- Email: [email protected]
Theoretical and experimental investigation of the performance of adsorption heat storage system
Applied Thermal Engineering 147 (2019) 10–28
Contents lists available at ScienceDirect
Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng
Research Paper
Theoretical and experimental investigation of the performance of adsorption heat storage system
T
⁎
Hesham O. Helalya, Mohamed M. Awadb, , Ibrahim I. El-Sharkawyb,c, Ahmed M. Hamedb a
Higher Institute of Engineering and Technology, New Damietta, Damietta, Egypt Mechanical Power Engineering Department, Faculty of Engineering, Mansoura University, Mansoura 35516, Egypt c Higher Future Institute of Engineering and Technology in Mansoura, Mansoura, Egypt b
H I GH L IG H T S
for simulating adsorption based energy storage processes is presented. • ACFDmodel simulation using COMSOL Multi-physics is carried out. • Experiments carried out to explore the storage system performance. • Two cases arearepresented to highlight the model validity. • A parametric study is performed to predict the thermal energy storage system behavior. •
A R T I C LE I N FO
A B S T R A C T
Keywords: Thermal energy storage Open adsorption COMSOL multi-physics Parametric study
Theoretical and experimental investigation of the thermal energy storage of an open adsorption system is presented. The theoretical model describes the mass and energy transfers in the system. The coupled energy and mass balance equations have been solved using the COMSOL™ software. The model was validated against laboratory experiments performed at varying conditions using AA13X as well as silica gel as adsorbent. Laboratory experiments have been conducted to study the effect of flow rate and inlet relative humidity on the amount of energy stored. Temperature and energy density profiles during the adsorption process have been obtained and analyzed for various conditions. Results showed that there is a trade-off between released energy and temperature output and an optimization is recommended before choosing the operating flow rate. It was found that a flow rate of 21 m3/hr has the best performance for a bed volume of 5. 09 × 10−4 m3 . Furthermore, the results show that the storage density increases with the increase of the air inlet relative humidity. For the predefined working conditions and assumptions, the numerical solution shows satisfied agreement with the experimental measurements. A parametric study was performed using silica gel as adsorbent to predict the behavior of the thermal energy storage system for varying operating conditions and parameters. The studied parameters include the system flow rate, relative humidity, regeneration temperature, and the particle diameter. It was found that a flow rate of 0.7 m3/min , regeneration temperature of 95 °C and average particle diameter of 1.0 mm gave the best performance for the bed with a volume of 1.57 × 10−3 m3 .
1. Introduction To face the challenge of energy crisis and pollution issues, it is necessary to develop and expand advanced technologies to reduce the energy demand, increase the energy supply while utilizing renewable energy effectively. One of the most promising developing technologies is energy storage, as it provides the benefit of capturing available energy for use at a later time. This study focuses on the technology of
thermal energy storage (TES), through adsorption. Thermal energy storage can be divided into sensible heat, latent heat and thermochemical or combination of these methods. A brief comparison between the different types of thermal energy storage systems is provided in Table 1. For the types of TES, sensible and latent TES were already developed for many years for commercial applications [1]. In recent years, sorption thermal energy storage systems are increasingly fast as they become promising options for thermal heat
⁎
Corresponding author. E-mail addresses: [email protected] (H.O. Helaly), [email protected] (M.M. Awad), [email protected] (I.I. El-Sharkawy), [email protected] (A.M. Hamed). https://doi.org/10.1016/j.applthermaleng.2018.10.059 Received 22 March 2018; Received in revised form 21 September 2018; Accepted 14 October 2018 Available online 16 October 2018 1359-4311/ © 2018 Elsevier Ltd. All rights reserved.
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Nomenclature
a a0 b b0 c Cg Cpg Cpp Di Dm Dp Ds0 Dz E Ea EEC f fo ΔHads hfd K keff Kg Kp L Ma Mw no Nu
Nuo Δp PH2 o Ppa q Q qe R R ep rc rp RH% Sc Sv t T Tg Tin To Tw Vg
calculated Toth parameter (mol/kg/kPa) fitted Toth parameter (mol/kg/kPa) fitted Toth parameter (1/kPa) fitted Toth parameter (1/kPa) fitted Toth parameter (K) water concentration in the gas phase (mol/m3) heat capacity of the gas stream (J/kg·K) heat capacity of the pellet (J/kg·K) inside diameter of the column (m) molecular diffusivity (m2/s) diameter of the pellet (m) pre-exponential constant (m2/s) axial dispersion (m2/s) fitted Toth parameter (K) activation energy (J/mol) efficiency evaluation criterion (–) resistance coefficient (–) resistance coefficient of the base case (–) heat of adsorption (J/mol) heat transfer coefficient in the bulk phase (W/m2·K) overall mass transfer coefficient (1/s) effective thermal conductivity of the solid-fluid system (W/m·K) thermal conductivity of the gas stream (W/m·K) thermal conductivity of the adsorbent matrix (W/m·K) column length (m) molar mass of air (g/mol) molar mass of water (g/mol) fitted Toth parameter (–) Nusselt number (–)
Greek symbols
∊c μg ρg ρp ρs (ρCp )eff
I. Charging (Regeneration): To store the thermal energy, a hot fluid passes through the bed to remove the adsorbate. This thermal energy can be provided by either waste heat sources or solar energy to avoid the depletion of natural resources and environmental pollution. II. Storing: The energy is permanently stored in the material for an indefinite period as long as it is sealed from the surroundings. Since the thermal energy is not sensibly stored, so theoretically, it is possible to store the material at ambient temperature without reducing the stored energy. III. Discharging: (Adsorption), Adsorption is an exothermic physical process where a gas diffuses into the pores of a porous solid material and is trapped into the crystal lattice which releases heat [6].
Table 1 Comparison of different types of TES based on various performance factors [30,31]. Sensible TES
Temperature range
Lifetime
Up to: 110 °C (Water tanks) 50 °C (Aquifers) 400°C (Concrete) Low 0.1–0.2 GJ/m3 Low
Technology status
Available commercially
Storage density
Latent TES
20 − 40 °C (Paraffins) 30 − 80 °C hydrates)
Thermochemical TES
So far, there is a brief background about open sorption thermal energy storage systems. In Munich, Hauer [7,8] reported a successful open adsorption system using 7000 kg zeolite 13X, which was installed in a school and connected to the local district heating network to offset the peak energy demands, performing at a storage density of 124 kWh/ m3. Within the “MonoSorp” project [9,10], the heat store is designed to be integrated into a building with a controlled ventilation system and with heat recovery. A major improvement was made by using zeolite honeycomb structures called “monoliths” instead of the ordinarily employed fills. These regularly shaped bodies have a large number of small, straight channels. Their main advantages being a low-pressure loss and better adsorption properties than fills [11]. In the “SolSpaces” project [12], special attention was given to the large heat capacities issue. The main idea is to divide the sorption store into several sectors,
20−200 °C (Salt
Moderate 0.2–0.5 GJ/m3 Often limited due to storage material cycling Available commercially for some temperature ranges and materials
void fraction of the column (m3void /m3column) gas stream viscosity (kg/m/s) density of the gas stream (kg/m3) bulk density of the pellet (kg/m3) solid density of the pellet (kg/m3) the effective volumetric heat capacity at constant (J/m3·K )
later use. In adsorption process, thermal energy can be stored through three steps:
storage [2]. Their advantages include relatively high storage densities and storing thermal energy for longer periods with limited heat loss as it can be stored at near ambient temperature. Thermochemical storage can be divided into chemical reaction and sorption [3]. “Sorption” is first proposed by McBain [4] as a general term which encompasses both absorption and adsorption. Adsorption systems can be divided into “closed” and “open” systems. The utilization of the adsorption-desorption cycle as a means of low-temperature energy storage was first proposed by Close [5] who had obtained warm and relatively dry air by using a silica gel adsorbent bed. Thermal energy storage allows storing energy to be used for heating or cooling for
Performance parameter
Nusselt number of the base case (–) pressure drop (Pa) water vapor partial pressure (Pa) column pressure (Pa) adsorbed water in the pellet (mol(H2o)/kg pellet) flow rate (m3/s) adsorbed amount in equilibrium (mol(H2O)/kg pellet) universal gas constant (J/mol·K) Reynolds number for the particles (–) inside radius of the column (m) pellet radius (m) relative humidity (%) Schmidt number (–) specific surface area (m2/kg) time (s) average temperature (K) temperature of the gas stream (K) inlet gas stream temperature (K) ambient air temperature (K) temperature of the wall (k) superficial velocity of the fluid (m/s)
Relatively high: 0.5–3 GJ/m3 Relatively high
Undergoing research
11
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
which can be adsorbed or desorbed separately. It is expected that over 90% of the heat demand for room heating and over 80% of the heat demand for residential hot water preparation can be covered by the solar thermal system combined with the sorption store. Hauer and Fischer [13] used zeolite 13X into a dishwasher to decrease the energy consumption by means of an open adsorption system. The reduction of the energy consumption compared to a traditional dishwasher from about 1.05 kWh to 0.80 kWh per washing cycle leads to energy savings of about 24%. To increase the storage density, Dicaire and Tezel [14,15] tested different adsorbents at varying flow rates and different values of relative humidity to determine the optimal operating conditions. A maximum energy density of 160 kWh/m3 has been achieved with the AA/13X as adsorbent. In addition, there are some studies that focused on the most essential operating conditions that affect the storage process. Dawoud et al. [16] tested zeolite 13X with water as the adsorbate, and obtained an energy density of 144 kWhr/m3 at a flow rate of 0.5 LPM and 165 kWhr/m3 at a flow rate of 2.0 LPM. The study concluded that higher flow rates result in a faster discharge of the storage unit and higher efficiencies. On the other hand, the lower the flow rate during the discharge phase the higher the temperature, at which the storage unit could be discharged. Li et al. [17] tested the mass flow rate effect on the storage performance of the adsorption bed. Results showed that as the mass flow rate increased, the energy storage density improved, but the exergy efficiency decreased due to the increased exergy destruction and loss caused by the pressure drop. They also varied the adsorption bed inlet temperature from 30 °C to 45 °C to test its effect on the storage performance. Results showed that as the adsorption temperature was increased, energy efficiency and loading difference decreased. Liu and Nagano [18] used CaCl2 as an adsorbent and analyzed the effect of air flow rate on the system’s performance. Results showed that more heat is released per time with a faster flow. Additionally, the discharge duration of the high temperature outlet air decreases when the air flow rate is high. They also tested the inlet relative humidity effect on the storage performance of the system. Results showed that the duration of the discharge temperature was extended when the relative humidity increases. Additionally, the equilibrium sorption amount increased with higher relative humidity according to the isothermal equilibrium sorption amount line. Although sorption thermal storage systems have some advantages, they seem to have more complexity than other storing methods such as; expensive investment, poor heat and mass transfer ability and low heat storage density in actual systems. To overcome these barriers, extensive academic efforts are now being carried out worldwide [19]. In the present study, it is aimed to investigate the adsorption process within a thermal energy storage unit using COMSOL Multi-physics software [20]. This simulation can be used to predict the temperature distribution of the thermal energy storage unit and the amount of energy stored. The simulated results will be validated against experimental temperature profiles for various operating conditions. A parametric study of the effect of operating and design parameters is objected.
