Estimation of loss caused by earthquakes and secondary ...

Nat. Hazards Earth Syst. Sci., 11, 3025–3033, 2011 www.nat-hazards-earth-syst-sci.net/11/3025/2011/ doi:10.5194/nhess-11-3025-2011 © Author(s) 2011. CC Attribution 3.0 License.

Natural Hazards and Earth System Sciences

Estimation of loss caused by earthquakes and secondary technological hazards N. I. Frolova1 , V. I. Larionov2 , S. P. Suschev2 , and J. Bonnin3 1 Seismological

Center of IGE, Russian Academy of Sciences, Moscow, Russia High Technical University named after Bauman, Moscow, Russia 3 Institute of Physics of the Earth, University of Strasbourg, France 2 Moscow

Received: 17 June 2011 – Revised: 16 September 2011 – Accepted: 28 September 2011 – Published: 14 November 2011

Abstract. Assessment of expected loss and damage caused by earthquakes and secondary technological accidents are of primary importance for the development and implementation of preventive measure plans, as well as for emergency management just after the disaster. The paper addresses the procedures for estimations of loss caused by strong events and secondary hazards with information technology application. Examples of individual seismic risk zoning at Russian federal and regional levels are given, as well as that of scenario earthquakes consequences estimation, taking into account secondary technological hazards.

1

Introduction

Earthquakes are becoming more devastating, especially when they occur in industrialized regions. Social and economic losses due to those events and secondary processes triggered by them, increase annually, which is definitely in relation with the evolution of society. Seismic and secondary technological hazards identification and analysis, as well as risk assessment and mapping, are the first steps in prevention strategy aimed at saving lives and protecting property against future events. The paper addresses methodological issues of risk assessment and mapping taking into account technological accidents at fire, explosion and chemical hazardous facilities triggered by strong seismic events. Special GIS environments are usually developed for risk assessment and mapping at different levels. Examples of individual seismic risk zoning for the population of the Russian Federation and for the Krasnodar region are given, as well as that Correspondence to: N. I. Frolova ([email protected])

of scenario earthquakes consequences estimation taking into account secondary technological hazards at urban and facility levels.

2

The methodological issues of risk assessment

The section describes the procedures for individual risk assessment created by earthquakes and seismic risk assessment taking into account secondary technological accidents. For estimation of risk indexes and risk mapping, the probabilistic approach is used. For seismic risk assessment the authors follow the concepts proposed and agreed upon by UN and other experts in the field (Karnik et al., 1978; Fournier d’Albe, 1982, 1986; Karnik, 1984; Boissonnade and Shah, 1984; Mitigating, 1991; Dolce et al., 1995; UNISDR, 2009; Risk, 2010; Ranguelov, 2011). Individual risk Re due to any hazard is determined as the probability of death and/or injuries and/or economic loss for persons due to potential hazard within one year at a given place. Individual seismic risk Rs is the product of hazard H and vulnerability Vs . Vulnerability of the population to seismic action of a given intensity is understood here as the ratio between the number of persons expected to be affected by fatalities, injuries, losses of property and the total number of persons living in a certain type of buildings (Larionov and Frolova, 2003a). Individual seismic risk Rs (Bonnin et al., 2002; Bonnin and Frolova, 2004; Frolova et al., 2003, 2011; Larionov et al., 2003b; Methods, 2000) may be determined through mathematical expectation of social losses, which include fatalities, injuries and persons who lost their property, M(N ), taking into account the number of inhabitants N in the considered settlement and probability of seismic event H

Published by Copernicus Publications on behalf of the European Geosciences Union.

