Determination of thermo-physical properties and stability testing
of high-temperature pha-change materials for CSP applications
Ming Liu a,n,J.C.Gomez b,C.S.Turchi b,N.H.S.Tay a,W.Saman a,F.Bruno a
a Barbara Hardy Institute,University of South Australia,Mawson Lakes Boulevard,Mawson Lakes,SA5095,Australia
b Concentrating Solar Power Program,National Renewable Energy Laboratory,15013Denver West Parkway,Golden,CO80401,USA
a r t i c l e i n f o
Article history:
Received19November2014
Received in revid form
3March2015
Accepted11March2015
Available online1April2015
一般献血多少毫升Keywords:
Concentrating solar power
Pha change materials
Thermo-physical properties
Thermal stability
Thermal cycling
Sub-cooling
a b s t r a c t
This paper prents the thermo-physical properties and stability testing results of six high-temperature
pha-change candidate materials for potential u as a cascaded storage system for concentrating solar
power applications.This type of storage is a promising technology becau it offers a higher utilization of
the possible pha change and a more uniform heat-transferfluid outlet temperature,compared with the
single pha-change material(PCM)storage system.The tested materials were inorganic eutectic PCMs
with reported pha-change temperatures between300°C and600°C.Four PCMs were made from
carbonate salts(Na2CO3,K2CO3,and Li2CO3)and two from chloride salts(NaCl,MgCl2,and KCl).The
pha-change temperature,pha-change enthalpy,and specific heat of the PCMs were measured
using a differential scanning calorimeter.Large material samples were tested in an oven subjected to
劳驾
multiple melt–freeze cycles.The results showed that the carbonate PCMs have a very high degree of sub-
cooling in the initial cycles,which decread in subquent cycles.The chloride PCMs have a negligible
degree of sub-cooling.There is some disagreement between the measured and reported thermo-physical
property values of the tested materials,which demonstrates the uncertainty associated with published
property values.One carbonate PCM and one chloride PCM were recommended as promising latent heat
storage materials.
&2015Elvier B.V.All rights rerved.
1.Introduction
Concentrating solar power(CSP)plants constitute an increasing
proportion of future power generation worldwide.CSP systems can
incorporate thermal energy storage(TES),which solves the time mis-
match between the solar energy supply and electricity demand pro-
files,and allows for more efficient u of the turbine and other power-
block components.The state-of-the-art TES system relies on nsible
storage and employs molten salts in cold and hot tanks.Compared to
nsible heat storage,latent heat storage using pha-change materials
(PCMs)allows large amounts of energy to be stored in relatively small
volumes,resulting in a smaller storage system and reduced cost of
energy for CSP plants[1,2].Latent heat storage has been applied in
numerous low-temperature applications[3–6]and is quite promising
for future high-temperature thermal storage applications[7,8].
The required PCMs must have a pha-change(melting and freez-
ing)temperature within the desired usage temperature range.For the
majority of parabolic trough and solar tower power plants,the oper-
ating temperature of the solarfield is290–390°C and290–565°C,
respectively.The currently available heat-transferfluids(HTFs)are
synthetic oil,water/steam,and molten nitrate or nitrite salts[9].
Alternative HTFs,such as supercritical carbon dioxide(s-CO2)and other
molten salts,are being investigated by rearch teams worldwide.
Higher operating temperatures of over650°C are achievable,and it is
known that higher HTF operating temperature generally allows for
higher turbine efficiency.Michels and Pitz-Paal[10]designed and
tested a cascaded PCM system using a ries of three PCMs,spanning
the operating temperature range from290°C to350°C.The experi-
mental results indicated that the cascaded system offers a higher uti-
lization of the possible pha change compared with the single PCM
system.Enormous works have been carried out by Abengoa Solar LLC
[11]to develop cascaded arrangement storage systems by using a
variety of eutectic salts and metal alloys as PCMs.The cascaded storage
system offers a more uniform HTF outlet temperature[10,12]and
higher exergy efficiency[13,14].Therefore,it is worthwhile to identify
candidate PCMs with melting temperatures in the range of300–650°C.
