Vous êtes sur la page 1sur 13

Visualizing Boreholes in 3D Using GIS

Erik Fjeldsted[1], Tyler Hansen[1], Craig Hebbert[1], Scott VanderDoes[1]


[1] Department of Civil and Environmental Engineering, Utah State University, 8200 Old Main Hill, Logan,
Utah, 84322-8200, USA (erik.fjeldsted@aggiemail.usu.edu, tyler.m.hansen@aggiemail.usu.edu,
craighebbert@gmail.com, scott.vanderdoes@gmail.com)

Abstract: In geotechnical engineering, visualization of subsurface soil properties can be greatly beneficial.
Visualizing raw or interpolated borehole data can better show spatial trends. These trends can then be used
to select new borehole locations and reduce the total number of boreholes needed. This can save time and
money in preparing geotechnical reports. Herein, we present methods of creating 3D representations of
bore log data in the Wasilla, Alaska area using ArcScene.

Keywords: Borehole data; Geotechnical Engineering; Subsurface Visualization; GIS

1. INTRODUCTION

Geotechnical reports describe the engineering characteristics of the soil for a particular site. They are used
by engineers in planning, site preparation, and footing design for new construction. Any new construction,
whether it is infrastructure such as roads, bridges and dams, or buildings from small residential houses to
large skyscrapers, will require information about the soil it is to be built on. Geotechnical reports are also
used in evaluating the soil conditions under existing structures. Such evaluations may be required when
making expansions or in remediation of post construction problems.

The key pieces of information in geotechnical reports are the boring logs. Borings are holes drilled into the
earth. The boring log shows a profile of the soil characteristics through the depth of the boring. It is these
soil characteristics such as soil type, strength, and water content, that engineers use in planning and
designing.

Field borings are expensive and sometimes can take several weeks to complete. As such, existing borings
can be useful information in making estimates during the early stages of design and in preparing bid
documents. It was determined that our goal was to create a geodatabase organizing the borings completed
by Mark Hansen, PE, in the Wasilla, Alaska area. Organizing the boring logs in ArcScene allows engineers
to visualize soil profile trends over a large area. Although a boring may not have been completed on a
particular site, initial estimates can be made from nearby borings.

Since the company was founded in 1987, Mark Hansen, PE, has completed hundreds of borings. For this
project not all of the borings could be imported into a database. Setting up the organization and structure
so that new borings can be easily added to the geodatabase and visualized in ArcScene was the desired
outcome.

1.1. Background

The data provided for each borehole includes location, soil type, moisture content, blow count, and frost
class. Soil type defines the type of soil at a given depth, whether that be sandy, clayey, silty, etc. Moisture

1
content is given in percentages which can range from 0 (completely dry) to over 100% (for soils that have
more water than solids).

Frost class refers to how much heave a particular soil will have if it freezes. This can cause problems for
most any structure. Frost class ranges from non frost susceptible (NFS), to the greatest heave potential
(F4). An example of frost heaving can be seen in Figure 1. (Chamberlain, 1981)

Figure 1. Example of frost heaving

The blow count data is generally obtained from a standard penetration test (SPT), where a thick-walled
sample tube is driven into soil using a standard procedure (ASTM D1586-11). The number of blows of the
weight that it takes to drive the tube into the soil a set distance is called the blow count. This information
gives an indication as to the density of the soil that the sampler is being driven into and is commonly used
in geotechnical formulas. (Kulhawy and Mayne, 1990)

2. METHODS

It was determined that the bore log data would be best represented in ArcScene. ArcScene is a similar
software to ArcMap with the added benefit of displaying data in 3D. Data from either program is cross-
compatible in the other. They both can display data from the same geodatabases. Again, the differences
start in the fact that ArcScene can display the data in 3D.

In order to set up the geodatabase, we needed to manipulate the raw data from bore logs and also work
with ArcScene. This proved to be the most challenging part of the process. Many hurdles had to be
overcome while attempting to compile the data into one spreadsheet. Also, problems arose as we started
attempting to model the data we had compiled.

2.1. Data and Related Issues

The bore log data to be visualized in ArcScene, data was obtained from Mark Hansen in PDF form. Several
logs were acquired, with multiple bores in each log, totaling 54 boreholes. Each log contains soil
2
classifications of the various soil layers in addition to frost class, moisture content, and blow count for
various layers. This information is given for each borehole at various depths. The locations of the boreholes
were given in latitude and longitude and are shown in Figure 2.

Figure 2. Locations of the boreholes which were modelled in 3D

The data provided for each property of the borehole were given for various depths. Many of the depths did
not correspond to depths for other attributes for the data; e.g., the data for soil class is given in ranges while
the blow count and moisture content are given as single data points at varying depths independent of each
other and soil class.

Obviously, manipulating the data and importing it into ArcScene would have been much easier had the data
already been in spreadsheet form instead of the provided PDF. After manually entering the data into a
spreadsheet, the latitude and longitude were entered for each borehole along with a unique ID number to
better identify each borehole.

Originally, because each set of data was independent of each other, the soil class, frost class, moisture
content, and blow count were placed in separate tables. Because of the differing depths of each piece of
data, it was determined that combining all of the soil attributes into one table would be advantageous. In

3
order to do this, we needed a uniform set of depths to record each attribute in the table for each borehole.
This resulted in 571 rows of data for the 54 boreholes. An example of the data obtained and finally
manipulated can be seen in Table 1.

Table 1. Example data from borehole 15006_1

Boring Begin End Soil Frost Moisture


Name Lat. Long. depth Depth Type Class Content Blow Count
GW-
15006_1 61.60538056 -149.3571111 0 1.5 GM F1
15006_1 61.60538056 -149.3571111 1.5 2.5 GM F2
15006_1 61.60538056 -149.3571111 2.5 4.5 GM F2 10.30% 46
15006_1 61.60538056 -149.3571111 4.5 5 GM F2
15006_1 61.60538056 -149.3571111 5 7 GW NFS
15006_1 61.60538056 -149.3571111 5 7 GW NFS 1.90% 47
15006_1 61.60538056 -149.3571111 7 9.5 GW NFS
15006_1 61.60538056 -149.3571111 9.5 10 GM F2
15006_1 61.60538056 -149.3571111 10 12 GM F2 5.50% 57
15006_1 61.60538056 -149.3571111 12 15 GM F2
15006_1 61.60538056 -149.3571111 15 16.5 GM F2 5.00% 65
15006_1 61.60538056 -149.3571111 16.5 22 GM F2
15006_1 61.60538056 -149.3571111 20 22 GM F2 3.60% 57

After many iterations of shuffling data into various table layouts, it was finally in a format and layout that
could be easily imported into a single geodatabase. With the data in multiple tables, multiple datasets would
have needed to be linked together to get the same results.

2.2. Technical Problems

After the data from the geotechnical reports was entered and organized in an Excel document, it then
needed to be input into ArcScene and manipulated into a 3D visualization. During initial research, it was
found that there were two potential tools that looked promising to convert the imported borehole data into
a 3D representation. One tool was a free custom developed toolbar specifically designed for 3D borehole
visualization, and the other was a tool commercially developed titled ArcHydro Groundwater. These tools
looked promising due to an easy user interface and tutorials were found for each of these tools. It was
thought that these tutorials would be extremely important and helpful as no team member had any
experience working with ArcScene.

The first tool which was tried was a custom toolbar titled 3D Borehole Tools developed by Jennifer Carrell
for the Illinois State Geological Survey. The tutorial found in research walked step-by-step through borehole
data import and 3D visualization and could then interpolate between borehole data to create a continuous
subsurface visualization. This was exactly what was desired as an outcome. The toolbar was found and
downloaded without any problems and can be seen in Figure 3.

4
Figure 3. Screenshot of how 3D Borehole Toolbar should look (USGS)

However, as the custom toolbar was imported into ArcScene, a problem was quickly discovered as seen in
Figure 4. This toolbar had been developed for a previous version of ArcScene using VBA code. The newest
version of ArcScene we used did not support VBA code and therefore, did not support this toolbar. After
some research, it was determined that the processes to obtain the necessary license and permissions to
run the VBA tool were complicated and time consuming. It was, therefore, concluded that another tool would
need to be used.

