两种大气扩散模型在核电厂址小尺度扩散建模中的对比

王朝阳; 董信文; 杨力; 徐宇涵; 方晟; 李新鹏; 陈义学

PDF(5413 KB)

中国环境科学 ›› 2025, Vol. 45 ›› Issue (7) : 3667-3680.

大气污染与控制

两种大气扩散模型在核电厂址小尺度扩散建模中的对比

王朝阳¹, 董信文², 杨力¹, 徐宇涵², 方晟², 李新鹏¹, 陈义学¹

作者信息 +

Comparison of two atmospheric dispersion models in Small-Scale dispersion modeling at nuclear power plant sites.

WANG Zhao-yang¹, DONG Xin-wen², YANG Li¹, XU Yu-han², FANG Sheng², LI Xin-peng¹, CHEN Yi-xue¹

Author information +

文章历史 +

摘要

对Micro-SWIFT SPRAY (MSS)模型与GPU加速的QES模型(GPU-QES)在复杂核电厂址环境中的小尺度大气扩散性能进行了对比.基于三门核电风洞实验数据,使用两种模型模拟了两种典型风向扩散场景,对比了风场和浓度场预测结果,包括地面水平风场与浓度场以及垂直风速剖面和浓度剖面.结果表明,GPU-QES在风速预测方面优于MSS,特别是在建筑中间区域,风速低估幅度减少了约53%.此外,GPU-QES生成三维风场的速度是MSS的450倍,能够在4s内生成覆盖9km²区域,包含3500万个网格点的风场数据.在羽流扩散预测中,两种模型均能有效再现高浓度区域,而在低浓度区域,GPU-QES模型的模拟值与真实值的一致性更高.统计分析显示,两种模型均满足了国际可接受模型指标,但GPU-QES在大多数指标上表现出更高的模拟精度,特别是在建筑地面区域,其模拟值在测量值2倍线以内的比例(FAC2)比MSS高出4.5%.此外,由于MSS与GPU-QES在建筑后方扩散行为的差异,将两模型结果取平均后,模拟精度得到了提高(FAC2达到71.8%).另一方面,由于浓度计算方式不同,GPU-QES生成羽流速度比MSS快约48倍.因此,在需要快速且精准预测的核应急响应场景中,GPU-QES相较于MSS显示出更明显的优势.

Abstract

This study evaluated the small-scale atmospheric dispersion performance of the Micro-SWIFT SPRAY (MSS) model and the GPU-accelerated QES model (GPU-QES) in the complex environment of a nuclear power plant site. Based on wind tunnel experimental data from the Sanmen Nuclear Power Plant, two typical wind direction dispersion scenarios were simulated by both models, and the results were compared in terms of wind field and concentration field predictions, including horizontal ground-level wind and concentration fields, as well as vertical wind speed and concentration profiles. The results indicated that GPU-QES outperformed MSS in wind speed prediction, especially in the areas between buildings, where the underestimation of wind speed is reduced by about 53%. Furthermore, GPU-QES simulated the three-dimensional wind field 450times faster than MSS, generating wind data over a 9km² area with 35million grid points in just 4s. In plume dispersion prediction, high-concentration areas were effectively reproduced by both models, but GPU-QES showed higher consistency with actual values in low-concentration regions. Statistical analysis revealed that both models meet internationally accepted model criteria. However, higher simulation accuracy was demonstrated by GPU-QES in most metrics, especially in the building ground area, where the fraction of predictions within a factor of two of the observations (FAC2) was 4.5% higher than that of MSS. Additionally, due to differences in the dispersion behavior behind buildings between MSS and GPU-QES, simulation accuracy was improved by averaging the results of both models (FAC2reached 71.8%). On the other hand, due to differences in concentration calculation methods, GPU-QES generated the plume approximately 48times faster than MSS. Therefore, a clear advantage was shown by GPU-QES over MSS in nuclear emergency response scenarios that required rapid and accurate predictions.

