- Email: [email protected]
Hybrid renewable energy-fuel cell system: Design and performance evaluation
Electric Power Systems Research 79 (2009) 316–324
Contents lists available at ScienceDirect
Electric Power Systems Research journal homepage: www.elsevier.com/locate/epsr
Hybrid renewable energy-fuel cell system: Design and performance evaluation Sonia Leva ∗ , Dario Zaninelli 1 Dipartimento di Elettrotecnica, Politecnico di Milano, Piazza Leonardo da Vinci, 32, 20133 Milano, Italy
a r t i c l e
i n f o
Article history: Received 6 December 2006 Received in revised form 21 February 2008 Accepted 2 July 2008 Available online 19 August 2008 Keywords: Fuel cells New/alternative energy resources Photovoltaic power systems System costs
a b s t r a c t The paper introduces hybrid photovoltaic-wind-diesel generation systems supplying a remote power load considering the advantages of sustainable energy from the economic point of view. In particular, a cost investment valuation is performed on a real plant showing the effect of sustainable economical saving. The possibility to introduce a fuel cell generation device in a photovoltaic-wind existing plant for supplying Telecommunication apparatus is also investigated and the results are reported and discussed in the paper. Furthermore, starting from measured data, a control system is realized in order to verify the functionality of the plant. © 2008 Elsevier B.V. All rights reserved.
1. Introduction A high integrated hybrid system (HIHS) can be seen as a way of using renewable sources, to produce electric energy without pollution in stand-alone applications supplying remote load. The photovoltaic (PV) energy production has been mainly used: • in industrialized countries both for grid connected applications and for supplying remote loads with low consumption; • in underdeveloped countries for generating electricity with reduced need of maintenance and with low operation costs. Research and development makes PV technology suitable for industrial applications related to telecommunications, water pumping, water purification, public lighting, etc. In Europe (where the weather conditions are very variable and the forecasts are available for a limited period ahead of time) the PV system needs to be oversized and integrated with batteries and diesel generator set (DGS) [1,2]. Because of these limits the PV has been rarely chosen as energy supply in off grid industrial application which requires more than a few kWh/day. The wind generator (WG) is applied in Europe in case of grid connected systems, but is not so diffused for supplying remote load, although its reduced watt peak cost compared to the PV solution.
∗ Corresponding author. Tel.: +39 02 23993709. E-mail addresses: [email protected] (S. Leva), [email protected] (D. Zaninelli). 1 Tel.: +39 02 23993721. 0378-7796/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.epsr.2008.07.002
The European weather conditions, in fact, do not allow the construction of PV only and/or WG only stand-alone plants, despite the availability of these technologies for energy supply in off grid systems. The HIHS instead allows us to project a power supply system capable of guaranteeing the continuity of the supply, mixing the different renewable energy resources – like PV, WG (even microhydro if possible) – limiting the use of DGS for short periods during winter and/or for backup purposes only. On the design point of view the choice of the generator sizes of the HIHS plant is very important, so that this leads to a good ratio between cost and performance [3]. The HIHS is usually designed to lower the fuel consumption of the diesel machine in the range of 70–90%, compared to only battery–diesel supply architecture overcoming the DGS drawbacks [4]. Therefore, the primary source of electric energy for the plant becomes the renewable one, while the diesel machine constitutes the auxiliary source in case of emergency or for battery charge. Furthermore the hybrid systems can be very effective in terms of sustainable energy development, permitting generation of electric energy with minor environmental interference. In this paper, a technical-economic analysis of a real plant is presented, taking into consideration both plant costs and sustainability costs [5–7], comparing three different possible plant configurations: PV only, PV and DGS, HIHS composed by PV, WG and DGS. The plant under study is used to supply a telecommunication system. This load imposes a fundamental constraint: the use of telecommunication apparatus for the community and emergency purposes makes possible black outs unacceptable. The main characteristic of this system is the need of a continuous power supply, nevertheless on the economic point of view, an oversize system would lead to an excessive increase of the cost of the energy produced.
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
317
Fig. 1. Hybrid system layout.
Furthermore, an improved HIHS composed of PV, WG, hydrogen fuel cell (FC), water electrolyser, batteries and a DGS – as backup device – is presented and analyzed from an economical and technical point of view. In fact, FC devices are a very interesting option to be used with intermittent sources of generation like the PVs because of high efficiency, modularity and fuel flexibility. In the future both DGS and batteries could be removed or reduced according to the experience by FC devices [7,8]. The technical literature in this field presents hybrid systems with FC together with only one renewable source (PV or WG) and economical analysis are seldom dealt with [3,4,9]. The HIHS, including PV generator, WG and DGS, without FC, is already operates. The FC and electrolyser implementation are at the moment in phase of study. An economical and environmental evaluation is necessary (Sections 4 and 5) before the experimentation on field and a preliminary study of the complete system has to highlight technical problems that can occur. For this reason a simple control system – starting from monthly/daily average data measured on the existing plant in 2005 – will be designed in Section 6. 2. Hybrid system
the wind diagram. Based on these diagrams, it is possible to calculate the PV module size and number, and the WG size. The system sizing is quite simple but the difficult part is making the best choice because of the number of possible choices. Focusing on system sizing, it is important to note that, in the considered site, PV and WG gives its peak production in different periods of the year. Typically the PV reaches the best performance during spring and summer while WG has its peak production during the winter. Starting from the load diagram shown in Fig. 2, from the local solar radiation diagram and from the wind map, taking into account the cost of installation and operation, the HIHS is composed by 1. PV modules sized for a 100% load supply from April to September. The result of this sizing brings to a PV array made by 64 modules, 75 Wp and 48 Vdc each. The total power is 4800 Wp. Fig. 3 shows average monthly PV energy production in kWh/ month. 2. WG sized to supply the lack of PV power from October to March. The WG is of 500 Wp and it produces power with wind speed
2.1. System layout The plant layout commonly used, according to Fig. 1, is made up of a set of photovoltaic (PV) modules, a WG, batteries and a DGS. The DGS works for short periods during winter and for backup purposes. Furthermore, its presence increases the overall system reliability. The batteries are used for the storage of energy for a short time and for realizing the decoupling of the energy production from the consumed one. The whole system is managed and controlled by a complete controller, developed for these purposes [10]. 2.2. System sizing The required inputs for the overall system sizing are substantially three: the load diagram, the solar exposition diagram and
Fig. 2. Load diagram.
318
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
Fig. 3. Load diagram and energy production of PV, WG and DGS of the HIHS.
Fig. 4. Energy surplus of the system.
lower than 2 m/s and higher than 150 km/h. Fig. 3 shows the average monthly WG energy production [9].2 3. DGS sized for backup purposes and expected to work for short periods during winter but also to guarantee the necessary reliability of the system. This leads to a DGS with 5 kW and fuel consumption of 3000 gkW/h. In this way, ignoring possible black outs, the DGS normal working time is about 47 h/year, with the need of 104 l/year of diesel, one maintenance operation and two refuelling operations per year. 4. Battery set of 2000 Ah 48 Vdc, C10 type.
be carried out both on site or by a remote station by means of a communication protocol. • The capability of the system to transmit warnings to a remote station, pointing out the anomalous operations of the plant. • The presence of a local memory card for environmental and electrical data collection and processing.
The main economic advantages of this sizing, not considering at the moment the environmental costs, are substantially related to the reduction of PV arrays size in respect to the PV-only solution,3 and the saving on DGS refueling and maintenance operations compared to the PV + DGS without WG solution.4 Fig. 3 shows the average monthly energy balance – based on 2005 data – between the energy production and consumption of the hybrid PV–WG–DGS system mentioned above. It is possible to see that, despite the reduction of the PV arrays, there is still an energy surplus for almost all the year. The use of this surplus and the control of all the operation parameters is done by an electronic control and monitoring system designed for the operation of an HIHS that has been presented in [10]. The main characteristics of the control system are summarized in the following: • The possibility to measure electric quantities (voltages, currents, frequency of the ac supply) on the plant and external quantities (temperature, wind speed) useful for the efficiency optimization. • The regulation of the battery power flow to avoid lifetime reduction or operating conditions that can decrease the battery reliability. • The definition of a protection and warning system able to prevent fault conditions. • The possibility to check the present operation of the plant and the previous data recorded in the memory, together with the possibility to set the warning thresholds. These operations can
2 Comparing the PV and WG energy production it is possible to see that even if the WG size is about 1/10 of the PV size, its contribution to the energy production is about 1/5 of the PV production. 3 For “PV-only” system the array is made by 116 modules, 75 Wp and 48 Vdc each, with a covered area of about 100 m2 and with a battery set of 2000 Ah 48 Vdc, C10. This system is over-sized but does not protect from possible black outs due to unexpected weather conditions then a DGS for buck up purpose is present. 4 For PV + DGS system the array is made as for HIHS, the DGS normal working time is about 465 h/year, with a fuel consumption of 1025 l/year. This implies 11 maintenance operations and 20 refuelling of the tank per year.
