This URL here was posted on AMIA-L.Nonprint Media Update: Longevity and Optical Media
"CD-ROMs only last five years!" For those of us concerned with long-term storage of scientific and technical information on optical media, this is a myth that deserves debunking. My research indicates that this idea started with a couple of misleading articles published in the popular press a few years ago, and has rapidly spread to become a sort of doomsday prediction for CD-ROM as a viable storage medium.
A real understanding of how long a given CD-ROM is going to be effectively usable requires a discussion of the physical life expectancy, or longevity, of optical media in general. Since the early 1980's, when optical discs were invented, manufacturers have been performing extensive life-expectancy studies, claiming results that varied from a low of five to a high of over 200 years. The primary caveat of these test results is how the optical media are handled and stored. It is now commonly understood that CD-ROMs are susceptible to environmental degradation in the form of light, heat, and humidity. The question is to what degree (literally!) and over what timeframe?
In 1992 the National Institute of Standards and Technology (NIST) published Development of a Testing Methodology to Predict Optical Disk Life Expectancy Values. The study methodology used "accelerated aging," which assumes that temperature and relative humidity (RH) are the crucial variables that, over time, affect optical media longevity:
"... with storage at nominal room temperature and 90 percent humidity [results indicate that] the most conservative estimate is 57 years, while a more liberal estimate is 121 years. In either case, a relative humidity between 40 and 50 percent should lead to an even longer life expectancy." (NIST, 1992)
In 1995, results were published in a report by Eastman Kodak scientists, whose tests had attempted close-conformance to the then-developing standards for optical media life estimation. Their results showed that, if stored in the dark at 25°C (80°F), 40%RH, with 95 percent confidence, 95 percent of the media tested will have a data lifetime of greater than 217 years (my emphasis). The same year, in an issue of Computer Technology Review, it was reported that "through accelerated life testing, the shelf life ... is estimated to exceed 36 years for 97.5% of the [optical] disk surfaces, under the rather harsh storage conditions of 30°C, 80%RH. Similarly, the archival life, the time during which data can be effectively read, is estimated to exceed 510 years under the same conditions."
In spite of these and other published longevity test results, a 1998 article in US News and World Report, which inaccurately cited results from tests at the National Media Laboratory, raised the five-year specter. That article was quickly followed by another in the same vein in Business Week, which perpetuated much of the same misinformation. Both of these articles claimed, among other things, that no standards for media longevity testing existed, when in fact the American National Standards Institute (ANSI) had at that time written and accepted standards for validating claims of longevity.
Though second-generation CD-ROM copies are somewhat less tolerant of environmental extremes, all optical media can be stored for significant periods of time under proper conditions. The DOD Handbook for DOD Produced CD-ROM Products, (Draft, 31 July 1995), states only that the "CD should ... not be exposed to excessive sun or heat."
At DTIC CD-ROMs are stored in a climate-controlled environment, well below "nominal room temperature...and humidity," and are never exposed to excessive sun or heat. For DTIC's storage conditions, current estimates for CD-ROM longevity are well in excess of 100 years.
If you have any questions about this subject or information on its sources, please do not hesitate to contact me directly. My email address is pclifton@dtic.mil.