Geographic information system (GIS)–based resource assessment is an important and relatively inexpensive tool for identifying areas that are of interest for geothermal power production. Of particular interest is the under-exploited industry of co-produced fluids and low-temperature formation waters in oil- and gas-producing basins. Obtaining bottom-hole temperature (BHT) data is now free and easily accessible due to the efforts of the National Geothermal Data System (NGDS). Oil- and gas-producing sedimentary basins in Colorado, Illinois, Michigan, and North Dakota contain formation waters of a temperature that is adequate for geothermal power production (90–150 °C) using existing binary power plant technology. While resource assessment gives a broad picture of the energy available in a basin, the problem remains of knowing where a power plant must go, and if it is economically feasible to do so in any given area. The Denver, Illinois, Michigan, and Williston sedimentary basins were evaluated using a play fairway analysis methodology to identify optimum locations for geothermal power production. These regions have been previously assessed for thermal energy in place, and geothermal gradients from that study, along with gravity anomaly information, magnetic intensity, and digital elevation models (DEMs) for slope analysis were incorporated into a geodatabase for map generation. Raster layers were created and then reclassified into nine classes each, with high geothermal gradient, low magnetic intensity, low Bouger anomaly, and low slope receiving the highest values. The layers were then weighted using a matrix weight assignment similar to that used in the Environmental Protection Agency’s DRASTIC water pollution model, and combined with the “Raster Algebra” tool in ArcGIS. Areas of greatest potential were identified and overlaid on a DEM layer. This shows locations where temperature will be highest at the shallowest depths in regions of soft sediments, refining the map creation process.

With increasing public awareness of our reliance on foreign fuels, it has become clear that energy independence is an issue of national security. We also face devastating climate change effects due, in part, to increased greenhouse gas emissions. Fossil fuels pose other threats to the environment, such as pipeline leaks and oil spills. Geothermal energy, therefore, has the potential to be an important part of our nation’s energy portfolio.

Co-produced fluids, the hot water brought to the surface along with oil during the pumping phase, are of special interest because this water is hot enough to flash a secondary working fluid in an organic Rankine cycle (ORC) binary power plant. Geothermal power plants built on enhanced geothermal system (EGS) reservoirs and using “closed-loop” cycles will produce near-zero carbon dioxide (CO2) emissions, one of the principal greenhouse gases (GHGs) implicated in global warming (Tester et al., 2006; Clark et al., 2012). In comparison with fossil-fueled, nuclear, or solar electric power plants, EGS plants require much less land area per megawatt (MW) installed or per megawatt hour (MWh) delivered (Tester et al., 2006). The benefits of EGS also apply to low-temperature geothermal and co-produced power production where existing wells are used, mitigating any potential environmental impacts from drilling. In addition, geothermal energy production has the advantage over other methods of renewable and/or sustainable sources in that it can be modulated to be either intermittent or base load, according to demand. Base-load power production means that energy is being provided to the infrastructure constantly, because the power-producing energy is always available.

Understanding sedimentary basin structure and the thermal energy contained within is an important first step to producing energy from co-produced and low-temperature resources; however, these analyses do little to show a potential producer or investor where appropriate well or plant locations exist. Other considerations that must be taken into account are: how deep is the appropriate reservoir, and how hot are those temperatures expected to be? Are these identified areas crystalline or softer sedimentary rocks, which affect potential drilling costs or extent of the reservoir? Are the sediments dense, which can be related to pore space, possible permeability, or lithology indicators? We propose using geothermal gradient calculated from corrected bottom-hole temperatures in conjunction with magnetic and gravity data and digital elevation models as a tool for assessment of heat distribution and to constrain locations where optimal conditions for heat extraction exist.

Resource assessments have been completed on four sedimentary basins in the mid-continent region: the Michigan and Illinois basins (Crowell and Gosnold, 2014), the Denver-Julesberg Basin (Crowell et al., 2013), and the Williston Basin (Crowell et al., 2011). These previous works have shown that, with current technology, geothermal production in the Michigan and Illinois basins is marginally feasible at best, but with the advent of future technology, power production may become economic.

