Statistics House Investigation
The performance of Wellington’s Statistics House during the Kāikoura earthquake focused on the design, construction and
land influences on the performance of the building.
Building and Construction Minister Nick Smith released the Independent Panel’s investigation into the performance of
Statistics House on 31 March 2017. It found a combination of four factors contributed to the partial failure of lower
floor segments.
Two of the factors – the flexible frames and style of floor construction – combined with significant shaking for up to
120 seconds, and localised amplification of the shaking, to compromise the support of the lower pre-cast concrete floor
units on their framing.
The Kāikoura earthquake was a highly complex earthquake and is challenging existing seismic hazard models. More than
twelve separate faults broke during the quake, including some that had not previously been mapped, and the frequency of
shaking that arrived in Wellington 240km from the epicentre most affected mid-rise Wellington buildings.
The combination of factors that led to the partial collapse of floor units in Statistics House was not anticipated by
the design standards in place, when it was built in 2005. Design standards are regularly updated to incorporate new
learnings. The soon to be published amended Concrete Structures Standard will contain specific provisions that
acknowledge the impact of beam elongation on frame buildings and how seating of pre-cast concrete floor units needs to
be enhanced to capture current learnings.
Questions and Answers can be found below.
MBIE, the Institution of Professional Engineers New Zealand (IPENZ) and IPENZ technical societies have produced
information for owners and building professionals responsible for assessing and designing multi-storey concrete moment
resisting frame buildings with precast concrete floor systems that may be vulnerable to loss of floor support during an
earthquake.
Framed buildings with precast concrete floor systems is available on the Building Performance website.
Recommendations
The Panel’s findings have made four recommendations:
Recommendation One:
• MBIE has written to other councils – Hutt City, Upper Hutt and Porirua - about how the 14 November 2016 earthquake may
also have affected similar buildings to Statistics House in their areas. This will enable them to contact the building
owners so they can undertake damage assessments similar to those performed in the targeted building damage assessment
programme undertaken by the Wellington City Council.
• The New Zealand Structural Engineering Society and the New Zealand Society for Earthquake Engineering have issued
guidance on how to assess damage to buildings which share the same design characteristics as Statistics House. This will
ensure consistency amongst the assessments. It is up to engineers/owners and their insurers to determine how best to
repair buildings.
Recommendation Two:
• MBIE will be working closely with IPENZ and its technical societies to ensure the engineering profession is aware of
the design issues with buildings similar to Statistics House.
• MBIE, in collaboration with IPENZ technical societies, will develop comprehensive guidance to assist with assessing
these types of buildings.
Recommendation Three:
• MBIE has contacted Standards NZ, requesting that the Earthquake Actions Standard NZS1170.5 and Concrete Structures
Standard NZS3101 standards committee meet to consider the findings from Statistics House investigation.
• An amendment to Standard NZS1170.5 was published in September 2016 and the committee will need to consider what
immediate modifications to the standard may be necessary to address findings from the Statistics House investigation
• Standard NZS3101 is currently under review and Standards NZ is working towards getting this updated and released
quickly and this can incorporate learnings from the Statistics House investigation.
Recommendation Four:
• There is limited understanding of the impact of basin-edge effects and duration of earthquakes for New Zealand. MBIE
will be commissioning further research to better understand basin-edge effects and the effects of duration. The research
will involve international collaboration with countries that have cities affected by basin-edge effects such as Kobe and
Los Angeles.
Questions & Answers
What did the Statistics House investigation look at?
The aim of the investigation into Statistics House was to understand what caused that building’s performance in the
Kāikoura earthquake and any implications for New Zealand’s building regulatory system, that is; it was a technical
investigation of the Design, the Construction, and the Land influences on the performance of Statistics House.
It did not look at any liabilities regarding the part of the building which failed, or the future of the building or
other buildings.
Why did part of the flooring on the lower levels of Statistics House fall down?
