1 Fall 2015 CAMP Meeting he Fall 2015 CAMP Meeting was held November 4 th through the 6 th in historic Gettysburg, PA. The CAMP members were welcomed to the meeting by Stephanie Coffin of the USNRC. Ms. Coffin thanked the CAMP participants for attending the meeting and acknowledged the success of the CAMP program, which has been operating for roughly 20 years. She discussed the importance of the CAMP program along with the benefits to both CAMP members and NRC. The new Office Deputy Director is Michael Weber who replaces Brian Sheron, and Ed Hackett who replaces Steve West. Ms. Coffin also discussed Project AIM 2020 which is designed to improve efficiency of operations. She provided the NRC’s workload forecast through 2020 along with projections of the NRC’s workforce competency areas. Donna-Marie Sangimino provided an update on the International Programs Team (IPT), and introduced the (Continued on page 2) TRACE User Problems his is a report on TRACE user problems and resolutions. Open trouble reports identified in the previous newsletters with no new progress are not discussed, but can be found in the TRACE trouble report system (TRACEZilla) on the NRC Codes website (https://www.nrccodes.com). For the time period between January 1, 2015 and December 31, 2015, 33 new trouble reports were submitted to TRACEzilla. 39 trouble reports were either resolved or closed during this same time period. As of December 31, 2015, there were 739 trouble reports in the TRACE bug reporting system. Three of those are for the PARCS code, two are bugs for the AVScript, 14 are bugs entered to test the system, and one bug number (452) was skipped. There were 719 TRACE-specific bugs in the system at the end of the reporting period; 33 of those were open, with the remainder resolved, closed, (Continued on page 4) T T Sponsored by the U.S. Nuclear Regulatory Commission Volume 16 / Number 1 / January 2016 NEW EMAIL FOR TRACE BUGS: Any new TRACE bugs can now be reported to the following email address: [email protected]Thermal-Hydraulic Code News Committed to the support of the T/H Codes User Community Contents: Fall 2015 CAMP Meeting TRACE User Problems Recent RELAP5 User Problems Upcoming Spring CAMP Meeting TRACE Validation Other Items of Interest Website: https://www.nrccodes.com
Transcript
1
Fall 2015 CAMP Meeting
he Fall 2015 CAMP Meeting was held November 4
th through the 6
th in historic Gettysburg, PA. The
CAMP members were welcomed to the meeting by Stephanie Coffin of the USNRC. Ms. Coffin thanked the CAMP participants for attending the meeting and acknowledged the success of the CAMP program, which has been operating for roughly 20 years. She discussed the importance of the CAMP program along with the benefits to both CAMP members and NRC. The new Office Deputy Director is Michael Weber who replaces Brian Sheron, and Ed Hackett who replaces Steve West. Ms. Coffin also discussed Project AIM 2020 which is designed to improve efficiency of operations. She provided the NRC’s workload forecast through 2020 along with projections of the NRC’s workforce competency areas.
Donna-Marie Sangimino provided an update on the International Programs Team (IPT), and introduced the
(Continued on page 2)
TRACE User Problems
his is a report on TRACE user problems and resolutions. Open trouble reports identified in the previous newsletters with no new progress are
not discussed, but can be found in the TRACE trouble report system (TRACEZilla) on the NRC Codes website (https://www.nrccodes.com). For the time period between January 1, 2015 and December 31, 2015, 33 new trouble reports were submitted to TRACEzilla. 39 trouble reports were either resolved or closed during this same time period. As of December 31, 2015, there were 739 trouble reports in
the TRACE bug reporting system. Three of those are for the PARCS code, two are bugs for the AVScript, 14 are bugs entered to test the system, and one bug number (452) was skipped. There were 719 TRACE-specific bugs in the system at the end of the reporting period; 33 of those were open, with the remainder resolved, closed,
(Continued on page 4)
T
T
S p o n s o r e d b y t h e U . S . N u c l e a r R e g u l a t o r y C o m m i s s i o n
Volume 16 / Number 1 / January 2016
NEW EMAIL FOR TRACE BUGS:
Any new TRACE bugs can now be reported to the following email address: [email protected]
Thermal-Hydraulic Code News
Committed to the support of the
T/H Codes User Community
Contents:
Fall 2015 CAMP Meeting
TRACE User Problems
Recent RELAP5 User Problems
Upcoming Spring CAMP Meeting
TRACE Validation
Other Items of Interest Website: https://www.nrccodes.com
from page 1 team members. She discussed the mission of IPT and described the benefits and results of international cooperative research. Ms. Sangimino discussed final provisions of Article VII, which addresses how agreements will be extended. An implementing agreement may be extended for an additional period of time upon mutual agreement of the parties. 9 months prior to expiration, IPT will inquire about interest in renewing with the goal being to put in place a timely renewal. 3 months prior to expiration, IPT will exercise Article VII and request a one year extension, with the renewal occurring as soon as possible thereafter. In addition, Ms. Sangimino addressed changes in the agreement invoicing. IPT has historically been sending invoices for membership dues. Due to internal changes, the NRC Office of the Chief Financial Officer (CFO) will now be sending invoices. There will be no letter, but instead a one page invoice that references the Agreement. Contact information for the IPT team was provided. Mr. Thurston stated that the general and TPC meeting minutes from the 2015 Spring CAMP Meeting were loaded onto NRCCodes.com and were transmitted to members via email. Dr. Bub Dong Chung motioned to accept the meeting minutes. Mr. Sean Roshan seconded the motioned, and the minutes were accepted.
The first technical presentation was given by Dr. Chris Hoxie of the USNRC. Dr. Hoxie gave a presentation on the Status of NRC Code Development. CAMP related contacts at the NRC were provided as follows:
RELAP5: [email protected] Dr. Hoxie described the major elements of the “Reactor Core and System Analysis Code Suite”, and provided a summary of improvements. The codes included in this suite include SNAP (platform/graphical user interface), Triton/Polaris/Helios/CASMO and GenPMAXS (cross section library generation), PARCS & PATHS (core physics and steady state thermal hydraulics) and TRACE (reactor system thermal hydraulic analysis). Dr. Hoxie provided an overview of NRC/ISL presentations at this meeting. He also presented some information on CTF, a two-fluid, three-field subchannel analysis code. CTF was originally developed for USNRC by PNL in the early 1980’s as COBRA-TF. It was recently modernized as part of the DOE/CASL program and was renamed CTF. NRC plans to use CTF to benchmark calculations and to assist implementation of a droplet field in TRACE. The CTF code maintenance & user group is managed by North Carolina State University. CAMP members wishing to use CTF are strongly encouraged to join the CTF User’s Group free of charge. Contacts at NCSU include: Prof. Maria Avramova: [email protected] Prof. Kostadin Ivanov: [email protected] Lastly, Dr. Hoxie reminded CAMP members that TRACE is the future of NRC safety systems thermal hydraulic analysis. To assist in NRC near term development efforts for TRACE, he would like member assessments on five key areas: (1) fuel rod models, (2) spacer grid models, (3) pre-CHF droplet field void fraction predictions, (4) testing of new features, e.g., higher order numeric, and (5) additional integral tests. Next, Mr. Christopher Murray gave a presentation on TRACE Code Development Status. Mr. Murray indicated that future focus will be on bug fixes, modeling improvements, and shortening the release cycle. He showed a summary of trouble reports being tracked, and indicated that the current developmental version is 5.960. He then reviewed major changes implemented since the last CAMP meeting, i.e., from developmental version 5.910. He concluded by discussing some of the future plans for TRACE. Mr. Doug Barber then reported on the RELAP5/MOD3.3 status. The latest full release is RELAP5/MOD3.3Patch04, which is Version 3.3iy, October 2010. The most recent developmental version
is 3.3kk, which fixes several bugs and includes new features such as the improved wall condensation in presence of noncondensable, addition of Methane to the list of noncondensable gases, and fluid option for specifying relative humidity. Mr. Barber stated that users should submit their RELAP5 user problems to either ISL or NRC directly. The user problems should not be posted to the NRCCodes.com site, since this is just a list tool and not a true bug reporting feature. Any bugs added directly to the Sharepoint site won’t be seen unless someone from ISL logs into the site and checks it. Therefore, if users do post bug reports to the website directly, they should then follow up with an email to Mr. Barber ([email protected]). A table showing the code names, version numbers and release dates for all RELAP5/MOD3.3 releases was provided. Six user problems that were resolved were discussed and one user problem that is in-work was also discussed. A complete list of user problems from 1998 to date is posted on the NRCCodes Sharepoint site (https://www.nrccodes/com). In addition, 4 new code versions have been created since the Spring 2015 CAMP Meeting, and these were described. Lastly, the update RELAP5 priority list was presented. Dr. Nathanael Hudson gave a presentation on the status of PARCS. He first discussed the recent PARCS/PATHS progress, and then provided some detail on PARCS development. He then discussed some PARCS assessment activity and concluded with the current development activities. Mr. Chester Gingrich reported to the group on the SNAP development status and updates. He began by reporting on SNAP “Core” updates and status. This was followed by a discussion of the animation changes, TRACE plug-in updates, TRACE Multi-Chan component, TRACE initial condition extraction in job streams, PARCS plug-in updates, RELAP5-to-TRACE plug-in updates, SCALE plug-in updates, and TRACE axial plots with AptPlot.
