Hydrogen Storage System Models
J une 4, 2012
2012 World Hydrogen Energy Conference
Bruce Hardy, Claudio Corgnale,
Savannah River National Laboratory
2
Adsorbent Models
• Solve conservation equations for mass, momentum, and energy in 2
or 3 dimensions
Use weakly compressible Brinkman equations in all flow domains
Include thermal radiation
Temperature dependent fit for adsorbent specific heat
Correlations for realgas hydrogen properties from the
NIST REFPROP 23 V9.0
©
database
Valid for 0.05 ≤ P ≤ 35.0MPa and 70 ≤ T ≤ 450K
Compressibility factor
Enthalpy
Viscosity
Thermal conductivity
• Maxsorb™
/MOF5™ thermodynamic models for absolute adsorption
and internal energy of adsorbed hydrogen obtained from:
Richard, Bénard and Chahine. “Gas Adsorption Process in Activated Carbon Over a
Wide Temperature Range Above the Critical Point. Part 1: Modified DubininAstakhov
Model.”
Richard, Bénard and Chahine. “Gas Adsorption Process in Activated Carbon Over a
Wide Temperature Range Above the Critical Point. Part 2: Conservation of Mass and
Energy.”
MOF5™ isotherm data obtained by Ford
Error between NIST data and
correlations is less than 0.63%
3
Governing Equations
( )
t
n
t
P
P t
T
T
a Ads
H
H H
∂
∂

.

\

− ⋅ −∇ =
∂
∂
∂
∂
+
∂
∂
∂
∂
ε
ρ
ρ
ρ ρ
s 2
2 2
v
( )
( ) ( )
t
T
Cp I T k
t
n u
t
U
M
hS
P
t
P
T
c
c
T
T
h
t
T
T
h
c
Ads Ads s s
d
s
T
s
a a
Ads
H
s
H
H
s H
∂
∂
− ∇
(
¸
(
¸
⋅ ∇


.

\

− − ∇ + ∇ + ∇ ⋅ ∇ =

.

\

∂
∂
+
∂
∆ ∂
+ + 
.

\

∇ ⋅ +
∂
∂
∂
∂
+ 
.

\

∂
∂
+
∂
∂
∂
∂
•
•
ρ
µ
η
ε
µ
ρ ε ε
n Dissipatio Viscous
v v
3
2
v v
Energy Sorption Work Pressure
v v
0
2
0 2
2
2
(
(
¸
(
¸

.

\



.

\

− =
P
P
ln
E
RT
exp n n
0 2
2
a
max a
adsorption of energy Free T E
a
= + = β α
Conservation of Mass (Hydrogen)
Conservation of Momentum
(Brinkman Equations)
( ) Porosity Effective V V
a void Ads
= − = ρ ε
Conservation of Energy
Absolute Adsorption
Relative Internal Energy
of Adsorbed Hydrogen
component Entropic
component Enthalpic
=
=
β
α
(
(
¸
(
¸


.

\

− − +
(
(
¸
(
¸


.

\



.

\

− − − = ∆
max
a
max
a max
a
n
n
ln RT na
n
n
ln erf 1
2
n
U α
π α
) , (
2
0
0
T P u u
u n U U
gas adsorbed of energy Internal U
atm H
a a a
a
=
− = ∆
=
( )
( ) ( ) g I v
1
3
2
v v P
v
S
v v
t
v
2 H s d
T
s s
s 2
0
s s 2
2 H s 2 H
ρ
ε
η
µ
ε
µ
ε κ
µ
ε
ρ
ε
ρ
+
(
¸
(
¸
⋅ ∇ 
.

\


.

\

− ⋅ ∇ −
(
¸
(
¸
∇ + ∇ ⋅ ∇ + −∇ =

.

