Engineering, Technology & Applied Science Research Vol. 8, No. 4, 2018, 3168-3171 3168 www.etasr.com Mavromatakis et al.: Photovoltaic Systems and Net Metering in Greece Photovoltaic Systems and Net Metering in Greece Fotis Mavromatakis Department of Electrical Engineering Technological Educational Institute of Crete Heraklion, Crete, Greece fotis@staff.teicrete.gr George Viskadouros Department of Electrical Engineering Technological Educational Institute of Crete Heraklion, Crete, Greece viskadouros@staff.teicrete.gr Hara Haritaki Department of Accounting and Finance Technological Educational Institute of Crete Heraklion, Crete, Greece haritaki@staff.teicrete.gr George Xanthos Department of Business Administration Technological Educational Institute of Crete Heraklion, Crete, Greece xanthos@staff.teicrete.gr Abstract—The latest measure for the development of photovoltaics in Greece utilizes the net-metering scheme. Under this scheme the energy produced by a PV system may be either consumed by the local loads or be injected to the grid. The final cost reported in an electricity bill depends upon the energy produced by the PV system, the energy absorbed from the grid and the energy injected to the grid. Consequently, the actual electricity consumption profile is important to estimate the benefit from the use of this renewable energy source. The state latest statistics in Greece for households reveal that the typical electrical consumption is 3750 kWh while 10244 kWh are consumed in the form of thermal energy. We adopt in our calculations the above amount of electrical energy but assume four different scenarios. These different hourly profiles are examined to study the effects of synchronization upon the final cost of energy. The above scenarios are applied to areas in different climate zones in Greece (Heraklion, Athens and Thessaloniki) to examine the dependence of the hourly profiles and the solar potential upon the financial data with respect to internal rate of return, payback times, net present value and the levelized cost of energy. These parameters are affected by the initial system cost and the financial parameters. Keywords-photovoltaic; net metering; modeling; financial I. INTRODUCTION One of the support mechanisms to promote photovoltaic (PV) technology and reduce the energy costs for residential and commercial customers is related to the local production of renewable energy (self-production). The net metering scheme involves grid connected systems and is available in many countries worldwide. Utilities can better manage peak loads since PV systems generate most of their power around noon time. Local power production allows reducing the strain in the electrical distribution system and the transmission and distribution losses. Part of the energy generated by the PV system may be consumed by the loads (self-consumption) of the owner of the PV system, e.g. lights, refrigerators, pumps, etc, while any surplus energy will be exported to the grid. The owner of the PV system will be billed for the net energy which is basically equal to the energy retrieved from the grid minus the energy injected to the grid. In Greece the relevant law was introduced in 2014. Recently, in 2017, the Greek state setup the details of net metering and introduced the concept of virtual net metering. Under this scheme the points of energy production and consumption may well differ electrically or spatially. In addition, it is foreseen that a customer may incorporate more points of consumption into the energy balance agreement. In this work we focus on the net metering scheme for residential customers. The total billing of energy in households involves basically two parts. The first part refers to the cost of the actual energy consumed by the customer (“energy supply charges”). The second part refers to a few regulated costs like those related to transmission, distribution, services of common wealth, reduction of gas pollution, special fees and other charges (“regulated costs”). The “energy supply charges” depend upon the specific energy supplier company, which may be other than the Greek Public Power Corporation (PPC), whereas the “regulated costs” are common to all energy suppliers. Both categories of costs are calculated in a specific way and in this work those defined by the PPC are adopted. We explore the implications of net metering at three different geographical areas in Greece. For each area, we assume a number of hourly profiles for the loads