According to The Huffington Post, more than 660,000 people lost power in Connecticut during Tropical Cyclone Sandy. A spokesperson at Connecticut Light and Power reported that 90% of these outages were caused by falling trees that downed power lines. Others spoke of heavier damage, such as flooded substations as the root cause.
Utility distribution operators and outage crews put in a tremendous effort to repair damages and restore power. In the aftermath, utilities are doubling down on grid resiliency to be better prepared for next time. Smart utilities are investing in looped power lines peppered with intelligent devices in an effort to reduce the outage impact of downed power lines and other faults. By “looped and intelligent,” I mean power systems that have automated redundancy, enabling faster restoration. System Average Interruption Frequency Index (SAIFI) is the IEEE metric for the number of sustained interruptions per customer over a year. The distinction between a shorter and a sustained outage varies from state to state but is most often 5 minutes, meaning a sustained outage is greater than 5 minutes.
In a recent interview, Don Schneider, who oversees grid modernization projects at Duke Energy, told me that, “As part of Duke Energy’s overall Grid Modernization program, we are deploying grid automation devices that help improve SAIFI.”
When dual- or multi-sourced lines with grid automation devices are in place, there will be significant reductions in affected customers during storms that cause heavier damage over a small area (for example one faulted substation in looped circuits), or moderate to extensive damage over a large area (for example a half a dozen randomly distributed faulted lines per dozens of substations, i.e., enough transformer capacity to reconfigure sourcing to some or all un-faulted sections).
“Self-healing teams, by means of two-way communication to fault-interrupting and switching devices, can locate a faulted section of line, isolate that section of line, and restore power from another source,” says Schneider.
Self-Healing Helps Many
He also describes how the investment in self-healing is creating returns for Duke customers in Ohio. “In our currently active Ohio Grid Modernization project we have installed 24 self-healing teams over the past 4 years. From these self-healing teams we have experienced 20 operations that have resulted in 30,000 customers that have avoided a sustained outage,” says Schneider.
Based on EPRI research, it’s reasonably conservative to assume that an avoided outage is worth $2.50 to residential customers and $250 to small commercial customers. With a typical 90:10 ratio between these two customer classes that comes out to an average value of $25 per outage. Using this conservative assumption, it’s fair to say that self-healing in the currently active deployment has likely created $750,000 in satisfaction value to Duke Ohio customers.
While self-healing systems can do little when power systems are faulted at the majority of sources within a service territory, it’s interesting to look into cases such as in Connecticut, where 90% of outages were reportedly line faults caused by trees. Assuming most of the substations were still standing, and if all lines were modernized, the utility could quickly restore 20%(severe) to 60% (moderate) of all affected customers. This means that 132,000 to 396,000 customers would have lights back within seconds or minutes, representing a value of $3.3 to $9.9 million for a single storm.
The performance of some reliability programs can be tracked, and improvements in SAIFI indicates that a deployment is successful. “Prior to the start of our program, our 2008 annual average Ohio SAIFI number was 1.33. Our 2012 annual average Ohio SAIFI number is expected to be 1.03,” says Schneider. He points out that the Grid Modernization program consists of other substation device automation and critical line device automation that also contributes to improved SAIFI.
Tags: Digital Utility Strategies, Grid Resilience, Natural Disasters, Smart Grid Practice
| No Comments »