Development of experimental programme on the PSB RBMK integrat test facility of thermal-hydraulic system codes aimed at safety analysis of NPPs with RBMK-type reactors.

The specific objective of this project was to contribute to the formation of an integral test database on the PSB-RBMK test facility for the supplementary validation of the best estimate Western thermalhydraulic codes for accident and transient analysis of RBMK reactors. GRS (Germany) was the contractor.
The PSB (abbreviation of the Russian name: Full Scale Safety Facility) RBMK test facility is a new integral TH facility under construction at EREC at the moment of the project implementation.
PSB-RBMK facility has been designed for experimental investigation of TH processes which may take place in the main circulation circuit during transient and accident conditions. The facility contains the models of all main circulation circuit elements (full scale models of six technological channels of the RBMK), a drum separator model, models of the suction and pressure headers and of the group distribution headers, models of valves and fitting elements of the main circulation circuit and simulators of main reactor systems (ECCS, feedwater system, emergency protection system). To investigate loss-of-coolant conditions it is planned to connect discharge lines with the anticipated break locations.
With the PSB-RBMK test facility it is expected to investigate the influence of various phenomena occurring in the course of RBMK specific accidents and transients. Experimental data obtained will be used for the validation of thermalhydraulic codes that are used for NPPs safety analysis.
The project activities were related to:

Assessment of the existing database with respect to their suitability and availability. Determination of modes, processes and phenomena to be investigated on the RBMK test facility.
Development of a programme of experimental scenarios (test-matrix) for the PSB-RBMK test facility.
Development of an ATHLET MOD 1.2 Cycle D code input for the PSB-RBMK test facility and pre-test calculations of two selected scenarios.

1) Identification of relevant Modes and Processes for the Safety of RBMKs:

The working group reviewed the supplementary validation matrices for Western thermalhydraulic codes, developed by NIKIET organisation. Experts of NIKIET took part in this review. After comprehensive discussion and small updating the NIKIET matrices for LOCAs and transients were accepted as validation matrices for this project. Two matrices were developed: validation matrix for accidents of the LOCA-type, validation matrix for transients.

Loca-type accidents included:
Pressure Header break (a not closing check valve can be considered as an additional failure in an adjacent Group Distribution Header, GDH). GDH break (the check valve non-fit in other GDH can be an additional failure). Critical break (partial break) of GDH. Downcomer break. Steam Line break. Loss of Flow in Fuel Channel (the accident is a consequence of the inadvertent closure of isolating control valve or its destruction).
The following conditions were included in the validation matrix for transients:
Coolant Flow Rate Decrease (Reasons can cause these accidents: stop of one or several operating MCPs. Breakage of the MCP main stop valve disk. Loss of flow at GDH inlet cause e.g. by blockage of filters). Loss of Feed Water. Loss of AC Power (the core is cooled under natural circulation conditions. The emergency feed pumps powered from diesel generator are actuated to supply feed water into the circuit). Reactivity Initiated Accidents (The accidents involve reactivity growth. They are caused by spontaneous withdraw of a single CPS rod or group of CPS rods). ATWS (They are BDBA. The above listed transients without emergency protection actuation are proposed to consider as emergency states).
The following typical phenomena are identified for accidents considered in the matrices:
Discharge from the Circulation Loop Break, Thermalhydraulics of the core and channels. Fuel Assembly Heat Transfer (Heat transfer prior Critical heat flux (CHF) onset. Heat transfer under steady state conditions and transients.CHF sharp decrease of heat transfer from fuel elements to coolant. Post-CHF heat transfer. Heat transfer to coolant after the CHF onset. Radiation heat transfer. Radiant heat transfer through layer of the wet or superheated steam between fuel assembly and fuel channel wall). Feedback between thermalhydraulic and neutronic processes. Natural Circulation Development and Breakdown (Natural circulation takes place after the MCP shut-off and coast-down, it evolves in the circuit when the motive pressure drop - difference between hydrostatic pressure drops across the downcomer and upward segments of the circuit - exceeds pressure losses in channels. Depressurization takes place in the circuit under the accidents of LOCA-type. As a consequence flashing in the downcomer occurs. It results in the decrease of the driving head and can lead to natural- circulation breakdown. Analogous situation can take place under accidents of loss of AC power as well as with the opening of the main safety valve and following their non-closure, failure of the valve). Steam-zirconium Reaction. MCP behaviour. Development and Degradation of the Flow Fluctuations in Channels of Different Power. Core Heat Fluxes (cross heat fluxes between the graphite stack and channels, heat exchange between PC and the graphite through the gaps and bushes are considered). Water Entrainment from Drum Separator (DS). DS Behaviour (Included: The boiling and condensation in DS. The structure and dinamics of the coolant flow in the DS varies with change of the pressure and feed water supply.Separation and water entrainment from DS into steam lines. The phenomenon is typical of regimes with considerable level increase above submerged perforated plate. Steam entrainment into downcomers. The phenomenon is typical of regimes with considerable decrease of total collapsed level and pressure in DS). Flow Instability in Channels of one GDH (The phenomenon is typical of the parallel steam generator channels. This can take place both at forced and at natural circulation. The flow oscillations in the channels are followed by other thermalhydraulic parameter variations. Under development of such oscillations the cladding temperatures can reach unsafe values).
Data sources used in the validation matrices include experimental data obtained from test facilities and from NPPs during both planned tests and incidents. For modes listed in the matrix for LOCA-type accidents only separate effect tests have been performed. Integral tests have not been performed. For modes listed in the matrix for transients integral tests data are available. These data are obtained from experimental studies with integral test facilities and from NPP incidents.

