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

net/publication/260317704

Analysis of stratified thermal storage systems: An overview

Article  in  Heat and Mass Transfer · January 2014

DOI: 10.1007/s00231-014-1302-8

CITATIONS READS

37 5,204

3 authors:

Howard Njoku Val Ekechukwu

University of Nigeria University of Nigeria
76 PUBLICATIONS   249 CITATIONS    86 PUBLICATIONS   1,295 CITATIONS   

SEE PROFILE SEE PROFILE

Samuel Onyegegbu
University of Nigeria
42 PUBLICATIONS   308 CITATIONS   

SEE PROFILE

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

Infrared Thermography Analysis of Energy Systems View project

Improved solar-thermoelectric thermal devices View project

All content following this page was uploaded by Howard Njoku on 18 February 2015.

The user has requested enhancement of the downloaded file.

Heat and Mass Transfer manuscript No.
(will be inserted by the editor)

Assessment of stratified thermal storage systems using energy and

exergy methods: A review
H.O. Njoku1 , O.V. Ekechukwu2 , S.O. Onyegegbu1
1
Department of Mechanical Engineering, University of Nigeria, Nsukka, Nigeria.
2
National Universities Commission, Maitama, Abuja FCT, Nigeria.

Published in Heat and Mass Transfer Vol. 50, No. 7, pp. 1017 – 1030.
The final publication is available at http://link.springer.com/article/10.1007/s00231-014-1302-8

Abstract The presence of stratification is well known energy demand. Energy storage may also be required in
to improve the performance of stratified thermal energy order to make energy available at locations remote from
storage systems (STESS). The major energy and exergy where it was generated.
methods for modeling and assessing the performance of Energy may be stored in several forms, including
STESS are reviewed in this presentation. Current ana- chemical energy (e.g. in biomass and electrochemical bat-
lytical and numerical methods for modeling STESS are teries), mechanical energy (e.g. in pumped hydro, com-
surveyed, with their strengths and weaknesses. An ex- pressed air and flywheels) and as heat [2]. When energy
tensive survey of performance assessment methods based is stored in the form of heat, this is known as thermal
on energy, entropy and exergy analyses is also presented. energy storage (TES) and such storage may further be in
This survey highlights entropy generation ratios, which the form of either Sensible TES or Latent TES. While la-
are based on second law considerations, as effective in tent thermal energy storage (LTES) refers to the storage
quantifying the improvements in STESS performances of heat as the latent heat of phase change of a material
that result from the presence of stratification unlike energy- as it undergoes a change of phase (usually from solid to
based measures. The entropy generation number, Ns is liquid or vice versa) at a fairly constant temperature,
suggested as an effective tool for future performance sensible thermal energy storage (STES) is the storage of
studies on STESS. heat in a material as it experiences a change of temper-
ature.
Key words stratified thermal storage–energy analysis– Sensible TES systems may utilize either water or
exergy analysis–entropy generation minimization–thermal rocks as suitable storage media. Those which utilize rocks
storage efficiencies–performance assessment parameters as the storage medium are better suited for air heat-
ing applications, while water-based ones are preferred
for water heating applications [3]. These materials are
relatively inexpensive and readily available, contribut-
ing significantly to the cost-effectiveness of constructing
1 Introduction
and operating STES systems when compared to their
LTES counterparts. Furthermore, because of the com-
Energy storage is an energy management technology that
paratively higher thermal conductivities of STES mate-
has found increased application due to heightened con-
rials, the heat storage and release rates in STES sys-
cerns about the need for energy/exergy conservation [1].
tems are high. In the case of water-based sensible TES
Due to the fact that periods of energy demand and sup-
systems, irrespective of whether the component supply-
ply are not usually coincident, energy storage is useful for
ing heat (e.g. solar collector, fossil fuel or biomass-fired
leveling peaks in energy demand and supply and thus for
boiler, etc) or the heat storage component is of interest
eliminating wastages due to excess capacity during pe-
in the assessment of performance, the presence of tem-
riods of off-peak demand, and for augmenting for short-
perature stratification in the storage medium has been
ages during peak-demand periods. This is also the case
shown to lead to improved performance [3, 4, 5, 6, 7, 8].
when energy is demanded for at a rate higher than that
The first and second laws of thermodynamics lead to
at which it may be supplied; stored energy may at such
three approaches to the analysis and performance assess-
times be relied upon to meet up with the higher rates of
ment of stratified STES systems (and indeed any other
Correspondence to: H.O. NJOKU (e-mail: modes of TES.) An energy analysis results from solely
[email protected]) first law considerations, in which energy balances are
2 H.O. Njoku et al.

carried out on the system of interest in order to deter- The importance of temperature stratification in sen-
mine rates of energy flows into and out of the system. sible storage has been known since the 1970s [4, 13, 14]
On the other hand, the analysis of entropy transfers as- and since then much work has been done on the sub-
sociated with mass flows and entropy-generation due to ject, which have been presented in the reviews by refs
energy transfer and transformation processes within a [5, 14, 15, 16]. These studies have shown that a stratified
TES unit depends solely on second law considerations. STES unit will deliver superior performance compared
Exergy analysis however, results from considering a com- to a fully-mixed one. In a STES unit connected to a solar
bination of the first and second laws. The inclusion of collector, for example, if stratification is present in the
second law considerations is essential to a proper evalu- unit during charging, it will be possible for more heat to
ation of the performance of TES systems, (either in the be extracted from the collector and for a longer period
form of entropy or exergy analysis) [9, 10, 11, 12]. of time than would be possible with a comparable fully-
A survey of the different approaches to the theoreti- mixed unit, because the lower temperature of the return
cal study of the operation and performance characteris- fluid from the bottom of the storage unit to the collector
tics of stratified TES systems is presented in this review. will create a higher temperature difference between the
fluid and the collector. Heat delivery from a stratified
storage tank to the load during discharge would also be
2 Stratification in STES systems and suggested for a longer period of time than from a comparable fully
means of sustaining stratification mixed unit because of higher temperatures at the top
of the stratified tank. In the study of Cristofari et al.
Stratification results in a hot (or cold) water storage [7] for example, a stratified storage tank was coupled to
tank when different regions in the tank are at different a solar collector of copolymer material. The presence of
temperatures. The density variations which result from stratification in the storage tank led to higher efficiencies
those temperature differences will cause the hotter wa- and higher mean daily collector heat output compared
ter masses to rise to the top and the cooler water masses to a system with a fully-mixed storage tank. Similarly,
to fall to the bottom of the storage tank, with a middle Alizadeh [15] reported that performance is improved by
layer transition zone (the thermocline), separating the having the contents of a storage tank initially stratified
hot upper zone from the cold lower zone. A typical tem- at the start of operation.
perature profile in a stratified hot water storage is shown For a heat storage unit, if hot storage fluid is let into
in Figure 1. the unit from its top and cold fluid withdrawn from its
bottom, a state of stratification will ensue within the
unit. This state will however not always persist due to
the action of several mechanisms which eventually result
in a fully mixed store. The loss of stratification in an ini-
tially stratified unit, also known as destratification, will
occur as a result of (a) mixing of the incoming stream
with the fluid already in the storage tank during charging
or discharge, (b) heat conduction along the storage tank
wall, (c) heat conduction between fluid layers at different
temperatures, and (d) heat losses across the tank walls
to ambient. Destratification occasioned by the mixing of
the inlet jet stream with the fluid in-situ was shown by
VanBerkel [17] to be a double stage process consisting of
fluid withdrawal from the thermocline by viscous drag
and a subsequent mixing of fluid particles, accompanied
by increased heat diffusion across fluid layers.
The rate of destratification can be reduced by elimi-
nating or diminishing the action of the processes afore-
mentioned. Mixing can be reduced by decelerating the
incoming fluid entering the storage tank introducing suit-
ably devised obstacles. Divergent inlet cones [15], inlet
and outlet diffusers [18, 19], porous inlet manifolds [20,
21, 22] and impingement plates of different configura-
tions [8, 16, 23, 24, 25, 26], have been introduced near the
inlets or outlets of stratified STES units to achieve this,
with varying degrees of success. Destratification due to
fluid mixing may further be reduced by partitioning the
Fig. 1 Typical temperature profile in a stratified STES unit. STES unit to physical separate fluid portions at different
Assessment of stratified thermal storage systems using energy and exergy methods: A review 3

