Multilayer geospatial analysis of water availability for shale resources development in Mexico

In 2013, Mexico passed eleven national structural reforms to modernize the country's economy. One of the most relevant is the energy reform approved in December 2013 [1]. Before the energy reform, Mexico's government monopolized the oil and gas industry for approximately 70 years, and the electricity sector for around 55 years. This reform ends this monopoly with the intent of boosting Mexico's energy industry by fostering competitiveness and private investment throughout the energy value chain [2]. In particular, in oil and gas projects, private investment is expected through service contracts, profit sharing agreements, and licenses [3]. Mexico's Oil Company (Pemex) will transform into a company of the state that will collaborate through profit and/or production sharing agreements [4].

Separately, the use of horizontal drilling and hydraulic fracturing (HF) in the United States enabled extraction from gas and tight oil shale reserves previously deemed uneconomic, and led to an increase in oil and gas production [5–7]. It is expected that the technologies could enable extraction of Mexico's shale resources, as well. Mexico is considered to be one of the top 10 countries in terms of technically recoverable shale resources in the world [8, 9], and its government expects to triple its current gas production by developing HF wells in its shale basins [10, 11]. Mexico's shale resources are located in 5 main basins which have combined risked technically recoverable reserves (RR) of 545 Trillion cubic feet (Tcf) of gas and 13.1 Billion barrels (Bbbl) of oil [12, 13], equivalent to approximately 636 Quadrillion British thermal units of energy (Quads) [14]. The 5 shale basins in Mexico, described in table 1 and shown in figure 1, are Burgos, Sabinas, Tampico, Tuxpan, and Veracruz. The Burgos and Sabinas basins have the largest RR in Mexico (69.3% and 20%, respectively). Furthermore, around 88% of the RR in the 5 basins are located in the shale gas areas.

Zoom In Zoom Out Reset image size

The use of HF carries environmental challenges due to (1) the water quantity it requires, and (2) the large volume of wastewater it produces [15, 16]. Impacts vary depending on the water availability and competing demands at a local level [17], but it is important to understand the sources and intensity of water used for HF to evaluate its full impacts on the water resources [18]. Water intensity for HF varies on several factors such as depth, type of shale resources, and thickness of each formation. In shale basins in the United States, HF has required between 1 × 10⁹ and 28 × 10⁹ gallons of water per Quads (gal/Quads) in gas areas, and between 1.5 × 10⁹ and 21.7 × 10⁹ gal/Quads in tight oil areas [19]. In the Eagle Ford Shale formation in Texas, which extends to Mexico, the average water consumption in the oil and gas shale areas, based on the wells estimated ultimate recovery, is 2.6 × 10⁹ and 1.5 × 10⁹ gal/Quads [20]. The estimated ultimate recovery (EUR) is defined as the approximate quantity of shale resource that could be recovered throughout a typical 20–30 year lifetime of a hydraulic fracturing well [20].

Due to the high water demand that HF could cause at a local level [21], different strategies should be analyzed to increase water availability and decrease water stress in the regions where shale resources are located. These strategies could include (1) shifting to less water intensive power plant technologies, (2) shifting to more efficient irrigation technologies, (3) using brackish groundwater as a source of water for HF, and/or (4) recycling and reusing wastewater (e.g. flowback and produced water) that will return to the surface over the lifetime of the HF wells, which in Texas shale areas ranges from 15% to 200% relative to the water demand of the HF wells [22, 23]. For instance, despite the potential increase of water consumption from HF in Texas, it has been shown that in switching from coal-fired to natural gas combined cycle, power plants could reduce water consumption by approximately 60% because of more efficient energy conversion and less water-intensive cooling systems at the power plant [24]. Also, it has been shown that it is possible to increase water availability in the shale regions by using the energy wasted in flared gas activities from HF to treat the flowback and produced water [15]. In 2012, these flared gas in Texas could have been used to treat between 180 and 540 million cubic meters of these wastewater, which is around 1% to 2.4% of Texas' annual water demand [23].