Information about features and bug fixes in SNAP releases can be found at: CORE changes: https://www.appliedprog.com/snap/docs/changes.jsp TRACE Plug-in changes: https://www.appliedprog.com/snap/plugins/trace/changes.jsp To report issues, feature requests, bugs, go to: https://www.appliedprog.com/snap/docs/support.jsp (or simply email [email protected]) To review previously reported issues and their status, go to: https://www.appliedprog.com/issues/index.jsp Additional presentations included:
Mr. Seppo Hillberg, VTT Technical Research Centre of Finland, Ltd., “CAMP Activities in Finland”
Dr. Andrej Prošek, Josef Stefan Institute, “Status of CAMP Activities in Slovenia”
Ms. Dulce Mejia, Comision Nacional de Seguridad Nuclear y Salvaguardias, “Status of CAMP Activities in CNSNS – Mexico”
Dr. Abdelkrim Sekhri, Kernkraftwerk Leibstadt AG, “Leibstadt Status CAMP Activities (Switzerland)”
Mr. Sean Roshan, Royal Institute of Technology, “Member Country Status Report on CAMP Activities in Sweden”
Dr. Ernest Staron, National Atomic Energy Agency (PAA), “Activities Related to CAMP in Poland in 2015”
Mr. Yuriy Vorobyov, SSTC NRS, Ukraine, “SSTC Experience of Modeling and Accident Sequences Analysis using TRACE Code”
Ms. Aeju Cheong and Dr. Kyung Won Lee, Korea Institute of Nuclear Safety (KINS), Korea,
“Assessment of Bennett’s Heated Tube Tests using TRACE and RELAP5 Codes”
Dr. Hyoung Kyu Cho, Department of Nuclear Engineering, Seoul National University, Korea, “Assessment of Wall Film Condensation Model in the Presence of NC Gas using RELAP and TRACE”
Scott Elkins, U.S. Nuclear Regulatory Commission, “TRACE Applications from a US NRC Perspective”
Hao-Chun Chang, Institute of Nuclear Engineering and Science, National Tsing Hua University, Taiwan (R.O.C.), “Fuel Rod Performance & Uncertainty Analysis during Over-Pressurization Transient of Kuosheng Nuclear Power Plant with TRACEFRAPTRAN Dakota Codes in SNAP Interface”
Hao-Chun Chang, Institute of Nuclear Engineering and Science, National Tsing Hua University, Taiwan (R.O.C.), “RELAP/MOD 3.3 Model Assessment and Hypothetical Accident Analysis of Kuosheng Nuclear Power Plant with SNAP Interface”
Duho Hong and Hansik Jung, Doosan Heavy Industries and Construction, Korea, “Steam Line Break Analysis using RELAP5 / MOD 3.3 Patch 4 for ATLAS Steam Generator SLB Test”
Hansik Jung and Duho Hong, Doosan Heavy Industries and Construction, Korea, “Feedwater Line Break Analysis using RELAP5 / MOD 3.3 Patch 4 for ATLAS SG FWLB Test”
Dr. Tomasz Kozlowski, University of Illinois, USA, “Inverse Uncertainty Quantification of TRACE Physical Models using Bayesian Analysis”
Dr. Peter Yarsky, US NRC, “Treatment of Bounding Mixing Uncertainty for BWR/5-6 Standby Liquid Control System Injection During ATWS”
Aleksandar Delja, Canadian Nuclear Safety Commission, Canada, “Research Reactors – NRU Independent Thermal-hydraulic Safety Analysis”
Dr. Steve Bajorek, US NRC, “TRACE Simulation of LTC with Core Inlet Blockage and Boric Acid Accumulation”
Dr. Joseph Staudenmeier, US NRC, “Using TRACE for Research Reactor Safety Analysis”
Dr. Fulvio Mascari, ENEA, Italy, “Analyses of TRACE/PARCS Coupling Capability by Simulating a MSLB Transient”
Dr. Guido Mazzini, CVRez – TSO Group, Czech Republic, “PARCS/TRACE Coupling for VVER 1000”
Dr. Surik Bznuni, Nuclear and Radiation Safety Center of Armenian Nuclear Regulatory Authority,
Armenia, “Preliminary Results of PARCS 3D Depletion Model Verification for WWER-440 Reactor”
TRACE User Problems …continued from
page 1 or duplicates. This implies that ~95% of the TRACE trouble reports have been resolved or closed or duplicates. A resolved trouble report indicates that it has been addressed with an update or documentation change that is pending. A closed trouble report indicates that it has either been addressed by successfully re-running the test problem(s) of interest with a current version of TRACE or it has been closed by incorporating a pending update into the NRC developmental version of TRACE. A pending update implies that it has been tested, documented, reviewed, and submitted to the NRC, but has not been included in the developmental version of TRACE. Inclusion of a pending update into TRACE typically requires additional review and testing by the NRC staff. The most current developmental version of TRACE is Version 5.970 as of December 31, 2015. Updates included in Version 5.960 are listed on the TRACE user support web site (https://www.nrccodes.com).
New Trouble Reports Still Open Trouble Report 713 – HEMSound can fail when pressure > pCrit if liquid density = gas density. Developer reported that HEMSound failed when pressure was greater than critical pressure and liquid phase density = gas phase density. When pressure is above the critical pressure the TRACE steam tables return liquid and gas phase properties that same. Depending upon the path as the pressure goes above the critical pressure, TRACE will attempt to force any two phase flow to single phase liquid or single phase gas. For choked flow where the upstream conditions are above the critical pressure, then flow at the choking plane may be two-phase. HEMSound needs to be modified so that the coding supports liquid density equal to gas density upstream from the choking plane and two-phase at the choking plane. This trouble report is not resolved. Trouble Report 717 – Data exchange between reflood HS and reflood fluid component not general.
