The KArlsruhe PROgram System (Extended) KAPROS(E)
and the version
KArlsruhe Neutronic EXtendable Tool KANEXT
History, status and perspectives,
a personal (re)view of C.H.M. Broeders.

Important references:

The first international presentation of the KAPROS system was in 1975 at the conference "ANS computational methods in nuclear energy; Charleston, South Carolina, USA; 15 Apr 1975".

In 1976 followed the reports
KFK2253: Das Karlsruher Programmsystem KAPROS. Teil I. Uebersicht und Vereinbarungen fuer Benutzer und Programmierer with detailed information about objectives of the KAPROS project and more specific information in
KFK2254: Das Karlsruher Programmsystem KAPROS. Teil II. Dokumentation des Systemkerns about system details and
KFK2317: Das Karlsruher Programmsystem KAPROS. Teil Ia. Kurzes KAPROS-Benutzerhandbucha short user manual.

The report in 1985
KFK3860: Die FORTRAN-77 Version des Karlsruher Programmsystems KAPROS describes the cleanup of the old KAPROS FORTRAN-4 code, including lots of IBM-assembler code, to a consistent FORTRAN-77 version.

In 1999 the report FZKA6280: Introduction to the UNIX-Version of KAPROS was the first external reference on the activities to adapt the original IBM Mainframe KAPROS codesystem for UNIX operating systems.

An exhaustive description of the status of the KAPROS system and its contents may be found in appendix B of the recent thesis.

The mostly used references for the current version of KAPROS/KAPROSE/KANEXT are a presentation at
KTG Reaktortagung Düsseldorf 2004 and this WEB information about history, status and perspectives.

The seminar presentations of Febrary 2007 (KAPROS IRS seminar) and May 2009 (KANEXT INR seminar)
give compact information about most of the developments, described in the following topics.


History of applications, code developments and available computer systems since 1967

Projects
Development aspects
Computer
From
(Steamcooled) Breeder Reactor
&
Tight Lattice LWR
to
Accelerator Driven Nuclear Waste Incinerator
&
Multi physics applications

  • History 1: The very beginning.
  • History 2: The next steps.
  • History 3: Other early developments in Europe.
  • History 4: KAPROS validation and utilization.
  • History 5: Impact of Tight Lattice Light Water Reactor research.
  • Transition from IBM mainframe to LINUX operating system.
  • Multi-group constant library development and application in KAPROS/KANEXT.
  • Development and application of tools for nuclear fuel cycle analysis.
  • Summary of available options in the current version of KAPROS/KANEXT.
  • On the future of the KAPROS/KANEXT modular code system.
  • Action plan for 2010.
  • Epilogue.
  • Status / perspectives 2016.

  • From

    IBM-7070
    IBM-360
    . . .
    IBM-AIX

    to

    LINUX PC


  • History 1: The very beginning.

    During the sixties/seventies of the past century, large national and international research and development (R&D) programs for the establishment of longtime exploitation of energy production with nuclear fission reactors existed. Around this time also a substantial enhancement of electronic devices for numerical calculations could be observed (development of analog and digital computers). So in many research institutions big efforts were ongoing to create computer programs (software) to support the design of various fission reactor proposals. At that time the research center Karlsruhe (FZK) was strongly involved in the development of fission reactors with breeding characteristics, i.e. producing more fissile fuel than it consumes (fast spectrum breeder reactors). For the numerical simulations only few computer equipment with adequate software was available worldwide. At FZK it was decided to utilize IBM hardware computers with dedicated proprietary software, using mainly FORTRAN and ASSEMBLER compiler. Based on these boundary conditions, in the Institute for Neutron Physics and Reactor Technology (Institut für Neutronenphysik und Reaktortechnik, INR) the code system NUSYS for consistent nuclear physics reactor calculations was created in the early sixties (scanned copy of final computer documentation of
    NUSYS input options). The code system NUSYS was utilized for many projects, e.g. in my diploma thesis at INR in 1967/1968, "The Influence of Nuclear Data Uncertainties of Reactor Materials on the main Safety and Stability Parameters of a Large Steam-Cooled Fast Reactor (D-1 Design)" (link to Diploma work). Observing the fast computer power development, allready in the early stage discussions started related to the follow-up code system for NUSYS. It was decided to try to avoid a solution with, at that time conventional, use of "overlay techniques", where different sections of a code occupy same fast storage regions of computer memory. In this solution all FORTRAN subroutines and functions must have unique names within the full system.
    (Go to top of document).

  • History 2: The next steps.

