Update and Review of Continental Conductive Surface Heat Flow Measurements in México: An Analysis of Deep Boreholes

Abstract

Conductive heat flow is an important parameter that is used to explain, directly or indirectly, several geological, geophysical and geochemical processes in the Earth´s interior. It is also one of the main input parameters for reliable estimations of resources related with geothermal and petroleum systems. That is because heat flow is used to describe subsurface temperature profiles and heat transfer mechanisms, thereby enabling the establishment of heat storage reserves in the case of geothermal systems and conditions of thermal maturation of organic matter in petroleum genesis. Since 2014, collection of data to estimate new continental conductive heat flow values in México has been an exhaustive scientific task. As a result, data from 4159 sites have been compiled, mostly from deep geothermal and petroleum boreholes. In this context, only 3,888 new geothermal gradient data were compiled and used to estimate new heat flow values. These new values complement the 702 continental heat flow values compiled and published between 1974 and 2021. Traditionally, all efforts to measure geothermal gradient in México have focused on the five high enthalpy geothermal fields under exploitation. Therefore, this continuous updating of the continental heat flow database would be an excellent input for Geothermal Play Fairway Analysis, enabling to define areas at a regional level with thermal anomalies and discovering new prospects, resulting in better knowledge of Mexican geothermal resources. Finally, the obtained data will help interested private and public entities to improve the geothermal exploration techniques in collaboration with academic institutions. Moreover, the scientific community interested in Earth science studies will benefit from this information with application to diverse research that involves the thermal evolution of the crust.

Introduction

For suitable estimation of energy resources in geothermal systems, a most realistic conceptual model must be developed, which depends on an assertive study and understanding of several geological, geophysical and geochemical parameters involved in such systems (e.g., Avellán et al., 2018; Jácome-Paz et al., 2019; Almaguer et al., 2020; Sena-Lozoya et al., 2020). A particularly important geophysical parameter is heat flow, which is required to estimate the heat stored in the subsurface, which in turn can be exploited to produce electrical power and/or for direct geothermal uses (Fuchs et al., 2020; Guerrero-Martínez et al., 2020).

Tectonic evolution makes México a privileged country regarding geothermal energy, i.e., large-scale geodynamic phenomena linked to subduction, extensional lithosphere activity, volcanism and major fault systems, among others, considered as some of the main originators of crustal thermal anomalies (e.g., Ferrari et al., 2007, 2012; Campos-Enríquez et al., 2019; Busby et al., 2020). Nowadays, only five developed Mexican high enthalpy geothermal fields (Cerro Prieto, Los Azufres, Los Humeros, Tres Vírgenes and Domo San Pedro) have an installed capacity of 962.7 MWe (Think GeoEnergy, 2023). However, the country can increase its potential to produce geothermal energy due to more than 4,000 thermal manifestations of low- and medium-enthalpy systems across the country (Iglesias et al., 2015), e.g., Los Cabos, Guaymas; Cuitzeo, Rancho Nuevo, (Hernández-Morales and Wurl, 2017; Almirudis et al., 2018; Gutiérrez-Negrín et al., 2020; Landa-Arreguín et al., 2021). Moreover, the Mexican geothermal potential probably can be accentuated by the existence of blind or hidden geothermal systems (Prol-Ledesma, 2000; Báncora and Prol-Ledesma, 2008; Faulds and Hinz, 2015; Dobson, 2016).

There is no trustworthy estimate of the actual geothermal potential of México, although some works have been published with related information through the study of heat flow, geothermal gradient or subsurface temperature maps (Blackwell and Richards, 2004; Iglesias et al., 2016; Prol-Ledesma & Morán-Zenteno, 2019). However, these records must be constantly updated. The heat flow pioneering works date back to more than 40 years ago (e.g., Smith, 1974; Smith et al., 1979; Reiter & Tovar, 1982; Ziagos et al., 1985). More recently, some Mexican research projects, focused on analyzing and estimating the geothermal reserves in México, have published heat flow measurements obtained from geothermal and petroleum exploration boreholes (e.g., Prol-Ledesma et al., 2016; Espinoza-Ojeda et al., 2017a, b; Prol-Ledesma et al., 2018).

Hence, this work´s main objective was to present an updated heat flow database for México, as well as updated geothermal gradient and heat flow maps. To accomplish this objective, 4,159 drilling reports of deep geothermal and petroleum boreholes were processed to compile thermal logs (transient borehole temperature (TBT) and bottom-hole temperature (BHT)) and rock formation stratigraphy. A rigorous quality assessment of the data produced 3,888 new heat flow values that were added to the 702 continental heat flow values compiled and published previously between 1974 and 2021 (Smith, 1974; Smith et al., 1979; Reiter & Tovar, 1982; Ziagos et al., 1985; Blackwell & Richards, 2004; Prol-Ledesma et al., 2016; Espinoza-Ojeda et al., 2017a, b; 2018; 2021).