\

+ + ∇ ⋅ +
∂
∂
( ) P 4
2
2
2 H
2
∇ + +
=
µ βρ α α
κ =Ergun Permeability
β
κ
1
= =BlakeKozeny Permeability
or
Plus Other Ancillary
Equations!
4
Validation With UQTR Charging Experiments (MaxSorb)
Volume of Dewar 19 L
Volume of Tank 2.5 L
Mass of MaxSorb™ 0.674 kg
Mass of Tank 3.786 kg
Mass of TC Assembl y 0.471 kg
Overview of Low
Temperature Experiment
P
0
=0.182 MPa T
0
=79K
Dewar is filled with liquid N
2
Large Dewar – kept pressure
vessel immersed in liquid N
2
H
2
added at 100 SLPM & ~80K
Transient temperatures in tank
and in Dewar are measured
N
2
is allowed to boil off, flowrate
and temperature are measured
Vessel surface temperatures
measured for use as boundary
conditions
5
Distribution of Stored Hydrogen
Outer Wall of the Vessel Contacted by Liquid N
2
at ~ 79K
6
UQTR Charging Experiments
TC1
Time (s)
0 200 400 600 800 1000 1200 1400 1600
T
e
m
p
e
r
a
t
u
r
e
(
K
)
75
80
85
90
95
100
105
Model
Data
TC2
Time (s)
0 200 400 600 800 1000 1200 1400 1600
T
e
m
p
e
r
a
t
u
r
e
(
K
)
70
80
90
100
110
120
130
Model
Data
TC3
Time (s)
0 200 400 600 800 1000 1200 1400 1600
T
e
m
p
e
r
a
t
u
r
e
(
K
)
60
80
100
120
140
160
Model
Data
TC4
Time (s)
0 200 400 600 800 1000 1200 1400 1600
T
e
m
p
e
r
a
t
u
r
e
(
K
)
60
80
100
120
140
160
180
Model
Data
7
Discharge – Central Heating Element
To extend dormancy hydrogen discharge is
effected by depressurization then heating
T
0
= 80 K
P
0
= 20 bar
Adsorbent Volume = 0.164m
3
Wall ρ*Cp=2.43e6 (J/m
3
K)
Heater on if P<50 bar
and Tavg of heater is < 420K
No Flow
A
d
i
a
b
a
t
i
c
B
o
u
n
d
a
r
i
e
s
No Flow
Pressure increases by 6.2 bar in 1800 seconds
Not very effective!
Midplane
8
FlowThrough Charging with MaxSorb™ and MOF5™ Data
T
0
= 180 K
P
0
= 5 bar
Ads Vol = 0.164m
3
Wall ρ*Cp=2.43e5 (J/m
3
K)
T
in
=80K
P
in
from 5 to 200 bar in 20 sec
Avg exit velocity
from 0 to 9 m/s from
3 to 5 sec
A
d
i
a
b
a
t
i
c
W
a
l
l
Thermal conductivity has little influence
for this cooling method
Available ⇒ Amount Released Upon Return to Initial State
2017 DOE Technical Target is to
charge 5.6 kg of usable H
2
in 198 sec
Initial Temperature is 180K
Initial Pressure is 5 bar
299
260
180
140
220
100
80
Temper atur e(K)
9
State of Exhaust Hydrogen
Case
Charge Time (s)
Mass of Exhaust
H
2
(kg)
Average
Temperature (K)
MaxSorb
Low Wall ρCp
140
17.19
133.67
MaxSorb
Nominal Wall ρCp
198
*
27.51
120.06
MOF5
Low Wall ρCp
95
11.61
132.42
MaxSorb
Low Wall ρCp
Radial Cooling
155
19.58
137.49
Flow through cooling is most efficient if the mass
and average temperature of exhaust hydrogen are
minimized (minimize total enthalpy)
* Had not reached full capacity
Each case loads approximately 8
kg of recoverable hydrogen
10
Heat Dissipation During Charging
E
n
d
o
f
P
r
e
s
s
u
r
e
T
r
a
n
s
i
e
n
t
t
P
T
c
c
T
∂
∂
∂
∂
−ε Generation by Pressure Work =
( )
(
¸
(
¸
∂
∂
− + ∆
∂
∂
−
t
n
h u n U
t
a
g a a Ads 0
ρ
Generation by Heat of Adsorption =
E
n
d
o
f
P
r
e
s
s
u
r
e
T
r
a
n
s
i
e
n
t
Total Pressure
Work (MJ)
Total Heat of
Adsorption (MJ)
MaxSorb 1.39 4.81
MOF5 2.03 2.14
Difference in pressure work for this case
is principally due to different porosities
Relative amount of pressure work is due
to:
• Fraction of gas adsorbed
• Adsorbent thermodynamics
11
UQTR FlowThrough Experiments (MaxSorb)
Heat exchangers
for cooling inlet gas
Pressure vessel
Purpose: Validation of the flowthrough cooling concept
Currently in preliminary stage
Preliminary tests use same pressure vessel as for
