Just another

Fuel Cell Formulary

by Dr. Alexander Kabza

This version is from November 9, 2016 and optimized for Smartphones.
Latest version see www.kabza.de. Don’t hesitate to send me any comments or failures or ideas by
email!
Contents 2

Contents

1 Fundamentals 4
1.1 Thermodynamic of H2 /O2 electrochemical device . . . . 4
1.2 Nernst equation . . . . . . . . . . . . . . . . . . . . . . . 7
1.3 Reactant consumption and feed . . . . . . . . . . . . . . 8
1.4 Hydrogen energy and power . . . . . . . . . . . . . . . 10
1.5 PolCurve (Current voltage dependency) . . . . . . . . . 12
1.5.1 Kinetic of the H2 /O2 electrochemical device . . . 12
1.5.2 Butler-Volmer equation . . . . . . . . . . . . . . 12
1.5.3 Example PolCurve . . . . . . . . . . . . . . . . . 15
1.6 Fuel cell stack power . . . . . . . . . . . . . . . . . . . . 16
1.7 Fuel cell efficiencies . . . . . . . . . . . . . . . . . . . . 18
1.7.1 Different efficiencies . . . . . . . . . . . . . . . . 18
1.7.2 Electric efficiency calculation . . . . . . . . . . . 23

2 Fuel cell system (FCS) 24

2.1 Anode subsystem . . . . . . . . . . . . . . . . . . . . . 24
2.2 Cathode subsystem . . . . . . . . . . . . . . . . . . . . 26
2.2.1 Cathode pressure controls . . . . . . . . . . . . 27
2.2.2 Stack pressure drop . . . . . . . . . . . . . . . . 30
2.2.3 Cathode air compression . . . . . . . . . . . . . 31
2.3 Coolant subsystem . . . . . . . . . . . . . . . . . . . . . 32
2.4 Fuel cell system efficiency . . . . . . . . . . . . . . . . . 33
2.4.1 FCS electric efficiency . . . . . . . . . . . . . . . 33
2.4.2 Auxiliary components . . . . . . . . . . . . . . . 34

c Alexander Kabza Fuel Cell Formulary

Contents 3

3 Example calculations 36
3.1 Stack operating parameters . . . . . . . . . . . . . . . . 39

4 Appendix 1: Fundamentals 41
4.1 Thermodynamic fundamentals . . . . . . . . . . . . . . 41
4.2 Energy and relevant energy units . . . . . . . . . . . . . 43
4.3 Temperature dependency of thermodynamic values . . 44
4.4 Standard temperature and pressure . . . . . . . . . . . 44
4.5 HHV and LHV . . . . . . . . . . . . . . . . . . . . . . . . 47
4.6 PolCurve parameter variation . . . . . . . . . . . . . . . 47

5 Appendix 2: Constant values and abbreviations 49

5.1 Relevant units based on SI units . . . . . . . . . . . . . 51
5.2 Used abbreviations . . . . . . . . . . . . . . . . . . . . . 51

6 Appendix 3: Air 52
6.1 Moist air . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
6.1.1 Water vapor saturation pressure . . . . . . . . . 52
6.1.2 Other formulas to express humidity of air . . . . 55
6.2 Energy and power of air . . . . . . . . . . . . . . . . . . 66
6.3 Air compression . . . . . . . . . . . . . . . . . . . . . . 67
6.3.1 Definitions . . . . . . . . . . . . . . . . . . . . . . 67
6.3.2 Thermodynamics . . . . . . . . . . . . . . . . . . 68
6.3.3 Ideal and real compressor . . . . . . . . . . . . . 73
6.3.4 Examples . . . . . . . . . . . . . . . . . . . . . . 74
6.3.5 Comment . . . . . . . . . . . . . . . . . . . . . . 75

7 Appendix 4: Fuel cell stack water management 76

7.1 Assumptions . . . . . . . . . . . . . . . . . . . . . . . . 76
7.2 Derivation . . . . . . . . . . . . . . . . . . . . . . . . . . 76

8 Appendix 5: FCS controls parameters 78

8.1 Fuel cell stack pressure drop . . . . . . . . . . . . . . . 78
8.2 Cathode dynamics - UNDER CONSTRUCTION *** . . . 79

c Alexander Kabza Fuel Cell Formulary

 4

1 Fundamentals
This is a collection of relevant formulas, calculations and equations
for designing a fuel cell system (FCS).

1.1 Thermodynamic of H2/O2

electrochemical device

In an H2 /O2 electrochemical device hydrogen is oxidized by oxygen to

water in an exothermic reaction:

1
H2 + O2 → H2 O(g) ∆r H < 0
2

The reaction enthalpy ∆r H is equal to the enthalpy of water forma-

tion ∆f H. The (chemical) energy content of any fuel is called heating
value (for more information on thermodynamics see chapter 4.1). The
heating value of hydrogen is equal to the absolute value of the reac-
tion enthalpy. Because product water is produced either as gaseous
or liquid phase, we distinguish between the lower heating value (LHV)
and the higher heating value (HHV) of hydrogen:

1 kJ
H2 + O2 → H2 O(g) − ∆f HH2 O(g) = LHV = 241.82
2 mol

c Alexander Kabza Fuel Cell Formulary

1.1 Thermodynamic of H2 /O2 electrochemical device 5

1 kJ
H2 + O2 → H2 O(l) − ∆f HH2 O(l) = HHV = 285.83
2 mol

Please note that LHV and HHV have positive signs whereas ∆H is
negative. All thermodynamic potentials are dependent on tempera-
ture and pressure, but are defined at thermodynamic standard condi-
tions (25◦ C and 100 kPa, standard ambient temperature and pressure
SATP, see also section 4.4).
The difference between LHV and HHV of 44.01 kJ/mol is equal to the
molar latent heat of water vaporization at 25◦ C.
The (thermodynamic) electromotive force (EMF) or reversible open
cell voltage (OCV) E 0 of any electrochemical device is defined as:

∆G
E0 =
n·F

where n is the amount of exchanged electrons and F is Faraday’s

constant (see chapter 5). For more information see chapter 4.1. For
the hydrogen oxidation or water formation n = 2. The free enthalpies
∆G of water formation is either

∆f GH2 O(g) = −228.57 kJ/mol or ∆f GH2 O(l) = −237.13 kJ/mol

Therefore the corresponding EMFs are:

−∆f GH2 O(g) −∆f GH2 O(l)
Eg0 = = 1.184 V El0 = = 1.229 V (1.1)
2F 2F

These voltages are the maximum voltages which can (theoretically)

be obtained from the electrochemical reaction of H2 and O2 . Beside
that it make sense to define two other voltages to simplify the effi-
ciency calculations (see section 1.7). If all the chemical energy of

c Alexander Kabza Fuel Cell Formulary

1.1 Thermodynamic of H2 /O2 electrochemical device 6

T [°C] -25 0 25 50 100 200 400 600 800 1000

LHV [kJ/mol] 241.3 241.6 241.8 242.1 242.6 243.5 245.3 246.9 248.2 249.3
HHV [kJ/mol] 287.4 286.6 285.8 285.0 283.4 280.0 265.8 242.3 218.8 195.3
G H2O(g) [kJ/mol] 230.8 229.7 228.6 227.5 225.2 220.4 210.3 199.7 188.7 177.5
G H2O(l) [kJ/mol] 245.4 241.3 237.1 233.1 225.2 210.0 199.0 230.4 267.1 308.3
300
290
280
HHV
270 LHV
Enthalpy [kJ/mol]

260 DG H2O(l)
DG H2O(g)
250
240
230
220
210
200
0 20 40 60 80 100 120 140 160 180 200
T [°C]

Figure 1.1: ∆H and ∆G as a function of temperature

hydrogen (i.e. its heating value) were converted into electric energy
(which is obviously not possible!), the following voltage equivalents
would result:

0 LHV 0 HHV
ELHV = = 1.253 V EHHV = = 1.481 V (1.2)
2F 2F
Those two values are the voltage equivalent to the enthalpy ∆H, also
called thermal cell voltage (see [1], page 27) or thermoneutral volt-
age.
The thermodynamic potentials ∆H and ∆G are dependent on tem-
perature; therefore also the corresponding voltage equivalents are
functions of temperature. The temperature dependency (of absolute
values) is shown in figure 1.1, values are calculated with HSC Chem-
istry 6.21.

c Alexander Kabza Fuel Cell Formulary

1.2 Nernst equation 7

1.2 Nernst equation

The theoretical cell porential or electromotive force (EMF) is not only

depending on the temperatue (as shwon in the previous section), but
also depending on the pressure. This dependency is in general de-
scribed by the Nernst equation. For the H2 /O2 electrochemical reac-
tion the Nernst equation is:
 
1/2
RT p p
 H2 O2 
E = E0 + ln (1.3)
2F pH 2 O

p is the partial pressure of H2 , O2 and H2 O.

Introducing a system pressure pSys and defining pH2 = α pSys , pO2 =
β pSys , pH2 O = γ pSys , the Nernst equation simplifies to:

αβ 1/2
!
RT
0 RT
E=E + ln + ln pSys (1.4)
2F γ 4F

If α, β and γ are constant, increasing the system pressure from p1 to

p2 influences the cell potential as follows:

RT p2
∆E = ln (1.5)
4F p1

Assuming a cell runs at 80◦ C, then doubling the system pressure re-
sults in a potential increase of ∆E = 7.6 mV ln 2 = 5.27 mV . It is
important to mention that due to other effects an increase in system
pressure results in much higher cell voltage increas than just given by
Nernst equation!

c Alexander Kabza Fuel Cell Formulary

1.3 Reactant consumption and feed 8

1.3 Reactant consumption and feed

The reactants H2 and O2 are consumed inside the fuel cell stack by
the electrochemical reaction. Based on the Faraday’s laws the molar
flows ṅ for reactant consumptions are defined as follows:

I ·N mol
ṅH2 = [ṅH2 ] = (1.6)
2F sec

I ·N 1 mol
ṅO2 = = · ṅH2 [ṅO2 ] =
4F 2 sec

I is the stack load (electric current, [I] = A), and N is the amount of
single cells (cell count, [N ] = −). 2 is again the amount of exchanged
electrons per mole H2 for the hydrogen oxidation; respectively 4 e-
per mole O2 for the oxygen reduction. F is the Faraday constant.
The stoichiometry λ defines the ratio between reactant feed (into the
fuel cell) and reactant consumption (inside the fuel cell). Due to fuel
cell design and water management issues etc. the stoichiometry must
always be more than one:

ṅfeed
λ= >1 [λ] = −
ṅconsumed

The reactant feed for H2 and air into the fuel cell stack are now defined
by an anode and cathode stoichiometry λ:

I ·N
ṅH2 , feed = ṅH2 · λanode = · λanode (1.7)
2F

ṅO2 I ·N
ṅair, feed = · λcathode = · λcathode
xO2 4F · xO2

c Alexander Kabza Fuel Cell Formulary

1.3 Reactant consumption and feed 9

where xO2 is the oxygen content in air.