Methodology

Technical reports from 4,159 sites were compiled, including geological and thermal geophysical data. All the information was provided by the Comisión Federal de Electricidad (CFE-Mexican Government Electricity Company) and Petróleos Mexicanos (PEMEX-Mexican Government Oil Company). The methodology proposed by Espinoza-Ojeda et al. (2017b) and Prol-Ledesma et al. (2018) which consists of data collection and processing to estimate new heat flow values was applied to the gathered data; the main tasks were as follows:

• (1) Update the stratigraphy database and the TBT and BHT data from drilled deep boreholes. The stratigraphic profile database was obtained from the drilling stage reports provided by CFE and PEMEX. TBT data include logs of temperature, depth and thermal recovery time. BHT was recorded at the deepest part of a borehole when the drilling process was stopped at different depths. BHT data were corrected using eight correction methods (Table 1): AAPG; FM; FMD; FMC; HC; KC; FM-SMU; and COG.

• (2) Update the continental heat flow database of México. The calculation of the new conductive surface heat flow values was carried out using the Bullard (1939) method. A statistical analysis to discard boreholes with convection disturbances was applied.

• (3) Elaborate new geothermal gradient and heat flow maps of México. The results of two interpolation methods and analysis of the interpolation error are presented here.

• (4) Geothermal Play Fairway Analysis. Application of this regional exploration method was attempted using the new heat flow data to define undiscovered prospective areas.

Table 1 Summary of BHT correction methods used in this work

Database

Figure 1 presents the locations of sites that comprise the new heat flow database created from the TBT and BHT records from 3,888 drilled deep boreholes. Also, 1107 published offshore sites (inside the exclusive economic zone) are included in Figure 1 to visualize their distribution and covered areas (e.g., Von Herzen 1963, 1964; Epp et al., 1970; Erickson et al., 1972; Henyey & Bischoff, 1973; Lawver et al., 1973, 1975; Lee & Henyey, 1975; Lawver & Williams, 1979; Williams et al., 1979; Becker, 1981; Lonsdale & Becker, 1985; Prol-Ledesma et al., 1989; Khutorskoy et al., 1990; Becker & Fisher, 1991; Sanchez-Zamora et al., 1991, 2013, 2021; Nagihara et al., 1996; Fisher et al., 2001; Blackwell & Richards, 2004; Rosales Rodríguez, 2007; Espinoza-Ojeda et al., 2017a; Neumann et al., 2017; Negrete-Aranda et al., 2022; Peña-Domínguez et al., 2022). The new database includes TBT and BHT measurements from 14 geothermal and 3857 petroleum boreholes, 17 of which are offshore; see Figure 2 for an illustration of the T–z (temperature–depth) logs of geothermal and petroleum boreholes used in this study. The rock lithology profile database was updated with the technical drilling reports that supplied 188 stratigraphic profiles: 29 from geothermal and 159 from petroleum boreholes. Figure 3 displays a histogram of the reported total depth of the compiled and analyzed boreholes in this study.

Figure 1

Updated continental heat flow database of México (orange filled triangles = new heat flow sites, blue filled circles = published heat flow sites) including 1,04 published offshore sites (gray filled circles) and 17 new sites (red-white triangles). Locations of the Mexican geothermal fields (Cerro Prieto, Las Tres Vírgenes, Domo San Pedro, Los Azufres and Los Humeros), currently under power exploitation

Figure 2

Plots of the temperature–depth profile (T–z) for six boreholes used for heat flow analysis. MEX0120-121 geothermal and MEX0152-0200-0309-1636 petroleum boreholes. BHT data corrected by different methods are included

Figure 3

Histogram of reported total depth in the compiled and analyzed boreholes in this study. n is number of data

BHT data are measurements logged during thermal disturbance generated by drilling process; therefore, they must be corrected to determine the formation temperature. The BHT collected data do not have an associated drill stem test (DST) record because formation temperature is obtained commonly from DST data; hence, calculation of formation temperature was through the application of few correction methods.

The correction methods applied for BHT (Table 1) are a function of depth and do not depend on other parameters (e.g., drilling mud circulation time, thermal recovery time, borehole size, some thermophysical properties of the formation rock or drilling fluid, etc.). These methods have already been evaluated and applied in various studies, and reliable results were obtained even in areas other than those for which they were determined (e.g., Deming, 1989; Förster et al., 1995; Crowell et al., 2012). In our case, we did not have enough data to calibrate those methods and it is likely that some methods would underestimate or overestimate the formation temperature after BHT correction. For that reason, we defined an "apparent" formation temperature as the average of corrected BHTs with the methods in Table 1 (see Fig. 2), thus:

$${\text{BHT }}\left( {{\text{corrected}}} \right) \, = {\text{ BHT }} + \Delta T$$

(1)

where BHT (°C) is the actual temperature logged during drilling process and ΔT (°C) is the differential of temperature to correct the BHT and, thus, to estimate formation temperature.