1 Table 1. Comparison of building vulnerability classes according to MMSK-86 and EMS-98. 1 Table Comparison of building vulnerability to MMSK-86 EMS-98.technological hazards N. I.1. Frolova et al.: Estimation of loss classes causedaccording by earthquakes and and secondary of buildings’ types according to EMS-98 1 Description Table 1. Comparison of building vulnerability classes accordingVulnerability to MMSK-86class and EMS-98. Description of buildings’ types according to EMS-98 1 Table 1. Comparison of building vulnerability classes accordingVulnerability to MMSK-86class and EMS-98. EMS-98 MMSK-86 Description of buildings’ types according to EMS-98 Vulnerability Table 1. Comparison of1building vulnerability classes according to MMSK-86 EMS-98. Table 1. Comparison of building vulnerability classesand according to MMSK-86class and EMS-98. EMS-98 Description of buildings’ types according to EMS-98 VulnerabilityMMSK-86 class Rubble stone, field stone A A EMS-98 Description of buildings’ types according to EMS-98 VulnerabilityMMSK-86 class Description offield buildings’ class Rubble stone, stone types according to EMS-98 EMS-98 AVulnerability A MMSK-86 Adobe A A Rubble(earth stone,brick) field stone A A EMS-98 MMSK-86 EMS-98 MMSK-86 Adobe A A Rubble(earth stone,brick) field stone A A Simple stone B A Adobe A A Rubble stone,brick) field stone stone A Rubble(earth field AA Simple stone brick) B A Adobe (earth (earth A A Adobe A Massive stonebrick) CA Б Simple (earth stone B A Adobe brick) A A Simple stone B A Massivestone stone C Б Simple B A Unreinforced BC Б Massive stone Massive stone(bricks/concrete blocks) C Б Simple stone B A Unreinforced BB Б Unreinforced (bricks/concreteblocks) blocks) Massive stone(bricks/concrete C Б Unreinforced (brick) with RC floors CC Б Unreinforced (bricks/concrete blocks) B Б Unreinforced (brick) with RC floors Massive stone C Б Unreinforced (brick) with RC floors C Б Unreinforced (bricks/concrete blocks) BD Б Reinforced or confined Reinforced or confined D В Unreinforced (brick) with RC floors C Б Unreinforced (bricks/concrete blocks) BC Б Reinforced without earthquake-resistant design (ERD) Reinforced or (brick) confined D В Unreinforced with RC floors C Б Reinforced with minimum level of ERD D C7 Reinforced without earthquake-resistant design (ERD) C В Reinforced or (brick) confined D В Unreinforced withlevel RC floors CE Б Reinforced with average of ERD C8 Reinforced or without earthquake-resistant design (ERD) C В Reinforced confined D В Reinforced without with minimum of ERD design (ERD) D С7 Reinforced with high levellevel of ERD C9 Reinforced earthquake-resistant CF В Reinforced or confined D В Reinforced with minimum level of ERD design (ERD) D С7 Timber structures B-C7 Reinforced without earthquake-resistant CD В Reinforced with average level of ERD E С8 Reinforced with minimum level of ERD design (ERD) D С7 Reinforced without earthquake-resistant C В Reinforced with average level ofof ERD E С8 Reinforced with minimum level ERD D С7 Reinforced with high level of ERD F С9 Reinforced with average level ofof ERD E С8 Reinforced with minimum level ERD D С7 Reinforced with level of ERD F С9 Reinforced with high average level of ERD E С8 Timber structures D В-С7 Reinforced with high level of ERD F С9 of seismic load with intensityС8 I ; −P (Cj |Bi ) is the probabilReinforced with average level of ERD E structures D В-С7 Reinforced with high level of ERD F 2 Timber ity of people to survive j levelС9of impact under the condition Timber structures D В-С7 Reinforced with high level of ERD (1) F survived the С9damage state i (five damage Rs = H × Vs (I ) = H × 2 M(N)/N that the buildingD Timber structures В-С7 2 states are considered Timber structures D here, from В-С7i = 1 (slight damage) to i = 5 where −Vs (I ) is the 2vulnerability of population in the con(total collapse)). sidered settlement; −H 2 is the probability of seismic event

3026

occurrence per one year; −N is the number of inhabitants in the considered settlement. The mathematical expectation of social losses M(Nj ) in certain j -type buildings for the considered settlement, taking into account inhabitant migration in the buildings of j -type during the day and night, is determined by equation ZZ Z Imax PCj (I ) × f (x,y,I ) M(Nj ) = f (t) Sc Imin

×9j (x,y) × dI dxdy

(2)

where – Imin and Imax are the maximum and minimum possible seismic intensities; Sc is the settlement area; PCj (I ) is the probability of fatalities, injuries and persons having lost their property, under the condition of damage to buildings of j -type due to an earthquake with intensity I ; ψj (x,y) is the density of population distribution within the considered area in buildings of j -type; f (x,y,I ) is the density function of earthquakes’ intensity probabilities within the unit area with coordinates x, y; f (t) is the function obtained on the basis of statistical analysis of data on the population migration over 24 h. Computations of PCj (I ) are carried out using equation PCj (I ) =

5 X

PBi (I ) × P (Cj |Bi )

(3)

i=1

where −PCj (I ) is the probability of people being impacted on during the earthquake of intensity I ; −PBi (I ) is the probability of definite damage state i of buildings under the action Nat. Hazards Earth Syst. Sci., 11, 3025–3033, 2011

The computations of PCj (I ) are usually done for building and structure types classified according to the MMSK-86 scale (Shebalin et al., 1986): buildings’ type A (from local materials); buildings’ type B (brick, hewn stone or concrete blocks); buildings’ type C (reinforced concrete, frame, large panels and wood); buildings’ types E7, E8, E9 (earthquake resistant which are designed and constructed to withstand earthquakes with intensity 7, 8, 9). MMSK-86 and EMS-98 originate from MSK-64 and the expert estimation of different building types according to MMSK-86 and EMS-98 was undertaken (Table 1) in order to have the possibility of comparison of different vulnerability functions. The mathematical expectation of social losses M(N ) due to earthquakes in damaged and collapsed buildings for the 12 considered settlement, taking into account inhabitant migra12 tion in the buildings of all types during12the day and night, is determined by equation 12 M(N ) =

n X

12

M(Nj )

(4)

j =1

where −n is the number of considered building types according to the MMSK-86 scale. Secondary consequences resulting from technological accidents (fires, explosions, release of chemical materials) triggered by earthquakes, are estimated in a few steps: (1) Critical facilities with storage of different hazardous materials (fire, explosion and chemical hazardous substances) are identified; (2) distribution of shaking intensity is simulated for given parameters of scenario earthquake; (3) field of impact factors, such as excessive pressure, combustion temperature, www.nat-hazards-earth-syst-sci.net/11/3025/2011/

