SolarEne,T¢~,Vol 10,No.4.pp 313-332.

t983 110384192X/83/04031~-2050300/0
PrintedinGreatBrilain ~ 1983PergamonPres~lhl

LOW TEMPERATURE LATENT HEAT THERMAL

ENERGY STORAGE: HEAT STORAGE
MATERIALS

A. ABHAT
Institut for Kernenergetik und Energiesysteme (IKE), University of Stuttgart, Stuttgart, FRO

(Receiced 6 November 1981; accepted 18 May 1982)

Abstract--Heat-of-fusion storage materials for low temperature latent heat storage in the temperature range
0-120°C are reviewed. Organic and inorganic heat storage materials classified as paraffins, fatty acids, inorganic
salt hydrates and eutectic compounds are considered. The melting and freezing behaviour of the various substances
is investigated using the techniques of Thermal Analysis and DifferentialScanning Calorimetry. The importance of
thermal cycling tests for establishingthe long-term stability of the storage materials is discussed. Finally, some data
pertaining to the corrosion compatibilityof heat-of-fusion substances with conventional materials of construction is
presented.

1. INTRODUCTION to cover a variety of low temperature applications, such

Efficient and economical heat storage is the key to the
effective and widespread utilization of solar energy for
low temperature thermal applications. Amongst the
various heat storage techniques of interest, latent heat
storage is particularly attractive due to its ability to provide
a high energy storage density and its characteristics to
store heat at a constant temperature corresponding to the
phase transition temperature of the heat storage substance.
The term "Latent Heat Storage", as we generally
understand it today, applies to the storage of heat as the
latent heat of fusion in suitable substances that undergo
melting and freezing at a desired temperature level.
Consequently it is also often called the "Heat-of-Fusion"
storage. Typical heat-of-fusion storage substances well-
known to all of us are ice, paraffin or Glauber Salt. The
term "latent heat storage" may also be applied to include
the heat stored in substances, such as Diaminopen-
taerythritol, wherein heat is stored as the heat of crystal-
lization, as the substance is transformed from one solid
phase to another. The stored heat is recovered in a
likewise manner as the original solid phase is regained.
Excluded in the present definition of "latent heat
storage" is, however, the heat stored in materials that
undergo a liquid-to-vapor phase transition, e.g. water-to-
steam. Although the latter phase transitions are asso-
ciated with a latent heat of phase transition that is almost
an order-of-magnitude higher than that for solid-to-liquid
or solid-to-solid phase change, the practical problems of
storing a gaseous phase and the necessity of pressurized
containers for this purpose rule out their potential utility.
The review article relates to the discussion of heat-of-
fusion storage, a technique which is also of the greatest
current practical value. A temperature range of 0-120°C
as domestic hot water production, direct or heat-pump
assisted space heating, green house heating, solar cool-
ing, etc. is considered. It should, however, be
emphasized here, that although heat storage in solid-solid
phase transitions is much less understood today, it does
hold out future promise.
2. BASIC REQUIREMENTSOF A LATENT HEAT
THERMAL ENERGY STORET
Any latent heat thermal energy storage system must
possess at least the following three components: (1) a
heat storage substance~ that undergoes a solid-to-liquid
phase transition within the desired operating temperature
range and wherein the bulk of the heat added is stored as
the latent heat of fusion; (2) a containment for the
storage substance; (3) a heat exchanging surface for
transferring heat from the heat source to the heat storage
substance and from the latter to the heat sink, e.g. from
the solar collector to the heat storage substance to the
load loop.
The development of a latent heat thermal energy
storage (LTES) system hence involves the understanding
of two essentially diverse subjects: heat storage materi-
als and heat exchangers, The flowchart in Fig. 1 provides
an overview of the different stages that may be involved
in the development of an L T E S system and of the speci-
alized problems that need to be tackled.
The present paper is confined to the discussion of
heat-of-fusion storage materials. The measurement
techniques are outlined and the short and long-term
behaviour of various heat storage substances is presen-
ted. Finally, some information pertaining to the com-
patibility of the storage substances with construction
materials is provided.

fThe following abbreviations are used in the text:
LTES = latent heat thermal energy storage.
PCM = phase change (heat-of-fusion) storage material.
3. HEAT-OF-FUSION STORAGEMATERIALS
Desired material properties
A large number of organic and inorganic substances
are known to melt with a high heat of fusion in any

SEVok30,No.4--8 3 13
314 A. ABHAT

LATENT HEAT THERMAL ENERGY STORAGE SYSTEMS

f I _
HEATEXCHANGERDEVELOPMENT ]
I MATERIALINVESTIGATION q

CHOICE OF MATERIALS I CHOICE OF

IN THE APPROPRIATE HEAT EXCHANGER
.... TEMPERATURE RANGE CONCEPT(S)

1
1
HEAT STORAGE I THERMAL EXPERIMENTAL 7
MATERIALS - - MATERIALS? ANALYSIS, INVESTIGATION]
PARAMETRIC
INVESTIGATION
F- I I LABORATORY
MODELS J
PROPERTY FREEZING
CHA-
DATA RACTERISTICS
1 INTEGRATION
WITH HEATING/
COOLING
SYSTEMS
PILOT UNITS FIELDTESTS
~Z~ [CO~;~I~I~ITYI I SIMULATION I L
n~,~S~
vKtu~,
i FORMANCEOFJ
yr.-l/
LJCOMPARISONS

HEAT STORE I FINAL COST

LONGLIFE ~l AND HEATING/l
I COOLING l
ANALYS S _J
CONSIDERATIONS - - L SYSTEM J
COMMERCIAL ~
PRODUCTI ON

Fig. 1. Flowchart providingan overviewof the various steps in the developmentof a latent heat storage system.

required temperature range, e.g. 0-120°C. However, for
their employment as heat storage materials in LTES
systems, these phase change materials must exhibit cer-
tain desirable thermodynamic, kinetic and chemical pro-
perties. Moreover, economic considerations of cost and
large-scale availability of the materials must be con-
sidered. The various criteria that govern the selection of
phase change heat storage materials are summarized in
Table I.
Candidate heat storage materials
It is quite apparent that no single material can fully
satisfy the long list of criteria listed in Table 1. Trade-
offs are hence made in the selection of candidate heat
storage materials. Within the operating temperature
range of 0-120°C, candidate PCMs are grouped into the
families of organic and inorganic compounds and their
eutectica, as seen in Fig. 2. Sub-families of organic
compounds include paraffin and non-paraffin organics.
Figures 3(a, b), respectively show the latent heat of
fusion per unit mass and per unit volume of some heat
storage materials of interest in the temperature range
0-120°C[1, 19, 23, 29]. Of particular interest is Fig. 3(b),
which provides information on the compactness of an
LTES system for a given amount of energy storage.
Figure 3(b) shows that the organic compounds have a
volumetric latent heat storage capacity in the range 125-
200 kJ/dm ', whereas that of the salt hydrates is almost
twice as much, between 250-400 kJ/dm 3.
Notwithstanding the disadvantages concerning
volume requirements, the organic substances serve as
important heat storage materials due to the several desir-
able properties they possess in comparison with in-
organic compounds. Some of these advantages include
their ability to melt congruently, their self-nucleating
properties, and their compatibility with conventional
materials of construction[3]. We shall now examine
some of the important heat storage material groups in
some detail.
Para~ns. Paraffins, as we normally understand them,
are substances having a waxy consistency at room tem-
perature. Chemically speaking, paraffin waxes consist
primarily of straight-chain hydrocarbons with only a
small amount of branching, such as 2-methyl groups,
near the end of the chain. Paraffins contain in them one
major component called alkanes, characterized by
C,H2,+z; the n-alkane content in paraffin waxes
usually exceeds 75 per cent and may reach lOOper
cent[13]. Depending on the chain length of the alkane in
the paraffin, paraffins may be even-chained (n-paraffin)
or odd-chained (iso-paraffin).
Pure paraffins contain only alkanes in them, for
example the well-known paraffin octadecane, C~,H38.
The melting point of the alkanes increases with the
increasing number of carbon atoms; alkanes containing
14--40 C-atoms possess melting points between 6 and
80°C and are generally termed as paraffins. Commercial
waxes, on the other hand, may have a range of about 8-15
carbon-numbers [13].
In their solid phase, paraffins are generally found in
two allotropic modifications that differ in their physical
properties and the crystal structure. The primary
modification existing at higher temperature, i.e. slightly
above the melting point of the substance, is soft and
 Low temperature latent heat thermal energy storage: heat storage materials 315

Table 1. Desired properties of phase change heat storage materials [3]

A. THERMODYNAMIC CRITERIA

THE PHASE CHANGE MATERIAL SHOULD POSSESS

O A MELTING POINT IN THE DESIRED OPERATING TEMPERATURE RANGE

O HIGH LATENT HEAT OF FUSION PER UNIT MASS, SO THAT A LESSER

AMOUNT OF MATERIAL STORES A GIVEN AMOUNT OF ENERGY

O HIGH DENSITY, SO THAT A SMALLER CONTAINER VOLUME HOLDS THE

MATERIAL

O HIGH SPECIFIC HEAT TO PROVIDE FOR ADDITIONAL SIGNIFICANT

SENSIBLE HEAT STORAGE EFFECTS

O HIGH THERMAL CONDUCTIVITY~ SO THAT THE TEMPERATURE GRADIENTS

REQUIRED FOR CHARGING AND DISCHARGING THE STORAGE MATERIAL
ARE SMALL

O CONGRUENT MELTING: THE MATERIAL SHOULD MELT COMPLETELY ~O

THAT THE LIQUID AND SOLID PHASES ARE IDENTICAL IN COMPOSI-
TION. OTHERWISE~ THE DIFFERENCE IN DENSITIES BETWEEN SOLID
AND LIQUID CAUSE SEGREGATION RESULTING IN CHANGES IN THE
CHEMICAL COMPOSITION OF THE MATERIAL /20/

O SMALL VOLUME CHANGES DURING PHASE TRANSITION, SO THAT A

SIMPLE CONTAINMENT AND HEAT EXCHANGER GEOMETRY CAN BE USED.

