Accepted Manuscript

A GIS System Integrated with Earthquake Vulnerability

Assessment of RC Building

M.C. Ningthoujam, Radhikesh P. Nanda

PII: S2352-0124(18)30077-8
DOI: doi:10.1016/j.istruc.2018.07.013
Reference: ISTRUC 307
To appear in: Structures
Received date: 8 February 2018
Revised date: 19 June 2018
Accepted date: 27 July 2018

Please cite this article as: M.C. Ningthoujam, Radhikesh P. Nanda , A GIS System
Integrated with Earthquake Vulnerability Assessment of RC Building. Istruc (2018),
doi:10.1016/j.istruc.2018.07.013

This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.
 ACCEPTED MANUSCRIPT

A GIS system integrated with earthquake

vulnerability assessment of RC building

M. C. Ningthoujam1† and Radhikesh P. Nanda2‡

1. Research Scholar, Department of Civil Engineering, NIT Durgapur, Durgapur-713209, India (E-mail:

T
monikachanu2012@gmail.com)

IP
2. Associate Professor, Dept. of Civil Engineering, NIT Durgapur, Durgapur-713209, India (E-mail:

CR
rpnanda2002@yahoo.co.in)

ABSTRACT

US
The paper presents methodology for developing rapid visual screening score sheet for seismic vulnerability

assessment of existing RC buildings based observed damage building stocks. This study takes into consideration the
AN
building attributes such as soft storey, substantial overhang, re-entrant corners, the age of buildings, apparent

construction quality, eccentric staircase location with respect to building plan, maintenance, type of soil and the
M

number of stories in the building. The proposed rapid visual screening score sheet, where the age of the building
ED

which is one of the important factors for strength deterioration of structure is considered besides structural and

construction deficiencies. The result of the vulnerability score sheet presented in term of damage grade ranging from
PT

grade 1 to grade 5 as per damage grade definition of Indian codal system. Application of the procedure is illustrated

with a case study example of Ward no.6 of Uripok constituency of Imphal city, Manipur of India and the results for
CE

the vulnerability assessment are presented in colour-coded maps on GIS platform. The present score sheet resides in

simple and robust assessment methodology, compatible with Indian condition and in the capacity of providing
AC

intuitive representations of the spatial distribution of damage results.

Keywords: Vulnerability, Age of building, Construction quality, Maintenance condition, GIS.

1
 ACCEPTED MANUSCRIPT
1. INTRODUCTION

Seismic vulnerability assessment is very important for earthquake risk mitigation programme. In general,

vulnerability may be defined as susceptibility to damage during an earthquake which depends on the quality of

exposure conditions of the existing buildings. Many researchers have proposed many seismic vulnerability

assessment methodologies. Different vulnerability assessment methods have many common features and various

authors have classify them into different groups based on their study. Many authors namely Calvi et al (4), Perrone

et al. (28), Yadollahi et al. (31), Lang et al. (17), NIDM (19), Pankaj and Manish (27) have done a proper review on

T
IP
available seismic vulnerability assessment methodologies. Calvi et al. (4) performed detailed study of available

vulnerability assessment methods and classify them broadly as empirical, analytical and hybrid method. In empirical

CR
vulnerability procedures, the damage scale is used in reconnaissance efforts to produce post-earthquake damage

statistics, whilst in analytical procedures, this is related to limit-state mechanical properties of the buildings, such as

US
inter-storey drift capacity (Calvi et al., (4)). The hybrid method is the combination of empirical and analytical

method. One of the empirical or hybrid method is rapid visual screening (RVS) which facilitates the assessment of
AN
huge buildings during a short period of time. Lourenco and Roque (18), Giulio et al. (9), Jason et al. (15), Naili and

Alzeria (20) worked on seismic vulnerability of masonry monumental structures. The methods are simple, fast and
M

based on a simplified geometric approach for immediate screening of the large number of buildings at risk.
ED

Alternate to RVS is detail evaluation. Since Rapid Visual Screening (RVS) is a simple procedure for quick

assessment of large building stock requiring only visual evaluation and limited additional information for
PT

preliminary vulnerability assessment. Hence, due to its simple and ability to assess large building stock in less time,

it is the most appropriate for the case of India with heavy papulation.
CE

According to the difference in building codes and construction practices, the scoring system and parameters taken

for assessing the vulnerability of buildings differ. In most of the RVS procedure, a basic score is assigned to
AC

building based on the seismicity of the region and type of the building. The basic score is then modified by the

presence of certain earthquake inducing features in the building to get the final score. The final score will be co-

related with the damage grade for the particular building. FEMA 154 (6) procedure for Rapid Visual Screening

(RVS) was first proposed in the U.S.A in the year 1988. However, the procedure was further modified by

incorporating latest technological advancements and lessons from earthquake disasters and published as FEMA 178

(7), FEMA 310 (8) and FEMA 154 (6). FEMA 154 (6) method assigns a basic structural score based on seismic

hazard intensity of the region, building type and its lateral load resisting system. Performance modifiers are

2
 ACCEPTED MANUSCRIPT
specified to take into account the effect of number of storeys, plan and vertical irregularities, pre-code or post-

benchmark code detailing, poor condition of the building and soil-type. A score of 2 is suggested as a cut-off and

score less than 2 require detailed analysis. Even though this RVS procedure was originally developed for typical

constructions in the U.S.A, it has been widely used in many other countries after suitable modifications. Some

modified RVS methodologies from FEMA code are New Zealand Society for Earthquake Engineering (NZSEE

(23)) which was proposed in 1996 and RVSP method (Nanda et al. (24)) which has been modified for Indian

conditions. Wallace and Miller (Wallace et. al. (30)) also followed the RVS methodology proposed in FEMA 154

T
(6) to identify potential seismic hazards for Oregon's public facilities, including schools, hospitals, fire stations, and