2) Assessment of the existing experimental data in respect to their suitability and availability:

The Russian project members have prepared general information on the existing RBMK experimental database and short descriptions of existing facilities and experimental results. The analysis of this information has confirmed that the existing test data are insufficient for code validation. It was found, that no integral test data for LOCA are available. Integral test data for transients obtained in NIKIET, RRC KI and EREC in the 70-ties and 80-ties are of comparatively poor quality and their benefit for code validation is limited.
Authors of the original NIKIET tables took part in the work on updating the tables. After some changes made to the original NIKIET tables they were used for compilation of table used in this project (see attached table).
�

A description of the following facilities was included in the project final report: KS (RRC KI), BM (NIKIET), Test Facility 108 (EREC), KT (NIKIET), KSB (RRC KI)
Review of available experimental data: This included the following phenomena/experiments:

Hydraulics of single and two-phase flow in RBMK Channels (Experimental Research of Thermalhydraulic Characteristics of a Full-scale RBMK-1000 FA Model in Test Facility KS, RRC KI 1974).
Heat Transfer Crisis and Post-CHF Heat Transfer in the RBMK-1000 and RBMK-1500 Rod Assemblies (Investigations of the Onset of the Heat transfer Crisis and Post-CHF Heat Transfer in the Electrically Heated RBMK FA Models, EREC 1977-1989).
Relative Motion of Phases in Channels of Complex Geometry (Experimental Investigation of Void Fraction Distribution throughout the RBMK FC HRS Height in BM facility, NIKIET 1974-1984).
CHF Conditions in the RBMK-1000 and RBMK-1500 Rod Bundles (Investigation of CHF Conditions in the RBMK FC Full-scale Models in the KS Test Facility, RRC KI 1974).
The Phase Counter-current Flow in Heated Channels at Low Flow Rates (Simulation of Stagnation Conditions during Cool-down of the RBMK-1000 under Natural Circulation or after GDH Partial Breaks in Test Facility KS, RRC KI 1974-1980), (Investigation of Thermal-hydraulic Processes Occurring in the RBMK-1000 FC at Circulation Stop under Repair Conditions in Test Facility KS, RRC KI 1974), (Influence of Coolant Flow Stop at the RBMK FC Inlet on Residual Heat Removal in Test Facility KS, RRC KI 1986-1989).
Critical Discharge of Steam-water Mixture from Long Pipes and Nozzles (Test Data of Experimental Studies on Critical Coolant Discharge through Circuit Elements of Steam Generating Units, EREC 1975-1991).
Auto-oscillations in the Channel (Experimental Study of the RBMK Main Coolant Circuit Thermalhydraulic Stability in Facility 108, EREC 1982-1989), (Investigation of the RBMK Starting Conditions in Test Facility BM, NIKIET 1972).
Development and Break-down of Natural Circulation (Experimental Study of Emergency Cool-down Modes under Loss of AC Power Conditions and Simultaneous Depressurisation in Facility BM at NIKIET 1972-1981), (Experimental Study of Natural Circulation under Steady-state and MCC Depressurisation Conditions in Facility 108, EREC 1982)
Condensation Front Propagation (Tests on Flood of the Heated SWP Pipe with Water in Test Facility "SWP pipe", NIKIET 1991)
Relative Motion of Phases in a Big Volulme (Experimental Studies of the DS Vessel Internals Model for RBMK-1000 in Facility 102, EREC 1987) (Commercial Tests of Drum Separators in the RBMK-1000 Units, data obtained in DSs at different RBMK-1000 NPPs) (Commercial Tests with Drum Separators in RBMK-1500 Units, data obtained during Ignalina NPP Unit 1 start-up tests)
Reflood (Data obtamed in the KSB Test Facility, RRC KI 1991)
Coast-down of MCP (Component Test of One MCP, data on the MCP obtained in the course of the pump component test performed by the designer SDBME, Special Design Bureau of Mechanical Engineering, Nischni Novgorod, Russia) (Tests on the MCP Coast-down at the Ignalina NPP Unit 1, data obtained in testing the MCP coast-down behaviour at the Ignalina NPP Unit 1 1984)
On the basis of the analysis performed the following conclusions were drawn:
- There are experimental data available, which are suitable for supplementary validation of Western thermalhydraulic codes for some specific RBMK processes and phenomena.
- There are no experimental data of integral test modeling LOCA-type accidents.
- Integral test for the study of transients were performed at BM (NIKIET) and 108 (EREC) test facilities. Experimental data obtained with these facilities in 70-ties and 80-ties are of low quality .The facilities have design shortcomings in modelling the main circulation circuits of RBMKs. For example, the channels of test facility 108 have no FA models and the BM test facility has simplified, scaled down FC models.