B, KINETIC CRITERIA

THE PHASE CHANGE MATERIAL SHOULD EXHIBIT

I LITTLE OR NO SUPERCOOLING DURING FREEZING, THE MELT SHOULD
CRYSTALLIZE AT ITS THERMODYNAMIC FREEZING POINT. THIS IS
ACHIEVED THROUGH A HIGH RATE OF NUCLEATION AND GROWTH RATE
OF THE CRYSTALS. AT TIMES, THE SUPERCOOLING MAY BE SUPPRESSED
BY INTRODUCING NUCLEATING AGENT OR A "COLD FINGER" IN THE
STORAGE MATERIAL /20/.

C. CHEMICAL CRITERIA

THE PHASE CHANGE MATERIAL SHOULD SHOW

I CHEMICAL STABILITY

I NO CHEMICAL DECOMPOSITION, SO THAT A HIGH LTES SYSTEM LIFE

IS ASSURED

I NON-CORROSIVENESS TO CONSTRUCTION MATERIALS

e THE MATERIAL SHOULD BE NON-POISONOUS, NON-FLAMMABLE AND

NON-EXPLOSIVE

D. ECONOMIC CRITERIA
THE PHASE CHANGE MATERIAL SHOULD BE

I AVAILABLE IN LARGE QUANTITIES

I INEXPENSIVE

Paraffins 1
Compounds
Non-parafhn 1
Orgamcs ~]
OrgonlCS

Eutectic~" )

I,~orgo~n,cs compounds I

Eutectlc~: }
Fig. 2. Families of phase change heat storage materials[3].
316 A. ABHAT

450 l i
A, PARAFFINS
1, PARAFFIN 5913
2, PARAFFIN 6106
400 3. PARAFFIN 5838
Paraffins
4, PARAFFIN 6035
Non-Paraffin Organics
-~ 350 5. PARAFFIN6403
• Inorganic Compounds 6. PARAFFIN6499
L21
[] Eutectic Compounds 7, OCTADECANE
o
~E
300 25 •
29
37 B. ORGANICCOMPOUNDS
~ 250 22 19 27 28 36 A
• 7 20 8, CAPRILIC ACID
24
33A ,&013 35 5 ~6 9. POLYGLYCOLE 600
c
.o
u'l
200 23• 19
26
•
r, N A18 39•
1U.
11,
CAPRICACID
LAURICACID
u_ 150 A A • 4 12. MYRISTIC ACID
8 10 30 E Ig
A
41 13. POLYGLYCOLE 6000
"6 9
A ~2 •
31• •32 17 38 14. LOXIOL G 32
~ 100 15. PAI241TICACID
T 16. STEARICACID
17. BIPHENYL
18. PROPIONAMIDE
19. NAPTHALENE
I I I I I
20. ACETAMIDE
0 10 20 30 40 50 60 70 80 90 100 110 120
(a) Melting Point [°C] C. IHQRGAN1CCOMPOUNDS

582 k J/din3 21. H20

450 f 22. LICLO3.3H20
' 371 ' ' 23. NAOH.31/2H20
E 25 24. KF.4H20
"E3 400 % 25. LINo3.3H20
22 19 Paraffins
3 27~ 34
Non-Paraffin Organics
26, CACL2,6H20
350 28 27, NA2SOh.IOH20
• Inorganic Compounds 28. NA2CO3,1OH20
E 24
33 [] Eutectic Compounds 29. NA2HPO4.12H20
300' b21 30. ZN(NO3)2.6H20
30
23 31, CA,BR2,6H20
c • 31 32. KF.2H20
250 26
3s%
41 17
A
39
33, NA2S203.5H20
34, NI(NO3)2.6H20
200
~2o 35. NACH3COO.3H20
c &
._(2 719 32 E3 36. NAOH.H20
/>,11 LO 15 18 37. BA(OH)2.8H20
LL 150 19 12 ? 38, MG(NOs)2.6H20
A 1 t0 19
8 14 39. MGCL2.6H20

®
100
"1- p, EUTEGTIC COMPOUNDS

50 - 40. 21,1 WT,% PROPIONAM1DE

+ 74.9 % PALMITIC ACID
3 41. 53 MOL-%MG(NO3)2.6H20
I L a I [ I
+ 47 MOL-%MGCL2.6H20
0 10 20 30 40 50 60 70 80 90 100 110 120
42. 27 MOL-% LINO3 + 68 MOL-%
(b) Melting Point [ °C] NH4NO3 + 5 MOL-%NH4CL

Fig. 3. Latent heat of fusion per unit mass and per unit volume of selected phase change heat storage materials
undergoing a phase transition within the temperature range 0-120°C.

plastic, the individual crystals are needle-shaped. The paraffin mixtures and are not completely refined of oil.
secondary modification existing at lower temperature, i.e. Manufacturers' data on their physical properties as well
below the melting point, is hard and brittle, the crystals as cost data are also included in Table 2.
are disc-shaped. The transition from one crystal form to Non-para#in organics. Although a number of non-
another is reversible [28]. paraffin organics have been included in Figs. 3(a)--(b), we
Paraffins qualify as heat-of-fusion storage materials shall restrict our discussion to the group of materials
due to their availability in a large temperature range and called "Fatty acids". Other non-paraffin organics may,
their reasonably high heat of fusion. Furthermore, they however, be assumed to have properties similar to those
are known to freeze without supercooling [4]. Due to cost of the fatty acids.
considerations, however, only technical grade paraffins Fatty acids. Fatty acids are organic compounds
may be used as PCMs in latent heat stores. Table 2 lists characterized by CH3(CH2h,COOH with heat of fusion
some technical grade materials, which are essentially values comparable to that of paraffins. Some fatty acids
 Low temperature latent heat thermal energy storage: heat storage materials 317

Table 2. Physical properties and cost data of some paraffins

PARAFFIN DISTRIBUTION OIL ~E~I~ HEAT OF FUSION DENSITY SPECIFIC THERMAL COST REF.
OF ~ENT ~I~/~ AT HEAT I Vl TV (1979)
COIk~)UCT
C-ATOMS 20 *C 70 *C AT 1(]0 *C (SOLIDPHASE)
% °C KJ/KG KJ/DM 3 KG/~ 3 KJ/KG.K W/M.K I]M/KG

1,20 2)
_ 1)
C14 4.5 165 20
CL5 - C16 8 153 0,50 2) 20
5913 3) C]_3 - C24 20 22 - 24 189 144 0.900 0.760 2.1 0.21 0.50 4
0CTADECANE C18 0 28 244 189 0.814 0.774 2.16 0,15 150,0 4
6106 3) C16 - C28 5 42 - 44 189 145 0,910 0.765 2.1 0.21 0.70 4
P].16 4) 45 - 48 210 165 0.817 0,786 2.5 0,44 2) 20
5838 3) C20 - C33 < 0.5 48 - 50 189 145 i0.91P 0.769 2.1 0.21 1,00 4
6035 3) C22 - C45 4 58 - 60 189 150 0.gL~O 0,795 2.1 0.21 0,60 4
6403 C23 - CLE5 < 0,5 62 - 6 4 189 150 0.915 0,790 2.1 0,21 1,00 4
6499 3) C2_].- C50 3 66 - 68 189 157 0.930 0.830 2.1 0.21 0,80 4

i) - IMPLIES DATA NOT AVAILABLE 3) MANUFACTURERS OF TECHNICALGRADEPARAFFINS5913, 6106, 5838, 6035,

6403 AA~)6499: 'TER HELLPARAFFIN, HI~'~JRG, FRG

2) COSTESTIMATESARE FOR 1974 (REF. 2(]) 4) MANUFACTURERS OF PARAFFINP116: SUN OIL COMPANY,USA

Table 3. Physical properties and cost data of some fatty acids

MATERIAL MELTING HEAT OF FUSION DENSITY S P E C I F I C THERMAL COST REF,
POINT/RANGE HEAT CONDUCTIVITY (1979)
*C KJ/KG KJ/DM3 KG/DM3 KJ/KG K W/M.K DM/KG

CAPRILIC 16,5 149 128 1.033(10 °C) - 1) 0.148(20 °C) 4

ACID
0.862(8O °C)
CAPRIC 31.5 153 136 0.886(40 *C) 0.149(40 °C) 4
ACID
LAURIC 42 - 44 178 155 0.870(50 *C) 1,6 0.147(50 *C) 2,50 4
ACID

MYRISTIC 54 187 158 0,844(80 *C) 1,6(s) 2) 2.50 4

ACID 2.7(L)
PALMITIC 63 187 159 0.847(80 °C) 0.165(70 *C) 2,30 4
ACID
STEARIC 70 203 191 0.941(40 *C) 2,35(125°C) 0,172(70 *C) 2,O0 4
ACID

1) -IMPLIES DATA NOT AVAILABLE

2) S - SOLID; L - LIQUID

of interest to low temperature LTES applications are
listed in Table 3.
Fatty acids are known to possess a reproducible mel-
ting and freezing behaviour and freeze with little or no
supercooling[4, 19]. They hence qualify as good PCMs.
Their major drawback, however, is their cost which is
2-2.5 times greater than that of paraffins.
Salt-hydrates. Salt hydrates, characterized by
M.nH20, where M is an inorganic compound, form an
important classs of heat storage substances due to their
high volumetric latent storage density (see Fig.
3(b)). In fact, their use as PCMs has been propagated as
early as 1947124]. Table 4 provides a list of some salt
hydrates melting in the temperature range 0-120°C along
with their thermophysical properties and cost data.
The major problem in using salt hydrates as PCMs is
that most of them melt incongruently, i.e. they melt to a
saturated aqueous phase and a solid phase which is
generally a lower hydrate of the same salt. Due to
density differences, the solid phase settles out and
collects at the bottom of the container, a phenomenon
called decomposition. Unless special measures are taken,
Ihis phenomenon is irreversible, i.e. during freezing, the
solid phase does not combine with the saturated solution
to form the original salt hydrate. Another important
problem with salt hydrates is their poor nucleating pro-
perties resulting in supercooling of the liquid salt hydrate
prior to freezing. Suitable measures must be adopted to
eliminate supercooling or reducing it to a minimum.
Typical methods suggested in the literature for this pur-
pose are: (1) addition of nucleating agents that have a
crystal structure similar to that of the parent
substance [25]: (2) using a "cold finger" in the PCM[20];
(3) promoting heterogeneous nucleation by using (rough)
metallic heat exchanging surfaces in contact with the salt
hydrate[4]. We shall examine the influence of these
techniques on the melting and freezing characteristics of
salt hydrates in greater detail at a later stage.
The selection of a salt hydrate as a PCM can be eased
by an understanding of the binary phase diagrams, M-
H20, for which the reader is referred to Ref. [30]. To
illustrate the use of the phase diagrams, two examples
318 A. ABHAT