IP
emergency response centers. The procedure of RVS used in India given by Arya (2) utilises a damageability grading

CR
system based on the primary structural lateral load-resisting system, and building attributes that modify the seismic

performance expected for this lateral load-resisting system along with non-structural components. The screening is

US
based on code based seismic intensity scale, building type and damageability grade as observed in past earthquake

and covered in Medvedev Sponheuer Karnik (MSK) and European macro-intensity. The RVS procedure developed
AN
by Greek in 2000 (OASP, (5)), needs to identify both the primary structural lateral load resisting system and

structural materials of the building to calculate the basic score. The basic score will then be modified to get the basic
M

structural hazard score by identifying both the seismic zone and three significant structure characteristics (weak
ED

story, short columns and regular arrangement of the masonry). Finally, this score will be modified by identifying

some modifiers related to the observed performance attributes to arrive at the final score. Italy method (Angeletti et
PT

al. (1)) developed in Italy is based on eleven assessed parameters. The eleven parameters are resisting system type

and organization, resisting system quality, conventional resistance, location and soil condition, diaphragms, plan
CE

configuration, vertical configuration, connectivity between elements, low ductility structural members, non-

structural elements and preservation state. RVS for Indian condition (Jain (14)) was developed based on database of
AC

damaged buildings in Bhuj earthquake of January 26, 2001. The parameters considered in this procedure are

building typology based on occupancy type (residential or non-residential building), presence of basement, number

of storeys, maintenance condition of building, asymmetric location of staircase with respect to plan, presence of re-

entrant corners, presence of open storey, presence of stub column, presence of substantial overhang and presence of

short column. Based on the type of seismic zone and soil type, base score is assigned to the buildings and later on

modified based on the parameters mentioned above. Turkey method (BU-ITU-METU-YTU (3)) developed a

preliminary seismic assessment procedure for reinforced concrete buildings in Turkey which take into account the

3
 ACCEPTED MANUSCRIPT
influence of structural configuration as well as the secondary factors. The Japanese procedure (JPDPA (16)) is based

on Seismic Index (Is) for total earthquake resisting capacity of a storey which is estimated as the product of a basic

seismic index based on strength and ductility indices, an irregularity index, and a time index. Canada method

(NRCC (22)) suggested by National Research Council, Canada (NRCC (22)) is based on a seismic priority index

which accounts for structural as well as non-structural factors including soil conditions, building occupancy,

building importance and falling hazards to life safety and a factor based on occupied density and the duration of

occupancy. Sucuoglu et al. (29) proposed a simple screening procedure for three- to six-story substandard

T
concrete buildings in Turkey; the procedure is calibrated with field data compiled after the 1999 Duzce Earthquake.

IP
The method assigns a basic Score to different RC-frame buildings depending on the number of stories and the

CR
seismic zone. Vulnerability coefficients consider the presence of soft stories, apparent quality, and heavy overhangs.

Hospital Safety Index by the PAHO (Pan American Health Organization) and the WHO (World Health

US
Organization) are specifically developed for hospital buildings (PAHO/WHO (26)). The Hospital Safety Index

estimation facilitates the determination of the hospital's ability to continue providing services after a natural disaster.
AN
The methodology is based on questionnaires. The checklist contains 145 variables, each of which has three safety

levels: low, medium, and high. The method is articulated in three principal modules, which consider the three main
M

parameters that affect the vulnerability of hospitals, the structural safety, nonstructural safety, and functional safety.
ED

In determining the Safety Index, a different weight is assigned to each module; in particular, the weights assign 50

% of the index for the module to the structural safety, 30 % to the nonstructural module, and 20 % to the functional
PT

capacity. Perrone et al. (28) proposed a RVS method for the evaluation of a Safety Index for hospital buildings. The

methodology is based on the Hospital Safety Index initially proposed by the PAHO which is modified to suit the
CE

peculiarities of the Italian context. The assessment of the vulnerability of structural elements, nonstructural elements

and plants as well as organizational aspects are considered in this method.

AC

The objective of this study is to develop a vulnerability assessment methodology in the form of RVS score sheet

using database of affected R.C.C. buildings compiled after the 4th January, 2016 Manipur earthquake. Finally, the

methodology will be implemented on a case study area of Ward number 6 of Uripok constituency of Imphal city,

India. The classified damage will be shown in spatial map on GIS (Geographical Information System) platform

which will be helpful in disaster management programmes.

2. OBSERVED AREA

4
 ACCEPTED MANUSCRIPT
Besides other damaged structures, only database of RC buildings that were affected during the 4 th January 2016

Manipur earthquake of Moment magnitude 6.7 were compiled consisting of 396 buildings in random for developing

a vulnerability assessment methodology in the form of RVS. Intensive field survey was conducted around the

Imphal city after the earthquake and a database was developed that included affected buildings ranging from grade 1

to grade 5 as per damage grade definition of Indian codal system (IS 1893 (13)). European Macroseismic Scale

(Grunthal (10)) also classify damage grade into five damage grade system. The comparison of Indian codal damage

classification and that of European Macroseismic Scale is given in table 1. The building database was formed from

T
IP
affected building of Uripok, Dewlahland, Keisampat, Naoremthong and Thangal bazaar of Imphal city. The damage

distribution of the investigated buildings is presented in table 2.

CR
Table 1. Comparison of Indian codal and that of European Macroseismic Scale damage classification

Classification of damage to reinforced concrete buildings Classification of damage to reinforced concrete

US
as per EMS (Grunthal (10) buildings as per IS: 1893 (12)

Grade 1: Negligible to slight damage Grade 1: Slight damage

AN
(No structural damage, slight non-structural damage) Fine cracks in plaster: fall of small pieces of plaster
M

Fine cracks in plaster over frame members or in walls at

the base, fine cracks in partitions and infills.

