Aires Etal 2018 BarragemMariana Landusechange

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/320947381

Changes in land use and land cover as a result


of the failure of a mining tailings dam in
Mariana, MG, Brazil

Article in Land Use Policy · October 2017


DOI: 10.1016/j.landusepol.2017.10.026

CITATIONS READS

0 87

5 authors, including:

Uilson Ricardo Venâncio Aires


Universidade Federal de Viçosa (UFV)
9 PUBLICATIONS 0 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Changes in land use and land cover as a result of the failure of a mining tailings dam in Mariana, MG,
Brazil View project

All content following this page was uploaded by Uilson Ricardo Venâncio Aires on 09 November 2017.

The user has requested enhancement of the downloaded file.


Land Use Policy 70 (2018) 63–70

Contents lists available at ScienceDirect

Land Use Policy


journal homepage: www.elsevier.com/locate/landusepol

Changes in land use and land cover as a result of the failure of a mining MARK
tailings dam in Mariana, MG, Brazil

Uilson Ricardo Venâncio Airesa, Bismarck Soares Matos Santosb, Clívia Dias Coelhoa, ,
Demetrius David da Silvaa, Maria Lúcia Calijurib
a
Universidade Federal de Viçosa, Departamento de Engenharia Agrícola, Campus Universitário, Viçosa, MG, 36570-000, Brazil
b
Universidade Federal de Viçosa, Departamento de Engenharia Civil, Campus Universitário, Viçosa, MG, 36570-000, Brazil

A R T I C L E I N F O A B S T R A C T

Keywords: The failure of the mining tailings dam in Mariana, Minas Gerais, Brazil, was the largest disaster of its type in
Environmental disaster Brazil, generating serious environmental and socioeconomic problems. The aim of this study was to identify
Geographic information system potential sites for the disposal of the tailings that accumulated along the watercourses affected by the dam failure
Multicriteria analysis in the municipalities of Mariana and Barra Longa, as well as to quantify the area affected by the tailings. To
quantify the area affected by the tailings dam failure, the Land Change Modeler (LCM) tool was used to verify the
changes in land use and land cover that occurred in the periods before and after the environmental impact. For
the identification of the suitable sites for tailings disposal, multicriteria analysis was performed using the Idrisi
Selva® software and considering three scenarios to identify potentially affected areas through the application of
the weighted linear combination (WLC) and ordered weighted average (OWA) techniques. In scenario 1, the
WLC technique was applied, considering average risk and high compensation. In scenarios 2 and 3, the OWA
technique was applied, with high risk and high compensation for scenario 2 and low risk and high compensation
for scenario 3. The highest suitability value for tailings disposal was identified in scenario 2, which obtained the
maximum value of 213 on a scale of 0–255, while for the scenarios 1 and 3, the suitability values were 158 and
124, respectively. The largest site for the tailings dam identified in the analysis for the municipally of Mariana
has a size of 0.25 km2 and obtained a suitability value of 207. In the municipally of Barra Longa, the largest site
identified has a size of 0.037 km2, with maximum suitability value of 209. In relation to the changes in land use
and cover, it was verified that the vegetated areas, composed mainly of riparian forest and pasture, presented
greater losses as a consequence of the dam failure, with reductions of 11.99 and 4.81 km2, respectively, de-
monstrating that the dam rupture disaster has caused serious damage to the environment and the population
living in the region.

1. Introduction dams (Azam and Li, 2010; Rico et al., 2008).


These dams are designed to last indefinitely; however, over recent
The waste generated in mining activities is of particular concern due decades, accidents related to large tailings ponds have caused serious
to the large quantities produced and the presence of toxic elements disasters, drawing public attention to the safety of this kind of facility
(Guerra et al., 2017). The storage of this waste generally occurs in (Sitharam and Hegde, 2016). The impacts caused by the failure of the
tailings dams, which are often built on steep slopes using the coarse dams present serious effects on the ecosystem and human health,
material derived from mining activities (Azam and Li, 2010). The safety mainly due to the rapid disruption to the site caused by the wastes and
of tailings dams is currently a challenge for engineering, mainly be- the long-term chronic contamination of the area, as well as the high
cause (i) the construction of dykes for tailings storage is done with the cost of recovering the affected areas (Dudka and Adriano, 1997; Gomes
residual material from mining activities; (ii) the height of the barrage et al., 2017). The disruption of waste dams frees millions of cubic
slopes is modified as the amount of waste increases; (iii) there is a lack meters of mining tailings to the environment (Gomes et al., 2017); for
of regulation for dam construction projects, especially in developing comparison, the 1998 dam failure in Aznalcollar, Spain, released 2
countries; (iv) there are high maintenance costs after the closure of the million m3 of toxic mud, mainly the chemical elements Zn and As, into
exploration area; (v) there is a lack of continuous monitoring of the the environment, resulting in the contamination of aquifers and surface


