Tornado Hazard Assessment for a Nuclear Power Plant in China

Available online at www.sciencedirect.com

ScienceDirect ScienceDirect Energy Procedia 00 (2016) 000–000

Availableonline onlineatatwww.sciencedirect.com www.sciencedirect.com Available Energy Procedia 00 (2016) 000–000

ScienceDirect ScienceDirect

www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia

Energy Procedia Procedia 00 127(2017) (2017)000–000 20–28 Energy www.elsevier.com/locate/procedia

International Youth Nuclear Congress 2016, IYNC2016, 24-30 July 2016, Hangzhou, China International Youth Nuclear Congress 2016, IYNC2016, 24-30 July 2016, Hangzhou, China

Tornado Hazard Assessment for a Nuclear Power Plant in China Tornado Hazard bAssessment for a Nuclear Power Plant in Chinac a, a a a

The 15th Heating and Cooling Hao Zhu *, Jiayi ChenInternational , Fengju LiSymposium , Xiaoping on BaiDistrict , Xiaoliang Wang , Hongyu Wang , a, b aand Wei Zhengaa a Hao Zhu *, Jiayi Chen , Fengju Li , Xiaoping Bai , Xiaoliang Wang , Hongyu Wangc, a Assessing the feasibility ofWei using the heat demand-outdoor and China Nuclear Power Engineering Co., Ltd., No.117, North of theZheng Western 3 Ring Road, Haidian District, Beijing100840, China a

temperature function for a long-term district heat demand forecast

College of Environmental Sciences and Engineering, Peking University, No.5 Yiheyuan Road, Haidian District, Beijing 100871, China a China Nuclear Power Engineering Co., Ltd., No.117, North of the Western 3 Ring Road, Haidian District, Beijing100840, China c Institute of Atmospheric Environment, China Meteorological Administration, No.388 South of Changbai Road, Heping District, Shenyang b College of Environmental Sciences and Engineering, Peking University, No.5 Yiheyuan Road, Haidian District, Beijing 100871, China 110016, China c Institute of Atmospheric Environment, China No.388 a,b,c a Meteorological Administration, a b South of Changbai Road, c Heping District, Shenyang c 110016, China b

I. Andrić

a

*, A. Pina , P. Ferrão , J. Fournier ., B. Lacarrière , O. Le Corre

IN+ Center for Innovation, Technology and Policy Research - Instituto Superior Técnico, Av. Rovisco Pais 1, 1049-001 Lisbon, Portugal b Veolia Recherche & Innovation, 291 Avenue Dreyfous Daniel, 78520 Limay, France c Département Systèmes Énergétiques et Environnement - IMT Atlantique, 4 rue Alfred Kastler, 44300 Nantes, France

