Development and Testing of an Ice Accumulation Algorithm

On the heels of a series of severe ice storm events in Oklahoma, the Oklahoma Association of Electric Cooperatives (OAEC) sought help for real-time assessment and forecasts of ice accumulation on their power lines. OAEC had partnered with the Oklahoma Climatological Survey (OCS) on development of a decision-support system that consolidated available information, but fell short of actual ice assessments and forecasts. OAEC worked with the National Weather Service (NWS) Forecast Office in Tulsa, Oklahoma to address those needs.

Using their experiences with previous ice storms, OAEC and NWS-Tulsa developed the following ice accumulation index to categorize threat levels:

• Sperry-Piltz Level 1 – Some localized utility interruptions possible … typically lasting only 1 or 2 hours maximum.

• Sperry-Piltz Level 2 – Scattered utility interruptions expected … typically lasting less than 8-12 hours maximum.

• Sperry-Piltz Level 3 – Numerous utility interruptions, with some damage to main feeder lines expected...outages lasting from 1-3 days.

• Sperry-Piltz Level 4 - Prolonged & widespread utility interruptions, with extensive damage to main distribution feeder lines and possibly some high voltage transmission lines...Outages expected to last more than 3 to 5 days.

• Sperry-Piltz Level 5 - Catastrophic damage to entire utility systems ... Outages could last from one week to several weeks in some areas.

The categories are based upon combinations of precipitation totals, temperatures and wind speed. Utility systems may be able to handle moderate ice accumulations, but stressed lines under wind forces are more likely to break. Therefore, one inch of ice may be a Level 2 or Level 3 ice event, but if wind speed exceeds 25 mph, it becomes a Level 5 event.

The algorithm was tested in several ice events during the winter of 2007-2008. Anecdotal evidence suggests that the algorithm performed exceptionally well, with observed damage consistent with the scale. In some cases, the algorithm out-performed local utility managers, either indicating more severe problems than they anticipated (prior to getting crews in the field) or areas of less damage within the overall storm pattern. Use of the algorithm has implications on deployment of repair crews during and after the event to restore power to customers as quickly as possible, and also has positive implications for use by the Oklahoma Department of Transportation to predict ice accumulation on streets, highways, and bridge surfaces.

The algorithm is undergoing a more systematic evaluation in collaboration with OCS. Students at OCS are calculating frozen precipitation accumulation and wind speed at each Mesonet site for ice events dating back to 1994, including both severe and less substantial events. These data will be used to calculate the ice accumulation index for each Mesonet site for each event, which will then be compared against damage and power outage reports to validate the algorithm. Preliminary results should be available for presentation at the meeting.