resulting to the same annual energy consumption. Adopting typical values for the financial data the payback times, net present value and the levelized cost of energy are calculated and compared. II. MODEL DATA The input model data may be distinguished to two categories. The first involves the data related to the photovoltaic system while the second involves the data related to the consumer load. The PV*SOL software was utilized for the hourly PV calculations [1]. This software makes use of the METEONORM database to retrieve solar irradiance and meteorological data for the selected sites: Heraklion – Crete, Ath pea con typ kW ran sum ave 37 am kW (84 mo fol fol on wi “re fro syn loa A. rat PV dif sys to com loa the dep am tha res the sce the Ath ado ava ado con ado pro be Engineerin www.etasr thens - Attica ak power of e nsumed energ pical commerc W while the m nge of 250-3 mmarizes the b Site Heraklio Athens Thessalon According to erage annual e 50 kWh, whi mounts to 139 Wh refers to th 4.9%), water ostly used to g llowed by wo llow with muc the electrical ll be conside egulated costs om the grid, nchronization ads. The hourly e Formally, the tio of self-cons V system. Th fference betwe stem minus th Formally, the 100%. The mi mpletely out o ads) while the e produced P pend upon the mount of energ at is adopted a sulting in the e consumer u enario (S2) uti e Centre of Re thens and the opted in the c ailable as a opted in this s nsumption in opts a load c ofile to increas a real profile ng, Technology r.com and Thessalo each PV system gy. In all case cial inverter w modules are ch 300W at ST basic data for TABLE I. GHI (kWh/m on 1870 s 1710 niki 1580 o the Greek S electrical energ le the corresp 94 kWh [3]. hermal energy heating (4.4% generate therm ood burning (2 ch smaller per l energy and it ered in the n s” depend ma it is importan between the energy profiles e degree of syn sumed energy e self-consum een the AC e e AC energy i e degree of sy inimum degre of phase with e maximum re PV energy. C e degree of sy gy consumed f assumes that worst-case sc under the net ilizes results f enewable Ener average hourl current work [ built-in profi study. The spe a block of fla curve centered se the degree e, but it is imp y & Applied Sci niki - Norther m was chosen s, the PV sys with a nomina haracterized by TC conditions all three sites. BASIC DATA Gt m2) (kWh/m 1960 1800 1670 Statistical Aut gy consumptio ponding total The additiona y which is use %) and cookin mal energy is 23.8%). Other rcentages. In t is the energy net metering ainly upon the nt to understa generation of s nchronization, y over the ener med energy is energy (Eac) p injected to the ynchronization ee refers to loa the solar ener efers to loads Consequently, ynchronization from the grid. all loads are o cenario (S1), i t metering sc from a short su rgy Sources (C ly profile of a [4]. The third ile in the sof ecific profile r ats. Finally, th d close to noo of synchroniz plemented to ience Research rn Greece [2]. n to fully matc tem comprise al power aroun y peak power s. Table I b Pp m2) (kW) 2.43 2.65 2.80 thority in 201 on of a househ energy requir al energy of ed for space h ng (9.7%). Th diesel oil (60 r sources of e this work we y of 3750 kW scheme. Sinc e energy cons and the conce f PV power an , S, is defined rgy produced b s calculated a roduced by th grid (Einj). n may vary fro ads that are bas rgy (e.g. only that exactly m , the billing n since it affec The first load on during the i.e. highest co cheme. The s urvey perform CRES) in the a all measureme load profile ( ftware and is refers to the e he last scenario on time with zation. This ma study the effe h V Mavro . The ch the s of a nd 2.5 in the below 13 the hold is ement 10244 eating he fuel 0.3%), energy focus Wh that ce the sumed ept of nd the as the by the as the he PV m 0% sically y night match costs cts the curve night ost for second med by area of ents is S3) is s also energy o (S4) a flat ay not ects of incr arb the 3,7 B. cus calc are in stat per eur the acc met cor scen arou scen scen pur pro than con of z eve calc und fina inv of s and of s the cos resi still ene C H Vol. 8, No. 4, 20 omatakis et al.: reased synchr itrary units. T total annual 50 kWh. Fi Financial dat As already m stomers and i culated with d issued every one calendar ted by the PP riod in Herak ros. This value PPC and ar cording to the tering (S1), w rresponding co nario S1, the und noon tim nario, concern nario (S4) is a rposes. It