Abstract

Abstract Tornado is one of the most hazardous external natural events which should be taken into account in the siting, design and safety evaluation of nuclear power plants. The tornado hazard has been assessed for a nuclear power plant located in the Southeast Tornado is one of the most hazardous external natural events which should be taken into account in the siting, design and safety China in this paper. Using the original Fujita scale and grading by a number of experts, intensities of the tornado segments have evaluation of nuclear power plants. The tornado hazard has been assessed for a nuclear power plant located in the Southeast Abstract been determined. The design-basis tornado wind speed, parameters and missiles generated have been evaluated following the China in this paper. Using the original Fujita scale and grading by a number of experts, intensities of the tornado segments have methodology listed in the Chinese Nuclear Safety Guide HAD101/10. The results show that the overall cumulative frequency and been determined. The design-basis tornadoaddressed wind speed, the parameters missiles generated have been evaluated the District heating networks are site commonly literatureand asrelationship. one of the most expected effective solutions for following decreasing the intensity of tornadoes in the area is not in a in logarithm-linear The annual tornado occurrence methodology listed in the Chinese Nuclear Safety Guide HAD101/10. The results show that thewhich overall cumulative frequency and greenhouse gas emissions from the building sector. These systems require high investments are returned through the heat frequencies of various F scales have been estimated by fitting higher intensity samples. The resulting wind speed corresponding intensityDue of to tornadoes in theclimate site area is not in logarithm-linear relationship. heat The demand expectedinannual tornado occurrence thefrequency changed conditions anda to building renovation the future could decrease, year seems be sensitive to thepolicies, fitting samples. Considering the uncertainties, the tosales. an exceedance of 10-7 per frequencies of various F scales have been estimated by fitting higher intensity samples. The resulting wind speed corresponding prolonging the investment return period. intensity and frequency of tornadoes in the survey area, the ultimate design-basis tornado wind speed is taken to be 70.4 m/s, -7 per year seems to ofbeusing sensitiveheat to the fitting samples.temperature Considering the uncertainties, the toThe an exceedance frequency of this wind paperspeed is 10 to assess the feasibility demand – outdoor demand whichmain is thescope loweroflimit of F3 scale. Moreover currentthe problems in tornado hazard assessmentfunction in Chinafor areheat discussed. intensity tornadoes located in the survey area,(Portugal), the ultimate design-basis windThe speed is taken to be 70.4ofm/s, forecast.and Thefrequency district ofofAlvalade, in Lisbon was used as a tornado case study. district is consisted 665 which is thethat lower limit wind construction speed of F3 scale. currentThree problems in tornado hazard assessment China arethree discussed. buildings vary in both periodMoreover and typology. weather scenarios (low, medium, inhigh) and district ©renovation 2016 The Authors. Published by Elsevier Ltd. scenariosPublished were developed (shallow, © 2017 The Authors. by Elsevier Ltd. intermediate, deep). To estimate the error, obtained heat demand values were Peer-review under responsibility of the the organizing organizing committee of IYNC2016 IYNC2016. compared with results from a dynamic heat demand model, previously developed and validated by the authors. Peer-review under responsibility of committee of © 2016 The Authors. Published by Elsevier Ltd. The results under showed that when only weather change is considered, the margin of error could be acceptable for some applications Peer-review responsibility of the organizing committee of IYNC2016. Keywords: relationship; combined vortex model; (the errordesign-basis in annual tornado demandwind wasspeed; lowerFujita thanscale; 20%event-intensity for all weather scenariosRankine considered). However, after missile introducing renovation scenarios, the error value increased up to 59.5% (depending on the weather and renovation scenarios combination considered). Keywords: design-basis tornado wind speed; Fujita scale; event-intensity relationship; Rankine combined vortex model; missile The value of slope coefficient increased on average within the range of 3.8% up to 8% per decade, that corresponds to the decrease in the number of heating hours of 22-139h during the heating season (depending on the combination of weather and renovation scenarios considered). On the other hand, function intercept increased for 7.8-12.7% per decade (depending on the coupled scenarios). The values suggested could be used to modify the function parameters for the scenarios considered, and improve the accuracy of heat demand estimations. © 2017 The Authors. by Elsevier Ltd. * Corresponding author.Published Tel.: +86-10-8802-2099. Peer-review under responsibility of the Scientific Committee of The 15th International Symposium on District Heating and E-mail address: [email protected] * Corresponding author. Tel.: +86-10-8802-2099. Cooling.

E-mail address: [email protected] 1876-6102 © 2016 The Authors. Published by Elsevier Ltd. Keywords: Forecast; Climate change Peer-review Heat underdemand; responsibility of the organizing committee of IYNC2016. 1876-6102 © 2016 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of the organizing committee of IYNC2016.

1876-6102 © 2017 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of the Scientific Committee of The 15th International Symposium on District Heating and Cooling.

1876-6102 © 2017 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of the organizing committee of IYNC2016 10.1016/j.egypro.2017.08.091

2

Hao Zhu et al. / Energy Procedia 127 (2017) 20–28 Hao Zhu et al./ Energy Procedia 00 (2016) 000–000