Reflood fuel rod HSs calculate quench front locations, which is used by the appropriate fluid component to determine consistent flow regimes. Currently, the reflood fuel rod HS that will be used by the fluid component for quench front location determination is identified in the fluid component NHSCA array. This implies that if there is more than one reflood fuel rod HS component with the fluid component fluid cells, then only the HS identified by the NHSCA array will be used to determine the quench front location for fluid component flow regime determination. A more general approach would be to calculate the average quench front location based on all of the fuel rod HS associated with the fluid component fluid cells. It is proposed that the average be based on the number of actual fuel rods represented by the fuel rod HSs within the fluid component fluid cell. This trouble report is still open. Trouble Report 730 – Control rods marked as stuck are not actually stuck when using PARCS. The problem was noted while performing a coupled TRACE/PARCS calculation that involved control rod withdrawals. Stand-alone PARCS calculations then showed that PARCS 2.7 performs as expected, where a case with “stuck” rods shows higher core power than the case with all rods active. PARCS 3.1 showed the same power vs. time for the “stuck” rod case as the active rod case. This trouble report is still open. Trouble Report 734 – Divide by zero error in PowerTask resulting from logic error in BWR power modeling. Developer reported an unexpected divide by zero error when attempting to model parallel uncoupled flow channels, one a CHAN and the other a VESSEL with a HTSTR (simulating a PWR channel). The error results from conflicts between the point kinetics, moderator heating and reactivity feedback models for the two paths. The resolution is not to simply restrict models to only one type of heat structure arrangement, since it is possible to have a CHAN component in a BWR model that also includes HTSTR components to simulate feedwater heaters. A fix for this problem has been developed that includes additional error checking that causes the code to fail if one or more CHAN components and a separate HTSTR component have active kinetics, moderator heating, or feedback in a single model. The update TR734 fixes this problem and will be submitted to the NRC when completed. This trouble report is still open. (As of January 20, 2016, it has been submitted to the NRC and marked resolved). Trouble Report 735 – Simple trip does not include non-zero delay time.
A user reported that the simple TRIP ignores non-zero delay times. The simple TRIP was added to TRACE to simplify conversions between RELAP5 and TRACE. The RELAP5 trip does not support a time delay, and the TRACE documentation does not clearly state that the time delay is not available for the simple TRIP. Adding a delay to the simple TRIP would have the effect of adding functionality that was not intended and is available using native TRACE TRIP types. The manuals have been revised to clarify that the delay time is not available for the simple TRIP, and input checking has been added that provides a warning message when a delay is included. The update TR735 fixes this problem and will be submitted to the NRC when completed. This trouble report is still open. (As of January 20, 2016, it has been submitted to the NRC and marked resolved). Trouble Report 737 – Built-in D2O property error. The built-in D2O property fits were not intended to extend to pressures above 10 MPa. The TRACE code now allows the use of fluid property tables that are included in a separate file. Use of D2O property tables that include properties above the 10 MPa is the preferred solution to this bug. The D2O property fits have been removed from the code, and appropriate warning messages and input checks have been added. The update D2Ofit fixes this problem and will be submitted to the NRC when completed. This trouble report is still open. (As of January 20, 2016, it has been submitted to the NRC and marked resolved). Trouble Report 738 – V5 patch 4 sensitivity coefficients not functioning. Testing of V5 patch 4 found several sensitivity coefficients that either did not work at all or affected the solution in unexpected ways. This trouble report is open. Trouble Report 739 – Separation is not properly achieved for all cases with a SEPARATOR component. This bug report is related to CAMP action item number 56. A simplified TRACE model designed to show the problem generated incorrect gas and liquid flows at the separator component outlets. This trouble report is still open.
Trouble Reports Closed During the time period from January 1, 2015 to December 31, 2015, the following trouble reports have been closed. A trouble report is closed by: 1) rerunning the test problem of interest successfully with the latest version of TRACE, 2) updating documentation to address the trouble report, 3) incorporating an update
Thermal-Hydraulics Codes News January 2016
6
into the NRC development version of TRACE. In some cases a combination of these three fixes may be used. Trouble Report 705 – Fuel pin average temperature used in TRACE internal fuel densification model in error when fine mesh is turned on. When name list input CALCSWELLDENS is true, then the TRACE internal fuel densification model is used to calculate fuel densification, which depends upon fuel temperature. The fuel pin average temperature calculation used in the fuel densification model was in error when fine mesh was turned on for the fuel rod HS. Update TwoDBurn fixes this logic error and was included in version 5.910 and closed this trouble report. Trouble Report 706 – Error in fuel rod to cladding gap calculation following cladding rupture. Development of a separate update related to graphing gas gap thermal conductivity showed that the gap thickness calculation was in error following a cladding rupture. The gas gap thickness is never allowed to go to zero, since the minimum gap size is the sum of the cladding and fuel roughness, which must be larger than zero..An error in the logic for calculating the gap heat transfer resulted in the gap relocation model adjustment being applied to the old gas gap thickness, but neglected the plastic and thermal deformations. The gap was then closed after a number of timesteps. TwoDBurn fixes this problem and was included in version 5.910, closing this trouble report. Trouble Report 707 – Thermal conductivity evaluated at the wrong spatial location when 2
nd order finite is on.
When the first order finite volume solver is turned on, the thermal conductivity is evaluated at the HS mesh between neighboring nodes. The conductivity is evaluated at an average temperature (between the nodes) using the material type between the nodes. The second order finite element methodology uses thermal conductivity and temperature at three points to develop a 2
nd order function. This function is evaluated at the
midpoint between the nodes to get the heat flux between them. Therefore, the thermal conductivity for each node (not mesh) needs to be evaluated based on the node temperature and the HS material that surrounds the node. The logic for the 2
nd order solver has been
corrected as part of update FxFEK, which went into version 5.951, closing this trouble report. Trouble Report 721 – Error signal assignment in ISVN=21. User reported problems with assigning a pressure signal to a CONTAINMENT component when using SNAP. The
job does not run when the pressure signal is assigned within SNAP, but does run outside the SNAP environment. This trouble report was determined to be a SNAP issue, independent of the TRACE code. No updates to TRACE were made. If it is determined that there is a problem with TRACE or the TRACE documentation, it can be reopened. This trouble report was marked closed. Trouble Report 732 – Reynolds number for annular film wall drag appears to be in error. Wall drag coefficients were added as plot variables as part of a different update. The plot variables uncovered a potential issue when spikes were noticed in the axial liquid phase wall drag when the void fraction was close to one. In annular film flow, the wall drag is based on the liquid phase volumetric flux. Because the liquid phase volumetric flux is used, the Reynolds number can be quite small at high void fractions (near one), even though the liquid phase velocity may be quite high, indicating turbulent flow in the film. This trouble report was added to investigate this issue, and confirm that the correct velocity information was being used when calculating the Reynolds number. Further investigation showed that the usual practice is to use the liquid phase volumetric flux when calculating Reynolds number for film flows. Since the TRACE coding is consistent with the typical experimental treatment, this trouble report was marked closed Trouble Report 736 – Testing for version 5.910_Special indicates non-null backups. Update TR727 was built on version 5.910 plus the updates 1)FxFmodDen, 2) FxPzrAcc, 3) Fix to XTV Auto, 4) PARCS v32m17, and 5) cleanTH. This base version was selected to allow for coupled TRACE/PARCS calculations in the testing of update TR727. The regression testing showed that both 5.910_special and 5.910_special + TR727 failed for a number of null backup calculations. One of the 5 updates (or a combination thereof) that went into the “special” version causes a null backup error. Separate testing for FxFModDen and FxPzrAcc indicates no null backup errors. It was determined that the XTV_Auto update caused the non-null backup errors. The issue arose from logic in the update that did not anticipate a subroutine call with a null input variable. The non-null backup issue was fixed before the XTV-Auto update was included in TRACE. The fixes were not incorporated into the code that was used in the “special” version. The final version of the update was called AUTOBreakBugFix and was included in TRACE version 5.921. This closes this trouble report.