Thanks to these properties on the collected data, it is possible to manage the HIHS and to modify, if necessary, the control algorithm in order to optimize the system and work with respect to the local conditions. 3. Fuel cell system The chosen layout leads to an energy production surplus (Fig. 4). This surplus – calculated neglecting the possible DGS production – is due to renewable production. Several energy storage techniques, like batteries, fly-wheels and hydrogen production are available for system integrators when energy storage is needed [2]. The main considered parameters during sizing are energy storage capacity, source availability, and response time of the energy source. Normally batteries present good modularity, fast time response and good energy–mass ratio storage, but they need maintenance and are subjected to fault. In addition the batteries performance depends on working temperature that has to be controlled during operation time. These characteristics make batteries suitable for hybrid systems. Fly-wheel presents little modularity, very fast time response, but the stored energy is available for the system for very a short time with respect to batteries; on the other hand a plant with a fly-wheel presents less system complexity and practically minimal or no maintenance [11,12]. FC, compared with the two previous techniques, presents modularity and the possibility of partial load working; good energy storage capability that can be simply obtained with correct hydrogen tank sizing. One main weak point of the FC is its very slow dynamics. Maintenance problems of fuel cell are still an open problem because no registered data of the working lifetime of the catalyst is available [13,14]. A way to increase the efficiency of the HIHS system consists in substituting the DGS backup set with FC. In this way, it is possible to use the surplus energy to produce, with a hydrolysis process, the hydrogen needed for the FC. 3.1. System layout The system layout remains the same as the previous one except for the presence of the fuel cell, hydrogen and water tank (see Fig. 5).
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
319
Fig. 5. Fuel cell system layout.
We assume that FC will substitute the DGS. The DGS presence is related to the lack of practical experience in systems with FC and their reliability. Using this plant as a test prototype it is possible to verify, from a practical point of view, the energy availability all over the year. A further step, after a data collection phase on the plant during a few years, will be a plant design without the DGS backup set. 3.2. System sizing The hydrogen can be produced, during the surplus of energy production, from water by an electrolyser and stored in a container for further use. Assuming a hydrogen production rate of about 0.018 m3 /kWh, the surplus energy can be used to produce about 18.6 m3 /year. With low temperature fuel cell, neglecting the use of thermal energy, its efficiency can be considered to be about 30%. The lower heating value of hydrogen is 10.72 MJ/m3 (38.6 kWh/m3 ) [14]. Therefore the total electric energy that can be produced by the FC is EH2 = VH2 LHVH2 = 215 kWh/year
(1)
where is the efficiency of the fuel cell, VH2 is the total net volume of hydrogen produced per year and LHVH2 is the lower heating value of hydrogen. In this way, although the hydrogen can be produced only during spring and summer and needs storage, the use of an FC can lead to a complete autonomy of the plant. The estimated efficiency of the complete process can be calculated as follows: ε = 0.018 × 0.3 × 38.6 = 0.208
(2)
corresponding to about 21%. In this paper, the whole system is analyzed only from the electrical point of view, neglecting the thermal energy produced by the FC. This energy can be used, during winter, either for avoiding ice formation in the water tank or for local warming of the apparatus.
The plant, including PV generator, WG and DGS with the mentioned characteristics, is already in function. The FC and electrolyser implementation are at the moment in phase of study. Before the experimentation on field an economical and environmental evaluation is necessary (Sections 4 and 5) and a preliminary study of the complete system that must clarify the technical problems that can occur. For this reason a simple control system – starting from monthly/daily average data measured on the existing plant in 2005 – will be realized (Section 6). The model allows studying the systems with FC and electrolyser shown in Fig. 5, even if with some approximation. 4. Environmental costs Environmental costs of energy production are normally treated as an externality. This means that the influence of the energy production on the environment does not find a corresponding cost on the bill. Until some years ago all the climate change warnings, mainly related to the greenhouse gases emission, involved studies made in order to reduce the gas production in energy cycles and calculate the economic value of the externalities. The aim of this evaluation is the quantification of an environmental fee related to the quantity of pollution emission, normally carbon-dioxide. One of the first studies made by the European Community on the externalities evaluation of energy production was related to the Non-Nuclear Energy (NNE) program, a research and development specific program within the European Union’s Fourth Framework Program for Research and Technological Development. The main issues of this program were • improving security of energy supply; • protecting the environment; • encouraging the rational use of energy. The NNE was structured in two components NNE-JOULE dealt with research and development in the field of innovative energy technologies, and THERMIE which focused on the demonstration
320
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
Table 1 Externalities values of energy production
HIHS and HIHS + FC type. The following data refer to real application on an existing plant.
Description
Externalities (centD /kWh)
PV WG DGS
0.112 0.050 3.376
of these energy technologies. The European project THERMIE [5] in 2002 lead to a cost for producing energy with fossil sources that can be estimated in 0.20 D /kWh. This esteem of the externalities value had a constant value in time [15]. Another externality evaluation method, based on the economic value proposed by THERMIE can be derived observing that the growth of the carbon dioxide emission over the last 50 years is quasi-linear [5,17]. It is possible to obtain a formula that evaluates, linearly, the externalities related to the emissions of the plant for the future years. Furthermore, evaluation of the externalities was developed with the ExternE Project, developed within the European Commission R&D program Joule II, which proposed a unified methodology for the quantification of the externalities of different power generation technologies. Under successive research programs this methodology has been developed along three different lines: the first, involved with the further development and updating of the methodology performance called ExternE Core; the second, mainly a data collection phase called ExternE National Implementation (in order to create an European data set); and the third, the applicative one, called ExternE Transport, in which the proposed methodology is applied to the energy related impacts from transportation. In more detail the ExternE project evaluated the environmental costs of several traditional fuel cycles and of several renewable energy cycles5 [16,18]. The recent evaluation technique, developed under the ExternE research program, is based on the impact pathway method [19]. Basing on the project results it is possible to evaluate these costs and associate them with the cost of energy production. Then, it is possible to compare the proposed plant layouts and calculate an environmental net present value (NPV), based on both the plant cost and externalities, determining the correspondent energy cost. The application of the externalities is done evaluating the ratio of the single energy source production with respect of the total energy production and then applying the economic values reported in Table 1. The values of Table 1 are obtained from ExternE summary for PV, WG and DGS [16,18]. These values and the one of European project THERMIE will be used in the economic analysis in order to derive a cost for the produced kWh. 5. Economic analysis Scope of this section is to estimate the kWh cost and therefore the economic convenience by using the Unit Electricity Cost (UEC) method referring to the analyzed different plants solution [6,20].
5.1.1. Diesel generator, fuel and maintenance costs The DGS has and indicative cost of about 5100 D . Focusing on the fuel costs, in the PV + DGS plant the generator works 465 h per year. The fuel consumption is about 1025 l and its cost is about 1.43 D /l. The yearly fuel cost amounts to 1463 D . In the HIHS type plant the diesel generator works only 50 h/year with a yearly fuel cost of 150 D . The diesel generator maintenance costs are calculated on the number of maintenance events per year. The PV + DGS type plant needs 11 maintenance operations per year with an overall cost of 5500 D , while HIHS type plant needs only two maintenances per year with an overall cost of 1000 D . 5.1.2. Photovoltaic modules The system under analysis is built of 64 modules of 75 Wp each. The average price of each module is 266 D . So the PV modules cost is about 17,000 D . The PV modules have to be installed on support structures in order to obtain the design inclination, for a total price of 878 D . 5.1.3. Batteries The batteries have a capacity of 2000 Ah and the cost is about 4080 D . The chemical protection system price is assumed equal to 1275 D . The battery charger on the basis of the market analysis is evaluated in about 850 D . 5.1.4. Wind generator The system is provided with a small sized WG with 500 W power. All the accessories needed for its installation and normal operations, like supports, are available on the market. The overall cost of the generator and accessories is valuated at about 4200 D [10]. 5.1.5. Fuel cell The fuel cell, which has 1 kW of rated power at 48 Vdc, is a PEM type supplied directly with hydrogen gas. Its cost is about 3000 D . The fuel consumption is 15 l/min of industrial grade hydrogen. The hydrogen is provided by an electrolyser fed by water which has been treated by a deionizer. The cost of the hydrogen generator and water deionizer is about 7000 D . The hydrogen is produced, directly at high pressure, by the electrolyser and stored in a storage tank, made of stainless steel, for further use. The hydrogen tank capacity is 120 l at 1 MPa at room temperature, its cost is about 1000 D . The hydrogen transmission pipes are completed with pressure regulators and safety valves [14]. 5.1.6. Installation The evaluation of the installation costs is done with the labor price taken from Milan’s Chamber of Commerce in 2005. It is assumed that three people form the working team involved in the installation of the plants. The installation of all the considered plants can be done in about 15 working days, ten of them in which the whole team is present and five of them in which the most qualified worker is not present.
