ASL Full Paper (Rini Mulyani)

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

SEISMIC RISK ASSESSMENT OF BUILDINGS IN PADANG

CITY, WEST SUMATRA


Rini Mulyani
Department of Civil Engineering, Bung Hatta University, Padang, Indonesia
Email: rinimulyani@bunghatta.ac.id

in 2011, have emphasized the significance of seismic


ABSTRACT The mega magnitude earthquakes risk assessment. It is well known that casualties and
such as the Mw 9.1 Indian Ocean Earthquake in 2004 fatalities during an earthquake are mainly associated
and the Mw 9.0 Japan earthquake in 2011 have with damaged and collapsed buildings. On 30
reiterated the significance of seismic risk assessment. September 2009, a major earthquake registering a
A key element for reliable seismic risk assessment is magnitude Mw 7.6 occurred off the coast of Padang
appropriate hazard identification, vulnerability in West Sumatra. This was followed by another
assessment and exposure in the investigated area. earthquake occurring inland and measuring Mw 6.6.
This paper presents seismic risk assessment of The main earthquake had a devastating effect on
buildings in Padang City, West Sumatra, Indonesia. many of the buildings and affected infrastructure and
The risk assessment utilizes a seismic risk tool so- communities. The Indonesian government estimated
called Earthquake Risk Assessment Framework the economic consequences of the earthquake to be
(ERA Framework). The framework incorporates a Rp. 21.58 trillion (approximately US $ 2.4 billion),
stochastic technique using historical and and BNPB (2009) reported 1,177 fatalities and
instrumental earthquake catalogues. The proposed approximately 3,000 people injured (1,688 minor and
method is designed to be relatively simple to suit the 1,214 major injuries).
immediate needs of developing countries, often
lacking of human resources and detailed
seismological data. The risk model developed in this
study is verified against building damage data
obtained from the Mw 7.6 of Padang earthquake. It is
found that the loss estimated by the risk model is
about 4.76% lower than that of the damage statistic
of the event. The results are reasonable considering
that the risk model does not take liquefaction into
account, which occurred in the event. The earthquake
risk model in this study estimates that the building
loss could have been reduced from £1.122 billion to
£564 million if the buildings were built according the
Indonesian earthquake resistant building standard;
thus lowering the loss by about 50%.

Key Words: earthquake; risk; vulnerability;


building; structures. Figure 1. Percentage of physical damage and
economic losses due to several types of hazards in
Indonesia (BNPB, 2015)

Based on the data obtained from BNPB (2015),


1. Introduction earthquakes contributes to 53.4% and 42,7% of total
physical damages and economic losses due to natural
Series of mega magnitude earthquakes occurring
hazards, respectively. The percentage is quite high
in the past decade, such as the Mw 9.1 great Sumatra
compared with those of other hazards such as floods,
earthquake in 2004 and the Mw 9.0 Japan earthquake
volcanic eruptions, landslides and tsunami as shown
in Figure 1. The physical damages include the
damage of buildings and infrastructures, which often
brings casualty and fatality. Therefore, earthquake
resistance structures are very important to minimize
seismic risk in the prone areas and the seismic
vulnerability of structures has to be determined for
risk identification. Risk assessment is particularly
required to determine the vulnerability of buildings in
the prone region with high population such as Padang
City, West Sumatra.

2. Method Figure 3. Correlation between ground


acceleration and damage state for existing building
Earthquake risk assessment of buildings requires stock in Padang (GESI, 2001)
several stages including earthquake hazard analysis,
the vulnerability analysis of buildings, building Table 1 shows the estimated mean damage ratio
inventory as well as the risk estimation of buildings (MDR) of existing RCI buildings in Padang subjected
in the investigated area (Figure 2). In this study, the to Mw 7.6 earthquake occurring in the area. The event
vulnerability of buildings in Padang City, West produced a MDR of 18% in the area.
Sumatra, is investigated using an Earthquake Risk
Assessment Framework (ERA Framework), Table 1. The estimation of mean damage ratio
previously developed at the University of Sheffield for RCI buildings in Padang based on the damage
(Khan, 2011; Kythreoti, 2002; Mulyani, 2013). The data of the Mw 7.6 earthquake
earthquake risk model is then validated against the
Central
damage data of the Mw 7.6 earthquake of Padang, Prob. of DS
Damage Damage Ratio
which is discussed in the following sections. (PGA 0.26g
State (DS) (CDR) % -
at Padang city)
HAZUS (1999)
Seismic Hazard None 0 0.606
Assessment Light 20 0.173
Moderate 55 0.125
Vulnerability Assessment Extensive 80 0.096
MDR = 0.180