Mexico's water availability varies across the country. Mexico's Water Commission (Conagua) classifies the water availability of the aquifers and watersheds based on four zones that depend on a groundwater availability index (GWAI) and a surface water availability index (SWAI), respectively [26]. These indices are used as thresholds to assign a water price to users [25]. For both indices, the areas classified as Zone 1 represent the areas with greater water scarcity and higher water prices, while the areas classified as Zone 4 have plenty of water available and the water prices range from 10% to 15% of those in Zone 1 areas, depending on the use (e.g. municipal, industrial, agricultural) [25]. Figure 2 shows the classification of Mexico's aquifers and watersheds based on the indices stablished by Conagua. In the northern part of Mexico, most watersheds are classified as Zone 1, while closer to the Gulf of Mexico watersheds are classified as Zone 4. On the other hand, most of the aquifers overlaying the shale areas are classified in a better category (Zones 3 and 4). The aquifers with saline or brackish groundwater are reported every year by Conagua, but the breakdown of the volumes are not reported [27]. Of the aquifers overlaying the shale basins, the Bajo Rio Bravo (No. 1 in figure 2), Cuatrocienegas-Ocampo (No. 2 in figure 2), Paredon (No. 3 in figure 2), and El Hundido (No. 4 in figure 2) have brackish groundwater [27].

Zoom In Zoom Out Reset image size

Even though HF is considered a water-intensive activity, when compared with other high-water users, such as irrigation, it has been relatively small in Texas [21, 28]. Nevertheless, HF has been the major water-intensive user in some counties in Texas [17]. Thus, it is relevant to analyze the potential impacts that it could have on the water resources in Mexico and to quantify the water availability and source (e.g. surface or groundwater) in the different areas where HF wells will be employed.

The main contribution of this study is to develop a methodology to identify the spatial location and volume of water available in the aquifers and watersheds that overlay the five shale basins in Mexico. Furthermore, this letter describes data, methods, and results used to identify the potential EUR energy that could be extracted over the 20–30 year lifetime of HF wells supplied with available surface and groundwater. This research intends to serve as a starting point analysis for the development of strategies for managing new HF demands in the context of constrained water resources.

Following Conagua's methodology [31], the mean annual water available in the watersheds (MASWA_i) that overlay the shale areas is estimated using equation 1:

$$\begin{equation} {\rm MASWA}_i = {\rm Inputs}_i - {\rm Outputs}_i \end{equation} \tag{ 1 }$$

where Inputs_i and Outputs_i are the inputs and outputs of water to the watershed i, as shown in table 2.

The MASWA_iis assumed to be evenly distributed along each watershed i. A ratio is estimated to determine the mean annual surface water available (SWA_i) in the intersection of the watersheds and the shale areas using equation 2:

$$\begin{equation} {\rm SWA}_i = \frac{{{\rm SA}_k {\rm B}_j \cap {\rm A}_i }}{{{\rm A}_i }}{\rm MASWA}_i \end{equation} \tag{ 2 }$$

where A_i is the area of the watershed i, and SA_kB_j∩A_i is the area of the intersection of the watershed i and the shale area k in the shale basin j.

Three different alternatives based on the four zones of WA that Conagua uses to classify Mexico's watersheds are analyzed using the SWA_i [26, 27]. The SWAI_i for each watershed is estimated using equation 3:

$$\begin{equation} {\rm SWAI}_i = \frac{{{\rm Inputs}_i }}{{{\rm Outputs}_i }} \end{equation} \tag{ 3 }$$

The ranges of the SWAI used to classify Mexico's watersheds by zones are shown in table 3. These zones are based on the WA in each watershed on ascending order from Zone 1 to Zone 4, and are used to determine the water prices. As shown in table 3, the price per cubic meter of water in the watersheds classified as Zone 1, Zone 2, and Zone 3 are 766%, 299%, and 31% more expensive that in Zone 4.

