INTEGRATING RENEWABLES FOR REMOTE FUEL SYSTEMS INTÉGRATION DES ÉNERGIES RENOUVELABLES DANS DES RÉGIONS ÉLOIGNÉES RONNY GLÖCKNER, IFE, NORWAY ØYSTEIN ULLEBERG, IFE, NORWAY RAGNE HILDRUM, STATKRAFT SF, NORWAY CATHERINE E. GRÉGOIRE PADRÓ, NREL, USA
1. INTRODUCTION - INTRODUCTION Hydrogen energy systems have been proposed as a means to increase energy independence, improve domestic economies, and reduce greenhouse gas and other harmful emissions from stationary and mobile sources. These systems, however, face technical and economic barriers that must be overcome before hydrogen can become a competitive energy carrier for the 21st century. A wind-H2 system for a remote location was studied to determine the feasibility of producing hydrogen for transportation applications. This case study was initiated by STATKRAFT SF, Norway’s largest producer of hydroelectric power and the second largest producer in the Nordic region. The modelling work was performed by the Institute for Energy Technology (IFE). The study has then been undertaken as a case study by the International Energy Agency Hydrogen Implementing Agreement’s Annex 13 – Design and Optimization of Integrated Systems. Eight participating countries (Canada, Japan, Lithuania, Netherlands, Norway, Spain, Sweden, and USA) are cooperating to develop an understanding of the technical, economic, and social issues related to the introduction of hydrogen energy systems. 2. BACKGROUND - PROBLÉMATIQUE In Norway, more than 99% of the electricity demand is supplied by hydropower. During the last 10 years, however, public resistance based on environmental issues has virtually brought the development of the remainder of these resources to a standstill. Because of the increase in power demand and good wind resources along a sparsely populated coastline (a potential of around 10 TWh/year is recognised), the focus on wind energy plants has grown over the last years. It seems that the public resistance, especially in the early phase of the introduction of large-scale wind energy development, is smaller than is the case for further hydropower development. Over the last year (2000), the Norwegian government has approved several applications for building of wind parks, in total amounting to some 300 MW peak power. The selected locations for Norwegian wind parks are remote areas, such as islands, where the best wind resources are found. The wind parks will produce power for the small communities in these areas, but also for the common electricity grid as a complement to the existing hydropower. In this study, the production and use of hydrogen in fuel cells in relation to such wind parks is discussed. A wind/electrolyser-based hydrogen refuelling station taking the base power from one of the large windmills in a wind park is modelled. The hydrogen produced is assumed to be used in buses operating the public transport on the remote location. At a later stage, hydrogen for operation of ferries would also be an attractive alternative. The existence of a low-voltage grid is assumed for this study. In practice, this enables testing under constant conditions during the commissioning period. In the modelling work, the assumption of a weak grid enables the identification of several cases of operation of the filling station. In addition to a grid-connected case, two cases where the power from the renewable energy source (wind) is used to run the filling station autonomously are identified and discussed. 3. SYSTEM DESCRIPTION - DESCRIPTION DU SYSTÈME As mentioned above, the wind-H2 energy system is connected to a low-voltage grid (22 kV). The H2 system receives power from a 2 MW variable speed Wind Energy Conversion System (WECS). The produced hydrogen is stored as compressed gas at 300 bar. The simulations were done using the computer modelling code; TRaNsient SYStem simulation program (TRNSYS). TRNSYS is a modular simulation tool where components are connected much like in an actual system. For this study, detailed technical models developed at IFE since 1995 where applied. The models comprise the
Proceedings of the 18th World Energy Congress, 21-25 October 2001, Buenos Aires, Argentina
Integrating Renewables for Remote Fuel Systems
Glöckner et al., 2001
components: Alkaline electrolysers, hydrogen storage, power conditioning, compressors, WECS and system control units [1]. A schematic drawing of the wind-H2 energy system is shown in Figure I.
