3.4.3 Liguid Nitrogen The HPGe detectors used in the IMPs operated with liquid nitrogen at a temperature of -196°C. In the early months of the program the liquid nitrogen was air lifted from Honolulu on scheduled MAC cargo flights. Two military surplus 500-gallon containers were used. Shipping regulations required that the pressurized containers be vented outside the aircraft cabin. The condition of the containers, combined with these regulations, resulted in excessive nitrogen loss before reaching Enewetak. The on-atoll transfer containers were military surplus, wheeled, horizontal 50-gallon liquid oxygen carts, all of which had a high liquid nitrogen loss rate. This system was rather expensive and inconvenient. An improved system was devised, and better containers purchased. A military surplus, trailer-mounted liquid oxygen/liquid nitrogen plant was obtained, and the base operating contractor had people trained to operate it. About every two weeks, the plant was activated and two of the three on-atoll liquid nitrogen containers were filled. The containers were Linde LS-160B models, each holding 160 liters. This scheme successfully supplied the IMP and Radiation Lab with liquid nitrogen. 3.4.4 Detector Performance Three detectors were purchased for use in the project and a fourth was ordered a few monthslater, when the effects of Enewetak conditions on the detectors were confirmed. Two other detectors had been procured for a similar measurement program at the Nevada Test Site (NTS). Detectors were assigned by DOE to Enewetak or NTS, based on priority and scheduling of the two projects. Detectors were transferred informally and expeditiously, in response to DOE direction. All six detectors were used at Enewetak at various times. All detectors used at Enewetak were initially calibrated in Las Vegas, as discussed in Section 3.2.3. Starting in July 1978, a calibrated 241 4m source was available on-atoll and periodic remeasurements of effective detector area were made. These were used to provide an effective area correction factor for data handling. Field calibration sources, consisting of 4lam, 187g and 60Co, were used for three-times-daily detector performance monitoring. Field calibration was performed to set the gain of the detector electronics, and to generally track detector behavior. Tech Notes 5.2 and 11 discuss effective area factor and field calibration. For the field calibration measurements, the percentage standard deviation for the 241 4m value was 2 to 5 percent. The mean error in a series of effective area measurements was 1.1 + 0.8 percent. In the first months of the project, gradual loss of detector resolution with usage was noted. This was traced to water vapor entering the liquid nitrogen dewar during refilling in the field, causing an ice layer to form at the bottom of the dewar. This in turn partially insulated the detector, causing higher than design operating temperature. The problem was solved by the following maintenance procedure. About once each month, the detector was brought to room temperature, and ethanol used to remove water from the detector dewar. The dewarinterior was then dried using a stream ofair. The dewar wasthen refilled with liquid nitrogen. Operational history of the detectors is summarized in Appendix D. The average detector life span when installed in an IMP was about four months, with a range of less than a month to over seven months. Causes or symptoms of failure were: preamp corrosion, vibration sensitivity, no signal transmission, wide peaks and noise at low energy, and the dewarfailure. The last three items listed can probably all be classed as dewar failure, and were ultimately traced near the end of the project to corrosion of the 22 mil beryllium entrance window, or the beryllium-aluminum epoxy seal. An all-aluminum window was ordered on repaired detectors, but was not available in time to be used on the Enewetak project. 106