Corresponding author.
E-mail address: [email protected] (C.D. Coelho).

http://dx.doi.org/10.1016/j.landusepol.2017.10.026
Received 24 April 2017; Received in revised form 1 September 2017; Accepted 13 October 2017
0264-8377/ © 2017 Published by Elsevier Ltd.
U.R.V. Aires et al. Land Use Policy 70 (2018) 63–70

Fig. 1. Location of the area affected by the tailings dam failure used in this study.

water and the death of several aquatic species (Davies, 2002). A major life, increasing biochemical oxygen demand, and contributing to diffuse
disaster of this type also occurred in 2003 in the Republic of Macedonia pollution and contamination by pathogens (Bilotta and Brazier, 2008;
(Europe), releasing approximately 100,000 m3 of damaging tailings Stutter et al., 2017).
with elevated amounts of heavy metals (Ag, As, Cd, Cu, Mo, Pb, Sb and Aiming toward the environmental recovery of the affected areas, it
Zn) through the Kamenica River valley, which affected the water is crucial to remove the mining tailings that have been accumulating
quality of Lake Kalimanci (Vrhovnik et al., 2013; Rico et al., 2008). along the river banks and other affected areas and ensure their ade-
There are historical records of environmental disasters involving quate disposal so that this material is not carried to the watercourses,
dams or tailings pile failure in Brazil; however, none of them had the mainly through surface runoff in the rainy season. In this context,
same impacts as the Fundão dam failure on November 5th, 2015, which multicriteria analysis is a tool that can help in the decision-making
belongs to the Germano mining complex in the municipality of process in identifying potential sites for tailings disposal, since it takes
Mariana, MG and was the largest environmental disaster of this type in into account several criteria that delimit the most appropriate areas.
the country (MMA, 2016; ANA, 2016). Such analyses have been used for several purposes, such as the identi-
The dam contained 50 million m3 of iron tailings, of which 34 fication of sites for the implementation of sanitary landfills, which takes
million m3 were released into the environment. This amount reached into account criteria that are similar to those of tailings disposal
663 km of rivers and streams in the Rio Doce river basin in Minas Gerais (Demesouka et al., 2014; Ersoy and Bulut, 2009; Felicori et al., 2016;
and Espírito Santo states (MMA, 2016). Approximately 80 species of Gbanie et al., 2013).
fish are identified as native to this basin, of which 11 are considered Multicriteria analysis is a mathematical tool used to investigate a
endangered and 12 are considered endemic to the affected region series of alternative answers to a given question based on competing or
(IBAMA, 2015). The endemic species are the most threatened because conflicting priorities, providing a structured method for comparative
their development is dependent on the natural characteristics of the analysis and assisting in the decision-making process (Martinelli et al.,
environment that they inhabit (Burlakova et al., 2011). 2014). In this analysis, the criteria are divided into constraints or fac-
In addition to environmental damage, the failure of the dam also led tors, of which the constraints allow the imposition of absolute limita-
to severe socioeconomic losses. The district of Bento Rodrigues, one of tions to a given geographic space according to Boolean logic, that is,
the most affected areas, situated 5 km downstream from the dam, was values of 1 or 0 represent yes or no, respectively, to identify appropriate
practically buried by water, mud and debris. Nineteen fatalities were areas and eliminate unfit areas. Factors are relative limitations that
identified, with 600 people living at the site (ANA, 2016), and an es- determine the degree of fitness of a region of interest. They are con-
timated 207 of the 251 properties in the area were destroyed (IBAMA, tinuous in nature and represent the variation in aptitude within an
2015). established range (Campos et al., 2014; Felicori et al., 2016).
The failure of the dam also resulted in impacts to water quality, In this context, the aim of this work was to apply multicriteria
which impaired the public water supply in municipalities and for sev- analysis for the definition of potential areas that can be used for the
eral other water users around the affected area. Other uses of water disposal of mining tailings resulting from the rupture of the Fundão
resources, such as electric power generation, industrial use, leisure and dam, which accumulated on the river banks of the Mariana and Barra
fishing, were also harmed (ANA, 2016). For example, at the water Longa municipalities in the state of Minas Gerais, Brazil. In addition, it
quality monitoring station in Linhares, in the state of Espírito Santo, the is intended to verify the extent of the area affected by the tailings
observed turbidity value was on the order of 2500 NTU (CPRM, 2015). through comparison in relation to the land use and cover prior to the
A high concentration of suspended sediments in watercourses sig- environmental impact.
nificantly reduces light penetration, affecting the dynamics of aquatic

