SBLOCA, 3-loop model, 900 MWe. Calculations code: RELAP5, CATHARE
During a Scenario 1 model of the typical 900 MW Nuclear Power Plant was developed. It consists of 3 primary loops with pressurized water (PWR) reactor as a heat source. The model was made both in RELAP5 and CATHARE codes. During transient calculation loos of coolant accident is occurred with a 6 inch break.
Model of the PWR made in CATHARE has been developed by experts from AREVA and shared with Warsaw University of Technology – SARWUT team as a reference model. Nodalization of the reactor consists of 2 loops. Intact loop has weighting factor equal to 2. Safety systems are modeled with use of gadget components which are attached to the cold leg of the intact loop at the length of 10.5 m from the reactor pressure vessel (RPV). Pressurizer is attached to the hot leg of the broken loop by a surge line at the length 3. 85 m from the RPV. Pressurizer is a one volume component. Nodalization of the pressurizer can by simplified and approximated by one volume due to the fact that pressurizer empties at the beginning stage of the transient calculations, therefore it has no impact on the later stage of the transient. Break is modeled with use of boundary conditions which is inactive during steady state conditions (there is no flow through the break).The break is connected to the cold leg by a 1 m axial component with cold leg at the length around 3.6 m from the RPV. RPV is modeled with use of 7 components. 4 of them are volume components and 3 of them are axial components. Coolant flows into the inlet plenum where it is divided into two streams. At the initial state, 2 % of coolant flows into the upper head volume which simulates volume of the upper plenum of the RPV and the free volume of the guide tubes. Rest of the water comes in the downcomer which is simulated by one axial component. From the downcomer water enters the lower plenum. 95% of the total flow of the downcommer during initial state flows in to the core region. Hydraulic part of the core is divided into 12 volume components with rod-bundle geometry. Fuel in the core is irradiated and the power of the reactor is calculated by a reactor kinetics models. Water from the core bypass and the core is collected in the outlet plenum. Steam generators (SG) are modeled with use of two components. Downcomer of the SG and the riser are connected by a very short horizontal pipe which increase its area from downcomer area to the riser area. All three components are modeled with one axial. Riser part of the axial is connected with volume which simulates steam dome and free volume of the steam separator. At the junction of the riser with volume, model of the separation of two phases is set. One of the SG (connected with intact loop by heat exchange) has weighting factor equal to 2. Steam pipes from both SGs are connected via volume with boundary condition to simulate the turbine.
On the other hand, the RELAP5 model consists of 3 separate loops with several break valves designed to simulate postulated accidents. One of the loops holds the break pipe, a 1.031 meter long pipe 6 inches in diameter with a trip valve at the end, corresponding to the AREVA nodalization.
For each loop the hot legs are modeled by pipe 300 (400, 500), branch 310 (410, 510) and single volume 320 (420 or 520). The horizontal part of hot leg is divided into two components in order to properly connect pressurizer in loop 2, and the proper simulation of hot leg breaks.
The steam generator inlet plenums are modeled by single volumes 330 (430, 530). Pipes 340 (440 and 540) represent the many thousands of steam generator U-tubes lumped into the pipe component. The 10-node pipe nodalization has been demonstrated to be sufficient to model, with enough fidelity, phenomena associated with reflux cooling and significantly reduced SG secondary side water levels.
Single volumes 350 (450 and 550 for loops 2 and 3, respectively) represent the steam generator outlet plenums. Pipes 460 (360 and 560) represent the pump suction cold leg. The pump components 466 (366, 566) are used for modeling the primary coolant pumps in loops 2, 1 and 3, respectively. The pump discharge cold legs are modeled with single volumes 470 (370, 570), branches 480 (380, 580) and pipes 490 (390, 590). The nodalization has been demonstrated to be sufficient to properly simulate horizontal stratification of fluid within the cold leg during loss-of-coolant accidents. For each loop the high and low pressure safety injection system functions are modeled with pairs of time dependent volumes and junctions. The fluid injection temperature is specified by the time dependent volumes while the fluid flow rate is specified as a function of pressure in the cold legs by time-dependent junctions. For the small break only intact loop injections are active and those are the HPIS and the Safety Pump with reference to the CATHARE model. An accumulator component 572 (372, 472) is used to simulate the injection of highly borated water from the nitrogen-charged accumulators. In loop 3, however, the accumulator is inactive, closed with a valve. For each primary loop all injection lines of the ECCS, for simplicity, are connected through single-volume component 578 (378, 478) to one branch component 580 (380 and 480 for loops 1 and 2, respectively) representing the ECC mixing part of the cold legs. The valve component 531 (331, 431) connecting the SG inlet plenum 530 (330, 430) with the SG secondary side riser 800 (600, 700) is used to simulate a double-sided break of a single SG U-tube. Break pipe 3 is used to simulate a 6inch break of the cold leg.
On the SG secondary side, main feedwater enters the downcomer annulus at branch 620 (720, 820) and is mixed with the recirculation flow returning from the separator component 610 (710, 810).The combined flow then flows downwards the downcomer annulus 630 (730, 830) and enters the riser modeled with a pipe 600 (700 and 800 for SGs 2 and 3, respectively). The riser pipe is subdivided into 11 axial nodes, where the first 8 volumes are thermally connected to the primary side U-tubes, while the other nodes simulate the channels feeding the separator swirls. The separator cylinders are modeled by the separator component 610 (710, 810). The volume between the outermost separator cylinders and the inside of the SG shell is modeled by single volume 612 (712, 812). The top of SG, including the dryer and the steam dome regions, are modeled by branches 642 (742, 842), 640 (740, 840) and single-volume 650 (750, 850).
The reactor core is modelled by pipe component 170 with 14 axial nodes, where nodes 02 through 13 represent the heated length of the core.
The reactor core consists of 157 fuel assemblies in a standard geometrical arrangement. The fuel assembly consists 264 fuel rods designed in the conventional 17X17 array. The fuel rods are supported by a bottom grid, 6 mid grids, 3 intermediate flow mixers and a top grid. The fuel assembly also consists of 24 guide thimble tubes and one instrumentation tube. The nominal reactor thermal power is equal to 2831.5 MWth. For the LOCA studies a conservative, chopped cosine axial power distribution is used. During the normal operation and slow transients the heat structure power is evaluated by means of the point reactor kinetic model. The reactor scram power curve is defined by general table.
downcomer annulus at branch 620 (720, 820) and is mixed with the recirculation flow returning from the separator component 610 (710, 810).The combined flow then flows downwards the downcomer annulus 630 (730, 830) and enters the riser modeled with a pipe 600 (700 and 800 for SGs 2 and 3, respectively). The riser pipe is subdivided into 11 axial nodes, where the first 8 volumes are thermally connected to the primary side U-tubes, while the other nodes simulate the channels feeding the separator swirls. The separator cylinders are modeled by the separator component 610 (710, 810). The volume between the outermost separator cylinders and the inside of the SG shell is modeled by single volume 612 (712, 812). The top of SG, including the dryer and the steam dome regions, are modeled by branches 642 (742, 842), 640 (740, 840) and single-volume 650 (750, 850).
For the 6-inches cold leg break, the results of both codes showed in the transient the same dominant physical phenomena of an IB-LOCA such as: single phase depressurization, creation of void and liquid stratification in the primary system, drop of core level by the reduction in water inventory,…but still different in details. The first attempt of the 6-inches cold leg break modeling demonstrated the capability of the WUT team to create an input decks for both codes in short time. Continuing work between WUT and AREVA would be needed to improve the results of comparison.