Inorganic salts are promising candidates for high-temperature
PCMs becau they produce a large enthalpy change without
substantial density change during solid–liquid pha transforma-
tion.Kenisarin[15]published a comprehensive review on PCMs in
the melting temperature range of120–1000°C.The PCMs include
fluorides,chlorides,bromides,hydroxides,nitrates,carbonates,
Contents lists available at ScienceDirect
journal homepage:/locate/solmat
Solar Energy Materials&Solar Cells
dx.doi/10.1016/j.solmat.2015.03.014
0927-0248/&2015Elvier B.V.All rights
rerved.
n Corresponding author.Tel.:þ61883025132;fax:þ61883023380.
E-mail address:ming.liu@unisa.edu.au(M.Liu).
Solar Energy Materials&Solar Cells139(2015)81–87
molybdates,and others.Binary and ternary eutectic compositions bad onfluorides and chlorides have high latent heat of fusion; also,the cost of the latter is low.Fluoride salts are less attractive becau of their relatively high cost and high corrosion to the containment materials[16].It has been reported that the melting temperature and heat capacity increa is in the following order: nitrates,chlorides,carbonates,andfluorides[1].The German aerospace center did intensive rearch t
o investigate sodium nitrate as a storage medium in direct steam generation technology [17–19].Gomez et al.[20]identified three PCMs with melting temperatures near320°C,350°C,and380°C.They evaluated the properties of the PCMs and found that the PCMs with chloride anions only were chemically unstable.
The storage capacity of a latent heat storage system is deter-mined by the specific heat of the storage material as well as the enthalpy of the pha change(latent heat of fusion),which can potentially result in a smaller and lower-cost alternative to the currently available nsible storage system.Therefore,it is neces-sary to investigate the thermo-physical properties of the candidate PCMs becau they have a direct impact on the thermal perfor-mance and cost analysis of the storage system in the CSP plants.In general,becau many of the properties given in the literature are usually calculated rather than experimentally verified values,there is a discrepancy between the two values[20].As an example,the measured latent heat of fusion determined by Gomez et al.[20]is only half of the theoretical value for some eutectic nitrate and chloride PCMs.The discrepancy occurs becau the mixtures and eutectics do not exhibit ideal mixing behavior[8].
初一英语听力The candidate PCMs should melt congruently and have insignif-icant sub-cooling;otherwi,the storage capacity of the system will be reduced.Also,good chemical and thermal stability are necessa
ry to ensure a reasonable life span of the storage system,which is another important criterion for the PCM lection.The life span of the storage system for CSP plants is considered to be between20 and30years.However,only limited rearch exists on the thermal stability of high-temperature PCMs.Laing et al.[18]tested a proto-type PCM storage unitfilled with sodium nitrate.The operation of 172cycles(more than4000h)proved that there was no change in the melting temperature and no decomposition of the PCM.Shin et al.[16]tested a salt eutectic(32wt%Li2CO3–35wt%K2CO3–33wt%Na2CO3)and found that it exhibits a distinct pha-change temperature(395–397°C)without pha paration or sub-cooling over50thermal cycles,and the corrosion of the salt within the SS316container was negligible.Petri et al.[21]evaluated four car-bonate salts(Li2CO3,Na2CO3,52.2wt%BaCO3–47.8wt%Na2CO3,and 81.3wt%Na2CO3–18.7wt%K2CO3)as PCMs and they cycled tho candidates for13,21,36,and38cycles,respectively.They found that the eutectic salt with52.2wt%BaCO3and47.8wt%Na2CO3showed stable performance,but the melting point measured was30°C higher than the published value(686°C).The salt mixture with 81.3wt%Na2CO3–18.7wt%K2CO3was not stable and it melted incongruently.When the pure salts(Li2CO3and Na2CO3)were tested in aled containers,they leaked through the weld andfittings.