    During the period ~1970 . . . ~1990 several versions of a NUSYS follow-up system called KAPROS for KArlsruhe PROgram System were developed on the local IBM mainframe computer environment at FZK. These developments were characterized by moving from FORTRAN-IV to FORTRAN-77 and by significantly reducing the ASSEMBLER code in the system (see e.g.
    KFK3860). The latest IBM mainframe KAPROS version only contained small ASSEMBLER code parts for tasks, not possible with FORTRAN coding: invocation of external executable modules and dynamic array allocation. In this period also significant efforts were devoted to the documentation of the system, both with respect to the kernel code and to the application modules. In a late phase the printed documentation was converted to an on-line electronic documentation system with ASCII text files. These files are mainly in german language with some english exceptions.
    (Go to top of document).

  • History 3: Other early developments in Europe.

    As mentioned above, in the early stage of the KAPROS development other scientific groups all over the world were also creating comparable software packages. In the framework of the European fast breeder development cooperation, it was decided to try to establish a common European code system for fast breeder reactor development. In the seventies four relevant European code packages were considered in detail:
    * the german codes IANUS from the commercial company INTERATOM,
    * KAPROS at research center Karlsruhe as discussed above,
    * the british code COSMOSS and
    * the french code CCRR.
    The Fast Reactor Physics Progress Report
    KFK1632 from 1973 gives some information about KAPROS and IANUS in this period.
    After a series of presentations of these code systems and discussions at the European management level, it was decided that the french partner should develop a new system. The result is the frequently used code ERANOS.
    Here it should be noted that the code ERANOS is still based on the basic FORTRAN principles of "overlay techniques". At end-user level several ERANOS code specific pre-processor options for input- and code- handling are introduced (ALOS, ESOPE, LU), sometimes hiding the FORTRAN code application. Documentation is a mixture of French and English language.
    (Go to top of document).

  • History 4: KAPROS validation and utilization.

    Traditionally at FZK/INR the validation of calculation tools for investigations of nuclear fission reactors had a major priority. Most of these validation activities could be linked to international projects. In this context, at FZK the KAPROS code system was utilized from the very beginning for various practical problems. Strong interaction was established between developers and users. Important application areas were in this period:
    * Support for the design of the german prototype fast breeder reactor (FBR) SNR300,
    * Support of analysis of french fast breeder activities,
    * Analysis of world-wide experiments related to fast breeder development,
    * Support of studies related to power enhancement in the german FBR experimental reactor KNK2 at FZK,
    * Creation of application dependent (PWR, SWR,..) libraries for the code KORIGEN for post irradiation analysis,
    * Support of tight lattice light water reactor (TLLWR) research
    (Go to top of document).

  • History 5: Impact of Tight Lattice Light Water Reactor research.

    Most of my personal contribution to the content of the code system KAPROS is related to the research at FZK for a special purpose Light Water Reactor
    (link to PhD work). Around 1980 international interest existed to investigate options to increase the production of plutonium in LWR to create the required plutonium stocks for an accelerated introduction of fast breeder reactors, based on (UPuO2) MOX fuel. For this purpose, LWR with tight lattices with epithermal spectra were investigated in several countries. At that time specific computer codes for the simulation of reactors with dominating thermal or fast neutron spectra existed. However, for reactor systems with "intermediate spectra" (neither well moderated, nor practically unmoderated) no dedicated qualified simulation codes were available. For this reason a procedure within the KAPROS code system was developed to enable reliable simulation of "intermediate spectra" reactor systems. In this module, KARBUS for KArlsruhe Reactor BUrnup System, special features of specific codes for the simulation of thermal and fast reactor system zones are included in one code, e.g. enabling treatment of neutron upscattering, required in thermal systems as well as a detailed description of high energy reactions like (n,xn) processes, usually applied for the simulation of fast neutron spectrum systems. In addition to the options for refined reactor zone cross section calculations, the KAPROS module KARBUS also enables to apply these results directly in full reactor system flux calculation for most common reactor geometries. For most cases diffusion and transport solutions can be obtained.
    During the TLLWR project at FZK, a close cooperation was established between the industry partner KWU (Erlangen), the Ben Gurion University (Jerusalem) and the experimental facility PROTEUS at PSI Würenlingen. First experimental results from dedicated TLLWR experiments in the PROTEUS facility played an important role in the KAPROS/KARBUS validation work. The test section of the KANEXT package contains an evaluation of important reaction rates comparisons of computed values with experimental ones (C/E values for reaction rate ratios).
    (Go to top of document).

  • Transition from IBM mainframe to LINUX operating system.