5.1. Cost evaluation of the plant hardware The first step for the economic analysis consists in the evaluation of the overall costs of the three plants under examination: PV + DGS,
5
The considered fuel cycle are both conventional sources like coal, lignite, peat, gas, oil, orimulsion, nuclear and renewable energy like biomass, hydro, wind, photovoltaic, and waste incineration.
Table 2 Plants costs Plant
Installation (D )
PV only PV + DGS HIHS HIHS + FC
5770 5770 5770 5770
Hardware (D ) 43,335 32,625 36,840 48,250
O&M (D ) 1040 6925 1150 4040
Total (D ) 50,145 45,360 43,760 58,060
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
Table 2 summarizes the costs of the considered plants. A comparison of Table 2 data shows that the Operation and Maintenance (O&M) cost of a HIHS is cheaper than the other plants. This is substantially due to the reduction of periodic maintenance and refuelling operations. In the case of FC, the costs are higher because of the absence of a real market for FC maintenance. FC costs should be reduced in the next years because of the increase of the FC production.
321
Table 3 UEC For different type of plants Plant
UEC (D /kWh)
UEC with environmental costs (D /kWh)
UEC with externalities (D /kWh)
PV only PV + DGS HIHS HIHS + FC HIHS + FC (no DGS)
0.507 1.562 0.440 1.047 0.849
0.514 1.580 0.443 1.053 0.849
0.509 1.566 0.442 1.049 0.850
5.2. Produced energy cost The evaluation of the cost of the produced energy by the real plant is made by using the life-cycle costing (LCC) [6,20]. The parameters for LCC evaluation are period n: the period of time during which the system is under analysis, that is, the greatest expected lifetime of the component; inflation i; discount rate d (the applied active rate d = 2–5%); cost of money and the overall hardware and installation costs; operation and maintenance costs; fuel costs; environmental costs: economic value due to pollution emitted in the atmosphere; substitution costs of every part at the end of its operative life. The LCC evaluation needs that all costs and benefits, present and future, are brought back to present value or present worth (PW). The present value of the amounts is calculated, as function of the period of the payment, using two different updating factors: single payment PWr = Cr
1 + i N 1+d
= Cr Pr
(3)
annual payment PWa = Ca
(1 + i/1 + d)(|(1 + i/1 + d)|N ) − 1 = Ca Pa ((1 + i)/(1 + d) − 1)
(4)
where Cr is the cost of year r and N is the expected lifetime of the part. The Ca is the recurrent annual cost. Pr and Pa are respectively the payment evaluated in the cash flow. The sum of all PW gives the LCC. The produced energy cost is calculated by means of the annualized life-cycle costing (ALCC): ALCC =
LCC = Pa
D
(5)
year
which leads to the UEC by means of the annual produced energy given by the ratio between the ALCC and the annual energy E produced by the plant: UEC =
ALCC = E
D
kWh
(6)
The UEC for the three considered plants are reported in Table 3. These values have been calculated only accounting the costs reported in Table 2, and adding the environmental costs indicated by European project THERMIE and obtained from ExternE summary (see Table 1). The UEC is a very significant indicator for performance evaluation of energy generation systems and gives, practically, the kWh generated cost. The UEC for different plant types, with a range of −20% around the normal produced energy, are evaluated and reported in Fig. 6. This allows to calculate possible UEC variations due to black outs.
Fig. 6. UEC of produced kWh for the analyzed plants in a range of −20% around the normal produced energy without sustainability costs.
Data reported in Table 3 shows that HIHS lead to lowest UEC. Anyway, the HIHS + FC plant, even if it is a prototype, is cheaper than the PV + DGS plant. Furthermore, comparing the data reported in Table 3 with the ones reported in [5], it is possible to assert that the environmental cost is strictly related to the DGS emissions. Finally, the UEC in the case of HIHS + FC without DGS, are evaluated. These costs are today only an attempt, but in the future, when FC will have a market spread with reliable operations, the economic impact will be more precise. 6. Control system and simulation results In this section, we briefly discuss the simulation model used in this paper. The simulation will be extended for a temporal window of 24 h in two different conditions: one regarding a typical day of May when the energy production is high, the other one corresponding to a typical day of November when the energy production is not enough to power the load. In the first case the energy surplus is deviated towards the aqua electrolyser, storing hydrogen in the tank. In the second case the FC will provide the missing energy taking the hydrogen from the tank. The system analysis will be done considering initially the plant including the battery, but without the DGS. Subsequently the possibility of eliminating the battery will be discussed. 6.1. Control and simulation system [8] The models of the different parts of the system are shown in Fig. 7 and it is composed of the following subsystems: load regulator, wind and FV generators, aqua electrolyser, FC and battery. The load regulator can be seen as a simple transfer function with a rapid response: Freg =
1 s + 0.1
322
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
Fig. 7. Control and simulation system.
The process, modelled as above, is controlled in retroaction by a PI regulator with parameter determined by the trial and error method. The regulator receives as input the error between the required power and the regulated one. The WG is modelled as a couple of random signals overlapped. The first one with a long period (1 h) and average equal to the estimated average power produced during the day. The second one, uniform type, with a short sample time (3 min) necessary to simulate sudden variation in wind direction and power. Properly setting the variation between the mean values, a good approximation with respect to typical wind generation performances is achieved. The photovoltaic source is modelled starting from measured values included in a. mat file that gives the value of the produced instant power for each simulation time. As expected, a sort of Gaussian curve is observed. For the model of the electrolyser it is considered a transfer function as follow: Freg =
1 s + 0.2
It takes into account, as a simply way, the slight delay introduced by the device. The electrolytic process is controlled by a PI regulator another time with characteristic set by the trial and error method. The reference for the regulator is null when the battery is charged within a certain value and equal to the energy in excess otherwise. The limit for the battery charge is determined by empiric method in such a way to obtain a decent value of charge at the end of the day. Finally the hydrogen produced is calculated as integral of the power transformed and multiplied by the following constant: VH2 ,elect = 0.018 (m3 /kWh) = 5 × 10−9 (m3 /J) Similarly the fuel cell is modelled by a single pole transfer function, but with a long period time constant that correspond to a
slower kinetic of the cell: Freg =
1 s+4
Also in this case the reference for the regulator is null when the battery is charged within a certain value and equal to the energy in deficit otherwise. The limit for the battery charge is determined by empiric method in such a way to obtain a decent value of charge at the end of the day. The consumed hydrogen is calculated as integral of the power transformed and multiplied by the following constant: VH2 ,cell = 0.3 × 38.6 (kWh/m3 ) = 41.7 × 106 (J/m3 ) The model of the battery is obtained integrating the current absorbed by the charger and limiting the total charge to a suitable value determined in such a way to obtain an acceptable steady state condition. This model represents a really simplified approach that introduces an approximation not negligible. In fact the charge of a battery depends also on the velocity on which the battery is charged, but taking into account all this variables leads to a really complex model not needed for the scope of the study. The power absorbed by (or fed from) the battery is equal to the sum of the produced one minus the one absorbed by the load: Pbatt = Pw + Psv − Pload From which it is possible to calculate the current simply dividing per the voltage (in this case 48 V): Ibatt =
Pbatt P = batt Vbatt 48
The charge accumulated is calculated integrating this value. It shall be noted the it is necessary to properly set a integration constant and limiting the value of the output in order to avoid the windup phenomena.
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
323
Fig. 8. Most significant data obtained in May.
Fig. 9. Most significant data obtained in November.
