November-December 2018 NPJ

Product & Service Directory–2019 www.NuclearPlantJournal.com 77 Clogging at either the strainer or in-vessel fuel channels can result in not meeting the Emergency Core Cooling System long-term cooling acceptance criteria required by 10CFR50.46. STPNOC’s risk-informed approach was truly innovative and creative. The new methodology is used to address the complexities of the physical phenomena and time-dependent progression of the long-term cooling issues associated with GSI-191 which greatly simplified implementation by STPNOC. Throughout the evolution of the project, these models were developed and refined to provide the necessary information. The technical design for the new framework was developed by STPNOC in collaboration with the University of Texas at Austin. A new method for time-dependent Emergency CoreCoolingSystem(ECCS)performance in the presence of fiber was developed at Los Alamos National Laboratory and further developed by Alion Science and Technology. ABS Consulting developed the GSI-191 PRA model methodology for the initial detailed model approach. Texas A&M University developed and validated a new coupled Reactor Coolant System (RCS) /Reactor Containment Building(RCB)thermal-hydraulicsmodel that became essential to success as the project matured. Other stakeholder plants collaborated with STPNOC to develop the risk-informed process. The project team created a software program (CASA Grande) central to the effort’s success. CASAGrande is a detailedmodel of RCB pipe break, debris-generation, and debris propagation phenomena. Also part of this process was the creation of a Computer-Aided Design (CAD) model of the RCB structures and commodities that models the location and geometry of welds, insulation, coatings, equipment, and robust barriers such as concrete structures. CASA Grande reads the CAD file and develops a spatial representation of each particle of debris material in theRCB. Eachdebris typehas a particularZoneof Influence (ZOI) defined as the spherical volume centered about the break site inwhich thefluid escaping from the break has sufficient energy to generate debris from the particular target material (insulation, coatings, or other materials). The spatial distributions are examined for various break sizes and break orientations at each weld location to calculate the type and amount of debris generated and transported for each scenario. To facilitate resolution of this issue in a timely and cost-effective manner, STPNOCproposed an innovative stream-lined risk-informedmethodology. The methodology, named Risk over Deterministic (RoverD), categorizes scenariosasrisk-informediftheyproducea debrisamountthatexceedstheamountused in a conservative test that gives acceptable strainer head loss results. Those scenarios that do not exceed the tested amount are categorizedas deterministic.This newand innovative method significantly reduces the complexity and scope of scenarios that require closer scrutiny to only the ones that are risk-informed. The RoverD process uses the amount of fine fiber debris from Low Density Fiberglass (LDFG) insulation from a successful, conservative, deterministic strainer head loss test as the criterion for deterministic success. The process relegates to core damage the break sizes that generate more fine fibrous debris (fine fiber amounts tested bound all other debris types). NUREG 1829 was used to determine the break frequency of the smallest break that results in core damage;thatfrequencywasconservatively identified as the change in Core Damage Frequency. RoverD uses an additional innovative tool (Risk Unified Frequency Functional) to evaluate combinations of break models (both Double Ended Guillotine Break and continuum). STPNOC evaluated the scenarios initially deemed deterministically acceptable to confirm that there are no downstream (in-vessel) effects for smaller breaks. Medium and Large cold leg breaks (CLB) were examined with the Fiber Diffusion Operation Engine tool to evaluate the time-dependent fiber accumulation on the ECCS strainers and in the core using different assumptions of bounding debris amounts, ECCS pump configurations, and models of ECCS strainer fiber capture efficiencies. Medium and Large hot leg breaks (HLB) are evaluated using RELAP5-3D thermal hydraulic simulations that showed adequate long-term core cooling for less than 16 breaks. Breaks that are larger than 16 are categorized as risk-informed in RoverD. The deterministic evaluations were made using the guidance in NEI 04- 07 Pressurized Water Reactor Sump Performance Evaluation Methodology. The risk results using this newandcreative methodology were found acceptable for STP Units 1&2 by the NRC for being within RG 1.174 Region III (very small) while maintaining adequate defense-in- depth and safety margin. Radiation Protection Safety Substantial radiation dose to plant workers was avoided. The estimated occupational dose for work to replace/ modify thefibrous insulation in the reactor containment building was estimated to be 176 rem total for both STPUnits. This is a factor of four higher than current industry dose/cycle. More importantly, this would be an actual dose to real people, not a hypothetical accident dose from an event now shown not to contribute significantly to nuclear safety risk. Nuclear Safety Because the analysis associated with the RoverD methodology shows that the effects of debris make no significant contribution to risk, its use will enhance plant safety by allowing operators to focus their resources and efforts on the systems and components most important to plant operations. The application of the methodology maintains the Operators’ abilitytoprioritizemaintenanceandability to allocate resources more efficiently and effectively. The debris-specific Technical Specification action also avoids unnecessary plant-mode changes that might otherwise be required for potential debris-related issues that would drive a plant shutdown. (Continued on page 78)