Table 4. Physical properties and cost data of some salt h rdrates

MATERIAL MELTING HEAT-OF-FUSION DENSITY SPECIFIC HEAT THERMAL COST REF,
POINT CONDUCTIVITY (1979)
*C KJ/KG KJ/DM3 KG/DM3 KJ/KGK W/M.K DM/KG

0.917 (0 *C) 2.09 (S)2) 2.2 (S) 2) .3)

H20 1) 0 333 306 0,998 (20"C 4.18 (L) 2) 0.6 (20"C) 23

KF,4H20 18.5 231 336 1,455 (18"C) 1.84 (s) _4) 23

1,447 (20"C) 2.39 (L)

CACL2,6H20 29.7 171 256 1.710 (25"C) 1,45 (s) 0.36 4

1,496 (L)
NA2S04,10H20 32,4 254 377 1.485 (S) 1,93 (S) 0,544 0,10 4

NA2HP04,12H20 35,0 281 405 1.520 (s) 1,70 (s) 0,514 (32%) 0,95 16
1.442 (L) 1,95 (L) 0,476 (49"C)

ZN(N03)2.6H20 36,4 147 304 2.065 (14"C) 1.34 (s) 2.40 4

2.26 (L)

NA2S203.5H20 48.0 201 322 1.73 (S) 1.46 (S) 0.30 4

1.67 (L) 2.39 (L)
BA(0H)2.8H20 78,0 267 581 2.180 (S) 1,17 (S) 1.75 4

MGCL2,6H20 116,0 165 239 1.57 (20"C) 1.72 (S) 0,20 23

1,442(78"C) 2.82 (L)
1) DATA FOR H20 IS INCLUDED FOR THE SAKE OF COMPARISONS
2) S - SOLID; L - LIQUID
3) NEGLIGIBLE
4) - IMPLIES DATA NOT AVAILABLE

80

L ~,m
c_)
o

e~ 2O
D

- - "1-

%
-20 z z

-40
0 20 t,0 60 80
Weight % Zn(NO3)2

Fig. 4. Phase diagram for Zn(NO0:-H20 system[15].

have been included here: (a) Zn(NO3)-H20 system; (b)

Na2SO4-H20 system. 50
(a) Zn(NO3)2-H20 system. Figure 4 shows the phase
diagram for Zn(NO3)2-H20 between the temperature 4O Liquid ~ Liquid .~
limits from - 4 0 to +80°C[15]. If a liquid mixture of 36' 5
T
~.
....... -r
Zn(NO~)2 and H20 is cooled below 80°C, 5 different 32.4-
o 30 Fp
hydrated salts of Zn(NO3)2, as shown in Fig. 4, crystallize
out of solution. Of these, Zn(NO0.9H,O has an in- B 7 ', ii - N°2SO~ ]
congruent melting point, whereas the remaining 4 20
hydrates have a congruent melting point. The 6-hydrate, E / .......... ....... i
with a melting point of 36.4°C, additionally has a high /-,so,.,°.2°i
latent heat of fusion and is hence well-suited as a heat
storage material.
t
(b) Na2SO4-H20 system. The second example we
wish to discuss here is the well-known storage material
sodium sulphate-10-hydrate or Glauber salt. This 100W/o H20
material has been extensively investigated so far for use Weight Percent No2SO~
in LTES systems. Figure 5 shows a partial phase Fig. 5. Partial phase diagram for Na:SO4-H20 system[9].
 Low temperature latent heat thermal energy storage: heat storage materials
diagram of the sodium sulphate-water system/9].
Na2SO4" 10H20 decomposes peritectically on heating to
32.4°C to yield anhydrous sodium sulphate and a
saturated solution of Na2SOn in water. The solid crystals
containing 44 per cent anhydrous Na2SO4 and 56 per
cent water by weight change to a mixture of 15 per cent
anhydrous Na2SO4 and 85 per cent saturated solution of
Na2SO4 in water (10H20). The 15 per cent anhydrous
Na2SO4 remains insoluble and settles down as a white
bottom sediment. It is not possible to dissolve this
material by the usual expedient of increasing the tem-
perature, because the highest solubility of Na2SO4 in
water occurs at the melting point with solubility decreas-
ing at higher temperatures. Due to the large density
difference between the saturated solution (1350kg/m 3)
and the anhydrous solid Na2SO4 (2550kg/m3), gravity
induced separation results. Consequently melting is
partly incongruent due to segregation/26].
Now, if the mixture of the saturated solution and
anhydrous salt is cooled below 32.4°C, sodium sulphate
can be absorbed in the solution only as rapidly as water
can diffuse through solid sodium sulphate decahydrate to
the anhydrous sodium sulphate particles on which the
decahydrate particles form. Since peritectic solidification
reactions are characteristically much slower than con-
gruent solidification or eutectic solidification, and
because the rate limiting process here is solid state
diffusion, even stirring cannot significantly affect the rate
of absorption of Na2SO4, which therefore remains sett-
led at the bottom of the container/9].
If the phase diagram for the Na2SO4-H20 system is
now carefully followed, it can be seen that the solid state
diffusion as a reaction step can be eliminated by using
excess water in the system. A recommended
319
composition/9] is 30 per cent Na2S04 and 70 per cent
H,.O by weight, corresponding to 68.2 per cent
Na2SO4 • 10H20 and 31.8 per cent excess water. At 15°C
(point 2), this mixture comprises 58.1 per cent
Na2SO4.10H~0 and 41.9 per cent solution of com-
position 10.6 per cent Na2SO4 and 89.4 per cent H20 by
weight. When the temperature of the mixture is raised
beyond 15°C, the solubility of the decahydrate increases
with the increase in temperature up to 32.4°C. At the
liquidus line, the decahydrate completely goes into solu-
tion and no segregation hence takes place. Now if the
solution is cooled from 36°C (point 5 in Fig. 5), deca-
hydrate crystals begin to separate as the solution reaches
the liquidus line (at 30°C). With further reduction in
temperature, the system enters into a two-phase region,
Na2SO4.10H20 and liquid. The amount of each con-
stituent can be easily calculated at any temperature by
using the lever rule. The stored thermal energy in the
system is released when the Na2SO4-10H:O crystals
separate from the solution/9].
Eutectics of organic and inorganic compounds.
Eutectics of organic or inorganic compounds qualify as
latent heat storage materials as they possess a fixed
melting/freezing point. Some candidate materials having
acceptable values of heat-of-fusion are listed in Table 5.
It should be mentioned here that the search for eutectics
is rather recent, and only limited data on thermophysical
properties and cost are presently available.
Other compounds. A material group that deserves
mention for use in LTES systems is the clathrate
hydrate. True clathrate hydrates are continuous solid
water structures containing closed cavities within which
are guest molecules that do not interact strongly with
water. These guest molecules act to stabilize the "ice"

Table 5. Physical properties and cost data of some organic and inorganiceutectic compounds
MATERIAL MELTING HEAT-OF-FUSION SPECIFIC COST REFERENCE
(WEIGHT ~ OF COMPOUND POINT HEAT
IN BRACKETS)
°C KJ/KG KG/DM
3 KJ/KG K DM/KG
NA2S04 (31%)
NACL (13 %) 4 234 1) _ 16
KCL (16 %)
H20 (40 %)

CACL2 (48 %)
NACL (4.3 %) 26.8 _1)
20
KCL (0.4 %)
H20 (47.3 %)

CA(N03)2.4H20 (67 %) 30 136 228 0,324)

MG(N03)2.6H20 (33 %) 20

PROPIONAMIDE (25.1%) 50 192 1,96 (S) 7, 19

PALMITIC ACID (74.9 %) 2.40 (L)
MG(N03)2,6H20 (53 %)2) 59.1 144 232 1.34 (s) 3) 0.445) 29
MGCL2.6H20 (47 %) 3.16 (L) 3)

MG(N03)2.6H20 (53 %)2) 61 148 249 - 20

AL(N03)2.9H20 (47 %)

LIN03 (27 %)2) 1.07 (S) 3) 3.405) 29

NH4N03 (68 %) 81.6 111 205 2.20 (L) 3)
NH4CL (5 %)

1) - IMPLIESDATA NOT AVAILABLE 4) COSTDATA IS FOR 1974

2) PROPORTIONSARE IN MOL 5) COSTDATA IS FOR 1977 (JAPANESE MARKET)
3) S - SOLID; L - LIQUID
320 A. ABHAT

Table 6. Details of measurement techniques employedfor the determination of melting and freezing behaviour of
phase change heat storage materials[6]
MEASURB'~ .~P/~qATI~ DESIGN+ QUANTITYOF ~F.ASLIRB~TS MADE DATA EVALL~TED
IECI-INIQUE HEAT STORAGE
SUIBST/~CE
IN THE TESTS
(ORDER OF
MAGNITUDE)

DIFFERENTIAL PERKIN-EU"ER 1-10 MG VARIATID~I OF THE 1) FORMOF D-I£ B ~ I C 0F E M 0 ~

SC.Pd~I NG CAI.O- CALORIMETER, 11-lERMAL ENERGY
MERRY (I~C) MODELDSC-2 WlTH IME S.aMPLETO ENABLE
IT TO ~ HEATINGOR
COOLINGAT A ~ T A N T PRE-
DETERMINEDRATE,
"n-IE OUTPUTOF THE MEASURE-
MBqTS IS THE ENERGY-TIME
I)IAGRAM~ ALSOCALLE)
MIC PEAKS ON THE
2) MELTINGPOINT/RANGE
3) FRE.ZIN6 eOmT/RAI~E
4) DEGREEOF ~ I I ~
5) HEATSOF FUSIONAND SOL[DIFICATION
6) SPECIFICHEAT AS A FUNCTIONOF
I~TLIE

