9. Grid-Connected of Photovoltaic Systems

Transcription

1 9. Grid-Connected of Photovoltaic Systems H. Boileau Savoie University, FR Learning outcomes After reading this chapter, the user should possess knowledge of: A core description of PV systems connected to the electrical network The additional components on the DC part of PV systems Photovoltaic inverters The additional components on the AC part of PV systems System design and forecast of the electrical production Profitability of photovoltaic installations (TCE method from Bernard Chabot/ADEME) Assessing the compatibility between the photovoltaic field and the inverter Description of PV frames connected to the electrical network Photovoltaic installations connected to the electrical network represent the majority of the photovoltaic installations currently installed in the world (in 2015). Indeed, these photovoltaic installations are the simplest possible, therefore the costs are lower and all of the produced electricity is injected on the grid to be used. These advantages have these photovoltaic systems offering a lower cost per produced kwh, which explains the strong market penetration and commercial interest. In the absence of radiance, there is, of course no energy production, but the network compensates. On the other hand, it should be known that, in absence of voltage, the photovoltaic installation is bypassed for safety reasons, even if there is a strong luminous radiance.

2 Figure 1 - Simplified diagram of a photovoltaic installation Fig. 1 display a simplified diagram of a photovoltaic installation connected to the network with the main elements The photovoltaic field, when under solar irradiation, produces electrical energy in the form of DC current. This DC current is transformed by an inverter into AC to be injected in the grid, with the same amplitude and phase of the AC voltage of the network (typically an amplitude of 230 V and a frequency of 50 Hz in Europe). There may be protection devices installed between the PV installation and the inverter as well, if they are necessary or compulsory due to local regulations. Protective devices can also be inserted between the inverter and the network (inevitably necessary but not imposed by the safety standards). Finally, there usually is an energy meter for the invoicing of produced photovoltaic kwh. It can be useful to insert a feedback system to check how the installation runs because, for minor issues or losses, nothing indicates if the system runs correctly or not. A display can be used for the continuous monitoring of useful information, such as the instantaneous power and the cumulated energy produced.

3 The photovoltaic field The photovoltaic field is the entirety of the photovoltaic modules of a PV installation. The entirety of these photovoltaic modules can be connected in various ways, to one or several inverters. The three principal possible configurations are shown in Figure 2. Figure 2: Principal possible configurations of PV fields. The explanation of Figure 2 follows: a) The entirety of the photovoltaic field is connected to one inverter, which is called the centralized inverter. This configuration is the least expensive but all of the modules must be of the same type, have the same angular position and direction, for the simple reason that the current and voltage produced by each chain of modules must be of the same value. If not, there is a production loss. Each chain obviously needs to have the same number of modules. The influence of a shade on one or more photovoltaic modules can be rather important on the electric production, because the modules of each chain are connected in series and it will change change of current and/or voltage in a chain of modules that is in parallel with the other chains. Other disadvantages include that a breakdown of the inverter causes the complete halt of the PV field, the breakdown of a single PV module is difficult to locate and the voltage of a chain of modules is often high, several hundreds of volts in DC current, which is dangerous for living organisms. b), The entirety of the photovoltaic field is divided into chains of PV modules, each one connected to an inverter. All of the PV modules of the same chain must be of the same type,

4 have the same angular position and direction. On the other hand, from one chain to another, the type and position of the modules can be different. For example, this could be a system on different roofs connected on a single main bus. The influence of shading is less important than in configuration a, as shaded modules will only affect the chain they are installed on. Among other advantages, a breakdown of an inverter causes the complete halt of a chain (but not of the whole PV field) and the breakdown of a single PV module is easier to locate. However, the voltage of a single chain of modules remain often high, several hundreds of volts in DC current, which is dangerous for living organisms. c), In this configuration, each PV module is connected to an inverter, said micro-inverter in this configuration. Here, all the modules can be of different type and positioned differently as well, as they are independent. The breakdown of an inverter causes the stop of only one module, therefore little production loss. The influence of shading is very limited, only to the modules concerned. The high cost is the primary disadvantage of this solution but offers the most advantages. Normative Aspects on the characteristics of PV modules and safety: For instance, we briefly mention the French norms. - NF 61215: skill of the design and approval of crystalline PV modules. - NF 61646: skill of the design and approval of PV modules in thin layer - NF 61730: Characterization of the performances: flash test, NOCT, coefficients, etc. - Mechanical Tests: charges, shock, spindly, etc. - Climatic Tests: hot-cold, UV, etc. - Electric Tests: dielectric, leakage current, etc. The respect of these standards guarantees the quality of the photovoltaic modules and their conditions of use as, for example, a voltage of insulation of volts specifies the maximum number of PV modules in a single chain.