ED

Grade 2: Moderate damage Grade 2: Moderate damage

(Slight structural damage, moderate non-structural Small cracks in plaster: fall of fairly large pieces of
PT

damage) plaster: Pantiles slip off cracks in chimneys parts of

Cracks in columns and beams of frames and in structural chimney fall down
CE

walls, cracks in partition and infill walls; fall of brittle

AC

cladding and plaster, Falling mortar from the joints of

wall panels.

Grade 3: Substantial to heavy damage (moderate Grade 3: Heavy damage

structural damage, heavy non-structural damage) Large and deep cracks in plaster: fall of chimneys

Cracks in columns and beam-column joints of frames at

the base and at joints of coupled walls. Spalling of

concrete cover, buckling of reinforced bars, large cracks

5
 ACCEPTED MANUSCRIPT
in partition and infill walls, failure of individual infill

panels.

Grade 4: Very heavy damage (heavy structural damage, Grade 4: Destruction

very heavy non-structural damage) Gaps in walls: parts of buildings may collapse:

Large cracks in structural elements with compression separate parts of the buildings lose their cohesion:

failure of concrete and fracture of rebars; bond failure of and inner walls collapse

beam reinforcing bars; tilting of columns.

T
IP
Collapse of a few columns or of a single upper floor.

CR
Grade 5: Destruction (very heavy structural damage) Grade 5: Total damage

Collapse of ground floor parts (e.g. wings) of the Total collapse of the buildings

US
building.
AN
Table 2. Damage distribution of the investigated building of Imphal city, Manipur, India

Damage Observed
M

Grade1(Slight Grade2(Moderate Grade3(Heavy Grade4(Destruction) Grade5(Total

damage) damage) damage) damage)

ED

235 59 50 39 13
PT

An investigation is conducted on the building database to check on the causal parameters of the damage distribution
CE

of buildings during the earthquake. The structural and construction deficiencies along with associated damages in

buildings during the earthquake were investigated. During the survey, soft storey failure were observed as shown in
AC

fig.1. Due to unavailability of spaces in this congested city area, the ground floor is left open for parking without

infill walls in the majority of the buildings. Enclosed projections of the upper storeys of a building (other than

balconies) creating an upper footprint that projects by more than 1.2 m from the ground floor footprints, on one or

more sides of the building have been considered as overhang (Jain et al. (14)). Substantial projection of the upper

storey of the building of more than 1.2 m from the ground floor footprints was observed in the affected buildings.

Most of the overhang was protruded in order to get space for room apart from architectural features and balconies.

This increases the vulnerability of the buildings. Floating columns from the first storey were also observed in the

6
 ACCEPTED MANUSCRIPT
affected buildings as shown in fig.2. Upper stories are constructed more in balconies in order to cover up maximum

area leading to floating column in the upper storey of the building. Due to lack of continuous vertical load path

distribution, the lateral forces are not effectively transferred to the foundation. The presence of re-entrant corners are

one the serious plan irregularities that result in the poor seismic performance of buildings. Plan irregularities such as

L, U and T-shaped as sown in fig.3. (IS 1893 (12)) are called re-entrant corners. Buildings with re-entrant corners

are common features that have been seen during the survey. A typical map with larger buildings showing the

different shapes with re-entrant corners of the building database is shown in fig.4. During the survey, buildings with

T
IP
regularity in plan and elevation were also observed and an example is shown in fig.5. Buildings remain for several

years without getting due maintenance. Water stagnation, paint pealing, plaster break-off, fungus growth, cracking

CR
of external rendering and cover concrete are common and widespread. Penetration of moisture into reinforced

concrete components promotes corrosion process and further damages the concrete cover. Addressing the issue of

US
deterioration, carbonation,chloride ingress, leaching,sulphate attack, alkali-silica reaction and freeze-thaw are the

known responsible natural causes. Out of these, the first three can all lead to corrosion of reinforcement. For the
AN
present study, the condition of the building with age greater than 35 years were analyzed for maintenance condition

effect. Based on the maintenance condition of the building, a decision can be made on the maintenance level of the
M

building as poor, moderate and good. Algae, fungus growth or efflorescence on the surface of wall, mesh pattern of
ED

cracks (suggesting drying shrinkage, surface crazing, frost attack or alkali-aggregate reaction) are taken due to

moderate maintenance condition. While water stagnation, cracking of external rendering and cover concrete, micro
PT

and macro cracking of concrete are caused due to poor maintenance condition. Both poor and moderate maintenance

condition were observed during the survey. Faulty construction practices and lack of quality control has played a
CE

main role in causing damage during the earthquake. Testing for proper material properties before construction was

seldom performed as buildings in Imphal city is mainly constructed under local mason. Minor and major cracks and
AC

falls of plaster were seen near the column-beam-wall junction. So, the bonding between various structural

components was found to be poor that may be due to the poor construction qualities during the construction time.

Some of the government market at Thangaal bazaar and residential buildings of Keisampat and Dewlaland that have

been constructed recently suffered damage grade 5 and poor construction quality are taken to be the main cause of

the failure. Buildings with age less than 35 years were observed and classified under good, moderate and poor

categories for the current study for construction quality. Spalling of concrete surface, fall of plaster, pattern cracks in

concrete, corrosion of reinforcement materials, honeycombs on the surface of concrete, incorrect alignment of beam,

7
 ACCEPTED MANUSCRIPT
column and slab, incorrect stirrup and tie hooks etc. are classified under poor construction quality while surface or

appearance defects on ceilings, floors, concrete walls, brick walls, concrete surface such as dips, uneven surfaces

with missing grout and mastic sealant, hit, scratches and efflorescence, etc. are classified under moderate

construction quality for the present study. Fig.6. shows maintenance condition and construction quality of the

building database. Staircases are areas of potential damage in buildings, if not accounted for in staircase design and

construction. The asymmetric location of the staircase can be effectively used as a vulnerability parameter in rapid

visual assessment as it has implications on the torsional behavior of the building and yet is easy to observe. If the

T
IP
staircase location was found symmetric with respect to either one (Fig.7. a, b, or c) or both the horizontal axes

(Fig.7. d or e), the building was considered to be symmetric (Jain et al. (14)). In most of the affected buildings, the

CR
staircase was found to be placed asymmetrically with respect to plan leading to torsional effect. During the survey

work, the position of the staircase (Eccentric staircase location with respect to plan) was checked and buildings with

US
eccentric staircase were analyzed. Fig.8. shows the staircase failure of the building database. The amplification of

ground motion due to local soil condition is one of the major factor that increases seismic damage. Pallav K et al.
AN
(25) carried out the soil classification based on standard penetration test (SPT-N) in 122 boreholes in Imphal city.

