MarcoMalagnino

Dear all, The exact forecast of the snow amount, in term of losses and periods, is quite impossible. There are some possibilities linked with meteo data which indicate also snow coverage, but as Mr. Mermoud has written, there are several different factor which influence the snow coverage effect: snow type, ambient temperature before, during and after snowing, module tilt angle, wind, humidity, irradiation, etc. I think that the evaluation of the expected losses due to snow coverage is responsibility of the user, as well as for normal soiling (also difficult to appreciate as the snow). For snow losses, probably could be a good approximation to define some days per month with the expected snow coverage of the month (a sort of unavailability of the system) and concentrate the expected monthly snow losses in the selected days of unavailability, without taking into account the partial covering effect. The hourly or daily behavior would be probably not realistic but the monthly values could be good enough to evaluate possible influences on the design of the pv-plant (narrow row distance, higher tilt angle, etc.). However I think that the snow losses, as well as the soiling losses, should be taken into account by PVsyst before the other losses (Horizon, Shading and IAM losses). Best regards

Dear Mermoud, I have asked to several different inverter manufacturers and all of them have answered that there will be an active power reduction due to a PF less than 1, only in case of overload (in a way, the PF<1 reduces only the overload limit of the inverter). This is due to the fact that the reactive power comes from the Grid. Best regards Marco

Dear Mermoud, according to your answer, the effect of a constant power factor on the PR of the system is only due to a possible rise of the clipping losses, thus having an effect only over the maximum nominal active power (in KW) delivered by the inverter. It means that during the time when the produced active power (AC side) is less then the Pnom of the inverter (adjusted according the given power factor) I will not get any energy losses. In a way, the AC active power at the inverter output will be like the product of the dc power (inverter input) per the inverter efficiency (if no clipping or voltage limitation occur) until the current apparent power is less than the maximum apparent power of the inverter, right? But in which way the inverter can produce an apparent power higher than the product of the dc power per the inverter efficiency? Where does the rest of the apparent power (the reactive power) come from? From the grid? Or from the inverter? Because as EPC company we have to give guarantee for the designed PR, I would like to be sure that I am not underestimating the effect due to the power factor. I am not really expert of what the inverter in the reality does, but my understanding was different from what PVsyst simulates: I thought that the inverter produces an apparent power (kVA) equal to the product of the DC input power per inverter efficiency; if the PF = 1, then the active power (kW) would be like to the apparent power and there would be no additional losses, but if the PF < 1, then the active power would be like the product of the apparent power per the cos(phi), and it would mean that PR would be lower (proportional to the PF). In my case for a project in USA we have to set the inverter with cos(phi) = 0,95 fixed. According to my design (DC/AC ratio), as effect of a fixed cos(phi) = 0,95 PVsyst simulates additional losses of about 0,3% but actually I am not sure if these losses should be 5%? Could you please clear this point, and if you have some references/articles about this theme, could you please be so kind to send them to me? Thanks in advance, Marco

Dear Mermoud, I have reviewed my simulation and I have noticed that, even if I have entered the time-shift correction (-29 minutes for this location), PVsyst has not changed it. After several attempts I was able to get the desired time-shift modification and now it seems that the irradiance levels on module plane are more plausible (now the max GlobInc is 1100 W/m²). Thanks for your time and disposability. Best regards Marco

Dear Mermoud, thanks for your quickly replay, but I am not sure that this is the only cause of that difference. I have simulated another location, this time in USA (CA). I have imported a TMY3 file from NREL (but using the Meteonorm format because the data comes from NSRDB Data viewer) with the hourly values of GlobHor, GlobDiff, Tamb and Wind. I am always using an horizontal tracking system and in this case the hourly values of the calculated GlobTilt is really higher than the respective hourly values of the entered GlobHor: e.g. At end of April the max GlobHor is about 745 W/m² and the respective GlobInc is 1235 W/m². Obviously the irradiance level plays an important role for such systems, above all for the clipping losses of the inverter. I will send you in a separate EMail the meteo file and the PVsyst Project. Could you please advise me which is and where is the mistake I make in generating the GlobInc values? Thanks in advance. Marco