Frolova etconsequences al.: Estimation of loss caused by earthquakes and secondary technological 1N. I.Secondary resulting from technological accidents (fires, explosions, release of hazards 2

3027

chemical materials) triggered by earthquakes, are estimated in few steps: 1. Critical facilities

the concentration of the chemical hazardous substance in the concentration of chemical hazardous materials, are con3structed with taking storageinto of account different wind hazardous materials (fire, explosion atmosphere and chemical hazardous at the point with coordinates (x,y). direction, air temperature and2.others factors;of(4)shaking probabil4and pressure, substances)wind are velocity identified; Distribution intensity isIntegrated simulatedindividual for given risk Re (x,y) due to earthquakes and secondary technological estimatedofthat the critical facility3.will survive the dam5ity isparameters scenario earthquake; Field of impact factors, such as excessive pressure, accidents (Methods, 2000, 2002; Frolova et al., 2007) is determined by equation age state above threshold value; (5) social loss and individual 6risk combustion temperature, of chemical materials, are constructed are estimated in the caseconcentration of technological accidentshazardous trign Y 7gered taking into accountaccording wind direction, velocity and [1 others and (6). and pressure, Rwind by earthquake, to Eqs.air(5)temperature − Rei (x,y)] (8) e (x,y) = 1 − Individual risk R in the case of an accident at fire and ei 8 factors; 4. Probability is estimated that the critical facility will survive damage state i=1 above explosion hazardous facilities (Methods, 2002) is determined 9 threshold value; 5. Social loss and individual risk are estimated inwhere the case – nof istechnological the number of considered emergencies; Rei (x,y) (fires, explosions, release of by 1 Secondary consequences resulting from technological accidents ZZ is the individual risk due to i-th emergency situation. 10 accidents triggered by earthquake, according to equations (5) and (6). Xmaterials) triggered by earthquakes, are estimated in few steps: 1. Critical facilities 21X chemical R = H E (x,y) × P (x,y) k jan accident accidents ei Individual 1 11 Secondary consequences fromoftechnological (fires, explosions, releasefacilities of in kj the case at fire and explosion hazardous risk Rei resulting 3 N with storage of different hazardous materials (fire, explosion and chemical hazardous k j S 2 12 chemical materials)2002) triggered by earthquakes, are estimated in few steps:3 1. Seismic Critical facilities risk assessment at different levels (Methods…, is determined by 4×ψ(x,y) substances) identified; 2. Distribution of shaking × dx are × dy. (5) intensity is simulated for given 3 with storage1 of different hazardous materials (fire, explosion and chemical hazardous of scenario 3.( xField ofaccordsuch as excessive pressure, Earthquakes are the hazardous natural processes in the R5ei–=Hparameters Hk ∑ , yof )earthquake; ⋅ Paccident , year y ) ⋅ dx ⋅impact dy . factors, 13where (5)most ∑ j ( x, y ) ⋅ψper is the probability ∫∫ Ekj2.( xDistribution 4 substances) of shaking intensity is simulated for given which may result in fatalities, injuries Nkarek identified; j S Russian Federation, combustion temperature, of other chemical ing to6scenario k (fires, fire balls, concentration explosions and phe-hazardous materials, are constructed 5 14 parameters of scenario earthquake; 3.accident Fieldevents); ofper impact factors, such as excessive pressure, and economic loss. The results of seismic risk assessment where H is the probability of year according topressure, scenario k (fires, fire balls, nomena may be considered as scenario E (x,y) is ences resulting from technological accidents (fires, explosions, release of k into account wind direction, air temperature kj 7 taking and wind velocity and others taking into account secondary hazards are essential (practi6 15 combustion temperature, of chemical hazardous materials,events); are constructed the probability ofestimated impactconcentration mechanism j1.inCritical the point (x,y) forscenario explosions phenomena may be E the above triggered by earthquakes, areand in fewissteps: facilities kj(x,y) is 8 factors; 4. other Probability estimated thatconsidered the criticalasfacility will survive damage state cal) input for planning and implementing preventive meathe accident scenario k (as an impact mechanism,and thepressure, follow- wind velocity and others 7 16 taking into account wind direction, air jtemperature probability of impact mechanism in the point (x,y) for the accident scenario k (as impact ifferent hazardous materials (fire, explosion and chemical hazardous sures at national and local authority levels, as well as ac9 threshold value; 5. Social heat loss and individual risk are estimated in the case of technological ing factors could be considered: effect on population, 8 17 factors; 4. Probability is estimated that the critical facility will survive damage state above tions to be taken by the Ministry of the Russian Federation mechanism the following factors be considered: heat effect population, shock wave, entified; 2. shock Distribution of shaking intensity iscould simulated for given wave, debris of buildings and constructions, oth10 accidents triggered by earthquake, according toand equations (5)on and (6). for Civil Defense, Emergencies and Elimination of Conse9 18 threshold value; Social lossconstructions, and individual risk are estimated inthe theprobability case of technological ers);3. PField is5. the probability of case fatality in the point with j (x,y) debris ofofbuildings and and others); Pj(x,y) of hazardous fatality in the ario earthquake; impact factors, excessive pressure, 11 the as of an accident at is fire andquences explosion facilities (EMERCOM of Russia) just Individual risk Rei insuch of Natural Disasters coordinates (x,y)byunder the condition that impact(5) mecha10 accidents triggered earthquake, according to equations and (6). rature, concentration chemical hazardous materials, are constructed 19 12 pointofwith coordinates (x,y) under the condition that impact mechanism is realized; ψ(x,y)ofisa strong earthquake. Taking into acafterj the occurrence (Methods…, 2002) is determined by nism j is realized; ψ(x,y) is the density of population dis11 Individual risk Rei in the case of an accident at fire and explosioncount hazardous facilities the fact that in Russia, existing and undert wind direction, air and pressure, windwith velocity and others 20tribution the temperature density distribution in the vicinity of(x,y); the point with coordinates (x, y);atNpresent, is in theofvicinity of the point coordinates 1population R2002) = H E ( x , y ) ⋅ P ( x , y ) ⋅ ψ ( x , y ) ⋅ dx ⋅ dy construction oil pipe routes cross the earthquake prone areas 13 . (5) 12 (Methods…, is determined by ∑ ei k ∑ ∫∫ kj j thethe number people within zone of risk; is the ity is estimated that critical facility will survive damage state 21N is the number of people within thethe zone of risk; Sabove isSthe area within which people may be Nof k j S with high level of seismicity, the above-described procedure areaindividual within which people mayinbetheimpacted in the case of an 1 Social13 loss 22 and risk are estimated case of technological impacted in the case of the accident (i.e. the zone of risk). Rei =14 H E ( x , y ) ⋅ P ( x , y ) ⋅ ψ ( x , y ) ⋅ dx ⋅ dy is used for risk assessment regarding possible damage to fa. (5) ∑k where ∑j -zone kthe j probability of accident per year according to scenario k (fires, fire balls, ∫∫SHk iskj oftherisk). accident N (i.e. cilities of TRANSNEFT JSC; the corresponding GIS enviby earthquake, to equations (5) and (6). 