André Mermoud

The ambient temperature may indeed be lower on the water surface. However I don't know any study evaluating this effect. Now for the albedo, if you have rows (sheds) arrangement only the first row will "see" the albedo. For all other rows, the water surface is not visible. Therefore there is a shading factor of (n -1)/n on the albedo factor (n = number of sheds, illimited sheds hypothesis). By the way, however surprising it can appear, the albedo contribution of a lake (during the full year) has been measured as very low (much lower than the 0.2 usual coefficient).

The main simulation variables are explained in the help "Project design > Results". However this may not be completely up to date, we may have added variables which are not listed.

I can't explain how you can get a positive value (gain) for the overpower loss. If so, please send your whole project to support@pvsyst.com, using "Files > Export projects" in the main menu.

At first sight ... When cells are shaded in one only sub-string, the module current will be limited to half the normal current. As soon as cells are shaded in both sub-strings of one submodule, the submodule will behave in the normal way. NB: we don't intend to implement such a specific case in PVsyst, unles this becomes a general layout adopted by a signigicant number of manufacturers.

There is indeed an error in versions 4.46 to 6.52: when defining a not-south azimuth for the Single axis tracking object in the 3D scene, this is not taken into account in the simulation. This was corrected in the V 6.53.

0.041667 = 1/24. This constant means one hour in units of days.

The calculation is done for the hourly step. If the sun is apparent during 20 minutes, and below the horizon during 40 minutes (theoretical sun position), then the simulation will account for 1/3 of the beam component for this hour.

The calculation "According to module strings" is an upper limit. You can modify its weight by the factor "Fraction for electical effect". Please see the help "Project design > Shadings > Near Shadings: Main dialog > Partition in module strings". You get it by the orange questionmark button next to this option. You are strongly advised to use these buttons (or F1) when you have a doubt.

If you are able to do this simulation without defining a shading scene, this is indeed a bug (the error is not taken into account), we have to correct it.

Yes of course. The horizontal irradiance (usual Meteo data) is the normal input for the simulation. The POA values are computed from a Transposition model (Hay or Perez) for any plane orientation.

From version 6.40 to 6.42, the IAM profile was indeed computed with a cubic spline between point#2 and point#3, which may lead to an artefact curve depending on how the points are aligned (if Point#1 and Point#2 are 1). This could give a discrepancy up to 0.3% to 0.4% in some cases. Now all the points interpolations up to the first <1 are computed with a linear interpolation.

If you trust the manufacturer's parameters you can use them. The IAM measurements are very difficult. Please see our FAQ How to determine the IAM profile ?

For your configuration, you should define 4 strings in the width of the shed. With sheds arrangement (a full row of cells shaded), the by-pass diodes don't recovery any power and the full string production (for beam component) is lost. Please see our FAQ How to evaluate the effect of by-pass diodes in shaded arrays ?

This requires that you avail of data about the yearly variability of the meteo data for your site. You have a complete explanation about the P50-P90 concept in the Help "Project design > P50 - P90 evaluations"