Out of 122 boreholes, 144 boreholes were classified as soil class E-type (N<15) while remaining 8 boreholes were
M

of soil class D-type (15<N<50) following the classification concept applied by IBC (11). The number of storey and
ED

age of building is also taken as vulnerability parameters for the current study.
PT
CE
AC

Fig.1. Soft storey failure in buildings

8
 ACCEPTED MANUSCRIPT

T
IP
Fig.2. Building with floating column

CR
US
AN
Fig.3. Re-entrant corners (IS 1893 (12))
M
ED
PT
CE
AC

Fig.4. A typical map with larger buildings showing the different shapes with re-entrant corners of the damage
building database

9
 ACCEPTED MANUSCRIPT

T
IP
CR
Fig.5. Building with regular plan and elevation.

US
AN
M
ED
PT
CE
AC

10
 ACCEPTED MANUSCRIPT

Fig.6. Maintenance condition, construction and material quality of the building database

T
IP
CR
US
AN
Fig.7. Symmetric plans with respect to staircase locations (Jain et al. (14))
M
ED
PT
CE
AC

Fig.8. Staircase failure of the damage building

The variation in the grade of damage of the buildings is assumed to be functionally related with the above mentioned

ten variables like soft storey, substantial overhang, floating column, re-entrant corners, age of building, apparent

11
 ACCEPTED MANUSCRIPT
construction quality, and asymmetry of staircase location with respect to plan, maintenance condition, type of soil

and number of storey of the building. These expected causal variables are treated to be the independent variables of

interest in the regression models. The measurements of the variables for statistical analysis are defined as the

following.

(a) Dependent variable (Y): Grade of Damage of the building such as 1 for slight damage; 2 for moderate

damage; 3 for heavy damage; 4 for destruction; and 5 for total damage.

(b) Independent variable (Xi): The description of independent variables are given in table 3.

T
IP
Table 3. Description of independent variables

CR
Sl. No. Predictor variable Type Measurement

1 Age of building (X1) Quantitative Number of years

US
(1, 2, 3 . . .)

2 Number of storey (X2) Quantitative Number of storey

AN
(1, 2, 3, … )
M

3 Apparent construction quality (X3) Ordinal Good=0, Moderate=0.5, Poor=1

4 Maintenance condition (X4) Ordinal Good=0, Moderate=0.5, Poor=1

ED

5 Type of soil (X5) Ordinal Poor=0, Medium=0.5, Hard=1

6 Soft storey (X6) Dummy Present=1, Absent=0

PT

7 Substantial overhang (X7) Dummy Present=1, Absent=0

CE

8 Floating column (X8) Dummy Present=1, Absent=0

9 Re-entrant corners (X9) Dummy Present=1, Absent=0

AC

10 Asymmetry of staircase location with respect Dummy Present=1, Absent=0

to plan (X10)

3. METHODOLOGY

Base on the mentioned ten predictor variables and one response variable, a multiple regression analysis is

performed.

12
 ACCEPTED MANUSCRIPT
The model developed for the damage grade of the buildings is highly significant with respect to ten variables as

witnessed by the F-test, 58.1 (P<0.001). The ten independent variables explain 60% of the total variations in the

grade of damage in the study area (M. C. Ningthoujam and Radhikesh P. Nanda (21)).

A regression model is fitted based on the ten independent variables (X i) and expected damage grade (Y) as given in

Eq. (1) below

Y = A + b1 X1 + b2 X2 + b3 X3 + b4 X4 − b5 X5 + b6 X6 + b7 X7 − b8 X8 + b9 X9 − b10 X10 (1)

T
IP
Where, A is the constant term and b1 to b10 are the coefficients of the respective independent variables. The values of

A and the bi are obtained such that the sum of the squared residual is minimum as given in Eq. (2).

CR
∑ e2 = ∑(Yi − ODGi )2 (2)
i i

US
Where, e2= sum of the squared residual; Yi= expected damage grade and ODGi= observed damage grade

The degree of explanation of the dependent variable by the independent variables is indicated by R2 value which is
AN
given by Eq. (3).
M

SSerr
R2 = 1 − (3)
SStot
ED

Where,

SStot = ∑(Yi − ̅̅̅̅̅̅̅

ODGi)² (4)
PT

And,
CE

SSerr = ∑(ODGi − Yi )² (5)

i
AC

After analysing with multiple regression analysis, the fitted regression model becomes Eq. (6) with F = 58.08, P

<0.001 and R2 = 0.60. Hence, as per F value with P <0.001 showing the fitted model to be statistically highly

significant. R2 value indicates the independent variables cab explain 60% of the variation in damage grade of the

building

Y = 2.87 + 0.02X1 + 0.12X2 + 1.08X3 + 0.65X4 − 2.42X5 + 0.03X6 + 0.48X7 − 0.11X8 + 0.22X9 (6)

− 0.30X10

13
 ACCEPTED MANUSCRIPT
To explore the most important determinants of the damage grade of the buildings under study, a stepwise regression

analysis is performed which is shown in Table 3. In this regression analysis, the value of regression coefficient (b)

with 95 % confidence interval (CI), probability level (P) of test statistics (t) of each independent variable has its own

interpretation apart from statistical significance. The analysis is carried out in six steps say from Model 1 to Model

6.