23 according in the case of an accident at chemical hazardous facility (Methods…, 2002) Individual risk R ei in the case of an accident at a chemical Individual risk Rei 15 explosions and other phenomena may be considered as scenario events); Ekj(x,y) is the ronment is further developed. Examples of risk estimations where Hk is the atprobability accident per year according to scenario k (fires, fire balls, in the14case of an accident fire and of explosion facilities hazardous facility 2002) is hazardous determined by 24 is -determined by(Methods, 16 probability of impact mechanism j in the point (x,y) for theand accident scenario k (as impact mapping at different scales are given below. 15 explosions and other phenomena may be considered as scenario events); E kj(x,y) is the 2π 2V is determined by ZZ Z Z πmax V max 17 H mechanism the following factors could be considered: heat effect on population, shock wave, H ((x,y) (ψx,(x,y) [ Дthe f (a,)V×) ⋅PjPin x, point y )] ⋅ψ y )for ⋅ dVthe⋅ da ⋅ dx ⋅ dyscenario , 16 25 probability mechanism accident k (as impact 3.1 Seismic risk (6) assessment and mapping at federal Rei =Rei = of impact ×(x,y) f∫ (a,V ∫∫ ∫ N 18 debris of buildings and constructions, and others); Pj(x,y) is the probability of fatality in the N E ( x , y ) ⋅ P ( x , y ) ⋅ ψ ( x , y ) ⋅ dx ⋅ dy 0 . (5) kj j S V ∫∫S 17 mechanism Sthe 0following level min factors could be considered: heat effect on population, shock wave, Vmin 19 point with coordinates (x,y) under the condition that impact mechanism j is realized; ψ(x,y) is -H per year; N isisthethenumber of inhabitants; the area (6) ×isdathe × probability dxconstructions, × dy, of accident 18 26 debriswhere of×dV buildings and and others); Pj(x,y) probability of fatalitySinisdescribed the The with procedure above was used for the computaprobability of accident according to scenario k (fires, balls, of the point 20 per the year density of population distribution in fire the vicinity coordinates (x, y); N is 27 within which the (x,y) people may beaccident impactedper in the case of an accident atof a given facility; π =for the territory of the Russian Fedtion individual risk where – H is the probability of an year; N is the 19 point with coordinates under the condition that impact mechanism j is realized; ψ (x,y) is her phenomena may considered scenario events); Ekj(x,y) is theS is the area within which people may be 21 bethe number ofas people the zone of risk; eration with application number inhabitants; S minimum is the within area within which the people 28 3,14;ofVof and Vmax are and maximum possible values wind velocity; f(a,V) min 20 the density population distribution in the vicinity of the point withofcoordinates (x, y); N isis the of special GIS environment. Valact mechanism j in22 point (x,y) for the scenario(i.e. impacted thecase caseaccident of an theaccident accident theimpact zone of risk). ues of seismic risk obtained for separate cities and settlemay betheimpacted ininthe of atka(as given facility; 29 number density of function probability of wind direction a and wind velocity V; ψ(x,y) ismay the be density peopleof within zone of risk; S is the area within which owing 21 factorsthe be considered: heat on population, shock wave, mentspeople were averaged within the administrative divisions of πcould = 23 3.14;V arethe minimum and maximum possimaxReffect min and V Individual risk ei in the case of an accident at chemical hazardous facility (Methods…, 2002) 30 of population distribution in(a,V the vicinity of of therisk). point with coordinates (x, y); P[Д(x,y)] is the of individual seismic risk were con22 impacted in the case of the accident (i.e. the zone the country. Three maps ble values of wind velocity; f ) is the density function of and constructions,24and others); Pj(x,y)byis the probability of fatality in the is determined structed: Rs 1 2002) – probability of fatalities; Rs 2 – probability of probability ofRwind direction a and wind velocity V ; ψ(x,y) 23 Individual risk ei in the case of an accident at chemical hazardous facility (Methods…, tes (x,y) under the condition that impact mechanism j is realized; ψ(x,y) is 2π V max fatalities and injuries; R4 s 3 – probability of fatalities, injuries is the density ofHthe population distribution in the vicinity 24 is determined by= (a, Vcoordinates ) ⋅PP[[ Д (x,y)] (x, y(x, )] ⋅ψy); xN, yis)prob⋅ dV ⋅ da ⋅ dx , lation distribution inpoint theRvicinity of the pointfwith 25 (6)population due to the occurrence of and⋅ dy economic loss for is(the of the coordinates (x,y); eiwith N ∫∫ ∫0 ∫ 2π population V V min max S earthquakes within one year. Figure 1 shows the map of inability of the to be impacted by toxic dose in the ple within the zoneH of risk; S is the area within which people may be ( ( Rei = with f (a,(x,y); V ) ⋅ P[ Д (x,y) x, y )] ⋅isψthe x, toxic y ) ⋅ dVdose, ⋅ da ⋅which dx ⋅ dy, 25 (6) dividual seismic risk zoning point coordinates ∫∫ ∫ ∫ 26Nthewhere -H is the probability of accident per year; N is the number of inhabitants; S is the area Rs 1 (probability of fatalities). e of the accident (i.e. zone of risk). S 0 under V min time-dependent concentration of chemValues obtained for individual seismic risk vary from negis determined 27 within whichhazardous the peoplefacility may be impacted in2002) the case of an ligible, accident close at a givenzero, facility; n the case of an accident at chemical (Methods…, up πto=rather high values: more than ical hazardous material at a point with coordinates (x, y) by 26 where - H is the probability of accident per year; N is the number of inhabitants; −5 S is thetoarea 28 3,14; Vmin and Vmax are minimum and maximum possible values of 10 wind velocity; f(a,V) is the of fatalities (map Rs 1); more 30 × for the probability equation 27 within which the people may be impacted in the case of an accident atthan a given facility; = the probability of fatalities and injuries −5πfor 100 × 10 29 density function of probability of wind direction a and wind velocity V; ψ(x,y) is the density Ztk ax 28 3,14; Vmin and Vmax are minimum and maximum possible values of wind (map velocity; is thethan 150 × 10−5 for the probability of faRs f(a,V) 2); more f (a, V ) ⋅ P[ Д ((x,y) x,30 y )] ⋅= ψof( xpopulation ,(x,y,t) y ) ⋅ dV ⋅ da dx ⋅ dy, in the vicinity (6) ×⋅ dt, distribution of the point(7) with coordinates (x, y); P[Д(x,y)] is the and economic loss to population caused by 29 density function of probability of wind direction a and wind velocity V;talities, ψ(x,y) isinjuries the density in tn earthquakes per year (map Rs 3). Table 2 shows the extent of 4 30 of accident of population distribution in the vicinity of the point witharea coordinates (x, y); P[Д(x,y)] is the obability thetime number of inhabitants; S is the zones with different levels of individual seismic risk accordwhere –per tn .year; . . tk N is isthe interval within which the concening to maps Rs 1,Rs 2 and Rs 3. of the chemical substance (x,y,t) eople may betration impacted in the case of an accident is at dangerous; a given facility; π = is 4

