Y2K and the HP

by Joyce P. Davis, CHP

Reprinted from the HPS Newsletter- September 1998 by permission of the author. © 1998

Once, when I was trying to explain radiation science to someone, he became concerned when, after talking about doses in rem, I switched to millirem. I was startled to realize that he assumed millirem was the larger unit. He reasoned that a millipede had a thousand legs, so a millirem must be a thousand rem. And a thousand years (translated from Latin) is a millennium. The world is about to close the second millennium of the Common Era and open the third. We, the scientifically literate, use "kilo" (from the Greek) or "k" for a thousand and save "milli" or "m" for one-thousandth. Thus, the problem of strange computer goings-on as the twentieth century swings into the twenty-first, known as the "millennium bug" to those who cling to Latin, is called the "Y2K Problem" by us technical types.

As almost everyone is aware by now, the Y2K problem is the legacy of early computer programmers who economized on use of machine memory by using only two digits to represent the year in date fields. That means that the year "99" will be followed by "00." The big question is: Will the computer act as if the year 1900 has come again? If so, what will happen? There are other dates besides 1/l/00 that may cause trouble. For example, because some programmers used 99 or 9999 to represent end-of-file, dates like 1/l/99, 4/9/99 (the 99th day of 1999), and 9/9/99 may have their own problems. Also, the intricacies of our Gregorian calendar may not have been fully addressed, i.e., unlike other years divisible by four, years divisible by 100, like 1900, are not leap years, unless they are divisible by 400, like 2000. Will the computer recognize that 29 February 2000, is a valid date? Some systems can be made to work properly by turning the clock back to 1972, the most recent leap year with the same correspondence of days of the week with dates as the year 2000.

At first, people focused on big old mainframe systems. Then, they began to realize that date-processing microchips are contained in all sorts of control and monitoring systems these days, from the "smart" HVAC system to the "smart" coffee pot; some of these so-called "embedded" chips are not ready for the year 2000. The problem of data interchanges also surfaced. Finally, we are becoming aware that, in today's interconnected world, a failure in one microprocessor in some obscure but vital component could bring down a much larger system, like a power grid or communications network. Planning for "business continuity" in case of such potential disruptions is now getting attention. The U.S. federal government, like governments around the world, has been addressing the year 2000 issue, both for its own affected systems and for regulated industries. The Nuclear Regulatory Commission (NRC) has issued two generic letters-one to nuclear power plant licensees and one to uranium processors-requiring information on Y2K readiness programs. The power plant licensees must also provide written certification that facilities are Y2K-ready with regard to compliance with NRC license and regulatory requirements. (The generic letters and related documents are posted on NRC's Y2K Web site, listed below). The letter also refers to an industry guidance document, "Nuclear Utility Year 2000 Readiness." In its examples of systems and equipment that may be affected by Y2K problems, NRC includes, inter alia: radiation monitoring systems, dosimeters and readers, emergency response systems, security systems, communications systems, and engineering programs. NRC is making Y2K-related upgrades to its own Emergency Response Data System, which receives data during declared power plant emergencies, and licensees must upgrade their ERDS links, if necessary.

How will the coming of the millenium affect the health physicist? The subject has been discussed on "Radsafe", the health physics mailing list. The Radsafe archives (romulus.ehs.uiuc.edu/cgi-bin/lwgate/RADSAFE) include discussions of the issue and of relevant Web sites. Many manufacturers and suppliers of equipment and instruments have Web sites that identify deficient and Y2K-compliant equipment. Among the problem areas reported by health physicists are the computers, operating systems, databases, and application software used in dosimetry systems; automatic irradiators that do decay corrections; calculators with date functions; and data stored using non-Y2K compliant database programs. Problems are also introduced into otherwise compliant programs by the need to rely on data supplied by outside systems. For example, some software that calculates dose from radionuclide emissions relies on NOAA weather data. The format change from two- to four-digit years can introduce the need for substantial re-programming. The readiness of the global positioning system, used by some to identify monitoring locations, is not clear. And, of course, if power, conununications, or transportation systems suffer significant disruptions at the turn of the century, health physicists, along with everyone else, could feel the effects.

Here are some Web sites that may be helpful:

• National Institute of Standards and Technology: (www.nist.gov/y2k)
• Year 2000 Impact on Biomedical Equipment: (www.fda.gov/cdrh/yr2000/year2000.html)
• NRC Year 2000: (www.nrc.gov/NRC/NEWS/year2000.html)
• NEI Nuclear Utility Year 2000 Readiness: (www.nrc.gov/NRC/Y2K/NRCNEI/NEI9707.html)
• DOE Year 2000 Web Site: (websparc.hr.doe.gov/year2000/)
• GAO Guidance on Year 2000: (www.gao.gov/y2kr.htm)
• Some Year 2000 links (www.year2000.com/NFlinks.html)
• Federal Year 2000 Commercial Products Software Database: (y2k.policyworks.gov/)
• Defense Information Systems Agency Y2K Products Compliance Catalog: (www.mitre.org/research/cots/COMPLIANCE_CAT.html)
• GSA Year 2000 Teleconnnunications: (y2k.fts.gsa.gov/)
• GSA Year 2000 Information Directory: (www.itpolicy.gsa.gov/mks/yr2000/y2khome.htm)
• British Standards Institution: (www.bsi.org.uk/disc/year2000.html)