Font Size: a A A

Study On Prediction Models Of Wind Fields And Plume Concentrations For Early Emergency Response To Nuclear Accidents

Posted on:2004-11-07Degree:DoctorType:Dissertation
Country:ChinaCandidate:X Q ChenFull Text:PDF
GTID:1102360095461998Subject:Radiation protection and environmental protection
Abstract/Summary:PDF Full Text Request
In order to protect the public and environment effectively, and reduce the damage to minimum, it is necessary to assess the consequential emergency radiation exposure arising from a nuclear accident and to make decision in time. The validity of the emergency response depends on the intervention timing. Before the radioactive plume arriving, if the emergency protective countermeasures could be implemented adequately, the avertable doses could be realized and the radiological hazards could be mitigated.In the early stage of emergency response for a nuclear accident, the assessment of consequential radiation exposure is the basis of intervention. Prognosticating local wind fields, predicting the trajectory and spread of the radioactive plume and estimating the radiation dose in time can present the characteristics of radioactive plume transportation and diffusion, as well as the range and degree of radiological consequence arising from plume exposure, which is very important in the decision making.This dissertation composes of two parts: the first one is the prognostication of local wind field; and the second is the prediction of radioactive plume. Considering the elaborate atmospheric diffusion model is dependent on the high quality of wind fields, the study focus on the aspect of wind field prognostication that is on the basis of a non-hydrostatic, full compressible dynamics model, and the initial wind field provided by diagnostic wind field model. The plume prediction is on the basis of a particle random walking model, in which the average wind transport speed and the parameters of turbulent fluctuation are source from the non-hydrostatic, full compressible dynamics model.The model development of wind field prognostication and plume prediction base on Windows 98. The codes are written by FORTRAN90. The wind field diagnostic model, the wind field prognostic model and the radioactive plume concentration prediction model are integral considered in the system development. The equations of wind field prognostic model are computed using spatial centered finite differences on the Arakawa staggered C-grid. The time integration follows the split time step approach. The terms evaluated by large time step include the advective, gravity, inertial wave modes, and diffusion terms. The terms evaluated by small time step represent the acoustic modes and include the pressure gradient and divergence terms. The pressure and vertical velocity are coupled in the vertical through an implicit numerical manner, which removes the vertical sound constraint on the small time step. A 1.5 order closure scheme is used to predict turbulent kinetic energy for determining the eddy viscosity.The system is applied to simulate the Qinshan and Daya Bay NPP sites respectively. The wind field prognostic result is tested in Qinshan site, and the validation of the predicted plume concentration is verified in Daya Bay site.The simulation of the Qinshan nuclear power site shows that the agreement of the prognostic wind speed is better than that of the prognostic wind direction in comparing with the observed results. In the first six hours, the results of prognostic wind field are coherent with meteorological observation data in the whole. The statistical analysis results show that the prognostic results comparing with the observed data, about 93% of the difference of wind direction within the 1.5 of azimuth angle (<33.75 degree), 80% of the wind speed radios are within a factor of 1.5 in the first two hours.The simulation of the Daya bay nuclear power site shows that the particle spreading image and the time integrated concentration distribution can perform the variation of pathway of the pollutant transport by plume concentration prediction model, as well as the topographic effects. The simulations of five trace tests in field shows that above 59.1% of the radios between the predicted results and the observed results are within a factor of 10, and 41% of the radios are within a factor 5 approximately.The study show that the system usin...
Keywords/Search Tags:nuclear accident emergency response, consequence assessment, non-hydrostatic, random walk
PDF Full Text Request
Related items