The molar flows of unconverted (dry) reactants at the stack exhaust
are given as:

I ·N
ṅH2 , out = ṅH2 , feed − ṅH2 = · (λanode − 1)
2F

!
I ·N λcathode
ṅair, out = ṅair, feed − ṅO2 = · −1
4F xO2

It is very common to give the reactant feed either as a mass or volume

flow. The mass flow ṁ is the product of molar flow ṅ and molar weight
M:

g
ṁ = ṅ · M [ṁ] =
sec

The mass flow for stack anode and cathode inlet are therefore:

IN
ṁH2 , feed = · λanode · MH2 (1.8)
2F

IN
ṁair, feed = · λcathode · Mair (1.9)
4F xO2

The dry mass gas flow at stack cathode outlet contains less oxygen
than air and is calculated as:

!
IN λcathode
ṁcathode, out = ṁair, feed − ṁO2 , consumed = · · Mair − MO2
4F xO 2
(1.10)

c Alexander Kabza Fuel Cell Formulary

1.4 Hydrogen energy and power 10

The volume flow V̇ is calculated by the gas density ρ or the molar gas
volume V0, mol :

ṁ l
V̇ = = V0, mol · ṅ [V̇ ] =
ρ min

Both the gas density ρ and the molar gas volume V0, mol are tempera-
ture and pressure dependent! Therefor the referring temperature and
pressure need to be defined. N or n indicates that the volume flow
is normalized to a specific reference temperature and pressure (e.g.
[V̇ ] = Nl/min or ln /min). Physical standard temperature and pressure
(STP) is defined as 0◦ C and 101.325 kPa (for more information see
Nl
section 4.4). The molar gas volume at STP is V0, mol = 22.414 mol .
The amount of product water is equal to hydrogen consumption (equa-
tion 1.6) and given by:

I ·N mol
ṅH2 O, prod = ṅH2 = [ṅH2 O ] =
2F sec

The product water mass flow is:

I ·N g
ṁH2 O, prod = · MH2 0 [ṁH2 O ] = (1.11)
2F sec

1.4 Hydrogen energy and power

Hydrogen is a chemical energy carrier with a specific (chemical) en-

ergy density, which is defined by either HHV or LHV. The mass spe-
cific chemical energy is:

kJ
HHV 285.83 mol kJ kWh
WH2 , HHV = = g = 141.79 ≡ 39.39
M 2.02 mol g kg

c Alexander Kabza Fuel Cell Formulary

1.4 Hydrogen energy and power 11

kJ
241.82 mol kJ kWh
WH2 , LHV = g = 119.96 ≡ 33.32
2.02 mol g kg

The volume specific chemical energy (at STP) is:

kJ
HHV 285.83 mol J kWh
WH2 , HHV = = ln = 12.75 = 3.54 3
V0, mol 22.414 mol mln mn

kJ
241.82 mol J kWh
WH2 , LHV = ln = 10.79 = 3.00 3
22.414 mol mln mn

The chemical power of an H2 flow is:

J
PH2 , HHV = HHV · ṅH2 [PH2 , HHV ] = =W (1.12)
sec
Now the (chemical) power of H2 consumed in a stack (with N cells at
the stack load I) using equations 1.2 and 1.6 can easily be expressed
as:

PH2 , HHV consumed = 1.481 V · N · I (1.13)

And accordingly the (chemical) power of H2 feed with equations 1.2

and 1.7 is:

PH2 , HHV feed = 1.481 V · N · I · λanode (1.14)

c Alexander Kabza Fuel Cell Formulary

1.5 PolCurve (Current voltage dependency) 12

1.5 PolCurve (Current voltage dependency)

The current voltage dependency of a fuel cell is the most important

property to express the performance of a fuel cell. This dependency is
nonlinear due to kinetics of the electrochemical reaction of hydrogen
and oxygen; it’s called current voltage curve (I-U curve) or polarization
curve (PolCurve) or performance characteristic.

1.5.1 Kinetic of the H2 /O2 electrochemical device

The kinetics of electrochemical reaction is relevant as soon as the

electric circuit of any electrochemical device is closed and it starts
providing electric power. As soon as electrons and ions start moving
we get voltage losses (overvoltage). This voltage losses are nonlinear
functions of the load (or current density).
Voltage losses are the result of
• Oxygen reduction reaction (ORR) kinetics,
• Ohmic losses due to membrane ionic conductivity,
• Mass transport limitations at high load and
• Hydrogen oxidation reaction (HOR) kinetics (less relevant).

1.5.2 Butler-Volmer equation

The kinetics of ORR can be described with the Butler-Volmer equa-

tion. The Butler-Volmer equation defines the relation between current
density j and overpotential ϕ of any electrochemical reaction:

c Alexander Kabza Fuel Cell Formulary

1.5 PolCurve (Current voltage dependency) 13

8 0.5

6 0.4
ensity j [mA/cm²] / j / j0

0.3
4
0.2
2
0.1

j / j0
0 0.0
-100 -50 0 50 100 -10 -5 0 5 10
-2 -0.1
current de

-0.2
-4
-0.3
-6
-0.4
-8 -0.5
overpotential [mV] overpotential [mV]

(a) Large overpotential (b) Small overpotential

Figure 1.2: Butler Volmer

" !#
αnF (1 − α)nF
 
j(ϕ) = j0 exp ϕ − exp ϕ
RT RT

This is the sum of an anodic and cathodic current j(ϕ) = ja (ϕ) + jc (ϕ)
with:
!
αnF (1 − α)nF
 
ja (ϕ) = j0 exp ϕ jc (ϕ) = −j0 exp ϕ
RT RT

Side information: ex − e−x = 2 sinh x

With α = 1/2 the Butler-Volmer equation can be simplified to:

j(ϕ) 1 nF
 
= 2 sinh ϕ
j0 2 RT

Figure 1.2a shows the anodic current (green line), the cathodic cur-
rent (blue) and the Butler-Volmer equation itself (red). In case of low

c Alexander Kabza Fuel Cell Formulary

1.5 PolCurve (Current voltage dependency) 14

ln ( j / j0) 1

0
0 50 100 150 200 250 300
-1

-2

-3

-4
overpotential [mV]

Figure 1.3: Butler Volmer logarithmic scale

overpotentials (|ϕ| < 10mV , α = 1/2) the Butler-Volmer equation sim-

plifies to:

j(ϕ) nF
= ϕ
j0 RT

Figure 1.2b shows this linear relation between current density j and
overpotential ϕ, with the slope m = nF/RT . In case of high overpo-
tentials (|ϕ| > 100mV , α = 1/2) the Butler-Volmer equation simplifies
to:

j(ϕ) 1 nF
ln = ϕ
j0 2 RT

This relation becomes linear by using a logarithmic y-axes like shown

in figure 1.3. The slope of the blue dotted line m = 1/2 nF/RT is the
Tafel slope in the Tafel equation ϕ = m · j/j0 .

c Alexander Kabza Fuel Cell Formulary

1.5 PolCurve (Current voltage dependency) 15

1.0
cell voltage thermal power (HHV) thermal power (LHV)
0.9
800mV
0.8
U [V] / P [W/cm²] / efficiency [%]

700mV electric power

0.7
64% 600mV
0.6 56%
electric efficiency 500mV
(LHV) 48%
0.5
40%
0.4

0.3

0.2

0.1

0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
current density [A/cm²]

Figure 1.4: Polarization curve

1.5.3 Example PolCurve

The PolCurve plots the cell voltage versus current density of a single
cell or stack. Figure 1.4 shows an theoretical PolCurve, including the
electric and thermal power based on LHV or HHV (equation 1.17),
and the electric efficiency based on LHV (1.18).
For simulation and modeling it is helpful to have a mathematical ex-
pression for the PolCurves. The following empirical equation with
physical background describes the voltage current dependency quite
well (see [2], [3]):

E = E0 − b · (log j + 3) − Rj − m · expnj (1.15)

The parameters are:

• E0 : (measured) open cell voltage, [E0 ] = V

c Alexander Kabza Fuel Cell Formulary

1.6 Fuel cell stack power 16

• j: Current density, [j] = A/cm2

• b: Tafel slope (voltage drop due to oxygen reduction reaction),
[b] = V/dec
• R: Internal resistor, [R] = Ω cm2
• m, n: (Empirical) coefficients defining the voltage drop due to
diffusion limitation, [m] = V, [n] = cm2 /A
The example PolCurve in the chart 1.4 is calculated with the following
parameters: E0 = 1.0 V, b = 0.08 V/dec, R = 0.1 Ω cm2 , m = 0.0003 V,
and n = 3.30 cm2 /A.

1.6 Fuel cell stack power

Electric stack power Pel is the product of stack voltage and stack
load, also called gross power:

Pel = UStack · I = AveCell · N · I (1.16)

where AveCell = UStack /N is the average single cell voltage.

Thermal stack power Ptherm is that part of the consumed chemical

fuel power which is not converted into electric power:

Ptherm = PH2 , HHV − Pel

For the hydrogen fuel cell it is defined based on HHV as:

1.481 V
 
Ptherm, HHV = (1.481 V − AveCell) · N · I = Pel −1 (1.17)
AveCell

c Alexander Kabza Fuel Cell Formulary

1.6 Fuel cell stack power 17

The voltage equivalent 1.481 V is defined by equation 1.2. To calcu-

late the thermal power based on LHV this voltage equivalent needs
to be replaced by 1.253 V.

Recovered heat The recovered heat is that part of the thermal stack
power that is actually converted into usable heat. This is e.g. the heat
transferred into the (liquid) coolant and can be calculated as follows:

Precovered heat = V̇ · cp · ∆T · ρ

Coolant parameters are: Volume flow V̇ , heat capacity cp , tempera-

ture increase ∆T and density ρ.
Due to technical issues not all thermal power can be transferred to
coolant; therefore:

Precovered heat < Ptherm, HHV

The unusable or waste heat:

Pwaste heat = Ptherm, HHV − Precovered heat

This waste heat consists of two parts:

1. Heat rejected to ambient via stack surface (radiant heat).
2. Heat loss by cathode exhaust enthalpy; or more precise by the
difference between air outlet and air inlet enthalpy.

c Alexander Kabza Fuel Cell Formulary

1.7 Fuel cell efficiencies 18

1.7 Fuel cell efficiencies

1.7.1 Different efficiencies

There seems to be no commitment how to define or name the different

efficiencies for fuel cells. Therefore the only thing I can do here is to
clarify the differences; well knowing that in literature other definitions
and wordings are used. Don’t talk about efficiency if you don’t know
what you are talking about!
All efficiencies can be based on either LHV or HHV of the fuel. There
is also no commitment to use LHV or HHV. Therefore be careful and
keep in mind that the efficiency is higher if it refers to LHV!
In general the energy conversion efficiency η is defined as:
(useful) energy output (useful) power output
η= =
energy input power input

Thermodynamic efficiency The thermodynamic or maximum or

ideal efficiency is the ratio between enthalpy (or heating value) ∆H
and Gibbs free enthalpy ∆G (reflecting the maximum extractible work)
of any electrochemical device:

∆G
ηel, max =
∆H
For the hydrogen fuel cell it is:

−∆f GH2 O(g) E0 1.184 V

ηel, TD, LHV = = 0g = = 94.5%
LHV ELHV 1.253 V

−∆f GH2 O(l) E0 1.229 V

ηel, TD, HHV = = 0l = = 83.1%
HHV EHHV 1.481 V

c Alexander Kabza Fuel Cell Formulary

1.7 Fuel cell efficiencies 19

The Handbook of Fuel Cells [1] calls this simply ”ideal efficiency”.
The Fuel Cell Handbook [4] calls this efficiency the ”thermal efficiency
of an ideal fuel cell operating reversibly on pure hydrogen and oxygen
at standard conditions”.
Fuel Cell Systems Explained [5] calls this efficiency “the maximum ef-
ficiency possible” or “maximum efficiency limit” which explains it also
very well.

Electric efficiency The electric efficiency of a fuel cell (stack) is

defined as:

Pel
ηel =
Pfuel, consumed

Pel is the stack electric (gross) power and Pfuel, consumed is the con-
sumed fuel power (see equations 1.16 and 1.13).
The electric efficiency can easily expressed as (more details see sec-
tion 1.7.2):

AveCell AveCell
ηel, LHV = or ηel, HHV = (1.18)
1.253 V 1.481 V
How this efficiency is named in literature and codes & standards:
1. “Load efficiency” in the Handbook of Fuel Cells [1], refers to both
LHV and HHV.
2. The Fuel Cell Handbook [4] refers to HHV and says “this effi-
ciency is also referred to as the voltage efficiency”.
3. “Cell efficiency” in Fuel Cell Systems Explained [5], and refers
to LHV.

c Alexander Kabza Fuel Cell Formulary

1.7 Fuel cell efficiencies 20

4. SAE J2617 [6] calls this efficiency “stack sub-system efficiency”

and refers to LHV.