6.2. Simulation results Considering a typical day in May, when the energy production is high because of the PV production, the controller is programmed to use the energy surplus towards the aqua electrolyser in order to store hydrogen in the tank. The simulation model used in this condition includes the load power regulator, the generators, the model of the battery charge and the aqua electrolyser. The energy supplied in the first hour of the day is not enough to cover the entire requirements of the load, therefore this will be supplied from the battery that in the meantime endures a partial discharge. When the PV production rises, the power generated becomes greater than the energy required so the battery can be recharged. When this process is finished the energy in excess (Fig. 8) is deviated towards the aqua electrolyser which produces the hydrogen for the FC: for the considered day the system can produce and store of almost 0.112 m3 of hydrogen. This means a corresponding energy per year estimated in 1148.74 kWh. It must be noted that the control system is studied to guarantee a high battery charge level in order to always supply the load for 4 or 5 days in case of breakdown or particularly unfavorable environmental conditions. In a typical November day, when the energy production is low, the controller will provide the energy deficit taking hydrogen from the tank and using it in the fuel cell. The simulation model used in the second condition is made up of the load power regulator, the generator, the model of the battery charger and the FC. The results obtained from the simulation are synthesized in Fig. 9. In the early morning the discharge of the battery is limited by the FC, in order to avoid the slow but progressive discharge of the battery, that in the arch of some days – above all if weather conditions are unfavorable to WG or FV – would lead to the impossibility of powering the load correctly. However, in a few hours the power production rises because of the PV energy so the FC can be turned off and the battery partially recharged. The hydrogen consumed during the entire day is approximately 0.07 m3 . The total annual energy consumed by the FC is approximately 54.24 kWh.
Therefore, we can see that the amount of hydrogen produced in the summer months exceeds the load needed, even if we must consider conditions of breakdown that could lead to different energetic balances. Moreover, it is useful to emphasize that using the FC leads to the same result as that of using DGS. In fact, DGS would work in the winter months in almost the same way as the FC, while in the summer months the energy surplus would be wasted. Finally, we analyzed the possibility of eliminating the battery group and using the FC also for the management of the energy on the short period. In this case the FC must be able to supply the load all by itself. This would mean an increase in the use of the more technologically advanced and expensive devices reducing the long-term costs. In Fig. 10, we can be seen, in comparative way, the necessary and produced hydrogen. Considering a typical day of May, the system needs almost 0.13 m3 more then the 0.112 m3 hydrogen produced, corresponding to energy excess from PV of 7.2 kWh/day. During a typical day of November, this amount increases to 2.2 m3 corresponding to an energy excess of 122 kWh/day. As we can observe the system demand is not satisfied. In order to cover the power demand, an excessive over-sizing of the PV plant would be necessary, with a consequent and inconvenient increase of the costs. It is important to underline that these results, from an energy balance point of view, are in agreement with that obtained by other
Fig. 10. Hydrogen produced and necessary in case of battery absence (m3 /month).
324
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
authors [3,4] that have analyzed a simpler system from the dynamic point of view. In fact a typical self-sufficient HIHS supplying remote plant must include short-term and long-term energy storage. The combination of a battery bank – commonly used for short-term energy storage – with long-term energy storage in the form of hydrogen can significantly improve the performance of stand-alone renewable energy systems. 7. Conclusions The paper deals with a technical and economical analysis of possible implementation of renewable energy (photovoltaic and wind energy) sources on an existing plant for supplying isolated telecommunication devices. In particular the possibility of substituting the diesel generator backup group with a fuel cell is investigated both through the energy balance and as an economic investment. The results show that the energy balance is satisfied and the plant can work properly with fuel-cell devices, too. The energy storage is done with a diesel oil tank for the first layout, and with hydrogen, and water tank, for the second one. The water tank is necessary only in order to improve the autonomy of the plant. Maintenance and refueling with the diesel generator are the major problems because of the periodic maintenance required by the diesel machine, which works only during winter time. Using the fuel cell technology, the need of periodic maintenance and refueling operation is strongly limited or removed. Furthermore the real case application, based on an existing remote telecommunication installation, allows us to compare different plants (PV only, PV + DGS, HIHS, HIHS + FC) economically and technically and show the advantages of the HIHSs. Considering the environmental cost, the photovoltaic-wind-fuel cell configuration, having near-zero emission, also permits an increase of the sustainability and so a decrease of the external costs in the electric power generation. Finally, the possibility of substituting the storage battery with fuel cell is analyzed. Unfortunately with respect to the present technologies and the efficiency that corresponds, this solution is not convenient both from the technological point of view and from the economic one. In spite of this, future developments appear feasible: it cannot be excluded that substantial progresses will be made in the next few years. References [1] T.M. Calloway, Autonomous photovoltaic-diesel power system design, in: Proceedings of IEEE Photovoltaic Specialists Conference, Las Vegas, Nevada (USA), October 1985, pp. 280–284. [2] A. Chaurey, S. Demi, Batter storage for PV power systems: an overview, Renewable Energy 2 (3) (1992) 227–235.
[3] K. Agbossou, M. Kolhe, J. Hamelin, T.K. Bose, Performance of a stand-alone renewable energy system based on energy storage as hydrogen, IEEE Transaction on Energy Conversion 19 (September (3)) (2004) 633–640. [4] Z. Jiang, Power management of hybrid photovoltaic—fuel cell power systems, in: Proceedings of IEEE Power Engineering Society General Meeting, Montreal, Canada, June 18–22, 2006. [5] F. Iannone, S. Leva, D. Zaninelli, Hybrid photovoltaic system and sustainability: economic aspects, in: Proceedings of IEEE PES General Meeting, Denver (CO), USA, June 6–9, 2004, pp. 1933–1938. [6] S. Sampattagul, S. Kato, N. Maruyama, A. Nishimuram, Comparison of coal-fired natural gas-fired power plants as economically viable and ecologically sustainable power generation systems, International Journal of Emerging Electric Power Systems 3 (2) (2006) (Article 1116). [7] F. Iannone, S. Leva, D. Zaninelli, Hybrid photovoltaic and hybrid photovoltaicfuel cell system: economic and environmental analysis, in: Proceedings of IEEE PES General Meeting, San Francisco (California), USA, June, 2005, pp. 1503–1507. [8] R. Contino, F. Iannone, S. Leva, D. Zaninelli, Hybrid photovoltaic-fuel cell system controller sizing and dynamic performance evaluation, in: Proceedings of IEEE PES General Meeting, Monreal (Quebec), CANADA, June, 2006, 6 p. [9] T. Seniyu, D. Hayashi, N. Urasaki, T. Funabashi, Oscillation frequency control on H∞ controller for a small power system, using renewable energy facilities in isolated island, in: Proceedings of IEEE Power Engineering Society General Meeting, Montreal, Canada, June 18–22, 2006. [10] W. Dalbon, S. Leva, M. Roscia, D. Zaninelli, Hybrid photovoltaic system control for enhancing sustainable energy, in: Proceedings IEEE PES Summer Meeting, Chicago (IL), USA, July 21–25, 2002, pp. 134–139. [11] http://www.igetcol.it/. [12] M.L. Lazarewicz, A. Rojas, Grid frequency regulation by recycling electrical energy in flywheels, in: Proceedings of IEEE PES General Meeting, Denver (CO), USA, June 6–9, 2004, pp. 2038–2042. [13] S. Williamson, A. Emadi, M. Shahidehpour, Distributed fuel cell generation in restructured power systems, in: Proceedings of IEEE PES General Meeting, Denver (CO), USA, June 6–9, 2004, pp. 2079–2084. [14] http://www.fuelcellstore.com/. [15] http://www.cordis.lu/joule/home.html. [16] European Commission Directorate-General XII Science, Research and Development, Externalities of Energy ExternE National Implementation Italy Final Report, Contract JOS3-CT95-0010 FEEM, October 1997. [17] Landau Network-Centre Volta School International UNESCO Science for the Peace, Energetic political and Strategies for the future: Scientific, Technological and Economic aspects, vol. 1, Como (Italy), Settembre 1998 (in Italian). [18] European Commission Directorate-General XII Science, Research and Development, Externalities of Energy ExternE National Implementation Germany Final Report, Contract JOS3-CT95-0010 JOU2-CT-0264 IER, November 1997. [19] European Commission, ExternE-Externalities of Energy, vol. 10, National Implementation, EUR 18528, Directorate-General XII, Science, Research and Development, Luxembourg, 1999. [20] T. Markvart, Solar Electricity, John Wiley & Sons, England, 1997. Sonia Leva, received M.S. degree (1997) and the Ph.D. degree (2000) in Electrical Engineering from the “Politecnico di Milano”, Milano, Italy. She is now Asistant Professor in Elettrotecnica in the Department of Energy of the Politecnico di Milano. Her current research interests are concerned with the electromagnetic compatibility, the power quality, Renewable Energy and Sustainable Development. Dr. Leva is a member of IEEE. Dario Zaninelli, received the Ph.D. degree in Electrical Engineering from the Politecnico di Milano, in 1989, and he is now Full Professor in the Department of Energy the Politecnico di Milano. His areas of research include power system harmonics and power system analysis. Dr. Zaninelli is a senior member of IEEE, a member of AEI and a member of the Italian National Research Council (C.N.R.) group of Electrical Power System.