64
U.R.V. Aires et al. Land Use Policy 70 (2018) 63–70

2. Methods 2.3. Factors and criteria used to identify areas suitable for tailings disposal

2.1. Study area Elevation data were extracted from the digital elevation model
(DEM), with a spatial resolution of 30 m, generated from satellite in-
The work was carried out in the municipalities of Mariana and Barra formation (ASTER sensor) from the Shuttle Radar Topography Mission
Longa, state of Minas Gerais, Brazil, covering a total area of 1578 km2 (SRTM) (USGS, 2016). The DEM was used to obtain the information
(Fig. 1). The Fundão dam was built in the subdistrict of Bento Ro- related to the slope and hydrography of the study area.
drigues, located 35 km from the center of Mariana city and 124 km To obtain the hydrography of the area, the software MapWindow
from Belo Horizonte city. The second affected municipality was Barra was used, which enabled the generation of the hydrologic conditioned
Longa, in the Zona da Mata region, 172 km from Belo Horizonte. digital elevation model (HCDEM) through the IPH – Hydrotools tool.
This procedure was necessary to validate the hydrography extracted
from the DEM (more detailed) with the official Brazilian hydrography.
2.2. Changes in land use and land cover due to the failure of the tailings The dates related to the vegetation information, the geological
dam faults, the soil maps and the conservation units were provided by the
Spatial Data Infrastructure of Minas Gerais state (IEDE, 2016) at a scale
The land use and land cover in the region affected by the tailings of 1:50,000.
dam failure were obtained from Landsat 8 sensor OLI (Operational Land The road and highway database was extracted from the TrackSourse
Imager) images with a spatial resolution of 30 m, provided by the Project (2016) and compared to Google Earth Pro® images to confirm
United States Geological Survey (USGS, 2016) (https://earthexplorer. the locations of the identified roads and highways.
usgs.gov/). In this way, the definition of the potential areas for tailings disposal
An image from October 2015 was used as a representation of the was conducted based on the combination of information regarding
area prior to the environmental impact (which occurred on November paved roads, hydrography, slope, urban areas, geological faults, soils,
5th, 2015), and an image from November 2015 was used as a re- vegetation and conservation units.
presentation of the area after the environmental impact. Road information was used to determine which areas were most
To improve the resolution of the Landsat images, the multispectral suitable for tailings disposal in order to reduce transport costs.
bands were fused with the Panchromatic band to obtain the final re- Consequently, the areas closer to the roads resulted in the greatest
solution of 15 m. This procedure allows the attainment of more detailed suitability in the analysis results.
information about the study area. In the case of hydrography, the restriction of potential areas to a
For the classification of the land use and land cover, the maximum minimum distance of 200 m (buffer) from watercourses according to
likelihood supervised classification method was applied. Six classes of the Brazilian Standard (NBR 13896–Solid Wastes Landfill – Project,
land use and land cover were identified: Rock Outcropping, Pastures, Installation and Operation Criteria Procedure) was adopted as a cri-
Water Resources, Tailings, Vegetation and Urban Area, which were the terion in order to reduce the risks of these materials again occurring in
most visible in the selected images and of the greatest relevance for this the rivers. In this way, the further away from these areas, the greater
study. the suitability of the results obtained (ABNT, 1997).
The Land Change Modeler (LCM) tool in the Idrisi Selva® 17.0 In terms of slope values, potential sites were restricted to sites with a
software was used to verify the changes in land use and land cover due maximum slope of 30% in accordance with NBR 13896. Consequently,
to the failure of the tailings dam. To verify the accuracy of the classi- the suitability of the sites increased with a decrease in slope (ABNT,
fication, the Kappa index (K) was calculated (Eq. (1)). The samples 1997).
considered in the assembly of the confusion matrix were verified with A buffer of 500 m was established around the urban area for tailings
Google Earth Pro®, disposal, since this distance is also required for the construction of solid
waste landfills (nonhazardous) (ABNT, 1997) to avoid public health
y y
problems through the inhalation of fine particles of these materials that
n. ∑ x ii − ∑ (x i +. x+i )
can be carried by the wind.
i=1 i=1
K= y According to the Brazilian Standard (NBR 8419–Introduction of
n2 − ∑ (x i +. x+i ) Projects for The Sanitary Filling of Urban Solid Wastes), it is re-
i=1 (1)
commended that the construction of sanitary landfills includes a buffer
y of 200 m from geological faults; thus, this distance was also adopted as
where ∑ x ii is the sum of the diagonal of the confusion matrix, n is the a restriction in the definition of potential areas for the disposal of
i=1 mining tailings (ABNT, 1992).
total number of samples representing the soil use and cover observed in
The NBR 8419 recommends the implementation of the sanitary
the field, and xi+ and x+i are the sum of the rows and columns of the
filling of urban solid wastes only in areas with a permeability of less
confusion matrix, respectively. The interpretation of the values ob-
than 10−6 cm s−1 and an unsaturated zone with a thickness greater
tained for K were made based on the classification proposed by Landis
than 3 m. There is no information about the groundwater depth in the
and Koch (1977) according to Table 1.
study area; for this reason, it was decided to use sites with latosol soil,
which are deeper, with previous compaction work to reduce soil per-
Table 1
Interpretation of K values for the classification of the land use and land
meability.
cover. For vegetation (native or planted) and conservation units, it was
Source: Landis and Koch (1977). determined that the mining tailings disposal areas should be located
outside of vegetation areas, aiming toward the preservation of the en-
Values of Kappa Interpretation
vironment.
<0 No agreement Table 2 presents, in a synthetic way, the criteria established for the
0.00–0.19 Poor agreement analysis.
0.20–0.39 Fair agreement In the comparison of criteria at different scales, it is necessary to
0.40–0.59 Moderate agreement
normalize the dataset through pertinence functions, which obey fuzzy
0.60–0.79 Substantial agreement
0.80–1.00 Almost perfect agreement logic, converting the data to a scale that ranges from 0 to 1. In this
paper, linear functions were used, as the authors consider that these