Fuel electric efficiency The fuel efficiency considers the amount

of hydrogen feed to the stack (and not only the amount of consumed
hydrogen). It is defined as:
Pel
ηfuel, el =
Pfuel, feed

Pel is the stack electric gross power and Pfuel, feed is the fuel feed power
(see equations 1.16 and 1.14).
According to equation 1.18 it is:

AveCell AveCell
ηfuel, el, LHV = or ηfuel, el, HHV =
1.253 V · λanode 1.481 V · λanode

Obviously, the relation between ηfuel, el and ηel is the anode stoic:

ηel
ηfuel, el =
λanode

How this efficiency is named in literature:

1. Fuel Cell Systems Explained [5] calls this efficiency simply “ef-
ficiency” (of either LHV or HHV).
2. Fuel Cell Handbook [4] calls it “net cell efficiency“ and defines
it as following: “To arrive at the net cell efficiency, the voltage
efficiency must be multiplied by the fuel utilization.”

c Alexander Kabza Fuel Cell Formulary

1.7 Fuel cell efficiencies 21

Voltage efficiency is the ratio between average and reversible cell

voltage E 0 (see equation 1.1):

AveCell
ηvoltage =
E0

Normally the voltage efficiency is based on ∆f GH2 O(l) :

AveCell AveCell
ηvoltage = 0
=
El 1.229 V

AveCell
ηvoltage = ηel, HHV · ηel, TD, HHV = · 83.1%
1.481 V

Thermal efficiency ηtherm can be calculated according the electric

efficiency:

Ptherm
ηtherm =
Pfuel, consumed

Based on HHV it is:

1.481 V − AveCell
ηtherm, HHV =
1.481 V

Recovered heat efficiency Due to the fact that not all thermal power
Ptherm is transferred into the coolant the recovery heat efficiency can
be calculated:

Precovered heat
ηrecovered heat =
Ptherm

c Alexander Kabza Fuel Cell Formulary

1.7 Fuel cell efficiencies 22

For HHV this is:

Precovered heat V̇ · cp · ∆T · ρ
ηrecovered heat, HHV = =
Ptherm, HHV (1.481 V − AveCell) · N · I

Overall stack efficiency can be calculated as:

Pel + Ptherm
ηoverall = =1
Pfuel, consumed

This (thermodynamic) efficiency needs to be ηoverall = 1, because the

consumed hydrogen is converted completely in electric and thermal
power:

Pfuel, consumed = Pel + Ptherm

Therefore it is also:

ηoverall = ηel + ηtherm

Total efficiency is maybe more reasonable and defined as:

Pel + Precovered heat

ηtotal = <1
Pfuel, feed

Obviously ηtotal < 1, because Precovered heat < Ptherm and Pfuel, feed >
Pfuel, consumed !

c Alexander Kabza Fuel Cell Formulary

1.7 Fuel cell efficiencies 23

1.7.2 Electric efficiency calculation

Pel AveCell
How to convert ηel, LHV = Pfuel, consumed
to 1.253 V
?
With equations 1.2 and 1.6 it is:

Pel
ηel, LHV = Pfuel, consumed
AveCell·N ·I
= ṅH2 ·LHV
= AveCell·2F
LHV
AveCell AveCell
= 0
ELHV
= 1.253 V

c Alexander Kabza Fuel Cell Formulary

 24

2 Fuel cell system (FCS)

IEC 62282-1:2005 [7] defines the main components of a fuel cell sys-
tem and draws the system boundary around, see figure 2.1.
The main components are the fuel cell stack, an air processing sys-
tem (e. g. a compressor and humidifier), the fuel gas supply, thermal
management system etc. Other peripheral components are valves,
pumps, electrical devices etc. Fuel like hydrogen (from a tank sys-
tem) and air (ambient) are feed into the FCS.
Main FCS inputs are fuel (e.g. hydrogen), air and water, but also
electric and thermal power. Main FCS outputs are the electric net
and thermal power, the reactant exhaust gases and product water.
The (electric) gross power is the stack electric power. The difference
between gross and net is the electric power consumed by the periph-
eral electric components inside the FCS:

Pnet = Pgross − Pparipheral

2.1 Anode subsystem

Hydrogen is feed into the fuel cell system from a tank system. The
consumed hydrogen inside the fuel cell stack is based on the Faraday
equation (equation 1.6) and dependent on the stack load. Actually

c Alexander Kabza Fuel Cell Formulary

2.1 Anode subsystem 25

Power inputs Recovered

Electrical heat
thermal Thermal
Management Waste heat
Fuel System
Fuel Processing
System Usable power
Fuel Cell Power Electrical (net)
Module (Stack) Conditioning
Air System
Air Processing
System
Water and/or Internal power Discharge
Ventilation needs water
Byproduct
Inert gas Management
Exhaust gases,
Water ventilation
Ventilation Automatic Onboard
System Control Energy
EMD, System storage
vibration,
wind, rain, EMI, noise,
temperature, etc. vibration
Source: IEC 62282‐1:2005 ‐ Fuel cell technologies, Part 1: Terminology

Figure 2.1: Fuel Cell System according to IEC 62282-1

Drawings
always more for the document
hydrogen „Fuel cells
is feed intoformulary“
the FCS than is consumed inside
Created: Alexander Kabza; May 28, 2008
the stack.Last modification: Alexander Kabza; July 14, 2015
The anode flow shall on one hand beFuel high enough to remove product
cell system
water, but shall on the other hand be low to increase efficiency. This
H 2

dilemma is often solved by anAiranode recirculation

Compressor
Fuelloop (see figure
cell
2.2). Coolant T
stack
T in out
Periphery

The anode recirculation is either real-

Recirculation pump
ized with an Anode Recirculation Pump
Purge valve
(ARP) or by one or more jet pumps. A
H Fuel Cell 2
water trapp shall remove liquid water at Stack
Water trap
the anode exhaust.
p
The purge valve anode

opens with a specific frequency

2 and du-
p 2 Figure 2.2: Anode loop
ration to remove inert gases accumu-
lated in the anode
p
loop (mainly nitrogen
1
1
from the cathode side). The anode loop ensures a high anode gas
V2 V1 V

c Alexander Kabza Fuel Cell Formulary

2.2 Cathode subsystem 26

flow (e.g. λgross > 2) through the stack and allows a nearly complete
utilization of hydrogen (λnet  1.1). This stack anode off gas is finally
purged either into the cathode exhaust or directly out of the FCS. This
purged hydrogen is obviously lost for the electrochemical conversion,
therefore the amout of anode purge losses shall be minimized.
The relation between hydrogen feed versus consumed hydrogen is
called FCS fuel stoichiometry:

fuel feed H2 feed

λnet = λfuel,FCS = = ≥1
fuel consumed H2 consumed

The fuel efficiency (or fuel utilization) of a FCS is defined as:

ηfuel,FCS = 1/λfuel,FCS

The anode recirculation loop and its anode purge controls needs to be
optimized regarding high fuel efficiency or low fuel losses. By doing
this the fuel stoichiometry can be close to 1. One controls concept
is the Amperehour counter. The anode purge is triggered when the
integral of stack load reaches an Ah-threshold of e.g. 50 Ah. If the
stack runs at 300 A constant load there is one purge every 10 min,
calculated as following:

300 A 300 A
= = 0.00166 s−1 = 0.1 min−1
50 Ah 180, 000 As

2.2 Cathode subsystem

An air processing system or cathode submodule is e.g. a compressor

and a gas-to-gas humidifier. Its purpose is to feed the conditioned
(humidified) and required amount of air into the stack.

c Alexander Kabza Fuel Cell Formulary

2.2 Cathode subsystem 27

Ambient air has a specific relative humidity. Depending on the FCS

design, this water input into the FCS may be neglected. More impor-
tant is the pressure drop inside the stack and inside the FCS, and
therefore the absolute pressure after the compressor and before the
stack. The cathode stoichiometry is dependent on the system power
level; typical values are between 1.8 (at higher load) to 3 (or more at
low load).

2.2.1 Cathode pressure controls

In case of (low temperature) PEM fuel cell stacks it is essential to re-

move all product water from the stack. The stack operating conditions
to fulfill this requirement can be calculated. Due to the fact that most
water is removed via cathode exhaust; therefore we assume no water
removal through anode exhaust. Based on this assumption we get
the following equation to calculate the specific humidity at cathode
exhaust (more details see chapter 7):

λ Mwater
x
· Yin + 2 · Mair
Yout = λ MO2 (2.1)
x
− Mair

Yin is the specific humidity at cathode inlet; according to equation 6.3

Y is a function of temperature, pressure and relative humidity at stack
cathode inlet.
This equation allows to answer the following question: Under which
stack operating conditions is water removal (via cathode side) maxi-
mized? Answer: The maximum amount of water can be removed if
a) the inlet air is dry, b) under minimum pressure and c) is cathode
exhaust is fully saturated.
• Criterion a) means no water input and therefore Yin = 0 kg/kg.
Now the specific humidity at cathode outlet Yout (or Y.So.C) just

c Alexander Kabza Fuel Cell Formulary

2.2 Cathode subsystem 28

 [‐] 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5
Y.So.C [kg/kg] 0.224 0.206 0.191 0.178 0.167 0.157 0.148 0.140 0.133 0.126 0.120 0.115
T.So.C [°C] 66.6 65.2 63.9 62.7 61.6 60.6 59.5 58.6 57.8 56.8 56.0 55.2

Table 2.1: Specific humidity and temp. at stack outlet versus cathode
stoichiometry

depends on cathode stoichiometry λ (all other parameters are

constant).
• Criterion b) menas lowest possible pressure at cathode exhaust,
which is of course ambient pressure, and therefore p.So.C =
101.325 kPa.
• Criterion c) means relative humidity at cathode exhaust shall be
RH.So.C = 100% (fully saturated at cathode exhaust temper-
ature T.So.C). In many fuel cell applications (e.g. automotive)
the product water exits the stack as vapor and liquid water. Nev-
ertheless this assumption allows first Fuel Cell System design
considerations.
With these three criteria we get the dependencies shown in table 2.1.
This table now defines the lowest stack cathode exhaust temperature
T.So.C under which (theoretically) all product water can be fully re-
moved as a function of cathode stoichiometry λ. If this temperature
gets lower, the cathode stoichiometry must be increased to remove
product water.
What happens if this temperature increases? One one hand the cath-
ode stoichiometry may be decreased; but this is of course not doable
for temperatures above 65◦ C corresponding to cathode stoichiometry
below λ < 1.5. On the other hand either the inlet air may be humidi-
fied, or the cathode exhaust pressure may be increased, or both.
The dependency of cathode outlet pressure and temperature is also
given by equation 2.1. The chart in figure 2.3 shows this pressure-
temperature-dependency at three different stack operating conditions.

c Alexander Kabza Fuel Cell Formulary

2.2 Cathode subsystem 29

5
x 10
4
In: 70°C, 150kPa, RH 50%, CStoic 2.0
CStoic 1.8
3.5 RHin 80%

2.5
pout [Pa]

1.5

0.5
70 75 80 85 90 95 100
Tout [°C]

Figure 2.3: Cathode pressure controls as a funtion of temperature

This chart shows the cathode exhaust pressure (pout or p.So.C) as a

function of cathode exhaust temperature (Tout or T.So.C). p.So.C is
calculated based on equation 6.5 as:
Mw
Mair
+ Yout
pout = · ϕout · ps (tout ) (2.2)
Yout

This cathode exhaust pressure (pout or p.So.C) in a function of stack

cathode inlet RH, temperature and pressure, cathode stoichiometry
λ, and cathode outlet temperature tout and RH ϕout .
Further optimization of stack operating conditions is needed, but this
may be a reasonable starting point for FCS pressure controls delop-
ment. Especially as long as the stack temperature did not reach its
desired operating temperature, product water needs to be removed
also as liquid water! This is e.g. the case for automotive applica-
tion during cold ambient temperatures (cold startup conditons). Un-
der certain conditions the stack never warms up enough to reach the
temperature needed for full vaporization!

c Alexander Kabza Fuel Cell Formulary

2.2 Cathode subsystem 30

The typically fully humidified cathode exhaust gas of a FCS contains

a lot of energy in form of enthalpy (see also section 6.2).
The increasing enthalpy of process air inside the FCS cathode sub-
system considers first the temperature increase (from cold inlet to
warm outlet) and second the product water uptake.
As a fist guess the energy loss via stack (not FCS!) cathode exhaust
seems to be the difference between thermal power based on HHV
and LHV. But in fact this energy loss is even higher! This is due to
the fact that the air temperature increases inside the stack, while LHV
and HHV are based on the thermodynamic standard temperature Tc .

2.2.2 Stack pressure drop

In the previous section the pressure at stack exhaust was calculated.

