عدم اطمینان و آنالیز حساسیت رویدادهای سفر MCPs در Ignalina NPP

Eight main circulation pumps (MCPs) are employed for the cooling water forced circulation through the RBMK-1500 reactor at the Ignalina nuclear power plant (NPP). There have been a few events when one or more MCPs were inadvertently tripped. This paper presents investigation of a one MCP trip event and all MCPs’ trip events at Ignalina NPP. Thermal-hydraulic analysis was conducted using the best estimate system code RELAP5/MOD3.3. Uncertainty and sensitivity analysis of flow energy loss in different parts of the main circulation circuit (MCC), initial conditions and code-selected models was performed. Such analysis allows to estimate the influence of separate parameters on the calculation results and find those modelling parameters that have the largest impact on the investigated events. Uncertainty analysis indicates that natural circulation provides adequate cooling in the case of all MCPs tripped, and that the reactor is reliably cooled by forced circulation in the case of a single tripped MCP. On the basis of this analysis, recommendations for the further improvement of model are developed.

The Ignalina nuclear power plant (NPP) is a twin-unit with two RBMK-1500, graphite moderated, boiling water, multichannel reactors. Several important design features of RBMK-1500 are unique and extremely complex with respect to western reactors (the single coolant loop has a very long flow path of more than 200 m and consists of 1661 parallel pressure tubes and numerous components such as headers, pumps, valves, etc.) (Almenas et al., 1998). The main circulation circuit (MCC) consists of two identical halves—the left and right loops. The schematic representation of one MCC loop is shown in Fig. 1. Eight main circulation pumps (MCPs) are employed for the cooling water forced circulation through the RBMK-1500 reactor at the Ignalina NPP. The MCPs (Fig. 1(5)) are joined in groups of four pumps each (three for normal operation and one on standby). The MCPs feed a common pressure header (Fig. 1(8)) on each side of the reactor. Each pressure header provides coolant to 20 group distribution headers (GDH) (Fig. 1(9)), each of which in turn feeds from 38 to 43 fuel channels (FCs) (Fig. 1(11)). The coolant flow rate through individual FCs is regulated by isolating and control valves (ICV) (Fig. 1(10)) mounted in the lower water communication lines. Coolant passing through FCs is boiled and part of the water is evaporated. Steam–water mixture flows through steam–water communication lines (Fig. 1(12)) to drum separators (DS). Steam, separated in the DS, is supplied to the turbines through steam lines (Fig. 1(13)). Full-size image (11 K) Fig. 1. Schematic representation of one loop of the RBMK-1500 MCC: (1) drum separtor; (2) downcomers; (3) MCP suction header; (4) MCP suction piping; (5) MCPs; (6) MCP discharge piping; (7) bypass line; (8) MCP pressure header; (9) GDHs; (10) lower water communication line; (11) fuel channel; (12) steam–water communication line; (13) steam lines. Figure options In the all pumps trip case, the coolant during the first few seconds was supplied to the reactor by pumps coastdown due to the high inertia of pump flywheel. Later natural circulation through the core was established. Reactor scram was initiated by the emergency protection system in response to multiple pump trips. In the case of one MCP trip the throughput of the two running pumps in the affected MCC loop is increased, however, the total coolant flow through the affected loop is decreased. The reactor power was decreased down to 60% from design (4800 MWth) in response to one MCP trip signal. For validation of the RELAP5 model of Ignalina NPP a benchmark analysis of natural circulation phenomenon was performed. Data of operational occurrences, measured at Ignalina NPP were compared with the RELAP5 calculations. This paper presents benchmark analysis of one and all MCP trip events. Similar analysis is presented in the preceding works (Kaliatka and Uspuras, 2000; Uspuras et al., 2001; Kaliatka et al., 2001; Vileiniskis et al., 2002). In contradistinction to preceding works, this paper presents uncertainty and sensitivity analysis of these events. Such analysis allows to estimate the influence of separate parameters to the calculation results. On the basis of this analysis the recommendations for the further improvement of model are developed.

Sensitivity analysis of flow energy loss in different parts of MCC, initial unit conditions and code-selected models was performed for one and all MCP trip events. The performed analysis shows that the largest impact on the calculation results has the selection of the homogeneous media model in coolant natural circulation case. Due to the fact that homogeneous media model is too conservative and could be too big source of uncertainties, in future during accident analyses the use of only the non-homogeneous model is proposed. Since in case of a single tripped MCP, the largest impact on the calculation results uncertainties was the flow energy loss in ICV, it is recommended in cases of forced circulation to model the reactor core with more details, i.e. the core must be represented by possibly more channel groups. Performed sensitivity and uncertainty analysis demonstrated that the reactor core is reliably cooled in case of a single tripped MCP and all MCP trip events even taking into account possible sources of uncertainties (measurement errors of initial plant conditions, code selection models).