Conductive Heat Flow

Fourier’s law describes conductive heat transport phenomena in solid materials; therefore, the heat flow is calculated as:

$$q = - k\left( {{\text{d}}T/{\text{d}}z} \right)$$

(2)

where q is heat flow (W/m²), k is thermal conductivity (W/m K) and dT/dz is temperature gradient (also known as geothermal gradient: °C/m). In an ideal case, when a purely conductive heat transfer exists between different materials, thermal conductivity variations are present. These variations are considered in the thermal resistance concept. Hence, the heat flow for a layered medium is calculated as:

$$q = \frac{{\left[ {T\left( {z_{i} } \right) - T_{0} } \right]}}{{R_{i} }}$$

(3)

where T(z_i) is temperature (°C) logged at a specific depth, and R_i is thermal resistance (m²K/W) that is defined as:

$$R_{i} = \mathop \sum \limits_{i}^{n} \frac{{\Delta z_{i} }}{{k(z_{i} )}}$$

(4)

where Δz_i is thickness (m) of each layer and k(z_i) is the thermal conductivity (W/m K) of that layer. From Eq. 3, the Bullard (1939) method is derived to estimate heat flow for an ideal case, thus:

$$T\left( {z_{i} } \right) = T_{0} + q_{0} R_{i}$$

(5)

where theoretically, a plot between T(z_i) measurements and the summed thermal resistance R_i (m²K/W) must show linear tendency. Hence, an extrapolation of the straight line allows the slope (surface heat flow: q₀) and intercept (surface temperature: T₀) values to be determined. If some values do not follow the linear pattern, the linear relationship between T(z_i) and R_i is not valid; therefore, those values are defined as outliers or rejected data, which could be thermally disturbed by local convective effects, and most assuredly located within hydrothermal systems.

Commonly, a convective thermal disturbance can be thought of as a sudden or predominant warming (excessive rise) or cooling (temperature drop) of the temperature gradient, which would graphically observe as a "sharp" and major change in the slope of the gradient. In addition, heat flow affected by heat refraction was neglected in the country scale of this work due to the non-existence of considerable contrasts in thermal conductivities and/or geological structures heterogeneities (e.g., Jobmann & Clauser, 1994; Fulton & Saffer, 2009; Norden et al., 2012). In addition, as a test of linearity, the analysis of the coefficient of determination (R²) to evaluate the variation between T(z_i) and R_i to determine if linear regression model fits these parameters is suitable (i.e., R² ≈ 1) (e.g., Miller & Miller, 2000; Bevington & Robinson, 2003). Therefore, if linearity deviation is present, the borehole is disturbed by a convection flow, and a low R² is calculated. Hence, undisturbed boreholes have correlation between T and z [i.e., T–z plot] with an R² above 90% of confidence and, consequently, were considered for the construction of the heat flow database (e.g., Richards et al., 2012). Therefore, cases with R² below 90% of confidence are identified as disturbed thermal data and were discarded. All borehole data used in this study to calculate heat flow by the Bullard (1939) method satisfied the parameter of reliability R² ≈ 1 in the T–z and T(z_i)–R_i plots.

For illustration of the above principle, Figure 4 shows three geothermal boreholes. Figure 4a represents the T–z and the Bullard plot for the borehole MEX0121, respectively. This case could be considered as purely conductive heat transfer because the linearity between T–z and T(z_i)–R_i is undisturbed; the R² value supports this statement. In the case of borehole MEX0128 (Fig. 4c, d), a semi-linearity in the T–z and T(z_i)–R_i plots is evident. The calculated R² and most data fall inside the 90% confidence limits, validating the predominant linearity in borehole MEX0128. Finally, borehole MEX0137 (Fig. 4e, f) is a representative example of a borehole perturbed by dominant convective heat transfer at depth. This case is representative of the 20 rejected boreholes that were excluded in the conductive surface heat flow database. However, surface temperatures of MEX0121 (Las Tres Virgenes, Figure 1), MEX0128 and MEX0137 (Los Humeros, Figure 1) were not reported. Based on surface temperature of 25 °C as reference datum that was assumed according to the National Institute of Statistic and Geography (INEGI, 2021), a high surface temperature is observed in Figure 4, except 4a, which fits the slope of the T–z and T(z_i)–R_i correlation. This indicates, despite having conductive thermal behavior of the thermal profiles, a thermal anomaly of a geothermal area.

Figure 4

Borehole MEX0121: (a) temperature–depth profile; (b) Bullard plot. Borehole MEX0128: (c) temperature–depth profile; (d) Bullard plot. Borehole MEX0137: (e) temperature–depth profile and (f) Bullard plot

Additionally, Figure 5 displays three different cases of petroleum boreholes. Figure 5a shows that temperature increases linearly with depth in borehole MEX0152 (central north of México); the T–z linear relationship seems to be interrupted in the 2000–3000 m interval; even then, the linear tendency of the T–z and T(z_i)–R_i plots predominates in the R² analysis, where just one datum was rejected. In the second case, borehole MEX0200 (northeastern México: Fig. 5c, d) could be considered an ideal example of conductive heat transfer, according to the T–z plot and the R² result of the Bullard plot. In the third case, borehole MEX0309 (central north of México: Fig. 5e, f) presents semi-linear temperature increases with depth. This semi-linear relationship is apparently preserved between the temperature and thermal resistance in the Bullard plot (Fig. 5f). In these cases, in relation to their location, a surface temperature of 25 °C was assumed (INEGI, 2021).

