PERFORMANCE MEASUREMENT AND

MATHEMATICAL MODELLING OF INTEGRATED

SOLAR WATER HEATERS
CELINE GARNIER
BEng in Energy and Environmental
A thesis submitted in partial fulfilment of the requirements of Edinburgh
Napier University for the award of Doctor of Philosophy.
arch !""#
ABSTRACT
$n a period of rapidly growing deployment of sustainable energy sources the
e%ploitation of solar energy systems is imperative. &older climates li'e those
e%perienced in (cotland show a good potential in addressing the thermal energy
requirement of buildings) particularly for hot water derived from solar energy. *he
result of many years of global research on solar water heating systems has outlined the
promising approach of integrated collector storage solar water heaters +$&(,(-./ in
cold climates. *his calls for a need to estimate the potential of $&(,(-. for the
(cottish climate.
*his research pro0ect aims to study and analyse the performance of a newly developed
$&(,(-. for (cottish weather conditions1 optimise its performance1 model its
laboratory and field performance together with its environmental impacts and analyse
its integration into buildings and benefits of such a heating system1 for the primary
purpose of proposing a feasible $&(,(-. prototype. 2aboratory and field e%periments
were performed to investigate the performance of the newly developed $&(,(-. and
the parameters affecting it which were fundamental to modelling its performance. *his
was followed by developing a thermal macro,model able to compare the temperature
variation in different $&(,(-. designs) including internal temperature and e%ternal
weather conditions for a given aspect ratio and to evaluate the performance of this $&(,
(-. for laboratory and field conditions. *his was followed by a three,dimensional
&omputational 3luid Dynamic +&3D/ analysis of the $&(,(-. in order to optimise the
fin spacing as a means of improving its performance. A 2ife &ycle Assessment +2&A/
and monetary analysis considering the whole life energy of the different $&(,(-.
designs were carried out using a previously developed thermal model in order to
establish the most viable $&(,(-. with the smallest carbon footprint. 3inally1 a study
to show how the $&(,(-. could be integrated into buildings and its potential benefits
to builders and households was underta'en.
*hrough this wor'1 important parameters for modelling laboratory and field
performance of $&(,(-. are established. *he innovative modelling tool developed can
predict the bul' water temperature of the $&(,(-. for any orientation and location in
the world with good accuracy. $mprovements of the $&(,(-. fin design were
suggested through the &3D analysis while 'eeping the costs to a minimum. *he $&(,
(-. prototype showed a high commercial potential due to its environmental and
monetary benefits as well as its potential for integration into commonly used solar water
heating installations and modern methods of construction such as roof panels which
could result in a viable commercialisation of the prototype.
i
ACKNOWLEDGEMENTS
$ would li'e to ta'e the opportunity to than' my supervisors r 4ohn &urrie and
Prof *ariq uneer for their support and guidance throughout the course of my research.
3urthermore1 $ am grateful for the assistance $ have received from the technician r $an
&ampbell and his practical 'now,how. 3urther1 $ would li'e to give my appreciation to
the PhD student .aroon 4unaidi and my friend Ale%ander (cott,*onge for providing
advice and support throughout my research. y coach and friend Ale%andra (tellatou
was supporting me everyday of my present tenure for which $ am sincerely grateful.
3inally1 this research would not have happened without the love and support of my
friends and family.
&5est dans l5effort que l5on trouve la satisfaction et non dans la r6ussite. Un plein effort
est une pleine victoire.
*he pleasure lies in ma'ing the effort1 not in its fulfilment.

7andhi1 E%trait des Lettres l'Ashram
ii
DECLARATION
$ hereby declare that the wor' presented in this thesis was solely carried out by myself
at Edinburgh Napier University1 e%cept where due ac'nowledgement is made1 and that
is has not been submitted for any other degree.
88888888888888888
CELINE GARNIER (CANDIDATE)
8888888..
Date
iii
TABLE OF CONTENTS
ABSTRACT i
ACKNOWLEDGEMENTS...........................................................................................ii
DECLARATION...........................................................................................................iii
TABLE OF CONTENTS..............................................................................................iv
LIST OF FIGURES.......................................................................................................ix
LIST OF TABLES.......................................................................................................xvi
NOMEMCLATURE.................................................................................................xviii
CHAPTER 1 INTRODUCTION...................................................................................1
9.9 Energy..........................................................................................................................9
9.9.9 Energy1 environment and society.....................................................................9
9.9.! &urrent energy scenario ...................................................................................!
9.! Energy issues and challenges......................................................................................:
9.!.9 $mplications set by the global energy scenario.................................................:
9.!.! Energy policy and prospects of renewable energy.........................................99
9.; Prospect of (-. in U< and (cotland.......................................................................9=
9.;.9 Energy and Environmental (cene in (cotland...............................................9=
9.;.! (olar mar'et and prospects.............................................................................9=
9.;.; Need for affordable (-. and data modelling +prospect of solar water
heating for (cotland/...............................................................................................9:
9.> *he Present ?esearch Pro0ect....................................................................................9#
9.>.9 Pro0ect aim......................................................................................................9#
9.>.! @b0ectives.......................................................................................................9#
9.>.; @utline of the thesis........................................................................................!"
iv
9.= &oncluding remar's...................................................................................................!9
CHAPTER 2 REVIEW OF THE RELEVANT LITERATURE.............................21
!.9 (olar water heaters.....................................................................................................!!
!.9.9 (olar hot water systems..................................................................................!!
!.9.! (olar thermal collectors..................................................................................!;
!.9.; (olar water heater parameters.........................................................................;9
!.! odelling the solar water heater...............................................................................;:
!.!.9 acro model A *hermal model......................................................................;:
!.!.! icro model A &omputational 3luid Dynamics +&3D/..................................>=
!.; *he integration of solar water heaters.......................................................................>B
!.;.9 2ife cycle assessment +2&A/..........................................................................>B
!.;.! -eather conditions in (cotland......................................................................>#
!.;.; Building integrated (-................................................................................=9
!.> &oncluding remar's...................................................................................................=;
CHAPTER 3 LABORATOR AND FIELD E!PERIMENTS..............................."#
;.9 ethodology..............................................................................................................=>
;.9.9 (olar water design and construction...............................................................=>
;.9.! Assessment and calibration of e%periment equipment...................................=#
;.9.; E%perimental considerations...........................................................................BC
;.! 2aboratory e%periments.............................................................................................B#
;.!.9 E%perimental set up........................................................................................B#
;.!.! E%periment results..........................................................................................:"
;.!.; &omparison of past and current research.......................................................:;
;.!.> Discussion of laboratory testing.....................................................................C9
;.; 3ield e%periments......................................................................................................C!
;.;.9 E%perimental test set up .................................................................................C!
;.;.! E%periment results..........................................................................................C=
v
;.;.; Discussion of field testing..............................................................................#B
;.> Uncertainties and errors associated with measurements...........................................#:
;.>.9 Equipment error and uncertainty....................................................................#:
;.>.! @perational errors...........................................................................................#C
;.= &oncluding remar's...................................................................................................#C
CHAPTER # MODELLING.......................................................................................$$
>.9 acro model A *hermal model...............................................................................9""
>.9.9 Purpose and 2anguage..................................................................................9""
>.9.! &apabilities and 2imitations.........................................................................9""
>.9.; Assumptions.................................................................................................9"9
>.9.> *hermal networ' and fundamental heat transfer analysis............................9";
>.9.= odelling stratification for laboratory conditions........................................99>
>.9.B Digital simulation flow chart........................................................................99:
>.9.: &omputational and e%perimental data comparison......................................99C
>.9.C Error analysis................................................................................................9;=
>.9.# Discussion of thermal models.......................................................................9;B
>.! icro model A &3D.................................................................................................9;C
>.!.9 Purpose and fin optimisation........................................................................9;C
>.!.! &apabilities and limitations..........................................................................9;#
>.!.; odel calibration..........................................................................................9;#
>.!.> odelling results..........................................................................................9>"
>.!.= Discussion on &3D.......................................................................................9>:
>.; &oncluding remar's.................................................................................................9>C
CHAPTER " LIFE CCLE ASSESSMENT (LCA)...............................................1"1
=.9 7oal and (cope .......................................................................................................9=9
=.! aterial inventory...................................................................................................9=;
=.!.9 (tainless,steel...............................................................................................9=;
vi
=.!.! Aluminium....................................................................................................9=;
=.!.; 7lass wool insulation....................................................................................9=>
=.!.> -indow glass...............................................................................................9==
=.!.= *imber..........................................................................................................9=B
=.; Energy analysis........................................................................................................9=B
=.> Environmental impacts............................................................................................9=#
=.= $nterpretation...........................................................................................................9B9
=.=.9 $dentification of Energy and Environmental issues .....................................9B9
=.=.! $&(,(-.s energy and carbon dio%ide savings............................................9B!
=.=.; Energy paybac' time A EPB*......................................................................9BB
=.=.> &arbon dio%ide emission paybac' time A E&PB*.......................................9B:
=.=.= &onclusions and recommendations..............................................................9B#
=.B onetary analysis +A/..........................................................................................9:"
=.B.9 &ollector material costs................................................................................9:"
=.B.! $&(,(-.s costs...........................................................................................9:!
=.B.; onetary paybac' time +PB*/ analysis....................................................9:;
=.: &omparison with other systems and limitations of the study..................................9:=
=.C &oncluding remar's.................................................................................................9:B
CHAPTER % INTEGRATION INTO HOUSING DESIGN..................................1&&
B.9 *owards Dero,carbon homes....................................................................................9::
B.9.9 U< 7overnment $nitiatives..........................................................................9::
B.9.! Eero,carbon housing for U<........................................................................9C"
B.! 3lat,plate collectors for low carbon construction....................................................9C9
B.!.9 E%isting flat,plate collectors.........................................................................9C!
B.!.! (uggested installation for the $&(,(-.......................................................9C:
B.; $ntegration of the $&(,(-. into roof structure......................................................9#"
B.;.9 odern ethod of &onstruction +&/.....................................................9#9
vii
B.;.! ($?( ?oofs...................................................................................................9#!
B.;.; (tructural issues............................................................................................9#;
B.;.> $ntegration of the $&(,(-..........................................................................9#=
B.> &oncluding remar's.................................................................................................9#=
CHAPTER & CONCLUSIONS.................................................................................1$&
:.9 (ummary of conclusions.........................................................................................9#:
:.! &ontribution to 'nowledge......................................................................................!">
:.; Potential future wor'F 2oo'ing bac'1 loo'ing ahead..............................................!"=
REFERENCES2'&
APPENDICES 2'&
viii
LIST OF FIGURES
Fi(. 1)1* +,- 2''% +.+-(/ 01,-+ 23 (425,4 3i.,4 +.+-(/ 62.0789:i2.; REN21
(2''<) 2
Fi(. 1)2* 2''% C2.0789:i2. 9+- 6,9i:, i. :2..+0 2i4 +=7iv,4+.:; BP (2''&)..............&
Fi(. 1)3* E.> 23 /+,- 2''% F200i4 37+4 -+0+-v+0 :2 9-2>76:i2. (R?P) -,:i20; >,:, 3-28
BP (2''&) @ WEC (2''&)................................................................................................<
Fi(. 1)#* +,- 2''& W2-4> ,.> EU i.0:,44+> 6,9,6i:i+0 23 024,- PV; 024,- :1+-8,4
,.> Ai.> 92A+- B D,:, 3-28 :1+ REN21 (2''&) ,.> EC (2''&)...............................1#
Fi(. 1)"* E7-29+,. 024,- 8,-C+:; D,:, 3-28 E7-O50+-vDER (2''&) ,.> E7-29+,.
C288i00i2. (EC) (2''&)...............................................................................................1&
Fi(. 2)%* A>v+-:i0+8+.: 32- :1+ 64i8,x 024,- A,:+- 1+,:+- i. 1<$2; B7::i ,.> P+-4i.
(1$<') 2%
Fi(. 2)&* M2.:14/ 8+,. :+89+-,:7-+ 23 :1+ 0:2-,(+ 347i> 32- 3744 8ix+> :,.C ,.>
0:-,:i3i+> :,.C; C-i0:23,-i +: ,4. (2''3).........................................................................32
Fi(. 2)<* H2: A,:+- >-,A)233 9-23i4+0 32- 0:7>i+0 23 M6L+..,. (2''%) ,.> UK)ISES
32- 2''4?>,/ ,.> 3''4?>,/ -+09+6:iv+ 42,>0..................................................................3%
Fi(. 2)$* E=7iv,4+.: 1+,: 342A 6i-67i: 32- C-7E +: ,4 (2''2) 6244+6:2-; i.074,:i2. ,.>
4200+0 3$
Fi(. 2)1'* H+,: :-,.03+- 8+61,.i08 i. ,. ICS)SWH.................................................#'
Fi(. 2)11* O-i+.:,:i2. ,.(4+ 32- ,i- 6,vi:/ -+(-+00i2.................................................#1
Fi(. 2)12* G+28+:-i6 9,-,8+:+- 23 C-7E +: ,4. (1$$$) :-,9+E2i>,4)01,9+>
024,-?+.+-(/ 0:2-+...........................................................................................................#3
Fi(. 2)13* Si894i3i+> 9-26+00+0 :/9i6,44/ 62.0i>+-+> i. LCA 32- , 9-2>76:.............#&
i%
Fi(. 2)1#* E4+8+.:0 23 , 3744 LCA; E.vi-2.8+.:,4)T+61.242(/)B+0:)P-,6:i6+)
P-2(-,88+ (2''').........................................................................................................#<
Fi(. 2)1"* +,-4/ :2:,4 (425,4 12-iE2.:,4 024,- i--,>i,:i2. i. CW1?82; UK; FG-i +:
,4. (2''&) "'
Fi(. 2)1%* Av+-,(+ ,85i+.: :+89+-,:7-+ -,.(+0 i. E7-29+ 32- :1+ 82.:1 23
H,.7,-/ ,.> :1+ 820: 9-25,54+ ,-+, 23 Ai.:+- 29+-,:i2. 32- :1+ ICSSWH >+0i(.;
S8/:1 +: ,4. (2''1)........................................................................................................."1
Fi(. 2)1&* S24,- 6244+6:2- 9200i54+ i.:+(-,:i2.; H,+1.i( +: ,4. (2''&)........................."2
Fi(. 3)1<* S+6:i2.,4 vi+A 23 :1+ ,478i.i78 ICS SWH 70+> i. :1+ 9-+0+.: A2-C....""
Fi(. 3)1$* ICS 8,i. 9,-:0 ,3:+- 67::i.(; 324>i.( ,.> >-i44i.( 29+-,:i2.0.................""
Fi(. 3)2'* SWH 8,i. 9,-:0 5+i.( ,00+854+> ,.> A+4>+>.........................................."%
Fi(. 3)21* SWH 3i.0 ,00+854+> ,.> A+4>+>................................................................"%
Fi(. 3)22* W,:+- 342A i. 8,.73,6:7-+> 6244+6:2- B 52::28 -i(1: i.4+: 9i9+ (5/9,00
342A) "<
Fi(. 3)23* W,:+- 342A i. CAD >-,Ai.(0 6244+6:2- B 52::28 4+3: i.4+: 9i9+..............."<
Fi(. 3)2#* C24> :+0: i.0:,44,:i2......................................................................................%'
Fi(. 3)2"* H2: :+0: i.0:,44,:i2.......................................................................................%1
Fi(. 3)2%* T1+-8262794+ 920i:i2.0 ,.> v2478+ 62.:-24 ,0026i,:+>; ,* A478i.i78
6244+6:2-; 5* S:,i.4+00)0:++4 C244+6:2- (H7.,i>i +: ,4 (2''")).......................................%2
Fi(. 3)2&* A6-/4i6 94,0:i6 :75+0 i.0+-:0 Ai:1 :1+-8262794+ 042:.................................%3
Fi(. 3)2<* A6-/4i6 94,0:i6 :75+ i.0+-:0 i. 9i9+0............................................................%3
Fi(. 3)2$* A00789:i2. 23 :1+-8262794+ 920i:i2. i. :1+ SWH...................................%#
Fi(. 3)3'* E.+-(/ >i0:-i57:i2. 23 :1+ +4+6:-i6,44/ 1+,:+> 0i4i62.+ -755+- 9,>..........%#
%
Fi(. 3)31* A502-5+- 01++: 23 :1+ ICS)SWH.................................................................%"
Fi(. 3)32* Si4i62. 9,0:+...................................................................................................%"
Fi(. 3)33* H+,:i.( 9,> :1+-82(-,91i6 i8,(+0* ,. Si4i62.+ 0i>+ (1+,:i.( 0i>+); 5.
F-2.: 0i>+; S6,4+* C......................................................................................................%%
Fi(. 3)3#* H+,:i.( 9,> :1+-82(-,91i6 i8,(+0* H2: 092:0 (,) ,.> 6224 092:0 (5)
-+9-+0+.: :1+ 0i4i62.+ A+4>0. S6,4+* C.........................................................................%%
Fi(. 3)3"* T+0: -i( Ai:1 62.:-24 ,.> 8+,07-+8+.: +=7i98+.:..................................%$
Fi(. 3)3%* Ci-67i: >i,(-,8 32- 62.:-24 ,.> 8+,07-+8+.: 23 :1+ 1+,: 347x...............&'
Fi(. 3)3&* T+89+-,:7-+ 0:-,:i3i6,:i2. 9-23i4+ 32- , 2''W?82 1+,: 347x...................&1
Fi(. 3)3<* B72/,.6/ 32-6+0 Ai:1 :i8+...........................................................................&2
Fi(. 3)3$* T+89+-,:7-+ 0:-,:i3i6,:i2. 9-23i4+ ,3:+- < 127-0 23 29+-,:i2..................&3
Fi(. 3)#'* Di8+.0i2.4+00 :+89+-,:7-+ IT1?THJ v0 >i8+.0i2.4+00 >i0:,.6+ I1?HJ ,:
#" i.64i.,:i2.; ,* 1''W?82 1+,: 347x i.97:; 5* #''W?82 1+,: 347x i.97:..............&"
Fi(. 3)#1* W,:+- :+89+-,:7-+ 9-23i4+0 23 3i..+> ,478i.i78 ,.> 0:,i.4+00)0:++4 ICS)
SWH0 ,: #''W; #" i.64i.,:i2.....................................................................................&%
Fi(. 3)#2* W,:+- :+89+-,:7-+ 9-23i4+0 23 3i..+> ,478i.i78; 3i..+> ,.> 7.)3i..+>
0:,i.4+00)0:++4 ICS)SWH0 32- +i(1: 127-0 23 +x9207-+ :i8+ ,: 2''W.......................&&
Fi(. 3)#3* ICS)SWH A,:+- :+89+-,:7-+ i.6-+,0+* 7.)3i..+> v+-070 3i..+>
0:,i.4+00)0:++4 &<
Fi(. 3)##* ICS)SWH A,:+- :+89+-,:7-+ i.6-+,0+* 7.)3i..+> 0:,i.4+00)0:++4 v+-070
3i..+> ,478i.i78...........................................................................................................&<
Fi(. 3)#"* A478i.i78 ,.> 0:,i.4+00)0:++4 ICS)SWH +33i6i+.6/ 6289,-i02. ,: #"
>+(-++ i.64i.,:i2.; 1''W?#''W....................................................................................<'
Fi(. 3)#%* C-200)0+6:i2.,4 -+9-+0+.:,:i2. 23 :1+ 3i+4> :+0:+> ICS)SWH....................<2
%i
Fi(. 3)#&* Ex9+-i8+.:,4 :+0: -i(....................................................................................<3
Fi(. 3)#<* B,6C i.074,:i2. 23 -223 0i874,:+> 52x........................................................<3
Fi(. 3)#$* Ex9+-i8+.:,4 :+0: -i( 0i>+ i.074,:i2...........................................................<#
Fi(. 3)"'* Fi+4> +x9+-i8+.: >,:,..................................................................................<"
Fi(. 3)"1* O.+ >,/ 3i+4> +x9+-i8+.: >,:,....................................................................<%
Fi(. 3)"2* SC/ -,>i,:i2. +33+6: 2. 62v+- :+89+-,:7-+ Ai:1 C6 ,0 :1+ 64+,- 0C/ i.>+x
<&
Fi(. 3)"3* E.+-(/ .+:A2-C 32- 64+,- 0C/ 62.>i:i2.0...................................................<<
Fi(. 3)"#* S:-,:i3i6,:i2. Ai:1 :i8+ ,.> 024,- -,>i,:i2................................................<<
Fi(. 3)""* Di8+.0i2.4+00 42.(i:7>i.,4 0:-,:i3i6,:i2.* T1 ? T52::28 v+-070 1?L.......<$
Fi(. 3)"%* E33i6i+.6/ 4i.+ 32- :1+ ICS)SWH 0:7>i+>...................................................$1
Fi(. 3)"&* S:-,:i3i6,:i2. i. :1+ ICS)SWH ,.> A,:+- :+89+-,:7-+ >-,A)233............$3
Fi(. 3)"<* D-,A)233 9-23i4+ ,.> 9+-32-8,.6+..............................................................$#
Fi(. 3)"$* A.:+ M+-i>i,. >i8+.0i2.4+00 42.(i:7>i.,4 0:-,:i3i6,:i2. Ai:1 >-,A)233*
T1 ? T52::28 V0 1?H.....................................................................................................$"
Fi(. 3)%'* P20: M+-i>i,. >i8+.0i2.4+00 42.(i:7>i.,4 0:-,:i3i6,:i2. Ai:1 >-,A)233*
T1 ? T52::28 V0 1?H.....................................................................................................$%
Fi(. #)%1* T1+-8,4 .+:A2-C 23 :1+ 0/0:+8.................................................................1'#
Fi(. #)%2* H+,: 347x +x61,.(+0 2667--i.( ,: :1+ ICS)SWH.....................................1'"
Fi(. #)%3* D+A)92i.: ,.> ,85i+.: :+89+-,:7-+0 -+4,:i2.01i9 32- 64+,- >,/0........1'$
Fi(. #)%#* D+A)92i.: ,.> ,85i+.: :+89+-,:7-+0 -+4,:i2.01i9 32- 2v+-6,0: >,/0. .1'$
%ii
Fi(. #)%"* G+28+:-i6 9,-,8+:+-0 23 :1+ ICS)SWH (547+ .2>+0 >+.2:+ 347i>
:+89+-,:7-+ 8+,07-+8+.:0 426,:i2.0).......................................................................11'
Fi(. #)%%* S:-,i(1: -+6:,.(74,- 3i.0 23 7.i32-8 6-200 0+6:i2. ,.> +.+-(/ 5,4,.6+ 32-
,. +x:+.>+> 07-3,6+.....................................................................................................111
Fi(. #)%&* Di8+.0i2.4+00 3i. :+89+-,:7-+ v+-070 3i. 4+.(:1.....................................113
Fi(. #)%<* Di8+.0i2.4+00 :+89+-,:7-+ 0:-,:i3i6,:i2. ,3:+- < 127-0.........................11"
Fi(. #)%$* Si874,:+> 3i. 1+,: :-,.03+- -,:+................................................................11<
Fi(. #)&'* Si874,:+> 3i. :+89+-,:7-+ >-29 A1+-+ i0 :1+ +x6+00 :+89+-,:7-+
,61i+v+> ,0078i.( $'K 23 1+,: i0 62.>76:+> :1-27(1 :1+ 3i.0. .............................11$
Fi(. #)&1* Si874,:+> 3i. 1+,: :-,.03+- 62+33i6i+.:.....................................................12'
Fi(. #)&2* Si874,:+> 3i. 1+,: :-,.03+- -,:+ Ai:1 :i8+ B A478i.i78 388 SWH. . .12'
Fi(. #)&3* Si874,:+> 3i. +33i6i+.6/..............................................................................121
Fi(. #)&#* ICS)SWH +33i6i+.6/ Ai:1 :i8+ 32- #''W?82 1+,: 347x ,994i+>.............122
Fi(. #)&"* C2897:,:i2.,4 ,.> +x9+-i8+.:,4 6289,-i02. 32- 2"'W?82 1+,: 347x. 123
Fi(. #)&%* C2897:+> v0 Ex9+-i8+.:,4 A,:+- :+89+-,:7-+0....................................123
Fi(. #)&&* C2897:+> v0 Ex9+-i8+.:,4 0:-,:i3i6,:i2. ,3:+- < 127-0 23 29+-,:i2. 32-
,44 1+,: i.97:0 12&
Fi(. #)&<* C2897:+> v0 Ex9+-i8+.:,4 0:-,:i3i6,:i2. ,3:+- 22 127-0 32- ,44 1+,:
i.97:0 12<
Fi(. #)&$* 2''W?82 >,:, ) C2897:+> v0 Ex9+-i8+.:,4 0:-,:i3i6,:i2. ,3:+- < 127-0
23 29+-,:i2. 12$
Fi(. #)<'* V,4i>,:i2. 23 :1+ 82>+4 ) C2897:+> v0 Ex9+-i8+.:,4 0:-,:i3i6,:i2. ,3:+-
< 127-0 23 29+-,:i2......................................................................................................13'
Fi(. #)<1* C2897:+> ,.> +x9+-i8+.:,4 >,:, 32- , 3iv+ >,/ 9+-i2> i. H74/ 2''&. . .131
%iii
Fi(. #)<2* C2897:+> v0 Ex9+-i8+.:,4 32- , 3iv+ >,/ 9+-i2> i. H74/ 2''&..............132
Fi(. #)<3* Si874,:+> +33+6: 23 Ai.> v+426i:/ 2. A,:+- :+89+-,:7-+ 32- , 3iv+ >,/
9+-i2> i. H74/ 2''& A1+-+ V'; V2; V# ,.> V% -+9-+0+.:0 :1+ Ai.> v+426i:/ v,47+0
23 '8?0; 28?0; #8?0 ,.> %8?0 -+09+6:iv+4/.................................................................133
Fi(. #)<#* C2897:+> v0 Ex9+-i8+.:,4 32- , "' >,/ 9+-i2> i. H74/.........................13#
Fi(. #)<"* T+89+-,:7-+ 0:-,:i3i6,:i2. 23 3i.0 ,.> 8i>>4+ 4i.+ 23 A,:+- 6244+6:2-...1#1
Fi(. #)<%* V+426i:/ 9-23i4+ B # 3i.0; :29 vi+A; ,3:+- 2' 8i........................................1#2
Fi(. #)<&* V+426i:/ 9-23i4+ B # 3i.0; 0i>+ vi+A; ,3:+- 2' 8i.......................................1#2
Fi(. #)<<* L2.(i:7>i.,4 A,:+- :+89+-,:7-+ 0:-,:i3i6,:i2. B # 3i.0; 3-2.: vi+A; ,3:+-
2'8i. 1#3
Fi(. #)<$* L2.(i:7>i.,4 A,:+- :+89+-,:7-+ 0:-,:i3i6,:i2. B # 3i.0; 0i>+ vi+A; ,3:+-
2'8i. 1##
Fi(. #)$'* V+426i:/ 9-23i4+ B " 3i.0; :29 vi+A; ,3:+- 2' 8i........................................1#"
Fi(. #)$1* V+426i:/ 9-23i4+ B " 3i.0; 0i>+ vi+A; ,3:+- 2' 8i.......................................1#"
Fi(. #)$2* L2.(i:7>i.,4 A,:+- :+89+-,:7-+ 0:-,:i3i6,:i2. ) " 3i.0; 0i>+ vi+A; ,3:+-
2'8i. 1#%
Fi(. #)$3* H+,: :-,.03+- ,.> A,:+- :+89+-,:7-+ 9-23i4+0 i. :1+ ICS)SWH............1#&
Fi(. #)$#* T2:,4 >,:, ) 8+,07-+> ,.> 62897:+> 32- , "' >,/ 9+-i2>* &:1 H7.+ B
2&:1 H74/ 2''& 1"'
Fi(. ")$"* S+6:i2.,4 vi+A 23 :1+ ,478i.i78 ICS SWH. N2:+* , >i8+.0i2. >-,Ai.( i0
012A. i. Fi(7-+ 3)1.....................................................................................................1"2
Fi(. ")$%* G4,00 A224 9-2>76:i2. 9-26+00; S27-6+* ERIMA (2''&).........................1""
Fi(. ")$&* Wi.>2A (4,000 9-2>76:i2. 9-26+00; S27-6+* M,>+ H2A (2''&).............1"%
Fi(. ")$<* U0+374 +.+-(/ 0,v+> ,..7,44/ 32- +,61 ICS)SWH...................................1%3
%iv
Fi(. ")$$* S24,- 3-,6:i2. 23 AL)3; AL)1." ,.> ST)1.". N2:+* 024,- 3-,6:i2. i0 :1+
,827.: 23 +.+-(/ 9-2vi>+> 5/ :1+ ICS)SWH >ivi>+> 5/ :1+ +.+-(/ -+=7i-+> :2
-,i0+ A,:+- :+89+-,:7-+ :2 ""2C................................................................................1%#
Fi(. ")1''* A..7,4 CO2 0,vi.(0 5/ 37+4 :/9+ 70i.( ICS)SWH0..............................1%"
Fi(. ")1'1* ICS)SWH ECPBT* C,:+(2-i0,:i2. Ai:1 -+09+6: :2 37+4 :/9+...............1%<
Fi(. ")1'2* C20: 6289,-i02........................................................................................1&'
Fi(. %)1'3* T2A,->0 E+-2 6,-52. 128+0* ,. R7-,4LED; 5. E621270+; 6. Li(1:1270+;
>. Si(8, 1<1
Fi(. %)1'#* S+9,-,:+ 024,- :,.C 9-+)1+,:i.( :1+ >A+44i.( 12: A,:+- :,.C..............1<3
Fi(. %)1'"* C285i.+> 024,- >A+44i.( 12: A,:+- :,.C...............................................1<#
Fi(. %)1'%* Ex:+-.,4 1+,: +x61,.(+-..........................................................................1<"
Fi(. %)1'&* C285i)52i4+- 024,- 6289,:i54+ i.0:,44,:i2.............................................1<%
Fi(. %)1'<* S24,-:Ai. 0/0:+8.......................................................................................1<&
Fi(. %)1'$* ICS)SWH i.0:,44,:i2................................................................................1$'
Fi(. %)11'* S:-76:7-,4 I.074,:+> R223 S/0:+8 ) SIRS; W22> (2''%).......................1$2
Fi(. %)111* S:-76:7-,4 I.074,:+> R223 S/0:+8; H,8+0 H21.0 -223 :-i,4...................1$3
Fi(. %)112* I.0:,44,:i2. 061+8,:i6 ,.> 3i.,4 i.0:,44,:i2. 23 VELU! Ai.>2A0;
VELU! (2''") 1$#
%v
LIST OF TABLES
T,54+ 1)1* E8i00i2. -,:+0 23 +.+-(/ 027-6+0 (DEFRA; 2''<,)....................................3
T,54+ 2)2* D+:,i40 23 >-,A)233 9,::+-.0 70+> 5/ H7.,i>i +: ,4. (2''<).......................3%
T,54+ 2)3* C-i:i6,4 ,.(4+ v+-070 ,09+6: -,:i2; H244,.>0 +: ,4. (1$&%).........................#1
T,54+ 3)#* A478i.i78 ,.> S:,i.4+00)0:++4 91/0i6,4 9-29+-:i+0 ,: 3''K..................."%
T,54+ 3)"* E=7i98+.: ,.> S9+6i3i6,:i2.0...................................................................."$
T,54+ 3)%* S:,i.4+00)0:++4 v0 ,478i.i78 2v+-,44 +33i6i+.6i+0 32- >i33+-+.: 1+,: 347x*
N2:+ 4,52-,:2-/ :+0:0.....................................................................................................<1
T,54+ 3)&* D,i4/ 12: A,:+- >+8,.> 9-23i4+.................................................................$2
T,54+ #)<* Ov+-,44 i89-2v+8+.: 3,6:2- 32- (iv+. 9,-,8+:+-0................................122
T,54+ #)$* S:,:i0:i6,4 i.>i6,:2-0 v,47+0 B T2:,4 D,:,................................................12"
T,54+ #)1'* S:,:i0:i6,4 i.>i6,:2-0 v,47+0 B T2:,4 D,:,..............................................12$
T,54+ #)11* S:,:i0:i6,4 i.>i6,:2-0 v,47+0....................................................................132
T,54+ #)12* S:,:i0:i6,4 i.>i6,:2-0 v,47+0....................................................................13#
T,54+ ")13* E852>i+> +.+-(/ 23 8,:+-i,40...............................................................1"&
%vi
T,54+ ")1#* M,00 5,4,.6+ 23 :1+ 6244+6:2-0................................................................1"&
T,54+ ")1"* E852>i+> +.+-(/ ,.> i.6i>+.6+ 23 8,:+-i,40 2. :1+ +.+-(/ 5,4,.6+.1"<
T,54+ ")1%* E.+-(/ ,.,4/0i0 32- , 382 i.0:,44,:i2...................................................1"<
T,54+ ")1&* E852>i+> 6,-52. >i2xi>+ 23 8,:+-i,40.................................................1"$
T,54+ ")1<* E852>i+> +.+-(/ ,.> i89,6:0 23 8,:+-i,40 2. :1+ +.+-(/ 5,4,.6+. . .1%'
T,54+ ")1$* T2:,4 CO2 +8i00i2.0 32- , 382 i.0:,44,:i2...........................................1%'
T,54+ ")2'* EPBT........................................................................................................1%%
T,54+ ")21* ECPBT.....................................................................................................1%&
T,54+ ")22* R,A 8,:+-i,4 620:0 9+- 0=7,-+ 8+:+- 23 6244+6:2-................................1&2
T,54+ ")23* C,0+ 1 ) C244+6:2-0 MPBT >+9+.>i.( 2. ,7xi4i,-/ 1+,:i.( 0/0:+8 32- ,
382 0/0:+8 1&#
T,54+ ")2#* C,0+ 2 B Ov+-,44 ICS)SWH0 MPBT......................................................1&#
%vii
NOMEMCLATURE
A55-+vi,:i2. ,.> S/85240
EPBD Energy Performance of Buildings Directive
EPDM Ethylene Propylene Diene ,class rubber
ICS)SWH $ntegrated &ollector (torage (olar -ater .eater
LLCT 2ow and Eero &arbon *echnologies
MBE ean Bias Error
MMC odern ethods of &onstruction
RMSE ?oot ean (quare Error
SWH (olar -ater .eater
SIRS (tructural $nsulated ?oofing (ystem
G-++C ,.> S7506-i9:0
wb
T
bul' water temperature <
h
T
water temperature at local point <
b
T
bottom collector water temperature <
t
T
top collector water temperature <
w
T
water temperature at tG i <
w
TH
water temperature at tG iI9 <
ini w
T
1 initial water temperature at tG " <
a
T
ambient temperature <
p
T
absorber plate temperature <
c
T
glass cover temperature <
Sky
T
s'y temperature <
dp
T
dew point temperature <
x f
T
1 fin temperature at fin position % <
w
dT
temperature difference of the water at a time interval dt <
m
T
rise in mean water temperature ini w w m
T T T
1

<
o

e%cess temperature at %G" G base of the fin + W p o

T T
/ <
dimensionless temperature +
( )
wb wb h
T T T J
/
D depth of the collector m
+
h vertical length in the collector + Sin H h
+
/ m
%viii
H total length of the collector m
h local length in the collector m
K h
dimensionless length + H h h J K /

tilt angle of the absorber plate rad

B

buoyancy force N
b

density of the water at the bottom of the collector 'gJm

;
t

density of the water at the top of the collector 'gJm

;
!
the standard gravity of #.C9NJ'g NJ'g
dt the time interval in second s
7+/
e
rate of incident solar radiation transmitted -
"
heat flu% applied to the +solar/ &ollector (urface -Jm
!
#sef$l
%
useful energy transferred to the water -
a w
%

water,ambient energy lost -
w p
%
absorber plate,water energy lost -
s c
%

glass cover,surrounding heat loss -
c p
%
absorber plate,glass cover energy lost -
ins
%
total energy transferred from the fins to the water -
f
%
fin heat transfer rate per stripes -
pc
#
absorber plate,glass cover overall U,value -J<
wa
#
water,ambient overall U,value -J<
ca
#
glass cover,ambient overall U,value -J<
pw
#
absorber plate,water overall U,value +G
w
h
/ -J<
&ns
'
resistance of the insulation material <J-
Wind
'
resistance occurring at the bo% surface in contact with the wind <J-
pc
h
absorber plate,cover convection coefficient
-Jm
!
. <
Wind
h
glass cover,ambient e%ternal convection coefficient -Jm
!
. <
w
h
absorber plate,water convection coefficient -Jm
!
. <
f
h
fin heat transfer coefficient per stripes -Jm
!
. <
fa
k
thermal conductivity of air -Jm. <
fw
k
thermal conductivity of water -Jm. <
w
(
thermal capacitance of the water 4J<
%i%
c
(
specific heat of the collector material 4J<
i
(
thermal capacitance of the insulation 4J<
wd
(
thermal capacitance of the wood bo% 4J<
!
(
thermal capacitance of the glaDing 4J<
m
(
overall thermal capacitance of the material 4J<
( overall thermal capacitance of the system 4J<
w
)
wind speed mJs
tilt angle of the absorber plate rad
c
A
cross sectional area m
!

f
A
fins area
/ + w x *
m
!
t
thic'ness of the fins m
w
width of the fins m
+ perimeter m
x
fins position m

s
x
strip length ta'en equal to m ""! . " for modelling m
d depth of the fins m
(tefan,BoltDmann constant 4Js.m
!
.<
>

