Hurricane Sandy in hindsight

How can we prevent the ravages of another superstorm?

By Massoud Amin, IEEE, University of Minnesota

A year has passed since Hurricane Sandy overwhelmed the eastern United States and knocked out power to millions for days and even weeks. We must ask ourselves: What can be done to avoid such widespread outages in the future?

The answer should be an optimistic one, though it's also multi-faceted and comes with a few caveats.

Outages happen, details matter

We have the technology to harden the grid, make it more reliable and bolster its resilience. But generalizations must be tempered by reality. Outages happen. No technology or amount of money can fully protect the entire grid from the impacts of extreme weather, occasional equipment failures or anomalous events.

Details matter, too. Every grid requires a thorough risk assessment to understand its strengths and weaknesses and to determine where cost-effective investments could be made. We must recognize and adapt to a changing risk portfolio, which now includes extreme weather and terrorism.

Generally, it's possible to modernize a grid in one to two years, though systems integration and maturation to achieve maximum value takes more time. The cost depends on the desired outcome. It's likely to be substantial for any individual utility and its customers, so investments must be analyzed for costs and benefits, in the context of a much bigger picture.

As Americans we must ask ourselves: What quality of life do we desire for the 21st Century? Are we willing to invest today for economic prosperity going forward? What are the consequences of taking no action?

Facing the facts

For the past 32 years, as a nation we've under-invested in our power infrastructure and overharvested the existing assets (See Figure 1.) A World Economic Forum report in 2011 ranked the United States below 30th in the world for the quality of our electric power sector. Electricity outages and power quality issues cost the U.S. economy about $119 billion to $188 billion annually, according to a landmark paper published by the Electric Power Research Institute in 2001. Between 2001 and 2010, major outages in this country doubled, according to the data from the U.S. Energy Information Administration. Today, 78 percent of outages affecting consumers are weather-related. Finally, New York, New Jersey and Maryland experience annual outages totaling 214 minutes, whereas Japan's cumulative outages total 4 minutes.

Click to enlarge

Caption: Figure 1.  Since the "cross over" point in about 1995 utility construction expenditures have lagged behind asset depreciation.  This has resulted in a mode of operation of the system analogous to "harvesting the farm far more rapidly than planting new seeds" (data provided by EEI and graph courstesy of EPRI)

Fortunately, my (and others') economic analyses demonstrate that investments in power infrastructure not only deliver value exceeding costs, but also contribute significantly to economic growth and job creation.

In short, we should make these investments on a case-by-case basis if we value our standard of living and desire a prosperous future and global competitiveness. The most significant variable in this scenario is public policy and the political will to shape our future.

The context

Let's first examine the threat. Hurricane Sandy was just the latest in a succession of unprecedented storms that included Tropical Storm Irene and "Snow-tober" in 2011. Rather than a tight focus on Sandy's effects alone, grid investments should be based on probabilistic risk assessments for operations on blue sky days as well as extreme weather, equipment failures and other anomalous events.  

Strengthening the grid has three basic components: hardening, reliability and resilience. ("Hardening" means protecting assets to withstand physical assault. "Reliability" is measured by traditional indices for the frequency and duration of outages. "Resilience" is the ability of the system, once impacted, to return to normal operations.)  

The to-do list

In terms of hardening, substation locations should be optimized. Design and construction standards for flood protection should be raised, particularly for underground substations. Design standards for feeders should be raised to approximate standards for higher voltage lines. Selective undergrounding, based on probabilistic risk assessment, may be cost effective. Data-driven vegetation and asset management can prevent unexpected equipment failures. New materials can make power poles sturdier.

The distribution system, of course, generally lacks the intelligence long applied to generation and transmission. Thus a range of measures typically referred to as "smart grid" may be applied to increase both reliability and resilience. 

Sensors from the substation down the feeders to the smart meter on businesses and homes will provide operators with useful data. Advanced communication networks will bring that data to operators and deliver commands to control field devices. Substation automation is a first step in improving sensing and command-and-control capabilities.  