This paper aims to characterize the thermo-physical properties of the six PCM candidates with melting temperatures between300°C and600°C.The temperature range was lected to allow a cascaded PCM storage system to be developed for current and future solar tower power plants.The composition of the carbonate and chloride salt eutectic PCMs was taken from the literature.The properties measured include the pha-change temperature,latent heat of fusion,and solid and liquid specific heat capacities.Also,the thermal stability of the salt eutectics has been examined by repeatedly cycling the candidates between300°C and600°C in a furnace.To our knowledge,this is thefirst time that the thermo-physical properties of the investigated PCMs have been measured by experiment and their thermal stability is evaluated.2.Materials and methodology
2.1.Materials
Table1provides the listing of the compositions and thermo-physical properties of the salt mixtures studied.Candidates CA1–CA4 are eutectic mixtures of carbonate salts and CH1and CH2are eutectic mixtures of chloride salts.The thermo-physical properties of candidate CA1were obtained by experiment,whereas the properties of the other candidates are calculated values.The individual carbo-nate and chloride salts were supplied by the Alfa Aesar and Sigma Aldrich with purity levels greater than99%.The salts were dried in a muffle furnace at120°C for24h.The single dried salts were
weighed in a balance with a resolution of0.1mg and mixed at a specific ratio as shown in Table1.Becau chloride salts are highly hygroscopic,the weighing and mixing procedures were performed inside a dry glove-box under UHP nitrogen atmosphere.Before performing the differential scanning calorimeter(DSC)test,the samples were placed in a furnace at120°C for2h to remove any moisture absorbed during sample handling and preparation.
如何修改网络密码2.2.Methodology
2.2.1.Differential scanning calorimeter
The specific heat capacity and latent heat of fusion and pha-change temperature were evaluated using a DSC(Netzsch DSC404), which operates according to the heat-flux principle.A sample and a reference are subjected to a controlled temperature program(heat-ing,cooling,or isothermal gment)and the DSC nsor measures the temperature of the sample and the difference between the sample and reference.Then the heat-flow difference between the sample and the reference can be determined.To measure the solid and liquid specific heat capacities for the PCM samples,the ASTM E1269-05standard was followed,in which a sapphire disk with known heat capacity is employed as the reference material.
Before performing the test,the chemical stability of the DSC container materials(aluminum,SS316,aluminum oxide,and gra-phite)with the PCMs was evaluated in a controlled-atmosphere furnace under UHP nitrogen gas.Aluminum oxide and graphite crucibles with lids were ud in the pha-change temperature and enthalpy measurement becau they showed good corrosion resis-tance.In the heat-capacity measurement,graphite crucibles were ud to obtain fast heat transfer into the sample due to the high thermal conductivity of the graphite.The analys were conducted under a constant stream of UHP nitrogen at aflow rate of20mL/min at a heating/cooling rate of10K/min for pha-change temperature and enthalpy measurements,and20K/min for specific-heat capacity measurements.A smaller heating/cooling rate of5K/min was also employed in the preliminary test,but no obvious discrepancy was found in the results;hence10K/min was ud for the temperature
简历特长怎么写比较好Table1
Composition of candidate PCMs and their melting temperatures(T m),latent heats of fusion(ΔH),and solid and liquid specific heats(C p,s and C p,l)[15].
Candidate Eutectic composition
(wt%)
Reported values
T m(°C)ΔH(J/g)C p,s(J/g K)C p,l(J/g K) CA132Li2CO3–35K2CO3–
33Na2CO3
397276 1.67 1.63
CA228.5Li2CO3–71.5K2CO3498316––
CA335Li2CO3–65K2CO3505344 1.34 1.76
CA422Li2CO3–62K2CO3–
16Na2CO3
580288 1.80 2.09
CH152MgCl2–48NaCl4504300.92 1.00
CH264MgCl2–36KCl4703880.840.96
M.Liu et al./Solar Energy Materials&Solar Cells139(2015)81–87 82
and enthalpy tests.The DSC has a calorimetric data resolution of 0.01μW and a temperature accuracy of 70.20K.DSC measurements were repeated four times for carbonate candidates CA1–CA4and the average values were taken as the results.