$P improvement factor
* number of fins
c

emissivity of the glass cover

+

emissivity of the absorber plate

B

bul' emissivity temperature of the absorber plate

s$r

bul' emissivity temperature of the glass cover

Sky

s'y emissivity
(
kA h+ m
!
%%
CHAPTER 1 INTRODUCTION
*his chapter reviews present day energy usage1 current energy issues and challenges1
prospects for renewable energy and in particular potential for solar energy. $t also
discusses the requirements for studies on solar water heaters +(-./ and specifically
introduces the potential for integrated collector storage solar water heaters +$&(,(-./
in the U< and conte%tualises the current research. $t also summarises the problem
statement and ob0ectives of the present research and provides an outline of the thesis.
1.1 E.+-(/
1.1.1 E.+-(/; +.vi-2.8+.: ,.> 026i+:/
Energy1 by definition1 is the capacity to perform wor'. (olar energy supports virtually
all life on Earth through photosynthesis and drives the climate and weather. *his energy
can be developed by means of a variety of natural and synthetic processes. Plants
capture solar radiation by photosynthesis and convert it to chemical form1 while direct
heating or electrical conversion is used by solar equipment to generate electricity or to
do other useful wor'. Even the energy stored in widely used energy resources li'e
petroleum and other fossil fuels was originally converted from sunlight by
photosynthesis over time.
Energy use and supply plays an essential role in society and human life. Economic
development and the wellbeing of society are closely lin'ed to energy. *he continual
growth in populations and venture for a better economy resulted in a dramatic increase
of energy consumption1 particularly in the last two centuries due to the increase of
available energy sources. -ith the evolution of society1 humans progressively gained
access to larger amounts of energy and increased significantly their energy consumption
primarily in the more developed nations to cover industrial1 transport1 space,heating1
lighting and refrigeration needs. odern societies use more energy every year for
industry1 services1 homes and transport and ta'e for granted easily available energy.
.owever1 the increasing concern of climate change and the effect of burning fossil fuels
as well as global awareness of the limited supply of fossil fuels such as oil1 coal or
natural gas shows that energy does define and constrain our progress. *hus1 to optimise
9
the future of our society the allocation of Earth5s energy resources needs to be
understood and prioritised.
1.1.2 C7--+.: +.+-(/ 06+.,-i2
-orld energy sources can be divided into two main categoriesF non,renewable and
renewable. Non,renewable energy sources define energies which cannot be replenished
in a short time period and that eventually become too e%pensive and too
environmentally damaging to recover. -orldwide1 non,renewable energy sources are
predominant including fossil fuels and uranium for nuclear power. 3ossil fuels range
from very volatile materials li'e natural gas1 to liquid oil1 to non,volatile materials such
as coal accounting for :#L of the world5s primary energy supply as shown in 3igure 9,
9 from ?EN!9 +!""C/ data. ?enewable energy sources can be replenished naturally in a
short period of time and will never run out. *hey include solar1 wind1 geothermal1
biomass energy as well as hydropower and ocean energy.
Fossil fuels
79%
Nuclear
3%
Renewables
18%
Biofuels 0.3%
Power generation 0.8%
Hot water/heating 1.3%
Large hydropower 3%
Traditional bioass
13%
Fi(. 1)1* +,- 2''% +.+-(/ 01,-+ 23 (425,4 3i.,4 +.+-(/ 62.0789:i2.; REN21 (2''<)
1.1.2.1 N2. -+.+A,54+ 027-6+0
(olid fossil fuels were the first energy resource used. &oal remains the world5s most
abundant and fastest,growing fossil fuel in !""B accounting for about !;L of the global
power usage based on ?EN!9 +!""C/ fossil fuel data and specification fuel data stated
by BP +!"":a/ and BP +!"":b/. &oal remains the leading source for electricity
generation as well as the largest worldwide source of carbon dio%ide emissions. @il is
the leading fuel in the transport sector therefore ma'ing it very vulnerable to any
disruption in oil price and supply as stated by Buc'lin +!"";/. @il accounted for ;>.=L
!
of the global power usage according to ?EN!9 +!""C/1 BP +!"":a/ and BP +!"":b/
data. Natural gas has been used for over a century for lighting and heating and is now
considered a very valuable resource) being more efficient and cleaner than other fossil
fuels releasing lower amounts of carbon dio%ide emissions per unit energy released as
shown in *able 9,9.
T,54+ 1)1* E8i00i2. -,:+0 23 +.+-(/ 027-6+0 (DEFRA; 2''<,)
F7+4
:/9+
E8i00i2. -,:+0
C(CO
2
?CW1
Electricity ".>;
Natural 7as ".9#
@il ".!=
&oal ".;"
Petrol ".!>
Nuclear ".""# to "."9>
?enewables "
-orld natural gas consumption accounts for !9.=L of the global power usage. BP
+!"":a/ states that European consumption decreased mainly due to large increases in
contracted prices in the U< and Eastern European countries resulting in large
consumption decline. *he ma0or disadvantage in the use of natural gas is its
transportation which is more complicated and e%pensive than other fossil fuels.
Nuclear energy accounts for ;L of the global power usage as stated by ?EN!9 +!""C/.
*he -orld Energy &ouncil +-E&/ +!"":/ stated that at the current rate of production1
using current reactor technology1 global nuclear reserves are estimated to last for almost
another C= years. $ts technology is sometimes promoted as a sustainable energy source
that reduces carbon emissions and can increase energy security for countries with access
to this technology by decreasing dependence on fossil fuel sources. .owever1 there are
political1 security and environmental concerns about nuclear reactor safety as well as
radioactive waste disposal and plant decommissioning.
Despite the world5s increasingly heavy dependence on this leading energy source1
depleting fossil fuel resources do not ma'e them reliable choices for the future and
provide impetus for moving the economy towards sustainable energy sources.
;
1.1.2.2 R+.+A,54+ +.+-(/
?enewable energy can be defined as energy flows which are replenished at the same
rate as they are MusedN as stated by (orensen +!""!/. (olar radiation is the principal
source of earth5s renewable energy sources. (ome of the common renewable energy
technologies and their current status in the global energy mi% are reviewed in the
following paragraphs.
N2.)024,- -+.+A,54+0
*idal and geothermal energy do not depend on solar radiation. *idal power traps water
in a basin activating turbines generating electricity as it is released through the tidal
barrage. *idal energy power generation is still in an early stage of development and
accounts for only "."9L of the world5s energy consisting of ".; 7- reported by
?EN!9 +!""C/. .owever1 based on a recent report from the (ustainable Development
&ommission +(D7/ Appleyard +!"":/ stated that U<5s tidal resources could provide
9"L of the U<5s electricity in the near future.
7eothermal energy meaning M.eat from the EarthN is the energy generated by the heat
contained within the Earth. 7eothermal energy is used for power and for heating.
*rapping geothermal energy for power generation can be achieved using hydrothermal
reservoirs1 hot dry roc'1 geopressure brines and magma1 each with engineering
challenges and constraints ma'ing them not always practical or economically feasible.
2ow temperature geothermal resources by heat e%traction from the near sub,surface of
the Earth have been widely used in the past and provide energy for space1 water heating1
district heating1 greenhouse heating or warming of fish ponds in aquaculture. *he
relatively constant temperature of the top 9= metres of the EarthHs surface can be used to
heat or cool buildings indirectly through the use of heat pumps. *he report published by
?EN!9 +!""C/ states by the end of !"": worldwide use for electricity had reached
9" 7-1 with an additional ;; 7- used directly for heating with half being geothermal
heat pump installations and accounted for ".!L of the global energy usage.
S24,- +.+-(/* I.>i-+6: 70+0
(olar energy can be converted indirectly to useful energy by other energy forms. (olar
radiation drives a hydrologic cycle by causing evaporation1 precipitation and surface
run,off. .ydropower is defined as the energy of moving water and has been e%ploited
>
for many years for irrigation purposes or watermills. *he largest hydropower in use
today generates electricity by transforming potential energy of water stored at an
elevation into 'inetic energy by the rotation of a turbine rotating the motor to produce
electricity. Based on ?EN!9 +!""C/ report hydroelectricity generates about 9BL of the
world5s electricity in !""B consisting of ::" 7- of large hydro plants and :; 7- of
small hydropower installation.
(olar radiation results in differences of temperature in the atmosphere and oceans in
such manner that the convective currents produce winds1 ocean currents and waves.
-ind power has been commonly used for centuries by windmills for pumping water or
crushing corn. -ind power can be produced on a large scale such as on land,based or
offshore wind farms connected to electrical grids or by individual wind turbines
providing electricity to isolated rural areas. -ind is now one of the most advanced
renewable energies due to its mass production1 improvement in quality1 reliability and
cost effectiveness. ?EN!9 +!""C/ reported that wind power capacity accounts for :>
7- or ".;L of global energy usage.
-ave power uses ocean surface motion caused by winds and is mainly used for
electricity generation. ost companies and infrastructures are concentrated along the
U< coastline according to &ameron +!"":/. A tremendous amount of energy is
available in ocean waves though initial technical problems associated with this
technology have delayed its development.
Biomass1 and especially the fuels derived from biomass named Biofuels1 is another
indirect manifestation of solar energy. Biofuels have recently been sub0ect to increasing
attention as interest in sustainable fuel sources grows. *raditional biomass accounts for
about #L of the 9;L of global biomass energy usage as stated by ?EN!9 +!""C/.
Biomass can be classified as a renewable resource if a sustainable balance is therefore
maintained between carbon emitted and absorbed. *he combustion of biomass fuel
emits &@
!
to the atmosphere1 however emission are no more than the amount it
absorbed during its lifetime growth. Biomass is used in power and heating with an
estimated capacity of >= 7- in !""B in Europe with two third used for heating.
S24,- +.+-(/* Di-+6: 70+0
(olar energy is often seen as the fuel of the future and represents the energy generated
directly from the sun. Applications vary from the residential1 commercial1 industrial1
=
agricultural and transportation sectors. (olar energy can be used in two main ways to
produce heat and electricity.
(olar electricity generation has been developed primarily through photovoltaic +PO/ and
solar thermal power generation. PO is generally used for small and medium,siDed
applications and can be grid,connected or autonomous. ?EN!9 +!""C/ reported that PO
connected grids represent the fastest growing global power generation technology
reaching an estimated cumulative installed capacity of :.C 7- at end of !"":. ?EN!9
+!""C/ estimated the cumulative e%isting solar PO at 9".=7- by the end of !"":
accounting for about ".";L of the global energy usage.
&oncentrating solar thermal power +&(P/ generation is more commonly built for large,
scale electricity generation. *his system uses direct solar radiation concentrated by
lenses or mirrors and trac'ing systems to provide high temperature heat for generating
electricity or directly generating electricity by using PO. *oday1 total installed &(P
capacity is estimated by ?EN!9 +!""C/ to be ".>7- accounting for about ".""!L of
global energy usage. .owever1 the -orld Energy &ouncil +!"">/ reported that on a
long term scenario1 the contribution of &(P could reach B;"7- by !">". *hese
technologies however have limited use in cloudy locations as they require direct solar
radiation.
(olar thermal applications are the most widely used solar energy technology and include
solar coo'ing1 deto%ification1 desalination or solar thermal systems. (olar distillation1
pasteurisation and desalination purify water using solar radiation by means of different
processes. (olar drying and solar ponds are different thermal methods to provide
process heat to dry agricultural products1 clothes or to reach high temperature for
chemical reactions and melting of metals. (olar coo'ers capture sunlight which is
converted to heat retained for coo'ing. (olar coo'ers are used for coo'ing1 drying or
even pasteurising water and mil'. $t can be a real solution for problems such as fuel
poverty1 impure mil' and drin'ing water and health problems caused by indoor air
pollution from combustion of hydrocarbon faced by developing countries often with
high solar energy potential.
3inally1 solar thermal collector applications are the most widely used solar energy
technology for domestic hot water and space heating using only sunlight to heat water.
Different designs and techniques are available depending on the application and
temperature required. 7laDed solar water heaters +(-./ such as flat plate collectors1
batch systems or evacuated tube and air collectors typically used for space heating are
the most common types while unglaDed collectors are generally used for heating
B
swimming pools. (-. systems are described in more detail in &hapter !. (-.
systems are efficient and reliable technologies compared to other solar technologies and
are gaining ground in a few countries. ?EN!9 +!""C/ reported that (-. e%isting
capacities accounted for 9"=7-*h in !""B or 9.;L of the global energy usage and was
estimated to reach 9!C7-*h in !"":) accounting for a total installed collector area in
use around the world of 9C; million square meters.
1.2 E.+-(/ i007+0 ,.> 61,44+.(+0
1.2.1 I894i6,:i2.0 0+: 5/ :1+ (425,4 +.+-(/ 06+.,-i2
*he global energy demand is increasing worldwide and will continue to rise as
developing nations reach developed status and developed nations maintain their
modernisation trends. 3igure 9,! from BP +!"":a/ reveals that the energy consumption
per capita is significantly higher in developed states than in less developed and
developing countries.
Fi(. 1)2* 2''% C2.0789:i2. 9+- 6,9i:, i. :2..+0 2i4 +=7iv,4+.:; BP (2''&)
$EA +!""B/ reports that energy demand is pro0ected to grow on average by 9.BL a year
thus an increase of 0ust over one,half between !""B and !";". @ver :"L of this increase
would come from developing countries such as $ndia and &hina which alone would
account for ;"L. *oday5s worldwide energy use is C"L fossil fuels which is to remain
the dominant source of energy until !";" in the ?eference (cenario reported by $EA
+!""B/.
:
" A 9.=
9.= A ;."
;." A >.=
>.= A =."
P".=
.owever1 the current global use of fossil and nuclear fuels has many consequences
including the depletion of natural resources1 threat to the world energy security and
global climate change caused by emissions of greenhouse gases from fossil fuel
combustion. *hese consequences have many environmental1 economical and political
impacts over the world.
1.2.1.1 N2.)-+.+A,54+ -+027-6+0 ,.> :1+i- 4i,5i4i:/
3ossil fuels stoc's are rapidly depleting around the globe. At current consumption rates1
BP +!"":/ and -E& +!"":/ state that proven world reserves should last between >" to
9=" years depending on the fuel as shown in 3igure 9,;.
147
85
63.3
40.5
0 20 40 60 80 100 120 140 160
Coal
Nuclear
Natural as
!il
R"# ratio
Fi(. 1)3* E.> 23 /+,- 2''% F200i4 37+4 -+0+-v+0 :2 9-2>76:i2. (R?P) -,:i20; >,:, 3-28 BP (2''&) @
WEC (2''&)
Pressure on e%isting reserves is increasing daily due to increased demand. onbiot
+!"">/ reported that the world currently consumes si% barrels of oil for every new barrel
discovered. Energy e%perts suggest that oil production will probably pea' sometime
between !""> and !"9" as stated by Asif et al. +!"":/1 followed by natural gas. *he
general decline of fossil fuel production will cause a global energy gap which could
result in serious international1 economic and political crises and conflicts detailed
further in section 9.!.9.;.
*he price of oil has been rising because demand is growing faster than a finite supply.
Based on data from Asif and uneer +!"":/ the cost of crude oil per barrel increased by
="L between !""> and !""=. $EA +!""B/ states that rising oil and gas demand could
C
accentuate the consuming countries vulnerability to severe supply disruption following
a price shoc'. Nuclear energy could be a possible route to reduce the energy gap
however environmental and political concerns associated with this technology do not
ma'e it a favourable approach especially with the current e%istence and growth of green
energies.
@ther concerns recently raised are the vulnerability of current energy infrastructures to
adverse weather conditions. Depletion of water could result in serious problems as vast
amounts of water are required for fuel processing and cooling in fossil fuel1 nuclear and
geothermal power plant as stated by 7leic' +9##>/. et @ffice +!"";/ and 4owit Q
EspinoDa +!""B/ outlined the !""; and !""B heat wave consequences on the operations
of several power plants. Plants were put at ris' and were shut down due to lac' of water
to cool the condensers. @ther components such as gas and oil pipelines or transmission
line could be affected by e%treme weather.
*his energy scene is facing another ma0or challenge and is raising serious
environmental concerns associated with fossil fuel consumption and production.
1.2.1.2 E.vi-2.8+.:,4 62.6+-.0
Energy production1 distribution and consumption raised serious environmental concerns
over the last century. &limate change is defined by the UN +9##!/ as Ma change in
climate that is attributed either directly or indirectly to human activity that alters the
composition of the global atmosphere and which is in addition to natural climate
variability observed over comparable time periods.N.
&limate change is caused by an increase of greenhouse gases by natural processes or by
human activities. 3or the last 9=" years1 atmospheric concentrations of greenhouse
gases have been steadily increasing partly due to the industrialisation revolution of
human activities. &arbon Dio%ide +&@
!
/ is one of the main greenhouse gases which is
primarily produced from burning fossil fuels and deforestation. @ther gases such as
methane and nitrous o%ide predominantly produced from agricultural activities and
changes in land use or halocarbons and sulphur he%afluoride released by industrial
processes contribute to climate change. aplecroft net 2td +!"":b/ reported that world
carbon dio%ide emissions in !"!= are pro0ected to reach ;C.C billion tons1 e%ceeding
9##" levels by C9L.
#
&limate change is already happening and its first signs can be witnessed around the
globe. $P&& +!"":/ reported measurements recording an increase of ".:> R ".9C& in
mean global surface temperature from 9#"B to !""= against the baseline of 9>& from
(ims +!"">/ and predicted a rise by 9.> to =.CS& during the !9st century. As a result1
mean sea level rose by 9" to !"cm with an average rate increase of about ;.9 R ".:mm
per year from 9##; to !"";. $P&& +!"":/ reported that since 9##; thermal e%pansion of
the oceans has contributed to =:L of the sea level rise while melting glaciers and ice
caps contributed for about !CL and polar ice sheets contributed for the remaining 9=L.
Asif et al. +!"":/ stated that sea level is pro0ected to rise by about ="cm during the !9st
century with a range of 9= to #=cm. *hese pro0ections have been bac'ed up by recent
satellite data from $P&& +!"":/ showing an annual Artic sea ice reduction of !.: R ".BL
per decade since 9#:C1 with larger decreases in summer of :.> R !.>L per decade.
ountain glaciers and snow cover on average have declined in both hemispheres with
snow cover decline of 9"L since the late 9#B"s in the mid and high latitudes of the
northern .emisphere. aplecroft net 2td +!"":a/ showed that there was a general
increase in precipitation in many regions of the world with a decline only observed over
Northern .emisphere sub,tropical regions while UNEP and UN3&&& +!""9/ described
scenarios of worsening intensity in parts of Africa and Asia. (ome e%treme weather
events have changed in frequency and intensity over the last =" years such as heat
waves1 heavy precipitation events1 intense tropical cyclone activities and incidence of
e%treme high sea,level ma'ing climate change one of the main challenges of today.
1.2.1.3 E62.28i6; 1+,4:1 ,.> 924i:i6,4 i89,6:0
*he impacts of depleting fossil fuels and climatic changes have been a ma0or source of
dispute and are all inter,lin'ed. &limate change impacts will lead to economic losses for
various sectors around the world. According to the (tern ?eview +!""B/ if action is not
ta'en to curb carbon emissions1 climate change could cost between = and !" percent of
the annual global gross domestic product. *hese increases in costs arise from e%treme
weather1 including floods1 droughts and storms.
.ealth impacts as a direct result of climate change are also immense. *he -orld .ealth
@rganisation +!""C/ stated that climate change led to people dying every year from its
side,effects through deaths in heat waves1 and in natural disasters such as floods1 as well
as influencing patterns of life,threatening vector,borne diseases such as malaria and an
increase in food insecurity1 water shortage and other multiple stresses. According to the
9"
$P&& +!"":/ people living in poverty would be worst affected by the effects of climate
change.
@ther stresses such as global economic and political instability are sub0ect to attention.
&onflicts and possible shortages of energy supply have already started to have important
implications. @E&D +!""C/ showed potential water related tensions emerging between
nations that share common freshwater reserves. Access to water1 its allocation and use
are becoming increasingly critical concerns that may have profound consequences for
political and social stability. -ith worldwide demand for energy increasing every day1
dependency on oil imports for most developed countries and therefore energy insecurity
is also becoming an increasing problem. *his results in profound economic and political
implications such as the 9#=; coup in $ran organised by the U( and Britain1 followed by
the (ueD crisis to the 7ulf -ar in 9##9 and the $raq war of !"";. any wars have been
conducted and are still fought all over the world to ensure corporate control over oil.
*ensions are increasing on new discovery of oil. Disputes between countries over oil
reserves represent another potential concern. *he recent e%ample of the ?ussian
e%pedition aiming at strengthening ?ussia5s claim of the oil and gas wealth beneath the
Arctic @cean by planting their country5s flag on the seabed shows the potential e%tent of
the conflict. &ountries bordering the Artic including ?ussia1 the U(1 &anada and
Denmar'1 have launched competing claims to the region.
-ith worldwide demand for energy increasing every day1 the development of new1
clean1 renewable energy sources is critical to EarthHs environment. -orldwide and in the
United <ingdom +U</1 wor' is under way on a variety of potential answers to the
global energy challenge. Energy policy and the potential of renewable energy for the
environment are discussed.
1.2.2 E.+-(/ 924i6/ ,.> 9-209+6:0 23 -+.+A,54+ +.+-(/
1.2.2.1 K/2:2 P-2:2624 ,.> :1+ +.+-(/ 924i6/ 23 UK
Political considerations over the security of supplies1 environmental concerns related to
global warming and sustainability are ma0or political issues and the sub0ect of
international debate and regulation.
99
As a result international treaties such as the United Nations 3ramewor' &onvention on
&limate &hange +UN3&&/ +!""C/ were formed at the United Nations &onference on
Environment and Development. *he treaty aimed to achieve the stabilisation of
greenhouse gas concentrations in the atmosphere at a level that would prevent
dangerous interference with the climate system. Parties have been meeting annually in
&onferences of the Parties +&@P/ to assess progress in dealing with climate change and
to establish legally binding obligations for developed countries to reduce their
greenhouse gas emissions. As a result1 the <yoto Protocol was adopted by &@P,; on
December 991 9##: and finally implemented on 3ebruary 9B1 !""= during the &@P,99.
*his international agreement legally binds most industrialised countries to reduce
emissions of gases contributing to climate change by an average of =.!L below 9##"
levels between the years !""C,!"9!1 defined as the first emissions budget period. *he
<yoto Protocol was the first step towards a truly global emission reduction regime.
.owever1 the commitment period of the <yoto Protocol ends in !"9! and a new
international framewor' needs to be negotiated and ratified to deliver the strict emission
reductions. *he recent &@P,9; held in Bali achieved agreement on a timeline
negotiation on the post,!"9! framewor' as a successor to the <yoto Protocol.
Negotiations will be held in Poland during December !""C. *o date1 9:C nations have
ratified the treaty1 only the U(A and <aDa'hstan have signed but not ratified the <yoto
Protocol. Under the international <yoto Protocol and European Union agreements the
U< must reduce its baseline greenhouse gases emissions by 9!.=L by !"9!.
3urther steps in this direction were added by the U< 7overnment which has set its own
international and domestic energy strategy to reduce carbon emissions in the Energy
-hite paper +!"":/1 the &limate &hange Bill +!"":/ and more recently in the &limate
&hange Act !""C. *he Energy -hite PaperF Meeting the Energy &hallengeN addresses
a commitment to reduce carbon dio%ide emissions1 with consequent constraints to its
energy policy. A mandatory B"L reduction in &@
!
emissions below the 9##" baseline
by !"="1 with an intermediate target of between !BL and ;!L by !"!"1 was published
within the &limate &hange Bill by DE3?A +!"":/. A review of the &limate &hange Bill
published by DE3?A +!""Cb/ resulted in the recently transition to the &limate &hange
Act !""C committing the U< to cut emissions by C"L by !"=". Along with this new
reduction1 the Energy -hite Paper outlines the importance to ensure secure1 clean and
affordable energy. *he (cottish E%ecutive +!"";/ agreed a more stringent overall target
compared to U< with its own energy policy aspiring to meet >"L of its energy through
9!
renewables by !"!". 3urther steps in support of this aspiration were proposed by the
(cottish 7overnment +!"": Q !""C/ such as the introduction of the ?enewables
@bligation (cotland +?@(/ providing an incentive to supply higher levels of renewable
energy and encourage developers to bring forward new renewable energy schemes and
the &limate &hange +(cotland/ Bill setting a C"L beneath 9##" baseline by !"=" target1
with an intermediate target of ="L by !";".
1.2.2.2 R24+ ,.> 9-209+6:0 23 -+.+A,54+ +.+-(/
Ensuring a secure1 clean and affordable supply of energy is essential towards the
sustainable development of a modern society. ?enewable energies are potential
solutions to provide energy with Dero or almost Dero emissions therefore reducing local
and global atmospheric emissions. (olar energy is the most abundant energy resource
on earth and is available for use. According to the -orld Energy &ouncil +!"":/ the
annual solar energy intercepted by the surface of the Earth is appro%imately 9C"P-1
therefore1 more than :=1""" times the world5s total annual primary energy consumption
of 9>*-h. @ther ma0or renewables were estimated at B9*- such as wind1 biomass1
hydro and geothermal. *here are therefore plentiful sources of energy that can meet the
present world energy demand with appropriate technology. An increasing share of
renewable energy in the energy portfolio would offer a certain degree of
manoeuvrability against sudden fossil fuel energy price increases and therefore create a
secure long,term sustainable energy supply.
,'enewable so$rces of ener!y ha-e a considerable potential for increasin! sec$rity of
s$pply in .$rope/ De-elopin! their $se0 howe-er0 will depend on extremely s$bstantial
political and economic efforts/ 123 &n the medi$m term0 renewables are the only so$rce
of ener!y in which the .$ropean #nion has a certain amo$nt of room for manoe$-re
aimed at increasin! s$pply in the c$rrent circ$mstances/ We cannot afford to ne!lect
this form of ener!y/4 7reen paper on the security of energy supply +!""9/.
-ith the recent energy policies and protocols described in section 9.!.!.9 and based on
energy policy goals of the Energy -hite Paper1 the role and prospects for renewables
appear to be good. ?enewables can enhance diversity in energy supply mar'ets and
secure reliable and sustainable energy supplies. *hey can therefore reduce occurrence of
energy poverty increasing with energy pricing and reduce atmospheric emissions.
9;
?enewables can also promote competitive new mar'ets in the U< and beyond to meet
specific needs for energy services.
According to aplecroft 2td +!""B/1 while emissions of developed countries are li'ely
to stabilise1 emissions from developing countries continue to rise steadily and are
e%pected to equal those of developed countries. *hese developing countries often
possess large amounts of renewable energy potential especially for solar applications.
*he 'nowledge transfer of such technologies to help developing countries reduce
emissions as their economies grow is vitally important to their economic development.
?enewables such as solar1 wind1 hydropower and biofuels are potential candidates to
meet global energy requirements in a sustainable way. 3igure 9,> shows a comparison
of world wide and EU installed capacities of three intermittent renewables) namely
photovoltaic +PO/1 solar thermal and wind power. *hese data include &hina which
dominates the solar thermal world mar'et with over B"L of the global installed
capacity.
93700
128000
56300
16176
7800
4500
0 20000 40000 60000 80000 100000 120000 140000
$in% #ower
&olar '(er)al
#(oto*oltaic
+$
!"
#orld
Fi(. 1)#* +,- 2''& W2-4> ,.> EU i.0:,44+> 6,9,6i:i+0 23 024,- PV; 024,- :1+-8,4 ,.> Ai.> 92A+- B
D,:, 3-28 :1+ REN21 (2''&) ,.> EC (2''&)
*he published comparison of PO1 solar thermal and wind capacities loo's different if
European data are considered. -ind power is the leader of the three technologies and is
already a mar'et reality. (olar thermal is a strong second place with many applications
for the direct use of solar thermal energy in space heating1 and cooling1 water heating1
crop drying and solar coo'ing. $t is most used for water heating hot water being a
9>
necessity in today5s lifestyle. $n some countries1 residential and commercial water
heaters consume a substantial portion of the average utility bill and are dependent
particularly upon the combustion of fossil fuels.
1.3 P-209+6: 23 SWH i. UK ,.> S62:4,.>
1.3.1 E.+-(/ ,.> E.vi-2.8+.:,4 S6+.+ i. S62:4,.>
An understanding of current energy supply and demand of the U< and (cotland is vital
to develop an appreciation of the opportunities of (-.. (cotland consumes 9:=*-h of
delivered energy consuming #.9L of U< energy with only C.=L of the U< population
showing greater energy consumption per capita than the average for the U< in total.
*he main end,use sectors can be divided in fourF domestic1 transport1 industry and
services. *he domestic sector is the largest consumer with an energy split of ;>L. A
more detailed picture of the energy used in the domestic sector show that 9".9L of the
total consumption in the U< is used by (cotland1 despite the fact that it represents C.=L
of the U< population.
*he colder climate in (cotland is a ma0or factor e%plaining this figure. (cotland
typically has 9"L more degree,days than the U< average consequently affecting the
requirements for space heating. *he main fuels used in the domestic space and water
heating account for over C"L of a householdHs energy consumption according to the
(cottish E%ecutive +!""B/ are gas1 along with coal and oil.
*he domestic sector consumes =B*-h in total. $f the C"L accounting for space and
water heating were to be replaced by renewable energy1 then >>.C*-h of energy would
be carbon neutral which would result in a decrease in &@
!
emission of 9!.9>7t&@
!
Jyear
based on the (cottish E%ecutive +!""B/ data.
*here is vast scope in addressing the energy requirement1 particularly for hot water from
renewable sources.
1.3.2 S24,- 8,-C+: ,.> 9-209+6:0
&ertain parameters play a 'ey role in the behaviour of (-. performance1 such as
geographic location1 solar potential1 types of (-. and others which could e%plain why
9=
they are more developed in some areas. .owever1 their potential are nowadays valuable
in any location.
*here was a solid growth of >>.;L in the European solar thermal mar'et in !""B
corresponding to more than ; million m
!
of newly installed surface of solar thermal
collectors operating in Europe. *he greatest growth occurred in the 3rench mar'et with
a growth rate of over C;.9L1 while the mar'et is still dominated by 7ermany with a
growth of !:L despite the consecutive decrease of investment subventions granted in
the framewor' of the 7erman mar'et stimulation program as stated by Eur@bserv5E?
+!"":/. *he European mar'et is dominated by 7ermany1 7reece1 Austria and 3rance
with over ::L of the mar'et represented by these countries. *he EU has reached an
average capacity in operation of ;".C '-thJ91""" inhabitants. *his indicator ma'es it
possible to directly compare different countries involvement in developing solar thermal
applications as shown in 3igure 9,=. National values range from =99 in &yprus and !>"
in Austria to less than 9 '-thJ91""" inhabitants in some Baltic countries. *he U<
position in the mar'et is towards the lower end with only !=!19B" m
!
of solar thermal
collectors in operation in !""B which accounts for only 9.!L of the mar'et. *his
indicator shows that the margin of progression is still very large in 7ermany and that
this potential has barely been tapped in 3rance1 (pain and the U<. 3igure 9,= also
shows that despite the higher available solar radiation of (pain1 3rance and the U< there
has been lower involvements in developing solar thermal compared with other
European countries with lower solar radiation availability.
9B
0
100
$00
300
%00
&00
'00
(00
800
0 1 $ 3 % &
,*ailabilit- of solar ener.- on an (ori/ontal 0lane1 2$(")
2
"%a-
&
$
3

i
n
s
t
a
l
l
e
%

c
a
0
a
c
i
t
i
e
s
1

)
2
"
1
0
0
0

c
a
0
i
t
a
s
"nited )ingdo
*erany
*ree+e
,ran+e
-pain
.ustria
/enar0
-weden
1alta
2yprus
Fi(. 1)"* E7-29+,. 024,- 8,-C+:; D,:, 3-28 E7-O50+-vDER (2''&) ,.> E7-29+,. C288i00i2. (EC)
(2''&).
&yprus is the European leader when it comes to ability to capture solar energy with over
#"L of the buildings equipped with solar collectors1 which is twice as much as Austria
and 7reece. .owever if the whole EU was at the same level per capita as Austria today1
the annual EU mar'et would be over 9:" millions m
!
1 with a capacity in operation of
about 99# 7-th. *his capacity would effectively substitute large amounts of
conventional fuels such as oil and gas or electricity and put Europe ahead of &hina in
terms of the global mar'et. $n their report ?EN!9 +!"":/ declared that the global mar'et
is dominated by &hina with annual domestic sales of well over 9" millions m
!
with the
EU holding only a small fraction of the world mar'et as second.
*his section shows that the U< position in the mar'et is towards the lower end and that
the U< potential for developing solar thermal applications in U< has barely been
tapped.
1.3.3 N++> 32- ,332->,54+ SWH ,.> >,:, 82>+44i.( (9-209+6: 23 024,- A,:+-
1+,:i.( 32- S62:4,.>)
$n a period of rapidly growing deployment of sustainable energy sources the
e%ploitation of solar energy systems is imperative. (olar water heating systems have
been well developed in (cotland with a large choice of systems available to suit many
applications. *he higher annual heating loads faced by (cotland allows longer and
9:
&#
+,
R
C4
,5
&$
FR
6N
57
8R
greater savings using solar thermal power ma'ing this type of technology more effective
in northern countries. *he viability of such systems has been proven as the number of
commercial manufacturers increased and the number of installers accredited by the
(cottish &ommunity and .ouseholder ?enewables $nitiative , (&.?$ +!""C/
registration lists increased from ;; in arch !"": to B: in 4uly !""C or =9L more
installers of (-..
*he Energy (aving *rust +!""=/ estimated commercial prices for an installed (-. in
the range of T!""" to T;""" depending on the type of collector and the capacity of the
system required1 with an average lifespan of !" years. According to the (cottish (olar
Energy 7roup , ((E7 +!""C/ about half of the hot water energy demand over the year
could be provided with an average collection efficiency of about >"L for a !,=m
!

collector based on the fairly cloudy (cottish climate getting 9"""'-hJm
!
. *here is then
a niche mar'et for lower cost (-. with a lifespan and annual thermal efficiency equal
or superior to current designs on the mar'et.
(olar water heaters also have the benefit of reducing impacts on the environment by
reducing carbon dio%ide emissions by ".>,".:= tonnes per year depending on the fuel
replaced according to the Energy (aving *rust +!""=/. .owever1 as energy efficiency
and better design of collectors can reduce energy use) the energy e%pended during the
manufacturing process1 also called Membodied energyN1 ta'es a more significant part of
the real energy savings and paybac' time. $t is therefore very important to consider the
whole energy life,cycle of the (-. as it allows for a better assessment of its
environmental impact.
&ertain parameters play a 'ey role in the success of (-. technologies such as their
monetary and energy pay bac' period. .owever1 their potential use or application are
valuable at any location and in any application as buying coal1 oil or gas fuelled
substitutes do not offer the customer any financial reward. $n order to encourage the
installation of renewable devices1 a ;"L grant is available for solar thermal
technologies by the (&.?$ +!""C/ thereby increasing the (-. return, on,investment.
*he result of many years of global research on solar water heating systems has outlined
the promising approach of integrated collector storage solar water heaters +$&(,(-./
in cold climates. $&(,(-. are the simplest type of (-. on the mar'et and a popular
9C
choice when choosing a design as they possess several advantages. *hese systems
incorporate the solar collector and thermal storage tan' in a single unit. *heir simplicity
of construction and installation ma'e them affordable with low costs. @ther advantages
such as their aesthetic value1 low maintenance1 ease of integration in roof structures and
good overall performance in overcast s'y conditions ma'e them suitable for northern
hemisphere applications where this research is ta'en. *his calls for a need to estimate
the potential of $&(,(-. for the (cottish climate.
1.# T1+ P-+0+.: R+0+,-61 P-2M+6:
1.#.1 P-2M+6: ,i8
&nte!rated collector stora!e solar water heaters5 *o study and analyse the performance
of a newly developed $&(,(-. for (cottish weather conditions1 optimise its
performance1 model its laboratory and field performance with its environmental impacts
and analyse its integration into buildings and benefits of such a heating system1 for the
primary purpose of proposing a feasible $&(,(-. prototype.
1.#.2 O5M+6:iv+0
*he ob0ectives of the pro0ect can thus be summarised as followsF
9. *o study the possible improvement of $&(,(-. under (cottish climatic
conditions through the literature review) considering the large potential mar'et
for solar thermal systems.
!. 3ollowing the analysis underta'en in +9/1 to test in laboratory conditions the
newly developed $&(,(-. in order to accurately compare its performance with
a previous study carried out for $&(,(-..
;. *o carry field tests of the same $&(,(-. to evaluate its performance under real
(cottish weather conditions.
>. *o develop a thermal macro,model able to compare the temperature variation in
different $&(,(-. materials) including internal temperature and e%ternal
weather conditions for a given aspect ratio and to evaluate the performance of
this $&(,(-. for laboratory and field conditions.
=. *o model the $&(,(-. using 3luent B.! &3D +&omputational 3luid Dynamics/
software to estimate an optimum fin spacing to ma'e further improvements on
the current $&(,(-..
9#
B. *o determine the monetary and energy paybac' considering the whole life
energy of the $&(,(-. using the thermal model previously developed to
determine its feasibility.
:. *o show the integration of $&(,(-. into buildings and its potential benefits for
builders and households.
1.#.3 O7:4i.+ 23 :1+ :1+0i0
&hapter 9 gives the introduction to the thesis1 starting from a very broad sub0ect of
energy it narrows down to renewable energy then to solar energy and then further to
solar water heater and more particularly to $&(,(-.. $t also provides the bac'ground1
statement and ob0ectives of the present research problem.
&hapter ! deals with the literature review covering the following topicsF +a/ solar hot
water systems1 thermal collectors and solar water heater parameters1 +b/ modelling the
solar water heater through a macro,model +thermal model/ and micro,model +&3D/1 +c/
life cycle assessment +2&A/1 weather conditions in (cotland and building integrated
(-..
&hapter ; deals with laboratory and field e%periments. *he methodology used is
introduced followed by the test regime1 measurement and data collection. $t also gives
laboratory e%perimental set,up and results with a comparison with previous wor'.
E%perimental set,up for the field tests and results of e%periments are also provided.
3inally an account of uncertainties and propagation of errors through both e%perimental
set,ups is given.
&hapter > presents the modelling procedure of the $&(,(-.. *he thermal model results
and e%perimental data are inter,compared for laboratory and field tests by means of
various statistical indicators while the micro model is used to determine optimum fin
spacing.
&hapter = performs a 2&A of different types of $&(,(-. for use in (cottish conditions
including the e%perimentally tested system using the previously developed macro
model. Using modelled bul' water temperature in the collector1 performance1 monetary1
energy and carbon savings comparison of three different $&(,(-.s was provided by
changing their material and thic'ness.
!"
&hapter B demonstrates the potential for $&(,(-. to be integrated into buildings and
the different considerations to be ta'en into account when choosing such (-.. Benefits
in installing such heaters and possible installations scheme into buildings are also
provided.
&hapter : draws up important conclusions from each aspect of the presented wor' and
also discusses the potential for future wor'.
1." C2.647>i.( -+8,-C0
*his chapter provides a detailed introduction to energy1 its various resources1 associated
issues and challenges and its current and future status. (ignificance of renewable energy
resources was established and various such forms discussed. (olar energy was
addressed in detail as this forms the basis of the present pro0ect. (-. and underlying
mar'et prospects for U< and (cotland were briefly discussed. *he need for new (-.
technologies was also emphasised. *his was followed by stating the research problem1
its aims and ob0ectives and finally the structure of the thesis.
CHAPTER 2 REVIEW OF THE RELEVANT LITERATURE
$n the conte%t of the present research pro0ect1 investigating solar water heaters has three
core aspects. *he first is studying the prospects and diversity of solar water heaters. $t
also includes a guidance to improve the configuration of solar water heaters and
e%perimental procedures. *he second aspect deals with the modelling methodology
!9
itself. *he third involves solar water heating in (cotland1 commercial designs and
weather considerations.
*his chapter is accordingly sub,divided into three sections. (ection !.9 reviews the
different type of solar water heaters1 their construction and the different influencing
characteristics. (ection !.! provides a comprehensive yet summarised coverage of
different methods of modelling solar water heaters. (ection !.; reviews the integration
of solar water heaters into buildings under (cottish weather considerations.
2.1 S24,- A,:+- 1+,:+-0
2.1.1 S24,- 12: A,:+- 0/0:+80
(-. systems are generally very simple using only sunlight to heat water. A wor'ing
fluid is brought into contact with a dar' surface e%posed to sunlight which causes the
temperature of the fluid to rise. *his fluid may be the water being heated directly1 also
called a direct system1 or it may be a heat transfer fluid such as a glycolJwater mi%ture
that is passed through some form of heat e%changer called an indirect system. *hese
systems are used for domestic hot water or space heating. *hey can be classified into
three main categoriesF +a/ Active systems which use pumps to circulate water or a heat
transfer fluid1 +b/ Passive systems +or *hermosyphon systems/ which circulate water or
a heat transfer fluid by natural convection and +c/ batch systems using a tan' directly
heated by sunlight.
Active systems are also called forced circulation systems and can be direct or indirect.
A system is called direct when the pump is used to circulate drin'ing water between the
hot water storage tan' and the collector. $n an indirect system the pump is used to
circulate a heat transfer fluid1 either water or an antifreeDe mi%ture1 between the
collector and a heat e%changer. .eat is transferred by the heat transfer fluid from the
collector to the drin'ing water through the heat e%changer in the storage tan'. *he use
of a heat e%changer reduces significantly the system efficiencies.
Passive systems simply circulate water or a heat transfer fluid by natural convection
between a collector and an elevated storage tan' +above the collector/. *he principle is
simple1 as the fluid heats up its density decreases. *he fluid becomes lighter and rises to
the top of the collector where it is drawn to the storage tan'. *he fluid which has cooled
down at the foot of the storage tan' then flows bac' to the collector. *his system can be
!!
operated as a direct or indirect system1 whereby a heat e%changer can allow the use of
antifreeDe to prevent frost damage during colder winter months. (ince no pumps or
controls are necessary1 system installation costs are reduced.
3inally1 batch systems have combined collection and storage functions. *hey are the
simplest type of (-. with flow of water throughout the system being set off whenever
there is water demand. .ot water is drawn to the top of the collector and replaced by
cold water at the bottom. $&( systems can supply preheated water directly to the end use
appliance or more commonly to a conventional water heater tan'. Depending on the
system1 there is no requirement for pumps or moving parts. .owever1 the system is
prone to freeDing1 therefore draining water in the collector and pipe is advised during
cold periods. *hese systems could also be integrated in to active or passive systems
depending on the design of the house.
Active and passive (-. systems remain more common than batch (-. systems. *o
date1 a small amount of wor' has given adequate consideration to the integration of
batch systems1 especially within domestic buildings. Batch (-. are normally used for
low temperature applications such as domestic or light industry due to the low capital
outlay. *his ma'es it possible to achieve realistic capital cost paybac's during the
lifetime of the systems. *he three common types of built,in,storage solar collectors are
described in section !.9.;.9 below.
2.1.2 S24,- :1+-8,4 6244+6:2-0
*here are three common designs of solar water heaters. $n order of decreasing cost they
areF concentrating collectors1 evacuated tube collector and flat,plate collectors. *he
choice of collector is determined by the heating requirements and the environmental
conditions in which it is employed.
&oncentrating collectors are mostly used in high temperature solar applications such as
steam production for the generation of electricity. *hey trac' the sun during the daytime
and use reflectors to concentrate the sun5s energy onto the absorber. *hese systems
however are most appropriate for applications where direct solar radiation is high. *hey
are best suited to climates that have a high percentage of clear s'y days and are
therefore not so common in (cotland
!;
Evacuated tube collector systems are principally used for domestic water heating
purposes as suggested by orrison et al. +!"">/. *he solar absorbing surface is placed
in the centre of a vacuum tube which minimises convection losses thereby increasing
efficiency. .owever1 improved efficiency is reflected in their price. *hey are most
suited to e%tremely cold ambient temperatures or in situations of consistently low,light.
*hey are also used in industrial applications1 where high water temperatures or steam
need to be generated where they become more cost effective.
3inally1 flat,plate collectors are used e%tensively for domestic hot water heating
applications due to performance advantages in relation to cost. *hey collect both direct
and diffuse radiation. *heir simplicity in construction reduces initial cost and
maintenance of the system. Although flat plate collectors are outperformed by
evacuated tubes collectors they are generally a more cost effective solution in most
climates and for the ma0ority of domestic hot water services. *here are variants of flat
plate collectors such as serpentine tubes1 parallel tubes or integrated collector storage
solar water heaters +$&(,(-./. *he simplicity in construction and low cost of $&(,
(-. ma'e them an attractive and popular choice when choosing a (-.. A more
detailed picture of these systems is of interest and is presented in the following section.
1.1.1.1 B-i+3 1i0:2-/ 23 ICS)SWH
$&( are one of the oldest solar water heater designs still in use. *he first record of $&(,
(-. systems was in the United (tates of America +U(A/ in the late 9C""s where
sufficient hot water for showering by the late afternoon on clear days was achieved.
$&(,(-. was the first patented and commercially manufactured (-. in 9C#9 under
the trade name U&lima% (olar -ater .eater5 designed by <emp +9C#9/. *he first
detailed study on $&( tan' systems was carried out in 9#;B by Broo's +9#;B/ at the
University of &alifornia Agricultural E%perimental (tation1 U(A. Although $&(,(-.
were a growing mar'et1 the discovery of natural gas and oil postponed investigations of
$&(,(-. until 9#:". After the oil crisis of @ctober 9#:;1 a renewed worldwide interest
in solar energy was revitalised including a renewed interest in $&(,(-.. *he first
developments in solar energy were underta'en by passive solar enthusiastic individuals1
0oined later by universities and research institutions. *oday solar heating and $&(,(-.
is a very popular area of research worldwide encouraged by global initiatives to
promote renewable energy technologies to limit greenhouse gas emissions.
!>
A long review of technical developments of $&(,(-. to benchmar' the performance
of all heaters using common performance indices was carried out by (myth et al.
+!""B/. -orld,wide popular $&(,(-. were reviewed as well as the regional1
environmental and economical factors that have influenced solar heaters designs. Based
on this study1 a review of the 'ey components of $&(,(-. are presented in the
following section
1.1.1.2 K+/ 62892.+.: 23 ICS)SWH
*here are many $&(,(-. designs but generally all consist ofF +a/ a storage tan'1 +b/ a
heat transport fluid flowing through the collector to remove heat from the absorber1 +c/ a
flat plate absorber which intercepts and absorbs the solar energy1 +d/ a heat insulating
bac'ing with baffle and insulation1 +e/ a transparent cover+s/ that allows solar energy to
pass through but reduces heat loss from the absorber.
S:2-,(+ :,.C
(iDe and shape of the storage tan' have a significant effect on the performance of the
(-.. (everal designs of $&(,(-. have been studied over the years. *he first
manufactured commercially U&lima% (olar -ater .eater5 designed by <emp +9C#9/
was composed of four small blac' painted cylindrical storage tan's interconnected and
mounted in a glaDed wooden bo% insulated with felt paper. 3igure !,9 shows the
advertisement of this (-. in 9C#! with a purchase price of V!=.
!=
Fi(. 2)%* A>v+-:i0+8+.: 32- :1+ 64i8,x 024,- A,:+- 1+,:+- i. 1<$2; B7::i ,.> P+-4i. (1$<')
$n 9#":1 .as'ell +9#":/ patented a shallow rectangular tan' design increasing the
surface area allowing a faster warm,up in the morning and hotter water on cloudy days.
*anishita +9#:"/ indicates that in 9#>: Wamamoto designed the first 4apanese
commercial (-. made of a stainless,steel piping. *he same authors states that later in
9#="s the closed membrane1 'nown as plastic bag $&(,(-.1 was successfully
developed and commercialised in 4apan. $t consisted of a large rectangular
polyvinylchloride bag1 with a blac' membrane as a bottom surface and a transparent
membrane on top. Another type of storage tan' was tested by &hinnappa and
7nanalingam +9#:;/ using square coil pipes.
7arg +9#:=/ conducted the first year long performance study of a rectangular design
$&(,(-. followed by uneer and .awas +9#C>/ and uneer +9#C=/ who proposed a
simple rectangular design $&(,(-.. uneer and .awas +9#C>/ compared $&(,(-.
and thermosyphon collector showing that the simple and low cost $&(,(-. narrowly
outperformed the thermosyphon collector. $n his later study1 uneer +9#C=/ studied the
effect of the collector depth on performances. ?esults showed that an increase of !cm in
depth enhanced the storage volume which made the system operate at lower
temperatures and so the losses are limited resulting in an increase of efficiency of CL.
(everal studies carried by Ecevit et al. +9#C#/1 Pra'ash et al. +9##!/ and <aushi' et al.
+9##>/ suggested that a triangular storage tan' can perform better due to increased
!B
natural convection. (o'olov and Oa%man +9#C;/ showed an efficiency of =;L for their
triangular $&(,(-. using a baffle plate to separate the absorbing plate from the storage
volume. Ecevit et al. +9#C#/ studied the effect of different volumes in triangular
collectors showing good overall performance in all designs. A side by side comparative
e%perimental study of a rectangular and triangular shaped $&(,(-. was carried out in
9##> by (oponronnarit et al. +9##>/ under identical conditions. *he thermal efficiencies
of the triangular and rectangular were found to be B;L and =#L respectively. ohamad
+9##:/ introduced a simple thermal diode to prevent reverse circulation at night time in
his triangular $&(,(-.. *he storage tan' efficiencies improved by 9=.;L using the
diode. $n the same year1 an innovative pyramid shape was developed for Egyptian
weather conditions by Abdel,?ehim +9##:/ providing temperature range of >",B" &.
$n his subsequent studies uneer et al. +!""Ba/ launched a new (-. design with the
introduction of fins. *he report details side,by,side testing of a system using a novel
rectangular finned $&( against a plain rectangular non,finned $&( system installed at the
Din *e%tile ills1 Bhai Pheru1 2ahore1 Pa'istan. .ot water demand for this study was
C"S& for the dyeing process. *heir research concluded that on clear days the non,finned
design achieved ma%imum temperatures in the region of :"S& at best1 whilst the finned
heater regularly achieved water temperatures higher than :"S&1 sometimes reaching
temperature in the region of C"S&. *he solar fraction of the finned heater was found to
be :;L compared to B;L for the non,finned system. Using the same design1 7rassie et
al. +!""B/ performed laboratory e%periments in order to optimise the collector design
under (cottish weather conditions.
A recent study developed by &ruD et al. +!""!/ introduced a trapeDoidal storage tan'
design for editerranean climates inducing a thermal stratification in the water store.
*he research indicated that ;"L to :"L of the daily load could be covered by the $&(,
(-..
H+,: :-,.092-: 347i>
*he collector fluid in an $&( system is potable water. -ater is the cheapest1 most
readily available and thermally efficient fluid but does freeDe and can cause corrosion.
A drain,bac' system can empty the collector to a bac',up hot water cylinder at times of
low temperatures and negative heat gains. Using softened water will help e%tend
operational lifespan.
!:
A502-5+- 94,:+
*he absorber plate material is a main parameter in an $&(,(-. as it has the principal
function to absorb solar radiation and transfer heat to the storage fluid. As mentioned by
(myth et al. +!""B/1 wood was one of the first materials used1 later systems using plastic
were developed and then replaced by new resistant material such as copper1 aluminium1
stainless steel and polymers with painted absorber surfaces. *he use of polymer
absorber has been studies by Oan Nie'er' et al. +9##B/ for parallel tubes collectors in
(outh Africa. Polymers1 compared to metals have a lower thermal conductivity and may
have high temperature limitations1 but they are relatively ine%pensive and can offer
good freeDe protection. *ransfer of absorbed solar energy to the storage fluid depends
on the collector materials thermal conductivity. aterials such as steel1 aluminium or
copper are more frequently used with copper and aluminium as the most effective.
Another parameter has to be loo'ed at when choosing the collector material. $t must be
of satisfactory strength and durability1 lend itself to manufacturing constraints and be of
an appropriate cost. &apral,Aluminium,2td +!""B/ stated that in most surroundings1
aluminium has a very high resistance to corrosion. Aluminium is covered by a thin
effective coating of impermeable o%ide which protects from further o%idation. .owever1
this material should be used by itself as in most combinations with other metals
galvanic corrosion occurs. (everal studies carried out by (tic'ney and Nagy +9#C"/1
Burton and Eweig +9#C9/1 Bainbridge +9#C9/ and *iller and -ochatD +9#C!/ and
recently *ripanagnostopoulos and Wianoulis +9##!/ suggested the use of an optional
selective coating for absorber plate. &oatings can be applied such as electroplated blac'
nic'el1 blac' chrome1 and copper o%ide or anodiDed aluminium. All studies showed that
coatings enhanced the heat transfer and protected the absorber plate. *o ma%imiDe
efficiency1 their absorptivity should be at least ".# and emissivity less than ".9.
I.074,:i2.
Back heat losses
Early systems were simply e%posed tan's of water left out to warm in the sun. *hese
systems suffered substantially from ambient heat losses1 particularly at night,time and
during non,collection periods. -ide theoretical and e%perimental studies on an $&(
(-. undertoo' by 7arg and ?ani +9#C!/ in $ndia proposed a baffle plate to reduce
night time losses. &ollector performance improvement was found during night time in
addition to during the day. *hey also observed that by using an insulation cover during
cooling hours1 the collector performance could be improved by :"L. <aushi' et al.
!C
+9##=/ carried out a comprehensive evaluation of a triangular $&( system with a baffle.
*hey concluded that the thic'ness and the material of the baffle had little impact on the
system performance and that the inclusion of the baffle had a significant benefit on the
system performance during non collection periods. Smyth et al. (2003) and
*ripanagnostopoulos and (ouliotis +!"">/ proved that collector performance
improvement is closely lin'ed with collector bo% efficacy. *he baffle should be
weather,proof1 guard the absorber1 insulate from the environment1 reduce convection
and conduction losses and provide a sturdy housing to hold the components. *he
collector bo% should also be designed to allow for roof integration. 7as'ets and seals
are further important parameters in maintaining a weather,tight seal between
components whilst allowing thermal e%pansion of the components. -ithout them1
moisture would fog the glaDing thereby reduce transmittance and possibly damage the
insulation and the absorber coating.
uneer et al. +!""Ba/ and 7rassie et al. +!""B/ e%plained that the insulation of the
storage tan' was essential to protect from conduction and convection heat losses. Both
authors proposed an easy solution using "."=m of glass wool insulation on all sides of
the collector housed in a wooden bo%. $nsulation is usually between !=mm and 9=" mm
and ideally would have reflective foil to reflect thermal radiation bac' to the absorber.
&ommon insulation materials are high,temperature resistant fibreglass batting1 semi,
rigid board or mineral wool. $nsulation should have a low thermal e%pansion coefficient
and should not melt or outgas at collector stagnation temperatures.
Top heat losses
Another way to decrease heat losses is the addition of a lid on top of the (-..
oveable insulated lid were introduced by Baer +9#:=/ resulting in a significant
improvement in the heat retaining properties of the $&(,(-.. Early attempts by
c&rac'en +9#:C/ using materials such as *ransparent $nsulating aterial +*$/ in
order to minimise heat losses caused decreases in the system efficiency. .owever1
recent studies of ?eddy and <aushi'a +9###/ and .enderson et al. +!""B/ showed that
the use of new *$ has led to greater collection performances.
T-,.09,-+.: 62v+-
!#
Bishop +9#C;/1 Duffie and Bec'man +9##9/ and (myth et al. +!""B/ established that the
use of one or more transparent covers reduced thermal losses by radiation and
convection by restricting air movement. 3urthermore1 covers protect the absorber and
selective coating +if applicable/ from harm1 dirt1 rain and other environmental factors.
7lass or plastic sheets are the most common materials used for covers. *he same
authors indicated that the most important property required of the transparent cover
material is high transmittance of solar radiation1 as any loss in transmittance will lead to
a direct reduction in collection efficiency. 7lass is considered more appropriate than
plastic as it is more durable and does not degrade in sunlight or at high collector
temperatures. *he only disadvantages of glass material are the more e%pensive costs and
the possibility of impact damage. Duffie and Bec'man +9##9/ suggested the use of
additional covers to reduce convective losses. .owever1 increase in glaDing would result
in higher transmission losses. Bainbridge +9#C9/ showed that using selective coating
glaDing could result in enhancing greenhouse effect and maintain high transmission.
7laDing with very high solar transmittance1 such as A? coatings used in PO application
by 7ombert et al. +9##C/1 could result in cost effective alternative for solar thermal
application.
(tudies conducted by uneer et al. +!""Ba/1 .enderson et al. +!""B/ and 7rassie et al.
+!""B/ using a >mm thic' glass sheet to cover the top of the bo% collector and
maintaining an air gap of ;=mm established good collector performance.
1.1.1.3 A>v,.:,(+0 23 ICS)SWH
?ectangular $&(,(-. have several advantages over other systems. *he direct contact
of the wor'ing fluid with the absorber plate increases the system efficiency by
eliminating the need of a heat e%changer used in most heating systems. *he absence of
conduits and connecting pipes to the storage tan' results in eliminating the bul' heat
losses appearing in other type of systems. @ther advantages of $&(,(-. are the
effective use of direct and diffuse radiation as well as the absence of solar trac'ing
requirements and minimal maintenance as well as their low cost and good overall
performance. *hey can be designed such that they seamlessly integrate in roof
structures giving them an additional aesthetical and architectural value.
;"
2.1.3 S24,- A,:+- 1+,:+- 9,-,8+:+-0
1.1.1.# O9:i878 ,.(4+ 23 i.64i.,:i2.
Different studies show different optimum angle of inclinations for Edinburgh city.
Duffie and Bec'man +9#:>/ stated an optimal angle to be ".# times the latitude of the
location. *his asserts the collector tilt can ta'e values from "S at the equator to #"S1 in
Polar regions and would be ="S in Edinburgh. ?eiss and Bainbridge +9#:C/ in their
research showed that systems in higher northern latitudes are often optimised for winter
collection and are therefore tilted = to 9" more than the angle of latitude. Another
research carried out by uneer et al. +!""Bb/ based on !: year irradiance data for
Edinburgh established that higher irradiation occurred at ;"S inclination for a south
facing wall. .owever1 irradiation is not the only factor influencing a solar water heater.
A more recent study conducted by .enderson et al. +!"":/ showed that an $&(,(-.
would perform better in Edinburgh at ="S compared to ;=S because of its higher Nusselt
number value. *he study covered the behaviour of the water cavity collector for
different inclination angles. *he effects were evaluated using two modes of analysis in
con0unction) &omputational 3luid Dynamic +&3D/ software and e%perimental testing.
*he authors found another advantage of having an angle of ="S compared to ;=S would
be the reduction of heat losses for higher inclination angles. 3or regions with lower
latitudes1 increasing the tilt angle initially improves the thermal efficiency. .owever1
any successive increase in tilt would reduce the efficiency as the solar gain value would
become very low. Based on these past studies1 an inclination angle of >= south was
implemented.
1.1.1." S:-,:i3i6,:i2.
*hermal stratification is an important parameter in the efficient operation of solar
collector systems. (everal authors XOan <oppen et al. +9#:#/1 3urbo and i''elsen
+9#C:/ and Oan Ber'el +9##B/Y highlighted the importance of stratification in detached
vertical storage tan's of solar water heating systems. *he merits of stratification were
then described by .ollands and 2ightstone +9#C#/ who demonstrated that (-. systems
operating with thermally stratified storage tan's have substantially improved
performance1 particularly at low flow rates. &omparisons between stratified and mi%ed
tan's were carried out by Oan Ber'el +9##B/ and &ristofari et al. +!"";/. Oan Ber'el
+9##B/ demonstrated that mi%ed water tan's performed disappointingly in terms of
output and collection efficiency compared to a well stratified tan'. &ristofari et al.
;9
+!"";/ studied the performance of a classical solar water heating installation with a flat
plate thermal collector and water tan' storage for storage in low flow conditions
installed at a editerranean site1 A0accio1 &orsica1 3rance. ?esults showed that a
stratified tan' has higher performance than a fully mi%ed tan' with monthly mean
temperatures varying from 9#.;& to B#.:& for a fully mi%ed tan' and from !".9& to
:9.!& for a high thermal stratified tan' illustrated in 3igure !,!.
Fi(. 2)&* M2.:14/ 8+,. :+89+-,:7-+ 23 :1+ 0:2-,(+ 347i> 32- 3744 8ix+> :,.C ,.> 0:-,:i3i+> :,.C;
C-i0:23,-i +: ,4. (2''3).
*hermal stratification in $&(,(-. has been observed in much research but poorly
analysed. @nly four studies on stratification in $&(,(-. were reported by uneer
+9#C=/1 Burns et al. +9#C=/1 (myth et al. +9###/ and .enderson et al. +!""B/. *he
uneer +9#C=/ study involved change of vessel depth and showed no significant
influence of performance with change of depth. Burns et al. +9#C=/ evaluated an
insulated single glaDed rectangular wooden bo% containing a !="l blac' painted drum at
a tilt angle of >" in (an Diego1 &alifornia1 U(A. .is research concluded that thermal
stratification was very important in the efficient operation of $&(,(-. systems1
however1 it was suggested that these systems should only be used for preheating. *his
research is limited1 however as tests were only carried out for eight days in April and
thus cannot predict the hot water production for a year period. (myth et al. +9###/
developed a heat retaining $&(,vessel. .is research showed that reduced vessel wall
conduction and thus convective motion and fluid mi%ing1 improved thermal
stratification. 3inally1 the last research in date of thermal stratification in $&(,(-. was
;!
performed by .enderson et al. +!""B/ who developed one of the first studies on thermal
stratification for a rectangular $&(,(-. at various inclinations and constant heat flu%
values. *he authors showed that stratification played an essential role in defining the
heat gain characteristics when the storage tan' was in direct contact with the heating
surface. E%periments and &3D analysis were performed and showed that the degree of
stratification in the $&(,(-. depends upon the angle of inclination1 the e%posure time
and the incident insolation. *he same author posited that stratification in solar water
heaters can be roughly divided into two classesF +a/ static mode1 +b/ dynamic mode. *he
static mode is a limited study of temperature for fully or partially charged heaters1 while
a dynamic mode deals with charging and discharging cycle.
(everal methods are suggested in the literature to maintain and promote stratification in
storage tan's by utiliDing sleeves partitions1 baffles1 porous inta'e manifolds and low
flow rates. A water storage tan' would naturally remain stratified until water is supplied
to the load. As the hot water leaves the tan' and fresh cold water is infused1 mi%ing
occurs and as a result of flow circulation leads to de,stratification. *he location of the
outlet and inlet pipes should then be designed in accordance to the stratification studies
to not destroy stratification into the water tan'. 7ertDos and &aouris +!""C/ investigated
the influence of ten different inlet and outlet port placement on the velocity field. .is
research concluded that the most suitable arrangement consisted in the inlet and outlet
ports placed sideways and in opposite position in order to not destroy stratification in
the $&(,(-..
odelling stratification has been sub0ect to much research. (everal multi,node stratified
tan' models have been developed and compiled by Duffie and Bec'man +9##9/. ost
recent research on stratification was carried using the transient simulation pac'age
*?N(W( to simulate thermal stratification in water storage tan's such as &ruic'shan'
and .arrison +!""B/ and 4ordan and 3urbo +!""=/. .owever1 e%isting models are only
applicable to vertical storage tan's as a separate component of the collector. *o the
'nowledge of the author1 no stratification model has been presently suggested for
inclined $&(,(-..
*he literature showed thermal benefit of high stratification and emphasised the use of
low flow rates as being essential for the good performance of a collector1 however
improved mi%ing in the tan' results in higher rates of heat transmission. *hen
;;
stratification is important but good mi%ing in some case would result in better
performances for the (-.. (tratification is significant when trying to optimise the
collector for draw off as it helps to provide temperatures closer to those required for
domestic hot water systems. (tratification is critical to evaluating the actual heat content
inside the heater1 which is vital for estimating the efficiency. *herefore1 further research
is required in modelling stratification for $&(,(-..
1.1.1.% D-,A)233 61,-,6:+-i0:i60
*he word MDraw,offN refers to the collection of hot water from the collector. *he study
of draw off patterns from an $&(,(-. is important in order to evaluate the practical
performance and thus the effective output of the collector. Draw,off characteristics can
be organised in four sections. 3irst1 hot water temperature to be achieved and hot water
consumption1 representing the total water drawn,off1 need to be identified for the U<.
*hen the flow rate of water draw,off varies and is of primary importance. 3inally1 the
time of the day the water is drawn,off1 thus the draw off pattern1 needs to be identified.
*he hot water consumption within a domestic property is dependent upon the property
siDe1 the number of occupants1 household income1 lifestyle habits and the type and
number of end,use appliances. *he primary domestic uses of hot water includes
bathingJshowering1 hand washing1 clothes washing and dish washing. uneer +!""9/
indicates that the appro%imate temperature requirements for domestic hot water are
>"S& for bathsJshowers and hand basins1 ="S& for dishwashing by hand and up to C"S&
for clothes and dish washing machines. &ourtney +9#::/ reports typical delivered hot
water temperature in the region of ="S& to B"S& which are then diluted with cold water
or electrically heated at the end,use appliance to achieve the desired temperature.
?eliable data on domestic hot water consumption is diverse and some of the available
information is dated. -hittle +9#:C/ indicates demand for hot water consumption can be
anywhere between ;B lJdayJperson to =B lJdayJperson at a delivery temperature of ==S&.
.owever1 in a 9##C Energy 3ollow Up (urvey conducted on behalf of D*$ +!""=/ and
DE3?A an average ;# litres per day for washing machines1 baths and showers with a
further 9" litres for assumptions about the use of hot water at sin's and basins was
suggested per person per day. According to EU?@PA +!""B/ an average European uses
;B litres of B"S& hot water per day. A recent survey conducted by c2ennan +!""B/
e%amined the water consumption through a typical day for (cottish households. *he
survey was carried out for ;! properties monitoring hourly activity for domestic hot
;>
water appliances over a one,wee' period. *he average number of people living in a
household was !.>> persons1 comparable to the U<,&ensus +!""9/ average of !.;B
persons per household in !""9. *he hot water consumption too' account of all
instantaneous heaters such as electric power showers1 water heated internally by
domestic appliances i.e. washing machines and dishwashers1 and hot water drawn from
boilers or hot water tan's that is not used for space heating. *he average wee'day hot
water consumption for all properties surveyed was >#.# lJdayJperson based on a delivery
temperature of ==S&.
As stated before1 hot water requirement varies between households. *he flow rate and
pattern of water drawn are important parameters to the performance of the heater.
.ollands and 2ightstone +9#C#/ suggested that low flow rates from the storage tan' can
improve the system performance. 3anney and <lein +9#C:/ also investigated the
influence of the collector and heat e%changer flow rates on the performance of the
(-.. 2owering the collector fluid flow rate for direct systems resulted in
improvements to the thermal performance of the heater. *his was attributed to better
stratification within the solar tan' and a significant reduction in mi%ing1 which occurred
between the solar and au%iliary heated portions in the single tan'. Until the 9#C"s flow
rates were commonly around "."9= 'gJm
!
.s as stated by Duffie and Bec'man +9##9/.
2ower rates are recently being used such as ".""! to ".""B 'gJm
!
.s in (weden where
flow rates are given per m
!
of collector area.
$t is important to review the different draw,off patterns1 utilised in the past in order to
determine the optimum draw off profile for the present research. -hen water is drawn
off the corresponding quantity of cold water enters the $&( at the bottom1 thereby
decreasing the temperature1 which ta'es time for the store temperature to recover.
(everal studies have been carried out on draw,off profiles in the U<. An investigation
by &ourtney +9#::/ showed that if the period of pea' water demand occurred in the
evening rather than the morning1 annual heat output would increase by appro%imately
".= 74 for a > m
!
solar collector delivering 9:" lJday. Evening pea' demands ensure
solar contributions are utilised and there are no significant overnight losses that may
occur with a poorly insulated hot water heating system1 however lower tan'
temperatures achieved through daytime draw,offs can allow collectors to operate more
effectively. &ruD et al. +!""!/ utilised the British (tandard domestic heating load profile
as the model for the simulation of water demand in his trapeDoidal $&(,(-.. Different
;=
draw,off profiles can be found for the U<. *ypical demand profile data from U<,$(E(
supplying ;""lJday are represented in 3igure !,; compared to a more recent study
carried out by c2ennan +!""B/ for ;! Edinburgh households. .is study showed hot
water draw,off profile for a > m
!
$&( system supplying !"" lJday for a property
occupied by four persons1 assuming =" litres per day per person. $n both hot water
draw,off profiles1 significant periods of hot water demand are between Cam and #am
and :pm to Cpm.
0
&
10
1&
$0
$&
30
3&
%0
%&
&0
1 $ 3 % & ' ( 8 3 10 11 1$ 13 1% 1& 1' 1( 18 13 $0 $1 $$ $3 $%
3our of %a-
9
i
t
r
e
s
1+Lennan deand profile
")45-!- deand profile
Fi(. 2)<* H2: A,:+- >-,A)233 9-23i4+0 32- 0:7>i+0 23 M6L+..,. (2''%) ,.> UK)ISES 32- 2''4?>,/
,.> 3''4?>,/ -+09+6:iv+ 42,>0.
Another more recent study for Edinburgh was carried out by 4unaidi et al. +!""C/ for an
$&(,(-. connected to a water tan'. 3our draw off patterns were studied shown in
table !,9 below.
T,54+ 2)2* D+:,i40 23 >-,A)233 9,::+-.0 70+> 5/ H7.,i>i +: ,4. (2''<)
Draw,off
pattern
9
st
Draw,off
Oolume J time
!
nd
Draw,off
Oolume J time
;
rd
Draw,off
Oolume J time
>
th
Draw,off
Oolume J time
A !=l "#F"" !=l 9"F"" !=l 9;F"" !=l 9:F""
B !=l "CF"" !=l "#F"" !=l 9!F"" !=l 9:F""
& ="l 9"F"" ="l 9#F""
D ="l 9!F"" ="l 9#F""
*he research suggested that solution A and B could achieve a temperature of >"& in
the tan's. (cheme B was suggested as the best in terms of total energy e%traction.
4unaidi et al. +!""C/ show the benefit in having four draw offs instead of two and
therefore classified the four draw,off scheme from best to worstF B1 A1 D1 &. .owever
this draw,off profiles negate the advantage of the integrated tan' within an $&(,(-..
;B
*herefore1 more realistic draw,off scheme such as c2ennan +!""B/ are generally
preferred.
1.1.1.& P+-32-8,.6+ 61,-,6:+-i0,:i2. 23 ,. ICS)SWH
*he various types of $&(,(-. developed over time have impelled the need for a
common criterion to measure the effectiveness of (-.. As a result1 organisations such
as the European (olar &ollector and (ystems *esting 7roup1 E(&(*7 +9#C#/1
developed a standard testing procedure for all forms of (-. used in the European
&ommunity +E&/. $n the U(A1 the A(.?AE (tandard,#= +9#C9/ also developed
standards procedures in order to assess short term testing performance of (-.. *he
widely used test procedure proposed by the E(&(*7 based on a combination of test
methods showed a realistic $&(,(-. performance representation allowing a direct
comparison with other e%perimentally investigated $&(,(-.. As stated by (myth et al.
+!""B/1 the collection efficiency is similar to the .ottel,-hillier,Bliss equation
developed by .ottel and -hillier +9#==/ and Bliss +9#=#/ for distributed type solar
collectors1 but applies to diurnal system performance. &ollection efficiency is plotted
against values of
a-e
& T J
where
a-e
&
is the average insolation on the collector aperture
and T is the difference between the water temperature of the heating period and
ambient temperature over the same period. *his performance method is applied to flat
plate collector wor'ing in steady state mode. $t can be applied to $&(,(-.) however it
is not representative of the actual system performance as it is not a steady state situation.
Direct comparison of measured system performances is relatively straight forward)
however1 predicting the performance of an $&(,(-. at another location or with a
different installation configuration is more challenging. *herefore a review of the
different methods to predict $&(,(-. performances was carried out in the following
section.
2.2 M2>+44i.( :1+ 024,- A,:+- 1+,:+-
2.2.1 M,6-2 82>+4 B T1+-8,4 82>+4
As stated above1 predicting the performance of an $&(,(-. is a challenging tas'. A
model able to predict water temperature in an $&(,(-. would be a useful tool to allow
any user to predict the performance of the installed (-.1 depending on the orientation
;:
and location. Builders could use such a tool to demonstrate the performance of the $&(,
(-. and the benefit of such installation.
*he first study to propose and use computer simulation to predict $&(,(-.
performance was conducted by 7arg et al. +9#:!/ using a single node system. Burns et
al. +9#C=/ developed an analytical model to predict the performance of a stratified $&(,
(-.. 7ood agreement between theoretical and e%perimental results was reported. 2ater
studies by &ruD et al. +!""!/ investigated the energy saving potential provided by a
trapeDoidal,shaped solar collector. *his was evaluated e%perimentally and by
computation1 using a thermal networ' analysis model based on 2uiD +9#C=/. *he
thermal networ' showing the equivalent heat flow circuit collector1 insulation and losses
is showed in 3igure !,>. A full scale e%perimental test rig was employed to validate
their modelling. Despite good agreement of simulation and e%perimental results1 the
authors failed to account for two factors. 3irst1 the heat capacity of the system was
neglected. (econd1 tests were carried out in a laboratory which is unrepresentative of
field conditions. *hus1 there is a need to develop a more accurate device to predict
performance for $&(,(-.. 4unaidi +!"":/ developed a similar program for fluid flow
and heat transfer generating bul' water temperatures for a rectangular shaped $&(,(-.
using specific parameter such as a constant heat flu% for the day. *he e%perimental and
computed temperatures were found with a deviation of 9"L. *hese models are limited
as they can not be applied to real weather conditions. *here is therefore a need to
develop an improved model able to predict the bul' water temperature inside the
collector for any weather condition and account for different $&(,(-. design
parameters. *o achieve such a program1 heat transfer mechanisms in the collector need
to be reviewed and assessed.
;C
Fi(. 2)$* E=7iv,4+.: 1+,: 342A 6i-67i: 32- C-7E +: ,4 (2''2) 6244+6:2-; i.074,:i2. ,.> 4200+0
*he three heat transfer mechanisms1 conduction1 radiation and convection1 occur inside
the (-. and result in heat losses or gains. An $&(,(-. is an assembly of two natural
convection cavities composed of an air cavity1 represented by the space between the
absorber plate and the glass cover1 sitting on top of a water cavity. &onvective
behaviour of both cavities are different for a given angle of inclination as they operate at
different ?ayleigh numbers and most li'ely have dissimilar geometrical dimensions as
their aspect ratio. Both cavities are insulated by a batch system where conduction is
occurring. A representation of an $&(,(-. heat transfer mechanism is represented in
3igure !,=.
;#
Fi(. 2)1'* H+,: :-,.03+- 8+61,.i08 i. ,. ICS)SWH
*he different mechanisms emerging from the system are detailed belowF
, &onvection from two parallel inclined plates at different temperatures +air cavity/
, &onvection from an inclined flat plate +water cavity/
, &onvection from cover to the environment
, &onvection from bo% to the environment
, ?adiation from plate to cover
, ?adiation from cover to s'y
, &onduction through bo%
, .eat capacitance of the system
A review of widely used equations for natural convection and radiation are detailed
below.
1.1.1.< N,:7-,4 62.v+6:i2. 3-28 :A2 9,-,44+4 i.64i.+> 94,:+0 B ,i- 6,vi:/
Natural convection in air cavities has been studied by many researchers and is a well,
established phenomenon. .ollands et al. +9#:B/ presented new e%perimental
measurements on free convective heat transfer rates through inclined air layers of high
aspect ratio1 heated from below. *he ?ayleigh number ranged from subcritical to 9"
=
)
the range of the angle of inclination MN was between " and :" degrees. Although it
was anticipated that the results might be identical to the results for the horiDontal layer if
one replaced M?aN by M 'a(os N1 significant departures from the introduction of the
angle were observed. As a result of their study1 .ollands et al. +9#:B/ recommended a
relationship giving the Nusselt number as a function of 'a(os e%pressed below. *his
e%pression has been since widely accepted.
>"