TEST TI.A~S OR i0 G VARIATION OF 1) SHAPE OF THE ~ ' n J R E - T I M E

ANALYSIS (TA) C.VJ~S AMPULES "I~MPERATIJRE AT SELECII~D DI~ MELTING~ I:I~.EZII~
LOCATIOI~WITHIN THE HEAT PLKr-r_ttS, HOMOGENEITYOF "WE GIRI-
FIAT Ca_ASS IOOG STORAGESLIBST/~ICERESIJL- GINAI. SI.IBST~NI~FOU.D~IK~MELTING,
CONTAINERS T[NG FR(~ ENERGYINPUT/OUT- ETC,
POT TO T~E S~c'tJE, 2) MELTINGP011~/R,~(~
THE OUTPUT OF THE MEASURE- 3) FREEZINGPOIN T / R ~
LAE~TORY MODELS 1-10 KS MENTS IS THE TE'PERATUR£- 4) DEGI~_EOF S I . ~ I N 6
OF PASSIVELY OPE- TIME DIAGR~ 5) CONGI~IBW/INCONQ~UENTMELTING
RATING LATENT 6) I~ITION OF THE ORIGINAL
HEAT STORAGE SIJIBST,qqC,I-,, ~M.TION OF NEW
SYSTEM IESle4S: Pf-I~.S, SE(g~C~,TIONEFFECTS
1) E i X L ~
FINNED
HEAT PIPE
HEAT EX-
olmccm/2/
2) FINNED-
ANNULL6
MEAT EX-

ELEMENT
/5/

DETAILS OF TEST APPARATUS DESIGN ARE PROVIDED IN APPENDIX I

structure. There are also other structures in which the
guest molecules participate in the water lattice directly,
these substances are known as semi-clathrates and in-
clude hydrates of amines and tetra alkylammonium
salts [20].
Examples of clathrate hydrates are[20]: SO2.6H20
(mpt), 7°C; AHF+ =247kJ/kg), C2H40" (6.9)H20 (rap,
ll.l°C).
Examples of semi-clathrate hydrate are [20]:
(CH3)3N.(10~/4H20) (mp, 5.9°C, AHF=239kJ/kg),
Bu4NOH.32H20 (mp, 30.2°C).
Like salt hydrates, clathrate hydrates may melt con-
gruently or incongruently. They also tend to supercool
and hence suitable nucleating techniques must be found
to promote crystallization.
4. MELTING AND FREEZING CHARACTERISTICS
By melting and freezing characteristics of phase
change heat storage materials, we understand the
behaviour exhibited by these materials during heating
and cooling, viz. melting and freezing ranges, congruency
of melting, nucleation characteristics, supercooling of the
melt and stability to thermal cycling. A comprehensive
knowledge of the melting and freezing behaviour of
heat-of-fusion storage substances, and particularly its
reproducibility as a consequence of repeated melting and
freezing of the substances, is hence essential for the
assurance of the long-term performance of a latent heat
store.
fmp = melting point.
AH~- = heat-of-fusion.
Measurement techniques
Primarily two measurement techniques are employed
for the determination of the melting and freezing
characteristics: (1) Differential Scanning Calorimetry
(DSC), and (2) Thermal Analysis (TA). Table 6 provides
details of the two measurement techniques and a listing
of data that may be evaluated.
The two techniques distinguish themselves in terms
of the type of measurements made, the quantity of the
sample used in the tests and the speed with which results
can be obtained. For example, the DSC provides quick
and reliable results in the form of Energy-Time diagrams
(Thermograms) using very small quantities of the sample
(ca. 1-10mg)). Evaluation of the thermograms yields
rather precise values of the phase transition tem-
peratures during melting and freezing of the sample, the
heats of fusion and solidification and the specific heat
variation as a function of temperature. The DSC is,
however, a severe test for substances that supercool,
since the supercooling tendencies are maximized due to
the small quantity of the samples and the poor nucleation
conditions in the DSC testing pans. Consequently, for
materials like salt hydrates, the DSC fails to provide
meaningful information on the degree of supercooling
and on the freezing point of the substance[7, 19].
The Thermal Analysis technique, on the other hand,
involves the determination of Temperature-Time (T-t)
diagrams, or the Heating and Cooling curves, recorded
during the melting and freezing of the sample. The
TA-technique uses about 10 g to a few kilograms of the
sample, depending on the apparatus used, and is hence
slower. With proper care, the rate of heating and cooling
 Low temperature latent heat thermal energy storage: heat storage materials 321

A. Heating curves B. Cooling curves

70 6o
(a) 50 N
.-~0 Tin Paraffin
61o5 3o ~' ~----____~,.,.
o ,Fr , ?, , I?o, " . . . 20 ~o Cn It~
J ,
0 20 40 ~ aO t00 120 140 0 20 40 60 80 100 12D 140
t I minl t (mini

60~ = -' , , , , , . . . . . . (b) eo ' '

_J.--g/ Louric _ 5o
Acid ~_
~ 4°rT, It
30 I I
2Oo "i'o' 11min]
't4o 20
0
I ,
20
, i
40
i i
60
i
~
J ~ i
100
J i
120
i~
140
rim=hi

60 , i- , , i , r ~

(c) ,-s o ~ ~
~ ]o Na2HPO4. V 4o
12 H20 30
10 2O
i
O t.O 80 120 160 2(10 240 280 0 2 t. 6 10 12
tlm~l ~lh]
Fig. 6. Typical heating and cooling curves for representative heat storage materials belonging to the material
families of (a) commercial paraffins, (b) fatty acids,and (c) inorganic salt hydrates [3].

can be well controlled and relatively accurate results can seen in Fig. 6(a), the paraffin exhibits two freezing
be gained. If the tests are undertaken in glass containers, ranges: a narrow freezing range characterized by QN
they suffer from the disadvantage that metallic surfaces (Region II) and a larger freezing range characterized by
that generally form the heat exchanging surface within a NP (Region III). The two freezing ranges respectively
latent heat store, or other means that are employed to signify a liquid-to-solid transition and a solid-to-solid
improve heat transfer in a latent heat store, are transition[3]. A part of the total latent heat of fusion
absent. Nucleating conditions different to those is stored in the substance during each of these trans-
in actual LTES systems hence exist. Experiments itions.
done in scaled-sown test models are the most accurate as Not all commercial waxes (all of these are essentially
conditions similar to those in large-scale LTES systems paraffin mixtures) display a freezing behaviour as shown
are simulated within them. These tests are also the in Fig. 6(a). For the sake of comparisons, the cooling
slowest in comparison to those in glass apparatuses due curves obtained during slow cooling of three different
to the large quantities of phase change materials they paraffins are presented in Fig. 7 [17]. The paraffins in-
require (1-10 kg). cluded in the diagram are seen to vary extensively in their
Thermal cycling of the sample to obtain data on the freezing interval. While the n-paraffin 6106 exhibits a
reproducibility of the melting and freezing behaviour of rather fiat freezing plateau during the time period of the
the substance may be undertaken using both measure- experiments, the iso-paraffins 6403 and 6035 undergo
ment techniques. Any variations observed in the form of freezing in a temperature range. In fact, the freezing
the endothermic or exothermic peaks on the DSC-ther- interval of paraffin 6035 is extremely large, rendering it
mograms, or alterations in the values of the ther- completely unfit for use in an LTES system.
mophysical properties of the thermally cycled heat
90 , , , , ,
storage materials, serve as a measure of the change in 1--
their melting and freezing characteristics. In the TA- 80~ -- Paraffin 61.03
measurements, on the other hand, changes in the form It \ - - - Paraffin 6035
~ 70I', ~.._ ~ - - - - P a r a f f i n 6106
of the temperature-time diagrams or in the heat transfer
rate into or out of the sample provide information on the ® 60 = \
influence of thermal cycling.
50 \ ~
We shall now discuss some results for the melting and
freezing characteristics of PCMs belonging to three ~o " ~ "" "" ~--~~ ~
materials groups: paraffins, fatty acids and salt hydrates.

Paraffins
13°
2o 4 2i ; ,i ~i
(a) TA-measurements. Figure 6(a) shows typical heat- Time [h]
ing and cooling curves obtained during TA-measure- Fig. 7. Comparison of cooling curves for three different com-
ments with a technical grade n-paraffin, paraffin 6106. As mercial paraffins[17].
 322 A. ABHAT

Table 7. Melting and freezing range of Paraffin 6106 obtained

from Thermal Analysis(TA) measurements[3]

]MANUFACTURER'S RECOMMENDED
FREEZING RANGE

MEASURED FREEZING RANGE

MEASURED MELTING RANGE

42 - 44 °C

38 - 43 *C

36 - 43 "C
10
PARAFFIN 6&O3
HEATING 2 CYCLE
RUN NO 1L2
HEATING RATE 5KtMIN

/
G2~
Another important observation with paraffin is the
large difference between the experimentally measured
b
freezing range and the manufacturer's data, as seen
L L I
through th6 example of Paraffin 6106 in Table 7. This 8 1'0 I~N]12 "~141
I iTIWE j
result is of particular importance to the design and 280 290 300 310 320 330 3~0 350
TEMPERATURE [K} ]
operation of a LTES system, which calls for an exact
knowledge of the phase transition temperature range of
the heat storage material. Further quantification of these
results can be carried out using the DSC-technique.
(b) DSC-measurements. Typical thermograms for n-
025
and iso-paraffins are presented through the example of
X
Paraffin 5838 and Paraffin 6403 in Figs. 8 and 9, PARAFFIN 6~03
respectively[7]. The n-paraffin exhibits a sharp peak COOLING 2 CYCLE
RUN NO 1~3
around 46°C and secondary peaks at temperatures be- I N G RATE 5K/MIN
tween 12 and 46°C, whereas the iso-paraffin exhibits one ua