5 The phenomenon of "hot spot'' To explain the phenomenon of hot spot, let us take an example with a photovoltaic module of 72 cells (characteristic CP = 250W c, V mpp = 36V dc, V oc = 45V dc, I mpp = STC). If a cell of this module is shaded (I cell = 0A) or this module is short-circuited (Figure 3): Figure 2: PV Module with one shaded cell. The shaded cell (I cell = 0A, thus it behaves as an open circuit) receives all the voltage in reverse. The opposite breakdown voltage of a cell PV is typically of 25 V dc (Zener voltage), but the operative voltage of the module is V mpp = 36 Vdc. Therefore, the cell is destroyed, because it receives a strong voltage and becomes conducting, resulting to destruction by overload. Let us consider a more realistic case of a photovoltaic field with three chains (or strings) of six modules in series, where one cell of one chain is shaded like before (Figure 4).

6 Figure 3 : Three chains with one shaded cell. The operating voltage of such a string is 36 V dc 6 modules = 216 V dc. Where the cell is shaded (I cell = 0A), this chain is an open circuit, with a no load voltage of 45V dc 6 modules = 270 V dc. The reverse voltage of the shaded cell is then 54 V dc. The Zener effect takes place in this photovoltaic cell and it becomes conductive. The power dissipated in the shaded cell PV is 25V 6A = 150W, but the photovoltaic cell is not designed to dissipate this levels of power (no more than few Watts). This power will be converted to thermal where the photovoltaic cell is the most resistive until carbonizing that point (from where the name of hot spot originates). Figure 5 displays a photograph of a PV panel destroyed by such an effect. Figure 4: Damaged PV Panel, hot spot. The solution to prevent the phenomenon of hot spot is to use diodes, known as bypass diodes, connected in reverse and per about 20 PV cells. In the example of a 12 Volts PV module with 36 cells, a bypass diode is connected per PV 18 cells, as shown in Figure 6:

7 Figure 5 : Operation of a bypass diode. For this 12 volts module, the shaded cell results to the lack of current generation by the cell. Without a bypass diode, the PV module produces a lower or no current, with the risk of a hot spot arising. With the bypass diode, the power generated by the group of 18 unshaded cells can go through the bypass diode but, in the case of our example, the output voltage of the module will be two times lower. However, the diode prevents from having a high voltage in reverse on the shaded cells. For PV modules of silicon crystalline technology, the bypass diodes are cabled in the connecting box (Figure 7). Figure 6 : Connection of a bypass diode. The number of bypass diodes depends on the number of cells in PV module. Generally, two bypass diodes for a module with 36 cells, three bypass diodes for a module with 60 cells (very common type of module) and four bypass diodes for a module with 72 cells. The influence of the shade on a standard 12 Volts module, according to the point of operation, is described by the two following figures:

8 Figure 7 : Current vs voltage for different shades on a cell Figure 8 : Power vs voltage for different shades on a cell If a PV cell on a module is shaded, the module will produce an electric current lower than the rated current. If the load requires a very weak current, the output voltage will be the nominal voltage, but if the current consumed increases higher than what the shaded cell can produce, the voltage will decrease by a factor of two because of the bypass diode. Figure 8 shows that if a cell is shaded, then two points of maximum capacity appear, variable according to the extent of the shading. This phenomenon will thus disturb the operation of the MPPT (Maximum Power Point Tracking) process of the inverters. On a PV module where a cell is defective (shades or similar defect), measurements of the noload voltage and short-circuit current are both good. To detect this kind of defect, it is necessary to generate a current/voltage graph.

9 Protection against reverse currents: Reverse currents are another phenomenon that can deteriorate PV modules besides the phenomenon of hot spot. The phenomenon occurs when the current of a string is reversed when several strings are connected in parallel. In this case, if one of the chains is shaded, the cells do not produce any current. The current produced by the parallel chains now may pass through the chain with the shaded modules, leading to its destruction. With respect to the norms, a module should withstand in reverse twice the current that it is able to produce under STC conditions. When there are more than three chains, it is imperative to put a protective component in series with each chain, like a diode, to avoid this reverse current. Figure 9: Wiring example of three strings of six PV modules with the bypass diodes at the edges of the PV modules and the serial diodes at the end of the strings to protect from reverse currents. The serial diode protects effectively but causes a loss in power (loss of 1 volt if connected in series). In France, UTE C guideline recommends the use of a fuse, which is causing a lower power loss but fuses are irreversibly damaged when actuated. It is important to accurately assess the value of this fuse. Often, the manufacturers of PV modules provide the recommended value of this serial fuse with the electric characteristics of the PV module.

