Surface and Underground Geomechanical Characterization of an Area Affected by Instability Phenomena in Zaruma Mining Zone (Ecuador)

: In the last decade, in the mining district of Zaruma-Portovelo, there has been signiﬁcant land subsidence related to uncontrolled mining activity. The purpose of this work was to carry out a surface and underground geomechanical characterization of a mining sector north of the city of Zaruma that allows the deﬁnition of potentially unstable areas susceptible to the mass movement. The methodology used consists of the following stages: (i) compilation of previous studies; (ii) surface and underground characterization of rocky material to establish its susceptibility to mass movement; (iii) interpretation of results; and (iv) proposal of action measures. Among the most relevant results, it stands out that 26.1% of the 23 stations characterized on the surface present conditions that vary from potentially unstable to unstable. In underground galleries, the studied mean values of the 17 stations indicate that the rock has a medium to good quality, representing a medium susceptibility to gallery destabilization. The results obtained for the surface areas (depths up to 50 m, where altered materials predominate) and the underground areas (depths > 50 m, where the alterations are speciﬁc) can be used to identify the areas with a more signiﬁcant potential for instability. For both cases, it has been possible to deﬁne speciﬁc monitoring, control, and planning actions for sensitive areas.

According to a number of studies [12,14,15], human activity also has a notable influence on this geological phenomenon type, mainly due to the induced modifications in the land. A particular case of these activities is mining, which, when properly developed, favors the economy and develops the regions where a resource is exploited. In uncontrolled mines or without the due technical considerations, adverse effects can be caused in environmental aspects and ground instability [16][17][18][19]. Subsidence is a frequent geological problem in extractive activities of solid resources [9,13,[20][21][22][23][24][25][26][27][28][29][30][31] and liquid resources [10,32,33]. If mining operations are developed in urban environments, it is necessary to carry out geological-geotechnical studies to establish this safety initiative. Thus, "urban geoscience" has been revealed as a new interdisciplinary field necessary in city planning and development to evaluate the geological environment and make possible the subsoil's strategic management [8].
The studied area corresponds to the north of Zaruma city (Ecuador). This locality is located in Southern Ecuador (Figure 1a), in the El Oro province, and it is part of the central gold-bearing districts [46][47][48]. This gold potential has allowed the development of intense and extensive mining activity through different types of exploitation. Thus, in the Zaruma environment, it is possible to identify both large legal mines and illegal mines (that are sometimes just a few dozen meters from the surface, without stabilization strategies). From the orographic point of view, the environment presents high slopes where mass movements linked to triggers such as heavy rainfalls, deforestation, seismic activity, civil works (e.g., roads), and mining are frequent. Since 2015, in addition to occasional landslides, there have been several subsidence events linked to the existence of illegal gold mining [16,38,[49][50][51]. One of the most relevant subsidence events was the one that originated the collapse of the La Inmaculada School in the center of Zaruma in 2017 [49,51]. This event caused an increase in restrictions on mining activity below the urban area of Zaruma (Figure 1b) defined in the decree of the mining exclusion zone (EZ) (Ministerial Agreement No. 2017-002 [52]), issued by the relevant body of the Ecuadorian state (Ministry of Mining) [53]. Due to the city's subsidence reports, in 2017, the exclusion area had to be increased, covering the entire city. Furthermore, to prevent illegal mining from weakening the area, the decree was modified again, prohibiting mining activity in approximately 2.15 km 2 around Zaruma city. The study area encompasses part of the last mining exclusion zone (Figure 1b). Zaruma was selected as the case study due to its importance as a gold deposit in Ecuador and its particular geologic, geomorphologic, climatic, and orographic conditions. The specific study area was chosen to consider the possibility of accessing underground galleries in current exploitation (outside the exclusion zone) and without exploitation (within the exclusion area) belonging to the BIRA S.A. mining company.
The aim of this study was to carry out a surface and underground geological-geomechanical characterization of an area of 0.7 km 2 in the north of Zaruma. The plan was to: (i) define areas susceptible to subsidence based on the geomechanical characteristics of the rocks; (ii) identify areas of high susceptibility where mining activity must optimize its exploitation procedure to guarantee stability underground and on the surface; and (iii) propose measures to be taken regarding mining activities in the underground and land use planning on the surface.