The function ‘Y’ for the stepwise regression model is calculated in the similar manner based on the R 2 value. The

forward selection method is adopted. The variable that has the highest R-Squared is selected. At each step, the

T
IP
variable that increases R-Squared the most is selected and the process is ended when none of the remaining variables

are significant. F-test is done when several parameters are involved at once. Table 4 summarizes the F-test with the

CR
F value and the corresponding P value. From F with P value less than 0.05 is taken as statistically significant while P

value less than 0.001 is taken as statically highly significant. Model 1 to model 6 are statistically highly significant

US
as shown by F-test. Also it reveals that only six predictor variables have a significant contribution to damage grade

of the buildings in the study area. They are the type of soil, apparent construction quality, maintenance, the age of
AN
the building, substantial overhang, and a number of stories. It reveals that the regression model can be fitted with the

important six independent variables. In the last model that is model - 6, all the variables except type of soil viz.,
M

apparent construction quality (b=1.09 with 95 % CI: 0.81-1.36; P<0.001), maintenance (b=0.65 with 95 % CI: 0.40-
ED

0.91; P<0.001), age of building (b=0.02 with 95 % CI: 0.01-0.02; P<0.001), substantial overhang (b=0.41 with 95 %

CI: 0.14-0.69; P<0.01), and number of storey (b=0.10 with 95 % CI: 0.02-0.18; P<0.05) have been observed to be
PT

positively and statistically influential on the damage of the buildings under observation. Only one, type of soil (b=-

2.43 with 95 % CI: -2.81 to -2.06; P<0.001) has negatively as well as high significant impacts on damage grade.
CE

The stepwise regression models are observed as the following (Eq. (7) to Eq. (12)).
AC

Model-1 (F=175.03, P<0.001, R2 = 31 %):

Y = 4.58 − 3.06X5 (7)

Model-2 (F=195.76, P<0.001, R2 = 49 %):

Y = 3.69 − 2.85X5 + 1.62X3 (8)

Model-3 (F=158.73, P<0.001, R2 = 55 %):

14
 ACCEPTED MANUSCRIPT
Y = 3.38 − 2.66X5 + 1.20X3 + 0.86X4 (9)

Model-4 (F=132.64, P<0.001, R2 = 58 %):

Y = 3.16 − 2.62X5 + 1.18X3 + 0.67X4 + 0.02X1 (10)

Model-5 (F=110.87, P<0.001, R2 = 59 %):

Y = 2.97 − 2.43X5 + 1.10X3 + 0.64X4 + 0.02X1 + 0.45X7 (11)

T
IP
Model-6 (F=94.72, P<0.001, R2 = 60 %):

CR
Y = 2.70 − 2.43X5 + 1.10X3 + 0.65X4 + 0.02X1 + 0.41X7 + 0.10X2 (12)

US
The value of R2 can be improved from Model-1 to Model-6 which is gradually from 31 % to 60 %. In the first

model (Model-1), type of soil can explain 31 % of the total variations in the average grade of damage of the
AN
buildings under study in the present analysis. It quantitatively reveals that, if a level changes upward in the type of
M

soil, the grade of damage to the building decreased by at least 3 levels in the damage grade irrespective of other

nine independent variables. The degree of explanation of the models can gradually be improved to be 49 % in the
ED

Model-2 with the two independent variables viz., type of soil and apparent construction quality. In this second

model without considering the eight variables, the average grade of damage is found to be 3.7 considering the
PT

effects of only two elsewhere variables. As increases in ordinal scale 0, 0.5, 1 level in apparent construction

quality, the level of damage is also correspondingly increased by 1.6 times while controlled the effect of type of
CE

soil. In a similar way, when a level changes in ordinal scale 0, 0.5, 1 in the type of soil, the grade of damage is

found to be decreased by 2.9 times when the effect of apparent construction quality is controlled (Table 5).
AC

The model fitting is thus progressing monotonically and hence the last sixth model has been achieved. In the last

model, after controlling the joint effect of five variables namely, type of soil, apparent construction quality,

maintenance condition, the age of the building, substantial overhang, the damage grade of the buildings is

observed to be increased by 0.1 times as an increment of one storey. This variation is found to be statistically

significant at 5 % probability level of significance.

15
 ACCEPTED MANUSCRIPT
Table 4. Model Summary

Model R2 R2-change F (P) Durbin Watson

1 0.308 0.308 175.03 (<0.001)

2 0.499 0.191 195.76 (<0.001)

3 0.548 0.049 158.73 (<0.001)

1.46
4 0.576 0.027 132.64 (<0.001)

T
5 0.587 0.011 110.87 (<0.001)

IP
6 0.597 0.010 94.72 (<0.001)