temperatures [8, 27, 28]. Generally, the presence of these 3 Energy analysis of stratified STES systems
obstacles have been shown to aid the sustenance of strat-
ification by reducing mixing between fluid layers within 3.1 Mathematical simulation of heat and fluid flow in
the storage tank [16, 23]. stratified STES systems
Theoretical studies on stratified STES systems have re-
In order to lower conduction heat transfer along the
sulted in mathematical models of varying degrees of com-
tank walls, low thermal conductivity materials such as
plexity. Transient one-, two- and three-dimensional mod-
fibreglass and HDPE can be used for storage tank con-
els have been developed and used to analyze the perfor-
struction [19], however, choosing an appropriate length
mance of STES units. The choice of the level of complex-
to diameter ratio may render the axial tank wall con-
ity is governed by a trade-off between simplicity and low
duction effects insignificant [29]. By sufficiently insulat-
computational cost on one hand, which obtains for 1-
ing storage tanks walls, heat losses to ambient may be
dimensional models, and on the other hand, how closely
reduced to insignificant levels [29, 30, 31].
the model describes the underlying physics of the prob-
Barba and Spiga [32] investigated the inclusion, into lem and reproduces the temperature and flow fields, as
a STES tank, of an encapsulated phase change material obtains with the models of higher dimensions. Further-
(PCM) whose melting point was close to the required more, whereas analytical or numerical solution techniques
discharge/utilization temperature and found that the have been applied in the implementation of the one-
tank’s performance was improved by this, as it received dimensional models, the models of higher dimensions
the benefits of thermal energy storage in both latent and have been resolved solely by numerical methods.
sensible heat forms. In the past decade, hybrid thermal Yoo et al. [37] analytically solved the 1-D energy
energy storage systems (HTESSs) in which appreciable equation for adiabatic finite size stratified storage tanks
and comparable quantities of heat are stored as both with fixed inlet temperature, using Laplace transform
latent and sensible heat in stratified TESs have been in- techniques. In a subsequent work [38], and using simi-
vestigated by the group of Cabeza et al. [33, 34, 35, 36]. lar techniques, they considered tanks with variable inlet
By inserting a PCM module into the upper region of a temperatures. Comparing their results with closed form
hot water storage unit [35], solutions, they showed theirs to be more accurate than
the previous solutions of Cabelli [13], who ignored tank-
end effects by modelling a tank of infinite size. The 1-
D energy equation was also solved by Abu-Abdou and
– the thermal storage density of the unit could be sig-
Hussein [39] for an adiabatic finite size storage tank with
nificantly increased (compared to that of a sensible
inlet fluid temperatures varying as step and ramp func-
heat storage unit), leading to a reduction in the size
tions of time to obtain temperature distributions in the
of storage units required;
tank at different operation instants and under different
– heat losses from the upper water layers in the unit
charging rates. The finite size tank configuration of Yoo
can be compensated for by the latent heat stored in
et al. and Abu-Abdou and Hussein was more realistic
the PCM, thereby delaying destratification; and
than the infinite size configuration of Cabelli. However,
– the upper water layer can be partially reheated by
the practical usefulness of both their results and those
the latent heat stored in the PCM after draw-offs.
of Cabelli is limited by the complexity of the techniques
they employed and by the fact that their models become
unreliable at Reynolds numbers greater than 200 [40]. In
Refs. [33, 34, 35] reported the outcomes of experiments the 1-D analyis of Li and Sumathy [8], a nodal temper-
conducted on a stratified hot water storage tank which ature model was developed and used to study the per-
had 3 PCM capsules inserted in the upper section of the formance of a stratified storage tank coupled to a solar
tank. The studies were conducted on a 150 litre cylin- air conditioning system. Two 1-D numerical models were
drical hot water storage tank into which 1.5 litre volume also developed by Alizadeh [15] for a horizontal strati-
cylindrical PCM capsules were inserted. The inclusion fied cold storage tank, the Turbulent Mixing Model and
of the PCM capsules led to improvements in thermal the Displacement Mixing Model. In the turbulent mixing
storage density of up to 45%, reduction in reheat time model, fluid entering the tank was assumed to be equally
of about 20 minutes, and increase in the heat retention fed into the initial layers in the tank, while layers above
time of up to 200% [35]. Mazman et al. [34] also reported the mixing layers experienced plug flow. In the displace-
improvements in the performance of a stratified hot wa- ment mixing model, however, it was assumed that the
ter storage when PCM mixtures—paraffin/stearic acid, fluid entering the tank mixes with the initial layers of the
paraffin/palmitic acids and stearic acid/myristic acid, tank, followed by a plug flow in subsequent layers. The
were added. Their comparison of the three showed that former gave closer agreement with experimental results
the paraffin/stearic acid mixture gave the most improve- than the latter.
ment based on the ratio of the quantity of heat stored In order to extend the applicability of 1-D models
in the system that could be recovered for use. to flows with high Reynolds numbers, empirical factors
4 H.O. Njoku et al.

have been introduced into the model equations to ac- Stratified TES systems with embedded PCMs cap-
count for turbulence effects. An eddy conductivity fac- sules have also been modeled by Mehling et al. [35] and
tor, , to account for turbulent mixing at the inlet of stor- Ibanez et al. [36]. The models of the latter were based on
age tanks, was incorporated by Al-Najem and El-Rafaee the TRNSYS software, while the former, using a nodal
[40] into their model which was based on a Galerkin fi- temperature approach, developed a pure conduction model
nite element technique. Their results showed very close by descretizing the storage tank into fluid layers of con-
agreement with experimental data. A mixing coefficient, stant temperatures, while ignoring convection effects within
Z was also incorporated by Nelson et al. [29] into their the tanks. They also modeled the PCM as a single phase
1-D nodal temperature model. They went on further to substance of uniform temperature and ignored the effects
conduct a parametric study based on finite difference of the phase change phenomenon in the PCM. These
solutions of their ensuing model equations to determine simplifications led to significant deviations from experi-
the effect of different operating conditions on the per- mental results.
formance of thermal storage tanks when operated in the
charge, discharge and static modes.
Though 1-dimensional models can give significant in- 3.2 CFD and ANN modeling of Stratified TES Systems
sight into the operation of storage tanks, they are limited
by their inability to satisfactorily describe situations in- Increased access to and huge improvements in the com-
volving high mass flow rates and to handle situations in putational ability available on PCs has resulted in the
which obstacles are present in the storage tank, which is more widespread use of computer-based simulation or
the case with many practical systems. To illustrate this, CFD (computational fluid dynamics) tools in the analy-
a 1-dimensional plug-flow model and a 2-dimensional sis of fluid flow and heat transfer effects in heat storage
model were implemented for a cylindrical hot water stor- systems. 2-D and 3-D analysis of fluid flow and heat
age tank by Ghaddar and AlMaarafie [41], and both were transfer in TES systems have thus become prevalent
compared with experimental results. The 1-D model was in the literature in the recent past. The descretization
solved using finite difference methods, while the 2-D methods used in CFD analysis include finite difference
model was solved using spectral element methods. A methods, finite volume methods, finite element meth-
comparison of the two showed that the 2-D model gave ods, spectral element methods [44], and very recently,
closer agreement with the experimental results than the the space-time conservation element and solution ele-
1-D model. However, by adding to the 1-D model, a mix- ment (CE/SE) method [45, 46, 47]. Of these, the finite
ing parameter that was similar to those used in refs [29, volume method is the most widely used and is the ba-
40], results comparable to those of the 2-D model were sis of the major commercial and non-commercial CFD
obtained. The parameter was obtained empirically by codes in use [44, 48].
adjusting its value to minimize the standard deviation A 2-D plug-flow CFD model was used by Shin et al.
error between measured and predicted outcomes. But [26] to study the effect of storage tank size, inlet diffuser
because this mixing parameter depends on the Peclet shape and Froude number (and thus inlet velocity) on
number, and thus the significance of convective effects the performance of a stratified TES tank, and obtained
and the progression of the storage process, we see once results that were acceptably close to experimental re-
again that the satisfactory use of such a model is limited, sults. A 2-D CFD analysis also was performed by Zachár
requiring the availability of detailed experimental data. et al. [23] of a stratified hot water storage tank with in-
Ismail et al. [42] developed two 2-D models for strati- let plate diffuser. Finite difference forms of the governing
fied hot water storage tanks – one based on the solution equations were derived and solved using a Matlab code,
of the continuity, energy and momentum equations and the results were validated with experimental results and
the other on a plug flow model, (assuming uniform ax- found to compare favourably with the output of a com-
ial flow velocity throughout the tank, zero radial velocity mercial CFD package. Other 2-D CFD studies on strat-
and absence of convective effects). A comparison of both ified storages have been conducted by Van Berkel et al.
models showed that their results were similar soon after [4] and Abdoly and Rapp [31], (implemented on self-
the start of charging but diverged appreciably after the developed codes,) and by Shah and Furbo [24], (using a
effects of convective mixing had become significant. A commercial package.)
transient 2-D model, accounting for convection in the A 3-D analysis permits the minimum level of simpli-
axial direction and conduction in both axial and radial fications and helps the investigator replicate, as much as
directions, was developed and solved analytically by Issa possible, the underlying physics of the case being stud-
and Al-Nimr [43] for solar collector STESS. Their results ied. A 3-D CFD analysis was employed by Altuntop et
were verified with experimental findings. At high charg- al. [16] in their investigation of the effect of obstacles
ing velocities, axial conduction effects in the storage tank on stratification in a hot water storage tank. Colour
were found to be negligible and stratification poorly de- contours of temperature distributions in the tank were
veloped. Stratification was also poorly developed at low obtained in addition to other results which have been
tank inlet – outlet temperature differences. discussed in previous a section. The vast capabilities of
Assessment of stratified thermal storage systems using energy and exergy methods: A review 5

