Developing a Strategic Ground Model for the Gold …oemg-global.com/.../2017/07/220_Errey_final-paper.pdfDeveloping a Strategic Ground Model for the Gold Coast Waterways Errey, J.

Coasts & Ports 2017 Conference – Cairns, 21-23 June 2017Developing a Strategic Ground Model for the Gold Coast WaterwaysErrey, J. Bourner, J. & McRae, B.

Developing a Strategic Ground Model for the Gold Coast Waterways

Jason Errey1 , Jessica Bourner2and Brian McRae3

1 Director, OEMG Global, Bellingen, Australia.email: [email protected]

2 Senior Advisor (Strategy), Gold Coast Waterways Authority, Australia.3 Manager (Strategy), Gold Coast Waterways Authority, Australia.

AbstractThe Gold Coast Waterways Authority (GCWA) has undertaken an ambitious effort to create a singleintegrated ground model, including advanced geophysics. It is envisioned that the model will incorporatespatial datasets including geophysics, geotechnical and environmental and become the primary decision andrisk management tool across the Authority. This paper tracks the acquisition, processing and utilisation of thedataset and its practical application as a strategic asset. This is a key part of an integrated approach to themanagement of the waterways and is designed to capture benefits including; risk reduction, better decisionsfor infrastructure placement, reduced design and construct costs and better environmental outcomes.

GCWA was established on December 1, 2012. The Authority is tasked with the management of the inlandwaterways, including rivers, canals, lakes and dams within the Gold Coast Local Government Area, as wellas the mouths of the Nerang River, Currumbin and Tallebudgera creeks. Management of the waterwayspresents unique challenges as they encompass sensitive environments (seagrass, mangrove and acidsulfate soils) and, support significant residential, tourism, industrial and commercial uses.

The Authority holds a strong archive of data regarding the waterways, including changes associated withdevelopment of the Gold Coast Seaway. Gaps in the understanding of the geological setting were identified,including the extents of indurated sand deposits (locally known as “coffee rock”). Understanding the depth,distribution and density of the coffee rock is important, particularly for commercial and recreational vesselaccess, as it is difficult and costly to dredge and pile. More broadly, the coffee rock is thought to influence thehydrodynamics and associated movement of sediments, including the net import/export regime through theSeaway. The application of Aquares geophysics during this project has made it possible to visualise relevantaspects of the Coffee Rock.

GCWA commissioned OEMG to create an Integrated Digital Ground Model based on data acquired utilisingthe Aquares Sub-Bottom Profiler. To test the potential applicably of the system, four proposed project areaswere identified and, linked along existing channels. Approximately 210 line kilometres of data were acquiredcreating an extensive ground model. The dataset has informed a sediment sampling plan that will be used tovalidate the Aquares data in terms of levels and calibration of geological predictions, such as coffee rock andacid sulfate soils, as well as environmental associations such as seagrass communities.

Keywords: advanced geophysics, maintenance dredging, coastal structures, coffee rock, Integrated DigitalGround Model.

1. IntroductionSince the early iterations of the Integrated DigitalGround Model (IDGM), a compelling story of thestrategic value of robust geological models hasemerged. The economic and environmentalbenefits are measurable; to date the IDGMcombined with the Aquares sub-bottom profilingmethod has realised ~$100 million of savings, at acost of ~1% of benefits [3, 4, 5]. Less visible is anoverall reduction in project risk. Evidence fromrecent projects suggests this net risk reductionincreases contractor participation and associatedinnovation [4]. However, as all major decisionssurrounding concept design are reliant on acompatible geological setting, the value of thesestudies to the proponent has been limited due to

the work being undertaken just before or duringconstruction.

It is likely therefore, that initial ground studiesbased on high resolution advanced geophysicalmethods, rather than point source methods (suchas boreholes), at the concept phase, or at thelatest, the Environmental Impact stage, wouldreturn greater benefits including; improvedeconomic and environmental surety and, greaterinvestor confidence for on time and on budgetproject delivery. The key however, is to providecomplex datasets in simple to understand, industryrecognised formats, that can be trusted bystakeholders. The creation and delivery of such adataset is the basis of this paper.