Figure I: Schematic presentation of the wind-H2 energy system - Diagramme du système de conversion d’énergie éolienne en hydrogène 3.1 Wind resources - Resources éoliennes
A three-year wind speed (m/s) data series with a time-step of 1 hour from a measurement station at a specific remote location along the Norwegian coastline was used as input to the simulation. The actual time period used in this study was: • • •
Start date/time: 1997-10-31, 00:00:00 Stop date/time: 2000-10-31, 23:00:00 No. of data points: 26304
Wind-speed data recorded mostly at a height of 50 meters (in smaller periods of time at 30 meters) where extrapolated to 67 meters, which is the hub height of the WECS used in the modelling. The wind data are presented in Figure II. 3.2 WECS power output - Rendement électrique du système de conversion d’énergie éolienne The WECS is of the type VESTAS V80 with a maximum power output of 2 MW. In order to simulate the power contribution from this windmill, a linear extrapolation to the measured data of wind speed and power output, as supplied by VESTAS Wind Systems A/S (Denmark), was made. A representation of the power output from the WECS as a function of wind speed is shown in Figure III.
2
Integrating Renewables for Remote Fuel Systems
Glöckner et al., 2001
35
vWIND (m / s)
30 25 20 15 10 5 0
1999
1998
1997
0
26
52
78
2000
104
130
156
time (weeks) Figure II:
Wind speed (in m/s) as a function of time (in weeks) for the 3 years period. In order to visualise summer and winter variations, a smoothed curve is also presented. Three winter seasons with good wind resources are seen - Vitesse du vent (en m/s) en fonction du temps (en heures). Le profil attenué est présenté pour visualiser les variations entre l’été et l’hiver. Trois saisons d’hiver avec des vents favorables sont montrés
2000 1750
vcutout = 25 m/s
PWECS (kW)
1500
vRATED = 14 m/s
1250 1000 750 500
vcutin = 4 m/s
250 0 0
5
10
15
20
25
v (m / s) Figure III: Power/wind speed characteristics for the VESTAS V80 (2MW) WECS. The line represents the linear extrapolation used in the simulations. No hysteresis loop is assumed upon reentering the valid wind speed range from wind speeds above 25 m/s - Énergie produite par le système VESTAS V80 en fonction de la vitesse du vent. La ligne continue représente le résultat de la simulation. Pas de cycle d’hystérese lorsque, une fois en dessus de 25 km/h, on revient dans la gamme de vitesses valables
3
Integrating Renewables for Remote Fuel Systems
Glöckner et al., 2001
3.3 Electroyser and storage - Electrolyseur et stockage d’hydrogène In this study electrolyser technology from Gesellschaft für Hochleistungselektrolyseure zur Wasserstofferzeugung mbH (GHW) is chosen [2]. GHW develops alkaline electrolysers with a nominal operating temperature of 70-80°C and a working pressure of 30 bar. The GHW electrolysers are especially advantageous with the unpredictability of wind resources as they can take variations in power ranging from 20% up to 100% of maximum rated power. Figure IV shows the voltage-current characteristics for the GHW electrolyser applied in the modelling.
2.0
Ucell, V/cell
1.8 Ucell Ucell,meas
1.6 GHW Electrolyzer p = 30 bar T = 80°C
1.4
1.2
0
200
400
600
800
1000
ic, mA/cm^2 Figure IV: Voltage-current (U-I) characteristics of the GHW electrolyser used in the simulation Relation entre la tension et le courant dans l’électrolyseur GHW utilisé pour la simulation Hydrogen is stored at a pressure of 300 bar. A compressor is used to increase the hydrogen pressure from 30 bar (the operating pressure of the electrolyser) to the storage pressure. The size of the storage unit depends on the operating scenario, as described in Section 4. 3.4 Mini-buses - Minibus The hydrogen load is represented by three PEM fuel cell powered mini-buses. The fuel demand and other technical information on the fuel cell mini-buses are listed in Table I.
