Energy balance analysis of combined photovoltaic–diesel powered telecommunication stations

Electrical Power and Energy Systems 33 (2011) 1739–1749

Contents lists available at SciVerse ScienceDirect

Electrical Power and Energy Systems journal homepage: www.elsevier.com/locate/ijepes

Energy balance analysis of combined photovoltaic–diesel powered telecommunication stations J.K. Kaldellis ⇑, Ioanna Ninou Lab of Soft Energy Applications & Environmental Protection, TEI of Piraeus, P.O. Box 41046, Athens 12201, Greece

a r t i c l e

i n f o

Article history: Received 20 August 2010 Received in revised form 23 December 2010 Accepted 12 August 2011 Available online 8 October 2011 Keywords: Hybrid stand-alone systems Remote areas Lead acid batteries Optimum sizing Fuel consumption Panel tilt angle

a b s t r a c t During the last years photovoltaic (PV) generators comprise a promising option for satisfying the electrification needs of both grid connected and isolated systems worldwide. At the same time, the mobile telecommunication (T/C) sector presents a vast growth that leads to the expansion of the respective networks even at the most remote of areas. In this context, there are numerous T/C stations that cannot appreciate connection to an electricity grid and thus cover their energy needs usually on the basis of oil-consuming diesel generators. On the other hand, replacement of the oil-based generation with PV-battery standalone configurations implies system over-sizing so as to obtain zero load rejections throughout the year. In an attempt to both reduce the amounts of oil consumed and downsize the PV-battery system, a hybrid PV-based stand-alone system, employing also a diesel engine and a battery bank, is currently proposed. System sizing is first undertaken for various scenarios of annual fuel consumption and panel tilt angle under the restriction of zero load rejection, while following, the main directions for obtaining optimum solutions are given. Finally, an extensive energy balance analysis of typical configurations also provided reflects the fact that such hybrid systems may support better utilization of the PV energy production and also reduce considerably the annual amounts of oil required for the diesel-only solution. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction During the last years, development of solar energy technologies, among other achievements [1,2], is much determined by the increase of the photovoltaic (PV) technology efficiency and the respective remarkable production cost reduction, making PV generators a promising option for handling the electrification requirements of both grid connected and isolated consumers worldwide [3–7]. Actually, in cases of remote installations, absence of moving parts and low maintenance needs make stand-alone PV stations a really competitive long-term energy solution in comparison with the diesel-electric generators adopted up to now. This is much more interesting in areas with high solar potential, like in most locations of the Greek territory, where the annual available solar energy at horizontal plane varies between 1300 and 1800 kW h/m2. In this context, the present study investigates the possibility to cover the energy demand of numerous remote telecommunication (T/C) stations (repeaters) on the basis of an appropriate PV generator. At this point it is important to note that recently the mobile T/C sector has experienced an exponential growth worldwide, contributing in the rapid data and information distribution [8]. This general trend has also taken place in Greece, where nowadays more than ⇑ Corresponding author. Tel.: +30 210 5381237; fax: +30 210 5381467. E-mail address: [email protected] (J.K. Kaldellis). URL: http://www.sealab.gr (J.K. Kaldellis). 0142-0615/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijepes.2011.08.016

5000 T/C stations existing cover almost 99% of the Greek territory [9]. Taking into consideration that the corresponding antennas are usually placed at high elevations, existing in rural areas and mountainous regions, it is quite rational that the respective installations should be in several cases found far away from the electrical grid. More precisely, more than 10% of the existing T/C stations mentioned above are not grid connected, thus they are covering their electricity needs on the basis of small diesel-electric generators. Unfortunately, this solution, adopted up to now, is responsible for high operational costs due to both the necessary fuel consumed (more than 12 tn of diesel-oil annually) by the existing internal combustion engines and the increased maintenance needs, with the diesel engines employed requiring continuous technical support and constant fuel transports. In order to face all these problems, several authors propose the installation of a PV-based stand-alone system, which may cover the electricity needs of a remote T/C station with the utilization of an appropriate energy storage device [10–13]. However, according to the results available, a similar installation normally requires a remarkable number of PV panels (over-sizing) along with a rather huge battery bank, due to the unavailability of solar energy during nights and cloudy days [14]. To overcome these problems, the authors propose the utilization of a hybrid PV-based installation, which guarantees the energy autonomy of a typical T/C station with the assistance of a relatively limited lead-acid battery array and a small dieselelectric generator.

1740

J.K. Kaldellis, I. Ninou / Electrical Power and Energy Systems 33 (2011) 1739–1749

The general idea of the proposed solution is to combine the advantages of all the system components, minimizing the dimensions and improving the reliability of the installation, reducing the operational and maintenance effort/cost and providing an electricity generation option with long-term generation cost that is definitely inferior to that of the PV-only and the diesel-only solutions [15]. More specifically, in the present study one investigates the appropriateness of a PV based hybrid power station for representative T/C stations on the basis of a detailed energy balance analysis. In this context and considering the high solar potential of most Greek territories, the proposed work is focused on analyzing the long-term operation of a typical PV-based T/C station located in South Greece on an energy balance basis, with similar installations being already in pilot operation at certain Greek locations, Fig. 1a and b. More specifically, the present analysis is configured so as to initially estimate the energy autonomous PV-based configurations of the typical T/C station under investigation for several diesel oil quantities consumed on an annual basis, including also the zero oil (or the PV-only) solution. Accordingly, the main directions for obtaining the optimum hybrid station dimensions are provided, taking also into consideration the energy performance of the installations under investigation. Finally, the proposed analysis is integrated through the investigation of the energy balance of several representative hybrid configurations, presenting the contribution of the solar energy in meeting the T/C energy demand and analyzing the disposal of the PV-generator energy production, i.e. load demand, system losses and energy surplus. As a result, contribution of the specific work lies – among other reasons – on:

Fig. 1. Pilot PV-based T/C stations in Greece.