CR
Table 5. Stepwise regression analysis on grade of the damage of buildings

US
95 % CI for B

Model Variable b SE t P Lower Upper

AN
1 Constant 4.7 -3.06 21.42 .000 4.16 5.00

Type of soil -3.06 -3.06 -13.23 .000 -3.52 -2.61

M

2 Constant 3.69 0.20 18.84 .000 3.31 4.08

ED

Type of soil -2.85 0.20 -14.41 .000 -3.24 -2.46

Apparent construction quality 1.62 0.13 12.26 .000 1.36 1.88

PT

3 Constant 3.38 0.19 17.54 .000 2.99 3.75

Type of soil -2.66 0.19 -13.98 .000 -3.03 -2.29

CE

Apparent construction quality 1.20 0.14 8.43 .000 0.92 1.48

Maintenance condition 0.86 0.13 6.55 .000 0.60 1.12

AC

4 Constant 3.16 0.19 16.51 .000 2.79 3.54

Type of soil -2.62 0.19 -14.15 .000 -2.98 -2.25

Apparent construction quality 1.18 0.14 8.59 .000 0.91 1.45

Maintenance 0.67 0.13 5.05 .000 0.41 0.93

Age of building 0.02 0.00 5.01 .000 0.01 0.02

5 Constant 2.97 0.20 15.03 .000 2.58 3.36

16
 ACCEPTED MANUSCRIPT
Type of soil -2.43 0.19 -12.67 .000 -2.80 -2.05

Apparent construction quality 1.10 0.14 7.95 .000 0.83 1.37

Maintenance condition 0.64 0.13 4.90 .000 0.39 0.90

Age of building 0.02 0.00 5.43 .000 0.01 0.02

Substantial overhang 0.45 0.14 3.27 .001 0.18 0.72

6 (Constant) 2.70 0.23 11.96 .000 2.25 3.14

Type of soil -2.43 0.19 -12.78 .000 -2.81 -2.06

T
IP
Apparent construction quality 1.09 0.14 7.88 .000 0.81 1.36

Maintenance condition 0.65 0.13 5.01 .000 0.40 0.91

CR
Age of building 0.02 0.00 5.63 .000 0.01 0.02

Substantial overhang 0.41 0.14 3.00 .003 0.14 0.69

US
No. of storey 0.10 0.04 2.52 .012 0.02 0.18
AN
Base on the statistical regression analysis, a rapid visual screening procedure is proposed taking into account the
M

building parameters, namely type of soil, apparent construction quality, maintenance condition, the age of the

building, substantial overhang and number of the storey as given by model 6 (Eq. (12)). The damage grade (Y) of
ED

model 6 (Eq. (12)) depends on the constant term, coefficients of the independent variables and the presence of the

vulnerable parameters which is the six independent variables in our study. Hence, based on the model 6, the
PT

equation for the final score (S) which indicates the damage grade for vulnerability analysis is generated and given in

Eq. (13). The constant term is denoted by basic score (BS) while the coefficients are denoted by Vulnerability Score
CE

Modifier (VSM). The basic score is the constant term which is 2.7 for zone V while for zone II to zone III, the basic

score is given by the constant term of the regression model normalized with the PGA (Peak Ground Acceleration)
AC

value of the respective zone i.e., 0.1, 0.16 and 0.24 for zone II, zone III and zone IV respectively as per IS 1893

(12). For the present study, the database is taken for zone V (IS 1893 (13)), hence a basic score of 2.70 is given to

Zone V for soft soil condition. Similarly, a basic score of 0.75, 1.2 and 1.8 are given to zone II, zone III and zone IV

respectively for soft soil condition. The vulnerability Score (VS) is based on the present vulnerability parameters

which is then modified by using Vulnerability Score Modifier (VSM). The VSM are the respective co-efficient for

each independent variable of the regression model 6 (Eq. (12)).The VS is taken as 1 for substantial overhang,

construction quality, maintenance condition and type of soil while the actual age and number of storey are taken for

17
 ACCEPTED MANUSCRIPT
the age of building and number of storey parameter respectively. Base on the final score, damage grade can be

assigned to the building which is given in table 6. A data collection form for the rapid visual screening of RC frame

buildings is given in table 7 based on this model.

𝑆 = 𝐵𝑆 + ∑(𝑉𝑆 𝑋 𝑉𝑆𝑀 ) Eq. (13)

Where S=Final score, BS=Basic Score, VS=Vulnerability Score and VSM=Vulnerability Score Modifier

Table 6. Assigned damage grade based on Expected damage score

T
IP
Expected damage grade Assigned damage grade

S<1 Damage Grade 1 or probability of Un-damage

CR
1≤S<1.5 Damage Grade 2

1.5≤S<2 Damage Grade 3

US
2≤S<2.65 Damage Grade 4

2.65<S Damage Grade 5

AN
M
ED
PT
CE
AC

18
 ACCEPTED MANUSCRIPT
Table 7. Data Collection Form for Rapid Visual Screening of RC Frame Buildings

Building Name
Address
Other Identifiers
GPS Co-ordinates (If available)
No. Stories Year Built
Total Floor Area (sq. Ft. /sq. m)
Seismic zone
Surveyor Date

T
IP
CR
PHOTOGRAPH
(OR SPECIFY PHOTOGRAPH NUMBERS)

US
AN
OCCUPANCY SOIL TYPE Age of the building:

Commercial Residential Educational Occupancy Type I Type II Type III Age of building (In years)
M

number
Hard Medium Soft

BASIC SCORE, MODIFIERS, AND FINAL SCORE, S

ED

Seismic zone Zone II Zone III Zone IV Zone V

Basic scores 0.75 1.20 1.80 2.70
Substantial overhang +0.41 +0.41 +0.41 +0.41
PT

Age of building : Increment per year +0.02 +0.02 +0.02 +0.02

Poor construction quality +2.20 +2.20 +2.20 +2.20
Moderate construction quality +1.10 +1.10 +1.10 +1.10
CE
Basic score modifier

Poor Maintenance condition +1.30 +1.30 +1.30 +1.30

Moderate Maintenance condition +0.65 +0.65 +0.65 +0.65
Type of soil:
AC

Soil type-I -4.86 -4.86 -4.86 -4.86

Soil type-II -2.43 -2.43 -2.43 -2.43
Soil type-III 0.00 0.00 0.00 0.00
Number of stories: Increment per storey +0.10 +0.10 +0.10 +0.10

FINAL SCORE, S=

Final score S<1 1≤S<1.5 1.5≤S<2 2≤S<2.65 2.65<S

Assigned damage grade Grade 1 Grade 2 Grade 3 Grade 4 Grade 5

RECOMMENDED ACTION FOR DIFFERENT DAMAGE GRADE: Recommended action to

be taken for the existing
Grade 1/2: Seismically safe but non-structural damage can be harmful to occupants so building:
repairing work is recommended for the building

Grade 3/4/5: Seismically not safe as structural damage will occur so further re-
evaluation and retrofitting is recommended for the building19
 ACCEPTED MANUSCRIPT
In order to check the correctness of the model, the damage grade of the building obtained from the model is cross-

checked with the observed damage grade. A detailed comparison of the damage grade of existing building obtained

by proposed model and observed damage grade from the survey is given in Table 8. From the comparison, it was

found that 64.65 % of the buildings were correctly classified by the proposed model.