Thermal-Hydraulics Codes News January 2016
7
Trouble Reports Resolved with Updates Pending During the time period from July 1, 2014 to December 31, 2014 the following Trouble Reports have been resolved with updates or documentation modifications pending. These Trouble Reports are awaiting NRC review and a decision to implement or not to implement the associated code updates or document modifications into TRACE. Please note that at the time of publication, these trouble reports may have already been closed. Trouble Report 653 – Power weighting of reactivity feedback parameters. Boron reactivity feedback is dependent on power weighting. The intent of the power weighting is to increase the importance of the boron in the regions where thermal neutron flux is relatively high, and decrease the importance in regions where the thermal neutron flux is low. The power weighting factor in the water rods and the core bypass for BWR is set to 1.0. It may not be correct to use a weighting factor of 1.0 in these regions. The thermal flux in the water rods and the bypass region may be the same or higher than the flux in the water around the fuel rods themselves. In general, the water rod and core bypass should get the same power weighting factor as the water around the fuel rods, and the power distribution should be normalized. Similar changes should be made for the other reactivity feedback parameters and for both BWR and PWR geometry. Update FxPowerWt makes these changes. It also includes some corrections to the units (ppm or density) for the solute (boron) as well as some additional scripts for comparing input files. Update FxPowerWt resolves this trouble report and is pending. Trouble Report 686 – TRACE HS to CONTAN fluid compartment failure. This report was initiated in June of 2014. A user provided a test problem where there was no heat transfer between a TRACE HS and the CONTAN fluid compartment. A logic error was found that caused a negative HTC to be computed. The HTC is ultimately set to zero. Update FxSphericalHSContan resolved the trouble report in August of 2014, but an additional error was found when FindTi is called. Update FxTi resolves this trouble report, and is pending. Trouble Report 690 – User guidelines for HS need to be developed to ensure that users do not develop unstable models. The heat conduction solution is heavily dependent on material type, the geometry of the material, and the nodalization used. If a material with low thermal inertia
(high conductivity, low specific heat) is used for a thin geometry with many radial nodes, the solution can become unstable. The temperature distribution in the material will be flat, or nearly so. As a calculation progresses, small changes in the surface heat flux can cause large changes in the surface temperature. If the temperature gets low enough, the material properties may become invalid and will introduce errors into the calculation, causing it to fail. Update FxBoronDensity fixed the problems caused by negative specific heat values, Improved user guidelines for heat structures that will help users avoid these types of problems have been developed. The guidelines were included in the SMass update, which is pending. Trouble Report 691 – The volume average velocity logic for PLENUM component is in error. During the testing of the NewPlots update, it was determined that the volume average velocity logic for the PLENUM component was in error. For the w4loop test problem in the Regression test problem set PLENUM component 16 has one inlet and four outlet junctions for a total 5 junctions The volumetric flow going through the plenum component is ~15 * 1.149 = 17.235 m**3/s. The cell average flow area is 1.3196 m**2. Therefore, the cell average velocity for the plenum component should be ~17.235 / 1.3196 = 13.06 m/s. TRACE calculated ~5 m/s. The NewPlots update resolves this trouble report and is pending. Trouble Report 692 – Regression test problems BWRMassErrTest has a mass error of ~9% after 1100 time steps. A mass error of 9% for the BWRMassErrTest problems is not expected. The error should be smaller than that. This test problem was re-run with version 5.190 and the mass error was much smaller (5.5E-8%) after 2466 timesteps. This resolves this trouble report. Trouble Report 695 – Initial value of signal variable for single junction. Single junctions were added to TRACE to help with conversion from RELAP5 input. The single junction is essentially a pipe component without any volume, just two edge faces. This bug results when signal variables are used to report the mass flow in a single junction. The mass flows that are reported come from edge 2 of the single junction, instead of edge 1. It was also noted that edge 2 mass flows were set to zero initially, while at the same time, edge 1 mass flows had the expected value. The code computes values for edge 1 and copies them to edge 2. It is unclear why they were different for the user. Update FxSVZero adds code that checks the input to give an error if a signal variable tries to reference
Thermal-Hydraulics Codes News January 2016
8
edge 2 in a single junction. In addition, the logic was modified to allow for the mass flow rate in the single junction to be initialized earlier in the calculation. Update FxSVZero resolves this error report and is pending. Trouble Report 697 – Solubility limit added as a graphics variable. The solubility limit was requested as a graphics variable, but was not part of any fluids data structure. Update FxFillContan adds the graphics variable and appropriate data structures. The update resolves this error report and is pending. Trouble Report 698 – Problem with moderator density feedback reactivity input. The R5FDBK option is set to 1 to activate the RELAP5 reactivity feedback calculation. If that option is active, TRACE should use volume-weighted moderator density in the feedback table. It was found that the core power was responding in the opposite direction as expected, as if the reactivity were opposite in sign from that intended. The cause of the bug was traced to an incorrectly converted reactivity feedback table. The update includes code that performs input checking to catch reactivity tables that have not been correctly converted. Update FxFBModDen resolves this error report and is pending. Trouble Report 699 – Duplicate junction number flag as programming error and should have been input error. A user copied a break and valve in SNAP. Although the break and valve each received new numbers, the junction between them did not. The code generated an error (as expected), but incorrectly identified it as a programming error. The update resolves the issue by correcting the identification of duplicate junction numbers to reflect an input error instead of a programming error. Update FxJunErrMess resolves this trouble report and has been submitted to the NRC and is pending. Trouble Report 700 – Fill mass flow rates in error under some conditions for IFTY = 10. The code correctly calculated feed water FILL velocities, but the mass flow rates were not consistent with those velocities. This can result in problems when the control system uses the mass flow rate as input. The problem resulted from a logic error in the FILL component with IFTY = 10. Update FxFBModDen corrects the logic error. The update resolves this trouble report and is pending. Trouble Report 703 – TRACE error check for gas gap material ID does not work.
A model was failing due to melting fuel. The problem was traced to a gas gap between the fuel and cladding that had two nodes identified as the gap. The input manual indicates that this situation is not allowed. Input checking was in place, though did not function as expected, because the fuel-clad interaction option was off. The input checking was modified to check all heat structures, regardless of the setting of fuel clad interaction. If no fuel-clad interaction is to be modeled, more than one gap can be included, as long as there is solid HS material between the gaps. In addition, input checking was added to require the radial power peaking in the cladding to be zero. This trouble report is resolved by the FxMultGap update, which is pending. Trouble Report 704 – TRACE CHF Table Evaluation for Pool Boiling with Counter-Current Flow. In pool boiling counter-current flow, the vapor boil off is replaced by water that is flowing in. At steady-state, the vapor mass flow is matched by the liquid mass flow, which suggests that the total mass flux should approach zero. However, in the counter-current flow model, it was found that TRACE was using the liquid mass flux when evaluating the CHF limit. The update FxPoolBoilCHF resolves this trouble report and is pending. Trouble Report 708 – Error in conc – solute concentration input description concerning units. The units in the Users’ Guide for the fluid component solute concentration input are incorrectly identified as kg-solute per kg-solution if ISOLUT = 2. In fact, the units are always kg-solute per kg-water, and do not depend on the setting of ISOLUT. The manual documentation has been updated with the update FxFillContan, which is pending. Trouble Report 709 – LBLOCA simulation using side junctions in error. A user reported problems with simulating LOCAs. Results seemed to be inconsistent with expectations. It was determined that the input models were (unknowingly) inconsistent. The desire was to test the impact of variations in L/D on the results of the LBLOCA. It was found that the variation in the L/D values was made by changing the length of the pipe. This had the undesired effect of moving the location of the break further and further from the pressure vessel. The correction is to modify the L/D by renoding the pipe instead of changing the physical distance of the break from the pressure vessel. The TRACE modeling guidelines provide suggestions as to how to perform this type of sensitivity study. This trouble report requires no updates, and will be marked closed.