21

1. Introduction Tornadoes are generally described as violently rotating columns of air, usually associated with a thunderstorm. They are one of the most hazardous severe weather phenomena, causing significant to catastrophic damage to manmade structures, vegetations as well as human injuries and fatalities. Thus they should be taken into account in the siting, design and safety evaluation of nuclear power plants and other important infrastructures. Wind speeds associated with tornadoes are rarely measured by meteorological instruments because of the relatively small spatial extent and short time duration of a tornado at a certain location. Intensity of tornadoes is commonly measured by the Fujita scale [1], which classifies tornado intensity in six F scale classes from F0 to F5 based on the damages caused by the tornado. Most significant tornadoes (F2 and greater) are related to supercell thunderstorms [2], and are accompanied by hailstorms, thunderstorms, rainstorms, and other violent convective weather systems [3]. Other factors such as topography and underlying surface also play an important role in the generation of tornadoes. In China, most of the significant tornadoes (greater than F2) occurred in the eastern plain and coastal areas, coincident with the convective system-prone areas [3], where all of the nuclear power plants in operation and in construction and most of the planning ones are located. According to IAEA and Chinese nuclear safety guides [4-7], information over as long a period of time as possible should be collected to determine whether there is a potential for the occurrence of tornadoes in the region. If the possibility of tornadoes in the region is confirmed, a detailed investigation should be performed to obtain suitable data for the evaluation of a design basis tornado. United States is the country where tornadoes occur most frequently in the world. Detailed tornado information including the location, path length, width, and maximum damage rating for every tornado segment is documented by the National Weather Service. Such data have been used to determine tornado strike probabilities and maximum wind speeds for use in nuclear power plant design [8]. In 2007, the National Weather Service of US has adopted the enhanced Fujita (EF) scale in place of the original Fujita (F) scale to improve the accuracy and precision of damage surveys. To reflect the changes, U.S. Nuclear Regulatory Commission updated regulatory guide 1.76 in 2007 [9], in which the contiguous United States is divided into three tornado intensity regions, and the design-basis tornado wind speeds at the 10-7 per year probability level are given for the three regions. The regulatory guide has facilitated the licensees and applicants in selecting the design-basis tornado and design-basis tornado-generated missiles for a nuclear power plant. However, no tornado database is available in China. The historical tornado records are scattered in some books, reports, newspapers, and also appear on internet after 2000 [3,10]. Tornado hazard assessment for a nuclear power plant is usually evaluated on a case-by-case basis using the principles in the Chinese Nuclear Safety Guide HAD101/10 [7]. In this paper, the method and process of tornado hazard assessment for a nuclear power plant site in Southeast China is described. Moreover the uncertainties and current problems are discussed. 2. Data acquisition The location of the nuclear power plant in study is shown in Fig.1. The site is located in the middle of the coast of Fujian Province, which is in the southeast of China. According to the requirement of Chinese Nuclear Safety Guide HAD101/10 [7], a climatically homogeneous region should be selected for the development of tornado inventories. Considering these requirements, a typical area used to evaluate the design-basis tornado is a three-degree latitude and longitude box centered at the site, which is illustrated in Fig.1. The land survey area is 41,743 km2. Tornado information was obtained from Collections of Meteorological Hazards in China [11], Annals of Meteorological Hazards [12], and other relevant books and reports, such as China Climate Impact Assessment [13], monthly reports of meteorological observation, local chronicles, et al. Due to the rapid development and popularity of internet since 2000, tornado videos, photos and onsite interviews have been widely reported on various meteorological websites and the government portal websites, which can be used as the supplement of tornado information. For most of the tornado records, only the locations, dates and damages were reported. Little information about tornado segment length, width, and area was available. From 1959 to 2012, a total of 103 tornado segments were collected.

Hao Zhu et al. / Energy Procedia 127 (2017) 20–28 Hao Zhu et al./ Energy Procedia 00 (2016) 000–000

22

3. Assessment of tornado hazard

Fig. 1. Location of the NPP (red dot) and the three-degree latitude and longitude box survey area of the tornadoes (blue rectangle).

Fig. 2. Flow chart of tornado hazard assessment for an NPP.

