Professional Documents
Culture Documents
net/publication/283473049
CITATIONS READS
10 2,129
3 authors, including:
Some of the authors of this publication are also working on these related projects:
Advanced methods of green hydrogen production and its transportation - H2LAB View project
All content following this page was uploaded by Ankica Kovač on 01 May 2018.
ScienceDirect
Article history: Solar hydrogen power system for supplying isolated passive house with utility power en-
Received 20 July 2015 ergy and hydrogen fuel for personal transport has been established and discussed.
Received in revised form Commercially available components such as photovoltaic module, proton exchange
14 September 2015 membrane fuel cell stack and electrolyzer stack were acquired. Appropriate control logic
Accepted 15 September 2015 was developed. MATLAB/Simulink software has been used to simulate control logic and
Available online 31 October 2015 system components. The proof of the control logic and components models has been
performed by simulations predefined scenarios, i.e. with and without solar irradiance and
Keywords: with given energy demands by the user. All of that can be used as valuable tool for any
Solar energy autonomous solar hydrogen system design of passive house.
Passive house Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights
Hydrogen reserved.
PEM fuel cell
Alkaline electrolyzer
Mathematical modeling
operating voltage; Rs/U is series resistance; m is ideality factor; According to the ideal gas law, the pressure p of gas storage
Ut/V is thermal voltage. can be calculated according to the equations (4) and (5).
In order to find the needed UI relationship, a systematic RT
pH2 ¼ nst (4)
way for obtaining the best possible curve fit was developed VH2
according to the PV module UI characteristic. The PV module where:
is represented with one block in Simulink scheme given Z
1
in Fig. 2. nst ¼ IEly IFC dt (5)
2F
Hydrogen storage
Wcomp/W is total compressor work; ngas/mol s1 is gas flow;
The hydrogen storage models were based on either the ideal
Wl/J mol1 and Wll/J mol1 are polytropic work; hcomp/% is
gas law or Van der Waals equation of state for real gases.
compressor efficiency.
Fig. 2 e Simulink model of PV module; Simulink model of an electrolyzer; Simulink model of a hydrogen storage; Simulink
model of the compressor; Fuel cell model inputeoutput signals; Simulink model of the battery; Simulink model of the load.
16004 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 6 0 0 1 e1 6 0 0 9
Because the process of refueling hydrogen storage in the Simulink models of PV module, an electrolyzer, H2 storage,
vehicle was manually controlled (this was not considered by compressor, PEM fuel cell and overall solar hydrogen power
controller simulated in Simulink), compressor model in Fig. 2 system are given in Fig. 2. PEM fuel cell was chosen because of
was not included in the power system simulation. its high efficiency and low operation temperature. According
to the performed calculations within selection process of the
Fuel cell system components it proved to be the most cost-effective as
Power data of selected PEM fuel cell stack are as follows: well.
electrical rated power ¼ 2.4 kW (2 1.2 kW); maximum elec-
tric power ¼ 3 kW (2 1.5 kW); open circuit voltage ¼ 2 ca. Power system components models
48 V; rated electric current ¼ 45 A; rated electric voltage ¼ 26 V;
maximum electric current per stack ¼ 50 A; hydrogen con- The main objective of the control strategy is to utilize
sumption at rated electric power ¼ <18.5 SLPM. excess RES in system to charge short-term and/or long-term
In order to model fuel cell stack there were calculated energy storage. And opposite, if there is deficit of RES
different voltage losses and consequently polarization curve. the systems control strategy decides whether to use long-
The first step in creating the polarization curve was to calcu- term or short-term energy storage to supply the load. In
late the Nernst voltage given in equation (8). order to fulfill these scenarios, the state of system must be
The Nernst voltage can be calculated using the following known all the time. A scheme of control system is given
equation: in Fig. 3.
! The control strategy includes hydrogen state of the charge
Gf ;liq RT PH2O
UNerst ¼ ln 1=2
(8) (SOC), battery SOC, in addition, weather/load prediction and
2F 2F PH2 PO2
actual power balance in system could be taken into account.