Risk Assessment

4. Verification of The Seismic Risk Model


Figure 2. ERA Framework scheme with A Real Event

3. The Selection of Vulnerability Functions To validate the seismic risk model developed in
for the Building Stock in Padang City this study, the loss due to the Mw 7.6 Padang
earthquake of 30 September 2009 is estimated and
To determine the vulnerability of buildings due to compared with the existing loss data as shown in
earthquake, vulnerability curves for each type of Table 2.
structures in Padang are required. The vulnerability The earthquake occurred at the intraplate of the
curves correlate the expected ground motions with subduction zone; and thus the focal depth was
the mean damage ratio of the existing building stock relatively deep at about 81 km. The epicenter of the
in the region. For earthquake risk, this study adopts event was around 60 km from the city of Padang. The
vulnerability curves proposed by GESI (2001), since predicted peak ground acceleration was estimated at
GESI provides flexibility in selecting the class of 0.27g, which was consistent with the strong ground
construction in terms of design, quality and material. motion records obtained from BMKG, the Indonesian
In addition, the GESI project includes data from meteorology and geophysics agency (Wilkinson et
many countries including Indonesia. The chosen al., 2012)
vulnerability curves are shown in Figure 3. GESI
vulnerability curves comprise 4 damage states, which
are defined in Table 1.
Table 2. Number of damaged buildings due to not significant given the approximate nature of both
the Mw 7.6 Padang earthquake of 30 September methods. The estimated ratio between the loss and the
2009 (BNPB, 2009) value of buildings is shown in Figure 5. Based on
Figure 5, it is observed that the residential areas are
Number of damaged buildings mostly affected by the earthquake. This finding is
Building
Function Lightly Moderately Heavily consistent with the BNPB data.
Damaged Damaged Damaged
1. Residential 37587 38485 40406
2. Schools 1606 1038 903
3. Hospitals 9 10 2
4. Offices 59 19 14
5. Religious 238 211 169
6. Commercials 5 10 5
Legend
Total Damage
(% of All Buildings)
0-3

4 - 13

14 - 23

24 - 30

31 - 40

±
41 - 50

51 - 70

71 - 100

Legend 0 2.5 5 10
PGA Km
PGA (g)
0.12 - 0.14
0.15 - 0.17 Figure 5. The predicted mean damage ratio
0.18 - 0.21 (MDR) for the buildings in Padang due to the Mw
0.22 - 0.24
7.6 earthquake based on GESI vulnerability curves
0.25 - 0.27

0 3 6 12 MapAction (2009) released a damage distribution


Km
map for Padang due to the earthquake. The map was
developed based on the data from Indonesian
Figure 4. The predicted PGA of the Mw 7.6 Statistical Agency, OCHA and the Indonesian
earthquake in Padang Ministry of Public Works. Considerable difference is
observed between the damage distribution in Figure
The Indonesian government estimated that the 5 and that obtained by MapAction (2009),
earthquake caused an economic loss of about particularly at the eastern and southern part of the
Rp. 21.58 trillion (approximately £ 1.4 billion) in city. MapAction (2009) estimated a damage level of
West Sumatra. The damaged data for Padang city was 13-100% in the areas. However, this study finds that
reported by the Indonesian mitigation agency most of the eastern and southern areas are
(BNPB, 2009). Based on the data, the building loss uninhabited (or have a very limited population), and
for the area can be assessed using the building therefore, they are categorised as “NA” in the
inventory information available for the area. An building inventory database. A comparable damage
estimated loss of £1,179 million is obtained for the distribution is observed throughout the main part of
region. the city with the percentage of damage about 3-30%.
However, MapAction (2009) shows higher values for
some areas near the coast. The field survey after the
earthquake revealed that liquefaction occurred near
3. Results and Discussion the coastal region, which exacerbated the level of
damage in the area. The liquefaction effect is not
Using the seismic hazard and risk model taken into account in this study, and therefore, the
developed in this study, it is estimated that the level of damage obtained is likely to be
damages in Padang for the event reach £ 1,122 underestimated.
million. This value is only 4.76% lower than that
obtained from the damage statistic. This difference is
Figure 6 shows the estimated loss per m2 of Mulyani, Alarcon, & Wilkinson, 2010; Wilkinson,
buildings in Padang subjected to the event. It is Alarcon, Mulyani, Chian, & Whittle, 2009;
observed that the buildings in the densely populated Wilkinson, Alarcon, Mulyani, Whittle, & Chian,
areas are more susceptible to the earthquake (red and 2012). Poor quality materials, lack of seismic
orange zones in Figure 6. An average loss of about detailing and poor seismic design are among the main
GBP 26-42 per square meter is predicted for most problems identified in the damaged structures.
parts of the city. The risk model in this study Although the Indonesian seismic design guideline for
estimates that about 37% of UBM structures are buildings was introduced since 1983, the use of
damaged during the event, something expected seismic design in building practices was not evident.
considering the poor seismic performance of the As a consequence, the structures were vulnerable,
UBM structures. However, the RCI buildings have and therefore, the seismic risk was considerably high.
the highest loss per square meter due to their higher
construction cost. The comparison of the damage and
building loss/m2 for every type of structure is shown
in Figure 7.