High resolution advanced geophysical datagathered for this project supports both the GCWAScientific Research and Management Programand the data needs specific to several projects inconcept phase. The acquisition of such a largedataset allowed the contractor and the client toinvestigate the usefulness of this type of data tothe organisation as a stand-alone tool, as well asthe advantages of combining additionalenvironmental and oceanographic datasets.Various delivery frameworks were investigated tomake the data available to a wide audienceincluding standard paper charts, CAD in variousformats and novel digital solutions such as 4D GIS,Google Earth and geoPDF. The overall goal is toleverage the geological data and the IDGM as astrategic asset for the Authority.

2. Project SettingThe benefits of ground modelling and the variousacquisition tools utilised to generate data arearguably under-appreciated. Practically, a groundmodel should be viewed as a framework fordecision making and risk management. It canobviously be used to inform design and feasibilityanalysis. But, just as importantly, ground modelscan be used to generate environmental andeconomic savings through the identification ofopportunity risks, such as: ’can this project be builton time and budget?’; ‘can this project be builtwithin the environmental constraints of theregulator?’; ‘can the project be modified to utilisethe existing geological structure?’ or ‘does theproponent sufficiently understand the sub-bottomto contractually share geological risks?’.Maximising opportunities associated with eachmajor project decision is key to minimising overallproject risk and returning value to stakeholders.From the location and type of structure or activitiesproposed, to the equipment anticipated; all majordecisions, consciously or otherwise, are based onsuppositions about the geological setting. It istherefore important for proponents and contractorsto have a good understanding of both the groundmodel and associated uncertainties.

2.1 Analysis of AlternativesAll ground models comprise various subsets oftechnology. From boreholes to geophysics eachacquisition method has unique advantages andapplicability when considering a given site.However, all acquisition methods must beconsidered remote sensing techniques that requireindependent validation of results. Indeed, atechnique well suited to the testing environmentbut poorly applied and/or not independently tested,and/or not made available to relevant projectteams, will compromise the realisation of projectobjectives. In addition, all ground models containuncertainties that are inherent to the acquisitiontechniques utilised.

Uncertainty and the implications of uncertaintymust be effectively communicated to the projectteam. For example, Borehole studies are the mostcommonly utilised ground modelling tool for civilconstruction, however vertical errors are common,as are misinterpretations in sediment or rock type(Section 2.1.1, Figure 1). A failure to communicateuncertainties inherent in the acquisition of boreholedata often conveys a false sense of security aboutthe overall findings, resulting in engineering issuesduring construction. The same outcome can arisewith geophysical studies where poor procurementstrategies fail to adequately consider value formoney over price, resulting in short-term costsavings, but the acquisition of low resolution data,poor quality data and analysis and, understandingof risks.

Figure 1 A comparison of borehole cross-sections andtwo and three dimensional geophysics. The top diagram(chart 2) shows three historical boreholes that are up to40m off track; no rock is evident. The two dimensionalquantitative geophysics is added; again, no rock isobvious. In the lower diagrams, three dimensionalAquares geophysics is added and a high resistivitystructure is seen that is subsequently drilled and foundto be rock (from [4].

The biggest failure in ground studies therefore, isnot necessarily what technology is utilised, but howthe technology is applied. If the various groundmodelling tools are applied in a staged manner,each designed to test the previous study, a poorlyapplied or inappropriate technology becomesapparent through contradictory results ornonsensical relationships (Figure 1).


2.1.1 BoreholesBoreholes are critical for the acquisition ofengineering data about a specific structure, butshould not be used to infer geological structures.Utilising Boreholes as a primary data source toderive the structural setting is problematic for tworeasons: 1, as point source data, the geologist hasvery little geological context to assist in theunderstanding and classification of, particularlydeeper, layers as they relate to a project(Figure 1); and 2, it is not statistically possible todemonstrate that a sampling pattern hasadequately sampled all geological structuresrelevant to a project (this includes the “I havesampled each pile location, what can go wrong?”sampling pattern). By examining the process ofundertaking boreholes, these risks can be betterunderstood.

Figure 2 Bore log along the Yamba pipe route (courtesyClarence Valley Council). Solid lines betweendescription of strata are seen when an interface occursin the SPT sample and dashed lines are seen when theinterface occurs between SPT samples, indicating theheight is guessed.