导出引用

王朝阳, 董信文, 杨力, 徐宇涵, 方晟, 李新鹏, 陈义学. 两种大气扩散模型在核电厂址小尺度扩散建模中的对比[J]. 中国环境科学. 2025, 45(7): 3667-3680

WANG Zhao-yang, DONG Xin-wen, YANG Li, XU Yu-han, FANG Sheng, LI Xin-peng, CHEN Yi-xue. Comparison of two atmospheric dispersion models in Small-Scale dispersion modeling at nuclear power plant sites.[J]. China Environmental Science. 2025, 45(7): 3667-3680

中图分类号： X591

参考文献

[1] 李幻婷,杨力,宋嘉悦,等.四种气象数据在福岛核事故¹³⁷Cs扩散模拟中的适用性 [J]. 中国环境科学, 2024,44(11):6025-6035. Li H T, Yang L, Song J Y, et al. Applicability of four types of meteorological data to fukushima nuclear accident ¹³⁷Cs simulations. [J]. China Environmental Science, 2024,44(11):6025-6035.
[2] 董龙翔,余晔,左洪超,等.WRF-Fluent耦合模式的构建及其对城市大气扩散的精细化模拟 [J]. 中国环境科学, 2019,39(6):2311-2319. Dong L X, Yu Y, Zuo H C, et al. Elaborated simulation of urban atmospheric dispersion using an established WRF-Fluent coupling model. [J]. China Environmental Science, 2019,39(6):2311-2319.
[3] 董龙翔,左洪超,杨宾,等.边界入流风场扰动对模拟城市大气扩散的影响 [J]. 中国环境科学, 2017,37(9):3210-3219. Dong L X, Zuo H C, Yang B, et al. Influence of boundary inflow wind fluctuation on simulated urban atmospheric diffusion. [J]. China Environmental Science, 2017,37(9):3210-3219.
[4] 韩素芹,蔡旭晖,李培彦,等.天津市突发事故条件下大气污染的应急模拟 [J]. 中国环境科学, 2009,29(9):919-923. Han S Q, Cai X H, Li P Y, et al. Numerical simulation of emergency response to atmospheric pollution accident in Tianjin. [J]. China Environmental Science, 2009,29(9):919-923.
[5] 葛宝珠,陆芊芊,陈学舜,等.放射性核素大气扩散数值模拟研究综述 [J]. 环境科学学报, 2021,41(5):1599-1609. Ge B Z, Lu Q Q, Chen X S, et al. A review of the numerical simulations of the atmospheric dispersion of radionuclides [J]. Acta Scientiae Circumstantiae, 2021,41(5):1599–1609.
[6] 田飞,伯鑫,薛晓达,等.基于复杂地形-气象场的二噁英污染物沉降研究 [J]. 中国环境科学, 2019,39(4):1678-1686. Tian F, Bo X, Xue X D, et al. Study on settlement of dioxin pollutants under complex terrain-weather conditions [J]. China Environmental Science, 2019,39(4):1678-1686.
[7] 杨力,张钰杰,方晟,等.气态放射性核素扩散模拟研究——分裂烟团模型的验证与应用 [J]. 清华大学学报(自然科学版), 2024, 64(12):2019-2030. Yang L, Zhang Y J, Fang S, et al. Research on the simulated dispersion of gaseous radionuclides: Validation and application for splitting puff dispersion model [J]. Journal of Tsinghua University (Science and Technology), 2024,64(12):2019-2030.
[8] Perry S G, Cimorelli A J, Paine R J, et al. AERMOD: A dispersion model for Industrial source applications. Part II: Model performance against 17 field study databases [J]. Journal of Applied Meteorology, 2005,44(5):694-708.
[9] Camelli F E, Hanna S R, Lohner R. Simulation of the must field experiment using the feflourban cfd model. [J]. American Meteorological Society, 2004,13(12):23-26.
[10] Patnaik G, Boris J P, Grinstein F F. Large scale urban simulations with the MILES approach (Invited) [C]//16th AIAA Computational Fluid Dynamics Conference, 2003:23-26.
[11] Hanna S, White J, Trolier J, et al. Comparisons of JU2003 observations with four diagnostic urban wind flow and Lagrangian particle dispersion models [J]. Atmospheric Environment, 2011,45(24): 4073-4081.
[12] Röckle R. Bestimmung der Strömungsverhältnisse im Bereich komplexer Bebauungsstrukturen [Z]. Technische Universität Darmstadt, Germany, 1990.
[13] Kaplan H, Dinar N. A Lagrangian dispersion model for calculating concentration distribution within a built-up domain [J]. Atmospheric Environment, 1996,30(24):4197-4207.
[14] Mortarini L, Carlino G, Tinarelli G, et al. Development of the microscale lagrangian particle dispersion model MicroSpray for the simulation of two-phase releases [C]//HARMO 2011-Proceedings of the 14th international conference on harmonisation within atmospheric dispersion modelling for regulatory purposes. University of Western Macedonia, 2011:704-707.
[15] Singh B, Hansen B S, Brown M J, et al. Evaluation of the QUIC-URB fast response urban wind model for a cubical building array and wide building street canyon [J]. Environmental Fluid Mechanics, 2008,8: 281-312.