is u ofile. During n the energy r ntribution to th zero net energy ery site and ev culated to dete der the net me ancial calcul estment. Anot synchronizatio d scenarios. Th synchronizatio other two citi st of energy w idence. It is c l worth cons ergy. TABL Heraklion Athens Thessaloniki Cost of energy eraklion (1 yr) 018, 3168-3171 Photovoltaic Sy ronization. All The software n energy consu ig. 1. The dif ta mentioned, th in this case, data provided b four months, i year. Accord PC, the cost o klion, without e does not incl re, then, attrib Greek law. In where all loads ost drops to best scenario me. The total c ning the same an ideal scena unlikely to co summer time equired by the he total cost. T y, there is a m very scenario, ermine the ann etering scheme lations conce ther interesting on which is sho he dependence on is evident. ies and are no without net m clear that even idering net m LE II. DEGRE S1 0.0% 3 0.0% 3 0.0% 3 285 € 2 1 Systems and Net l profiles are normalizes the umed by the h fferent hourly prof his work focu the typical by the PPC [5 i.e. there are th ding to the r of a typical fo net metering ude charges th buted to the n the worst-ca s are active du around 92 e is S4 where t cost rises to 2 four-month su ario and is sho onsider a resid the PV energ e loads but stil This is becaus minimum cost s the four-mon nual electricity e. These costs erning the g parameter is own in Table e of the saving Similar costs ot repeated her metering rises t n in the worst- metering to r EE OF SYNCHRON S2 S 7.8% 44 7.3% 43 6.7% 42 204 € 20 3169 t Metering in G shown Figure ese profiles, so household loa files uses on reside cost of energ 5]. The energy hree clearance esidential tari our-month sum g, amounts to hat are collecte local municip ase scenario o uring the nigh euros. Contrar the loads are a 28 euros under ummer period. wn for compa dence with su gy is usually h ll there is a non se even in the set by the PPC nth billing cost y and savings are adopted i viability of s the annual d II for all three gs upon the d s are calculate re. The total an to 664 euros -case scenario educe the co IZATION S3 S4 4.5% 71.3 .4% 70.1 2.4% 68.2 00 € 132 Greece e 1 in o that ads is ential gy is y bills e bills iff as mmer o 220 ed by pality of net ht, the ry to active r this . This arison uch a higher nzero e case C. For ts are costs in the f the egree e sites egree ed for nnual for a o it is ost of 4 3% 1% 2% 2 € Engineering, Technology & Applied Science Research Vol. 8, No. 4, 2018, 3168-3171 3170 www.etasr.com Mavromatakis et al.: Photovoltaic Systems and Net Metering in Greece III. DISCUSSION The use of net metering by residential customers in three different areas and climate zones in Greece is considered in this work. The cost of energy for each scenario is practically the same for all sites explored, while the annual degree of synchronization shows minor variations (Table II). The financial metrics used to evaluate the viability of the investment of such domestic PV systems involve the levelized cost of electricity (LCOE), the simple payback time (PBT), the net present value (NPV) and the internal rate of return (IRR). These are calculated following the formulation provided by the National Renewable Energy Laboratory (NREL) [6]. The LCOE represents the cost of energy per kWh produced by the PV system over the investment horizon, while the simple PBT represents the time it takes for the net revenues to equal the initial investment cost. The NPV represents the savings over the same period taking in terms of the current value of the money. The IRR represents the deflated cost of capital at which the NPV becomes zero during the life time of the project. The basic financial parameters used in the calculations are summarized in Table III. In this work the investment time interval is set to 25 years and basically, reflects the duration of the contract signed with the Hellenic Electricity Distribution Network Operator S.A. under the net metering scheme. However, it must be made clear that this doesn’t represent the actual life time of a PV module. It formally represents the time interval, at the end of which, a module may not produce less than 80% of its initial power rating (guaranteed by the manufacturers). A module will keep producing energy although degraded by about 0.4%/yr [7, 8]. TABLE III. FINANCIAL DATA Initial cost (€/kW)1 System degradation2 Inverter replacement (€/kW)3 WACCnom4 & Inflation5 Scenarios 1,000.0 0.4% 350 7% & 4% S1, S2, S3, S4 