where: The electric power produced by the PV modules is managed
Gf,liq/228,170 J mol1 is Gibbs free energy of hydrogen in by a control of electrical part of power system. It sends the
liquid state; Tk/K is temperature; i/A cm2 is current density. power to the different components of the system following a
For the purpose of the model simplicity at this level, predefined priority order. The power demand of the user has
theoretical value of UNernst ¼ 1.23 V was assumed. Then the the first priority (batteries). The excesses of power are sent to
fuel cell voltage was calculated by following equation: the electrolyzer stack to produce hydrogen (storage). Actu-
ally all power systems must include a control strategy that
UFC ¼ UNerst ðEAct þ EOhmic þ EConc Þ (9)
describes the interactions between its components. Also
The losses were composed of activation, concentration and controller has been checked using three scenarios. In the first
Ohmic losses. Activation losses were given in the equation (10). scenario it was considered the daily PV system operation to
supply electricity to an electrolyzer stack to produce
DEAct ¼ a þ blogðiÞ (10) hydrogen while other components were not working. In the
where: second scenario night period when PV system and an elec-
trolyzer stack are not in function, storage had enough
RT RT hydrogen to supply PEM fuel cell stack and battery SOC is in
a ¼ 2:3 logi0 ; b ¼
aF aF 30% level for critical situation. In the third scenario only
io ¼ 3 106 A cm2 is exchange current density; a ¼ 1 is battery was running and operating to meet the load Other
transfer coefficient. parts of the system were not in function.
Ohmic losses were given according to the equation (11).
Fuel cell
PV System
Control System
Electrical
Load
Mannual control
Electrolyzer
Car
Battery
H2
H2
Control line
From Sensors
H2
Storage
Electricity flow
Hydrogen flow
The nature of fuel cell polarization curve can help in H2 storage tank
identification of particular problems. In order to see the result of hydrogen storage simulation the
Polarization curve and efficiency were given for a single flow rate of hydrogen is much higher than the real as well as
cell. There are additional power losses and resulting efficiency input data in storage model just to show whether simulation
loss due to auxiliary components in the fuel cell stack (fans, model worked properly.
pumps, relays and controllers).
Fig. 4 e Current density vs. ohmic, concentration and Fig. 5 e Current density vs. potential losses related to
activation losses in fuel cell. polarization curve.
16006 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 6 0 0 1 e1 6 0 0 9
Period 0 am e 6 am
During this time period of the day very small electricity con-
sumption existed (0.001) that is quite visible on the curve of
Electricity consumption; solar irradiance did not exist. This is
visible on curve of Solar irradiance. During this period pressure
in the hydrogen storage was equal to 4 bar. This is visible on
the curve of Hydrogen storage tank pressure and shows that fuel
cell stack did not produce electricity. Battery covered small
electricity consumption because it had higher priority than
Fig. 6 e Simulation results of H2 storage. fuel cell stack. SOC of the battery slightly dropped that is
visible in the curve of SOC battery.
As it can be seen from Fig. 6 it was operating as expected.
That is why the pressure of the storage tank was increased Period 6 am e 7 am
very fast to the maximum level. Electricity consumption sharply jumped from 0.001 to 0.1 and
during period of 1 h had the constant value (0.1). At the same
time solar irradiance started to rise. Because of higher priority
Functioning control system embedded in the power system
battery supplied the requested larger amount of energy. Due
(proof)
to that curve of SOC battery sharply dropped in this period. At
the same time electrolyzer stack started to produce hydrogen
The most important values of the solar hydrogen power sys-
that was stored in hydrogen storage. Hydrogen pressure
tem are: PV modules surface, battery and hydrogen storage
started to rise as it can be seen from the curve of Hydrogen
size. Solar energy is directly converted into electrical energy
storage pressure.
through PV modules. These modules originally have 36 cells in
series. Simulations were performed with the following sizes:
Period 7 am e 4 pm
number of PV modules was 51; battery capacity was 7751.9 A s;
Electricity consumption sharply dropped at 7 am and reached
the volume of hydrogen storage was 0.1 m3. Rather small
the constant value (0.02). During this period solar irradiance
storage volume enables short simulation time. Simulated time
continued to rise up to 1000 W m2 at 12 pm and then grad-
of 24 h was chosen only for components models and simula-
ually decreased to 0 W m2 at 6 pm. Hydrogen pressure
tion procedure verification. Generally speaking, user should
continued to rise and when it reached storage limit of 30 bar it
adjust simulated time and step size to available computer
stayed constant. Battery SOC continued to gradually decrease
speed and memory. The initial conditions also influence the
up to the value of 30% what was allowable minimum. Within
results of the simulation. Their values are as follows: battery
this period there was an interesting situation between 12.5 pm
SOC was 1.00 (100%); hydrogen storage pressure was 4 bar. In
and 16 pm related to battery charging and discharging and
order to prove the controller system function, different sce-
operation of electrolyzer stack at the same time as the fuel cell
narios have been performed. The function of controller sys-
stack (refueling and emptying of the hydrogen storage).