Table I: Technical data for H2 fuelled PEMFC vehicles - Données techniques pour des vehicules avec pile à combustible Vehicle type
16-seat mini-bus
No. of vehicles
3
Onboard storage capacity
120 Nm3 @ 200 bar = ∼10.7 kg
Fuel consumption
0.6 Nm3.km-1 [3]
Driving range
200 km
Filling frequency
day-1
4
Integrating Renewables for Remote Fuel Systems
Glöckner et al., 2001
4. DEFINITION OF CASES - DÉFINITION DES CAS Three distinct cases with regard to the operation of the electrolyser are described: one case with constant operation (drawing electricity from the grid); and two cases with variable operation of the electrolyser (true stand-alone systems). 4.1 Electrolyser operated at constant power (Case 1) - Électrolyseur opéré à puissance constante (le Cas 1) The electrolyser can be operated at constant power. For a wind-H2 system, however, this will require that power be supplied from the grid during times of poor wind resources. Based on the information in Table I, 32 kg of hydrogen are required each day. We assume that the filling of all three mini-buses is possible within 1 hour (fast-fill). The size of H2 storage unit and the electrolyser may then be easily calculated. Based on manufacturers’ data, the energy required by the electrolyser, when operated constantly at maximum power, is 4.4 kWh/Nm3 H2 [2]. This gives a total energy demand per filling of 120 Nm3 x 3 x 4.4 kWh/Nm3 = 1584 kWh. In 24 hours this equals an electrolyser of capacity 1584 kWh / 24 h = 66 kW. The size of the storage is also easily estimated by taking the total amount of H2 needed per day, 32 kg / 0.08988 kg/Nm3 = 360 Nm3. At 300 bar, this is around 1.7 m3 physical gas volume (adding a lower and upper dead-band of 20% + 20% = 40%). For comparison, this is equivalent to about 50 x 50-liter bottles of commercially available H2 at 200 bar. The electrolyser and H2 storage unit sizes for Case 1 are summarised in Table II. The state of charge (SOC) of the H2 storage unit is shown in Figure V. Table II: Dimensions of electrolyser and H2-storage unit for Case 1 - Dimensions de l’électrolyseur et de l’unité de stockage d’hydrogène, le Cas 1. Component
Value
Unit
Electrolyser
66
kW
Compressed H2 storage unit
1.7
m3 (tank volume)
500
Nm3 H2
45
kg H2
SOC
upper dead-band
lower dead-band time Figure V: Schematic representation of state of charge (SOC) of H2 storage in the case where the electrolyser is operated at constant power (Case 1) - Diagramme représentant le niveau de charge (SOC) du stockage d’hydrogène dans le cas où l’électrolyseur opère à puissance constante (le Cas 1)
5
Integrating Renewables for Remote Fuel Systems
Glöckner et al., 2001
Continuous operation of the electrolyser requires that power be supplied from the grid at times of low wind resources. For Case 1, 19% of the energy needed to operate the electrolyser continuously at 66 kW must be supplied from the grid for the 3-year period under study here. The power from the WECS is insufficient to operate the electrolyser at full power for about 24% of the time. For sake of illustration, the electrolyser power profile for a critical wind resource period (04.08.98-21.08.98) is shown in Figure VI.
from WECS from GRID 70
Power (kW)
60 50 40 30 20 10 0 7200
7250
7300
7350
7400
7450
7500
7550
time (h) Figure VI: Power distribution for the electrolyser in Case 1 for a critical wind resource period. Approximately 73 kW are directed to the electrolyser (includes 10% loss in the inverter) Distribution de puissance électrique pour l’électrolyseur dans le cas 1 pour la période de vent faible
4.2 Stand-alone operation (Cases 2 & 3) - Opération de l’electrolyseur isolé (les Cas 2 et 3) In Cases 2 and 3, where the electrolyser receives power from the WECS only, the storage and electrolyser have to be somewhat over-sized. Figure VII illustrates the fraction of the power from the WECS that is directed to the electrolyser. If the power output from the WECS is between the lower and upper power limit for the electrolyser (20 - 100% of maximum power), then all power from the WECS is directed to the electrolyser. If, on the other hand, the WECS produces more power than the electrolyser can utilise, then the over-shooting power is distributed to the grid. Also if the power from the WECS is lower than needed for operating of the electrolyser at 20%, the power is instead distributed to the grid. 4.2.1
Direct-connect Scenario (Case 2) - Scénario de connection directe (le Cas 2)
In Case 2, the electrolyser is operated whenever there are wind resources and the storage unit is over-dimensioned accordingly. To determine the appropriate electrolyser and storage unit sizes, optimisation calculations were performed to achieve a design where no hydrogen is dumped by ventilation to the surroundings and where the storage unit is never entirely drained. The electrolyser and storage unit dimensions are given in Table III. Figure VIII shows the distribution of power from the WECS to the electrolyser in a selected ∼4 days period. The optimal solution for Case 2 shows that the
6
Integrating Renewables for Remote Fuel Systems
Glöckner et al., 2001
Power from WECS (arbitrary units)
electrolyser will only need to be over-dimensioned by ∼45% (from 66 to 98 kW). The SOC of the storage unit will reflect the variations in wind resources as a function of time and show peaks during wintertime and minima during summertime. The resulting storage unit SOC is shown in Figure IX.