Figure 5

Borehole MEX0152: (a) temperature–depth profile, (b) Bullard plot. Borehole MEX0200: (c) temperature–depth profile, (d) Bullard plot. Borehole MEX0309: (e) temperature–depth profile and (f) Bullard plot

As final examples, Figure 6 represents ideal cases of conductive heat transfer, where linear relationship predominates in the T–z and T(z_i)–R_i plots, and the results are confirmed by the R² values. The petroleum boreholes MEX1636, MEX2801 and MEX3360 are located in central eastern coast of Mexico; hence, a surface temperature of 17.5 °C was assumed (INEGI, 2021).

Figure 6

Borehole MEX1636: (a) temperature–depth profile and (b) Bullard plot. Borehole MEX2801: (c) temperature–depth profile and (d) Bullard plot. Borehole MEX3360: (e) temperature–depth profile and (f) Bullard plot

In summary, Figure 4 represents the geothermal sites with predominant semi-linear and nonlinear relationships between T–z and T(z_i)–R_i, indicating that most of them are perturbed by convection. As a result, 20 of the 37 analyzed geothermal sites were rejected. Petroleum borehole MEX0309 (Fig. 5e, f) is a representative example of petroleum cases with predominant semi-linear relationship between T–z and T(z_i)–R_i. Finally, Figure 6 illustrates the T–z and T(z_i)–R_i plots from the MEX1636, MEX2801 and MEX3360 petroleum boreholes. Most of these examples represent the petroleum sites analyzed in this work, which describe an evident linear correlation between T–z and T(z_i)–R_i.

Figure 7 displays the histograms of the percentage error for the geothermal gradient and heat flow estimations. It is shown that the relative errors are predominantly under 8% for the geothermal gradient estimations (Fig. 7a) and under 10% for the heat flow estimations (Fig. 7b). This empirical analysis supports the accuracy of the linearity analysis (R²) because the results obtained show very low relative errors, meaning that in most studied sites conductive heat transfer predominates.

Figure 7

Histograms of percent error between estimated slope and its standard deviation: (a) geothermal gradient relative error (GG) and (b) heat flow relative error (HF)

Interpolation

The development of the updated geothermal gradient and heat flow maps of México was achieved with interpolation data to describe different isotherms obtained from these two important parameters, according to their location and magnitude. Two interpolation methods were used to obtain the best results: inverse distance weighting (IDW) and empirical Bayesian kriging (EBK). The produced interpolation maps were constructed with SAGA software using 500 × 500 m cells.

IDW interpolation estimates the value of a point with a linearly weighted combination of a group of points (Bartier & Keller, 1996). IDW interpolation assumes that the variable being mapped decreases in influence at a greater distance from its sample location (Bartier & Keller, 1996; Setianto & Triandini, 2013). The IDW equation can be written as:

$$q_{x,y} = \frac{{\mathop \sum \nolimits_{i = 1}^{n} q_{i} d_{x,y,i}^{ - w} }}{{\mathop \sum \nolimits_{i = 1}^{n} d_{x,y,i}^{ - w} }}$$

(6)

where q_x,y is the point (unknown heat flow value) to be estimated at location x,y; q_i represents the control value (known heat flow value) for the ith sample point; d_x,y,i is the distance between q_x,y and q_i; w is the total number of predictions for each validation case defined by the user; and n is the number of control points nearest to the point being interpolated. By setting a high-power value, the closest points will have more influence for a better texture of the surface. As the power increases, the interpolated and the closest sample point values begin to approach each other. On the contrary, by specifying a lower power value, the surrounding points will take on more influence than those further away; this produces a smoother surface. Arguably, the optimal value for power is when the lowest minimum mean absolute error is considered (Bartier & Keller, 1996; Setianto & Triandini, 2013).

The EBK method provides an estimate at an unsampled site q_x,y (unknown heat flow value), based on the weighted average of adjacent observed sites z_i (known heat flow value) within the study area (Pilz & Spöck, 2008; Setianto & Triandini, 2013). The estimate of the weighted average given by the EBK predictor at q_x,y is defined as:

$$q_{x,y} = \mathop \sum \limits_{i = 1}^{n} \lambda_{i} q_{i}$$

(7)

where λ_i is the weight assigned to each q_i and n represents the number of sample locations.

Finally, the R² was also used to evaluate interpolation error. This parameter determines the amount of variation between actual heat flow measurements and values estimated by interpolation, and it allows to determine the most suitable interpolation (i.e., when R² approaches 1).