peak which is, however, smeared over a large tem-

-175
perature range of about 43 K. Evaluation of the peaks
shows that both paraffins undergo solid-liquid and solid-
solid phase transitions. The D$C technique additionally I I
provides a quantitative measurement of the energy con-
J ~ I I ITINE [MIN!
tent associated with each of the phase transitions. 3/,0 330 320 310 300 290 280
TEMPERATURE[KI
Some results from the DSC-measurements of the
phase transition temperatures and enthalpies of com- Fig. 9. Typical DSC-thermogramsfor iso-paratfinsshown through
mercial paraffins are presented in Table 8. For the sake of the example of Paraffin6403[7].
comparisons, results from measurements with 99 per
cent pure octadecane, a research grade paraffin, are also
I I~-- ] I
included in the Table (from [51).
For all commercial paraffins tested, the deviation be-
PARAFFIN 5838
12 HEATING 1CYCLE tween the manufacturer's data and the measured values
RUN NO 73 is noteworthy. All paraffins exhibited two phase tran-
09 HEATING RATE 25 K/MIN
sitions. Whereas the n-paraffins experienced the primary
phase transition (solid-liquid) in a narrow temperature
06
range of about 2 K, the corresponding temperature range
03 for iso-paraffins was large (-> 14 K). The research grade
n-octadecane, on the other hand, exhibited only one
0 I
2 &
I t
6
I
8 TIME MIN I solid-liquid phase transition in a narrow temperature
J I
tz: 300 305 310 315 320 TEMPERATURE [K] range of 1.5 K. Evaluation of the thermograms further-
more showed that while the total phase transition
enthalpy of the commercial paraffins agreed closely
with the manufacturer's data, the enthalpy associated with
the solid-solid phase transition was rather significant (ca.
30 to 50 per cent of the total).
RAFFIN 5838
OLING 3 CYCLE
The above results are of particular importance for the
-08 choice of paraffins for low temperature applications,
LINNOGRA;8E 25 K/MIN
wherein the temperature excursions of the store are
(.9 -12 generally limited to 10-15 K about the melting point.
Only n-paraffins may hence be selected. Although the oil
a content of the paraffins does not play any significant role,
6 9 TIME
J_ [ MINI
15 the amount of heat stored in a a narrow temperature
1_
33o~ 320 310 TEMPERATURE { K ]
range is dependent on the phase transition enthalpy of
Fig. 8. Typical DSC-thermogramsfor n-paraffinsshown through the primary solid-liquid transition, which should be
the example of Paraffin 5838[7]. measured for the paraffin in question.
 Low temperature latent heat thermal energy storage: heat storage materials 323
Table 8. Phase transition temperatures and associated phase transition enthalpies for paraffins, obtained from
Differential Scanning Calorimetry (DSC) measurements[5]
PARAFFIN PW~UFACTURER'S
DATA MEASUREDVALUES
TYPE C~IN OIL- FREEZING HEATOF ~ "Ig,AN- SOLII)-LIQUID SOLID-SOLID ~E TRANSITION
STRUC- CON'- RANGE FUSION SITIONTEMRE- PHASE
TURE TENT RANGE ~ITION TRANSITION

Fr er1 all.l_ 1) pT2 aH2 i) TOTAl. PROPORT

IONS

/%/ /'C/ /KJ/KC-,/ I'C/ I'C/ IKJIKG/ /'C/ /KJ/V.C,/ IKJI~cV AH AH %

6106 N 5 LQ - 44 189 19 - 44 40.7 - 44 129.8 19- ~.7 49.2 179 72.5 27.5
N 0.5 48 - 50 189 122 - 48.3 06.2 - o,8.3 134.4 12.7 - 46.2 63.0 i~.4 68.0 32.0
6035 tSO 4 58 - 60 189 -8 - 64.4 39 - 64,5 ~5.7 -8 - 39 83.2 168.9 50.7 49.3
6A133 tSO 0,5 62 - 64 189 23 - 66.6 51,7 - 66,6 129.8 23 - 51,7 59.2 189 682 31.3
6499 iso 3 66 - 68 189 -6 - 71.6 2) 2) 2) 2) lt6 2) 2)

OCTADECm~3) 28 - 29 246 27 - 28.5 27 - 28,5 230 230 100

1) AHI ~ AH2 REPRESENTTHE ~ I E S ASSOCIATE/)WITH THE SOLID-LIQUID ~ SOtlD-SOLIDPHASETRANSITIONS RESPECTIVELY

2) PARAFFIN~ EXHIBITED A PEAKsPREADOVERA ~ TEMPERATURE~ OF ABOUT78 K, SO THATTHE ~ PHASETRANSITIONS
COULDNOT BE DISTINGUISHEDFROMEACHOTHER
3) N-OC-TAI~EC~/~E(99 ~ PURE) WAS EMPLOYEDAS A REFERENCEPARAFFINDURINGTHE INVESTIGATIONS

evident, though in some cases, a small degree of super-

z,
LAURIC ACID
cooling (-0.5 K) has been measured[4]. This behaviour
HEATING 2 CYCLE is representative of all fatty acids as well as of polyethy-
3 RUN NO 15
HEATING RATE 2
lene glycols with melting points between 15 and
70°C[19].
2
(b) DSC-measurements. Fatty acids possess good mel-
1
ting and freezing characteristics, as may be seen in Fig.
10 from the sharp peaks traced on the DSC-thermograms
0 " I
during the heating and cooling of lauric acid.
3 TIME( MIN]
I I I Results from the DSC-measurements with two fatty
312 31& 316 31B TEMPERATUREI K ]
acids [6] are summarized in Table 9. The difference be-
tween the melting and freezing points given in the Table is
a measure of the supercooling of the substances. As
~-~ LAURIC
mentioned earlier, yet smaller amount of supercooling (ca.
: I C L,. ] CYCLE I / 0.5 K) were observed in TA-tests undertaken using
-'1- Ro..o 1, I / somewhat larger quantities of the substance. The small

il I/
I/
amount of supercooling during the freezing of fatty acids
does not hamper their potential use as heat storage
substances.
Although fatty acids are good heat storage sub-
stances, they are somewhat too expensive for large-scale
IO 15 20 2S TIME [MINI
heat storage applications. They are, however, far cheaper
) 317 31~ 31--15 31"J& I. T E M P E R ~ K"'~
|
than research grade paraffins and are recommended for
Fig. 10. DSC-thermograms of lauric acid, representative of fatty the function tests of new "passively-operating" heat-of-
acids [6]. fusion storage system designs. Recent discussions with
manufacturers indicate that the cost of some fatty acids
Fatty acids can be brought down by about 50 per cent as a result of
(a) TA-measurements. The melting and freezing large scale production. For LTES systems operating
characteristics of fatty acids are presented through the within a narrow temperature range, the cost of heat
example of lauric acid in Fig. 6(b), which contains storage material per kilojoule of stored energy may then
the heating and cooling curves for this material [3]. The compare fairly well for fatty acids and commercial
freezing plateaus are long and flat and no supercooling is paraffins.

Table 9. DSC-measurements with fatty acids [6]

HEAT STORAGE NUMBER OF MEASURED VALUES
SUBSTANCE THERMAL CYCLES
PERFORMED MELTING FREEZING HEAT OF HEAT OF
POINT POINT FUSION SOLIDIFICATION
1%1 1%1 /KJ/KG/ /KJ/KGI
LAURIC ACID 3 43,5 -'I"0,05 39,9 * 0.05 169.3 -'!" 2 . 0 168,8 + 2.5

PALIMITIC ACID 5 61,2 +- 0.07 59.9 _+ 0.13 196.1 + 2.0 197.0 -'! 3 . 0
324 A. ABHAT

apparatus has been found to have a significant influence

\ Table 10 compares results from TA-measurements with
50"
Crystals N°2C0310H20
No2SQ 10H20
2O
Time [hi
Fig. l l. Freezing (cooling) curves for various inorganic salt
hydrates[26].
Salt hydrates
(a) TA-measurements. Typical heating and cooling
curves for salt hydrates are shown through the example
of Na:HPO4 • 12H20 in Fig. 6(c). Melting of the material
takes place at a constant temperature, whereas freezing
is associated with appreciable supercooling[3]. Freezing
curves for several other salts obtained from tests in glass
capsules[26] are presented in Fig. 11. For all these
materials, the melt does not freeze at its thermodynamic
freezing point, but is supercooled by several degrees below
the freezing point. The supercooled liquid hence exists in a
highly metastable state. Formation of or introduction of
a single crystal nucleus into the melt causes a spon-
taneous crystallization of the whole melt.
In the case of salt hydrates, the geometry of the test
on the melting and freezing behaviour of the material[6].
the salt hydrates carried out in three different experi-
mental apparatuses. Poor nucleation conditions in the
glass test tube apparatus (Apparatus A)--which is also
the most commonly suggested apparatus in the lit-
erature-cause substantial supercooling of the salt
hydrates. In fact, the degree of supercooling observed
during the cooling of Na2S203"5H20 was so large
( > 40 K) that no freezing of the substance occurred within
the temperature range used in the tests. As a con-
sequence, the tests were terminated after merely 2
thermal cycles.
The amount of supercooling could be reduced by using
glass containers (Apparatus B) containing a shallow bed
of the salt hydrate. The larger glass surface aided in
somewhat improving the nucleation conditions in the
melt and the shallow bed assisted in eliminating segre-
gation effects in the molten material.
Both Apparatuses A and B are, however, far remote
from practical latent heat store designs. Experiments to
investigate the melting and freezing characteristics of
salt hydrates were hence undertaken in laboratory models
(Apparatus CI, C2) of two latent heat stores that have
been recommended for large scale applications [2, 5]. The
heat storage concepts selected employ large metallic heat
exchanger surfaces (fins) for heat transfer into the poorly
conducting heat-of-fusion storage substance, and are
suitable for "passively-operating" latent heat stores.
Results contained in Table 10 now indicate a very
favourable freezing behaviour of the salt hydrates,
characterized by their small degree of supercooling. This
improvement, in comparison to the values measured
using Apparatus A and B, results from the presence of
the aluminum heat exchanger surfaces contacting the salt
hydrate, which strongly promote heterogenous nuclea-
tion in the molten salt[4].
(b) DSC-measurements. The calorimetric measure-
ments with salt hydrates[7] are illustrated here through
the example of the incongruently melting calcium chloride

Table 10. TA-measurementswith inorganicsalt hydrates[6]

SALT HYDRATES LITERATURE TEST + NO, OF M E A S U R E DVALUES
VALUE OF APPARATUS THERMAL (AVERAGES)
FREEZING USED CYCLES
POINT PERFORMED
DEGREE OF SUPERCOOLING
I°CI /K/
CACL2,6H20 29,7 A 10 25
B 5O 5
C1 100 i
ZN(N03)2.6H20 36,4 A 10 8
C2 20 4
NA2S203.5H20 48.0 A 2 40
B 40 ii
c2 20 5

+ DETAILS OF TEST APPARATUS DESIGN ARE PROVIDED IN APPENDIX

A - GLASS TEST TUBES

B - GLASS CONTAINERS
C1 - LABORATORY MODEL OF THE FINNED HEAT PIPE
HEAT EXCHANGER DESIGN / 2 /
C2 - LABORATORY MODEL OF THE FINNED ANNULUS
HEAT EXCHANGER ELEMENT / 5 /
 Low temperature latent heat thermal energy storage: heat storage materials
6-hydrate. Experimentation with this substance was
undertaken in two operation modes: Mode 1--tests in
hermetically sealed pans, and Mode 2--tests in non-
hermetically sealed or "open" pans, wherein a hole was
punched in the hermetically sealed pans.
Figures 12 and 13, respectively present typical ther-
mograms obtained during the heating and cooling of
CaCI2.6H20 in hermetically sealed and open pans[6,
21]. Tests in the hermetically sealed pans (Mode 1)
delivered thermograms with a single endothermic or
exothermic peak which was reproducible in form even
after thermal cycling up to 11 cycles. Operation in open
pans (Mode 2), on the other hand, yielded thermograms
with a single endothermic peak during the first melting
325
process but with two or more peaks during subsequent
cooling and heating processes. The salt hydrate
CaCI2.6H20 thus undergoes decomposition during tests
in the open pans. By the fifth thermal cycle, a clear
secondary peak is observed at 45°C, signifying the partial
decomposition of CaCI2.6H20 into its lower form
CaCI2.4H_,O.
Results pertaining to the melting point and the heat of
fusion of four salt hydrates tested [6] are summarized in
Table 11. The values of the thermophysical properties of
the substances as obtained from measurements with
various samples tested in hermetically sealed pans (Mode
1) are seen to fall in a small range and the average values
agree fairly well with literature values. On the other

0 ~
2,5

uq
2,0
RUN NO : 183 RUN NO. : 184.
I COOLING : /,,.CYCLE COOLING : t,.CYCLE
RATE : 2.5 KIMIN RATE " 2.5 K/NIN
.!

1,5
tr
o 0

=: 1.0
/_
ud

<[ 0.5
B

0 I I 2
290 3O0 3~0 250 2/*0 230
TEMPERATURE [K] TEMPERATURE [K]

I i !

20

o3
u RUN NO.: 197 RUN NO. : 198
H E A T I N G : 11 CYCLE COOLING : 11. CYCLE
RATE : 2.5 K / M I N RATE 2.5 K / M I N
z 1.5
o
I..-
&.
r,,"
0

1.0

w
z
w

0.5

0 t I I 1
290 300 310 260 250 2t,0
TEMPERATURE [K] TEMPERATURE [ K ]

Fig. 12. DSC-thermogramsfor calciumchloride 6-hydrate, tested in hermeticallysealed pans[6].

326 A. ABHAT

r 1 1 T r q i

I•
|
I.,I(~
AT I ~ I.CYCLE
RUN NO. 106
HEATINO RATE 1.2~" KIMIN COOLING I.CYCLE
RUN NO 10g
CI8 I COOLING RATE 1.25 KI HIN

0.t.

i i
I 2 3 t, 5 8 i • - TIME IMIN]
2~78 TEMPERATt.h~.IK] 258

f.__----
"0.6 ~1.21 HEATING 5 CYCLE
RUN NO. 116
COOLING 3 CYCLE HEATING RATE 5.0K/MIN
RUN NO 113
COOLING RATE 2.5K! MIN m~08

-2.2

-3£I } L ~5 TIME[ MIN]

0 I t J
2 e, T kME [ I~IN]
290 30(] 310 TEMPERATURE
[K]

Fig. 13. DSC-thermogramsfor calcium chloride 6-hydrate, tested in non-hermetically sealed (or "open") pans [6].

Table 11. DSC-measurementswith inorganic salt hydrates [6]

,SALT LITERATUREVALUE OPERATION ME/k,~IRE]) VALUES
HYI~TE OF
MF.LTINGI.r~kT-OF- NO. OF TOTAl. NO. ~ OF
POINT FUSION ,SAI'I~ES THER- VALLF.S
M~. CYCLES - I
MELTING HEAT-OF- I
MELTING NE~T--~-
POINT FI.~1ON POINT FIEION
(PRIMARY
I TR~SITION)
/'C/ /KJ/KG/ I
I
i.~ /~,,~ /'C/ i /KJ/Kr.,/
c~.~0 29.7
~,5-~.0
ZNt%b.r~2o ~.4 Z~7 5 I 29 36.0-57,0 136 - 175 ]6,5
N,~S,~.~H,~ ~.o 201 1 1 i 4 a,8.5-49.0 190 - 20O
2 1 I 2 +) +) I +)
N,(%>2.F~H20~.7 188 1 3 13 53.5-57.0 145 - 160 "t 55,2 J 152
i 1 , 1 I
NOTES; + OPERATIO~I M01DE: ,].-TESTS IN HB~4ETICAU_y ~ P~WWS
2-1"IE~ IN OPEN PANS
") ,~CAI'IF.RTOO ~ TO QBI"AIN MEANINGFUL AVERAGES
+) NO EVALUATION ~ I ~
, AVB~6E OF 8 ~ CYCLES

hand, results from measurements in open pans show a
large scatter as a consequence of the decomposition of
the salt hydrates.
In practical systems, the use of salt hydrates is hence
recommended only in hermetically sealed or encap-
sulated heat-of-fusion storage system designs. It is fur-
thermore recommended to use the salt hydrates in an
atmosphere free of air[6, 7].
Modified salt hydrates. The discussion in the preceding
section has been restricted to the investigation of the
behavioural characteristics of "pure" salt hydrates i.e. in
which no foreign substances are intentionally added to
prevent segregation and/or to eliminate supercooling of
the substance prior to crystallization. There have been
several attempts in the literature to "modify" salt
hydrates to obtain a reproducible melting and freezing
behaviour. Although only limited success has been
reported with the methods employed, we shall discuss
them here briefly for the sake of completeness.
(a) Suspension media. Addition of suspension media or
thickening agents to the salt hydrate to prevent separa-
tion of the solid and liquid phases has been recom-
mended [l l, 27]. The use of a thickener also assists
in suspending the nucleating agents within the heat stor-
age medium bulk, which otherwise tend to collect at the
container bottom due to density differences. Thickening
agents, however, displace a part of the salt hydrate in the
heat store, thus reducing the volumetric heat storage
capacity of the heat store. Furthermore, they work
towards a lowering of the melting point of the heat
 Low temperature latent heat thermal energy storage: heat storage materials 327
Table 12. Some suspension media for use with inorganic salt hydrates
HEAT OF FUSION MATERIALS ~USPENSlON MEDIA REFERENCE
CACL2,GH20 HYDROXY ETHYL CELLULOSE 11
CA(NO3)2,4H20 POLYACRYLIC ACID 11
NACO3.1OH20 POLYETHYLENE OXIDE 11
NA2HPOh.12H20 STARCH 11
NA2S203.SH20 wOOD PULP 11
NA2SO4.10H20 CLAY (BENTONITE, ATTAPULGITE) 11, 26

Table 13. Some nucleatingagents for use with inorganic salt hydrates
HEAT STORAGE NUCLEATINGAGENT DEGREE OF SUPERCOOLING REFERENCE
SUBSTANCE (PCM) (SEEDCRYSTAL)
WITHOUT SEED WITH SEED
CRYSTAL CRYSTAL
K K
LICLO3.3H20 KCLOq, NA2SIF6, K2SIF6, BASIF6 2 - 8 1) 0.5 23
KF.4H20 PUMICE STONE 15 - 24 9.5 23
CACL2.6H20 BACO3, SRCO3, BAF2, SRF2 10 - lq 3 - 5 1) Ii, 23
NA2SO4.10H20 BORAX 14 2 - 5 8, 23, 26
ZN(NO3)2.BH20 ZNO, ZN(OH)2 2 - 7 i - 6 4, 19
KF.2H20 AL203 25 - 35 5 - 8 23

i ) MINIMUM AND MAXIMUM VALUES OBSERVED

storage substance. Table 12 lists some suspension media
recommended for different salt hydrates.
(b) Nucleating agents. A nucleating agent is a material
having a crystal structure similar in lattice spacing to that
of the heat storage substance. They serve as nuclei for
the PCM crystals to grow on them during freezing of the
PCM and are also termed as "seed-crystals". For good
results, measures should be taken whereby the nucleat-
ing agents are homogeneously dispersed within the bulk
PCM and/or they are in contact with the heat transfer
surface. Table 13 provides a list of some of the com-
monly recommended nucleating agents for use with salt
hydrates and the reduction in the degree of supercooling
attained with their use.
(c) Extra water principle. Extra water may be added to
a salt hydrate to allow dissolution of the anhydrous salt in
the water at the melting point of the salt hydrate, so that
the heat storage medium becomes a saturated salt solu-
tion at the melting point. During cooling, the solubility
of the salt in water for temperatures below the melting
point decreases with decreasing temperatures resulting in
crystallization of the salt hydrate. Thus at temperatures
below the melting point, the storage medium consists of
a salt hydrate solid phase and solution. Upon repeated
heating and cooling, the solid phases of the salt hydrate
then completely go into solution or crystallization out of
solution takes place. Soft stirring of the medium may,
however, be necessary to overcome the density differences
between the salt solution and the solid phases/23/. The
stirring effect has been achieved in some practical systems
by employing the direct contact heat exchange principle,
wherein cold oil sprayed through fine pores directly in the
molten phase of the heat storage substances simultaneously
cools down the liquid and creates a mixing effect/12/.
Using a mixture of 68.2 wt per cent Na2SO4 • 10H20 and
31.8wt per cent H20, Biswas[9] reports good results in
comparison to pure Glauber salt. Nucleation of the deca-
hydrate occurred readily, even without the addition of
borax. Furbo[14] reports some preliminary results with
several different salt hydrates and finds them attractive in
use in comparison to sensible heat storage in water.
The major disadvantage of the extra water principle is,
however, the loss in volumetric heat storage capacity in
relation to that of the pure salt hydrate. For example, with
Biswas' composition of Na2SO4.10H20 and n20, a
mass about 50 per cent larger and a volume 70 per cent
larger than for an ideally efficient system based on pure
Na2SO4.10H20 would be required to store the same
amount of heat [9].
(d) Other techniques. Another technique to avoid phase
separation in incongruent melting salt hydrates is that
suggested by Carlsson et al.[lO], who propose chemical
modification of an incongruent heat-of-fusion system to
make it behave as a congruent system. Through the
addition of 2 wt. per cent SrCI2 • 6H20, the melting point
of the resultant salt hydrate was altered so that the
melting point maximum for CaCI2.6H20 coincided with
the point where equilibrium between CaCI2.6H20
CaCIE.4H20 and the solution exists. Upon melting of the
modified salt hydrate, the peritectic point was thus by-
passed and the formation of the lower salt hydrate
(CaCI2.4H20 in this case) avoided.
5. THERMALCYCLING
One of the most severe tests that phase change heat
storage materials must undergo is thermal cycling in-
volving repeated melting and freezing of the heat storage
materials. For example, for a 20-yr life of a one-day heat
328 A. ABHAT