Table 8. Comparison of the damage grade of existing building obtained by proposed model and observed damage

grade

Number of Degree of fitness of the % Error

T
IP
buildings model

Correctly classified 256

CR
In-correctly classified 140

US
Grade 1 as
13
Grade 2
AN
Grade 2 as
35
Grade 1
M

Grade 2 as
02
Grade 4
ED

Grade 2 as
Incorrectly classified grade

10
64.65 % 35.35%
Grade 5
PT

Grade 3 as
05
CE

Grade 1

Grade 3 as
02
AC

Grade 2

Grade 3 as
04
Grade 4

Grade 3 as
34
Grade 5

Grade 4 as
04
Grade 1

20
 ACCEPTED MANUSCRIPT
Grade 4 as
04
Grade 2

Grade 4 as
26
Grade 5

Grade 5 as
01
Grade 4

T
IP
4. A CASE STUDY

The proposed model has been applied to estimate the vulnerability of RCC buildings of ward no.6 of Uripok

CR
constituency (longitude and latitude corresponding to 93⁰92′ E and 24⁰80′ N) of Imphal city. Intensive field survey

was conducted by filling the proposed RVS questionnaire. The actual damage grade observed after the earthquake

US
was also observed during the survey. The case study area consists of 434 RCC buildings and a building experienced

damage grade ranging from damage grade 1 to damage grade 4. Out of the 434 buildings, 257 buildings experienced
AN
damage grade 1, 150 buildings experienced damage grade 2, 26 buildings experienced damage grade 3 and 1
M

building experienced damage grade 4. The location of the case study area is shown in fig.9. From the analysis, 363

RCC buildings experience damage grade 1, 61 buildings experience damage grade 2, 8 buildings experience damage
ED

grade 2 and 2 buildings experience damage grade 2. The comparison of the damage grade of existing building

obtained by proposed model and observed damage grade are done and is shown in table 9. From the comparison, it
PT

was found that 74 % of the buildings were correctly classified by the proposed model. Vulnerability maps has been

developed for the case study on GIS platform and are shown in fig.10. From the map, the damage grade distribution
CE

can be easily seen which also facilitates in easy interpretation of the output. The building that has been wrongly

classified by the proposed model can also be identified by comparing the vulnerability maps shown in fig.10. which
AC

has been produced based on the proposed model and fig.11. which is based on the actual damage experienced by the

buildings due to Manipur earthquake. Hence similar methodology can be adopted based on damage study for other

areas.

21
 ACCEPTED MANUSCRIPT

T
IP
CR
US
AN
M
ED

Fig.9. Location of case study (Ward no.6, Uripok Constituency, Imphal, Manipur)
PT

Table 9. Comparison of the damage grade of existing building obtained by proposed model and observed damage

grade
CE

Number of buildings Degree of fitness of the model

AC

Correctly classified 320

In-correctly classified 114

Grade1 as Grade 2 84
Incorrectly classified

Grade 1 as Grade 3 03 74 %
grade

Grade 2 as Grade 1 20

Grade 2 as Grade 3 06

Grade 3 as Grade 4 01

22
 ACCEPTED MANUSCRIPT

T
IP
CR
US
AN
Fig.10. Vulnerability assessment of RCC buildings of Imphal city, Manipur, India (Ward no. 6, Uripok

constituency)
M
ED
PT
CE
AC

Fig.11. Actual damage grade of Ward no. 6 of Uripok constituency, Imphal city due to M 6.7 Manipur earthquake,

2016.

23
 ACCEPTED MANUSCRIPT

5. CONCLUSION

Distribution of earthquake damage grade due to presence of different earthquake inducing features are presented and

managed by relational database management system in GIS. The developed procedure can be readily applied to any

region of India in order to assess the necessity of more detailed investigation for earthquake damage scenario

prediction. The methodology can be implemented to estimate seismic vulnerability of different types of existing

building stock in Indian Cities and further extended to investigate the impact of mitigation measures on the

T
IP
consequences of an earthquake. Applications of the procedure are illustrated with an example of Ward no.6 of

Uripok constituency of Imphal city of Manipur and the results for the vulnerability assessment are presented in

CR
colour-coded maps on GIS platform. The building database which was used to formulate the proposed model has

been classified with 64.65 % of accuracy by the proposed model while for the case study area, 74 % of the damage

US
grade of buildings were classified correctly. Hence similar methodology can be adopted based on damage study for

other areas. The main advantage of the methodology resides in the simple and robust seismic assessment
AN
methodology, compatible with Indian condition and in the capacity of providing intuitive representations of the

spatial distribution of loss results. The GIS platform facilitates generation and display of various thematic maps.
M

Since the assessment results, such as the extent of damaged buildings can be used in identifying areas with
ED

potentially heavy damage in the case of an earthquake, the disaster management plans in those areas can be suitably

strengthened. Therefore, Earthquake vulnerability methodology integrated with an open source GIS package,
PT

provides a solution for both forecasting and vulnerability estimation.

CE
AC

24
 ACCEPTED MANUSCRIPT
REFERENCES

1. Angeletti P. Comparison between Vulnerability Assessment and Damage Index, Some Results.

Proceedings of the 9th World Conference on Earthquake Engineering, Tokyo, Kyoto, Japan; 1998.p. 181–

18.

2. Arya AS. Rapid structural and non-structural assessment of school and hospital buildings in SAARC

countries. SAARC Disaster Management Centre, New Delhi; 2011.