CFD analysis were explored by Cónsul et al. [49] in the The Richardson number, Ri, the ratio of buoyancy to
analysis of a horizontal stratified storage tank. They un- inertia forces in the storage tank, has been widely used
dertook a 3-D analysis using the finite volume descretiza- to measure stratification. It is expressed as
tion method and implemented it on a Beowulf cluster.
gβH(Tin − T0 )
Energy and exergy assessments of the storage unit were Ri = 2 (1)
done using results of the CFD analysis. A major fea- Uin
ture of their study was the drastic reduction in comput- where β, the compressibility of the storage fluid, is com-
ing time that was achieved—the use of a parallel multi- puted at the average temperature in the tank, Tav =
block algorithm to divide the computational domain into (Tin + T0 )/2. Storage tanks with higher Ri experience
blocks for solution on the separate PCs that made up the greater stratification because the momentum of the in-
cluster, meant that the computation was completed in coming fluid stream is insufficient to overcome the tem-
about 16 hours as opposed to about 3 12 days when the perature difference induced buoyancy forces, leading to
computation was done on a single PC. reduced mixing of the inlet stream with the resident
Recently, an artificial neural network (ANN) scheme fluid, and causing the incoming fluid to rise to the top
has been applied to the 1-D modeling of a stratified stor- of the storage tank. In tanks with low Ri, the incoming
age tank by Gećzy-Vı́g and Farkas [50]. The development fluid is entering with a momentum that is comparable or
of the ANN model required that the model to be trained even greater than the buoyancy forces, and sufficient to
with sufficient experimental data in order for the be- push it through the colder layers, leading to mixing, and
haviour of the system to be well captured by the model. in some cases direct discharge of the hot fluid. Van Berkel
Their model predicted the performance of the storage
‰
[17] and Van Berkel et al. [4] showed analytically that
tank to within less than 1 of experimental data. The the Richardson number had the most significant effect on
use of ANN modelling is only applicable to situations stratification in storage tanks. This was supported by the
where measured data already exists, since pre-existing recent comparative study of Castell et al. [51], in which
data is needed for training the ANN model, limiting its several dimensionless parameters popularly used to char-
usefulness for assessing novel or proposed designs. acterize stratification – the MIX number, Richardson
number, aspect (height to diameter) ratio, discharging
efficiency ratio, Peclet number, and Reynolds number,
3.3 Performance measures based on energy analysis were assessed in order to make qualitative statements
on their suitability for assessing stratification in STES
In order to quantify the extent to which the performance units. (The definitions of these dimensionless parameters
of a thermal storage device approaches the ideal, rational are presented in subsequent sections.) Stratification was
measures of performance are needed. These performance shown in that study to clearly increase with Richardson
measures are also needed for comparing both similar and number.
dissimilar thermal storage devices so as to make intelli- The other significant dimensionless parameter is the
gent choices. An ideal performance measure should be Reynolds number of inlet charging fluid expressed as
able to distinctly relate the performance of the system
being assessed to that of an ideal, i.e., it should show Uin din
Rein = (2)
how closely the performance of the system approaches ν
or departs from that of an ideal system. Since the extent Rein is the ratio of inertia to viscous forces in the flow-
of stratification in a storage tank changes in the course ing fluid and gives an indication of the presence or ab-
of its operation, such a measure should also be able to sence of turbulence in the flow. The influence of Reynolds
capture the time-dependent nature of the performance number on internal conditions, via convective mixing in
of the systems being assessed. The parameters suggested storage tanks, was highlighted in Van Berkel et al. [4].
for measuring the performance of stratified TES systems The conclusion of the Castell et al. study, however, was
can be grouped into a) dimensionless numbers, b) effi- that it was unsuitable for characterizing stratification
ciency ratios, and c) miscellaneous figures of merit. Mea- in storage tanks. High Rein charging in storage tanks
sures based on energy (first law) considerations are dis- will promote mixing and thus a deterioration of strati-
cussed in the following sections, while those based on fication. But because the influence of turbulence is not
exergy and entropy generation (second law) considera- isolated to the thermocline, and may not extend beyond
tions are discussed later on. it for high values of Ri, Rein values alone may not be
sufficient to characterize the extent of stratification in a
3.3.1 Dimensionless numbers Some dimensionlesss pa- storage tank.
rameters that are significant in the operation and per-
formance of stratified sensible heat storages have been 3.3.2 Efficiency ratios The dimensionless numbers men-
identified and their use as measures for performance as- tioned in the previous section are not direct measures of
sessment of stratified STESs have been attempted sev- performance or of the extent of stratification in a stor-
erally. age unit. Instead, they give indirect indications of the
6 H.O. Njoku et al.