    In the nineties, a small team realized in a concentrated cooperation a transition of the KAPROS system from the IBM mainframe environment to the UNIX workstation application. For this purpose a stepwise procedure was applied. In a first step the IBM mainframe application was adapted to a solution on IBM workstation computers with the UNIX variant operating system AIX, containing IBM propiatary C- and FORTRAN-compiler with very good comptability with IBM mainframe software. In 1999, a first KAPROS version could be implemented on a LINUX operating system using the Fujitsu FORTRAN compiler. After some major conceptual changes in view of 64-bit application the name KAPROS was replaced by KANEXT ( KArlsruhe Neutronics EXtendable Tool) recently. This KAPROS/KANEXT kernel and most of the modules are compatible with the commercial LaHey Fujitsu FORTRAN compiler version 6.2 for 32-bit and version 8.1 for 64-bit architecture, with the open-source FORTRAN compiler g95 and gfortran and with the INTEL ifort compiler with special licensing procedures. For the few C-programs in the system the open-source code compiler gcc with nearly all versions may be used, as well as the INTEL icc compiler. Here it should be noted that in the course of about fifteen years of experience, the backward comptability of the new gcc versions was not satisfactory. The transfer of variable number of arguments and application of "shared memory allocation" required adaptations for using the new gcc version releases. Nearly all original options described above are available in the current version of the system. Moreover, powerful archive and restart options enable the use of KAPROS/KARBUS results in coupled stand-alone codes. For this purpose several KAPROS modules for the creation of international standard interface files are available.
    (Go to top of document).

  • Multi-group constant library development and application in KAPROS/KANEXT.

    From the very beginning the creation and qualification of multi-group constant libraries was a major objective of the fast reactor research at FZK. As a starting point the basic principles of the russian ABBN (
    KFK-TR-144) multi-group library were adopted. The report KFK770 gives interesting information of these early developments. This work started at Research Center Karlsruhe with the creation of an evaluated nuclear data library KEDAK with own storage formats. For the creation of multi-group data from the KEDAK file, the code MIGROS3 was developed. In the first stage of multi-group library creation at FZK the well-known, widely used, KFKINR set was completed. Structure and management of the so-called GRUBA libraries is documented in the KFK1815 report. For the determination of macroscopic cross section data of reactor zones from microscopic material dependent data on the GRUBA libraries a unique solution method was developed. The calculation rules for the cross section types of interest are sepated from the calculation code (see e.g. the original reports GRUMA,GRUCAL) and stored on a dedicated file ( "Steuerfile") with its own interpretation rules. This feature enabled the extention of the first relatively simple fast reactor applications to more complicated nuclear systems without major code modifications. Some interesting extentions were:
    * Refinement of the fast reactor methods of the KFKINR set to the KFKINR2 specifications
    The first relatively simple calculation rules based on the ABBN documents were replaced by more sophisticated approximations. The resulting restructured "Steuerfile" is basis for subsequent applications
    * Introduction of gamma heating in the cross section calculations
    A special version of the GRUBA library with corresponding "Steuerfile" enabled coupled neutron/gamma calculations
    * Introduction of up-scatter capabilities for thermalized systems
    During the TLLWR work up-scatter capabilities were introduced in the KAPROS/KANEXT cross section processing procedures. For this purpose also small modifications in the cross section processing of the code were introduced.
    * Extention to higher energies for source driven systems (ADS)
    For application in deterministic studies of accelerator driven systems (ADS) the typical energy region of multi-group libraries for fission reactors was extended from 10 to 250 MeV (Diploma thesis Oberle). Nuclear data types on the library and dedicated computation rules were introduced, without need for code modifications.
    * Introduction of refined methods for transient analysis
    For application in reactor transient analysis the required data for spectra and fractions of delayed neutrons were included in detail in the library and in the corresponding "Steuerfile".

    In the current version of the KAPROS/KANEXT modular system the following libraries and "Steuerfiles" are included:

    Library "Steuerfile" Comments
    KFKINR F26 First complete 26 group constant set, based on ABBN basic principles
    KFKINR2 F26TN First complete 26 group constant set, based on FZK developments
    G69P1V03 F69V2STR Final version of 69 group constant set developed for TLLWR work
    G69P5JEFF31FBR2 FENDFJEF Actual 69 group constant set based on JEFF3.1 data for FBR research,
    extended "Steuerfile" for transient analysis
    G69P5JEFF31LWR FENDFJEF Actual 69 group constant set based on JEFF3.1 data for LWR research,
    extended "Steuerfile" for transient analysis
    G69P5E70BFBR FENDFJEF Actual 69 group constant set based on ENDF/B-7 data for FBR research,
    extended "Steuerfile" for transient analysis
    G69P5E70BLWR FENDFJEF Actual 69 group constant set based on ENDF/B-7 data for LWR research,
    extended "Steuerfile" for transient analysis
    G85P5E70BFBR FENDFJEF Actual 85 group constant set based on ENDF/B-7 data for FBR research,
    extended "Steuerfile" for transient analysis,
    16 additional groups up till 20MeV above 69 WIMS group structure
    G85P5E70BLWR FENDFJEF Actual 85 group constant set based on ENDF/B-7 data for LWR research,
    extended "Steuerfile" for transient analysis,
    16 additional groups up till 20MeV above 69 WIMS group structure
    G350P5JEFF311E FENDFJEF Actual 350 group constant set based on JEFF3.11 data,
    extended "Steuerfile" for transient analysis, 350 groups up to 20MeV
    G350P5E70 FENDFJEF Actual 350 group constant set based on ENDF/B-7 data,
    extended "Steuerfile" for transient analysis, 350 groups up to 20MeV
    Summary of KAPROS/KANEXT multi-group library information