Table 2 Simulated adsorbents characteristics. Type
Dso [m2/s]
Ea [J/mol]
Reference
Silica gel Type A Zeolite 13X
2.54E −4 5.45E −11
4480 100
[32] [33]
Table 3 Temperature-dependent Toth isotherm parameters for water vapour adsorption on different adsorbents [22]. Parameter
(Zeolite 13X)
(Silica gel)
a o (mol/kg Pa) bo (1/ Pa) n o (−) c (K) E (K)
3.63E-09 2.408E-10 0.3974 −4.199 6852
0.1767 2.787E-08 −0.00119 22.13 1093
Table 4 Input parameters used for simulation. Parameter (symbol)
Silica gel
Zeolite 13X
Bulk density (ρp)
750 kg/m3) ( 0.36 (–)
730 (kg/m3) 0.395 (–)
Porosity (εc) Specific heat capacity (Cpp)
1130
( ) J kg·K
Average particle radius (rp)
0.6∗10−3 (m)
Thermal conductivity (Kp)
0.173
Specific surface area (Sv )
5.3∗105 (m2/kg)
( ) W m·K
1080
( ) J kg·K
1.15∗10−3 (m)
( )
W m·K 5.5∗105 (m2/kg)
3.3
Table 5 Data inputs for “Transport of diluted species in porous media”. Interface input
Value
ux
Q/0.25 πD2i 0 Dz From heat transfer interface ∊c
uy Diffusion Temperature Porosity
2. Theoretical model The model, which describes an open flow system, is schematically presented in Fig. 1. During adsorption, moist air with a specified inlet moisture concentration c(g,in) and temperature Tin flows with a velocity vg through a porous cylinder (adsorbent) of initial moisture content qin, length L and radius rc. The moisture concentration in the adsorbent q. Assumptions made before modeling the process are mentioned below: 1. The airflow passing through the bed is one-dimensional and steady. 2. The transport processes within the adsorbent bed are transient
Fig. 1. Schematic representation of the model. 12
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Interface input
Value
(LDF), which was first proposed by Gleuckauf [21], expressed by the difference of the amount adsorbed, q , and the amount adsorbed in equilibrium qe .
ux
Q 0.25πDi2
dq = K (qe−q) dt
uy Porosity Fluid type Concentration Porous media density Porous media specific heat capacity Porous media thermal conductivity Reaction rate
0 ∊c Moist air From transport of diluted species interface ρp c pp
Table 6 Data inputs for “Heat transfer in porous media”.
(1)
where K is the overall mass transfer coefficient and is calculated using Eq. (2).
( ) E
K=
15∗Dso exp − RTa rp2
Kp
(2)
Interface input
Value
where rp is the adsorbent particle radius, Dso the pre-exponential constant, Ea is the activation energy and R refer to the universal gas constant. The values of activation energy Ea and the pre-exponential constant Dso are provided in Table 2. The equilibrium water concentration was calculated using the Temperature-Dependent Toth isotherm [22].
Initial value q F
q in
qe =
k(q e−q) ∗Δhads ∗ρp
Table 7 Data input for “convection-Diffusion equation”.
k (qe−q) 0 0 1
Β C da
Table 8 Characteristics of the adsorption column and operating conditions for thermal energy storage model validation [14,15]. Parameter
Value
Column length (L) Column inner diameter (Din) Flow rate (Q) Relative humidity (ϕ) Ambient temperature (Tin) Regeneration temperature Adsorbent
7 cm 3.4 cm 20, 24, 28 (LPM ) 80% 25 °C 250 °C Activated alumina and zeolite (AA13X)
aPH2 o [1 + (bPH2 o )n]1/ n
(3)
a = ao exp(E / T )
(4)
b = bo exp(E / T )
(5)
n = no + c / T
(6)
Eqs. (3)–(6) include the partial pressure of water (PH2 o ), the gas temperature (T), as well as the fitted Toth parameters (ao , bo , E, no , c ). The fitted parameters were obtained from experimental data from Wang and LeVan [22] and are presented in Table 3. The initial condition for the water mass balance in the pellet is given by Eq. (7), which is calculated using Temperature-Dependent Toth isotherm at the regeneration conditions.
q|t = 0 = qin
(for all z and r values)
(7)
2.2. Gas side mass balance convection and diffusion of heat and water vapor. 3. The properties of the adsorbent particles are homogeneous. 4. The linear driving force (LDF) model is used to predict the sorption rate by considering both the solid side and the gas side mass transfer resistances. 5. Instantaneous adsorption on pellet surface [14].
The mass balance in the bulk phase of the column is described by Geankoplis [23] as follows:
∊c
∂cg ∂t
= ∊c Dz
∂ 2c g ∂z2
−∊c vg
∂cg ∂z
−ρp
∂q ∂t
(8)
The parameters in Eq. (8) include the column void fraction (∊c ) , the bulk gas concentration in the column (cg) , the axial dispersion coefficient (Dz ) , the superficial gas velocity (vg ) , the pellet density ( ρp ), and
2.1. Solid side mass balance
the rate of water adsorption ∂q . The corresponding initial and boundary ∂t conditions for the material balance in the bulk phase of the column
Adsorption uptake is approximately given by the linear driving force
Fig. 2. Outlet column temperature and water content with respect to time graphs (Q = 20 LPM). 13
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 3. Outlet column temperature and water content with respect to time graphs (Q = 24 LPM).
Fig. 4. Outlet column temperature and water content with respect to time graphs (Q = 28 LPM).
Fig. 5. Schematic diagram of experimental set-up.
2.3. Energy balance in the column
include:
cg |t = 0 = cg, inlet
(for all z and r values )
(9)
cg |z = 0 = cg, inlet
(for all r and t values )
(10)
∂cg ∂z
|z = l = 0
(for all r and t values )
The temperature of the gas stream (bulk phase in the column) is calculated by doing an energy balance around the column. Energy balance is done by considering the energy accumulation, conduction and convection in the column [24]. The energy balance in the column is represented using Eq. (12).
(11) 14
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 6a. Pictorial view of the adsorption-desorption system.
(vg ), the effective thermal conductivity (keff ), the heat transfer coefficient in the bulk phase (hfd ), the column radius (rc ), the column wall temperature (Tw ), and the heat of adsorption (ΔHads ). The effective volumetric heat capacity of the solid-fluid system (ρCp )eff is given by:
(ρCp )eff = (1−∊c )ρp CP,p + ∊c ρg CP,g
(13)
Moreover, the effective thermal conductivity of the solid-fluid system, (keff ) is given by:
keff = (1−∊c ) kp + ∊c k g
Here the subscripts p and g refer to the porous matrix and gas phases, respectively. ρ is the density, CP is the specific heat at constant pressure and k is the thermal conductivity, respectively. The corresponding initial and boundary conditions for the energy balance in the column include:
Fig. 6b. Pictorial view of the adsorption-desorption system.
(ρCp )eff
(14)
2hfd ∂q ∂T + ρg CP, g vg ·∇T = ∇ ·(keff ∇T )− (Tg−Tw ) + ρp ΔHads rc ∂t ∂t (12)
The parameters in Eq. (12) include the effective volumetric heat capacity at constant pressure (ρCp )eff the density of the gas stream ρg , the heat capacity of the gas stream (CP, g ) the superficial gas velocity
Tg |t = 0 = Tin
(15)
Tg |z = 0 = Tin
(16)
∂Tg ∂z
|z = l = 0
Fig. 7. Thermocouples positions along the bed. 15
(17)
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 8. Variation of bed temperature with time at different positions (Q = 17 m3/h) .
Fig. 9. Variation of bed temperature with time at different positions (Q = 19 m3/h) .
2.4. Additional equations
Sc =
Axial dispersion coefficient (Dz ) is calculated based on Reynolds and Schmidt numbers which are calculated using physical properties of the gas stream [25–27] by using the following equations.
Rep =
Dz =
(19)
Dm (20 + 0.5(Rep )(Sc )) εc
(20)
Packed beds cause high-pressure drop as a result of high turbulence of fluid flow through it. This pressure drop depends mainly on the particle size, depth of the bed, fluid viscosity, bed porosity, and flow velocity. The pressure drop increases the blower power used to overcome the friction through the bed. The Ergun Equation [28], can be
ρg vg Dp μg
μg ρg Dm
(18)
16
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 10. Variation of bed temperature with time at different positions (Q = 23 m3/h) .
Fig. 11. Variation of bed temperature with time at different positions (Q = 28 m3/h) .
3. CFD model and setup
used to calculate pressure drop through the bed.
μg vg (1−∊c )2 ρg vg2 (1−∊c ) Δp = 150 2 + 1.75 3 L Dp ∊c Dp ∊3c
The modeling for the current adsorption study was performed using the COMSOL Multi-physics [20] software in which the system of differential equations were solved.
(21)
where, Δp is the pressure drop, L is the length of the packed bed, vg is the superficial velocity, ρg is the fluid density, μg the fluid viscosity, Dp the effective particle diameter and ∊c the void fraction.
3.1. Physics interface “Transport of Diluted Species in Porous Media” interface was used to solve gas side mass balance. “Heat transfer in porous media” interface 17
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 12. Variation of bed temperature with time at different positions (ϕ = 70%) .
Fig. 13. Variation of bed temperature with time at different positions (ϕ = 75%) .
was used to solve overall energy balance. “Classical PDEs convection diffusion Equation” interface was used to solve solid side mass balance. Moreover, properties of moist air as a function of temperature and relative humidity from COMSOL model library have been used.
3.3. Geometry and mesh From the first assumption, geometry was two dimensional. The end point at 0 was the inlet during adsorption phase while the end point at L was the outlet as shown in Fig. 1. Mesh has been set as “physics-controlled mesh” for Sequence Type and “Extra fine” for Element size.
3.2. Parameters, variables and functions 3.4. Transport of diluted species in porous media “TDS”
Model parameters have been defined as given in Table 4. Eqs. (2)–(6) and (18)-(22) were defined as variables.
To solve gas side mass balance, i.e. Eq. (8) for adsorption phase, “TDS” was used. Because of the symmetry of the geometry around midX-Axis, a symmetry was used to decrease the time needed to solve the 18
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 14. Variation of bed temperature with time at different positions (ϕ = 80%) .