 The examination of PV-diesel hybrid energy systems for the satisfaction of remote T/C stations as well, on top of the numerous studies investigating the use of similar systems at the residential and community level, or the use of PV-only systems (using battery storage only) for serving T/C stations.  The determination of a detailed electricity demand profile for a typical T/C station operating in the Greek territory, considering that energy consumption of such stations may be quite important – even at a national level [16] – due to the remarkable expansion of T/C networks, and the fact that the respective demand pattern varies considerably in comparison with the respective of residential-type applications.  The use of an analytical calculation sizing algorithm that incorporates extra features (such as the variable tilt angle input) and takes into account hourly input data, thus resulting to the production of several energy autonomous system configurations and the detailed determination of their energy performance. 2. Proposed solution for the electrification of remote T/C stations To configure the size of the energy autonomous PV-based configurations that may satisfy the annual electricity needs of a representative medium-sized remote T/C station (e.g. 30 MW h annual electricity consumption) under the condition of zero load rejections, certain problem inputs need to be determined. In this context, after an extensive local market survey a representative daily electricity consumption profile [13,17] is adopted, being also dependent on the specific month analyzed (i.e. winter, summer, other), see also Fig. 2. More precisely, the electrical load varies between 2.4 kW and 4.2 kW, while the corresponding annual electricity consumption approaches 27.2 MW h. Note that up to now one needs approximately 12.2 tn of diesel-oil (specific heat content, Hu = 40 MJ/kg) in order to cover the above mentioned electrical load demand using a 7.5 kW (10PS) diesel-electric generator (round trip electricity generation efficiency gd  20%). Thus, the annual electricity consumption – on an hourly basis – is the first input of the present analysis (Fig. 2), while one should keep in mind that a typical T/C station (repeater) requires longterm zero load rejection operation [18]. Additionally, the corresponding detailed solar radiation and ambient temperature are also necessary to integrate the energy balance calculations [19], see for example Fig. 3. Note that most Greek areas possess quite high solar potential, especially during the summer period, where the solar energy available at the horizontal plane is higher than 220 kW h/ mo m2. On the contrary, during winter the values observed are less than 70 kW h/mo m2. Finally, the operational characteristics of all the components (e.g. power curve of the PV panels at standard day conditions, inverter efficiency, battery bank characteristic, diesel-electric generator specific fuel consumption, etc.) composing the hybrid system under investigation, are also required. Recapitulating, the problem to be solved in the current paper is first to determine the appropriate size and accordingly carry out an extensive energy balance analysis of a hybrid PV-based power station that is able to cover the energy requirements of a remote T/C station, without load rejection during an entire year period. The installation also includes an appropriate lead-acid battery bank in order to improve the system reliability [20–22] and a small dieselelectric generator, used as a back up solution [23,24] for long periods of low solar irradiance. As a result, the two extreme cases of the proposed solution are either a diesel only system (with no PV modules) or a stand-alone PV generator with zero diesel-oil contribution. In this context, emphasis is given in the proposed analysis so as to present the detailed energy balance of several representative hybrid PV-based power configurations. More specifically, according to the analysis undertaken, the solar energy disposal for covering the load

1741

J.K. Kaldellis, I. Ninou / Electrical Power and Energy Systems 33 (2011) 1739–1749

LOAD DEMAND OF THE REMOTE T/C STATION 4,5

January March May July September November

Power Demand (KW)

4,0

February April June August October December

3,5

3,0

2,5

2,0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hours of Day Fig. 2. Load demand curves of a T/C station.

Winter and Summer Solar Potential for the T/C Station Area December

1000

July 900

Solar Radiation (W/m2)

800 700 600 500 400 300 200 100 0 0

100

200

300

400

500

600

700

Hours of Month Fig. 3. Solar irradiance time-evolution for typical winter and summer months.

demand, the system losses as well as any energy surplus appearing during system operation, are all determined and analyzed. For this purpose, the main parameters to be estimated are the PV generator peak power, the lead-acid batteries storage capacity and the annual diesel-oil consumption. Finally, as already mentioned, the main directions for obtaining the optimum hybrid station dimensions are provided, taking also into consideration the energy performance of the installations under investigation. As already mentioned the proposed (Fig. 4) hybrid power station is based on: i. A PV system of ‘‘z’’ panels (‘‘N+’’ maximum power of every panel) properly connected (z1 in parallel and z2 in series) to feed the charge controller to the voltage required. ii. A lead-acid battery storage system for ‘‘ho’’ hours of autonomy, or equivalently with total capacity of ‘‘Qmax’’, operation voltage ‘‘Ub’’ and maximum discharge capacity ‘‘Qmin’’ (or equivalently maximum depth of discharge ‘‘DODL’’). iii. A DC/DC charge controller of ‘‘Nc’’ rated power, charge rate ‘‘Rch’’ and charging voltage ‘‘UCC’’.

Fig. 4. Proposed hybrid PV-based solution.

iv. A DC/AC inverter of maximum power ‘‘Np’’ able to meet the consumption peak load demand, taking also into account an appropriate security margin.