1
1
]
1

,
_

+
1
]
1

1
]
1

+ 9
=C;"
cos
cos
/ C . 9 +sin 9:"C
9
cos
9:"C
9 >> . 9 9
; J 9
B . 9

L
L L
a
'a
'a 'a
*$ +!.9/
*he notation K implies that1 if the quantity in brac'ets is negative1 it must be set equal to
Dero. ?estriction in usage of this relation are that .J2 Z 9! and "[\ ]
('

1 where
('

1 is the critical tilt angle which varies with the aspect ratios as represented in *able
!,!.

T,54+ 2)3* C-i:i6,4 ,.(4+ v+-070 ,09+6: -,:i2; H244,.>0 +: ,4. (1$&%)
.J2 9 ; B 9! 9!
('

!= =; B" B: :"
*he ?ayleigh number is e%pressed belowF

;
/. .+ . L T T !
'a
c p

+!.!/
Equation +9/ has been found to give e%cellent results when applied inside the regime
described above. *he angle depends on the orientation of the hot surface as shown in
3igure !,B.
Fi(. 2)11* O-i+.:,:i2. ,.(4+ 32- ,i- 6,vi:/ -+(-+00i2.
1.1.1.$ N,:7-,4 62.v+6:i2. 3-28 i.64i.+> 34,: 94,:+0 B A,:+- 6,vi:/
any studies have been carried out on inclined heated plates X3u0i and $mura +9#:!/1
&hurchill and &hu +9#:=/1 3ussey and -arneford +9#:C/1 &ruD et al. +!""!/Y providing
a good breadth of conditions with appropriate regressions. 3u0i and $mura +9#:!/
studied natural heat transfer convection from a plate with arbitrary inclination. *he heat
>9

9C"

"
&ool surface
.ot surface
was transferred from one side surface of two plates. *he authors focused on natural
convection from an isolated1 vertical plate1 and e%amined the influence of the tilt angle
over the range of

C# " and
99 =
9" 9" < < 'a(os . 3or downward facing
heating plate the authors suggested the following regressionF
( )
> J 9
=B . " 'a(os *$ +!.;/
*he value of the e%ponent
( ) > J 9
suggest a laminar flow up to values of the parameter
'a(os of
99
9" . 2ater studies from &ruD et al. +9###/ investigated buoyancy,driven
convective heat e%change in a trapeDoidal,shaped solar water heater. Using results from
3u0i and $mura +9#:!/1 &ruD et al. +9###/ developed a new formula of the Nusselt
number for inclined water cavities with an estimated uncertainty of :L e%pressed
belowF
( )
B J 9 # J 9
> J 9
H
=B . "
,
_

,
_

#
H
#
D
'a(os *$ +!.>/
-here D is the +front,to,bac'/ bottom depth of cavity floor1 .5 is the inclined height of
the heater) characteristic length and U is the +front,to,bac'/ top depth of cavity floor
with ?ayleigh number of equationF
!
;
H
/. .+ . . Pr

,
_

H T T !
'a
p
+!.=/
*his e%pression was found in good agreement with the correlation of 3u0i and $mura
+9#:!/1 when

>= 1 indicating that the geometric parameters +DJU and .5JU/ have
only a small influence in comparison with the modified ?ayleigh number1 'a(os .
*he parameter .5 represents the inclined height of the heater1 characteristic length. *he
geometric parameters of &ruD et al. +9###/ are shown in 3igure !,:.
>!
Fi(. 2)12* G+28+:-i6 9,-,8+:+- 23 C-7E +: ,4. (1$$$) :-,9+E2i>,4)01,9+> 024,-?+.+-(/ 0:2-+
1.1.1.1' R,>i,:i2. 3-28 94,:+ :2 62v+-
Duffie and Bec'man +9#:>/ suggested a standard equation for heat e%change radiation
between two parallel plates such as the absorber plate and the glass cover of a collector.
B
c p
c p r
T T
%

/ +
> >
1

+!.B/
where
B

is the bul' emissivity of equation

9
9 9
+
+ c

1
+
T
and
+

are the
temperature and emissivity of the absorber plate1
c
T
and
c

are the temperature and

emissivity of the glass cover temperature1 is the (tefan,BoltDmann constant.
1.1.1.11 R,>i,:i2. 3-28 62v+- :2 0C/
?adiation losses from the cover to the environment represent an important fraction of
the total heat losses) therefore they need to be determined precisely. $n a laboratory
environment1 only ambient conditions apply. .owever1 in real weather conditions the
system reacts differently to the environment. ?adiation losses from cover to s'y can be
determined such asF
s$r
Sky c
s c r
T T
%

/ .+
> >
1

+!.:/
where
s$r

is the bul' emissivity of equation

9
9 9
+
Sky c

1 Sky
T
and Sky

are the
temperature and emissivity of the s'y.
>;
*he s'y5s temperature is assumed to be equal to the ambient temperature for laboratory
conditions. .owever1 for field tests the s'y acts as a blac' body during a clear night s'y.
*he s'y emissivity is therefore an important parameter for inclusion in models in order
to simulate real weather conditions. any correlations have been reported in literature
for calculating the coefficient of emissivity inside the atmospheric transparency window
such as $dso +9#C9/1 Berdahl and artin +9#C>/ or &ucumo et al. +!""B/. Equations for
s'y emissivity are generally valid only for clear s'y conditions) as in equation +C/ below
developed by &ucumo et al. +!""B/ for s'y emissivity in the atmospheric transparency
window1
(S

.
> . :;=#> C . BC"
/ 9 + 9":#=!
9
!
+

+
a a
(S
T T

+!.C/
-ith
!
9""
BBBC . "
9""
B""9 . " :>: . "

,
_

+
dp dp
T T
+!.#/
-here
(S

is the atmospheric emissivity in the transparency window for clear s'y

conditions and

is the global emissivity.

*he value of s'y emissivity obtained from equation +C/ is valid for clear s'y conditions
only. *o,date only a few studies have been carried out to e%amine the s'y emissivity
under overcast s'y conditions. Under an occluded or cloudy s'y the emissivity1 Sky

1can be calculated from (antamouris and Asima'opoulos +9##B/ general relationship

e%pressed asF
/ """!C . " "";= . " "!!> . " 9 +
; !
n n n
(S Sky
+ +
+!.9"/
-here n is the total opaque cloud inde%1 " for clear s'y and 9 for overcast s'y.
*he s'y temperature is also an important parameter regarding radiation losses from the
cover. any studies have been carried out on s'y temperatures by Duffie and Bec'man
+9##9/1 (winban' +9#B;/ or Berdahl and artin +9#C>/. (winban' +9#B;/ suggested a
s'y temperature regression ofF
= . 9
==! . "
a Sky
T T +!.99/
>>
*he more recent study by Berdahl and artin +9#C>/ used e%tensive data from the
United (tates to relate the s'y temperature1 Sky
T
1 to the dew point temperature1 dp
T
1
ambient temperature and hour from midnight1 t. *he equation developed is e%pressed asF
[ ]
> J 9 !
/ 9= + "9; . " """":; . " ""=B . " :99 . " t (os T T T T
dp dp a Sky
+ + +
+!.9!/
*he emission within the atmospheric window1 and thus the s'y temperature1 depends on
the local weather1 and varies throughout the day as well as from one location to another.
*he s'y temperature shall therefore be related to weather and time parameters to
enhance the accuracy of the results) which resulted in using the Berdahl and artin
+9#C>/ model.
By using the defined equations of the heat transfer mechanism occurring in the system1
a thermal networ' analysis of $&(,(-. can be performed. Another approach to model
$&(,(-. is the use of ;,D simulation programs.
2.2.2 Mi6-2 82>+4 B C2897:,:i2.,4 F47i> D/.,8i60 (CFD)
&3D in simple terms is the use of computer and numerical methods to solve and analyse
problems involving fluid flow. &3D has grown from a mathematical curiosity to an
essential tool in almost every branch of fluid dynamics1 from aerospace propulsion to
weather prediction. Although &3D plays an important role in providing complementary
information about environmental performances1 it is usually unable to give an accurate
prediction due to errors involved in numerical calculations. *herefore1 relying solely on
&3D is not recommended and results should be validated through e%periments.
Although &3D did not eliminate the need of e%periments1 it has significantly reduced
the amount needed. &3D is very often used in product concept1 product development
and also virtual prototypes such as (-.. l$ent software is of interest for this study. $t
is a general purpose &3D code based on the finite volume method that offers a wide
array of physical ! and ;, dimensional +;,D/ models. *his software is widely used in
several academic institutes and commercial research and development as it saves time
and effort while maintaining a high degree of accuracy.
>=
?ecent studies indicate an increase in the use of &3D for analysing and optimising solar
collector5s design and performance. *he pioneering study of 7ill +9#BB/ on cavities
used a finite element method to solve the flow in a square cavity. $n 9#C; the
'nowledge of flows was e%tended by numerical investigation of ;,D aspects by @Doe
and 3u0i +9#C;/. A finite difference technique was employed to calculate the
longitudinal roll pattern and changes with angles for a cavity heated from the bottom.
Based on his wor'1 Ehong et al. +9#C=/ inspected the influence of varied fluid properties
and radiation. A comprehensive review on inclined cavities was reported by Wang
+!"";/ whose study was based on laminar natural convective flow in inclined
rectangular glaDing cavities. *he wor' on flow pattern inside the storage tan' of $&(,
(-. is still largely une%plored. @nly two studies from the author5s 'nowledge have
been recorded. 7roenhout et al. +!""!/ conducted a !,D &3D analysis involving both
radiation and convection in a solar collector to compare with e%perimental results. &3D
model results underestimated e%perimental data. $ntensive ;,D &3D analysis on $&(,
(-. was carried out by .enderson et al. +!"":/. .is research loo'ed at air and water
cavities and at the $&(,(-. as an entity under different type of boundary conditions.
*he specification of boundary conditions is essential before starting any simulation.
*his specification consists of two elementsF +a/ the boundary conditions type1 and +b/
the location where it is applied. Boundary conditions applied to the system vary with the
purpose of the simulation. .enderson et al. +!"":/ studied the flow behaviour of air and
water cavities for different angles and different boundary conditions in order to optimise
the design of the $&(,(-.. $n order to compare with the results of .enderson et al.
+!"":/ the same model boundary conditions and similar designs were used.
@nce the optimal virtual design is achieved1 a prototype is manufactured for testing and
the technology is sub0ected to other assessment such as economic1 energetic and
environmental which can be evaluated by using life cycle assessment.
2.3 T1+ i.:+(-,:i2. 23 024,- A,:+- 1+,:+-0
2.3.1 Li3+ 6/64+ ,00+008+.: (LCA)
*oday5s economy is made of strict environmental controls and changing mar'ets in
which society is becoming progressively more aware of environmental matters. 3ield et
al. +9##;/ outlined that 2&A is one of the most useful tools for the study and analysis of
strategies to meet the challenges of tomorrow. 2&A is a method used for assessing the
>B
environmental impact by identifying the material1 energy and waste flows of a product1
process or activity over its entire life,cycle. *he 2&A of a product includes each stage
from raw material e%traction1 manufacturing1 retailing1 use and maintenance to disposal
or recycling. (implified processes typically considered in 2&A for a product as shown
in 3igure !,C.
Fi(. 2)13* Si894i3i+> 9-26+00+0 :/9i6,44/ 62.0i>+-+> i. LCA 32- , 9-2>76:
*his tool can help companies to identify necessary changes to their operations and
products and as a result improve environmental performance and reduce cost. 2&A
methodology is a recent concept which has developed as a consequence of an enhanced
awareness of environmental pressures by industry1 the public and governments. 2&A
was initially developed by the (ociety of Environmental *o%icology and &hemistry
+(E*A&/ and was later optimised by the $nternational @rganiDation for (tandardiDation
+$(@,9>">" +!""B//. 2&A became progressively a more important tool for engineers1
scientists1 designers1 managers and environmentalists.
A full 2&A study includes four different phasesF 7oal and (cope Definitions1 2ife
&ycle $nventory Analysis1 2ife &ycle $mpact Assessment and finally the $nterpretation.
Numerous studies have been carried out on 2&A identifying the four phases of 2&A for
different products X-eir and uneer +9##C/1 Ardente et al +!""=a and !""=b/1
*silingiridis et al +!"">/1 &rawford at al +!"">/1 *aborians'i and Prado +!"">/1 *an and
<hoo +!""=/ or Asif et al. +!""B/Y. *he relationship of these phases is illustrated in
>:
3igure !,#. *he goal and scope phase is the point at which decisions need to be made
concerning the purpose of the 2&A and what it is going to cover including the
identification of the studyHs limits or system boundaries. @nce the first phase is set1 the
inventory analysis phase provides data on environmental burdens associated with the
product or system. Energy and raw material used1 solid1 gaseous and liquid wastes
generated are e%amined and measured. *he impact assessment estimates how the
product or system affects the environment. $t typically classifies the environmental
burdens with the same environmental impact and then estimates their contribution to
that particular environmental impact. At this stage1 the main factor contributing to an
environmental impact may be identified1 allowing improvements to be processed.
3inally1 the interpretation phase can be described as an iterative process. Although
issues might be identified at any stage1 this phase loo's ob0ectively to the whole life
cycle in order to assess improvement procedures to reduce environmental burdens
associated with the product or system.
Fi(. 2)1#* E4+8+.:0 23 , 3744 LCA; E.vi-2.8+.:,4)T+61.242(/)B+0:)P-,6:i6+)P-2(-,88+ (2''').
A full 2&A for a product1 particularly for a complicated product can be a time,
consuming and costly e%ercise) therefore the principle of streamlined 2&A is often used
to simplify the study. *his type of study identifies the product or process characteristics
that are critical in the conte%t of the study and includes only these in the streamlined
2&A. any studies have used streamlined 2&A as it offers the potential to deliver
results more quic'ly and at a lower cost. *hese have been mainly conducted by Asif et
>C
al. +!""=/ and Asif et al. +!"":b/1 and more particularly on $&(,(-. by Ardente et al
+!""=a and !""=b/1 Asif and uneer +!""B/ and Asif et al. +!"":a/. $n both the latter
studies1 the author5s research only included an account of the embodied energy and
environmental impacts associated with the materials involved in the construction of the
$&(,(-.. Environmental impacts were only e%pressed in terms of carbon dio%ide
emission values of the involved materials. &arrying out a 2&A on a new product is of
primary importance to achieve a better product. .owever1 another parameter introduced
by Asif et al. +!""=/ is the cost effectiveness of products and services. *his parameter
has become very important in today5s mar'et) therefore1 a monetary analysis +A/ can
be carried out to estimate the overall costs of a pro0ect over its life. &osts such as
capital1 running1 energy and others such as unpredicted maintenance and repair costs are
normally included.
An 2&A and A were adopted to evaluate the energy usage1 environmental impact and
economic monetary performance of an $&(,(-.. Both assessments differ depending
on the country in which they operate according to varying energy prices and local
weather conditions.
2.3.2 W+,:1+- 62.>i:i2.0 i. S62:4,.>
(cotlandHs climate and topography provide a great diversity for potential renewable
energy resources1 whose use could reduce fossil fuel dependence. $t is surrounded by
coast apart from in the south where it shares a land border with England. *he 7ulf
(tream significantly affects the climate. (cotland is warmed by this stream and as such
is much warmer than it would otherwise be. Because the sea surrounds (cotland1 there
is a great opportunity to generate energy with wind farms or offshore wave power as
shown by Bruce +9##>/. (cotland has an abundant on, and offshore wind resource but
its potential is limited by the environmental impact of large numbers of wind farms.
-indy conditions increase heat losses from the top cover of (-.. *herefore1 ways to
limit heat losses have to be considered.
(olar energy in the U< and (cotland is far greater than is commonly perceived as
illustrated in 3igure !,9" showing the vast potential mar'et for solar systems.
>#
Fi(. 2)1"* +,-4/ :2:,4 (425,4 12-iE2.:,4 024,- i--,>i,:i2. i. CW1?8
2
; UK; FG-i +: ,4. (2''&)
ac7regor and Balmbro +9#C>/ showed that (cotland has one of the best climates in
Europe for solar water heating in buildings due to the regular heating profile of (cotland
with a heating season lasting between nine and ten months. ac7regor and Balmbro
+9#C>/ believe (cotlandHs first solar priority should be for providing heat1 not electricity.
(cotland is a cold country where people use much more heat than electrical energy)
therefore solar heating is much more cost,effective than solar electricity generation.
*his brings us to the main conclusion that there is much scope for renewable energy in
(cotland. *ests were located in Edinburgh city +=BN1 ;.!-/. *he city is located ;=m
above sea level in an oceanic climate with large differences between summer and winter
radiation levels. Past studies by uneer et al. +!""Bb/ show a pea' value of solar
radiance for Edinburgh of around #""-Jm
!
on a bright sunny summer day. .owever in
winter conditions the ma0ority of time the values for solar radiance are around
>""-Jm
!
.
Ambient temperatures of Edinburgh can reach freeDing point and below. .owever1 most
$&(,(-.s have large thermal masses and simple compact structures which give good
general resistance to freeDing climates. 3igure !,99 based on the study of (myth et al.
+!""9/ on performance and freeDe resistance capabilities of the $&(,(-. in winter
conditions shows that the U< is within the range of operation for $&(,(-..
="
Fi(. 2)1%* Av+-,(+ ,85i+.: :+89+-,:7-+ -,.(+0 i. E7-29+ 32- :1+ 82.:1 23 H,.7,-/ ,.> :1+ 820:
9-25,54+ ,-+, 23 Ai.:+- 29+-,:i2. 32- :1+ ICSSWH >+0i(.; S8/:1 +: ,4. (2''1).
Nevertheless1 $&(,(-. should ta'e this parameter into account when design and
integration into building phases occurs using careful construction1 good thermal
insulation1 fair thermal mass1 properly controlled ventilation and good integration as
described by $mbabi and usset +9##B/.
2.3.3 B7i4>i.( i.:+(-,:+> SWH
Architectural quality of integration of solar thermal technologies is a ma0or issue which
is a critical factor to their success. (olar thermal systems are now becoming increasingly
used in architectural design of roofs facades or balconies as shown by 4aehnig et al.
+!"":/. (-. are typically integrated into roof designs as they can be simply and
quic'ly installed into the cladding of the roof 'eeping costs relatively low. $ndividual
collectors are now made in a wide range of siDe and shapes so they can fit the e%act
shape of the roof as shown in 3igure !,9! a and c. 3acade integrated solar collectors are
less usual but are increasingly appearing as an architectural design as well as generating
energy. A more unusual set up is their recent integration in balconies such as in Oal
*horens1 3rance and 2iepDig1 7ermany whose facade was renovated with added
insulation and nine pre,fabricated balcony bloc's on which the solar collector were
mounted. *he 3rench and 7erman buildings were renovated in !""> and !""9
=9
respectively with the latter shown in 3igure !,9! d. 3igure !,9! shows the different
possible solar collector5s integrations.
a, Berlin1 7ermany
b, Dornbirn1 Austria
c, athoyJ*isun1 Austria d, 2eipDig1 7ermany
Fi(. 2)1&* S24,- 6244+6:2- 9200i54+ i.:+(-,:i2.; H,+1.i( +: ,4. (2''&).
unari Probst and ?oec'er +!"":/ conducted a survey on the different perception of
architects and engineers on the integration quality of building integrated solar thermal
technologies. *he study addressed glaDed and unglaDed collector solar thermal
collectors1 with blac' and coloured absorbers with different methods of integration.
3our were roof integrated while si% were on fa^ades comprising a variety of building
types1 materials and colours. *his study showed how essential teamwor' is between
architects and engineers in order to achieve both energy efficiency and architectural
integration and how the system characteristics affecting the building appearance have an
impact on integration quality. .estnes +9###/ suggested that collectors should be
developed to respond to technical constraints1 but should furthermore become an
architectural element1 conceived to be integrated into the buildings s'in1 to ease
designer5s integration efforts and to reduce the overall costs.
=!
odern ethod of &onstruction +&/ could help to achieve this purpose. English,
Partnership +!"":/ describes & as ma'ing use of more effective materials1 speeding
up housing delivery1 enabling high standards of design quality and possibly helping to
reduce resource consumption in the U<. & are increasingly regarded as a means of
improving quality1 reducing time spent on site1 improving on,site safety and
overcoming s'ills shortages in the construction of housing. & and advanced
technologies are coming together to deliver higher performing1 more sustainable and
smarter buildings in order to achieve sustainable low energy houses1 as @ffsite !"":
showed in their last event of 4une !"":. *here are therefore prospects for $&(,(-.
integration with &.
2.# C2.647>i.( -+8,-C0
A review of the three core aspects of the research was presented herein. 3irst prospects
and diversity of solar water heater systems and collectors and more precisely of $&(,
(-. was given. (olar water heater considerations such as angles of inclination1
stratification1 draw off characteristics and performances of (-. were also discussed.
*he second aspect gave an account of different methods of modelling (-.. .eat
transfer mechanisms in (-. for thermal modelling presently of interest to this research
pro0ect were discussed. &3D modelling as a tool for optimisation of (-. designs was
also addressed. *he third reviewed the importance towards further stage integration of
(-. and the steps of a 2ife &ycle Assessment. A review of geographical and weather
conditions in (cotland was provided. 3inally1 a brief review of building integrated (-.
and & was also provided.
*o sum up1 all three steps e%plore above are deeply inter,related because each of them
are crucial for the complete development of this research on $&(,(-. systems.
=;
CHAPTER 3 LABORATOR AND FIELD E!PERIMENTS
E%perimental research is an essential element in validating the performance of any
proposed advancements of solar heating design technology particularly where
significant improvements have been identified the $&(,(-. prototype. *his chapter
has been divided into four parts. *he first part deals with the methodology adopted for
the e%periments by introducing the preliminary design of the solar collector followed by
the test regime1 measurement and data collection adopted for laboratory and field
e%periments. *he second section e%amines laboratory e%perimental set,up and results
and ma'es a comparison with previous wor' of 7rassie et al +!""B/. (ection three
describes the e%perimental set,up for the field e%periments and provides e%periments
results. 3inally1 the fourth section evaluates uncertainties and propagation of errors
through both e%periment set,ups.
3.1 M+:12>242(/
An appropriate methodology must be applied to generate confidence in e%perimental
design and results. *his wor' builds on earlier research by 7rassie et al +!""B/ which
was based on a stainless,steel $&(,(-. design and focussed around the use of &3D
modelling techniques. 2essons were learned from the methodology employed in their
test regimes and these have been significantly improved to enhance the research rigour.
*he 'nowledge so acquired also resulted in the use of a different material1 aluminium1
for the collector manufacture and optimisation of e%tended surfaces within the collector
design in order to enhance the collector performance.
3.1.1 S24,- A,:+- >+0i(. ,.> 62.0:-76:i2.
1.1.1.12 P-+4i8i.,-/ D+0i(.
*he design of the studied aluminium $&(,(-. is based on previous studies of uneer
et al +!""Ba/1 .enderson et al +!""B/ and finally the laboratory study of 7rassie et al
+!""B/ on a stainless,steel constructed $&(,(-. design.
S24,- A,:+- 1+,:+- >+0i(.
*he flat,plate type collector was designed1 from ;mm thic' aluminium sheets1 of
internal dimensions 9"""%9"""%="mm. -ater inlet and outlet pipes1 each of !!mm
internal diameter1 were provided on the top and bottom right side of the bo%. *hree
=>
pipes were provided to insert thermocouples in the $&(,(-. in order to determine the
internal temperature profile. *o protect from convection heat losses the water tan' was
insulated with a 9""mm thic' layer of fibreglass wool on the bottom and all sides and
placed inside a hard wooden bo%. A gap of ;=mm between the absorber plate and the
>mm thic' glaDing reduces heat loss by restricting air movement and gave an aspect
ratio of !C.=:. 3our fins were designed using aluminium sheets of C""%="%;mm) length1
height and thic'ness respectively. 3ins are a good means of improving performance in
an ine%pensive $&(,(-. and to achieve an increase in thermal and construction
efficiency of the system. *hese fins were designed to be fi%ed inside the water tan'
0oining the top and bottom surfaces and parallel to its sides1 leaving a distance of l""
mm from both the top and bottom ends. Basic design and composition of the (-. are
presented in 3igure ;,9.
Fi(. 3)1<* S+6:i2.,4 vi+A 23 :1+ ,478i.i78 ICS SWH 70+> i. :1+ 9-+0+.: A2-C
*he top surface of the collector was covered with an electrically heated silicone rubber
pad used to control power input during laboratory tests. $t was also painted blac' to
improve the solar radiation collection for the field e%periments.
3igures ;,!1 ;,; and ;,> show pre,assembly1 welding and post,welding1 respectively of
the collector during manufacture.
Fi(. 3)1$* ICS 8,i. 9,-:0 ,3:+- 67::i.(; 324>i.( ,.> >-i44i.( 29+-,:i2.0
==
-ooden container
-ater tan'
.eated surface J Absorber plate
7lass cover
Air cavity
3ins
7lass wool insulation
9!;= mm
9C" mm
Fi(. 3)2'* SWH 8,i. 9,-:0 5+i.( ,00+854+> ,.> A+4>+>
Fi(. 3)21* SWH 3i.0 ,00+854+> ,.> A+4>+>
C244+6:2- 8,:+-i,4
*he thermal conductivity of the collector material is an intrinsic parameter for effective
$&(,(-. designs. *he material must allow the absorption of solar radiation and heat
transfer into the storage fluid. Aluminium is very effective1 since its thermal
conductivity of !;: -Jm.< is higher than other materials traditionally used for $&(,
(-.s and was therefore chosen for the construction of the collector. Physical
properties of aluminium are given in *able ;.9 referring to $ncropera and De-itt +!""9/
data.
T,54+ 3)#* A478i.i78 ,.> S:,i.4+00)0:++4 91/0i6,4 9-29+-:i+0 ,: 3''K
Properties Aluminium (tainless,steel Units
=B
Oalues Oalues
Density !:"! :#"" 'gJm;
(pecific heat
#"; >::
4J'g.<
*hermal conductivity
!;: 9>.#
-Jm.<
An additional 'ey property of aluminium is its low density which is only one,third that
of stainless,steel1 thus ma'ing aluminium collectors lighter and safer to install on roof
structures. *his material is also easily formed1 machined and cast1 significantly
decreasing the final product price. Another main advantage is its high resistance to most
forms of corrosion. Aluminium produces a natural thin surface layer of aluminium
o%ide which provides a highly effective barrier to the corrosion caused by air1
temperature1 moisture and chemical attac'. 3inally1 almost every aluminium product
can be recycled at the end of its useful life1 without loss of metal quality or properties as
identified by the $nternational Aluminium $nstitute +!""C/.
1.1.1.13 C244+6:2- 8,.73,6:7-+ ) >i33+-+.6+0 Ai:1 9-+4i8i.,-/ >+0i(.
Discrepancies in the as built collector compared to the design drawings were observed.
&ritical internal dimensions were not achieved. $nitial critical internal dimensions were
fi%ed at ="%9"""%9"""mm providing an internal storage volume of =" litres. *he
manufactured collector achieved dimensions of ##9%#C#%>:mm giving a :.#L
difference in volume corresponding to a four litres loss. @ther divergences in the
positioning of the inlet pipe were found. 2aboratory tests where however not affected
by this change in inlet pipe position due to non,flow tests) however1 this modification
can have ma0or influence on a full flow field tests. A visualisation of the problem is
shown in 3igures ;,= and ;,B.
=:

Fi(. 3)22* W,:+- 342A i. 8,.73,6:7-+> 6244+6:2- B 52::28 -i(1: i.4+: 9i9+ (5/9,00 342A)
Fi(. 3)23* W,:+- 342A i. CAD >-,Ai.(0 6244+6:2- B 52::28 4+3: i.4+: 9i9+
*he construction errors have been addressed with the manufacturer in order to discuss
ways and methods to improve design criteria for future test units. *he main problems
encountered in manufacture were due to the material being prone to distortion during
the welding process using *$7 +*ungsten $nert 7as/. *he use of a $7 +etal $nert
7as/ welding process to reduce the heat input and reduce distortion will be considered
for the construction of future units. According to the manufacturer the dimensions
attained were realistic for a one,off prototype.
$nlet and outlet pipes of !!mm internal diameter were chosen as the nearest
commercially available products which can be tapped to suit a standard valve. Pipe
=C
diameters were changed to another standard 9;mm diameter by the manufacturer. *his
is important when considering the integration of $&(,(-. into buildings.
3.1.2 A00+008+.: ,.> 6,4i5-,:i2. 23 +x9+-i8+.: +=7i98+.:
A rigorous calibration of e%perimental devices assures that the system will produce
accurate and repeatable results with a specified degree of confidence.
1.1.1.1# T1+-8262794+0
<,type thermocouples using nic'el,chromium and nic'el,aluminium alloys to generate
voltage were used in the e%periments as they fulfilled the application temperature range
and were the most accessible type.
T1+-8262794+ C,4i5-,:i2.
*he calibration process consisted of a two point cold and hot test. *he equipment used
for the calibration and their respective specifications are summarised in *able ;.!
below.
T,54+ 3)"* E=7i98+.: ,.> S9+6i3i6,:i2.0
Nu)ber 8:ui0)ent &0ecification !ri.in
! Data logger type 93C
7rant $nstrument (quirrel !"!" series
*ype 93C
?esolution up to B significant digits
?esolutionF ".9&
7rant $nstrument
9 *hermos 3las'
$nsulated flas' for cold test. 3illed with
water Napier University
9 &onstant temperature bath
O!>"91 ;""-1 !""J!="O1 =""ml
a%imum temperatureF 9="S&
2iquidF -ater
Electrothermal
! -itness *hermometers
Accurate thermometers calibrated by
British (tandard. *wo range usedF ",9"S&
and =",#"S&
&alibrationF British
(tandard
!: *hermocouples
(ingle1 < type1 AccuracyF t".=&
*emperature rangeF ,!"" to 9;="& Napier University
9 (oftware
(quirrelOiew Plus1 !"!" logger set,up1
download and data e%port application for
-indows and e%port data direct to E%cel.
7rant $nstrument
9 &omputer
=#
C24> :+0:*
All thermocouples were placed in a thermos flas' filled with ice. *heir temperature
readings were recorded by the data logger after stabilisation of the temperature. A
reference British (tandard thermometer of range ",9"& was placed in the ice to act as a
witness measure. 3igure ;,: shows the installation.
Fi(. 3)2#* C24> :+0: i.0:,44,:i2.
H2: :+0:*
A constant temperature bath was used to provide uniform water temperature
distribution. *he bath was filled with water within which the thermocouples were
placed. *hermocouple temperature readings were recorded by the data logger.
?eference British (tandard thermometers of range =",#" & and :",9"" & were
positioned in the hot bath to calibrate the thermocouples. 3igure ;,C shows the
installation.
B"
R+3+-+.6+
T1+-828+:+-
T/9+ K)0i.(4+
T1+-8262794+0
D,:, L2((+-
C2897:+-
I.0:,44,:i2.
I6+
T1+-820
34,0C
Fi(. 3)2"* H2: :+0: i.0:,44,:i2.
O50+-v,:i2.0*
@ut of these tests corrective equations were developed by plotting reference British
(tandard thermometer readings against thermocouple measured temperatures. Each
thermocouple had an associated equation. &omparison between linear and polynomial
trends were underta'en and showed that a linear regression was sufficiently accurate to
be applied. 3or each new value read by the thermocouple1 the correct value was
determined by its equation.
Errors were quantified statistically by using the coefficient of determination1 ?
!
1 mean
bias error1 BE1 root mean square error1 ?(E1 as well as ('ewness and <urtosis of
the error histograms and showed acceptable margin of errors. *hese statistical factors
are further e%plained in &hapter >.
T1+-8262794+ 920i:i2.0 i. :1+ 6244+6:2-
*he calibrated thermocouples were used to monitor water temperature at different
positions in the collector. A good distribution of the thermocouples in an inclined $&(,
(-. is essential to determine the temperature profile inside the collector as
stratification causes a variable temperature distribution inside the water tan'. Past
studies on $&(,(-. by 7rassie et al +!""B/ suggested a positioning scheme of
thermocouples in $&(,(-.. *emperature profiles observed by 7rassie et al +!""B/
showed that further thermocouples should be added at the collector5s e%tremities to
evaluate accurately phenomenon occurring in these Dones. A new thermocouple
distribution was therefore developed. *hermocouples were doubled at the top and
bottom part of the middle acrylic clear tube. Both distributions are represented in 3igure
;,# below.
B9
R+3+-+.6+
T1+-828+:+-
T/9+ K)0i.(4+
T1+-8262794+0
D,:, L2((+-0
H2: A,:+-
C2.0:,.:
:+89+-,:7-+
5,:1
Fi(. 3)2%* T1+-8262794+ 920i:i2.0 ,.> v2478+ 62.:-24 ,0026i,:+>; ,* A478i.i78 6244+6:2-; 5*
S:,i.4+00)0:++4 C244+6:2- (H7.,i>i +: ,4 (2''")).
&ontrol volumes1 represented by dashed lines1 were assigned to each of the
thermocouples in order to effectively evaluate an average temperature for the water
tan'. *he temperature profile in the (-. could then be developed using monitored
temperature data.
T1+-8262794+D0 i.0+-:0 ,.> 920i:i2.0
&lear acrylic tube inserts were used in order to provide support for the thermocouples.
*he high resistance to heat distortion and simplicity to insert thermocouples along
acrylic clear tubes made it the best option for this application. *hermocouples were
running through the acrylic plastic tube inserts with multiple slots to e%pose the
thermocouple tips to the water as shown in 3igure ;,9". *ips were fi%ed in the slot by
silicone.
B!
Fi(. 3)2&* A6-/4i6 94,0:i6 :75+0 i.0+-:0 Ai:1 :1+-8262794+ 042:
Acrylic plastic tubes were placed in the collector through three pipes previously
integrated in the collector design as shown in 3igure ;,99.
Fi(. 3)2<* A6-/4i6 94,0:i6 :75+ i.0+-:0 i. 9i9+0
P20i:i2.i.( 23 :1+-8262794+ :i90 i. :1+ 6244+6:2-
*he positioning of thermocouple tips might affect the readings if the thermocouples
appear to be in the flow regime as shown in 3igure ;,9!. Acrylic tube inserts were
turned to ensure that the tips of the thermocouples were positioned down the central
plane. *his was identified as a potential error score when performing a critical review of
previous test regimes employed in $&(,(-. development.
B;
*hermocouple tip
Fi(. 3)2$* A00789:i2. 23 :1+-8262794+ 920i:i2. i. :1+ SWH
1.1.1.1" T1+ +4+6:-i6,44/ 1+,:+> 0i4i62.+ -755+- 9,>
An electrically heated silicone rubber pad was used to control power input simulating
the distribution of solar radiation on the top surface of the collector during laboratory
tests.
E.+-(/ >i0:-i57:i2. 23 :1+ +4+6:-i6,44/ 1+,:+> 0i4i62.+ -755+- 9,>
*he heat distribution was assumed to be uniform on the collector surface1 however1 due
to the seasonal variance in sun elevation and positioning1 the distribution of solar
radiation on the collector surface is non,uniform. All heat losses due to conduction were
ignored during laboratory e%periments. Energy distribution of the electrically heated
silicone rubber pad is represented in 3igure ;,9;.
Fi(. 3)3'* E.+-(/ >i0:-i57:i2. 23 :1+ +4+6:-i6,44/ 1+,:+> 0i4i62.+ -755+- 9,>
B>
O9:i8i0i.( :1+ 62.:,6: 5+:A++. :1+ +4+6:-i6,44/ 1+,:+> 0i4i62.+ -755+- 9,> ,.>
6244+6:2- 07-3,6+
Aluminium has good heat absorption properties as well as heat transfer properties.
.owever1 the contact between the electrically heated silicone rubber pad and the top
surface of the collector should have minimum air gaps in order to ma%imise heat
transfer. $n order to compare with the methodology used by 7rassie et al +!""B/1 tests
reproducing the same conditions were conducted using silicone nodes to fi% the
electrically heated silicone rubber pad to the collector5s top surface. 3igures ;,9> and ;,
9; show the surface of the pad after the e%periment.
Fi(. 3)31* A502-5+- 01++: 23 :1+ ICS)SWH
Fi(. 3)32* Si4i62. 9,0:+
*he adherence between the electrically heated silicone rubber pad and the collector5s
surface was not optimal and therefore affected the heat transferred to the water.
$nfra?ed *hermographic images of the pad were ta'en before removing the silicone
node using a 32$? *herma&A thermal camera. (ilicone and non,silicone surfaces
were pictured in order to evaluate the effect of silicone on heat distribution and are
represented in 3igures ;,9Ba and b. -hile a uniform flu% through the heating pad was
B=
(ilicone
Electrically heated
silicone rubber pad
(ilicone nodes
e%pected on both sides1 colder spots observed at various positions on the silicone side of
the pad indicated an uneven heating pattern and uniform heat distribution to the non,
silicone side of the pad.
, 5

,
Fi(. 3)33* H+,:i.( 9,> :1+-82(-,91i6 i8,(+0* ,. Si4i62.+ 0i>+ (1+,:i.( 0i>+); 5. F-2.: 0i>+; S6,4+*
C
3igure ;,9:a and b showed a more detailed picture of the uneven heating and cooling
pattern of the silicone sided pad. *he silicone retains heat when cooled and acts as
insulation when heated.
Fi(. 3)3#* H+,:i.( 9,> :1+-82(-,91i6 i8,(+0* H2: 092:0 (,) ,.> 6224 092:0 (5) -+9-+0+.: :1+ 0i4i62.+
A+4>0. S6,4+* C
BB
, 5
, 5
$n order to improve the initial heating methodology the silicone nodes were removed
and replaced by heat sin' compound paste) improving the overall heat transfer between
the two surfaces by two means.
3irst1 a uniform layer of the compound could be applied over the electrically heated
silicone rubber pad attached to the collector top surface1 thus providing a more uniform
heat distribution and decreasing air interstices between the two surfaces. A second 'ey
factor was the higher thermal conductivity of the heat sin' compound of ".# -Jm.<
compared to the silicone sealant +".!-Jm.</ resulting in a ::L improvement in heat
transferred to the collector.
*he electrically heated silicone rubber pad can withstand e%posure temperature of up to
!;"& and has a high emissivity of value ".CB. *he fibreglass reinforced silicone rubber
layers sandwich a pre,formed heating element made from fine resistance wires. *he
silicone rubber resistant heating pad dimensions were 9""" #B"mm.
1.1.1.1% D,:, L2((+-0
7rant $nstrument (quirrel !"!" series1 type 93C1 data loggers were used for recording
temperature. $n the presence of a temperature difference between the 0unctions of the <
type thermocouple a small temperature dependent current flows around the circuit. *he
measured voltage is converted to a temperature reading through the use of the digital
data loggers using a polynomial relationship for current conversion to temperature
defined by the British (tandards +B( EN B"=C>.9/. *he 9! Bit data loggers have a
sensitivity specification of ".9&.
3.1.2.1 P/-,.28+:+-
A <ipp and Eonen B3; pyranometer was used during the field e%periment to measure
hourly horiDontal global solar radiation in -Jm
!
1 received by the $&(,(-.. $t consists
of an array of photodiodes with a computer,generated shading pattern to measure
incident solar radiation mounted on a base covered by a hemispherical glass dome.
$ncoming solar radiation lights the photodiodes converting it to a voltage proportional to
the irradiance on that surface which is then read by the computer. Based on &hapter !1 a
>= inclination of the $&(,(-. was used for field tests. *he Msolar radiation and
daylight odelN developed by &lar'e and uneer +!""C/ was used to get ad0usted solar
radiation for the required inclination.
B:
3.1.3 Ex9+-i8+.:,4 62.0i>+-,:i2.0
*he aim of this section is the identification of the errors due to the assumptions and test
regime that have been made during the e%perimentation phase.
H+,:+> 07-3,6+
*he heat distribution of the electrically heated silicone rubber pad was assumed to be
uniform on the top surface of the heater. All heat losses due to conduction were ignored
during the laboratory e%perimental tests1 however the effects of the heat losses from the
electrically heated silicone rubber pad to the ambient air1 the thermal capacitance of the
water storage unit1 the glass reflectivity and all other associated losses were ta'en into
account.
L+,C,(+ ,.> 9-+007-+ i.0i>+ :1+ 6244+6:2-
*he e%tremities of the pipes from which the acrylic plastic tubes with thermocouples
were inserted were sealed with silicone to avoid any lea's. *his might have caused
minor problems due to the build up of pressure inside the collector and thus e%pansion
of the collector andJor drive the water at the top of the pipes. An e%pansion water tan'
was installed on top of the inlet pipe to overcome such problems.
D+9,-:7-+ :+89+-,:7-+
*ests were carried out for different controlled heat flu%es provided by the electrically
heated silicone rubber pad. $n order to achieve comparable results starting water
temperatures had to be the same for each e%periment. *herefore1 water in the collector
was purged and replaced by fresh water at appro%imately 9>& after each e%periment.
Wi.> >,:,
-ind speeds were recorded hourly by an anemometer for field tests) however1 only an
average wind speed for the location was utilised in calculations as directions were not
recorded. *he location of the $&(,(-. was partly sheltered from wind1 therefore
decreasing its effect on the glass cover.
BC
3.2 L,52-,:2-/ +x9+-i8+.:0
3.2.1 Ex9+-i8+.:,4 0+: 79
*he equipment set up for the laboratory tests attempted to replicate the previous
e%perimental schemes of 7rassie et al +!""B/ in order to compare their respective
performance and is described in this section. A total of !! thermocouples were used to
measure water temperature inside the collector as shown previously in 3igure ;,99.
Ambient temperatures inside the laboratory were recorded by two thermocouples while
three others were placed on the glass cover. *he electrically heated silicone rubber pad
was designed to simulate a constant and uniform heat flu% over the flat plate collector
reproducing Edinburgh solar conditions. A variac transformer controlled the power
supply to obtain a measured amount of heat. *ests were carried out for heat flu%es from
9"" to >"" -Jm
!
with increasing intervals of =" -Jm
!
at different angles including " A
;" A ;= A >= A B". *emperatures were recorded at 9" minute intervals over !> hours1
giving time to reach a state of quasi,equilibrium. *he test rig is shown in 3igure ;,9C
with the circuit diagram of the electrical equipment for control and measurement of the
heat flu% in 3igure ;,9#.
Fi(. 3)3"* T+0: -i( Ai:1 62.:-24 ,.> 8+,07-+8+.: +=7i98+.:
B#
Fi(. 3)3%* Ci-67i: >i,(-,8 32- 62.:-24 ,.> 8+,07-+8+.: 23 :1+ 1+,: 347x
3.2.2 Ex9+-i8+.: -+074:0
1.1.1.1& G+.+-,4 3i.>i.(0
*he e%perimental data obtained replicated and validated 7rassie et al5s +!""B/ wor' on
stainless,steel $&(,(-.. Equilibrium was e%perienced after a period of !> hours when
the pea' value of mean water temperature was achieved and all the input heat flu%
balanced with net losses. At this stage the water tan' temperatures and collector
efficiency tended to a constant value resulting in an asymptotic shape in their profiles.
Every heat flu% value has an associated water temperature when equilibrium is reached.
*he higher the heat flu% imposed1 the higher the temperature reached by the collector
but the lower is the overall efficiency as this is the ratio of power output to power input.
$n accordance with the results of 7rassie et al +!""B/1 an increase of the temperature
stratification with time was observed in the $&(,(-.. *he system behaviour also
varied with the angle of inclination. 7rassie et al +!""B/ showed that system
performance differs at various inclination angles and showed that thermal optima
increased by increasing angle of inclination. .owever1 it was suggested that an angle of
>= gave the best overall results when analysing temperatures reached over time1 thus
achieving high efficiencies. ?esults of the research confirmed that the suggestion of an
inclination angle of >= for the $&(,(-. position was appropriate for Edinburgh.
3urther e%periments at this inclination showed that the $&(,(-. can wor' successfully
in a (cottish climate.
1.1.1.1< A478i.i78 ICS)SWH 0:-,:i3i6,:i2.
(tratification in a water tan' is an essential factor for good performance of an $&(,
(-.. $t refers to the degree of temperature difference between the top and the bottom
:"
2oad J Electrically
heated silicone
rubber pad
-attmeter
Ammeter
Ooltmeter
Oariac
portions of the collector. $n order to determine dimensionless stratification1
e%perimental temperature profiles were scaled and hence the ratio *
h
J *
bottom
was
introduced where *
h
is the local and *
bottom
is the minimum recorded temperature in the
water tan'. (tratification profiles for different hours are plotted for a heat flu% of
!""-Jm
!
in 3igure ;,!". *he overall collector temperature increases with time until
reaching steady,state. *he first three hours show a rapid increase in temperature
stratification1 decreasing later with time until reaching equilibrium.
0.38
1.00
1.0$
1.0%
1.0'
1.08
1.10
1.1$
1.1%
0.0 0.1 0.$ 0.3 0.% 0.& 0.' 0.(
3&in1 )
'
(
"
'
b
o
t
t
o
)
1hr $hr
3hr %hr
&hr 'hr
(hr 8hr
10hr 1'hr
Fi(. 3)3&* T+89+-,:7-+ 0:-,:i3i6,:i2. 9-23i4+ 32- , 2''W?8
2
1+,: 347x
-ater at a given temperature will settle down at an appropriate height in the $&(,(-.
in accordance with the prevalent density of the fluid. .ot water of low density will
naturally sit in the upper layers while cold water of high density will drop down to the
bottom layers. During the first three hours of the charging process1 high temperature
stratification occurs in the $&(,(-.. -ith time1 the temperature of the upper layers is
established and reaches an equilibrium which results in the lower layer achieving a
similar temperature. *his therefore decreases the density gradient inside the $&(,(-..
A temperature stratification profile is thus built up.
*his variation in density results in differential buoyancy forces in the fluid causing
convection by gravity. Pressures increases with depth1 therefore the pressure within the
collector will e%perience different pressures at the top and bottom1 with the pressure on
the bottom being higher. *his difference in pressure causes an upward buoyancy force.
:9
*he buoyancy force1
B

1 is thus the difference between the forces at the top and

bottom in the collector for one square meter and can be e%pressed as belowF
+
h !
t b B
/ +
+;.9/
-here
b

is the density of the fluid at the bottom of the collector +'gJm

;
/
t

is the density of the fluid at the top of the collector +'gJm

;
/
+
h is the vertical length from top to bottom of the collector +m/
!
is the standard gravity of #.C9NJ'g
?esulting buoyancy forces within the collector were then calculated. During the
charging period forces increase within the cavity from ;.!N to C.;N generating
buoyancy driven flows within the cavity. *his resulted in an increase in thermal
stratification within the cavity from !.B& to ;.B&. -ith time the temperature
difference between top and bottom decreases. *his reduction can be associated with a
decrease in buoyancy forces operating within the cavity therefore reducing fluid motion
and natural convection occurring in the collector. After 9Bhrs1 the temperature gradient
decreased to 9.#& associated with a buoyancy force of >.BN. *hese results are
presented in 3igure ;,!9.
0.0
1.0
$.0
3.0
%.0
&.0
'.0
(.0
8.0
3.0
1 $ 3 % & ' ( 8 3 10 11 1$ 13 1% 1& 1'
'i)e1 (rs
;
u
o
-
a
n
c
-

f
o
r
c
e
1

N
0.0
0.&
1.0
1.&
$.0
$.&
3.0
3.&
%.0
'
e
)
0
e
r
a
t
u
r
e

.
r
a
%
i
e
n
t
1

<
C
Buoyan+y for+e
Teperature gradient
Fi(. 3)3<* B72/,.6/ 32-6+0 Ai:1 :i8+
*he stratification profile was plotted for an average day length of eight hours of $&(,
(-. operation for different wattages. @ne such profile is shown in 3igure ;,!!. A
definite relationship between the temperature at any point of the collector and the
:!
applied heat flu% was observed. *he higher the heat flu% imposed the higher the water
temperature ratio *
h
J *
bottom
.
1.00
1.0$
1.0%
1.0'
1.08
1.10
1.1$
1.1%
0.0 0.1 0.$ 0.3 0.% 0.& 0.' 0.(
3&in1 )
'
(
"
'
b
o
t
t
o
)
100#
$00#
%00#
Fi(. 3)3$* T+89+-,:7-+ 0:-,:i3i6,:i2. 9-23i4+ ,3:+- < 127-0 23 29+-,:i2.
*emperatures are thus dependent on the longitudinal position within the collector1
applied heat flu%1 and time. A polynomial function for the stratification profile for each
hour was then developed ta'ing heat flu% and longitudinal location in the collector as
the main parameters.
3.2.3 C289,-i02. 23 9,0: ,.> 67--+.: -+0+,-61
A comparative analysis between the stainless,steel $&(,(-.s previously developed by
7rassie et al +!""B/ alongside the newly developed aluminium $&(,(-. was
performed in order to validate e%periments and to show the benefit of the new $&(,
(-.. *he methodology adopted in both studies were first compared followed by
comparisons of e%perimental results.
1.1.1.1$ T+0: 0+:79 6289,-i02.
ethodologies adopted in the research were compared with those used by 7rassie et al
+!""B/ to give a viable comparison. *he newer collectors were designed and
manufactured with ;mm thic' aluminium sheets as compared to the 9.=mm thic'
stainless,steel ;">,!B $&(,(-. of the same overall dimensions. 7rassie et al +!""B/
developed two $&(,(-.s) one with and one without fins.
:;
*wenty,two <,type thermocouples were calibrated and used to monitor the temperature
profile inside the aluminium collector compared to twelve thermocouples used in the
stainless,steel collectors. $nformation on the calibration of the thermocouples was not
available for the stainless,steel $&(,(-. e%periments.
$n order to demonstrate the influence of the number of thermocouples on the accuracy
of results1 a first set of calculations using 7rassie et al5s +!""B/ thermocouple positions
for the aluminium collector was implemented. &alculations were then repeated but
ta'ing into account the !! thermocouples. $t was established that the number of
thermocouples did not influence the overall efficiency results.
*he second comparison consisted of the methodology for calculating efficiencies. *he
heat capacity +&
p
/ and density +/ of the water were assumed to be constant by 7rassie
et al +!""B/ contrary to the more accurate method adopted in this research where &
p
and
were calculated as a function of the water temperature in the collector. An accuracy
difference of 9L on the overall efficiencies was found by comparing the two methods.
3inally1 materials used to attach the electrically heated silicone rubber pad to the
collector top surface were different. (ilicone was used by 7rassie et al +!""B/ rather
than a heat sin' compound. *wo tests were carried out using both materials in order to
compare the conductivity of the two materials resulting in a efficiency difference of 9L
between the two methods. @ther parameters which might have affected the accuracy of
the results were the calibration of the thermocouples and data logging equipment.
Based on this methodology the thermal performances of the two finned $&(,(-.s and
benefits of using fins were assessed1 compared and ad0usted. ?esults are reported in the
ne%t section.
1.1.1.2' S:-,:i3i6,:i2. ,.,4/0i0
*he first comparison consisted of a stratification analysis. A dimensionless study of
water temperature according to the distance from bottom to top portions of the collector
was developed with the aim of comparing the thermal stratification of both $&(,(-.s.
Dimensionless temperature of ratio M
H h
T T J
N was plotted in 3igure ;,!; against the
dimensionless distance ratio M H h J N where
h
T
is the water temperature at the local
:>
point +</1
H
T
is the water temperature at the top of the collector +</1 h is the local
length in the collector +m/1 H is the total length of the collector +m/.

3igure ;,!; a and b confirm findings from section ;.!.!.! for the aluminium material
only. $t validates that the higher the heat flu% applied the higher the stratification in the
$&(,(-. and that the longer the e%posure time is the lower is the stratification due to
the buoyancy. *hese findings are consistent with those of 7rassie et al +!""B/.

0.8%
0.8'
0.88
0.30
0.3$
0.3%
0.3'
0.38
1.00
0 0.$ 0.% 0.' 0.8 1
6i)ensionless %istance
6
i
)
e
n
s
i
o
n
l
e
s
s

t
e
)
0
e
r
a
t
u
r
e
-tainless4steel 41hr
-tainless4steel 48hr
.luiniu 41hr
.luiniu 48hr

0.('
0.(8
0.80
0.8$
0.8%
0.8'
0.88
0.30
0.3$
0.3%
0.3'
0.38
1.00
0 0.$ 0.% 0.' 0.8 1
6i)ensionless %istance
6
i
)
e
n
s
i
o
n
l
e
s
s

t
e
)
0
e
r
a
t
u
r
e
-tainless4steel 41hr
-tainless4steel 48hr
.luiniu 41hr
.luiniu 48hr
Fi(. 3)#'* Di8+.0i2.4+00 :+89+-,:7-+ ITh/THJ v0 >i8+.0i2.4+00 >i0:,.6+ Ih/HJ ,: #" i.64i.,:i2.; ,*
1''W?8
2
1+,: 347x i.97:; 5* #''W?8
2
1+,: 347x i.97:
Another important finding from this figure is the effect of the collector material on
stratification. (teeper stratification curves were found in the stainless,steel $&(,(-.
:=
,
5
compared with the aluminium profiles demonstrating a better mi%ing in the latest $&(,
(-.. *his decrease in stratification is e%plained by the higher conductivity of
aluminium material allowing a more rapid transfer of heat from the top to bottom water
layers. 3igure ;.,!> support this finding by depicting the rise in temperature registered
by the thermocouples placed at the top and bottom in the stainless,steel and aluminium
$&(,(-..
10
$0
30
%0
&0
'0
(0
80
0 1 $ 3 % & ' ( 8 3 10 11 1$ 13 1% 1& 1' 1( 18 13 $0 $1 $$
'i)e1 (r
'
e
)
0
e
r
a
t
u
r
e
1

<
C
Fi(. 3)#1* W,:+- :+89+-,:7-+ 9-23i4+0 23 3i..+> ,478i.i78 ,.> 0:,i.4+00)0:++4 ICS)SWH0 ,: #''W;
#" i.64i.,:i2..

.igher water stratification was achieved by the stainless,steel $&(,(-. compared to
aluminium with #."S& and >.=S& difference respectively from the bottom to top of the
collector after an e%posure time of eight hours. *he aluminium temperature profile
clearly shows that the heat transferred to the water was improved and generated higher
overall water temperatures.
*hese results indicate that stratification depends upon the incident solar radiation +heat
flu% applied to the collector surface/1 the e%posure time and collector material for a
given angle of inclination of the collector.
1.1.1.21 B+.+3i:0 23 i.:-2>76:i2. 23 3i.0 i. :1+ 6244+6:2- >+0i(.
$n order to quantify the benefit of integrating fins in the collector1 a comparison of the
actual rise in temperature of the finned aluminium1 finned and un,finned stainless,steel
:B
A478i.i78 6244+6:2-
52::28 +.>
A478i.i78 6244+6:2-
:29 +.>
S:,i.4+00)0:++4 6244+6:2-
52::28 +.>
S:,i.4+00)0:++4 6244+6:2-
:29 +.>
$&(,(-.s was used. 3igure ;,!= compares the results obtained from preliminary
e%periments where ini w w m
T T T
1

. ini w
T
1 is the initial mean water temperature and
w
T
is the mean water temperature at any time t.
A rise in temperature of 9#.!&1 9B.C& and 9;.C& was observed after eight hours of
e%posure time at !""-Jm
!
for the finned aluminium1 finned stainless,steel and un,
finned stainless steel $&(,(-.s respectively. *his corresponded to an improvement of
9CL by using fins in the stainless,steel $&(,(-.1 and an additional 9;L improvement
using the more conductive aluminium.
0
$
%
'
8
10
1$
1%
1'
18
$0
0 1 $ 3 % & ' ( 8
'i)e. (rs

'
)

1

<
C
.luiniu6,inned
-tainless4steel6,inned
-tainless4steel6"nfinned
Fi(. 3)#2* W,:+- :+89+-,:7-+ 9-23i4+0 23 3i..+> ,478i.i78; 3i..+> ,.> 7.)3i..+> 0:,i.4+00)0:++4
ICS)SWH0 32- +i(1: 127-0 23 +x9207-+ :i8+ ,: 2''W.
$n order to support the results identified above the overall water temperature of the un,
finned stainless,steel collector was plotted as a function of the finned stainless,steel
$&(,(-. in 3igure ;,!B.1 and as a function of the finned aluminium $&(,(-. in
3igure ;,!:. *hese scatter plots have the very specific purpose of demonstrating how
much the MfinN affects water temperature achieved in the collector. $t is apparent from
3igure ;,!B and ;,!: that the finned heaters outperform the un,finned by achieving a
higher temperature rise
m
T
illustrated in 3igure ;,!B and ;,!: by the deviation from
the Mone to oneN straight line relationship. 3igure ;,!B and ;,!: also shows that over
time1 data points converge) approaching to the Mone to oneN straight line relationship.
7radually1 un,finned and finned water temperature in the collector tends to reach the
same value) approaching an equilibrium temperature observed for the 9""-Jm
!
heat
flu% profile.
::
0
10
$0
30
%0
&0
0 10 $0 30 %0 &0
Finne% stainless=steel '
)
1 <C
5
n
f
i
n
n
e
%

s
t
a
i
n
l
e
s
s
=
s
t
e
e
l

'
)
1

<
C
%00#
$00#
100#
Fi(. 3)#3* ICS)SWH A,:+- :+89+-,:7-+ i.6-+,0+* 7.)3i..+> v+-070 3i..+> 0:,i.4+00)0:++4
0
10
$0
30
%0
&0
0 10 $0 30 %0 &0
Finne% alu)iniu) '
)
1 <C
5
n
f
i
n
n
e
%

s
t
a
i
n
l
e
s
s
=
s
t
e
e
l

'
)
1

<
C
%00#
$00#
100#
Fi(. 3)##* ICS)SWH A,:+- :+89+-,:7-+ i.6-+,0+* 7.)3i..+> 0:,i.4+00)0:++4 v+-070 3i..+> ,478i.i78
:C
*hese findings suggest that incorporating fins in the collector results in an increase in
heat transfer1 and that using aluminium generates higher overall water temperatures as
shown in 3igure ;,!B and ;,!:. $n a real world application water will be regularly
drawn from the $&(,(-.1 not allowing enough time for the water to achieve
equilibrium temperature1 early differences in temperature rise is critical for the good
performance of the $&(,(-..
1.1.1.22 E33i6i+.6/ i89-2v+8+.:
*a'ing into account all the parameters discussed in section ;.!.;.91 efficiencies were
calculated by the following equation and then corrected. *he collector efficiency is the
ratio of the bul' heat collected to the heat available as shown in equation ;.!.
&
"
c
+;.!/
-here
c
"
is the heat collected at each control volume +-Jm
!
/
& is total irradiation from the heating pad +-Jm
!
/
Also e%pressed as
j j
j i
i
i i p
j
P
T V C
wi

.
. N . .O
;
22
1
-ith
/ + F
9 +
>
6 6 6
t t
+;.;/
@verall efficiencyF
j j
n
j
j i
i
i i p
n
j
P
T V C
wi

.
. N . .O
1
;
22
1 1
+;.>/
-here
i
)
is the control volume associated of the thermocouple MiN +m
;
/1 6
+
is
the heat flu% applied to the system +-/1
wi
p
C
is the specific heat capacity of water
associated with thermocouple temperature +4J'g.</1
i

is the water density associated
with the thermocouple temperature +'gJm
;
/1 6 i
T
1

is the temperature difference

between tG0 and tG"s +</ and 6

is the time step +s/.
:#
*o assess the improvement in efficiency of the new aluminium $&(,(-. results of the
above calculations are shown in 3igure ;,!C.
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
'i)e1 (rs
8
f
f
i
c
i
e
n
c
-
.l4100#
-t4100#
.l4%00#
-t4%00#
Fi(. 3)#"* A478i.i78 ,.> 0:,i.4+00)0:++4 ICS)SWH +33i6i+.6/ 6289,-i02. ,: #" >+(-++ i.64i.,:i2.;
1''W?#''W
3igure ;,!C illustrates two phases in the efficiency curves. Phase one shows a fast
increase in efficiency followed by a regular decrease in phase two. *his trend is due to
the thermal mass of the system. $t corresponds to the heat stored in the collector material
and the water for a given power applied to the system. 3igure ;,!C shows that a longer
period was needed for lower heat flu%es as it ta'es longer for the material to charge. *he
charging stage1 associated with the first phase is then dependent on the heat flu%
applied. *he same figure outlines that the thermal mass of the system is strongly
dependent on the type of material used for the collector. *he aluminium collector has a
heat capacity of 9:.! '4J< compared to 9;.;'4J< for the stainless,steel. *he !;L
difference e%plains the fact that the aluminium collector needs more time to charge for
lower heat flu% and vice,versa.
@verall efficiencies obtained from the preliminary analysis for aluminium and stainless,
steel $&(,(-. are presented in *able ;.;. *otal efficiencies vary according to the heat
flu% in a similar way to that described by 7rassie et al +!""B/. *he higher the heat flu%
applied to the system1 the lower the overall efficiency.
C"
1 2
T,54+ 3)%* S:,i.4+00)0:++4 v0 ,478i.i78 2v+-,44 +33i6i+.6i+0 32- >i33+-+.: 1+,: 347x* N2:+ 4,52-,:2-/
:+0:0
C244+6:2- :/9+ H+,: 347x ,994i+> Av+-,(+ 2v+-,44
+33i6i+.6/
1''W 2''W 3''W #''W
A478i.i78 ::L ::L BBL B;L :9L
S:,i.4+00)0:++4 BCL BCL B;L B!L B=L
$t was also identified that global efficiencies for the aluminium collector were higher at
all heat input regimes compared to stainless,steel. A drop in efficiency was noted
between !""-Jm
!
and ;""-Jm
!
. *ests carried out for !="-Jm
!
showed an
intermediate value of efficiency.
3.2.# Di06700i2. 23 4,52-,:2-/ :+0:i.(
*he purpose of this section was to determine whether the performances of an $&(,(-.
using aluminium as a material were improved by comparison with a previous study by
7rassie et al +!""B/ utilising a stainless,steel construction device.
*his chapter sets out ad0ustments in the methodology adopted by 7rassie et al +!""B/ in
order to ma'e an appropriate comparison between the two $&(,(-.s. @bservations
showed that a generally more accurate and rigorous methodology was adopted in the
e%periment of the aluminium $&(,(-..
*he e%perimental results clearly demonstrated that using aluminium generated higher
overall collector water temperatures and improved heat transfer to the water. *he
stratification analysis indicated that this depends upon the incident solar radiation1 the
e%posure time and collector material1 given a constant angle of inclination of the $&(,
(-..
3indings suggested that incorporating fins in the collector resulted in an increase in heat
transfer of 9CL in the stainless,steel $&(,(-.. Using a more conductive material such
as aluminium made an improvement of 9;L in water temperature. An overall increase
in the aluminium $&(,(-. efficiency of BL was found by comparing performances of
both $&(,(-. over an eight hour time period.
*he benefit of selecting aluminium for the $&(,(-. was clearly supported by the
current findings showing it out,performing the stainless,steel $&(,(-.. Although the
C9
collector was tested in a laboratory with calibrated instruments1 the actual performance
of the $&(,(-. in real weather conditions would yield important results on assessing
the viability of the aluminium $&(,(-.. *here was therefore a strong imperative for
conducting field tests with the aluminium $&(,(-..
3.3 Fi+4> +x9+-i8+.:0
*o investigate the influence of (cottish weather conditions on the performance of the
aluminium $&(,(-.1 e%tensive field measurements were conducted in Edinburgh
during the summer of !"":.
3.3.1 Ex9+-i8+.:,4 :+0: 0+: 79
*he test rig was located on the roof of Napier University at 9=.=m above ground,level
or 99;m above sea,level. Due to technical constraints the $&(,(-. was placed facing
9>; degrees from north. *he electrically heated silicone rubber pad was replaced by a
blac' paint applied to the absorber plate to improve the solar radiation collection. 3igure
;,!# shows the cross,sectional representation of the field tested $&(, (-. while an
illustration of the adopted tests rig is shown in 3igure ;,;"
Fi(. 3)#%* C-200)0+6:i2.,4 -+9-+0+.:,:i2. 23 :1+ 3i+4> :+0:+> ICS)SWH
C!
Fi(. 3)#&* Ex9+-i8+.:,4 :+0: -i(
A simulated roof,structure was used to insulate the sides and bac' of the $&(,(-.. $t
protected the $&(,(-. from the effects of weather by creating a micro,environment
and by decreasing the heat losses from the bac' of the $&(,(-.. 7in!span insulation
closed,cell polyurethane material was used partially to insulate the bottom of the $&(,
(-. and the bac' side of the simulated roof,structure as shown in 3igure ;,;9.
Fi(. 3)#<* B,6C i.074,:i2. 23 -223 0i874,:+> 52x
C;
(imulated
M?oof structureN
&old water tan'
ineral fibre wool was then placed inside the simulated roof,structure and rigid
insulation was siDed and placed on the sides. -ood was then fi%ed on top of the
insulation as well as a flush system on top of the collector to avoid any water ingress.
Pipes were insulated in order to avoid possible freeDing during cold days. 3igure ;,;!
shows the final installation.
Fi(. 3)#$* Ex9+-i8+.:,4 :+0: -i( 0i>+ i.074,:i2.
*he same !! <,type thermocouples used for laboratory e%periments were connected to
a data,acquisition system to monitor the water temperature inside the aluminium
collector covering middle longitudinal and lateral lines. @ther thermocouples were used
to record the absorber plate1 the glass cover1 the ambient1 and the inlet and outlet water
temperatures. A calibrated <ipp and Eonen B3; pyranometer was used to record solar
radiations at the location of the $&(,(-. while an average wind speed of ;.: mJs was
recorded ne%t to the location of the $&(,(-. using a -(;"! Oaisala anemometer.
*his average data was ta'en as the reference wind speed in the study. A cold water tan'
was placed above the $&(,(-. to feed the collector with fresh cold water and a valve
was used when draw,off of water occurred. A timer was placed to simulate water draw,
off profiles. *he angle of inclination of the $&(,(-. was 'ept at >=1 hourly radiation
and ten minutes interval temperatures were recorded for an initial period of three
months.
C>
3.3.2 Ex9+-i8+.: -+074:0
1.1.1.23 Wi:127: D-,A)O33
3ield tests started on B
th
4une !"": for an initial period of three months. $n order to
identify the performance of the $&(,(-.1 data management and analysis were
performed for a typical month of 4uly. ?ecorded data during the period of the 99
th
,!B
th

of 4uly are represented in 3igure ;,;; where *w is the bul' water temperature1 *a is the
ambient temperature1 *p is the plate temperature and *c is the cover temperature.
0
10
20
30
40
50
60
70
80
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
6ates
'
e
)
0
e
r
a
t
u
r
e
1
<
C
0
100
200
300
400
500
600
700
800
900
1000
&
o
l
a
r

r
a
%
i
a
t
i
o
n
1

$
"
)
2
'w
'a
'0
'c
lobal solar ra%iation
Fi(. 3)"'* Fi+4> +x9+-i8+.: >,:,
Each day is characterised by a heating and a cooling profile which are dictated by solar
radiation. *he higher the incident solar radiation received by the absorber plate1 the
higher the water temperature achieved. *he decrease in incident solar radiation results
in heat losses from the collector to the surroundings leading to a subsequent decrease in
water temperature. *he gradual heating and cooling profiles seen are due to the thermal
mass of the system previously introduced during the efficiency study of laboratory
e%periments. *he thermal mass is the capacity of a material to store heat. $t plays an
important role in the effectiveness of the heating charge and cooling discharge
operations of the collector and the amount of the stored thermal energy. *he water
stored in the collector accounts for most of the thermal mass of the system. *he high
specific heat >.9C'4J'g.< and moderate density of 9""" 'gJm
;
of the water results in a
C=
relatively high volumetric capacity of >9C" '4Jm
;
.<. *his shows the ability of a given
water volume to store internal energy while undergoing a given temperature change. $ts
low thermal conductivity of ".=C-Jm.< also contributes to this effect by slowly
storing or releasing heat. *he collector material also contributes to the thermal mass
through its moderate volumetric capacity of !>!> '4Jm
;
.<1 assuming the density of
aluminium at !::" 'gJm
;
and a specific heat of C:=4J'g.<. *his results in a thermal
mass of !!>.: '4J< for the collector.
A more detailed picture of a typical day of 4uly is illustrated in 3igure ;,;>.*he
collector is warmed passively by the solar radiation during the day. .eat stored in the
mass is retained and gradually released bac' during the night illustrated by the cooling
discharge profile.
0
10
$0
30
%0
&0
'0
(0
80
00 01 0$ 03 0% 0& 0' 0( 08 03 10 11 1$ 13 1% 1& 1' 1( 18 13 $0 $1 $$ $3
'i)e1 (rs
'
e
)
0
e
r
a
t
u
r
e
1
<
C
0
100
$00
300
%00
&00
'00
(00
800
300
1000
&
o
l
a
r

r
a
%
i
a
t
i
o
n
1

$
"
)
2
Tw
Ta
Tp
T+
*lobal solar radiation
Fi(. 3)"1* O.+ >,/ 3i+4> +x9+-i8+.: >,:,
3igure ;,;> demonstrates the close relationship between the absorber plate and water
temperatures and supports findings on the effect of thermal mass. *he absorber plate
absorbs incident solar radiation during the charging operation resulting in an increase of
absorber plate temperature. *he absorber plate transfers the heat to the water resulting in
similar temperature profiles. A decrease in incident solar radiation on the absorber plate1
due to shade1 resulted in a decrease in absorber plate and water temperatures. .owever1
as the water volume retained more heat than the absorber plate1 higher water
CB
temperature occurred during discharge. $t was also observed that a significant decrease
in incident solar radiation for a relatively long period of time during the day can lead to
significant losses. .owever1 a short time period would not affect the system
considerably as shown 3igure ;,;> between the hours of 9! and 9Bhrs.
SC/ -,>i,:i2. +33+6: 2. 62v+- :+89+-,:7-+
3urther analysis of the cover and ambient temperatures in 3igures ;,;= showed
significant outcomes. $t can be seen that recorded cover temperatures were lower than
ambient temperatures during daytime of the 99
th
and 9=
th
of 4uly. A possible e%planation
for this might be that the effective s'y temperature can be about 9",!" degrees &elcius
below ambient temperature at ground level with clear s'y conditions and close to or 0ust
below the ambient temperature during cloudy conditions. $n order to assess this possible
e%planation the clear s'y inde% for the location of the $&(,(-. for those dates were
ta'en from the BAD& +British Atmospheric Data &entre/ and plotted against both
temperatures in 3igure ;,;=.
0
&
10
1&
$0
$&
30
11 1$ 13 1% 1&
6a-
'
e
)
0
e
r
a
t
u
r
e
1

<
C
0.0
0.&
1.0
3
o
u
r
l
-

a
*
e
r
a
.
e
%

c
l
e
a
r

s
2
-

i
n
%
e
>
1

2
c
Ta
T+
0+
Fi(. 3)"2* SC/ -,>i,:i2. +33+6: 2. 62v+- :+89+-,:7-+ Ai:1 C6 ,0 :1+ 64+,- 0C/ i.>+x
Both days were recorded as clear s'y days from the BAD& +!"":/. *he s'y acts as a
blac' body during night,time and during an early morning clear s'y1 resulting in a cover
temperature lower than the ambient temperature. *he energy networ' illustrating the
e%changes between the cover M&N and the absorber plate MPN as well as the ambient air
MAN and the M('yN node is shown in 3igure ;,;B.
C:
Fi(. 3)"3* E.+-(/ .+:A2-C 32- 64+,- 0C/ 62.>i:i2.0
S:-,:i3i6,:i2. ,.> 024,- -,>i,:i2.
$n order to identify the influence of incident solar radiation on stratification the rise in
temperature registered by thermocouples placed at different locations along the length
of the collector is illustrated in 3igure ;,;:. (trong evidence of the importance of the
shading effect was found for the hours 991 9! and 9; where a decrease in stratification
rate was recorded1 supporting previous findings. *his is due to a sudden decrease in
incident solar radiation on the collector.
0
10
$0
30
%0
&0
'0
(0
0 1 $ % & ' ( 8 3 11 1$ 13 1% 1& 1' 18 13 $0 $1 $$ $3
'i)e1 (rs
'
e
)
0
e
r
a
t
u
r
e
1

<
C
0
100
$00
300
%00
&00
'00
(00
800
300
&
o
l
a
r

r
a
%
i
a
t
i
o
n
1

$
"
)
2
T$ T& T3 T1$ T13 T1' *lobal solar radiation
Fi(. 3)"#* S:-,:i3i6,:i2. Ai:1 :i8+ ,.> 024,- -,>i,:i2.
$n order to support these findings e%perimental temperature profiles were scaled to the
ratio *
h
J *
bottom
where *
h
is the local and *
bottom
is the minimum recorded temperature in
CC
the water tan'. Dimensionless stratification profiles for different hours are represented
in 3igure ;,;C for the same typical day.
1.00
1.0$
1.0%
1.0'
1.08
1.10
1.1$
1.1%
1.1'
1.18
1.$0
1.$$
1.$%
0.0 0.$ 0.% 0.' 0.8 1.0
6i)ensionless %istance
6
i
)
e
n
s
i
o
n
l
e
s
s

t
e
)
0
e
r
a
t
u
r
e
0hr $hr
%hr 'hr
(hr 8hr
3hr 10hr
11hr 1$hr
13hr 1%hr
1&hr 1(hr
13hr
Fi(. 3)""* Di8+.0i2.4+00 42.(i:7>i.,4 0:-,:i3i6,:i2.* T1 ? T52::28 v+-070 1?L
*emperature stratification decreases gradually over night until Bam1 at which time
incident solar radiation starts being collected by the absorber plate. -hile the overall
collector temperature increases from :am to !pm1 stratification increases from :am to
#am then decreasing later with time. *his is e%plained by the buoyancy effect already
discussed in section ;.!.!.!. 3rom :am high temperature stratification occurs in the
$&(,(-.. At #am the temperature of the upper layers is established resulting in the
lower layer achieving a narrower temperature range1 therefore decreasing the
stratification in the collector. A decrease in de,stratification rate was observed at the 991
9! and 9;hr suggesting that the collector was shaded. *his is supported by a sudden
decrease in incident solar radiation on the collector for this period of time shown in
3igure ;,;:.
C244+6:i2. +33i6i+.6/
$t should be noted that the following procedure can be applied to $&(,(-. evaluation)
however it is not truly representative of actual system performance due to its transient or
dynamic nature.
*he common standard procedure developed by the European (olar &ollector and
(ystems *esting 7roup1 E(&(*7 +9#C#/ was used to measure the effectiveness of the
$&(,(-.. Assuming the collector is operating in a steady state condition1 an energy
C#
balance represented by equation ;.= enables the calculation of the rate of useful energy
gain.
( )
a w
T T # & %
+;.=/
-here

is the optical efficiency +transmission,absorption product/ of the single

glaDed float glass cover
T
w
is the bul' water temperature +</
T
a
is ambient air temperature +</
# is the combined collector heat loss coefficient +-Jm
!
.</.
& is the incident solar radiation +-Jm
!
/
Based on equation ;.! already described in section ;.!.;.>1 the collector efficiency is the
ratio of the bul' heat collected to the heat available.
Using equations +;.!/ and +;.=/1 an e%pression for efficiency was developed in equation
;.B below.
( )
&
T T
#
a w

+;.B/
*o avoid negative values of efficiency occurring at early morning or late afternoon a
critical irradiance
c
&
was introduced. Negative values can occur when incident solar
radiations are lower than the heat losses of the $&(,(-.. *his method allowed setting
the heat collection to nil in order to sustain a net energy gain.
( )
( )
a w
c
T T #
&

+;.:/
As discussed by E(&(*7 +9#C#/ equation +;.B/ is useful because it can be used to plot
an informative efficiency curve1 which predicts collector performance for any operating
point.
A constant heat loss coefficient would result in a straight line if efficiency is plotted
against the 'nown operating point. 3igure ;,;# illustrates the efficiency curve for the
#"
$&(,(-. e%periment data. Based on equation ;.B1 the heat loss coefficient of the $&(,
(-. was found to be >.= -Jm
!
.<.
0
0.1
0.$
0.3
0.%
0.&
0.'
0.(
0.8
0 0.0$ 0.0% 0.0' 0.08 0.1 0.1$ 0.1% 0.1'
?'w='a@ " A 1 7.)
2
" $

Fi(. 3)"%* E33i6i+.6/ 4i.+ 32- :1+ ICS)SWH 0:7>i+>

As the temperature difference between the bul' water and ambient air increases the
efficiency of collection falls. 3or any given operating point the efficiency could be
obtained from 3igure ;,;# and used to predict the thermal performance of the $&(,
(-.. .owever1 limitations occur by using this procedure for $&(,(-. as it applies
+only/ to diurnal system performance. *his performance method is applied to the flat
plate collector wor'ing in a steady state mode.
1.1.1.2# Wi:1 D-,A)O33
3ield tests with draw,off were carried out for the month of August in order to identify
how the $&(,(-. would operate in a building. Based on the literature review1
developed in &hapter !1 c2ennan +!""B/ water consumption for a typical day in
(cotland was ta'en as reference for an average household of !.>> persons. *he $&(,
(-. was designed as =" litres or the consumption of one person. *herefore c2ennan
+!""B/ draw,off profiles were ad0usted for one person in order to ma'e an appropriate
analysis. Due to e%periment restriction caused by the low accuracy of the timer a
minimum flow rate of ".""ClJm
!
.s was established delivering an average daily hot water
consumption of :#lJdayJperson. c2ennan +!""B/ and the e%perimental draw,off
pattern established are shown in *able ;,> below.
#9
T,54+ 3)&* D,i4/ 12: A,:+- >+8,.> 9-23i4+
.7erage hot water deand
profile by 1+Lennan for $.%%
person
.7erage draw4off profiles
used in e8perient
assuing 1
$
/person
3ours 9itres"(rs 9itres"(rs
1 0 0
$ 0 0
3 0 0
% 0 0
& 1 0
' 1 0
( % 3
8 $3 1&
3 10 (
10 ' %
11 ' %
1$ % 3
13 & 3
1% $ 1
1& $ 1
1' 1 1
1( $ 1
18 % 3
13 1' 11
$0 13 3
$1 ( &
$$ 3 '
$3 & 3
$% 1 0
'otal water consu)0tionB 1$$ Litres/day (3 Litres/day/pers
*he flow rate achieved for the e%periment of ".""ClJm
!
.s was found acceptable when
compared with previous research of Duffie and Bec'man +9##9/ suggesting flow rates
of "."9='gJm
!
.s or lower flow rates of ".""! to ".""B 'gJm
!
.s recently used in (weden.
?ecorded data for two typical days of August are represented below in 3igure ;,>".
#!
0
10
$0
30
%0
&0
'0
(0
80
0 3 ' 3 1$ 1& 18 $1 0 3 ' 3 1$ 1& 18 $1
'i)e1 (rs
'
e
)
0
e
r
a
t
u
r
e
1

