Electrostatic Discharge (ESD), Humidity and the Data Center

I was reading DataCenterKnowledge's post on Yahoo's Data Center revolution, and one area that I want to elaborate on is electrostatic discharge, humidity and IT equipment.

It’s been a long journey from third-party colocation space to the Yahoo Computing Coop. The journey started by addressing the accepted range of 45 to 55 percent humidification for data center space.  “Where did this belief come from?” Noteboom wondered.

Noteboom went searching for references, and says he found only a few academic scenarios warning of humidity issues. For practical arguments for tightly controlled humidity for computer rooms, Noteboom said he had to go back to the era of punch-card data entry, when concerns focused on whether excess humidity might wilt the punch cards and make them difficult to process properly.

“In 2005 we built a data center without humidification control,” said Noteboom. “It was filled with 25,000 servers. We were worried about static, and had all these protocols for handling equipment. We don’t have them anymore. We’ve operated through an entire lifecycle in this data center, and we’ve seen no impact.”

In 1980 as a summer intern I worked at HP on the idea that electrostatic discharge was causing component failures.  I was working with guys like Dick Moss who was an ESD expert.

Richard Y. Moss
Hewlett-Packard Co.,
Palo Alto, CA, USA

Two of the questions most asked by manufacturing managers are "How good must my ESD-control program be?" and "How much money can it save me?" These are reasonable questions, considering that managers must justify their decisions in terms of return-on-investment or payback period.

Unfortunately, the first question is difficult to answer and the second almost impossible. One rarely knows the extent of ESD damage in a process until after it has been eliminated.

There is a bunch of stuff I learned 30 years ago, zapping parts with ESD, testing parts for failure, digging holes in plastic IC components, poring fuming sulfuric acid to burn the plastic away, and looking at integrated circuits to see where failures were occurring.  We knew places like Fort Collins, CO had a higher electrostatic discharge risk due to low humidity, and you addressed the problem with wrist grounding straps, anti-static work surfaces, anti-static floor treatments, and air ionizers to remove the statically charged particles.

Quite frankly I've thought it dumb that the Data Center industry is obsessed about humidity which was most likely caused by a concern over electrostatic discharge, and potentially paper punch cards.  There are much more cost effective ways to remove ESD risk than humidity.  And, even a high humidity doesn't completely eliminate ESD.

If you want to address ESD then following guidelines like this.

Principles of electrostatic safe handling

There are two simple principles we can use to protect ESD sensitive components from ESD damage:

  • Only handle sensitive components in an ESD Protected Area (EPA) under protected and controlled conditions
  • Protect sensitive devices outside the EPA using ESD protective packaging

How many other dumb things are being done because of its been done in the past?

OK this is rant for the day.  I am calmer now.