Measurement of molten chloride PCMs at high temperatures is dif ficult becau the chloride salts in the liquid state strongly wet the graphite containers.As a result,the salt creeps toward the crucible wall and the salt in the container is thicker near the wall and thinner in the middle.This condition leads to a temperature gradient in the sample that affects the accuracy of the measure-ment.Therefore,in pha-change temperature and enthalpy measurements,three samples were taken and tested for each chloride candidate,two melt –freeze cycles were carried out for each sample,and the average values were taken as the results.2.2.2.Thermal cycling test using high-temperature furnace
A furnace with a maximum operating temperature of 1200°C was ud to conduct the stability test (melt –freeze cycling).Gen-erally,the sub-cooling problem is more vere in small samples,such as t
ho ud for DSC testing [22].To avoid this problem and to get more realistic values of the pha-change temperature range,larger PCM samples (50–150g)were ud in the stability test.The experimental tup in the furnace is prented in Fig.1.The furnace has an internal dimension of 0.25m Â0.25m Â0.3m.Carbonate and chloride PCM samples were contained in SS316L and aluminum oxide crucibles,respectively.Type K omega TM thermocouples with an accuracy of 70.75%were covered with aluminum oxide sheaths to protect the thermocouple from the corrosion of the salt.Then the sheathed thermocouples were inrted into the PCM samples to measure their temperatures during both melting and freezing.The temperature of the oven was controlled at 300°C and 600°C during the freezing and melting process,respectively.The temperatures inside the PCM samples and the furnace were recorded every 10s by using a data acquisition system.The thermal stability of candidate CA1as shown in Table 1was tested by Shin et al.[16]and the eutectic exhibits a distinct pha-change temperature (395–397°C)without pha paration or sub-cooling over 50thermal cycles.Therefore,it was deemed unnecessary to include CA1in this test.2.3.DSC analysis
The software Proteus s from Netzsch was ud to measure and evaluate the DSC data.After the measurement,the DSC signals (heat flow)of the sample and the reference under the same controlled temperature program were plotted against temperature.The pha-change temperature (melting and freezing)was taken as the
baline ont.The latent heat of fusion was estimated by the integration of the DSC heat flux under the pha transformation peak.The speci fic heat capacity of the PCM was evaluated bad on the known heat capacity of the sapphire,as shown in Eq.(1),
C m m C DSC DSC DSC DSC (1)
p,sample sapphire sample
sample baline sapphire baline
p,sapphire
=
×
−−×where C p,sample and C p,sapphire are the speci fic heat capacities of the sample and the sapphire;m sample and m sapphire are the mass of the sample and the sapphire;and DSC sample ,DSC sapphire ,and DSC baline are the DSC heat flux signals for the sample,sapphire,and baline measurements,respectively.
3.Results and discussion
3.1.Pha-change temperature and latent heat of fusion
The pha-change temperatures and latent heat of fusion of the PCMs under investigation were evaluated and the results are sum-marized in Table 2.One sample for each of the candidates CA1–CA4was tested and each subjected to four melt –freeze cycles.Due to the creeping behavior of the chloride salt on the crucible wall,three samples for candidates CH1and CH2were tested and each of the subjected to two melt –freeze cycles.The sample weights are listed in Table 2.The thermo-physical properties of CA1measured in this study are very clo to the reported values,which are the experi-mental values.For the other samples,there is a discrepancy between the measured value and the calculated value,shown in Table 1.For PCMs made from carbonate salts,the freezing temperatures are usually much lower than the melting temperatures,which indicate that tho PCMs have higher degrees of sub-cooling.Compared to carbonate salts,chloride candidates CH1and CH2have much smaller degrees of sub-cooling.Sub-cooling is undesirable becau it reduces the usability of PCMs and can also completely prevent heat recovery if too vere [23].Some PCMs show sub-cooling in the DSC test becau of the small amount of sample ud [24].However,they could show no or a small degree of sub-cooling in large-scale plants [24].