<
C
Tw Tout
T$ T&
T13 T1'
Fi(. 3)"&* S:-,:i3i6,:i2. i. :1+ ICS)SWH ,.> A,:+- :+89+-,:7-+ >-,A)233
*he manufacturing error addressed in section ;.9.9.! concerning the positioning of the
outlet pipe might affect the flow inside the $&(,(-. when cold water is introduced)
however no specific problems were observed. ?esults showed that the low flow rate
used for e%periments maintained a high level of temperature stratification. A sudden
drop in temperature at the bottom of the tan' is recorded at pea' demand hours of 9#F""
and !"F"". *he hot water is tapped and replaced by cold inlet water at the bottom of the
collector therefore decreasing the lower tan' water temperature. (trong evidence
suggesting that draw,off increases the overall stratification is shown in 3igure ;,>". A
more detailed picture for day two encompassing an interesting set of results is shown in
3igure ;,>9.
#;
0
$
%
'
8
10
1$
1%
1'
1 $ 3 % & ' ( 8 3 10 11 1$ 13 1% 1& 1' 1( 18 13 $0 $1 $$ $3 $%
'i)e1 (rs
6
r
a
w
=
o
f
f
1

l
0
10
$0
30
%0
&0
'0
(0
'
e
)
0
e
r
a
t
u
r
e
1

<
C
/raw off profile
Tw
Tout
Tin
Fi(. 3)"<* D-,A)233 9-23i4+ ,.> 9+-32-8,.6+
$t is noticed that the water collected is equal to or above the mean water temperature.
*his is merely due to the stratification inside the tan'. *he lower mean water
temperature achieved after draw,off suggests once more an increase of the overall
stratification in the collector. *he average temperatures of the water drawn,off were
recorded as ;".>&1 ;B.!&1 >=.=& and finally ;#.:& at pea' demand hours of CF""1
#F""1 9#F"" and !"F"" respectively. $n order to support the statement that Mdraw,off
increases the overall stratificationN of the collector stratification analysis was carried out
at the two daily pea' demand periods and is illustrated in 3igure ;,>! and ;,>;.
#>
1.00
1.0&
1.10
1.1&
1.$0
1.$&
1.30
1.3&
1.%0
0.0 0.1 0.$ 0.3 0.% 0.& 0.' 0.( 0.8 0.3 1.0
6i)ensionless %istance
6
i
)
e
n
s
i
o
n
l
e
s
s

t
e
)
0
e
r
a
t
u
r
e
&900
(900
8900
03900
10900
11900
Fi(. 3)"$* A.:+ M+-i>i,. >i8+.0i2.4+00 42.(i:7>i.,4 0:-,:i3i6,:i2. Ai:1 >-,A)233* T1 ? T52::28 V0 1?H
*emperature stratification decreases gradually by night until =am. *he temperature
stratification then increases with time1 from Bam to 99am while the overall collector
temperature increases until !pm. *his increase in stratification is due to cold water
being introduced at the bottom of the collector lowering bottom water temperature
therefore increasing the temperature stratification inside the collector. (tratification then
decreases from 9!am to >pm during which time there is no significant water draw,off.
*emperature of the upper layers then starts to be established resulting in the lower layer
achieving a closer equivalent temperature. *he temperature stratification then increases
from =pm to :pm due to an increase in water being used. An interesting result comes
from the decrease in stratification at Cpm1 despite the water being drained,off. *his can
be e%plained by the combination of the large amount of water used at :pm and the very
low or nil incident solar radiation received by the absorber plate at Cpm in August
resulting in a lower top collector temperature rise.
#=
1.0
1.1
1.$
1.3
1.%
1.&
1.'
1.(
1.8
0.0 0.$ 0.% 0.' 0.8 1.0
6i)ensionless %istance
6
i
)
e
n
s
i
o
n
l
e
s
s

t
e
)
0
e
r
a
t
u
r
e
1'900
13900
$0900
$$900
Fi(. 3)%'* P20: M+-i>i,. >i8+.0i2.4+00 42.(i:7>i.,4 0:-,:i3i6,:i2. Ai:1 >-,A)233* T1 ? T52::28 V0 1?H
3.3.3 Di06700i2. 23 3i+4> :+0:i.(
1.1.1.2" Wi:127: >-,A)233
@ne anticipated finding showed that the higher the incident solar radiation received by
the absorber plate1 the higher the water temperature achieved. $t was demonstrated that
the thermal mass plays an important role in the effectiveness of the system as it affects
the heating charge and cooling discharge operations of the collector. $t was also
observed that a significant decrease in incident solar radiation for a relatively long
period of time can lead to significant losses1 therefore positioning of the $&(,(-. is of
primary importance to achieving good performances. $t was assessed the cloud cover
should not be neglected as it affects the cover temperature and therefore the final water
temperature output of the $&(,(-.. 3inally1 the common standard procedure
developed by the E(&(*7 +9#C#/ applied to the $&(,(-. suggested that there was a
need for new methods to predict water temperatures achieved and performances of $&(,
(-..
1.1.1.2% Wi:1 >-,A)233
?esults demonstrated that draw,off increases the overall thermal longitudinal
stratification in the collector. *he manufacturing error of the positioning of the outlet
pipe showed no adverse effect on temperature stratification inside the collector. Average
draw,off temperatures recorded at pea' demand hours suggest that the $&(,(-. could
#B
cover ==L1 BBL1 C;L :!L of the hot water demand at CF""1 #F""1 9#F"" and !"F""
respectively based on a delivery temperature of ==S&. A strong relationship between
water drawn,off1 water temperature stratification1 solar radiation and outlet water
temperature achieved in the collector was reported. $t was shown that the $&(,(-.
maintained temperature stratification when draw,off occurred and even improved it.
3.# U.6+-:,i.:i+0 ,.> +--2-0 ,0026i,:+> Ai:1 8+,07-+8+.:0
$nvestigating uncertainties and propagation of errors in e%perimental procedures is
essential to determine the quality of the data presented. Any li'ely sources of errors or
problems related to measurement may be categorised under the following two main
categoriesF equipment error and uncertainty and operation related problems and errors.
A brief analysis of both categories is detailed below.
3.#.1 E=7i98+.: +--2- ,.> 7.6+-:,i.:/
-ith any measurement there e%ist errors1 some of which are systematic and others
inherent to the equipment employed. *he uncertainty in the resolution is often referred
to as the accuracy of the equipment. Equipment uncertainty and error were analysed and
are reported.
*emperatures in the collector were measured using <,type thermocouples. *he
uncertainty in thermocouple measurement after calibration was found to be t".=&.
Uncertainties associated with calibration were neglected as accurate methods were used
to define the temperature of each thermocouple. An additional uncertainty of ".9& was
associated with data logger conversion due to its limitation in resolution.
*he electrically heated silicone rubber pad and electrical equipment used in the
laboratory e%periment for control and measurement of the heat flu% had associated
uncertainties. *he ammeter and voltmeter used in the laboratory were chec'ed.
Uncertainty in measuring voltage and current were found to be t;L and t9L of reading
values. A wattmeter was used to chec' the heat flu% already measured through the
ammeter and voltmeter. *he uncertainty in the power output associated to the wattmeter
was found to be t>.=L. *he uncertainty associated with the conversion efficiency from
electrical to heat is difficult to assess as no technical data were provided by the
electrically heated silicone rubber pad manufacturer. An overall uncertainty of t9"L in
applied heat flu% was found acceptable.
#:
3ield e%periments equipment used for control and measurements have associated
uncertainties. *he B3; has an estimated uncertainty in global radiation measurement of
t=-Jm
!
at #=L confidence limits. *his suggests that #=L of individual readings will be
within the stated limits under normal climatic conditions. *he accuracy of the
anemometer was suggested at t".;mJs for a wind speed less than 9"mJs1 therefore
appropriate for the $&(,(-. location. *he timer used for e%periments showed
limitations associated with the draw off profiles. *he shortest switch time was of 9
minute therefore restraining the draw,off profile able to be achieved and has an
accuracy of t > minutes per year.
3.#.2 O9+-,:i2.,4 +--2-0
$n addition to the above sources of equipment,related errors1 care should be ta'en to
avoid operational errors highlighted belowF
, @peration related problems and errors
, &omplete or partial shade of the B3; from direct sunlight by buildings
, Dust1 snow1 dew1 water,droplets1 bird droppings etc.
, $ncorrect sensor levelling
Processing errors1 human errors or calibration errors are amongst others.
3." C2.647>i.( -+8,-C0
A review of the methodology adopted during laboratory and field e%periments was
presented herein. An account of the laboratory e%perimental set up with its results was
given. $&(,(-. performance was found to be enhanced by using aluminium material.
Determination of stratification dependent parameters1 presentably of interest to this
research pro0ect was also discussed in order to model stratification for laboratory
conditions. A field e%periment was advised and carried out to show the actual
performance of the $&(,(-. in real weather conditions. Parameters such as thermal
mass of the system1 incident solar radiation1 cloud cover and s'y temperature were
found to influence significantly the $&(,(-. performance and are of high interest in
order to model performance of such $&(,(-. for field conditions. &urrent procedures
measuring the effectiveness of $&(,(-. emphasise a need to predict water
temperatures and performances achieved of $&(,(-.. $t was also noted that the $&(,
(-. is able to maintain temperature stratification when draw,off occurred1 a crucial
factor to maintain good performance of the $&(,(-.. $t was also observed that draw,
#C
off actually increased the overall stratification1 thus improving the overall performance
of the $&(,(-.. 3inally an account of uncertainties and propagation of errors through
both e%perimental set,ups were addressed.
*he wor' demonstrated in this chapter shows the different parameters influencing the
$&(,(-. and accentuates the need for the development of a thermal program
predicting actual performance of such $&(,(-..
CHAPTER # MODELLING
odelling is very useful to represent the evolution of fluid flow1 temperature and
stratification processes in $&(,(-.s. icro and macro models were developed to
##
assess these parameters. *he macro model assessing fluid flow and heat transfer was
developed using Oisual Basic for Applications +OBA/. *he micro model assessed the
optimum fin spacing using &3D +&omputational 3luid Dynamics/ 3luent B.! software
while meshing was done using pre,processor 7ambit 2.2. *he macro and micro
models were both used through an iterative time step scheme. Numerical calculations
were developed for the macro model.
#.1 M,6-2 82>+4 B T1+-8,4 82>+4
As was previously stated in &hapter5s ! and ;1 the common criterion used to measure
the effectiveness of flat,plate (-. using the steady state test procedure proposed by the
E(&(*7 +9#C#/ cannot be applied to $&(,(-. as this involves a transient regime.
Direct comparison of measured system performances is relatively straight forward)
however1 predicting the performance of an $&(,(-. at another location or with a
different installation configuration is more challenging.
#.1.1 P7-920+ ,.> L,.(7,(+
A macro model assessing fluid flow and heat transfer in the $&(,(-. was developed
using Oisual Basic for Applications +OBA/. *his macro model is a thermal energy
simulation model which analyses combined heat flu% environmental networ's within
the $&(,(-.. $t was developed to compare the temperature variation in different $&(,
(-. materials1 internal water temperature and e%ternal weather conditions for a given
aspect ratio. *he model generates corresponding bul' water temperature in the collector
with a given hourly incident solar radiation1 ambient temperature and inlet water
temperature) thus predicting $&(,(-. performances. Depending on the conditions
applied A laboratory or field A input data will differ.
#.1.2 C,9,5i4i:i+0 ,.> Li8i:,:i2.0
OBA is functionally rich and e%tremely fle%ible but it does have some ma0or
limitations. $t e%tends the capabilities of icrosoft -ord or E%cel through macro
programming language and is widely used by solar energy professionals and engineers.
$t can also be used to control almost all functions of the host application +E%cel1
-ord8 functions/ and the graphic interface utilisation1 li'e menus and toolbars. OBA
is similar to Oisual Basic +OB/. .owever OBA can only run via a host application
ma'ing it impossible to operate as a standalone e%ecutable. E%cel combines a tabular
data acquisition with an inbuilt OBA. $n addition to this1 some functions in E%cel are
9""
particularly suited for engineering modelling as described by 2iengme +!"";/ and are
available in OBA as stated by 4acobson +!""9/. .owever E%cel has two ma0or
limitations as it cannot tabulate more than B=1""" rows1 nor graph more than !!1"""
data points. Any function of the application or even of the whole operating system is
accessible to the user running a document containing OBA macros ma'ing
improvements of the program possible on a continual basis as the code is accessible.
*his model was built from scratch hence 'nowledge based ownership is whole. *he
short training time and the lac' of special licensing requirements of this 'ind of
application ma'e the model attractive.
#.1.3 A00789:i2.0
A number of assumptions were made in developing the macro model for instance from
lac' of data or in order to reduce computation. Assumptions applying to both laboratory
and field conditions are e%pressed belowF
+a/ *ransient performance of the system was assumed to be steady state during the small
time steps chosen to solve the thermal networ'
+b/ *hermophysical properties were ta'en at the film temperature based on uneer et al
+!"";/ water and air properties.
+c/ .eat loss through the absorber plate to the glass cover was assumed to be equal to
the heat loss from the cover to ambient conditions due to steady state time step.
+d/ .eat flow was assumed to be one dimensional.
+e/ *he absorber plate and glass cover were assumed to be isothermal for the small time
steps chosen.
+f/ *he plate temperature was started at two degree &elsius above the water temperature
in order to initialise the simulation. *his was later recalculated by the programme
through energy balance equations.
+g/ *he thermal analysis was underta'en for a >= collector inclination
+h/ Dust on the glass cover was assumed negligible
1.1.1.2& F2- 4,52-,:2-/ 62.>i:i2.0
Assumptions made to develop the laboratory macro model are given belowF
+a/ *he thermal capacity of the transparent cover was neglected.
+b/ -ind velocity was assumed to be ".9mJs for laboratory conditions.
+c/ *he s'y temperature was assumed equal to ambient temperature.
+d/ (tratification inside the collector was first neglected and water temperature was
considered as uniform. (tratification was then modelled.
9"9
$nput parameters required for the system to predict $&(,(-. performance areF
+9/ .eat flu%
+!/ Ambient air temperature
+;/ $nlet water temperature
+>/ Average wind velocity
+=/ $&(,(-. characteristics +materials1 dimensions1 etc8/
+B/ $nclination angle of the collector
1.1.1.2< F2- 3i+4> 62.>i:i2.0
Assumptions made to develop the field macro model are given belowF
+a/ *he absorber plate was assumed to have a constant rate of heat flu% for the small
time increments ta'en by the programme.
+b/ An average wind velocity of ;.:mJs was used for the field e%periment based on data
recorded on the roof where the $&(,(-. was previously tested.
+c/ ?adiative heat losses from the glass cover accounted the glass cover and s'y
emissivities.
+d/ *he dry,bulb temperature for part,overcast conditions was assumed as the average
between the overcast and clear s'y conditions.
+e/ $f overcast condition data were not available then the cloud cover
/ +((
was
assumed as part,overcast
/ > + ((
+f/ (tratification inside the collector was neglected for field e%periments) it was thus
assumed that water temperature inside the collector was uniform.

$nput required for the system to predict $&(,(-. performance areF
+9/ .oriDontal global and diffuse irradiation J (loped global irradiation
+!/ Ambient temperature
+;/ $nlet water temperature
+>/ -ind velocity
+=/ $&(,(-. characteristics +materials1 dimensions1 etc8/
+B/ $nclination angle of the collector
+:/ &loud cover
&lear1 overcast and part,overcast conditions were differentiated. ('y conditions were set
asF &lear s'y conditionsF ! " ((
@vercast s'y conditionsF C B ((
9"!
Part,overcast s'y conditionsF = ; ((
*he more detailed and complete the weather information inputs1 the fewer error sources1
and thus confidence in the macro model results1 might be achieved.
Effects of assumptions were carefully considered in order to ensure that the analysis
was appropriate. *a'ing these into account1 a thermal networ' and fundamental heat
transfer analysis was developed and then coded through OBA.
#.1.# T1+-8,4 .+:A2-C ,.> 37.>,8+.:,4 1+,: :-,.03+- ,.,4/0i0
1.1.1.2$ T1+-8,4 .+:A2-C
*he model developed differentiates between unfinned and finned1 collector materials
and dimensions of collectors1 and was developed on a steady state approach as the water
temperature does not change significantly in very short time intervals. 2aboratory and
field macro models have differences as they are sub0ect to different assumptions and
methods of calculations. .owever1 the global thermal networ' of the $&(,(-. system
will remain the same and is shown in 3igure >,9 below including the water body M-N1
the absorber plate MPN1 the cover M&N 1 the ambient air MAN and the M('yN nodes.

9";
c
ca
#
r
ca
#
r
pc
#
c
pc
#
Sky A
(
+
W
A
m
(
w
(
pw
#
wa
#
( ) 8
*op &over
Air cavity
-ater cavity
Fi(. #)%1* T1+-8,4 .+:A2-C 23 :1+ 0/0:+8
1.1.1.3' F7.>,8+.:,4 1+,: :-,.03+- ,.,4/0i0
*ransient performance of the system was predicted for a ;B" second time interval by
solving the mathematical model consisting of energy balance equations. An iterative
process was developed to solve the thermal networ'.
*he energy balance on the absorber plate at a given instant1 assuming steady state
conditions1 is given byF
( ) ( ) ( )
w w a w c p #sef$l
T T ( % % 8 %

H
+>.9/
( )
e w p c p
8 % % +
at tG" +>.!/
s c c p
% %


+>.;/
-here #sef$l
%
is the useful energy1 7+/

is the rate of incident solar radiation
transmitted through the glass cover to the absorber plate1 w p
%
is the energy lost from
9">
the absorber plate to the water1
s c
%

is the heat loss from the cover to the surroundings1
c p
%
is the energy lost from the absorber plate to the glass cover1 and
a w
%

is the
energy lost by conduction and convection from the water to the ambient. & is the overall
thermal capacitance of the system composed of
w
(
1
i
(
1
wd
(
1 !
(
1
c
(
which are
the water1 the insulation1 the wood1 the glass cover1 and collector material respectively
and varies for field and laboratory e%periments.
w
TH
is the final water temperature at
tGiI9 and
w
T
is the initial water temperature at tGi.
A schematic diagram of heat flu% e%changes occurring at the $&(,(-. is shown in
3igure >,! below.
Fi(. #)%2* H+,: 347x +x61,.(+0 2667--i.( ,: :1+ ICS)SWH
Heat loss from the plate to the cover
*hese can be e%pressed byF
( ) / +
/ +
> >
c p pc
B
c p
c p pc c p
T T #
T T
T T h %

+>.>/
9"=
.eat losses .eat gains
T
A
Air cavity
-ater cavity
T
,
P
4200+0
T
6
T
9
P
U0+374
7lass cover
$nsulation
Absorber plate
where
B

is the bul' emissivity defined by equation

9
9 9
+
p c
B

1
+
T
is the
temperature of the absorber plate1
c
T
is the temperature of the glass cover temperature1
pc
#
is the overall U,value between the plate and the cover1 is the (tefan,BoltDmann
constant and pc
h
is the convective heat transfer coefficient between inclined parallel
plates with a slope from the absorber plate to the glass cover. Based on the wor' of
.ollands et al +9#:B/ the following relationship of pc
h
for inclined cavities was
developedF
L
k
'a
'a 'a
h
fa
L
L L
pc

,
_

1
1
]
1

,
_

+
1
]
1

1
]
1

9
=C;"
cos
cos
/ C . 9 +sin 9:"C
9
cos
9:"C
9 >> . 9 9
; J 9
B . 9

+>.=/
-here fa
k
is the thermal conductivity of air at film temperature and L is the
characteristic length. *he notation K implies that1 if the quantity in brac'ets is negative1
it must be set equal to Dero.
Heat loss from the cover to the surroundings
2osses to surroundings will differ whether the models compute laboratory or field
conditions. Both are e%pressed below respectivelyF
2aboratory conditionsF
( ) / + / +
> >
. a c ca Sky c c a c Wind s c
T T # T T T T h % +

+>.B/
3ield conditionsF
( ) ( )
Sky c cs a c ca
s$r
Sky c
a c Wind s c
T T # T T #
T T
T T h % +

/ +
/ .+
> >

+>.:/
9"B
where
Wind
h
is the e%ternal convection from the glass cover to ambient1 Sky
T
is the s'y
temperature1
a
T
is the ambient temperature1
c

is the glass cover emissivity and

s$r

is the bul' emissivity of equation
9
9 9
+
Sky c

where Sky

is the emissivity of the s'y.

*he following correlation was suggested by Duffie and Bec'man +9#:>/ for convection
occurring on a collector surface where
w
)
is the wind speed.
w Wind
) h ; C . ! +
+>.C/
According to &$B(E guide A +!""B/1 a value of ".9mJs of
w
)
can be used to represent
convective velocity in a closed room adequate for laboratory tests. A value of ;.:mJs1
based on e%perimental observations1 for field tests based in Edinburgh was assumed.
As stated in the literature1 Sky
T
is the s'y temperature and is assumed equal to
a
T
for
laboratory conditions. .owever1 for field tests the s'y is acting as a blac' body during
night and during clear s'y conditions in early morning.
*he s'y temperature was then determined by using the e%pression developed by Berdahl
and artin +9#C>/ in equation +>.#/ below where the s'y temperature is related to the
dew point temperature1 dp
T
1 ambient temperature1
a
T
1 and time relative to midnight1
t .
[ ]
> J 9 !
/ 9= + "9; . " """":; . " ""=B . " :99 . " t (os T T T T
dp dp a Sky
+ + +
+>.#/
*he s'y emissivity was calculated by integrating &ucumo et al5s +!""B/ equation +>.9"/
of clear s'y emissivity1
(S

1 into (antamouris and Asima'opolous5s +9##B/ equation

+>.99/ for s'y emissivity in the transparency window1 Sky

.
> . :;=#> C . BC"
/ 9 + 9":#=!
9
!
+

+
a a
(S
T T

+>.9"/
-ith
!
9""
BBBC . "
9""
B""9 . " :>: . "

,
_

+
dp dp
T T

9":
-here
(S

is the atmospheric emissivity in the transparency window for clear s'y

condition and

is the global emissivity.

*he s'y emissivity in the transparency window1 Sky

1 was then calculated by using the

final e%pression by (antamouris and Asima'opolous +9##B/ e%pressed belowF
/ """!C . " "";= . " "!!> . " 9 +
; !
n n n
(S Sky
+ +
+>.99/
-here n is the total opaque cloud amount) " for clear s'y and 9 for overcast s'y.
A bul' emissivity1
s$r

1 ta'ing into account the s'y and glass cover emissivity of

equation
9
9 9
+
Sky c

was then introduced to calculate the radiative heat losses from
the glass cover to the surrounding for field e%periments.
Dew,point and ambient temperatures relationship were developed for Edinburgh1 based
on one year data from the British Atmospheric Data &entre +BAD&/. *his data is
plotted in 3igures >,; and >,> from which the following relationships were developed
for the different types of s'y.
3or clear daysF
:# . " :9 . "
a dp
T T
(
+>.9!/
3or overcast daysF
C9 . " C! . "
a dp
T T
9
+>.9;/
$t was assumed that part,overcast dew,point temperature was the average between clear
and overcast conditions resulting in a dew,point e%pression belowF
!
9 (
+9
dp dp
dp
T T
T
+
+>.9>/
9"C
y : 0.(18 4 0.(3
;
$
: 0.88
410
4&
0
&
10
1&
$0
410 4& 0 & 10 1& $0 $&
,)bient te)0erature1 <C
6
e
w
=
0
o
i
n
t

t
e
)
0
e
r
a
t
u
r
e
1

<
C
Fi(. #)%3* D+A)92i.: ,.> ,85i+.: :+89+-,:7-+0 -+4,:i2.01i9 32- 64+,- >,/0
y : 0.8$8 4 0.81
;
$
: 0.80
410
4&
0
&
10
1&
$0
$&
410 4& 0 & 10 1& $0 $&
,)bient te)0erature1 <C
6
e
w
=
0
o
i
n
t

t
e
)
0
e
r
a
t
u
r
e
1

<
C
Fi(. #)%#* D+A)92i.: ,.> ,85i+.: :+89+-,:7-+0 -+4,:i2.01i9 32- 2v+-6,0: >,/0
Heat loss from the absorber plate to the water
*he heat transfer to the water is mainly a convective process thus conduction was
neglected. *he heat loss from the absorber plate to the water is e%pressed belowF
( ) ( )
p w pw p w w w p
T T # T T h %
+>.9=/
9"#
where
+
T
is the temperature of the absorber plate1
w
T
is the water temperature at t G
i1 w pw
h #
is the overall U,value between the absorber plate and the water. A Nusselt
e%pression for trapeDoidal,shaped $&(,(-. based on 3u0i and $mura +9#:!/ and
adapted by &ruD et al +9### Q !""!/ to evaluate
w
h
e%pressed asF
( )
L
k
#
H
#
D
'a(os
h
fw
w

,
_

,
_

,
_

B J 9 # J 9
> J 9
=B . "
+>.9B/
-here m H 9 and m # D "= . " 1

>= for Edinburgh conditions1 fw

k
is the
thermal conductivity of water at film temperature and L is the characteristic length. A
visual description is shown in 3igure >,= below.
Fi(. #)%"* G+28+:-i6 9,-,8+:+-0 23 :1+ ICS)SWH (547+ .2>+0 >+.2:+ 347i> :+89+-,:7-+
8+,07-+8+.:0 426,:i2.0)
Heat loss from the water to the ambient
( )
a w wa a w
T T # %
.
+>.9:/
99"
where
w
T
is the water temperature at t G i1
a
T
is the ambient temperature and
wa
#
is
the overall U,value between the water to the ambient through the bo%1 developed asF
&ns Wind
wa
' '
#
+

9
+>.9C/
-here
&ns
'
is the resistance of the insulation material which will vary between
laboratory and field installations1 and
Wind
'
is the resistance occurring at the bo%
surface in contact with the wind.
Heat loss from the fins to the water
*he inclusion of e%tended heat transfer surfaces +fins/ into the design resulted in an
increase in heat transfer to the water. $deally the fin material should have a large thermal
conductivity to minimiDe temperature variations from its base to tip. 3our fins of
straight1 rectangular uniform cross sectional area were used for the design. Each fin was
attached to a base surface of temperature
/ " + T T
p

called plate temperature and
e%tends into fluid temperature
w
T
as described in 3igure >,B.
Fi(. #)%%* S:-,i(1: -+6:,.(74,- 3i.0 23 7.i32-8 6-200 0+6:i2. ,.> +.+-(/ 5,4,.6+ 32- ,. +x:+.>+>
07-3,6+
Based on the wor' of $ncropera and Dewitt +!""9/1 a new model was developed using
an elemental strip method to determine the heat transferred from the fin to the water
assuming an adiabatic tip condition + d x /F
"
d x
o
dx d
. *he fin temperature
distribution along the length
x
was determined by the following formulaF
999
T
p
w o x f
T
md
x d m
T +

cosh
/ + cosh
1

+>.9#/
-here w p o
T T
c
kA h+ m
!
c
A
G cross sectional area +
w t A
c

/
+ G perimeter
x
G strip position
d G depth of the fins
*he fin heat transfer rate f
%
per strip can be e%pressed asF
/ + .
w f f f f
T T A h %
+>.!"/
-here

f
A
fin area
w x *
s

* number of fins

s
x
strip length ta'en equal to m ""! . " for modelling.
*he heat transfer to the water is mainly a convective process. uneer et al +!"";/ used
the equation of &hurchill and &hu +9#:=/ incorporating the following regression
assuming a uniform heat flu%F

( ) [ ]
# J >
9B J #
> J 9
Pr J >;: . " 9
. B: . "
BC . "
+
+
L
'a
*$
+>.!9/
-here
/ + (os d L
with

>= as the inclination angle of the collector and d as the

depth of the fins.
*he effective length of the fin was defined when #" percent of the actual W p o
T T

was achieved , as only #" percent of the heat was assumed to conduct through the fins
and is represented in 3igure >,: below.
99!
0.0
0.1
0.1
0.$
0.$
0.3
0.3
0.%
0.%
0.&
0.0 0.1 0.$ 0.3 0.% 0.& 0.' 0.( 0.8 0.3 1.0
6i)ensionless fin len.t(1 x?>
6
i
)
e
n
s
i
o
n
l
e
s
s

f
i
n

t
e
)
0
e
r
a
t
u
r
e
1

T
p
!
T
f
"
x
#
/

T
p
!
T
w
#
.luiniu +olle+tor43
.luiniu +olle+tor41.&
-tainless4steel +olle+tor41.&
Fi(. #)%&* Di8+.0i2.4+00 3i. :+89+-,:7-+ v+-070 3i. 4+.(:1
*he total heat transferred through the fins to the water
ins
%
was calculated using the
effective length of the fin. *he quantity
ins
%
1 e%pressed below was calculated by
adding up the fin heat transfer rate f
%
per strip until reaching the limit of effective
length.

L #"
"
x x x ins
A h %
+>.!!/
An improvement factor +&/ was introduced representing the improvement achieved by
using fins in the $&(,(-. over an un,finned collector and is e%pressed in equation >.!;
below.
finned #n
w p
ins
%
%
&

+ 9
+>.!;/
Calculation of the bulk water temperature
Using the equations developed above1 an iterative computer program was developed
using OBA with initial conditions of p
T
and
c
T
set asF
= . " +
w p
T T
99;

,
_

;
a p
p c
T T
T T
Using equations +>.!/ and +>.;/1 e%pressions for p
T
and
c
T
for tG" were developed
below in equations +>.!>/ and +>.!=/. *he glass cover temperature at tG" is a function of
the assumed value of absorber plate temperature.
pc ca
a ca p pc
c
# #
T # T #
T
+
+

for laboratory conditions +>.!>/

cs pc ca
Sky cs a ca p pc
c
# # #
T # T # T #
T
+ +
+ +

for field conditions

( )
w pc
w w c pc
p
# #
T # T # 8
T
+
+ +

+>.!=/
$terations were made until a balance was obtained between heat losses. *he program
finally output the glass cover temperature U
c
T
5 and the absorber plate temperature U
p
T
5. .eat flu%es in the networ' are calculated in the macro and using equation +>.9/ a
transformed equation +>.!B/ can be developed to calculate the water temperature riseF


+
+ +
(
% % %
T
(
%
T T
a w ins w p
w
#sef$l
w w
H
+>.!B/
An interval of one hour was ta'en to display results. After each interval the value of the
cover1 absorber plate and water temperature were calculated and used as the input for
the ne%t time step. *he cycle was repeated for !> hour time increments.
$f the option MN@N fins is activated1 then
x
A
is automatically set to Dero1 then fins are
not ta'en into account into the calculations.
#.1." M2>+44i.( 0:-,:i3i6,:i2. 32- 4,52-,:2-/ 62.>i:i2.0
(tratification in a water tan' is an essential factor for the good performance of an $&(,
(-.. *his section e%plores the modelling of stratification inside an aluminium $&(,
(-. for laboratory conditions.
99>
-ater temperatures within the collector were observed1 in &hapter ;1 to be dependent
on the longitudinal position1 imposed heat flu%1 and e%posure time. Using e%perimental
data1 a polynomial function for the stratification profile at each hour was developed
using heat flu% and longitudinal position in the collector as the main parameters. $n
order to obtain a generalised equation for any heat flu%1 dimensionless temperature1 1
and dimensionless length1 H h h J K 1 were calculated and plotted for each temperature
measurement location for the aluminium $&(,(-. on an hourly interval. *his is shown
in 3igure >,C.
40.08
40.0'
40.0%
40.0$
0.00
0.0$
0.0%
0.0 0.1 0.$ 0.3 0.% 0.& 0.' 0.( 0.8 0.3 1.0
(CD("3

D
?
'
(
=
'
w
b
@
"
'
w
b
100#
$00#
%00#
Fi(. #)%<* Di8+.0i2.4+00 :+89+-,:7-+ 0:-,:i3i6,:i2. ,3:+- < 127-0
Polynomial functions were developed from 3igure >,C for each wattage as defined
belowF
c bh ah + + K K
!
+>.!:/
A single function was then developed incorporating the change of stratification with
position and heat flu%. *he following equations were developed to produce a
temperature stratification profile for an average eight hour day. Polynomial functions
for heat flu% of 9""1 !"" and >""-Jm
!
are e%pressed belowF
"=>" . " K 9CB: . " K 9!9! . "
!
9"" 1 C
+ h h
W hrs

+>.!C/
"CC: . " K ;9!! . " K !":= . "
!
!"" 1 C
+ h h
W hrs

+>.!#/
99=
99"! . " K ;C>C . " K !=;9 . "
!
>"" 1 C
+ h h
W hrs

+>.;"/
A final overall e%pression was then developed for each hour as a function of the heat
flu% input and the longitudinal position in the collector. *he e%pression for the stratified
temperature at the end of eight hours of operation is given asF
c bh ah
hrs
+ + K K
!
C

+>.;9/
-here a1 b and c are the variation of the polynomial coefficient with heat flu% for each
hour. &oefficients were noted and modelled in a more generalised function and are
provided belowF
; ; ! B
9" > . : 9" 9= 9" !

+ " " a
+>.;!/
; ; ! B
9" : . 9 9" 9 . ! 9" ;

+ + " " b
+>.;;/
; ; ! B
9" ; . ; 9" B . " 9" C . "

" " c
+>.;>/
*he same approach was used for each hour.
Using the previously developed macro model for laboratory conditions) polynomial
functions were added to the macro model to compute water temperature stratification
with height. *he initial laboratory macro model was able to generate corresponding bul'
water temperatures in the collector with a given hourly incident solar radiation1 ambient
temperature and inlet water temperature.
*he new model was able to compute the bul' water temperature variation in different
$&(,(-. collectors for a given aspect ratio and to evaluate temperatures at various
points in the collector +temperature stratification/ based on the mean water temperature
given by the macro model and the total time of e%posure.
?esults comparing computed with e%perimental stratification are discussed in section
>.9.:.!.
99B
#.1.% Di(i:,4 0i874,:i2. 342A 61,-:
$nput .eat 3lu% or
(olar radiation1
7+/
&alculation of *
w

from _
Useful
Update time step1
tF i G i I 9
i G 9"K
END
&hec'
i
Update *
p
and *
c

i G i I 9
&alculation of U
ca

and U
pc
$nlet water
temperature1 *
w
$nitialisation of
*
p
and *
c
Ambient
temperature1 *
a
&loud cover1 &&
&alculation of
*
dp
&alculation of
*
('y
and
('y
Air properties at
film temperature1
*
fa
-ind velocity O
w
&over and plate
emissivity1
c
1
p
-ater properties at
film temperature
*
fw
&alculation of
_
3ins

&alculation of
_
Useful
&alculation of
*
p
and *
c

+eq. >.!> A >.!=/
&alculation of
U
pw
&alculation of
U
wa
&hec'
i G 9"
i G 9"
i \ 9"
i \ 9"K
99:
#.1.& C2897:,:i2.,4 ,.> +x9+-i8+.:,4 >,:, 6289,-i02.
1.1.1.31 L,52-,:2-/ -+074:0
R+074:0 32- 3i..+> 6244+6:2-0
3igures >,# and >,9" show the simulated heat transfer rate through the fins and the fin
temperature through the fin length respectively. *he fins were used to enhance heat
transfer between the material and the ad0oining fluid. *he heat transfer rate was
increased by employing different fins configuration as demonstrated in 3igure >,#. *he
thermal conductivity of the fin material has a strong effect on the temperature
distribution along the fins +3igure >,9"/ and therefore influences the degree to which the
heat transfer rate is enhanced as shown in 3igure >,#. $deally the fin material should
have a large thermal conductivity to minimise temperature variations from its base to its
tip.
0.0
0.1
0.$
0.3
0.%
0.&
0.'
0.(
0.8
0 % 8 1$ 1' $0 $% $8 3$ 3' %0 %% %8
Fin len.t( >1 ))
F
i
n

(
e
a
t

t
r
a
n
s
f
e
r

r
a
t
e
1

$
.luiniu +olle+tor43
.luiniu +olle+tor41.&
-tainless4steel +olle+tor41.&
Fi(. #)%$* Si874,:+> 3i. 1+,: :-,.03+- -,:+
99C
&1.30
&1.3&
&1.%0
&1.%&
&1.&0
&1.&&
&1.'0
0 & 10 1& $0 $& 30 3& %0 %& &0
Fin len.t( >1 ))
F
i
n

t
e
)
0
e
r
a
t
u
r
e
1

<
C
.luiniu +olle+tor43
.luiniu +olle+tor41.&
-tainless4steel +olle+tor41.&
0.1

01,9=3
0.1

0,
,9=1.5
0.1

0
1&'=1.5
Fi(. #)&'* Si874,:+> 3i. :+89+-,:7-+ >-29 A1+-+
) ( 1 . ' 1 . '
$ p %
T T
i0 :1+ +x6+00
:+89+-,:7-+ ,61i+v+> ,0078i.( $'K 23 1+,: i0 62.>76:+> :1-27(1 :1+ 3i.0.
3in length effectiveness1 and consequently heat transfer rate are enhanced by choosing a
material with high thermal conductivity) such as aluminium. $t was also enhanced by
increasing the thic'ness of the fin1 increasing at the same time the surface in contact
with water. *he optimum effective length of the fin was defined earlier when #" percent
of the actual $ p %
T T
was achieved. *he optimum effective length for the
aluminium collector fin of ;mm and 9.=mm thic'ness1 and stainless,steel of 9.=mm
thic'ness were found to be ;;.B:1 ;9.=" and 9!.B!mm respectively) equating to an
effective fin utilisation of B:.;L1 B;L and !=.;L1 based on their respective optimum
effective length.
(imilarly1 a higher fin heat transfer coefficient is achieved using the higher conductivity
material and greater material thic'ness. $ndeed1 by increasing the thic'ness a greater
amount of heat was transferred through the fins5 length to achieve improved rates of
heat transfer to the bottom of the collector as shown in 3igure >,99.
99#
%8.0
38.0
1%8.0
138.0
$%8.0
$38.0
3%8.0
338.0
%%8.0
0 % 8 1$ 1' $0 $% $8 3$ 3' %0 %% %8
Fin len.t( >1 ))
F
i
n

(
e
a
t

t
r
a
n
s
f
e
r

c
o
e
f
f
.
1

$
"
)
2
.
7
.luiniu +olle+tor43
.luiniu +olle+tor41.&
-tainless4steel4+olle+tor 1.&
Fi(. #)&1* Si874,:+> 3i. 1+,: :-,.03+- 62+33i6i+.:
?esults showed that the higher the heat flu% applied to the system1 the higher the fin
heat transfer rate achieved. .owever1 the fin heat transfer rate decreases along the
length until reaching an asymptotic value corresponding to the heat flu% applied. 3igure
>,9! reveals another typical trend showing that with time the fin heat transfer rate
decreases. *his is because with time the collector water temperature increases thus
reducing heat transfer from the fins as the water approaches a temperature equilibrium
with the e%ternal heat flu% applied to the system
0.0
0.&
1.0
1.&
$.0
$.&
0 % 8 1$ 1' $0 $% $8 3$ 3' %0 %% %8
Fin len.t( >1 ))
F
i
n

(
e
a
t

t
r
a
s
n
f
e
r

r
a
t
e
1

$
Fi(. #)&2* Si874,:+> 3i. 1+,: :-,.03+- -,:+ Ai:1 :i8+ B A478i.i78 388 SWH
9!"
1 (r
5(rs 10(rs 15(rs 20(rs
A fin efficiency f

was then calculated to evaluate the fin performance. ?eferring to

the wor' of $ncropera and Dewitt +!""9/1 for a straight fin of uniform cross section and
an adiabatic tip condition the fin efficiency was defined byF
mL
mL thanh
f
.

+>.;=/
-here
C
k& hP m
.
*he fin efficiencies were plotted in 3igure >,9; for the three $&(,(-. aluminium and
stainless,steel fins to see the improvement in efficiency from using a higher
conductivity material and by increasing the fin thic'ness. .igher efficiencies were
achieved for the aluminium material with a ;mm fin thic'ness.
0
$0
%0
'0
80
100
0 & 10 1& $0 $& 30 3& %0 %& &0
Fin len.t( >1 ))

f
,

%
.luiniu +olle+tor43
.luiniu +olle+tor41.&
-tainless4steel +olle+tor41.&
Fi(. #)&3* Si874,:+> 3i. +33i6i+.6/
*he improvement factor +$3/ decreases with time while the water temperature increases.
$ndeed
ins
%
decreases while w p
%
increases with time and are related to the water
temperature and the $3. *hese results are interrelated. A comparison between different
types of fins was made for an input temperature of !!.9& and a heat flu% of 9""-Jm
!

applied to the system. ?esults are presented in *able >.9 below.
9!9
T,54+ #)<* Ov+-,44 i89-2v+8+.: 3,6:2- 32- (iv+. 9,-,8+:+-0
aterial 3in thic'ness1 mm @verall $3
Aluminium ; 9.9>
Aluminium 9.= 9.9"
(tainless,steel 9.= 9.">
Unfinned , 9.""
By increasing the thic'ness of the fins the overall $3 improved by ;.=L1 while by
changing the material to a more conductive material the $3 improved by =.=L. *he
improvement factor measuring the fin performance showed that1 overall1 a higher
performance was achieved for the aluminium $&(,(-. with a ;mm fin thic'ness.
C289,-i02. 23 :1+ >i(i:,4 0i874,:i2. ,.> 4,52-,:2-/ -+074:0
A comparison was underta'en between the computational results from the model and
the actual e%perimental data. (imulated and computed efficiency of the aluminium $&(,
(-. achieved with time for >""-Jm
!
was plotted in 3igure >,9> and show a good
empirical correlation.
*he comparison of the water temperatures achieved e%perimentally and numerically
was plotted in 3igure >,9= for the three $&(,(-.s. 7ood correlations were observed
for all the $&(,(-. types.
0
10
$0
30
%0
&0
'0
(0
80
0 $ % ' 8 10 1$ 1% 1' 18 $0 $$
'i)e1 (rs

1

%
%00#6!8periental
%00#6-iulation
Fi(. #)&#* ICS)SWH +33i6i+.6/ Ai:1 :i8+ 32- #''W?82 1+,: 347x ,994i+>
9!!
0
&
10
1&
$0
$&
30
3&
%0
0 $ % ' 8 10 1$ 1% 1'
'i)e1 (rs

'
)

1

<
C
,inned .uiniul6!8periental
,inned .luiniu6-iulation
,inned -tainless4steel6!8periental
,inned -tainless4steel6-iulation
"nfinned -tainless4steel6!8periental
"nfinned -tainless4steel6-iulation
Fi(. #)&"* C2897:,:i2.,4 ,.> +x9+-i8+.:,4 6289,-i02. 32- 2"'W?8
2
1+,: 347x
&omputed data were plotted as a function of e%perimental data in 3igure >,9B to
e%amine the validity of the simulation. $t was demonstrated that e%perimental and
computer simulation results are in good agreement as they follow a 9F9 relationship with
a 9"L over prediction in the modelled value.
0
$0
%0
'0
80
0 $0 %0 '0 80
' )easure%1 <C
'

c
o
)
0
u
t
e
%
1

<
C
Fi(. #)&%* C2897:+> v0 Ex9+-i8+.:,4 A,:+- :+89+-,:7-+0
9!;
.owever1 a proper validation of the results by statistical methods and e%treme range
tests was carried out in order to improve confidence in the findings.
'tatistical measures used for evaluation of the model
$n order to evaluate the performance of each set of models1 the following statistical
indicators were employedF
, (lope of the best,fit line
, &oefficient of determination +?
!
/
, ean bias error +BE/
, ?oot mean square error +?(E/
, ('ewness
, <urtosis
The slope of the best:fit line between the computed and measured variable is desirably
equal to one. (lope values e%ceeding one indicate over,estimation) while slope values
less than one indicate under,estimation of the computed variable.
The coefficient of determination0 '
;
0 is the ratio of e%plained variation to the total
variation. $t lies between Dero and one. A high value of ?
!
indicates a lower une%plained
variation and is desirable. ?
!
is often used to 0udge the adequacy of a regression model
but it should not be the only criterion for choosing a particular model.
( )
( )

!
!
!
< 9
< (
= =
= =
'
+>.;B/
-here
(
=
G computed value
<
=
G mean value
9
=
G measured value
n
G number of points
The mean bias error0 <B.1 provides a measure of the trend of the model1 whether it has
a tendency to under,predict or over,predict its modelled values. Positive values of BE
indicate under estimation while negative values imply over estimation by the proposed
model. A BE nearest to Dero is desired. $t is given by the following equationF
9!>
n
= =
<B.
n
i
( 9

9
/ +
+>.;:/
-here
n
G number of points
The root mean s%$are error0 '<S.1 gives a value of the level of scatter that the model
produces by providing a comparison of the actual deviation between the predicted and
the measured values. (ince it is a measure of the absolute deviation1 ?(E is always
positive. A lower absolute value of ?(E indicates a better model.
! J 9
9
!
/ +