Geothermal Play Fairway Analysis

In the petroleum industry, Play Fairway Analysis can be defined as a methodology that incorporates geological data at regional or local scale to improve knowledge for locating and prospecting a probable hydrocarbon reservoir, or as a combination of petroleum fields that are characterized by homologous geological conditions (e.g., Eguiluz-de Antuñano, 2009; Gutiérrez-Negrín, 2015; Shervais et al., 2017). Thus, a Geothermal Play Fairway Analysis (GPFA) can be defined as an exploration approach that integrates regional data to demarcate geothermal plays and it is an important tool for locating blind geothermal systems. The combination of geoscientific parameters for defining the heat source of geothermal energy systems may include heat flow, faults and fractures network related with permeability, and tectonic evolution (Moeck, 2014; Moeck et al., 2015). GPFA incorporates parameters as layers in a geographic information system (GIS). This approach reduces the risk of failure by defining local areas, which have high probability of hosting exploitable geothermal systems (Moeck, 2014; Moeck et al., 2015; Shervais et al., 2017, 2020).

In this study, we proposed the analysis of an area with a high density of heat flow values located in a region controlled by the same set of geological conditions. We compiled geodynamic elements that indicate the presence of geothermal potential (Shervais et al., 2020): subsurface thermal anomaly or thermal regime (possible heat source), fracture-related reservoir permeability, and accessible fluids. Heat flow data, fault systems, geological and tectonic events (recent volcanism) on a regional scale were compiled in a GIS with a data layer per indicator. In this work, recent volcanism (age ≤ 3 Ma) was considered because geothermal systems are frequently linked to regions of both active tectonism and volcanism, young plutonism (≤ 3 Ma) and crustal thinning (Moeck, 2014; Moeck et al., 2015).

Results and Discussions

The Mexican heat flow database was updated with the newly estimated 3871 heat flow values from 702 previously published continental heat flow data; their locations are shown in Figure 1. As a result, the new database is an integration of 4573 continental heat flow measurements and 4337 geothermal gradient values (see Fig. 1). As important remark, the updated Mexican heat flow database contains 702 published (plus 1,084 offshore sites) and 3,888 new data (including 17 offshore sites). This database is compiled in a worksheet file, which contains two classification sheets, the “previous data” and the “new data.” An electronic worksheet file named “Mexico_HF_database (Espinoza-Ojeda et al., 2023)” can be downloaded from: https://github.com/omespinozaojeda/Mexico_HF_database.git.

Figures 8 and 9 show the histograms for the available geothermal gradient estimations and the published and new heat flow values, respectively. Due to the large amount of data used for this work, the data were grouped in two energy level subsets to enable visualization of the data variation. In Figures 8 and 9, the non-uniform distribution of the updated database for geothermal gradient and heat flow values is evident. In Figure 8a, the most numerous geothermal gradient estimations are located in a 25–55 °C/km interval, which is broadly related to the 40–140 mW/m² heat flow interval (Fig. 9a) due to the direct proportionality between geothermal gradient and conductive heat flow. The histograms illustrated in Figures 8a and 9a, once again, demonstrate that México has a large number of sites considered of low and medium heat flow. The high and very high geothermal gradient (Fig. 8b) and heat flow estimations (Fig. 9b), considered high heat flow sites, are located mainly in geothermal zones where convective heat transport occurs, as has already been shown in the works of Espinoza-Ojeda et al. (2017b) and Prol-Ledesma et al. (2018). The large amount of data in the updated heat flow database covers, alas unevenly, all the geologic provinces in México (see Fig. 1).

Figure 8

Histograms of updated continental geothermal gradient database of México. (a) Horizontal lines: values of 0–100 °C/km. (b) Diagonal lines: values of 100–500 °C/km. n is number of data

Figure 9

Histograms of updated continental heat flow database of México. (a) Horizontal lines: compiled heat flow values of 0–200 mW/m². (b) Diagonal lines: compiled heat flow values of 200–1300 mW/m². n is number of data

The realization of the updated geothermal gradient and heat flow map was enabled by the application of two interpolation methods. The reliability of the interpolation by IDW and EBK was measured through the correlation plots between the actual data values (heat flow and geothermal gradient) and the values estimated by each method (Fig. 10). In Figure 10, the R² is indicated for each analyzed case to identify the best interpolation method. From the correlation plots in Figure 10, it appears that the IDW method produced better interpolation results, with R² of 0.9605 and 0.9673, while EBK yielded R² of only 0.7935 and 0.8834 for heat flow and geothermal gradient, respectively. Thus, according to the results shown in Figure 10, the interpolation of all the available continental geothermal gradient and heat flow measurements in México was performed with the IDW method. Considering all the points within search distance, the weighting function Inverse Distance to a Power was used, and the selected power was 2. These parameters provided the best results in the evaluation; therefore, the model resolution of 500 m appears sufficient to resolve the continental trends.