Fig. 15. Energy density as a function of inlet flow rate.
model. Main interface inputs are shown in Table 5.
da
∂q + ∇ ·(−c∇q) + β ·∇q = f ∂t
(22)
3.5. Heat transfer in porous media “HT” 3.7. Model validation
To solve global energy balance, i.e. Eq. (7) for adsorption phase, “HT” was used. Because of the symmetry of the geometry around midX-Axis, a symmetry was used to decrease the time needed to solve the model. Main interface inputs are shown in Table 6.
In the following, simulation results of three cases are presented here to highlight the applicability and validity of the model. The input parameters for this case are selected from a previous experiment by Ugur [14,15] at varying conditions for AA13X as adsorbent, which is a hybrid mixture of activated alumina and zeolite 13X. To simulate the results of the experiments. Table 8 presents the properties of the adsorption column used to validate the model. In Figs. 2–4a comparison between temperature of air leaving the
3.6. Classical PDEs, convection diffusion equation To solve solid side mass balance, i.e. Eq. (1) this physics was used. The physics main equation is provided in Eq. (22). Table 7 shows the main interface inputs. 19
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 16. Outlet temperature as a function of time for varying flowrates.
system and the outlet moisture content according to the developed model and the results presented in literature [14] at different flow rates is illustrated. It can be noticed that the model results are in good agreement with that presented in literature.
Table 9 Base case values used in the parametric study. Parameter
Value
Column length (L) Column inner diameter (Din) Relative humidity (ϕ) Ambient temperature (Tin) Flow rate (Q)
20 cm 10 cm 80% 20 °C
Regeneration temperature Adsorbent
4. Experimental study The objective of the following experiments is to explore the performance of the storage system and the influence of varying the main operating parameters affecting its performance. Water was identified as an adsorbate and Silica gel as an adsorbent. Thermo-physical properties of silica gel used are given in Table 4. The experimental setup of the adsorption-desorption system, as shown in Fig. 5, provides a detailed
1 m3/min 120 °C Silica gel
Fig. 17. Power extracted during the discharging phase as a function of time for varying flowrates. 20
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 18. Column outlet temperature as a function of time for varying flowrates.
Fig. 19. Energy densities for varying flowrates.
flow diagram to the various sub-components in the adsorption-desorption system.
4.2. Instrumentation Testo 435 indoor air quality meter is used to measure the inlet temperature and relative humidity of air in addition the inlet air velocity using different probes. The device has been calibrated in accredited Testo laboratories. The accuracy of the temperature probe is ± 0. 2 °C and that of relative humidity is 0.1%RH . The accuracy of the velocity is 0.01 m/s . Temperatures have been measured using K-type thermocouples at different positions in the bed with an accuracy of ± 0.2 °C.
4.1. Experimental set-up Figs. 6a and 6b shows a pictorial view of the thermal energy storage unit. It is composed of the adsorption packed bed, two centrifugal blowers, evaporative bed, electric heater and four ball valves. The packed bed is stainless steel cylinder. Its volume is found to be 5.0 × 10−4 m3 with an inner diameter of 101.6 mm. To ensure that the adsorbent particles wouldn’t leave the bed during the tests, the inlet and the outlet is blocked with narrow stainless-steel net. The bed is insulated using glass wool. The air comes from two air blowers and goes through the wet and dry cylinders. The 1200-W electric heater is used to control the inlet temperature during adsorption and regeneration processes. A set of fittings and 1-in PVC plumping have been used for connections between the packed bed and the two inlet cylinders.
4.3. Experimental procedure Experimental procedure is composed of three steps: regeneration, storing and adsorption. 1. The regeneration process (charging) is done by introducing a hot air through the bed to remove the adsorbate. Inlet airflow is heated using the electric heater. Hot dry air at 120°C with humidity ratio of 21
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 20. Effect of flow-rate on pressure drop per length.
Fig. 21. Variations of Nu/ Nuo ratio, f / fo ratio, and EEC with flow rate.
d. By combining different flows from the two paths, the desired relative humidity and flow rate can be reached. e. And after that, valve (3) is closed and valve (4) is set totally open and the humid air passes through the bed to start the adsorption process. f. During the adsorption process, the change of the temperature values with the time across the bed are recorded. In addition, the values of the flow rate and the relative humidity are monitored, using Testo 435, to be sure that their values are around the desired values, as shown in Fig. 7.
10 g/kg is allowed to pass through the bed for two hours. 2. The storing process is conducted by sealing the column directly after the regeneration process and the column was allowed to cool to room temperature. 3. The adsorption process (discharging) is done by introducing air with high relative humidity through the bed. Adsorption is an exothermic physical process, which releases the stored heat. Adsorption process has been conducted according to the following steps: a. Before turning on the centrifugal blowers, valve (4) is set totally closed and valves (1, 2, and 3) are set totally open. b. The submerged pump is stared to circulate water over the evaporative pads. c. The two centrifugal blowers are turned on. The evaporative pad path produces air with high relative humidity (RH) whereas the other path air produces air with low RH.
4.4. Results and discussion In the following, multiple tests were performed to understand how the operating conditions could affect the energy released from the bed. 22
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 22. Power extracted during the discharging phase as a function of time for varying inlet relative humidity.
Fig. 23. Column outlet temperature as a function of time for varying inlet relative humidity.
4.4.2. Effect of inlet relative humidity The inlet relative humidity was varied to investigate its effect on the system performance. Relative humidity ranging from 70% to 80% were used during the tests. The temperature values along the bed with respect to time at different positions were compared with the experimental results, for varied values of relative humidity, and shown in Figs. 12–14.
The results from the experimental data were compared with the results from the theoretical model. 4.4.1. Effect of inlet flow rate The effect of flow rate on the Silica gel packed bed was examined by varying the flow rate from 17 to 28 m3/h while the temperature and the inlet relative humidity were held constant at 25 °C and 80%, respectively. The temperature values along the bed with respect to time are obtained at different positions. These values are compared with the experimental results as shown in Figs. 8–11. Theoretical and experimental results show that the maximum temperatures of bed surface occur ate bed exit (x/L = 1). On the other hand, the temperature profile demonstrates a sudden increase in temperature during the first few minutes, then a gradual decrease in temperature with time.
4.4.3. System energy density The system performance can be calculated based on the energy storage density. The amount of energy density is calculated using the following equation:
Energy Density = 23
Total energy released Bed volume
(23)
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 24. Energy densities for varying relative humidity.
Fig. 25. Power extracted during the discharging phase as a function of time for varying regeneration temperature.
amount of energy density is found to be 87.86, 90.08 and 92.1 kWh/m3 at relative humidity 70%, 75% and 80%, respectively.
where the total energy released is given by ṁ ∗CP, a (Tout −Tin ) ∗ (tn−tn − 1). The term (tn−tn − 1) represents the time step between the recorded values of temperature. The packed bed with adsorbent has a volume of 5.09∗10−4 m3 . The energy density released from the packed bed as a function of the flow rate is shown in Fig. 15. Flow rate affects the amount of released energy from the packed bed. As the flow rate decreases, the released energy decreases. That is due to the increasing of dissipated energy to the surroundings. As decreasing the flow-rate will increase the bed temperature, as shown in Fig. 16. So, an optimization, between the amount of energy needed to be stored and the temperature lift, needed to be done before choosing the operated flow rate. As a result, a flowrate of nearly 20.5 m3/h was determined to be the optimum flow-rate as increasing the flowrate above this value causes a limited increasing in the energy density at the cost of temperature lift. The inlet relative humidity of the column was varied to observe its effect on the thermal energy storage system. Results show that the energy released increases as the relative humidity increases. The
4.5. Error analysis After modeling water vapor adsorption process in COMSOL, obtained results are compared with the experimental results to estimate the accuracy of the model. This accuracy calculation is done for the outlet bed temperature by using “Mean sum of the percent errors” method, as shown below.
Mean sum of errors =
100 n
n
∑ i=1
modeled data−Experiment data Experiment data
i
(24) The model estimates the bed outlet temperature with 4.4%, 5.2%, 6.2% and 7.1% mean error, when compared to the laboratory experiments at flow rates 17, 19, 23 and 28 m3/h, respectively. Furthermore, the model estimates the outlet column temperature with 24
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 26. Column outlet temperature as a function of time for varying regeneration temperature.
Fig. 27. Energy densities for varying regeneration temperature.
5.8%, 7.3% and 5.8% mean error, when compared to experiments, at inlet relative humidity 70%, 75% and 80%, respectively.
2
2
ωx =
⎛ ∂X ⎞ ω X 2 +⋯+⎛ ∂X ⎞ ω X 2 1 n ⎝ ∂X1 ⎠ ⎝ ∂Xn ⎠ ⎜
⎟
⎜
⎟
(25)
where ωx is the uncertainty of the variable X , ω Xn is the uncertainty of parameter and Xn is the parameter of interest. Analysis of experimental error shows that the maximum uncertainties in the energy density is 10.4%.
4.5.1. Uncertainty analysis In this study, the measured experimental parameters are bed temperatures, flow rate, relative humidity and released energy density. Testo 435 indoor air quality meter is used to measure the inlet temperature and relative humidity of air in addition the inlet air velocity using different probes. The device has been calibrated in accredited Testo laboratories. The accuracy of the temperature probe is ± 0.2 °C. The accuracy of the relative humidity probe is 0.1%RH . The accuracy of the velocity probe is 0.01 m/s . Three thermocouples are used to measure the silica gel temperature at different positions in the bed. Temperatures have been measured using K-type thermocouples with an accuracy of ± 0.2 °C. The maximum uncertainties of the evaluated parameters are estimated by using the following expression;
5. Parametric study A parametric study was performed using the model to predict the behavior of the thermal energy storage system for varying operating conditions and parameters using the theoretical model. The adsorbent used to apply the parametric study was Silica gel. The effects of the different parameters studied were observed by independently changing one parameter while keeping all other system parameters constant and equal to the values given in Table 9, which are referred to as the base case values. To calculate energy density, Eq. (23) is used. The studied 25
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 28. Column outlet temperature as a function of time for varying particle diameters.
Fig. 29. Energy density as a function of particle diameter.
parameters include system flowrate, inlet relative humidity, regeneration temperature and particle radius.
equation (Eq. (21)) and is shown in Fig. 20. Liu et al. [29] used the efficiency evaluation criterion (EEC) to evaluate the comprehensive performance of a heat transfer unit or device.
