7] Demirci E, C - uhadaroğlu B. Statistical analysis of wind circulation
formula for the thermal conductivity, l ¼ 0:884  and air pollution in urban Trabzon. Energy and Buildings
1:722 102 (hazelnut shell ratio) with %R2 ¼ 77:2; 2000;31(1):49–53.
which indicates strong but not linear relation with the [8] C 177-97 Standard test method for steady heat flux measurements
values of 0.05 of significance level of the F. The variation and thermal transmission properties by means of the guarded-hot-
of the thermal conductivity in Fig. 3 also shows that the plate apparatus. PA, USA: ASTM; 1997.
[9] E 1225-99 Standard test method for thermal conductivity of solids
12.5% of hazelnut shells additive by weight improves by means of the guarded-comparative-longitudinal heat flow
the thermal conductivity approximately 30% compared technique. PA, USA: ASTM; 1999.
to the specimen without the hazelnut shells additive. [10] Norusis MJ. SPSS Base system user’s guide, 1990.

View publication stats