|Abstract or Summary
- Improved methods of determining temperature transients in reactor systems are
desired because of recent interest in Pressurized Thermal Shock (PTS) issues.
The research presented herein was performed in support of the Nuclear Regulatory
Commission's effort to re-evaluate its existing PTS rules. These rules are
particularly important to the re-licensing of aging nuclear power plants. The much advanced
computational power available to industry may offer a tool that allows
the accurate calculation of temperatures inside the reactor vessel while not being
inaccessibly expensive. It is proposed that an off-the-shelf Computational Fluid
Dynamic (CFD) code, STAR-CD, can be a competitive tool in solving the thermal
hydraulic domain of a reactor system. A comparison of the methodology and
accuracy of the code types that have been previously used in PTS and one that has
not been used extensively, CFD, is provided.
A review of the literature shows that computer codes have been validated
for solving PTS scenarios. The highly specialized program, REMIX, has been
utilized extensively from 1986 to 1991 to interpret accident scenarios in reactor
systems. Other programs are also available that can calculate downcomer
temperatures including system and CFD type codes. Three codes representing the
three different types of programs available are described in detail in the literature
Data appropriate for assessing a program's ability to calculate the response
of a system to a PTS scenario is available from the current matrix of PTS tests
being completed at the APEX-CE facility of the Oregon State University Nuclear
Engineering department. The facility is a reduced scale integral test facility
originally built for modeling the then-proposed AP-600 plant designed by
Westinghouse. For the current test series, the facility was modified to model the
Palisades nuclear power plant, a Combustion Engineering Pressurized Water
Reactor (PWR). Two of the tests were chosen for their PTS typical conditions to
compare with calculations of STAR-CD, REMIX, and RELAP.
The computer models in each of the programs were either created, modified
from a previous version, or the calculations for the comparisons were contributed.
The downcomer temperatures at several locations and cold leg temperature
gradients, where available, were extracted from the data and calculations and
compared. Comparisons are presented in chapter 5 with graphs, along with some
interpretation of the comparisons. It was found that STAR-CD agreed best with the
data set in the downcomer and is the only program that calculated the temperature
gradient in the cold legs. The agreement of STAR-CD with the cold leg data is also
very good. REMIX and RELAP calculations agreement with data for downcomer
temperatures are found to be good for all comparisons made, qualitatively more
than quantitatively when contrasted with the STAR-CD calculations.