10 Additional Components on the DC part of photovoltaic installations Cables and electric connectors The electric cables used in photovoltaic installations must meet specific criteria (standard UTE C ). Some of these criteria are: a double insulation (class II), resistance to 1000V, resistance to UV and resistance up to a temperature of 90 C. The cross section of these wires is standardized. For example, commercial wires may have a cross section of 1.5 mm², 2.5 mm², 4 mm², 6 mm², etc. The cross section of the wires is selected according to the intensity of the current and the length of wiring. Figure 10 : Cables commonly used in PV installations. The electric connectors are secure (standards UL1703, VDE126-3, etc.). This means that they offer protection against direct contact, they can be lockable (type MC4) or not (MC3) depending on their accessibility, with a good behaviour towards UV and bad weather (IP54). Figure 11: Examples of connectors used in PV installations. Each photovoltaic module has two connectors, male and female, which facilitates the connection in series of those. Additional cables are used to make inter-connections between the strings in parallel and between the PV fields and the inverters. TAKE HEED, because of the continuous solar irradiation during the day, the voltage at the boundaries of the PV field has a magnitude of several hundreds of volts. This voltage can be dangerous during the implementation or the maintenance of a PV installation and, in the case of a circuit breakdown, a maintained electric arc is being created, because there no passage through 0 volts like in the case of the AC voltage. For that reason, the use of circuit breakers specific to photovoltaic is often required (Guideline UTE C in France).

11 Figure 12 : Electric Arc in a PV circuit. Protection against lightning of PV installations The photovoltaic field is by default exposed to the sun and subject to all weather conditions, including direct and indirect lightning impacts. For that, safety standards recommend or impose the use of lightning protector with suitable installation of the ground cable, with its specifications depending on the area where the PV installation is installed (in France, guideline UTE C15 712). Figure 13:Example of a PV installation grounding (source: Diagram Dehn, online: To protect from direct lightning impacts, there is almost no other solution than a lightning protector. For indirect lightning impacts, there are various solutions to decrease the destruction risk of the PV installation components. For example, the wiring of the PV modules can be done in such a manner so as to decrease the surface of the loops (Figure15), reducing the electric field induced in the loop by the strong magnetic variation caused by the amperage of a flash striking the ground in the vicinity.

12 Figure 14: Wiring examples of four PV modules. DC Junction Box A PV field generally is composed of PV modules cabled in series between them, creating strings. This is done to reach a high enough voltage. When the desired voltage has been reached, multiple chains can be connected in parallel (as long as their voltage remains the same), preserving the voltage and increasing the amperage (basic principle of electricity). Dimensioning requires adaptation through a thoughtful arrangement of the PV modules in series and parallel to surface available (on a roof, for example), but especially on the voltage and current intensity specifications of the inverter. Between the PV field and the inverter(s), a junction box is used to connect the chains of modules in parallel between them. The junction box also includes the protection components, such as the lightning protectors, the fuses, DC switches, etc. Figure 15:Example of a junction box with four chains in parallel, two lightning protectors and a DC switch.

13 Photovoltaic Inverters DC to AC inverters used by PV field convert electrical energy from the DC that the PV field generates to AC, compatible in terms of voltage and frequency with the network. The electric symbol of PV inverter is: Various types of inverters: Figure 16 : Inverters for PV fields (Source: SMA) Inverters for PV fields (Figure 17) are typically used for large PV installations on ground or on mounts, rated for several hundreds or thousands of kwc. Their AC output usually is threephase and the input DC voltage is up to a few hundred volts. Figure 17 : Inverters for PV installations (Source: SMA) Inverters for PV installations (Figure 18) are used in small or medium size projects. They are rated from some kwc up to a few hundreds of kwc. The inverter may be connected to one or several strings of PV modules, depending on the model and the size of the installation.