5.1. Effect of inlet flow rate
EEC = The inlet stream flowrate was varied to observe its effect on the system performance. Flowrates ranging from 0.5 m3/min to 6 m3/min were used and the resulting power-extracted curve and temperature lifts are presented in Figs. 17, 18. By analyzing Fig. 18, we found that increasing the flowrate reduces the output temperature. As shown in Fig. 19, lower flow rates result in lower energy densities, as reduced flowrate do not allow for optimal energy output from the adsorption column. So we need to optimize the operated flow rate as increasing the energy density is at the cost of the temperature lift and duration. On the other hand, the results show an increase in pressure drop with the increase of the flow rate. This increase will lead to a rise in the blower power used to overcome the friction through the bed. The effect of flow rate on the system pressure drop is calculated using Ergun
Nu/ Nuo f / fo
(26)
where Nu and Nuo represent the Nusselt numbers, and f and fo represent resistance coefficients. Here, the subscript o refers to the base case values. Fig. 21 shows the variations of Nu/ Nuo ratio, f / fo ratio, and EEC with the flow rate. Fig. 21 shows that the optimized flow rate, which has the maximum efficiency evaluation criterion, is found to be 0.7 m3/min . 5.2. Effect of inlet relative humidity The inlet relative humidity of the column was varied to observe its effect on the thermal energy storage system. Relative humidity ranging 26
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 30. Effect of particle diameter on pressure drop per length.
Fig. 31. Variation of EEC with flow rate for different particle diameter.
26. The regeneration temperature affects the amount of water remained in the adsorbent after the regeneration process, as the equilibrium state is a function of the temperature and the water partial pressure. The amount of the energy density as a function of regeneration temperature is provided in Fig. 27. It was found that as the regeneration temperature increased above 90 °C causes a limited increasing in the energy density at the cost of regeneration temperature.
from 60% to 100% were used and the resulting power-extracted curve and temperature lifts are presented in Figs. 22, 23. It is notable that the water vapor content of the inlet air has an important impact on the outlet air temperature during the heat release process due to the high sorption amount at high relative humidity. The amount of the energy density as a function of inlet relative humidity is provided in Fig. 24. It was found that the storage density increases with the increase of the relative humidity.
5.4. Effect of particle diameter 5.3. Effect of regeneration temperature The particle diameter was varied from 0.3 mm to 2.5 mm for the base case column dimensions. For a fixed column diameter, changes in the pellet diameter not only affect the adsorption kinetics, they also affect the bed void fraction, as smaller particles generally result in an adsorption column with a lower void fraction. The column outlet
The regeneration temperature of the column was varied to observe its effect on the thermal energy storage system. Regeneration temperatures ranging from 40 °C to 200 °C were used and the resulting power-extracted curve and temperature lifts are presented in Figs. 25, 27
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
temperatures generated by varying the particle radius are provided in Fig. 28. Increasing the particle diameter also has a negative effect on the overall energy density, as shown in Fig. 29. In the lower range of adsorbent particle diameters, up to 1 mm, the energy density does not significantly change with increasing particle diameter. However, as the particle diameter is increased above 1 mm, the energy density significantly decreases due to the difficulty of the adsorbate reaching the adsorption sites. These effects are not significant at low particle diameters, but as seen in Fig. 29, are extremely significant at larger particle diameters. Pressure drop cross the bed will highly decrease with the increase of the particle diameter, based on Eq. (21). The effect of particle diameter on the pressure drop is shown in Fig. 30. Based on the EEC formula, Eq. (26), the flow rate is varied for different particle diameters to obtain the optimal flow rate for each particle diameter, as shown in Fig. 31.
[6] A. Ülkü, M. Mobedi, Adsorption in energy storage, Energy Storage Systems, Springer, 1989, pp. 487–507. [7] A. Hauer, Thermal energy storage with zeolite for heating and cooling applications, Proceedings of 3rd workshop of annex 17 ECES IA/IEA, 1–2 October 2002, Tokyo, Japan, (2002). [8] A. Hauer, Adsorption systems for TES-design and demonstration projects, Springer Netherlands, Dordrecht, 2007, pp. 409–427. [9] H. Kerskes, B. Mette, F. Bertsch, S. Asenbeck, H. Drück, Development of a thermochemical energy storage for solar thermal applications, Proceedings of 30th biennial ISES Solar World Congress, 28 August – 2 September 2011, Kassel, Germany, (2011). [10] C. Bales, P. Gantenbein, D. Jaenig, H. Kerskes, K. Summer, M. Van Essen, R. Weber, Laboratory tests of chemical reactions and prototype sorption storage units. A Report of IEA Solar Heating and Cooling programme – Task 32: Advanced Storage Concepts for Solar and. Low. Energy. Buildings, 2008, Report B4 of Subtask B. Available from http://task32.iea-shc.org/Data/Sites/1/publications/task32-b4.pdf. [11] J. Hammer, H. Fritz, Extrusion of Zeolitic Honeycomb Structures using Thermoplastic Polymers as Plasticising Aid and Binder, Institut für Kunststofftechnologie der Universität Stuttgart, 2002. [12] F. Bertsch, H. Kerskes, H. Drück, H. Müller-Steinhagen, Materialuntersuchungen für chemische Langzeitwärmespeicher (Material investigations for chemical long-term heat storage), Proceedings of OTTI20, Symposium Thermische Solarenergie, 5.–7 May 2010, Bad Staffelstein, Hannover, Germany, (2010). [13] A. Hauer, F. Fischer, Open adsorption system for an energy efficient dishwasher, Chemie Ingenieur Technik 83 (1–2) (2011) 61–66. [14] B. Ugur, Thermal Energy Storage in Adsorbent Beds, Master's Thesis, Chemical and Biological Engineering Departmentm University of Ottawa, Ottawa, ON, Canada, 2013. [15] D. Dicaire, F.H. Tezel, Regeneration and efficiency characterization of hybrid adsorbent for thermal energy storage of excess and solar heat, Renewable Energy 36 (3) (2011) 986–992. [16] B. Dawoud, E.H. Amer, D.M. Gross, Experimental investigation of an adsorptive thermal energy storage, Int. J. Energy Res. 31 (2) (2007) 135–147. [17] G. Li, S. Qian, H. Lee, Y. Hwang, R. Radermacher, Experimental investigation of energy and exergy performance of short term adsorption heat storage for residential application, Energy 65 (2014) 675–691. [18] H. Liu, K. Nagano, Numerical simulation of an open sorption thermal energy storage system using composite sorbents built into a honeycomb structure, Int. J. Heat Mass Transf. 78 (2014) 648–661. [19] N. Yu, R.Z. Wang, L.W. Wang, Sorption thermal storage for solar energy, Prog. Energy Combust. Sci. 39 (5) (2013) 489–514. [20] COMSOL Multiphysics®, Version 5.0, 2014, Burlington, MA, USA. Available from https://www.comsol.com. [21] E. Gleuckauf, J. Coates, Theory of chromatography. Part IV. The influence of incomplete equilibrium on the front boundary of chromatograms and the effectiveness of separation, J. Chem. Soc. (1947) 1315–1321. [22] Y. Wang, M.D. LeVan, Adsorption equilibrium of carbon dioxide and water vapor on zeolites 5A and 13X and silica gel: pure components, J. Chem. Eng. Data 54 (10) (2009) 2839–2844. [23] C.J. Geankoplis, Transport Processes and Separation Process Principles, fourth ed., Prentice Hall, 2003. [24] M. Hashi, J. Thibault, F. Tezel, Recovery of ethanol from carbon dioxide stripped vapor mixture: adsorption prediction and modeling, Ind. Eng. Chem. Res. 49 (18) (2010) 8733–8740. [25] C.J. Geankoplis, Solutions Manual to Accompany Transport Processes and Separation Process Principles: (includes Unit Operations), fourth ed., Prentice Hall, 2003. [26] T.L. Bergman, F.P. Incropera, Introduction to Heat Transfer, sixth ed., John Wiley & Sons, 2011. [27] D.A. Anderson, J.C. Tannehill, R.H. Pletcher, Computational Fluid Mechanics and Heat Transfer, first ed., Hemisphere Publishing Corp., Mcgraw-Hill Book Company, 1984. [28] S. Ergun, Fluid flow through packed columns, Chem. Eng. Prog. 48 (2) (1952) 89–94. [29] W. Liu, P. Liu, J.B. Wang, N.B. Zheng, Z.C. Liu, Exergy destruction minimization: a principle to convective heat transfer enhancement, Int. J. Heat Mass Transf. 122 (2018) 11–21. [30] A.H. Abedin, M.A. Rosen, A critical review of thermochemical energy storage systems, Open Renew. Energy J. 4 (1) (2011) 42–46. [31] S. Kalaiselvam, R. Parameshwaran, Thermal Energy Storage Technologies for Sustainability: Systems Design, Assessment and Applications, Elsevier, 2014. [32] K. Chihara, M. Suzuki, Air drying by pressure swing adsorption, J. Chem. Eng. Jpn. 16 (4) (1983) 293–299. [33] S. Pal, M.R. Hajj, W.P. Wong, I.K. Puri, Thermal energy storage in porous materials with adsorption and desorption of moisture, Int. J. Heat Mass Transf. 69 (2014) 285–292.
6. Conclusion Thermal energy storage using an open adsorption system has been theoretically and experimentally investigated. The main remarkable points of the present study can be summarized as follows: I. A detailed CFD model of heat and mass transfer in a packed bed has been presented using COMSOL Multi-physics. II. The model was validated against two different laboratory experiments with various operating conditions which used AA13X and silica gel as adsorbents. III. According to the flow rate parametric study, we point out that an optimization, between the amount of energy needed to be stored and the temperature lift, needed to be done before choosing the operated flow rate. IV. The parametric study shows that, the higher the relative humidity, the higher the outlet temperature and the energy released from the column. V. According to the regeneration temperature parametric study, increasing the regeneration temperature above 90 °C has a limited effect on the system storage density. VI. It was found that for the given system, in the lower range of adsorbent particle diameters, up to 1 mm, the energy density does not significantly change with increasing particle diameter. Appendix A. Supplementary material Supplementary data to this article can be found online at https:// doi.org/10.1016/j.applthermaleng.2018.10.059. References [1] B. Zalba, J.M. Marı́n, L.F. Cabeza, H. Mehling, Review on thermal energy storage with phase change: materials, heat transfer analysis and applications, Appl. Therm. Eng. 23 (3) (2003) 251–283. [2] W.E. Wentworth, E. Chen, Simple thermal decomposition reactions for storage of solar thermal energy, Sol. Energy 18 (3) (1976) 205–214. [3] J. Hadorn, Advanced storage concepts for active solar energy—IEA SHC Task 32 2003-2007, Paper No. (009), Proceeding of 1st International Conference on Solar Heating, Cooling and Buildings (EuroSun 2008), 7-10 October 2008, Lisbon, Portugal, (2008). [4] J.W. McBain, XCIX. The mechanism of the adsorption (“sorption”) of hydrogen by carbon, Philos. Mag. Ser. 6 18 (108) (1909) 916–935. [5] D.J. Close, R.V. Dunkle, Use of adsorbent beds for energy storage in drying of heating systems, Sol. Energy 19 (3) (1977) 233–238.