Figure 4. Screenshots showing VBA code error and how the toolbar looks in ArcScene without the
required VBA license

The second tool implemented was a tool called ArcHydro Groundwater. This toolbar was developed
commercially by Aquaveo and had a tutorial walkthrough. The toolbar was downloaded and imported into
ArcScene without any problems. The tutorial then started to be worked through by the team. The first step
was to step up a Geodatabase in a specific format with specific features classes. After the Geodatabase
was set up, data from our Excel document was then imported and organized. The locations of the various
boreholes were imported relatively easily using borehole specific latitude and longitudes; however,
problems arose when the data from the geotechnical reports (soil type, water content, soil layer thicknesses,
SPT blow counts, frost classes, etc.) started to be imported as it had to be in a very specific format. After
the data was then reformatted, the data was imported and linked to the borehole locations. As the depths
of the boreholes were being extracted using the ArcHydro Groundwater toolbar, it was discovered that the
version of the toolbar that was free to download did not include the feature to visualize subsurface
conditions. Aquaveo’s website was consulted and it was found that in order to visually represent borehole
subsurface conditions in 3D, it would cost $2000 for a single-user license as seen in Figure 5.

5
Figure 5. Screenshot showing the price of the subsurface analyst tool (Aquaveo)

Many hours had been spent using this toolbar and getting the data imported and organized, only to discover
that in order to finish the project, either a large sum of money was needed to be spent or the team needed
to start from square one. Given the lack of funding for this project, only one option was viable, to start over
again. From the initial research done, these were the only tools found that could be used for 3D borehole
visualization. Now, the tools included with ArcScene would need to be used to manually manipulate the
data to visualize it. This was problematic as none of the team had experience using ArcScene and minimal
help was found online to assist in the project.

2.3. What Worked

After many different attempts to try and get the information in a presentable format, the following procedure
was used with the help of Caleb Buahin. The first thing done, was to add a latitude and longitude for all of
the depths for all of the boring points. At first it was not clear how to do this most efficiently. Then the idea
presented itself to use the vlookup function in excel, since a table with the boring number, latitude, and
longitude was previously created, then it was saved as a .csv.

After getting the table of data formatted correctly, a digital elevation model (DEM) for the project area was
imported into ArcScene. Then the borehole data was imported by going into ArcCatalog and selecting
“create feature from XY table”. When bringing in the borehole data, the coordinate system that was used
was “Projected Coordinate System → World→WGS 1984”. The WGS 1984 uses latitude and longitude for
projecting the data. In ArcScene, 3D elevation relief is able to be represented. When going into the
properties for the layer there is a tab titled “Base Heights”. In this tab the DEM was set to be a floating layer.
By adding relief to the map, the 3D features of Arcscene were able to be used. The Add Z information tool
locates where a vector is, finds that point on the DEM, then assigns an elevation value to the point. This
6
tool is located in the Arctoolbox→ 3D Analyst → 3D Features → Add Z information. This was done to be
able to give the borehole data a ground height value and be attached to the ground surface. The Buffer tool
was used to make the boreholes look larger and easier to read.

The bore log information then needed to be extruded downward to show the different engineering
properties. This was done by going to Properties → Base Heights → selecting No elevation values from
surface and use a constant value or expression. The expression used was (initial height) - (end boring
height)*50. The reason the data was multiplied by 50 was to exaggerate the depths to be able to see the
layers more clearly. Then in the extrusion tab, the function ([End_Depth]- [Begin_depth])*50 was used. This
expression was multiplied by 50 as well to exaggerate the layers to better be able to see, spatially, what is
taking place, as can be seen in Figure 6.