Figure 10

Accuracy comparison between IDW and EBK interpolation methods through correlation plots between actual data (heat flow and geothermal gradient) and modeled data by each method. Obtained R² is indicated for each analyzed case

The Mexican geothermal gradient map is described in Figure 11, where zones with geothermal gradient above the continental average (≈30 °C/km) can be observed. Excluding the current geothermal fields under exploitation, known as zones with thermal anomalies, several zones with increased continental-scale thermal regime can be found, providing extensive areas for regional exploration. These zones contain gradients of > 40 °C/km, as illustrated in central and northeastern México like “hot spot” anomalies. In northwestern México, large zones with gradients of > 45 °C/km were highlighted where interesting prospects may be located.

Figure 11

Continental geothermal gradient map of México (updated: 2022). Locations of the Mexican geothermal fields (Cerro Prieto, Las Tres Vírgenes, Domo San Pedro, Los Azufres and Los Humeros) are shown as references of areas with known thermal anomalies

Lucazeau (2019) reported that the heat flow arithmetic mean varies from 62.4 to 68.4 mW/m² in the continental domain for oil data and non-oil data, respectively. These values clearly highlight the importance of high heat flow values that characterize large parts of México. The heat flow map (Fig. 12) shows anomalously high and very high heat flow values (≥ 200 mW/m²) that are assumed to be representative of the TransMexican Volcanic Belt (center of México), among which Domo San Pedro, Los Azufres and Los Humeros geothermal fields are currently under exploitation, while other geothermal prospects such as Cerritos Colorados and Las Derrumbadas are being evaluated. This thermal anomaly is related to local processes (e.g., volcanism and tectonics) and a significant convective component in the heat transfer. As a consequence, thousands of volcanic Middle Miocene to Holocene structures (some still active: ~ 5 Ma) and some intrusive bodies can be found (Ferrari et al., 2012) as well as active intraplate tectonic structures, resulting in magmatic intrusions as heat sources of the current geothermal fields, and lastly also several geothermal prospects (Gutiérrez & Aumento, 1982; Jácome-Paz et al., 2019; Reyes-Orozco et al., 2019; Mazzoldi et al., 2020; Olvera-García et al., 2020a, b; Sourisseau et al., 2020).

Figure 12

Continental heat flow map of México (updated: 2022). Locations of the Mexican geothermal fields (Cerro Prieto, Las Tres Vírgenes, Domo San Pedro, Los Azufres and Los Humeros) are shown as references of areas with known thermal anomalies

In southern (e.g., Sierra Madre del Sur) and southeastern México (e.g., Southeast basin and Yucatan Peninsula), no geothermal resources have been prospected. That is because low heat flow values (≤ 50 mW/m²) are predominant in the back-arc region of the Middle America trench and local hydrothermal activity related with El Chichón and Tacaná active volcanoes is too restricted to increase the regional heat flow. The arc magmatism has reflected differences in the southern thermal structure due to the predicted cold slab-mantle interface temperatures, which are lower than one-half of the initial mantle temperature. Along the slab-mantle interface, underthrusting removes heat from the upper plate and interface temperatures decrease (e.g., Ziagos et al., 1985; Prol-Ledesma et al., 1989; Peacock, 2003; Manea et al., 2005; Minshull et al., 2005; van Keken & King, 2005; Currie & Hyndman, 2006). There are medium heat flow sites in southeastern México (>60 mW/m²) and, in the Yucatan Peninsula, some could be characterized as geopressurized resources associated with oil deposits (e.g., Southeast basin) (Shann, 2020); these local anomalies might have originated from a combination of factors like radiogenic activity in the crust and upper mantle, and the Chicxulub crater formation could raise the mantle lithosphere causing an increased surface heat flow (e.g., Flores-Márquez et al., 1999; Wilhelm et al., 2004; Šafanda et al., 2005; Espinosa-Cerdeña et al., 2016).

Zones with geothermal resources from medium to high heat flow are predominantly concentrated in northern México (>70 mW/m²). The tectonic province of the Sierra Madre Oriental and the Sabinas and Burgos basins is located in northeastern México. The thermal alterations observed in this territory are derived from processes of sedimentation and subsidence in a basin, i.e., thermal history of the sedimentary strata is perturbed during the development of the basin, which is directly related to the predominant conductive heat flow coming from deeper strata in most sedimentary basins, and also related to the increase in burial depth with time (Cuevas Leree, 1985; Gray et al., 2001; Eguiluz-de Antuñano, 2009; Davison & Cunha, 2017). Some high heat flow sites that are widespread in the Sabinas and Burgos basins (> 100 mW/m²) are related with recent intraplate volcanism (Aranda-Gómez et al., 2005).