Dear all, I have a question about the calculation of the irradiance on module plane for single axis horizontal trackers. Comparing the hourly values of the GlobHor and GlobInc I have noted that at the noon the GlobInc is slightly higher (in my case between 0,5% and 1,8%, avg 1%) than the respective GlobHor at the same hour. Actually I expected that at noon, because the tracker is in the horizontal position, the GlobHor and GlobInc have the same value. But it is not the case. Why there is this difference between them? Thanks in advance

Dear Mermoud, I have a question about the calculation of P90, P75 and P99: according the PVsyst Help, in oder to calculate P90, P95 and P99 the following coefficients will be applied to sigma: "NB: In the Gaussian distribution, P90 represents a shift of -1.28 sigma, P95 => -1.64 sigmas, and P99 => -2.35 sigmas." but in literature (for example Wikipedia https://fr.wikipedia.org/wiki/Loi_normale) I have found other coefficients: 68 % d'entre elles sont dans l'intervalle [{x} -sigma; {x}+sigma]; 95 % d'entre elles sont dans l'intervalle [{x} -2sigma; {x} + 2sigma]; 99,7 % d'entre elles sont dans l'intervalle [{x} - 3sigma;{x} + 3sigma]. According to the literature, I should use the following coefficient: P68 = -1 sigma, P90 = -1.645 sigma, P95 = -1.96 sigma, P99 = -2,58 sigma. Could you please clarify this inconsistency? Thanks in advance. BR Marco

Dear Mermoud, I have noticed that the ac cable losses, as well as the external transformer losses, depend in PVsyst on the Pdc of the pv generator defined in the simulation through the definition of the loss fraction coefficient. In order to calculate this losses you consider the Pac as follow: Pac(STC) = Pdc(STC) x european efficiency of the inverter. Now my questions: which is the Pac(STC) I have to use for the definition of these losses? For what I know, the design of the ac cable is done considering the maximum power from the inverter (or from the transformer, if it is a central inverter with transformer) and the ac voltage. Why don't you take into account the inverter limit and then the inverter maximum power as Pac(Stc) of the system? I know that it is just a definition problem but it can cause different losses according to the different interpretation of the Pac(STC). Thanks in advance. Best reegards, Marco

Dear Mermoud, I am trying to understand the modelling of the inverter effciency in PVsyst. I have some problems to undestand the way PVsyst uses in order to build the efficency curve starting from the max and Euro efficiency. In the Help "Grid inverters, Efficiency curve" you write: ------------------------------------------------------ Interpretation of the efficiency curve: - Up to the power threshold (often named "Starting operating at ..." in datasheets), the input power corresponds to internal consumption for the internal needs of the device, and doesn't produce power. - From this threshold, the device is supposed to produce an AC power proportionally to (Pin - PThresh). - This production is penalized by an ohmic loss of the internal components (transformer and transistors), which increases quadratically with power (like R · I²). When translated into efficiency, this gives a maximum efficiency usually around 50-60% of the nominal power. PVsyst model for automatic construction of the efficiency profile In most of the specifications, the manufacturers or databases give the so-called European Efficiency (or CEC efficiency in the US) which is intended to provide an average efficiency over yearly operating conditions. From these two values specified by the user - Maximum efficiency and European efficiency - PVsyst constructs a default profile, with the following hypothesis: - The AC production is proportional to the DC available energy, minus the Pthresh "internal consumption", - The efficiency is penalized by a resistive loss of the output circuitry (transistors and transformer), in a quadratic way as function of the power (R * I²), - This transfer curve has to match the specified maximum power value, at a DC power arbitrarily fixed at a value of 60% of Pnom, - The Pthresh "effective" value (=> slope of the curve) is adjusted in order to match the Euro (or CEC) average efficiency. The contribution of the Resistive loss is fixed according to a normalized resistance factor, proportionally to the difference between the given Max- and Euro- efficiency. The normalized resistance factor commands the losses at high powers (above the 60% point), with respect to the losses due to Pin_threshold. We choose a usual value of 3.0 (arbitrary units) as default. -------------------------------------------------- Could you please explain in an easier way this modelling method and why, if the threshold is 8000W (imposed by the manufacturer), do you calculate a Pthresh "effective" value? How do you calculate this "effective" value? Thanks in advance