3

4

Hao Zhu et al. / Energy Procedia 127 (2017) 20–28 Hao Zhu et al./ Energy Procedia 00 (2016) 000–000

23

A complete tornado hazard assessment for a nuclear power plant (NPP) includes tornado information collection, tornado intensity classification, and determination of tornado striking probability-wind speed relationship, designbasis tornado parameters and tornado-generated missile characteristics. Fig.2 shows the flow chart of the assessment. 3.1. Intensity Classification In China, the original Fujita–Pearson scheme is used for the classification of a tornado [7]. This system is a combination of the Fujita scale rating for wind speed, the Pearson scale for path length and the Pearson scale for path width. Table 1 shows Fujita-Pearson intensity classification scheme. It should be noted that the classification of tornado intensities is subjective, depending on the assessment of damages to structures and vegetation, and on construction quality, population density, location and witnesses’ narration [14]. Description of damages caused by tornadoes is sometimes ambiguous, which brings additional difficulties for an accurate intensity classification. To better quantify the F-scale of a tornado, grading by a number of experts is an acceptable method. The steps are as following [10]: (1) Listing the tornado cases according to the occurrence dates; (2) Inviting 20 meteorological experts and graduate students to classify the tornado cases independently based on the damages caused by the tornadoes; (3) Calculating the total scores graded by all of the experts and students for each of the tornado cases, and removing the top three and the minimum three total scores, then calculating the average score for each case. It should be noted that Fujita scales are integers, to better analyze the tornado data and for the convenience of the following data fitting, the intensities were rounded by 0.2 levels at the averaging step. Using the steps above, Fscales of the 103 tornado events were determined and listed in table 2. There were 14 tornadoes with intensity stronger than F2.0, and the most intense tornado in the area was estimated to be F2.6. Based on the tornado damage information collected, the tornado with intensity F2.6 occurred suddenly in Nanan county in March 1981 and lasted for 5 to 10 minutes. Two stone gates of more than 500 kg weight were blown down by the strong wind. The top of an ancient tower was blown away. 145 electronic and radio poles and more than 2300 fruit trees were broken down. Table 1. Fujita-Pearson intensity classification scheme. Intensity F0 F1 F2 F3 F4 F5

Description Light damage Moderate damage Considerable damage Severe damage Devastating damage Incredible damage

Wind speed (m/s) Less than 33 33 to 49 50 to 69 70 to 92 93 to 116 117 to 140

Path length(km) Less than 1.6 1.6 to 5.0 5.1 to 16.0 16.1 to 50.9 51 to 160 161 to 507

Path width (m) Less than 16 16 to 50 51 to 160 161 to 509 510 to 1600 1601 to 5070

Table 2. Intensity classification of the tornado events in the site area. Intensity F0

F1

F2

F scale 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6

Lower limit wind speed (m/s) 17.8 20.6 23.4 26.4 29.5 32.7 36.1 39.5 43.0 46.7 50.4 54.2 58.1 62.2

Frequency 13 15 12 5 5 7 15 8 3 6 6 4 3 1

Cumulative frequency 103 90 75 63 58 53 46 31 23 20 14 8 4 1

Hao Zhu et al. / Energy Procedia 127 (2017) 20–28 Hao Zhu et al./ Energy Procedia 00 (2016) 000–000

24

5

Although tornadic wind speeds have a significantly lower probability of occurrence at a point than the probability for basic wind speed, if a tornado of such intensity occurs and strikes a nuclear power plant, the buildings and structures not properly designed and exposed to such an event may be damaged and a local collapse may occur [15]. Light-weight objects and debris from nearby damaged structures lying within the path of the tornado may become missiles and impact with the structures and equipments. If these exposed structures and equipments are not well designed or protected, damages may be caused. Another hazard may be induced is a loss of offsite power. To avoid potential destructions caused by tornadoes, tornado-resistant design for above-ground buildings that house essential facilities should be implemented [15]. First the design basis tornado wind speed and other parameters should be assessed. Tornado loads including combinations of tornado wind effects, atmospheric pressure change effects, and tornado-generated missile impact effects are considered in the design of safety-related structures. For the design of seismic Category I structures, load combinations of the most severe tornado load and the loads resulting from normal and accident conditions should be included. Missile barriers should also be designed. With the above countermeasures, it can be ensured that the structural integrity of the plant structures designed for tornado loads will not be damaged and safety-related systems and components located within these structures will be protected [16]. 3.2. Methodology Four steps are involved in the evaluation of design-basis tornado wind speed [7]: (1) determining the relationship of area-intensity for the total survey area; (2) determining the relationship of event-intensity for the local site area; (3) calculating the probability of a point experiencing a particular wind speed within specified wind speed ranges; (4) determining the annual frequency of exceedance at which a particular site will experience tornado wind speeds in excess of a specified value. Since the tornado information in China is not sufficient to establish a credible area-intensity relationship, for most of sites in China, the area of damage is calculated by the product of the Pearson path length and path width. The lower limit of Pearson path length Lpi is as following:  Lpi 1.609  100.5(i 1)