It is unusual to see a continuous (push) corethrough sediments. The cheapest and fastestmethod of sampling sediments is destructive “washboring”, with physical sampling occurring every fewmetres when a “Standard Penetration Test” (SPT)is undertaken. The SPT sample is recovered in atube that is 500mm long by 50mm in diameter. If achange in sediment type occurs between SPTsamples, for example the change between silty

and gravelly sand at 2 meters (Figure 2), the actualheight of the transition between strata is guessedby the geologist and indicated by a dashed line. Inthe absence of additional data, that dashed linewould change to a solid line and become a designdepth, with indications of uncertainty lost.However, regardless of the sampling methodutilised, the act of recovering a sample will disturbthat sample. Compressive layers are likely to bemissed or distorted, as is intra-layer variability andcomplex geology (Figure 1). This results in a lackof detail that will increase risk and reduce the valueof the ground study, if this is the sole source ofdata about structural context.

Figure 3 Aquares data visualised in 4 Dimensions (X,Y, Z, Quality) of within the Eden Port.

2.1.2 Geophysical methodsGeophysical methods are broadly broken intoquantitative and qualitative methods [3].Quantitative geophysics are reflection basedseismic methods. They are used to define depth tolayer, but cannot be used to resolve qualityvariation between or within layers. Qualitativegeophysics, usually seismic refraction or resistivitymethods, are used to define structural variation inthe geology at a site. Refraction methods do notreadily resolve the level and depth of structures orcomplex geology due to both technical limitations[3] and procurement issues (Section 2.1). Aquaressub-bottom profiling is advanced in that it canreadily resolve qualitative and quantitative data(Figure 1 [3, 4, 5]). Aquares is also unique in that itis a digital acquisition technique and processedgrids can be utilised for 4 dimensionalvisualisations (Figure 3).

The accuracy and resolution of advanced methodshowever, varies widely with outcomes dependenton the resolution and data density (Figure 4) and,relates directly to the procurement process. Theprice of methods relates to how the acquisitioncable is towed (Figure 5) and the shot density. Theimportance of these critical elements is dilutedhowever, when the tender simply asks for a“refraction survey”. In the case where technology isspecified in place of outcomes, the cheapest


survey technique is proposed and usually selectedfor use. It must be recognised however, that nogeophysical technique can provide engineeringdata about the sub-bottom. Engineering data maybe inferred from advanced geophysical data, butsite specific data is only gained from geotechnicalsampling. Advanced geophysical methods areutilised to map the structures and target boreholesto ensure a site is adequately sampled.

Figure 4 A comparison of resolution of the three maingeophysical methods, note the solid black line is theproposed tunnel alignment. The rock layer as defined byreflection (red dashed line) and refraction seismics isseen in the top section and Aquares and reflection in thebottom. Note the reflection and refraction completelymisses the palaeochannel to the right of the section.

2.1.3 The Aquares SystemThe Aquares system is a resistivity based sub-bottom profiler that has been designed, built andoperated by DEMCO NV and OEMG Global overthe last 20 years. The system comprises a heavilyarmoured bottom towed cable (Figure 5), a “bangbox” capable of delivering approximately 20A witha duty rate of 1.2Hz, a signal modulator andlaptops for recording Aquares data, water depthand position (Figure 6). The system relies on abottom towed cable to transmit an electrical signaland receive the ground response. A bottom towedcable can be more expensive to operate, but thequality of the results, considering both thehorizontal and vertical resolution, is much higher(Figure 1, Figure 4) as compared to surface tows.

Figure 5 General cable setup of a resistivity basedsystem. Note the receiving array must be in contact withthe seafloor to achieve high resolution data.

2.1.4 Insights from Eden Port Re-DevelopmentA project planning consideration was an analysisof various data sets undertaken by OEMG duringthe creation of an IDGM for the Port of Eden [4](Figure 1). Here, there was no relationshipbetween historical boreholes 40m off track, andrecently acquired quantitative (reflection)

geophysics. This was due to the complex geologyat the site and the low resolution of the chosengeophysical method (Charts 1 & 2, Figure 1), alongthe construction alignment. Equally an uncertainrelationship was seen between the reflectiongeophysics and the qualitative (Aquares)geophysics (Chart 3, Figure 1).