Contents lists available at ScienceDirect
Electric Power Systems Research journal homepage: www.elsevier.com/locate/epsr
Hybrid renewable energy-fuel cell system: Design and performance evaluation Sonia Leva ∗ , Dario Zaninelli 1 Dipartimento di Elettrotecnica, Politecnico di Milano, Piazza Leonardo da Vinci, 32, 20133 Milano, Italy
a r t i c l e
i n f o
Article history: Received 6 December 2006 Received in revised form 21 February 2008 Accepted 2 July 2008 Available online 19 August 2008 Keywords: Fuel cells New/alternative energy resources Photovoltaic power systems System costs
a b s t r a c t The paper introduces hybrid photovoltaic-wind-diesel generation systems supplying a remote power load considering the advantages of sustainable energy from the economic point of view. In particular, a cost investment valuation is performed on a real plant showing the effect of sustainable economical saving. The possibility to introduce a fuel cell generation device in a photovoltaic-wind existing plant for supplying Telecommunication apparatus is also investigated and the results are reported and discussed in the paper. Furthermore, starting from measured data, a control system is realized in order to verify the functionality of the plant. © 2008 Elsevier B.V. All rights reserved.
1. Introduction A high integrated hybrid system (HIHS) can be seen as a way of using renewable sources, to produce electric energy without pollution in stand-alone applications supplying remote load. The photovoltaic (PV) energy production has been mainly used: • in industrialized countries both for grid connected applications and for supplying remote loads with low consumption; • in underdeveloped countries for generating electricity with reduced need of maintenance and with low operation costs. Research and development makes PV technology suitable for industrial applications related to telecommunications, water pumping, water purification, public lighting, etc. In Europe (where the weather conditions are very variable and the forecasts are available for a limited period ahead of time) the PV system needs to be oversized and integrated with batteries and diesel generator set (DGS) [1,2]. Because of these limits the PV has been rarely chosen as energy supply in off grid industrial application which requires more than a few kWh/day. The wind generator (WG) is applied in Europe in case of grid connected systems, but is not so diffused for supplying remote load, although its reduced watt peak cost compared to the PV solution.
∗ Corresponding author. Tel.: +39 02 23993709. E-mail addresses: [email protected] (S. Leva), [email protected] (D. Zaninelli). 1 Tel.: +39 02 23993721. 0378-7796/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.epsr.2008.07.002
The European weather conditions, in fact, do not allow the construction of PV only and/or WG only stand-alone plants, despite the availability of these technologies for energy supply in off grid systems. The HIHS instead allows us to project a power supply system capable of guaranteeing the continuity of the supply, mixing the different renewable energy resources – like PV, WG (even microhydro if possible) – limiting the use of DGS for short periods during winter and/or for backup purposes only. On the design point of view the choice of the generator sizes of the HIHS plant is very important, so that this leads to a good ratio between cost and performance [3]. The HIHS is usually designed to lower the fuel consumption of the diesel machine in the range of 70–90%, compared to only battery–diesel supply architecture overcoming the DGS drawbacks [4]. Therefore, the primary source of electric energy for the plant becomes the renewable one, while the diesel machine constitutes the auxiliary source in case of emergency or for battery charge. Furthermore the hybrid systems can be very effective in terms of sustainable energy development, permitting generation of electric energy with minor environmental interference. In this paper, a technical-economic analysis of a real plant is presented, taking into consideration both plant costs and sustainability costs [5–7], comparing three different possible plant configurations: PV only, PV and DGS, HIHS composed by PV, WG and DGS. The plant under study is used to supply a telecommunication system. This load imposes a fundamental constraint: the use of telecommunication apparatus for the community and emergency purposes makes possible black outs unacceptable. The main characteristic of this system is the need of a continuous power supply, nevertheless on the economic point of view, an oversize system would lead to an excessive increase of the cost of the energy produced.
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
317
Fig. 1. Hybrid system layout.
Furthermore, an improved HIHS composed of PV, WG, hydrogen fuel cell (FC), water electrolyser, batteries and a DGS – as backup device – is presented and analyzed from an economical and technical point of view. In fact, FC devices are a very interesting option to be used with intermittent sources of generation like the PVs because of high efficiency, modularity and fuel flexibility. In the future both DGS and batteries could be removed or reduced according to the experience by FC devices [7,8]. The technical literature in this field presents hybrid systems with FC together with only one renewable source (PV or WG) and economical analysis are seldom dealt with [3,4,9]. The HIHS, including PV generator, WG and DGS, without FC, is already operates. The FC and electrolyser implementation are at the moment in phase of study. An economical and environmental evaluation is necessary (Sections 4 and 5) before the experimentation on field and a preliminary study of the complete system has to highlight technical problems that can occur. For this reason a simple control system – starting from monthly/daily average data measured on the existing plant in 2005 – will be designed in Section 6. 2. Hybrid system
the wind diagram. Based on these diagrams, it is possible to calculate the PV module size and number, and the WG size. The system sizing is quite simple but the difficult part is making the best choice because of the number of possible choices. Focusing on system sizing, it is important to note that, in the considered site, PV and WG gives its peak production in different periods of the year. Typically the PV reaches the best performance during spring and summer while WG has its peak production during the winter. Starting from the load diagram shown in Fig. 2, from the local solar radiation diagram and from the wind map, taking into account the cost of installation and operation, the HIHS is composed by 1. PV modules sized for a 100% load supply from April to September. The result of this sizing brings to a PV array made by 64 modules, 75 Wp and 48 Vdc each. The total power is 4800 Wp. Fig. 3 shows average monthly PV energy production in kWh/ month. 2. WG sized to supply the lack of PV power from October to March. The WG is of 500 Wp and it produces power with wind speed
2.1. System layout The plant layout commonly used, according to Fig. 1, is made up of a set of photovoltaic (PV) modules, a WG, batteries and a DGS. The DGS works for short periods during winter and for backup purposes. Furthermore, its presence increases the overall system reliability. The batteries are used for the storage of energy for a short time and for realizing the decoupling of the energy production from the consumed one. The whole system is managed and controlled by a complete controller, developed for these purposes [10]. 2.2. System sizing The required inputs for the overall system sizing are substantially three: the load diagram, the solar exposition diagram and
Fig. 2. Load diagram.
318
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
Fig. 3. Load diagram and energy production of PV, WG and DGS of the HIHS.
Fig. 4. Energy surplus of the system.
lower than 2 m/s and higher than 150 km/h. Fig. 3 shows the average monthly WG energy production [9].2 3. DGS sized for backup purposes and expected to work for short periods during winter but also to guarantee the necessary reliability of the system. This leads to a DGS with 5 kW and fuel consumption of 3000 gkW/h. In this way, ignoring possible black outs, the DGS normal working time is about 47 h/year, with the need of 104 l/year of diesel, one maintenance operation and two refuelling operations per year. 4. Battery set of 2000 Ah 48 Vdc, C10 type.
be carried out both on site or by a remote station by means of a communication protocol. • The capability of the system to transmit warnings to a remote station, pointing out the anomalous operations of the plant. • The presence of a local memory card for environmental and electrical data collection and processing.
The main economic advantages of this sizing, not considering at the moment the environmental costs, are substantially related to the reduction of PV arrays size in respect to the PV-only solution,3 and the saving on DGS refueling and maintenance operations compared to the PV + DGS without WG solution.4 Fig. 3 shows the average monthly energy balance – based on 2005 data – between the energy production and consumption of the hybrid PV–WG–DGS system mentioned above. It is possible to see that, despite the reduction of the PV arrays, there is still an energy surplus for almost all the year. The use of this surplus and the control of all the operation parameters is done by an electronic control and monitoring system designed for the operation of an HIHS that has been presented in [10]. The main characteristics of the control system are summarized in the following: • The possibility to measure electric quantities (voltages, currents, frequency of the ac supply) on the plant and external quantities (temperature, wind speed) useful for the efficiency optimization. • The regulation of the battery power flow to avoid lifetime reduction or operating conditions that can decrease the battery reliability. • The definition of a protection and warning system able to prevent fault conditions. • The possibility to check the present operation of the plant and the previous data recorded in the memory, together with the possibility to set the warning thresholds. These operations can
2 Comparing the PV and WG energy production it is possible to see that even if the WG size is about 1/10 of the PV size, its contribution to the energy production is about 1/5 of the PV production. 3 For “PV-only” system the array is made by 116 modules, 75 Wp and 48 Vdc each, with a covered area of about 100 m2 and with a battery set of 2000 Ah 48 Vdc, C10. This system is over-sized but does not protect from possible black outs due to unexpected weather conditions then a DGS for buck up purpose is present. 4 For PV + DGS system the array is made as for HIHS, the DGS normal working time is about 465 h/year, with a fuel consumption of 1025 l/year. This implies 11 maintenance operations and 20 refuelling of the tank per year.