    Application of a 350 group library is recommended

    (Go to top of document).

  • Development and application of tools for nuclear fuel cycle analysis.

    The analysis of the consequences of energy production in nuclear fission reactors is a key issue at Research Center Karlsruhe (FZK) since a long time. The principle of this energy resource is based on the fact that in case of a reaction of a neutron with isotopes of the existing heavy metal uranium, there is a probability that the resulting neptunium isotope is splitting into two medium weight isotopes and a few neutrons, together with release of a large amount of "excess energy". The produced neutrons enable construction of a critical reactor with self-sustaining neutron generation. The "excess energy" is converted in the reactor in useable thermal and electric power, to be distributed for various applications. A disadvantage of this technology is that the primary fission reactions produce fission product pairs (FPP). Most of the isotopes of these FPP are short-lived and highly radioactive. Mean life times are in the order of less than seconds to several hundreds of years. This means that after about 500 years most of the radioactivity of the FPP is disappeared. Moreover, together with these primary neutron fission reactions with uranium, other unavoidable absorption reactions lead to the creation of articifial heavy metal elements like plutonium, americium and curium. Several isotopes of these articifial heavy metal elements have very large half lives and consequently significant radiotoxicity for a very long period (hundred thousands of years).
    The prediction of the evolution of the composition of nuclear reactor fuel during and after irradiation in a nuclear reactor system is the major task of tools for nuclear fuel cycle analysis. The research in this area started at FZK in an early stage on the basis of the original
    ORIGEN code. The extentions of ORIGEN by FZK resulted in the widely used code KORIGEN ( validation , code description). The codes ORIGEN and KORIGEN consist of two major components with main tasks:
    - determination of time dependent material inventories
    - evaluation of consequences of these inventory build-up, like γ-heating, radiotoxicity, etc.
    The first component, material inventory build-up, was adopted as basic part of the module BURNUP for burnup analysis in KAPROS/KANEXT. The module BURNUP is the central component of several KAPROS/KANEXT modules (procedures) for nuclear fuel cycle investigations, e.g.:
    - BURN0D: reactor burnup investigations, based on a fundamental (zero-dimensional) approximation for the reactor system
    - KARBUS: reactor burnup investigations with options for various reactor models, using zone-wise burn-up
    - VABUSH: sophisticated reactor burnup investigations with options for fuel assembly shuffling in a reactor with hexagonal fuel assemblies. This module was successfully applied for first analysis of the fuel loading problem in the small sub-critical source driven system MYRRHA, being investigated in the framework of EU research.
    The module BURNUP was validated at several stages for various problems. Information may be found in the TLLWR documentation. Recently, PWR burn-up validation work was organized on the basis of the ICE experiment in the NPP Obrigheim in the 1970-ies. A main resource for this IAEA coordinated research project (CRP) for analysis of accelerator driven systems (ADS) is the recent diploma thesis. Preliminary evaluation of the results of several contributions, based on deterministic and Monte Carlo methods, show good agreement for most comparisons. Final results will be linked here, as soon as possible.
    (Go to top of document).

  • Summary of available options in the current version of KAPROS/KANEXT.

    The main purpose of the modular code system KAPROS/KANEXT is to provide a tool for deterministic multi-group neutron physics simulations for a wide variety of nuclear reactor systems, with special emphasis on the validation of methods and data for design investigations of innovative systems. For this purpose a number of independent modules to perform specific tasks is available. Task sequencing is possible by input specification and by creation of dedicated modules, usually in FORTRAN programming language. Several existing modules enable performing complicated sequences for specific tasks. In the following most of the options available within KAPROS/KANEXT are discussed.
    For most of these options typical examples are provided in the test section of KAPROS/KANEXT.