65
U.R.V. Aires et al. Land Use Policy 70 (2018) 63–70

Table 2 combination of results from the analysis, so that the site with the best
Variables considered in the analysis and the respective criteria and constraints. suitability for tailings disposal has the highest score (Blachowski, 2015;
Source: Adapted from ABNT/NBR 13896: 1997–Solid Wastes Landfill: Project,
Romano et al., 2015).
Installation and Operation Criteria Procedure and ABNT/NBR 8419: 1992–Introduction
of Projects for Sanitary Filling of Urban Solid Wastes. The WLC technique does not always provide an adequate analysis
because it assumes average risk, not considering factors with a greater
Theme Function Control Points degree of limitations than others and presenting deficiencies when
applied in more complex areas. On the other hand, the OWA technique,
a b c d
in addition to allowing for the compensation of factors through the
Roads Decreasing 15 m 15 m 15 m 17,750 m application of weights, also controls the level of risk (Eq. (2)) of the
linear analysis and the degree of compensation (Eq. (3)), since it involves two
Hydrography Increasing linear 200 m 17,870 m 17,870 m 17,870 m sets of weights: factor importance criterion weights and order weights.
Slope Decreasing 0% 0% 0% 30%
In this way, risk levels depend on the position of the weights ordered in
linear
Urban Area Decreasing 500 m 500 m 500 m 30,560 m the ranking and the magnitude of their values, with higher values lo-
linear cated in the first positions representing lower risk and, on the other
Geological Faults Increasing linear 60 m 17,800 m 17,800 m 17,800 m hand, higher values in the last positions resulting in greater risks
Soil – Scale (0–255)
(Romano et al., 2015; Calijuri et al., 2007).
Vegetation – Scale (0–255)
Conservation – Scale (0–255) n
1 ⎡ ⎤
Units Risk = . ∑ (n − 1). Oi⎥
n − 1 ⎢ i=1 (2)
⎣ ⎦