1,250.0 0.4% 350 7% & 4% S1, S2, S3, S4 1,500.0 0.4% 350 7% & 4% S1, S2, S3, S4 1Initial cost of PV system 2System degradation per year 3Inverter replacement per ten years 4WACCnom: nominal weighted average cost of capital 5Inflation: the annual inflation rate A replacement time of string inverters of 10 years, considering a good quality inverter, is adopted. In such small PV systems the owner of the system may carry out the operation and maintenance costs (O&M) and thus, do not pay any costs for the cleaning of the modules and the inverter, the inspection of cable connections, etc. Finally, a fixed cost of 300 euros is incorporated in year 0 costs as a grid connection fee foreseen by the state law. A single-phase grid connection is assumed since the examined PV peak power is small. To examine the viability of the investment it is crucial to integrate local economy parameters such as the inflation and the nominal weighted average cost of capital (WACC). Inflation is an increase in the cost of goods and services per unit time. For convenience, inflation is customary to refer to a year. Based on the performance of economic indicators, an average annual inflation rate of 4% was selected [9]. On the other hand, WACC is associated with the amount of profit that obtained from saving capital. Net Metering is nothing more than an agreement with a power company, enabling the consumer to install a solar power system to meet part or all of the energy consumption. The power company compensates the energy generated by the solar modules with the power consumed by the owner of the photovoltaic system. When there is excess energy because the consumption is low it will be supplied to the grid. On the other hand, when the PV system does not produce enough energy (e.g. clouds or night time) then energy will be consumed from the grid. When the demand for electricity is consistent with production, the compensation price increases and thus, depreciation of capital spent for the installation of the system becomes faster. It should be noted that the excess energy does not lead to income growth due to lack of agreement in tariff for net metering systems. The analysis of a household in Heraklion shows that, in the worst-case scenario S1 (just night loads), the breakeven point is 13 years for a PV system price of 1,250 €/kWp. This parameter ranges from 9 to 16 years considering system prices of 1,000 and 1,500 €/kWp and all possible scenarios. The cash flow analysis for this household meets the agreement mentioned earlier for the electrical energy demand and shows that in the consumption profile S2 (mixed daily and nights loads) the breakeven point is 8 years for a PV system price of 1,250 €/kW. Generally, in the case of scenarios S2 and S3 which better represent everyday profiles, the breakeven points drop to 7, 8 and 12 years for the corresponding initial system costs reported in Table III. The financial savings for these consumption profiles in Heraklion are very close to 3,600 euros (NPV) while the internal rate of return is close to 11-12% for the price of 1,250 €/kW which is quite satisfactory since the real average cost of capital in the calculations is 2.9%. If the real cost of capital is below the IRR value of 11-12%, then the investment is viable (Table IV). The analysis for the cities of Athens and Thessaloniki also confirms that the worst-case scenario is S1 while scenario S4 is the best one as expected. However, both are unlikely to occur under typical conditions. Scenarios 2 and 3 are more realistic and provide very similar results for all cities despite the different origins of the hourly profile data. The financial savings range from around 2,300 € (1,500 €/kWp) to 3,700 € (1,000 €/kWp) which are, of course, lower than the corresponding values in Heraklion due to the increased cost of the PV system (2.43 kWp vs 2.80 kWp). It was necessary to increase the installed PV power since the solar irradiance in northern Greece is reduced by around 17% with respect to the solar irradiance in Heraklion, Crete while the household electrical loads remain the same among all cities. The internal rates of return for the typical scenarios S2 and S3 are around 12% for Heraklion, 10% for Athens and 9% for Thessaloniki adopting the unit cost of 1,250 €/kWp. Since the real cost of capital of 2.9% (7% nominal and 4% inflation) is less than these IRRs, the investment is viable with a net present value ranging from around 3,000 to 3,600 €. The breakeven times