You are on page 1of 8

6th International Conference on Seismology and Earthquake Engineering

LESSONS FROM THE FEBRUARY 2011 M6.3 CHRISTCHURCH EARTHQUAKE


G. A. MacRae1, R. P. Dhakal1
1

Department of Civil and Natural Resources Engineering, University of Canterbury, Christchurch, New Zealand (gregory.macrae@canterbury.ac.nz) (rajesh.dhakal@canterbury.ac.nz)

ABSTRACT
This paper describes NZ schools, the decision making activities that are required to retrofit schools, and recent decisions regarding the tertiary school, the University of Canterbury, after recent earthquakes.

Keywords: Earthquake Sequence, Lessons Learned 1. THE SHAKING


In September 2010, a M7.1 earthquake focussed toward Christchurch caused considerable damage and loss to Christchurch, north Christchurch and its surrounds. Effects were severe on some unreinforced masonry structures, some houses in the region of high shaking, on structures on liquefiable soil over a large region. The shaking and the effects was summarized by MacRae (2011). The 22 February 2011 earthquake which occurred at 12:55pm was M6.3 with a dip of 65o and the shaking in the south-eastern suburbs, and in Christchurch itself was much greater than from the M7.1 earthquake as shown in Figure 1. Based on discussions with Paul Somerille and John Berrill, there are a number of reasons for this: 1) The earthquake hypocentre and rupture occurred much closer to Christchurch than did the 2010 M7.1 earthquake. 2) There was a near fault directivity effect. The shaking was focussed up toward Christchurch and Lyttleton. The steeply dipping surface generated shear waves with a strong vertical component. 3) There was a hill crest effect shown in Figure 2 causing focussing of shaking near the crest. It focuses in places and cancels out in other places giving variable shaking. Because the hill is hard rock, here is little opportunity for attenuation of the high frequency components.

SEE6 / 1 / IIEES

Figure 1. Christchurch Botanical Gardens Records (From Brendon Bradley, U. Canterbury, New Zealand)

Figure 2. Ridge Effect (Focussing of rays)

Figure 3. Hanging Wall/Wedge Effect (From Berrill 2011)

There was a trampoline effect caused by the strong vertical shaking moving down and up again throwing the soil, structure or contents into the air. 2) There was a basin effect near Christchurch as the waves moved between the different layers of soil. According to Berrill (2011), the same energy causes greater deformations in the softer basin soils than in the stiffer soils. In Christchurch, the main layers are: The full depth of the alluvial gravel deposits of 700-1000m, which has a horizontal period of resonance of about 2.5-3.0s. The upper 10-30m of very soft soil, laid down loosely in the rapid sea-level rise following the last glaciation, about 10,000 years ago. The resonant period of this layer varies markedly depending on its thickness and density, but is in the range of 0.5-1.0s. 3) There was a hanging wall effect, which amplifies the high frequency shaking. This is illustrated by the sketch from Berrill (2011) and it affected the structures on the hills. 4) The rupture was subsurface. These ruptures which dont break the surface generally have significantly stronger shaking affecting most structures than those that do break the surface according to observations from previous earthquakes (Somerville 2003). It is not clear whether or not there was a basin edge effect which focuses longer period waves from a number of angles to specific area just inside the edge of the basin.

1)

SEE6 / 2 / IIEES

There was considerable discussion about whether or not the M6.3 shaking was an earthquake in its own right, or whether it was an aftershock which was part of the earthquake sequence. According to GNS Science, an aftershock rupture must be both close in terms of proximity and time to the main shock rupture. In addition it must be of lower magnitude than the main shock rupture. The M6.3 rupture satisfied all three criteria, as it occurred near the end of the M7.1 rupture fault, and it was close in time being about 5 months after the main shock, which is less than the 2-3 years expected for a M7.1 rupture (Cousins, 2011). The location of the 2011 M6.3 fault is shown by the yellow line in Figure 4 while that of the 2010 M7.1 fault is shown by the red line. It should be noted that the fault lines do not meet up, so investigations have been undertaken by geologists and seismologists about the possibility of a further aftershock there, or from a fault beneath Christchurch.

