Richard Perez

ASRC-CESTM, The University at Albany


The January 1998 ice storm in Northern New York, Quebec, Vermont, New Hampshire and Maine caused severe damage to electric power transmission and distribution lines. Several million customers were affected by lengthy power outages. Some outages lasted several weeks.

Several of the costliest consequences of the outage could have been avoided if only a fraction of the electrical power normally available from the utility power grid had been available to customers.

In this report we present strong evidence that customer owned grid connected photovoltaic PV systems equipped with emergency battery back-up storage could have prevented much of the above collateral ice-storm damage.

Reasonably sized PV systems (i.e., 1-2 kW for individual residences) would have supplied enough energy to power refrigerators, emergency lights, computer and communication equipment as well as [non-electric] heating systems.

This assertion is based on a site time specific monitoring of the solar resource following the ice storm, using cloud cover data from geostationary weather satellites. The viability of this methodology has been thoroughly assessed and presented in numerous publications by the author [1,2,3,4]


The satellite image in figure 1 shows that there were large areas of cloudless conditions in the regions affected by the storm, in the days that followed its onset.

  Figure 1: Satellite view of the Northeastern US on January 12, five days after the onset of the storm

  A quantitative analysis of a specific point near Plattsburgh, New York (Figure 2), shows that the solar resource, although not optimal every day, was far from zero in the week that followed the onset of the storm on January 7-8.

 Figure 2: Global irradiance near Plattsburh, New York

Regionally, the map presented in Figure 3 shows that the average hourly output of a 1 kW-ac PV system in the area affected by the storm would have been of the order of 120-160 Watt, that is 2.9-3.8 kWh per day. This regional map is based on the analysis of 150 satellite images. The uncertainty in the average PV output is of the order of ± 20%.

 Figure 3: Average hourly output of a PV array throughout the northeastern United States for the 2-week period following the onset of the storm

Therefore, a typical 2 kW residential PV system with a 10 kWh emergency battery storage (2 average days' output) would have been able to sustain a 250-300 Watt continuous load. This is more than enough to sustain all critical loads. Meeting these critical loads would have considerably lessened the financial impact of the storm for the affected customers.


Photovoltaic is considered by many as an attractive, environmentally safe and renewable source of electrical power. The advent of net metering legislation and other existing or pending incentives (e.g., New York's 25% tax credit) are bringing residential PV systems closer to economic feasibility, particularly if the systems can be financed through mortgages.

The added benefit of user-side power output reliability, demonstrated through this case study (as well as other power outage case studies by the author and colleagues [5,6]), should further increase the economic attractiveness of PV to all concerned parties: end-users, the insurance industry, governments and utilities (power distributors). The concerned parties should work toward incorporating these benefits into tangible financial decision-making elements: e.g., subsidies, enhanced tax incentives, or enhanced net metering schemes such as have been pioneered in several European locations.

 Acknowledgement: This study was made possible thanks to the ongoing support of NREL for the acquisition of satellite data archives (NREL contract XAH-515-222-01) and their analysis to investigate the impact of PV on regional and local electric loads (NREL contract XAD-817-671-01). Thanks to Marek Kmiecik for processing of satellite data.


  1. Zelenka, A., Perez R, Seals R. and Renné D., (1998): Effective Accuracy of Satellite-derived irradiance, Theoretical and Applied Climatology (in press -- accepted 3/98)
  2. R. Perez and R. Seals, (1997): Production of Site-Time Specific Irradiances from Satellite and Ground Data. Final report to NYSERDA and NREL. NYSERDA Report No 98-3 35 pp., NYSERDA, Albany, NY / NREL Golden, CO.
  3. R. Perez, R. Seals and A. Zelenka, (1997): Comparing Satellite Remote Sensing and Ground Network Measurements for the Production of Site/Time Specific Irradiance Data. Solar Energy 60, 2, 89-96.
  4. Richard Perez and Eugene Maxwell Editors, (1996): Proceedings of Satellites for Solar Energy Resource Information Workshop (Washington, DC). 2 Volumes, 135 pp.
  5. R. Perez, R. Seals, H. Wenger, T. Hoff and C. Herig, (1997): PV as a Long-Term Solution to Power Outages. Case Study: The Great 1996 WSCC Power Outage. Proc. ASES Annual Conference, Washington, DC.
  6. R. Perez, (1998): Solar Resource Availability in the Wake of the May 1998 Mechanicsville Tornado. Results presented at the ASES forum on satellite-based Solar Resource Assessment, ASES Annual Meeting, Albuquerque, NM.