Differential Analysis of the Angle of Incidence Response of Utility-Grade PV-Bruce King-Sandia National Laboratories

P R E S E N T E D B Y Sandia National Laboratories is a multimission laboratory managed and operated by National Technology day-to-day and site-to-site variability increase uncertainty in these differencesAOI (q) Introduction5 • In this presentation, we explore a differential method for visualizing and quantifying the differences in the reflective (AOI) properties of several modules and the potential impact on system power• Several commercial utility grade modules and a module with an experimental ARC are used to demonstrate the method• A key point, all testing was performed simultaneously, making direct comparisons possibleMFG Model PSEL ID First Solar FS4 (No ARC) 3262First Solar FS4 (non-prod ARC 1) 3261JA Solar JAM6(k)-72-4BB 345W 3268Yingli YL330P-35b 3267 AOI (q) Test Method Outdoor Angle of Incidence Characterization Method7 Equipment:• Azimuth-Elevation solar tracker capable of rotating the test plane to solar incident angles between 0° and 90°• Global Pyranometer in the test plane measuring diffuse POA irradiance (Ediff)• Pyrheliometer on a separate weather tracker measuring Direct Normal Irradiance (E DNI)• Current-Voltage (IV) sweep system• Module temperature measurement systemEnvironmental Conditions:• High Irradiance, low diffuse• Low variation in Irradiance during test• Low wind speed/changes in ambient temperature during test Bruce H. King, Clifford W. Hansen, Dan Riley, Charles D. Robinson, Larry Pratt, “Procedure to Determine Coefficients for the Sandia Array Performance Model (SAPM),” SAND2016-5284, Sandia National Laboratories, Albuquerque, NM, 2016. 0° 5° 10° 15° 20° 25° 30° 35°40° 44° 48° 52° 56° 60° 64° 67°70° 73° 76° 79° 82° 85° 87° 89°Typical Incident Angles Outdoor Angle of Incidence Characterization Method8 Procedure:• Initiate IV scans, 2 scans/minute typical• Hold module normal to the Sun for a minimum of 10 minutes. Ensure Short Circuit Current (Isc) is stable• Index tracker off sun • Dwell for several minutes at each AOI, collect 4-5 IV curves per condition. Analysis9 • Correct measured Isc for temperature and spectrum• Find reference Iscr at AOI = 0°• Find normalized I sc (Nisc)• Plot Nisc vs AOI to visualize function, f2(q) Test Results and Differential Analysis Results – Standard Reporting11 • All modules relatively flat (pure cosine response) out to ~ 55°• Minor apparent differences beyond 55°• Yingli and First Solar ARC1 appear to be similar and consistently outperform all other modules commercial modules• Performance assessment is typically visual and subjective (“better” or “worse”) New Approach to Quantifying Performance – Differential Analysis12 • Determine simple differential between test device and a reference• For this example, we use a plain glass module with no ARC• Reference and test device must be measured simultaneously to eliminate differing environmental conditions between tests• Resulting differential is independent of diffuse light and only dependent on DNI Differential Analysis - Examples13 • Examples for two commercial modules and a First Solar module with experimental ARC• Divergence from plain glass behavior can be seen as low as 30°• All modules showed a boost at higher AOI. Degree of boost appears to be correlated with peak q.• Differential for commercial modules went negative at high AOI.• Differential for one module showed a dip at intermediate AOI, ~35°. Application to PV System Performance Application to PV System Performance15 Goals: • Demonstrate applicability of differential analysis to simulate differences in system performance• Apply for multiple modules, multiple system configurationsSystems• Fixed Tilt 35°, Fixed Tilt 10°, single axis tracker (SAT)• Modules: JA Solar, Yingli, First Solar Non-Production ARC1Inputs• 2017 weather from on-site weather station at Sandia (Albuquerque, NM) • High temporal resolution, 1 minute samples• f2(q), Df2(q) for each module from simultaneous outdoor testing Modeling Steps16 • Determine AOI (q) for each time step, each system orientationpvl_getaoi.m, pvl_singleaxis.m• Calculate diffuse irradiance for each AOI (only used for % gain calculations)pvl_haydavies1980.m, pvl_extraradiation.m• Determine Df 2(q) for each time step, each system orientationlookup table with spline interpolation• Calculate Net Irradiance difference, Net Irradiance for plain glass module • Determine Daily Average Difference in Net Irradiance• Determine Daily % Difference in Irradiance, compared to plain glass (directly comparable to % power gain or loss) Difference in Net Irradiance17 • Results are normalized for length of day – reveal seasonal differences that are dependent on AOI only• Performance differences can clearly be seen, both between modules and systems• For 35° Tilt, gains for the two modules with the most pronounced differential are seasonally flat• For both 10° Tilt and Single Axis Tracker, gains for these same two modules show strong seasonality (better relative performance in Winter) % Difference, Relative to Plain Glass18 • Results are totaled for each day, includes seasonal differences due to length of day• Performance differences relative to the reference module (plain glass) can be quantified• For 35° Tilt, modules with seasonally flat gains in differential provide a higher % gain in summer due to longer day• For both 10° Tilt and Single Axis Tracker, seasonal gains are more pronounced.• For 10° Tilt, Winter gains up to 1% are observed. Annual Gain19 Module Orientation35° Tilt 10° Tilt SATJA Solar 0.02 0.04 0.03Yingli 0.33 0.44 0.16First Solar ARC1 0.33 0.46 0.10Annual % Gain in Effective Irradiance • Annual gains (approaching 0.5%) were highest for 10° Tilt orientation• Annual gains were modest for Single Axis Tracker• Module with lowest differential response showed negligible annual gains in any orientation Summary20 • Differential Analysis is an effective approach to visualize and quantify the effectiveness of ARCs at non-normal incidence angles• “Better” performing modules show minimal differential response at low incidence angles and strong peaks at higher angles• “Weaker” modules may show dips in response at lower angles. This may negate gains seen at higher angles. • Differential Analysis can be extended to demonstrate effectiveness of different ARCs in different deployment scenarios• Of the scenarios investigated, 10° Fixed Tilt benefitted the most from good ARCs and Single Axis Trackers benefitted the least. Reminder: Gains or losses in Incident Angle response due to an ARC are IN ADDITION TO gains in normal transmission 21 Thank You!bhking@sandia.gov