,
_

n
= =
'<S.
n
i
( 9
+>.;C/
Skewness represents a measure of the lac' of symmetry in a distribution.
A distribution is symmetric or normal if it loo's the same to the left and right of the
centre point1 resulting in s'ewness statistic of Dero value for perfect symmetry. *his
would indicate a robust model. A positively s'ewed distribution tails off to the high end
of the scale while negative s'ew tails off the low end of the scale.
7$rtosis is defined as a measure of the degree of pea'edness in the distribution1 relative
to its width. *he 'urtosis statistic will be Dero for a normal distribution1 positive for
pea'ed distributions and negative for flat distributions. A pea'ed distribution of the
errors is highly desirable. A high positive value of 'urtosis suggests that there are fewer
outliers in the estimation.
*he si% statistical indicators detailed above were used in the research to quantitatively
evaluate the performance of the model. *hese were calculated using a macro within
OBA and are summarised in *able >,!. E%treme range tests were also used to evaluate
the reliability of the laboratory macro model to predict bul' water temperature in the
$&(,(-..
T,54+ #)$* S:,:i0:i6,4 i.>i6,:2-0 v,47+0 B T2:,4 D,:,
&tatistical in%icator Ealues
9!=
-lope 1.01
;
$
0.33'
1B! <2= 40.30
;1-! <2= 1.00
-0ewness <2= 40.(3
)urtosis 3.$1
*he slope of the best fit line between computed and measured temperature was found to
be equal to 9."9 suggesting a slight over estimation of the computed variable. *he high
value of ?
!
indicates a low une%plained variation showing a good capability of the
regression model. *he BE value of ,".;" shows that the model has a tendency to over
predict its computed values. *he low ?(E value of 9." shows an acceptable actual
deviation. *he low negative s'ewed distribution tends to the lower end of the scale.
.owever1 its low value of , ".:# indicates a robust model. *he positive 'urtosis of ;.!9
indicates a pea'ed distribution of the errors suggesting that there are low outliers in the
estimation.
3inally e%treme range tests were underta'en to show the credibility of the model. 3or a
period of eight hours at constant heat flu% of =" and B""-Jm
!
good close fitting curves
of computed and e%perimental data were achieved with very low BE and ?(E
values of ,".>; and ".=; respectively suggesting a robust model.
(tatistical and e%treme range results validated the accuracy of the model to compute
bul' water temperature.
1.1.1.32 S:-,:i3i6,:i2. ) 82>+4 -+074:0
(esults
*his section e%plores the validity of the model predicting stratification inside the $&(,
(-. by comparing computed results with e%perimental data. *he simulation was
carried out for a wide range of heat inputs +="1 9""1 9="1 !""1 !="1 ;""1 ;="1 >"" and
B""-/ for a period of eight hours simulating a typical day of operation. &omputed
water temperatures at any given height were plotted as a function of e%perimental data
to e%amine the validity of the simulation. $t was demonstrated that e%perimental and
computer simulation results were in good agreement as they followed a 9F9 relationship
shown in 3igure >,9:.
9!B
0
10
$0
30
%0
&0
'0
(0
80
0 10 $0 30 %0 &0 '0 (0 80
' )easure%1 <C
'

c
o
)
0
u
t
e
%
1

<
C
Fi(. #)&&* C2897:+> v0 Ex9+-i8+.:,4 0:-,:i3i6,:i2. ,3:+- < 127-0 23 29+-,:i2. 32- ,44 1+,: i.97:0
$t is observed in 3igure >,9C that variation between e%perimental and computed results
data occurs during e%tended +!!hr/ model validation and e%perimental testing regime
displayed graphically by a deviation from the uniform 9F9 relationship. $t suggests that
as temperatures increase towards boiling1 bul' fluid movement is brea'ing down the
stratification. Based on the assumption that water temperatures to be achieved in a
(cottish climate are unli'ely to e%ceed :"& and that a typical solar heating day would
be of eight hours or less ma'e this model suitable for (cottish weather conditions.
9!:
0
10
$0
30
%0
&0
'0
(0
80
30
100
0 10 $0 30 %0 &0 '0 (0 80 30 100
' )easure%1 <C
'

c
o
)
0
u
t
e
%
1

<
C
Fi(. #)&<* C2897:+> v0 Ex9+-i8+.:,4 0:-,:i3i6,:i2. ,3:+- 22 127-0 32- ,44 1+,: i.97:0
*he stratification model was developed using e%perimental data for 9""1 !"" and
>""-Jm
!
. A detailed picture of the stratification model results for a heat input of
!""-Jm
!
applied to the system for up to eight hours is showed in 3igure >,9#. Each
wattage followed a 9F9 relationship showing good agreement between e%perimental and
computed data.
9!C
0
10
$0
30
%0
&0
0 10 $0 30 %0 &0
' )easure%1 <C
'

c
o
)
0
u
t
e
%
1

<
C
Fi(. #)&$* 2''W?8
2
>,:, ) C2897:+> v0 Ex9+-i8+.:,4 0:-,:i3i6,:i2. ,3:+- < 127-0 23 29+-,:i2.
'tatistical measures used for evaluation of the model
*he previously detailed statistical indicators were used to evaluate the performance of
the model to predict temperature stratification in the $&(,(-. and are summarised in
*able >,;.
T,54+ #)1'* S:,:i0:i6,4 i.>i6,:2-0 v,47+0 B T2:,4 D,:,
-tatisti+al indi+ator >alues
-lope 1.00
;
$
0.33
1B! <2= 40.%(
;1-! <2= 0.83
-0ewness <2= 40.1(
)urtosis 3.33
*he slope of the best fit line was found to be equal to one suggesting no over or under
estimation of the computed variable. *he high value of the coefficient of determination
+?
!
/ indicates a low une%plained variation and thus good correlation. *he BE value of
,".>: shows that the model has a tendency to over predict its computed values by nearly
half of one degree &elsius. *he low ?(E value of ".C; suggests a satisfactory actual
deviation. *he low s'ewness negative value of ,".9: is close enough to Dero to indicate
a robust model. *he positive 'urtosis value of ;.;# indicates a pea'ed distribution of the
errors suggestive of fewer outliers in the estimation.
9!#
Different heat inputs outside the normal e%perimental range were then simulated in
order to validate the model. ?esults for ="1 9="1 !="1 ;=" and B""-Jm
!
were plotted for
a typical day of eight hours in 3igure >,!". *he figure confirmed a good agreement
between e%perimental and computed data following a 9F9 relationship with an BE and
?(E of 9 and 9.> respectively and a slope and coefficient of determination of ".#:
suggesting an under estimation of the computed variable1 validating the stratification
method used.
0
10
$0
30
%0
&0
'0
(0
0 10 $0 30 %0 &0 '0 (0
' )easure%1 <C
'

c
o
)
0
u
t
e
%
1

<
C
&0#
1&0#
$&0#
3&0#
'00#
Fi(. #)<'* V,4i>,:i2. 23 :1+ 82>+4 ) C2897:+> v0 Ex9+-i8+.:,4 0:-,:i3i6,:i2. ,3:+- < 127-0 23
29+-,:i2.
1.1.1.33 Fi+4> ) 82>+4 -+074:0
*his section e%plores the validity of the model to generate and predict water bul'
temperature inside the $&(,(-. for field weather conditions. (imulations were
underta'en for an aluminium collector of ;mm thic'ness tilted at an inclination angle of
>= in order to compare computed results with e%perimental results underta'en on the
roof of Napier University as described in &hapter ;.
9;"
(esults
$n accordance with laboratory computed data1 the role of e%tended heat transfer fins was
demonstrated and showed good improvement by achieving higher bul' water
temperatures.
A first analysis carried out for a period of five days +99
th
, 9=
th
4uly !"":/ e%plores the
validity of the model to predict bul' water temperature inside the $&(,(-. by
comparing computed results with e%perimental data. ?esults are plotted in 3igure >,!9.
0
10
$0
30
%0
&0
'0
(0
1 ( 13 13 1 ( 13 13 1 ( 13 13 1 ( 13 13 1 ( 13 13
'i)e1 (rs
'
e
)
0
e
r
a
t
u
r
e
1

<
C
0
100
$00
300
%00
&00
'00
(00
800
300
1000
&
o
l
a
r

r
a
%
i
a
t
i
o
n
1

$
"
)
2
'w=8>0eri)ental 'w=Co)0ute%
'0=8>0eri)ental '0=Co)0ute%
'a &olar ra%iation
Fi(. #)<1* C2897:+> ,.> +x9+-i8+.:,4 >,:, 32- , 3iv+ >,/ 9+-i2> i. H74/ 2''&
3igure >,!9 shows that similar to e%perimental field results1 computed results show that
each day is characterised by a heating and a cooling profile which are dictated by solar
radiation. *he higher the incident solar radiation received by the absorber plate1 the
higher the water temperature computed. $t was observed that gradual heating and
cooling profiles due to the thermal mass of the system were e%perienced due to the
consideration of thermal mass in the macro model. .owever1 a lower cooling rate of the
modelled values was observed each day. *his could be e%plained by the propagation of
errors and uncertainty by the use of regression analysis and assumed weather data)
resulting in a decrease in accuracy of the model.
$t was also observed in 3igure >,!9 that e%perimental water temperature was followed
very closely by predicted water temperature although it showed a slight over,estimation
9;9
by the proposed model. $n order to e%amine the validity of the simulation computed data
were plotted as a function of e%perimental data in 3igure >,!!.
0
10
$0
30
%0
&0
'0
(0
0 10 $0 30 %0 &0 '0 (0
' )easure%1 <C
'

c
o
)
0
u
t
e
%
1

<
C
Fi(. #)<2* C2897:+> v0 Ex9+-i8+.:,4 32- , 3iv+ >,/ 9+-i2> i. H74/ 2''&
$t was demonstrated that e%perimental and computer simulation results were in good
agreement as they followed a 9F9 relationship shown in 3igure >,!!. (tatistical
indicators were then used to evaluate the performance of the model for those five days
and are resumed in *able >,>.
T,54+ #)11* S:,:i0:i6,4 i.>i6,:2-0 v,47+0
-tatisti+al indi+ator >alues
-lope 1.03
;
$
0.3(
1B! <2= 41.$(
;1-! <2= $.13
-0ewness <2= 0.8'
)urtosis 0.%$
*he slope of the best fit line suggests a slight over estimation of the computed variable.
*he high value of ?
!
indicates a low une%plained variation showing a satisfactory
accuracy for the regression model. *he BE value of ,9.!: shows that the model has a
tendency to over predict its computed values. *he ?(E value of !.9# shows an
acceptable actual deviation. *he low positive s'ewed distribution trends to the high end
9;!
of the scale and indicates a robust model. *he positive low 'urtosis indicates a pea'ed
distribution of the errors suggesting that there are low outliers in the estimation.
A study of the effect of wind velocity on the bul' water temperature was carried out and
is illustrated by 3igure >,!; where O"1 O!1 O> and OB represents the wind velocity
values of "mJs1 !mJs1 >mJs and BmJs respectively. *he higher the wind velocity1 the
higher the heat losses resulting in lower bul' water temperature showing the influence
of wind speed on overall performance. *hus1 having accurate wind speeds would result
in more accurate bul' water temperatures. *he more data input1 the more accurate
becomes the model.
$0
$&
30
3&
%0
%&
&0
&&
'0
'&
(0
0 8 1' 0 8 1' 0 8 1' 0 8 1' 0 8 1'
'i)e1 (rs
;
u
l
2

w
a
t
e
r

t
e
)
0
e
r
a
t
u
r
e
1

<
C
>0 >$
>% >'
Fi(. #)<3* Si874,:+> +33+6: 23 Ai.> v+426i:/ 2. A,:+- :+89+-,:7-+ 32- , 3iv+ >,/ 9+-i2> i. H74/ 2''&
A1+-+ V'; V2; V# ,.> V% -+9-+0+.:0 :1+ Ai.> v+426i:/ v,47+0 23 '8?0; 28?0; #8?0 ,.> %8?0
-+09+6:iv+4/
A second simulation was carried out over a period of =" days +:
th
4une , !:
th
4uly !"":/
in order to determine the validity of the model to predict bul' water temperature for an
e%tended period. *otal computed and e%perimental data are plotted in 3igure >,;> at the
end of the chapter.
&omputed data were then plotted as a function of e%perimental data to e%amine the
validity of the simulation. $t was demonstrated that e%perimental and computer
simulation results were in good agreement as shown in 3igure >,!>.
9;;
0
10
$0
30
%0
&0
'0
(0
0 10 $0 30 %0 &0 '0 (0
' )easure%1 <C
'

c
o
)
0
u
t
e
%
1

<
C
Fi(. #)<#* C2897:+> v0 Ex9+-i8+.:,4 32- , "' >,/ 9+-i2> i. H74/
*he same statistical indicators were then used to evaluate the performance of the model
to predict bul' water temperature in the $&(,(-. for the e%tended =" day test and are
given in *able >,=.
T,54+ #)12* S:,:i0:i6,4 i.>i6,:2-0 v,47+0
-tatisti+al indi+ator >alues
-lope 1.0%
;
$
0.3(
1B! <2= 41.%3
;1-! <2= $.$$
-0ewness <2= 0.$1
)urtosis 0.1'
*he slope of the best fit line suggests an over estimation of the computed variable. *he
high value of ?
!
indicates a low une%plained variation showing a good adequacy of the
regression model. *he BE value of ,9.># shows that the model has a tendency to over
predict its computed values by nearly 9.=S&. *he ?(E value of !.!! shows an
acceptable actual deviation. *he low positive s'ewed distribution tails off to the high
end of the scale while the positive low 'urtosis indicates a pea'ed distribution of the
errors suggesting that there are low outliers in the estimation.
9;>
?esults validated the field macro model as a robust model to predict bul' water
temperature in the $&(,(-. under different weather conditions.
#.1.< E--2- ,.,4/0i0
$nvestigating uncertainties and propagation of errors in the modelling procedure is
necessary to determine the quality of the data presented. Any sources of errors may be
categorised under the following two main categoriesF modelling uncertainties and
statistical disadvantages. A brief analysis of both categories is detailed below.
1.1.1.3# M2>+44i.( +--2-0
*he model was coded using numerous regressions that have a band of error. *he use of
a series of regressions results in the propagation and intensification in uncertainty of the
final result.

@ther uncertainties associated with assumptions made for the different type of model
and data used also intensify the uncertainty of the results. Assumption and data
uncertainties were analysed and are highlighted belowF
, Uncertainties of weather data recorded for simulation
, Uncertainty in material properties for calculation
, Uncertainty associated with the solar model of &lar'e et al +!"":/ used to incline
horiDontal solar radiation
, Uncertainty of estimation of the total enthalpy of the water when highly stratified
, Uncertainties due to air lea'age from the air cavity
, Uncertainty in calculation due to lateral stratification
1.1.1.3" Li8i:,:i2.0 23 0:,:i0:i6,4 +--2- i.>i6+0
Each statistical indicator has associated disadvantages which are important to address1
some of which are described belowF
, -ithin a dataset overestimation of one observation can over,compensate for another
+underestimate/1 ma'ing it difficult for BE to represent the true nature of the
accuracy of the model.
, ?(E fails to give a complete assessment of a model as only a few large errors can
increase its value substantially.
9;=
, ?
!
does not always measure the appropriateness of the model1 since it can be
artificially inflated by adding another parameter or higher order polynomial terms in
the regression equation.
, ('ewness does not account for the magnitude of errors) it only gives information on
the normality of their distribution.
, <urtosis can be very sensitive to outliers1 particularly if the data set involves
measured values. *hus1 only a few errors in the tails of the distribution1 resulting as a
deviation from erroneous or irrelevant observations1 can affect its value.
#.1.$ Di06700i2. 23 :1+-8,4 82>+40
1.1.1.3% L,52-,:2-/ 82>+4
A macro,model for fluid flow and heat transfer was developed using Oisual Basic for
Applications +OBA/. *he role of e%tended heat transfer fins was modelled and showed
good improvement by achieving higher bul' water temperatures. Effective fin heat
transfer rates were enhanced by the choice of a material of high conductivity such as
aluminium. *he effective use of the fin was found to be B:.;L and B;L for the
aluminium of ;mm and 9.=mm thic'ness and !=.;L for the 9.=mm thic' stainless,steel.
*he fin efficiency f

was then calculated to measure the fin performance and showed a

similar results pattern) with an improvement in efficiency achieved by using higher
conductivity material. Another measure of the fin performance was introduced by using
an improvement factor +$3/ representing the improvement achieved by using fins in the
$&(,(-.. Using a material of a high conductivity resulted in an overall improvement
of #.BL. *he computational results from the model were found to be in close agreement
with the e%perimental data. (atisfactory fitting curves were observed for the water
temperatures achieved e%perimentally and numerically for all the $&(,(-. types.
*esting of the results by statistical methods and e%treme range tests validated the
findings and the ability of the model to predict the water bul' temperature in the $&(,
(-..
*he laboratory macro,model was able to compare and predict the temperature variation
in diverse $&(,(-. configurations with different geometry1 number of fins1 internal
temperature and constant heat flu% for a given aspect ratio.
9;B
1.1.1.3& S:-,:i3i6,:i2. 82>+4
*he ob0ective of this study was to model stratification profiles with time in an $&(,
(-. sub0ected to laboratory conditions. *hree month full,scale e%perimental
laboratory tests were employed to validate the model.
*he stratification profile changes with time and therefore needed to be modelled to
effectively predict the collector performance. $mprovements to a previous laboratory
macro,model were incorporated. *his allowed longitudinal stratification of temperature
within the solar collector to be determined based on the mean water temperature given
by the macro model1 the given heat flu% and the total time of e%posure.
&omputational results from the model were found to be in close agreement with the
e%perimental data with a tendency of the model to over,predict computed values by
nearly half of one degree &elsius at high heat flu%es. ?esults were validated by using
the model with wattages not used in the development phase of the program. E%treme
range simulations showed the suitability of the programme to display temperature
stratification.
1.1.1.3< Fi+4> 82>+4
3urther wor' on the initial laboratory thermal model resulted in the development of a
thermal model suitable to predict bul' water temperature in real weather conditions.
(imulations for a = and =" day period were underta'en. Both simulations showed that
computational results were found to be in close agreement with the e%perimental field
measurements. *his statement was validated by statistical methods suggesting that the
field macro,model is a robust tool to compute bul' water temperature in the $&(,(-.
for any weather conditions.
*his model was able to compare the temperature variation and predict the bul' water
temperature in diverse $&(,(-. configurations1 with different geometry1 number of
fins and e%ternal weather conditions for a given aspect ratio.
Although this model only gives mean values of water temperature1 while in reality this
varies along the longitudinal height of the collector1 it gives a good estimation of the
$&(,(-. performance. ?esults for the =" day period also demonstrated that the
simulation can be e%tended which could result in a useful tool to estimate the yearly
performance of the $&(,(-..
9;:
#.2 Mi6-2 82>+4 B CFD
A &omputational 3luid Dynamic +&3D/ study was underta'en to model flow and heat
transfer in the three,dimensional +;D/ (-. geometry to provide useful information
about performance of the collector and evaluate design improvements. A first analysis
was underta'en for a four,fin collector designed as a mean of improving its
performance.
#.2.1 P7-920+ ,.> 3i. 29:i8i0,:i2.
*he &3D analysis of the $&(,(-. was underta'en in order to improve the four,fin
$&(,(-. performance by optimising the fin spacing. Prior to simulation1 issues
regarding the parameters influencing fin optimisation need to be determined1 constraints
stated1 heat transfer parameters outlined1 and the type of &3D analysis established.
3in material1 length1 thic'ness1 and the number of fins in the $&(,(-. are the four
main parameters to be defined when analysing fin optimisation. *he material thermal
conductivity is an intrinsic parameter for an effective $&(,(-.. *he low density1 high
thermal conductivity and recyclable properties of aluminium highlighted the suitability
of this material as already discussed in section ;.9.9.
*he fin length was fi%ed at a ma%imum of C""mm. 3ins are used to increase the heat
transfer from the heated surface by increasing the effective surface area. *he
effectiveness of the fin is enhanced by increasing the ratio of the perimeter to the cross,
sectional area +
c
A
/ as stated by $ncropera and Dewitt +!""9/ therefore the use of thin1
but closely spaced fins is preferred1 with the provision that the fin gap is not reduced to
a value where flow between the fins is severely impeded1 thereby reducing the
convection coefficient. *he study investigated the improvement in performance of the
$&(,(-. by increasing the number of fins.

*hree main constraints were ta'en into consideration for design improvement and
includedF cost1 volume +=" litre tan' siDe/ and manufacturing ability.
*hree main heat transfer parameters influence the $&(,(-. performance and can be
recapitulated asF the shape of fins1 the angle of inclination of the heater1 and time of
e%posure to incident solar radiation. Due to manufacturing constraints1 simple
rectangular shapes of fins were considered. A >=S inclination angle and a ;""-Jm
!
heat
flu% were ta'en as the reference conditions.
9;C
3inally1 the type of &3D analysis1 !D or ;D1 is an important parameter to consider. A
previous study of uneer et al +!""Ba/ outlined that !D analysis was sufficient for a
satisfactory analysis of the system. .owever1 !D analysis would only suffice for a
horiDontal inclination of the heater. 3or any angle above Dero a gradient e%ists in the
longitudinal direction ma'ing a !D analysis insufficient. .ence1 based on a >=S
inclination of the collector1 a ;D analysis was underta'en. As the quiescent fluid is
unavailable in &3D simulation1 the process was assumed transient.
#.2.2 C,9,5i4i:i+0 ,.> 4i8i:,:i2.0
&3D in simple terms is the use of computer and numerical methods to solve and analyse
problems involving fluid flow. $t is very often used in product concept1 product
development and also virtual prototyping as it provides complementary information
about environmental performances and significantly reduces the amount of
e%perimental wor' needed. &3D,3luent software1 based on the finite volume method
offering a wide array of physical !D and ;D models1 was used for this study. ?ecent
studies indicate an increase in the use of &3D tools for analysing and optimising design
and performances of solar collector5s. *his is despite some inaccurate predictions due to
the assumptions involved in modelling calculations stated by 7roenhout et al +!""!/.
*his model provided information about collector performance and was a useful tool for
design improvements providing detailed pictures of fluid movements and stratification
occurring in the collector.
Despite these advantages the &3D micro,model requires considerable s'ill to operate
thus requiring an e%tended training time and incurs high annual license and maintenance
costs. ?egardless of an increased speed of computation available1 &3D still cannot be
used for real time computation. Numerical analyses require considerable time to set up
and perform. 3inally1 software codes are loc'ed which ma'e them unavailable for
improvements
#.2.3 M2>+4 6,4i5-,:i2.
*he modelling process was carried out for ;D analysis using &3D,3luent B.! software.
7ambit !.! was used to mesh the model. *he simple and regular geometry of the model
suggested the use of a quad structured mesh of siDe 9mm based on .enderson et al
+!"":/. *wo different boundary condition types were applied on the absorber plate for
tests1 depending on the purpose of the simulation. @ptimisation simulations used a
constant heat flu% of ;""-Jm
!
as the boundary condition on the absorber plate. A
9;#
constant absorber plate temperature boundary condition was applied when comparing
the heat absorbed by the two $&(,(-. designs. According to these conditions the
program calculated the temperature field in the geometry and as a result values at the
nodes were displayed. ?esults are presented in section >.!.>.
#.2.# M2>+44i.( -+074:0
(tratification is essential for good performance of a solar collector and is significant
when trying to optimise the collector for draw,off as was observed in &hapter ;)
however1 better mi%ing in the tan' results in higher rates of heat transfer. 3urthermore1
the addition of fins might give a differential of temperature A lower from the top to the
bottom of the $&(,(.- A but will optimise the transfer of energy into water. A &3D
stratification analysis was developed for the four,fin $&(,(-. design and then
underta'en for a second design using five fins to compare both water stratification and
velocity magnitude for a given time and heat transfer through time.
1.1.1.3$ Fi. 1+,: :-,.03+-
A good understanding of the influence of e%tended surfaces is important when see'ing
performance enhancement of the $&(,(-.. 3igure >,!= represents the temperature of
the fins after an elapsed time of B" minutes and a simulated heat flu% on the absorber
plate of ;"" -Jm
!
. .eat is transferred through the fins from the top of the collector to
the bottom where the water temperature is cooler. *his results from gravity and
buoyancy effects. -ater with different temperatures will settle at a height in the $&(,
(-. corresponding to the density of the fluid. .ot water of low density will naturally
settle in upper layers while cold water of high density will fall to the bottom layers.
During the first three hours of the charging process1 high temperature stratification
occurs in the $&(,(-.. -ith time1 temperatures in the upper layers become fully
established and reach an equilibrium causing the lower layer to increase in temperature
and therefore decrease the density gradient of water inside the $&(,(-..
9>"
Fi(. #)<"* T+89+-,:7-+ 0:-,:i3i6,:i2. 23 3i.0 ,.> 8i>>4+ 4i.+ 23 A,:+- 6244+6:2-
$n order to ease the comprehension of the analysis process a discussion of the initial
four,fin collector design is discussed below.
1.1.1.#' F27-)3i. 6244+6:2- >+0i(.
*he display of the water velocity profile was completed in 3igure >,!B. Oelocity was
observed to increase with the longitudinal length and then decrease as it gets closer to
the top of the collector. *his can be e%plained by the accumulation of hot water at the
top of the collector reducing the velocity boundary layer and promoting diffusion of
heat with time. $t is observed in 3igure >,!B that an increase1 followed by a decrease1
pursued by an increase in velocity occurs between the fins. *his decline in velocity
between the fins suggests that additional fins could be incorporated in the design to
increase the velocity magnitude.
9>9
Fi(. #)<%* V+426i:/ 9-23i4+ B # 3i.0; :29 vi+A; ,3:+- 2' 8i.
*he horiDontal plane view of velocity profile in 3igure >,!: supports this observation
and shows that high velocities occur at the absorber plate where higher temperatures are
e%perienced.
Fi(. #)<&* V+426i:/ 9-23i4+ B # 3i.0; 0i>+ vi+A; ,3:+- 2' 8i.
9>!
.igher velocities
.igher
velocities
?esults showed that the overall velocity increased with time until it reached a pea'
value from which velocity magnitude declines with the passage of time. *he increase is
due to a high heat flu% through the fins to the water occurring after the initial
capacitance effect of the system. -ith time1 hot water builds up at the top of the
collector as shown in 3igure >,!C resulting in a decline of heat transfer and therefore
heat is diffused from the system.
Fi(. #)<<* L2.(i:7>i.,4 A,:+- :+89+-,:7-+ 0:-,:i3i6,:i2. B # 3i.0; 3-2.: vi+A; ,3:+- 2'8i.
3igures >,!C and >,!# represent the longitudinal temperature stratification in the
collector inclined at >=. As predicted1 higher temperatures occur in the top of the
collector. (tratification occurs for a ma%imum water temperature of !#=.#< at the top
of the collector to a minimum temperature of !#;.;< at the bottom) a !.B<
temperature gradient.
9>;
Fi(. #)<$* L2.(i:7>i.,4 A,:+- :+89+-,:7-+ 0:-,:i3i6,:i2. B # 3i.0; 0i>+ vi+A; ,3:+- 2'8i.
$n an attempt to improve the heat transfer one fin was added to the original $&(,(-.
design. New ;D velocity and stratification analysis were underta'en for the five,fin
$&(,(-. configuration.
1.1.1.#1 Fiv+)3i. 6244+6:2- >+0i(.
3igures >,;" and >,;9 display the velocity profile of the new $&(,(-.. An increase in
velocity1 compared to the four,fin design was observed resulting in an increase of !L in
the ma%imum velocity. *he main velocity pattern occurs between 9.#"mmJs to ;.CmmJs
for the five,fin collector compared with a 9.CBmmJs to ;.:!mmJs for four,fin design.
9>>
Fi(. #)$'* V+426i:/ 9-23i4+ B " 3i.0; :29 vi+A; ,3:+- 2' 8i.
Fi(. #)$1* V+426i:/ 9-23i4+ B " 3i.0; 0i>+ vi+A; ,3:+- 2' 8i.
*he longitudinal temperature stratification of the five fin collector is shown in 3igure >,
;!. (tratification occurs from a ma%imum water temperature of !#=.#< at the top to a
9>=
minimum temperature of !#;.C< at the bottom of the collector1 resulting in a !.9<
temperature difference. *his corresponds to a 9#L decrease in stratification.
Fi(. #)$2* L2.(i:7>i.,4 A,:+- :+89+-,:7-+ 0:-,:i3i6,:i2. ) " 3i.0; 0i>+ vi+A; ,3:+- 2'8i.
1.1.1.#2 I89-2v+8+.:0
Previous observation showed that the five,fin collector achieved an increase in velocity
of !L compared to the four,fin design1 while longitudinal water temperature
stratification decreased by 9#L. *his could be e%plained by the addition of the fins
responsible for an increased transfer of heat inside the collector1 thus increasing velocity
and overall temperature in the collector.
$n order to demonstrate that five fins is a more suitable geometry arrangement for
optimal collector design1 a ;D &3D analysis was carried out assuming a constant
temperature boundary condition on the absorber plate set at ;9;<. *he fluid departure
temperature was set at !#;<. Despite the fact that the system is under a constant heat
flu%1 using a constant input temperature allowed the heat absorbed by both collectors to
be derived as shown in 3igure >,;;.
9>B
$3%
$3'
$38
300
30$
30%
30'
308
310
31$
31%
0 10 $0 30 %0 &0 '0
'i)e1 )n
,
*
e
r
a
.
e
%

w
a
t
e
r

t
e
)
0
e
r
a
t
u
r
e
1

<
C
0
&00
1000
1&00
$000
$&00
3000
3&00
%000
%&00
$
a
t
t
a
.
e

t
r
a
n
s
)
i
t
t
e
%
1

$
"
)
2
% fins 4 a7erage Tw
& fins 4 a7erage Tw
% fins 4 heat transittan+e
& fins 4 heat transittan+e
Fi(. #)$3* H+,: :-,.03+- ,.> A,:+- :+89+-,:7-+ 9-23i4+0 i. :1+ ICS)SWH
Assuming a constant temperature boundary condition on the absorber plate resulted in
high rates of heat transmittance. *his figure allows a comparison of the rate heat is
absorbed by both collectors. $nitially1 for the first ten minutes1 more heat was absorbed
by the five,fin collector compared to the four,fin. *his is lin'ed to a frenDy of heat
transfer activity as the fins are surrounded by cooler water. After ten minutes the five
fins are surrounded with warmer water than the four fin design. *he difference in
temperature between the four fins and the water is higher resulting in a higher heat flu%.
Both profiles reach a pea' value at the equilibrium temperature ;9;< set as the
boundary condition.
#.2." Di06700i2. 2. CFD
*his section reported the implementation of a four,fin $&(,(-. concept utilising &3D,
3luent software through 7ambit in order to optimise its design performance. $nitial
results of the four,fin $&(,(-. indicated that one fin could be added to the original
design to improve the heat transfer. A ;D &3D simulation was then underta'en for a
five,fin $&(,(-.. *wo boundary conditions were applied to the systems in order to
compare the water temperature stratification achieved and the heat transferred to the
water body by each collector.
*he first boundary condition applied to the systems consisted of a constant heat flu% on
the absorber plate. *his method was used to characterise the temperature stratification
9>:
and the velocity magnitude within both collectors. Despite a minor decrease in
temperature gradient between the top to bottom in the five,fin $&(,(-.1 it was clearly
observed that stratification remained. $t is important to state that the addition of fins
should not impede the flow between the fins +as was e%perienced when trialling a multi,
fin design in an earlier study/1 thereby reducing the convection coefficient as stated by
$ncropera and Dewitt +!""9/. ?esults showed that the addition of one fin in the collector
increased the velocity in the collector which has a corresponding increase on the Nusselt
number) raising the heat transfer coefficient in a predictable manner as stated by @Doe
+9#:>a/. *he velocity magnitude was also observed to decrease with time for both
collectors with the water becoming warmer.
A second boundary condition applied to the systems consisted of a constant temperature
on the absorber plate which was used to characterise the heat absorbed by each
collector. ?esults revealed that the addition of one fin accounts for the increased transfer
of heat inside the collector. $t was observed that the five fins supplied more energy to
the collector than four at the beginning of the charging process. *he intermittent
availability of incident solar radiation in (cotland shows high potential for this type of
improvement as it is advantageous to have a rapid heating process.
$n light of the results presented in this paper1 the five,fin collector performed generally
better than the original four,fin collector. *herefore1 this new design could be suggested
as a new arrangement of the collector assuming that the cost associated with this
improvement is negligible.
#.3 C2.647>i.( -+8,-C0
*his chapter e%plored the modelling of $&(,(-.5s in order to assess their performance.
*o start with1 two macro,models for fluid flow and heat transfer were developed using
Oisual Basic for Applications +OBA/. Both generated corresponding bul' water
temperature5s in the collector with given hourly incident solar radiation +-Jm
!
/1
ambient temperature and inlet water temperature and therefore predictable $&(,(-.
performances. *he first model was developed for laboratory conditions and modelled
bul' water temperature and stratification profiles over time in an $&(,(-. while the
second model predicted the bul' water temperature in the $&(,(-. for field weather
conditions.
(tatistical methods and e%treme range tests suggested that both models were robust and
could predict the performance of an $&(,(-. at any location and with different
installation configurations. *he field macro,model also demonstrated that a simulation
9>C
could be e%tended which could result in a useful tool to estimate the perennial
performance of the $&(,(-.. A life cycle assessment +2&A/ based on this statement
was then carried out and is presented in &hapter =. An account of uncertainties and
propagation of errors associated with modelling and the accuracy of statistical methods
used was addressed.
A &3D,3luent analysis of the four,fin $&(,(-. concept was simulated in order to
optimise the design performance of the original concept and resulted in a new
simulation for a five,fin $&(,(-.. *he new design showed high potential for (cottish
applications due to its rapid heating process which is advantageous in locations where
availability of incident solar radiation is intermittent. *herefore1 this design could be
suggested as a new arrangement of the collector assuming that the cost associated with
this improvement is negligible.
$n conclusion1 this section of the thesis opens an avenue for future $&(,(-. design
optimisation but also widens the utilisation of the field macro,model developed by
showing its potential to provide data for a full year5s performance and therefore provide
an effective life cycle assessment of the $&(,(-..
9>#
0
10
$0
30
%0
&0
'0
(0
'i)e
'
e
)
0
e
r
a
t
u
r
e
1

<
C
Tw6!8periental
Tw62oputed
Fi(. #)$#* T2:,4 >,:, ) 8+,07-+> ,.> 62897:+> 32- , "' >,/ 9+-i2>* &
:1
H7.+ B 2&
:1
H74/ 2''&
9="
CHAPTER " LIFE CCLE ASSESSMENT (LCA)
$ndustries apply 2ife &ycle Assessment +2&A/ as a technique to identify significant
environmental impacts of their products in order to improve their environmental
performance during their whole life cycle. *his chapter performs a streamlined 2&A of
different types of $&(,(-. designed for use in a (cottish environment. 3irstly1 the goal
and scope of the 2&A were described. A material inventory providing general
information regarding material production processes used in the $&(,(-.s was
provided. An energy analysis of the $&(,(-.s using embodied energy of materials1
complemented by their associated environmental impacts were then dealt with. 3inally1
a monetary analysis +A/ was provided enabling a cost comparison of the different
$&(,(-. alternatives identifying their value per pound (terling spent.
".1 G2,4 ,.> S629+
*he first step is to define the reference metric used in this study to quantify performance
of the $&(,(-.. $nternational @rganiDation for (tandardiDation , $(@ +!""B/
demonstrates that a reference unit in any 2&A needs to be determined in order to
compare collected data from different studies of the same product category. *his study
is carried out to determine the best $&(,(-. alternative between three $&(,(-.s
presented for analysis. Based on the wor' of Ardente et al +!"";/1 a reference unit +or
3unctional Unit/ equal to the entire equipment was used to benchmar' this analysis.
*he three $&(,(-.s systems assessed in this study correspond to two aluminium
construction of 9.= and ;mm thic'ness +A2,9.= and A2,;/ and one of stainless,steel of
9.=mm +(*,9.=/ incorporating e%tended heat transfer surfaces. Each $&(,(-. is
composed of three main components identified below and shown in 3igure =,9F
9. *he collector +including the absorber plate and the water tan'/
!. *he glass wool insulation
;. *he glass cover
>. *he wooden bo%
9=9
Fi(. ")$"* S+6:i2.,4 vi+A 23 :1+ ,478i.i78 ICS SWH. N2:+* , >i8+.0i2. >-,Ai.( i0 012A. i.
Fi(7-+ 3)1
*he collector was categorised as a passive $&(,(-.. *he water tan' and absorber plate
are integrated into one unit. *his type of collector is recommended for medium to low
temperature domestic water heating applications. *he $&(,(-.s were assumed to be
directly installed on sloping roofs of >= angle of inclination. $mpacts associated with
each collector component were analysed in order to identify the main life cycle issues
affecting each component of the $&(,(-.s.
*he 2&A of each $&(,(-. type will provide an analysis of the environmental benefit
of operating these heaters in (cottish weather conditions in order to appreciate if this
tool can identify a best $&(,(-.s alternative. A material inventory introducing the
main component production processes was first carried out to understand their potential
harms to the environment and their energy requirements during production. An energy
and environmental impact analysis for each component was then performed.
$nterpretation of results by introducing energy savings and carbon emission reduction
followed by a monetary saving analysis were determined. Assumptions for the analysis
included a typical water inlet temperature of 9!& for (cotland and a domestic hot
water +D.-/ demand temperature of ==&1 assuming a daily draw,off regime of water
at >pm. @n days of poor insolation either electric or hydrocarbon bac',up sources
elevate the water temperature to ==S&1 this is accounted in the 2&A and cost analysis.
*he D*$ +!""!/ stated that the average population of a U< home is appro%imately three
people1 with each consuming =" litres of hot water per day as identified in a later survey
carried on behalf of D*$ +!""=/ and DE3?A. 7iven that the capacity of the $&(,(-.s
is =" litres with 9m
!
of collector plate1 all interpretations were based on the installation
of ;m
!
of collector plate or 9=" litres to meet typical domestic demand from an
installation.
7lass &over
7lass wool insulation &ollector
-ooden bo%
9=!
".2 M,:+-i,4 i.v+.:2-/
*he following section discusses the processes involved in production of the main
materials used in the construction of the $&(,(-.s.
".2.1 S:,i.4+00)0:++4
*here are a number of stages involved in producing stainless,steel for commercial use.
(tainless,steel is produced from ore and scrap metal. *he principal ore used is hematite
which contains almost :"L iron. (crap iron and stainless,steel and various alloys of
chromium +and nic'el1 molybdenum etc. depending on the stainless type/ are used in
stainless,steel ma'ing today. *he main process used at present for refining stainless,
steel is the electric arc furnace used to produce carbon and alloy steels. A current is
passed through an electrode resulting in a temperature increase to a point where the
scrap and alloys melt. *he molten material is then transferred into an Argon @%ygen
Decarbonisation container1 where the carbon levels are reduced and the final alloy
additions are made to ma'e the composition chemistry depending on the type of
stainless,steel.
(tainless,steels are a group of high alloy steels that are designed to provide high
corrosion resistance. *his is provided by the thin o%ide film which forms due to the
chromium element in the stainless,steel reacting with o%ygen in the atmosphere. *he
stainless,steel ;"> used in this study is classified as an austenitic steel with an alloy
containing 9C,!"L of chromium and C,9"L of nic'el resulting in a very ductile
material and a good corrosion resistance as stated by 7roover +!""B/.
".2.2 A478i.i78
$n a similar fashion1 aluminium is produced from ore +Bau%ite/ and scrap metal. Bau%ite
is mined and refined in order to e%tract aluminium for use. *he ore is first washed and
crushed into a fine powder. *he aluminium is then e%tracted from the bau%ite as
aluminium o%ide +alumina/ which is subsequently bro'en down and separated into
aluminium and o%ygen gas using an electrolysis process also called smelting. *he
aluminium is then alloyed with other materials and is then processed through casting1
rolling or e%trusion for further fabrication of aluminium products.
Aluminium is one of the newer materials among the principal metals on the mar'et. $ts
low density made it an attractive choice use in engineering. Aluminium also has high
electrical and thermal conductivity and an e%cellent corrosion resistance due to the thin
9=;
layer of aluminium o%ide that forms on its surface when e%posed to o%ygen. $t is a very
ductile material and although it is 'nown to be relatively low in strength it can be
alloyed and heat treated to allow it to compete with steels.
".2.3 G4,00 A224 i.074,:i2.
7lass,wool is produced from raw materials such as sand1 recycled glass1 soda ash1
limestone and glass,wool scrap from the production process. ?eusing of waste off,cuts
into the production process reduces raw material input and therefore energy
requirements.
?ecycled glass from building windows1 car glass and bottle glass is increasingly used in
the manufacturing of glass wool accounting for ;"L to B"L of the raw material input.
*he raw material mi% is then melted at high temperature ranging from 9;""S& to
9=""S& in an electric or gas furnace while fumes and dusts are cleaned by filters to
minimise environmental impact.

*he melted glass is then converted into fibres with its passage through a spinner drilled
with holes and through a fibre attenuation using high temperature gas 0et streams. (mall
quantities of binding agents are then added to the fibres in order to give the desired
structure and density required for its end,use application. *he glass wool mat finally
goes to a curing oven heated at around !""S& where it gets its final shape1 strength and
stability1 and where it is laminated and polymerised. *he glass wool is then cut in width
and length at the required dimensions according to its final usage. *he product is then
tightly compressed in order to reduce the product volume in the pac'aging resulting in
an easier and cheaper storage and transport of the product. 3igure =,! represents the
glass wool production process illustrated by the European $nsulation anufacturers
Association , E?$A +!"":/.
9=>
Fi(. ")$%* G4,00 A224 9-2>76:i2. 9-26+00; S27-6+* ERIMA (2''&)
".2.# Wi.>2A (4,00
*he window glass is produced using the float glass method. ?aw materials such as soda
lime glass which principally contains silica sand1 soda ash and limestone are mi%ed and
weighed and then flow into a furnace heated to 9=""S&. elting1 refining and
homogenising ta'e place simultaneously in the furnace and determine the glass quality.
&ompositions of the glass can be modified during this process to change the properties
of the finished product such as adding magnesia to help reduce devitrification as stated
by 7roover +!""B/.
*he molten glass is then fed onto a float bath where it is floated on top of a bath of
molten tin in a continuous ribbon and starts to cool slowly. *he ribbon then goes
through a coating stage through chemical vapour deposition laying down a variety of
coatings to vary the optical properties of float glass depending on the final application.
*he ribbon then passes through an annealing process where glass is slowly cooled in a
2ehr oven to prevent build up of stress. *he glass is cooled under controlled conditions
9==
until it reaches room temperature and is then cut to the desired shape and siDe. 3igure =,
; represents the window glass production process illustrated by ade .ow +!"":/.
Fi(. ")$&* Wi.>2A (4,000 9-2>76:i2. 9-26+00; S27-6+* M,>+ H2A (2''&)
".2." Ti85+-
*he timber frame was produced from U< grown mi%ed conifer logs through a sawmill
process. $n the first stage the grading process pre,sorts conifer logs into siDe and end
use1 before input to the sawmill. *he logs then go through a debar'er prior to sawing to
remove bar' from logs. *he bar' waste is then reused as raw material for horticulture.
(ubsequently debar'ed logs go through the sawmill to produce dimensional lumber.
*he cut products are then sorted through infra,red scanning equipment ensuring that
only product of the required grade is included in the finished pac'. *he timber is then
cut to the desired lengths by multiple cross cut machinery for the pac'aging process.
3inally1 preservative treatment is applied to timber products to ensure long term
protection and service life of the timber.
".3 E.+-(/ ,.,4/0i0
*he energy analysis concerns energy flows occurring during the life cycle of the
product. *he energy consumption can be split in two phasesF Direct energy and
Embodied energy. Direct energy corresponds to the energy used during a life cycle step
including electricity or heat energy employed during the production of the collector or
9=B
fuel for transport. Embodied energy corresponds to the energy consumed by all the
processes associated with the production of the individual materials employed.
*he total energy consumption is obtained by summing embodied and direct energy
contributions. .owever1 Ardente et al +!""=a/ showed that direct energy consumption
only account for 9"L of the overall energy consumption. *herefore1 the streamlined
2&A only accounts for embodied energy involved in the production processes of the
$&(,(-.s materials.
Embodied energy assesses the total primary energy consumed over the whole life time
of a product with the boundaries ideally being set from the e%traction of raw materials
to the end of the product5s lifetime) a boundary condition which is 'nown as Mcradle to
graveN. Data regarding the embodied energy of materials refers to the $nventory of
&arbon and Energy report issued by .ammond and 4ones +!""B/. Embodied energy of
each material presented in *able =,9 is based on average U< mar'et data.
T,54+ ")13* E852>i+> +.+-(/ 23 8,:+-i,40
M,:+-i,4 E852>i+> +.+-(/ (MH?C()
(tainless,steel =9.=
Aluminium 9=".!
7lass 9;.=
*imber C.:
7lass wool !C."
*he aluminium embodied energy production was found to be 9=".! 4J'g
corresponding to an average of ;;L recycled and B:L raw aluminium material. 3uture
scenarios could be imagined of C=L recycled and 9=L raw aluminium material which
would reduce the embodied energy of aluminium production process to =;.:4J'g.
A brea'down of the materials used to produce the final $&(,(-.s was quantified in
order to calculate their respective overall embodied energy. *able =,! reports the mass
balance of a 9m
!
collector summarising the quantity of materials required for each $&(,
(-..
T,54+ ")1#* M,00 5,4,.6+ 23 :1+ 6244+6:2-0
C244+6:2- :/9+ M,:+-i,4 M,00 (C()
S:,i.4+00)0:++4 (tainless,steel +9.=mm sheet/ !C."
1."88 7lass 9!.B
*imber ;".:
9=:
7lass wool ;.=
A478i.i78 Aluminium +9.=mm sheet/ #.C
1."88 7lass 9!.B
*imber ;".:
7lass wool ;.=
A478i.i78 Aluminium +;mm sheet/ 9#.B
388 7lass 9!.B
*imber ;".:
7lass wool ;.=
*he embodied energy and effect on the energy balance of materials involved in the
construction of the $&(,(-.s were then calculated) hence the overall embodied energy
of the $&(,(-.s was determined. *able =,; provides information gathered on each
$&(,(-..
T,54+ ")1"* E852>i+> +.+-(/ ,.> i.6i>+.6+ 23 8,:+-i,40 2. :1+ +.+-(/ 5,4,.6+
C244+6:2- :/9+ M,:+-i,4 E852>i+> +.+-(/ (MH) I.6i>+.6+ (K)
S:,i.4+00)0:++4 (tainless,steel +9.=mm sheet/ 9>>!." :;."
1."88 *imber !B=.: 9;.=
7lass 9B#.: C.B
7lass wool #:.! >.#
T2:,4 1$&#.% 1''.'
A478i.i78 Aluminium +9.=mm sheet/ 9>:!." :;.>
1."88 *imber !B=.: 9;.;
7lass 9B#.: C.=
7lass wool #:.! >.C
T2:,4 2''#.% 1''.'
A478i.i78 Aluminium +;mm sheet/ !#>;.# C>.:
388 *imber !B=.: :.B
7lass 9B#.: >.#
7lass wool #:.! !.C
T2:,4 3#&%." 1''.'
Based on the earlier discussed D*$ +!""!/ study in section =.91 results have to be
translated to ;m
!
of collector plate or 9=" litres1 to reproduce an appropriate study for
U< homes. Primary energy consumption1 measured in '-h1 for the production of ;m
!
$&(,(-.s during the complete life cycle phase1
L(A
.
1 is represented in *able =,>.
T,54+ ")1%* E.+-(/ ,.,4/0i0 32- , 38
2
i.0:,44,:i2.
C244+6:2- :/9+ M,:+-i,4 E852>i+> +.+-(/ (CW1)
S:,i.4+00)0:++4
1."88
(tainless,steel +9.=mm sheet/ 9B>=.=
A478i.i78
1."88
Aluminium +9.=mm sheet/ 9B:".>
9=C
A478i.i78
388
Aluminium +;mm sheet/ !C#:.9
*he energy analysis provided the total energy and material consumption for the
production of each $&(,(-.. *his analysis enables the evaluation of comparative
environmental impacts and potential improvements which are provided in the following
sections.
".# E.vi-2.8+.:,4 i89,6:0
*his section aims to provide an understanding and evaluation of the magnitude and
significance of the potential environmental impacts of each $&(,(-.. ?esource
consumption1 air emissions1 water emissions1 waste and solid pollutants are the main
environmental impacts. .owever1 the streamlined 2&A only accounted for the
environmental impact of air emissions in order to ease the system analysis.

