Development of Advanced Direct Perception Displays for Nuclear Power Plants to Enhance Monitoring, Control, and Fault Management
B. G. Jones,* N. Moray,* P. Sanderson,* S. Shaheen, D. Reising
U.S. Department of Energy, DE-FG02-92ER75781; University of Illinois (In conjunction with the Department of Mechanical and Industrial Engineering)

A direct perception interface (DPI) integrates information into a unified animated diagram that supports fault diagnosis more strongly than conventional displays. Building on our earlier work on a DPI for nuclear thermal hydraulics, this project will lead to a complete suite of DPI displays for an entire nuclear plant, from nucleonics to power generation. It will also take into account the need for teams of operators to extract different types of information from a DPI. A special case of reactor start-up has been examined that demonstrates the effectiveness of the display.

Infrastructure Support for Nuclear Engineering Research and Education
B. G. Jones,* J. F. Stubbins*
Commonwealth Edison Co.; U.S. Department of Energy, DE-FG02-95NE38111

This multiyear project has supported departmental infrastructure including: partial funding of the TRIGA digital console upgrade, startup funding for three new professors to establish new experimental facilities on the TRIGA reactor: a small-angle scattering and reflectometry system for material studies, thermal/epithermal neutron beam system for medical diagnosis and treatment, and a temperature-controlled neutron irradiation facility in the central thimble location in the TRIGA. Instrumentation and computational equipment have also been acquired for upgrading existing and equipping new undergraduate nuclear engineering laboratories, as well as the student computing facilities in the department. Computational equipment to provide startup for a new faculty member was also provided.

Single-Purpose and Dual-Purpose Nuclear Desalination Plants
M. Ragheb,* A. Siddique
University Illinois

Exergetic efficiency analysis of a dual-purpose (DP) electricity and water production plant is considered. The total system cycle efficiency is the sum of the electrical power efficiency and the water production efficiency weighted by the ratio of the heat addition temperature of the water cycle to the heat addition temperature of the electrical cycle. Analyses show that product streams represent competing processes from the perspective of overall cycle efficiency-i.e., improvement of one process occurs at the expense of the other. Thus, thermoeconomic analysis of a DP plant is considered for optimization of the water-electricity ratio under minimized water costs.

Neural Network Controller for Nuclear Power Plants
M. Ragheb,* S. R. Ray* (Comput. Sci.), O. Uluyol
University of Illinois

This multidisciplinary research is aimed at developing neural network-based power level control strategies for nuclear reactors. A 10th order mathematical model of a PWR is simulated using the MATLAB/Simulink environment. The model is formulated based on point kinetics with six delayed neutron groups and feedback from lumped fuel and coolant temperature calculations. The neural network controller is structured in the form of a local output gamma feedback neural network (LOGF-NN) which utilizes digital gamma memories to provide temporal context in varying time scales. This novel feature lends the LOFG-NN to modeling complex engineering systems where the system is composed of disparate members with varying time constants.

Local Output Gamma Feedback Neural Network
M. Ragheb,* S. R. Ray* (Comput. Sci.), O. Uluyol
University of Illinois

A theory is introduced for a multilayered local output gamma feedback neural network (LOGF-NN) within the locally recurrent globally feedforward neural network paradigm. It is developed for the classification and prediction tasks for spatiotemporal systems and allows the representation of different time scales through the incorporation of a digital gamma memory. As a demonstration, it is applied to the benchmark problem of sunspot series prediction and is compared to other neural network (weight elimination neural network) and statistical (linear and threshold autoregressive) methods. Overall, the proposed LOGF-NN approach's performance is comparable to the TAR method and outperforms the WNET approach.

Energy Economics
C. E. Singer,* M. Khanna
U.S. Department of Energy, DE-FG03-97SF21281

Energy economics studies related to the future of nuclear energy in South Asia and elsewhere include (1) calibration of models of demography and economic development, (2) analysis of regional resources of uranium and other depletable fuels, (3) development of regionally disaggregated models of the competition between various depletable and renewable energy resources, (4) examination of prospects for conclusion and implementation of agreements on greenhouse gas emissions, (5) examination of nuclear waste disposal in various countries, and (6) implications for the future of nuclear energy and nuclear fuel cycles.

Global Management of Weapons-Usable Special Nuclear Materials
C. E. Singer*
University of Illinois

Institutional factors that influence global cooperation on management of special isotopes used in nuclear weapons are under investigation with emphasis on: (1) interaction of civilian and military programs in India and Pakistan, (2) Sino-Indian relations, (3) prospects for correlating agreement between the U.S. and Russia on holdings of nuclear explosives with policies and plans of other self-declared nuclear weapons states, and (4) policy and technical questions related to building confidence in the absence of assembled nuclear explosives from prohibited areas.

Study of Contaminant Transport in Geo-Media in the Context of Underground Storage of Waste
R. Uddin,* A. J. Toreja, S. W. Tam (ANL)
Argonne National Laboratory, DOE ANL 962262404

One of the main concerns involved with the design of a high-level waste repository is the leakage of radioactive contaminants to the environment through groundwater flow. To examine the effects of such leakage, a model which will simulate fluid flow through porous media of varying heterogeneity is currently under development. A discretized form of the convective-diffusion equation will be used to describe the contaminant transport. Simulations using Monte Carlo renormalization group methods will be performed.

Subcooled Boiling Analysis in Pressurized Water Reactors
R. Uddin,* F. Wang, D. Rock
Duke Power Co.

Deposition of boron (negative reactivity) in crud formed on fuel elements has been suggested to be the reason for the discrepancy between the predicted and measured axial power distributions in reactors cores. Formation of crud is directly related to the subcooled boiling that occurs in the top portions of most pressurized water reactor cores. This project aims at establishing the link among subcooled boiling, crud formation, and the deviation between measured and predicted axial flux distribution. Means of reducing the boron deposition in crud will be explored in the second phase of the project.

Subchannel Analysis Using CFD Code
R. Uddin,* D. Rock
Duke Power Co.

Significant safety margins are imposed on reactor operations because computer codes used to design and simulate these systems are often based on "lumped" analyses. We are performing CFD calculations to determine the subchannel details of the velocity and temperture fields in a fuel rod bundle. This will help us identify conservative operational boundaries that can be relaxed.