Modeling PV system performance involves a number of steps that are shown in the figure below. A more detailed explanation of each step is presented below the figure.
- Irradiance and Weather – This step involves choosing a source for defining the site’s expected irradiance and weather conditions. Typical sources include: typical meteorological years (TMY), satellite-derived data, on-site ground measurements, etc. Modelers can choose numerous possible weather input approaches for their performance modeling studies. Uncertainty in irradiance data is one of the most significant factors affecting uncertainty in PV system performance.
- Incidence Irradiance – This step translates irradiance measured at standard orientations (horizontal, plane of array, and normal to the sun) to beam and diffuse components on the plane of the array. Numerous algorithms are available to perform these translations.
- Shading and Soiling – If the array is partially shaded or the modules are soiled, then incident irradiance available for conversion to electrical energy is reduced. Various algorithms exist to calculate shading and its effect on the system. Fewer methods exist to predict the amount of soiling on the array with time. Soiling varies greatly by region and site. Usually this step is treated with a constant derate factor that may vary by month.
- Cell Temperature – Many factors influence PV cell temperature: module materials and construction, mounting and racking configurations, the incident irradiance (modified by shading and soiling), the wind speed at the array level, and ambient temperature, among other variables. Many methods that have been proposed to estimate cell temperature from these variables. Most models assume that cell temperature is constant over the modules and array. Field data indicates that temperature gradients can occur across the array.
- Module Output – This step involves predicting the module’s IV curve under the previously described conditions: irradiance (including spectral content) and cell temperature. Various model forms have been developed to estimate the IV curve (single diode, semi-empirical, etc.).
- DC and Mismatch Losses –This step involves estimating the losses on the DC circuit(s) due to resistive wire losses and mismatch between series-connected modules and parallel strings. Few performance modeling applications include the option to rigorously represent these effects. Most simply include a scalar derate factor to represent these losses.
- DC to DC Max Power Point Tracking – Most, if not all, modeling applications assume that the DC voltage on the array can be held at the array’s maximum power point at all times. In fact, differences exist between max power point tracking (MPPT) algorithms, and sometimes PV systems may operate off of the maximum power point. Few models are currently set up to represent performance in such a scenario. Sometimes a derate factor is set to represent MPPT efficiency.
- DC to AC Conversion – This step accounts for the inverter’s conversion efficiency. This efficiency can vary with environmental parameters, such as temperature, as well as electrical conditions, such as the DC power level.
- AC Losses – Finally, once the power has been converted to AC, it must be transmitted to an interconnection point. This step accounts for any losses along this transmission path (wire losses, transformer losses, etc.).