    A typical nuclear reactor system is composed of several sub-zones, e.g. power producing zones, shielding zones, control rod zones, reflector zones, upper/lower plenum etc.
    * Typical tasks during a deterministic full core simulation.
    For a full core simulation with determistic multi-group solvers, adequate multi-group zone-wise data is required. Depending on the composition and geometry of the sub-zones different approximations for the determination of this data are applicable, see e.g. documentation of Tight Lattice LWR
    (TLLWR) work. Reactor zones without fissile material (fuel zones) and without strong absorbers (control rod zones) usually can be handled space-homogenized. For fuel zones, spectrum dependent solutions (fast/thermal) are available. Special care is required for control rod zones.
    - After preparation of the required zone-wise cross section data, full core calculations may be performed with suitable neutron flux solvers, based on specific approximations of the governing Boltzmann balance equation in its various formulations. The KAPROS/KANEXT modular code system offers several options for this purpose. Some of them are integrated in KAPROS/KANEXT (strongly coupled), others are stand-alone codes with automatic KAPROS/KANEXT interfaces (weakly coupled).
    - After successful single zone or full core calculations, evaluation of the results is an important task to be able to analyse and judge the results. The modular code system KAPROS/KANEXT offers for this purpose very powerful archiving features. Standard KAPROS/KANEXT archives may be created (initiated and filled) by simple directives in the modules and in the input files. Subsequently, simple directives may retrieve these results for analysis in KAPROS/KANEXT evaluation modules, being available for various applications.
    - For an overall analysis of energy production in nuclear fission reactors, the behaviour of nuclear reactor fuel under irradiation, i.e. during energy production by fissions, is another key issue, being under consideration during KAPROS/KANEXT development. The achievements of the development of the KORIGEN code at FZK, (KORIGEN) were effectively included for burnup and depletion analysis in KAPROS/KANEXT. In the module BURNUP modified KORIGEN libraries may be used. For faster access, rearrangement of the data with an auxilliary module is applied.
    - For the analysis after burnup and depletion calculations with BURNUP, the available archive options enable detailed analysis of results, e.g. with the module MIXIMA with automatized coupling with the public domain graphics tool xmgrace.
    * Specification of reactor zone compositions.
    Generally, the specification of reactor zone compositions is not well standardized in nuclear reactor simulation codes. Required are at least information about nuclide identification, about mean temperatures and about concentrations (number densities). Fast reactor zone specifications are traditionally based on homogenized representation. For thermal reactor systems heterogenity information is required. In KAPROS / KANEXT data block specifications are defined for this purpose. The required data blocks may be specified in the input stream or may be created by dedicated KAPROS/KANEXT modules. Currently two modules support creation of these data blocks:
    - The module NDCALC creates all required input data blocks based on parametric zone specifications as convenient for parametric reactor system investigations. The module was developed during the TLLWR research. For most practical problems input preparation for NDCALC is possible, specification of complicated experiments may cause problems.
    - The module NDWIMS was developed in an early stage to enable use of WIMS input files within the KAPROS code system. All relevant data blocks for KAPROS are created, based on the original specifications for the WIMS code. In the KAPROS/KANEXT test section an example using NDWIMS is given for the evaluation of an experiment in the PROTEUS reactor facility at PSI Wuerenlingen Switzerland in the framework of Tight Lattice LWR work. Reaction rate ratios of MOX fuel experiments in an epithermal neutron spectrum may be easily compared for different multigroup constant libraries in KAPROS/KANEXT.
    * Specification of reactor geometries.
    Also the specification of reactor geometries is generally not well standardized. Most of the commonly used codes have their own rules. In KAPROS/KANEXT some options are available to handle this problem, especially with coupled codes with CCCC interface files.
    * Creation of multi-group cross section data in reactor zones.
    The modular code system KAPROS/KANEXT is a very flexible powerful tool for the creation of multi-group cross section data in reactor zones. The module KARBUS enables the preparation of such data for an arbitrary number of zones, stored in a KAPROS-own format (the SIGMN structure). For each zone the most convenient calculation method may be selected from several options (see also TLLWR work):
    - GRUCAL: homogenized zone calculation as applied in standard fast breeder solutions.
    - GRUCAH: cell heterogenity corrections within GRUCAL solution.
    - GRUCEL: three zone collision probabilty calculations in cylinderized Wigner-Seitz representation of reactor lattices.
    - DANSYS enables cross section zone homogenization for reactor zones based on flux calculations with the coupled S-N transport codes DANTSYS or PARTISN. A typical application of this option is fuel assembly homogenization.
    - KAPER4: fast reactor cell code for neutronic calculations of heterogeneous lattices of fast reactors, as described in KFK4435
    * Creation of interface files for stand-alone codes with multi-group and geometry data for reactor zones.
    In accordance with the original design objectives, the KAPROS/KANEXT code system contains well defined effective specifications for most of the fundamental results within nuclear reactor simulations, e.g. for zonewise material specifications, macroscopic cross sections and neutron flux distributions. For coupling with external codes like DANTSYS, PARTISN, DOORS, TORT, DIF3D, VARIANT the CCCC specifications as defined in CCCC interface specifications are applied as far as possible. For this purpose a number of auxilliary conversion modules are available:
    - The module HXFILE, developed for TLLWR research, creates a code specific coupling of KARBUS with the nodal flux solver HEXNODK, a KWU code development, licensed for FZK.
    - The module CRGIP creates the GIP cross section interface file, e.g. for use in the DOORS code related programs.
    - The module GCMPXS creates the CCCC interface file COMPXS as specified in the DIF3D/VARIANT documentation
    - The module VAPROC creates the CCCC interface files COMPXS and GEODST, e.g. for use in the DIF3D code. For the creation of COMPXS, the module GCMPXS is invoked.
    - The module CTFILE, developed in an early stage of KAPROS coupling with external codes, handles cross section transfer to the CITATION code, without using a standard interface definition.
    * Determination of neutron flux and power distributions.
    Direct determination of neutron flux and power distributions is supported in several modules, either fully integrated in KARBUS/KANEXT or accessible as "coupled codes".
    - KAPROS/KANEXT modules of the package DIXY2 with evaluation options