Study Area
The morphology of Zaruma is characterized by steep slopes, rounded mountain ranges, and heights between 1150 and 2800 m a.s.l. The predominant geoforms of the area corresponds to heterogeneous slopes, with medium to strong slopes [49].
The lithology within the study area (Figure 2a,b) consists of metamorphic, igneous, and sedimentary rocks ranging from the Precambrian-Paleozoic to the Quaternary age [55]. The basement rocks are metamorphic rocks (Triassic). Above them, massive andesitic lavas are unconformably intruded by small plutons of diorite to granodiorite composition (Lower Cretaceous). Felsic volcanic lavas, pyroclastics, and rhyolitic flows (Tertiary-Miocene) unconformably cap all of these units. Finally, Quaternary alluvial and colluvial deposits can be recognized along the Amarillo and Calera rivers [56]. According to Van Thournout et al., 1996, the mineralization in this area was originated by collapse and post-collapse rhyolitic activity. The Zaruma area (Figure 2b) is part of the epithermal vein system (Zaruma-Portovelo). This epithermal vein system is the result of hydrothermal processes close to a Miocene volcano that produced an andesitic to dacitic sequence. The identified ore and gangue mineral assemblages are typical for intermediate sulfidation epithermal gold vein deposits (Au ± Ag ± Cu) associated with Early Miocene continental arc magmatism [55,57]. The gold exploitation in Zaruma is trough underground mining used in epithermal deposits of in the vein system. The galleries are horizontal excavations into the side of hills or mountains, and levels are excavated horizontally off the decline to access the ore body. The distance between exploitation levels is approximately 30 m. In general, the deposit is extracted using the chambers and pillars method or a cut and fill method in which empty spaces are filled as the mineral is extracted. Geomorphologically, the most prominent feature is the rugged relief corresponding to the Southern Andes of Ecuador, characterized by the marked fluvial incision and the absence of stratovolcanoes [60]. Zaruma's climate is classified as subtropical, dry from May to November, and humid from December to April. In the study area, as occurs in sub-tropical environments, high rainfall during six months of the year originates essential chemical processes that generate deep weathering profiles. Regarding the vegetation cover, the area initially corresponded to a premontane forest. Currently, many of these areas are used for agriculture and livestock. The alteration and transformation of the rock into saprolite and residual soils represent an essential determining factor when evaluating land stability. Zaruma city occupies an area of approximately 1 km 2 and 10,000 inhabitants. The main economic activities are mining, agriculture, livestock, and tourism.

Materials and Methods
The procedure followed in this study is divided into three main different phases ( Figure 3): (i) information processing and systematization, (ii) surface and underground survey, and (iii) interpretation and evaluation of results.

Stage I: Information Processing and Systematization
The study began with the collection and review of previous studies carried out in the area [61][62][63][64], specifically, with a collection of data from boreholes, topography, and general geological maps provided, in most cases, by BIRA S.A. mining company. The fieldwork campaign was designed in order to complete the data needed to determine the area's state of stability at the surface and underground level. The characterization methods are detailed in Section 3.2.

Stage II: Surface and Underground Survey
This stage included the geological and geomechanical characterization of rocks on the surface (outcrops) [65] and underground (galleries) [34,35]. During the first sampling campaign, 23 stations were selected for geologic and geomechanical characterization through a study of outcrops. In the second sampling campaign, 17 stations were taken for geologic and geomechanical characterization through a study of galleries. The number of stations studied was limited due to budget constraints, access permits, terrain, and gallery conditions. This aim of this phase of the work was to obtain susceptibility maps of instability at the surface and underground level.

Surface Characterization
The surface study addressed the prominent outcrop characterization and 23 stations in the study area ( Figure 4). The number of stations and location depend on the favorable terrain conditions (existing outcrops and accessibility). At first, the primary lithology to be characterized in each sampling station was defined from the general lithology identified from the outcropping study. Next, through specific theoretical-practical methodology application used in landslides by various authors [65][66][67][68][69], the rocks' geomechanical characteristics were defined (Table 1) by assigning evaluations made following expert criteria. The selected parameters were lithology, geological structure, morphometry, discontinuity, water, vegetable cover, seismic activity, and weathering rank, which were evaluated from 0 to 4 (Table S1) based on the conditions observed in the field. This method allows the susceptibility level to landslide (Table 2) to be defined. The susceptibility level is a qualitative value (I to V), and it is related to the susceptibility coefficient (SC) estimated [70,71]. Thus, the grade or susceptibility level are: I for SC < 5; II for 5 < SC < 10; III for 10 < SC > 15; IV for 15 < SC < 20; and V for SC > 20. As a result, susceptibility maps of instability were prepared from the geomechanical evaluations obtained by using the geostatistical kriging tool that allows the values to be interpolated.

Underground Characterization
The underground characterization was carried out in mining galleries at depths greater than 50 m, with the permission of Bienes Raices S.A. mining company (BIRA S.A.). Specifically, this stage focused on the geological and geomechanical analysis of the rock samples from the 17 sampling stations ( Figure 5) located in the mining galleries. The location of the stations was selected based on the accessibility conditions and galleries that are outside the mining exclusion zone. As in the surface characterization phase, the stage was to identify the most representative rock types in each of the defined stations and their main characteristics (physical and chemical alterations). Next, an instability evaluation of the galleries was carried out using two evaluation methodologies applied for the 17 underground stations: (i) Q-Barton Index [34] and (ii) Mining Rock Mass Rating [35], which allow the comparison and validation of the rock quality results. The Mining Rock Mass Rating (MRMR) method was applied in order to validate Q-Barton method. Both methods allow zoning of the rock quality and its susceptibility level, based on the valuation interpolation using the kriging tool.