1742

J.K. Kaldellis, I. Ninou / Electrical Power and Energy Systems 33 (2011) 1739–1749

v. A small internal combustion engine of ‘‘Nd’’ (kW), able to meet the consumption peak load demand ‘‘Npeak’’ (i.e. Nd P Npeak), where ‘‘Npeak’’ is the maximum load demand of the consumption, including a safety margin (e.g. 30%). For estimating the appropriate size of the proposed PV-diesel hybrid system, three governing parameters should be defined: the peak power ‘‘No’’ of the PV generator used (or equivalently the number ‘‘z’’ of the panels required, No = z  N+), the battery maximum necessary capacity ‘‘Qmax’’ and the annual diesel-oil consumption ‘‘Mf’’. During the long-term system operation, the following energy production scenarios exist: – Energy (DC current) is produced by the PV generator and is sent directly to the consumption via the inverter of the system. In the case that there is energy surplus, the energy output of the PV panels not absorbed by the consumption is stored at the batteries via the charge controller. – The output of the PV generator cannot meet the corresponding load demand. The battery is used to cover the energy deficit via the charge controller and the DC/AC inverter. Note that in that case the battery depth of discharge should not be higher than a predefined value, in order to prolong the operational life period of the batteries. – Energy is produced (AC current) by the small diesel-electric generator and is forwarded to the consumption, in cases that the PV generator cannot meet the load demand and the system batteries are near their maximum depth of discharge value. 3. Energy autonomous configurations – proposed numerical algorithm As already mentioned, the present study is concentrated initially on investigating the energy needs of remote T/C stations, while emphasis is given on recording the corresponding energy consumption profile, see also Fig. 2. Accordingly, taking into account the available solar irradiance as well as the corresponding ambient temperature [25,26], one may calculate the optimum PV generator peak power and the battery capacity required in order to minimize the imported oil consumption or the life cycle operational cost of the entire installation. To confront similar problems, a computational algorithm ‘‘PHOTOV-DIESEL III’’ is developed. The specific numerical code is the extension of the already presented ‘‘PHOTOV-III’’ numerical code [14,15] and is used to carry out the necessary parametrical analysis on a given time step, e.g. on an hourly energy production-demand basis. More precisely, given the ‘‘Mf’’ value for each ‘‘z’’ and ‘‘Qmax’’ pair, the ‘‘PHOTOV-DIESEL III’’ algorithm is executed for the entire time-period selected (e.g. 1 month, 6-months, 1 year or even for 3 years), while emphasis is laid on obtaining zero-load rejection operation. After calculating the appropriate (Mf, z, Qmax) combinations that guarantee the stand-alone system energy autonomy, one may proceed to analyze the proposed PV-diesel hybrid installation energy balance in detail. In this context, the main steps of the ‘‘PHOTOV-DIESEL III’’ algorithm (see also Fig. 5) are: i. For every ‘‘Mf’’ value analyzed, select a number of ‘‘z’’ panels. ii. Accordingly select a ‘‘Qmax’’ value. iii. For every time point of a given time period (with a specific time step, e.g. 1 h) estimate the energy produced ‘‘NPV’’ by the PV generator, taking into account the existing solar radiation, the ambient temperature and the selected PV panel power curve.

iv. Compare the energy production with the isolated consumer energy demand ‘‘ND’’, increased to take into account any energy transformation losses. If any energy surplus occurs (NPV > ND), the energy surplus is stored to the battery bank and a new time point is examined (i.e. proceed to step vi). Otherwise, proceed to step (v). v. The energy deficit (ND–NPV) is covered by the energy storage system, only if the battery is not near the lower capacity limit (Q > Qmin). Accordingly proceed to step (vi). In cases that the battery is practically empty (Q 6 Qmin), the load is covered by the diesel-electric generator and the algorithm continues to step (vi). If the available diesel-oil quantity has been already consumed the battery size is increased (by a given quantity) and the complete analysis is repeated, starting from step (ii). vi. If the time period analyzed is completed proceed to step (i), increasing the number of PV panels ‘‘z’’. Otherwise proceed to step (iii). while the algorithm main inputs include the following: i. Detailed solar radiation ‘‘G’’ measurements for a given time period (usually 1 year), at horizontal plane. ii. Ambient temperature ‘‘h’’ data for the entire period analyzed so as to produce the PV panels’ surface temperature and assess its impact on the respective energy yield. iii. Operational characteristics (current ‘‘I’’, voltage ‘‘U’’, efficiency ‘‘gPV’’ and ‘‘N+’’) of the PV modules selected, usually in the form of I = I(U; G) (see also Fig. 6). iv. Operational characteristics of the diesel engine generator, e.g. efficiency/specific fuel consumption curve. v. Operational characteristics of all the other electronic devices of the installation, i.e. inverter efficiency, battery cell operational (Q–U; h) curve, etc. vi. Electricity consumption profile of the remote T/C station on an hourly basis for an entire year. Following the integration of the energy balance analysis, a (z–Qmax) curve is predicted under a given diesel-oil quantity ‘‘Mf’’. To get an unambiguous picture, keep in mind that for every pair of (z–Qmax) the stand-alone hybrid PV-based system is energy autonomous for the period investigated, using however a predefined diesel-oil quantity ‘‘Mf’’. Finally, the optimum pair may be selected from every (z–Qmax) curve, on the basis of a specific optimization criterion. One of the most widely used criteria for PV-based stand-alone systems is the definition of the minimum first installation cost configuration for every annual diesel-oil quantity used. In this context, the initial cost ‘‘ICo’’ of a PV stand-alone system can be approximated as:

IC o ¼ IC PV þ IC d þ IC bat þ IC elec þ f  IC PV

ð1Þ

where ‘‘ICPV’’ is the PV modules ex-works cost, ‘‘ICd’’ is the diesel engine purchase cost, ‘‘ICbat’’ is the battery bank buy-cost and ‘‘ICelec’’ the cost of the major electronic devices. Finally, the BOS (balance of system) cost is expressed via the first installation cost coefficient ‘‘f’’ and the ex-works cost of the PV panels. Using previous analysis by the authors [15,23,27], one may finally get:

IC o ¼ f  z  Pr  No  ð1 þ f Þ þ /  Nd þ cb  Q max þ a þ b  ðz  Nþ Þ

ð2Þ

where ‘‘f’’ is a function of ‘‘z’’ (i.e. f = f(z)), expressing the scale economies for increased number of PV panels utilized. In the present case (z  100) f is taken equal to one. Subsequently ‘‘Pr’’ is the specific buy-cost [28] of the PV panels (generally Pr = Pr(No)) expressed in €/kWp. Similarly, ‘‘cb’’ (€/Ah) is slightly dependent on the battery capacity. Thus for the local market – after a market survey [29]

J.K. Kaldellis, I. Ninou / Electrical Power and Energy Systems 33 (2011) 1739–1749

Fig. 5. PHOTOV-DIESEL III numerical algorithm.