The strongest correlation noted was betweenAquares data and newly acquired, Aquarestargeted boreholes (Chart 4, Figure 1). Thesetargeted boreholes confirmed, vertically andhorizontally, the predictions of the Aquares data. Inthis case, the geological setting was too complexfor quantitative geophysics or randomly scatteredboreholes to resolve. Ultimately Aquares data wasutilised by the project team to tailor the proposedinfrastructure and activities to the site and realise a20% cost saving over business as usual groundmodelling practice [4].

Figure 6 System diagram of a typical Aquares spread.Note for this survey, the shallow water spread wasutilised.

2.1.5 GCWA Data Catalogue and GapsThe GCWA has a catalogue of sediment data forGold Coast waterways that extends back to 1978(pre- Seaway construction). Hydrographic surveysrecords go back to 1919, with extensive recordsfrom the 1980s. In 2014 as part of the GCWA’sSand Management Plan, a Scientific Research andManagement Program (SRMP) was created. Sincethen, several strategic projects have beendelivered under the SRMP including large scalecaptures of bathymetry data using a topo-bathyLiDAR and seagrass community mapping. The firstphase of the Strategic Sediment Sampling projectcreated a single dataset with all known sedimentsampling data and developed a strategic samplingand analysis program to fill knowledge gaps andidentify risk hotspots associated with dredging,


including areas of high fines, Acid Sulfate Soils(ASS) and indurated sands. The Integrated DigitalGround Model represents the fourth phase of thisproject.

An early driver for this project was the inability toachieve channel design depths on the westernshore of the Broadwater. Borehole recordsindicated sufficient sand cover, but duringdredging, coffee rock was encountered at depthsclose to one-half of the channel design depth.Intended outcomes for the dataset include possiblerealignment of this channel to achieve designdepths without having to remove coffee rock.Similar planning application is intended fordredging on the Coomera River, where previousdredging also fell short of target due to unexpectedhard shoals.

Figure 7 Extents of the Aquares data collected withinthe Authority’s area of management. Here Aquaresresults are seen at 4m Below Seabed (BSB). Note thesub-surface at this depth appears harder in the southand in the North-Western corner. Two colour keys areprovided, Aquares in Ohm meters (Ohmm) relates to theresistivity results and the Borehole Key, both relate toGCWA data provided throughout this paper.

2.1.6 The Integrated Digital Ground ModelThe Integrated Digital Ground Model (IDGM) wasoriginally conceived in response to a studypublished by Engineering Australia indicating poorground studies result in up to 50% of infrastructureprojects exceeding time or financial budgets [1].The IDGM is designed as a tool to generate and

visualise robust ground models and realise thebenefits. As discussed, there are four main issuessurrounding ground studies: An understanding of technology and

perception issues of what can and cannot beachieved with each technology;

How data is gathered and in what order; How data is visualized and communicated to

the project team and stakeholders; and When ground studies are undertaken in the

project cycle.Each of these four points are not only critical to thesuccess of the ground studies, but also to thesuccessful use of the ground studies, and areaddressed in the design of the IDGM.

For this project the primary aim was to visualisethe depth, extent and characteristics of theindurated sands across the project site. Aquareswas chosen as the preferred investigation tool tosuit the aims and scale of the project. Refractionmethods are not readily able to distinguishinduration and cannot acquire data quickly enough,and in sufficient detail to be economic on theproposed scale. Visualisation of the geophysicaland historical data would normally be provided insoftware package Encom PA [4], but the sheervolume of data and variation in knowledge andskills of the users, made this problematic, and willbe discussed later (Section 4). Finally, anduniquely to this project, the value of this type ofdata can be assessed during the design phase ofprojects.

3. Data AcquisitionOEMG Global was originally approached to gatherdata about the depth, extent and characteristics ofthe indurated sands in four areas within theAuthority’s management area and, to provide aframework to confirm the findings of thegeophysics. The four areas being: Coomera River between the rail bridge and

Sanctuary Cove; The northern Broadwater between Salacia

Waters and Wave Break Island; The Seaway and selected offshore areas; and Southern Broadwater from Wave Break Island

to just south of the Sundale Bridge.