are nine (9) years for Heraklion and twelve (12) years for the other two cities for the same scenarios suggesting that a residential customer will benefit from the net Engineering, Technology & Applied Science Research Vol. 8, No. 4, 2018, 3168-3171 3171 www.etasr.com Mavromatakis et al.: Photovoltaic Systems and Net Metering in Greece metering scheme. Furthermore, depending on the size of the PV system the savings in CO2 emissions range from 53,000 to 62,000 kg [10]. Finally, the Levelized Cost of Energy (LCOE) is also calculated, although it is not directly related to the net metering measure. The high solar potential in Greece offers the opportunity to establish low LCOE values with respect to other European countries. Under the data given, it is calculated that the LCOE ranges from the minimum of 0.052 €/kWh in Heraklion and for a system cost of 1,000 €/kWp to a maximum of 0.081 €/kWh in Thessaloniki and for a system cost of 1,500 €/kWp. TABLE IV. ECONOMIC METRICS FOR A HOUSEHOLD IN HERAKLION Investment Area Heraklion Consumption profile S1 S2 S3 S4 Synchronization (%) 0.0% 37.8% 44.5% 71.3% LCOE (€) 0.0613 Breakeven (year) 13 9 9 8 IRR 8.4% 11.5% 11.6% 14.0% NPV (€) 2,164 3,593 3,625 4,776 Simple PayBack (year) 8.8 7.3 7.2 6.3 Pp=2.43 kWp, Cost 1,250 €/kWp These results agree with studies of the LCOE [11]. It is estimated that the final savings will increase since the formal clearance time interval for the energy consumed and produced is three years while the simulations are conducted on an annual timeline. Future work involves the simulation of solar data with 1-minute resolution to explore the effect upon the degree of synchronization since hourly data smooth out any shorter time variations. Furthermore, it is interesting to examine the case where the energy required to supply all kind of residential loads may be in the form of electrical energy excluding the use of wood, natural gas, LP gas or diesel oil for e.g. heating, cooking. Heat pumps with a high coefficient of performance can account for the heating loads (space heating, hot water). IV. CONCLUSIONS In this paper we present the results from the economic analysis of PV systems under the net metering scheme in different cities and climate zones in Greece (Heraklion, Athens and Thessaloniki). Several input parameters like the initial system cost, inflation, cost of capital and other parameters are used to examine different scenarios for potential investors of small residential PV systems. Residential customers may adopt the net metering scheme to reduce the cost of energy bills. The lifetime earnings (NPV) for a residence in Heraklion amounts to around 3,600 € while the breakeven time occurs at the ninth (9th) year of operation for a system cost of 1,250€/kWp. An IRR of 12% is calculated for the same site. The degree of synchronization is around 40% for typical household hourly profiles and it affects the final energy cost. Under the specific scenarios the breakeven point is 12 years, the IRR is around 9% and the NPV is a bit less than 3,000 € for a site in northern Greece. REFERENCES [1] Valentin Energy Software, PV*SOL Simulation Program for Photovoltaic Systems, Berlin, 2018 [2] METEONORM Global Meteorological Database for Engineers, Planners and Education, Version 4.0.95, Switzerland [3] Greek Statistical Authority, Yearly Report, 2013, http://www.statistics.gr [4] Center for Rewenable Energy Sources, http://www.cres.gr [5] PPC, Residential Tariffs, https://www.dei.gr/en [6] E. Drury, P. Denholm, R. Margolis, The impact of different economic performance metrics on the perceived value of solar photovoltaics, NREL Technical Report/TP-6A20-52197, 2011 [7] R. M. Smith, D. C. Jordan, S. R. Kurtz, Outdoor PV Module Degradation of Current-Voltage Parameters, NREL, World Renewable Energy Forum Denver, Colorado, May 13–17, 2012 [8] F. Vignola, J. Peterson, R. Kessler, F. Lin, B. Marion, A. Anderberg, F. Mavromatakis, PV module performance after 30 year without washing, 43rd Conference of the American Solar Energy Society (SOLAR 2014), San Francisco, California, July 06-10, 2014 [9] Interest Rates of Deposits and Loans, Bank of Greece, https://www.bankofgreece.gr/ [10] European CO2 Emission Data, https://www.eea.europa.eu/data-and- maps/indicators/overview-of-the-electricity-production-2/assessment [11] C. Kost, S. Shammugam, V. Juelch, H. T. Nguyen, T. Schegl, Fraunhofer Institute for Solar Energy Systems ISE, Levelized Cost of Electricity, Renewable Energy Technologies, 2018 .