(1)

where i is the Fujita-Pearson scale, Lpi is in kilometres. The path width is the average width of the damage area measured perpendicular to the tornado path. The lower limit of Pearson path width Wpi (in kilometres) is expressed as:  W pi 1.609  100.5( i  5)

(2)

Thus the area of damage (in km2) can be denoted as ai  L pi  W pi 1.6092  10(i  3)

(3)

The Chinese Nuclear Safety Guide HAD101/10 [7] recommends that the cumulative frequency of tornadoes above a particular intensity can be described as following, which is called the event-intensity relationship: ln N i  c 'Vi  k '

(4)

where Ni is the cumulative frequency of tornadoes whose maximum wind speed exceeds a particular threshold wind speed Vi; Vi is the lower limit tornado wind speed of Fujita scale i, which is denoted as Vi=6.30(i+2)1.5; c’ and k’ are coefficients obtained by least square linear fitting of historical data. Based on the event-intensity relationship, the expected annual occurrence frequency of Fi scale tornado in the survey area denoted as λi can be calculated as following:

6

Hao ZhuEnergy et al. /Procedia Energy Procedia 127 (2017) 20–28 Hao Zhu et al./ 00 (2016) 000–000

i 

25

N i  N i 1 . E  S 1

(5)

where E and S are the end and starting year for the tornado records, respectively. Combining the above area-intensity and event-intensity relationship with a single Rankine combined vortex model, which takes variation of the rotational speed in a tornado into account [9], the probability of a point experiencing wind speed of Fi scale per year can be expressed as: n

P V j ,V j 1  

 a

i ij

i j

A

(6)

.

where A is the survey area; n is the highest scale of tornadoes in the area, usually set to be F6 or F7 formally; aij is the path area of a tornado with Fi scale and within the wind speed range (Vj,Vj+1), which can be expressed as: 2  R j  R j 1  Li  33.5ai aij 

V j 1  V j V jV j 1

(7)

.

where R is the radius of a tornado with wind speed V. Here, it is assumed that the average damage area is caused by wind with wind speed higher than 33.5 m/s. Finally the probability of a point experiencing wind speed in excess of Fk scale is: n

PE Vk    P V j ,V j 1 .

(8)

j k

The design-basis tornado wind speeds for nuclear power plants should correspond to the exceedance frequency of 10-7 per year [7,9]. From the probability-wind speed curve, the design-basis tornado wind speed can be determined. 3.3. Results and uncertainties (a)

(b)

-5

4

-6

3

-7

log10(PE)

ln(Ni)

5

-8

2

≥F2

-9

1 0

-10

-1

-11 15

25 35 45 55 65 wind speed(m/s)

75

≥F1.8 ≥F1.6 20

40 60 80 100 120 wind speed(m/s)

Fig. 3. (a) Relationship between cumulative frequency and intensity; (b) probability of a point experiencing wind speed in excess of a specified value.

Hao Zhu et al. / Energy Procedia 127 (2017) 20–28 Hao Zhu et al./ Energy Procedia 00 (2016) 000–000

26

7

Table 3. Fitting coefficients and the expected annual tornado occurrence frequency for various fitting samples. Fitting samples ≥F1.6 ≥F1.8 ≥F2.0