14 Their input DC voltage usually is rated for a few hundreds of volts, while be single or three phase. the AC output may Figure 18: Micro Inverter. The micro inverter (Figure 19) is connected to a single or up to a few PV modules. These inverters are not rated for more than a few hundred watts, while their input voltage usually is not higher than a few tens of volts. The AC output is single phase. Because of the low DC voltage input, these inverters are interesting because of their safety and can be used for small, medium or even large photovoltaic fields. All types of inverters have in common the seeking of the maximum capacity operation point (MPPT Maximum Power Point Tracking), a conversion efficiency from DC to AC of about 95% and automatic disconnection if no AC network voltage is detected (Standard VDE ) to avoid the electrocution of workers in cases of network maintenance. Principle of operation of a PV inverters: The purpose of the PV inverter is to convert the electric output generated by the PV field to an AC output compatible with the network. To that purpose, the MPPT system seeks the point of operation where the power is maximum from all of the possible points of operation (DC voltage) at exit of the PV field. Then a second electronic system, the inverter, converts the DC voltage into AC, compatible in terms of magnitude and phase with that of the network. The inverter's efficiency depends on how well it can match the voltage, frequency and phase of the network, as the portion of power injected to the network to the produced power from the PV field determines the electric losses in the inverter. The electronics of the PV inverter must also take into account the safety standards, such as the disconnection from the network in the event of temporary absence of AC voltage, or avoid the insertion of harmonics to the network that can disturb the operation of sensitive electrical appliances. Part of the performance feedback from a PV installation can also be assured by the inverter, by recording or forwarding to a server its operating information. Electric Characteristics of a photovoltaic inverter: The photovoltaic inverters have electric characteristics to be consulted for their correct operation (Figure 20). The main features are: On the DC (input) side:

15 Maximum capacity input: Maximum P in in Watts Maximum voltage input: Maximum V in in Volts Range of voltage of operation MMPT input: from V mppt min to V mppt max Maximum intensity input: Maximum I in in Ampere On the AC (output) side: Maximum output power: P out in Watts Typical voltage and the operating output range: V ac typ, from V ac min to V ac max Maximum output intensity: I ac max Conversion yield at the nominal output Figure 19 : Photovoltaic example of inverter and its characteristic (Source: SMA) European standardized output of a PV inverter: The PV inverter consumes a small part of the electric energy generated by the PV field (or from the network at night), inducing losses. The output of a PV inverter is generally defined at 100% of its nominal output. However, as solar irradiance during the day varies greatly, the electric production of the PV field will vary accordingly and, ultimately, the power point of operation of the PV inverter will also vary from zero (at night) to a value close to its nominal output under the best conditions of solar irradiance (if correctly dimensioned). To calculate the electric production of a PV installation in a more realistic way, a European average output was defined according to various points of operation with a coefficient for each one of these points of operation. This European output was defined according to the formula found in Figure 21.

16 Figure 20 : European standardized output of a PV inverter. Safety and standard for the photovoltaic inverters: The PV inverters must conform with various standards, the most important of which certainly is the standard VDE0126 that obliges the decoupling of the PV inverter when there is no network voltage. Indeed, if a worker locally disconnects part of the network for interventions/repairs, it is essential that the PV inverter stops the injection of electricity to avoid electrocuting the maintenance workers. Figure 21 : The conditions imposed by standard VDE0126 for France. These conditions can slightly vary from one country to another.

17 Additional Components on the AC side of the photovoltaic installations Network side box with the components of protection: (C Standard for France, consumer side) Figure 22: Network side box with the components of protection. In this box, we find components such as differential circuit breakers (usually rated at 30 ma) lightning arrest switches, disconnecting switches (Figure 24). Figure 23 : Differential circuit breaker (30 ma) lightning arrest switch, disconnecting switch. It is important to affix a descriptive label indicating the connection of a photovoltaic installation inside the electric box, network side, to indicate a potential danger to a technician (Figure 25). Figure 24 : Descriptive label indicating the connection of a photovoltaic installation

18 Electric meters and circuit breaker network side: (Standard C in France, distributer side of the network) The production of photovoltaic electricity must be routed through an electric meter to allow an invoicing of the electrical energy injected in the network, just like the electricity consumption by an individual. These metering elements, along with protection devices like the differential circuit breaker, are generally installed within the property but belong to the electrical energy distribution company. Figure 25 : Example of an electric meter and differential circuit breaker The electric meter allows the counting of the kwh produced by the photovoltaic installations that were injected into the network, as shown in Figure 27: Figure 26 : Dual electric meter connection. Usually, a second electric meter is put in series, but oppositely connected, to meter electric kwh that may be potentially consumed from the network (by the inverters during night time, for example). From the aforementioned setup, a consumer of electrical energy can connect and consume electrical energy locally. In this case, part of the photovoltaic

