The California Solar Initiative's RD&D Program has now concluded after operating since 2008 through 2016. All solicitations and research projects are now complete.



 Program overview + Research briefs       

 [25 MB ]
 Program overview  [13 MB ]

 Research briefs  (high resolution)                     

 [44 MB ]

For more information on the research projects and resulting reports please review the section below. All completed final reports are also available separately for review and download. 

FINAL REPORTS for completed projects along with individual research briefs can be found here.


CSI RD&D Program - Program Evaluation 



 Report Summary       

 [277 KB ]
 Final Report  [3 MB ]


 [4 MB ]


Funded Projects

This is a database of projects funded by the CSI RD&D. The project summaries are listed alphabetically below. You may also choose a subset of projects by clicking a keyword, or using the search function.

High-fidelity Solar Forecasting Demonstration for Grid Integration

Current Highlights:
  • FINAL PROJECT WEBINAR- slides.  PDF  (3.2 MB)
  • FINAL PROJECT WEBINAR - audio with slides. WMV (68.7 MB)
  • Final Project Report:  February 2016  PDF  (4 MB)

Project Information

Solicitation: Solicitation 3
Target Area: Grid Integration
Awardee: University of California San Diego

Green Power Labs, Clean Power Research, SCE (letter of support), SMUD, National Renewable Energy Lab (NREL), Power Analytics

CSI Funding: $1,548,148
Match Funding: $1,548,148
Start Date: July 2012
End Date: July 2015
Status: Complete

Contact Information

Principal Investigator Name: Jan Kleissl
Principal Investigator Email: jkleissl @ ucsd.edu
Principal Investigator Phone: 858-534-8087
Grant Manager Name: Stephan Barsun
Grant Manager Email: stephan.barsun @ itron.com
Grant Manager Phone: 509-891-3187

Project Summary

This project will demonstrate that solar resource forecasting can be the most cost effective strategy for integrating large amounts of photovoltaic (PV) generation into distribution systems. Initially, historical aggregate PV ramp rates will be analyzed and a baseline of solar forecast performance under the extreme ramps will be created. Additionally, sites for additional measurements to improve solar forecast accuracy will be proposed based on statistical analysis. For resource adequacy applications, the UCSD team will then improve and demonstrate forecasts for the marine layer meteorological conditions that affect a large fraction of solar PV in San Diego Gas and Electric territory on summer mornings. High resolution numerical weather prediction and statistical models will be developed and applied to improve forecast accuracy. Local applications of solar forecasting using sky imagery will also be demonstrated on five typical feeders with variations in PV penetration, location / meteorology, and voltage regulation equipment. On these feeders, fast demand response potential based on demand and solar forecasting and dynamic loading will be demonstrated. Through increased granularity in modeling distribution feeder voltages and their reaction to fluctuations in solar PV output, geographic diversity is expected to reduce previously observed impacts at high solar PV penetration.

Deliverables and Documents


  • Final Project Report: High-fidelity Solar Forecasting Demonstration for Grid Integration. Kleissl, J. UC San Diego. February 2016. PDF (4 MB)
  • Final Report for Subtask 4.4: Model Predictive On-Load Tap Changer Control for High Penetrations of PV Using Sky Imager Solar Forecast. Disfani, V., Ubiratan, P., and Kleissl, J. Dept of Mech. & Aerospace Engr., Jacobs School of Engr., Centers for Renewable Resource Integration & Energy Research. For CSI RD&D. UC San Diego. October 2015.  PDF  (1.6 MB)

  • Subtask 4-3 Report: Impact Research of High Photovoltaics Penetration Using High Resolution Resource Assessment with Sky Imager and Power System Simulation. Nguyen, Dung, Ubiratan, P., Velay, M., Hanna, R., Kleissl, J., et al. UC San Diego. November 2015.  PDF  (4.6 MB)

  • Subtask 4.2 Report. Sky Imager Forecasts for SDG&E Distribution Feeders. June 2015. Murray, K., B. Kurtz, A. Nguyen, et al. UC San Diego, CA.  PDF (1.5 MB)

  • Task 5 Final Report: Substation Net Load Operational Forecasts for Fast Demand Response Implementation. Kaur, A., Nonnenmacher, L., and Coimbra, C. Center for Renewable Resources and Integration, UCSD. Jan 27, 2015. PDF (741 KB)

  • Net Load Forecasting with Distributed Generation. Green Power Labs. Dec 16, 2014. PDF (3 MB)

  • CSI Subtask 2.3 Final Report: Recommend placement and operation of SDG&E weather stations to maximize forecast effectiveness. Coimbra Energy Group, UCSD. Mar 2014. PDF (1.5 MB)

  • Ensemble Forecasting of Solar Irradiance in Coastal Southern California. Ozge Akyurek, B., Zhong, X., and Kleissl, J. UCSD. Jan 24, 2015. PDF (2 MB)

  • Aggregate Ramp Rates Analysis of Distributed PV Systems in 5 Locations at San Diego County. Bosch, Juan L., Jamaly, M., and Kleissl, J. Dept. of Mech & Aerospace Eng., University of California, San Diego. February 6, 2013.  PDF  (8 KB)

  • Marine Layer Forecasting Report. Green Power Labs. May 30, 2014.  PDF  (5 MB)

The following three reports were developed as part of California Energy Commission funded work and cross posted here as cost-share:

  • Observability of Microgrid Operations by the California Independent System Operator. Final Project Report. Torre, William V. Center for Energy Research Department of Electrical and Computer Engineering University of California, San Diego. April 2014. PDF (1.7 MB)

  • Improved Modeling Tools Development for High Penetration Solar. UCSD: Byron Washom, Byron, Kleissl, J., Nottrott, A., Hanna, R., and Yang, H. Power Analytics Corp: Meagher, Kevin, Zhang, Y, and Koopman, R. Prepared for the U.S. Dept. of Energy. March 2014.  PDF  (6 MB)

  • Intra-hour Solar Power Forecasting at Utility-scale Power Plants in the Los Angeles Basin. Final Project Review Draft. Babacan, Oytun, Wang, G., Yang, H., and Kleissl, J. Prepared for the California Energy Commission Public Interest Energy Research PIER Program. 2012.  PDF  (2 MB)



  • FINAL PROJECT PRESENTATION-audio with slides WMV (68.7 MB)

  • Initial Project Presentation. Aug 9, 2012. PDF


Following is a ZIP file with code that contains analogue ensemble forecasts developed as part of this grant. The principal idea of the analog ensemble forecast is to use a clustering method to identify time periods in the past (“analogues”) that were similar to the time immediately preceding and during the current forecast horizon. These time periods represent an “ensemble” that can be used to

(i) correct the deterministic forecast given past forecast errors, and to

(ii) generate probabilistic forecasts.

To better understand the code files, please read the READ ME FIRST! file prior to reviewing the code files.


Forecasting Code File  ZIP  (566 K)

Project Activities

July 2, 2012 Project started

Project Links

Tags: Solar Forecasting