tem was to drive all three scenarios without manual
intervention during simulation. Therefore, the requirements
Period 6 pm e 10 pm
of the scenario should be consistent with the strategy and the
Electricity consumption rised by step of 0.2 what was quite
internal logic controllers. The scenarios have been defined
high consumption; battery was between 30 and 40% SOC; fuel
due to operating status of individual system elements and
cell stack used hydrogen; storage pressure dropped gradually
external conditions such as solar irradiance and energy
Fig. 7 e a) Hourly solar irradiance in island of Hvar; b) Equivalent DC electric load power profile for specified day.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 6 0 0 1 e1 6 0 0 9 16007
Period 10 pm e 12 pm
Again electricity consumption was very small (0.001); battery
was empty; electrolyzer stack didn't work, but fuel cell stack Table 1 e Summary of energy system in the passive
worked and draw only very small hydrogen quantity resulted house.
with extremely slow drop of hydrogen pressure. Components and quantities Dimension Value
Space heating system
Summary of household energy system Total heating load kWh per year 4395.7
In Table 1 it is given summary of energy system used in the Solar collectors area m2 7
presented passive house. In term of definition passive house is Solar thermal collectors area m2 3
Water tank volume m3 0.750
a concept for ultra-low energy building that require little en-
Space cooling system
ergy for space heating or cooling and save energy up to 90%
Type of system: solar desiccant cooling
compared with typical building and over 75% compared with Total cooling load kWh per year 5204.1
average new buildings. It requires very little energy to achieve Cooling power capacity kW 10
a comfortable temperature. A building based on this concept Power of motor for desiccant wheel kW 0.12
is energy efficient, comfortable, sustainable and environ- Power of cooling power pump kW 0.3
mental friendly at the same time. Ventilation rate of conditioned space m3/h 500
Power system
In the table above are listed all required parameters of en-
Photovoltaic array total area m2 46.94
ergy system needed for design of presented passive house, Number of modules pcs 51
according to climate conditions at the island Hvar. The results Nominal power of module kW 0.12
presented in this table give their contribution to the future work Total power of PV array kW 6.12
of other researchers as well whose work whether it is based on Fuel cell stack
designing a particular energy system or passive house in its all. Installed power of fuel cell stack kW 2.4
Maximum hydrogen consumption SLPM* 36
Maximum water production l/h 1.6
Maximum heat production kW 4.8
Conclusion Maximum fuel cell stack % 50
system efficiency
This work presents part of Matlab/Simulink based simulation Electrolyzer stack
tool required for energy system design of passive house and Power of electrolyzer stack kW 5
Number of electrolyzer cells in stack pcs 22
the need for personal transport of its residents, based on using
Rate of hydrogen production Nm3/h 1
solar energy and hydrogen entirely. Furthermore, it presented
Overall energy consumption kWh/Nm3 H2 4.9
mathematical model of the solar hydrogen power system to Surface area cm2 300
provide electricity load for grid disconnected passive house
*SLPM e standard litter per minute.
consisting of PV modules, battery, electrolyzer, hydrogen
16008 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 6 0 0 1 e1 6 0 0 9
storage, fuel cell and controller. The model has been tested WI, WII Polytropic work, J mol1
using specific selected commercial components. It was Z (p) Compressibility factor
designed logical structure of controller for the whole system.
Greek symbols
This logical structure has been tested using three selected
a Charge transfer coefficient, 0e1
scenarios. This model can be used as a tool to test different
hcomp Compressor efficiency
scenarios corresponding to different purposes as follows:
[19] Maroufmashat A, Sayedin F, Khavas SS. An imperialist [21] Garcia-Valverde R, Miguel C, Martinez-Bejar R, Urbina A.
competitive algorithm approach for multi-objective Optimized photovoltaic generator-water electrolyzer
optimization of direct coupling photovoltaic-electrolyzer coupling through a controlled DC-DC converter. Int J
systems. Int J Hydrogen Energy 2014;39:18743e57. Hydrogen Energy 2008;33:5352e62.
[20] Atlam O, Barbir F, Bezmalinovic D. A method for optimal A. Autonomous hydrogen production system. Int J
[22] Ðukic
sizing of an electrolyzer directly connected to a PV module. Hydrogen Energy 2015;40:7465e74.
Int J Hydrogen Energy 2011;36:7012e8.