The three different alternatives analyzed were:

• Alternative 1. This alternative evaluates the case in which all the SWA_i is supplied for HF in the intersection of the watersheds and the shale areas. Under this alternative, new water users could not be added, nor could current users could increase their demand. Also, the water price for all users would be more expensive. This alternative provides an upper bound.
• Alternative 2. This alternative evaluates the case in which part of the SWA_i is supplied for HF, but leaving a SWAI>3 (Zone 3). Under this alternative, new users could still be added or current users could increase their water demand, and the water price for all users would either remain the same or increase by around 31%.
• Alternative 3. This alternative evaluates the case in which part of the SWA_i is supplied for HF, but leaving a SWAI>9 (Zone 4). Under this alternative, new users could still be added or current users could increase their water demand, and the water price for all users would remain in the cheapest price.

Equation 4 was used to determine the volume supplied for HF in alternatives 2 and 3:

$$\begin{equation} {\rm x}_i = \frac{{{\rm Inputs}_i }}{{{\rm SWAI}}} - {\rm Outputs}_i \end{equation} \tag{ 4 }$$

where x_i is the water available to be supplied in watershed i by leaving a SWAI greater or equal to 3 or 9 for alternatives 2 and 3, respectively.

Following Conagua's methodology [31], the mean annual groundwater available from the aquifers (MAGWA) that overlay the shale areas were estimated using equation 5:

$$\begin{equation} {\rm MAGWA}_i = {\rm R}_i - {\rm NCD}_i - {\rm GWE}_i \end{equation} \tag{ 5 }$$

where R_i is mean annual recharge in aquifer i, NCD_i is the natural committed discharge of aquifer i, and GWE_i is the extraction of groundwater for other uses. The NCD_i is defined as the volume of water from the aquifer committed to springs and rivers, while the GWE_i is determined by the volume of groundwater assigned to other users by Conagua [31].

The MAGWA_i was assumed to be evenly distributed throughout each aquifer i. A ratio was estimated to determine the mean annual groundwater available (GWA_i) in the intersection of the aquifers and the shales areas using equation 6:

$$\begin{equation} {\rm GWA}_i = \frac{{{\rm SA}_k {\rm B}_j \cap {\rm AG}_i }}{{{\rm AG}_i }}{\rm MAGWA}_i \end{equation} \tag{ 6 }$$

where AG_i is the area of the aquifer i, and SA_kB_j∩AG_i is the area of the intersection of the aquifer i and the different shale area k in the shale basin j.

In similar fashion as for surface water, three different alternatives were analyzed using the GWA_i. These alternatives are based on the four zones of groundwater availability that Conagua uses to classify Mexico's aquifers using a groundwater availability index (GWAI) [25, 32]. The GWAI_i for each watershed is estimated using equation 7:

$$\begin{equation} {\rm GWAI}_i = \frac{{{\rm MAGWA}_i }}{{{\rm R}_i - {\rm NCD}_i }} \end{equation} \tag{ 7 }$$

The ranges of the GWAI used to classify the aquifers by zones are shown in table 4. These zones are based on the WA in each aquifer in ascending order from Zone 1 to Zone 4 and are used to determine the water prices. As shown in table 4, the price per cubic meter of water in the aquifers classified as Zone 1, Zone 2, and Zone 3 are 921%, 295%, and 38% more expensive than in Zone 4.

The three different alternatives analyzed were:

• Alternative 1. This alternative evaluates the case in which all the GWA_i is supplied for HF in the intersection of the aquifers and the shale areas. Under this alternative, new water users could not be added, nor could current users could increase their demand. Also, the water price for all users would be more expensive. This alternative provides an upper bound.
• Alternative 2. This alternative evaluates the case in which part of the GWA_i is supplied for HF, but leaving a GWAI > 0.1 (Zone 3). Under this alternative, new users could still be added or current users could increase their water demand, and the water price for all users would either remain the same or increase by around 38%.
• Alternative 3. This alternative evaluates the case in which part of the GWA_i is supplied for HF, but leaving a GWAI > 0.8 (Zone 4). Under this alternative, new users could still be added or current users could increase their water demand, and the water price for all users would remain in the cheapest price.