are minimum and maximum possible values of wind velocity; f(a,V) is the probability ofwww.nat-hazards-earth-syst-sci.net/11/3025/2011/ wind direction a and wind velocity V; ψ(x,y) is the density

bution in the vicinity of the point with coordinates (x, y); P[Д(x,y)] is the

Nat. Hazards Earth Syst. Sci., 11, 3025–3033, 2011

3028

N. I. Frolova et al.: Estimation of loss caused by earthquakes and secondary technological hazards

Table 2. Values of individual seismic risk and extent of zones with different risk levels. Risk ranges 10−5 yr−1

< 0.1 0.1–1.0 1.0–5.0 5.0–10.0 10.0–30.0 30.0–100.0 100.0–150.0 > 150.0

Qualitative risk characteristics

Extent of zones, map Rs 1



106 km2

%

106 km2

%

106 km2

%

8.8 2.5 2.4 1.2 1.2 0.5 – –

53 15 14 7 7 3 – –

8.1 2.9 1.5 1.4 1.5 1.1 0.1 –

49 17 9 8 9 7 1 –

7.6 2.2 1.9 0.9 1.8 1.6 0.2 0.4

46 13 11 5 11 10 1 2

small moderate average high rather high extremely high

Fig. 1. Map of individual seismic risk Rs 1, 10−5 yr−1 , for the territory of the Russian Federation.