Q-Barton Index
The method proposed by Barton, Lien, and Lunde [34], evaluates the rocks' geomechanical behavior (Equation (1)) and establishes a qualitative classification of their quality (Table 3) through the use of six parameters: rock quality designation (RQD) [72,73], joint roughness number (Jr), joint set number (Jn), joint alteration number (Ja), joint water reduction factor (Jw), and stress reduction factor (SFR) [34] (Table S2). These parameters were obtained from field surveys in existing tunnels, which allow the issuance of criteria for the sizing of new tunnels. In order to be able to compare the surface and underground characterization, a parameter defined in the surface rocks (susceptibility level, Table 2) is also included in Table 3. The susceptibility level is a qualitative value (I to V) included in the analysis and inspired by previous work [74][75][76], the authors of which proposed a direct relationship between Q-Barton RQD and susceptibility. In addition, due to the lack of coincidence in the number of classes between Q- Barton (9) and the levels of susceptibility (5), the value of susceptibility III (3/5) was assigned to the fair value (5/9). Q = (RQD/Jn)*(Jr/Ja)*(Jw/SRF) (1) The MRMR method proposed by Reference [35] is used to perform rock mass valuation including aspects related to mining and combines several parameters for the mass valuation (RMR [77]), considering such criteria as mining-induced stresses (Cs), blasting effects (Cb), weathering (Cw), and joint orientation (Co) ( Table S3). Using the weighted sum of different parameters, and with Equation (2), the MRMR value was obtained. MRMR = RMR*Cw*Co*Cs*Cb, From the MRMR values obtained, the massif's behavior can be qualitatively characterized (Table 4). A parameter defined for the surface rocks was added to the qualitative qualification of the rock quality (susceptibility level, Tables 2 and 4) in order to enable comparisons between the surface and underground characterization. The susceptibility level is a qualitative value (I to V) included in the analysis and inspired by previous work [78,79], the authors of which proposed a direct relationship between MRMR and susceptibility. In this case, 5 levels of susceptibility were assigned to 5 levels of MRMR analysis.

Stage III: Interpretation of Results and Stabilization Proposals
Based on the susceptibility level to the mass movement (surface analysis) and rockfall (underground analysis), the obtained results (susceptibility level maps) were evaluated to establish recommendations for stabilization strategies in the most susceptible sectors.

Geological Characterization
The data collected in surface outcrops made it possible to determine the primary lithology. Thus, (i) altered rhyolitic tuff to the NE of the study area, (ii) porphyritic andesite and siliceous breccia to the S, and (iii) intrusive diorite to the NW were identified. The predominant structural trend in the area is NS (Figure 6a). The lithological reconstruction ( Figure 6b) using the data of previously carried out perforations in the study area shows that there are potent layers of andesite and porphyritic andesite at the base of this regular column without relevant alteration (thicknesses greater than 45 m). Overlaying these layers, an intercalation of altered tuff/tuff/tuff agglomerate (approximately 45 m) is identified. Over this intercalation, a layer of white fractured breccia (5 m) is identified, and finally, in the upper 5 m, weathered rock (saprolite) of brown color appears.

Lithological Sampling and Characterization
The detailed reconnaissance of the 23 stations allowed the detailed geological characterization through the identification, sampling, and analysis of rock samples (defined most often as rhyolitic tuff and porphyritic andesite), although with characteristics of saprolite and weathered rock. Table 5 presents the lithological characterization and the main description of these surface samples. The results obtained from evaluating susceptibility level in rocky massifs and altered rock indicate that: 26.1% of the stations present conditions that vary from potentially unstable to unstable. The remaining 73.9% are stable, and they need monitoring (Table 6 and Figure 7b). Figure 7a shows a zoning map of the studied area based on the obtained susceptibility conditions.

Geological Characterization
The lithologies identified in the 17 underground stations evaluated are described in Table 7. The primary lithologies identified, in addition to the ore materials (quartz veins of approximately 2 m), are andesite, diorite, microdiorite, and rhyolitic tuff. It was also possible to define the degree of fracturing and alteration of these rocks (Table 7) in this study. Table 7. Lithological characterization of underground samples.

1,6 Andesite
Slightly weathered and fractured andesite, with a greenish-grey hue. The rocky massif is fractured into blocks, with rough surfaces and little evidence of deterioration.

Microdiorite
Slightly deteriorated and fractured microdiorite, with rough surfaces without deterioration. Greenish-grey rock.