Bottom-hole temperature data were obtained from the National Geothermal Data System (NGDS, 2014), corrected using the Harrison method (Harrison et al., 1983), and imported into ArcGIS. Included in the analysis were 36,861 wells for the Denver-Julesberg Basin (some well locations shown in Fig. 1), 6269 wells for the Illinois Basin, 9298 wells for the Michigan Basin, and 9332 wells for the Williston Basin. The wells were then sorted according to depth, and interpolation surfaces were generated for every 500 m down the stratigraphic column (Fig. 2). This information was used for assessment of energy in place, and was kept in the geodatabase for future reference. Geothermal gradients were then calculated using the following formula:

The average surface temperatures used for each state were obtained from the National Climate Data Center (NOAA, 2014), listed as: Colorado at 7.3 °C, Illinois at 10.97 °C, Michigan at 6.89 °C, and North Dakota at 4.68 °C. Once calculated, geothermal gradient values for each basin were interpolated using the kriging method and analyzed for areas in which hotter temperatures could be found at relatively shallow depths (Fig. 3).

Gravity and magnetic data were obtained from the Gravity and Magnetic Database of the U.S. hosted by the University of Texas at El Paso (University of Texas at El Paso, 2014). The robust data set included 232,129 magnetic and 46,535 gravity data points for Colorado, 191,079 magnetic and 106,420 gravity data points for Illinois, 376,256 magnetic and 68,092 gravity data points for Michigan, and 172,604 magnetic and 20,933 gravity points for North Dakota. In a gravity survey, low-gravity values indicate likely areas of thick sediments, whereas areas of high-gravity values indicate denser igneous or metamorphic rocks. High magnetic intensity values indicate rocks containing magnetite, indicating dense mafic rock, whereas low values indicate sediments or granite. Sediments have the lowest magnetic intensity values because few to no mafic grains are present. The data sets were added as layers into the ArcGIS geodatabase and were interpolated using the kriging method (Figs. 4 and 5). Slope was calculated from digital elevation model (DEM) layers (Fig. 6), comprising another geodatabase layer. Slope was considered because while geothermal power plants have a small footprint, attempting to construct a power plant on an area with any significant slope means increased construction costs for levelling the area.

The geothermal gradient, gravity, magnetic, and slope layers were reclassified using the ranges for all four basins on a scale of 1–9, with 9 being most desirable, to ensure compatibility between the basins and to prepare for raster algebra. Reclassification values are shown in Table 1. A weighting matrix, similar to that proposed by the Environmental Protection Agency’s DRASTIC model for evaluating water pollution (Babiker et al., 2005), was then calculated to determine what values to use within the raster algebra tool (Table 2). The raster algebra tool was run, and overall desirability values were calculated. The combination of weighted attributes, as shown in Table 2, results in raster cells indicating where slope, magnetic intensity, and gravity are lowest, and geothermal gradient is highest. The warmer colors in the resulting raster indicate areas where geothermal power plant placement would be optimal given these four variables. Once the desirability value, a unitless number, was calculated, the clip tool was used to remove areas with values below 3 on a scale of 1–9. The desirability cutoff of 3 was chosen because although geothermal gradient is not extremely high, it indicates where a temperature of 110 °C can be found within the first two kilometers of sediment in addition to favorable values for the other three attributes.

The Denver Basin

The Denver Basin is an asymmetric foreland basin that trends north-south, parallel to the Rocky Mountains. The entire basin, which spans Wyoming, Nebraska, and Colorado, has a surface area of ∼155,000 km2 (Martin, 1965; Curtis, 1988). The areas of high geothermal gradient appear to be fault controlled and lithologically controlled (Figs. 7 and 8), and these regions are surrounded by basement faults and outcrops of crystalline rock. Comparing the geothermal gradient to the magnetic intensity map (Fig. 9) reveals that the locations of interest are located above regions of relatively lower magnetic intensity for the region; however, the magnetite content in the nearby crystalline rock exposures is obvious. Examination of the regional gravity anomaly indicates less dense rock than the surrounding area, especially in the pink “hot spot.” The areas west of Denver are of interest because they are located near high population centers, and costly infrastructure is already in place. These high-population areas are near the depocenter of the basin and the Golden fault along the Front Range of the Rocky Mountains, where the hottest temperatures in the basin are recorded (Fig. 10). Colorado is required to provide 13% of its energy from renewable energy by the year 2020, and although it is currently providing 14% from renewable sources, none of the existing sources are base load (U.S. EIA, 2014). Geothermal power production is therefore of considerable interest in this state.