PGRID PELECTROLYSER
PMAX (ELECTROLYSER)
PMIN (ELECTROLYSER)
time (arbitrary units) Figure VII: Schematic presentation of power distribution from the WECS. The grey area represents power distributed to the electrolyser while the hatched areas represent the power sent to the grid - Diagramme de la distribution d’énergie électrique venant du système de conversion d’énergie éolienne. La surface grise représente l’énergie fournie à l’électrolyseur, la surface hachurée représente l’énergie envoyée au réseau
Table III:
Summary of electrolyser and storage unit capacity for Case 2 - Données sur l’électrolyseur et sur l’unité de stockage d’hydrogène pour le Cas 2
Component
Value
Unit
Electrolyser
98
kW
Compressed H2 storage unit
50
m3 (tank volume)
11800
Nm3 H2
1060
kg H2
This control strategy – electrolyser operation whenever there are wind resources – results in an operation profile as presented in Figure X. Here we see that the electrolyser is operated at 80-100% of its maximum power for ~75% of the time and that it is idling ~10% of the time.
7
Integrating Renewables for Remote Fuel Systems
Glöckner et al., 2001
2000
PWECS PELECTROLYSER
1800
Power (kW)
1600 1400 1200 1000 800 600 400 200 0
760
780
800
820
840
time (h) Figure VIII: The distribution of power to the electrolyser from the WECS for a 100-hour period for Case 2. The horizontal dotted lines represent the upper and lower limits of electrolyzer operation - Distribution de puissance du système de conversion d’énergie éolienne (WECS) pendant une période de 100 heures pour le Cas 2. Les lignes horizontales pointillées représentent les limites d’opération de l’électrolyseur
1.0
State of Charge (SOC)
0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 1997 0.0
0
26
2000
1999
1998
52
78
104
130
156
time (weeks) Figure IX: Storage state of charge (SOC) for Case 2 for a 50 m3 storage unit at 300 bar (11800 Nm3). The dashed line indicates fulfilment of energy conservation over the 3-year period - Niveau de charge de l’unité de stockage d’dydrogéne de capacité de 50 m3 à la pression de 300 bars (11800 Nm3) pour le Cas 2
8
Integrating Renewables for Remote Fuel Systems
Glöckner et al., 2001
80
% of maximum power
70 60 50 40 30 20 10 0
IDLE
20-40%
40-60%
60-80%
80-100%
Power fraction Figure X: Operation profile for the electrolyser of Case 2 - Profil d’opération de l’électrolyseur pour le Cas 2 4.2.2
Top-charging Scenario (Case 3) - Opération du système à haut niveau de charge (le Cas 3)
In Case 3, the electrolyser operation is guided by the SOC of the H2-storage unit. This case represents a so-called top-charging scenario. The electrolyser and storage unit are dimensioned so as to keep the storage SOC within the dead-band when possible. This type of control routine is commonly used in stand-alone systems based on energy storage in batteries [1]. The lower and upper dead-bands were set to 0.80 and 0.94, respectively, for Case 3. Hence, the electrolyser is set to idling when the upper limit is reached and set to operate when the lower limit is reached. The aim of this strategy was to minimize the storage unit size. The storage unit size was set to 7.35 m3 at 300 bar. This is comparative to a commercially available storage solution comprised of a 20-feet container with 147 x 50 l bottles at 300 bar. In order to keep this storage unit from being exhausted at any time during the 3-year period, the electrolyser capacity had to be increasing to some 230 kW, an electrolyser capacity that is nearly 3.5 times that of the constant operation situation of Case 1. The electrolyser and storage unit dimensions for Case 3 are summarised in Table IV. Table IV: Summary of electrolyser and storage unit size for Case 3 - Données sur l’électrolyseur et sur l’unité de stockage d’hydrogène pour le Cas 3 Component
Value
Unit
Electrolyser
230
KW
Compressed H2 storage unit
7.35
m3 (tank volume)
1743
Nm3 H2
156
kg H2
9
Integrating Renewables for Remote Fuel Systems
Glöckner et al., 2001
Figure XI shows the distribution of power from the WECS in a selected one-week period. Figure XII shows the SOC for the H2 storage unit of Case 3. Dips in SOC will appear at times of continuously low wind resources and will set the storage unit size for a given electrolyser capacity.