    - KAPROS/KANEXT modules D3D/D3E with dedicated evaluation module AUDI3
    - KARBUS coupled with CITATION, VARIANT/DIF3D, HEXNOD, DANTSYS/PARTISN, DOORS.
    * Burnup and depletion calculations.
    In KAPROS/KANEXT the burnup and depletion calculations are performed in the module BURNUP, developed in an early stage from the code ORIGEN. The local improvements of the code KORIGEN were fully adopted. In addition, ordering of library data was improved. Comptabilty with KORIGEN is obtained with available auxilliary modules.
    * Reactivity control in nuclear reactors.
    During power generation in a nuclear reactor system the neutron multiplication factor must be controlled. In a critical reactor the value has to be close to one. In source driven systems a relation exists between level of sub-criticality and power production. For this purpose reactivity control is applied in such systems. In KAPROS/KANEXT the boron concentration in water lattices can easily be modified. In the KAPROS/KANEXT test section boron variation in a light water reactor lattice is demonstrated for an irradiation experiment in the PWR Obrigheim (KWO). A refined application of boron concentration variation is described in diploma thesis Send. Absorber rod movement for reactivity control can be handled by geometry input specification changes.
    * Fuel management in nuclear reactors.
    Satisfactory fuel utilization in power producing nuclear reactor systems needs adequate management of the available fuel. Nuclear fuel management is a complicated task and strongly design dependent. In (link to PhD work) early fuel management applications are presented. The test section of KAPROS/KANEXT contains a recent application for the design of fuel element shuffling in a source driven subcritical reactor (module VABUSH applied for XT-ADS project).
    * Determination of safety and performance parameters of nuclear reactor systems.
    Systematic analysis of safety and performance parameters is an important task in nuclear reactor system engineering. During the design work for an Advanced Pressurized Water Reactor (APWR) (see TLLWR work) specific modules were created within KAPROS. Important modules for this purpose are available in the current version of KAPROS/KANEXT. E.g. the module GRUMIX for easy modification of temperatures and densities in selected zones is a very good tool for the determination of safety coefficients like Doppler and voiding. Here it may also be noted that the standardized archiving options in KAPROS/KANEXT enable the evaluation of the results of complicated calculations in subsequent small dedicated modules.
    * Consistant perturbation calculations.
    In the early stage of nuclear reactor engineering large efforts were devoted to the fast and reliable determination of reactivity changes in a close to critical reactor system due to small changes in a fixed system state. Several approximations exist for this so-called perturbation theory. As described in (perturbation theory) support), KAPROS/KANEXT is a good environment for perturbation theory solutions. The first 2-dimensional diffusion code DIXY2 in KAPROS has its own module for perturbation theory application. For 3-dimensional evaluations, including perturbation calculations, the dedicated module AUDI3 was created. In the test area of KAPROS/KANEXT a coupling of the 3-dimensional diffusions code D3E with AUDI3 for perturbation investigations is provided. Coupling of AUDI3 with the coupled code TORT for 3-dimensional transport calculations is in development.
    * Graphical and tabular representation of calculation results.
    Graphical representation of results of computational codes is a key issue in research and development (R&D) work. From the very beginning, strong efforts were devoted to support this incentive. Current supported options are:
    - Options of the included PLOTEASY package modules.
    At the time the KAPROS development started, practically no standardized graphics software and only few plotting devices were available. Graphs were usually produced manually in a special office. At FZK a public IBM assember programm (PLOTA) was developed to produce curve plots on available IBM devices. Later PLOTA was simulated by the more general CALCOMP PLOTS software solution with elementary plotter movement definitions. Finally, for LINUX/UNIX the PLOTA/PLOTS code was adapted to the PSPLOT package of NSU Oceanographic Center to produce postscipt files from PLOTA applications. Based on the PLOTA specifications, a number of stand-alone codes and KAPROS modules were developed to produce automatically relevant graphs. These codes generate quite simple graphs for giving fast overviews. For improved graphical representation interface files with standard data tabulations are available.
    - Data for free source UNIX tool xmgrace.
    Several evaluation modules enable the creation of input data files for the open source graphics code xmgrace.
    - Data for commercial graphics tool TECPLOT.
    A few dedicated evaluation modules support creation of input data files for the commercial visualisation code TECPLOT.
    - Data for commercial tool MATLAB and compatible open source code OCTAVE .
    Recently, a KANEXT evaluation module was created to produce graphs with the commercial tool MATLAB. Most of the options can be handled with the latest versions of the open source codes OCTAVE and GNUPLOT.
    * KAPROS/KANEXT and multi-physics applications.
    The design goal "full modularity" of KAPROS/KANEXT makes the system very well suited for solving multi-physics problems like consistent coupling of neutron physics, thermo-hydraulics and transient solutions. In the following one example from the early KAPROS development and some recent applications are discussed. Documentation related to the recent work may be found on http://inrwww.webarchiv.kit.edu
    * In an early stage of the main-frame KAPROS version the module KINTIC was developed (see KFK1632 section 3.6) for FBR transient investigations. However, this module is not implemented in the current version of KAPROS/KANEXT.
    * In the framework of the R&D support of recent EC projects for accelerator driven systems (ADS), within the LINUX version of KAPROS, a coupling was realized of available neutron physics solutions with the fast reactor transient code SAS4, developed for many years at FZK. This module SAS4ADS was successfully applied for several important project tasks. The module VABUSH of the test section of KAPROS/KANEXT, mentioned before as fuel management application for the EC XT-ADS project, is based on this early LINUX SAS4ADS code.
    * In the framework of the R&D support in the first stage of the EC project for an High Performance Light Water Reactor (HPLWR) a coupling was realized between KAPROS and the system code RELAP5 with very interesting results (ICAPP 2003).
    * During students work in the section Neutron Physics and Reactor Dynamics (NR) a number of couplings between KAPROS and versions of the thermohydraulics subchannel code COBRA were created and successfully applied. One objective of these activities was to support utilization of KAPROS in Light Water Reactor (LWR) projects like NURESIM and follow-up proposals.
    Here it should be noted that presently neutron physics codes have better provisions for information exchange by interface files than thermohydraulic codes.
    (Go to top of document).

