Final Report GEOG-594B Jon Williams

Realizing the Bakken Regions Full Potential: Using GIS to Determine Suitable Areas for Additional Oil Wells in North Dakota
Summary Geospatial analysis shows that the most favorable areas for additional oil well development are located in the western quarter of the state, while the least favorable areas are located along the eastern border. Approximately 11,500 square miles, or 16% of North Dakotas total land area, falls into the most favorable category. The key assumptions that were taken into account when developing and executing this analysis was that the criteria; meaning the data used, and the rankings and weights associated with that data, were in line with what oil companies and geologists use as they approach a similar question. The analysis also assumes the fidelity of the data, namely the accuracy and currency of it. The third assumption that was made was that the owners of the mineral rights in the region are willing to sell those rights to allow for drilling to take place. This also assumes that there arent any existing agreements with other individuals or entities that would prevent them from doing so. The current breakdown of the mineral rights ownership are 82% are owned by private parties, 6% are owned by the state, and the remaining 12% are owned by the federal government1. Background
1

By 2040, oil and natural gas are expected to meet about 60 percent of the global energy demand, with North America likely transitioning from a net importer of energy to a net exporter by 20252. Additionally, global oil supply will struggle to keep pace with demand3. This is partly due to declining output from mature fields, which could decline perhaps by 80 percent by 2035. Just to replace this lost output, the world needs new production capacity equivalent to one Saudi Arabia every three years. The Bakken region shows its relevance by helping to address this need, as it is one of the largest contiguous deposits of oil and natural gas in the United States4. A recent report released by the United States Geological Survey (USGS) estimates that the formation has 7.4 billion barrels of oil5. For comparison, the largest oil field in the United States, the Arctic National Wildlife Refuge in Alaska6, is estimated to hold up to 10 billion barrels of oil; however due to environmental issues, a drilling ban is presently in place. When the Bakken region was surveyed in 1995, there was an estimated 151 million barrels of oil7. The current estimate is approximately 50 times more than that. The large difference between the estimates within that 18 year time frame can be attributed to better technology being used now to locate oil reserves and also better extraction methods. There are three drilling methods that are used in the Bakken region: vertical drilling, horizontal drilling, and hydrologic fracturing (fracking). Vertical drilling is the traditional, straight down drilling where the oil is literally pumped straight out of the ground8. However, oil doesnt always collect in pools, and can seep through the rocks and other porous materials. Vertical drilling doesnt drain all of the oil, as rocks and sediment break up the patches of oil. Horizontal

drilling, sometimes called directional drilling, allows for the drilling hole to be turned horizontally; therefore, exploiting more oil than the vertical wells can. Though this method was developed in the late 1920s, it only became cost-effective and technically feasible in the early 2000s9. Fracking, which is the most common method used in the Bakken region, utilizes the horizontal drilling method and then introduces the use of water pressure to create fractures in the rocks, which then allow the oil to escape and flow out of the well10. The costs needed to build a new well in the Bakken region are what will potentially hold off development with the exception of larger oil companies with capital on hand. Currently, ConocoPhillips, Shell, ExxonMobil, and British Petroleum (BP), and other larger oil companies, are taking interest in the region, with some of them already in production. The building costs are high, but the rewards more than offset the cost. Each well is expected to cost $10 million to build, and is expected to generate $20 million in pure profits11. In addition to the $20 million profit, each well is expected to pay out $4.4 million in taxes, $1.6 million in salaries and wages, and $7.6 million in royalties. There are both positive and negative impacts to the Bakken region from this oil boom. Presently, there are over 8,000 active wells producing approximately 800,000 barrels of oil per day12. This increase has taken North Dakota from being the number 8 oil producing state in 2005, to the number 2 oil producing state, providing about 11% of the U.S. oil production. With additional wells being constructed at a rate of 3,000 per year13, big oil companies estimate that by 2015 total daily oil production could reach 1 million barrels14.

Unemployment is extremely low in North Dakota, especially in the oil region. The jobs that employ the area are not strictly from jobs related to oil and drilling, but support jobs such as building construction, road construction, over-the-road trucking, restaurants, and other local businesses also benefit from the drilling15. Oil production accounts for an estimated $34.4 billion of the states economy. While some states are dealing with fiscal shortfalls and constraints, North Dakota has a surplus of over $1 billion, approximately 25% of its budget. Real estate values are skyrocketing, with vacant homes and apartments scarce or non-existent, though 2,000-3,000 housing units are being constructed yearly16. In addition to housing trouble, there are also other infrastructure issues. There is only one hospital in the oil producing region, which sometimes has a 2 hour wait time, and the current road network is inadequate to handle the influx of traffic, though the road is being expanded from two lanes to six. These issues, along with plenty others will need to be addressed to compensate for the estimated population increase of 50% within the next 20 years. Data Research/Gathering One of the most important features that oil companies and geologists look for is the geologic composition of the area in question. The rock types that are being studied for potential oil reserves or traps are sedimentary rocks17, with an increasing emphasis on shale, particularly oil shale. The sedimentary rocks that are favorable for oil reserves are sandstone and siltstone. These are classified as sands, silts, clays/mud, which refers to the size of the grains rather than the composite of the rock itself. Not all shales contain oil, and with the ones that do, there are different types. Oil shale is a mixture of organic chemical compounds from which shale oil can
4