present the best fit to the proposed analyses, and employed by Felicori n


1 2
and Marques (2017) in the identification of potential areas for the n. ∑
i=1
(O − )
i n
disposal of solid waste. Trade off = 1 −
The linear function has four control points (Fig. 2): the first position n−1 (3)
(a) indicates that the function starts to rise above 0, and the second where n is the total number of factors, i is the rank of the factor, and Oi
position (b) indicates that the function has reached the maximum value is the weight order for order factor i.
(1); the third position (c) indicates when the function starts to again fall Fig. 3 represents the decision strategy triangle of the techniques
below 1, and the fourth position (d) shows that the function has reached employed in the multicriteria analysis.
0 again. It is noteworthy that identified locations near the value of 0 in Fig. 3 indicates that the closer a value is to 1, the lower the risk used
the function present worse or better suitability depending on whether in the analysis. It is also noted that the WLC technique presents the
the function is increasing or decreasing, respectively. average risk. In this paper, the analysis was conducted taking into ac-
The analysis was performed for the purpose of identifying the best count three scenarios to define suitable areas for tailings disposal.
areas for the construction of tailings disposal, with a minimum size of Scenario 1 was established using the WLC technique, considering
0.02 km2. This will allow the disposal of the tailings in the most feasible average risk (0.5). For scenarios 2 and 3, the OWA technique was ap-
location from an environmental and economic point of view. plied, with high risk (0.36) and high tradeoff (0.80) for scenario 2 and
low risk (0.61) and high tradeoff (0.82) for scenario 3.
2.4. Analysis used to identify the suitable areas for tailings disposal
3. Results and discussion
The software Idrisi Selva® version 17.0 was used for the identifica-
tion of potential sites for the disposal of tailings derived from the dam 3.1. Change in land use and land cover due the failure of the mining tailings
rupture. The aggregation processes were conducted in the Decision Tool dam
Wizard environment using the hierarchical analytical process (HAP)
proposed by Saaty (1980) for the pairwise comparison and evaluation Using the Landsat 8 OLI images, it was possible to define six classes
of factors. of land use and cover in the study area: Rock Outcropping, Pasture,
In the criteria aggregation, two procedures were used, the weighted Water Resources, Areas Composed of Rejects, Urban Area and
linear combination (WLC) and the ordering weighted average (OWA). Vegetation. The latter group consists of both native and planted forests.
The WLC technique allows for the total compensation of the factors The results of the land use and land cover classification for the
through the application of weighted weights, for which an average risk municipalities of Mariana and Barra Longa in the period before and
is assumed. Using this method, the factors are combined by applying a after the dam failure can be found in Fig. 4.
weight to each of them, generating a map of suitability from the The classification corresponded adequately to the study area, with

Fig. 2. Fuzzy membership function; increasing (a),


descending (b).

66
U.R.V. Aires et al. Land Use Policy 70 (2018) 63–70

affected by the mining tailings in the municipalities of Mariana and


Barra Longa. The Urban Area was affected by 1.55 km2, and the re-
ductions occurred mainly in the district of Bento Rodriguez, which is
located near the dam (Fig. 1), and in the city of Barra Longa. On the
other hand, the gain in the Tailings category area was 21.02 km2, oc-
curring mainly around the banks of the affected watercourses and in the
district of Bento Rodrigues. The regions classified as Rock Outcropping
did not change during the analysis period.
Fig. 5b represents the percentage of area loss or gain in each class
for different periods before and after the tailings dam failure. The
Tailings category area had a gain of 33.33% in relation to its area in the
period before the rupture of the dam. On the other hand, the class that
presented the greatest reduction was Water Resources, with a decrease
of 8.7%, because the river channel is the preferred route for the flow
Fig. 3. Decision strategy triangle.
and consequently the deposition of the mining tailings. The Urban Area
class was the second in terms of suffering the most changes compared to
Kappa indexes of 0.83 and 0.86 in the images before and after the the period before the environmental impact, with a retraction of 4.52%.
environmental impact, respectively, which are considered excellent The Vegetation and Pasture categories presented the smallest changes
according to Lands and Koch (1997). The value obtained for the period in their areas, with decreases of 1.37% and 0.92%, respectively.
before the dam failure indicates that the classification represents 83% It was observed that the area of disturbed vegetation (11.99 km2)
of the land use of the region for that period. After the disruption, the obtained through the analysis of the change in land use and occupation
accuracy of the classification indicates that the current land use and represents 81.62% of the value reported (14.69 km2) by the Brazilian
cover in the study area is 86%. The biggest errors in the analysis were Institute for the Environment and Renewable Natural Resources
identified in the classification of the water resources because the pixels (IBAMA, 2015), referring to the area of affected vegetation in the entire
that represent the watercourses are more difficult to identify. Rio Doce river basin. The result shows that most of the vegetation area
To evaluate the changes in land use and land cover as a result of the that has undergone alteration is concentrated in the municipalities of
dam failure (Fig. 5), we used the LCM tool, aiming to verify the losses of Mariana and Barra Longa, because the propagation energy of the tail-
area in km2 (Fig. 5a), the percentage of land use and land cover changes ings after the rupture of the dam was higher in these localities.
within a class (Fig. 5b) and the percentage of the modified area in each As verified, the rupture of the tailings dam resulted in large im-
class (Fig. 5c). mediate impacts to the environment. In this context, the quantification
The largest loss of area found was in the Vegetation category of the change in land cover and use is essential for the adoption of
(Fig. 5a), with a reduction of 11.99 km2. The second most affected area measures aiming the recovery of the affected areas. This information
was Pasture, with a reduction of 4.81 km2. The reduction in Water can provide parameters to assist in the current waste management plan
Resources was 2.66 km2 and represents the watercourses that were of the tailings in the Rio Doce river basin, which suggests dividing the

Fig. 4. Classification of the land use and the land cover: (a) before and (b) after the tailings dam failure.