OEMG worked with the Authority to ensure anacquisition plan that was cost-effective andprovided the best coverage for upcoming projects(Figure 7). Three of the areas are centred on theBroadwater, to improve the understanding of thegeological setting and the extent of the geologicalstructures present, the fourth includes an upstreamreach of the Coomera River, supporting access tothe Coomera Marine Precinct. At this stage,dredging works are envisaged for the Coomeraarea and trenching is planned for the area nearWave Break Island and the Sundale Bridge.


To achieve these aims, OEMG works included thecollection of some 210 line kilometres of data overa 5-day period, including mobilisation anddemobilisation of the vessel. Line spacing wasgenerally less than 20m and over 100,000individual shot points were collected andprocessed to create a 4-dimensional model of thesurvey area (Figure 8). Data was successfullycollected through closely spaced moorings in theCoomera area and, in close co-ordination with theSeaway Control Tower, in the busy Seaway andBroadwater areas, with no incidents. The systemwas deployed on a small single engine craft(Figure 9) and no time was lost to weather orbreakdown.

Figure 8 A screen capture of the IDGM in Encom PAfor the Middle section. Here a vertical cross section isseen close to the western shoreline along with ahorizontal section at -10m LAT and a 4D view includingavailable Boreholes. Legends are seen in Figure 7.

4. ResultsData for the entire survey was very clean (noisefree) and required very little processing, with goodconfidence that the Aquares could resolve theextents, depth and characteristics of the induratedsands (Figure 7) and, the geological setting to 12mbelow seabed (BSB). Good distinctions were notedbetween harder regimes (red and yellow) to thesouth of Wave Break Island and in the Coomeraarea and the softer regimes to the north of WaveBreak Island. Aquares data were gridded on aregular grid at a resolution of 5m x 5m in thehorizontal and 0.25m in the vertical. Colour codingis seen in Figure 7. 3D grids are provided relativeto Seabed (BSB) and Lowest Astronomical Tide(LAT). Horizontal grids can be extracted atrequired levels and Vertical grids can be extractedalong any alignment (Figure 8).

Based on limited historical geotechnical data(Section 5.1), it is assumed that: resistivity’s above3 Ohmm represent harder material, likely induratedsands; resistivity’s between 1.5 and 3 Ohmm arelikely sediments that are weakly indurated;resistivity values between 0.6 and 1.5 Ohmm aremedium to course sands; and values below 0.6Ohnm are likely fine sands, silts and muds.

To visualise the data in 4D utilising EncomPA, thesurvey had to be broken up into the four areas(Coomera, North, Middle and South). Even so, theprocessed ASCII (text) grids for each section wereapproximately 300 megabytes in size,approximately twice the size of “normal” surveys.The large file sizes proved cumbersome due toinherent limitations in the software. Unfortunately,EncomPA is still 32bit and is not utilising theavailable system resources of the computer andgraphics card to full potential. Additional issueswere noted when creating long vertical sections.However, PA proved adequate if file sizes werecarefully managed and visualised data was kept toa minimum (Figure 8).

Figure 9 The survey was undertaken utilising a 6.5m,single engine vessel.

5. DiscussionThe collection of this data provided both difficultiesand opportunities. The scale of the project and thevolume of data are very large, but also, it turns out,the potential applications are extensive. Despite anumber of past projects being successfullydelivered in Encom PA by OEMG, systemmanagement within the software compromises theability to use large datasets. While the softwareprovides stunning visuals, it requires a dedicated,fast machine and a trained user. When consideringground modelling for a specific project, much of theanalysis and visualisations is normally preparedduring reporting, however this project required thevisualisation of geological structures over a widearea for an undetermined audience. To this end,after the initial models were produced, efforts wereundertaken to develop a program to test the datato increase surety and, understand the potentialusers of the data and their requirements.

5.1 Data TestingTo test the Aquares results, archived data wasprovided including grab samples and historicalboreholes. The most relevant data was noted as aseries of boreholes undertaken for the RecycledWater Long Term Release Program [2]. Althoughthe boreholes appear to largely line up withpredictions associated with the Aquares data(Figure 8), several potential issues were noted withthe borehole data, namely the lack of accurateposition data and supporting field notes concerningvertical levelling.