1743

1744

J.K. Kaldellis, I. Ninou / Electrical Power and Energy Systems 33 (2011) 1739–1749

I-U CURVES OF THE KYOCERA LA361 K51S PANEL 1000W/m2

800W/m2

600W/m2

400W/m2

200W/m2

3.5 3.0

Current (Amp)

2.5 2.0 1.5 1.0 0.5 0.0 0

2

4

6

8

10

12

14

16

18

20

22

24

Voltage (Volt) Fig. 6. Power curve of a typical PV panel of 51 Wp under standard panel surface temperature of 25 °C.

concerning lead-acid batteries of 24 V – the following semiempirical relation may be used:

cb ¼

5:0377 Q 0:0784 max

ð3Þ

Additionally, the parameters ‘‘a’’ (in €) and ‘‘b’’ (in €/kW) describe the cost of the major electronic devices, being generally a function of the peak load demand (e.g. inverter) and the PV modules rated power (e.g. charge controller), respectively. Recapitulating, the initial installation cost of a stand-alone PV-based hybrid system is a function of ‘‘z’’ and ‘‘Qmax’’ (if ‘‘N+’’ is defined) as well as of the purchase and installation cost of the diesel electric generator and the installation inverter, being both functions of the peak load demand of the system. Taking into consideration that the peak load demand of the T/C station is given one may express the corresponding initial cost as a function of the number of the PV panels ‘‘z’’ and the battery capacity ‘‘Qmax’’, i.e.:

IC o ¼ IC o ðz; Q max Þ

ð4Þ

Hence, by using the initial cost function (Eq. (2)) it is possible to estimate, for any annual fuel consumption value corresponding to a specific (z–Qmax) curve, the minimum initial cost solution. Finally, it is important to note that the Greek State and the European Union strongly subsidy small PV systems [30], while the subsidization percentage ‘‘c’’ varies between 40% and 70% [31]. 4. Application results The proposed analysis is accordingly applied to a representative remote T/C station located in the Rhodes island. This S. Greece area possesses high solar potential, see also Fig. 3, since the annual solar energy per square meter exceeds 1800 kW h. More specifically, in Fig. 3 one may observe the solar irradiance profiles for all the months of a typical year, on the basis of the experimental data [32]. Accordingly, using the ‘‘PHOTOV-DIESEL III’’ algorithm on a PV-only mode (i.e. by setting Mf = 0) and for various tilt angles ‘‘b’’, one has the opportunity to estimate various energy autonomy configurations, Fig. 7. According to the results obtained, for all the ‘‘b’’ values examined, for every energy autonomy (Qmax–z) curve one may distinguish three separate regions. Thus, in the first part an abrupt battery capacity reduction is encountered as the PV

panels’ number is slightly increased. Subsequently, in the second part the battery capacity drop is decelerated, while in the third part the battery capacity remains almost constant achieving the same asymptotic value (approx. 9000 Ah) for all the ‘‘b’’ values examined [14]. Another important issue to be discussed is the quite bulk dimensions of the PV-only solution required in order to guarantee the energy autonomy of the T/C station on an annual basis. Actually, the PV panels are normally more than 700 (i.e. z > 700), while the corresponding lead-acid batteries’ capacity varies between 15,000 Ah and 30,000 Ah. Finally, the impact of the ‘‘b’’ angle on the proposed configuration dimensions is obvious, since for constant battery capacity there is a considerable ‘‘z’’ diminution as ‘‘b’’ increases from 0° to 45°, while the situation is inversed as ‘‘z’’ increases from 60° to 75°. On the basis of the results obtained, the minimum dimensions of the proposed installation seem to be realized for panels’ tilt angle around 60° [15]. In order to obtain a more quantitative evaluation, in the same Fig. 7 one may find the corresponding constant initial cost curves (including 40% initial investment subsidization). On the basis of the analysis described, the minimum initial cost solution that guarantees the T/C station energy autonomy without the utilization of a diesel-electric generator should be based on 740 PV panels (51 Wp each) and a battery array of approximately 33000 Ah (24 V), while the initial capital to be invested approaches 162,500€. Accordingly, by introducing a small diesel-oil quantity during the annual operation of the hybrid power station (i.e. Mf = 1000 kg/year or M f  8% M f , where M f ¼ 12:2 tn=year) significant battery capacity and number of PV-panels decrease is encountered in comparison with the PV-only solution, Fig. 8. Actually, Fig. 8 presents the energy autonomous PV panel and battery capacity combinations for several panel tilt angles ‘‘b’’, varying from 0° to 75 °, under the condition that the maximum available annual diesel-oil quantity is 1000 kg, i.e. only 8% of the corresponding one required for the diesel-only solution (Mf ¼ 12185 kg=year). Opposite to the PV-only solution presented above, one may also state that the proposed installation dimensions are quite rational, since by using battery (24 V) capacity less than 10,000 Ah and PV-generator peak power demand in the order of 25 kWp one may guarantee an entire year energy autonomy of the T/C station, using also only 1 tn of diesel oil, thus saving almost 92% of the up to now annual fuel consumption. Finally, the influence of the ‘‘b’’

1745

J.K. Kaldellis, I. Ninou / Electrical Power and Energy Systems 33 (2011) 1739–1749

STAND-ALONE HYBRID PV-BASED CONFIGURATION FOR A REMOTE T/C STATION IN SOUTH GREECE (Mf=0Kg/y)

β=0

BATTERY CAPACITY Qmax (Ah)

60000

β=15 β=30

50000

β=45 β=60

40000

β=75 ICo=150000

30000

ICo=162500 ICo=175000

20000 10000 0 650

750

850

950

1050

1150

1250

1350

1450

1550

NUMBER OF PANELS (z) Fig. 7. PV-only configuration for a typical remote T/C station.