67
U.R.V. Aires et al. Land Use Policy 70 (2018) 63–70

Fig. 5. Changes in land use and land cover due to the


failure of the tailings dam: losses of area (km2) in
each class (Fig. 4a), the percentage of land use and
land cover changes within a class (Fig. 4b) and the
percentage of the modified area in each class
(Fig. 4c).

basin into stretches according to the degree of environmental impact where the maximum suitability value was 213. Scenarios 1 and 3 ob-
(Renova, 2017). tained suitability values of 158 and 124, respectively. The low value of
The impacts caused by the rupture of the tailings dam indicate the suitability identified in scenario 3 is related to the weights assigned to
great need for the increased surveillance and monitoring of dams; each factor, indicating that the low risk condition is not suitable for the
Morgenstern et al. (2016) note several failures in the construction analysis of the study area; this is because the best solution is not always
mechanisms of the Fundão dam as it increased in volume, resulting in associated with lower risk (Eastman, 2001).
its failure. It is estimated that there are approximately 126 tailings For the definition of the suitable areas for the disposal of the tail-
dams in Brazil that may present structural problems, evidencing the ings, the criterion employed by Calijuri et al. (2002) in the identifica-
need for effective management strategies in large-scale mining opera- tion of areas for the disposal of urban solid waste was used, in which
tions (Garcia et al., 2017). areas with suitability greater than 200 fulfilled the needs of the project.
Thus, we selected areas with a suitability value higher than 200, which
3.2. Multicriteria analysis used to identify the suitable areas for tailings are found in scenario 2.
disposal In Fig. 6, it is possible to verify the locations of the suitable areas for
the tailings disposal in the municipalities of Mariana and Barra Longa,
Based on the criteria established in Table 2, the areas suitable for with an emphasis on the larger areas.
tailings disposal were obtained considering a scale of suitability ranging The site (Area 1) identified for the disposal of the tailings dam in the
from 0 to 255, where the closer to the maximum value, the more sui- analysis is located approximately 1.0 km of the old Bento Rodrigues
table is the site of interest. district and 2.5 km from the site where the failure of the tailings dam
The best result was found using the information for scenario 2, occurred. The area selected in this municipality has a size of 0.25 km2,

68
U.R.V. Aires et al. Land Use Policy 70 (2018) 63–70

Fig. 6. Suitable areas for tailings disposal obtained in the municipalities of Mariana and Barra Longa using the analysis performed with the scenario 2 conditions.

and the maximum suitability reached at this location was 207. The conditions of the material, such as the water content, and the possibility
selected area is also located in one of the places most affected by the of carrying out compacting operations, considering the lack of areas for
tailings dam failure. residue disposal, particularly because of the difficulty in obtaining en-
The site (Area 2) selected for tailings disposal in the municipality of vironmental licenses (Nery et al., 2014). An alternative to this type of
Barra Longa is also the one that presented the largest size, at 0.037 km2. storage can be dry stacking, which is a method used to replace the
This site is located approximately 3 km from the urban area of Barra construction of dams. This methodology aims to use a filter to remove
Longa and obtained a maximum suitability of 209. the moisture from the tailings. In a study by Gomes et al. (2016), the
The removal of the remaining material in the river channels and in use of this methodology demonstrated the viability of the dehydration
the most affected areas is essential to avoid the continuous con- of iron ore tailings, in which a final moisture content of 15% was ob-
tamination of the watercourses. For comparison, the accumulation of tained. The results are applicable to dry stacking, in which the final
the remaining tailings along the river banks of up to 40 cm thick was moisture content of these materials is recommended to be between 20%
observed (Gomes et al., 2017). In the preliminary analysis of the che- and 25% (Bowker and Chambers, 2015).
mical and cytotoxic profile, highly mobilized fractions of Al, As, Ba, Fe,
Mn, Pb, and Sr were observed in the tailings samples, which had the 4. Conclusions
potential for cytotoxicity and DNA damage (Segura et al., 2016).
The definition of suitable areas for the disposal of waste can help in Based on the objectives proposed in this study and the results ob-
the current proposal for the management plan for the waste in the Rio tained, the following can be concluded:
Doce river basin. The management plan for the tailings proposed by
Renova (2017) includes the use of techniques to remove the tailings • It was possible to successfully identify potential areas for the dis-
from affected areas, such as the excavation, dredging and manual re- posal of mining tailings using multicriteria analysis.
moval of the tailings in the most sensitive areas, including the affected • The analysis of changes in land use and occupation due to the
region in the Bento Rodrigues district. In this context, the suitable areas rupture of the tailings dam provides information that can help in the
identified in the multicriteria analysis can provide essential information quantification of the impact caused and the adoption of measures to
for this project. recover the affected areas.
The appropriate disposal of the tailings that accumulated in the • The sites that presented the greatest potential for the disposal of the
affected areas was also one of the main measures for the recovery of the mining tailings were identified near the urban area of Barra Longa
areas affected by the Aznalcóllar dam rupture in Spain in 1998, with the city and in the Bento Rodrigues district in the municipality of
disposal of the tailings in suitable places that allowed the treatment of Mariana.
this waste (Olíes et al., 2012).
We recommended that the tailings disposal be conducted using the
stacking procedure, where the volume to be stored depends on the