Thanks to these properties on the collected data, it is possible to manage the HIHS and to modify, if necessary, the control algorithm in order to optimize the system and work with respect to the local conditions. 3. Fuel cell system The chosen layout leads to an energy production surplus (Fig. 4). This surplus – calculated neglecting the possible DGS production – is due to renewable production. Several energy storage techniques, like batteries, fly-wheels and hydrogen production are available for system integrators when energy storage is needed [2]. The main considered parameters during sizing are energy storage capacity, source availability, and response time of the energy source. Normally batteries present good modularity, fast time response and good energy–mass ratio storage, but they need maintenance and are subjected to fault. In addition the batteries performance depends on working temperature that has to be controlled during operation time. These characteristics make batteries suitable for hybrid systems. Fly-wheel presents little modularity, very fast time response, but the stored energy is available for the system for very a short time with respect to batteries; on the other hand a plant with a fly-wheel presents less system complexity and practically minimal or no maintenance [11,12]. FC, compared with the two previous techniques, presents modularity and the possibility of partial load working; good energy storage capability that can be simply obtained with correct hydrogen tank sizing. One main weak point of the FC is its very slow dynamics. Maintenance problems of fuel cell are still an open problem because no registered data of the working lifetime of the catalyst is available [13,14]. A way to increase the efficiency of the HIHS system consists in substituting the DGS backup set with FC. In this way, it is possible to use the surplus energy to produce, with a hydrolysis process, the hydrogen needed for the FC. 3.1. System layout The system layout remains the same as the previous one except for the presence of the fuel cell, hydrogen and water tank (see Fig. 5).
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
319
Fig. 5. Fuel cell system layout.
We assume that FC will substitute the DGS. The DGS presence is related to the lack of practical experience in systems with FC and their reliability. Using this plant as a test prototype it is possible to verify, from a practical point of view, the energy availability all over the year. A further step, after a data collection phase on the plant during a few years, will be a plant design without the DGS backup set. 3.2. System sizing The hydrogen can be produced, during the surplus of energy production, from water by an electrolyser and stored in a container for further use. Assuming a hydrogen production rate of about 0.018 m3 /kWh, the surplus energy can be used to produce about 18.6 m3 /year. With low temperature fuel cell, neglecting the use of thermal energy, its efficiency can be considered to be about 30%. The lower heating value of hydrogen is 10.72 MJ/m3 (38.6 kWh/m3 ) [14]. Therefore the total electric energy that can be produced by the FC is EH2 = VH2 LHVH2 = 215 kWh/year
(1)
where is the efficiency of the fuel cell, VH2 is the total net volume of hydrogen produced per year and LHVH2 is the lower heating value of hydrogen. In this way, although the hydrogen can be produced only during spring and summer and needs storage, the use of an FC can lead to a complete autonomy of the plant. The estimated efficiency of the complete process can be calculated as follows: ε = 0.018 × 0.3 × 38.6 = 0.208
(2)
corresponding to about 21%. In this paper, the whole system is analyzed only from the electrical point of view, neglecting the thermal energy produced by the FC. This energy can be used, during winter, either for avoiding ice formation in the water tank or for local warming of the apparatus.
The plant, including PV generator, WG and DGS with the mentioned characteristics, is already in function. The FC and electrolyser implementation are at the moment in phase of study. Before the experimentation on field an economical and environmental evaluation is necessary (Sections 4 and 5) and a preliminary study of the complete system that must clarify the technical problems that can occur. For this reason a simple control system – starting from monthly/daily average data measured on the existing plant in 2005 – will be realized (Section 6). The model allows studying the systems with FC and electrolyser shown in Fig. 5, even if with some approximation. 4. Environmental costs Environmental costs of energy production are normally treated as an externality. This means that the influence of the energy production on the environment does not find a corresponding cost on the bill. Until some years ago all the climate change warnings, mainly related to the greenhouse gases emission, involved studies made in order to reduce the gas production in energy cycles and calculate the economic value of the externalities. The aim of this evaluation is the quantification of an environmental fee related to the quantity of pollution emission, normally carbon-dioxide. One of the first studies made by the European Community on the externalities evaluation of energy production was related to the Non-Nuclear Energy (NNE) program, a research and development specific program within the European Union’s Fourth Framework Program for Research and Technological Development. The main issues of this program were • improving security of energy supply; • protecting the environment; • encouraging the rational use of energy. The NNE was structured in two components NNE-JOULE dealt with research and development in the field of innovative energy technologies, and THERMIE which focused on the demonstration
320
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
Table 1 Externalities values of energy production
HIHS and HIHS + FC type. The following data refer to real application on an existing plant.
Description
Externalities (centD /kWh)
PV WG DGS
0.112 0.050 3.376
of these energy technologies. The European project THERMIE [5] in 2002 lead to a cost for producing energy with fossil sources that can be estimated in 0.20 D /kWh. This esteem of the externalities value had a constant value in time [15]. Another externality evaluation method, based on the economic value proposed by THERMIE can be derived observing that the growth of the carbon dioxide emission over the last 50 years is quasi-linear [5,17]. It is possible to obtain a formula that evaluates, linearly, the externalities related to the emissions of the plant for the future years. Furthermore, evaluation of the externalities was developed with the ExternE Project, developed within the European Commission R&D program Joule II, which proposed a unified methodology for the quantification of the externalities of different power generation technologies. Under successive research programs this methodology has been developed along three different lines: the first, involved with the further development and updating of the methodology performance called ExternE Core; the second, mainly a data collection phase called ExternE National Implementation (in order to create an European data set); and the third, the applicative one, called ExternE Transport, in which the proposed methodology is applied to the energy related impacts from transportation. In more detail the ExternE project evaluated the environmental costs of several traditional fuel cycles and of several renewable energy cycles5 [16,18]. The recent evaluation technique, developed under the ExternE research program, is based on the impact pathway method [19]. Basing on the project results it is possible to evaluate these costs and associate them with the cost of energy production. Then, it is possible to compare the proposed plant layouts and calculate an environmental net present value (NPV), based on both the plant cost and externalities, determining the correspondent energy cost. The application of the externalities is done evaluating the ratio of the single energy source production with respect of the total energy production and then applying the economic values reported in Table 1. The values of Table 1 are obtained from ExternE summary for PV, WG and DGS [16,18]. These values and the one of European project THERMIE will be used in the economic analysis in order to derive a cost for the produced kWh. 5. Economic analysis Scope of this section is to estimate the kWh cost and therefore the economic convenience by using the Unit Electricity Cost (UEC) method referring to the analyzed different plants solution [6,20].
5.1.1. Diesel generator, fuel and maintenance costs The DGS has and indicative cost of about 5100 D . Focusing on the fuel costs, in the PV + DGS plant the generator works 465 h per year. The fuel consumption is about 1025 l and its cost is about 1.43 D /l. The yearly fuel cost amounts to 1463 D . In the HIHS type plant the diesel generator works only 50 h/year with a yearly fuel cost of 150 D . The diesel generator maintenance costs are calculated on the number of maintenance events per year. The PV + DGS type plant needs 11 maintenance operations per year with an overall cost of 5500 D , while HIHS type plant needs only two maintenances per year with an overall cost of 1000 D . 5.1.2. Photovoltaic modules The system under analysis is built of 64 modules of 75 Wp each. The average price of each module is 266 D . So the PV modules cost is about 17,000 D . The PV modules have to be installed on support structures in order to obtain the design inclination, for a total price of 878 D . 5.1.3. Batteries The batteries have a capacity of 2000 Ah and the cost is about 4080 D . The chemical protection system price is assumed equal to 1275 D . The battery charger on the basis of the market analysis is evaluated in about 850 D . 5.1.4. Wind generator The system is provided with a small sized WG with 500 W power. All the accessories needed for its installation and normal operations, like supports, are available on the market. The overall cost of the generator and accessories is valuated at about 4200 D [10]. 5.1.5. Fuel cell The fuel cell, which has 1 kW of rated power at 48 Vdc, is a PEM type supplied directly with hydrogen gas. Its cost is about 3000 D . The fuel consumption is 15 l/min of industrial grade hydrogen. The hydrogen is provided by an electrolyser fed by water which has been treated by a deionizer. The cost of the hydrogen generator and water deionizer is about 7000 D . The hydrogen is produced, directly at high pressure, by the electrolyser and stored in a storage tank, made of stainless steel, for further use. The hydrogen tank capacity is 120 l at 1 MPa at room temperature, its cost is about 1000 D . The hydrogen transmission pipes are completed with pressure regulators and safety valves [14]. 5.1.6. Installation The evaluation of the installation costs is done with the labor price taken from Milan’s Chamber of Commerce in 2005. It is assumed that three people form the working team involved in the installation of the plants. The installation of all the considered plants can be done in about 15 working days, ten of them in which the whole team is present and five of them in which the most qualified worker is not present.