Equation 8 was used to determine the volume supplied for HF in the alternatives 2 and 3:

$$\begin{equation} {\rm y}_i = {\rm MAGWA}_i - {\rm GWAI}\;{\rm (R}_i - {\rm NCD}_i ) \end{equation} \tag{ 8 }$$

where y_i is the water available to be supplied from aquifer i by leaving a GWAI greater or equal to 0.1 or 0.8, respectively. For the aquifers with saline water (brackish or marine intrusion) the fresh and saline water are lumped together in the MAGWA. In these aquifers, the water available for new water intensive users, such as HF, could increase since the demand for saline water is typically smaller than for freshwater. Future research has to be conducted to estimate specific availability of saline water in the aquifers overlaying the shale areas.

Three scenarios of WA, described in table 5, for the shale areas in the shale basins were analyzed by intersecting the surface and groundwater availability alternatives.

Figure 4 shows the EUR energy that could be extracted in the 3 scenarios analyzed in the shale areas of the Burgos Basin. The shaded areas represent the areas where at least 80% of the WA is located in each of the shale areas for each scenario. For instance, in Scenario 1 and 2 the WA is equally distributed in the oil area, whereas in Scenario 3 there is no WA at all in this area. For all the scenarios, the WA for the Oil, Wet Gas 1 (northern) and Dry Gas areas is from groundwater, while around 95% of the WA for the Wet Gas 2 area (southern) is from surface water. In Scenario 3, on average an EUR of around 6 Quads could be extracted in this basin while leaving the classification of the water sources in Zone 4. Most of this energy could be extracted in the wet gas area closer to the Gulf of Mexico (Wet Gas 2), and a fewer from the dry gas and northern wet gas areas.

Zoom In Zoom Out Reset image size

Figure 5 shows the EUR energy that could be extracted in the 3 scenarios analyzed in the shale area of the Sabinas Basin. The shaded areas represent the areas where at least 80% of the WA is located in the Dry Gas area for each scenario. There is no surface water available in any scenario analyzed. In Scenario 3, on average an EUR of around 0.14 Quads could be extracted in this basin with the WA in the aquifers while leaving its classification as Zone 4.

Zoom In Zoom Out Reset image size

Figure 6 shows the EUR energy that could be extracted in the 3 scenarios analyzed in the shale areas of the Tampico Basin. The shaded areas represent the areas where at least 80% of the WA is located in each shale area for each scenario. For Scenario 1, most of the WA is from surface water, while in Scenario 2 and 3 the main sources of water varies for each shale area (table 7).

Zoom In Zoom Out Reset image size

As figure 6 shows, in Scenario 1 the maximum, minimum, and average EUR energy that could be extracted is the same in each shale area, which means that the average water that could be supplied to HF wells in one year could extract all the RR in these areas over the lifetime of these wells. In Scenario 3, on average an EUR of around 6 Quads could be extracted, while leaving the watersheds and aquifers classification as Zone 4. Most of this energy could be extracted in the Oil area closer to the Gulf of Mexico, followed by the Wet Gas and Dry Gas areas.

Figure 7 shows the EUR energy that could be extracted in the 3 scenarios analyzed in the different shale areas of the Tuxpan Basin. The shaded areas represent the areas where at least 80% of the WA is located for each scenario. For all the scenarios, more than 90% of the WA is from surface water. In Scenario 1, the maximum, minimum, and average EUR energy that could be extracted in the shale oil area of the Tuxpan Basin is the same, which means that the average water that could be supplied to the HF wells in one year could extract all the RR in these areas over the lifetime of these wells. In Scenario 3, on average an EUR of around 0.77 Quads could be extracted, while leaving its watersheds and aquifers classification as Zone 4.

Zoom In Zoom Out Reset image size

Figure 8 shows the EUR energy that could be extracted in the 3 scenarios analyzed in the shale areas of the Veracruz Basin. The shaded areas represent the areas where at least 90% of the WA is located in each shale area for each scenario. For all the scenarios, more than 95% of the WA in both shale resources type areas (Oil and Dry Gas) is from surface water. In the 3 scenarios, the maximum, minimum, and average EUR energy that could be extracted is the same in each shale area, which means that the average water that could be supplied to the HF wells in one year could extract all the RR in these areas over the lifetime of these wells. The main reason is due to the high WA in the region and the small amount of RR located in these areas. Most of the water is located in the south of the shale areas.

Zoom In Zoom Out Reset image size