28
Contents lists available at ScienceDirect
Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng
Research Paper
Theoretical and experimental investigation of the performance of adsorption heat storage system
T
⁎
Hesham O. Helalya, Mohamed M. Awadb, , Ibrahim I. El-Sharkawyb,c, Ahmed M. Hamedb a
Higher Institute of Engineering and Technology, New Damietta, Damietta, Egypt Mechanical Power Engineering Department, Faculty of Engineering, Mansoura University, Mansoura 35516, Egypt c Higher Future Institute of Engineering and Technology in Mansoura, Mansoura, Egypt b
H I GH L IG H T S
for simulating adsorption based energy storage processes is presented. • ACFDmodel simulation using COMSOL Multi-physics is carried out. • Experiments carried out to explore the storage system performance. • Two cases arearepresented to highlight the model validity. • A parametric study is performed to predict the thermal energy storage system behavior. •
A R T I C LE I N FO
A B S T R A C T
Keywords: Thermal energy storage Open adsorption COMSOL multi-physics Parametric study
Theoretical and experimental investigation of the thermal energy storage of an open adsorption system is presented. The theoretical model describes the mass and energy transfers in the system. The coupled energy and mass balance equations have been solved using the COMSOL™ software. The model was validated against laboratory experiments performed at varying conditions using AA13X as well as silica gel as adsorbent. Laboratory experiments have been conducted to study the effect of flow rate and inlet relative humidity on the amount of energy stored. Temperature and energy density profiles during the adsorption process have been obtained and analyzed for various conditions. Results showed that there is a trade-off between released energy and temperature output and an optimization is recommended before choosing the operating flow rate. It was found that a flow rate of 21 m3/hr has the best performance for a bed volume of 5. 09 × 10−4 m3 . Furthermore, the results show that the storage density increases with the increase of the air inlet relative humidity. For the predefined working conditions and assumptions, the numerical solution shows satisfied agreement with the experimental measurements. A parametric study was performed using silica gel as adsorbent to predict the behavior of the thermal energy storage system for varying operating conditions and parameters. The studied parameters include the system flow rate, relative humidity, regeneration temperature, and the particle diameter. It was found that a flow rate of 0.7 m3/min , regeneration temperature of 95 °C and average particle diameter of 1.0 mm gave the best performance for the bed with a volume of 1.57 × 10−3 m3 .
1. Introduction To face the challenge of energy crisis and pollution issues, it is necessary to develop and expand advanced technologies to reduce the energy demand, increase the energy supply while utilizing renewable energy effectively. One of the most promising developing technologies is energy storage, as it provides the benefit of capturing available energy for use at a later time. This study focuses on the technology of
thermal energy storage (TES), through adsorption. Thermal energy storage can be divided into sensible heat, latent heat and thermochemical or combination of these methods. A brief comparison between the different types of thermal energy storage systems is provided in Table 1. For the types of TES, sensible and latent TES were already developed for many years for commercial applications [1]. In recent years, sorption thermal energy storage systems are increasingly fast as they become promising options for thermal heat
⁎
Corresponding author. E-mail addresses: [email protected] (H.O. Helaly), [email protected] (M.M. Awad), [email protected] (I.I. El-Sharkawy), [email protected] (A.M. Hamed). https://doi.org/10.1016/j.applthermaleng.2018.10.059 Received 22 March 2018; Received in revised form 21 September 2018; Accepted 14 October 2018 Available online 16 October 2018 1359-4311/ © 2018 Elsevier Ltd. All rights reserved.
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Nomenclature
a a0 b b0 c Cg Cpg Cpp Di Dm Dp Ds0 Dz E Ea EEC f fo ΔHads hfd K keff Kg Kp L Ma Mw no Nu
Nuo Δp PH2 o Ppa q Q qe R R ep rc rp RH% Sc Sv t T Tg Tin To Tw Vg
calculated Toth parameter (mol/kg/kPa) fitted Toth parameter (mol/kg/kPa) fitted Toth parameter (1/kPa) fitted Toth parameter (1/kPa) fitted Toth parameter (K) water concentration in the gas phase (mol/m3) heat capacity of the gas stream (J/kg·K) heat capacity of the pellet (J/kg·K) inside diameter of the column (m) molecular diffusivity (m2/s) diameter of the pellet (m) pre-exponential constant (m2/s) axial dispersion (m2/s) fitted Toth parameter (K) activation energy (J/mol) efficiency evaluation criterion (–) resistance coefficient (–) resistance coefficient of the base case (–) heat of adsorption (J/mol) heat transfer coefficient in the bulk phase (W/m2·K) overall mass transfer coefficient (1/s) effective thermal conductivity of the solid-fluid system (W/m·K) thermal conductivity of the gas stream (W/m·K) thermal conductivity of the adsorbent matrix (W/m·K) column length (m) molar mass of air (g/mol) molar mass of water (g/mol) fitted Toth parameter (–) Nusselt number (–)
Greek symbols
∊c μg ρg ρp ρs (ρCp )eff
I. Charging (Regeneration): To store the thermal energy, a hot fluid passes through the bed to remove the adsorbate. This thermal energy can be provided by either waste heat sources or solar energy to avoid the depletion of natural resources and environmental pollution. II. Storing: The energy is permanently stored in the material for an indefinite period as long as it is sealed from the surroundings. Since the thermal energy is not sensibly stored, so theoretically, it is possible to store the material at ambient temperature without reducing the stored energy. III. Discharging: (Adsorption), Adsorption is an exothermic physical process where a gas diffuses into the pores of a porous solid material and is trapped into the crystal lattice which releases heat [6].
Table 1 Comparison of different types of TES based on various performance factors [30,31]. Sensible TES
Temperature range
Lifetime
Up to: 110 °C (Water tanks) 50 °C (Aquifers) 400°C (Concrete) Low 0.1–0.2 GJ/m3 Low
Technology status
Available commercially
Storage density
Latent TES
20 − 40 °C (Paraffins) 30 − 80 °C hydrates)
Thermochemical TES
So far, there is a brief background about open sorption thermal energy storage systems. In Munich, Hauer [7,8] reported a successful open adsorption system using 7000 kg zeolite 13X, which was installed in a school and connected to the local district heating network to offset the peak energy demands, performing at a storage density of 124 kWh/ m3. Within the “MonoSorp” project [9,10], the heat store is designed to be integrated into a building with a controlled ventilation system and with heat recovery. A major improvement was made by using zeolite honeycomb structures called “monoliths” instead of the ordinarily employed fills. These regularly shaped bodies have a large number of small, straight channels. Their main advantages being a low-pressure loss and better adsorption properties than fills [11]. In the “SolSpaces” project [12], special attention was given to the large heat capacities issue. The main idea is to divide the sorption store into several sectors,
20−200 °C (Salt
Moderate 0.2–0.5 GJ/m3 Often limited due to storage material cycling Available commercially for some temperature ranges and materials
void fraction of the column (m3void /m3column) gas stream viscosity (kg/m/s) density of the gas stream (kg/m3) bulk density of the pellet (kg/m3) solid density of the pellet (kg/m3) the effective volumetric heat capacity at constant (J/m3·K )
later use. In adsorption process, thermal energy can be stored through three steps:
storage [2]. Their advantages include relatively high storage densities and storing thermal energy for longer periods with limited heat loss as it can be stored at near ambient temperature. Thermochemical storage can be divided into chemical reaction and sorption [3]. “Sorption” is first proposed by McBain [4] as a general term which encompasses both absorption and adsorption. Adsorption systems can be divided into “closed” and “open” systems. The utilization of the adsorption-desorption cycle as a means of low-temperature energy storage was first proposed by Close [5] who had obtained warm and relatively dry air by using a silica gel adsorbent bed. Thermal energy storage allows storing energy to be used for heating or cooling for
Performance parameter
Nusselt number of the base case (–) pressure drop (Pa) water vapor partial pressure (Pa) column pressure (Pa) adsorbed water in the pellet (mol(H2o)/kg pellet) flow rate (m3/s) adsorbed amount in equilibrium (mol(H2O)/kg pellet) universal gas constant (J/mol·K) Reynolds number for the particles (–) inside radius of the column (m) pellet radius (m) relative humidity (%) Schmidt number (–) specific surface area (m2/kg) time (s) average temperature (K) temperature of the gas stream (K) inlet gas stream temperature (K) ambient air temperature (K) temperature of the wall (k) superficial velocity of the fluid (m/s)
Relatively high: 0.5–3 GJ/m3 Relatively high
Undergoing research
11
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
which can be adsorbed or desorbed separately. It is expected that over 90% of the heat demand for room heating and over 80% of the heat demand for residential hot water preparation can be covered by the solar thermal system combined with the sorption store. Hauer and Fischer [13] used zeolite 13X into a dishwasher to decrease the energy consumption by means of an open adsorption system. The reduction of the energy consumption compared to a traditional dishwasher from about 1.05 kWh to 0.80 kWh per washing cycle leads to energy savings of about 24%. To increase the storage density, Dicaire and Tezel [14,15] tested different adsorbents at varying flow rates and different values of relative humidity to determine the optimal operating conditions. A maximum energy density of 160 kWh/m3 has been achieved with the AA/13X as adsorbent. In addition, there are some studies that focused on the most essential operating conditions that affect the storage process. Dawoud et al. [16] tested zeolite 13X with water as the adsorbate, and obtained an energy density of 144 kWhr/m3 at a flow rate of 0.5 LPM and 165 kWhr/m3 at a flow rate of 2.0 LPM. The study concluded that higher flow rates result in a faster discharge of the storage unit and higher efficiencies. On the other hand, the lower the flow rate during the discharge phase the higher the temperature, at which the storage unit could be discharged. Li et al. [17] tested the mass flow rate effect on the storage performance of the adsorption bed. Results showed that as the mass flow rate increased, the energy storage density improved, but the exergy efficiency decreased due to the increased exergy destruction and loss caused by the pressure drop. They also varied the adsorption bed inlet temperature from 30 °C to 45 °C to test its effect on the storage performance. Results showed that as the adsorption temperature was increased, energy efficiency and loading difference decreased. Liu and Nagano [18] used CaCl2 as an adsorbent and analyzed the effect of air flow rate on the system’s performance. Results showed that more heat is released per time with a faster flow. Additionally, the discharge duration of the high temperature outlet air decreases when the air flow rate is high. They also tested the inlet relative humidity effect on the storage performance of the system. Results showed that the duration of the discharge temperature was extended when the relative humidity increases. Additionally, the equilibrium sorption amount increased with higher relative humidity according to the isothermal equilibrium sorption amount line. Although sorption thermal storage systems have some advantages, they seem to have more complexity than other storing methods such as; expensive investment, poor heat and mass transfer ability and low heat storage density in actual systems. To overcome these barriers, extensive academic efforts are now being carried out worldwide [19]. In the present study, it is aimed to investigate the adsorption process within a thermal energy storage unit using COMSOL Multi-physics software [20]. This simulation can be used to predict the temperature distribution of the thermal energy storage unit and the amount of energy stored. The simulated results will be validated against experimental temperature profiles for various operating conditions. A parametric study of the effect of operating and design parameters is objected.