Fig. 8. Energy autonomous PV-based hybrid configurations for a typical T/C station.

angle on the hybrid configuration dimensions is also significant. More precisely, there is a considerable ‘‘z’’ diminution as ‘‘b’’ increases from 0° to 45°, while accordingly the ‘‘z’’ number is remarkably increased as ‘‘b’’ takes values higher than 60°. In this context, there are several rational (Qmax–z) combinations between 45° and 60° that guarantee energy autonomy of the T/C station, see for example the b = 52.5° curve. The next set of Fig. 9a–f describe all the energy autonomous configurations of the PV-based hybrid station as a function of the PV panels’ tilt angle and for a wide range of diesel-oil annual consumption values (i.e. from Mf = 0 to Mf = 5000 kg/year or from 0% up to 40% of the diesel-only solution). As it results from this figure, for all the ‘‘b’’ values examined there is a considerable ‘‘z’’ number decrease (for constant Qmax = ct values) as the Mf value increases. Note that this decrease is much more important when one passes from the PV-stand-alone mode to the hybrid PV-based mode, i.e. by introducing only 1000 kg of diesel-oil on annual basis. Besides, the asymptotic Qmax value of all ‘‘b’’ values examined is gradually reduced approaching the zero value (Qmax ? 0) as the Mf value tends to 5 tn/year. This is quite rational, since the energy storage

need is substituted by the direct consumption of increased diesel-oil quantities. Finally, one should also mention the remarkable slope change of the energy autonomous curves, which becomes more abrupt as Mf increases. The same behavior is valid for all the ‘‘b’’ values investigated, Fig. 9a–f.

5. Energy balance analysis As already stated, one of the main targets of the present study is to extensively analyze the energy balance of the proposed hybrid PV-based installation for the complete time period investigated. To get a representative picture of the proposed system performance, Fig. 10 presents a typical winter week (January) energy balance profile based on the experimental solar irradiance and ambient temperature data (Fig. 3), resulting from the operation of the minimum initial cost configuration (Mf = 1000 kg/year, b = 60°) under the no-load rejection restriction. In the same figure one may find the corresponding battery bank state of charge values ‘‘Q’’. Using the available solar radiation measurements, a 2-day low

70000

Mf=0 Mf=1000 Mf=2000 Mf=3000 Mf=5000

60000 50000 40000 30000 20000 10000 0 200

400

600

800

1000

1200

1400

1600

Mf=0 Mf=1000 Mf=2000

50000

Mf=3000

40000

Mf=5000

30000 20000 10000 0 600

800

1000

1200

1400

1600

20000 10000 0 0

200

400

600

800

1000

1200

70000

Mf=1000 Mf=2000

50000

Mf=3000 Mf=5000

40000 30000 20000 10000 0 600

800

1000

1200

1400

1600

1600

Mf=0 Mf=1000 Mf=2000 Mf=3000 Mf=5000

60000 50000 40000 30000 20000 10000 0 0

200

400

600

800

1000

1200

(e)

Mf=0

1400

ENERGY AUTONOMOUS HYBRID PV-BASED CONFIGURATION FOR A REMOTE T/C STATION IN SOUTH GREECE (β=60ο)

(b)

60000

400

30000

NUMBER OF PANELS (z)

70000

200

Mf=3000 Mf=5000

40000

NUMBER OF PANELS (z)

ENERGY AUTONOMOUS HYBRID PV-BASED CONFIGURATION FOR A REMOTE T/C STATION IN SOUTH GREECE (β=30ο)

0

Mf=1000 Mf=2000

50000

(d)

60000

400

Mf=0

60000

(a)

70000

200

70000

NUMBER OF PANELS (z)

ENERGY AUTONOMOUS HYBRID PV-BASED CONFIGURATION FOR A REMOTE T/C STATION IN ο SOUTH GREECE (β=15 )

0

ENERGY AUTONOMOUS HYBRID PV-BASED CONFIGURATION FOR A REMOTE T/C STATION IN ο SOUTH GREECE (β=45 )

NUMBER OF PANELS (z)

BATTERY CAPACITY Qmax (Ah)

0

BATTERY CAPACITY Qmax (Ah)

BATTERY CAPACITY Q max (Ah) BATTERY CAPACITY Qmax (Ah)

ENERGY AUTONOMOUS HYBRID PV-BASED CONFIGURATION FOR A REMOTE T/C STATION IN ο SOUTH GREECE (β=0 )

BATTERY CAPACITY Qmax (Ah)

J.K. Kaldellis, I. Ninou / Electrical Power and Energy Systems 33 (2011) 1739–1749

BATTERY CAPACITY Q max (Ah)

1746

1400

1600

ENERGY AUTONOMOUS HYBRID PV-BASED CONFIGURATION FOR A REMOTE T/C STATION IN ο SOUTH GREECE (β=75 ) 70000 60000

Mf=0 Mf=1000

50000

Mf=2000 Mf=3000

40000

Mf=5000

30000 20000 10000 0 0

200

400

600

800

1000

1200

NUMBER OF PANELS (z)

NUMBER OF PANELS (z)

(c)