[16] Bozorgmehr B, Willemsen P, Margairaz F, et al. QES-Winds v1. 0: theory and user’s guide [Z]. University of Utah: Salt Lake City, UT, USA, 2021.
[17] Tinarelli G, Brusasca G, Oldrini O, et al. Micro-Swift-Spray (MSS): a new modelling system for the simulation of dispersion at microscale. General description and validation [C]//Air Pollution Modelling and Its Applications XVII, 2007:449-458.
[18] Oldrini O, Armand P, Duchenne C, et al. Description and preliminary validation of the PMSS fast response parallel atmospheric flow and dispersion solver in complex built-up areas [J]. Environmental Fluid Mechanics, 2017,17:997-1014.
[19] Cox R M, Cogan J, Sontowski J, et al. Comparison of atmospheric transport calculations over complex terrain using a mobile profiling system and rawinsondes [J]. Meteorological Applications, 2000,7(4): 285-295.
[20] Cox R M, Sontowski J, Dougherty C M. An evaluation of three diagnostic wind models (CALMET, MCSCIPUF, and SWIFT) with wind data from the dipole pride 26field experiments [J]. Meteorological Applications, 2005,12(4):329-341.
[21] Gomez F, Ribstein B, Makké L, et al. Simulation of a dense gas chlorine release with a lagrangian particle dispersion model (LPDM) [J]. Atmospheric Environment, 2021,244:117791.
[22] Liu Y, Li H, Sun S, et al. Enhanced air dispersion modelling at a typical chinese nuclear power plant site: Coupling RIMPUFF with two advanced diagnostic wind models [J]. Journal of Environmental Radioactivity, 2017,175:94-104.
[23] Dong X, Zhuang S, Fang S. Local-and small-scale atmospheric dispersion modeling towards complex terrain and building layout scenario using Micro-Swift-Spray [C]//International Confernece Pacific Basin Nuclear Conference. Singapore: Springer Nature Singapore, 2022:133-145.
[24] Dong X, Zhuang S, Fang S, et al. Validation and sensitivity study of Micro-SWIFT SPRAY against wind tunnel experiments for small-scale air dispersion modeling between mountains and dense building at a nuclear power plant site [J]. Progress in Nuclear Energy, 2021, 142:104007.
[25] Wang S, Li X, Fang S, et al. Validation and sensitivity study of Micro-SWIFT SPRAY against wind tunnel experiments for air dispersion modeling with both heterogeneous topography and complex building layouts [J]. Journal of Environmental Radioactivity, 2020, 222:106341.
[26] Margairaz F, Singh B, Gibbs J A, et al. QES-Plume v1. 0: a Lagrangian dispersion model [J]. Geoscientific Model Development, 2023,16(20):5729-5754.
[27] Bozorgmehr B, Willemsen P, Gibbs J A, et al. Utilizing dynamic parallelism in CUDA to accelerate a 3D red-black successive over relaxation wind-field solver [J]. Environmental Modelling & Software, 2021,137:104958.
[28] Moody M J, Gibbs J A, Krueger S, et al. QES-Fire: a dynamically coupled fast-response wildfire model [J]. International journal of wildland fire, 2022,31(3):306-325.
[29] Ulmer L, Margairaz F, Mahaffee W F, et al. A fast-response model of turbulence and passive scalar transport in row-organized canopies [J]. Agricultural and Forest Meteorology, 2024,349:109919.
[30] Williams M D, Nelson M A, Brown M J. Condensed QUIC-PLUME theory guide for QUIC-FST [Z]. Los. Alamos. National Lab.(LANL), Los Alamos, NM (United States), 2004.
[31] Margairaz F, Eshagh H, Hayati A N, et al. Development and evaluation of an isolated-tree flow model for neutral-stability conditions [J]. Urban Climate, 2022,42:101083.
[32] Kaplan H, Dinar N. A Lagrangian dispersion model for calculating concentration distribution within a built-up domain [J]. Atmospheric Environment, 1996,30(24):4197-4207.
[33] Kolmogorov A N. The local structure of turbulence in incompressible viscous fluid for very large Reynolds [J]. Numbers. In Dokl. Akad. Nauk SSSR, 1941,30:301.
[34] Rodean H C. Stochastic lagrangian models of turbulent diffusion [M]. Boston: American Meteorological Society, 1996.