Figure 6. Finished product in ArcScene showing the locations of and properties of the boreholes while
also having a photo overlay

2.4. Results

After the previously described steps were completed, a 3D representation of the boreholes could be shown
representing soil type, frost class, moisture content, and blow counts. These four parameters are each
attached to separate layers and therefore can be viewed independently of each other. These parameters
can be extremely useful to help better understand overall subsurface conditions over a wide spread area.
By comparing the 3D representations we can more easily interpolate between boreholes. This helps us to
be able to better understand the variability in a specific location as well as understand the associated level
of risk for a location. If multiple boreholes show similar subsurface conditions a much higher level of
confidence can be used for geotechnical design. However, if there is a lot of variability between individual
3D borehole representations, there is a much higher level of risk due to the high level of unknown
7
subsurface conditions. This variability and uncertainty in subsurface conditions can significantly impact the
outcome and design of a project. This uncertainty can lead to unconservative soil design parameters being
used for the project design. This can lead to significant problems such as differential foundation settlement
and cracked foundations. These problems can significantly reduce the life expectancy of a project or lead
to very costly repairs. Therefore, depending on the nature and requirements of a project, if a high level of
variability is seen in the 3D representation, either more boreholes should be drilled to better understand
subsurface conditions or very conservative design parameters should be chosen for design.

In addition to helping with specific, current projects these 3D borehole representations could also be used
to assist in preliminary site selection for future projects without performing costly subsurface explorations
or significantly decreasing the amount of subsurface investigation that must be performed. It would be very
useful to continue to expand this Geodatabase to include more boreholes from other projects in the area.
This would help us to more fully understand subsurface conditions and make better interpolations between
borehole locations due to the decreased distance between boring locations.

Based upon the data provided in the geotechnical reports two types of 3D profiles were developed. First,
for the moisture content and SPT blow counts a profile consisting of separate floating cylinders with data
breaks was created (See Appendix). The second type of profile developed for the soil type and frost class
is a continuous profile consisting of a single cylinder (See Appendix). The difference in the two types of
profiles is due to how data is collected and recorded for subsurface explorations. For the moisture content
and SPT blow counts, tests are performed and samples taken over specific depth intervals (usually every
3 to 5 feet). This leads to gaps in the subsurface profile where no data is available. Therefore, the floating
type of profile gives us an accurate representation of the actual known conditions. In other words, the space
in between the floating cylinders is due to unknown actual conditions as continuous testing wasn’t
performed. However, for the soil type and frost class the actual conditions for the entire profile are much
better known. As each borehole is drilled, the auger brings the soil to the surface. The soil can then be
identified and linked to the approximate depth of the drilling auger. The frost class can then be linked to the
identified soil type. This leads to a much more continuous subsurface profile.

3. CONCLUSION

The completed geodatabase is a useful tool for visualizing the changes in soil throughout the Wasilla,
Alaska area. Soil type varies greatly over the large sample area. When the sample area is limited to an
individual property, trends in the soil profile can be more accurately interpolated between boreholes. Being
able to visualize the soil layers helps plan for site preparation and estimate how much ground improvement
will be required. When considering a new site for construction, engineers will be able to use ArcScene to
simply look at the soils in the area around the new site to predict ease of construction.

Entering new data is greatly simplified now that there is a set format for entering the data and an established
geodatabase with layers for soil type, frost class, moisture content, and blow counts. As more borehole
data is added to the geodatabase, it will become a more effective tool for engineers.

ACKNOWLEDGEMENTS

Special thanks needs to be given to Caleb Buahin and Dr. Jeff Horsburgh for assisting us in getting the
project off the ground. Thanks should also be given to Mark Hansen, PE, for providing us with all of the
data used for the project.

8
REFERENCES
Aquaveo, http://www.aquaveo.com/software/ahgw-archydro-groundwater-introduction
ASTM D1586-11. “Standard Test Method for Standard Penetration Test (SPT) and Split-Barrel Sampling
of Spoils.” West Conshohocken: ASTM International, n.d.
Chamberlain, Edwin J. “Frost Susceptibility of Soil.” U.S. Army Cold Regions Research and Engineering
Laboratory (1981): 1-121.
Kulhawy, F. H. and P. W. Mayne. Manual on Estimating Soil Properties for Foundation Design. Research
Project 1493-6. Ithica: Cornell University, 1990.
USGS, Carrell, http://pubs.usgs.gov/of/2014/1167/pdf/ofr2014-1167_carrell-tools-and-techniques.pdf.

9
10
11
Soil Classification for Wasilla, Alaska

12
13

Vous aimerez peut-être aussi