5.1. Cost evaluation of the plant hardware The first step for the economic analysis consists in the evaluation of the overall costs of the three plants under examination: PV + DGS,
5
The considered fuel cycle are both conventional sources like coal, lignite, peat, gas, oil, orimulsion, nuclear and renewable energy like biomass, hydro, wind, photovoltaic, and waste incineration.
Table 2 Plants costs Plant
Installation (D )
PV only PV + DGS HIHS HIHS + FC
5770 5770 5770 5770
Hardware (D ) 43,335 32,625 36,840 48,250
O&M (D ) 1040 6925 1150 4040
Total (D ) 50,145 45,360 43,760 58,060
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
Table 2 summarizes the costs of the considered plants. A comparison of Table 2 data shows that the Operation and Maintenance (O&M) cost of a HIHS is cheaper than the other plants. This is substantially due to the reduction of periodic maintenance and refuelling operations. In the case of FC, the costs are higher because of the absence of a real market for FC maintenance. FC costs should be reduced in the next years because of the increase of the FC production.
321
Table 3 UEC For different type of plants Plant
UEC (D /kWh)
UEC with environmental costs (D /kWh)
UEC with externalities (D /kWh)
PV only PV + DGS HIHS HIHS + FC HIHS + FC (no DGS)
0.507 1.562 0.440 1.047 0.849
0.514 1.580 0.443 1.053 0.849
0.509 1.566 0.442 1.049 0.850
5.2. Produced energy cost The evaluation of the cost of the produced energy by the real plant is made by using the life-cycle costing (LCC) [6,20]. The parameters for LCC evaluation are period n: the period of time during which the system is under analysis, that is, the greatest expected lifetime of the component; inflation i; discount rate d (the applied active rate d = 2–5%); cost of money and the overall hardware and installation costs; operation and maintenance costs; fuel costs; environmental costs: economic value due to pollution emitted in the atmosphere; substitution costs of every part at the end of its operative life. The LCC evaluation needs that all costs and benefits, present and future, are brought back to present value or present worth (PW). The present value of the amounts is calculated, as function of the period of the payment, using two different updating factors: single payment PWr = Cr
1 + i N 1+d
= Cr Pr
(3)
annual payment PWa = Ca
(1 + i/1 + d)(|(1 + i/1 + d)|N ) − 1 = Ca Pa ((1 + i)/(1 + d) − 1)
(4)
where Cr is the cost of year r and N is the expected lifetime of the part. The Ca is the recurrent annual cost. Pr and Pa are respectively the payment evaluated in the cash flow. The sum of all PW gives the LCC. The produced energy cost is calculated by means of the annualized life-cycle costing (ALCC): ALCC =
LCC = Pa
D
(5)
year
which leads to the UEC by means of the annual produced energy given by the ratio between the ALCC and the annual energy E produced by the plant: UEC =
ALCC = E
D
kWh
(6)
The UEC for the three considered plants are reported in Table 3. These values have been calculated only accounting the costs reported in Table 2, and adding the environmental costs indicated by European project THERMIE and obtained from ExternE summary (see Table 1). The UEC is a very significant indicator for performance evaluation of energy generation systems and gives, practically, the kWh generated cost. The UEC for different plant types, with a range of −20% around the normal produced energy, are evaluated and reported in Fig. 6. This allows to calculate possible UEC variations due to black outs.
Fig. 6. UEC of produced kWh for the analyzed plants in a range of −20% around the normal produced energy without sustainability costs.
Data reported in Table 3 shows that HIHS lead to lowest UEC. Anyway, the HIHS + FC plant, even if it is a prototype, is cheaper than the PV + DGS plant. Furthermore, comparing the data reported in Table 3 with the ones reported in [5], it is possible to assert that the environmental cost is strictly related to the DGS emissions. Finally, the UEC in the case of HIHS + FC without DGS, are evaluated. These costs are today only an attempt, but in the future, when FC will have a market spread with reliable operations, the economic impact will be more precise. 6. Control system and simulation results In this section, we briefly discuss the simulation model used in this paper. The simulation will be extended for a temporal window of 24 h in two different conditions: one regarding a typical day of May when the energy production is high, the other one corresponding to a typical day of November when the energy production is not enough to power the load. In the first case the energy surplus is deviated towards the aqua electrolyser, storing hydrogen in the tank. In the second case the FC will provide the missing energy taking the hydrogen from the tank. The system analysis will be done considering initially the plant including the battery, but without the DGS. Subsequently the possibility of eliminating the battery will be discussed. 6.1. Control and simulation system [8] The models of the different parts of the system are shown in Fig. 7 and it is composed of the following subsystems: load regulator, wind and FV generators, aqua electrolyser, FC and battery. The load regulator can be seen as a simple transfer function with a rapid response: Freg =
1 s + 0.1
322
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
Fig. 7. Control and simulation system.
The process, modelled as above, is controlled in retroaction by a PI regulator with parameter determined by the trial and error method. The regulator receives as input the error between the required power and the regulated one. The WG is modelled as a couple of random signals overlapped. The first one with a long period (1 h) and average equal to the estimated average power produced during the day. The second one, uniform type, with a short sample time (3 min) necessary to simulate sudden variation in wind direction and power. Properly setting the variation between the mean values, a good approximation with respect to typical wind generation performances is achieved. The photovoltaic source is modelled starting from measured values included in a. mat file that gives the value of the produced instant power for each simulation time. As expected, a sort of Gaussian curve is observed. For the model of the electrolyser it is considered a transfer function as follow: Freg =
1 s + 0.2
It takes into account, as a simply way, the slight delay introduced by the device. The electrolytic process is controlled by a PI regulator another time with characteristic set by the trial and error method. The reference for the regulator is null when the battery is charged within a certain value and equal to the energy in excess otherwise. The limit for the battery charge is determined by empiric method in such a way to obtain a decent value of charge at the end of the day. Finally the hydrogen produced is calculated as integral of the power transformed and multiplied by the following constant: VH2 ,elect = 0.018 (m3 /kWh) = 5 × 10−9 (m3 /J) Similarly the fuel cell is modelled by a single pole transfer function, but with a long period time constant that correspond to a
slower kinetic of the cell: Freg =
1 s+4
Also in this case the reference for the regulator is null when the battery is charged within a certain value and equal to the energy in deficit otherwise. The limit for the battery charge is determined by empiric method in such a way to obtain a decent value of charge at the end of the day. The consumed hydrogen is calculated as integral of the power transformed and multiplied by the following constant: VH2 ,cell = 0.3 × 38.6 (kWh/m3 ) = 41.7 × 106 (J/m3 ) The model of the battery is obtained integrating the current absorbed by the charger and limiting the total charge to a suitable value determined in such a way to obtain an acceptable steady state condition. This model represents a really simplified approach that introduces an approximation not negligible. In fact the charge of a battery depends also on the velocity on which the battery is charged, but taking into account all this variables leads to a really complex model not needed for the scope of the study. The power absorbed by (or fed from) the battery is equal to the sum of the produced one minus the one absorbed by the load: Pbatt = Pw + Psv − Pload From which it is possible to calculate the current simply dividing per the voltage (in this case 48 V): Ibatt =
Pbatt P = batt Vbatt 48
The charge accumulated is calculated integrating this value. It shall be noted the it is necessary to properly set a integration constant and limiting the value of the output in order to avoid the windup phenomena.
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
323
Fig. 8. Most significant data obtained in May.
Fig. 9. Most significant data obtained in November.