50 , ,
To ~ 2 0 0 Sample A. Not Thickened
--~-rJo I Sample B, Thickened
S
f- -- ~ Ambient Air Temperature

o I,
T ~ ~ I:'

o Cycle No
,s ,._..-

~ ~200

201 200
1000

I 1 i I
100 Hours 1 2 3 Time [ h] /.
~ Heating ._L.
- ~- Cooling
Fig. 14. Temperatureprofilesmeasured during thermal cyclingtests with two samples of Glauber salt plus 3 wt. per
cent Borax. Sample B uses in addition 8 per cent thickener as suspensionmedia for the salt hydrate and nucleating
agent[27].

store, the phase change material experiences one melting-
freezing cycle dailp or a total of 7365 cycles during the
system life.
The influence of thermal cycling on the phase change
material characteristics must be measured experiment-
ally. Laboratory measurements comprising at least 1000-
2000 cycles are recommended, particularly with the in-
organic salt hydrates, to establish the long-term thermal
stability of the PCM. The measurements should include
the temporal variation of temperature within the PCM and
the temporal variation of the heat transfer rate during the
charging and discharging of the storage substance. The
thermal energy stored in the PCM during each cycle
should be computed and the variation of the stored energy
with the number of thermal cycles should be studied.
A large gap exists today in the area of thermal cycling
tests. Limited thermal cycling (120 cycles) carried out
with paraffins and lauric acid exhibited no degradation of
the materials[4]. Short-term thermal cycling (up to 90
cycles of Na2HPO4 • 12H20 showed that the material melts
congruently when it is not contaminated with the hepta-
hydrate, i.e. when the formation of the heptahydrate has
been prevented by appropriate nucleation[27].
Test results describing the temperature variation in
Na2SO4' 10H20 over 1000 heating-cooling cycles are
profile following cycling. However, the material without
the thickener contained approximately 30-35 per cent
liquid at the end of the cooling cycle. A sediment layer
formed which remained throughout the testing period. As
a consequence, the non-thickened material attained a
higher temperature during its heating cycle, as seen in
Fig. 14.
The heat stored in Glauber salt as a function of ther-
mal cycling was measured using standard calorimetric
techniques[22]. Two different samples consisting of
Na2SO4" 10H20 with 3 per cent Borax and thickened
material comprising 88wt. per cent Glauber salt+
2.64wt per cent Borax+9.36 wt per cent attapulgite
clay were included in the tests. The samples weighed
150 g each and were cycled within a temperature range
27-38°C. Figure 15 shows the measured thermal energy
per gram of mixture (AH) as a function of thermal
Glaubers SaLI+ 3 % Borax
6O •¢/•TheoreticaJ-Pure 251 0
J
5O
AH
A. 4O t kJ/kg
30 1255

available[27]. Two material samples of Na2SO4.10H20 ½ t

20
l[
were used in these tests--one sample comprising the ;!; • = Pure Glaubet's Sa# + 3 % Borax
10
phase change material and 3 per cent Borax by weight as • • Glouioers Sort + Borax ~"Aftoputqtte Clay

nucleating agent, and the second sample comprising the o 2b 6b ,6o ,,o ,8o 21o 2~o 36o 3~o ~ ,.&o
same constituents as above plus 8 per cent thickener that No. of Cycles

formed a thixotropic gel. The results of these thermal

Fig. 15. Thermal energy stored in unthickened and thickened
cycling tests are presented in Fig. 14. The material with mixtures of Glauber salt and Borax as a function of the number
thickener showed no degradation in the temperature of thermal cycles[22].
 Low temperaturelatent heat thermal energy storage: heat storage materials
cycling for the two materials. The results in Fig. 15 reveal
that the thermal capacity of the pure salt declines quickly
from an initial value of 238 kJ/kg to 63 kJ/kg after 40
cycles. The thermal capacity of the thickened salt also
declines--albeit slowly--to 125kJ[kg in the 140th
cycle[22]. Both the pure Glauber salt and the thickened
mixture are thus unfit for long-term use as latent heat
storage materials, despite the fact that the results in Fig.
14 for the temperature profiles showed the thickened
material to possess a reproducible behaviour upon ther-
mal cycling.
6. COMPATIBILITY WITH MATERIALS OF
CONSTRUCTION
Knowledge regarding the compatibility of phase
change heat storage materials with conventional materi-
als of construction is of particular importance for the
assurance of the life of an LTES system. Only limited
compatibility data is, however, presently available and
further work in this direction is recommended.
The choice of the operating conditions and evaluation
techniques plays an important role in the corrosion in-
vestigations. In most tests, the samples of the con-
struction materials are immersed in the liquid phase
(melt) of the heat storage material contained in air tight
bottles, held at a constant temperature, uusally 5-20 K
above the melting point of the corresponding PCM.
Several samples are used, one to a bottle, and they are
removed from the bottles after predetermined time in-
tervals, cleaned and analysed. More realistic operating
conditions, however, require that the PCM be thermally
cycled during the corrosion tests.
Gravimetric analysis, optical and scanning electron
microscopy techniques, and chemical anlysis of the cor-
rosion products are the conventional evaluation methods
employed. Gravimetric tests prior to and following the
329
corrosion tests after predetermined contact times pro-
vide the sample mass loss, from which the reduction in
sample thickness with time (for uniform surface cor-
rosion) and the corrosion rate may be computed. In
general, according to accepted norms, a metal
experiencing a thickness reduction of <-0.1 mm/a is ter-
med corrosion resistant. For cases where the corrosion
rate attains a linear value, an estimate of the lifetime (e.g.
20-yr life) of a metal in contact with a PCM is possible.
Microscopy techniques, on the other hand, provide in-
formation on the type of corrosion, for example, pitting
corrosion, cracks, etc.
The results from various corrosion studies undertaken
with phase change materials for latent heat storage ap-
plications are presented in Table 14[17, 18, 23, 29]. The
organic materials are seen to be compatible with the
metals tested. With the salts, however, one needs to be
careful as preferential compatibility may exist.
Stainless steel is the only metal that was found com-
patible with all phase change materials tested. Mild steel
is a fairly good material being corrosion resistant to most
PCMs in Table 14, with the exception of Zn(NO3)2.6H_,O
and the Mg(NO3)2-MgCI2 eutectic. The corrosion
investigation with mild steel joined by the tin solder and
several inorganic salt hydrates and their eutectics[23]
showed that the mass loss is rather low and decreases
with time to negligible values.
Copper was found compatible with all heat storage
materials tested, except sodium thiosulphate 5-hydrate.
When immersed in the melt of Na2S~O3.SH20, copper
exhibited rapid corrosion and a black layer of CuS was
seen to form within just 10 days of contact. The mass
loss of the sample after 300 days of contact was found to
be 8.17g and the thickness reduction 610#m. A photo-
graph of the corroded sample taken after 50 days con-
tact is presented in Fig. 16[17].

Table 14. Corrosion investigations with selected phase change heat storage materials
HEAT STORAGE PHASE TEMPERATURE R£F.
MATERIAL OF CONSTRUCTION
MATERIAL TRANSITION LEVEL OF MELT
TEMPERATURE DORING TESTS STAINLESS MILD TIN PLATED COPPER AL.qg,5 ALMG3
OF P(24 (APPROX.) STEEL STEEL MILDSTEEL
(°C) (°C) 1.4301 1.0330

i, LAURIC ACID 44 65 + + 0 + + + 17, 18

2. WAX ESTER, 80 + 0 o + 0 + 17
58
LDXIQL G 32

3. LICLD3.3H20 8.1 20 0 0 + O 0 0 23

4. CACL2,~%0 29.7 50 + + O + - 18

5. NA2S04.10H20 32.4 50 + 0 0 + O + 8

6. NA2HM34.12H20 35.0 55 + + O + 1.7, 18

7. Z~(N%)2,e~20 36.4 55 + O + 18

8. N~203.5~0 48.0 70 + + 0 + + 17, 18

c)
9, ~ . 3 H 2 0 58.0 60 0 0 + 0 0 0 25

10, MG(N03)2,Q'i20 116,0 140 0 0 + 0 0 0 23

60 0 0 + O O 0 23

59.1 70 + A) _ B) O 0 O 20
- ~t~CL2.~ 2 0 (4?MOL%)

NOTATIQ[~L: + Q~RR0610~ RESISTANT A) TESTS WITH STAINLESS STEEL AISI-403

- UNSUITABLE B) TESTS WITH ] ~ ~ STEEL
O r"ETAL-PCMPAIR NOT INVESTIC~TED

SE Vol. 30, No.