be produced; however, extracting shale oil is more costly than the production of conventional crude oil, as the shale needs to be heated to a sufficiently high temperature to separate the oil from the shale18. This differs from oil-bearing shales, which are shale deposits that have become impregnated with petroleum that can be produced from drilled wells 19. The Bakken Formation is an example of oil-bearing shales20. Another important feature that is highly desired among petroleum prospectors is a geologic formation called an anticline21. An anticline is a concave fold structure in which the rock beds tend to trap rising hydrocarbon fluids, allowing oil and gas to build up in the pore spaces of the rocks. Some of the other geospatial data that I gathered was land cover, active and inactive oil wells, federal and state land, national parks, and aquifer locations. Data Preparation To get the data in a useable format for the analysis, some of the data had to be created and modified. The creation of the anticlines and the Bakken boundary for North Dakota was accomplished by referencing a recent publication by the USGS. The federal and state land dataset was created by merging the Bureau of Land Management, Federal Forest Service, National Grasslands, National Wildlife Refuge, Reclamation Lands, State Forest, State Parks, and Wildlife Management datasets. The data was then reprojected to WGS-1984_UTM_Zone_13N. Most of the data was originally in GCS_North_American_1927. After the data was reprojected, the active and inactive oil wells dataset was created by separating them by attribute from the parent dataset. There were various well status attributes, so the wells with the status of: active, drilling, and location permitted to drill were extracted for the active wells. Wells
5

with the status of: abandoned, dry hole, inactive (shut in the last 3 -12 months), plugged and abandoned, and temporarily abandoned were extracted for the inactive wells. Wells with the status of expired permit, permit now cancelled, or permit now suspended were not included as it was determined that these attributes described an administrative issue, and not an oil production issue. The final step that was completed before beginning the analysis was converting all of the vector data to raster data. This was completed on the bedrock geology, surface geology, federal and state land, national parks, and aquifer locations. This didnt need to be performed on the anticlines or well locations as the first analysis step that was used for these datasets created a raster output, and the land cover was already in raster format. Raster Analysis The analysis that was selected for this process was raster analysis, more specifically the weighted overlay tool in ArcMap. The first analysis step was to perform a Euclidean distance analysis on the anticlines, active wells, and inactive wells. This tool calculates, for each cell, the straight line distance outward from the input data. The subsequent steps involved reclassifying, or changing the values, of all the raster data to assign ranks to each attribute. The following table is how each dataset was ranked:

The attributes were ranked on a scale of 1 to 9, with 1 being the most desirable attributes. In the bedrock geology dataset, there were attributes with multiple rock types. For the ranking on these, if there were multiple types in which they would have different weights, they were valued in between the two ranks. For example, shale was ranked as 3 and limestone as 9; so the attribute titled shale w/ local limestone, was ranked as 6. The rankings for surface geology were ranked the same as the corresponding attributes in the bedrock geology data. Land cover was ranked to provide preference to agriculture, scrub/shrub and grasslands, as these are easier to build on than forest or urban areas. Water and wetlands were classified as NoData. The NoData ranking will effectively render this area as non-existent to building wells, no matter what other layer is used. The proximity to anticlines was determined by finding out the typical width of the anticlines. This information was only located for one of the anticlines in the Bakken region, which stated it was 6 miles wide22. This was used as a base for

ranking the proximity to the anticlines. The proximity to the active and inactive wells was determined by ranking the proximity to be more favorable further away, with a greater distance needed on the inactive wells to receive the most favorable rank. Rankings for federal and state lands along with aquifers were determined by looking at the location of the current wells to see if there were any located in these areas. This showed that a few did exist over aquifers and inside federal and state lands, so it received a lower ranking, since there were some wells, but not many, located there. National parks will not have any oil wells located on them, so they were ranked as NoData where the National Parks are, and 1 for all other areas. Once the data had been reclassified, the final step was to use the weighted overlay tool. The purpose of this tool is to overlay several rasters using a common measurement scale and weights each according to its importance. The following table is how each dataset was weighted:

Research showed that geology and anticlines were the most important features of petroleum discovery, so these were weighted the highest at 20% each. The other weights were arbitrary and were weighted according to perceived importance. Note; however, that the National Parks dataset is weighted at 1%, as the NoData ranking would erase these areas from the final output. It didnt need to be weighted any higher to fulfill that result. Limitations to Analysis As with any type of analysis there were a few limitations, which if they had been resolved, could potentially yield a different result. For starters, research showed that rock core samples and seismic plots are usually obtained prior to any drilling. Rock core samples are valuable sources of information as they let the geologist find out exactly what rocks are present, and to identify the boundary between the various rock types23. Rock core samples can help determine the likelihood of oil being present underground. Seismic plots are one of the most important tools available24. It uses a source of energy, such as striking the ground with an iron plate or using dynamite, and then a receiver picks up the reflected sound waves. This method is used to determine the depth and orientation of rock formations. Another limitation is that only the bedrock and surface geology data were located. Having data of the geology between these two layers could have potentially been useful, as it would have shown the permeable and impermeable rock layers at different depths. This may have been an extremely difficult dataset to incorporate into the analysis, but it would have been interesting to see the data for possible inclusion.

As previously mentioned, the analysis relied on the assumption that the data, ranks, and weights are similar to what geologists use when prospecting for petroleum. The incorporation of additional data or changes in the ranks and weights which were used could potentially yield very different results, and changing the symbology on the map graphic can yield to different visual interpretations. Key Take-Aways Geospatial analysis concluded that the most suitable areas for developing additional wells are in the western portion of North Dakota, in close proximity to the anticline formations. Increased global demand and declining output from mature oil fields; along with continued unrest in the Middle East will strain oil resources; as production is reduced and there is increased difficulty and cost getting the oil to refineries for global shipment. The amount of suitable land and the estimated oil reserves in the Bakken region can help stem this strain, though it might be another few years before a real contribution is noticed domestically. This region is primed to enable North America to become a net exporter of energy; helping to reverse the current course of importing energy from a variety of countries worldwide.