charging tests with granular activated carbon produced
at UQTR
Tests conducted with helium, nitrogen, and hydrogen
T
0
=298 K
P
0
=0.037 MPa
P
f
≈8.5 MPa
12
Experimental Results for Hydrogen FlowThrough Cooling
The decrease in the average bed temperature demonstrates that flow
through cooling can effectively cool a cryoadsorbent storage vessel.
• Discharge temperature trace indicates that local bed and/or
wall temperatures exceed the average values
• Average of the lower wall temperatures exceeds the average
of the upper wall temperatures – Suggests Channeling
• Average bed temperature drops markedly after the start of
gas outflow (when flow through cooling occurs)
13
Experimental Validation of FlowThrough Cooling
t = 100s
319.23
112.71
300
250
200
150
t = 550s
318.75
99.158
300
250
200
150
Instantaneous Temperature Profiles
Good comparison with data
Application of Numerical Models
• Design of test rigs
• Development of test matrix
• Evaluation of design concepts
• Determination of flowrate and other
operating parameters
• Scaleup: Extension to other designs or
operating conditions
14
Thermodynamic Considerations: Limits to Pressurization With Adsorbent
Chahine, Richard and Cossement noted that at sufficiently high pressure, the
amount of hydrogen stored by compression alone exceeds that stored with
adsorbent. This applies to both MaxSorb and MOF5.
There are limits to the ability of increased pressure to improve storage in
adsorbent based systems.
15
Powder Form
3 to 6 mm
Pellet Form
50 mm
Large Compressed Form
“ Hockey Puck”
MOF5 Form, Tank, and Tank Internals / Heat Exchanger Design
MOF5 Form Selection:
• Powder Form
• Pelletized Form
• Large Compressed Form
(Hockey Puck)
• ENG or other thermal
enhancement
Tank Selection:
• Aluminum Type I
• Stainless Steel Type I
• Composite Fiber Type III
• Composite Fiber Type IV
Tank Internals / HX Selection:
• Resistance Heater
o Fin and tube
o Wire mesh
o Hexpack / Honeycomb
• MATI / IsolatedH
2
insert
Modular Adsorption
Tank Insert (MATI)
16
Conclusions
• Flow Through Cooling is a Viable Concept for Nominal Form of MaxSorb
Optimize vessel design & operation
Thermally isolate vessel wall from bed or reduce ρCp
Alternative, novel, heat transfer technologies being pursued by OSU, UQTR
and SRNL will likely be needed for compacted adsorbents
Optimize charging conditions to minimize total exhaust gas enthalpy
Flow through cooling not likely to work for compacted media
Need permeability data
• Charging Conditions and Vessel Geometry Affect Heat Release
Result of pressure work
Can be significant
This was noted by HermosillaLara, et. .al. (2007)
1
and Momen, et. al.
(2009)
2
who claimed pressure work accounted for more than 70% of the
energy released during the charging process based on their model and
experiments.
• Need Better Way to Heat Bed
Low bed thermal conductivity requires short thermal diffusion length
Flow through heating can work, but requires pump, valves and possibly
combustion of hydrogen
1 HermosillaLara G, Momen G, Marty PH, Le Neindre B, Hassouni K. Hydrogen storage by adsorption on activated carbon:
Investigation of the thermal effects during the charging process. Int J Hydrogen Energy 2007;32:154253.
2 Momen G, Hermosilla G, Michau A, Pons M, Firdaus M, Marty PH, Hassouni K. Experimental and numerical investigation of the
thermal effects during hydrogen charging in a packed bed storage tank. Int J Hydrogen Energy 2009;52:14951503.
17