The difficulty associated with directly measuring the

likelihood of stratification in a storage unit. Energy effi-
extent of stratification in a stratified STES unit was rec-
ciencies also fall into this category of measures, and have
been used widely to assess stratified STES systems andognized by Hahne and Chen [53], and based on the as-
sumption that stratification will improve the efficiency of
are usually defined as the ratio of a useful energy output
from a storage tank to the total energy input. Variousstorage in a store, they resorted to the use of a charging
efficiency, ηc defined as
authors have come up with different expressions of these
energy efficiencies, differing in their approach towards R tc
defining the energy inputs and useful energy outputs of ṁin cp [ϑin − ϑo ]dt
ηc = 0 (8)
the systems considered. The different possible expres- mcp (ϑin − ϑini )
sions have been summarized into four by Rosen [9] and Rt
where 0 c ṁin cp [ϑin − ϑ0 ]dt is the net energy stored at
Dincer et al. [52] based on the following energy balance
on a storage unit: the end of charging period, tc , while mcp (ϑin − ϑini )
is the thermodynamic maximum energy that may be
Energy Accumulation =Energy Input − [Energy Recovered stored. ṁin is charging mass flow rate, cp , the constant
pressure specific heat, ϑin and ϑo , temperatures at the
+Energy Loss]
inlet and outlet of the store, respectively, ϑini , the initial
(3)
temperature of the store and m, the mass of storage fluid
in the tank. Efficiency expressions similar to Equation 8
which is illustrated in Figure 2. have also been used [53] by Abdoly and Rapp [31], Yoo
and Pak [54] and Wildin and Truman [55].
An energetic efficiency, η was used by Solé et al. [56]
to assess stratified thermal storage tanks containing wa-
ter and water/PCM as storage media. The energetic ef-
ficiency was defined as
Ptend
∆Etank
η = Ptendtini (9)
tini m(hin − hout )

where the numerator represents the energy stored in the

tank, while the denominator is the energy supplied dur-
ing the whole charging period. η is equivalent to ηB in
equation 5 with the ‘energy recovered” set equal to zero.
Jannatabadi and Taherian [57] used a discharging effi-
ciency, η given as
Eout
η= (10)
Est
where Eout and Est are the energy withdrawn from the
Fig. 2 Energy balance on a thermal energy storage unit.
storage tank and the energy content of the storage tank
at the start of discharging, respectively. Recognizing that
The possible definitions for energy efficiencies, de- there is no energy input in a discharge process, this effi-
noted by ηA , ηB , ηC and ηD , are as follows: ciency is the same as ηC in equation 6, with the ‘Energy
input’ set equal to zero.
Energy recovered An extraction efficiency had earlier been used by La-
ηA = (4) van and Thompson [6], defined as the ratio
Energy input
Energy recovered + Energy accumulation η = Q̇t∗ /V (11)
ηB = (5)
Energy input
Energy recovered where Q̇t∗ is the volume of charging fluid into the tank
ηC = (6) that is needed to reduce the inlet-exit temperature dif-
Energy input + Initial energy in the store
Energy recovered + Final energy in the store ference to a pre-assigned value (taken to be 10 per cent
ηD = (7) drop of the initial value) and V the total volume of the
Energy input + Initial energy in the store
tank. A close correlation of the extraction efficiency, η
with Reynolds number, Red , Grashof number, GrD , and
The appropriateness of any of these efficiency definitions
aspect ratio, L/D, was obtained as
depends on the case under consideration and the aspect
of performance being assessed [9], i.e., whether the pro- η = 1 − exp −0.067Red−0.55 GrD 0.35
(L/D)0.58

(12)
cess of interest is the charging, discharging or storage
process. as shown in figure 3.
Assessment of stratified thermal storage systems using energy and exergy methods: A review 7

storage tanks. P CR was defined as:

P CR = Ht /H0 (14)
where Ht is the cooling capacity available at any time t
while H0 is the cooling capacity available initially at time
0. P CR is equivalent to ηD in equation 21, with both
‘energy recovered’ and ‘energy input’ set equal to zero.
P CR was found to improve appreciably with aspect ratio
(H/D) up to a limit of H/D = 3. This corresponds with
the result of Lavan and Thompson [6], that optimum
H/D would lie between 3 and 4. P CR also increased as
the flow rates and initial temperature difference between
inlet and outlet increased.
The thermocline thickness, h has also been used as
a measure of stratification in storage tanks [49, 61]. It is
defined as the thickness of the region in the storage tank
for which 0.9 ≤ Θ ≤ 0.1, where
Fig. 3 Correlation of extraction efficiency with Red , Grd
Θ = (T − Tc )/(Th − Tc ) (15)
and L/D (Ref. [6].)
and Tc and Th are the cold inlet and hot outlet temper-
atures, respectively. Cónsul et al. [49] used the dimen-
3.3.3 Miscellaneous measures of stratification The strat- sionless thermocline thickness h∗ (= h/H) to assess the
ification number, Str was proposed by Fernandez-Seara level of stratification in a horizontal storage tank and
et al. [58, 59] as a suitable dimensionless parameter for concluded that because of the multi-dimensional nature
directly quantifying the extent of stratification in strati- of temperature stratification in the horizontal tank stud-
fied stores. An extensive experimental study was under- ied, it was unsuitable for measuring stratification.
taken aimed at accessing the performance of a stratified It was shown by Rosen [11] that different tempera-
domestic hot water storage tank in the static and dy- ture profiles (and thus different states of stratification)
namic modes of operation. Str was defined as the ratio could exist in different heat storage tanks of identical ge-
of the mean of the temperature gradients at each time ometry, that possessed equal amounts of energy. Energy-
interval to that at the beginning (t = 0) (for an initially based stratification measures are generally incapable of
stratified store) and is stated mathematically as satisfactorily differentiating between different states of
stratification in thermal storage tanks since the same
(∂T /∂z)t results may be obtained when they are used to assess
Str = (13) similar storage tanks containing storage fluids at differ-
(∂T /∂z)t=0
ent states of stratification. The Richardson number, for
example, compares the contending forces that aid or de-
Str was shown to decrease as flow rates through the
stroy stratification (bouyancy and inertia, respectively,)
storage tank increased. Limitations in the use of this
in a storage tank. However, it will be unable to differ-
number for assessing performance are immediately ev-
entiate between the different temperature distributions
ident: firstly, it becomes indeterminate for an initially
that will be present in the tank at different time instants
fully mixed storage tank and secondly it will always be
during the the operation of the tank if operating under
equal to zero when a tank is stratified symmetrically
a fixed temperature difference and inlet mass flow rate.
about its mid-height.
Davidson et al. [24, 60] developed the MIX number
which similarly described the overall behaviour of the 4 Exergy analysis of stratified TES systems
stratified storage tank rather than local behaviour in
the thermocline or entrance effects. The MIX number When the assessment of TESs is on the basis of both the
characterizes the extent of mixing in solar water stor- first and second laws of thermodynamics, exergy flows
age tanks based on the weighted energy or moment of in the TESs are investigated. Since exergy quantifies the
energy, and ranges from 0 to 1 where zero represents a amount of work that can be extracted from a system
perfectly unmixed tank. as it is brought to equilibrium with its environment re-
The percentage cold recoverable, P CR was used to versibly, exergy analysis uses as benchmark, the actual
quantify the state of stratification in cold water storage theoretical limits of performance of the system. Unlike
tanks by Nelson et al. [19, 29] in their extensive experi- an energy-based analysis which compares either the en-
mental study of factors (aspect ratios, flow rates, insu- ergy stored or the energy output with the energy input,
lation and inlet mixing) that affect stratification in cold exergy analysis recognizes, on the one hand, the fact that
8 H.O. Njoku et al.

not all the input energy is recoverable, and that not all ‘irreversibilities’) and exergy losses which accompanied
the recoverable energy will be useful – that the higher the energy transfers across the system boundaries to the
the temperature at which the energy may be recovered, environment. These are not discernable from an energy
the higher it usefulness (or quality) [12]. Exergy analy- analysis [64]. An exergy analysis, in addition to an en-
sis therefore gives more insight into the actual level of ergy analysis, is therefore needed in order to correctly
performance of a TES system. assess the performance of a stratified TES system [11].
Further arguments in favour of exergy analysis have been
well outlined in Dincer and Rosen [1].