Changing the design of the distribution network from radial to loop arrangements will allow greater sectionalizing, which aids the granularity of fault detection, isolation and restoration (FDIR) -- sparing unaffected businesses and homes and focusing finite resources on specific trouble spots. Automated switches and reclosers can be programmed to speed the FDIR response beyond human capabilities.

As Americans we must ask ourselves: What quality of life do we desire for the 21st Century? Are we willing to invest today for economic prosperity going forward? What are the consequences of taking no action?

Software programs including outage management systems (OMS), advanced distribution management systems (A-DMS), geographic information systems (GIS), smart meters and advanced metering infrastructure (AMI) can help optimize the system. These programs must be integrated and the results delivered in dashboards that support human decision-makers to remove the fog of uncertainty during emergencies.

Keeping customers informed of estimated time of restoration (ETR), via social networks, should be improved to reduce the impact of major outages. If customers know what to expect, they can plan their own responses to prevailing conditions.

Data analytics will increasingly improve everyday operations and emergency responses and, ultimately, will generate predictive insights that bolster proactive practices.

A strategic approach might prioritize power reliability for public infrastructure such as street lights, police, fire and hospital facilities to maintain civil order and essential operations during chaotic weather events. Microgrids with distributed generation could enable the grid to shed significant loads and increase customers' self-reliance. Distributed generation and energy storage will also aid end users' own resilience.

Improving the indices

While we cannot eliminate outages, we can reduce the frequency and duration of outages, as measured by the traditional indices of SAIDI, SAIFI and CAIDI. (System Average Interruption Duration Index, System Average Interruption Frequency Index and Customer Average Interruption Duration Index, respectively.)

Based on IEEE models, as well as actual implementations at military bases with 20,000-50,000 inhabitants and cities with 500,000 to one million population, we've seen that the low end estimate for improvements in SAIDI and SAIFI from grid modernization is 12-14 percent, with CAIDI holding steady. At the higher range of possibilities, SAIDI and SAIFI might be improved as much as 30-40 percent and CAIDI by 17-18 percent. (This is for blue sky days.)

Though I've studied the blackouts of 1996 and 2003, I have not performed a similar review of potential reliability improvements based on Hurricane Sandy. But the data from 1996 and 2003 indicate that, in extreme weather, we could improve these indices by an order of magnitude. (Please see my past work on weather-related reliability improvements here and here.)
Two utility proposals

We'll soon see how two specific responses to Sandy's destruction fare. Consolidated Edison (ConEd) Co. of New York City and Public Service Electric & Gas (PSEG) based in Newark, New Jersey, both experienced massive destruction and their customers endured extended outages, as did many neighboring utilities.

ConEd is proceeding with a $250 million hardening program for Orange and Rockland counties that relies on many of the aforementioned measures. Readers can review that plan by Internet searching for "Post Sandy Enhancement Plan," dated June 20, 2013.  

In contrast, PSEG on Feb. 20, 2013, floated a nearly $4 billion proposal dubbed the "Energy Strong Program Petition" to the New Jersey Board of Public Utilities. PSEG's proposal includes ambitious hardening measures as well as smart grid-related work affecting reliability and resilience. The five-year program budgets $2.8 billion for electric infrastructure and $1.2 billion for natural gas. PSEG's proposal was actually initiated in 2011 after Tropical Storm Irene and "Snow-tober," when regulators required a response plan.

The failure to take cost-effective steps to modernize our grid for our quality of life, productivity and competitiveness -- let alone extreme weather events -- would be to miss a historic opportunity.  Merely coping is a defeatist mindset. We cannot just sit on our hands. We have the technology.  We know how to apply it. Now we need to build strong business cases, forge public-private partnerships and make the public aware of what is at stake -- nothing less than the future of this great country.

About the Author
Dr. Massoud Amin, is the Director of the University of Minnesota's Technological Leadership Institute and Professor of Electrical and Computer Engineering. A senior member of IEEE, he chairs the IEEE Control Systems Society's Technical Committee on Smart Grids, and serves as the chairman of the IEEE Smart Grid Newsletter.