PREFACE TO APPENDIX B: TECH NOTES The Tech Notes in this Appendix are an accumulation of papers, each documenting how or why something was done, or the results of special investigations. Generation of Tech Notes was begun in November, 1977, at the suggestion of Phil Nyberg, EPA, who was serving in his first tour of duty as Technical Advisor to the DOE/ERSP Manager. This use of Tech Notes as a special form of documentation is patterned after a similar technique utilized by the EPA and some other organizations. The original intent was for each Tech Note to document actions and results at the time a task was performed so the basis for actions, and any decisions of consequence which might follow, would be available for review by staff members following later in the rotation schedule. While continuing to fulfill this purpose, preparation of a Tech Note also became a means of transmitting data results, or conclusions and recommendations of special investigations, to the Commander, Joint Task Group, and his staff. Most Tech Notes were distributed to contractor agencies involved in the cleanup operation as well as to the JTG, but there were some exceptions to the usual pattern of distribution. In general, the Tech Notes prepared since August, 1979, have been reviewed only by members of the Editorial Committee working on this Final Report, and the DOE/ERSP Project Managers. Tech Notes are numbered by subject matter. All Notes dealing with the same subject have the same number in front of the decimal point. Thus, Tech Notes numbered 2.n all deal with the determination of the ratio of total transuranics (TRU) to americium-241, while n takes on the values from 0 through 24 to include all islands for which this determination was made (with the exceptions noted in the Contents of this Appendix). Each Tech Notein the 2 series describes the methods and results for estimating the ratio of TRU to 41 am fora single island. At the start of the cleanup project the ratio and error were estimated by the sample mean and standard deviation of the ratios from individual samples. In those cases where more than one population of ratios was present on an island, a separate analysis was performed to determine the boundaries between the populations of ratios. The statistical assumption on which use of the sample mean is based is that the variance of the TRU value is proportional to the square of the 24lAm value. As more data were collected, it became clear that a more accurate assumption would be that the variance of the TRU is proportional to the 241 am value. An estimator based on the latter assumption, described in Doctor and Gilbert (1978), was therefore used from February 1978 until the end of the project. In the process of changing the computer programs on-island to use the new method, a typographical error was made on entering a program into the computer. Although the error did not affect the estimate of the ratio of TRU to “41Am, it made the estimate of the standard deviation too large. This in turn caused the propagated standard deviation on the final TRU values to be too large. The 0.5 s upper bounds on the area average estimates, where s is the standard deviation of the kriging error, were therefore also too large. The standard deviation estimate on the ratio has been corrected in the text of the final report. The incorrect original estimate has been left intact in the Tech Notes, but an appropriate footnote has been added. While it is true that certain error terms were incorrectly computed on the high side, in no case was the magnitude of the difference between correct and incorrect numbers large enough to affect soil removal decisions or final categorization for certification purposes. This approach is taken here because the Tech Notes present information upon which decisions were made at the time. While the standard deviation estimate on the ratio was alone not of great importance to decisionmakers, the situation represents the philosophy followed throughout the Tech Notes; namely, that a Tech Note written early in the cleanup program should not be modified by knowledge gained later in the program since this would give an improper picture of the information available at the time decisions were made. Knowledge gained later is, in a few instances, presented in a follow-up Tech Note bearing the same numberin front of the decimal as the original Note. For ease of reference, the Tech Note number follows the B in the pagination.