4.1 Analysis of exergy flows in stratified TES systems

Based on an exergy analysis, Aghbalou et al. [65] op-

timized a solar energy installation utilizing a hot wa-
ter storage tank with embedded PCM slabs. They found
the optimum PCM melting temperature, Topt , PCM slab
height, and PCM slab thickness for the system by an-
alytical solution of a 1-dimensional model of the stor-
age tank coupled to the equations for the solar collector.
The optimum PCM melting temperature obtained from
the exergy analysis was identical with the one previously
gotten for solely PCM-based storage systems, viz. [66, 67,
68]:
Topt = (Th T∞ )0.5 (16)
where Th and T∞ are the heat source temperature and
ambient temperature, respectively.
An exergy-based analysis was performed by Jack and
Wrobel [69] in order to determine optimum charging
times for stratified storage tanks. Two idealized limiting
cases representing extremes in the operation of strati-
Fig. 4 A realistic (actual) vertically stratified temperature fied tanks were defined: the fully mixed and the per-
distribution, and some temperature-distribution models used
fectly stratified cases, and by varying a heat diffusion
to approximate it (linear, continuous-linear, general-linear,
coefficient, ξ and a heat conduction coefficient, κ, the
stepped with 2 zones, and stepped with 20 zones). The
continuous-linear distribution shown is equivalent to a gen- effects of mixing and conduction exergy losses on opti-
eral three-zone distribution. (Ref. [11].) mum charging times were studied. The presence of strat-
ification was shown to lead to increases in the second
law efficiency and to shorter optimum charging times.
Six models for describing the variation of tempera- An exergy-based parametric study was conducted by
ture with height in a stratified hot water storage tank Farmahini-Farahani [70], on the effect of tank aspect ra-
were developed in the studies of Rosen [11, 62] and Rosen tios (AR), inlet/outlet diameters and positions and inlet
and Hamzeh [63], viz.: the linear, stepped, continuous- port inclination on thermal stratification in cold stor-
linear, general-linear, basic three-zone, and the general age tanks. Commercial CFD packages were used to cre-
three-zone temperature distribution models, which are ate the mesh for the computational domain and solve
shown in figure 4 [11]. Based on these models, the quanti- the unsteady 2-D conservation equations, respectively.
ties of thermal energy and exergy that could be stored in Stratification was shown to increase by increasing the
such tank were computed and compared with the quanti- tank AR, reducing the inlet/outlet diameters, and mov-
ties of thermal energy and exergy that could be stored in ing the inlet/outlet ports closer to the top/bottom of
a fully mixed storage tank. It was found that only an ex- the store. Increased stratification at low charging mass
ergy analysis could point out the improvement in stored flow rates was also reported, when the inlet/outlet ports
exergy (and thus the quality of energy stored) when heat were inclined away from the top/bottom of the tank. In
is stored in a stratified tank compared to storage in an the CFD study of Cónsul et al. mentioned earlier [49],
identical fully mixed tank. A purely energy-based com- thermal stratification in the storage tanks modeled were
putation showed no difference in the quantity of heat also assessed, using three parameters for quantifying the
that could be stored in either of both tanks. The re- level of stratification – the MIX number, the dimension-
sults of an exergy analysis also provide vital information less thermocline thickness, h∗t and the non-dimensional
on the exact nature and location of degradations in en- exergy, ξ ∗ . A comparison of the three, showed that the
ergy quality which result from ‘exergy destructions’ (or use of ξ ∗ resulted in the most rational assessments.
Assessment of stratified thermal storage systems using energy and exergy methods: A review 9

An exergy analysis of a hybrid (water/PCM) stor- denoted by A , B , C and D , are defined as follows:
age tank was carried out by Solé et al. [56]; the energy
Exergy recovered
and exergy efficiencies of the water storge tank, with A = (18)
Exergy input
and without PCM included, were gotten experimentally
Exergy recovered + Exergy accumulation
and compared. The energy efficiency of the tank was B = (19)
Exergy input
higher when it had no PCM inserted (98% as against
Exergy recovered
94% with PCM.) The exergy efficiency was however in- C = (20)
Exergy input + Initial exergy in the store
creased (from 85% to 89%) by the insertion of PCM
Exergy recovered + Final exergy in the store
capsules. The lower energy efficiency of the storage tank and D = (21)
Exergy input + Initial exergy in the store
with the PCM was compensated for by its higher ther-
mal storage capacity which was between 3–7% higher Based on these efficiency definitions, appropriate ef-
than that of the tank without the PCM. ficiency expressions for the charging, storage and dis-
charge phases of TES system operations were also pre-
sented. As mentioned before, the appropriate expression
to use depends on the situation under study and the
4.2 Measures of performance based on exergy analysis
aspect of operation being assessed.
The storage exergy efficiency, ηII (t) was used by Jack
A number of assessment methods incorporating second and Wrobel [69] in their optimization study of a stratified
law considerations have been developed for stratified ther- TES, and defined as
mal storage tanks [71]. Similar to the energy based per-
formance measures discussed earlier, entropy/exergy based Estored (t)
ηII (t) = R t (22)
measures may be separated into entropy/exergy efficiency 0
Ėnet (s) ds
measures and miscellaneous dimensionless measures.
where Estored (t) is the stored exergy in the storage at
Rt
time t, while 0 Ėnet (s) ds is the cumulative exergy in-
4.2.1 Entropy/Exergy Efficiency ratios The exergy bal- put into the storage up to that time. ηII (t) was also
ance on a typical energy storage unit may be expressed re-written in terms of the exergy loss rate as
by equation 17 which is illustrated in Figure 5. Rt
Ėloss (s) ds
1 − ηII (t) = R0t (23)
Exergy Accumulation = Exergy Input − [Exergy Recovered Ėnet (s) ds
0
+Exergy Loss] − Exergy Consumption Rt
(17) where 0 Ėloss (s) ds is the cumulative exergy loss from
the storage up to time t. Both ηII (t) and 1 − ηII (t) gave
performance values lying between 0 and 1, and showed
distinctly the effects of internal mixing and conduction
losses on the performance of the storage units. An iden-
tical storage exergy efficiency was also used by Solé et
al. [56].
The entropy and exergy efficiencies of Shah and Furbo
[24] were applied to the discharge phase of the STES
operation. The efficiencies were obtained by comparing
the discharge phase exergy (and entropy) changes in an
actual STES unit with the discharge phase exergy (and
entropy) changes in an ideal STES unit in which internal
mixing is absent. The efficiencies were defined as follows:
(s0 − s1 )ideal
ηs = 0 < ηs < 1 (24)
(s0 − s1 )actual
ξ1−0,actual
ηξ = 0 < ηξ < 1 (25)
ξ1−0,ideal
where s and ξ are the total entropy and the exergy con-
Fig. 5 Exergy balance on a thermal energy storage unit.
tents of the stores, respectively, subscript ‘1’ refers to
any state during discharge, while subscript ‘0’ refers to a
Based on the exergy balance of equation 17, Rosen state of total mixing in the store. The efficiencies showed
[9] and Dincer et al. [52] also summarized existing ex- a strong correlation with Richardson number Ri, as de-
ergy efficiencies into four corresponding to the energy sired, when used to analyse the results of experiments
efficiencies of equations 4 - 7. These exergy efficiencies, on a stratified storage tank with different inlet designs.
10 H.O. Njoku et al.

The stratification efficiencies of Haller et al. [72] also will lead to zero and 100% efficiencies are not imme-
compared the second law performance of a hot water diately obvious from the SEN definitions—even though
storage tank with that of a fully mixed storage tank these will be hypothetical situations, they should be dis-
of similar configuration. Two efficiencies were also pre- cernible from an efficiency definition.
sented, one based on entropy generation, ηst,S and the A non-dimensional exergy, ξ ∗ was used by Cónsul et
other on exergy ηst,ξ , and expressed as follows: al. [49] and defined as
exp
∆Sirr,int Ξst − Ξ Ξ − Ξmix
ηst,S = 1 − (26) ξ∗ = ≡1− (30)
mix,hl
∆Sirr,int Ξst − Ξmix Ξst − Ξmix