Furthermore,the latent heat of solidi fication (freezing)is 30–60J/g less than the latent heat in the melting process for carbonate candi-dates CA1,CA3,and CA4.In contrast,ΔH f and ΔH m values differ by only 1–14J/g for chloride candidates CH1and CH2and carbonate candidate CA2.Figs.2–4show the DSC curves of samples CA1,CA3,and CA4for four concutive melt –freeze cycles.In addition to the major exothermic pha change where the materials relea most
of
Fig.1.Schematic diagram of the experimental tup for the thermal cycling test in the furnace.
M.Liu et al./Solar Energy Materials &Solar Cells 139(2015)81–8783
their latent heat of solidi fication there is another noticeable exother-mic pha change (shown in the circles of Figs.2–4)that occurs mostly at 100K below the melting temperature of candidates CA1,CA3,and CA4during heating.During this process,the absorbed energy is roughly equal to the dif
ference of the pha-change enthalpy mea-sured in the melting and freezing process.For candidate CA1,this process sometimes occurs just before the melting process starts,as prented in Fig.2.In addition,the measured melting temperature of CA4is 422.5°C,which is dramatically lower than the reported value (580°C)[15].
Fig.5shows the DSC curves of sample CA2for four concutive melt –freeze cycles.The melting process for cycles 1–4are not consistent,in which the pha-change temperatures and latent heats of fusion vary randomly from 468.7°C to 500.4°C and 234.1J/g to 265.5J/g,respectively.However,all of the freezing process occurred at a similarly sharp and narrow temperature range,and the measured freezing temperature and latent heat of fusion are 440.473.0°C and 233.371.5J/g,respectively.The average values of the latent heats obtained during both melting and freezing process are quite similar.
The measured pha-change temperatures of candidates CH1and CH2are slightly lower than the reported values [15],but the measured enthalpies are around half of the reported values.Gomez [25]tested three eutectic PCMs (34.81wt%NaCl –32.29wt%KCl –32.9wt%LiCl,80.69wt%KNO 3–11.87wt%KBr –7.44wt%KCl,and 60wt%MgCl 2–20.4wt%KCl –19.6wt%NaCl)using the DSC and found that the latent heat of fusion was almost half or less than the expected values from the literatur
e.3.2.Speci fic heat capacity
The speci fic heat capacity curves for carbonate PCM samples in the solid and liquid states,excluding pha transition,are shown in Fig.6.Generally,the speci fic heat capacities of the solid
carbonate PCMs increa slightly as temperature ris,and they are relatively constant in the liquid state.The average speci fic heat capacities on the solid and liquid states are listed in Table 3.The results for candidates CA1and CA3were quite clo to the pre-viously reported values [15].There is a sizeable discrepancy between the measured and reported values for CA4.However,it is believed that the result from this study is more reliable becau the speci fic heat capacities of all the other carbonate eutectics are in a similar range.
Becau of the creeping problem of the liquid chloride salt,only the speci fic heat capacities for solid chloride PCM samples are shown in Fig.7and Table 3.The eutectic of MgCl 2and NaCl (CH1)has a slightly incread speci fic heat capacity with temperature,
Table 2
DSC test results of pha-change temperature and latent heat of fusion (T m and T f :melting and freezing temperature;ΔH m and ΔH f :latent heats of melting and freezing).Candidate Eutectic composition (wt%)Weight (mg)T m (°C)T f (°C)ΔH m (J/g)ΔH f (J/g)Sub-cooling (°C)CA132Li 2CO 3–35K 2CO 3–33Na 2CO 313.7396.7374.0278.9248.722.7CA228.5Li 2CO 3–71.5K 2CO 311.9479.9440.4234.3233.339.5CA335Li 2CO 3–65K 2CO 3
10.9503.7450.1295.2269.058.6CA422Li 2CO 3–62K 2CO 3–16Na 2CO 314.5
425.5412.8274.7211.712.7CH152MgCl 2–48NaCl 12.6/12.4/9.3439.1430.7204.2190.48.4CH2
64MgCl 2–36KCl
8.3/15.9/15.7
459.1
450.8
210.2
199.3
8.3
Fig.2.DSC curves of CA1(32Li 2CO 3–35K 2CO 3–33Na 2CO 3,wt%)for four melt –freeze
cycles.