Table 2 Simulated adsorbents characteristics. Type
Dso [m2/s]
Ea [J/mol]
Reference
Silica gel Type A Zeolite 13X
2.54E −4 5.45E −11
4480 100
[32] [33]
Table 3 Temperature-dependent Toth isotherm parameters for water vapour adsorption on different adsorbents [22]. Parameter
(Zeolite 13X)
(Silica gel)
a o (mol/kg Pa) bo (1/ Pa) n o (−) c (K) E (K)
3.63E-09 2.408E-10 0.3974 −4.199 6852
0.1767 2.787E-08 −0.00119 22.13 1093
Table 4 Input parameters used for simulation. Parameter (symbol)
Silica gel
Zeolite 13X
Bulk density (ρp)
750 kg/m3) ( 0.36 (–)
730 (kg/m3) 0.395 (–)
Porosity (εc) Specific heat capacity (Cpp)
1130
( ) J kg·K
Average particle radius (rp)
0.6∗10−3 (m)
Thermal conductivity (Kp)
0.173
Specific surface area (Sv )
5.3∗105 (m2/kg)
( ) W m·K
1080
( ) J kg·K
1.15∗10−3 (m)
( )
W m·K 5.5∗105 (m2/kg)
3.3
Table 5 Data inputs for “Transport of diluted species in porous media”. Interface input
Value
ux
Q/0.25 πD2i 0 Dz From heat transfer interface ∊c
uy Diffusion Temperature Porosity
2. Theoretical model The model, which describes an open flow system, is schematically presented in Fig. 1. During adsorption, moist air with a specified inlet moisture concentration c(g,in) and temperature Tin flows with a velocity vg through a porous cylinder (adsorbent) of initial moisture content qin, length L and radius rc. The moisture concentration in the adsorbent q. Assumptions made before modeling the process are mentioned below: 1. The airflow passing through the bed is one-dimensional and steady. 2. The transport processes within the adsorbent bed are transient
Fig. 1. Schematic representation of the model. 12
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Interface input
Value
(LDF), which was first proposed by Gleuckauf [21], expressed by the difference of the amount adsorbed, q , and the amount adsorbed in equilibrium qe .
ux
Q 0.25πDi2
dq = K (qe−q) dt
uy Porosity Fluid type Concentration Porous media density Porous media specific heat capacity Porous media thermal conductivity Reaction rate
0 ∊c Moist air From transport of diluted species interface ρp c pp
Table 6 Data inputs for “Heat transfer in porous media”.
(1)
where K is the overall mass transfer coefficient and is calculated using Eq. (2).
( ) E
K=
15∗Dso exp − RTa rp2
Kp
(2)
Interface input
Value
where rp is the adsorbent particle radius, Dso the pre-exponential constant, Ea is the activation energy and R refer to the universal gas constant. The values of activation energy Ea and the pre-exponential constant Dso are provided in Table 2. The equilibrium water concentration was calculated using the Temperature-Dependent Toth isotherm [22].
Initial value q F
q in
qe =
k(q e−q) ∗Δhads ∗ρp
Table 7 Data input for “convection-Diffusion equation”.
k (qe−q) 0 0 1
Β C da
Table 8 Characteristics of the adsorption column and operating conditions for thermal energy storage model validation [14,15]. Parameter
Value
Column length (L) Column inner diameter (Din) Flow rate (Q) Relative humidity (ϕ) Ambient temperature (Tin) Regeneration temperature Adsorbent
7 cm 3.4 cm 20, 24, 28 (LPM ) 80% 25 °C 250 °C Activated alumina and zeolite (AA13X)
aPH2 o [1 + (bPH2 o )n]1/ n
(3)
a = ao exp(E / T )
(4)
b = bo exp(E / T )
(5)
n = no + c / T
(6)
Eqs. (3)–(6) include the partial pressure of water (PH2 o ), the gas temperature (T), as well as the fitted Toth parameters (ao , bo , E, no , c ). The fitted parameters were obtained from experimental data from Wang and LeVan [22] and are presented in Table 3. The initial condition for the water mass balance in the pellet is given by Eq. (7), which is calculated using Temperature-Dependent Toth isotherm at the regeneration conditions.
q|t = 0 = qin
(for all z and r values)
(7)
2.2. Gas side mass balance convection and diffusion of heat and water vapor. 3. The properties of the adsorbent particles are homogeneous. 4. The linear driving force (LDF) model is used to predict the sorption rate by considering both the solid side and the gas side mass transfer resistances. 5. Instantaneous adsorption on pellet surface [14].
The mass balance in the bulk phase of the column is described by Geankoplis [23] as follows:
∊c
∂cg ∂t
= ∊c Dz
∂ 2c g ∂z2
−∊c vg
∂cg ∂z
−ρp
∂q ∂t
(8)
The parameters in Eq. (8) include the column void fraction (∊c ) , the bulk gas concentration in the column (cg) , the axial dispersion coefficient (Dz ) , the superficial gas velocity (vg ) , the pellet density ( ρp ), and
2.1. Solid side mass balance
the rate of water adsorption ∂q . The corresponding initial and boundary ∂t conditions for the material balance in the bulk phase of the column
Adsorption uptake is approximately given by the linear driving force
Fig. 2. Outlet column temperature and water content with respect to time graphs (Q = 20 LPM). 13
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 3. Outlet column temperature and water content with respect to time graphs (Q = 24 LPM).
Fig. 4. Outlet column temperature and water content with respect to time graphs (Q = 28 LPM).
Fig. 5. Schematic diagram of experimental set-up.
2.3. Energy balance in the column
include:
cg |t = 0 = cg, inlet
(for all z and r values )
(9)
cg |z = 0 = cg, inlet
(for all r and t values )
(10)
∂cg ∂z
|z = l = 0
(for all r and t values )
The temperature of the gas stream (bulk phase in the column) is calculated by doing an energy balance around the column. Energy balance is done by considering the energy accumulation, conduction and convection in the column [24]. The energy balance in the column is represented using Eq. (12).
(11) 14
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 6a. Pictorial view of the adsorption-desorption system.
(vg ), the effective thermal conductivity (keff ), the heat transfer coefficient in the bulk phase (hfd ), the column radius (rc ), the column wall temperature (Tw ), and the heat of adsorption (ΔHads ). The effective volumetric heat capacity of the solid-fluid system (ρCp )eff is given by:
(ρCp )eff = (1−∊c )ρp CP,p + ∊c ρg CP,g
(13)
Moreover, the effective thermal conductivity of the solid-fluid system, (keff ) is given by:
keff = (1−∊c ) kp + ∊c k g
Here the subscripts p and g refer to the porous matrix and gas phases, respectively. ρ is the density, CP is the specific heat at constant pressure and k is the thermal conductivity, respectively. The corresponding initial and boundary conditions for the energy balance in the column include:
Fig. 6b. Pictorial view of the adsorption-desorption system.
(ρCp )eff
(14)
2hfd ∂q ∂T + ρg CP, g vg ·∇T = ∇ ·(keff ∇T )− (Tg−Tw ) + ρp ΔHads rc ∂t ∂t (12)
The parameters in Eq. (12) include the effective volumetric heat capacity at constant pressure (ρCp )eff the density of the gas stream ρg , the heat capacity of the gas stream (CP, g ) the superficial gas velocity
Tg |t = 0 = Tin
(15)
Tg |z = 0 = Tin
(16)
∂Tg ∂z
|z = l = 0
Fig. 7. Thermocouples positions along the bed. 15
(17)
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 8. Variation of bed temperature with time at different positions (Q = 17 m3/h) .
Fig. 9. Variation of bed temperature with time at different positions (Q = 19 m3/h) .
2.4. Additional equations
Sc =
Axial dispersion coefficient (Dz ) is calculated based on Reynolds and Schmidt numbers which are calculated using physical properties of the gas stream [25–27] by using the following equations.
Rep =
Dz =
(19)
Dm (20 + 0.5(Rep )(Sc )) εc
(20)
Packed beds cause high-pressure drop as a result of high turbulence of fluid flow through it. This pressure drop depends mainly on the particle size, depth of the bed, fluid viscosity, bed porosity, and flow velocity. The pressure drop increases the blower power used to overcome the friction through the bed. The Ergun Equation [28], can be
ρg vg Dp μg
μg ρg Dm
(18)
16
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 10. Variation of bed temperature with time at different positions (Q = 23 m3/h) .
Fig. 11. Variation of bed temperature with time at different positions (Q = 28 m3/h) .
3. CFD model and setup
used to calculate pressure drop through the bed.
μg vg (1−∊c )2 ρg vg2 (1−∊c ) Δp = 150 2 + 1.75 3 L Dp ∊c Dp ∊3c
The modeling for the current adsorption study was performed using the COMSOL Multi-physics [20] software in which the system of differential equations were solved.
(21)
where, Δp is the pressure drop, L is the length of the packed bed, vg is the superficial velocity, ρg is the fluid density, μg the fluid viscosity, Dp the effective particle diameter and ∊c the void fraction.
3.1. Physics interface “Transport of Diluted Species in Porous Media” interface was used to solve gas side mass balance. “Heat transfer in porous media” interface 17
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 12. Variation of bed temperature with time at different positions (ϕ = 70%) .
Fig. 13. Variation of bed temperature with time at different positions (ϕ = 75%) .
was used to solve overall energy balance. “Classical PDEs convection diffusion Equation” interface was used to solve solid side mass balance. Moreover, properties of moist air as a function of temperature and relative humidity from COMSOL model library have been used.
3.3. Geometry and mesh From the first assumption, geometry was two dimensional. The end point at 0 was the inlet during adsorption phase while the end point at L was the outlet as shown in Fig. 1. Mesh has been set as “physics-controlled mesh” for Sequence Type and “Extra fine” for Element size.
3.2. Parameters, variables and functions 3.4. Transport of diluted species in porous media “TDS”
Model parameters have been defined as given in Table 4. Eqs. (2)–(6) and (18)-(22) were defined as variables.
To solve gas side mass balance, i.e. Eq. (8) for adsorption phase, “TDS” was used. Because of the symmetry of the geometry around midX-Axis, a symmetry was used to decrease the time needed to solve the 18
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 14. Variation of bed temperature with time at different positions (ϕ = 80%) .