c’ 0.1582 0.1877 0.2202

k’ 10.336 12.017 13.909

λ1 3.02 6.34 14.75

λ2 1.88E-1 2.33E-1 3.04E-1

λ3 8.01E-3 5.47E-3 3.71E-3

λ4 2.43E-4 8.59E-5 2.83E-5

λ5 5.40E-6 9.37E-7 1.41E-7

Based on the intensity classification of tornado segments, the relationship between the cumulative frequency and intensity is illustrated in Fig.3(a), which is not logarithm-linear distributed as expressed in equation (4). Since the design-basis tornado wind speed only relies on the distribution of samples of higher intensity and samples of lower intensity do not directly impact the result, the event-intensity relationship is derived from fitting of higher intensity samples. Here, we choose fitting samples of intensity greater than F1.6, F1.8 and F2.0, respectively, to calculate the parameter λi. The fitting coefficients of the event-intensity relationship and the expected annual occurrence frequency of Fi scale tornado λi for the three fitting samples are listed in table 3. For the three kinds of fitting samples, the probability of a point experiencing wind speed in excess of Fk scale is shown in Fig.3(b), from which the design-basis tornado wind speed corresponding to the exceedance frequency of 10-7 per year can be derived. Table 4 shows the results for the three kinds of fittings. The maximum wind speed corresponding to a frequency of 10-7 per year is 69.4 m/s, which is close to the lower limit wind speed of F3 scale. Considering the uncertainties in the survey, intensity classification, and in the evaluation method, and the relatively low frequency and intensity of tornado cases in the area, the ultimate design-basis tornado wind speed is set to be 70.4 m/s, i.e. the lower limit wind speed of F3 scale. Table 4. Tornado wind speed corresponding to the exceedance frequency of 10-7 per year. Fitting samples ≥F1.6 ≥F1.8 ≥F2.0

Wind speed at 10-7 per year (m/s) 69.4 63.5 61.1

3.4. Design-basis tornado parameters HAD101/10 uses the same model as RG 1.76 suggests calculating the design-basis tornado parameters [7,9], which is based on a single Rankine combined vortex model. In this model, it is assumed that the flow is in cyclostrophic balance, i.e. the inward-directed pressure gradient force is in balance with the outward-directed centrifugal force. It can be derived that the maximum rate of pressure drop is as following: dP VT  Vm2 dt Rm

(9)

where ρ is the air density, taken as 1.226 kg/m3, VT is the translational speed of the tornado, Rm is the radius of maximum rotational speed Vm. The total pressure drop from a normal atmospheric pressure to the center of the Rankine combined vortex is: (10)

P  Vm2

For an intense tornado, Rm is assumed to be 50 m and the ratio of Vm and VT is: Vm/VT=290/70. From the above equations, the design-basis tornado parameters are calculated and listed in table 5. Table 5. Design-basis tornado parameters. Translational speed (m/s) 13.7

Maximum rotational speed (m/s) 56.7

Pressure drop (hPa) 39.4

Rate of pressure drop (hPa/s) 10.8

8

Hao Zhu et al. / Energy Procedia 127 (2017) 20–28 Hao Zhu et al./ Energy Procedia 00 (2016) 000–000

27

3.5. Design-basis tornado-generated missiles HAD 101/10[7] suggests to use a standard spectrum of tornado missiles, which includes an automobile of 1800 kg, an armor-piercing shell of 125 kg weight and 20 cm length, and a solid steel sphere of 2.5 cm diameter. For these three missiles, 35% of the maximum horizontal wind speed of design-basis tornado is considered as the collision velocity. The collision momenta of the three kinds of missiles are listed in table 6. Table 6. Collision momentum of the design-basis tornado-generated missiles. Missile type automobile of 1800 kg armor-piercing shell of 125 kg solid steel sphere of 2.5 cm diameter