3,4,8 Andesite
Andesite without evidence of deterioration, of a greenish-grey hue. The rocky massif is moderately fractured, with fresh and rough surfaces.

5,7 Microdiorite
Microdiorite without deterioration, with a greenish-grey hue, presents few fractures with very rough surfaces and without deterioration.

9
Andesite with dioritic dike Slightly deteriorated rock with a light green hue. The rocky massif has block fracturing and rough fracture surfaces with ferrous spots.
10,13,16 Porphyritic andesite Slightly deteriorated rock with a greenish-grey hue, showing areas of mylonites. The rocky massif presents fracturing in blocks with rough surfaces and high deterioration.

11
Diorite Slightly deteriorated rock with a greenish-gray hue, the rocky massif presents intense fracturing with rough surfaces without deterioration.

12,17
Rhyolitic tuff Porous rocks slightly altered and deteriorated, with a pink hue.

Geomechanical Characterization
The evaluation of the underground rocky massif quality, through the analysis of the different mine stations using the Q-Barton method, indicates that 41.18% of the evaluated stations correspond to the rocky massif of good to very good quality (detachment susceptibility of II in Table 8), while 35.29% correspond to fair quality rock (detachment susceptibility III in Table 8), and finally 23.53% represent rock of poor quality (detachment susceptibility in Table 9) (Figure 8b). Figure 8a shows a map zoning the studied area based on the rock mass quality obtained, based on the Q-Barton method.   On the other hand, 64.71% of the rock mass evaluated with the MRMR method presents fair rock quality (detachment susceptibility III in Table 9), while 35.29% corresponds to good rock quality (detachment susceptibility II in Table 9) (Figure 9b). Figure 9a shows a map zoning the studied area based on the rock mass quality obtained by the MRMR method.

Interpretation of Results and Discussion
The geological characterization carried out, both in the outcrops and surface stations and in the underground stations in the study area, revealed two geological scenarios: first of all, an area denominated superficial (up to about 50 m deep, Figure 6b) where the materials present an evident weathering as a consequence, mainly, of an intense chemical alteration. This alteration does not prevent the original rock's recognition, although its physical properties are profoundly modified (Table 5), compromising its strength and stability. Just below this zone appears the so-called underground zone (depths greater than 50 m). In this deeper zone, the rocks do not present significant alterations, except in some specific areas, due to the hydrothermal activity that produced the deposit's mineralization. It should be noted that the depth of 50 m mentioned corresponds to the maximum depth at which, in this study, the presence of generally altered rock was detected. In some areas, it was found that the altered rock layer has small thicknesses.
In the surface stability analysis, the applied land stability evaluation methodology [65][66][67][68][69] made it possible to define a susceptibility map of instability in the study area (Figure 7). This contribution is considered very relevant because it allows characterization of the low land stability and, in particular, its low potential to support underground mining operations.
In the case of underground materials, the two methods used (Q-Barton Index and MRMR) could be used to obtain comparable results to zone the areas susceptible to instabilities (Figures 8 and 9). Thus, the Q-Barton method presents a land categorization with lower values (23.53% of the stations rock of poor quality; 35.29% fair quality; 41.18% good-very good quality) than the MRMR method (64.71% of the stations rock of regular quality; 35.29% good). Underground stations classified with the worst qualities in both methods are characterized by presenting conjugate fracture systems and water presence. As shown in Figures 8 and 9, the most detailed results are provided by the Q-Barton method.
Given the stability conditions defined in the zoning maps, it is advisable to propose actions that guarantee the safety of the area, preventing instabilities from affecting the population and infrastructures. These proposals are summarized below:

•
In surface areas with high susceptibility to instability (susceptibility levels IV and V), it is recommended to improve the surface drainage system, prepare a reforestation plan, and restrict urban development. Illegal mining activities must be stopped to avoid endangering the population.

•
In underground areas susceptible to instabilities (susceptibility level above III), it is recommended to improve the water drainage systems in all galleries, use better methods and materials to support gallery ceilings, and use the filling of sterile chambers in addition to the support pillars.

•
In the closest areas to the urban area, it is recommended to extend the study with additional surface and underground geomechanical stations to obtain a metric scale susceptibility model.

Conclusions
The application of the evaluation processes used allowed the zoning of areas with a potential risk of land instability in the mining environment of Zaruma. The information collected on the surface (outcrops and sampling stations) and in underground galleries (sampling stations) made it possible to verify that there is an area up to 50 m thick in which the altered rock predominates. It does not present guarantees of stability to carry out extractive works.
The susceptibility maps of instability phenomena (subsidence and landslides) generated in this work are adequate spatial planning tools. Thus, they can be used as a basis for the proposal of measures to limit the subsidence within the city's urban area and safeguard its population's lives.