330
Fig. 16. Surface of Cu99.9 sample after 50 days contact with the
melt of sodium thiosulphate 5-hydrate[17].
Table 14 reveals that aluminum (A199.5) and aluminum
alloy AIMg3 are incompatible with the salt hydrates
investigated, with the exception of Na2S203.5H20. With
Na2HPO4.12H20, samples of both metals experienced
surface thickness reduction as well as surface attack, with
a white layer of aluminum hydroxide AI(OH)3covering the
metal surface after a short period of contact[17]. Figure 17
shows results for the variation of the surface thickness
reduction and the mass loss (corrosion) rate with time for
A199.5 and A1Mg~ samples immersed in molten
Na2HPO4.12H20. Initially the rate of corrosion is fairly
high, but drops with time and attains an asymptotic value
after about 100 days of contact. Scanning electron micro-
scopy investigations indicated pitting corrosion with trans
and intercrystalline cracks on the A199.5 surface after 80
days of contact, as seen in the photograph in Fig. 18. The
aluminum alloy AIMg3,on the other hand, exhibited fewer
cracks and tended more towards shallow pit formation[17].
From these and other results, it may be concluded that
mild steel and stainless steel are essentially compatible
with most heat storage materials, whereas aluminum and
copper or their alloys are only preferentially compatible.
Furthermore, common plastics are stated to be generally
corrosion resistant to most inorganic salt hydrates and
their eutectic compounds[23].
AS v and long-term performance of a latent heat store. Using
AI 9 9 . 5 ---O---- --'-t~-.-
c AI M g 3 ~ -.4~-.--
o
30
-o
~ 20
l "~--- •~ . ~.
• ~ 'E-- "-'-t
i i
i oi
2o ~ ' ,~o
Time [ d ]
Fig. 17. Temporal variation of the surface thickness reduction
and corrosionrate of A199.5and AIMg3samples in contact with
sodium hydrogenphosphate 12-hydrate[17].
A. ABHAT
Fig. 18. Scanning electron microscope photograph of AIMg3
sample surface showing pitting corrosion after 80 days contact
with the melt of sodium hydrogenphosphate 12-hydrate[17].
7. S U M M A R Y
Latent heat storage in the temperature range 0-120°C
is of interest for a variety of low temperature ap-
plications, such as space heating, domestic hot water
production, heat-pump assisted space heating, green
house heating, solar cooling, etc. The development of
dependable heat storage systems for these and other
applications requires an understanding of two essentially
diverse subjects: heat-of-fusion storage materials and
heat exchangers. The present paper reviews the state-of-
the-art of heat storage materials.
Inexpensive commercial paraffins, fatty acids, in-
organic salt hydrates, and eutectica of organic and in-
organic compounds are the important materials families,
to which candidate heat-of-fusion storage materials with
phase transition temperatures within the temperature
range 0-120°C belong. While the inorganic salt hydrates
are generally preferred due to their higher volumetric
heat-of-fusion in comparison to the organics, they usu-
ally suffer from the disadvantages of supercooling and
decomposition upon melting.
The knowledge of the melting and freezing charac-
teristics of the heat storage materials, their ability to
undergo thermal cycling and their compatiblity with
4 construction materials is essential for assessing the short
-o
two different measurement techniques--differential
n E
scanning calorimetry and thermal analysis--the melting
w
and freezing behaviour of the heat storage materials is
determined. Results from measurements undertaken with
representative organic and inorganic substances are dis-
2
...i cussed. Commercial paraffins are characterized by two
phase transitions--a solid-liquid and a solid-solid phase
transition--which may be spread over a large tem-
perature range depending on the paraffin concerned, n-
Paraffins are preferred in comparison to their iso-coun-
terparts, as the desired solid4o-liquid phase transition is
L i
,6o
' 0
200 generally restricted to a narrow temperature range. Fatty
acids are organic materials with excellent melting and
freezing characteristics and may have a good future
scope, if their cost can be brought down. Inorganic salt
hydrates on the other hand, must be carefully screened
 Low temperature latent heat thermal energy storage: heat storage materials
for congruent, "semi-congruent", or incongruent melting
substances with the aid of phase diagrams. Incongruent
melting salt hydrate may be "modified" to overcome
decomposition by adding suspension media, or extra
water, or other substances that shift the peritectic point.
The use of salt hydrates in hermetically sealed containers
is recommended. Moreover, the employment of metallic
surfaces to promote heterogeneous nucleation in a salt
hydrate is seen to reduce the supercooling of most salt
hydrates to a considerable extent.
Finally, some results from thermal cycling tests and
corrosion investigations are presented to illustrate their
importance in the appropriate choice of materials from
the standpoint of the life of a latent heat store.
Acknowledgements--The article forms part of a lecture given by
the author at the Ispra course "Thermal Energy Storage"
organized by the Commission of European Communities (CEC)
in June 1981. The full lecture is included in the course proceed-
ings published by D. Reidel Publishing Co., Holland on behalf of
CEC.
REFERENCES
1. A. Abhat, S. AbouI-Enein and G. Neure, Latent heat storage
for application to solar energy systems in dwellings. (In
German). Verein Deutscher Ingenieure, VDI-Berichte Nr.
288, pp. 97-104 (1977).
2. A. Abhat, Performance studies of a finned heat pipe latent
heat thermal energy storage system. In SUN, Mankind's
Future Source of Energy (Edited by F. de Winter and M.
Cox), Vol. I. pp. 541-546. Pergamon Press, New York
(1978).
3. A. Abhat, Short term thermal energy storage. Revue Physique
Applique 15, 477-501 (1980).
4. A. Abhat et al., Development of a modular heat exchanger
with an integrated latent heat store (in German). Report
No. BMFT FB-T 81-050, Ger/nan Ministry for Science and
Technology, Bonn, FRG (1981).
5. A. Abhat, S. AbouI-Enein and N. A. Malatidis, Heat-of-
fusion storage systems for solar heating applications. In
Thermal Storage of Solar Energy (Edited by C. den Ouden),
pp. 151-171. Martinus Nijhoff Publishers, The Hague, Hol-
land (1981).
6. A. Abhat and N. A. Malatidis, Determination of properties of
heat-of-fusion storage materials for low temperature ap-
plications. In New Energy Conservation Technologies and
Their Commercialization (Edited by J. P. Millhone and E. H.
Willis), Vol. 1, pp. 847-856. Springer-Verlag, Berlin, FRG
(1981).
7. A. Abhat, S. Aboul-Enein and N. A. Malatidis, Latent heat
thermal energy storage--Determination of properties of
storage media and development of a new heat transfer sys-
tem (in German). Report No. BMFT-FB-T 82-016, German
Ministry for Science and Technology, Bonn, FRG (1982).
8. S. Aboul-Enein, Investigation of a latent heat store for solar
space heating systems (in German). Diplomarbeit, Report
No. 5 D-0, IKE, University of Stuttgart, Stuttgart FkG
(1977).
9. D. R. Biswas, Thermal energy storage using sodium sulphate
decahydrate and water. Solar Energy 19, 99-100 (1977).
331
10. B. Carlsson, H. Stymne and G. Wettermark, An incongruent
heat-of-fusion system--CaCI2.6H20--made congruent
through the modification of the chemical composition of the
system. Solar Energy 23, 343-350 (1979).
11. D. Chahroudi, Suspension media for heat storage materials.
Proc. Workshop on Solar Energy Subsystems for the Heat-
ing and Cooling of Buildings. Charlottensville, Virginia,
U.S.A., 56--59 (1975).
12. D. D. Edie et al., Latent heat storage using direct contact
heat transfer. In Sun H (Edited by K. W. Boer and B. H.
Glenn), pp. 640-644. Pergamon Press, New York (1979).
13. Encyclopedia of Polymer Science and Technoh~gy, Vol. 14,
pp. 768-769. Wiley, New York (1971).
14. S. Furbo, Heat storage with an incongruently melting salt
hydrate as storage medium. In Thermal Storage of Solar
Energy (Edited by C. den Ouden), pp. 135-146. Nijhoff, The
Hague (1981).
15. Gmelin's Handbuch der anorganischen Chemie, Verlag
Chemie GmbH, Berlin, FRG, Various Editions (1927-1966t.
16. D. V. Hale, M. J. Hoover and M. J. O'Neill, Phase Change
Materials Handbook. NASA Contractor Report NASA CR-
61363, NASA Marshall Space Flight Centre, Alabama I1971).
17. D. Heine and A. Abhat, Investigation of physical and chem-
ical properties of phase change materials for space heat-
ing/cooling applications. In SUN: Mankind's Future Source
of Energy, (Edited by F. de Winter and M. Cox), Vol. 1, pp.
500-506. Pergamon Press, New York (1978).
18. D. Heine, The chemical compatibility of construction
materials with latent heat storage materials. Proc. Int. Conf.
on Energy Storage, pp. 185-192. Brighton (1981).
19. G. A. Lane and D. N. Glew, Heat-of-fusion systems for solar
energy storage. Proc. Workshop on Solar Energy Storage
Subsystems for the Heating and Cooling of Buildings, pp.
43-55. Charlottensville, Virginia (1975).
20. H. G. Lorsch, K. W. Kaufmann and J. C Denton, Thermal
energy storage for solar heating and off-peak airconditioning.
Energy Conversion 15, 1-8 (1975).
21. N. A. Malatidis and A. Abhat, Investigation of the ther-
mophysical behaviour of calcium chloride 6-hydrate for use
as heat storage material in latent heat stores (in German).
Forschung Ingenieur-Wesen 48, 1-26 (1982).
22. S. Marks, An investigation of Glauber's salt with respect to
thermal cycling. Solar Energy 25, 255-258 (1980).
23. J. Schr6der, R. and D systems for thermal energy storage in
the temperature range from -25°C to 150°C. Proc. Seminar
New Ways to Save Energy, 495-504. Reidel, Dordrecht,
(1980).
24. M. Telkes, Solar house heating--A problem of heat storage.
Heating and Ventilating 44, 68-75 (1947).
25. M. Telkes, Nucleation of supersaturated inorganic salt solu-
tions. Industrial and Engineering Chemistry 44, 1308-10
(1952).
26. M. Telkes, Solar energy storage. ASHRAE J. 16, 38 (1974).
27. M. Telkes, Thermal storage for solar heating and cooling.
Proc. Workshop on Solar Energy Storage Subsystems for the
Heating and Cooling of Buildings, pp 17-23. Charlottesville,
Virginia, U.S.A. (1975).
28. J. Teubel, W. Schneider and R. Schmiegel, Erd61paraffine,
VFB Deutscher Verlag fiir Grundstoff, Leipzig, GDR (1%5).
29. N. Yoneda and S. Takanashi, Eutectic mixtures for solar heat
storage. Solar Energy 21, 61-63 (1978).
30. M. Zief and W. R. Wilcox, Fractional Solidification, Vol. 1.
Marcel Dekker, New York (1967).
332 A. ABHAT

APPENDIX

CONiROL I
TEMPERATURE
RECORDER I
TEST TUBE
THERMOSTAT / THERHOCOUPLE
HEAT STORAGE MATERIAL

A. Glass test tube apparatus[4]

I~r moco~q~e

II II II /r Seal

/, / __1 __ .~1 ~ -- g'~'~/ ~ed,~

910

B. Glass container apparatus[4]

THERMOCOUPLES
TEl TE2 TE3 TEe TE5

STORAGE CHAMBER

, }
" ~ C H A N G E R

ALUMINIUM ~ P I P E

FINS HEAT STORAGECONTAINER

MATERIAL

C1. Laboratory model of the finned heat pipe heat exchanger[2]

CALORIMETER

THERMOCOUPLE
WIRING
C2. Laboratory model of the finned annulus heat exchanger element [5]