Collision velocity (m/s) 24.64 24.64 24.64

Collision momentum (kg m/s) 44352 3080 1.65

4. Conclusions and discussion In this paper, tornado hazard for a nuclear power plant in Southeast China is assessed. Using the original Fujita intensity classification scheme, the strongest tornado in the area is estimated to be F2.6. Overall, the cumulative frequency is not logarithm-linear distributed, thus the event-intensity relationship is derived from the fitting of higher intensity samples. The resulting wind speed corresponding to an exceedance frequency of 10-7 per year seems to be sensitive to the fitting samples. Considering the uncertainties, the intensity and frequency of tornadoes in the survey area, the ultimate design-basis tornado wind speed is taken to be 70.4 m/s, which is the lower limit wind speed of F3 scale. Using a single Rankine combined vortex model, the other design-basis tornado parameters are estimated, and the horizontal wind speed and collision momenta of design-basis tornado-generated missiles are calculated. In China, tornado information is recorded in some books and appears on internet in recent years. Collection and analysis of tornado samples is a time-consuming work. It is suggested that a tornado database should be established in future to facilitate the tornado hazard estimation. Information on tornado location, occurring time and date, path length, width, and damages caused should be collected as detailed as possible. The reliability of the tornado records should be evaluated by professionals. Since the intensity classification of tornado cases are subjective and involve many difficulties and uncertainties in practice such as distinguishing the damages caused by tornadoes from the damages due to other accompanying severe weather phenomena, reasonably interpreting the damages from witnesses’ narration etc., professional teams with meteorological background are suggested to carry out the work to give more reliable results. Satellite and aerial imagery-based damage surveys can supplement the traditional groundbased surveys. Geographic information system (GIS) is suggested to be used for the storage, maintenance, and analysis of tornado data, which will also be a great help for the future data share among various agencies and corporations. USNRC has adopted the Enhanced-Fujita scale in the latest version of RG1.76, while Chinese Nuclear Safety Guide has not yet updated the intensity classification scheme. The impact of shifting the original Fujita scale to the Enhanced-Fujita scale on the determination of design-basis tornadoes for the nuclear power plants in China should be evaluated. Uncertainties in the whole process of assessment have not been paid much attention to in the engineering applications by now. In future, the uncertainty and sensitivity analysis of design-basis tornado wind speed should be enhanced. The collision wind speeds of tornado-generated missiles are assumed to be the same for different missiles. The suitability of this assumption should be further investigated.

References [1] Fujita TT. Tornadoes and downbursts in the context of generalized planetary scales. Journal of Atmospheric Science 1981; 38: 1511-1534. [2] Trapp RJ, Stumpf GJ, Manross KL. A reassessment of the percentage of tornadic mesocyclones. Wea. Forecasting 2005; 20: 680-687. [3] Fan W, andYu X. Characteristics of spatial-temporal distribution of tornadoes in China. Meteorological Monthly 2015; 41:793-805.

28

Hao Zhu et al. / Energy Procedia 127 (2017) 20–28 Hao Zhu et al./ Energy Procedia 00 (2016) 000–000

9

[4] IAEA. Meteorological and hydrological hazards in site evaluation for nuclear installations. Safety Standards Series No. SSG-18, Vienna; 2011. [5] IAEA. Site evaluation for nuclear installations. Safety Standards Series No. NS-R-3, Vienna; 2003. [6] IAEA. Meteorological events in site evaluation for nuclear power plants. Safety Standards Series No. NS-G-3.4, Vienna; 2003. [7] National Nuclear Safety Administration. Extreme meteorological events in the siting of nuclear power plants (excluding tropical cyclones). HAD101/10, Beijing; 1991. [8] USNRC. Tornado climatology of the contiguous United States. NUREG/CR-4461, Washington, DC; 2007. [9] USNRC. Design-basis tornado and tornado missiles for nuclear power plants. RG 1.76, Rev.1, Washington, DC; 2007. [10] Chen J, Yang H, Zhu Y, et al. The survey and evaluation of tornadoes. Journal of Natural Disasters 1999; 8: 111-117. [11] Wen KG, Song DZ, Cai SH, et al. Collections of meteorological hazards in China: Fujian Volume. Beijing: China Meteorological Press; 2007. [12] China Meteorological Administration. Annals of meteorological hazards (2005-2012). Beijing: China Meteorological Press. [13] National Climate Center/ CMA. China climate impact assessment (2001-2012). Beijing: China Meteorological Press. [14] Verbout SM, Brooks HE, Leslie LM, et al. Evolution of the U.S. tornado database: 1954-2003. Wea. Forecasting 2006; 21: 86-93. [15] American Society of Civil Engineers/Structural Engineering Institute. Minimum design loads for buildings and other structures. ASCE/SEI 7-05, American Society of Civil Engineers, Reston, Virginia; 2006. [16] USNRC. Standard review plan, section 3.3.2. NUREG-0800, Washington, DC; 2007.