Thermal-Hydraulics Codes News January 2016
9
Trouble Report 710 – Inverted annular film boiling radiation heat transfer may be accounted for twice. It was discovered that a BWR assembly in annular film boiling with active radiation heat transfer enclosure modeling may account for the radiation heat transfer to the liquid twice. Once for the heat transfer from the wall, and once in the radiation enclosure logic. The code was updated to include a logic check that turns off the radiation heat transfer if the heat structure is part of a radiation enclosure that transfers heat to the coolant with a non-zero path length. This trouble report is resolved by update ZrEmiss, which is pending. Trouble Report 711 – Differences in TRACE Predictions for Two Similar Inputs. Two steady state models were used to evaluate SBLOCA cases. One model included “dead end” pipes in multiple locations to allow for ease in testing multiple LOCA locations using one steady-state case. It was found that the results were different between the cases, in spite of the fact that the dead end pipes should not impact the results. The two cases differed in the location of the dead end pipes and in the hydraulic diameters at the dead end edges of those pipes. TRACE averages the edge hydraulic diameters to calculate the cell hydraulic diameter. The average cell hydraulic diameter is used to calculate the heat transfer in that cell. Even though the dead end cells have a zero flow area, the differences in the hydraulic diameters of the two cases caused the changes in the results. Update FxPzrAcc includes fixes to the logic errors that contributed to the differences. This pending update resolves this trouble report. Trouble Report 712 – Error in format for error message. A format error in one of the error message routines was identified that causes a FORTRAN runtime error. Update HydrogenClad includes the fix for the format error. The update is pending, and this trouble report is resolved. Trouble Report 714 –Spacer grid model implementation for CHAN components does not appear to be complete. The pressure loss for 1D spacer grids was not being calculated. The code was still expecting spacer grid input from the heat structure input, when it should have been from the fluid component input. Update FxGridSpacers resolves this trouble report and is pending. Trouble Report 715 – Stratified flow in pressurizer and accumulators.
Horizontal stratified flow model should not be active for pressurizers and accumulators. This was correct for those component types, but not for pipe components with a type of pressurizer or accumulator pipe. This can result in dead end pipes that can be in horizontal stratified flow for a cell edge with zero flow area. This can affect the solution in the cell adjacent to that edge. Update FxPzrAcc resolves this trouble report and is pending. Trouble Report 716 – Grid Spacer Logic for old ROD HS input not working when NCRX > 1. ROD components spawn additional heat structures during the calculation. If grid spacers are to be used in the spawned heat structures, the user is unable to provide sufficient input, since the numbers associated with the new heat structures are unknown at the time of input. The code was updated to make the numbers for the spawned heat structures available to the grid spacer routines. The GridInp update resolves this trouble report and is pending. Trouble Report 718 – Grid Spacer HT enhancement logic for up flow and down flow. The enhancement logic for the heat transfer following a grid spacer was found to be incorrect. For up flow, the logic correctly calculates the heat transfer enhancement starting at the upper (trailing) edge of the grid spacer. In down flow, the enhancement is still computed from the upper (now, leading) edge of the grid spacer. The enhancement is due to the mixing vanes on the grid spacer, and does not take effect until after the flow has passed the spacer. The GridHT update is pending and resolves this trouble report. Trouble Report 719 – No fluid in cell. Arrays arlS, arvS, arelS, and areyS are initialized to the new timestep macroscopic densities using SetSysMacroDen. The arrays are used to normalize the residials in the mass and energy conservation equations in routine ResidualCheck. During the first timestep after backing up, the right hand side of the mass and energy equations is stored in arrays arlS, arvS, arelS and arevS. This error report results from normalizing the residuals of the mass and energy equations with the wrong values. Update FxSMass resolves this trouble report and is pending. Trouble Report 720 – Helical Coil Model. User requested an update to the vessel model that would allow for modeling bundle crossflow and a reactor core within a single vessel component. It was determined that the current vessel component has this
Thermal-Hydraulics Codes News January 2016
10
capability. The bundle crossflow correlations are only active in user-identified vessel cells. The documentation was not clear on this point, and has been corrected. The updated documentation is included in update FxPowerWt, which is pending. Trouble Report 722 – Inaccurate error message – reference to pGapInit. A user reported an error message that referred to input variable pGapInit. This input variable is not mentioned in the documentation or in the SNAP help information. The corresponding variable name that is actually mentioned in the manuals is PGAPT. The value input for PGAPT is eventually stored within the code as variable pGapInit. The error message was modified to refer to PGAPT and remove the confusion. Similar errors (those not consistent with the variables listed in the input manuals) were located and corrected. The ErrMess update resolves this trouble report and is pending. Trouble Report 723 – Incorrect error message when BREAK is connected directly to a VESSEL. A BREAK component may not be connected directly to a VESSEL component, The error message that results when the BREAK is connected directly to a VESSEL incorrectly identifies the situation as a programming error. The section of coding that identifies the problem is after the input checks are performed. At that stage of the calculation, most errors are programming errors. In order to correctly identify the neighboring component types, the entire input file must have been read in. This occurs slightly after the normal input checks are performed. Coding was added to the ErrMess update that checks for this situation after the neighboring components can be determined. The ErrMess update resolves this trouble report and is pending. Trouble Report 724 – User wanted to get fuel average temperature with fuel rod with no gap and no cladding. The capability to calculate fuel average temperature requires a single fuel arrangement – fuel surrounded by a gas gap and a layer of cladding material. A user desired additional capability to have the code compute an average fuel temperature without the gas gap. Code was included to compute the average fuel temperature based on the nodes that had adjacent mesh cells that were identified as having a fuel material type. If the default fuel type is not used, the user must identify the fuel material number as part of the NAMELIST input. These changes were included in update PlotAvgFRT, which is pending and closes this trouble report. Trouble Report 725 – Material properties causing programming error during restart.