10

Works Cited

Hansen, Brian. "North Dakota's Oil/gas Boom to Get Even Bigger, Official Says." Platts.com. Platts, McGraw Hill Financial, 05 Aug. 2012. Web. 29 July 2013. http://www.platts.com/latest-news/oil/Washington/North-Dakotasoilgas-boom-to-get-even-bigger-3246174. 2 "2012 Summary Annual Report." ExxonMobil.com. ExxonMobil, n.d. Web. 29 July 2013. http://www.exxonmobil.com/Corporate/Files/news_pub_sar-2012.pdf. 3 Williams, Mark. "Energy Demand, Supply and the Downstream Infrastructure Challenge." Shell.com. Shell Global, 03 May 2012. Web. 29 July 2013. http://www.shell.com/global/aboutshell/media/speeches-and-webcasts/2012/markwilliams-paris-03052012.html. 4 "Bakken Formation: News, Map, Videos and Information Sources." Geology.com. Geoscience News and Information, n.d. Web. 29 July 2013. http://geology.com/articles/bakken-formation.shtml. 5 Gaswirth, Stephanie B., et. al. "USGS Fact Sheet 20133013: Assessment of Undiscovered Oil Resources in the Bakken And Three Forks Formations, Williston Basin Province, Montana, North Dakota, and South Dakota, 2013." USGS.gov. United States Geologic Survey, 30 Apr. 2013. Web. 29 July 2013. http://pubs.usgs.gov/fs/2013/3013/. 6 "Bakken - Bakken Shale - Bakken Oil Map - Bakken North Dakota." Oilshalegas.com. N.p., n.d. Web. 29 July 2013. http://oilshalegas.com/bakkenshale.html. 7 "3 to 4.3 Billion Barrels of Technically Recoverable Oil Assessed in North Dakota and Montanas Bakken Formation25 Times More Than 1995 Estimate." USGS.gov. United States Geologic Survey, 10 Apr. 2008. Web. 29 July 2013. http://www.usgs.gov/newsroom/article.asp?ID=1911. 8 Huebsch, Russell. "Types of Oil Drilling | EHow." EHow. Demand Media, 16 Oct. 2009. Web. 29 July 2013. http://www.ehow.com/about_5539715_types-oil-drilling.html. 9 Svenvold, Mark. "How They Did It: Drilling a Horizontal Well." CNNMoney. Cable News Network, 13 Oct. 2008. Web. 29 July 2013. <http://money.cnn.com/2008/10/07/smallbusiness/horizonal_well.fsb/index.htm>. 10 "What Is Fracking?" Energyfromshale.com. Energy From Shale, n.d. Web. 29 July 2013. http://www.energyfromshale.org/hydraulic-fracturing/what-is-fracking. 11 "North Dakota's Oil Boom: 11 Cool Facts." MSNMoney. Microsoft Network, n.d. Web. 29 July 2013. http://money.msn.com/investing/north-dakotas-oil-boom-11-cool-facts. 12 Perry, Mark J. "AEIdeas." AEI-ideas.org. American Enterprise Institute, 27 Apr. 2013. Web. 29 July 2013. http://www.aeiideas.org/2013/04/north-dakota-oil-facts/. 13 Hansen, Brian. "North Dakota's Oil/gas Boom to Get Even Bigger, Official Says." Platts.com. Platts, McGraw Hill Financial, 05 Aug. 2012. Web. 29 July 2013. http://www.platts.com/latest-news/oil/Washington/North-Dakotasoilgas-boom-to-get-even-bigger-3246174. 14 "The Bakken Shale Oil Field - North Dakota." Bakkenshale.net. Bakkenshale.net, n.d. Web. 29 July 2013. http://bakkenshale.net/. 15 "North Dakota's Oil Boom: 11 Cool Facts." MSNMoney. Microsoft Network, n.d. Web. 29 July 2013. http://money.msn.com/investing/north-dakotas-oil-boom-11-cool-facts. 16 "North Dakota's Oil Boom: 11 Cool Facts." MSNMoney. Microsoft Network, n.d. Web. 29 July 2013. http://money.msn.com/investing/north-dakotas-oil-boom-11-cool-facts. 17 "Geologists Start With Sedimentary Rocks." Geomore.com. Oil On My Shoes, n.d. Web. 29 July 2013. http://www.geomore.com/sedimentary-rocks/. 18 "Oil Shale." Wikipedia. Wikimedia Foundation, 28 July 2013. Web. 29 July 2013. http://en.wikipedia.org/wiki/Oil_shale. 19 "Oil-bearing Shale." Oil-bearing Shale. Webref.org, n.d. Web. 29 July 2013. http://www.webref.org/geology/o/oilbearing_shale.htm. 20 "Oil Shale." Wikipedia. Wikimedia Foundation, 28 July 2013. Web. 29 July 2013. http://en.wikipedia.org/wiki/Oil_shale. 21 "Anticline." About.com. About.com Education, n.d. Web. 29 July 2013. http://geology.about.com/library/bl/images/blanticline.htm.

11

22

"Cedar Creek Anticline." High Plains Group. High Plains Group, n.d. Web. 29 July 2013. http://www.highplainsgroup.com/project-1/cedar-creek-anticline/. 23 "Rock Cores." Geomore.com. Oil On My Shoes, n.d. Web. 29 July 2013. http://www.geomore.com/rock-cores/. 24 "Seismic." Geomore.com. Oil On My Shoes, n.d. Web. 29 July 2013. http://www.geomore.com/seismic/.

Geospatial Data Sources

1) "Oil and Gas: ArcIMS Viewer." Dmr.nd.gov. North Dakota Dept. of Mineral Resources, n.d. Web. 29 July 2013. https://www.dmr.nd.gov/OaGIMS/viewer.htm. 2) "North Dakota Geologic Map Data." Mineral Resources On-Line Spatial Data. United States Geologic Survey, n.d. Web. 29 July 2013. http://mrdata.usgs.gov/geology/state/state.php?state=ND. 3) North Dakota GIS Hub Data Portal. State of North Dakota, n.d. Web. 29 July 2013. https://apps.nd.gov/hubdataportal/srv/en/main.home. 4) North Dakota Geographic Information Systems. State of North Dakota, n.d. Web. 29 July 2013. http://www.nd.gov/gis/apps/DataDownload/?clipping=Full. 5) Gaswirth, Stephanie B., and Et. Al. "Assessment of Undiscovered Oil Resources in the Bakken and Three Forks Formations, Williston Basin Province, Montana, North Dakota, and South Dakota, 2013." National Assessment of Oil and Gas Fact Sheet. United States Geologic Survey, n.d. Web. 29 July 2013. http://pubs.usgs.gov/fs/2013/3013/fs20133013.pdf. 6) "United States - Nationwide Data." United States - Nationwide Data. GeoCommunity, n.d. Web. 29 July 2013. http://data.geocomm.com/catalog/US/datalist.html. 7) North American Transportation Geospatial Data . Princeton University Library, n.d. Web. 29 July 2013. http://www.princeton.edu/~geolib/gis/canada.html.

12