In the case of central northern México, high heat flow isotherms (> 100 mW/m²) are predominant, even though some low heat flow areas (≤ 60 mW/m²) can be observed (Fig. 12). The development of the Gulf of California is the result of the cessation of subduction of the Farallon plate beneath North America when the Farallon plate was consumed, and the spreading centers collided with the North American plate (Larson et al., 1968; Moore & Buffington, 1968; Fletcher & Munguía, 2000; Ferrari et al., 2017). This event produced geologic activity described as volcanism and faulting (Angelier et al., 1981; Suarez-Vidal et al., 1991; Nourse et al., 1994; Mora-Klepeis & McDowell, 2004; Busby et al., 2020). The opening of the Gulf of California produced extensional tectonics in northwestern México, creating the southern extension of the Basin and Range tectonic province by normal faulting (Hamilton, 1987; Henry & Aranda-Gomez, 1992; Aranda-Gómez et al., 2000; Ferrari et al., 2013). Some models of continental extension indicate that the great variety of styles may reflect different thermal states during the initiation of rifting (Buck, 1991). The estimated high heat flow patterns could be similar to those of the southern Basin and Range and the Rio Grande rift (Decker & Smithson, 1975; Reiter et al., 1978; Smith & Jones, 1979; Mareschal & Bergantz, 1990; Lachenbruch et al., 1994; Wisian & Blackwell, 2004).

Finally, we observe evident variations from low to high heat flow isotherms predominate in northwestern México (Fig. 12). These are consistent with several studies that conclude that the regions around the Gulf of California (Baja California Peninsula and Sonora State) have many sites with possible significant geothermal potential, and some areas to be developed as geothermal projects (Campbell-Ramirez et al., 1993; Barragán et al., 2000, 2001; Quintero et al., 2005; Arango-Galván et al., 2015). Four regions can be considered the most remarkable according to their energy levels described in Figure 12: Guaymas, Sonora (> 150 mW/m²; adjacent to the Guaymas basin) (Williams et al., 1979; Prol-Ledesma, 1991; Almirudis et al., 2018); tip of the Baja California Peninsula (> 150 mW/m²) (López-Sánchez et al., 2006; Hernández-Morales & Wurl, 2017; Prol-Ledesma et al., 2021); areas close to the Tres Vírgenes geothermal field (> 150 mW/m²) (Prol-Ledesma et al., 2004; Villanueva-Estrada et al., 2012; Leal Acosta et al., 2013; Leal-Acosta & Prol-Ledesma, 2016; Sena-Lozoya et al., 2020; Hernández-Morales et al., 2021); the northern part of the Baja California Peninsula (> 100 mW/m²), from the Pacific coast to areas around Cerro Prieto and close to the Salton Through (Lee & Henyey, 1975; Vidal et al., 1978; 1981; Chavez, 1987; Reyes-Lopez et al., 1993; Elders, 1996; Beltran-Abaunza & Quintanilla-Montoya, 2001; Arango-Galván et al., 2011; González-García et al., 2018; Aguilar-Ojeda et al., 2021). These thermal anomalies could be attributed to hydrothermal convection phenomena related with structures produced by the development of the Gulf of California. Within naturally permeable structures, the presence of convective cells allows to generate temperature anomalies at economically interesting depths (e.g., Fig. 4: MEX0137). Recent studies showed that the presence of fluid in an abnormally permeable area is a factor that allows the development of a geothermal resource, from a natural geothermal heat flow (Guillou-Frottier et al., 2020; Duwiquet et al., 2020, 2021, 2022). As a side note, various studies have found that variation in continental heat flow is related to the age of the lithosphere (Polyak & Smirnov, 1968; Vitorello & Pollack, 1980). Nevertheless, the age of continents is more complicated to estimate than the oceanic lithosphere, as it involves an overlap of thermal processes (rifting or crustal extension, magmatic intrusions, erosion, etc.) (Sclater et al., 1981; Stein & Stein, 1992).

In this study, we focused on a reliable selection of deep boreholes with thermal data and analyzed their influence on the current heat flow database mostly using data from sedimentary basins, where heat flow is estimated mainly from depleted/exploration petroleum boreholes. This confirms that data from petroleum boreholes allow for improved statistical analysis, as high-quality estimates of geothermal gradients are achieved with greater depth and/or more temperature measurements (e.g., Figs. 6, 7). With increase in the number of heat flow sites within sedimentary basin types, thermal anomalies are better defined, as most high heat flow sites are located in active extensional areas (rifts) and the lowest values are found in coastal plain areas (along the coast of the Gulf of México, excluding the Gulf coast central region that coincides with the TransMexican Volcanic Belt).

Although there are still large areas without available measurements, this had no negative effect on the statistics of the numerical prediction (interpolation), which showed the smooth variation in the thermal regime, e.g., the interpolation provided reliable heat flow estimates in the NW part of México because there are no significant variations in lithology of the region and the conductive heat transfer component is predominant. Unquestionably, the availability of more thermal data in those “gap” areas might improve the statistics for regional heat flow prediction. Figure 10 demonstrates that distinct choices of features can lead to quite different predictions even for regions without any constraining measurements.