To also calculate the pressure at stack inlet the cathode pressure drop
through the stack is needed. Any gas flow through a pipe causes a
specific pressure drop as a function of gas flow. A simplified assump-
tion for the cathode air pressure drop of an air flow through a stack
can be made with the orifice equation (more details see chapter 8).
For a given stack coefficient and cross sectional area αA (obtained
by one mearured pressure drop), the pressure drop ∆p can be calcu-
lated as a function of air mass flow ṁ:

q
p2 − p0 (αA ρ0 )2 · (p0 + p2 )2 + 2 ṁ2 p0 ρ0
∆p = + − p2 ∆p ∝ ṁx
2 2 αA ρ0
(2.3)
Here p2 is the gauge pressure at the stack exhaust, and p0 and ρ0 are
absolute pressure and air density at STP.

c Alexander Kabza Fuel Cell Formulary

2.2 Cathode subsystem 31

2.2.3 Cathode air compression

The theoretical work for gas compression can be calculated as fol-

lowing (for detailed information see chapter 6.3):

 κ−1
 p2
wcomp = cp, m T1 · Π κ −1 Π=
p1

The gas is compressed from pressure p1 to pressure p2 , temperature

before compression is T1 . Constant values for air: Heat capacity at
constant pressure cp, m, air = 1005.45 kgJK , (κ − 1)/κ = 0.285.
The units for the compression work can be expressed as follows:

kJ W
[wcomp ] = =
kg g/sec

For dry air at room temperature (25◦ C) this compression work wcomp
as a function of pressure ratio Π looks as follows:
Finally the required power for compression is:

Pcomp = ṁ · wcomp

Example: The theoretical (isentropic) power request to compress

an (dry) air mass flow of 92 g/sec (from ambient temperature and pres-
sure) to 250 kPaabs is:

W
Pcomp = ṁ · wcomp = 92 g/sec · 88 = 8.1 kW
g/sec

c Alexander Kabza Fuel Cell Formulary

2.3 Coolant subsystem 32

160
143
140
Power per mass flow [W / g/sec]

127

120
Isentropic work [kJ/kg]

109

100 88

80
65
60
36
40
19
20
0
0
1.0 1.5 2.0 2.5 3.0 3.5 4.0
Pressure ratio  [‐]

Figure 2.4: Isentropic work for air compression

2.3 Coolant subsystem

Coolant (typically liquid, e.g. DI water or WEG) is feed by the coolant

pump into the fuel cell stack. Other peripheral components (e.g. elec-
tric converters, compressor, etc.) inside the FCS may also be cooled
in parallel to the stack or serial before or after the stack.
The coolant enters the stack at the stack inlet temperature and leaves
the stack at the stack outlet temperature; leading to a specific temper-
ature increase ∆T depending on stack load. Typical the temperature
increase is ∆T ≤ 10 K. According to given ∆T requirements depend-
ing on stack load the coolant flow needs to be a function of the stack
load, too.
The required coolant flow rate V̇ at a given stack recovered heat
Precovered heat and coolant ∆T is calculated as follows:

Precovered heat
V̇ = [V̇ ] = l/min
cp · ∆T · ρ

c Alexander Kabza Fuel Cell Formulary

2.4 Fuel cell system efficiency 33

Temp. 0°C 20°C 40°C 60°C 80°C

cp ρ cp ρ cp ρ cp ρ cp ρ
ratio
kJ/kg K g/cm³ kJ/kg K g/cm³ kJ/kg K g/cm³ kJ/kg K g/cm³ kJ/kg K g/cm³
0% 4.218 0.9998 4.182 0.9982 4.179 0.9922 4.184 0.9832 4.196 0.9718
10% 4.09 1.018 4.10 1.011 4.12 1.002 4.13 0.993 4.15 0.983
30% 3.68 1053 3.72 1044 3.76 1034 3.80 1023 3.84 1010
50% 3.28 1.086 3.34 1.075 3.40 1.063 3.46 1.050 3.53 1.036

Table 2.2: Coolant properties

where cp is the specific heat capacity, ∆T = Tcool, out − Tcool, in and ρ

the coolant density.
Table 2.2 shows the specific heat capacity cp and density ρ of WEG
at different temperatures and mixing ratios (vol%). 0% is pure water
here.
This means that the required coolant volume flow for WEG (50vol%,
60◦ C) needs to be 12% higher compared to pure DI water.

2.4 Fuel cell system efficiency

2.4.1 FCS electric efficiency

The overall FCS efficiency considers the fuel, electric and thermal
input and electric and thermal output (see [8], 5.1.9.1):

electric and thermal output

ηFCS =
fuel and thermal input

When the thermal output is used in the vehicle and the electric and
thermal inputs are negligible, this equation simplifies to:

Pnet output
ηFCS, el =
Pfuel input

c Alexander Kabza Fuel Cell Formulary

2.4 Fuel cell system efficiency 34

“Fuel power input” Pfuel input is the fuel (chemical) energy content fed
into the FCS (see equation 1.12). It can be expressed based on LHV
or HHV; both [1] and [8] refer to LHV.
The electric net power output Pnet output is the stack gross power minus
the total electric power consumption of all auxiliary FCS components
(e.g. compressor or blower, pumps, valves, etc.):

X
Pnet = Pgross − Paux Paux = Pcomponent i
i

FCS efficiency can also be expressed as the product of different (sys-

tem related) efficiencies (see also [1], page 718):

ηFCS el = ηstack · ηfuel · ηperipheral

with

AveCell 1 Pnet
ηstack = , ηfuel = and ηperipheral =
1.253 V λfuel Pgross

2.4.2 Auxiliary components

The power consumption Paux of the auxiliary components inside the

FCS is typically not constant; but depending on the FCS power level.
Depending on the process air (cathode) pressure drop of the entire
FCS, the air compression process maybe the most dominant power
consumer.
The power consumption for air compression can be estimated with
the following given parameters (see also chapter 6.3):
• Stack load [A] and cell count

c Alexander Kabza Fuel Cell Formulary

2.4 Fuel cell system efficiency 35

• Cathode stoichiometry λcathode (I) (as a function of stack load)

• Compressor pressure ratio Π(I) as a function of stack load, inlet
temperature (constant)
• Compressor efficiency ηcompressor (Π, ṁ), as a function of pres-
sure ratio and air mass flow (see compressor efficiency map)

I ·N
ṁ = ṁair (I) = · λcathode (I) · Mair
4F · xO2

 κ−1
 1
Pcompressor = ṁ · cp, m T1 · Π(I) κ −1 ·
ηcompressor (Π, ṁ)

Under the assumption that the power consumption of all other auxil-
iary components is constant, the total aux power consumption can be
estimated.

c Alexander Kabza Fuel Cell Formulary

 36

3 Example calculations
Hydrogen flow power: A hydrogen flow transports (chemical) power.
According to equation 1.12 the hydrogen flow of e.g. 1 g/sec is equiv-
alent to:

g ln m3
1 ≡ 667.1 ≡ 40.0 n ≡ 120.0 kWLHV ≡ 141.8 kWHHV
sec min h

See chapter 1.4 for more information.

Stack efficiency calculation: A fuel cell stack produces 10 kW elec-

tric power, the hydrogen feed is 150 ln /min, and the stack runs at
600 mV average cell voltage. Calculate the electric efficiency, the fuel
efficiency and the anode stoichiometry (based on LHV).

AveCell 0.6 V
ηel, LHV = 0
= = 47.9%
ELHV 1.253 V

ln
150 min mol
ṅH2 feed = ln sec = 0.1115
22.414 mol · 60 min sec

kJ
Pel 10 sec
ηfuel, el, LHV = = = 37.1%
ṅH2 feed · LHV 0.1115 mol
sec
· 241.82 kJ
mol

c Alexander Kabza Fuel Cell Formulary

 37

ηel, LHV 47.9%

λanode = = = 1.29
ηfuel, el, LHV 37.1%

Average cell voltage and hydrogen consumption: A 1 kW fuel

cell stack is running at 50% electric efficiency and 40% fuel electric
efficiency. What is the average cell voltage, and how much hydrogen
is feed to the stack (based on LHV)?
According to equation 1.18 the electric efficiency is 50% at an aver-
age cell voltage of:

1.253 V · 50% = 0.6265 V

At 1 kW electric power and 40% fuel electric efficiency the hydrogen

feed is equivalent to 2.5 kW (chemical power).

Fuel cell system efficiency Here is an example for a FCS effi-

ciency estimation. The auxiliary power consumption is calculated as
follows:
X
Paux = Pcompressor + Pcomponent i
i

Pcomponent i is the summary of all other electric consumers inside

P
i
the FCS, beside the compressor. This value can be assumed to be
constant, e.g. 1 kW here in this example.
In this example the compressor power Pcompressor is a function of cath-
ode stoichiometry, cathode inlet pressure (pressure ratio Π), and com-
pressor efficiency. The cathode stoic λcathode is 5.0 to 2.0 for low load

c Alexander Kabza Fuel Cell Formulary

 38

100

90

80 gross
net
70 aux
Electric power [kW]

60

50

40

30

20

10

0
0 50 100 150 200 250 300 350 400 450
Stack load [A]

Figure 3.1: FCS power

and 1.8 above 0.15 A/cm2 . The cathode inlet pressure is linear be-
tween 130 kP a at low and 200 kP a at high load. The compressor effi-
ciency is linear between 70% at low and 80% at high air mass flow.
Based on the polarization curve shown in chapter 1.5, and using the
above parameter to calculate Paux , we get the electric power showed
in figure 3.1:
With an assumed net anode stoichiometry of 1.3 at low and 1.05 at
higher loads above 0.15 A/cm2 , the FCS efficiencies can be calcu-
lated as shown in figur 3.2.
The FCS peak efficiency is 55% for this example. The zero power
level is called idle point; here the FCS net power is zero, but the FCS
is still operating and consumes fuel.

c Alexander Kabza Fuel Cell Formulary

3.1 Stack operating parameters 39

70% stack el. eff. (gross) [%]

stack fuel el. eff. (net) [%]
60% FCS_net

50%
Efficiency [%]

40%

30%

20%

10%

0%
0 10 20 30 40 50 60 70 80 90
FCS net power [kW]

Figure 3.2: FCS efficiency

3.1 Stack operating parameters

Example H2 /air fuel cell stack with 200 cells and 300 cm2 active area:

c Alexander Kabza Fuel Cell Formulary

3.1 Stack operating parameters 40

Stack electric
Current density * A/cm2 0.05 0.10 0.20 0.40 0.60 1.00
Ave cell voltage * mV 839 797 756 703 661 588
Current (gross) A 15 30 60 120 180 300
Voltage V 168 159 151 141 132 118
Power (gross) kW el 3 5 9 17 24 35

Stack anode
Stoichiometry * - 3.0 2.0 1.3 1.3 1.3 1.3
Hydrogen flow Nl/min 62.7 83.6 108.7 217.4 326.2 543.6
Hydrogen flow g/sec 0.09 0.13 0.16 0.33 0.49 0.81

Stack cathode in
Stoichiometry * - 2.2 2.0 1.8 1.8 1.8 1.8
Air flow Nl/min 109.5 199.1 358.4 716.8 1075.2 1792.1
Air flow g/sec 2.4 4.3 7.7 15.4 23.2 38.6
Air flow kg/h 8.5 15.4 27.8 55.6 83.4 138.9

Stack thermal
P therm LHV kW 1.2 2.7 6.0 13.2 21.3 39.9
P therm HHV kW 1.9 4.1 8.7 18.7 29.5 53.6
delta HHV - LHV kW 0.7 1.4 2.7 5.5 8.2 13.7
Coolant inlet temp * °C 60 60 60 60 60 60
Coolant outlet temp * °C 61 63 66 67 68 70
Coolant delta T K 1 3 6 7 8 10
Coolant flow LHV 2 l/min 17.9 13.1 14.3 27.1 38.3 57.4
Coolant flow HHV 2 l/min 27.7 19.7 20.8 38.3 53.1 77.0

Stack cathode out

RH * % 100% 100% 100% 100% 100% 100%
pws at coolant outlet temp kPa abs 20.84 22.83 26.12 27.30 28.52 31.12
Pressure out kPa abs 120 120 125 131 137 149
Product water out g/sec 0.3 0.6 1.1 2.2 3.4 5.6
Consumed oxygen g/sec 0.2 0.5 1.0 2.0 3.0 5.0
Dry cathode out g/sec 2 4 7 13 20 34
Wet cathode out g/sec 2 4 8 16 24 39
Enthalpy out kJ/kg 405 445 497 499 500 503
P enthalpy out kW 1 2 3 7 10 17

Stack power in (hydrogen)

Hydrogen feed mol/sec 0.05 0.06 0.08 0.16 0.24 0.40
P chem H2 LHV kW 11 15 20 39 59 98
P chem H2 HHV kW 13 18 23 46 69 116

Stack efficiency (LHV)

Electric efficiency % 67% 64% 60% 56% 53% 47%
Fuel efficiency % 22% 32% 46% 43% 41% 36%

* given values
1
20°C and 50% compressor efficiency
2
WEG 50vol% at 60°C

Table 3.1: Example stack operating conditions

c Alexander Kabza Fuel Cell Formulary

 41

4 Appendix 1: Fundamentals

4.1 Thermodynamic fundamentals

The four thermodynamic potentials are:

Internal energy U is the total energy of an thermodynamic system
and the sum of all forms of energies. In an isolated system the
internal energy is constant and can not change. As one result of
the first law or thermodynamics the change in the inner energy
can be expressed as the heat dq supplied to the system and the
work dw performed by the system: dU = dq + dw
Free or Helmholtz energy F (or A): is the ”useful” work obtainable
from a thermodynamic system at isothermal and isobaric con-
ditions.
Enthalpy H: The enthalpy is a measure for the energy of a thermo-
dynamic system and the sum of internal energy and the work
for changing the volume: H = U + pV . For each chemical reac-
tion the entropy change defines the reaction to be endothermic
(∆H < 0) or exothermic (∆H > 0).
Free or Gibbs enthalpy G: The Gibbs or thermodynamic free en-
thalpy is the work that a thermodynamic system can perform:
G = H − T S. For each chemical reaction ∆G defines the reac-
tion to be exergonic (∆G < 0) or endergonic (∆G > 0).

c Alexander Kabza Fuel Cell Formulary

4.1 Thermodynamic fundamentals 42

The thermodynamic (or Guggenheim) square is a mnemonic for the

relation of thermodynamic potentials with each other.