6.2. Simulation results Considering a typical day in May, when the energy production is high because of the PV production, the controller is programmed to use the energy surplus towards the aqua electrolyser in order to store hydrogen in the tank. The simulation model used in this condition includes the load power regulator, the generators, the model of the battery charge and the aqua electrolyser. The energy supplied in the first hour of the day is not enough to cover the entire requirements of the load, therefore this will be supplied from the battery that in the meantime endures a partial discharge. When the PV production rises, the power generated becomes greater than the energy required so the battery can be recharged. When this process is finished the energy in excess (Fig. 8) is deviated towards the aqua electrolyser which produces the hydrogen for the FC: for the considered day the system can produce and store of almost 0.112 m3 of hydrogen. This means a corresponding energy per year estimated in 1148.74 kWh. It must be noted that the control system is studied to guarantee a high battery charge level in order to always supply the load for 4 or 5 days in case of breakdown or particularly unfavorable environmental conditions. In a typical November day, when the energy production is low, the controller will provide the energy deficit taking hydrogen from the tank and using it in the fuel cell. The simulation model used in the second condition is made up of the load power regulator, the generator, the model of the battery charger and the FC. The results obtained from the simulation are synthesized in Fig. 9. In the early morning the discharge of the battery is limited by the FC, in order to avoid the slow but progressive discharge of the battery, that in the arch of some days – above all if weather conditions are unfavorable to WG or FV – would lead to the impossibility of powering the load correctly. However, in a few hours the power production rises because of the PV energy so the FC can be turned off and the battery partially recharged. The hydrogen consumed during the entire day is approximately 0.07 m3 . The total annual energy consumed by the FC is approximately 54.24 kWh.
Therefore, we can see that the amount of hydrogen produced in the summer months exceeds the load needed, even if we must consider conditions of breakdown that could lead to different energetic balances. Moreover, it is useful to emphasize that using the FC leads to the same result as that of using DGS. In fact, DGS would work in the winter months in almost the same way as the FC, while in the summer months the energy surplus would be wasted. Finally, we analyzed the possibility of eliminating the battery group and using the FC also for the management of the energy on the short period. In this case the FC must be able to supply the load all by itself. This would mean an increase in the use of the more technologically advanced and expensive devices reducing the long-term costs. In Fig. 10, we can be seen, in comparative way, the necessary and produced hydrogen. Considering a typical day of May, the system needs almost 0.13 m3 more then the 0.112 m3 hydrogen produced, corresponding to energy excess from PV of 7.2 kWh/day. During a typical day of November, this amount increases to 2.2 m3 corresponding to an energy excess of 122 kWh/day. As we can observe the system demand is not satisfied. In order to cover the power demand, an excessive over-sizing of the PV plant would be necessary, with a consequent and inconvenient increase of the costs. It is important to underline that these results, from an energy balance point of view, are in agreement with that obtained by other
Fig. 10. Hydrogen produced and necessary in case of battery absence (m3 /month).
324
S. Leva, D. Zaninelli / Electric Power Systems Research 79 (2009) 316–324
authors [3,4] that have analyzed a simpler system from the dynamic point of view. In fact a typical self-sufficient HIHS supplying remote plant must include short-term and long-term energy storage. The combination of a battery bank – commonly used for short-term energy storage – with long-term energy storage in the form of hydrogen can significantly improve the performance of stand-alone renewable energy systems. 7. Conclusions The paper deals with a technical and economical analysis of possible implementation of renewable energy (photovoltaic and wind energy) sources on an existing plant for supplying isolated telecommunication devices. In particular the possibility of substituting the diesel generator backup group with a fuel cell is investigated both through the energy balance and as an economic investment. The results show that the energy balance is satisfied and the plant can work properly with fuel-cell devices, too. The energy storage is done with a diesel oil tank for the first layout, and with hydrogen, and water tank, for the second one. The water tank is necessary only in order to improve the autonomy of the plant. Maintenance and refueling with the diesel generator are the major problems because of the periodic maintenance required by the diesel machine, which works only during winter time. Using the fuel cell technology, the need of periodic maintenance and refueling operation is strongly limited or removed. Furthermore the real case application, based on an existing remote telecommunication installation, allows us to compare different plants (PV only, PV + DGS, HIHS, HIHS + FC) economically and technically and show the advantages of the HIHSs. Considering the environmental cost, the photovoltaic-wind-fuel cell configuration, having near-zero emission, also permits an increase of the sustainability and so a decrease of the external costs in the electric power generation. Finally, the possibility of substituting the storage battery with fuel cell is analyzed. Unfortunately with respect to the present technologies and the efficiency that corresponds, this solution is not convenient both from the technological point of view and from the economic one. In spite of this, future developments appear feasible: it cannot be excluded that substantial progresses will be made in the next few years. References [1] T.M. Calloway, Autonomous photovoltaic-diesel power system design, in: Proceedings of IEEE Photovoltaic Specialists Conference, Las Vegas, Nevada (USA), October 1985, pp. 280–284. [2] A. Chaurey, S. Demi, Batter storage for PV power systems: an overview, Renewable Energy 2 (3) (1992) 227–235.
[3] K. Agbossou, M. Kolhe, J. Hamelin, T.K. Bose, Performance of a stand-alone renewable energy system based on energy storage as hydrogen, IEEE Transaction on Energy Conversion 19 (September (3)) (2004) 633–640. [4] Z. Jiang, Power management of hybrid photovoltaic—fuel cell power systems, in: Proceedings of IEEE Power Engineering Society General Meeting, Montreal, Canada, June 18–22, 2006. [5] F. Iannone, S. Leva, D. Zaninelli, Hybrid photovoltaic system and sustainability: economic aspects, in: Proceedings of IEEE PES General Meeting, Denver (CO), USA, June 6–9, 2004, pp. 1933–1938. [6] S. Sampattagul, S. Kato, N. Maruyama, A. Nishimuram, Comparison of coal-fired natural gas-fired power plants as economically viable and ecologically sustainable power generation systems, International Journal of Emerging Electric Power Systems 3 (2) (2006) (Article 1116). [7] F. Iannone, S. Leva, D. Zaninelli, Hybrid photovoltaic and hybrid photovoltaicfuel cell system: economic and environmental analysis, in: Proceedings of IEEE PES General Meeting, San Francisco (California), USA, June, 2005, pp. 1503–1507. [8] R. Contino, F. Iannone, S. Leva, D. Zaninelli, Hybrid photovoltaic-fuel cell system controller sizing and dynamic performance evaluation, in: Proceedings of IEEE PES General Meeting, Monreal (Quebec), CANADA, June, 2006, 6 p. [9] T. Seniyu, D. Hayashi, N. Urasaki, T. Funabashi, Oscillation frequency control on H∞ controller for a small power system, using renewable energy facilities in isolated island, in: Proceedings of IEEE Power Engineering Society General Meeting, Montreal, Canada, June 18–22, 2006. [10] W. Dalbon, S. Leva, M. Roscia, D. Zaninelli, Hybrid photovoltaic system control for enhancing sustainable energy, in: Proceedings IEEE PES Summer Meeting, Chicago (IL), USA, July 21–25, 2002, pp. 134–139. [11] http://www.igetcol.it/. [12] M.L. Lazarewicz, A. Rojas, Grid frequency regulation by recycling electrical energy in flywheels, in: Proceedings of IEEE PES General Meeting, Denver (CO), USA, June 6–9, 2004, pp. 2038–2042. [13] S. Williamson, A. Emadi, M. Shahidehpour, Distributed fuel cell generation in restructured power systems, in: Proceedings of IEEE PES General Meeting, Denver (CO), USA, June 6–9, 2004, pp. 2079–2084. [14] http://www.fuelcellstore.com/. [15] http://www.cordis.lu/joule/home.html. [16] European Commission Directorate-General XII Science, Research and Development, Externalities of Energy ExternE National Implementation Italy Final Report, Contract JOS3-CT95-0010 FEEM, October 1997. [17] Landau Network-Centre Volta School International UNESCO Science for the Peace, Energetic political and Strategies for the future: Scientific, Technological and Economic aspects, vol. 1, Como (Italy), Settembre 1998 (in Italian). [18] European Commission Directorate-General XII Science, Research and Development, Externalities of Energy ExternE National Implementation Germany Final Report, Contract JOS3-CT95-0010 JOU2-CT-0264 IER, November 1997. [19] European Commission, ExternE-Externalities of Energy, vol. 10, National Implementation, EUR 18528, Directorate-General XII, Science, Research and Development, Luxembourg, 1999. [20] T. Markvart, Solar Electricity, John Wiley & Sons, England, 1997. Sonia Leva, received M.S. degree (1997) and the Ph.D. degree (2000) in Electrical Engineering from the “Politecnico di Milano”, Milano, Italy. She is now Asistant Professor in Elettrotecnica in the Department of Energy of the Politecnico di Milano. Her current research interests are concerned with the electromagnetic compatibility, the power quality, Renewable Energy and Sustainable Development. Dr. Leva is a member of IEEE. Dario Zaninelli, received the Ph.D. degree in Electrical Engineering from the Politecnico di Milano, in 1989, and he is now Full Professor in the Department of Energy the Politecnico di Milano. His areas of research include power system harmonics and power system analysis. Dr. Zaninelli is a senior member of IEEE, a member of AEI and a member of the Italian National Research Council (C.N.R.) group of Electrical Power System.