Christchurch

Figure 4. Approximate Locations of Fauls and Aftershocks (Courtesy GNS Science)

The response spectra at some city CBD (central business district) sites is given in Figure 5. The dashed black line is the 500 year design spectra, while the solid black line is 1.8 times this value. This is nominally considered to be the maximum credible earthquake (MCE). It may be seen that the CBD shaking was often considerably greater than both of these curves being often significantly more than 2 times the design level. In an design level earthquake we generally expect structures to stand up and life to be preserved, but in a MCE event, we generally hope that things will be OK given the levels of shaking that were experienced. One of the largest levels of shaking occurred at a school in a valley beside the port hills as shown in Figure 6. The two horizontal components of peak acceleration of 1.68 and 1.27g are high. According to Berrill (2011), for about 20 years, the strongest horizontal peak acceleration recorded was the 1.25g from Pacoima Dam, in the epicentral region of the 1971, M6.4 San Fernando, California Earthquake in northern Los Angeles. It indicates that the shaking occurring was very strong. The vertical accelerations were very high in the positive direction but less than 1g in the negative direction. The peak acceleration in this direction is likely to occur when the instrument is at the peak upward displacement. It indicates that the equipment and the slab (and building) it was connected to probably lifted up in the air and come and bounced on the ground causing impact with the soil. It is like a trampoline effect with a hard trampoline. On the nearby hills to this school, some major rocks rolled down the hill onto peoples houses. There was also some damage at the school itself, with masonry walls in 2 buildings which had moved, and a unreinforced masonry heating building and chimney that had come down. Apart from that the school buildings looked to be in very good shape.
SEE6 / 3 / IIEES

1.8

1.6

NZS1170 2500-yr Class D

1.4

NZS1170 500-yr Class D Deep or Soft Soil CHHC_MaxH_FEB

1.2
CCCC_MaxH_FEB

CBGS_MaxH_FEB

REHS_MaxH_FEB

0.8
GM_Larger_FEB

0.6

0.4

0.2

0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Period, T (s) Figure 5. 22 February 2011 M6.3 Response Spectra (Courtesy Geonet - GNS Science)

2. DAMAGE
The shaking in this earthquake caused much of the same types of damage observed in the earthquake, as well as 2010 earthquake, including liquefaction damage to houses, chimney damage to houses, and collapse of unreinforced masonry structures but it also caused some other types of damage: 1) Houses on the hills spurs were sometimes shaken to bits. That is, the bricks and veneers came off, and the internal linings were so much cracked that they provided no strength. Only the framing, which could sustain significantly greater displacements than the rest of the interenal and external cladding remained relatively undamaged. 2) Houses on the side of hills often had some foundation work during construction. Any foundations which were not robust suffered sliding, thereby distorting the house. 3) Many houses at the bottoms of slopes were bombarded by rocks which had been resting above them for many years. 4) Cliffs near houses often gave way, sometimes throwing large boulders, or rocks onto buildings below. A number of people died as a result of this. Also, the cracks at the top of the cliffs meant than many houses were considered too risky to be inhabited. 5) As land moved laterally on the hillsides, it often burst underground pipes. They water/liquid flowing from these broken pipes as well as rain makes the likelyhood of more sliding likely. 6) The scale of damage to unreinforced masonry structures, like Christchurch Cathedral, and other heritage structures. Some of these structures collapsed onto cars or busses killing people. Older URM structures retrofitted to 33% of code which were still brittle behaved poorly under the levels of shaking experienced. 7) Industrial buildings fared relatively well, but in a number of tilt-slab structures, the steel connections (particularly the welded ones) performed poorly. 8) Most of the central business district considered of reinforced concrete buildings. From about 1975, design for ductility was considered in all buildings, from 1985 capacity design considerations mean that ductility occurred in the ductile elements only. From 1992, the extra detailing was required to ensure that gravity columns were designed to undergo the lateral displacements expected in the seismic frames and that stairs would slide and not carry significant interstorey shear forces. In general well designed and detailed commercial buildings had no major structural damage if they were on a good foundation. Some had significant beam elongation and crack opening in the plastic hinges. While many structures are designed to have hinges at the column bases, this seldom occurred due to the flexibility of the foundations. Unfortunately some structures were not on a sufficient foundation, and while there was no structural damage, they had