  • On the future of the KAPROS/KANEXT modular code system.

    Although the KAPROS/KANEXT modular code system is based on one of the oldest tools for the simulation of nuclear reactor systems, the current implementation for LINUX operating system enables reliable nuclear reactor calculations for a wide variety of applications. The basic feature of independent modules, only coupled by a few well-defined kernel functions, is still unique in the area of nuclear reactor simulations.
    The main application areas of KAPROS/KANEXT are related to the flexible preparation of reactor zone wise effective cross sections, including fuel burnup and depletion, to the application of powerful archive and restart options, and to full core investigations with original neutron flux solvers like DIXY and D3D and with coupled external codes by means of standard interface files. The KAPROS/KANEXT system already contains several modules for the creation of such standard interface files, e.g. for exchange of information of reactor cross sections, of reactor geometries and of thermo hydraulic reactor zone characteristics with the codes DIF3D/VARIANT, DANTSYS, PARTISN, DOORS, TDTORT, CITATION, SAS4, RELAP, COBRA. The evaluation module AUDI3 offers unique perturbation theory calculations with interfaces to the stand-alone codes mentioned before. The current version of KAPROS/KANEXT contains quite well documented examples for several typical applications.
    These features enabled the successful application of the KAPROS/KANEXT modular code system during its long lifetime for various research projects with permanently a few maintainers/developers and at least 3-5 users. In the past few years young students and staff members consolidated the code stability with respect to new computer hardware and software developments. Moreover, multi-group cross section library preparation and management and effective zone cross section calculations have been systematically improved and harmonized.
    The documentation and quality assurance of the system had high priority from the very beginning. As a consequence, most of the written documents from the early phase (without options for electronic documents) were converted to computer ascii files. These ascii files can be viewed by most standard browser programs (e.g. the open source code firefox). For the modification of the source code or of the data libraries formal administrative procedures were applied for a long period. In the framework of the KANEXT development, the version control software Subversion (SVN) has been established for the purpose of quality assurance and documentation. Moreover, an internet based "WIKI" information exchange system is being established in this environment. The description of the few poorly documented modules with significant application potential is ongoing with high priority.
    As main outlook it may be concluded that the KAPROS/KANEXT modular code system still presents a unique research tool for the investigation of innovative nuclear reactor systems with various types of fuels, coolants and neutron spectra, including research reactors, generation 4 reactors (GEN-IV) and subcritical source driven systems for nuclear waste incineration (ADS). The problem of validation of calculation tools always played an important role, especially for innovative reactor systems. In the past, many dedicated integral experiments have been performed in view of code and data validation. Although such experiments are still of high importance, the availability of dedicated Monte Carlo codes, together with powerful computer systems, gives an alternative, usually faster and cheaper, option for preliminary validation of the deterministic multi-group solutions.
    As a final comment it should be pointed out that several previous applications, especially in the area "students work", have shown that the full modularity of the system facilitates solutions for multi-physics problems like consistent coupling of neutron physics and thermo-hydraulic solutions.
    Unfortunately, the proposal to include the KAPROS/KANEXT/KARBUS code system in European R&D projects for multi reactor physics like NURESIM and follow-up could not be realized.
    (Go to top of document).

