Sensor Scoring Methodology for Subterranean Structure Confirmation, 10-9418

Printer Friendly Version

Principal Investigators
Darrell S. Dunn
Mark R. Lesher
Theodore A. Kramer
Stephen L. Wiedmann
Jeremy K. Zoss

Inclusive Dates:  09/15/03 - 03/15/05

Background - One of the primary goals for the Center for Nuclear Waste Regulatory Analyses (CNWRA) is to evaluate the subterranean monitoring methods for the Yucca Mountain hazardous waste site during the 100-year pre-closure period. The site is characterized as having extreme environmental challenges that include high temperature (more than 180° F), high radiation (greater than 1,400 rem/hr), elevated humidity (approximately 90 percent), and remote access. As part of the requirements for this site, a number of remote sensors will be deployed to measure the factors deemed necessary to monitor the integrity of the hazardous waste emplacement drifts. Monitoring the performance of sensors that are remotely deployed in hazardous locations is not a Yucca-specific problem. Instead, hazardous waste facilities would benefit from being able to assess the performance of a particular sensor before it is deployed, potentially reducing the costs associated with sensor extraction and waste disposal.

Approach - The objective of this project is to develop a sensor-scoring methodology for long-term deployment of sensors in hazardous environments. Development of the sensor-scoring methodology begins by first identifying the important sensor performance metrics (characteristics) deemed critical to long-term deployment of environment-monitoring devices. The research team will then select a subset of sensor classes from the monitoring requirements for initial development of the sensor-scoring methodology to ensure that the methodology is general in scope. The research team will then begin the task of collecting relevant data to be utilized by the scoring methodology. The proposed sensor technology and classification study will establish a database of sensor performance characteristics both in and out of the sensor-operating range. These sensors will hail from regular industry and nuclear environments.

Upon completion of the sensor technology and classification study, the research team will either create a truly novel sensor scoring methodology or alter existing methodologies. The scoring methodology will then be validated with a sensor that has known degradation over time. It is anticipated that historical data for this sensor will be acquired from prior temperature or radiation aging studies.

Accomplishments - While this project was terminated before completion of all project goals, significant progress was made toward development of a sensor-scoring methodology. A sensor requirements definition was developed and was used as a guideline for later data collection and methodology-development tasks. Significant efforts were spent conducting a sensor technology and classification study, although this study eventually proved inadequate. A draft methodology was developed and underwent several iterations, but was ultimately terminated because of lack of supporting data from the classification study.

The draft sensor-scoring methodology that was developed is comprised of two qualitative scoring functions. The first is derived from commercially available reliability software, in which the user-supplied sensor performance data and environmental conditions are analyzed for reliability based on a fault-tree model that matches the user-specified sensor components and construction. Existing commercial reliability analysis software treats the entire sensor as a reliability model and is not capable of addressing degradation of sensor-performance characteristics, such as accuracy, sensitivity, and so on. Augmentation of the reliability analysis is necessary to address the full scope envisioned for the sensor-scoring methodology. 

The second component of the methodology involves the rating of sensor performance metrics (e.g., accuracy, sensitivity) based on order of importance and comparison against historical sensor degradation data. Though the research team conducted a sensor technology and classification study to obtain relevant historical data, insufficient data were obtained to support this methodology component. This lack of data was the primary cause for terminating the project.

Further development of the methodology should be delayed until such data becomes available or can be obtained from future work that involves the generation of data from experimental aging studies. If this work is not feasible, then the methodology can be altered to consist of a reliability assessment of the entire sensor assembly, where the sensor materials have known physical degradation from accelerated aging studies. It is also possible to construct custom modules for use with commercially available reliability analysis software to assess the degradation of individual sensor metrics. This methodology will allow the user to account for degradation of specific sensor characteristics deemed critical by the user.

2005 Program Home