SEE6 / 4 / IIEES

Figure 5. Heathcote Valley School Damage (from Berrill (2011) using Geonet data)

different column settlements or the structures were out-of-plumb after the earthquake. Some buildings suffered partial collapse (like the PGC Building) or total collapse (like the CTV Building). Many people died especially in the CTV building. Other buildings behaved well, but lost their stairs, like the Forsyth Barr building. It should be noted that

SEE6 / 5 / IIEES

it was known that some buildings in use were vulnerable to earthquake before these events. This include not only older URM structures but some of the more modern concrete structures too. 9) Many reinforced block masonry buildings behaved very well, but those not reinforced did not behave well. 10) Steel buildings generally behaved as, or better than, expected given the level of shaking. Some examples of damage in steel buildings were damage to non-sliding stairs in a 1986 building, and yielding of EBF shear links (as was expected), and foundation settlement causing relative displacement of the floor. There was one very interesting example of a parking structure with fractures to the flange of the beam containing the shear link. This was a result of poor construction. Also, in this building there was damage to reinforced concrete gravity columns which had not been detailed to undergo the frame displacements. 11) Again, non-structural damage was in many/most structures in the affected region. 12) The total death toll from this earthquake was just under 200 people. Because Christchuch is a small city (400,000 people, many people know the people who died, or knew someone who was a friend of one of these people).

3. POST-EARTHQUAKE ISSUES
After the earthquake Urban Search and Rescue (USAR) from many countries assisted in finding people and removal of bodies. Over 10,000 homes were uninhabitable. These were mainly houses around the hills or in highly liquefied suburbs such as Bexley. The CBD, in which more than 50,000 people work, and many more shop every day, was closed and put under military lockdown. This meant that looting was almost totally eliminated. Volunteer, as well as city, engineers performed by rapid damage assessments (both Level 1-exterior and Level 2interior). Many damaged heritage structures were pulled down. The city has been opened in stages. Some structures which are beneath damaged structures, like the Grand Chancellor, structure, have been out-of bounds until these structures are stabilized. A lot of care has been taken to ensure that people go back into the CBD safely. Even at the time of writing this paper, more than 3 months after the earthquake, there is still a RED zone into which ordinary people are not permitted. This has approach been commended and it has been criticized. People are safe and property is not stolen. On the other side, while people cannot work, many are moving away to other centres and starting a new life. Every day the city is closed, less people will return. The bright side in the longer term is that the city is very well insured. Even when one of the local insurance companies, AMI, ran out of money after 2 earthquakes, the government has agreed to provide payments. It is expected that over $15 billion will flow into the region over the next few years and this will enhance jobs. Immediately after the M6.3 earthquake, insurance companies stopped providing loans on houses. Without insurance, banks would not lend. Without money to borrow, most people could not buy. Insurance is becoming more available, but the companies generally want houses inspected by an engineer first to ensure that new insurees do not claim on previous earthquake damage in a future aftershock. There are issues about whether structures will be rebuilt on some liquefaction areas and near steep hill areas. While it is possible to build almost anywhere, it is likely that the some areas will not be rebuilt. Some of this is likely to be decreed by the government. In other cases, insurance companies may have very high rates making it uneconomical to rebuild. One of the things that engineers wanted to know after the earthquake is whether or not the seismic zone for Christchurch has changed. Several groups have worked together to provide
SEE6 / 6 / IIEES