The computed values of individual seismic risk Rs 1 are more than 30 × 10−5 yr−1 for all administrative divisions within the Sakhalin area, Republic of Altay, Tuva, Dagestan and Northern Osetia. The highest values of individual seismic risk Rs 3 are obtained for Kamchatka, near lake Baikal, Republic of Buryatia, Irkutsk region, Altay kray, as well as for the Krasnodar region and Chechen Republic.


3.2

Seismic risk assessment and mapping at regional level

The Krasnodar region is characterised by a high density population and a rather high level of seismic hazard. According to maps of review seismic zoning of the Russian Federation territory, earthquakes with intensities I = 6 − 10 according to the MMSK-86 scale may occur here. The loss computations for this region were done taking into account secondary man-made accidents at fire, explosion and chemical www.nat-hazards-earth-syst-sci.net/11/3025/2011/


3029

Fig. 2. Map of seismic risk Rst 1, 10−5 yr−1 zoning for the territory of the Krasnodar region, taking into account technological accidents triggered by earthquakes.

hazardous facilities. For risk assessment and mapping at a regional scale, information about the population and building stock distribution in the region was updated. The detailed inventory data obtained during the last years and data on the built environment provided by the Regional Departments of EMERCOM of Russian Federation, was used to develop the building stock models for 83 settlements with a population more than 10 000 inhabitants, 281 ones with a population from 2000 up to 10 000 inhabitants and 1647 rural settlements with the number of inhabitants less than 2000 persons each. The models are characterised by the percentage of buildings of different types according to the MMSK-86 scale and their height. The parameters of regional vulnerability functions for different building types, as well as vulnerability functions for population, were checked for the Krasnodar region. Estimation of individual risk for the population of the Krasnodar region was carried out for the worst scenarios when earthquakes occurred during the night time. In order to estimate expected social losses within cities and towns, they were divided into unit sites. Then indexes obtained for each www.nat-hazards-earth-syst-sci.net/11/3025/2011/

unit site were summed up. The regional map of risk zoning (Fig. 2) includes two elements: risk for settlements with the number of inhabitants less than 1000, shown by “hypsometric” contours, and risk for settlements with the number of inhabitants more than 1000 shown by symbols (circles of different sizes and colours). The “hypsometric” scale is used to represent both elements on the map. Values of risk obtained for the Krasnodar region (Fig. 2) vary from negligible values up to rather high ones equal to 30.0 × 10−5 . On the whole, for more that 60 % of the Krasnodar region territory, the values of seismic risk computed taking into account the secondary technological accidents exceed the value of 1.0 × 10−5 (Table 3). High level of individual seismic risk results from relatively high seismic activity of the area under consideration, lack of proper earthquake resistant measures for the existing building stock, as well as the presence of fire, explosion and chemical hazardous facilities in the settlements’ territory. Comparison of seismic risk values Rs 1 averaged for administrative divisions of the Krasnodar area (Fig. 1), which were obtained at federal level, with risk values Rst 1 (Fig. 2) Nat. Hazards Earth Syst. Sci., 11, 3025–3033, 2011

3030

N. I. Frolova et al.: Estimation of loss caused by earthquakes and secondary technological hazards Table 4. Parameters of scenario earthquakes for the most hazardous source zones. No. of event, see Fig. 4

No. 1 No. 2 No. 3 No. 4 No. 5 Fig. 3. Distribution of seismic risk Rst 1, taking into account secondary technological accidents triggered by earthquakes for the settlements of the Krasnodar region. Table 3. Extent of zones with different risk levels for the Krasnodar area. Number

Risk ranges, 10−5 yr−1

Extent of zones, km2

1 2 3 4 5 6 7 8 9

< 0.5 0.5 × 10−5 . . . 1.0 1.0. . . 5.0 5.0. . . 10.0 10.0. . . 15.0 15.0. . . 20.0 20.0. . . 25.0 25.0. . . 30.0 > 30.0

19 651 5920 18 334 1681 8944 2110 12 802 6331 7456

for separate settlements of the Krasnodar area, which were obtained at regional level, allows us to draw the conclusion that average contribution of secondary technological accidents to the seismic risk is 10–15 %. The largest contribution of possible accidents triggered by earthquakes, equal to 20– 25 %, is obtained for cities Novorossijsk and Tuapse. Here the contribution of technological risk (Fig. 3 – dark colour) is connected mainly with the oil pipe facilities. 3.3

Scenario earthquake consequences

The Stavropol region is also characterised by a high density of population and a rather high level of seismic hazard. According to maps of review seismic zoning of the Russian Federation territory, earthquakes with intensities I = 6 − 9 according to the MMSK-86 scale may occur in the region. The expected loss computations due to scenario events were done for the Stavropol area taking into account possible earthquake source zones (Fig. 4). The parameters of scenario earthquakes are shown in Table 4.