19 production will be self-consumed, while the surplus will be injected to the network (Figure 28). Figure 27 :Dual electric meter connection (self-consuming setup). It should be noted that to know the quantity of photovoltaic energy self-consumed, it is necessary to add another electric meter in serial to the photovoltaic installation. This meter is not essential for the distributer of electrical energy, it only allows the user to know the electrical energy consumption that the photovoltaic installation provides. In the case where the producer of photovoltaic electricity wishes to inject all the production in the electrical network (in the case of an advantageous tariff agreement), it is necessary to have a second electric network connection for the consumption of electricity from the electrical network (Figure 29). Figure 29: nework configuration for injection Dimensioning or calculation of the photovoltaic yield Depending on the case and purpose of the PV installation, the designer can either assess the electric production of the PV system or to size the installation according to certain power

20 requirements. Very often, especially with PV installations that will be connected to the electrical grid, the peak power P c of the installation is limited by external conditions, such as, for example, the installation surface available. In this case, the peak power P c of the PV installation is known and the annual electricity production is what needs to be assessed. To that end, it is essential to possess data on the annual average irradiation in the vicinity of the PV installations, and to calculate the optimal tilt and orientation of the PV panels. The typical data required for the sizing of PV installations is the annual global irradiation (I GPan ) on the horizontal plane. Software, such as Meteonorm, can determine this value for any site in the world with fair precision, for any given installation tilt and orientation. There are also free data resources, such as PVGIS ( which calculate and offer such data for almost any position in Europe, Africa and Asia by using climatic maps (Figure 30). Figure 30 :PVGIS output example for the site of the INES in Le Bourget du Lac in France The value for the average overall annual irradiation on a plane with a tilt of 30 towards the south is 3.98 kwh/m² per day (BDD classic PVGIS), which is a global irradiation received by the modules equal to: I GPan = = 1452 kwh/m² per annum. The third important data figure is the performance ratio (PR). The PR represents the entirety of the electric losses, which include the losses from the Joule effect in the wiring (~1%), the

21 losses of the PV inverter (~5%) and, most importantly, the losses from the rising temperature of the PV modules under solar radiance (~10% to ~15%). Different types of installations affect the airing of the PV modules and, in extent, their operating temperature under solar irradiation. For example, roof-mounted PV modules may have higher operating temperatures than ground installations on mounts. Depending on the location of the installation and the mounting type, it has been empirically estimated that the PR is close to the following values: Well ventilated Installation (e.g. PV installation on the ground) PR = 0.8 Fairly badly ventilated (e.g. super-installed on roof): PR = 0.75 Badly ventilated (e.g. fully integrated on roof): PR = 0.70 From these three figures, P c, I GPan and PR, the annual electricity production E a of a PV installation can be assessed: E a (in kwh per annum) = Pc (W c ) I GPan (in kwh/m² per annum) PR (dimensionless) This formula appears inhomogeneous but it should be noticed that the peak power P c expresses the electric output of a module under an irradiance of 1000W/m 2 and is not a SI unit. As I GPan expresses the annual solar irradiation in kwh/m 2, it can be assumed that it shows the number of hours that the modules will operate under a theoretical irradiance of 1000W/m 2. In other words, the PV installation will function this number of hours at its P c rating. Remarks: 1) The performance ratio PR represents the ratio of the electric output of the installation, or the electrical energy that is supplied to the grid (E a ) divided by the theoretical electrical energy produced by the PV installation (which is P c I GPan ) 2) The energy conversion output is calculated through the power peak P c, which depends on the surface and the output of the PV module under STC conditions (STC for Standard Test Condition, i.e. a radiance of 1000W/m², a temperature of 25 C and a solar spectrum AM1.5) 3) The above formula is usable only if the electric power output of the PV installation is a linear function of the irradiance, which is not completely true in real-world conditions, but it is sufficient for a quick assessment. To obtain better precision on the produced PV energy, it is necessary to use a simulation software, such as PVsyst or PVSOL that will take into account this nonlinearity. Similarly, it is possible to modify this formula for other time intervals, like a month or a day (in energy): E M (in kwh per month) = P c (W c ) I GPm (in kwh/m ² per month) PR (dimensionless)