−S U V
H A
−p G T

Table 4.1: Thermodynamic Square

How to get the total differential on a thermodynamic potential?

1. Select one thermodynamic potential in the middle of one border,
e.g. U .
2. Take the two coefficients in the opposite edge, this is −p and T
for U . As a first step we get: dU = −p{} + T {}
3. Take now the opposite edge of both coefficients, here −S is
opposite to T and V is opposite to −p. Put those in the brackets
to get the following result here: dU = −pdV + T dS
The following total differentials of the four thermodynamic potentials
can be determined:

dU = −pdV + T dS dH = V dp + T dS

dA = −SdT − pdV dG = V dp − SdT

How to determine the Maxwell relations?

1. Select two units in the edge of one border, e.g. T and V .
2. Select the corresponding units in the opposite border, here it is
−p and −S.
3. Now the Maxwell relation is here: ∂T /∂V = −∂p/∂S

c Alexander Kabza Fuel Cell Formulary

4.2 Energy and relevant energy units 43

The following four Maxwell relations can be determined:

(∂T /∂V )S = −(∂p/∂S)V

(∂T /∂p)S = (∂V /∂S)p

(∂p/∂T )V = (∂S/∂V )T
(∂V /∂T )p = −(∂S/∂p)T

And finally the following dependencies can be determined by the

Guggenheim square:

(∂U/∂S)V = T and (∂U/∂V )S = −p

(∂H/∂S)p = T and (∂H/∂p)S = V

(∂A/∂V )T = −p and (∂A/∂T )V = −S
(∂G/∂p)T = V and (∂G/∂T )p = −S

4.2 Energy and relevant energy units

Energy is the product of power and time: E = P · t

Units for energy:

[E]= J = kg m2 /s2 = Nm = Ws

Joules and kilowatt-hour:

1 kWh = 1000 Wh ≡ 3.600.000 Ws = 3.600.000 J = 3.600 kJ = 3.6 MJ

c Alexander Kabza Fuel Cell Formulary

4.3 Temperature dependency of thermodynamic values 44

T 0 25 50 100 200 300 400 500 600

cp [J/molK] 29.06 29.10 29.13 29.25 29.67 30.25 30.93 31.64 32.29
cp [J/kgK] 1003.4 1004.6 1005.9 1009.9 1024.4 1044.5 1067.9 1092.3 1114.8

Table 4.2: Air heat capacity versus temperature

4.3 Temperature dependency of

thermodynamic values

Heat capacity of air cp, air of air as a function of temperature (◦ C),

calculated with HSC Chemistry 6.21, air with 78 Vol% N2 , 21% O2
and 1% Ar, ist shown in table 4.2.

Heat capacity of other gases cp of different gases as a function

of temperature (◦ C), calculated with Calculated with HSC Chemistry
6.21, shown in table 4.1:

4.4 Standard temperature and pressure

Standard temperature and pressure (STP) is needed for many fuel

cell related calculations; e.g. to calculate the volumetric reactant flow
for a stack. Therefore it seems to be a simple question to ask for the
correct reference temperature and pressure for STP. But the answer
is not simple at all, as the following table shows [9]:

c Alexander Kabza Fuel Cell Formulary

4.4 Standard temperature and pressure 45

Molar heat capacity cp [J/molK] vs t [°C]

Species 0 25 50 100 200 300 400 500 600
N2(g) 29.12 29.13 29.14 29.19 29.47 29.95 30.57 31.26 31.92
O2(g) 29.26 29.38 29.50 29.87 30.83 31.83 32.76 33.56 34.21
H2(g) 28.47 29.05 28.90 28.75 28.77 28.99 29.29 29.64 30.00
CO2(g) 36.04 37.12 38.31 40.39 43.81 46.61 48.97 50.97 52.62
H2O(g) 33.56 33.60 33.71 34.04 34.97 36.07 37.25 38.48 39.74
Specific heat capacity cp [J/kgK] vs t [°C]
Species 0 25 50 100 200 300 400 500 600
N2(g) 1039.3 1039.7 1040.3 1042 1052 1069 1091 1116 1140
O2(g) 914.4 918.2 922.0 934 963 995 1024 1049 1069
H2(g) 14124 14409 14335 14260 14271 14380 14531 14703 14884
CO2(g) 818.9 843.5 870.4 918 995 1059 1113 1158 1196
H2O(g) 1863.1 1865.3 1871.0 1889 1941 2002 2068 2136 2206
50 CO2(g)
H2O(g)
45 O2(g)
N2(g)
40 H2(g)
cp [JJ/mol K]

35

30

25

20
0 50 100 150 200
T [°C]

Figure 4.1: Molar and specific heat capacity of different gases

c Alexander Kabza Fuel Cell Formulary

4.4 Standard temperature and pressure 46

T [◦ C] / [K] pabs [kPa] Publishing or establishing entity

0 / 273.15 100.000 IUPAC (present)
0 / 273.15 101.325 IUPAC (former), NIST, ISO 10780
15 / 288.15 101.325 DIN IEC 62282-3-2 [10]
20 / 293.15 101.325 EPA, NIST
25 / 298.15 101.325 EPA
25 / 298.15 100.000 SATP
Thermodynamic values are defined at 25◦ C and 100.000 kPa. There-
fore SATP (standard ambient temperature and pressure) is used to
express that within this document.

Which “standard” is to choose now? To calculate a gas volume

flow V̇ from a molar flow ṅ or mass flow ṁ, the reference temperature
and pressure need to be defined, also is mass flow controllers are
used!
The purpose of Mass flow controllers (MFCs) is to feed a required liq-
uid or gas flow. From their measurement principle those components
do measure a mass flow (e.g. kg/h or g/sec). But often volumetric val-
ues (e.g. l/min or m3 /h) are more common. To transfer gas mass flows
into gas volumetric flows, the gas density (kg/m3 ) must be known.
Since the gas density itself is dependent on temperature and pres-
sure, a reference temperature and pressure is needed. Well known
MFCs manufacturers like Brooks, Bronkhorst, Bürckert, Aera or Omega
do use the following temperature and pressure as reference for stan-
dard volume flows:

T0 = 273.15 K and p0 = 101.325 kPa

Using such types of MFCs require using this reference temperature

and pressure to ensure correct calculations like described above!

c Alexander Kabza Fuel Cell Formulary

4.5 HHV and LHV 47

By the way The well known molar volume of an ideal gas is given
at this temperature and pressure:

T0
V0, mol = R · = 22.414 ln /mol
p0

4.5 HHV and LHV

The higher heating value (HHV) of any fuel can be measured within
a calorimeter. The result of this measurement is the energy of the
chemical (oxidation) reaction starting at 25◦ C and ending at 25◦ C;
therefore product water is condensed completely. Due to that the
HHV includes the latent heat of water vaporization; but the specific
amount of product water produced by the fuel oxidation reaction needs
to be considered.
The lower heating value (LHV) itself can not be measured directly;
therefore the LHV needs to be calculated from HHV minus latent heat
of water vaporization.

4.6 PolCurve parameter variation

Charts 4.2 show parameter variations for the empirical PolCurve equa-
tion 1.15. The baseline parameters are: E0 = 1.0 V, b = 0.08 V/dec,
R = 0.1 Ω cm2 , m = 0.0003 V, and n = 3.30 cm2 /A.

c Alexander Kabza Fuel Cell Formulary

4.6 PolCurve parameter variation 48

Variation E0: 1.05 − 1.0 − 0.95 V Variation b: 0.06 − 0.08 − 0.1 V/dec
1 1

0.8 0.8
Cell voltage [V]

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0 0.5 1 1.5 2 0 0.5 1 1.5 2

Variation R: 0.07 − 0.01 − 0.13 Ω cm2 Variation m: 1e−4 − 3e−4 − 5e−4 cm2/A
1 1

0.8 0.8
Cell voltage [V]

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0 0.5 1 1.5 2 0 0.5 1 1.5 2
2
Current density [A/cm ] Current density [A/cm2]

Figure 4.2: Polarization curve model parameter variation

c Alexander Kabza Fuel Cell Formulary

 49

5 Appendix 2: Constant
values and abbreviations
Constant values
Faraday’s constant F = 96 485.3399 Amol
sec
[11]
J
Molar gas constant R = 8.314472 K mol [11]
Magnus equation const C1 = 610.780 Pa
C2 = 17.08085
C3 = 234.175 ◦ C
Thermodynamics
Standard temp. and pressure T0 = 273.15 K [12]
(STP, see section 4.4) p0 = 101.325 kPa [12]
Molar volume of ideal gases
(at STP) V0, mol = RT0 /p0 = 22.414 ln /mol
Standard ambient T and p Tc = 298.15 K [13]
(SATP) pc = 100.000 kPa [13]
kJ
LHV of H2 (at SATP) −∆HH2 O(g) = 241.82 mol [13], [14]
=119.96
ˆ kJ/g [15]
kJ
HHV of H2 (at SATP) −∆HH2 O(l) = 285.83 mol [13], [14]
=141.79
ˆ kJ/g [15]
kJ
Gibbs free enthalpy of H2 O(g) −∆GH2 O(g) = 228.57 mol [13]
kJ
and H2 O(l) (at SATP) −∆GH2 O(l) = 237.13 mol [13]

c Alexander Kabza Fuel Cell Formulary

 50

Molar weights
Molar weight water Mwater = 18.0153 g/mol
hydrogen MH2 = 2.01588 g/mol [11]
oxygen MO2 = 31.9988 g/mol [11]
nitrogen MN2 = 28.01348 g/mol [11]
CO2 MCO2 = 44.0095 g/mol [11]
air Mair = 28.9646431 g/mol [11]
Molar fraction of O2 in air xO2 = 0.21
Other relevant constants
Individual gas constant of air Rair = R/Mair = 287.06 kgJK
Individual gas const. water vapor Rw = R/Mw = 461.52 kgJK
Density of air (at STP) ρair = 1.293 kg/m3n
Density of H2 (at STP) ρH2 = 0.0899 kg/m3n
Specific heat capacity of water (20◦ C) cp, m, water = 4182 kgJK
Density of water (20◦ C) ρwater = 998.2 kg/m3
Specific heat cap. air (p = const) cp, m, air = 1005.45 kgJK
Specific heat cap. air (V = const) cV, m, air = cp, m, air − Rair
= 718.39 kgJK
Ratio of specific heats for air κ = cp /cV = 1.40
(κ − 1)/κ = 0.285
Specific heat capacity of water vapor cp, water = 1858.94 kgJK
Enthalpy of water vaporization (0◦ C) hwe = 2500.827 gJ

c Alexander Kabza Fuel Cell Formulary

5.1 Relevant units based on SI units 51

5.1 Relevant units based on SI units

Physical unit Name (symbol) SI unit non SI units
Force F Newton (N) kg m/s2 -
Pressure p Pascal (Pa) kg/(m s2 ) N/m2
Energy/work E/W Joule (J) kg m2 /s2 Nm, Pa · m3
Power P Watt (W) kg m2 /s3 J/s
Charge Q Coulomb (C) As -
Voltage U Volt (V) kg m2 /(s3 A) W/A, J/C
Resistance R Ohm (Ω) kg m2 /(s3 A2 ) V/A

5.2 Used abbreviations

FCS Fuel cell system
LHV Lower heating value
HHV Higher heating value
OCV Open circuit voltage
EMF Electromotive force
Nl (or ln ), Nm (or m3n )
3
Norm liter or norm cubic meter
(normalized to standard conditions)
RH Relative humidity
STP Standard temperature and pressure
SATP Standard ambient temp. and pressure
TD Thermodynamic
WEG Water ethylene glycol
EMC Electromagnetic Compatibility
EMI Electromagnetic Interference
CAN Controller Area Network

c Alexander Kabza Fuel Cell Formulary

 52

6 Appendix 3: Air

6.1 Moist air

6.1.1 Water vapor saturation pressure

The water vapor partial pressure at saturation pws (t) is a function of

temperature t [◦ C]. Its values are listed in tables (e.g. in [16], see
also table 6.1). There are several formulas to fit this table values
and to calculate pws (t) as a function of temperature. Maybe the most
simple and therefore popular one is the Magnus equation. But other
equations fit with a higher accuracy to the table values.