The IAM (Incidence Angle Modifier) is the transmission deficit (up to the solar cell) due to the incidence angle. The transmission loss is a general phenomenon, due to the reflexion and transmission of the sun's ray at each material interface (air-glass, glass-EVA, EVA-cell), as well as some absorption in the glass. The IAM only concerns the angular dependency of this effect, i.e. it is normalized to the transmission at perpendicular incidence (0° incidence angle). PVsyst uses an IAM function, which describes the deficit of transmission as a function of the incidence angle. This function is applied either to the beam component, and to the diffuse, using an integral over all "seen" directions, supposing an isotropic distribution of the diffuse irradiance. The IAM function may be determined in different ways: - By a physical calculation using the laws of optics (Fresnel laws), - By measurements, either indoor (flash test) or outdoor (at sun), - By a rough approximation proposed by the ASHRAE (old default in PVsyst until 2017). Fresnel's laws The Fresnel's laws describe the transmission and reflexion of a light ray at the interface of materials characterized by their Index of Refraction n. This calculation should be performed at each interface: Air-Glass (the dominant effect), Glass-EVA and EVA-Cell (less reflexion as the refraction indices are close to each other), secondary reflexion in the glass and its transmission to air, etc. Globally, the next figure shows the Fresnel's laws for a single glass cover, as well as a glass with antireflective (AR) coating. This is the present default in PVsyst since 2017. On this figure, we also show the old ASHRAE parametrization. We see that this approximation is not satisfactory, it is here for historical reasons (chosen since the beginning of PVsyst in 1992). Fresnel's laws and Ashrae parametrization Measurements Now the measurements proposed by diverse Laboratories (manufacturers) are very difficult to interprete and understand. We have received many measurements, which are contradictory. Normalized measurement procedures are described in the IEC 61853-2 directive. This document specifies 2 methods, either indoor measurements (flash tests) or outdoor measurements (at sun). Among all measurements that we have received: - Some indoor measurements seem of very good quality, they usually give results which are very close to the physical Fresnel's model. The next figure gives an example. However we also receive sometimes measurements with much higher values, often erratic. We can't believe in these values, and we don't accept them for the database if they are too different than the fresnel's laws. IAM measurements, Indoor ans Outdoor - The outdoor measurements are much more erratic, and in most cases give much higher values than the Fresnel's laws. The methodology is not so "clean" as the indoor method, as the measurement is perturbated by the unavoidable diffuse and albedo parts. The reference beam component is indirectly evaluated by a pyrheliometer (beam) and a pyranometer (global). The diffuse is also suffering of the IAM effect, but this is not taken into account in the methodology. Moreover: - The angle determination is a source of uncertainty. The IEC 61863-2 requires an accuracy of +/- 1°. Now such a discrepancy involves an uncertainty on IAM of 1.4% at i = 70°, and over 5% at I = 80°. - Remember that the sun moves by 1° every 4 minutes, and that the apparent diameter of the sun is 0.53°. - There may also be an effect of the light polarization when the sun is low on the horizon: the Fresnel's laws show that the parallel and perpendicular components have very different angular behaviours (+/- 20% at 70°, with respect to the average). When analysing the rough measurements of different modules (and +/- incidences), we notice a dispersion of the individual measurements of several percents, indicating a basic uncertainty in each measurement. Such a systematic discrepancy between indoor and outdoor measurements lead to significant differences in the IAM loss calculations, which to our opinion is not justified. Therefore since January 2017 (version 6.53) we don't accept outdoor IAM measurements for new modules of the PVsyst database. Only the Indoor data are retained, as far as they are not too high. Effect on the simulation The IAM loss affects the 3 components of the irradiance, with almost half the contribution for the diffuse and albedo. - The Diffuse and albedo contributions depend mainly on the geometry (plane tilt and shadings). - The Beam contribution is related to the climate, namely the beam quantity along the year, and is computed for each simulation step. The next figure (issued from the PVsyst tool "Detailed Losses > IAM > Detailed study") shows the contribution of different profiles, for the meteo of Marseille and a plane tilt of 20°. - The Ashrae parametrization with bo = 0.05 corresponds about to the Fresnel's laws for single glass. - The Ashrae parametrization with bo = 0.04 is close to the Fresnel's laws with AR coating, which is very close to the indoor measurement, and is probably the "modern" standard. - The "User defined IAM profile" is the outdoor measurement mentioned above. We see that it overperforms the indoor measurement by more than 1%, and the PVsyst default by 1.8%. IAM losses for different profiles NB: Antireflective coatings are relatively fragile. Nothing ensures that they remain intact, especially in hard conditions (deserts, cleaning, etc). Therefore the real performance may derive from the nominal (outdoor measurements) to the single glass (PVsyst present default) along the life of the plant.

I agree that a model would be very useful. However this would require several input informations, namely in the meteo file, but not only. The effect of snow is not only the falling snow, but also the remaining snow on the PV modules (how much time will is stay? Partially or completely? Effect of the temperature, insolation, plane tilt, module's base geometry, etc. We will think about implementing a snow coverage ratio defined in hourly values, to be specified by the user. The difficulty will be to evaluate the electrical mismatch effect in case of partial covering.

Please carefully read the help "Project design > Shadings - General" and "Project design > Shadings > Horizon - Far shadings".

This limitation is for the parameter's definition safety. We think that a higher slope of the PNom variation as function of the inverter temperature doesn't make much sense.

The best way will be to measure the irradiance in the horizontal plane. Then the program will do the transposition to any plane. If you can register the diffuse irradiance, the transposition accuracy will still be much better !

This is a development that we should do rather shortly indeed. However it is not straightforward, this represents significant work for doing correctly. This will not be in the next main release.

In PVsyst you cannot specify different modules in a sub-array. Therefore you cannot define your systemwithout using Multi-MPPT. Now you can also redefine your inverter as a (dummy) Multi-MPPT device, and define 2 different sub-arrays. In this case you should use the button "Adjust" for correclty sharing the nominal powers of each MPPT input.

In PVsyst and the Northern hemisphere, the azimuth zero corresponds to the south. Therefore your 215° correspond to 35°. Please see our FAQ http://"How is defined the plane orientation ?"

It is extremely difficult to find an algorithm satisfying any situation and anybody. The tables are filled one after the other. If their size is changing, the number of modules on each table will vary and the result will be different for each table. What we could do would be to fill a selection of tables. We will think about that for a next version.

The module area is used essentially for the efficiency definition. PVsyst proposes 2 definitions for the area taken as reference for efficiency: - The rough module size, which gives the "usual" module efficiency, - The cell's area, which leads to an efficiency at the sensitive level, which characterizes the cell's technology (quality). The frame mentioned in the "Table" definition is an extra area representing eventual mechanical supports, which could produce shades on the next table. It is indeed null in many cases.

The Imp and Vmp values are used for establishing the one-diode model. Therefore they affect the PV production of the simulation. Now manufacturers use to take the positive sorting into account by altering the definition of Imp and Vmp. If you account for the positive sorting in the "Module Quality Loss" factor, you will account for this gain twice. Choosing one or the other option is your choice. PVsyst tries to stay coherent and relies to the PNom value for the "official" module efficiency.