The PSB-RBMK thermalhydraulic test facility integral tests for the investigation of LOCA-type accidents and transients can be performed. Design of the test facility makes it possible to model accidents and transients identified in the matrices. Most of the processes and phenomena that are characteristic of those modes can be reproduced. The test facility does not allow to model cross heat fluxes between channels, steam-zirconium reaction and neutronic processes.

3) General description of the PSB-RBMK facility

The PSB RBMK test facility models one loop of the RBMK main circulation circuit including models of all important MCC componets (full-scaled models of six fuel channels of RBMK, drum separator model, models of suction, pressure and group distributions headers, models of MCC valves). The test facility is also provided with simulators of the main systems of the reactor plant (ECCS, feed water system, emergency protection system). To investigate LOCA regimes a leak line is connected to the modelled break points. The second half of the reactor is not modelled in the test facility and also no steam flow is modelled, which could simulate the influence of the intact half on the accident half under LOCAs as well.

The main parameters of PSB RBMK:
Scale: Elevations 1:1
Volume-power scale for one FC 1:1
Volume-power scale for one reactor loop 1:140
Maximal electrical power of test facility, MW 11
Maximal coolant flow rate, kg/s 66.7
Maximal pressure in drum separator, Mpa 10
Feed water temperature, �C 170
Number of loops 1
Number of FC in a loop 6

The test facility has some significant limitations and simplifications, i.e. 6 FC models instead of 830 FCs, 2 GDH instead of 22, MCP coastdown is not modelled, steam pipelines are not modelled, cross heat flux between channels is not modelled. Therefore experimental data obtained with the test facility cannot be directly used for reactor accident analysis. However the test facility circuit includes models of all main equipment of RBMK MCC. The test facility includes models of main systems of the reactor. The test facility makes it possible to investigate influence of different phenomena to the course of accidents and transients. Therefore it is reasonable to obtain with the test facility experimental data for thermalhydraulic codes validation. Tests scenarios are developed on the basis of information on the NPPs accidents course of development.