Magnus equation:
C2 · t
pws (t) = C1 · exp [pws (t)] = Pa, [t] = ◦ C (6.1)
C3 + t

Magnus equation converted to temperature versus water vapor satu-

ration pressure:
C3 · ln pws
C1
(t)
t=
C2 − ln pws
C1
(t)

c Alexander Kabza Fuel Cell Formulary

6.1 Moist air 53

100
101.32
Water vapor saturation pressure [kPa]

80
70.04

60

47.29
40
31.12

19.90
20
12.33
7.38
4.25
0.61 1.23 2.34
0
0 10 20 30 40 50 60 70 80 90 100
Temperature t [°C]

Figure 6.1: Water vapor saturation pressure (Magnus equation, 0 to

100◦ C)

1600
1596
1400
Water vapor saturation pressure [kPa]

1284
1200
1023
1000

805
800
627
600
481
400 364
272
200 199
143
101
0
100 110 120 130 140 150 160 170 180 190 200
Temperature t [°C]

Figure 6.2: Water vapor saturation pressure (Magnus equation, 100

to 200◦ C)

c Alexander Kabza Fuel Cell Formulary

6.1 Moist air 54

WMO Technical Regulations (WMO-No. 49) [17]:

T1
 
lg pws (T ) = 10.79574 · 1 − (6.2)
T
T
−5.02800 · lg
T1
 !
T
−4 −8.2969· T1
−1
+1.50475 · 10 · 1 − 10
 T1

+0.42873 · 10−3 · 104.76955(1− T ) − 1
+0.78614 [pws (T )] = hPa, [T ] = K

T is the temperature in Kelvin and T1 = 273.16 (the triple point of

water).

Equation used by Vaisala (documented in old user manuals):

3
Ci T i
X
Θ=T−
i=0

With T = temperature in K, C0 = 0.4931358, C1 = −4.6094296 · 10−3 ,

C2 = −1.3746454 · 10−5 , and C3 = −1.2743214 · 10−8 .

3
bi Θi + b4 ln Θ
X
ln pws (T ) =
i=−1

With b−1 = −5.8002206 · 103 , b0 = 1.3914993, b1 = −4.8640239 · 10−2 ,

b2 = 4.1764768 · 10−5 , b3 = −1.4452093 · 10−8 , and b4 = 6.5459673.

c Alexander Kabza Fuel Cell Formulary

6.1 Moist air 55

Antoine equation:
B
log10 pws (t) = A −
t+C

With t = temperature in ◦ C, A = 10.20389, B = 1733.926, and C =

233.665. Other parameters are defined depending on the temperature
range.

Comparison of accuracy: Between 25 and 100◦ C the WMO equa-

tion has the best accuracy compared to table values 6.1. The max-
imum relative failure is 0.07%, followed by the Vaisala equation with
0.10%, Magnus with 0.17% and Antoine with 0.36%. Due to the sim-
plicity of Magnus this is maybe the most preferable equation to fit the
water vapor saturation pressure.

6.1.2 Other formulas to express humidity of air

Specific humidity or humidity (mixing) ratio Specific humidity Y

is the ratio between the actual mass of water vapor mw in moist air to
the mass of the dry air mdry air :

mw Rair pw (t) Mw pw (t)

Y (p, t) = = · = · [Y ] = kg/kg
mdry air Rw pair − pw (t) Mair pair − pw (t)
(6.3)
With Rw /Rair = 0.622 and pw (t) = ϕ · pws (t) (see below) we get:

ϕ pws (t)
Y (p, t, ϕ) = 0.622 · (6.4)
pair − ϕ pws (t)

c Alexander Kabza Fuel Cell Formulary

6.1 Moist air 56

2.0

Specific humidity aat saturation Ysat [kg/kg]

1.8
101.3 kPa
1.6
120 kPa
1.4
150 kPa
1.2 200 kPa
1.0

0.8
RH is 100% (saturation)
0.6

0.4

0.2

0.0
0 20 40 60 80 100
Temperature t [°C]

Figure 6.3: Specific humidity at saturation versus temperature

The maximum amount of water vapor in the air is achieved when

pw (t) = pws (t) the saturation pressure of water vapor at the tempera-
ture t.
For other gases the molar weight has to be considered in the equation
above. The constant value 0.622 is only valid for moist air!
Since the water vapor pressure is small regarding to the atmospheric
pressure, the relation between the humidity ratio and the saturation
pressure is almost linear at temperature far below 100◦ C.

Relative humidity RH (or ϕ) is the ratio of the partial pressure of

water vapor pw (t) to the partial pressure of water vapor at saturation
pws (t):

pw (t) aw (t) mw
ϕ= = = ϕ = 0.0 − 1.0
pws (t) aws (t) mws

RH can also be calculated with Y (t):

c Alexander Kabza Fuel Cell Formulary

6.1 Moist air 57

2.0

1.8
absolute pressure is 101.325 kPa 100% RH
1.6
75% RH
midity Y [kg/kg]

1.4
50% RH
1.2 25% RH
1.0
Specific hum

0.8

0.6

0.4

0.2

0.0
0 20 40 60 80 100
Temperature t [°C]

Figure 6.4: Specific humidity at constant pressure versus tempera-

ture

Y (t) p Y (t) p
ϕ= Mw · = · (6.5)
Mair
+ Y (t) ps (t) 0.622 + Y (t) ps (t)

Absolute humidity or water vapor density Absolute humidity a(t)

is the actual mass of water vapor present in the moist air and a func-
tion of t:

Mw pw (t) 1 pw g K pw (t)
a(t) = · = · = 2.167 ·
R T0 + t Rw (t) T0 + t J T0 + t

[a(t)] = g/m3 , [t] = ◦ C

Molar fraction The molar fraction y(t) of water vapor is calculated

as:

c Alexander Kabza Fuel Cell Formulary

6.1 Moist air 58

600

absolute humidity at saaturation asat [g/m3] 500

RH is 100% (saturation)
400

300

200

100

0
0 20 40 60 80 100
Temperature [°C]

Figure 6.5: Absolute humidity

pw (t)
y(t) = y = 0.0 − 1.0
pa

At standard pressure and 100◦ C y = 1 because there is only water in

vapor phase (and no air).

Dew point temperature The dew point temperature (DPT) Tdp is

the temperature at which water vapor starts to condense out of the air,
the temperature at which air becomes completely saturated. Above
this temperature the moisture will stay in the air.
If the DPT is close to the air temperature, the relative humidity is high;
and if the dew point is well below the air temperature, the relative
humidity is low.
The following equation allows to calculate the DPT Tdp based on RH
(ϕ = RH/100%) and temperature t:

c Alexander Kabza Fuel Cell Formulary

6.1 Moist air 59

1.0
1E-2
0.9

0.8 1E-4

0.7 1E-6
0.6
Mol fraction y

1E-8
0.5 -100 -80 -60 -40 -20 0
0.4

0.3
101.325 kPa
0.2 120 kPa
150 kPa
0.1 200 kPa
0.0
0 20 40 60 80 100 120
Dew point temperature Tdp [°C]

Figure 6.6: Molar fraction versus dew point temperature

 
C2 ·t
C3 · ln ϕ + C3 +t
Tdp = ·t
C2 − ln ϕ − CC32+t

[Tdp ] = ◦ C, [t] = ◦ C, ϕ = 0.0 − 1.0

To calculate RH from temperature and DPT:

C ·T
pw (t) exp C32+Tdpdp
ϕ= = ·t
pws (t) exp CC32+t

Enthalpy The air enthalpy h is needed to calculate energy and power

of air. The enthalpy of moist air consists of sensible heat and latent
heat. Since moist air is a mixture of dry air and water vapor, the en-
thalpy includes the enthalpy of the dry air (the sensible heat hair ) and
the enthalpy of the evaporated water (the latent heat hwater ).

c Alexander Kabza Fuel Cell Formulary

6.1 Moist air 60

100

80 100% RH
80% RH
60% RH
mperature [°C]

60 40% RH
20% RH
40
Dew point tem

20

-20
0 20 40 60 80 100
Temperature t [°C]

Figure 6.7: Dew point temperature versus temperature

100%

90%

80%

70%
midity RH [%]

60%

50%
relative hum

40%

30% DPT = t - 2K
20% DPT = t - 5K
DPT = t - 10K
10% DPT = t - 20K
0%
0 20 40 60 80 100
Temperature t [°C]

Figure 6.8: Relative humidity versus temperature

c Alexander Kabza Fuel Cell Formulary

6.1 Moist air 61

Specific enthalpy h [kJ/kg] of moist air is defined as the total enthalpy

of the dry air and the water vapor mixture per kg of moist air.
The enthalpy consists of three parts: a) warming up dry air from the
reference temperature to the air temperature t, b) evaporating water
inside the moist air and c) warming up water vapor from the reference
temperature to the air temperature t:

h = hair + Y · hwater (6.6)

with

hair = cp, air · t and hwater = cp, water · t + hwe

h = cp, air · t + Y · (cp, water · t + hwe ) [h] = J/g

where Y is the specific humidity of moist air, cp, air is the specific heat
capacity of dry air, cp, water is the specific heat capacity of water vapor
and hwe is the specific enthalpy of water vaporization. In case of dry
air (Y = 0), the enthalpy is a linear function of temperature (with slope
cp, air ). Heat capacities are dependent on temperature (see chapter
4.3), but between 0 and 100◦ C this may be neglected.
Because the enthalpy is not an absolute value; only the enthalpy dif-
ference to a reference temperature can be calculated. Depending on
the unit of temperature, the reference temperature is 0◦ C if [t] = ◦ C;
or 0 K if [t] = K. If the temperature is given in ◦ C, the enthalpy is 0 kg
kJ

at 0◦ C. If temperature is given in K, the enthalpy is 0 kg

kJ
at 0 K.
Due to the fact that only the enthalpy difference is used, it doesn’t
matter if temperature is given in K or ◦ C! It’s just important not to
switch between both units.

c Alexander Kabza Fuel Cell Formulary

6.1 Moist air 62

500

450

400
Specific entthalpy h [kJ/kg]

350

300
100% RH
250 75% RH
200 50% RH
25% RH
150 dry (0%)
100

50
absolute pressure is 101.325 kPa
0
0 10 20 30 40 50 60 70 80 90 100
Temperature t [°C]

Figure 6.9: Specific enthalpy versus temperature at constant pres-

sure

2000

1800
101.325 kPa
1600 120 kPa
halpy h [kJ/kg]

1400 150 kPa

200 kPa
1200

1000
Specific enth

relative humidity is 100% (saturation)

800

600

400

200

0
0 10 20 30 40 50 60 70 80 90 100
Temperature t [°C]

Figure 6.10: Specific enthalpy versus temperature at saturation

c Alexander Kabza Fuel Cell Formulary

6.1 Moist air 63

The following chart show the enthalpy based on 0◦ C reference:

For dry air there is another empirical fit to calculate the enthalpy. This
equation is used in DIN IEC 62282-3-2 [10]:

h(t) = (A · t + B · t2 + C · t3 )/Mair [h(t)] = kJ/kg, [t] = ◦ C or K

with A = 27.434, B = 6.18/2000, C = −0.8987/3 · 106 .