The user manual states that setting the NAMELIST variable NMAT to -1 in a restart will cause TRACE to pull the material data from the dump file. If NMAT is greater than 0, then the code should find that number of materials included in the input file. This was found to be incorrect for some models, where the code was attempting to read the material data from the dump file even when NMAT was not equal to -1. When the input file is in native TRACB format, the correct code response is to attempt to read the material data from the restart file, regardless of the setting of NMAT. The code had a logic error that used an incorrect flag to identify the input file as TRAC-B input. The code occasionally misinterprets the input file as one in TRAC-B format as a result of the incorrect flag. An additional issue was identified in that the TPR data structures are not read in from the restart file before the input check is performed. The documentation also incorrectly stated that the material property data must always be included in the input file, since it cannot be read in from the dump file. These three issues were corrected as part of update TR725, which is pending. The update resolves this trouble report. Trouble Report 726 – Issue reading in material properties in native TRAC-B format. This trouble report is related to trouble report 725. An input deck designed to test the ability of TRACE to read in TRAC-B input files failed due to missing material properties. The failure was incorrect, since the material in question was provided as part of the input. Several additional test cases were generated to test various aspects and options for reading in materials. It was found that the cases run successfully in version 5.960 and version 5.910+TR725. Update TR725 is pending, and this trouble report is resolved. Trouble Report 727 – Hot Rods for coupled TRACE/PARCS calculations is not getting zero power. User indicated that the centerline temperature for a hot rod in a coupled TRACE/PARCS calculation indicated that the rod was getting zero power. Code review indicated that the average rods were getting a power distribution, but there was no logic to provide a power distribution in the hot rods when the rods were powered by PARCS. The logic was corrected in update TR727. The hot rods now use the same power shape that is used for the associated average rod. Update TR727 is pending and closes this trouble report. Trouble Report 728 – Programming Error in Solver. A user reported a case where TRACE indicated that there was a programming error. The development team was unable to reproduce the error, but did note some
Thermal-Hydraulics Codes News January 2016
11
concerns with the input model. The material properties for the rod materials had densities of the solid material that were several orders of magnitude less than what would be expected, and the thermal conductivity values were higher than would be expected. In addition, the radial node sizes are very small for the timestep size that was used. This trouble report was marked as resolved, since the development team could not reproduce the problem. Trouble Report 729 – Error in Adjusted Flow Losses feature of TRACE. A user identified four issues in the code. The first is that the adjusted flow loss controllers do not function during a transient – this is by design, so no changes were required. The user had problems when attempting to restart from a steady state with adjusted flow losses active. If the NADJLOSSES namelist variable is not set (it is not included at all) in the transient input, the code will fail with a run time error. If not included in the restart input, the code reads the value of NADJLOSSES from the dump file, but fails to allocate the flow loss array. The third issue is related to the application of flow losses in VESSEL components. The flow losses are specified for each location in the VESSEL, but the code fails to recognize differences in the axial direction. This has been corrected separately as part of the updates to version 5.960. The fourth issue relates to the use of FRIC values or K factors, and how they are reported. The internal conversions between the two factors were unclear, and the results printed to the plot file were also confusing. During the development of this update, it was found that the 3D radial additive flow loss factors were included twice. Update FxAddLoss resolves the issues with NADJLOSSES and the differences in how the FRIC and K factors are represented and reported. Update FxXRLoss resolves the double counting of additive flow loss factors. Both these updates are pending, and resolve this trouble report. Trouble Report 731 – Vent Valve Input Description in Error. Work on trouble report 729 uncovered an inconsistency in the vent valve flow loss documentation. The documentation indicated that the vent valve flow loss input should be unitless; however, when the values are loaded into the arrays at the appropriate VESSEL edge, the code expects the arrays to be loaded with FRIC values divided by the hydraulic diameter. Thus, the vent valve input is inconsistent with what the code expects. Coding was added to make the vent valve input dependent upon the setting of the IKFAC namelist input. If IKFAC is 0, the vent valve flow loss factors should be FRIC factors and if IKFAC=1, the factors should be K-loss factors. The code was also modified to expect
dimensionless values as input. Some errors were also corrected in the initialization of the uncertainty multipliers on the wall drag. However, these multipliers do not appear to be used in the current version of the code. These updates have been applied as part of update FxXRLoss, which is pending. This trouble report is resolved. Trouble Report 733 – HS input processing for oxLayerAr and burn2D is in error when fine mesh is turned on. Ken Jones reported a problem with the fine mesh extra axial levels at the top and bottom of a fine mesh HS. The fine mesh levels did not have the correct oxide layer values. When fine mesh is turned on for a given HS, the top and bottom HS coarse mesh nodes are split into two nodes and the original input for the first and last HS node is duplicated for the new nodes. A scan of the coding in HtStrInput module and HtStrUtil module indicated that the oxide layer and burn2d arrays were getting adjusted twice. The fix will be to remove the extra node adjusted for the oxLayerAr and burn2d arrays from the HtStrInput routines. Update FxFMOL is pending, and resolves this trouble report.
Recent RELAP5 User Problems
ELAP5 user problems reported or resolved are summarized in each issue of the newsletter. If you encounter a problem with RELAP5, please
report it to [email protected] complete list of RELAP5 user problems, including a description of the problem, status (resolved, in work, on hold or unresolvable) and, if resolved, the manner of resolution, is available on the https://www.nrccodes.com web site. Since the last TH newsletter was published eleven new user problems were submitted. All of these problems were resolved, along with one additional user problem previously submitted. A description of these resolved user problems is provided below in chronological order. No. 2014-05 (8/11/2014), Resolved Code Versions Affected: RELAP5/Mod3.3 Patch04 The user was running LBLCOA cases for the US-APWR using input decks provided by ISL. RELAP5 version 3.3iy (Patch04) is exiting on thermodynamic property failure in an accumulator, which occurs approximately 2 minutes after the accumulators start injecting (at approximately 650 seconds). The user attempted to adjust the time step, but it did not resolve the issue. The case ran successfully using RELAP5 version 3.3id.
Initial Response (08-28-14): It appears update 3.3if introduced the error seen in 3.3Patch04. This version attempted to fix spikes in accumulator pressure, along with spikes in the nitrogen mass, as the accumulator empties. With this update, the saturation temperature follows the sublimation line as the gas temperature drops below the triple point. That's what's happening in the user’s case in which the gas temperature goes to 171K. Resolved (04-15-15): Added coding to nth2x1.ff to limit s(stp) to the minimum pressure value in the steam tables. This was done in order to keep the US-APWR LBLOCA accumulator from failing in a subsequent call to getsnglatpt. No. 2015-01 (3/27/2015), Resolved Code Versions Affected: RELAP5/Mod3.3 Patch04 The HF critical flow model in patch 4 has been corrected as follows: “The gamma = Cp/Cv ratio was hardwired to 1.3 for steam. This was probably done because the programmer did not know how to easily get Cv from Cp and other variables in the steam tables. This was changed so that now Cv is computed and the gamma is now the correct value for steam.” It was first added to the 3.3gx version. The HF critical flow of patch 4 has been increased significantly than the previous at the high pressure steam conditions (~15Mpa). However, the user obtained an incorrect trend of critical flow from the two-phase condition to pure steam flow. The flowrate of pure steam flow has been higher than the two phase flowrate with no apparent reason. Resolved 04-23-15: In version 3.3gx, the value of gamma was changed from a hardwired value of 1.3 to Cp/Cv. The ratio of Cp/Cv is only applicable for ideal gases and shouldn’t have been used for steam. For steam, an isentropic correction factor is needed, which when derived is 1/(kappa*P). Therefore, gamma should be computed as (Cp/Cv)*(1/(kappa*P)). No. 2015-02 (3/27/2015), Resolved Code Versions Affected: RELAP5/Mod3.3 Patch04 The manometer oscillation problem with NC gas, which was one of level 1 QA test problems, has been tested using the old steam table, tpfh2o. The problem ran successfully with RELAP5/MOD3.3/Patch3 (3.3gl) and previous versions, but failed with the Patch 4 version (3.3iy). The patch 4 version stopped with the message "in withair, should not be here". The problem may be caused by the patch4 modification for restrictive convergence criteria of pressure, where the steam partial pressure is very low with NC gas.
Resolved (04-24-15): The user’s problem report was fixed with version 3.3jf. In this version, there was some logic added to the noncondensable iteration that reduced the pressure correction with each successive noncondensable iteration. This logic allowed the user’s problem to run. No further code modifications were required. No. 2015-03 (4/24/2015), Resolved Code Versions Affected: RELAP5/Mod3.3jma The user was trying to model the discharge of argon from a gas bottle. It passes through a short run of pipe and then through a control valve, followed by approximately 500 ft of 1in tubing and a discharge into a 12" Pipe. There’s a time dependent volume upstream of the tank and the valve opens slowly (over 30 seconds). The model includes a 15 seconds steady state period to initialize the model before the valve is opened. A few seconds after the discharge begins, the code fails with a "VapErr1" reason. It appears that small amounts of liquid are being generated as the argon passes through the valve. Resolved (04-24-15): To avoid the error, the user needed to be running as pure noncondensable. The input model was modified to set W5 of Card 120 to "2". This tells the code that the problem is pure NC and to not perform any of the liquid/vapor equations. In essence, it prevents liquid or vapor from appearing when there is none. The user’s case ran out to 90sec following this change. No. 2015-04 (5/18/2015), Resolved Code Versions Affected: RELAP5/Mod3.3ke The user noted that the edit of the HS input for the metal-water reaction from card 1CCCG003 still says “Cathcart/Pawel” is used even if Card 1 Option 81 is specified for “Baker-Just”. In addition, if the user inputs coefficients on 1CCCG003, the edit of the HS input still states “Cathcart/Pawel” is used. The manual discussion of Card 1 Option 81 needs to be clarified to state that the coefficients are modified, not the equations. Lastly, Vol. I, Eq. 4.13-2 has an error in the description of equation variables; it should be dr_n-1, not (dr_n-1)^2. Resolved (05-19-15): Modifed the input edit for HS card 1CCCG003 to display “Cathcart/Pawel”, “Baker-Just”, or “User-Defined”, along with the correct coefficients. Added clarification to the Card 1 Option 81 message printed to the output. Updated the input manual to clarify the use of Card 1 Option 81 and corrected Eq. 4.13-2 in Vol. I.