Fig. 15. Energy density as a function of inlet flow rate.
model. Main interface inputs are shown in Table 5.
da
∂q + ∇ ·(−c∇q) + β ·∇q = f ∂t
(22)
3.5. Heat transfer in porous media “HT” 3.7. Model validation
To solve global energy balance, i.e. Eq. (7) for adsorption phase, “HT” was used. Because of the symmetry of the geometry around midX-Axis, a symmetry was used to decrease the time needed to solve the model. Main interface inputs are shown in Table 6.
In the following, simulation results of three cases are presented here to highlight the applicability and validity of the model. The input parameters for this case are selected from a previous experiment by Ugur [14,15] at varying conditions for AA13X as adsorbent, which is a hybrid mixture of activated alumina and zeolite 13X. To simulate the results of the experiments. Table 8 presents the properties of the adsorption column used to validate the model. In Figs. 2–4a comparison between temperature of air leaving the
3.6. Classical PDEs, convection diffusion equation To solve solid side mass balance, i.e. Eq. (1) this physics was used. The physics main equation is provided in Eq. (22). Table 7 shows the main interface inputs. 19
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 16. Outlet temperature as a function of time for varying flowrates.
system and the outlet moisture content according to the developed model and the results presented in literature [14] at different flow rates is illustrated. It can be noticed that the model results are in good agreement with that presented in literature.
Table 9 Base case values used in the parametric study. Parameter
Value
Column length (L) Column inner diameter (Din) Relative humidity (ϕ) Ambient temperature (Tin) Flow rate (Q)
20 cm 10 cm 80% 20 °C
Regeneration temperature Adsorbent
4. Experimental study The objective of the following experiments is to explore the performance of the storage system and the influence of varying the main operating parameters affecting its performance. Water was identified as an adsorbate and Silica gel as an adsorbent. Thermo-physical properties of silica gel used are given in Table 4. The experimental setup of the adsorption-desorption system, as shown in Fig. 5, provides a detailed
1 m3/min 120 °C Silica gel
Fig. 17. Power extracted during the discharging phase as a function of time for varying flowrates. 20
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 18. Column outlet temperature as a function of time for varying flowrates.
Fig. 19. Energy densities for varying flowrates.
flow diagram to the various sub-components in the adsorption-desorption system.
4.2. Instrumentation Testo 435 indoor air quality meter is used to measure the inlet temperature and relative humidity of air in addition the inlet air velocity using different probes. The device has been calibrated in accredited Testo laboratories. The accuracy of the temperature probe is ± 0. 2 °C and that of relative humidity is 0.1%RH . The accuracy of the velocity is 0.01 m/s . Temperatures have been measured using K-type thermocouples at different positions in the bed with an accuracy of ± 0.2 °C.
4.1. Experimental set-up Figs. 6a and 6b shows a pictorial view of the thermal energy storage unit. It is composed of the adsorption packed bed, two centrifugal blowers, evaporative bed, electric heater and four ball valves. The packed bed is stainless steel cylinder. Its volume is found to be 5.0 × 10−4 m3 with an inner diameter of 101.6 mm. To ensure that the adsorbent particles wouldn’t leave the bed during the tests, the inlet and the outlet is blocked with narrow stainless-steel net. The bed is insulated using glass wool. The air comes from two air blowers and goes through the wet and dry cylinders. The 1200-W electric heater is used to control the inlet temperature during adsorption and regeneration processes. A set of fittings and 1-in PVC plumping have been used for connections between the packed bed and the two inlet cylinders.
4.3. Experimental procedure Experimental procedure is composed of three steps: regeneration, storing and adsorption. 1. The regeneration process (charging) is done by introducing a hot air through the bed to remove the adsorbate. Inlet airflow is heated using the electric heater. Hot dry air at 120°C with humidity ratio of 21
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 20. Effect of flow-rate on pressure drop per length.
Fig. 21. Variations of Nu/ Nuo ratio, f / fo ratio, and EEC with flow rate.
d. By combining different flows from the two paths, the desired relative humidity and flow rate can be reached. e. And after that, valve (3) is closed and valve (4) is set totally open and the humid air passes through the bed to start the adsorption process. f. During the adsorption process, the change of the temperature values with the time across the bed are recorded. In addition, the values of the flow rate and the relative humidity are monitored, using Testo 435, to be sure that their values are around the desired values, as shown in Fig. 7.
10 g/kg is allowed to pass through the bed for two hours. 2. The storing process is conducted by sealing the column directly after the regeneration process and the column was allowed to cool to room temperature. 3. The adsorption process (discharging) is done by introducing air with high relative humidity through the bed. Adsorption is an exothermic physical process, which releases the stored heat. Adsorption process has been conducted according to the following steps: a. Before turning on the centrifugal blowers, valve (4) is set totally closed and valves (1, 2, and 3) are set totally open. b. The submerged pump is stared to circulate water over the evaporative pads. c. The two centrifugal blowers are turned on. The evaporative pad path produces air with high relative humidity (RH) whereas the other path air produces air with low RH.
4.4. Results and discussion In the following, multiple tests were performed to understand how the operating conditions could affect the energy released from the bed. 22
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 22. Power extracted during the discharging phase as a function of time for varying inlet relative humidity.
Fig. 23. Column outlet temperature as a function of time for varying inlet relative humidity.
4.4.2. Effect of inlet relative humidity The inlet relative humidity was varied to investigate its effect on the system performance. Relative humidity ranging from 70% to 80% were used during the tests. The temperature values along the bed with respect to time at different positions were compared with the experimental results, for varied values of relative humidity, and shown in Figs. 12–14.
The results from the experimental data were compared with the results from the theoretical model. 4.4.1. Effect of inlet flow rate The effect of flow rate on the Silica gel packed bed was examined by varying the flow rate from 17 to 28 m3/h while the temperature and the inlet relative humidity were held constant at 25 °C and 80%, respectively. The temperature values along the bed with respect to time are obtained at different positions. These values are compared with the experimental results as shown in Figs. 8–11. Theoretical and experimental results show that the maximum temperatures of bed surface occur ate bed exit (x/L = 1). On the other hand, the temperature profile demonstrates a sudden increase in temperature during the first few minutes, then a gradual decrease in temperature with time.
4.4.3. System energy density The system performance can be calculated based on the energy storage density. The amount of energy density is calculated using the following equation:
Energy Density = 23
Total energy released Bed volume
(23)
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 24. Energy densities for varying relative humidity.
Fig. 25. Power extracted during the discharging phase as a function of time for varying regeneration temperature.
amount of energy density is found to be 87.86, 90.08 and 92.1 kWh/m3 at relative humidity 70%, 75% and 80%, respectively.
where the total energy released is given by ṁ ∗CP, a (Tout −Tin ) ∗ (tn−tn − 1). The term (tn−tn − 1) represents the time step between the recorded values of temperature. The packed bed with adsorbent has a volume of 5.09∗10−4 m3 . The energy density released from the packed bed as a function of the flow rate is shown in Fig. 15. Flow rate affects the amount of released energy from the packed bed. As the flow rate decreases, the released energy decreases. That is due to the increasing of dissipated energy to the surroundings. As decreasing the flow-rate will increase the bed temperature, as shown in Fig. 16. So, an optimization, between the amount of energy needed to be stored and the temperature lift, needed to be done before choosing the operated flow rate. As a result, a flowrate of nearly 20.5 m3/h was determined to be the optimum flow-rate as increasing the flowrate above this value causes a limited increasing in the energy density at the cost of temperature lift. The inlet relative humidity of the column was varied to observe its effect on the thermal energy storage system. Results show that the energy released increases as the relative humidity increases. The
4.5. Error analysis After modeling water vapor adsorption process in COMSOL, obtained results are compared with the experimental results to estimate the accuracy of the model. This accuracy calculation is done for the outlet bed temperature by using “Mean sum of the percent errors” method, as shown below.
Mean sum of errors =
100 n
n
∑ i=1
modeled data−Experiment data Experiment data
i
(24) The model estimates the bed outlet temperature with 4.4%, 5.2%, 6.2% and 7.1% mean error, when compared to the laboratory experiments at flow rates 17, 19, 23 and 28 m3/h, respectively. Furthermore, the model estimates the outlet column temperature with 24
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 26. Column outlet temperature as a function of time for varying regeneration temperature.
Fig. 27. Energy densities for varying regeneration temperature.
5.8%, 7.3% and 5.8% mean error, when compared to experiments, at inlet relative humidity 70%, 75% and 80%, respectively.
2
2
ωx =
⎛ ∂X ⎞ ω X 2 +⋯+⎛ ∂X ⎞ ω X 2 1 n ⎝ ∂X1 ⎠ ⎝ ∂Xn ⎠ ⎜
⎟
⎜
⎟
(25)
where ωx is the uncertainty of the variable X , ω Xn is the uncertainty of parameter and Xn is the parameter of interest. Analysis of experimental error shows that the maximum uncertainties in the energy density is 10.4%.
4.5.1. Uncertainty analysis In this study, the measured experimental parameters are bed temperatures, flow rate, relative humidity and released energy density. Testo 435 indoor air quality meter is used to measure the inlet temperature and relative humidity of air in addition the inlet air velocity using different probes. The device has been calibrated in accredited Testo laboratories. The accuracy of the temperature probe is ± 0.2 °C. The accuracy of the relative humidity probe is 0.1%RH . The accuracy of the velocity probe is 0.01 m/s . Three thermocouples are used to measure the silica gel temperature at different positions in the bed. Temperatures have been measured using K-type thermocouples with an accuracy of ± 0.2 °C. The maximum uncertainties of the evaluated parameters are estimated by using the following expression;
5. Parametric study A parametric study was performed using the model to predict the behavior of the thermal energy storage system for varying operating conditions and parameters using the theoretical model. The adsorbent used to apply the parametric study was Silica gel. The effects of the different parameters studied were observed by independently changing one parameter while keeping all other system parameters constant and equal to the values given in Table 9, which are referred to as the base case values. To calculate energy density, Eq. (23) is used. The studied 25
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 28. Column outlet temperature as a function of time for varying particle diameters.
Fig. 29. Energy density as a function of particle diameter.
parameters include system flowrate, inlet relative humidity, regeneration temperature and particle radius.
equation (Eq. (21)) and is shown in Fig. 20. Liu et al. [29] used the efficiency evaluation criterion (EEC) to evaluate the comprehensive performance of a heat transfer unit or device.