3. BU-ITU-METU-YTU. Earthquake Master Plan for Istanbul, Bogaziçi University, Istanbul Technical

T
University, Middle East Technical University, and Yildiz Technical University”. Metropolitan Municipality

IP
of Istanbul, Planning and Construction Directorate, and Geotechnical and Earthquake Investigation

CR
Department, Turkey; 2003.

4. Calvi GM, Pinho R, Magenes G, Bommer JJ, Restrepo-Velez LF and Crowley H. Development of seismic

US
vulnerability assessment methodologies over the past 30 years. Journal of Earthquake Technology; 2006:

43 (3).p. 75-104.
AN
5. Earthquake Planning and Protection Organization (OASP). Provisions for Pre-Earthquake Vulnerability

Assessment of Public Buildings (Part A). Athens. Greece; 2000.

M

6. FEMA 154. Rapid Visual Screening of Buildings for Potential Seismic Hazards.Federal Emergency
ED

Management Agency, Washington, D.C.; 2002.

7. FEMA 178, NEHRP. Handbook for the Seismic Evaluation of Existing Buildings. Federal Emergency
PT

Management Agency, Washington, D.C.; 1992.

8. FEMA 310. Handbook for the Seismic Evaluation of buildings – A Prestandard. Federal Emergency
CE

Management Agency, Washington, D.C.; 1998.

9. Giulio C, Antonio B. Alessandro DM, Marco C and Romina S. Seismic vulnerability assessment of a
AC

monumental masonry building. Engineering Structures; 2007: 136 .p. 454-465

https://doi.org/10.1016/j.engstruct.2017.01.035.

10. Grunthal G. European Macroseismic Scale. European Seismological Commission (ESC); 1998.

11. IBC-2006. International Building Code 2006 Edition. International Code Council; 2006.

12. IS 1893: 2002. Criteria for earthquake resistance design of structure part 1 general provisions and building.

Indian Standard. Bureau of Indian Standards, New Delhi; 2002.

25
 ACCEPTED MANUSCRIPT
13. IS 1893: 2016. Criteria for earthquake resistance design of structure part I general provisions and building.

Indian Standard. Bureau of Indian Standards, New Delhi; 2016.

14. Jain SK, Mitra K and Shah M. A proposed rapid visual screening procedure for seismic evaluation of RC

frames buildings in India. Earthquake Spectra; 2010, 26 (3).p. 709-729.

15. Jason MI, Paulo BL, Joao L, Saahil C and Edlyn C. Using simplified indices to forecast the seismic

vulnerability of New Zealand unreinforced masonry churches. Australian Earthquake Engineering Society

Conference, Gold Coast, Qld; 2012.

T
IP
16. JPDPA. Seismic Evaluation and Retrofit. The Japan Building Disaster Prevention Association, Japan;

2001.

CR
17. Lang DH, Verbicaro MI, Wong DD, Gutierrez M. Structural and nonstructural seismic vulnerability of

schools and hospitals in Central America. Final Report, NORSAR, Kjeller, Norway; 2009.

US
18. Lourenco PB and Roque JA. Simplified indexes for the seismic vulnerability of ancient masonry buildings.

Construction and building materials; 2006, 20 (4).p. 200-208,

AN
https://doi.org/10.1016/j.conbuildmat.2005.08.027

19. NIDM. Seismic Vulnerability Assessment of Building Types in India, Technical Do cument (Tech-
M

Doc) on Seismic Vulnerability Assessment Methods fo r Buildings”, National Disaster

ED

Management Authority Government of India; 2013.

20. Naili M and Algeria ZM. Seismic Vulnerability Assessment of the Old Old Casbah of Mostaganem City
PT

existing – Algeria”, 15WCEE, Lisboa; 2012.

21. Ningthoujam MC and Nanda RP. Rapid Visual Screening Procedure of Existing Building Based on
CE

Statistical Analysis. International Journal of Disaster Risk Reduction; 2018,

https://doi.org/10.1016/j.ijdrr.2018.01.033 .
AC

22. NRCC. Manual for Screening of Buildings for Seismic Investigation. Institute for Research in

Construction, National Research Council Canada, Ottawa; 1993.

23. NZSEE. Assessment and Improvement of the Structural Performance of Buildings in Earthquake.

Recommendations of a NZSEE Study Group on Earthquake Risk Buildings, New Zealand; 2006.

24. Nanda RP and Majhi DR. Rapid Seismic Vulnerability Assessment of Building Stocks for Developing

Countries. KSCE Journal of Civil Engineering; 2004, 18(7).p. 2218-2226.

26
 ACCEPTED MANUSCRIPT
25. Pallav K, Raghukanth STG., Konjengbam DS. Estimation of Seismic Site Coefficient and Seismic

Microzonation of Imphal City, India, Using the Probabilistic Approach. Acta Geophysica; 2015, 63(5).p.

1339-1367.

26. PAHO/WHO. Hospital safety index, guide for evaluators. Pan American Health Organization and World

Health Organization, Washington D C.; 2008.

27. Pankaj A. and Manish S. Earthquake resistant design of structure. Prentice Hall of India Private Limited,

New Delhi; 200.

T
IP
28. Perrone D, Aiello MA, Pecce M and Rossi F. Rapid visual screening for seismic evaluation of RC hospital

buildings. Structures; 2015.p. 57-70.

CR
29. Sucuoglu H, Yazgan U, Yakut A. A screening procedure for seismic risk assessment in urban building

stocks. Earthq Spectra; 2007.p. 441–58.

US
30. Wallace NM, Miller TH. Seismic screening of public facilities in Oregon's western countries”, Pract Period

Struct Des Constr; 2008.p. 189–97.

AN
31. Yadollahi M, Adnan A, and Zin RM. Seismic vulnerability functional method for rapid visual screening of

existing buildings”, Archives of civil engineering; 2012, 8 (3).p. 363-377.

M
ED
PT
CE
AC

27