Thermal-Hydraulics Codes News January 2016
13
No. 2015-05 (6/10/2015), Resolved Code Versions Affected: RELAP5/Mod3.3 Patch04 The user found an error in the funk.ff subroutine. The “s” vector was not dimensioned properly. In addition, the user reported an issue with the code cutting the time step. In the user’s problem the flow velocities are low and courant limitation should not affect the time step. It may have been a limitation due to another process, but it showed as courant (boiling or condensation?). There was also an issue in simulating evaporation from the free surface when the heat is not added to the water. It comes out that Relap is not able to calculate evaporation. The mass of water in the pool does not change. The user tried different combinations of temperatures, humidity or even the equilibrium option. Evaporation is diffusion process driven by humidity gradient, so the user tried the new interphase mass transfer model (option 74), but it did not work. Resolved 06-18-15: Both of the user’s cases run into Courant limit problems when run with version 3.3kg. But they can be made to run successfully simply by attaching junction 4 to the top of cell 15-1 instead of to the side. Even though they run successfully the output is essentially garbage because the mass errors are huge. When one models such large volumes of liquid with heat input to some of them, there is always a potential problem with small amounts of vapor being generated in a big cell so that the void fraction increase is very small and is truncated by the truncation logic in subroutine EQFINL. RELAP5 was never intended to be used to address large volumes as seen in these input files. With regard to the evaporation issue, no evaporation occurs in this test problem because there is no liquid/vapor surface at which evaporation can occur. Cells 16-04, 15-1 and 6-1 each contain dry air and all other cells are all liquid. RELAP5 can only calculate evaporation when a liquid/vapor surface exists WITHIN a cell; i.e., a cell with a void fraction between zero and one. If the input is changed so that X=0.1 in cells 16-04, 15-1, and 6-1 there will be vapor generation occurring in those cells. However, even though RELAP5 calculates evaporation, it does not calculate it correctly – it calculates it to occur very rapidly and unrealistically. The vapor generation model drives the gas phase to have steam at saturation for the temperature in the cell very rapidly. One would have to incorporate some sort of rate constant to model evaporation realistically. No. 2015-06 (6/11/2015), Resolved Code Versions Affected: RELAP5/Mod3.3 Patch04 The user reported a problem trying to get stable initial conditions for moist air: pressure=1 bar, T=306.95 K,
humidity=47%. He tried initialization flags t=4,5,6, but could not get the problem to initialize properly. He also asked how to calculate ug for flag t=6. A calculation like Xs*Us+Xa*Ua=Ug gives slightly superheated steam. The problem is that the parameters are not stable. Also variables usatf and usatg give strange values when plotted. Resolved (06-18-15): The only way to simulate relative humidity in the RELAP5 input is to use initial condition option 6 (P, Uf, Ug, Xs, Xa). The conditions in the deck received from the user were: 50200 106 1.e5 1.415628e5 9.e5 1.0 0.6802721. With this tempf=306.94, and calculated tempg=306.94. These conditions result in the steam being nearly 50 K below the saturation value. For all NC at a relative humidity of about 47% then following input is appropriate: 50200 106 1.e5 1.415628e5 4.09e5 1.0 0.985. But note that if this input is changed so that liquid is present in the cell (Xs<1.0), then RELAP5 will very rapidly evaporate enough water to bring the steam to a saturated condition. One cannot use RELAP5 to simulate evaporation from the surface of a pool of liquid. No. 2015-07 (6/15/2015), Resolved Code Versions Affected: RELAP5/Mod3.3 Patch04 The user was interested in modeling methane in RELAP5, and suggested to modify the R, Co, Do, and Uo values of a gas in Table 3.2-1 (Vol. 1) to match the properties of methane. The user developed the methane data based on the NIST webbook for methane between 200 K and 600 K and between 0.1 MPa and 0.6 MPa. Resolved (08-03-15): This change to add methane gas as a noncondensable was implemented in version 3.3kj. The properties used are as follows: Rn = 8314.3 / 16.043 = 518.251 J/kg-K Do = 3.172778 J/kg-K2 Co = 1575.186 J/kg-K Uo = 278,058.5 J/kg a = 2.34401E-5 b = 1.28372 muo = 1.011E-6 Tref = 165 K No. 2015-08 (6/15/2015), Resolved Code Versions Affected: RELAP5/Mod3.3jz The user discovered a bug in the input procedures. RELAP does not read the laminar shape factor (input record CCC0131) for single volume component. It does read it correctly for the branch components as well as the pipe components. In case of a single volume component the value of the shape factor is interpreted as zero, independently of the value entered in the record
Thermal-Hydraulics Codes News January 2016
14
CCC0131. The entered value is interpreted as the viscosity ratio, which should be the second entry in this record. Resolved (06-18-15): The code was processing the first single volume correctly, but subsequent single volumes had problems. The resolution to this problem involves a modification to the input processing logic for Single Volume Card CCC0131 in rsngv.ff. The number of words on Card CCC0131 had to be reset to 1 prior to reading each single volume input to ensure that the 1st item on the card is stored properly. No. 2015-09 (6/17/2015), Resolved Code Versions Affected: RELAP5/Mod3.3 Patch04 During assessments against wall condensation experiments, RELAP5/MOD3 predictions were usually not successful, and sometimes very low at high noncondensable quality. Through the critical review of wall condensation models of RELAP5, MELCOR, and TRACE, it was found that vapor density used in RELAP5 vapor condensation mass flux has been determined based on the saturation property at steam partial pressure. However the MELCOR vapor density has been determined based on the saturation property at total pressure. The density term derived by the diffusion layer theory from the recent journal paper (P. F. Peterson & T. Kageyama (1993)) is similar to MELCOR. The derived term from TRACE model also show the deficiency of the density term of RELAP5. It is suggested that the density term "rhovb" in function pintfc of RELAP5/MOD3.3/P4 should be determined based on the "saturated vapor density based on total pressure". Resolved (07-22-15): The user’s change to pintfc is correct and was implemented in version 3.3ki. No. 2015-10 (7/8/2015), Resolved Code Versions Affected: RELAP5/Mod3.3 Patch04 The user submitted an input file that failed around 225 seconds. The model was a core spray line break for a BWR with jetpumps. The choked phasic velocities are unrealisticly high at the failure point due critical pressure in the jetmixer volume (140- 01). It’s not clear what the choking (H-F) model is doing or if it is even contributing to the failure. The user did try the Ransom-Trapp model and it failed around the same time and in the same manner. The increase in pressure in volume 140-01 is causing the problems, but it’s not clear why the pressure is increasing, or why it "blows up" all of a sudden. The problem may be with the jetmixer component.
Resolved (12-07-15): This user problem was resolved through the addition of Card 1 Option 35 which ramps the jetmixer component momentum equations to be the same as the branch component momentum equations as the void fraction in the jetmixer goes from 0.3 to 0.4. No. 2015-11 (7/8/2015), Resolved Code Versions Affected: RELAP5/Mod3.3 Patch04 The user submitted a problem where the flow through a valve with the Ransom-Trapp model suddenly drops to zero without any apparent reason. The problem does not occur with the Henry-Fauske choking option. Resolved (07-22-15): The user’s error was reproducable with Patch 4 executables built with an earlier version of the Intel Fortran Compiler on Windows. However, with more recent executables built with IFC 13.1, the issue was not observed. The user also noted that the executables provided on the Patch4 distribution failed on Windows Server 2003. The reason for this was that a compatibility option is needed in the linker options in order to build a WIN32 application that WS2003 recognizes. The linker option is “/subsystem:console,5.01”. New executables for 3.3iy and 3.3kg, which were linked with this option, were sent to the user.