69
U.R.V. Aires et al. Land Use Policy 70 (2018) 63–70

References mine, Brazil. J. Mater. Res. Technol. 5, 339–344.


Gomes, L.E.O., Correa, L.B., Sá, F., Neto, R.R., 2017. The impacts of theSamarco mine
tailing spill on the Rio Doce estuary, Eastern Brazil. Mar. Pollut. Bull. 120, 28–36.
ABNT, 1992. NBR 8419 – Presentation of Sanitary Landfill Urban Solid Waste Projects. Guerra, M.B.B., Teaney, B.T., Mount, B.J., Asunkis, D.J., Jordan, B.T., Barker, R.J.,
ABNT, Rio de Janeiro (in portuguese). Santos, E.E., Schaefer, A.E.G.R., 2017. Post-catastrophe analysis of the fundão tailings
ABNT, 1997. NBR 13896–Solid Wastes Landfill: Project, Installation and Operation dam failure in the doce river system, southeast Brazil: potentially toxic elements in
Criteria Procedure. ABNT, Rio de Janeiro (in portuguese). affected soils. Water Air Soil Pollut. 228–252.
ANA, 2016. Special Note on the Rio Doce Basin: Dam Rupture in Mariana – MG/National IBAMA, 2015. Preliminary Technical Report: environmental impacts resulting from the
Water Agency. Superintendence of Information Management, Brasília, Brazil (in disaster involving the rupture of the foundation dam, in Mariana, MG. Brasília (in
Portuguese). portuguese).
Azam, S., Li, Q., 2010. Tailings dam failures: a review of the last one hundred years. IEDE, 2016. Database of Metadata Geospatial of Minas Gerais. Retrieved from: http://
Waste GEO Techn. 50–53. www.iga.mg.gov.br/IEDE/ (Accessed 29 March 2016) (in portuguese).
Bilotta, G.S., Brazier, R.E., 2008. Understanding the influence of suspended solids on Landis, J.R., Koch, G.G., 1977. The measurement of observe agreement for categorical
water quality and aquatic biota. Water Res. 42, 2849–2861. data. Biometrics 31, 159–174.
Blachowski, J., 2015. Methodology for assessment of the accessibility of a brown coal MMA, 2016. Understand the accident of Mariana and its consequences for the environ-
deposit with analytical hierarchy process and weighted linear combination. Environ. ment. Retrieved from: http://www.brasil.gov.br/meio-ambiente/2015/12/entenda-
Earth Sci. 74, 4119–4131. o-acidente-de-mariana-e-suas-consequencias-para-o-meio-ambiente (Accessed 02
Bowker, L.N., Chambers, D.M., 2015. The risk, public liability, & economics of tailings May 2016) (in portuguese).
storage facility failures. Research Paper. Stonington, ME. Martinelli, L., Battisti, A., Matzarakis, A., 2014. Multicriteria analysis model for urban
Burlakova, L.E., Karatayev, A.Y., Karatayev, V.A., May, M.E., Bennett, D.L., Cook, M.J., open space renovation: an application for Rome. Sustain. Cities Soc. 14, 10–20.
2011. Endemic species: contribution to community uniqueness effect of habitat al- Morgenstern, N.R., Vick, S.G., Viotti, C.B., Watts, B.D., 2016. Fundão tailing dam review
teration, and conservation priorities. Biol. Conserv. 144, 155–165. panel. Report on the immediate causes of the failure of the fundão dam. (88 p).
CPRM, 2015. Special Monitoring of the Rio Doce Basin, Report 1: Follow-up of the flood Nery, K., Ávila, J., Scarmínio, M., Bittar, L., Moreno, R., Seno, R., 2014. Study of alter-
wave, first field campaign. Belo Horizonte (in Portuguese). native Technologies for residue disposal (Red Mud). Miner. Metals Mater Soc.
Calijuri, M.L., Melo, A.L.O., Lorentz, J.F., 2002. Identification of areas for sanitary 161–163.
landfills with use of strategic decision analysis. Informática Pública 4, 231–250 (in Olíes, M., Moral, F., Galván, L., Cerón, J.C., 2012. Groundwater contamination evolution