Also here the temperature t can be given in ◦ C or K, see above. But
this equation is only valid for dry air!

c Alexander Kabza Fuel Cell Formulary

6.1 Moist air 64

The following table lists the water vapor saturation pressure over liq-
uid water (between 0 and 200 ◦ C) according to [16].

c Alexander Kabza Fuel Cell Formulary

6.1 Moist air Source: 65
Handbook of Chemistry and Physics, 66th edition (1985‐1986), CRC
Vapor pressure of water below 100°C (page D189f)

t [°C] p [kPa] t [°C] p [kPa] t [°C] p [kPa] t [°C] p [kPa]

0 0.61129 51 12.97 101 104.99 151 488.61
1 0.65716 52 13.623 102 108.77 152 501.78
2 0.70605 53 14.303 103 112.66 153 515.23
3 0.75813 54 15.012 104 116.67 154 528.96
4 0.81359 55 15.752 105 120.79 155 542.99
5 0.8726 56 16.522 106 125.03 156 557.32
6 0.93537 57 17.324 107 129.39 157 571.94
7 1.0021 58 18.159 108 133.88 158 586.87
8 1.073 59 19.028 109 138.5 159 602.11
9 1.1482 60 19.932 110 143.24 160 617.66
10 1.2281 61 20.873 111 148.12 161 633.53
11 1.3129 62 21.851 112 153.13 162 649.73
12 1.4027 63 22.868 113 158.29 163 666.25
13 1.4979 64 23.925 114 163.58 164 683.1
14 1.5988 65 25.022 115 169.02 165 700.29
15 1.7056 66 26.163 116 174.61 166 717.83
16 1.8185 67 27.347 117 180.34 167 735.7
17 1.938 68 28.576 118 186.23 168 753.94
18 2.0644 69 29.852 119 192.28 169 772.52
19 2.1978 70 31.176 120 198.48 170 791.47
20 2.3388 71 32.549 121 204.85 171 810.78
21 2.4877 72 33.972 122 211.38 172 830.47
22 2.6447 73 35.448 123 218.09 173 850.53
23 2.8104 74 36.978 124 224.96 174 870.98
24 2.985 75 38.563 125 232.01 175 891.8
25 3.169 76 40.205 126 239.24 176 913.03
26 3.3629 77 41.905 127 246.66 177 934.64
27 3.567 78 43.665 128 254.25 178 956.66
28 3.7818 79 45.487 129 262.04 179 979.09
29 4.0078 80 47.373 130 270.02 180 1001.9
30 4.2455 81 49.324 131 278.2 181 1025.2
31 4.4953 82 51.342 132 286.57 182 1048.9
32 4.7578 83 53.428 133 295.15 183 1073
33 5.0335 84 55.585 134 303.93 184 1097.5
34 5.3229 85 57.815 135 312.93 185 1122.5
35 5.6267 86 60.119 136 322.14 186 1147.9
36 5.9453 87 62.499 137 331.57 187 1173.8
37 6.2795 88 64.958 138 341.22 188 1200.1
38 6.6298 89 67.496 139 351.09 189 1226.9
39 6.9969 90 70.117 140 361.19 190 1254.2
40 7.3814 91 72.823 141 371.53 191 1281.9
41 7.784 92 75.614 142 382.11 192 1310.1
42 8.2054 93 78.494 143 392.92 193 1338.8
43 8.6463 94 81.465 144 403.98 194 1368
44 9.1075 95 84.529 145 415.29 195 1397.6
45 9.5898 96 87.688 146 426.85 196 1427.8
46 10.094 97 90.945 147 438.67 197 1458.5
47 10.62 98 94.301 148 450.75 198 1489.7
48 11.171 99 97.759 149 463.1 199 1521.4
49 11.745 100 101.32 150 475.72 200 1553.6
50 12.344

Table 6.1: Water vapor saturation pressure

c Alexander Kabza Fuel Cell Formulary

6.2 Energy and power of air 66

6.2 Energy and power of air

Energy and power of dry and wet air are important to calculate e.g.
the power demand for air humidification in a fuel cell system or to
calculate efficiencies. See also ASME PTC 50-2002 Fuel Cell Power
Systems Performance [18] and DIN IEC 62282-3-2 Fuel cell technolo-
gies - Part 3-2: Stationary fuel cell power systems [10].
The energy of air contains the air enthalpy and the pressure:

Eair = hair + Ep, air [Eair ] = kJ/kg

The air power is given as:

Pair = Eair · ṁair [Pair ] = kW

The air enthalpy hair is defined by equation 6.6. Please be aware

that always enthalpy differences are calculated and not the absolute
enthalpy!
The energy of compressed air is given by:
p
Ep, air = Rair · T0 · ln
p0

T0 and p0 are the reference temperature and pressure, see section 4.4.

Example 1: How much power is required to heat dry air from 20 to

60◦ C with a mass flow of 30 kg/hr?

∆h = h60◦ C/0% − h20◦ C/0% = 40.218 kJ/kg

1
P = ∆h · ṁair = 40.218 kJ/kg · 30 kg/hr · hr/sec = 335 W
3600

c Alexander Kabza Fuel Cell Formulary

6.3 Air compression 67

Example 2: How much power is required to heat and humidify air at

ambient pressure from 20◦ C and 50% RH to 60◦ C and 80% RH? Air
mass flow is 30 kg/hr.

h20◦ C/50% = 38.56 kJ/kg

h60◦ C/80% = 363.18 kJ/kg

∆h = 326.61 kJ/kg

1
P = ∆h · ṁair = 326.61 kJ/kg · 30 kg/hr · hr/sec = 2722 W
3600

6.3 Air compression

6.3.1 Definitions

Adiabatic Process where no heat is transferred to or from the sur-

roundings (dq = 0).

Reversible Process can be reversed without loss or dissipation of

energy (dS = 0).

Isentropic Process with equal entropy

Any reversible adiabatic process is an isentropic process (dq = dS =
0). But not vice versa!

c Alexander Kabza Fuel Cell Formulary

6.3 Air compression 68

6.3.2 Thermodynamics

Compression and expansion

A gas mass flow ṁ1 is compressed or

expanded from pressure p1 and temper-

Wel
ature T1 to pressure p2 and temperature
T2 . p2 > p1 in case of compression and
Gas in Compressor Gas out
p2 < p1 in case of expansion. The pres-
sure ratio Π is defined as:

Wloss
p2
Π=
p1 Figure 6.11: Gas compres-
sion
For compression Π > 1 and for expan-
sion Π < 1.
A compression process consumes energy or work, an expansion pro-
cess produces energy or work.
To calculate the efficiency of a compression process the theoretical
energy demand needs to be calculated based on an isentropic pro-
cess. In a real compression process there is always heat transferred
to the surrounding (or cooling), and also gas mass flow losses can
occur inside the compressor.

Thermodynamic of gases

Internal energy U The first law of thermodynamics [13] defines in-

finitesimal changes in internal energy dU of an ideal gas as changes
in added heat dq and provided work dw:

dU = dq + dw

c Alexander Kabza Fuel Cell Formulary

6.3 Air compression 69

Enthalpy H is defined as H = U + pV . Infinitesimal changes dH

are therefore defined as:

dH = dU + pdV + V dp

Heat capacities C describe the temperature dependencies of inter-

nal energy U and enthalpy H:

CV = (∂U/∂T )V Cp = (∂H/∂T )p

CV is the heat capacity at constant volume and Cp the heat capacity

at constant pressure of an ideal gas:

dU = CV dT and dH = Cp dT (6.7)

With pV = nRT it is dH = dU + nRdT and therefore:

Cp − CV = nR

For ideal gases the specific heat capacities cp and cV with [c] =
J/mol K are independent of temperature! For ideal one atomic gases
it ist:
Cp 5 CV 3 5
cp = = R cV = = R κ = = 1.67
n 2 n 2 3

And for ideal two atomic gases:

7 5 7
cp = R cV = R κ= = 1.40
2 2 5

The specific heat capacities cp, m and cV, m are given as:
cp, m = cp /M and cV, m = cV /M
[cp, m ] = [cV, m ] = J/kg K

c Alexander Kabza Fuel Cell Formulary

6.3 Air compression 70

Polytropic process For all thermodynamic processes of ideal gases

it is in general:

pV n = const n = polytropic index (constant)

The energy or work in a volume change process is in general:

dw = dU = −p dV = CV dT

With pV = nRT it is:

CV dT /T = −nR dV /V

The integral over start (1) and end state (2) is:
Z 2 Z 2
CV dT /T = −nR dV /V
1 1

CV ln (T2 /T1 ) = −nR ln (V2 /V1 )

Isentropic process The isentropic process is a special case of a

polytropic process, where the polytropic index becomes the isentropic
index or heat capacity ratio κ:

cp
n=κ= >1
cV

CV cV 1
With cp − cV = R, cV = CV /n and therefore nR
= cp −cV
= κ−1
it is:

κ−1
T1 V2

=
T2 V1

c Alexander Kabza Fuel Cell Formulary

6.3 Air compression 71

The following ratios are valid for an isentropic process:

κ−1 ! κ−1
κ κ
T1 V2 T1 p1 p1 V2
 
= = =
T2 V1 T2 p2 p2 V1

Volume work and technical work

The volume (change) work wV,12 is the work that is done on a piston
to reduce the start volume V1 to the end volume V2 . It is equivalent in
the change of inner energy. The volume work for an adiabatic process
(dq = 0) is defined as:

Z V2 Z 2 Z T2
wV, 12 = − p dV = dU = CV dT = CV ∆T
V1 1 T1

The technical work wt,12 is the work that is transferred from or mass
flow via a technical system (for instance a mechanical axle). The
technical work is equal to the enthalpy of the kinetic and potential
energy of a stationary mass flow. The technical work wt,12 for com-
pression is the work given to the drive shaft of the compressor; it
contains the displacement work, but not the friction [19]!
The technical work for an adiabatic process (dq = 0) is defined as:

Z p2 Z 2
wt, 12 = + V dp = dH = Cp ∆T
p1 1

Each compression process (polytropic,

isentropic, isothermal etc.) follows a p
2
(similar but not the same) curve the pV p2
chart (see figure 6.12) from the start (1)
to the end (2) point. Adding the different
p1 1

c Alexander Kabza VFuel

2 Cell Formulary
V1 V
6.3 Air compression 72

process areas shown in the pV chart

gives:
Z V1 Z p2
p2 V 2 + p dV = p1 V1 + V dp
V2 p1

Therefore:
wt, 12 = wV, 12 + p2 V2 − p1 V1

Comment: For liquids ∆V ≈ 0 and therefore wV, 12 ≈ 0; then it is:

wt, 12 ≈ (p2 − p1 ) · V
The thermodynamic derivation (see above) defines the following ra-
tios for the isentropic compression or expansion:
κ κ−1
V1 T1 V2
 
κ−1
Π= and = =Π κ
V2 T2 V1

Therefore for the volume work it is:

κ−1
 
wV, 12 = cV, m · (T2 − T1 ) = cV, m T1 · Π κ −1

And similar for the technical work:

κ−1
 
wt, 12 = cp, m · (T2 − T1 ) = cp, m T1 · Π κ −1

wt, 12 = h2 − h1 with h = cp, m · T

wt, 12 = κ · wV, 12 with κ = cp /cV

The technical work is always more than the volume work, because
κ > 1.

c Alexander Kabza Fuel Cell Formulary

6.3 Air compression 73

6.3.3 Ideal and real compressor

Ideal The ideal compression is an isentropic process. There is no

(energy or air mass flow) exchange with the surroundings.

Real Due to friction the gas compression in a real compressor al-

ways dissipates energy to the surroundings or to the cooling system.
This compression process is polytropic. After a polytropic compres-
sion the actual end temperature T2, act is always higher compared to
an ideal isentropic compression T2, s :

T2, act > T2, s

The isentropic efficiency ηs of a compressor is the ratio between isen-

tropic technical work ws, 12 and actual technical work wact, 12 :

ws, 12 T2, s − T1 h2, s − h1

ηs = = =
wact, 12 T2, act − T1 h2, act − h1

wact, 12 is ws, 12 plus specific friction work (dissipation).