exp
where Ξ, Ξst and Ξmix are the exergies in the tank
∆ξirr,int under consideration, a perfectly stratified tank, and a
ηst,ξ = 1 − (27)
mix,hl
∆ξirr,int fully mixed tank, respectively. ξ ∗ returns a value of 1
exp exp
for a fully mixed tank and 0 for a perfectly stratified
where ∆Sirr,int /∆ξirr,int are the entropy generation/exergy tank. As was mentioned earlier, in comparison with two
mix,hl
losses in the stratified tank under study, while ∆Sirr,int / other parameters used in their study, ξ ∗ gave more dis-
mix,hl
∆ξirr,int are the entropy generation/exergy losses in a tinct results. The non-dimensional exergy was also used
fully mixed tank. Two extremes in the operation of a by Farmahini-Farahani [70] for measuring the state of
thermal storage tank were identified in their derivation. stratification in the thermal storage tanks he studied.
The first, when the tank is fully mixed, for which case In addition to these, Rosen et al. [74] used the ratio of
exp mix,hl exp mix,hl
∆Sirr,int = ∆Sirr,int and ∆ξirr,int = ∆ξirr,int , and the exergy of a stratified storage, Ξ, to the exergy of
thus, ηst,S = ηst,ξ = 0. And the second, when it is per- the same storage when fully mixed, Ξm , i.e. Ξ/Ξm , to
exp exp
fectly stratified, for which case ∆Sirr,int = ∆ξirr,int = 0 study the effect of thermal stratification on the energy
and thus ηst,S = ηst,ξ = 1. This stratification efficiency and exergy capacities of thermal storage systems.
therefore places a thermal storage unit’s performance on
a scale which ranges from 0 to 1, giving an idea of how 4.2.3 Entropy generation numbers, Ns It may be ob-
close its operation is to either of the two extremes. served that the efficiencies of Shah and Furbo [24] (equa-
tions 24 and 25) and Haller et al. [72] (equations 26 and
4.2.2 Miscellaneous dimensionless ratios In an attempt 27) do not correspond with any of the conventional effi-
to develop an assessment criteria for assessing the oper- ciencies defined by equations 18 - 21. Instead of compar-
ation of stratified thermal storage tanks that is not lim- ing inputs with outputs, these efficiencies compare the
ited to considerations of only one of either the first or entropy generation/exergy in the storage units being as-
second law of thermodynamics, but which incorporates sessed with the entropy generation/exergy that will exist
both, Panthalookaran et al. [73] came up with two char- in ideal units. Such comparisons of the entropy genera-
acterization numbers, the stratification evaluation num- tion rate in an actual system, Sgen,actual with the en-
bers (SEN), SEN,1 for the charging/discharge stages and tropy generation rate in an ideal system (usually the
SEN,2 for the storage stage of stratified tanks’ operation. maximum entropy generation rate, Sgen,max ) have been
The SEN efficiencies, ηSEN,1 and ηSEN,2 for the charg- called the entropy generation number, Ns . The entropy
ing/discharge and the storage stages, respectively were generation number is expressed as either
expressed as follows, Sgen,actual Sgen,max
Ns = or Ns = (31)
S S
 
REG gen,max gen,actual
ηSEN,1 = 1 − 100% (28)
ER and has been used extensively in the literature for the
assessment of thermal systems (see e.g. ref [10] for ex-
ηSEN,2 = 1 − [REG ER ] 100% (29)
tensive applications of Ns ). The use of Ns is at the heart
where REG is the ratio of entropy generation during pro- of the entropy generation minimization method, which
cesses in the real storage tank to that during processes was popularized by Bejan [10, 75, 76, etc]. In the perfor-
in a perfectly stratified tank, while ER is the energy re- mance analysis of stratified TESS, Badescu [77] used the
sponse factor, a ratio of the energy changes in the tank entropy generation number to compare and optimize the
during a real process to energy changes during an isen- performances of stratified and fully mixed storage tanks.
tropic process. The effect of changes in the aspect ratio, The influence of inlet temperatures and charging mass
the shape and size of the storage tank and the pres- flow rates on performance were clearly indicated by Ns
ence of structures in the tanks on these efficiencies were and optimum charging times based on Ns were obtained.
studied using a CFD-based analysis. The need to switch Consider three tanks of identical geometry, charged
between the two different expressions for the different (or discharged) at the same inlet (or outlet) mass flow
modes of operation of the storage tank makes the use of rate and temperature. The first operates under real life
this method complicated. Secondly, the conditions which conditions, but the second and the third are idealized
Assessment of stratified thermal storage systems using energy and exergy methods: A review 11

Table 1 Equivalence of some entropy/exergy assessment parameters.

Parameter (Ref ) Equivalent entropy generation ratio
Ξst − Ξ Sgen,actual
ξ∗ = [49, 70] ξ ∗ = Ns =
Ξst − Ξmix Sgen,max
(s0 − s1 )ideal Sgen,max
ηs = [24] ηs = Ns =
(s0 − s1 )actual Sgen,actual
exp
∆Sirr,int Sgen,actual
ηst,S = 1 − mix,hl
[72] ηst,S = 1 − Ns = 1 −
∆Sirr,int Sgen,max
exp
∆ξirr,int Sgen,actual
ηst,ξ = 1 − mix,hl [72] ηst,ξ = 1 − Ns = 1 −
∆ξirr,int Sgen,max

as operating fully mixed and perfectly stratified, respec- an expression of the entropy generation number defined
tively. In the fully mixed tank, incoming fluid instan- by Sgen,actual /Sgen,max . The stratification efficiencies of
taneously mixes with the fluid resident in the tank to Haller et al. [72], ηst,S and ηst,ξ , are equavalent to 1−Ns ,
attain a uniform temperature in the tank, while in the for Ns = Sgen,actual /Sgen,max . However, the entropy effi-
perfectly stratified tank, it is assumed that absolutely ciency, ηs of Shah and Furbo [24] corresponds to the en-
no mixing or heat transfer occurs between the incom- tropy generation number defined as Sgen,max /Sgen,actual .
ing fluid and the resident fluid such that the fluid in the These parameters have been identified as rational per-
tank is divided by an infinitesimal thermocline into two formance measures for STESS, capable of providing dis-
segment at two uniform temperatures. At any instant in tinct quantifications of the time-dependent performance
the operation of the tanks, because the perfectly strat- of these systems under different operating conditions.
ified tank operates isentropically, the difference in the Following from this, further assessment studies based on
exergies contained in the actual tank, Ξa and the per- the entropy generation number can be recommended to
fectly stratified tank, Ξst will be equal to the entropy improve on current knowledge of the conditions neces-
generation in the actual tank, i.e., sary for the optimum performance of STESS.

Ξst − Ξa = Sgen,actual (32)

5 Summary
Furthermore, under these identical charging (discharg-
ing) conditions, the entropy generation in the fully mixed
In this review, the methods of analyzing the operation
tank is the maximum possible, hence, the difference in
and performance of stratified thermal storage units have
the exergies contained in the fully mixed tank, Ξmix and
been surveyed. The presentation of the factors that in-
the perfectly stratified tank is equal to the maximum en-
fluence stratification in storage units and the means of
tropy generation, i.e.,
improving and sustaining stratification shows that the
Ξst − Ξmix = Sgen,max (33) presence of obstacles in storage tanks generally enhances
stratification. Numerous mathematical models of STESS
Again, since the entropy generated in the perfectly strat- have been implemented. Their varying complexities are
ified tank is zero, the difference in the entropies gener- determined by a compromise between simplicity and ease
ated in the actual tank’s operation, Sa and in the per- of solution, on one hand, and a detailed accounting of
fectly stratified tank’s operation, Sst is equal to the en- the underlying forces driving heat and fluid flow, on the
tropy generation in the actual tank, i.e., other. Current computational advances are resulting in
a more widespread use of CFD analysis. The application
Sa − Sst = Sgen,actual (34) of artificial neural network analysis was also highlighted.
The diverse measures proposed for assessing strati-
Also, the difference in the entropies generated in the op-
fied storage systems were grouped broadly into energy
eration of the fully mixed tank, Smix and the perfectly
and exergy/entropy analysis methods, and their inher-
stratified tank is equal to the maximum entropy gener-
ent advantages and shortcomings discussed. Definitive
ation possible, i.e.,
evaluations of the performance of STESS cannot be ob-
Smix − Sst = Sgen,max (35) tained with energy-based measures because they are un-
able, on the basis of the state of stratification present
We see at once, by comparing these expressions with in a tank, to distinguish between the energy contents of
the definitions of some of the exergy/entropy based pa- the storage tank at different instants. The same also ap-
rameters discussed earlier, that a number of them are plies to dimensionless numbers, such as the Reynolds and
in fact variants of Ns . As shown by their equivalent Richardson numbers. On the other hand, exergy/entropy
Ns expressions in Table 1, the non-dimensional exergy, analysis methods were shown to provide clearly distinct
ξ ∗ of Cónsul et al. [49] and Farmahini-Farahani [70] is patterns in the performance of stratified thermal storage
12 H.O. Njoku et al.