(f)

1400

1600

Fig. 9. Annual diesel-oil consumption and panels’ tilt angle impact on the energy autonomous configuration of a remote PV-based hybrid power station.

energy production period is encountered, leading to a significant battery ‘‘DOD’’ increase or equivalently ‘‘Q’’ decrease. For this purpose the diesel-generator is activated to cover the energy deficit. A similar problem appears also during the 6th day of the week, leading also to considerable diesel-oil consumption. Generally speaking, the battery bank charge condition is quite low for the entire week, achieving ‘‘DOD’’ values in the order of 60% and approaching the first battery protection limit. However, the proposed configuration manages to cover the energy demand of the T/C station for the entire year without any load rejection, fulfilling also the minimum initial cost constraint. Proceeding now to the energy balance analysis of the proposed solution one may compare in Fig. 11 the corresponding energy balance results for the three representative points (A, B and C) of the proposed solution, see also Fig. 8. Actually, the specific case analyzed corresponds to Mf = 1000 kg/year and b = 60° autonomy curve of Fig. 8. In this context, the solution ‘‘A’’ uses the minimum

PV-panels number (z = 478) but quite bulk lead-acid batteries. Note that the PV generator covers almost 89% (Fig. 11a) of the annual energy production (i.e. 26.2 MW h/year), while the diesel generator contribution is only 3.3 MW h on an annual basis. As a result there is almost no energy surplus (Fig. 11b), i.e. the PV-generator and the diesel-electric generator hardly cover the T/C station load demand, while the corresponding system losses are approximately 2.5 MW h. Next, by slightly increasing the number of PV panels (point B, z = 536) a considerable battery capacity reduction is encountered (i.e. 7:1). More precisely, for this case the annual PV production is increased by 3.2 MW h, which is mainly transformed to system energy surplus, while the system losses approach 3.6 MW h. As in the previous case, the contribution of the diesel-generator is hardly reaching 10% on annual basis. Finally, for the last point ‘‘C’’ examined here, the ‘‘z’’ value is significantly increased (z = 800), while the battery capacity is fairly reduced. Due to the greater number of PV panels used, the energy

1747

J.K. Kaldellis, I. Ninou / Electrical Power and Energy Systems 33 (2011) 1739–1749 st

HYBRID T/C STATION ENERGY ANALYSIS: 1 WEEK of JANUARY (z=526, Qmax=8100, β=60ο, Mf=1000kg/y) PV-Power

8000

Load 7000

Diesel Production

16000

6000

ENERGY (Wh)

Battery Limit 11000

5000 4000

6000 3000

Battery Capacity (Ah)

Battery Capacity

2000 1000 1

11

21

31

41

51

61

71

81

91

101

111

121

131

141

151

161

1000

Hours -4000

0

Fig. 10. Energy analysis of a stand-alone PV-based hybrid T/C station during winter period.

ENERGY BALANCE OF A REMOTE T/C STATION AT PANELS' TILT ANGLE OF 60º

ENERGY (MWh)

50 45

Epv

40

Ediesel

35 30 25 20 15 10 5 0

A

B

C

Mf (1000kgr/Year) ENERGY BALANCE OF A REMOTE T/C STATION AT PANELS' TILT ANGLE OF 60º 50

ENERGY (MWh)

45 40

Eloss Esurpl Econsumption

35 30 25 20 15 10 5 0

A

B

C

Energy Autonomy Points (Mf=1000kg/y) Fig. 11. Energy balance of a typical remote PV-based hybrid T/C station.

production of the system not only covers the T/C station energy requirements and the corresponding system losses but also produces a significant energy surplus that approaches 15.3 MW h/ year. Considering the energy balance analysis undertaken, the optimum system dimensions should be expected near the point B area (for a given annual fuel quantity). This conclusion is quite rational, since the battery capacity of point B is considerably less than the one of point A and because for point B there is almost zero energy surplus, which is not the case for point C (due to the quite greater number of PV panels required). As it is obvious, for the optimum

sizing of a similar PV-based solution one needs a complete life cycle cost-benefit analysis, taking also into consideration the annual fuel cost, which however is the subject of a forthcoming study [23]. Accordingly, in Fig. 12a and b one may find the impact of the PV-panels’ tilt angle [33] on the installation energy balance, given the annual diesel-oil consumption equal to Mf = 1000 kg/year. Note that for comparison purposes the points selected for each b = ct angle correspond to the minimum initial cost solution. According to the results presented in Fig. 12a there is a gradual PV-generator annual energy yield decrease as the ‘‘b’’ angle is increased, mainly due to the lower number of the PV-panels required to guarantee the energy autonomy of the installation. As a result of this evolution, the energy surplus of the T/C station is continuously decreasing (Fig. 12b), while the system loss remains almost constant around 3.8 MW h. Finally, Fig. 13a and b demonstrate the energy balance of the remote T/C station under investigation for constant panels’ tilt angle (b = 60°) and for annual diesel-oil consumption varying from zero (PV-only solution) up to 5000 kg/year. At this point it is important to mention that for the PV-only solution the oversized PV-generator (64.3 kWp) produces almost 70 MW h/year (Fig. 13a), although the annual electricity consumption of the T/C station is slightly higher than 27 MW h. On the other hand, by introducing only 1000 kg/year of diesel-oil the annual energy production of the hybrid station drops to 30 MW h, enough to cover the system energy requirements, Fig. 13b. Actually, by a closer examination of Fig. 13b one may observe a significant reduction of the system energy surplus (meaning that the system is not oversized), while also the PV-generator loss is considerably decreased (i.e. only 1.3 MW h). On the other hand, generally speaking, there is a remarkable primary energy loss due to the diesel-electric generator wider operation, taking into consideration that its round trip efficiency is around 20%. This conclusion is obvious by examining the results of Fig. 13a and comparing the PV-generator contribution for Mf = 1000 kg/year (i.e. 89.5%) and for Mf = 5000 kg/year (i.e. 41.3%). Overall, by contrasting the energy balance results obtained for the specific application with the results obtained from studies considering more common type applications (e.g. residential or community level applications), suitability of the proposed PV-based hybrid solution for the satisfaction of T/C stations may be reflected. In fact, according to the results of representative studies [34–37], the solar energy penetration is in most cases limited (even at the levels of 20%), this deriving either from certain problem restrictions (e.g. battery capacity restrictions) or from the economic