Coordinates of epicentre ϕ0 N

λ0 E

43.75 43.92 43.98 43.71 44.98

43.08 42.49 43.22 42.44 41.97

Depth of source, km

M

20 15 15 20 10

7 6 6 7 5

Expected damage to the building environment and human casualties were estimated for the worst scenarios under the condition that 95 % of the population is inside buildings. Computations were carried out with special GIS code, which was developed for the region under consideration. The most severe damage to building stock was obtained for events no. 1 and no. 4 (Table 4) with magnitude M = 7 and source depth h = 20 km. In the majority of cities and settlements located within a radius of 30–40 km from the epicentres of events no. 1 (Fig. 5) and no. 4 the buildings and constructions may survive an average damage state of d = 4 (partial collapse), and within a radius of 50–100 km the average damage state d may be equal to 3 (heavy damage). Within the settlements, separate buildings of different types, classified according to MMSK-86, may undergo damage state from d = 5 (total collapse) up to slight damage d = 1, depending on their seismic vulnerability and distance from the epicentre. Table 5 lists the results of expected damage computations for large settlements in the Stavropol area for scenario events no. 1 and no. 4. In the case of scenario events no. 2 and no. 3 (Table 4) with magnitude M = 6 and source depth h = 15 km, the built environment in the majority of settlements may undergo in average slight damage (d = 1). The effect from such events may be felt in 400 settlements of the Stavropol area. Macroseismic effect from scenario event no. 5, with a magnitude M = 5 and source depth h = 10 km (Table 4) may reach I0 = 6 − 7 (MMSK-86 scale) in the epicentre area. The building stock in average may undergo slight damage (d = 1). Such an earthquake may be felt in 99 settlements of the Stavropol area. In Stavropol City located 5 km from the epicentre of the scenario event, the seismic intensity may reach I = 6.5 − 7 (MMSK-86 scale). In this city about 30 % of buildings may undergo moderate damage (d = 2), about 20 % – heavy damage (d = 3) and about 2 % – partial collapse (d = 4). When estimating scenario event consequences at urban and facility levels, the damage and loss resulting from secondary accidents triggered by earthquakes is also taken into account. Figure 6 shows the results of possible consequences simulation of chemical hazmat release at industrial facility. www.nat-hazards-earth-syst-sci.net/11/3025/2011/


3031

Fig. 4. Map of possible source zones for Stavropol region, according to Sobolev et al. (1996); numbered circles 1, 2, 3, 4 and 5 are scenario events’ epicentres (Table 4).

Fig. 5. Simulation of expected damage due to scenario event no. 1. Distribution of average damage states to building stock in settlements: black colour – total collapse, brown – partial collapse, red – heavy damage, yellow – moderate, green – slight, blue – no damage.

www.nat-hazards-earth-syst-sci.net/11/3025/2011/


3032


Table 5. Expected damage to building stock in settlements of the Stavropol area. Scenario event no. 1

Settlement name 1, km Essentuki Pyatigorsk Lermontov Kislovodsk Cherkessk

36 30 41 33 98

Scenario event no. 4

Percent of buildings with different d

dav

d =1

d =2

d =3

d =4

d =5

4 4 6 2 15

10 10 13 8 31

18 18 22 15 30

27 27 28 24 16

38 38 28 50 5

4 4 4 4 3

Percent of buildings with different d

1, km 52 61 62 31 64

d =1

d =2

d =3

d =4

d =5

10 13 13 4 8

19 23 23 10 26

27 28 28 18 32

24 20 20 27 22

14 8 8 39 9

dav 3 3 3 4 3

Table 6. Possible consequences of chemical hazmat release for two scenario accidents: most dangerous and probable events. Facility

Sochi Fishery Plant

Adler Water Supply System

Type and amount of chemical hazmat Estimated frequency of accidents, /year Size of the contaminated area, km2 Expected fatalities, persons Expected casualties, persons Expected damage, thousands rubles

Ammonia, 2 tons 1 × 10−5 0.82 28 224 120 000

Chlorine, 0.3 tons 1 × 10−4 0.3 2 16 321

Fig. 6. Simulation of possible contaminated zones in the case of an accident with chemical material release: 1 – source of release of 3.0 tons liquid chlorine; 2 – after 1 h; 3 – after 3 h and 4 – after 4 h.

By different colours the contaminated zones’ boundaries are shown for different time intervals from 1 h up to 4 h. Table 6 shows the results of possible consequences simulation of chemical hazmat release for two most dangerous and probable events in the case of strong earthquake in Sochi City. Figure 7 shows the results of individual risk estimation for oil reservoirs storage (tank farm) and pump station of oil pipe line system belonging to TRANSNEFT JSC for the case of technological accidents triggered by earthquakes. Nat. Hazards Earth Syst. Sci., 11, 3025–3033, 2011

Fig. 7. Zoning of the oil storage tank farm’s territory according to the level of individual risk.

4

Conclusions

The present paper describes the methodological procedure used for the assessment of seismic risk using special GIS environment. Examples of seismic risk assessment at different levels (urban and facility level), as well as scenario earthquake consequences’ assessment taking into account possible secondary accidents at fire, explosion and chemical hazardous facilities, are given.


