solar consulting services, usa - projects tracker...

13
Chris A. Gueymard Solar Consulting Services, USA CSP Today Seville 2014

Upload: phungkhanh

Post on 21-May-2018

218 views

Category:

Documents


0 download

TRANSCRIPT

Chris A. Gueymard Solar Consulting Services, USA

CSP Today Seville 2014

๏ Founded 2003, based on expertise acquired since the early ‘70s

๏ Provides high-end solar resource services to the solar industry:

๏ On-site measurement with advanced weather stations

๏ Satellite-derived solar resource data

๏ Aerosol data and analysis (SOLSUN)

๏ TMY data

๏ Specialized spectral irradiance modeling (SMARTS code)

๏ Circumsolar radiation

๏ Atmospheric attenuation (solar tower plants)

๏ Collaborates with major research institutions (NREL, NCAR, CNRS,

DLR, KISR, Universities…)

๏ Prepares aerosol data for major solar resource data providers

๏ Offers capacity-building workshops on resource assessment, etc.

๏ Involved in the development of solar irradiance standards (ASTM, IEC)

Solar Consulting Services

2

SOLSU N

๏ DNI: Direct Normal Irradiance, measured with pyrheliometer (≈2.5° acceptance radius)—may be different than what is actually usable by CSP!

๏ AOD: Aerosol Optical Depth, measured with sunphotomer/spectrometer

๏ CSI: CircumSolar Irradiance, measured with aureolemeter

๏ CSR: CircumSolar Ratio,

CSR =[CSI(2.5°) + DNI(0.25°)]/DNI(2.5°)

๏ Collaborative SE paper (2014) provides definitions and results: DNI, CSI, CSR, as affected by aerosols and clouds

Know Your Acronyms: DNI, AOD, CSI, CSR

CS

R

DNI

3

๏ New version of the U.S. National Solar Radiation Data Base (NSRDB) expected shortly (Dec. 2014)

๏ Will use the GSIP physics-based satellite model at 4 km resolution, in replacement of the 10-km SUNY/CPR empirical model; TMY/TBY files for each grid cell

๏ Solar Consulting Services (SCS) provides (i) calibrated aerosol data for North America, an important input of GSIP; and (ii) the best possible clear-sky radiation model, REST2

๏ NREL will help CSIRO improve the Australia solar resource maps, with aerosol data from SCS; new products expected in 2015

๏ Solar resource maps and databases were prepared for India in 2010 and 2012, with aerosol data also from SCS

๏ Another update expected for 2015/2016

๏ Getting aerosols right is a challenge over India; major source of uncertainty in DNI!

Solar Resource Developments at NREL

4

DNI

๏ Global NASA SSE dataset (coarse 1x1° resolution currently) being revised

๏ Will use GIS-style web server, improved algorithm and 0.5x0.5° resolution, offer smartphone app—availability mid-2015

๏ Further developments (budget permitting): 10-km resolution, to be integrated into NREL tools for solar analysis and decision support

๏ Continued support of IRENA’s Global Atlas

Solar Resource Developments at NASA

5

๏ Global Atlas for Renewable Energy portal, hosted by Masdar

๏ Various gridded maps of DNI, GHI, wind, population, infrastructures etc.; data integration possible (GIS-style)

๏ Variety of data providers, spatial resolutions, etc.

๏ Risk of inconsistency between databases from different sources

๏ Mostly for energy policy, preliminary potential studies and education

๏ Solar resource training workshops; first one to be held in Kuwait jointly with KISR, Nov. 16–20: 15 selected participants from MENA

Solar Resource Developments at IRENA

6

๏ ESMAP program created by the World Bank to help emerging countries with their renewable energy resource assessments

solar and wind

modeled databases and resource maps

ground measurement campaigns

๏ Limited budget, provided by various donors

๏ Countries already being supported: Pakistan, Zambia, Malawi, Tanzania, Maldives

๏ Candidate countries (in process): Indonesia, Papua New Guinea, Vietnam, Morocco, Tunisia, Niger, East Asia Pacific region

๏ In discussion: Namibia, Somalia

๏ Namibia: First solar resource assessment done in 2012, based on aerosol data from SCS

๏ Showed very high potential for CSP!

Solar Resource Developments at ESMAP (World Bank)

7

>2900 kWh/m2

Solar Forecasting Principles

๏ Irradiance forecasting becomes necessary, due to more variable generation

๏ Range of forecast horizons for various analyses: 1 min to 7 days; all methods actively researched: <1 min to 15+ min: sky imagers with cloud motion software

30 min to X hours: cloud motion vectors

Hours to days ahead: NWP modeling

AI, machine learning

8

VALUE CHAIN

Solar Forecasting Developments—USA

9

๏ Major DOE funding for 2 projects piloted by NCAR (physical forecasting) and IBM (AI, machine learning), period 2013–2015

๏ NCAR project (SunCast) is a public-private-academic partnership including many players (national labs, research groups, consultants, electric utilities, etc.); SCS is involved

Objectives: Irradiance forecasting (DNI and GHI), delivery mechanisms, validation, uncertainty quantification, real-world applications…

Development of an advanced WRF-Solar NWP model with improved cloud and irradiance modeling for high spatial/temporal resolution

Sophisticated data assimilation with AI blending, use of analog ensembles

๏ Many other research groups involved at various universities

• EU FP7 project with 12 participating groups from 7 countries, led by DLR

• 4-year project, 2013–2016; international Advisory Board

• Focused scope: forecast lead times of up to 4 hours (nowcasting)

• Looks into both temporal and spatial issues of DNI nowcasting

• Industry partners for direct applications and onsite validation

• Specific tasks on aerosol forecasting and circumsolar radiation prediction

Solar Forecasting Developments—DNIcast (EU)

10

Among the very active groups in solar resource and forecasting: MATRAS (U. Jaen)

• Operational weather forecasts for Andalucia: 5-km spatial resolution, 72 hours ahead: T, precip, wind and GHI

• Solar nowcasting with sky imagers and advanced cloud-tracking algorithm

• DNI and GHI forecasts using WRF in Southern Spain, 3-km res.

• Independent evaluation for all seasons and sky conditions

• Professional forecasting services provided by spinoff SynerMet

Solar Forecasting Developments—Spain

11

Solar Forecasting Developments—Japan

12

๏ Many research groups involved at various universities

๏ Extremely competitive funding process for a national “Energy Management System”

๏ First phase, 2012–2015, 26 teams in competition; one of them is the TEEDDA group (Tokai U., Chiba U. and U. Tokyo), heavily relying on satellite data and physical models (spectral irradiance!); method to be ultimately applicable globally

๏ Second phase, 2015–2020, for one surviving team

Conclusion

13

๏ New solar resource maps, databases and GIS products recently

proposed or in development, but effective “quality” still unknown other

than in broad/vague terms

๏ Lack of sufficient validation or benchmarking of these resource data

products; serious issues (>15% bias) over arid areas still unresolved;

better aerosol databases and products needed

๏ Lack of interest from the industry? Funding difficult!

๏ Strong developments in solar forecasting, many active teams around

the world

๏ Forces a win-win collaboration with the meteorology world; new

powerful products expected in 2016

๏ Day-ahead forecasting of clouds and aerosols still challenging

๏ Advanced ensemble forecasts combined with AI statistical methods

should bring improvements