tanks. This is especially so with the different variants of 17. J. Van Berkel. Mixing in thermally stratified energy
the entropy generation number, Ns . They are capable of stores. Sol. Energy, 58(4–6):203–211, 1996.
giving clear assessments of the time-dependent perfor- 18. M.W. Wildin. Diffuser design for naturally stratified
mance of stratified thermal storage systems as opposed thermal storage. ASHRAE Trans., 96(Part 1B):1094–
to the energy based measures, leading to the possibility 1102, 1990.
19. J.E.B. Nelson, A.R. Balakrishnan, and S. Srinivasa
of optimizing the duration of charge, storage and dis-
Murthy. Experiments on stratified chilled-water tanks.
charge operations. These qualities suggest the entropy Int. J. Refrig., 22:216–234, 1999.
generation numbers as basis of further studies on strat- 20. C.K. Yee and F.C. Lai. Effects of a porous manifold
ified thermal storage systems. Such studies may include on thermal stratification in a liquid storage tank. Sol.
CFD analysis, and optimization studies of geometry, flow Energy, 71(4):241–254, 2001.
rate, insulation and tank material property. 21. E. Andersen, S. Furbo, and J. Fan. Multilayer fabric
stratification pipes for solar tanks. Sol. Energy, 81:1219–
1226, 2007.
References 22. L.J. Shah, E. Andersen, and S. Furbo. Theoretical and
experimental investigations of inlet stratifiers for solar
1. Ibrahim Dincer and Marc A. Rosen. Exergy: energy, en- storage tanks. Appl. Therm. Eng., 25:2086–2099, 2005.
vironment and sustainable development. Elsevier Science, 23. A. Zachar, I. Farkas, and F. Szlivka. Numerical analyses
Oxford, UK, 2007. of the impact of plates for thermal stratification inside
2. Ibrahim Dincer and Marc A. Rosen. Thermal energy
a storage tank with upper and lower inlet flows. Sol.
storage: systems and applications. John Wiley & Sons,
Energy, 74:287–302, 2003.
Ltd, West Sussex, UK, 2nd edition, 2011.
24. L.J. Shah and S. Furbo. Entrance effects in solar storage
3. J.A. Duffie and W.A. Beckman. Solar Engineering of
tanks. Sol. Energy, 75:337–348, 2003.
Thermal Processes. John Wiley and Sons, New York,
25. U. Jordan and S Furbo. Thermal stratification in small
2nd edition, 1980.
solar domestic storage tanks caused by draw-offs. Sol.
4. J. Van Berkel, C.C.M Rindt, and A.A. Van Steenhoven.
Energy, 78:291–300, 2005.
Modelling of two-layer stratified stores. Sol. Energy,
26. Mi-Soo Shin, Hey-Suk Kim, Dong-Soon Jang, Sang-Nam
67(1–3):65–78, 1999.
Lee, Young-Soo Lee, and Hyung-Gi Yoon. Numerical and
5. Y.M. Han, R.Z. Wang, and Y.J. Dai. Thermal strati-
experimental study on the design of a stratified thermal
fication within the water tank. Renewable Sustainable
storage system. Appl. Therm. Eng., 24:17–27, 2004.
Energy Rev., 13:1014–1026, 2009.
6. Z. Lavan and J. Thompson. Experimental study of ther- 27. Z.F. Li and K. Sumathy. Experimental studies on a so-
mally stratified hot water storage tanks. Sol. Energy, lar powered air conditioning system with partitioned hot
19:519–524, 1977. water storage tank. Sol. Energy, 71(5):285–297, 2001.
7. C. Cristofari, G. Notton, P. Poggi, and A. Louche. Influ- 28. J. Rhee, A. Campbell, A. Mariadass, and B. Morhous.
ence of the flow rate and the tank stratification degree Temperature stratification from thermal diodes in solar
on the performances of a solar flat-plate collector. Int. hot water storage tank. Sol. Energy, 84:507–511, 2010.
J. Therm. Sci., 42:455–469, 2003. 29. J.E.B. Nelson, A.R. Balakrishnan, and S. Srinivasa
8. Z.F. Li and K. Sumathy. Performance study of a parti- Murthy. Parametric studies on thermally stratified
tioned thermally stratified storage tank in a solar pow- chilled water storage systems. Appl. Therm. Eng., 19:89–
ered absorption air conditioning system. Appl. Therm. 115, 1999.
Eng., 22:1207–1216, 2002. 30. S.S. Murthy, J.E.B. Nelson, and T.L.S. Rao. Effect of
9. M.A. Rosen. Appropriate thermodynamic performance wall conductivity on thermal stratification. Sol. Energy,
measures for closed systems for thermal energy storage. 49(4):273–277, 1992.
J. Sol. Energy Eng., 114:100–105, 1992. 31. M.A. Abdoly and D. Rapp. Theoretical and experimen-
10. Adrian Bejan. Entropy Generation Minimization. CRC tal studies of stratified thermocline storage of hot water.
Press, Boca Raton, Florida, 1996. Energy Convers. Manage., 22:275–285, 1982.
11. Marc A. Rosen. The exergy of stratified thermal energy 32. A. Barba and M. Spiga. Discharge mode for encapsulated
storages. Sol. Energy, 71(3):173–185, 2001. PCMs in storage tanks. Sol. Energy, 74:141–148, 2003.
12. M.A. Rosen and I. Dincer. Exergy methods for assessing 33. Luiza F. Cabeza, Manuel Ibanez, Cristian Solé, Joan
and comparing thermal storage systems. Int. J. Energy Roca, and Miquel Nogues. Experimentation with a wa-
Res., 27:415–430, 2003. ter tank including a PCM module. Sol. Energy Mater.
13. A. Cabelli. Storage tanks a numerical experiment. Sol. Sol. Cells, 90:1273–1282, 2006.
Energy, 19:45–54, 1977. 34. Muhsin Mazman, Luisa F. Cabeza, Harald Mehling,
14. K.G.T. Hollands and M.F. Lightstone. A review of low- Miquel Nogues, Hunay Evliya, and Halime O. Paksoy.
flow, stratified-tank solar water heating systems. Sol. Utilization of phase change materials in solar domestic
Energy, 43:97–105, 1989. hot water systems. Renewable Energy, 34:1639–43, 2009.
15. Shahab Alizadeh. An experimental and numerical study 35. H. Mehling, L.F. Cabeza, S. Hippeli, and S. Hiebler.
of thermal stratification in a horizontal cylindrical solar PCM-module to improve hot water heat stores with
storage tank. Sol. Energy, 66(6):409–421, 1999. stratification. Renewable Energy, 28:699–711, 2003.
16. Necdet Altuntop, Mevlut Arslan, Veysel Ozceyhan, and 36. Manuel Ibanez, Luisa F. Cabeza, Cristian Sole, Joan
Mehmet Kanoglu. Effect of obstacles on thermal strati- Roca, and Miquel Nogues. Modelization of a water tank
fication in hot water storage tanks. Appl. Therm. Eng., including a PCM module. Appl. Therm. Eng., 26:1328–
25:2285–2298, 2005. 1333, 2006.
Assessment of stratified thermal storage systems using energy and exergy methods: A review 13