To improve confidence, further jet probes havebeen added to a planned sediment sampling


campaign to determine the presence or absence ofcoffee rock and the height to target at selectedareas (Figure 10). As data is collected foradditional projects, this will also be added to themodel. While the final form of the database is stillunder development, it is believed that that theshortcomings of Encom PA make this pathuntenable. To this end, Geosoft has agreed toassist in the testing of Oasis Montage. Thesoftware is similar, but has already beentransitioned to 64 Bit. A preliminary review of thesoftware suggests it will likely manage the largerfile sizes, but the software seems to lack thefriendlier user interface of PA. However, it isassumed that work flows can be developed to workaround these issues and, improved usability is adevelopment goal for Geosoft.

Figure 10 Here points labelled “OEMG_XX” are jetprobe targets designed to test Aquares data.

5.2 Data Rollout to End UsersIn addition to using the models created to supporttraditional works programs, it is likely that they willbe beneficial for environmental programs such ashabitat monitoring and hydrological studies. Asdiscussed, a chart has been produced relative toseabed. This effectively creates a waterway widemap of the morphology of the seabed. Areas ofsand, silt, clay and rock outcrop are seen whichare likely to prove useful when consideringseagrass habitat, sediment sources or ASSmanagement.

In addition, to assist in early stage projectplanning, geoPDF documents and Google Earthcharts have been produced (Figure 7). It isenvisaged these will be useful to the GCWA worksdepartment when replacing or repairing navigationpiles and project managers when conceptualisingprojects.

6. ConclusionThis project involved one of the largest speculativedata collections of high resolution advancedgeophysics by a management authority, seen inAustralia. The net result was over 200 line

kilometres of data collected over a large area ofthe Gold Coast waterways. Detailed data wascollected to 12m below the seabed. Seabedstructures, including indurated sands and siltyareas were inferred and, predictions of theAquares data appear to conform to historicaldatasets. Studies are ongoing to verify andcalibrate the data and develop products suitable tovarious end users.

The biggest obstacle in the use of the IDGM for theGCWA is integrating the model into operations,including the project delivery framework. As withany decision support tool; it takes time for staff tofully comprehend the use and benefits. Movingforward, it is imperative that GCWA ensures theoutputs are easily accessible and digestible byproject managers and potential contractors. Insome respects, the complexity – richness of thedata set – is an impediment to uptake. However,simplified geoPDF and Google Earth outputsappear to assist and the data is already provinguseful for risk management of underway projects.

Once use of the strategic IDGM becomes anintegral part of the project planning process, thebenefits are enormous, allowing more effectivedecision making over the entire target area andacross different departments about infrastructurepriorities and placement. Project risk will be greatlyreduced and significantly decrease long termexpenditure. It is envisaged that this process canbecome a model for the management ofwaterways and major infrastructure across thecountry.

7. References[1] Baynes, B.J. (2010). Sources of geotechnical risk,Quarterly Journal of Engineering Geology andHydrogeology, 43, 321-331.

[2] Butler Partners (2016). Geotechnical Investigation;Recycled Water Long Term Release Stage 1;Broadwater, Southport; Project No. 016128A, August2016. Prepared for Gold Coast Water Authority.

[3] Errey, J.C. & Brabers, P.M. (2013). UsingGeophysics as a Tool for Mitigating Project Risk, Coastsand Ports 2013: 21st Australasian Coastal and OceanEngineering Conference and the 14th Australasian Portand Harbour Conference, Manly Australia, 10 - 13September 2013

[4] Errey J, Dooley A. Brabers P, 2015. CombiningHistorical Geotechnical and Geophysical Data withRecent Aquares Geophysical Survey Data into a GISFramework for the Eden Port Development. AustralianCoast and Ports Conference 2015, Auckland, NewZealand.

[5] Errey, J.C. and Stewart-Rattray, R.J. (2015). Reviewof Sustainability Outcomes for Infrastructure ProjectsGained Through the use of Modern QualitativeGeophysical Methods. IPWEA IFME joint InternationalConference. Rotorua N.Z. 8 – 10 June 2015

Developing a Strategic Ground Model for the Gold …oemg-global.com/.../2017/07/220_Errey_final-paper.pdfDeveloping a Strategic Ground Model for the Gold Coast Waterways Errey, J.

Documents