Upcoming Spring CAMP Meeting
he Spring 2016 CAMP Meeting is scheduled for April 26
th – 28
th, and will be hosted by the UAE.
The meeting will be held at the Khalifa University in Abu Dhabi. CAMP members are encouraged to register by April 12
th if they plan to attend. Details and
registration information is available on the NRCCodes Sharepoint site (https://www.nrccodes.com). Please contact Carl Thurston ([email protected]) or the meeting host, Ms. Ayesha Al Khadouri ([email protected]), if you have any questions regarding logistics for the meeting.
TRACE Validation
here is a common misconception amongst thermal-hydraulics code users that TRACE has not been fully validated and/or that RELAP5
has a larger validation database. The purpose of this article is to set the record straight and provide a location with relevant links and resources to help address this misconception. TRACE has indeed been fully validated for the
applications to which the NRC utilizes the code. Before the NRC applies TRACE to any new regulatory application, the code undergoes a project to assess its applicability or adequacy for modeling that specific reactor design. In addition, with each new official code release, the code is assessed against an extensive set of test cases. It is worth noting that TRACE’s developmental validation base is much larger than RELAP5 in terms of the sheer number of cases that are exercised, i.e., RELAP5 has about 50 cases, whereas TRACE has over 500 cases. RELAP5, of course, has been exercised to a greater degree within the CAMP community due to its popularity and the sheer number of years the code has been in use. Nonetheless, the validation of TRACE has been greatly expanded over the last decade. Below are some public links to the NRC’s TRACE Assessment Manual: https://adamswebsearch2.nrc.gov/webSearch2/main.jsp?AccessionNumber=ML120060208 https://adamswebsearch2.nrc.gov/webSearch2/main.jsp?AccessionNumber=ML120060187 https://adamswebsearch2.nrc.gov/webSearch2/main.jsp?AccessionNumber=ML120060172 https://adamswebsearch2.nrc.gov/webSearch2/main.jsp?AccessionNumber=ML120060191 For every Patch release, the NRC re-runs most of the entire DA set documented in the links above and the results are then documented in an “Addendum” report. With regard to TRACE V5.0P4, the finalization of the DA Addendum report is currently in progress. However, the Addendum reports through V5.0P3 can be downloaded from the below link on the NRCCodes website: https://www.nrccodes.com/TRACE/Documentation/default.aspx It is also worth noting that since the assessment of V5.0, the assessment base has been expanded to include the following: MIST facility PKL facility CONTAN component (HDR, Marviken, PSTF, CVTR) Several ROSA-VI cases ATLAS facility Oskarshamn Christensen Void Experiments Additional FIST experiments Additional FRIGG experiments BFBT Ringhals Cofrentes Cycle 18
SPERT Peach Bottom Turbine Trip Three Mile Island MSLB PANDA facility PUMA facility Giraffe Facility Panthers Facility GENESIS Freon Facility In the future, these additional cases will get folded into the DA manual and into the automated assessment framework so that they can be exercised more regularly. TRACE has also undergone a documented peer review. Links to these documents are provided below: https://adamswebsearch2.nrc.gov/webSearch2/main.jsp?AccessionNumber=ML090760541 https://adamswebsearch2.nrc.gov/webSearch2/main.jsp?AccessionNumber=ML090770759 https://adamswebsearch2.nrc.gov/webSearch2/main.jsp?AccessionNumber=ML090770760 https://adamswebsearch2.nrc.gov/webSearch2/main.jsp?AccessionNumber=ML090770758 https://adamswebsearch2.nrc.gov/webSearch2/main.jsp?AccessionNumber=ML090770761 Lastly, additional resources that may be of interest can be found in the below links: https://adamswebsearch2.nrc.gov/webSearch2/main.jsp?AccessionNumber=ML091940352 https://adamswebsearch2.nrc.gov/webSearch2/main.jsp?AccessionNumber=ML092290166 https://adamswebsearch2.nrc.gov/webSearch2/main.jsp?AccessionNumber=ML113350073
Other Items of Interest
ince the Spring 2015 CAMP meeting, there have been several NUREG/IAs completed or published. The list of recently completed NUREG/IAs include:
NUREG/IA-0456, BEPU Analysis and Benchmark with IIST 2% SBLOCA Experiment Using TRACE/DAKOTA (Rep of China)
NUREG/IA-0455, Analysis of the Control Rod Drop Accident (CDRA) for Lungmen ABWR (Rep of China)
NUREG/IA-0454, Modeling of ROCOM Mixing Test 2.2 with TRACE v5.0 Patch 3 (Finland)
NUREG/IA-0453, Benchmarking of a Generic CANDU Reactor with PARCS MCNP and RFSP (Canada)
NUREG/IA-0457, Assessment of Critical Subcooled Flow through Cracks in Large and Small Pipes using TRACE and RELAP5 (Canada)
NUREG/IA-0460, Model 3D Cores for PWR Using Vessel Components in TRACEv5.0p3 (Spain)
NUREG/IA-0458, RELAP5/MOD3.3 Analysis of Event with Actuation of Safety Injection System at the Lost of External Power (Slovenia)
NUREG/IA-0459, EPR Medium Break LOCA Benchmarking Exercise Using RELAP5 and CATHARE (Poland)
NUREG/IA-0461, TRAC-BF1 to TRACE Semi-Automatic Model Conversion, PBTT Example (Spain)
NUREG/IA-0462, Uncertainty and Sensitivity Investigations with TRACE-SUSA and TRACE-DAKOTA by Means of Post-test Calculations of NUPEC-BFBT Experiments (Germany)
NUREG/IA-046x, RELAP5/MOD3.3 Model Assessment and Hypothetical Accident Analysis of Kuosheng NPP with SNAP Interface (Rep of China)
NUREG/IA-046x, Fuel Rod Performance Uncertainty Analysis during Overpressurization Transient for Kuosheng Nuclear Power Plant with TRACE/FRAPTRAN/DAKOTA codes in SNAP interface (Rep of China)
NUREG/IA-046x, Analysis of the OSU-MASLWR
001 and 002 tests by using the TRACE code (Italy)
CAMP members should also be aware that the updated PWR guidelines (Rev. 1) have recently been posted to the NRCCodes sharepoint site. The link is provided below: https://www.nrccodes.com/CAMP/Shared%20Documents/Modeling%20Guidelines/TRACE_PWR_Modeling_Guidance_Rev1.pdf In other TRACE news, a user recently encountered problems using the old D2O steam table curve fits (see Trouble Report 727). These tables were never meant to be used above 10 MPa, and thus the proposed solution is to remove the old curve fits from the code. In addition, the old H2O curve fits may also be removed from the code. If any users require the old H2O or D2O curve fits for their analyses, please contact Christopher Murray: ([email protected]). You are encouraged to visit the SharePoint site, https://www.nrccodes.com. You can join in discussions, download relevant documents, access TRACE (Bugzilla) and RELAP5 User Problem descriptions and, for CAMP
members, access information on the CAMP program including status of proposed and active in-kind contributions, announcements and a calendar of upcoming events. The discussions area supports asking questions and sharing experiences. As a reminder, TRACE bugs may now be reported to the following email: [email protected]. Christopher Murray is the contact point for the SharePoint site. If you have any problems accessing the site, or if you have any questions or would like additional information on the NRC TH codes, please contact Christopher Murray at [email protected].