22 E j (in kwh per day) = P c (W c ) I GPd (in kwh/m ² per day) PR (dimensionless) Likewise, it is possible to write this formula for one moment t, therefore the operating power P e from the radiance IGP (which is in W/m ²): P e (in kw) = P c (W c ) I GP (in kw/m ²) PR (dimensionless) Or P e (in W) = P c (W c ) I GP (in W/m ²) PR (dimensionless) Example of dimensioning assessment Previously, for the site of INES in Le Bourget du Lac, we calculated the global irradiation on a 30 inclined plane towards south. It is I GPan = 1452 kwh/m² per annum (equivalent to hours with a solar irradiation of 1 kw/m²). For a photovoltaic installation with a power peak of P c = 3 kw c (surface of 30 m² and 10% energy conversion losses), with the installation integrated on a roof and the performance ratio considered equal to 0.7. The annual photovoltaic production E a estimated is: Remark: E a = = kwh 1) The average consumption in specific electricity (without neither heating nor warm water) of a household in France is approximately 3000 kwh per annum, roughly equivalent to the electric production of a photovoltaic installation of 3 kw c. 2) The load factor (number of operating hours at nominal output of the installation, therefore the power peak) is 3049 kwh/3 kw = 1016 hours. This load factor depends on many parameters, the principal one being the solar irradiation. It varies from 800 hours in northern Europe to 1500 hours in southern Europe. The load factor is lower compared to that of wind, approximately 2000 hours, and that of the nuclear power plants, approximately 7000 hours. The better the load factor is, the closer the production of electricity is to the peak output installed. The load factors mentioned above for PV installations are not favourable for their economic profitability, which indicates that the initial investment per power unit will have to be low to remain competitive. Compatibility study between the photovoltaic field and the inverter The dimensioning of a photovoltaic installation usually begins by setting a certain photovoltaic modules number on the supporting mounts (building roof or site on the ground or structure) with a certain tilt and orientation. This process will determine the number of photovoltaic modules to be used, according to the available area, but also reveal the various possible wiring configurations of the modules (number of modules per string and number of strings). The number of photovoltaic modules per string and the number of strings condition the electric output, but also the voltage and the current at exit of the photovoltaic

23 installation. External parameters, such as the solar irradiance and the ambient temperature, should also be taken into account. The inverter connected to the photovoltaic field has operating ranges for input and output voltage, current and power, all of which will have to be compatible with the electric production of the photovoltaic field and the specifications of the load/grid. Compatibility in Power: Due to the fact that solar irradiance under the European latitudes goes up to approximately 1000 W/m² and that the performance ratio generally is about 0.8, the inverter's power is usually selected to fall between 80% and 100% of the photovoltaic field's power peak. To ideally estimate the power of the inverter correctly, the power bar chart at exit of the photovoltaic field is required. This bar chart can be simulated by specialized software, like PVsyst. Let us examine two examples: Example 1. Figure 31 shows the power bar chart of a 3.18 kw c PV field, facing south on a 30 tilt, in Geneva (Swiss). Figure 31 Power bar chart of a 3.18 kwc PV field, facing south on a 30 tilt, in Geneva (Swiss). If the dimensioning of the inverter is equal to 80% of the field's power peak, which is 2.55 kw, the bar chart shows that the plant is under-dimensioned because the power of the field between 2.55 kw and 3 kw is not fully used. It would be more judicious to choose an inverter of 3 kw or 3.2 kw (close to the 100% of power peak of this field).

24 Example 2. Figure 32 shows the power bar chart of a 3.18 kw c PV field, facing south on a 90 tilt, in Geneva (Swiss). Figure 32 : Power bar chart of a 3.18 kwc PV field, facing south on a 90 tilt, in Geneva (Swiss). If the dimensioning of the inverter is equal to 100% of the power peak of the field, that is to say 3.2 kw, it can be seen in Figure 32 that the plant is over-dimensioned because the field does not deliver any power between 2.5 kw and 3.2 kw. It would have been more judicious to choose an inverter of 2.4 to 2.5 kw that is capable of taking advantage of the field's full output. Thus, the ideal is to know the power bar chart, but that is not always possible. Generally, an inverted equal to 80% of the field's power peak is used when the conditions of tilt and orientation are unfavourable (vertical, eastern or western orientation), the average ambient temperatures are rather high and there are conditions of poor solar irradiation, such as in downtown area (pollution) or at the coastline. Similarly, inverters rated at 100% of the field's power peak are selected when the conditions of tilt and orientation are favourable, the average ambient temperatures rather low and there are good solar irradiation conditions, such as in the countryside. It will be perhaps even necessary to over dimension an inverter in locations where the sky is very clear, such as high in the mountains (lower atmosphere density, less pollution, low average temperature and an high albedo). Compatibility in current intensity: The solar irradiation from a clear sky is about 800 to 900 W/m², but under certain conditions with a strong direct radiance and an overcast sky of white clouds (important diffusion), the