The partial-collapse of three pre-cast floor units at Statistics House affected only a very small part of the overall
floor area. It was caused by a combination of factors:
• the high level of flexibility in the building’s design;
• the style of floor construction;
• the magnitude and duration (120 seconds) of the earthquake; and
• the amplification of ground shaking by the geological basin beneath Wellington (basin-edge effects).
This combination of factors meant the building moved considerably and over many cycles of shaking. The building frame
was designed to flex during earthquakes but this caused the end frames of the building to stretch. The less flexible
ends of the floor units were also affected by this, resulting in some breakages to the ends of the floor units. The
combination of the frames stretching and the floors shortening compromised the support of three pre-cast floor units and
resulted in them falling onto lower floors.
Other nearby buildings, which performed better than Statistics House, did not have all of the same features as
Statistics House and therefore were not as vulnerable to the characteristics of this earthquake.
What did the investigation find?
With regard to the first two factors under investigation; the Design and Construction of Statistics House, the
investigation found the building was generally designed and constructed to the requirements of the design standards in
place in 2004/05.
With regard to the third factor; the Land Influences, the fact the building is on reclaimed land was not a significant
factor. The ground shaking from the earthquake was amplified because seismic waves reflected/refracted off of the edges
of the geological basin beneath Wellington. This phenomena is referred to as a basin-edge effect. At present there is
not sufficient international or national research to conclusively say where a basin-edge effect may occur.
Was the nature of the Kāikoura earthquake unusual?
The Kāikoura earthquake is challenging the understanding of how earthquakes are expected to behave. According to the
April 2017 write up in the Science Journal it is one of the most complex earthquakes ever studied globally. An important
issue is the way the quake was able to rupture so far along its path (over 200km from its epicentre) on so many
different faults.
Subsequent investigations have found that at least 12 separate faults broke during the quake, including some that had
not previously been mapped.
John Hare, of the Holmes Group, explained in his You Tube clip that the high frequency shaking was largely filtered out in the Wellington region leaving lower frequencies that most
affected our mid-rise buildings (and not low-rise buildings such as unreinforced masonry).
What are basin-edge effects?
Central Wellington is situated on the edge of a geologic basin in which softer alluvial soils have been deposited. This
is called a sedimentary basin; not to be confused with reclaimed land. During earthquakes there can be an amplification
of the ground shaking at the edge of the basin. Basin-edge effects resulted in concentrated areas of damage during the
1994 Northridge (California) and 1995 Kobe (Japan) earthquakes.
There is not sufficient international or national research to say conclusively where we can predict a basin-edge effect
exists, which is why we will be commissioning further research and working with and sharing findings with other
countries where basin-edge effects may impact buildings and infrastructure.
Does the Building Code factor in earthquake design?
Yes. Specific design standards are used by structural engineers to demonstrate compliance with the Building Code. These
include the Earthquake Actions Standard (NS1170.5) which provide the specific earthquake demand criteria to be used in
the design of buildings and material standards such as the Concrete Structures Standards (NZS3101) which provide methods
for determining the capacity of structural elements.
The Building Code and the supporting design standards permit buildings to be designed to undergo controlled damage in
earthquakes but still protect life. To construct buildings that are not damaged during large earthquakes would be
uneconomic or just not practical.
The combination of factors that led to the partial collapse of floor units in Statistics House was not anticipated by
the design standards in place, when it was built.
Design standards are regularly updated to incorporate new learnings. The soon to be published amended Concrete
Structures Standard will contain specific provisions that acknowledge the impact of beam elongation on frame buildings
and how seating of pre-cast concrete floor units needs to be enhanced to avoid what occurred at Statistics House.
MBIE will now be recommending further changes be considered, in particular for the Earthquake Actions Standard
(NZS1170.5) , based on what the investigation has taught us. See Recommendation Three and Recommendation Four.
Can buildings like Statistics House be improved to an acceptable standard?
Potential performance issues had been identified by the owner of Statistics House, taking into account lessons learnt
from the Canterbury earthquakes. Remedial work on the fourth floor of the building had been completed and this performed
well during the Kāikoura earthquake.