N. I. Frolova et al.: Estimation of loss caused by earthquakes and secondary technological hazards The estimations of individual seismic risk obtained are used by EMERCOM of the Russian Federation, as well as by other federal and local authorities, for planning and implementing preventive measures, aimed at saving lives and protecting property against future disastrous events. The results also allow to develop effective emergency response plans taking into account possible scenario events. Taking into consideration the size of the oil pipe line systems located in the highly active seismic zones, the results of seismic risk computation are used by TRANSNEFT JSC. Acknowledgements. The work was funded by the Ministry of Education and Science of the Russian Federation. The authors thank their colleagues for continuing support, discussion of different steps of loss estimation procedures and uncertainties introduced by databases and mathematical models. Special thanks are to the staff of Extreme Situations Research Center for their contribution to GIS environment development at different levels. Edited by: E. Petrova Reviewed by: F. Sabetta and another anonymous referee

References Boissonnade, A. C. and Shah, H. C.:. Seismic Vulnerability and Insurance Studies, The Geneva Papers, 9, 223–254, 1984. Bonnin, J. and Frolova, N.: Near Real-Time Loss Assessment Due To Strong Earthquakes: The State of The Art, Proc. XXIXth General Assembly of European Seismological Commission (ESC2004), Potsdam, Germany, 2004. Bonnin, J., Frolova, N., Kozlov, M., Larionov, V., Sushchev, S., and Ugarov, A.: Experience of “Extremum” System Application for Operative Earthquake Loss Assessment, Proc. 28th General ESC Assembly, Genoa, Italy, 2002. Dolce, M., Kappos, A., Zuccaro, G., and Coburn, A. W.: Report of the EAEE Working Group 3: Vulnerability and Risk Analysis, Proceedings of 10th European Conference on Earthquake Eingineering, 3049–3077, 1995. Fournier d’Albe: An Approach to Earthquake Risk Management, Eng. Struct., 4, 145–152, 1982. Fournier d’Albe: The Assessment of Seismic Risk, Proceedings of UNDRO/UNESCO/USSR Seminar on Earthquake Prediction and Mitigation of Earthquake Losses, Dushanbe, UNDRO, Geneva, 1986. Frolova, N., Kozlov, M., Larionov, V., Suschev, S., and Ugarov, A.: Extremum System for Earthquake Risk And Loss Assessment, Proc. of SE-40EEE, Skopje-Ohrid, Macedonia, 2003. Frolova, N., Larionov, V., and Bonnin, J.: Simulation-based information systems for multi-hazard risk and near real time loss estimations due to strong earthquakes, Proc. TIEMS2007 Conference, Trogir, Croatia, 2007.


3033

Frolova, N., Larionov, V., and Bonnin, J.: Earthquake Casualties Estimations in Emergency Mode, in: Human Casualties in Earthquakes: Progress in Modelling and Mitigation, Adv. Nat. Technol. Hazards Res., Springer, 29, 107–123, 2011. Karnik, V.: Existing Recommendations and Definitions on Disaster Insurance, The Geneva Papers, 9, 3–7, 1984. Karnik, V. and Algermissen, S. T.: Seismig Zoning: The Assessment and Mitigation of Earthquake Risk, UNESCO, Paris, 11– 47, 1978. Larionov, V. and Frolova, N: Peculiarities of seismic vulnerability estimations. In: Natural Hazards in Russia, volume 6: Natural Risks Assessment and Management, Publishing House “Kruk”, Moscow, 120–131, 2003a (in Russian). Larionov, V., Suschev, S., Ugarov, A., and Frolova, N.: Seismic risk assessment with GIS-technology application, in: Natural Hazards in Russia, volume 6: Natural Risks Assessment and Management, Publishing House “Kruk”, Moscow, 209–231, 2003b (in Russian). Methods of earthquake consequences forecast: Moscow, VNII GOChS, Extreme Situations Research Center, Seismological Center of IGE RAS, 2000, 27 pp., 2000 (in Russian). Methods of integrated natural and technological risk assessment: Moscow, VNII GOChS, Extreme Situations Research Center, Seismological Center of IGE RAS, 2002, 35 pp., 2002 (in Russian). Mitigating Natural Disasters: Phenomena, Effects and Options, A Manual for Policy Makers and Planners, UNDRO, New York, UN, 164 pp., 1991. Ranguelov, R.: Natural Hazards – Nonlinearities and Assessment, Acad. House M. Drinov, Sofia, 237 pp., 2011. Risk Assessment and Mapping Guidelines for Disaster Management: available at: http://register.consilium.europa.eu/pdf/en/10/ st17/st17833.en10.pdf, 2010. Shebalin, N., Gekhman, A., and Shestoperov, G.: Development of improved version of seismic intensity scale (MMSK-86) on the bases of MSK-64 scale and scale of Joint Council on Seismology and Earthquake Engineering – 73, Report on scientific research study within the Federal Program 0.74.03, number of state registration 01814003271, Joint Council on Seismology and Earthquake Engineering, Academy of Sciences of the USSR, Moscow, 1986 (in Russian). Sobolev, G., Rogozhin, E., Bogachkin, B., Ioganson, L., Nechaev, Yu., Reisner, G., and Sidorin, A.: Development of seismogenic zones map using over regional method. Institute Physics of the Earth Report “Compiling the map of detailed seismic zoning of the Stavropol region territory”, vol. 3, Moscow, 1996 (in Russian). UNISDR Terminology on Disaster Risk Reduction: available at: http://www.unisdr.org/eng/terminology/ UNISDR-terminology-2009-eng.pdf, 2009.