25 solar radiance can reach 1300 W/m² for a few minutes some times per annum. Although that is not detrimental for the inverter if it features over current protection, it is recommended to have a safety margin on the acceptable maximum current intensity of the inverter (attention to the fuses of protection that also have to be over dimensioned to avoid having to change them too often. Usually, the ideal is to have a 30% margin (according to the country's legislative recommendations) compared to the STC operation conditions of the photovoltaic modules. Compatibility in no-load voltage: Without any grid voltage, the inverter is disconnected and therefore the current intensity is zero. However, the inverter remains connected to the PV field, where now the voltage is the highest possible. The worst case scenario for the maximum voltage at the input of the inverter needs to take into account that the solar irradiance might reach up to 1300 W/m². This voltage must remain lower than the maximum acceptable voltage by the PV inverter, or there is a risk of damaging it. This is one of the leading causes of inverters breakdowns, especially when the high no-load voltage has not been checked in mountain installations). For that reason, it is recommended to have a safety margin on the maximum no-load voltage, ideally 15% compared to the photovoltaic modules STC conditions (according to the countries legislative recommendations). The photovoltaic modules usually have a maximum operating voltage, often about 1000 volts (check the characteristics of the photovoltaic modules). This value allows to calculate how many modules it is possible to put in a single serial string. Compatibility in operating voltage: Under operation, the inverter adjusts the point of operation at the boundaries of the photovoltaic field to seek the point where the power is maximum (MPPT), selecting the best possible pair of voltage and current intensity. The selected pair varies according to the solar radiance and the cell temperature of the photovoltaic modules. The current intensity under operation varies between zero and the maximum, whereas the operating voltage will vary between a U pvmin value (as soon as the irradiance reaches some tens of W/m²) and a maximum U pvmax value (for a strong irradiation). For good compatibility between the photovoltaic field and the inverter, these two voltage values must be within the V mppt-min and V mppt-max voltages of the photovoltaic inverter (see the electric characteristics of PV inverter). Software such as PVsyst or PVSOL make it possible to simulate all of the operation points of the voltage, current intensity (and, thus, power) over a year. The output can be hourly, using

26 an average weather file, allowing to check compatibility between the PV field and the photovoltaic inverter. For a summary check out, it is possible to take the following conditions: V mppt-min inverter < 80% of the typical STC voltage of the PV field V mppt-max inverter > 115% of the typical STC voltage of the PV field Explanation: 80% of typical STC voltage of the PV field represents the voltage of the PV field for a radiance of 100 W/m² (for a silicon crystalline PV field) and 130% of the typical STC voltage of the PV field represents the voltage of the PV field for a radiance of 1300 W/m² (for a silicon crystalline PV field). Dimensioning example of a small photovoltaic installation Let us assume that we have a roof upon which we want to install PV modules. Eight Photowatt PW modules (Figure 33) can fit on the roof. Figure 33 :Photowatt PW module characteristics. These eight photovoltaic modules are connected to a SMA Sunny Boy 1700 inverter (Figure 34).

27 Figure 34 : SMA Sunny Boy 1700 inverter characteristics By choosing a wiring of two strings in parallel with four modules in series per string attached to the inverter (Figure 35), let us see if this combination is compatible or not. Figure 35 :Two strings in parallel with four modules in series per string attached to an inverter. Let us check the various points: Power check: The photovoltaic installation includes eight Photowatt PW Wc modules, so the total peak power is 1880 W c. The inverter SMA Sunny Boy 1700 has a maximum power input of 1850 Watts. The ratio inverter/field power is equal to 0.98, is between 0.8 and 1 (adapted for a photovoltaic field in good conditions, typically tilted to 30 and directed towards the south). Thus, the selection is OK in terms of power. Current intensity check: The amperage output of the PV installation in the worst case scenario is equal to the typical intensity (STC conditions) multiplied by two (because there are two strings in parallel) and again multiplied by 1.3 (assuming a radiance of 1300 W/m²). This equals to 7.86 A =