Figure 8. The estimated average risk in the area


due to the Mw 7.6 earthquake
Figure 6. The estimated average loss/m2 for 4. Conclusions
Padang City due to the Mw 7.6 earthquake
▪ The risk model developed in this study is verified
against building damage data obtained from the
Mw 7.6 of Padang. It is found that the loss
estimated by the risk model is about 4.76% lower
than that of the damage statistic. The results are
reasonable considering that the risk model does not
take liquefaction into account.
▪ Earthquake risk assessments are conducted for
Padang city for four building categories. It is found
that the existing unreinforced brick masonry
buildings are the most vulnerable, followed with
Figure 7. The predicted damage and unit loss for the confined brick masonry, reinforced concrete
each type of structure in Padang structures with masonry infill and steel structures.
The earthquake risk decreases considerably (about
Figure 8 shows the average loss of buildings per
80%), if the buildings are designed according to
unit grid. It appears that the worse affected regions
modern seismic design standard.
are located near the city centre of Padang with total
economic consequences up to £138 million per unit ▪ The great economic loss associated with the Mw
grid. This outcome is comparable with the findings of 7.6 earthquake in Padang could have been
an EEFIT field survey in Padang in which the author minimised, if the vulnerable structures in the area
participated as a team member (Chian, Whittle, were designed to satisfy the seismic design
criteria. The earthquake risk model in this study
estimates that the building loss could have been
reduced from £1.122 billion to £564 million; thus
lowering the loss by about 50%. As a consequence,
the fatality reduction would have been much
higher. The predicted loss per unit grid for
seismically designed building stock due to the Mw
7.6 event.

5. Acknowledgements
The authors would like to acknowledge the
Indonesian Ministry of Higher Education and Bung
Hatta University for the financial support received in
this project.

6. References
BNPB. (2009). Laporan Harian Pusdalops BNPB Minggu
19 Oktober 2009. Retrieved from
BNPB. (2015). inaRISK, Bagaimana Risiko Wilayah Kita?
Retrieved from http://inarisk.bnpb.go.id/
Chian, S. C., Whittle, J., Mulyani, R., Alarcon, J. E., &
Wilkinson, S. M. (2010). Post Earthquake Field
Investigation of the Mw 7.6 Padang Earthquake
of 30th September 2009. Paper presented at the
14ECEE, Macedonia.
GESI. (2001). Final Report: Global Earthquake Safety
Initiative (GESI) Pilot Project. Retrieved from
Khan, S. A. (2011). An Earthquake Risk Assessment
Framework for Developing Countries: Pakistan
A Case Study. Retrieved from Sheffield:
Kythreoti, S. (2002). Earthquake Risk Assessment and
Management. Case Study: Cyprus. Retrieved
from Sheffield:
MapAction. (2009). The damage level of houses due to the
earthquake in Padang City. In MA077-IDN-Sit-
150k_PadangCityDamage-21Oct2009-A3-v02-
300dpi.jpg (Ed.), (Vol. 3.5 Mb). Saunderton,
Buckinghamshire, UK: MapAction.
Mulyani, R. (2013). Extended Framework for Earthquake
and Tsunami Risk Assessment: Padang City A
Case Study. (PhD), The University of Sheffield,
United Kingdom.
Wilkinson, S. M., Alarcon, J. E., Mulyani, R., Chian, D., &
Whittle, J. (2009). The Padang Sumatra-
Indonesia Earthquake of 30 September 2009, A
Field Report by EEFIT. Retrieved from London:
Wilkinson, S. M., Alarcon, J. E., Mulyani, R., Whittle, J.,
& Chian, S. C. (2012). Observations of damage
to buildings from M w 7.6 Padang earthquake of
30 September 2009. Natural Hazards, 63(2),
521-547.

You might also like