4) Development of the Experimental Program for PSB RBMK Test Facility

The EREC specialists provided a list of experiments planned on the PSB RBMK test facility. Brief descriptions of the emergency modes and suggestions on their modeling at the PSB RBMK test facility were presented. Four tests were found to be of the highest priority for computer code validation.
The analysis carried out has shown insufficiency of the volume of experimental data from integral tests, which reproduce accidental modes. PSB RBMK test facility design would make it possible to perform integral test investigation of most of the accidents listed in the Validation Matrices. LOCA type accidents are studied to a lesser extent among them.
The list of emergency modes selected for experimental investigation is shown:
Investigation of thermal-hydraulic processes that occur in NPP with RBMK-type reactor transients:

Shut-down of one or several operating MCPs ("coolant flow rate decrease")
One GDH flow rate decrease due to destruction of GDH check valve plate or blocking of mechanical filter in the GDH inlet ("coolant flow rate decrease")
Failures in feed water supply system ("loss of feed water")
Core cooling under natural circulation with loss of AC power, opening and non fit of MSV ("loss of AC power")
Investigation of thermalhydraulic processes that occur during LOCAs

FC flow rate decrease (including the case of zero flow rate) due to ICV blocking ("FC loss of flow")
Double-end MCP PH break ("PH break")
Double-end GDH break downstream of the check valve ("GDH break")
Double-end GDH break downstream of the check valve and failure of the neighbouring check valve ("GDH break with CV failure")
Partial break of GDH downstream of the check valve (“partial critical break")
Steam line break beyond the accident localization rooms (“steam line break")
Downcomer pipeline break close to pumps suction header, including the case of failure of GDH CV ("downcomer break")
Investigation of thermalhydraulic processes that occur during accidents with safety systems failures

Residual heat removal under unit blackout
Double-end MCP PH break and ECCS pumps and valves failures
Double-end GDH break and: ECCS pumps and valves failures, Scram actuation delay
On the basis of above it is possible to conclude that validation of computer codes is the mostly required for the following modes:

�"Critical" partial break of GDH.
Mechanical filter blockage in the inlet of GDH (GDH loss of coolant flow).
Double-end GDH break downstream of the check valve and failure of the neighbouring check valve ("GDH break with CV failure").
Residual heat removal under unit blackout.
The high priority of these tests is based on the fact that according to estimations performed the heat release and heat removal are highly unbalanced in the core, which causes or may cause damage of physical safety barriers.
First of all, it is necessary to obtain experimental data for these modes, which are needed for validation. The other experimental investigations should be performed after those of the first priority according to possibility and available funding.

5) Pre-Test analysis of experiments at the PSB-RBMK Test facility with the TH code ATHLET

The purpose of the first analysis was to assess test facility capabilities and quality of the input deck for the PSB RBMK test facility.
First pre-test: Core cooling under natural circulation with loss of AC power, opening and non-fit of MSV ("loss of AC power") was selected, but failure of complete MSV valve closing was not assumed, so the mode studied is: Core cooling under natural circulation with loss of AC power. This transient does not invoke serious consequences.
Second pre-test: it should demonstrate the capability of the code and the input deck in simulating a LOCA scenario. Partial break of GDH downstream of the check valve (“partial critical break"). In this test coolant boiling in the circulation circuit will occur and as a result steam may enter the MCPs. Since it is not allowed to run the MCP pumps in the PSB RBMK test facility under two phase flow conditions, the pumps are to be cut off from the circuit, when in course of the experiments water temperatures at pump inlet come close to saturation temperature. In such a case the mode studied is: Partial GDH break downstream of the check valve with loss of AC power.
The test facility was under construction during the project implementation. Therefore the input data base was compiled mainly on the basis of design documentation. It was necessary to analyse available documentation and to consider experience obtained during operation of similar thermalhydraulic test facilities built and tested earlier.
Russian experts developed a zero-version of the input deck for code ATHLET MOD1.2-Cycle B. This zero-version input deck was used to model initial steady state operation of the test facility. The mode parameters applied for the simulation of steady state in course of the computer calculation corresponded to the normal parameters of the reactor in operation.
For the performance of the final analyses it was decided to use the new version of the code, which is ATHLET MOD1.2-Cycle D. It was necessary to install the new code version on the EREC computers along with the service software package, and to instruct personnel for working with this software. A GRS expert visited EREC to install the new version of ATHLET code to the EREC computers and to instruct EREC personnel in working with all supplementary software (e.g., input graphics software), which facilitates computer analysis. First the zero-version of the input deck was adapted to the requirements of the new code version.