BI and Blackouts

Data analysis plays a role in disaster prevention and recovery

The cascading blackout that disrupted business over much of the northeastern United States and adjoining areas for up to several days in August was a painful reminder for many businesses that risk management and disaster recovery programs are worth investing in. To others, it brought an even more painful realization that the programs they had invested in weren't adequate. When this second group asks what they could have done differently, the answer may lie less in diesel generators and tape backups, and more in collection of operations data, information analysis, and software.

Businesses for which electricity is an input — in other words, all of them — suffer when that essential utility is unavailable. Production lines stop, retail outlets have to close, perishables spoil, and suddenly downed computers lose everything in temporary memory and are unavailable to process transactions. The Dow Jones Newswires reported 10 days after the blackout that the loss of power cost millions of dollars to the New York economy alone and that Michigan employers lost about $700 million in net output.

But let's not forget the power companies, which lost a huge quantity of their product, one that has been described as the "ultimate perishable." The chief cause of the blackout had to do not with generation capacity but with transmission and distribution. Therefore, according to a Reuters report, software purchases by the power utilities should increase by much more than gas turbine orders will.

ICF Consulting is a managing, technology, and policy consulting firm. Its August 14 report, "The Cascading Blackout: Why Wasn't the Power Outage Contained?," states, "As markets have grown, grid operators experience increasing difficulty in ensuring reliability and frequently lack necessary information." Needed information can come from historical analysis (BI) and demand planning software. Although power utilities do use demand planning, it's often outdated — still based on the days before the power market became complicated by "open access."

ICF continues, "Operators seek to constantly maintain balance on the grid by matching supply and demand in real time." Such a goal is often best met by expert or AI systems, which can be bolstered by information from historical analysis.

Demand planning and monitoring help avert service disruption. But not all disasters can — or should — be prevented. ICF further writes, "A key question for operators is how much risk are they willing to accept — especially when the probability of an unplanned transmission outage is probably less than three percent a year."

Ali Samad-Khan, director of global operational risk strategy at SAS, explains that sometimes risk management consists of estimating expected loss amounts from unexpected, uncontrollable loss events and setting aside capital to survive when those catastrophes happen. This practice is established in the financial services and banking world. "Some degree of preparedness is important; it's really a question of how much — and there's a cost-benefit decision involved," says Samad-Khan.

So BI can help you prevent disasters and better prepare you to accept those you can't prevent. Can it also help you to recover what you've lost? Yes.

Although 70 percent of companies recently surveyed by Imation Corp. said they had a disaster recovery plan including regular data backups, that doesn't mean they're safe from catastrophic data loss. According to Mark Silverman, CEO of "storage intelligence" software vendor Bocada, it's common for half of all backup attempts to fail. He says some companies have a failure rate as high as 80 to 100 percent. A survey conducted on companies just after the blackout by Info-Tech Research Group shows 76 percent hadn't been adequately prepared to protect their data.

Silverman is frustrated that so many companies seek to solve this problem by swapping out their existing backup system for another one. "None of them are bad, they're just all hard to manage," he says. Silverman identified a niche opportunity to apply BI techniques to the management of backup and recovery systems. An agentless tool that helps collect and analyze data specific to enterprisewide, heterogeneous backup and recovery systems, Bocada helps ensure that when the lights do go out, your mission-critical data won't be extinguished along with them.

— Jeanette Burriesci

In this Issue: