The following publications are works either authored by our staff or, in some cases, co-authored with people from outside the company.This selection of conference papers and journal articles can be accessed by requesting individual items from our Tonkin + Taylor Ltd Library (firstname.lastname@example.org) or by clicking on the button beside the item. There is no charge for this service. However, please note that our Library follows Library Association (LIANZA) guidelines (link to their guidelines here) and reserves the right not to supply any item if these conditions are not met.
The 2016 Meinong Taiwan Earthquake: learning from earthquakes report
The Mw 6.4 Meinong earthquake occurred on 6 February 2016 in the southern region of Taiwan. The
earthquake caused significant damage in and around Tainan city, with a number of collapsed and severely
damaged buildings and 117 deaths. A five-member Learning from Earthquakes (LFE) team visited Taiwan
approximately one month after the earthquake, with particular focus on learning from changes to design
practice and seismic mitigation efforts following the 1999 Chi-Chi earthquake in Taiwan. Land damage was
generally modest with liquefaction and slope-failures observed in a limited number of locations. Some
notable instances of liquefaction-related foundation settlement and tilting occurred in areas associated with
historical filling. Following the earthquake, the Taiwanese government publically released liquefaction
hazard maps that will have a significant impact on public awareness and land values. The observed structural
damage was characteristic of non-ductile and poorly configured buildings. The collapsed buildings all
contained irregularities and soft-storeys. The majority of older mixed-use buildings performed adequately,
but severe column failures were observed in several taller apartment buildings constructed in the 1990s. The
performance of schools and district offices provided valuable insight into the successful implementation of
seismic assessment and strengthening programmes. A comparison of existing and strengthened buildings
showed that efficient retrofit solutions can reduce the risk posed by critical structural weaknesses and improve
the safety and resilience of these buildings. A similar strategy could be implemented for common critical structural weaknesses in New Zealand buildings.
Liquefaction hazard mapping - liquefaction vulnerability mapping for a given return period versus return period mapping for a given severity of liquefaction vulnerability
Liquefaction hazard maps are typically developed by collating geotechnical investigation data and undertaking simplified liquefaction analyses. Liquefaction vulnerability parameters are commonly calculated using a simplified liquefaction triggering method, a given groundwater level and a given set of earthquake ground motions, corresponding to a particular return period of earthquake shaking. The results at each investigation location are then typically interpolated and subsequently, smoothing might be applied. A more robust methodology involves dividing a study area into smaller Similar Expected Ground Performance (SEGP) areas as a result of earthquake shaking. Liquefaction consequence parameter values for a wide range of earthquake scenarios are then calculated using the available geotechnical investigation data and grouped according to SEGP areas in which they are located. Each SEGP area then has its own unique liquefaction vulnerability distribution fitted to the data as a function of earthquake magnitude (Mw) and Peak Ground Acceleration (PGA). Using these functions, a variety of liquefaction hazard maps can be produced. A typical mapping approach is to present the median or mean liquefaction vulnerability for each SEGP area for a given level of earthquake shaking. A variant to this approach is to present the expected spatial variability of liquefaction. This approach provides greater insight into how a study area is expected to behave spatially, which is especially relevant for risk modelling. An alternative mapping approach is to determine the level of earthquake shaking required to attain a given level of liquefaction vulnerability. This approach identifies SEGP areas where more frequent, smaller levels of earthquake shaking are likely to result in liquefaction damage and other SEGP areas where less frequent, larger levels of earthquake shaking are required for liquefaction-related damage to occur. This alternative approach helps improve the communication of the liquefaction hazard to non-technical audiences and presents the results in a similar way to other natural hazards that are assessed for land-use planning and hazard management purposes.
Pipeline damage predictions in liquefaction zones using LSN
Liquefaction is a major concern regarding earthquake damage to infrastructure. Recent earthquakes in New Zealand and resulting liquefaction caused significant damage to buried pipeline systems. Following the 4 September 2010 Mw=7.1 Darfield earthquake, five earthquakes (22 February 2011, Mw=6.2, 13 June 2011, Mw=5.3 at 1 p.m. and Mw=6.0 at 2:20 p.m. and 23 December 2011, Mw=5.8 at 1:58 p.m. and Mw=5.9 at 3:18 p.m.) and thousands of aftershocks have been recorded in the area of Christchurch, NZ. These earthquakes termed the Canterbury Earthquake Sequence (CES) are unprecedented in terms of repeated earthquake shocks with substantial levels of ground motion affecting a major city with modern infrastructure. This study focuses on the effects of 22 February 2011 Christchurch earthquake induced liquefaction on buried pipelines. Correlations were developed between pipe damage, expressed as repairs/km, and a recently developed parameter called liquefaction severity number (LSN). Cone Penetration Test (CPT) based liquefaction triggering procedures were used to calculate LSN values. Studies by Tonkin and Taylor [1,2] and van Ballegooy et al. [3, 4, 5, 6] have shown that LSN provides a good correlation with land and esidential house foundation damage observations recorded in Canterbury. According to results obtained in this study for buried pipelines, LSN has reasonably good correlation with asbestos cement (AC), cast iron (CI) and polyvinyl chloride (PVC) pipeline damage.
How the best ideas win: the role of collaboration in successful innovation
Canterbury Earthquake Sequence : increased liquefaction vulnerability assessment methodology - Appendices
Planning for the NOW society - smart water and wastewater systems and their implementation in New Zealand
Modern society has created a culture which expects instant information. Digital media and applications have adapted to fuel and feed this desire, which can be seen in everything from the Fitbit to our demand for up to the minute news and sports information streams. Yet, with a few notable exceptions, as infrastructure providers we frustratingly continue to provide our planners and customers with outdated information.
What would it take for our customers to be able to view their water, gas and power meter readings in real time on a phone app? For entire wastewater networks to become ‘smart’ enough to tell operators that the sewer is blocked or about to surcharge? Some such capabilities are already available in other countries but for many New Zealand Councils this level of customer service and asset management capability may feel like decades away.
Technologies exist here in New Zealand that will shortly enable a step change in the way utilities and customers capture, use and disseminate information. This paper will draw on international case studies and showcase emerging technologies such as Celium, a low cost long range low power wireless network, which have the capability to bring customer service and asset management into the NOW.
Catchment level modeling of green roofs using InfoWorks CS
Green roofs are vegetation installed on top of buildings to provide flow control by attenuation,
storage and losses due to evapotranspiration. A green roof consists of several-layered materials to achieve the
desired vegetative cover and drainage characteristics. An attempt has been made to use the different runoff and
infiltration models available in the widely used hydraulic modeling software - InfoWorks CS to model runoff
from green roofs during storm events and over a longer continuous simulation period. The most suitable model
was then applied to test the benefits across 03 catchments in InfoWorks CS considering a range of percentage
uptake of green roofs within the catchments. The benefits of green roofs implemented on a catchment level are
assessed in terms of Combined Sewer Overflows (CSO) performances.
Influence of geometric, geologic, geomorphic and subsurface ground conditions on the accuracy of emprical models for prediction of lateral spreading
Liquefaction-induced lateral spreading can result in significant damage to the built environment, as observed in
Christchurch during the 2010 to 2011 Canterbury Earthquake Sequence (CES). Predicted Lateral Displacements (LD) from
published empirical models have been shown to vary from those measured in parts of Christchurch during the CES by a
factor of <0.5 to >2. A widely used empirical method for predicting LD is that proposed by Zhang et al. (2004). Based on a
few selected transects along the Avon River in Christchurch, the Zhang et al. (2004) model has been shown by some
researchers to provide better agreement between the measured and predicted magnitude and extent of lateral spreading
compared to other LD prediction models. Conversely, based on a different set of selected transects along the Avon River,
other researchers have shown that the Zhang et al. (2004) empirical model does not provide a good fit between the
measured and predicted LD compared to other LD prediction models. The reasons for these apparent contradictory
conclusions may result from the varied transect locations and associated geometric, geologic, geomorphic variability and
subsurface ground conditions. The objective of this study is to evaluate the combinations of these factors for which the
Zhang et al. (2004) empirical model predicts the LD reasonably well and also the conditions for which it does not predict
the LD very well. Combining the available datasets outlining horizontal ground surface displacements during the CES, the
maximum extent of lateral spreading and the magnitude of maximum displacement has been estimated along the Avon
River. By using the extensive Cone Penetration Test (CPT) dataset available, a regional lateral spreading assessment has
been undertaken, based on the Zhang et al. (2004) empirical model, to assess the predicted LD along a reach of the Avon
River eastward of the Central Business District (CBD). The results have been compared to the measured LD that occurred
for the 22 February 2011 earthquake. The results show that the Zhang et al. (2004) model tends to over predict LD more
in the older river terrace deposits when compared to the younger reworked river floodplain deposits.
A Review of Shoreline Response Models to Changes in Sea Level
Assessment of current and future coastal hazards is now a legislative requirement in New Zealand
and most parts of Australia. Methods for assessment of erosion hazard are well established, and uncertainty
in the present hazard can be reasonably well estimated. However, uncertainty in defining future
climate-change associated erosion/recession hazard increases due to both the assumptions
surrounding sea-level rise (SLR) as well as limitations of the models used to evaluate the
associated shoreline response. The most widely used methods for defining the coastal erosion hazard
extent utilise a modular approach whereby various independent components are quantified and summed
to provide a final value (e.g. see ). The SLR response component is based on the well-accepted
concept that an elevation in sea level will result in recession of the coastline. This component is
often the largest contributor to erosion hazard zones, so understandably this term is often the
subject of intense debate, media scrutiny and a focus in litigation. With the trends of increasing
populations on the coast this controversy is only likely to escalate. A range of models for
estimating coastal response to changes in sea level have been developed over the past 50 years.
These methods range from the application of basic geometric principles to more complex
process-based assessment. While some methods are used more widely than others, none have been
proven to be categorically correct or adopted universally. While most attention has focussed on the
response of open coast beaches to SLR, other shoreline types including gravel beaches and low
energy coastlines such as lagoons and estuaries are also affected. This paper briefly reviews
existing shoreline response models including the process assumptions, limitations, development and
application history. While most models are based on similar underlying process assumptions,
variation in the definition of model parameters (e.g. closure depth) can produce significant
differences in predicted recession values. As such, robust and informed selection of model
parameters are required to derive defensible conclusions.
Havelock North water supply Campylobacter outbreak - source and ingress
In August 2016, the Havelock North public water supply suffered a significant
contamination event (Campylobacter), resulting in an outbreak of gastroenteritis in the
Havelock North community. Groundwater bores within the Brookvale Road bore field were
suspected by Hastings District Council (HDC) to be the source(s) of contamination of the
drinking water supply. Following the outbreak, HDC, Hawkes Bay Regional Council (HBRC),
the Ministry of Health and their supporting agencies launched full scale investigations to
understand what caused the contamination of the bore(s). Bores 1 and 2 were immediately
decommissioned following the outbreak, Bore 3 had already been shut down following an
earlier E. coli transgression. A Government Inquiry into Havelock North Drinking-Water
was subsequently convened to investigate the outbreak and its cause.
T+T was engaged by HDC to undertake investigations to evaluate the potential sources of
the Campylobacter contamination within the Brookvale bore field catchment, and, if
possible, determine the source of the Campylobacter that caused the outbreak and the
means by which the contamination entered the water supply bores.
Results of the T+T investigations are detailed in this paper. The scope of the investigations
was modified as new information came to hand, and an understanding was developed
regarding the source of the contamination, and the mechanism by which contamination of
the water supply occurred. Genotyping of the Campylobacter undertaken by ESR
determined an ovine (sheep) source and its location, and groundwater modelling and dye
tracer tests confirmed the contamination pathway was from a ponded area within the
nearby Mangateretere Stream via groundwater to the Bore 1 screens. Bore head and
defective casing ingress scenarios were largely discounted by the further investigations.