Fig.3.DSC curves of CA3(35Li 2CO 3–65K 2CO 3,wt%)for four melt –freeze
鱼羊鲜的做法
cycles.
Fig.4.DSC curves of CA4(22Li 2CO 3–62K 2CO 3–16Na 2CO 3,wt%)for four melt –freeze cycles.
M.Liu et al./Solar Energy Materials &Solar Cells 139(2015)81–87柯基狗狗
84
and the eutectic of MgCl 2and KCl (CH2)has an approximately constant speci fic heat capacity between 0.70and 0.75J/(g K)over the measured temperature range.3.3.Thermal cycling stability
As previously mentioned,the chemical and thermal stabilities of a PCM are important criteria for material lection.Any degra-dation of the PCM would decrea its latent heat of fusion and alter its pha-change temperature range [22],which will reduce the life span of the storage system and reduce the ef ficiency of the power block if the pha-change temperature drops.
3.3.1.Carbonate PCM candidates
Candidates CA2–CA4(about 150g)have undergone 100melt –freeze cycles.Figs.8–10show the temperature pro files of CA2–CA4during different melt –freeze cycles.The pro files prent the
typical three stages of PCM melting and freezing [26].A sub-cooling phenomenon was obrved in all the carbonate samples,and it is more vere in samples CA2and CA3,which can reach as high as 35K.This value is smaller compared to that in the small-sample DSC test.The aluminum oxide sheath in sample CA2was broken after 50cycles;therefore,the temperature pro file after 50cycles is not prented in Fig.8.In sample CA2,the sub-cooling is nearly 35K in the first and tenth cycles.It reduces to 25K at the 30th cycle and further reduces to 16K at the 50th cycle.
In sample CA3,the sub-cooling occurs frequently in the first 70cycles.However,sub-cooling was rarely detected afterwards.As shown in Fig.9a,the temperature pro files for the first and the one-hundredth cycles are quite clo in the melting process and it does not show any degradation.The freezing temperature where sub-cooling occurs is about 30K higher than that without sub-cooling.Further rearch is warranted to investigate the caus of this phenomenon.
Fig.10a shows that sample CA4melts over a very wide tem-perature range,which is also proved by the DSC results in Fig.4.Also,as the number of cycles increas,the range of the melting temperature increas.After 20cycles,the pha-change phenom-enon becomes unnoticeable as the temperature where melting commences cannot be clearly en.Fig.10b shows that the tem-perature pro file shows a plateau at around 420°C when the sample CA4experienced the first solidi fication.A
s the number of cycles increas,the solidi fication temperature range becomes
wide.
Fig.5.DSC curves of CA2(28.5Li 2CO 3–71.5K 2CO 3,wt%)for four melt –freeze cycles.
11.2
1.4
1.6
1.8
250
300
350
400
450
500
550600650700
S p e c i f i c h e a t c a p a c i t y (J /(g ·K ))
Temperature (°C)
Fig.6.Results of speci fic heat capacities for carbonate PCM candidates in the solid and liquid phas.
Table 3
Averaged results of DSC speci fic heat capacities of PCM candidates (C p,s and C p,l :solid and liquid speci fic heat).Candidate
Weight (mg)
Reported values Measured values C p,s (J/g K)
C p,l (J/g K)C p,s (J/g K)C p,l (J/g K)CA133.6 1.67 1.63 1.42 1.68CA237.8–– 1.29 1.66CA335.0 1.34 1.76 1.43 1.73CA434.9 1.80 2.09 1.25 1.56CH136.50.92 1.00 1.07–CH2
35.0
0.84
格莱美直播0.96
0.73
–
0.60.811.21.4
S p e c i f i c h e a t c a
p a c i t y (J /(g ·K ))
Temperature (°C)
Fig.7.Results of speci fic heat capacities for chloride PCM candidates in the solid pha.
Fig.8.Temperature of sample CA2(28.5Li 2CO 3–71.5K 2CO 3,wt%)during (a)melting process and (b)freezing process.
M.Liu et al./Solar Energy Materials &Solar Cells 139(2015)81–87
85