information on this. The pre-earthquake zone factor was 0.22. This corresponds to the spectral acceleration at a period of 0.0s in Figure 5. Studies of previous earthquakes, such as the M7.1 1968 Inangahua earthquake have indicated that significant aftershocks should occur for several tens of years after the main shock. Therefore, the risk of similar shaking in Christchurch as a result of the M6.3 event was considered to be 6% in 50 years, whereas before it was considered to be less than 1/10 of that. GNS recommended that an appropriate value for design should therefore be a zone factor of 0.30 (McVerry 2011, personal communication) for structures with a fundamental period up to 1.5s. Several other considerations were made by representatives of the NZ earthquake loading code. These include the very good (no structural damage) of the modern structures, whereas damage should have been expected under the very high levels of shaking. Also, the fact that retrofit of older URM structures after the M7.1 2010 earthquake was increased from 33% to 67% of the code. For diligent owners who had retrofitted to 100% of the code, 67% of the new code values would still be less than what these have been retrofitted to, so they will not need to be retrofitted again. Engineers were happy to have a value to design to, regardless of its background and the value of 0.3s was adopted. For structures with a period greater than 1.5s, a special study is required. It seems to the authors that something was not considered as it was not explicitly discussed at the meetings related to this. That is, it is assumed that all future structures designed to the current code specifications will be the same as those in the past. Also, reasons for the good behaviour of modern structures has not been fully understood. It is the view of the authors that this may be partially due to the soft soils providing some isolation of the structures, but this is taken into account to some extent by the earthquake records. The other effect is that of the partitions, and facades with provide extra stiffness to the structures. As research progresses and non-structural elements are either explicitly excluded from the contributing to the lateral stiffness through special gaps, or are included directly as part of the lateral force resisting system by appropriate design, the total strength of future structural systems is likely to decrease. Open structures, like parking structures may therefore most likely represent the behaviour of future buildings. Whether or not this should affect the code force is a matter which depends on the desired performance objectives under specified levels of shaking. It is anticipated that they may low-damage technologies being developed in NZ will assist in limiting the loss in the rebuild (E.g. MacRae et al. 2009). It should be noted that no recommendations have been made to change the level of vertical acceleration considered in design. A number of decisions need to be made for structures with different types of damage. This includes masonry structure damage, cracks in plastic hinge zones, and fracture of EBF links. The decisions for different structures are (1) are they OK as is? (2) Should they be repaired? If so, how? (3) Should the part or the whole structure be removed? Some guidance is available in documents such as FEMA306, however the final decision involves many parameters and the cost is a significant one. Engineers are still struggling with these issues which they have not had to deal with previously. There is a desire amongst many engineers to introduce a tagging system for each building to inform occupants, and future building owners about the structure susceptibility to damage. It is believed by many that this may be a more effective method of causing change than trying to legislate it with building owners who claim that they do not have money for retrofit. Time will tell if this is effective or not. Amenities such as power, water, and sewage have taken time to restore in the liquefaction affected areas, affecting the lives of many, as well as in the CBD. Continued aftershocks have made repair difficult after the M6.3 event, and chimneys and other brittle elements are still falling. With relation to school buildings, they experienced the same issues as regular buildings. In general the damage to school buildings was slight as not many had significant reinforced masonry, or many stories. However, liquefaction, lack of amenities, as well as limited damage a

SEE6 / 7 / IIEES

few structures meant that there were many student refugees. These students either moved with their families, or were bussed to different schools, until their home schools were opened.

4. CONCLUSIONS
This paper described some lessons from the 2011 Christchurch Earthquake including: a) The fact that aftershocks may have more severe effects than main shocks if a number of conditions, such as proximity, near fault focussing, etc, are satisfied. b) Damage was really what could be expected (or less) given our current understanding of earthquake engineering. c) A number of issues need to be addressed after an earthquake. Some of these were described.

5. REFERENCES
1. Berrill J. (2011) Some Aspects of the M6.3 February 22nd Earthquake, Unpublished document, Civil Engineering, University of Canterbury. Referenced in Friday news, 29 April 2011. 2. Cousins J. (2011), GNS Science, Email correspondence following the Feb 2011 earthquake. 3. MacRae G. A., Clifton G. C., Mackinven H., Mago N., Butterworth J. and Pampanin S., The Sliding Hinge Joint Moment Connection, New Zealand Society for Earthquake Engineering, December 2010. 4. MacRae G. A. The 2010 Canterbury Earthquake, Sixth International Conference of Seismology and Earthquake Engineering, 16-18 May 2011 Tehran, Iran. 5. Somerville, P.G. (2003). Magnitude scaling of the near fault rupture directivity pulse. Physics of the Earth and Planetary Interiors 137, 201-212.

SEE6 / 8 / IIEES

You might also like