Very stringent technical standards were required by the Government Inquiry into the
Havelock North outbreak. The paper emphasises the very high level of technical
collaboration between HDC management, its consultants and legal advisors in order to
meet these requirements.
Methodology for developing microzonation maps of predicted liquefaction vulnerability severity
There are a range of indices and parameters available for estimating the liquefaction vulnerability of a particular site for a particular level of shaking. These indices and parameters are typically derived from Cone Penetration Test (CPT) investigations and based on simplified liquefaction triggering frameworks, which have uncertainties associated with them. As such, application of these indices and parameters to a range of different soil types and stratified soil profiles leads to uncertainties in the accuracy in assessing liquefaction vulnerability severity. Furthermore, the simplified frameworks used as the basis for the indices and parameters are typically developed from case histories where the soil layering is generally more straightforward. This may bias the simplified frameworks to non-heterogeneous spatial stratigraphy situations where the pore water pressure dissipation following liquefaction is likely to be one-dimensional. In Canterbury, New Zealand, where a sequence of earthquake events between 2010 and 2011 had significant liquefaction effects, there is considerable spatial heterogeneity in some areas, potentially resulting in three-dimensional (3D) effects that influence the liquefaction vulnerability severity at a particular site. Comparison of liquefaction vulnerability indices and parameters with observed liquefaction-related land damage in a range of earthquake events with varying levels of shaking in the 2010-2011 Canterbury Earthquake Sequence (CES) show that in some areas the indices agree with the observations and in other areas there are inconsistencies (e.g. liquefaction is predicted yet nothing occurred for a given level of shaking).
This paper describes the methodology used for developing microzonation maps of predicted liquefaction vulnerability severity in Canterbury by combining observations of performance at particular levels of shaking with an empirically calculated liquefaction parameter. Liquefaction-related land damage observations from previous earthquake events are combined with analytical predictions while considering other pertinent factors such as geology, groundwater depth, topography, and stratigraphy to assess liquefaction vulnerability severity. Examples of applying this methodology for the CES are presented and show that in some areas liquefaction indices can be used without any need for manual adjustment but in other areas adjustments are required to predict liquefaction vulnerability severity. The methodology utilizes an area-wide assessment approach, as characterizing liquefaction vulnerability severity should consider surrounding ground investigation data in areas of geologic similarity. Classifications of liquefaction vulnerability severity are introduced and the implications of a given severity classification for design of residential buildings are discussed.
Applying agile process management to flood hazard modelling
Fast model builds are essential if Auckland Council is to keep pace with rapidly developing greenfield areas and adequately plan stormwater infrastructure in catchments like the Hingaia Stream. With no detailed catchment model available for the Hingaia Stream catchment, Auckland Council applied Agile Process Management techniques to develop a Flood Hazard Model (FHM) that would service both the needs of the developer and Council planners for this Plan Change Variation in just seven weeks. Council’s Flood Planning Team leveraged off their Modelling Project Office to use multiple consultants and Council modellers to develop a fast and detailed 1D-2D coupled catchment model,
utilising the latest LiDAR data. The FHM is suitable for Plan Change purposes and testing gross landform changes. The FHM model also provides the platform for subsequent model refinement for flood hazard mapping.
The Agile Management approach demonstrates that fit for purpose model builds can be achieved within significantly shorter timeframes and in a manner that provides better value to Auckland ratepayers, whilst maintaining quality. These achievements help to meet the city’s objectives relating to enabling growth and helping making Auckland the world’s most liveable City. In summary, by using Agile, Auckland Council achieved:
• A fit for purpose model in short time frame.
• New model build and review tools that will be usable on other model builds.
• An outcome-focused collaborative working environment.
This document "" is copyright therefore we are required to ask for your name and we will email the document to you.