5.1. Effect of inlet flow rate
EEC = The inlet stream flowrate was varied to observe its effect on the system performance. Flowrates ranging from 0.5 m3/min to 6 m3/min were used and the resulting power-extracted curve and temperature lifts are presented in Figs. 17, 18. By analyzing Fig. 18, we found that increasing the flowrate reduces the output temperature. As shown in Fig. 19, lower flow rates result in lower energy densities, as reduced flowrate do not allow for optimal energy output from the adsorption column. So we need to optimize the operated flow rate as increasing the energy density is at the cost of the temperature lift and duration. On the other hand, the results show an increase in pressure drop with the increase of the flow rate. This increase will lead to a rise in the blower power used to overcome the friction through the bed. The effect of flow rate on the system pressure drop is calculated using Ergun
Nu/ Nuo f / fo
(26)
where Nu and Nuo represent the Nusselt numbers, and f and fo represent resistance coefficients. Here, the subscript o refers to the base case values. Fig. 21 shows the variations of Nu/ Nuo ratio, f / fo ratio, and EEC with the flow rate. Fig. 21 shows that the optimized flow rate, which has the maximum efficiency evaluation criterion, is found to be 0.7 m3/min . 5.2. Effect of inlet relative humidity The inlet relative humidity of the column was varied to observe its effect on the thermal energy storage system. Relative humidity ranging 26
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
Fig. 30. Effect of particle diameter on pressure drop per length.
Fig. 31. Variation of EEC with flow rate for different particle diameter.
26. The regeneration temperature affects the amount of water remained in the adsorbent after the regeneration process, as the equilibrium state is a function of the temperature and the water partial pressure. The amount of the energy density as a function of regeneration temperature is provided in Fig. 27. It was found that as the regeneration temperature increased above 90 °C causes a limited increasing in the energy density at the cost of regeneration temperature.
from 60% to 100% were used and the resulting power-extracted curve and temperature lifts are presented in Figs. 22, 23. It is notable that the water vapor content of the inlet air has an important impact on the outlet air temperature during the heat release process due to the high sorption amount at high relative humidity. The amount of the energy density as a function of inlet relative humidity is provided in Fig. 24. It was found that the storage density increases with the increase of the relative humidity.
5.4. Effect of particle diameter 5.3. Effect of regeneration temperature The particle diameter was varied from 0.3 mm to 2.5 mm for the base case column dimensions. For a fixed column diameter, changes in the pellet diameter not only affect the adsorption kinetics, they also affect the bed void fraction, as smaller particles generally result in an adsorption column with a lower void fraction. The column outlet
The regeneration temperature of the column was varied to observe its effect on the thermal energy storage system. Regeneration temperatures ranging from 40 °C to 200 °C were used and the resulting power-extracted curve and temperature lifts are presented in Figs. 25, 27
Applied Thermal Engineering 147 (2019) 10–28
H.O. Helaly et al.
temperatures generated by varying the particle radius are provided in Fig. 28. Increasing the particle diameter also has a negative effect on the overall energy density, as shown in Fig. 29. In the lower range of adsorbent particle diameters, up to 1 mm, the energy density does not significantly change with increasing particle diameter. However, as the particle diameter is increased above 1 mm, the energy density significantly decreases due to the difficulty of the adsorbate reaching the adsorption sites. These effects are not significant at low particle diameters, but as seen in Fig. 29, are extremely significant at larger particle diameters. Pressure drop cross the bed will highly decrease with the increase of the particle diameter, based on Eq. (21). The effect of particle diameter on the pressure drop is shown in Fig. 30. Based on the EEC formula, Eq. (26), the flow rate is varied for different particle diameters to obtain the optimal flow rate for each particle diameter, as shown in Fig. 31.
[6] A. Ülkü, M. Mobedi, Adsorption in energy storage, Energy Storage Systems, Springer, 1989, pp. 487–507. [7] A. Hauer, Thermal energy storage with zeolite for heating and cooling applications, Proceedings of 3rd workshop of annex 17 ECES IA/IEA, 1–2 October 2002, Tokyo, Japan, (2002). [8] A. Hauer, Adsorption systems for TES-design and demonstration projects, Springer Netherlands, Dordrecht, 2007, pp. 409–427. [9] H. Kerskes, B. Mette, F. Bertsch, S. Asenbeck, H. Drück, Development of a thermochemical energy storage for solar thermal applications, Proceedings of 30th biennial ISES Solar World Congress, 28 August – 2 September 2011, Kassel, Germany, (2011). [10] C. Bales, P. Gantenbein, D. Jaenig, H. Kerskes, K. Summer, M. Van Essen, R. Weber, Laboratory tests of chemical reactions and prototype sorption storage units. A Report of IEA Solar Heating and Cooling programme – Task 32: Advanced Storage Concepts for Solar and. Low. Energy. Buildings, 2008, Report B4 of Subtask B. Available from http://task32.iea-shc.org/Data/Sites/1/publications/task32-b4.pdf. [11] J. Hammer, H. Fritz, Extrusion of Zeolitic Honeycomb Structures using Thermoplastic Polymers as Plasticising Aid and Binder, Institut für Kunststofftechnologie der Universität Stuttgart, 2002. [12] F. Bertsch, H. Kerskes, H. Drück, H. Müller-Steinhagen, Materialuntersuchungen für chemische Langzeitwärmespeicher (Material investigations for chemical long-term heat storage), Proceedings of OTTI20, Symposium Thermische Solarenergie, 5.–7 May 2010, Bad Staffelstein, Hannover, Germany, (2010). [13] A. Hauer, F. Fischer, Open adsorption system for an energy efficient dishwasher, Chemie Ingenieur Technik 83 (1–2) (2011) 61–66. [14] B. Ugur, Thermal Energy Storage in Adsorbent Beds, Master's Thesis, Chemical and Biological Engineering Departmentm University of Ottawa, Ottawa, ON, Canada, 2013. [15] D. Dicaire, F.H. Tezel, Regeneration and efficiency characterization of hybrid adsorbent for thermal energy storage of excess and solar heat, Renewable Energy 36 (3) (2011) 986–992. [16] B. Dawoud, E.H. Amer, D.M. Gross, Experimental investigation of an adsorptive thermal energy storage, Int. J. Energy Res. 31 (2) (2007) 135–147. [17] G. Li, S. Qian, H. Lee, Y. Hwang, R. Radermacher, Experimental investigation of energy and exergy performance of short term adsorption heat storage for residential application, Energy 65 (2014) 675–691. [18] H. Liu, K. Nagano, Numerical simulation of an open sorption thermal energy storage system using composite sorbents built into a honeycomb structure, Int. J. Heat Mass Transf. 78 (2014) 648–661. [19] N. Yu, R.Z. Wang, L.W. Wang, Sorption thermal storage for solar energy, Prog. Energy Combust. Sci. 39 (5) (2013) 489–514. [20] COMSOL Multiphysics®, Version 5.0, 2014, Burlington, MA, USA. Available from https://www.comsol.com. [21] E. Gleuckauf, J. Coates, Theory of chromatography. Part IV. The influence of incomplete equilibrium on the front boundary of chromatograms and the effectiveness of separation, J. Chem. Soc. (1947) 1315–1321. [22] Y. Wang, M.D. LeVan, Adsorption equilibrium of carbon dioxide and water vapor on zeolites 5A and 13X and silica gel: pure components, J. Chem. Eng. Data 54 (10) (2009) 2839–2844. [23] C.J. Geankoplis, Transport Processes and Separation Process Principles, fourth ed., Prentice Hall, 2003. [24] M. Hashi, J. Thibault, F. Tezel, Recovery of ethanol from carbon dioxide stripped vapor mixture: adsorption prediction and modeling, Ind. Eng. Chem. Res. 49 (18) (2010) 8733–8740. [25] C.J. Geankoplis, Solutions Manual to Accompany Transport Processes and Separation Process Principles: (includes Unit Operations), fourth ed., Prentice Hall, 2003. [26] T.L. Bergman, F.P. Incropera, Introduction to Heat Transfer, sixth ed., John Wiley & Sons, 2011. [27] D.A. Anderson, J.C. Tannehill, R.H. Pletcher, Computational Fluid Mechanics and Heat Transfer, first ed., Hemisphere Publishing Corp., Mcgraw-Hill Book Company, 1984. [28] S. Ergun, Fluid flow through packed columns, Chem. Eng. Prog. 48 (2) (1952) 89–94. [29] W. Liu, P. Liu, J.B. Wang, N.B. Zheng, Z.C. Liu, Exergy destruction minimization: a principle to convective heat transfer enhancement, Int. J. Heat Mass Transf. 122 (2018) 11–21. [30] A.H. Abedin, M.A. Rosen, A critical review of thermochemical energy storage systems, Open Renew. Energy J. 4 (1) (2011) 42–46. [31] S. Kalaiselvam, R. Parameshwaran, Thermal Energy Storage Technologies for Sustainability: Systems Design, Assessment and Applications, Elsevier, 2014. [32] K. Chihara, M. Suzuki, Air drying by pressure swing adsorption, J. Chem. Eng. Jpn. 16 (4) (1983) 293–299. [33] S. Pal, M.R. Hajj, W.P. Wong, I.K. Puri, Thermal energy storage in porous materials with adsorption and desorption of moisture, Int. J. Heat Mass Transf. 69 (2014) 285–292.
6. Conclusion Thermal energy storage using an open adsorption system has been theoretically and experimentally investigated. The main remarkable points of the present study can be summarized as follows: I. A detailed CFD model of heat and mass transfer in a packed bed has been presented using COMSOL Multi-physics. II. The model was validated against two different laboratory experiments with various operating conditions which used AA13X and silica gel as adsorbents. III. According to the flow rate parametric study, we point out that an optimization, between the amount of energy needed to be stored and the temperature lift, needed to be done before choosing the operated flow rate. IV. The parametric study shows that, the higher the relative humidity, the higher the outlet temperature and the energy released from the column. V. According to the regeneration temperature parametric study, increasing the regeneration temperature above 90 °C has a limited effect on the system storage density. VI. It was found that for the given system, in the lower range of adsorbent particle diameters, up to 1 mm, the energy density does not significantly change with increasing particle diameter. Appendix A. Supplementary material Supplementary data to this article can be found online at https:// doi.org/10.1016/j.applthermaleng.2018.10.059. References [1] B. Zalba, J.M. Marı́n, L.F. Cabeza, H. Mehling, Review on thermal energy storage with phase change: materials, heat transfer analysis and applications, Appl. Therm. Eng. 23 (3) (2003) 251–283. [2] W.E. Wentworth, E. Chen, Simple thermal decomposition reactions for storage of solar thermal energy, Sol. Energy 18 (3) (1976) 205–214. [3] J. Hadorn, Advanced storage concepts for active solar energy—IEA SHC Task 32 2003-2007, Paper No. (009), Proceeding of 1st International Conference on Solar Heating, Cooling and Buildings (EuroSun 2008), 7-10 October 2008, Lisbon, Portugal, (2008). [4] J.W. McBain, XCIX. The mechanism of the adsorption (“sorption”) of hydrogen by carbon, Philos. Mag. Ser. 6 18 (108) (1909) 916–935. [5] D.J. Close, R.V. Dunkle, Use of adsorbent beds for energy storage in drying of heating systems, Sol. Energy 19 (3) (1977) 233–238.
28