28 20.43 A. The maximum current input of the inverter is 12.6 A, which is lower than the current produced by the photovoltaic field. Thus there is a problem with the current intensity. No-load voltage check: The no-load voltage of a module is 37.2 V (STC conditions). With a radiance of 1300W/m², this no-load voltage is 37.2 V 1.15 = V. Knowing that the maximum voltage of the circuit cannot exceed 1000 VDC, thus 1000 V/42.78 V = 23.37, so we can have a maximum kf 23 PV modules in series. In our case, we have only four modules in series. With four PV modules in series, the maximum voltage of the photovoltaic field is 37.2 V = V. This value is lower than the maximum input voltage of the inverter that is 400V. Thus, the maximum voltage no load voltage is OK. Typical voltage operation (MPPT point) check: The typical voltage (not MPPT) of a module is 29.9V (STC conditions). With four PV modules in series and a radiance varying from 100 W/m² to 1,300 W/m², the voltage under operation of the photovoltaic field varies from 29.9V = V to = V. As the operating range of the inverter varies between 147V with 320V, the output voltage of the photovoltaic field is too low for the inverter. Thus, there is a problem with the MPPT operation voltage. To conclude this example of wiring in parallel two strings of four modules each and connecting them to a specific SMA inverter, this combination is not compatible. The voltage is too low and the current too high. A solution would be to seek another inverter or to change the type of the photovoltaic modules, but an obvious solution would be to change the wiring of two strings of four modules into a single string of eight modules in series, increasing the voltage and reducing the current output of the PV field. Let us investigate whether this association is compatible. Power check: The same as before. Thus, the selection is OK in terms of power. Current intensity check: The amperage output of the PV installation in the worst case scenario is equal to the typical intensity (STC conditions and only one string) multiplied by 1.3 (Radiance of 1300 W/m²),

29 which is 7.86 A 1.3 = A. The input current of the inverter is 12.6 A, which is higher than the current produced by the photovoltaic field. Thus, the current intensity is OK. No-load voltage check: The no-load voltage of a module is 37.2 V (STC conditions). With a radiance of 1300W/m², this no-load voltage is 37.2 V 1.15 = V. Knowing that the maximum voltage of the circuit cannot exceed 1000V DC, thus 1000 V/42.78 V = 23.37, so we can have a maximum kf 23 PV modules in series. In our case, we have only eight modules in series. With eight PV modules in series, the maximum no load voltage of the photovoltaic field is 37.2V = V. This value is lower than the maximum input voltage of the inverter that is 400V. Thus, the maximum voltage no load voltage is OK. Typical voltage operation (MPPT point) check: The typical voltage (not MPPT) of a module is 29.9V (STC conditions). With eight PV modules in series and a radiance varying from 100 W/m² to 1300 W/m², the voltage under operation of the photovoltaic field varies from 29.9V = V to = V. As the operating range of the inverter varies between 147V with 320V, the output voltage of the photovoltaic field is compatible with the inverter. Thus, there is no problem with the MPPT operation voltage. Therefore, with the wiring of the eight modules Photowatt in series at the input of an SMA Sunny Boy 1700 inverter, this configuration is compatible. The most important point to check is not to exceed the maximum voltage at the input of the inverter because that can be destructive for the inverter. For the other cases, generally the risk is only to have a lower production than that expected. In this example, compared to the point of operation under STC conditions (radiance of 1000 W/m², spectrum AM1.5 and temperature of 25 C), the maximum intensity is taken with a factor of 1.3, the maximum voltage is taken with a factor of 1.15, and the minimum voltage with a factor of 0.8. These factors are relatively arbitrary but make it possible to simplify the study. For different studies, these variations are calculated by using the temperature coefficients of the photovoltaic module while varying the temperature (e.g. from 0 C to 70 C). This method gives results rather similar to the factors described before. On the other hand, an experienced engineer should be able to realistically adapt these factors for various conditions, such as a desert or the mountain landscape with extreme temperatures and radiances. For the compatibility check between the PV field and the inverter, inverter manufacturers often provide free software..

30 Figure 36 : PV field to inverter compatibility check software (Source: SMA Sunny Design software) Bibliography Weiss, Johnny. "Photovoltaics Design and Installation Manual." (2007): Balfour, John R., and Michael Shaw. Advanced photovoltaic system design. Jones & Bartlett Publishers, 2011.