37. Hoseon Yoo and Ee-Tong Pak. Analytical solutions to a 55. M.W. Wildin and C.R. Truman. Performance of strat-
one-dimensional finite-domain model for stratified ther- ified vertical cylindrical thermal storage tanks. Part I:
mal storage tanks. Sol. Energy, 56(4):315–322, 1996. Scale model tank. ASHRAE Trans., 89(2):1086–1095,
38. Hoseon Yoo, Charn-Jung Kim, and Chang Wook Kim. 1989.
Approximate analytical solutions for stratified therml 56. C. Solé, M. Medrano, M. Castell, A.and Nogués,
storage under variable inlet temperature. Sol. Energy, H. Mehling, and L. F. Cabeza. Energetic and exergetic
66(1):47–56, 1999. analysis of a domestic water tank with phase change ma-
39. K. Abu-Abdou and T. Hussein. Temperature distribu- terial. Int. J. Energy Res., 32:204–214, 2008.
tion in a storage tank. Wärme- und Stoffübertragung, 57. M. Jannatabadi and H. Taherian. An experimental study
15:271–277, 1981. of influence of hot water consumption rate on the ther-
40. N.M. Al-Najem and M.M. El-Refaee. A numerical study mal stratification inside a horizontal mantle storage tank.
for the prediction of turbulent mixing factor in thermal Heat Mass Transfer, 48:1103–1112, 2012.
storage tanks. Appl. Therm. Eng., 17(12):1173–1181, 58. Jose Fernandez-Seara, Francisco J. Uhya, and Jaime
1997. Sieres. Experimental analysis of a domestic electric hot
41. N.K. Ghaddar and A.M. Al-Maarafie. Study of charging water storage tank. Part I: Static mode of operation.
of stratified storage tanks with finite wall thickness. Int. Appl. Therm. Eng., 27:129–136, 2007.
J. Energy Res., 21:411–427, 1997. 59. Jose Fernandez-Seara, Francisco J. Uhya, and Jaime
42. Kamal A.R. Ismail, Janaina F.B. Leal, and Mauricio A. Sieres. Experimental analysis of a domestic electric hot
Zanardi. Models of liquid storage tanks. Energy, water storage tank. Part II: Dynamic mode of operation.
22(8):805–815, 1997. Appl. Therm. Eng., 27:137–144, 2007.
43. M. Issa and M. Al-Nimr. Temperature distribution inside 60. J.H. Davidson, D.A. Adams, and J.A. Miller. A coeffi-
hot-water storage tanks of solar collectors. Wärme- und cient to characterize mixing in solar water storage tanks.
Stoffübertragung, 23:297–305, 1988. J. Sol. Energy Eng., 116:94–99, 1994.
44. H.K. Versteeg and W. Malalasekera. An introduction to 61. W.P. Bahnflet and A. Musser. Thermal performance of
computational fluid dynamics: The finite volume method. a full-scale stratified chilled water thermal storage tank.
Longman Group Ltd, 1995. ASHRAE Trans., 104(2):377–388, 1998.
62. M.A. Rosen. Application of temperature-distribution
45. S.C. Chang. The method of space-time conservation el-
models to evaluate the energy and exergy of stratified
ement and solution element — A new approach for solv-
thermal storages. Int. J. Exergy, 3(3):239–254, 2006.
ing the Navier-Stokes and Euler equations. J. Comput.
63. M.A. Rosen and H Hamzeh. Emperical comparison of
Phys., 119:295–324, 1995.
temperature-distribution models for energy and exergy
46. S.C. Chang, X.Y. Wang, and Chow C.Y. The space-
analyses of stratified thermal energy storages. Int. J.
time conservation element and solution element method:
Green Energy, 3(3):291–307, 2006.
A new high-resolution and genuinely multidimensional
64. H.O. Njoku and O.V. Ekechukwu. Analysis of the exer-
paradigm for solving conservation laws. J. Comput.
getic performance of the reverse-side absorber-plate shal-
Phys., 156:89–136, 1999.
low solar pond. Int. J. of Sust. Energy, 30(6):336–352,
47. S. C. Chang and W. M. To. A New Numerical
2011.
Framework for Solving Conservation LawsThe Method
65. F. Aghbalou, F. Badia, and J. Illa. Exergetic optimiza-
of Space-Time Conservation Element and Solution El-
tion of solar collector and thermal energy storage system.
ement. NASA TM 104495, August 1991.
Int. J. Heat Mass Transfer, 49:1255–1263, 2006.
48. J. Tu, G.H. Yeoh, and C. Liu. Computational Fluid Dy- 66. H. Bjurström and B. Carlsson. An exergy analysis of
namics: A Practical Approach. Butterworth-Heinemann, sensible and latent heat storage. Heat Recovery Syst.
2007. CHP, 5:233–250, 1985.
49. R. Consul, I. Rodryguez, C.D. Perez-Segarra, and M. So- 67. G.A. Abebiyi and L.D. Russell. A second law analysis of
ria. Virtual prototyping of storage tanks by means of phase-change energy storage systems. HTD (Am. Soc.
three-dimensional CFD and heat transfer numerical sim- Mech. Eng.), 80:9–20, 1987.
ulations. Sol. Energy, 77:179–191, 2004. 68. M. De Lucia and A. Bejan. Thermodynamics of energy
50. P. Gećzy-Vı́g and I. Farkas. Neural network modelling storage by melting due to conduction or natural convec-
of thermal stratification in a solar DHW storage. Sol. tion. J. Sol. Energy Eng., 112:110–116, 1990.
Energy, 84:801–806, 2010. 69. Michael W. Jack and Jan Wrobel. Thermodynamic op-
51. A. Castell, M. Medrano, C. Sol, and L.F. Cabeza. Di- timization of a stratified thermal storage device. Appl.
mensionless numbers used to characterize stratification Therm. Eng., 29:2344–2349, 2009.
in water tanks for discharging at low flow rates. Renew- 70. Moien Farmahini-Farahani. Investigation of four geomet-
able Energy, 35:2192 – 2199, 2010. rical parameters on thermal stratification of cold water
52. I. Dincer, S. Dost, and X. Li. Performance analysis of tanks by exergy analysis. Int. J. Exergy, 10(3):332–345,
sensible heat storage systems for thermal applications. 2012.
Appl. Therm. Eng., 21:1157–1171, 1997. 71. M.Y. Haller, C.A. Cruickshank, W. Streicher, S.J. Har-
53. E. Hahne and Y. Chen. Numerical study of flow and heat rison, E. Andersen, and S. Furbo. Methods to determine
transfer characteristics in hot water stores. Sol. Energy, stratification efficiency of thermal energy storage pro-
64(1-3):9–18, 1998. cesses review and theoretical comparison. Sol. Energy,
54. H. Yoo and E. Pak. Theoretical model of the charging 83:1847–1860, 2009.
processs for stratified thermal storage tank. Sol. Energy, 72. M.Y. Haller, E. Yazdanshenas, E. Andersen, C. Bales,
51(6):513–519, 1993. W. Streicher, and S. Furbo. A method to determine
14 H.O. Njoku et al.

stratification efficiency of thermal energy storage pro-

cesses independently from storage heat losses. Sol. En-
ergy, 84:997 – 1007, 2010.
73. V. Panthalookaran, W. Heidemann, and H. Muller-
Steinhagen. A new method of characterization for strat-
ified thermal energy stores. Sol. Energy, 81:1043 – 1054,
2007.
74. Marc A. Rosen, Raymond Tang, and Ibrahim Dincer.
Effect of stratification on energy and exergy capacities
in thermal storage systems. Int. J. Energy Res., 28:177–
193, 2004.
75. Adrian Bejan. Entropy Generation through Heat and
Fluid Flow. Wiley, New York, 1982.
76. Adrian Bejan. Method of entropy generation minimiza-
tion, or modeling and optimization based on combined
heat transfer and thermodynamics. Rev. Gén. Therm.,
35:637–646, 1996.
77. Viorel Badescu. Optimal operation of thermal energy
storage units based on stratified and fully mixed water
tanks. Appl. Therm. Eng., 24:2101–2116, 2004.

View publication stats