The Illinois Basin

The Illinois Basin is an asymmetrical cratonic basin, with a northeast- to southwest-trending axis. Located primarily within the state of Illinois, it has margins in Kentucky and Indiana, spanning a surface area of 155,400 km2 (Macke, 1995). Areas of high geothermal gradient appear in predominantly Cambrian to Pennsylvanian aged surface rock and do not appear to be fault controlled (Figs. 11 and 12). Relatively low magnetic intensity is found in the areas of interest, indicating little to no magnetite and therefore little crystalline rock (Fig. 13). In three of the five areas, the gravity anomaly is higher, indicating denser rocks at depth (Fig. 14) and perhaps shallower sediments in these areas. The depth required to reach temperatures usable with current power-generating technologies in Illinois is too great to make geothermal power production feasible; however, temperatures are sufficient for district heating and greenhouse applications. Even though the state has a desirability rating that reaches 8, these areas are small and isolated, possibly the result of localized fracturing. The state of Illinois has an average energy usage that is 44% higher than the U.S. average (U.S. EIA, 2014). Using geothermal reserves can provide considerable contributions to offset current energy needs.

The Michigan Basin

The Michigan Basin is a roughly symmetrical cratonic basin. The surface area includes the entire state of Michigan, Indiana, Ohio, and parts of Canada, over 308,210 km2 (Dolton, 1995). High geothermal gradient regions surround the rim of the basin, in predominantly Devonian and Mississippian age rocks and do not appear to be fault controlled (Figs. 15 and 16). The magnetic intensity of the study areas appears to be average, which would be expected in areas with thick sediment and little to no magnetite (Fig. 17). Analysis of the gravity anomaly map clearly shows the rift that runs through the middle of the state and, therefore, the high ferromagnesian mineral content in the area (Fig. 18). On either side of the rift, the gravity anomaly drops off in intensity. In Michigan, much like in Illinois, the depth at which temperatures can be found that are adequate for geothermal power production are too great to be feasible with current technology, but the potential for district heating and direct use is adequate. The mid-continent rift system, which runs through the center of the state, appears to be the major control of geothermal gradient. The weather is cooler in Michigan than most parts of the country (U.S. EIA, 2014), so energy savings with district heating, in particular, are beneficial.

The Williston Basin

The Williston Basin is an asymmetric cratonic basin that trends roughly north-south (Heck et al., 2010). The surface area includes primarily North Dakota, and to a lesser degree, Montana, Saskatchewan, and Manitoba, for a total of 133,644 km2 (Carlson and Anderson, 1965). Not much is revealed when comparing the geothermal gradient to the geology and surface expression of the basin (Figs. 19 and 20), indicating that the “hot spots” are most likely not fault controlled. The magnetic intensity map (Fig. 21) does show the areas of interest in soft sediments with little to no magnetite presence, showing a lack of near-surface crystalline rock. The gravity map (Fig. 22) is much more informative because it shows the presence of the mafic pipe-shaped intrusion in the north-central part of North Dakota, as well as the north-south trend of the granite greenstone terrain in the large area of interest. The oil boom in the North Dakota portion of the Williston Basin is straining the infrastructure, and with the energy needed to pump the oil wells increasing rapidly, the need for more produced power exists. The Nesson, Billings, and Little Knife anticlines appear to be the most significant structural controls on the geothermal gradient. North Dakota is also among the coldest states in the country (U.S. EIA, 2014); therefore, significant energy requirements exist before accounting for the excess need for oil drilling and pumping. The geothermal gradient is conveniently highest in the region where geothermal power production would be most needed. While the gradient points to low-temperature resources, these resources are sufficient for supplying smaller, portable ORC binary plants as discussed by Gosnold et al. (2013).

The Williston and Denver-Julesberg basins not only show promise for further feasibility studies but appear to have excellent potential. These reports and the following research assume an adequate, sustainable water flow is available.

Using the desirability value to help select optimal power plant locations is known as a play fairway analysis. Cost is one of the greatest barriers to geothermal exploration; therefore, desirability was determined by what variables influence the lowest cost. The state of Colorado clip using a desirability value cutoff of 3 included most of the state, which isn’t useful in determining power plant placement. In this case, a new raster using only values of 6 and above was included in the Colorado map (Fig. 23), which ranges from 6 to 9. Illinois has a desirability range from 4 to 8 (Fig. 24), Michigan has a desirability range from 3 to 6 (Fig. 25), and North Dakota has a desirability range from 3 to 4 (Fig. 26).

Assuming an adequate, sustainable water supply, the Denver-Julesberg Basin has the highest capacity for large-scale geothermal power production near population centers where infrastructure currently exists. While not all oil-producing basins have the capability to produce electricity from geothermal brines with current technology, such as Illinois and Michigan, the need exists for offsetting energy production from power plants, many of which run on coal. Other forms of energy use, such as district heating and direct use, have the capacity to offset present power production and should not be ignored. Low-temperature power production, such as that found in North Dakota, is also of use with smaller binary power plants that can generate power locally for use in drilling and pumping oil.