  • Action plan for future

    In order to preserve the unique features of the KAPROS/KANEXT modular code system as "restricted open source tool" for a broad range of interested parties, the current activities are to prepare and to improve:
    - a frozen export version of KANEXT to be offered to the OECD/NEA data bank Paris for distribution to the OECD community, according to the correponding rules.
    - a dynamic, Subversion version control (SVN) based, complete KANEXT repository at an open internet server to cooperate with approved partners.
    For both objectives first versions are available now and ready for application.
    (Go to top of document).

  • Epilogue

    In the retrospective of nearly fifty years it is remarkable that my profesional life is strongly focused on scientific research for innovative nuclear reactor systems. It started for my diploma thesis with research for a steam-cooled breeder concept to combine existing light water technology with nuclear fuel breeding in a reactor. Finally, this concept was not realized in favour of heavy metal cooled breeder designs. In my second project, ending with my PhD thesis, the goal was to enhance the fuel utilization in LWR by increase of fuel conversion ratios. This work ended with a satisfactory design proposal, which was not realized because of delays in the development of nuclear energy production. The projects after 1992 were related to research for various aspects of a closed nuclear fuel cycle, including analysis of accelerator driven systems for nuclear waste incineration, to neutron physics analysis of a High Performance LWR, with water coolant at super-critical conditions, and recently to research for the new liquid metal cooled GEN-IV concepts. The main simulation tool for these investigations, covering a period of more than 40 years, was the modular system KAPROS/KAPROSE/KANEXT and its preceding versions.
    In this period, large efforts were devoted to the development of simulation tools for the investigation of nuclear reactor systems. The computer support for such simulations expanded at FZK from the first 10000 word IBM-7070 computer, via various IBM-360 mainframe installations with a few hundreds kilobyte fast memory storage, to current powerful workstations with gigabyte fast memory. In this environment, the software tools for such simulations were continuously improved and extended, resulting in the current unique KAPROSE modular code system. In recent years the KAPROS system has been applied in projects for most of the challenging proposals for innovative solutions for problems in the area of nuclear reactors research. A typical example are the investigations related to accelerator driven sub-critical reactors for nuclear waste incineration on powerful LINUX PC systems.
    Reviewing the history with close to 100 man-years support for KAPROS and observing the current status of KAPROSE, it may be concluded that, from a scientific point of view, this modular system with its powerful options for linking to any application, is very well suited for deterministic nuclear reactor simulations in the future, especially for innovative designs and for closed fuel cycle analysis, including consequences of nuclear phase out strategies.
    (Go to top of document).

  • Status / perspectives 2016

    After the
    Action plan for 2010, the scientific work in the area of nuclear engineering was strongly influenced by the Fukushima reactor accident. In the early stage after the earth quake, analysis within the KAPROS modular system could provide valuable documents concerning the radiological source terms after the accident to support the related ad-hoc KIT expert group (KANEXT application note 2011/01). A summary of these investigations was presented at Reaktortagung 2012 Stuttgart (KTG 2012). The consequence of this accident was, that phase-out of nuclear energy production in Germany was decided. This decission leads to considerable reduction of national support for reactor physics related work. Unfortunately, it was not possible to find sufficient funding for manpower for KAPROS related work. As a consequence the main active KAPROS developers have left the project and the support at KIT of the KAPROS system is strongly decreased, compared to the expectations in 2010. A main goal to establish an open subversion (svn) repository for approved external scientists can not be realized anymore because the actual KANEXT svn repository is only accessable by scientists with KIT SCC account.

    (Go to top of document).


    Last update 03.02.2017 by C.H.M. Broeders.