2000
PWECS PELECTROLYSER
1750
Power (kW)
1500 1250 1000 750 500 250
0 300 310 320 330 340 350 360 370 380 390 400
time (h) Figure XI: The distribution of power to the electrolyser from the WECS for a 100-hour period for Case 3. The horizontal dotted lines represent the upper and lower limits of electrolyzer operation - Distribution de puissance du système de conversion d’énergie éolienne (WECS) pendant une période de 100 heures pour le Cas 3. Les lignes horizontales pointillées représentent les limites d’opération de l’électrolyseur
1.0
State of Charge (SOC)
0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 1997
0.1 0.0
1999
1998
0
26
52
78
2000
104
130
156
time (weeks) Figure XII: Storage state of charge (SOC) for Case 3 for a 7.35 m3 storage unit at 300 bar (1743 Nm3) Niveau de charge de l’unité de stockage d’hydrogène de capacité de 7.5 m3 à la pression de 300 bars (1743 Nm3) pour le Cas 3
10
Integrating Renewables for Remote Fuel Systems
Glöckner et al., 2001
Figure XIII shows the operation profile for Case 3. In the top-charging strategy of Case 3, the electrolyser operates at near full power ∼30% of the time, while it is idling ∼60% of the time.
% of maximum power
60 50 40 30 20 10 0
IDLE
20-40%
40-60%
60-80%
80-100%
Power fraction Figure XIII: Operation profile for the electrolyser of Case 3 - Profil d’opération de l’électrolyseur pour le Cas 3 5. CONCLUSIONS - CONCLUSIONS Based on calculated component sizes, Case 1 will be the least costly in terms of capital investment. This is because both the electrolyser and the storage unit are held at the lowest possible size/capacity for this case. Case 2 seems unrealistic in that it requires a storage unit with a capacity of 11800 Nm3, which is both practically difficult to achieve and also associated with high investment costs. Case 3 also requires an over-dimensioning of the electrolyser and storage unit. To make Case 3 more realistic and avoid prolonged periods of idling, the electrolyser should be operated such that it is never switched off, but regulated between 20% and 100% power. This will, however, not contribute to a reduction in the electrolyser size given the size of the storage unit or vice versa. The storage-unit size will still be dependent on the length of the periods of low wind resources and the electrolyser size set. The storage-unit size of Case 3 can be somewhat reduced by setting the lower limit of the dead-band to a higher value. This will have an impact on the average time of operation/idling for the electrolyser. With a very narrow bandwidth, the electrolyser will be switched between idling and H2 production at a higher frequency. With regard to start-up time and handling of fast transients, this may not be an acceptable electrolyser operation control strategy. Depending on the electrolyser solution chosen, these possible restraints should be added to the modelling code. 6. FUTURE WORK - TRAVAIL FUTUR In this work, a known hydrogen load (3 mini-buses) was selected, but models with less stringent requirements should also be developed. One scenario of interest would include adding a small- to medium-sized fuel cell to the system, enabling distribution of electrical power to an application or to
11
Integrating Renewables for Remote Fuel Systems
Glöckner et al., 2001
the grid in times of low consumption of hydrogen. This would make the system more flexible. In addition, cost models should be included in the model library to enable users to identify break-even points with regard to control strategies and type of technology installed. Finally, the potential for hydrogen energy systems for load levelling in weak grids should also be investigated. 7. ACKNOWLEDGEMENTS - REMERCIEMENTS The authors wishes to acknowledge Kjeller Vindteknikk AS (KVT - wind consulting company) for the fruitful discussions and aid in constructing the complete wind data series. 8. REFERENCES - RÉFÉRENCES [1] [2] [3]
Øystein Ulleberg, ”Stand-alone power systems for the future: Optimal design, operation & control of solar-hydrogen energy systems,” PhD thesis at Norwegian Technical University, Trondheim (1998) Technical data communicated by H. Hofmann of MTU, Friedrichshafen-GmbH, Germany Communicated by R. Hamelmann at Proton Motor GmbH, Germany
12