The preceding discussions were the motivation for including the possible contribution of GPFA to this work. Figure 13 shows the geologic framework, which helps to describe the location of three different types of geothermal plays: regional geology to define sedimentary, metamorphic, plutonic and volcanic plays, and the main fault systems (Fig. 13a); correlation between thermal regime, recent volcanism (≤ 3 Ma) and play type (Fig. 13b); and correlation of thermal regime with the main fault systems (Fig. 13c) (Garrity & Soller, 2009; Padilla y Sánchez et al., 2013), and recent volcanism (≤ 3 Ma) (Duffield et al., 1984; Moran-Zenteno, 1986; Aranda-Gómez et al., 2000, 2005; Ferrari et al., 2005, 2007; Valdez Moreno et al., 2011; Global Volcanism Program, 2013; Macías & Arce, 2019; Prol-Ledesma & Morán-Zenteno, 2019).

Figure 13

Geothermal plays in México: (a) regional geology, tectonics (Recent volcanism), and fault system; (b) correlation between regional geology and thermal regime (conduction dominated heat transport); and (c) correlation between regional fault system and thermal regime (conduction dominated heat transport)

The northeast of México was selected to analyze the application of the GPFA, where the Sabinas and Burgos basins are located; this area is characterized by sedimentary rock formations, few recent volcanisms, but with an outstanding regional fault system. According to the works by Aranda-Gómez et al. (2005), Valdez Moreno et al (2011) and Padilla y Sánchez et al. (2013), it was possible to characterize the intraplate volcanism, regional geology and fault systems in Figure 14. From Figure 14, we observe that the relative high heat flow values (80–120 mW/m²) are directly linked to recent volcanism in the Sabinas basin (Aranda-Gómez et al., 2005; Valdez Moreno et al., 2011) and to the regional fault system of the Burgos basin (Eguiluz-de Antuñano, 2009; Padilla y Sánchez et al., 2013). Some of the intraplate volcanism is aligned along regional normal faults, suggesting that magma could have ascended through regional faults (Aranda-Gómez et al., 2005; Valdez Moreno et al., 2011). The GPFA application to the geological and heat flow data provides the highest likelihood of success in the search for prospects to exploit geothermal resources, even when data are sparse or incomplete. Additionally, the GPFA aids in the discovery of new geothermal resources. Therefore, the application of the GPFA to new heat flow data and the integration described in Figure 14 revealed that oil/gas producing areas in northeastern México can be considered as prospective zones with medium- to high-enthalpy geothermal resources.

Figure 14

Geothermal plays in northeastern México (Sabinas and Burgos basins): correlation between thermal regime (conduction dominated heat transport) with regional tectonics (associated with volcanism intraplate) and fault system. Red dotted lines represent limits of the northeastern México basins

While the heat flow maps published by Prol-Ledesma et al. (2018) and Prol-Ledesma and Morán-Zenteno (2019) show no differences, and although the analysis and description of the results were carried out for different purposes, the integration of Curie temperature data and hydrothermal manifestations data, together with direct measurements from the boreholes, might yield consistent maps for regional geothermal prospecting. The outcome, presented here, could be classified as a very precise map with values obtained only from direct thermal measurements in the boreholes. Therefore, this work provides almost 4,000 additional data to better define low, medium and high heat flow isotherms areas. Besides, the previously compiled and newly estimated thermal data are important contributions for discovering unknown low- and medium-enthalpy sites and provides the basis for the application of GPFA, thereby increasing the probability to identify new geothermal prospects. In addition, this work has a direct impact on both geothermal and petroleum industries, considering that most of the evaluated sites are petroleum boreholes and support the proposal of simultaneous geothermal/oil resources development. Therefore, it is a synergy of petroleum and geothermal industry.

Conclusions

The update of the Mexican heat flow database consists in the integration of 3,888 new data (3,871 continental and 17 offshore), the largest compilation carried out to date. This work represents a noteworthy increased number in the compilation of published and new geothermal gradients and heat flow measurements in México. The combination of these data improved the corresponding contouring maps. The updated geothermal gradient and heat flow database will be indispensable to carry out rigorous estimates of the heat reserves stored in the crust. The thermal anomalies reported in this work allow to locate promising areas for exploitation and direct use of geothermal energy. The increase in valuable data in the heat flow database can be combined with GPFA to discover new geothermal resources; the new prospects can be cataloged in an alternative way according to their geological characteristics and thermal regime. From the results, the predominant interval values for geothermal gradient and heat flow were 25–55 °C/km and 40–140 mW/m², respectively. In the north and central east coast of Mexico, there are large areas with wide variations in heat flow, from low to high values, which are an energetic attraction to be characterized by geothermal prospecting. In addition, the results obtained confirm that Mexico has great geothermal potential to be exploited through direct uses. As future work, constant update and analysis of geothermal gradient and heat flow measurements are necessary to reveal low-, medium- and high-enthalpy resources in large areas that remain uncharted.