1748

J.K. Kaldellis, I. Ninou / Electrical Power and Energy Systems 33 (2011) 1739–1749

performance of the configurations investigated (e.g. optimization on the basis of minimum electricity production cost), or from the energy demand patterns studied. Contrariwise, although the first installation criterion is currently adopted, minimization of the cost implying also – for certain configurations – maximum exploitation of the energy production (i.e. minimization of the solar energy surplus) and rational use of diesel oil (reduced even at the levels of 90% in comparison with the diesel only solution) indicates the appropriateness of the proposed solution for the satisfaction of such applications. Besides, as it accrues from all cases where comparison is attempted between the hybrid configuration and the PV-only and diesel-only solutions [34–37], the clear advantage of coupling solar energy with diesel power is illustrated, irrespective of the evaluation criterion and the resulting optimum energy balance mix each time decided.

ENERGY BALANCE for a REMOTE T/C STATION Mf=1000 kgr/year (Minimum ICo points) 45

Epv

ENERGY (MWh)

40

Ediesel

35 30 25 20 15 10 5 0 0

15

30

45

52,5

60

75

ANGLE (Degrees)

ENERGY (MWh)

ENERGY BALANCE for a REMOTE T/C STATION Mf=1000 kgr/year (Minimum ICo points) 6. Conclusions

45

Eloss

40

Esurpl EConsumption

35 30 25 20 15 10 5 0 0

15

30

45

52,5

60

75

ANGLE (Degrees) Fig. 12. Panels’ tilt angle impact on the energy balance of a typical remote PV-based hybrid T/C station.

ENERGY BALANCE OF REMOTE T/C STATION AT PANELS' TILT ANGLE OF 60º 80

Epv

ENERGY (MWh)

70

Ediesel

60 50 40 30 20 10 0

0 kg

1000 kg

2000 kg

5000 kg

Mf (kgr/Year) ENERGY BALANCE OF REMOTE T/C STATION AT PANEL TILT ANGLE OF 60º 80

ENERGY (MWh)

70 Eloss 60

Esurpl

50

Econsumption

40 30

In the present study, the energy autonomy configurations of a typical T/C station located in S. Greece are determined on the basis of a PV-based hybrid power station, using extensive solar potential and ambient temperature data. The proposed installation includes also an appropriate battery bank and a small diesel-electric generator, operating as a back up engine. The results obtained are based on experimental long-term measurements and operational characteristics by the proposed system components manufacturers. According to the results obtained, there is a remarkable system size and initial cost discrepancy between the zero diesel-oil (PV-only) and the 5000 kg/year (corresponding to 41% of the diesel-only consumption) solutions, strongly supporting the adoption of a hybrid solution instead of a diesel-only or a PV-only option. Subsequently, the desired panels’ tilt angle considerably affects the necessary system dimensions (battery capacity and PV-generator peak power), especially for low diesel-oil contribution scenarios. Actually, for every ‘‘Mf’’ value examined there is an optimum panels’ tilt angle value that minimizes the system initial cost. Note that for higher ‘‘Mf’’ values the panels’ tilt angle value impact is limited. Finally, an extensive energy balance analysis is carried out for several hybrid system configurations. Detailed results on an hourly and annual basis are demonstrated for various system panels’ tilt angles and representative annual diesel-oil quantities available. According to the results obtained (excluding the PV-only solution), the vast majority of the PV generator production is finally transferred to the consumption, while 11–12% is the PV-based system components energy losses. In this context, an optimum sized PVbased hybrid system presents also minimum energy surplus. Consequently, the energy balance performance of a PV-based hybrid system applied at a representative T/C station for a longterm period, along with its high reliability and low maintenance needs, reflect the advantages of the proposed solution – in comparison with other available alternatives – for the energy demand problem of numerous existing isolated T/C stations. On top of this, using the above-described methodology, one has the opportunity to select the desired hybrid power system optimum dimensions and the PV panels’ tilt angle, appreciating at the same time the necessary capital to be invested.

20 10

References

0

0 kg

1000 kg

2000 kg

5000 kg

Mf (kgr/Year) Fig. 13. Annual diesel-oil consumption impact on the energy balance of a typical remote PV-based hybrid T/C station.

[1] Gangotri KM, Bhimwal MK. The photochemical conversion of solar energy into electrical energy: Eosin–Arabinose system. Int J Electr Power Energy Syst 2010;32:1106–10. [2] Bhati KK, Gangotri KM. Photogalvanic conversion of solar energy into electrical energy by using NaLS–xylose–methylene blue system. Int J Electr Power Energy Syst 2011;33:155–8.