[35] Thomson D J. Criteria for the selection of stochastic models of particle trajectories in turbulent flows [J]. Journal of Fluid Mechanics, 1987, 180:529-556.
[36] Lodge J. Handbook on atmospheric diffusion [J]. Atmospheric Environment, 1983,17(3):673–675.
[37] Speckart S O, Pardyjak E R. A method for rapidly computing windbreak flow field variables [J]. Journal of Wind Engineering and Industrial Aerodynamics, 2014,132:101-108.
[38] Hayati A N, Stoll R, Pardyjak E R, et al. Comparative metrics for computational approaches in non-uniform street-canyon flows [J]. Building and Environment, 2019,158:16-27.
[39] Pardyjak E R, Brown M. QUIC-URB v. 1.1: Theory and user’s Guide [Z]. Los Alamos National Laboratory, Los Alamos, NM, 2003.
[40] Singh B, Hansen B S, Brown M J, et al. Evaluation of the QUIC-URB fast response urban wind model for a cubical building array and wide building street canyon [J]. Environmental Fluid Mechanics, 2008,8: 281-312.
[41] Ulmer L, Margairaz F, Bailey B N, et al. A fast-response, wind angle-sensitive model for predicting mean winds in row-organized canopies [J]. Agricultural and Forest Meteorology, 2023,329:109273.
[42] Sasaki Y. Some basic formalisms in numerical variational analysis [J]. Monthly Weather Review, 1970,98(12):875-883.
[43] Sherman C A. A mass-consistent model for wind fields over complex terrain [J]. Journal of Applied Meteorology and Climatology, 1978,17 (3):312-319.
[44] Pope S B. Turbulent flows [J]. Measurement Science and Technology, 2001,12(11):2020-2021.
[45] Alfonsi G. Reynolds-averaged Navier-Stokes equations for turbulence modeling [J]. Applied Mechanics Reviews, 2009,62(4):040802.
[46] Smagorinsky J. General circulation experiments with the primitive equations: I. The basic experiment [J]. Monthly Weather Review, 1963,91(3):99-164.
[47] Bailey B N. Numerical considerations for Lagrangian stochastic dispersion models: Eliminating rogue trajectories, and the importance of numerical accuracy [J]. Boundary-Layer Meteorology, 2017,162: 43-70.
[48] Du S. Universality of the Lagrangian velocity structure function constant (C0) across different kinds of turbulence [J]. Boundary-Layer Meteorology, 1997,83:207-219.
[49] Yee E, Wilson J D. Instability in Lagrangian stochastic trajectory models, and a method for its cure [J]. Boundary-layer Meteorology, 2007,122:243-261.
[50] Thomson D J. Criteria for the selection of stochastic models of particle trajectories in turbulent flows [J]. Journal of Fluid Mechanics, 1987, 180:529-556.
[51] Yang L, Fang S, Zhuang S, et al. Atmospheric ¹³⁷Cs dispersion following the fukushima daiichi nuclear accident: Local-scale simulations using CALMET and LAPMOD [J]. Annals of Nuclear Energy, 2024,195:110137.
[52] Dong X, Zhuang S, Fang S, et al. Multi-scenario validation of CALMET-RIMPUFF for local-scale atmospheric dispersion modeling around a nuclear powerplant site with complex topography [J]. Journal of Environmental Radioactivity, 2021,229:106547.
[53] Chang J C, Franzese P, Chayantrakom K, et al. Evaluations of CALPUFF, HPAC, and VLSTRACK with two mesoscale field datasets [J]. Journal of Applied Meteorology and Climatology, 2003,42(4): 453-466.
[54] Hanna S, Chang J. Acceptance criteria for urban dispersion model evaluation [J]. Meteorology and Atmospheric Physics, 2012,116(3): 133-146.
[55] 杨力,王存友,陈义学,等.四种拉格朗日粒子浓度计算方法的评估——箱式计数法,高斯核,均匀核和抛物线核 [J]. 中国环境科学, 2023,43(7):3404-3415. Yang L, Wang C Y, Chen Y X, et al. Evaluation of four lagrangian particle concentration calculation methods-Box counting, Gaussian kernel, Uniform kernel and Parabolic kernel [J]. China Environmental Science, 2023,43(7):3404-3415.
[56] Tinarelli G, Mortarini L, Castelli S T, et al. Review and validation of MicroSpray, a lagrangian particle model of turbulent dispersion [J]. Lagrangian Modeling of the Atmosphere, 2012,200:311-328.