*his environmental impact can be classified as either a direct or indirect emission.
Based on Ardente et al5s +!""=a/ study1 direct emissions were neglected as they
generally represent an overall incidence of 9"L. *herefore1 only indirect emissions
were considered. $ndirect emissions are mainly related to the raw materials production
process.
Based on the $P&& +!"":/ report1 increased industrial activity and other human
activities have increased &@
!
concentrations in the atmosphere. &@
!
being a greenhouse
gas1 this could result in an increase in mean global temperature. Based on this statement
&@
!
was chosen as the indicator of this category and embodied &@
!
associated with the
raw material production process of each $&(,(-.s was determined. Embodied &@
!

refers to the amount of carbon dio%ide emissions that result from the processes required
to produce the finished product. Data regarding the embodied carbon of materials refers
to the $nventory of &arbon and Energy report issued by .ammond and 4ones +!""B/.
Embodied &@
!
emissions of each material are presented in *able =,> and refer to U<
average mar'et data.
T,54+ ")1&* E852>i+> 6,-52. >i2xi>+ 23 8,:+-i,40
M,:+-i,4 E852>i+> CO2 (C(CO2?C()
(tainless,steel B.!
Oirgin aluminium C.>
7lass ".C
9=#
*imber ".=
7lass wool 9.>
*he embodied carbon U< average data of aluminium production was found to be C.;=
'g&@
!
J'g. 3uture scenarios having C=L recycled and 9=L raw aluminium material
would bring the embodied carbon down to ;.9: 'g&@
!
J'g.
Based on the figures from *able =,! the embodied carbon and effect on the energy
balance of materials involved in construction of the $&(,(-.s was then calculated)
hence the overall embodied carbon of the $&(,(-.s was determined. *able =,=
provides information gathered on each collector.
T,54+ ")1<* E852>i+> +.+-(/ ,.> i89,6:0 23 8,:+-i,40 2. :1+ +.+-(/ 5,4,.6+
C244+6:2- :/9+ M,:+-i,4 E852>i+> CO2 (C(CO2) I89,6:0 (K)
S:,i.4+00)0:++4 (tainless,steel 9:!.! C=.B
1."88 *imber 9>.B :.;
7lass #.: >.C
7lass wool >.: !.;
T2:,4 2'1.2 1''
A478i.i78 Aluminium C9.C :;.C
1."88 *imber 9>.B 9;.!
7lass #.: C.C
7lass wool >.: >.!
T2:,4 11'.< 1''
A478i.i78 Aluminium 9B;.: C=."
388 *imber 9>.B :.B
7lass #.: =."
7lass wool >.: !.>
T2:,4 1$2.& 1''
Based on the D*$ +!""!/ study the environmental impacts were considered for ;m
!
of
collector plate or 9=" litres. *otal &@
!
emissions1
L(A
.(
1 related to the production of
;m
!
$&(,(-.s are represented in *able =,:.
T,54+ ")1$* T2:,4 CO2 +8i00i2.0 32- , 38
2
i.0:,44,:i2.
C244+6:2- :/9+ M,:+-i,4 E852>i+> CO2 (C(CO2)
S:,i.4+00)0:++4
1."88
(tainless,steel +9.=mm sheet/ B";.B
A478i.i78
1."88
Aluminium +9.=mm sheet/ ;;!.>
A478i.i78
388
Aluminium +;mm sheet/ =::.#
9B"
"." I.:+-9-+:,:i2.
".".1 I>+.:i3i6,:i2. 23 E.+-(/ ,.> E.vi-2.8+.:,4 i007+0
An analysis of the energy and environmental impacts developed in sections =.; and =.>
was performed in order to identify and evaluate the significant issues based on the
above results. $t is a valuable step in order to compare the different production
alternatives and which material input has the most impact on $&(,(-. construction.
1.1.1.#3 E.+-(/ i007+0
*he embodied energy related to the collector production depends on the material type
and mass and is the main contributor to the $&(,(-.s overall embodied energy. *he
embodied energy associated with each material depends on the quantity needed and the
energy intensity of the material5s production.

Although the embodied energy per 'ilogram of aluminium is BBL higher than stainless,
steel1 the overall embodied energy of the collector1 for an identical material thic'ness1 is
of the same magnitude as shown in *able =,;. *his is e%plained by the B=L lower
aluminium density of !::"'gJm
;
compared with :#""'gJm
;
for stainless,steel.
$t was also observed from *able =,9 and =,; that from the three remaining materials
+glass1 timber and glass,wool/1 glass wool has the highest embodied energy per
'ilogram but has the smallest effect on the overall $&(,(-. embodied energy due to
its low mass. Oice versa the overall embodied energy of timber was found much higher
than glass,wool. *his is e%plained by the timber mass being nine times that of glass,
wool. (imilar observations were made for the glass compared with timber and glass,
wool materials.
1.1.1.## E.vi-2.8+.:,4 i007+0
*he embodied carbon emissions associated with each material depends on the &@
!

released during their production) depending essentially on the type of fossil fuels used
and the amount of material needed.
Although the embodied carbon emissions per 'ilogram of aluminium are !BL higher
than the stainless,steel) as shown in *able =,=1 the overall embodied carbon emissions
of each collector1 for an identical material thic'ness1 is >=L lower by using aluminium
9B9
as shown in *able =,B. $t was also noted that even by doubling the thic'ness of the
aluminium +A2,;/1 the embodied carbon emissions are still >L lower than the (*,9.=.
*his can be e%plained by the recyclable property of aluminium. *his verifies the
statement made in the above section =.=.9.91 showing that despite the higher embodied
carbon emissions per 'ilogram of aluminium1 the lower mass of aluminium needed to
produce an identical collector resulted in lower overall embodied carbon emissions.
Each material was then weighed to calculate their contribution to the overall carbon
emissions of each $&(,(-.. Again1 it was observed from *able =,= and =,B that from
the three remaining materials +glass1 timber and glass,wool/1 glass wool had the highest
embodied carbon emissions per 'ilogram but resulted in the smallest effect on the
overall $&(,(-. embodied carbon emissions due to its low mass. Oice versa timber
had the lowest embodied carbon emissions per 'ilogram despite having the biggest
effect on the overall $&(,(-. embodied carbon emissions due to its higher mass.
".".2 ICS)SWH0 +.+-(/ ,.> 6,-52. >i2xi>+ 0,vi.(0
A cost,benefit analysis of the energy and environmental parameters of each $&(,(-.
was performed in this section. 3or this analysis the useful energy saved annually1
S
.
1
and &@
!
emission savings1
S
.(
1 need to be determined.
1.1.1.#" S/0:+8 27:97:
*he macro,model developed in &hapter > for field weather conditions was used to
generate corresponding bul' water temperature in the collector for each $&(,(-.. Data
regarding the hourly solar radiation and other climatic parameters necessary for the
simulation come from the European &ommission 4oint ?esearch &entre1 $nstitute for
Environment and (ustainability +$E(/ +!"":/ for Edinburgh location. *he daily1
monthly and annual energy savings and &@
!
emission savings were then determined
assuming a typical inlet water temperature of 9!& for (cotland and a required
temperature of ==& supposing a draw,off of water at >pm daily.
*he daily potential energy outputs from using the $&(,(-.s were calculated using
equation =.9.
w
p w
( T m "
+=.9/
9B!
-here
w
T
is the inletJoutlet temperature difference of the water +&/
in w
T
1 is the assumed inlet temperature of the water of 9!& +&/
o$t w
T
1 is the modelled output water temperature at >pm +&/
m
is the daily water flow rate assumed at 9="lJday for ;m
!
collector area
w
p
(
is the specific heat capacity of the water +4J'g.&/
*he variation of the specific heat capacity of water over this temperature range is
negligible with a ma%imum variation of about ".>L and was therefore assumed constant
at >.!'4J'g.&.
onthly and useful energy saved annually1
S
.
1 were then calculated for each $&(,
(-.. 3igure =,> presents useful energy saved annually by each $&(,(-.. *his figure
was based on a ;m
!
collector area installation heating 9=" litres of water per day from a
starting temperature of 9!S& to the required temperature of ==S& in Edinburgh weather
conditions.
1109
1107
1041
0 $00 %00 '00 800 1000 1$00
)
s
; CW1
,lu)iniu)
collector = 3))
,lu)iniu)
collector = 1.5))
&tainless=steel
collector = 1.5))
Fi(. ")$<* U0+374 +.+-(/ 0,v+> ,..7,44/ 32- +,61 ICS)SWH
*he annual energy saving from using A2,;1 A2,9.= and (*,9.= $&(,(-.s were found
to be 99"#'-h1 99":'-h1 9">9'-h respectively. Based on the previous assumption1
!:>:'-h is annually required to raise the daily 9=" litres of water from 9!S& to ==S&.
*he $&(,(-.s did not raise the water by the required amount but partially fulfil the
daily domestic hot water needs. A2,;1 A2,9.= and (*,9.= $&(,(-.s could save up to
>".>L1 >".;L and ;:.#L of the required energy. Using a more conductive material
9B;
resulted in a !.>L annual increase of energy savings while increasing the material
thic'ness resulted in only ".9L annual increase.
A detailed picture of the monthly energy coverage +E&/ of each $&(,(-. is provided
in 3igure =,=. *he energy coverage represents the percentage ratio of the energy saved
and energy required to raise the 9=" litres of water from 9!S& to ==S&.
0
10
$0
30
%0
&0
'0
(0
80
?an ,eb 1ar .pr 1ay ?un ?ul .ug -ep @+t Ao7 /e+
+ont(
&
o
l
a
r

f
r
a
c
t
i
o
n
1

%
.luiniu +olle+tor 4 3
.luiniu +olle+tor 4 1.&
-taniless4steel +olle+tor 41.&
Fi(. ")$$* S24,- 3-,6:i2. 23 AL)3; AL)1." ,.> ST)1.". N2:+* 024,- 3-,6:i2. i0 :1+ ,827.: 23 +.+-(/
9-2vi>+> 5/ :1+ ICS)SWH >ivi>+> 5/ :1+ +.+-(/ -+=7i-+> :2 -,i0+ A,:+- :+89+-,:7-+ :2 ""
2
C
3or each month it is evident that the (*,9.= is under,performing compared with
aluminium $&(,(-.s. *he A2,9.= is slightly out,performed by the A2,; $&(,(-.1
however A2,9.= uses half the amount of material therefore suggesting A2,9.= as the
best alternative. Each annual energy saving has an associated annual &@
!
emission
saving presented in the ne%t section.
1.1.1.#% CO
2
0,vi.(0
Domestic water heating is primarily accomplished using natural gas or electricity with
natural gas accounting for B"L of the U< water heating systems. $n order to quantify
the annual &@
!
emission savings1
S
.(
1 detailed fuel conversion factors were
introduced. (uch factors provide &@
!
emissions e%pressed per unit activity such as fuel
or electricity consumed or distance travelled. Electricity and natural gas conversion
9B>
factors represents the fuel mi% used for electricity generation1 quality of coal consumed1
composition of the vehicle fleet and 0ourney ta'en in that year. Based on data from the
U< inventory of 7reenhouse 7as emissions produced for Department for Environment1
3ood and ?ural Affairs A DE3?A +!""=/1 electricity and natural gas conversion factors
were ta'en as ".>;'g&@
!
J'-h and ".9#'g&@
!
J'-h respectively. *he annual &@
!

emission savings of each $&(,(-.s were then calculated depending on the fuel used in
the household and are presented in 3igure =,B.
$11 $10
138
%(( %('
%%8
0
&0
100
1&0
$00
$&0
300
3&0
%00
%&0
&00
!le+tri+ity *as
Fuel
,
n
n
u
a
l

C
!
2

s
a
*
i
n
.
s
1

2
.
.luiniu +olle+tor 4 3
.luiniu +olle+tor 4 1.&
-tainless4steel +olle+tor 4 1.&
Fi(. ")1''* A..7,4 CO2 0,vi.(0 5/ 37+4 :/9+ 70i.( ICS)SWH0
Depending on the type of $&(,(-. installed there is a potential for saving of 9#C,!99
'g&@! per year if assisting a gas water heating system. -hile this may seem an
insignificant figure in terms of the overall targets set under the <yoto protocol1 if half of
the != million homes in the U< had an $&(,(-. system installed to assist their daily
hot water needs1 around !.B t&@! per year could be saved.
*hese results show that installation of $&(,(-.s in U< homes can contribute to
significant carbon dio%ide reduction. $n order to determine which of the three $&(,
(-.s would be the most viable solution1 an energy paybac' time +EPB*/ and carbon
dio%ide emission paybac' time +E&PB*/ were carried out in the ne%t sections.
9B=
".".3 E.+-(/ 9,/5,6C :i8+ B EPBT
*he energy paybac' time +EPB*/ is defined by Ardente et al +!""=b/ as the time
necessary for an $&(,(-. to collect the energy equivalent to that used to produce it.
*herefore the EPB* can be e%pressed in equation =.!.
#S. S
L(A
. .
.
.+BT

+=.!/
-here
L(A
.
is the energy input during the $&(,(-. life cycle +'-h/1
S
.
is the
annual useful energy saved by using the $&(,(-. +'-hJyear/ and
#S.
.
is the annual
energy employed during the use of the $&(,(-. +'-hJyear/.
*he use of a passive collector system or a solar driven pump was therefore assumed for
#S.
.
to be null. .ence EPB* were calculated for A2,;1 A2,9.= and (*,9.=. ?esults
are presented in *able =,C.
T,54+ ")2'* EPBT
C244+6:2- :/9+
L(A
.( (CW1)
S
.( (CW1?/+,-)
EPBT (/+,-0)
ST)1." 9B>B 9">9 9.B
AL)1." 9B:" 99": 9.=
AL)3 !C#: 99"# !.B
*he energy paybac' times related to the studied $&(,(-.s were found to be lower than
three years. *his value shows the great energy potential of such technology. A common
thought would be that the aluminium paybac' period should be significantly longer
because of its more intensive production process resulting in embodied energy value per
'ilogram of material being three times the stainless,steel one. .owever1 the recycling
property of aluminium decreases its embodied energy1 bringing down the energy
paybac' time considerably. 3uture scenario of C=L recycled and 9=L raw aluminium
material would lower the embodied energy of aluminium production therefore reducing
the EPB* of A2,9.= to 9" months.
9BB
".".# C,-52. >i2xi>+ +8i00i2. 9,/5,6C :i8+ B ECPBT
*he E&PB* was calculated based on the previously calculated annual emission savings
of &@
!
1
S
.(
1 in section =.=.!.!. *he E&PB* can be defined as the time necessary for
an $&(,(-. to equal the &@
!
emissions released during the production and use of the
installation itself as stated by Ardente et al +!""=b/. $t summarises the total impacts
during the life cycle and the emission savings. *he E&PB* is e%pressed in equation =.;.
#S. S
L(A
.( .(
.(
.(+BT

+=.;/
-here
L(A
.(
is the total &@
!
emissions related to the production of a ;m
!
$&(,(-.
during its life cycle +'g&@!/ represented in *able =,:1
S
.(
is the annual &@
!
emission
saving +'g&@
!
Jyear/ represented in 3igure =,B and
#S.
.(
is the annual &@
!
emissions
related to the $&(,(-. use +'g&@
!
Jyear/.
S
.(
represents the emissions saved by
using the $&(,(-. compared to the au%iliary system used in the household to deliver
hot water and therefore depends on the type of au%iliary heater employed.
*he use of a passive collector system or a solar driven pump was assumed therefore
#S.
.(
was set to null. E&PB* for A2,;1 A2,9.= and (*,9.= were calculated and
results are presented in *able =,# and 3igure =,:.
T,54+ ")21* ECPBT
C244+6:2- :/9+
L(A
.( (C(CO2)
S
.( (C(CO2?/+,-)
ECPBT (/+,-0)
G,0 E4+6:-i6i:/ G,0 E4+6:-i6i:/
ST)1." B"> 9#C >>C ;.9 9.;
AL)1." ;;! !9" >:B 9.B ".:
AL)3 =:C !99 >:: !.: 9.!
9B:
1.$
0.(
1.3
$.(
1.'
3.1
0.0
0.&
1.0
1.&
$.0
$.&
3.0
3.&
*as !le+tri+ity
Fuel
8
C
#
;
'
1

-
e
a
r
s
.luiniu +olle+tor 4 3
.luiniu +olle+tor 4 1.&
-tainless4steel +olle+tor 4 1.&
Fi(. ")1'1* ICS)SWH ECPBT* C,:+(2-i0,:i2. Ai:1 -+09+6: :2 37+4 :/9+
*he carbon dio%ide paybac' period depends on the au%iliary system used by the
household. 2i'e the energy paybac',time1 the &@
!
paybac',time results were generally
lower than three years.
*he per unit &@
!
evolution of electricity is much higher than natural gas resulting in
shorter environmental paybac' period for households using electricity as an au%iliary
heating system. *his is due to electricity having a high environmental impact due to the
fuel mi% used in its generation. E&PB* ranges from 9.B1 !.: to ;.9 years for the A2,9.=1
A2,; to (*,9.= respectively using gas while using electricity would result in a decrease
of E&PB* to C months1 9.! and 9.; years respectively) corresponding to a =BL
difference in E&PB* depending on the fuel type.
A considerable variation in the paybac' time of collectors of identical dimensions1 Al,
9.= and (*,9.=1 was observed. *his is e%plained by the recyclable property of
aluminium diminishing its embodied carbon and by its higher thermal conductivity
resulting in higher energy gain. A paybac' period based on future scenarios of C=L
recycled and 9=L raw aluminium material would result in lowering the E&PB* to 9"
and = months when gas and electricity au%iliary heating systems are used.
9BC
"."." C2.6470i2.0 ,.> -+6288+.>,:i2.0
An 2&A was carried out for each $&(,(-. type in order to establish the $&(,(-.
with the smallest carbon footprint. $t was found that despite the high amount of energy
to produce aluminium1 the MlightN and recyclable properties of the material resulted in
similar embodied energy when compared with stainless,steel. *his resulted in a smaller
environmental impact due to lower carbon emissions released to the atmosphere during
the production process.
Energy and carbon dio%ide savings associated by using those $&(,(-.s were modelled
and showed that although the A2,9.= $&(,(-. was out,performed by ".9L by the A2,
; $&(,(-.1 the fact that it only uses half the amount of material in producion ma'es it
more viable. *he carbon dio%ide savings showed that depending on the $&(,(-.
installed there is a potential for saving BL more 'g&@
!
per year by using an Al,9.=
compare with a (*,9.= and only ".=L by increasing the thic'ness of a same $&(,(-.
material.
*he EPB* of the $&(,(-.s was found to be consistently lower than three years for all
$&(,(-.s with the lowest EPB* value of 9 year : months achieved by the A2,9.=
$&(,(-.. *his value shows the great energy potential of such a $&(,(-. design. *he
calculations of the carbon dio%ide emission paybac' times confirm the higher potential
of using an A2,9.= $&(,(-. by decreasing the E&PB* by >!L and >CL when
compared with A2,; and (*,9.= $&(,(-.s respectively.
All results point to the A2,9.= $&(,(-. as the best alternative to operate in (cottish
weather conditions. $t would therefore be recommended to use an A2,9.= $&(,(-. for
further investigations due to its energy and environmental benefits. $ncreasing the
recyclable percentage in the A2,9.= could significantly reduce the environmental
impact of the collector. &opper and brass could be realistic alternatives to use in the
manufacture of the $&(,(-. in terms of embodied energy1 malleability and thermal
conductivity. .owever1 the cost of these materials being so much higher than that of
stainless,steel or aluminium would not ma'e them economically viable. Economics
plays a central role in any decision to purchase a solar water heater system. *he
customer is unli'ely to buy an $&(,(-. if they 'now that its only benefits are to the
environment. *herefore1 a monetary analysis was developed to consider the economics
of each system to ensure that such $&(,(-.s will be profitable.
9B#
".% M2.+:,-/ ,.,4/0i0 (MA)
*he monetary analysis +A/ can be defined as an economic evaluation technique to
determine the total cost of owning and operating a product with time. *he value of a
A lies in the ability to compare the cost of product alternatives and to determine
which one provides the best value per (terling pound spent. $n order to achieve an
appropriate A1 an insight of the collector material costs is provided1 followed by the
actual cost of the $&(,(-.s and a monetary paybac' time +PB*/ assessment.
".%.1 C244+6:2- 8,:+-i,4 620:0
$t is difficult to find a definite figure for the price of aluminium and stainless,steel as
they commonly fluctuate. Average figures obtained for the last year from ay !"": to
ay !""C would be T!#"" per tonne of stainless,steel and T9>>= per tonne of
aluminium. *he large price difference can have several possible e%planations. *he main
factor of the cost difference between the two metals is the added cost of machining
stainless,steel compared with aluminium) which is ductile and more easily machined.
-holesale metal suppliers EP( and 2E could guarantee costs up to three to four
months but not further than that period which reflects the frequency with which prices
vary. A comparison of the price trend of aluminium and stainless,steel over the last
three years is shown in 3igure =,C.
1000
$000
3000
%000
&000
'000
1
a
r
4
0
&
?
u
n
4
0
&
-
e
p
4
0
&
/
e
+
4
0
&
1
a
r
4
0
'
?
u
n
4
0
'
-
e
p
4
0
'
/
e
+
4
0
'
1
a
r
4
0
(
?
u
n
4
0
(
-
e
p
4
0
(
/
e
+
4
0
(
1
a
r
4
0
8
6ate
,
l
u
)
i
n
i
u
)

#
r
i
c
e
1

F
5
&
"
t
o
n
n
e
.luiniu -tainless4steel
Fi(. ")1'2* C20: 6289,-i02.
9:"
(tainless,steel has been the more e%pensive of the two materials but both materials have
been increasing at the same rate until recently. @ver the course of the last three years the
price of stainless,steel has risen beyond levels that anyone inside or outside the
stainless,steel industry could have forecasted. *here are a number of contributing
factors that have lead to these increased costs.
> : (hinese stainless:steel cons$mption
*here has been a recent and important increase in &hinese consumption of stainless,
steel which has caused a global price increase in spite of a consistent consumption for
the rest of the world. &hina has e%perienced phenomenal growth in the last five years
resulting in being the world5s largest stainless,steel consumer and has not yet reached
its pea'.
; : 'aw material shorta!e
(crap metal comes from waste stainless,steel items. oll +!""B/ announced a decline
from ;>.CL in !""> to ;!L by !"9" of the worldwide share of scrap in stainless,steel
production suggesting that scrap is being consumed at a far greater rate than it is being
created. *his current scrap shortage results in a notable increase in scrap prices. $ron ore
is another material which is currently in short supply due to the significant lac' of
investment in iron ore mines anywhere in the world e%cept for &hina and (outh
America. Another main reason is the shortage of nic'el ore supply1 a ma0or raw material
for stainless,steel production resulting in a high nic'el price. *hese raw material
shortages result in rising stainless,steel production costs.
; : Stainless:steel prod$ction
According to the $nternational (tainless,steel 3orum , $((3 +!"":/ global stainless,steel
production decreased by !.#L in !"":. *his happened across most regions and
countries e%cept for &hina and the Asian region accounting for B"L of the worldHs
stainless,steel production. A strong increase in stainless,steel production occurred in the
first half of !"": followed by a collapsed in the second half of the year after nic'el
prices dropped resulting in stainless,steel being withdrawn from the first half of !"":
stoc's.
? : Shippin!
9:9
Another item influencing stainless,steel prices is a shortage of shipping capacity. @ne of
the factors behind the growth in international trade over the past decade is a reduced
shipping cost but now the trade has grown to such an e%tent that there is a shortage of
shipping available. *he higher cost of shipping is hampering the movement of raw
materials around the globe.
@ : .ner!y costs
Energy costs are another factor increasing stainless,steel production costs. Production
cost increases due to higher energy input costs +i.e. Electricity is an output from many
energy resources and is used to power most equipment and facilities/. Energy mar'ets
are e%periencing high demand and limited supplies1 resulting in volatile mar'et and high
prices. (ince energy is a 'ey input in the production process1 sharp increases per unit in
goods occur as stated by OelaDqueD +!""B/. (tainless,steel being an energy intensive
process uses significant energy input for their production resulting in higher costs.
".%.2 ICS)SWH0 620:0
*he price of installing a domestic flat plate collector solar water heating system in
the U< today is anywhere from T!1""" to T;1""" which would include delivery1
installation and associated equipments costs as stated by E(* +!"":/. *he ma0ority of
solar water heating systems installed would come with a = to9" year warranty and
require very little maintenance. *he costs of the materials used in the $&(,(-.s
investigated for this pro0ect are detailed in *able =,9". Prices were updated to most
recent material cost data collected in ay !""C.
T,54+ ")22* R,A 8,:+-i,4 620:0 9+- 0=7,-+ 8+:+- 23 6244+6:2-
C244+6:2- :/9+ C244+6:2- 620: (Q) M,:+-i,4 620: (Q) ICS)SWH 620: (Q)
S:,i.4+00)0:++4
1."88
;=! 9"= >=:
A478i.i78
1."88
C9 9"= 9CB
A478i.i78
388
9B9 9"= !BB
*he difference in price of the two materials is a reflection of the recent surge in
global stainless,steel prices. Aluminium prices were ta'en from 2ondon etal
E%change +2E/ while stainless,steel ;">b prices were ta'en from EP( $nternational
2td. -hile the cost of materials may not be very high1 the additional labour and
9:!
delivery costs1 along with other required parts such as a pump1 controller1 valves and
piping would raise the price of parts up to the region of T91""".
*o encourage the public to become involved in the effort to reduce carbon emissions1
government grants for homes that install these heating systems are made available. *he
U< government 2ow &arbon Buildings Programme +2&BP/ offers a T>"" grant for
every professionally installed solar water heating system as stated by the D*$ +!"":/.
Another grant available in (cotland from the (cottish &ommunity .ouseholder
?enewables $nitiative +(&.?$/ provides ;"L of the installed cost up to a value of
T>1""" as stated by the Energy (aving *rusts +!"":/.
".%.3 M2.+:,-/ 9,/5,6C :i8+ (MPBT) ,.,4/0i0
*he monetary paybac' time +PB*/ can be defined as the time necessary for the $&(,
(-. to achieve the monetary equivalent to its capital cost and is e%pressed in equation
=.>.
< S
((
< <
<
<+BT
+

+=.>/
((
<
is the capital cost associated with the $&(,(-. life cycle +T/1
S
<
is the annual
monetary saved by using the $&(,(-. +TJyear/ and
<
<
is the annual associated cost
of using $&(,(-.s such as operational and maintenance costs +TJyear/.
7ull $ndustries +!""C/ also suggested that $&(,(-.s are virtually maintenance free
although the glaDing needs to be cleaned and the unit to be drained once a year1 as is
recommended for au%iliary water heaters. $t was then assumed for this study that annual
associated costs could be neglected1 therefore setting
<
<
to null.
*he potential annual monetary savings1
S
<
1 of a ;m
!
$&(,(-.s installation vary
considerably between homes using au%iliary gas fuelled or electric heating systems. *he
roller,coaster profile of energy prices in the U< will have a recognised impact on the
study findings. Data of 9!.9pJ'-h for electricity and ;.BpJ'-h including OA* for gas
prices refers to domestic standard prices of (cottish Power and British 7as prices
respectively in (eptember !""C. Based on this reference prices1 savings achieved by
9:;
using each $&(,(-.s were calculated. Annual savings1
S
<
1 were calculated from
equation =.=.
+ S S
. <
+=.=/
-here
S
.
is the useful annual energy saved by using the $&(,(-. +'-hJyear/ and
+

is the price of the fuel replaced in the household +TJ'-h/.

.ence PB* were calculated for A2,;1 A2,9.= and (*,9.= collectors. &ase 9 presented
in *able =,99 only accounts for the collectors capital cost1 assuming that the $&(,(-.
was mounted directly into the roof structure.

T,54+ ")23* C,0+ 1 ) C244+6:2-0 MPBT >+9+.>i.( 2. ,7xi4i,-/ 1+,:i.( 0/0:+8 32- , 38
2
0/0:+8
C244+6:2- :/9+ A478i.i78
388
A478i.i78
1."88
S:,i.4+00)0:++4
1."88
&ollector cost +T/ >C; !>; 9"=B
A7xi4i,-/ 1+,:i.(* G,0
Annual savings +T/ >" >" ;:
PB*+years/ 9!.! B.9 !C.>
A7xi4i,-/ 1+,:i.(* E4+6:-i6i:/
Annual savings +T/ 9;> 9;> 9!B
PB* +years/ ;.B 9.C C.>
Partially replacing a gas heating system with an $&(,(-. would result in a ma%imum
annual saving of T>"1 while replacing an electric heating system would result in savings
up to T9;> as shown in *able =,99 above. As e%pressed earlier1 the PB* depends on
the au%iliary heating system replaced by the $&(,(-.. ?eplacing an electric1 compared
to a gas heating system1 would be more profitable) increasing the PB* by :"L for
each $&(,(-..
.owever1 assuming $&(,(-. capital costs including all the components for
installation1 suggested in &ase !1 resulted in increasing PB* of !;1 ;# and =:L for the
(*,9.=1 A2,; and A2,9.= respectively as shown in *able =,9!.
T,54+ ")2#* C,0+ 2 B Ov+-,44 ICS)SWH0 MPBT
C244+6:2- :/9+ A478i.i78
388
A478i.i78
1."88
S:,i.4+00)0:++4
1."88
9:>
&ollector price +T/ :#C ==C 9;:9
A7xi4i,-/ 1+,:i.(* G,0
Annual savings +T/ >" >" ;:
PB* +years/ !".9 9>.9 ;B.C
A7xi4i,-/ 1+,:i.(* E4+6:-i6i:/
Annual savings +T/ 9;> 9;> 9!B
PB* +years/ =.# >.! 9".#
$t was observed that the lower the collector capital cost) the higher becomes the impact
on the PB* by adding further costs for each $&(,(-.. ?eplacing an electric
compared to a gas heating system suggested the same PB* improvement of :"L for
each $&(,(-..
$n both PB* studies the A2,9.= $&(,(-. was found as the most viable solution
reducing PB* by ="L and :#L compared to an A2,; or a (*,9.= system respectively
for &ase 9 and ;"L and B!L for &ase !. An estimated !" years life time would ma'e
the (*,9.= $&(,(-. a less attractive or viable alternative with an gas heating system)
as the customer would not see a monetary paybac' until after ;B.C years.
*he paybac' time detailed in *ables =,99 and =,9! only accounted for capital costs.
(ome $&(,(-.s may e%perience operational and maintenance costs associated with
certain component failures over the !" year period1 such as deficient valves and
problems related to the $&(,(-. itself such as glaDing brea'age1 seal failure1 insulation
degradation or connection failure. *he PB* when these costs are included would be in
the range of 9= to ;" years depending on the aluminium $&(,(-. installed and the fuel
being replaced.
".& C289,-i02. Ai:1 2:1+- 0/0:+80 ,.> 4i8i:,:i2.0 23 :1+ 0:7>/
any authors have investigated the benefits of 2&A of solar collectors and comparative
analysis of different collectors types by Asif et al +!""B/1 *silingiridis et al +!"">/1
irasgedis et al +9##B/1 &rawford et al +!"">/1 Ardente et al +!"";/ and Ardente et al
+!""=a Q !""=b/. .owever1 assumptions and data references are often unclearly stated
such as Battisti and &orrado +!""=/ ma'ing comparison between studies difficult.
3urthermore1 this is compounded by use of aggregated indices such as *silingiridis et al
+!"">/.
9:=

$n addition1 the same collector used in this study would have a different eco,profile
depending on its location. (cotland having a specific solar climate1 a location with
greater solar radiation potential would result in a better eco,profile of the $&(,(-. by
achieving more energy savings and therefore further reducing carbon emissions.
".< C2.647>i.( -+8,-C0
A comparative study between a stainless,steel and two aluminium $&(,(-.
using 2&A techniques and a monetary analysis were underta'en in this study. @f the
two identical collector units +(*,9.= Q A2,9.=/ investigated during this study the
aluminium $&(,(-. was found to be more efficient at transferring heat to the water.
@f the two materials1 aluminium was the better option mainly because of its high
thermal conductivity resulting in higher energy savings but also because of its lower
environmental impact due to its lightness and the spiralling price of stainless,steel. *he
potential energy savings associated with using these $&(,(-.s are quite substantial
and could reduce annual energy consumption in the home by up to >"L in a (cottish
climate by using aluminium $&(,(-.s. *he monetary analysis demonstrated that the
(*,9.= $&(,(-. was not a viable alternative as the customer would not have any
significant economic benefit over its estimated life time of !" years. .owever1 it
confirmed that the A2,9.= $&(,(-. was the most financially viable alternative. $t
should also be noted that prototype prices are always more e%pensive than large scale
manufacturing costs. Another advantage of recommending the A2,9.= as the best
alternative would be its low weight of 9"'g apt to fulfil the U< manual handling limits
of !='rJperson under the .ealth and (afety. $ssues about the durability of aluminium
compared with stainless,steel might arise. .owever1 based on the $&(,(-. estimated
life time of !" years1 aluminium can satisfactorily last this time period.
$t can be seen that A2,9.= $&(,(-. definitively offers a significant potential for
energy savings and &@
!
mitigation. Although the energy and the &@
!
paybac' times for
present,day systems are still relatively high1 it is generally lower than their e%pected life
time. 3uture wor' could be carried through &3D analysis for a multi,fin aluminium
$&(,(-. of 9.=mm thic'ness based on results from &hapter5s > and = to further
improve performance.
3inally1 the $&(,(-. studied being at its prototype stage1 it would be of interest to
research how it could be integrated into recent advances in roof structures in (cotland.
9:B
CHAPTER % INTEGRATION INTO HOUSING DESIGN
$n order to understand how to best integrate the $&(,(-. prototype into dwellings1 an
insight of government incentives towards the incorporation of 2E&* and their
development into legislative provision and Dero,carbon houses is given. Best
construction practice of such dwellings is also presented to show how technology such
as (-. can help in achieving Dero,carbon homes. *he integration of (-.5s depends
on the type of application. *ypical (-. installations into domestic hot water systems
are described in order to suggest the most suitable installation practice. 3inally1 the
integration of the $&(,(-. into roof structures is discussed in order to give an
overview of the potential integration of the system with modern modular methods of
construction +&/.
%.1 T2A,->0 E+-2)6,-52. 128+0
*his section will focus on the different government initiatives and legislative provision
towards the development of Dero,carbon houses.
%.1.1 UK G2v+-.8+.: I.i:i,:iv+0
1.1.1.#& E.+-(/ P+-32-8,.6+ 23 B7i4>i.( Di-+6:iv+ ) EU
*he Energy Performance of Buildings Directive +EPBD/ came into force on 4anuary
!""; with the ob0ective to improve the energy efficiency of buildings within the EU
through cost,effective measures and has therefore important implications for the
owners1 operators and developers of all buildings in the U<.
*he EPBD required all EU members to ta'e measures to ensure that minimum energy
performance requirements were set for new buildings and for large e%isting buildings
when they are refurbished. *his influenced U< Building ?egulations limiting the
overall annual &@
!
emissions of new buildings to a ma%imum target level) established
by reference to the calculated emissions from a notional gas,heated building of the same
siDe and shape as the proposed building.
*he most visible impact of the EPBD results in the implementation of energy
performance certificates1 which need to be prepared when buildings are constructed1
sold or rented since @ctober !""C. *he certificates classify buildings on a scale from A
+best/ to 7 +worst/ based on annual &@
!
emissions per unit floor area therefore
providing an indicator of energy performance of the building.
9::
Another aspect of the EPBD would be the provisions for regular maintenance
inspections of heating and cooling plant in order to provide adequate advice for the
operation of the plant.
$ncorporation of 2E&* is not compulsory for compliance with the regulations1 however
the target being e%pressed in terms of &@
!
emissions1 the choice of fuel is significant as
different fuels give rise to different &@
!
emissions for a given quantity of heat produced.
*his would results in encouraging the development of 2E&* for buildings to reduce the
building5s net demand for fuel.
*he Energy Performance of Buildings Directive +EPBD/ is thus a driving legislation
with an ob0ective to reduce building energy use and encouraging the development of 2E&*
for buildings.
1.1.1.#< C2>+ 32- 070:,i.,54+ 128+0
*he M&ode for (ustainable .omesN is the legislative bac'bone in the construction of
new homes in the U< which was implemented in ay !""C in England and -ales and
will form the framewor' for (cottish building standards. Prior to this an Eco.omes
Building ?esearch Establishment Environmental Assessment ethod +B?EEA/ was
in use although with no legislative mandate. Both tools are used to assess the
environmental performance of new buildings over their entire service life while only
Eco.omes B?EEA assess e%isting buildings.
*he &ode measures the overall sustainability performance of a new home using a 9 A B
star rating system based on the &ommunities and 2ocal 7overnment +&27/ +!""B/
publication. *he code divides the house into nine different categories such as
EnergyJ&@
!
1 -ater1 aterials1 (urface -ater ?un,off1 -aste1 Pollution1 .ealth Q
-ellbeing1 anagement1 and Ecology) each category has points awarded. *he higher
the score in the rating system1 the higher the sustainability rating of the house.
*he step,by,step approach of the &ode ma'es it an effective tool towards sustainable
housing. 2evel 9 is set 0ust above the current !""B building regulations while level B
would have very high mandatory heat loss parameter standards resulting in an increased
demand on the building envelope such as insulation1 glaDing and shading and how these
operate with the technological systems of the house to achieve Dero carbon. 2evel ;
standard of the code will become mandatory at the beginning of !"9"1 while level > and
level B will be phased,in in !"9; and !"9B respectively.
9:C
2ong term government ob0ectives for this &ode are to ma'e it the single national
standard for the design and construction of homes and for it to drive a step,change in
sustainable home building practice. *he 7overnment has also indicated its intention to
use the &ode as the basis of future Building ?egulations as stated by &27 +!""B/. *he
&ode became mandatory in ay !""C for all new private sector homes and will be a
ma0or help in trying to achieve the government5s ob0ective of Dero,carbon dwellings by
the year !"9B.
1.1.1.#$ S:,89 D7:/ L,.> T,x (SDLT) -+4i+3
3ollowing the government5s ob0ective of achieving Dero,carbon homes by the year
!"9B1 last year5s Budget ?eport published by .?& +!"":/ announced that all new
Dero,carbon homes would be e%empt from stamp duty land ta% on the house. *he (D2*
is a transaction ta% payable on the purchase of land and property to be paid by the buyer
at a rate which depends on the purchase price of the property. *he Dero,carbon home
standard is measured by use of the 7overnment5s (tandard Assessment Procedure for
the energy rating of dwellings +(AP/. *he relief is only available to the first newly built
Dero,carbon home sale and does not apply to the second or subsequent property sales. $t
is only valid for private residential individuals and not for commercial businesses and
has a time limit of five years1 after which time the scheme will be reviewed. *he relief
provides e%emption from the (D2* ta% when a house costs less than T=""1"""1 and
provides a T9=1""" reduction to all homes worth more than T=""1""".
*he main ob0ectives of introducing stamp duty relief on Dero,carbon homes is to help
launch their concept and create a mar'et for them1 to support micro,generation
technologies and enhance public awareness of the benefits associated with such homes.
$t also provides an incentive for developers to build Dero,carbon homes through
stimulating consumer demand by advertising them as free of stamp duty land ta%. *his
incentive would also help the U< to achieve a B"L cut in carbon emissions by !"=" as
stated by .?& +!"":/.
*he (D2* relief therefore enhances and encourages the construction of homes
incorporating specific energy efficiency standards in the fabric of the building and
micro,generation technology5s at the design stage.
9:#
%.1.2 L+-2)6,-52. 1270i.( 32- UK
7overnment initiatives and incentives offer developers the possibility to integrate
sustainable construction and renewable energies into new developments. A Dero,carbon
home is defined by .?& +!"":/ asF
,9ne that does not cons$me fossil f$els for heat and power/ &t is hi!hly
ins$lated and $ses renewable ener!y to power its needs o-er a year thro$!h
micro !eneration/ &t will draw from the !rid when the micro !eneration is
ins$fficient b$t co$ld sell excess !eneration back to the !rid/4
7ood and best practice e%amples of low and Dero,carbon developments were identified
and are introduced. An organisation1 Eero Energy Development +EED/ factories
recently introduced their innovative housing system '$ral A.D to the Dero,carbon
housing sector at Ecobuild +!""C/. '$ral A.D is designed as a three bed detached
dwelling that catches fresh air1 rain and sunlight allowing a full integration of renewable
micro,generation technologies which brings it to a Eero Energy Development.
Electricity is generated from solar photovoltaic +PO/ panels on the roof1 and where the
site is suitable1 a roof wind turbine. -ater heating is generated from roof mounted
(-.1 bac'ed,up in winter by a biomass boiler. '$ral A.D achieves &ode level >1 but
provides an upgrade path to level B1 depending on the budget available. *he higher code
levels will necessarily be achieved at a higher cost.
B?E $nnovation Par' @33($*E +!"":/ demonstrated various types of near Dero and
Dero,carbon homes designed and built by various companies including <ingspan1
(tewart ilne and .anson. *he 2ighthouse1 created by <ingspan @ffsite was the U<5s
first Dero,carbon house to meet &ode level B. 2i'e the ?ural EED1 the <ingspan
2ighthouse was designed as a two or three bed detached dwelling using efficient
construction methods1 energy use to achieve low &@
!
emissions and carbon footprint.
*he pitched roof at >"[ possess an inbuilt wind catcherJlight funnel and accommodates
solar PO generating most of the electricity needs while a (-. generates hot water. *he
building fabric and building envelope delivers a thermally insulated and air tight home
that will only need heating during winter months. *he (igma .ouse of (tewart ilne
7roup and .anson Ecohouse were designed as a four and three bed detached dwellings
and met &ode level = and level > respectively. *hey incorporate a roof light and micro,
generation technologies similar to other Dero,carbon homes. 3igure B,9 show the
different low and Dero,carbon developments described above.
9C"