J.K. Kaldellis, I. Ninou / Electrical Power and Energy Systems 33 (2011) 1739–1749 [3] Albrecht J. The future role of photovoltaics: a learning curve versus portfolio perspective. Energy Policy 2007;35:2296–304. [4] Hoffmann W. PV solar electricity industry: market growth and perspective. Sol Energy Mater Sol Cells 2006;90:3285–311. [5] Kazmerski LL. Solar photovoltaics R&D at the tipping point: a 2005 technology overview. J Electron Spectrosc Related Phenomena 2006;150:105–35. [6] Lazou AA, Papatsoris AD. The economics of photovoltaic stand-alone residential households: a case study for various European and mediterranean locations. Sol Energy 2000;62:411–27. [7] Kaldellis JK, Zafirakis D, Kondili E. Optimum sizing of photovoltaic–energy storage systems for autonomous small islands. Int J Electr Power Energy Syst 2010;32:24–36. [8] Lam PL, Shiu A. Economic growth, telecommunications development and productivity growth of the telecommunications sector: evidence around the world. Telecommun Policy 2010;34:185–99. [9] Koufomichalis C, Bargotakis C, Kaldellis JK, Tsoutsos Th. Examples of RES applications in the telecommunications. In: 7th National conference on the soft energy resources, Patras, Greece; November 2002. [10] Kaldellis JK, Vlachos GTh, Kavadias KA. Optimum sizing basic principles of a combined photovoltaic–wind–diesel hybrid system for isolated consumers. In: EuroSun 2002 international conference, Bologna, Italy; June 2002. [11] Liu S, Dougal RA, Solodovnik EV. Design of autonomous photovoltaic power plant for telecommunication relay station. IEE Proc – Gener Transm Distrib 2005;152:745–54. [12] Medrano A. Photovoltaic experiences in the electric supply for relay telecommunication station in Spain. In: 15th International telecommunications energy conference/intelec 93, Paris, France; September 1993. [13] Ninou I, Kaldellis JK. Energy balance analysis of a hybrid photovoltaic based solution for remote telecommunication stations. In: Xth World renewable energy congress, Glasgow-Scotland, UK; July 2008. [14] Kaldellis JK, Koronakis P, Kavadias K. Energy balance analysis of a stand-alone photovoltaic system, including variable system reliability impact. Renew Energy 2004;29:1161–80. [15] Kaldellis JK. Optimum techno-economic energy-autonomous photovoltaic solution for remote consumers throughout Greece. Energy Convers Manage 2004;45:2745–60. [16] Tu R, Liu XH, Li Z, Jiang Y. Energy performance analysis on telecommunication base station. Energy Build 2011;43:315–25. [17] Kaldellis JK, Kavadias KA, Magoulas Chr. Fuel saving in telecommunication stations by using photovoltaic generators. In: 1st National conference of mechanical and electrical engineers, Athens, Greece; March 2005. [18] Kaldellis JK. An integrated feasibility analysis of a stand-alone wind power system, including no-energy fulfillment cost. Wind Energy 2003;6:355–64. [19] Huld Th, Gottschalg R, Beyer HG, Topicˇ M. Mapping the performance of PV modules, effects of module type and data averaging. Sol Energy 2010;84: 324–38.

1749

[20] Jossen A, Garche J, Sauer DU. Operation conditions of batteries in PV applications. Sol Energy 2004;76:759–69. [21] Kottick D, Blau M. Operational and economic benefits of battery energy storage plants. Int J Electr Power Energy Syst 1993;15:345–9. [22] Kaldellis JK, Zafirakis D, Kavadias K, Kondili E. An optimum sizing methodology for combined photovoltaic-energy storage electricity generation configurations. Sol Energy Eng – Trans ASME 2009;131:210101–12. [23] Kaldellis JK, Kavadias KA. Cost-benefit analysis of remote consumers’ electrification on the basis of hybrid wind-diesel power stations. Energy Policy 2007;35:1525–38. [24] Muselli M, Notton G, Louche A. Design of hybrid-photovoltaic power generator with optimization of energy management. Sol Energy 1999;65:143–57. [25] Kaldellis JK, Spyropoulos G, Kavadias K. Computational applications of soft energy resources: solar potential-photovoltaic applications-solar heat systems. 1st ed. Athens: Stamoulis; 2007. [26] Luque A, Hegedus St. Handbook of photovoltaic science and engineering. 1st ed. West Sussex: John Wiley & Sons; 2003. [27] Kaldellis JK, Zafirakis D, Kaldelli El, Kavadias K. Cost benefit analysis of a photovoltaic-energy storage electrification solution for remote islands. Renew Energy 2009;34:1299–311. [28] Haas R. Building PV markets: customers and prices. Renew Energy World 2002;5:98–111. [29] Kaldellis JK, Tsesmelis M. Integrated energy balance analysis of a stand-alone wind power system, for various typical Aegean Sea regions. Wind Energy 2002;5:1–17. [30] Sandén BA. The economic and institutional rationale of PV subsidies. Sol Energy 2005;78:137–46. [31] Papadopoulos AM, Karteris MM. An assessment of the Greek incentives scheme for photovoltaics. Energy Policy 2009;37:1945–52. [32] Public Power Corporation (PPC). Wind speed measurements for Greece, 1980– 1985. 1st ed. Athens: PPC; 1986. [33] Chang YP. Optimal tilt angles for photovoltaic modules in Taiwan. Int J Electr Power Energy Syst 2010;32:956–64. [34] Shaahid SM, Elhadidy MA. Economic analysis of hybrid photovoltaic–diesel– battery power systems for residential loads in hot regions—a step to clean future. Renew Sust Energy Rev 2008;12:488–503. [35] Lau KY, Yousof MFM, Arshad SNM, Anwari N, Yatim AHM. Performance analysis of hybrid photovoltaic/diesel energy system under Malaysian conditions. Energy 2010;35:3245–55. [36] Hrayshat ES. Techno-economic analysis of autonomous hybrid photovoltaic– diesel–battery system. Renew Sust Energy Rev 2009;13:143–50. [37] Shaahid SM, El-Amin I. Techno-economic evaluation of off-grid hybrid photovoltaic–diesel–battery power systems for rural electrification in Saudi Arabia—a way forward for sustainable development. Renew Sust Energy Rev 2009;13:625–33.