cams ga solar resource and forecasting needs by kazantzidis
TRANSCRIPT
The DNICast team
A.Kazantzidis
Laboratory of Atmospheric Physics, University of Patras, Greece
Solar resource and forecasting needs
Presentation outlook
• DNICast at glance
• Needs for solar resource and forecasting: Aerosol optical depth, Cloud properties, Enhancements due to cloudiness
• A tip for future use of ultraviolet radiation products (in liaison with health scientific community)
DNICast: Direct Normal Irradiance Nowcasting methods for optimized operation of concentrating solar technologies
DNICast at a glance
DNICast at a glance
DNICast at a glance: the testbed
The basic steps for solar resource and
forecasting
Water vapor?
Aerosol optical properties: why needed?Concentrated Solar Technologies are dependent from DNI. Aerosols variability can be the major source ofDNI variability.
Differences (%) between the mean DNI for each year (2000-2012) and the average for the 13-year period. Only data for May-September are considered
Nikitidou et al., Renewable Energy, 2014
Difference (%) between the daily DNI and the corresponding monthly mean, for 5 areas in Europe
Aerosol optical properties: why needed?
What aerosol properties can we get?
Goodness of Fit Statistics (AERONET vs MACC)
AOD values from AERONET and MACC are compared in terms ofMBE, RMSE and CC (550nm)
|MBE|<20%, RMSE<30%
Trend Analysis
Highest Trend
Sig. Trend for AERONETNon Sig. Trend for MACC
Estimation of impact on clear-sky DNI
Relative MBE (%)
Estimation of impact on clear-sky DNI
Relative MBE (%) Including corrections on AODs
23/7/2014 24/7/2014
AOD(500) = 0.09 AOD(500) = 0.43
The DNICast approach to estimate AOD
440nm 500nm 675nm
Mean difs -0.009 0 -0.01
Median difs -0.004 0 -0.01
Std 0.03 0.02 0.02
The DNICast approach to estimate AOD
Cloud properties for solar forecasting
Previous study has determined homogenous spatial clusters ofsimilar CCI variability using cluster analysis and cluster validityassessment methodologies (Zagouras et al., Solar Energy,2013,2014).
CCI forecast: Seasonal Analysis (1)
Spatial distribution of the seasonal average mean MSE error (per pixel) of CMF between the predicted and the measured CCI values during winter
∑=
−=
n
i
iiPP
nMSE
1
2)ˆ(1
• Measures “fit-quality”
• Squaring emphasizes larger differences
Mean Squared Error
CCI forecast: Seasonal Analysis (2)Spatial distribution of the seasonal average mean MSE error (per pixel) of CCI between the predicted and the measured CMF values during summer
• Smaller error than in winter
• Error distinguished between land-sites and sea
• Meteotest method using GFS based Weather Research and Forecasting model for wind fields
DNICast: Cloud properties for solar resource
and forecasting
• DLR-PA method using Meteosat Rapid-Scan-Modus HRV channel
• DLR-DFD method using a sectoral method based on MeteosatSecond Generation imagery
DNICast: Cloud properties for solar resource
and forecasting
Enhancements due to cloudiness
A typical sky image (left panel) and the three selected areas (upper, middle, low) that correspond to different parts of the sky and solar zenith angles (right panel).
The CRE as a function of the cosine of the solar zenith angle and the ratio of upper cloud cover to the total one. The upper clouds correspond to zenith angles 0 to 45o (cos45o=0.707) and this area is highlighted.
Tzoumanikas et al., Renewable Energy, 2016
Enhancements due to cloudiness
Calculate end summer 25(OH)D (A)
required for 95% to remain ≥ 25 nmol/L
by end winter
Calculate monthly spend of 25(OH)D
(B)
Calculate UV dose (C) required to increase
25(OH)D from winter low to A, account for
spend B
Determine safe midday exposure time (no
sunburn), D
Calculate dose in time D for every day March
- Sept. Integrate to summer total E.
Is E ≥ C across UK?
A tip for future use of ultraviolet radiation products
(in liaison with health scientific community)
• UV risks and benefits are highly correlated to ambient UV exposure• UV effects are dependent of human behavior, skin type and age
Acknowledgements
All Colleagues from the DNICast project (www.dnicast-project.net/)
The Hellenic Network of Solar Energy
Lab. of Atmospheric Physics
University of Patras, Greece
www.atmosphere-upatras.grThank you!