6
>
Fi(. %)1'3* T2A,->0 E+-2 6,-52. 128+0* ,. R7-,4LED; 5. E621270+; 6. Li(1:1270+; >. Si(8,
Efforts have been made by developers to integrate sustainable construction and micro,
renewable energies into new developments. icro,renewable technologies have a
strong role to play in the development of Dero carbon housing. (olar water heaters have
been installed in all best practice e%amples of low and Dero,carbon developments shown
above) suggesting that there is a strong potential for this type of technology to
participate towards Dero,carbon homes. An insight into this technology1 and particularly
that of flat,plate collectors is provided in the ne%t section.
%.2 F4,:)94,:+ 6244+6:2-0 32- 42A 6,-52. 62.0:-76:i2.
$n order to understand the principles of operation and integration of this technology the
predominant types of installation of flat,plate collectors are presented.
, 5
9C9
%.2.1 Exi0:i.( 34,:)94,:+ 6244+6:2-0
(i% recognised companies1 AE( (olar 2td1 Amplifaire1 Energy limited1 (olar Power
(cotland1 ?enewable Devices Energy (olutions 2td and (olar Energy systems1 were
approached to provide typical installation details for (-. in the U<1 however only
three gave an insight of the type of installation in use. *heir installation systems are
described below.
1.1.1."' S24,- E.+-(/ S/0:+80
A frequently adopted alternative used by (olar Energy (ystems is to use a solar pre,heat
tan' which has no input from the bac',up heat source +boiler/. *he system uses an
open,vented system resulting in an un,pressurised installation. &old water from the
main enters the water reservoir in which a ball valve controls the level of water. &old
water is then fed to the bottom of the solar pre,heat tan' each time hot water is drawn,
off by the household. *he water is pre,heated by the indirect (-. primary circuit in
which the heat transfer fluid passes and circulates to the heat e%changer and bac'. *he
pre,heated water then feeds the downstream hot water tan' connected to the boiler to
achieve the desired hot water temperature supply as shown in 3igure B,!.
9C!
Fi(. %)1'#* S+9,-,:+ 024,- :,.C 9-+)1+,:i.( :1+ >A+44i.( 12: A,:+- :,.C
*he (-. primary circuit used for pre,heating is isolated from the consumed water
hence allowing the use of anti,freeDe. *his system is more e%pensive) requiring two
tan's1 has a high space requirement and ta'es a longer to install.
1.1.1."1 AES
AE( is one of (cotland5s most recognised solar system manufacturers and provides a
number of different (-. installation types. All their systems use an indirect sealed
primary circuit in which the heat transfer fluid that passes through the (-. is isolated
from the consumed water by a form of heat e%changer and hence allows the use of anti,
freeDe. A sealed primary system results in a significant rise in static pressure by the
filling process1 typically for solar D.- between 9 and ; Bar over atmospheric pressure
as stated by &$B(E +!"":/.
*he first proposed installation consists of a double,coil installation or combined solar,
dwelling hot water tan'. A schematic of the system is shown in 3igure B,;. &old water
from the mains enters the bottom of the water tan' which is fitted with two indirect
9C;
coils both wor'ing as heat,echangers1 one heated by the (-.1 the other heated by a
boiler. @utputs of the bac',up +boiler/ and the (-. systems are both into one water
tan'. *his type of installation is another frequently adopted alternative in the U<.
Fi(. %)1'"* C285i.+> 024,- >A+44i.( 12: A,:+- :,.C.
Another method would be to use an e%ternal heat e%changer. As previously1 cold water
from mains enters the bottom of the water tan' but also goes to a secondary pre,heated
circuit. *he secondary circuit goes into an e%ternal heat,e%changer in transferring the
heat from the primary (-. to the secondary circuit going bac' to the water tan'. *he
water uses a bac',up circuit to deliver the hot water supply. Each circuit requires a
pump resulting in higher final costs. A schematic of the system is shown in 3igure B,>.
9C>
Fi(. %)1'%* Ex:+-.,4 1+,: +x61,.(+-
Another installation proposed by AE( consists of a combi,boiler solar compatible
installation shown in 3igure B,=. &old water from the mains enters the bottom of the
water tan' to be pre,heated by the (-. primary circuit used as a heat,e%changer. Each
time hot water is drawn from the household the pre,heated water goes to an
instantaneous heating appliance such as a combi,boiler which uses a thermostatic
blending valve to protect the cold supply of the combi,boiler.
9C=
Fi(. %)1'&* C285i)52i4+- 024,- 6289,:i54+ i.0:,44,:i2.
1.1.1."2 S24,-:Ai. 0/0:+80
*he (olartwin system is a more recently adopted alternative using a direct solar system
installation. *he (-. system shares the same open,vented safety system and cold
water reservoir with the water tan'.
*he (-. system employs a fle%ible pipe in the absorber plate resisting the volumetric
e%pansion due to freeDing. *he (-. system is therefore freeDe,tolerant and can use
mains water as a heat transfer fluid. *he collector and circulation system can therefore
be plumbed directly into the e%isting hot water system. *he use of a small pump
powered by a photovoltaic module provides a low and variable flow enhancing storage
tan' stratification as stated by 7rassie et al +!""9/ and results in a net Dero carbon
system. A schematic of the system is provided in 3igure B,B.
9CB
Fi(. %)1'<* S24,-:Ai. 0/0:+8
*his system reduces costs as it requires minimal plumbing and additional components
required to prevent frost damage in the above systems are no longer necessary. $t is also
easier and quic'er to install and has an additional thermal benefit through heating the
domestic hot water directly.
%.2.2 S7((+0:+> i.0:,44,:i2. 32- :1+ ICS)SWH
Based on the typical commercial (-. installation detailed above1 constraints for the
installation of the $&(,(-. under study were considerated and an optimised
installation proposed.
1.1.1."3 I.0:,44,:i2. 62.0i>+-,:i2.0
*he solar water heater under study stores the solar heat directly in its integrated tan'
allowing the heat to build,up slowly during the day. Using the $&(,(-. as both the
collector and storage tan' significantly reduces the space requirements and costs
associated with typical (-. installation as it would use main water.
9C:
*he primary issues concerning the $&(,(-. installation would be the potential
freeDing problems and pressurisation of the system. *he system uses mains water as the
heat transfer fluid and could potentially freeDe during winter months. Using a sealed
system would result in a significant increase static pressure between 9 , ; Bar over
atmospheric pressure.
$n e%treme weather conditions1 freeDing could damage the $&(,(-.. 7rassie +!""9/
suggested the use of fle%ible ethylene propylene diene ,class rubber or EPD rubber
instead of the more commonly used copper pipes. Using this material allows pipes to
e%pand during freeDing periods therefore reducing potential damage. *he $&(,(-.
could be designed to use EPD rubber inlet and outlet pipes where the greatest
potential for freeDing occurs. An additional benefit in using EPD rubber inlet and
outlet pipes would be the avoidance of galvanic corrosion occurring between aluminium
and commonly used materials for water heating pipes such as copper1 brass and
stainless,steel.
Using a sealed direct system would result in pressures of up to ; Bar that the $&(,(-.
may not withstand. A pressure regulation valve and a pressure relief valve could then be
integrated in the installation to regulate the water pressure on entry to the $&(,(-. and
relieve e%cess pressure due to e%pansion.
1.1.1."# P-2920+> ICS)SWH i.0:,44,:i2.
*he system would consist of a direct sealed installation using mains water as a heat
transfer fluid that passes through the (-. incorporating a bac',up heat source to
provide required domestic hot water final delivery temperatures. A combi,boiler solar
compatible installation or direct electric heater1 such as *itan BB tan'less hot water
heater1 would be then integrated into the system as a bac',up heat source. (myth et al.
+!""9/ suggested that most $&(,(-.s have large thermal masses and simple compact
structures which give good general resistance to freeDing climates. 3urther studies in
order to integrate tolerance to freeDing are required for this system. $nitial discussion
suggest a flow regime induced within the $&(,(-. and the use of fle%ible inlet and
outlet pipes capable of resisting the volumetric e%pansion due to freeDing and reducing
the ris' of galvanic corrosion.
9CC
-ater quality is of primary importance if the water is being used directly from the $&(,
(-.. 2imescale is a common phenomenon occurring in hot water storage systems
indirectly or directly connected to the solar water heater. *he main component of
limescale deposited in water heaters consists of calcium carbonate which is soluble in
water. .owever1 at temperatures above :"& the soluble bicarbonate is converted to
poorly soluble carbonate leading to deposits in the water heater usually resulting in a
life and performance reduction of the system. *here are many ways to prevent
limescale. Products such as MEcoflowN are reported to prevent limescale at the source. $t
consists of a magnetic water conditioner fitted on the rising main where water comes
into the home providing a constant supply of magnetically treated water. *he
sterilisation of bacteria such as the legionella bacteria in hot water systems is also of
primary importance for .ealth and (afety approval of the $&(,(-.. *emperatures of
consumable water are generally between !"& and >B& according to &$B(E +!""!/
when storing heat1 thus providing adequate conditions for the multiplication of bacteria.
*he treatment of the pre,heated water from the $&(,(-. therefore cannot be
considered suitable for distribution as domestic hot water until fully treated. At B"&1
bacteria can survive for only minutes whereas at :"& the bacteria in the water are
'illed instantaneously. A flow limiter would be required in the case of an instantaneous
heating appliance such as the $&(,(-. in order to maintain the target treating
temperature. A thermostat interloc'ed to the instantaneous heating appliance will also
ensure that all water draw,off reached the target temperature. A schematic of the system
is shown in 3igure B,: including the components belowF
9. Pressure regulator valve
!. $&(,(-.
;. Pressure relief valve J Air vent
>. *hermostatic mi%ing valve
=. $nstantaneous heating appliance
B. *hermocouple
9C#
Fi(. %)1'$* ICS)SWH i.0:,44,:i2.
-hen hot water is drawn off from the $&(,(-.1 mains cold water controlled by a
pressure regulated valve is fed to the bottom of the $&(,(-. flushing out the hot water
from the top of the collector passing through a pressure relief valve. *he hot water then
goes to an instantaneous heating appliance such as a combi,boiler or electric heater to
deliver the desired temperature of hot water supply.
*his system would reduce costs as it requires minimal plumbing) it is of simple
construction and heats the domestic hot water directly using no separate water tan'.
(upposing that the $&(,(-. was integrated in a Dero,carbon house1 the electric
instantaneous heating device could be powered by micro,renewable technologies)
therefore resulting in a Dero,carbon system.
%.3 I.:+(-,:i2. 23 :1+ ICS)SWH i.:2 -223 0:-76:7-+
$ntegrating solar water heaters into roofs depends considerably on the roof type and the
design of the solar water heater installed. ost solar water heaters in U< are mounted
above the roof covering using brac'ets to hold the collector to the roof structure. *he
roof is therefore altered from its original design loading which might cause structural
and durability issues. (olar water heaters can also be vertically mounted or integrated
directly into the roof covering but this would require specialist structural advice.
Penetrations for brac'etry and water connection also compromise the roof integrity.
1
2
3
#
"
%
9#"
&urrent estimates from the Parliamentary @ffice of (cience and *echnology A P@(*
+!"";/ assert that there will be three million new U< households by !"9B. *he U<
government is encouraging a modern modular method of construction +&/1 which it
says can achieve Ma step change in the construction industry to produce the quantity and
quality of housing we needN as stated by P@(* +!"";/.
$n order to provide the best practice to integrate the $&(,(-. in future households1 an
overall insight of & and more specificallyan e%ample of this1 the structural
insulated roofing system +($?(/ developed by a (cottish based company Donaldson
*imber Engineering +!"":/ is presented to demonstrate its integration.
%.3.1 M2>+-. M+:12> 23 C2.0:-76:i2. (MMC)
*he recent interest in & for housing construction is driven by a growth in the
number of households in the U<. &1 also 'nown as prefabrication1 consists of the
manufacturing of house parts off,site in factories. -hile issues have arisen regarding
the quality of houses1 recent technical improvements in prefabrication including
innovative site,based methods and the wide range of material used1 suggests that &
has many advantages with potential benefits described by P@(* +!"";/ such asF
, faster construction
, fewer housing defects
, energy efficient houses
, reductions in energy use +less transport/
, less waste production
, safer wor'ing conditions
, lower impact on local residents during construction
*he two main products of & consist of modules and panels. odules1 also 'nown
as pods1 consists of ready,made rooms1 usually bathrooms or 'itchens where all the
fittings and services are added in the factory. ?eady,made rooms are then assembled
together to produce a whole dwelling. Panels consist of ready made walls1 floors or
roofs. *hese are transported to the site and can often be assembled within a day as
stated by English Partnership +!"":/.
9#9
%.3.2 SIRS R2230
*he ($?( roof +(tructural $nsulated ?oof (ystem/1 manufactured by Donaldson *imber
Engineering +D*E/ is an e%ample of modular offsite roof design1 employing modern
methods of construction &. *he ($?( roof is a roofing system that has been
developed to provide a safe and cost effective solution in delivering additional space
such as room in the roof in a domestic dwelling.
*he ($?( roof design parameters e%ceed current building regulations in terms of
insulation1 and comply with government directives on modern methods of construction.
*he ($?( system suits a variety of roof shapes and is manufactured
using timber from sustainable sources. It also contributes to Low
arbon !uild and has been approved by the "#! ("ational #ouse$
!uilding ouncil) last %uly 200&. *he ($?( roof assembly is shown in 3igures B,
C and B,#.
Fi(. %)11'* S:-76:7-,4 I.074,:+> R223 S/0:+8 ) SIRS; W22> (2''%)
9#!
Fi(. %)111* S:-76:7-,4 I.074,:+> R223 S/0:+8; H,8+0 H21.0 -223 :-i,4.
*he design fle%ibility of the ($?( roof would allow the direct integration of the $&(,
(-. on the roof panel. *his would result in a faster and safer integration of renewable
technology through off,site construction and high insulation of the $&(,(-. due to the
e%cellent insulation performance of ($?(.
%.3.3 S:-76:7-,4 i007+0
*here are many structural issues when considering the installation of a (-. system.
*he additional load of the mounted $&(,(-. requires specialist structural advice.
.owever1 directly integrating the $&(,(-. into the roof covering replaces tiles thereby
reducing weight issues. 3or instance1 the (andtoft company +!""C/ provides a range of
clay tiles weighing between ;>,:!'g for a 9m
!
area resulting in an average weight of
=;'gJm
!
. -hen integrated into the roof covering the filled 9m
!
A2,9.= $&(,(-.s
weighs >9.;'g which would result in a !!.9L weight reduction over clay tiles.
A structural engineer would normally be responsible for calculating such loads. As a
result of high wind speed1 steeper pitched roofs are generally installed in (cotland to
diminish the potential for tiles lifting.
9#;
-hether the installation is mounted above the roof or directly integrated into the roof
covering1 a flashing system would be required to ma'e the installation weatherproof. An
overview of OE2U` +!""=/ flashing systems is given as suitable method for the $&(,
(-. integration in roofs. OE2U` flashings ensure weathertight installation of all
OE2U` roof windows by ensuring safe water drainage from the roof above the
window. OE2U` +!""=/ uses prefabricated flashing systems which overlap with the
roof covering to provide a weathertight finish around roof windows. OE2U` flashing
systems are compatible with most types of roofing material and allow a discrete
installation as the roofing material covers the top and side flashing components.
3lashing materials are weather,resistant. A collar ensures a tight connection between the
roof window and the membrane and weather tight 0oints are created by overlapping the
flashing components whilst the foam gas'et creates a seal between the flashing and the
roofing materials. Another feature of OE2U` flashings is the incorporation of tile
support ensuring that the roofing material remains in the right inclination above the roof
window. A pleated apron is then dressed to form a tight seal underneath the window in a
profiled roof. 3igure B,9" shows a OE2U` installation.
Fi(. %)112* I.0:,44,:i2. 061+8,:i6 ,.> 3i.,4 i.0:,44,:i2. 23 VELU! Ai.>2A0; VELU! (2''")
'he general good practice introduced by I!S( (200&) in respects of
pitched roof guidance is)
, 'o maintain access to the roof covering below the solar water
heater
, 'o report loads onto the roof covering
9#>
, 'o design the loads onto substantial roof structures such as
trusses
, 'o provide sufficient lap of flashing where tiles are cut or raised
, 'o consider negative pressure lifting ad*acent components such
as tiles

!ased on these parameters+ the integration of the IS$S,# was
considered and simulated.
%.3.# I.:+(-,:i2. 23 :1+ ICS)SWH
*he integration of any (-. has a direct impact on a building including its performance
and aesthetics. $ntegrating an $&(,(-. should not devalue the property and should
comply with building regulations and planning.
Using OE2U` methods for weatherproofing the $&(,(-. in a ($?( roof structure
could ensure a good integration of the $&(,(-. in a roof covering. .owever there are
issues concerning the direct integration of the $&(,(-. into the roof covering. *he
viability of such integration should be studied further in order to ensure its feasibility.
%.# C2.647>i.( -+8,-C0
*his chapter e%plored the potential installation and integration of the $&(,(-. under
study into recent advances in constructions practice. ?ecent governmental initiatives
towards the e%pansion of Dero,carbon development enhanced the integration of micro,
generation technologies into buildings. *he best practice e%amples of low and Dero,
carbon developments such as the <ingspan 2ighthouse or ?uralEED identified the
potential for roof mounted (-.s to generate a dwellings water heating with a heating
bac' up and their contribution towards such Dero,carbon developments. Best (-.
installation practices of three (cottish companies were studied in order to provide an
optimum practice for the $&(,(-. installation. *he suggested installation consisted of
a direct sealed system using mains water as a heat transfer fluid and an instantaneous
heating bac',up source such as a *itan BB tan'less hot water heater.
*he integration of the $&(,(-. in future potential (cottish household is a final
milestone of the research pro0ect. $ntegrating an $&(,(-. should not devalue the
property and should contribute to complying with housing codes and legislations.
odern methods of construction and more precisely of structural insulated roofing
9#=
system +($?(/ were recently introduced as contributing to Low arbon !uild.
'he design fle-ibility of the SI.S roof could allow the simulated direct
integration of the IS$S,# inside the roof panel. Such installation
results not only in faster and safer integration through off$site
construction+ but also in higher performance of the IS$S,# due to
the high insulation performance of SI.S. (uch installation would also provide a
substantial value to developers proposing such pre,integrated micro,renewable systems
to customers resulting in achieving higher levels of the &ode for (ustainable .omes.
$n conclusion1 this section of the thesis opens an avenue for future wor' on $&(,(-.
integration but also shows the potential of $&(,(-. in the solar water heating and
Dero,carbon home mar'ets.
9#B
CHAPTER & CONCLUSIONS
Energy plays an essential role in society and human life1 driving economic development
and the wellbeing of society. odern societies use more energy every year for industry1
services1 homes and transport resulting in increased environmental concerns associated
with fossil fuel consumption and production. Ensuring a secure1 clean and affordable
supply of energy is essential towards the sustainable development of a modern society.
*he domestic sector is the largest consumer of energy in the U< with C"L accounting
for space and water heating. .ence1 there is good potential scope for addressing
domestic energy requirements1 particularly for hot water1 from renewable sources. (olar
water heating systems have real potential in (cotland due to the higher annual heating
loads allowing longer and greater savings using solar thermal energy. A review of many
years of research on solar water heating systems has outlined the promising approach of
integrated collector storage solar water heaters +$&(,(-./ in northern hemisphere
applications due to their good overall performance in overcast s'y conditions. $&(,
(-.s are one of the simplest types of (-. on the mar'et incorporating the solar
collector and thermal storage tan' in a single unit ma'ing them relatively affordable
with low costs. *here was therefore a need to estimate the potential of $&(,(-. for the
northern climates.
*his research was conducted with the primary goals of developing an affordable1
optimised and feasible $&(,(-. prototype for (cottish weather conditions and to
develop a new tool to model $&(,(-. performance. *he approach involved laboratory
and field tests of the $&(,(-.1 performance modelling for laboratory and field
conditions1 performance optimisation through &3D modelling1 life cycle assessment and
monetary analysis of the $&(,(-. and finally the analysis of potential benefits and
integration into buildings of this $&(,(-. prototype.
&.1 S788,-/ 23 62.6470i2.0
(ignificant points that can be drawn from the research analyses are as followsF
*aborator+ and field experiments
, *he methodology adopted in this study was generally more accurate and rigorous than
previous studies.
9#:
, *he incorporation of e%tended surfaces in the $&(,(-. was analysed and suggested
an increase in heat transfer and water stratification in the $&(,(-. as well as providing
structural strength to the collector. $ncorporating fins in the collector resulted in an
increase in heat transfer of 9CL in the stainless,steel $&-,(-.. Using a more
conductive material1 such as aluminium1 resulted in a 9;L increase in water
temperature.
, Analysis suggested that the temperature stratification profile results in differential
buoyancy forces in the fluid causing convection by gravity. $t also indicated that this
depends upon the incident solar radiation1 the e%posure time and collector material1
given a constant angle of inclination of the $&(,(-.. A polynomial function for the
stratification profile for each hour could then be developed ta'ing heat flu% and
longitudinal location in the collector as the main parameters.
, *he benefit of selecting aluminium for the $&(,(-. was clearly supported by the
current laboratory findings) showing it out,performing the stainless,steel $&(,(-.
with an overall increase in the aluminium $&(,(-. efficiency of BL over an eight
hour time period.
Although the collector was tested in a laboratory with calibrated instruments1 the actual
performance of the $&(,(-. under real weather conditions yielded important results
on assessing the viability of the aluminium $&(,(-. and are summarised below.
, Parameters such as thermal mass of the system1 orientation1 cloud cover and s'y
temperature were found to significantly influence the $&(,(-. performance in field
weather conditions1 and should not be neglected.
*he thermal mass was found to play an important role in the effectiveness of the system
as it affected the heating charge and cooling discharge operations of the collector and
the amount of stored thermal energy. *he water stored in the collector chiefly
contributed towards the thermal mass of the system with a value of !"C.;'4J< compared
with the aluminium and stainless,steel material of heat capacities of 9:.! '4J< and
9;.;'4J< respectively. *his shows the ability of a given water volume to store internal
energy while undergoing a given temperature change.
9#C
$t was also observed that a decrease in incident solar radiation for a relatively long
period of time can lead to significant losses and decrease in stratification rate1 therefore
positioning of the $&(,(-. is of primary importance to achieving good performances.
*he effective s'y temperature can be about 9",!" degrees &elsius below ambient
temperature at ground level with clear s'y conditions and close to1 or 0ust below1 the
ambient temperature during cloudy conditions. $t was therefore concluded that cloud
cover should not be neglected as it affects the cover temperature through radiation
losses and therefore the final water temperature output of the $&(,(-..
, *he draw,off study showed that the $&(,(-. was able to maintain temperature
stratification when draw,off occurs1 a crucial factor in maintaining good performance of
the $&(,(-.. $t was also observed that draw,off actually increased the overall thermal
longitudinal stratification1 thus improving the overall performance of the $&(,(-..
Average draw,off temperatures recorded at pea' demand hours suggest that the $&(,
(-. could cover ==L1 BBL1 C;L :!L of the hot water demand at CF""1 #F""1 9#F""
and !"F"" respectively based on a delivery temperature of ==S&. A strong relationship
between water drawn,off1 water temperature stratification1 solar radiation and outlet
water temperature achieved in the collector was reported.
, $n addition to the relative importance of different parameters influencing the $&(,
(-. performance developed in this study1 the current conventional standard procedure
used by the E(&(*7 +9#C#/ for predicting the effectiveness of $&(,(-. was found
not to be representative of the actual system performance as thse assume steady state
conditions. *his emphasised a need for new methods to predict water temperatures and
performance of $&(,(-..
,odelling -C'!'$H performances
, *wo macro models , laboratory and field , were developed in this study to predict the
performance of $&(,(-.s at any location and with different installation configurations.
Each macro model resulted in a thermal energy simulation model which analysed
combined heat flu% environmental networ's within the $&(,(-..
, *he laboratory macro,model was able to predict the bul' water temperature in
different $&(,(-. configurations using different construction materials for a given
9##
heat flu%1 ambient temperature1 inlet water temperature and resultant stratification
profile over time.
*he role of e%tended heat transfer surfaces +fins/ were modelled and showed good
improvement by achieving higher bul' water temperatures. Effective fin heat transfer
rates were enhanced by the choice of a material e.g. aluminium over stainless steel. *he
improvement was found to be B:.;L and B;L for aluminium fins of ;mm and 9.=mm
thic'ness and !=.;L for 9.=mm thic' stainless,steel.
An improvement factor +$3/ measuring the improvement achieved by using fins in the
$&(,(-. was developed. By increasing the thic'ness of the fins the overall $3
improved by ;.=L1 while changing to a more conductive material the $3 improved by
=.=L. *he greatest improvement factor was achieved for the aluminium $&(,(-. with
a ;mm fin thic'ness.
*he computational results from the model were found to be in close agreement with the
e%perimental data. (atisfactory curve fits were obtained for the water temperatures1
achieved e%perimentally and numerically for all the $&(,(-. types. *esting the results
using statistical methods resulted in very low BE and ?(E values) validating the
ability of the model to predict the water bul' temperature in the $&(,(-..
2ongitudinal stratification of temperature within the solar collector was modelled using
the mean bul' water temperature computed by the macro model1 the given heat flu% and
the total time of e%posure. *he stratification model was developed using polynomial
functions previously developed via e%perimental data for 9""1 !"" and >""-Jm
!
.
Different heat inputs outside the normal e%perimental range were then simulated in
order to validate the model. ?esults for ="1 9="1 !="1 ;=" and B""-Jm
!
over a period of
eight hours +simulating a typical day of operation/ demonstrated that e%perimental and
computer simulation results were in good agreement. *he BE value of 9 &elsius
showed that the model had a tendency to slightly under predict its computed value.
(tatistical results validated the accuracy of the model to compute bul' water
temperature.
, *he field macro thermal model developed was found to be suitable for predicting bul'
water temperatures under real weather conditions and therefore more predictable $&(,
(-. performance.
Analysis carried out using data for five and fifty day testing periods between the month
of 4une and 4uly validated the model to predict bul' water temperature inside the $&(,
(-. by comparing computed results with e%perimental data.
!""
$t was observed that gradual heating and cooling profiles due to the thermal mass of the
system were e%perienced in the macro model. .owever1 a lower cooling rate of the
modelled values was observed for each day. *his could be e%plained by the propagation
of errors and uncertainty by the use of regression analysis and assumed weather data)
resulting in a decrease in the accuracy of the model.
E%perimental water temperature was followed very closely by predicted water
temperature although it showed a slight over,estimation of 9.; and 9.= &elsius of its
computed values by the proposed model for five and fifty day periods respectively. *he
study also demonstrated the effect of wind velocity on the bul' water temperature and
the importance of having accurate wind speeds. $ncreasing the velocity from "mJs to
BmJs resulted in an average decrease of =L of the bul' water temperature.
Although this model only gives mean values of water temperature which1 in reality1
varies along the longitudinal height of the collector1 it gave a good estimation of the
$&(,(-. performance. *he simulation could be e%tended resulting in a useful tool to
estimate the yearly performance of the $&(,(-..
, Both simulation models showed that computational results were found to be in close
agreement with the e%perimental field measurements. (tatistical methods suggested that
both models were robust and could compute bul' water temperature in the $&(,(-.
and predict the performance of an $&(,(-. at any location and with different
installation configurations.
%ptimisation through C./ modelling
, A &omputational 3luid Dynamic +&3D/ study was underta'en to model flow and heat
transfer in the three,dimensional +;D/ (-. geometry to provide information about
performance of the collector and evaluate design improvements. $t was found that !D
analysis was insufficient for any angle of inclination of the $&(,(-. above horiDontal
as a gradient e%ists in the longitudinal direction. .ence1 based on a >=S inclination of
the collector1 a ;D analysis was underta'en to estimate an optimum fin spacing to ma'e
further improvements on the original four,fin $&(,(-..
, $t was found that one further fin added to the original four,fin $&(,(-. design would
improve the heat transfer by !.;L. $t was observed that the five fins supplied more
energy to the collector than four at the beginning of the charging process. *he
!"9
intermittent availability of incident solar radiation in (cotland shows high potential for
this type of improvement as it is advantageous to have a rapid heating process.
, $t was also observed that while the five fins supplied more energy to the collector than
four at the beginning of the charging process1 additional fins seriously impeded the
water flow with a decrease of the velocity in the (-.. *his decrease in velocity has a
corresponding reduction on the Nusselt number with a consequence that the heat
transfer coefficient falls in a predictable manner as outlined by &hurchill +9#:>/.
*ife c+cle assessment and monetar+ pa+back
, A 2&A and monetary analysis of two aluminium $&(,(-.s of 9.= and ;mm
thic'ness +A2,9.= and A2,;/ and one stainless,steel $&(,(-. of 9.=mm +(*,9.=/
incorporating e%tended heat transfer surfaces were carried out in order to establish the
most commercially viable $&(,(-. with the smallest environmental footprint.
, $t was found that despite the high amount of energy to produce aluminium1 the MlightN
and recyclable properties of the material resulted in similar embodied energy when
compared with stainless,steel. *his resulted in a smaller environmental impact due to
lower carbon emissions released to the atmosphere during the production process.
Energy and carbon dio%ide savings associated by using those $&(,(-.s were modelled
and showed that although the A2,9.= $&(,(-. was marginally out,performed +".9L/
by the A2,; $&(,(-.1 the fact that it only uses half the volume of material in
producion ma'es it more viable. *he carbon dio%ide savings showed that1 depending on
the $&(,(-. installed1 there is a potential for saving BL more 'g&@
!
per year by using
an Al,9.= compared with a (*,9.=1 and only ".=L by increasing the thic'ness of a
similar $&(,(-. material.
*he energy paybac' times +EPB*/ of the $&(,(-.s were found to be consistently
lower than three years for all $&(,(-.s with the lowest EPB* value of 9 year :
months achieved by the A2,9.= $&(,(-.. *his value shows the great energy potential
of such an $&(,(-. design. A calculation of the carbon dio%ide emission paybac'
times confirms the higher potential of using an A2,9.= $&(,(-. by decreasing the
E&PB* by >!L and >CL when compared with A2,; and (*,9.= $&(,(-.s
respectively.
!"!
*he potential annual monetary savings of a ;m
!
$&(,(-.s installation vary
considerably between homes using fossil fuelled or electric heating systems.
Partially replacing a gas heating system with an $&(,(-. would result in a ma%imum
monetary paybac' period +PBP/ of !C.>1 9!.! and B.9 years for the (*,9.=1 A2,; and
A2,9.= respectively considering the collectors5 capital cost only. Assuming $&(,(-.
capital costs including all the components for installation resulted in increasing PB*
of !;1 ;# and =:L for the (*,9.=1 A2,; and A2,9.= respectively. $n both PB* studies
the A2,9.= $&(,(-. was found as the most viable solution. *he (*,9.= $&(,(-. was
a less attractive or viable alternative when installed in con0unction with a gas central
heating system as the customer would not realise any significant economic benefit over
its estimated life time of !" years.
, All results point to the A2,9.= $&(,(-. as the most viable alternative to operate in
(cottish weather conditions. $t would therefore be recommended to use an A2,9.= $&(,
(-. for further investigations due to its energy and environmental benefits.
-ntegration and potential benefits
*his section e%plored the potential installation and integration of the $&(,(-. under
study into recent advances in constructions practice and its potential benefits for
builders and households.
, $t was found that recent government initiatives in the U< help support developers
towards the e%pansion of Dero,carbon development and enhanced the integration of
micro,generation technologies into buildings. ?oof mounted (-.s were found to be
installed in many best practice e%amples of low and Dero,carbon developments such as
the <ingspan 2ighthouse or ?uralEED suggesting that there is a strong potential for this
type of technology to participate towards Dero,carbon homes.
, Based on the predominantly installed commercial types of flat,plate collector in
(cotland1 best practice for installing the $&(,(-. appeared to be a direct sealed
installation. *his installation would use mains water supply as the heat transfer fluid
bac'ed,up by an Uinstantaneous5 in,line heat source for periods of low insolation or as
top,up heating provision.
!";
, Potential for integrating the $&(,(-. in future (cottish dwelling design was a final
milestone of this research pro0ect. $t was found that modern methods of construction
and1 more specifically1 e%amples of structural insulated roofing system +($?(/ or
(tructural $nsulated Panels +($P(/ were recently introduced as contributing to Low
arbon !uild and could allow the direct integration of the IS$S,#
inside the roof panel. 'hus contributing to complyiance with planning and
building regulations. Such installation would result not only in faster and
safer integration+ through off$site construction+ but also in higher
performance of the IS$S,# due to the high insulation performance of
such roof systems. *his would also provide a substantial value to developers
proposing such pre,integrated micro,renewable systems to customers resulting in
achieving higher levels of the &ode for (ustainable .omes.
&.2 C2.:-i57:i2. :2 C.2A4+>(+
@riginal contribution within the conte%t of the presented research can be identified in
the following sub,sectionsF
&nno-ation5
*he collector investigated is a new prototype enhancing the thermal efficiency of the
$&(,(-. through the use of more efficient materials while lowering the capital costs
down to the order of T;"" compared to commercially available collectors priced at over
T!="".
<ethodolo!y and feasibility5
2aboratory and field e%periments provided detailed information of the $&(,(-.
behaviour in different situations. *he field e%periments showed the feasibility of such
heaters in (cottish weather conditions. *ypical annual savings of 99"B '-h can be
achieved for a ;m
!
domestic system corresponding to a ma%imum annual saving of T>"
while partially replacing a gas heating system with an $&(,(-.1 and up to T9;> while
replacing an electric heating system.
*ew predictin! tool5
*he innovative modelling tool developed can predict the bul' water temperature of the
$&(,(-. for any location in the world and therefore predict the $&(,(-.
!">
performance. $&(,(-. design parameters can be changed by the user for their own
research resulting in an evolving programme.
Desi!n impro-ement B (osts5
$mprovements of the $&(,(-. design were suggested through the &3D analysis while
'eeping the costs to a minimum.
)iable prototype5
*he $&(,(-. prototype showed a high commercial potential due to its environmental
and monetary benefits resulting in a viable commercialisation of the prototype.
$mprovements to the $&(,(-. design were also suggested through the 2&A while
lowering the capital cost of the final unit.
)iable installation5
*he $&(,(-. prototype showed potential for integration into commonly used solar
water heating installations which could result in a viable installation. *his type of
installation could use an electric instantaneous hot water heater as an instantaneous
bac',up heater.
<arket -al$e5
*he $&(,(-. was shown to have a real mar'et value. *o date1 there is no commercial
$&(,(-. on the solar thermal mar'et resulting in a niche product. *he confidence in
the design resulted in patenting the $&(,(-.. *he $&(,(-. could also have a high
mar'et value when integrated into modern methods of construction such as roof panels.
&.3 P2:+.:i,4 37:7-+ A2-C* L22Ci.( 5,6C; 422Ci.( ,1+,>
*he avenues for possible future research can be classified as followsF
Thermal macro model potential impro-ements
*he thermal macro model developed could be implemented in the Oisual Basic program
and could provide a better user interface. 3urther improvements of the thermal model
could be incorporated. *he thermal stratification model could be developed for real
weather conditions. @ther improvements such as the modelling of water draw,off could
be integrated.
!"=
$rther (D modellin!
$n light of the &3D modelling results presented1 the five,fin collector performed
generally better than the original four,fin collector. *herefore1 this design could be
suggested as a new arrangement for the collector assuming that the addition of one fin
has a very low additional cost. *he 2&A study also recommended the use of an
aluminium $&(,(-.s of 9.= thic'ness incorporating e%tended heat transfer surfaces for
further investigations due to its energy and environmental benefits. 3uture wor' could
then be carried out through &3D analysis for a five fin aluminium $&(,(-. of 9.=mm
thic'ness. 3urther study assessing the structural integrity of the aluminium $&(,(-.s
of 9.= thic'ness should also be underta'en.
Life cycle assessment C !oin! f$rther
*he eco,profile of an $&(,(-. would be different for any given location in the world.
(cotland being a marginal weather location1 somewhere with greater solar radiation
potential would result in a better eco,profile of the $&(,(-. by achieving greater
energy savings and therefore further reducing carbon emissions. *his $&(,(-. would
therefore have better applications in a sunnier climate.
*he variability between 2&A studies ma'es the comparison between studies difficult.
.owever1 it would be interesting to aggregated indices such as *silingiridis et al +!"">/
in future 2&A research in order to compare with other studies.
+otential inte!ration C !oin! f$rther
.ealth and safety issues should be addressed in order to show the actual potential of the
$&(,(-. integration into dwellings and the fulfilment of the health and safety
constraints towards such systems. @ther issues such as planning permissions constraints
could be assessed.
*he integration of any (-. has a direct impact on a building including its performance
and aesthetics. $ntegrating an $&(,(-. should not devalue the property and should
contribute to achieve planning and regulatory requirements. *herefore further study on
integration of such heaters in ($?( roof structure would be recommended.
*he integration of the $&(,(-. into a trial house would be a good step in order to
conduct seasonal performance with draw,off. (uch a study of the $&(,(-. could be
conducted in (cotland to assess the actual potential of the collector throughout the year.
!"B
REFERENCES
APPENDICES
!":
APPEND$` AF 72@((A?W @3 *E?(
APPEND$` BF 2$(* @3 PUB2$&A*$@N(
APPEND$` &F *.E?A2 @DE2(
!"C
APPENDI! A
GLOSSAR OF TERMS
)nerg+ mix* *he combination of sources used to provide energy at any given time and
place.
Traditional biomass0 $t represents unprocessed biomass including forest product waste1
collected fuel wood1 agricultural waste and animal dung that are burned in stoves or
furnaces to provide heat energy for coo'ing1 heating and agricultural and industrial
processing1 typically in rural areas.

/egree!da+s0 *he rate of heat loss from a building1 +related to the building fabric/ and
the temperature difference between the inside and outside of a building , the greater the
temperature difference the more heat will be lost.
&tmospheric transparenc+ window0 *he atmosphere behaves li'e a blac' body
throughout the infrared spectrum with the e%ception of the wavelength interval between
C ,9>m. *he so,called Atmospheric *ransparency -indow is located between these
two e%tremes and is characterised by a high transparency to infrared thermal radiation
and therefore by low energy emission.
)nvironmentF *he term environment represents the e%ternal conditions including
temperature1 wind speed1 irradiation1 cloud cover etc with which the (-. interacts.
Clear sk+ index0 &orresponding to the ratio of the global horiDontal irradiance to the
global horiDontal irradiance under clear s'y conditions.
'olar fraction0 *he solar fraction is the amount of energy provided by the solar
technology divided by the total energy required.
!"#
APPENDI! B
LIST OF PUBLICATIONS
&urrie1 4.$.1 7arnier1 &.1 uneer1 *.1 .enderson1 D.1 7rassie1 *. +!"":/ &omparison of
3inned Aluminium and (teel $ntegrated &ollector (torage (olar -ater .eater1
(@2A?$( !"":1 ;rd $nternational &onference on (olar ?adiation and Day 2ighting1
New Delhi.
&urrie1 4.$.1 7arnier1 &.1 uneer1 *.1 7rassie1 *.1 .enderson1 D. +!""C/ odelling bul'
water temperature in integrated collector storage systems. Building (ervices
Engineering ?esearch and *echnology +Published/
7arnier1 &.1 &urrie1 4.$.1 uneer1 *. and 7irard1 A. +!""C/ &omputational study of an
integrated collector storage solar water heater1 EU?@(UN !""C1 $nternational
&onference on (olar .eating1 &ooling and Buildings1 2isbon.
7arnier1 &.1 &urrie1 4.$. and uneer1 *. +!""C/ $ntegrated collector storage solar water
heaterF *emperature stratification. Applied Energy +Accepted for publication/
7arnier1 &.1 &urrie1 4.$. and uneer1 *. +!""C/ $ntegrated collector storage solar water
heaterF 2ife &ycle Assessment. 4ournal of &leaner Production +&ommunicated for
publication/
7arnier1 &.1 uneer1 *. and &urrie1 4.$. +!""C/ athematical modelling of integrated
collector storage systems under real weather conditions. American (ociety of
echanical Engineers +&ommunicated for publication/
!9"
APPENDI! C
SOLAR THERMAL ENERG SSTEM FOR HOUSING (STESH)
FIELD THERMAL MODELS
'T)'H 1 .ield model program
Fi(. C1* C,9:i2. 23 STESH 9-2(-,88+ A+4628+ 9,(+
(ub compute"9+/
&ollsiDe G (heets+a(ite,(ystem $nputa/.?ange+a7!!a/.Oalue
Depth G (heets+a(ite,(ystem $nputa/.?ange+a7!;a/.Oalue
*hic'ness G (heets+a(ite,(ystem $nputa/.?ange+a7!>a/.Oalue
2ongueur G (heets+a(ite,(ystem $nputa/.?ange+a7!=a/.Oalue
Oolume G &ollsiDe K Depth
Ac G 2ongueur K *hic'ness
Peri G ! K +*hic'ness I 2ongueur/
finarea G "
H Depends on choice
$f (heets+a(ite,(ystem $nputa/.?ange+a7!:a/.Oalue G aNoa *hen finarea G "
$f (heets+a(ite,(ystem $nputa/.?ange+a7!:a/.Oalue G aWesa *hen finarea G > K ".""! K
2ongueur
!99
$n*w G (heets+a(ite,(ystem $nputa/.&ells+9#1 :/.Oalue I !:;.9=
(heets+a(ite,(ystem $nputa/.&ells+!"1 :/.Oalue G $n*w
*w G (heets+aeteorological $nputa/.&ells+C1 !9/.Oalue H(ite,(ystem $nput water temp.
3or % G C *o 9!:
*am G (heets+aeteorological $nputa/.&ells+%1 B/.Oalue
Oel G (heets+aeteorological $nputa/.&ells+%1 :/.Oalue
&loud&over G (heets+aeteorological $nputa/.&ells+%1 C/.Oalue
?efidnight G (heets+aeteorological $nputa/.&ells+%1 ;/.Oalue
Ep G ".#=
Aw G 9 Hm!
Angle G (heets+a(ite,(ystem $nputa/.?ange+a.9>a/.Oalue
7in G ".C K (heets+aeteorological $nputa/.&ells+%1 !"/.Oalue
mcp G (heets+a(ite,(ystem $nputa/.&ells+!#1 :/.Oalue
'material G (heets+a(ite,(ystem $nputa/.&ells+;"1 :/.Oalue
mcwater G !"C.;!; H'4J<
mcinsul G !.C#: H'4J<
mcwood G ;:.;! H'4J<
mcglaDing G 9".=B H'4J<
mcin G ;.;:B;: H'4J<
*a G *am I !:;.9=
H*s'y G *a
hwind G C.== I !.=B K Oel
Eg G ".CC
H $nitialisation of *p Q *c
*p G *w I ".=
*c G *p , ++*p , *a/ J ;/
deltime G ;B"
3or iminute G 9 *o 9"
H(tart iterative loop
3or i G 9 *o 9"

*dpclear G +".:9"= K +*a , !:;.9=/ , ".:CB>/
*dpovercast G +".C!;9 K +*a , !:;.9=/ , ".C99#/
!9!

$f " \G &loud&over \ ; *hen
*dp G *dpclear
Emis( G ".:>: I ".=#> K +*dp J 9""/ , ".==9 K ++*dp J 9""/ b !/
$f ; \G &loud&over \G = *hen
*dp G +*dpclear I *dpovercast/ J !
Emis( G +".:>: I ".=#> K +*dp J 9""/ , ".==9 K ++*dp J 9""/ b !// K +9 I "."!!> K ".=
, "."";= K +".= b !/ I "."""!C K +".= b ;//
$f = \ &loud&over *hen
*dp G *dpovercast
Emis( G +".:>: I ".=#> K +*dp J 9""/ , ".==9 K ++*dp J 9""/ b !// K +9 I "."!!> ,
"."";= I "."""!C/
End $f
End $f
End $f

Es'y G 9 I ++9":#=! K +9 , Emis(// J ++*a b !/ , BC".C K *a I :;=#>.#/
3s'y G 9 J ++9 J Eg/ I +9 J Es'y/ , 9/

&oordoA G 9= K ?efidnight
&oordoB G ".:99 I ".""=B K *dp I "."""":; K *dp b !
&oordo& G &os++&oordoA/ b +9 J >//

*s'y G *a K +&oordoB I "."9; K &oordo&/

Ucar G +3s'y K "."""""""=B: K +*c b > , *s'y b >// J +*c , *s'y/
Ucac G hwind
Uca G Ucac I Ucar
BEmm G +9 J Eg/ I +9 J Ep/ , 9
Upcr G +"."""""""=B: K +*p I *c/ K +*p b ! I *c b !// J BEmm
H&alculation of ?ayliegh number for air cavity
*fag G +*p I *c/ J !
(heets+aAirPropa/.?ange+a49;a/.Oalue G *fag
visair G (heets+aAirPropa/.?ange+a_9Ca/.Oalue
'fag G (heets+aAirPropa/.?ange+a?9Ca/.Oalue
visair G visair K "."""""9
!9;
?aag G +#.C9 K +9 J *fag/ K +*p , *c/ K 9 b ;/ J +visair/

H&alculation of Nu for air cavity by .olland ?egression
term9 G 9.>> K +9 , +9:"C J ?aag K &os+Angle///
term! G 9 , ++9:"C K (in+9.C K Angle/ b 9.B/ J +?aag K &os+Angle///
term; G ++?aag K &os+Angle/ J =C;"/ b +9 J ;// , 9

term> G Application.a%+term91 "/
term= G Application.a%+"1 term;/
Nuag G 9 I term! K term> I term=
Hcalculation of total conductance form plate to glass cover
Upcc G Nuag K 'fag K ".""9 J +"."!=/
Upc G Upcc I Upcr
H&alculation of ?ayleigh Number for -ater
(heets+a-aterPropa/.?ange+a_!>a/.Oalue G ".= K +*w I *p/
(peheatw G (heets+a-aterPropa/.?ange+aAA!#a/.Oalue
Betaw G (heets+a-aterPropa/.?ange+aA4!#a/.Oalue
Prw G (heets+a-aterPropa/.?ange+aA<!#a/.Oalue
Oisw G "."""""9 K (heets+a-aterPropa/.?ange+aA&!#a/.Oalue
denw G (heets+a-aterPropa/.?ange+a-!#a/.Oalue
'w G ".""9 K (heets+a-aterPropa/.?ange+aAE!#a/.Oalue
?aw G Prw K #.C9 K Betaw K "."""""9 K +9 b ;/ K +*p , *w/ J ++Oisw J denw/ b !/

$f +?aw \G "/ *hen ?aw G (qr++?aw/ b !/

H&alcuating water conductance
Nuw G ".=B K ++?aw K &os+Angle// b ".!=/ K ++9 J "."=/ b +9 J B//
hw G Nuw K 'w J 9
Uw G hw
HBac' losses
UAbloss G ".;=
H $terations for *p Q *c
*p G +7in I Upc K *c I Uw K *w/ J +Upc I Uw/
*c G +Upc K *p I Ucac K *a I Ucar K *s'y/ J +Upc I Ucac I Ucar/
Ne%t i
!9>
_pw G Uw K +*p , *w/
_pc G Upc K +*p , *c/
H3in lengh calculation and improvement factor
smallm G ;;= K ! K +2ongueur I *hic'ness/ J +'material K 2ongueur K *hic'ness/
smallm G (qr+smallm/
ml G smallm K Depth

Big G (qr+;;= K Peri K 'material K Ac K +*p , *w//
HAf G ! K +Depth K *hic'ness/ I ! K +Depth K 2ongueur/ I +2ongueur K *hic'ness/
3or 0 G : *o ;!
2ength G (heets+a3in Effecta/.&ells+01 9/.Oalue
tetab G Application.&osh+ml/
teta G Application.&osh+smallm K +Depth , +2ength K ".""9///
teta% G +teta J tetab/ K Abs+*p , *w/
*fin G teta% I *w
?afin G Prw K #.C9 K Betaw K "."""""9 K ++Depth J &os+Angle// b ;/ K +*fin , *w/ J
++Oisw J denw/ b !/
Nufin G ".BC I +".B: K ?afin b ".!C/ J +9 I +".>#! J Prw/ b +# J 9B// b +> J #/
hfin G Nufin K 'w J +Depth J &os+Angle//
_%fin G hfin K +*fin , *w/ K finarea
(heets+a3in Effecta/.&ells+01 !/.Oalue G teta%
(heets+a3in Effecta/.&ells+01 ;/.Oalue G *fin , !:;.9=
(heets+a3in Effecta/.&ells+01 >/.Oalue G hfin
(heets+a3in Effecta/.&ells+01 =/.Oalue G _%fin
tetao G (heets+a3in Effecta/.&ells+:1 ;/.Oalue , (heets+a3in Effecta/.&ells+;!1 ;/.Oalue
limittetao G ".9 K tetao
(heets+a3in Effecta/.&ells+:1 B/.Oalue G tetao
(heets+a3in Effecta/.&ells+:1 :/.Oalue G limittetao

H3in Effeciency
2c G +2ength K ".""9/
m2c G smallm K 2c
Ap G 2c K *hic'ness
!9=
3ineffi G Application.*anh+m2c/ J +m2c/
(heets+a3in Effecta/.&ells+01 9=/.Oalue G 3ineffi

Ne%t 0
Dim D+!=/
Dim E+!=/
D+"/ G (heets+a3in Effecta/.&ells+:1 9/.Oalue
D+9/ G (heets+a3in Effecta/.&ells+C1 9/.Oalue
D+!/ G (heets+a3in Effecta/.&ells+#1 9/.Oalue
D+;/ G (heets+a3in Effecta/.&ells+9"1 9/.Oalue
D+>/ G (heets+a3in Effecta/.&ells+991 9/.Oalue
D+=/ G (heets+a3in Effecta/.&ells+9!1 9/.Oalue
D+B/ G (heets+a3in Effecta/.&ells+9;1 9/.Oalue
D+:/ G (heets+a3in Effecta/.&ells+9>1 9/.Oalue
D+C/ G (heets+a3in Effecta/.&ells+9=1 9/.Oalue
D+#/ G (heets+a3in Effecta/.&ells+9B1 9/.Oalue
D+9"/ G (heets+a3in Effecta/.&ells+9:1 9/.Oalue
D+99/ G (heets+a3in Effecta/.&ells+9C1 9/.Oalue
D+9!/ G (heets+a3in Effecta/.&ells+9#1 9/.Oalue
D+9;/ G (heets+a3in Effecta/.&ells+!"1 9/.Oalue
D+9>/ G (heets+a3in Effecta/.&ells+!91 9/.Oalue
D+9=/ G (heets+a3in Effecta/.&ells+!!1 9/.Oalue
D+9B/ G (heets+a3in Effecta/.&ells+!;1 9/.Oalue
D+9:/ G (heets+a3in Effecta/.&ells+!>1 9/.Oalue
D+9C/ G (heets+a3in Effecta/.&ells+!=1 9/.Oalue
D+9#/ G (heets+a3in Effecta/.&ells+!B1 9/.Oalue
D+!"/ G (heets+a3in Effecta/.&ells+!:1 9/.Oalue
D+!9/ G (heets+a3in Effecta/.&ells+!C1 9/.Oalue
D+!!/ G (heets+a3in Effecta/.&ells+!#1 9/.Oalue
D+!;/ G (heets+a3in Effecta/.&ells+;"1 9/.Oalue
D+!>/ G (heets+a3in Effecta/.&ells+;91 9/.Oalue
D+!=/ G (heets+a3in Effecta/.&ells+;!1 9/.Oalue

E+"/ G (heets+a3in Effecta/.&ells+:1 =/.Oalue
E+9/ G (heets+a3in Effecta/.&ells+C1 =/.Oalue
!9B
E+!/ G (heets+a3in Effecta/.&ells+#1 =/.Oalue
E+;/ G (heets+a3in Effecta/.&ells+9"1 =/.Oalue
E+>/ G (heets+a3in Effecta/.&ells+991 =/.Oalue
E+=/ G (heets+a3in Effecta/.&ells+9!1 =/.Oalue
E+B/ G (heets+a3in Effecta/.&ells+9;1 =/.Oalue
E+:/ G (heets+a3in Effecta/.&ells+9>1 =/.Oalue
E+C/ G (heets+a3in Effecta/.&ells+9=1 =/.Oalue
E+#/ G (heets+a3in Effecta/.&ells+9B1 =/.Oalue
E+9"/ G (heets+a3in Effecta/.&ells+9:1 =/.Oalue
E+99/ G (heets+a3in Effecta/.&ells+9C1 =/.Oalue
E+9!/ G (heets+a3in Effecta/.&ells+9#1 =/.Oalue
E+9;/ G (heets+a3in Effecta/.&ells+!"1 =/.Oalue
E+9>/ G (heets+a3in Effecta/.&ells+!91 =/.Oalue
E+9=/ G (heets+a3in Effecta/.&ells+!!1 =/.Oalue
E+9B/ G (heets+a3in Effecta/.&ells+!;1 =/.Oalue
E+9:/ G (heets+a3in Effecta/.&ells+!>1 =/.Oalue
E+9C/ G (heets+a3in Effecta/.&ells+!=1 =/.Oalue
E+9#/ G (heets+a3in Effecta/.&ells+!B1 =/.Oalue
E+!"/ G (heets+a3in Effecta/.&ells+!:1 =/.Oalue
E+!9/ G (heets+a3in Effecta/.&ells+!C1 =/.Oalue
E+!!/ G (heets+a3in Effecta/.&ells+!#1 =/.Oalue
E+!;/ G (heets+a3in Effecta/.&ells+;"1 =/.Oalue
E+!>/ G (heets+a3in Effecta/.&ells+;91 =/.Oalue
E+!=/ G (heets+a3in Effecta/.&ells+;!1 =/.Oalue

limdist G (heets+a3in Effecta/.&ells+:1 C/.Oalue

*otal G E+"/
distance G D+"/
&nt G "
Do -hile +distance \ limdist/
&nt G &nt I 9
distance G D+&nt/
*otal G *otal I E+&nt/
2oop
*otal G *otal , E+&nt/
_fin G *otal

!9:
$mprovfactor G 9 I +_fin J _pw/

3in* G (heets+a3in Effecta/.&ells+;>1 ;/.Oalue

_useful G ".""9 K deltime K +Uw K Aw K +*p , *w/ I _fin , UAbloss K +*w , *a//
*w G *w I _useful J +mcwater I mcp I mcinsul I mcwood I mcglaDing I mcin/

Ne%t iminute
(heets+aeteorological $nputa/.&ells+%1 !!/.Oalue G Uca
(heets+aeteorological $nputa/.&ells+%1 !;/.Oalue G Upc
(heets+aeteorological $nputa/.&ells+%1 !>/.Oalue G Uw
(heets+aeteorological $nputa/.&ells+%1 !=/.Oalue G *p
(heets+aeteorological $nputa/.&ells+%1 !B/.Oalue G *c
(heets+aeteorological $nputa/.&ells+%1 !:/.Oalue G _pw
(heets+aeteorological $nputa/.&ells+%1 !C/.Oalue G _pc
(heets+aeteorological $nputa/.&ells+%1 !#/.Oalue G _fin
(heets+aeteorological $nputa/.&ells+%1 ;"/.Oalue G _useful
(heets+aeteorological $nputa/.&ells+%1 ;9/.Oalue G $mprovfactor
(heets+aeteorological $nputa/.&ells+% I 91 !9/.Oalue G *w

Ne%t %
End (ub
!9C