The following letter was received in response to the March 2002 feature, “Industrial-Strength Sag Correction,” by Dr. Deepak Divan, Robert Schneider, and Daniel Bielinski.

Dear Editor,

On page 26, the authors state that “the utility reliability level for normal utility events is 99.99% (4 nines, or 2107 sec per year).” This number is used again in the table on page 28.

This got me thinking about how many seconds there are in a year. When I do the arithmetic for a 365-day year, I get 31,536,000 sec. When I calculate 4 nines (× .9999) on that number, I get 31,532,846. Subtracting that from the original sec-per-year comes to 3154 sec. This number doesn't agree with the figure given in the article. As I figure it, 2107 sec per year comes out to .9999332, not .9999.

The same thing occurs when I do the math for the 7-nines number (2.5 sec per year). Seven nines (.9999999) times the number of sec in a year comes out to 31,535,996.85. The subtraction comes out to 3.15 sec, not 2.5 as stated on page 26. How did the authors come up with their numbers for 4 nines and 7 nines?

Later in the magazine, on page 46, the article titled “The Power of Energy Storage” refers to 3 nines and 9 nines. This seems to conflict with the previously mentioned article. Is there some consensus in the industry for these numbers of 9s?

This was my first reading of your magazine. A colleague has a subscription. Aside from these minor details, I enjoyed the articles and will send in an application for a subscription.
David Rowinski
ManTech QA

Author's Response

Mr. Rowinski certainly raises a good question, one that engineers everywhere will surely agree with.

The concept of high nines has really come from attempts to translate a reliability requirement into simpler terms. Precursors to this concept include the six-sigma reliability program popularized by GE, the recent ad campaign by Microsoft on delivering 5-nines reliability (99.999%), and the assessment of IT power system reliability by the Gilder Group's Mark Mills and Peter Huber.

To answer the question, we were following popular nomenclature in representing a 99.9932% reliability level as 4 nines. As every additional nine represents an order of magnitude improvement in overall reliability, we felt this was an easier way for readers to visualize the significant differences in performance levels.
Dr. Deepak Divan
SoftSwitching Technologies

Web Focus

I would like to see an area in PQ News Beat that identifies training available in such areas as fuel cells, battery maintenance, power quality, power equipment, power testing equipment, grounding, and overcurrent protection. The area could include training available from colleges, training groups, and equipment providers.
Michael Lloyd
Delphi Auto

Editor's Note

Currently, we inform readers of upcoming industry events through the Web site's industry calendar, the magazine's “Meeting Place” page, and our annual calendar. However, we tend to limit these events to larger conferences and meetings primarily sponsored by associations. If anyone would like to announce a company's training opportunities, please use the PQ discussion forum.
Phil Musser
Power Quality magazine

From the Forum:

I am trying to locate a listing of power companies and/or product suppliers that offer whole home surge protection. If someone has an opinion or experience as to the top/best providers, please post your response via the PQ discussion forum.
Jeff Tuel

Power Quality magazine appreciates your comments. Direct your letters to Phil Musser at Letters, if published, may be edited for space and clarity. To join the PQ discussion form, visit our Web site at