Modeling on Unconventional PVs and Their Standardization – Curved Surface,-Kenji Araki-Toyota Technological Institute

Modeling on Unconventional PVs and Their Standardization – Curved Surface, Vehicle-Integration, Multi-Junction Cells, and Static ConcentratorsKenji Araki1 , Yasuyuki Ota2 , Kan-Hua Lee1 , Kensuke Nishioka2 , and Masafumi Yamaguchi11 Toyota Technological Institute, cpvkenjiaraki@toyota-ti.ac.jp, Nagoya, Japan2 University of Miyazaki, Miyazaki, Japan PV needs to grow diversityVIPV(Car-roof PV)BIPVTracking PVLow cost III-V & tandem PV modeling needs diversityVIPV(Car-roof PV)BIPVTracking PVLow cost III-V & tandem Spectrum Prediction 3 -D Curved Surface Unwanted Orientation Partial shading Loss✔✔ ✔ ✔✔ ✔✔✔ ✔✔✔✔ Can we rely on the golden procedure? (irradiance with correction of temperature and IAM) Yes, for the conventional installation, but … Yes, for the conventional installation, but …Possibly, NO for diversified new PV installation Yes, for the conventional installation, but …Possibly, NO for diversified new PV installation VIPV and BIPV for ZEB require high energy yield under the unwanted operation and installation conditions (orientation, partial shading, colored PV and etc.), thus tandem cells will be preferred when the cost and other technological issues are solved.Key parameters are dependent each other. For example, different IAM, different orientation means different mixture of the direct and diffused sunlight, and thus varies spectrum. This impact was unexpectedly huge in our observation (not in this presentation).R&D of the new integrated model that can applied to every type of installation, including tandem cells. Basic model construction Option with unexpected and unwanted disturber (like CPV)Modulation factor by Monte Carlo calculation Monte Carlo approach Achievement High performance but relatively high temperature Valley depth varies by the type of concentrator cellsIsc drops sharply (not Voc)Quick decline, why?Consideration on spectrum (water and turbidity-dependent) and IAM successfully, explained strange behavior of III-V multi-junction cells PV module using 3 J tandem cellArea: 9 8 3 cm2STC: 3 0 .7 %2 0 1 6 /0 8 – up to now Selected topics around our integrated model challenging new type of installations and new cells1 . Challenge to the curved surface2 . Challenge to the mismatching loss by partial shading 1 ,0 0 0 cuts to the car-roof PVCar-roof PV may be killed by 1 ,0 0 0 small cuts, not by a sharp cut.Curved correction factor (local cosine loss)Higher incident angleFrequently shadedCar side is typically less than 5 0 % of the car-roofMismatching loss by uneven illuminationPartial shading by frequent drop of the small shadows (including mismatching loss associated by the partial shading)It is essential to acknowledge the impact of each damage. Give him 1 ,1 0 0 life points. But, not 3 ,0 0 0 life points… Issues of the curved moduleLocal cosine lossSignificance of the local loss varies by the incident anglesSelf-shading lossIf the curve is a simple convex body, it may be the extension of the cosine loss with Front OnlyDefinition of apertureAperture mask or Collimated solar simulatorAperture mask: Side reflection? Height of the window? Specification of the mask itself?Collimated solar simulator: Practical? Required collimation angle?IAM (Incident Angle Modifier) Varies by the curve shapeVaries by orientationsMore weight on the response from the higher incident angles. Definition of areas Curve profile is made by connection of power functions symmetry to y-direction but asymmetry (front-back) to x-direction.Irradiance onto both projected and curve-surface were calculated by ray-tracing.This explains how effective the SK’s construction of the aperture to the 3 -D curved object. The reflector inside the aperture volume help entire rays that passes aperture illuminate onto the 3 -D curved surface. A proposal of the curve correction factorP AIf A: Flattened area of the panel (Not projected area)I: Irradiance of the car-roof (Not GHI)f: Curve-correction factorη: EfficiencyWhat is f?Unique value depending on the 3 -D curve shape of the panel (affected by AOI of the panel and the standard car-roof irradiance model).Calculated by numerical calculation (calculation algorithm should be open, transparent, and repeatable within some acceptable numerical errors, ie. use of Monte Carlo method). Applicable to the 3 -D CAD interface (Most of the car-roof 3 -D shapes were not simple polynomials but segmented smooth functions like NURBS). Parameters of the curves were given by random numbers. Curve correction factor was calculated by the ray-tracing simulation using 1 E8 rays distributing by the direction weighted by the annual distribution of the incident angle on the horizontal plane. Quantitative approach Full shading Partial shadingCells in a module may be categorized into 3 types.1 . In the sun2 . Full shaded3 . Partial shadedShaded area: (Diffused sunlight) + (Reflected sunlight) Cells in the sun: All the sunlight including direct sunlight Modeling partial shading and its mismatching loss using random numbersParameter Distribution Range RemarksDate (Day number) U n i f o r m distribution 0 - 3 6 4 Repeat throwing a dice until the horizontal global sunlight given by the database is more than 1 Wh/m2 to avoid inclusion of the trial in the night time. The bissextile day is removed.Time U n i f o r m distribution 0 – 2 3 (1 hr)Number of cells partially shaded U n i f o r m distribution 0 – (Cells in the strings) Random numbers are given to each string. The number of the partial and full shaded cells must be less than the number in the strings.Number of cells fully shaded U n i f o r m distribution 0 – (Cells in the string)Shading ratio of each partially-shaded cell U n i f o r m distribution 0 - 1 Ratio of the partial shading (0 to 1 ) Car orientation U n i f o r m distribution 0 °- 3 6 0 ° Isc of each cell N o r m a l distribution Voc of each cell N o r m a l distribution Diode ideality of each cell N o r m a l distribution Representing the shape of I-V curve Typical result (Car-roof PV) The ratio of the power output of the partially shaded module is less than the ratio of the area in the sun.The reduction ratio corresponds to the mismatching loss.The ratio of this mismatching loss decreases with increase of the number of the strings.For Si cell in the normal size, it is difficult to construct with multiple strings considering the area of the individual cells per the area of the car-roof.Cutting cells into triangle will be effective to increase the number of strings, thus reduction of the mismatching loss by the partial shading.Triangle cut is also effective to coverage of the curved surface (less than 3 0 0 0 mm of the radius of curvature). The loss factor is a function of the number of strings, curvature of the car-roof and latitude.Increase the number of strings, correspondingly reduction of the cell size. Typical result (Urban CPV)Urban CPV is a good application to the EV charging station by RE (high efficiency and white module face). The common problem is the frequency of the partial shading, especially shading by the power line.CPV only utilize the DNI and the impact by the partial shading is large. However, high sensitivity implies it can be improved relatively easily. Number of strings X2 Invitation to the international web meeting for standardization of the car-roof PVBackground: IEC TC8 2 seriously considers the standardization of the car-roof PV. Before official organization is consolidated, I will call for the preliminary web meeting inviting academic, PV and car industries.Purpose: Establishment of the common language between car and PV industiries for standardization of the car-roof PV technology, like the standardization of BIPV (PV and architecture industries).Questions: Rating of the curved PV, Definition of aperture, Definition of the acceptance angle, Qualification and safety, Effective irradiation on the car-roof, Irradiation on the car-surface (Local coordinates), and etc., If you are interested in and eager to contribute Send me the mail for the invitation letter. Currently, 4 9 people from 1 4 countries are registered, including 8 from China, 1 1 from USA, and 3 from Germany.cpvkenjiaraki@toyota-ti.ac.jp • New-type installation may needs advanced model (preferably, unifying model, treating different type of installation fairly and with the same scale)• Spectrum prediction, 3 -D curved surface, unwanted orientation, partial shading loss, and etc.,• These key parameters are dependent each other, small correction sometimes does not work.• The unifying model was proposed.• Some factors are difficult to predict, then Monte Carlo approach is effective.• The new unifying model successfully explained strange behavior of these new PVs.• 3 -D curved issue may be convenient to treat by a curve correction factor.• Partial shading can be treated by the Monte Carlo method. • Invitation of the web meeting group. AcknowledgmentsPart of this work was supported by the Incorporated Administrative Agency New Energy and Industrial Technology Development Organization (NEDO) under the Ministry of Economy, Trade and Industry (METI). 24 取 法 乎 上 得 乎 其 中 取 法 乎 中 得 乎 其 下