To calculate the isentropic work, first of all the isentropic end temper-
ature T2, s needs to be calculated:

κ−1
T2, s = T1 · Π κ

Now the isentropic technical work can be calculated as:

κ−1
 
ws, 12 = cp, m · (T2, s − T1 ) = cp, m T1 · Π κ −1

c Alexander Kabza Fuel Cell Formulary

6.3 Air compression 74

6.3.4 Examples

Example 1: Isentropic efficiency

Air is compressed from 1 bara at 30◦ C (303.15 K) to 3 bara at 150◦ C

(423.15 K). In the compressor there is no energy or air mass flow loss
to the surroundings.
The actual technical work that is done at the air is:

J kJ
wact = cp, m · (T2, act − T1 ) = 1005 · 120 K = 120.6
kg · K kg

Isentropic temperature:
!0.285
κ−1 3 bara
T2, s = T1 · Π κ = 303.15 K · = 414.6 K
1 bara

Isentropic technical work:

ws = cp, m · (T2, s − T1 )
J kJ
= 1005 kg·K · (414.6 K − 303.15 K) = 112.4 kg

Isentropic efficiency:

ws 112.4 kJ/kg
ηs = = = 0.932 or 93.2%
wact 120.6 kJ/kg

Example 2: Electric efficiency

27 kg/h air is compressed from 2.4 to 6 barabs, the electric power

consumption of the compressor is 1305 W:

c Alexander Kabza Fuel Cell Formulary

6.3 Air compression 75

ṁ = 27 kg/h, p1 = 2.4 barabs and T1 = 18 ◦ = 291K

p2 = 6 barabs and Pel = 1305 W
Air mass flow losses due to gaps and other heat losses to the sur-
roundings are neglected! But of course heat is transferred to the
cooling!
The isentropic technical work given to the air is:
κ−1
 
ws = cp, m T1 · Π κ −1
 0.285 
J 6 bar kJ
= 1005 kg·K
· 291 K · 2.4 bar
− 1 = 87.6 kg

Isentropic technical power:

kJ
Ps = ṁ · ws = 87.6 · 27 kg/h = 2365.2 kJ/h = 657.0 W
kg

The electric efficiency of this compressor is:

Ps 657.0 W
ηel = = = 0.50 or 50%
Pel 1305 W

6.3.5 Comment

The technical work equation is often given in literature with different

indices. The next two equations show the relation between κ, R, cp
and cV :

κ cp /cV cp
·R= · (cp − cV ) = · (cp − cV ) = cp
κ−1 cp /cV − 1 cp − cV

1 1 cV
·R= · (cp − cV ) = · (cp − cV ) = cV
κ−1 cp /cV − 1 cp − cV

c Alexander Kabza Fuel Cell Formulary

 76

7 Appendix 4: Fuel cell stack

water management

7.1 Assumptions

Simple calculation of fuel cell stack water management, based on

calculation of specific humidity at cathode exhaust with the following
assumptions:
1. The product water is removed via cathode side; that means
there is no water removal via anode side.
2. Cathode air is heated inside the stack to a specific temperature
t (which does not necessarily have to be the stack coolant outlet
temperature).
3. Condensation is not considered! All water exits the stack as
water vapor.
4. All gases (air, hydrogen, water vapor) are treated as ideal gases.

7.2 Derivation

Stack inlet: Humidified air with specific humidity Y (calculated by

equation 6.4)

c Alexander Kabza Fuel Cell Formulary

7.2 Derivation 77

NI λ
ṁair in = · Mair · see equation 1.9
4F x

ṁwater in = ṁair in · Yin see equation 6.3

Stack outlet: Water inet and product water exit the cathode
!
IN λcathode
ṁcathode, out = · · Mair − MO2 see equation 1.10
4F xO 2

ṁwater, out = ṁwater, in + ṁwater, prod

The specific humidity at cathode outlet is:

λ Mwater
mwater, out ṁwater, in + ṁwater, prod x
· Yin + 2 · Mair
Yout = =   = MO2
ṁcathode, out IN
· λcathode
· Mair − MO2
λ
−
4F xO2 x Mair

Relative humidity can be calculated using equation 6.5.

c Alexander Kabza Fuel Cell Formulary

 78

8 Appendix 5: FCS controls

parameters

8.1 Fuel cell stack pressure drop

The reactant pressure drops through the fuel cell stack anode and
cathode are important stack parameters for system design consid-
erations. Especially the pressure drop on the cathode side needs
to be considered, because air needs to be compressed accordingly.
For simplification the stack can be seen as an orifice with a specific
pneumatic resistance, see figure 8.1
p1 (p.Si.C) and p2 (p.So.C) are gauge pressures before (upstream)
and behind (downstream) the stack or orifice. For gases the mass
flow ṁ is given by the following orifice equation:
q
ṁ = αA · 2 ρ1 ∆p (8.1)

p p
Si.C So.C

pD
T S.C T
Si.C So.C

F
F.Air.C (MFC) V1 Stack = orifice V2 p.PC.C

Figure 8.1: Cathode pressure drop schematic

c Alexander Kabza Fuel Cell Formulary

8.2 Cathode dynamics - UNDER CONSTRUCTION *** 79

ρ1 = ρ0 p1p+p
0
0
is the gas density at pressure p1 (temperature is not con-
sidered here!), and ∆p = p1 −p2 is the pressure drop at the orifice. αA
is an orifice specific parameter containing the orifice flow coefficient α
and the orifice cross-sectional area A, [αA] = m2 . This coefficient is
hardly accessible by theory, but it can easily be measured if all other
parameters of equation 8.1 are known.
Example: The air flow through a single cell is V̇ = 10 Nl/min (ṁ =
0.216 g/sec). Pressures are measured to be p1 = 15 kPa and ∆p =
10 kPa (and therefore p2 = 5 kPa), then αA = 1.251 · 10−6 m2 .
This orifice equation 8.1 can be solved to ∆p with the following result:

q
p2 − p0 (αA ρ0 )2 · (p0 + p2 )2 + 2 ṁ2 p0 ρ0
∆p = + − p2 ∆p ∝ ṁx
2 2 αA ρ0
(8.2)
Figure 8.2 shows the pressure drop ∆p as a function of air mass flow
ṁ with three three different values of αA. According to equation 8.1
αA = 2.9 · 10−4 m2 for a given pressure drop of 25 kPa at the air mass
flow of 115 g/sec and cathode exhaust pressure of 130 kPaabs.

8.2 Cathode dynamics - UNDER

CONSTRUCTION ***

The dynamic behavior of the cathode subsystem can be discussed

based on the simplified schematic in figure 8.3, where the stack again
is considered as a pneumatic resistor RStack . In case of steady state
conditions the air mass flow into the system is equal to the air mass
flow out of the system:

during steady state: ṁin ≡ ṁout

c Alexander Kabza Fuel Cell Formulary

8.2 Cathode dynamics - UNDER CONSTRUCTION *** 80

25
2.60e-04
2.80e-04
20 3.00e-04

Pressure drop [kPa]

15

10

0
20 40 60 80 100
Air flow [g/sec]

Figure 8.2: Pressure drop versus air mass flow

But during air mass flow transients the volume V between the air feed
and the fuel cell stack needs to be considered, as described here
[20]. The reason therefore is that each individual air flow corresponds
to a specific pressure drop (see equation 8.2), and therefore also the
pressure p1 before the stack is a function of air flow. An increasing air
flow ṁin also increases the pressure in the volume V , and therefore
ṁC > 0:

during transients: ṁin = ṁC + ṁout

The volume V can be describes as a pneumatic capacity CVol . Dur-

ing air flow up (or down) transients additional air flow ṁC goes into (or
comes out of) this capacity. Corresponding to an electric RC-circuit
(∆p corresponds to electric voltage, ṁ to electric current), the follow-
ing equations are valid also for this pneumatic RC-circuit:

∆p(ṁ)
τ = RStack CVol and RStack = (non-linear)
ṁStack

The relation between volume V and hydraulic capacity CVol is given

by the individual gas constant Rair and the temperature T :

c Alexander Kabza Fuel Cell Formulary

8.2 Cathode dynamics - FUNDER CONSTRUCTION *** 81
F.Air.C (MFC) V1 Blende V2 p.PC.C

p p
Si.C So.C
RStack
m in m out
m C m Stack

CVol

Figure 8.3: Pneumatic RC schematic

V V ∆p
CVol = and therefore τ ∝
Rair T ṁ

With the information on CVol and RStack the dynamic response of ∆p

(and therefore also air flow) transients for any cathode subsystem can
be calculated as follows:

t
∆p(t) = ∆pstatic · exp (− ) (8.3)
τ
On fuel cell teststands often the air flow ṁ is lower and the volume
V is higher than in the final FCS application. Especially there the
dynamic response needs to be considered! During electric load tran-
sients it needs to be considered that the reactant flows through the
stack may be significantly delayed by several seconds (depending on
τ )! Anode and cathode starvation issues may occure.

c Alexander Kabza Fuel Cell Formulary

Bibliography 82

Bibliography
[1] Arnold Lamm Wolf Vielstich and Hubert A. Gasteiger. Handbook
of Fuel Cells Vol. 1. John Wiley & Sons Ltd, 1 edition, 2003.
[2] S. Srinivasan et al. Advances in solid polymer electrolyte fuel
cell technology with low platinum loading electrodes. J. of Power
Sources, 22:359 – 375, 1988.
[3] J. Kim et al. Modeling of pemfc performance with an empirical
equation. J. Electrochem. Soc., 142(8):2670 – 2674, 1995.
[4] U.S. Department of Energy. Fuel Cell Handbook. National Tech-
nical Information Service, U.S. Department of Commerce, 7 edi-
tion, 2004.
[5] James Larminie and Andrew Dicks. Fuel Cell Systems Ex-
plained. John Wiley & Sons Ltd, 2 edition, 2004.
[6] Performance Working Group within the SAE Fuel Cell Stan-
dards Committee. SAE J2617: Recommended Practice for test-
ing Performance of PEM Fuel Cell Stack Sub-system for Auto-
motive Applications. SAE International, Nov2007.
[7] Fuel Cell Technologies BSI Technical Committee GEL/105. DD
IEC TS 62282-1:2005 Fuel cell technologies - Part 1: Terminol-
ogy. BSI British Standards, 18 April 2005.
[8] Performance Working Group within the SAE Fuel Cell Stan-
dards Committee. SAE J2615: Testing Performance of Fuel

c Alexander Kabza Fuel Cell Formulary

Bibliography 83

Cell Systems for Automotive Applications. SAE International,

Jan2005.
[9] Wikipedia. Standard conditions for temperature and pressure.
http://en.wikipedia.org/wiki/Standard conditions for temperature
and pressure, July 2010.
[10] DIN. DIN IEC 62282-3-2 Fuel cell technologies - Part 3-2: Sta-
tionary fuel cell power systems. DIN Deutsches Institut fuer Nor-
mung e. V., July 2009.
[11] http://physics.nist.gov/cuu/Constants/index.html. NIST Funda-
mental Physical Constants.
[12] DIN. DIN 1343 Reference conditions, normal conditions, nor-
mal volume; concepts and values. DIN Deutsches Institut fuer
Normung e. V., 1990.
[13] Peter W. Atkins. Physikalische Chemie, volume 2. VCH Verlags-
gesellschaft GmbH, 1990.
[14] DIN. DIN 51857 Gaseous fuels and other gases - Calculation of
calorific value, density, relative density and Wobbe index of pure
gases and gas mixtures. DIN Deutsches Institut fuer Normung
e. V., 1990.
[15] SAE. SAE J2572: Recommended Practice for Measuring Fuel
Consumption and Range of Fuel Cell and Hybrid Fuel Cell Vehi-
cles Fuelled by Compressed Gaseous Hydrogen. SAE Interna-
tional, 2008-10.
[16] David R. Lide. CRC Handbook of Chemistry and Physics, In-
ternet Version 2005. http://www.hbcpnetbase.com, CRC Press,
2005.
[17] Resolution 26 (EC-64). WMO Technical Regulations, Basic Doc-
uments No. 2, Volume I General Meteorological Standards and

c Alexander Kabza Fuel Cell Formulary

Bibliography 84

Recommended Practices. World Meteorological Organization

(WMO), july 2012 edition, 2012.
[18] A. J. Leo (Chair). ASME PTC 50-2002 Fuel Cell Power Systems
Performance. The American Society of Mechanical Engineers,
November 29, 2002.
[19] Prof. Dr. Dieter Winklmair. Script Energie- und Wärmechnik,
http://www.winklmair.de/ewt/folien-1.pdf, August 2010.
[20] Michael A. Danzer. Dynamik und Effizienz von Polymer-
Elektrolyt-Brennstoffzellen. PhD thesis, Universität Ulm, 2009.

c Alexander Kabza Fuel Cell Formulary