portuguese). in the Guadiamar and Agrio aquifers after the Aznalcóllar spill: assessment and en-
Calijuri, M.L., Loures, S.S.P., Santiago, A.F., Schaefer, C.R.G.R., Lugão, W.G., Alves, vironmental implications. Environ. Monit. Assess. 184, 3629–3641.
J.E.M., 2007. Identification of alternative areas for solid waste disposal in the Baixo Projet Tracksource, 2016. Download of Maps. Retrieved from: http://tracksource.org.br/
Ribeira do Iguape region – SP. Eng. Ambiental e Sanitária 12, 335–342 (in portu- (Accessed 23 April 2016) (in portuguese).
guese). Renova, 2017. Renova foundation define proposal for the rejects management in the Rio
Campos, P.B.R., Elmiro, M.A.T., Nobrega, R.A.A., 2014. Systematization of tacit knowl- Doce river basin. Retrieved from: http://www.fundacaorenova.org/release/
edge: identification of preferred corridors for transmission lines. Bol. Ciênc. Geod. 20, fundacao-renova-define-proposta-para-o-manejo-dos-rejeitos-na-bacia-do-rio-doce/
765–783 (in portuguese). (Accessed 10 August 2017) (in Portuguese).
Davies, M.P., 2002. Tailings Impoundment Failures: Are Geotechnical Engineers Rico, M., Benito, G., Salgueiro, A.R., Díez-Herreiro, A., Pereira, H.G., 2008. Reported
Listening? Waste GEO Technics, pp. 31–36. tailings dam failures A review of the European incidents in the worldwide contexto. J.
Demesouka, O.E., Vavatsikos, A.P., Anagnostopoulos, K.P., 2014. GIS-based multicriteria Hazard. Mater. 152, 846–852.
municipal solid waste landfill suitability analysis: a review of the methodologies Romano, G., Sasso, P.D., Liuzzi, G.T., Gentile, F., 2015. Multi-criteria decision analysis for
performed and criteria implemented. Waste Manage. Res. 32, 270–296. land suitability mapping in a rural area of Southern Italy. Land Use Policy 48,
Dudka, S., Adriano, D.C., 1997. Environmental impacts of metal ore mining and pro- 131–143.
cessing: a review. J. Environ. Environ. 26, 590–602. Saaty, T.L., 1980. The Analytic Hierarchy Process: Planning, Priority Setting, Resource
Eastman, J.R., 2001. Decision Support: Strategy Analysis. Idrisi 32: Guide to GIS and Allocation. Mcgraw-Hill, New York.
Image Processing, vol. 2 Clark Labs, Clark University, Worcester (22p). Segura, F.R., Nunes, E.A., Paniz, F.P., Paulelli, A.C.C., Rodrigues, G.B., Braga, G.U.L.,
Ersoy, H., Bulut, F., 2009. Spatial and multi-criteria decision analysis-based methodology Pedreira Filho, W.R., Barbosa Jr., F., Cerchiaro, G., Silva, F.F., Batista, B.L., 2016.
for landfill site selection in growing urban regions. Waste Manage. Res. 27, 489–500. Potential risks of the residue from Samarco’s mine dam burst (Bento Rodrigues,
Felicori, T.C., Marques, E.A.G., 2017. Multicriteria decision analysis applied in the se- Brazil). Environ. Pollut. 218, 813–825.
lection of suitable areas for disposal of solid waste in Zona da Mata Minas Gerais, Sitharam, T.G., Hegde, A., 2016. Stability analysis of rock-fill tailing dam: an Indian case
Brazil. J. Solid Waste Technol. Manage. 42, 47–64. study. Int. J. Geotech. Eng. 11, 332–342.
Felicori, T.C., Marques, E.A.G., Silva, T.Q., Porto, B.B., Bravin, T.C., Santos, K.M.C., 2016. Stutter, M., Dawson, J.J.C., Glendell, M., Napier, F., Potts, J.M., Sample, J., Vinten, A.,
Identification of suitable areas for the construction of sanitary landfills and sorting Watson, H., 2017. Evaluating the use of in-situ turbidity measurements to quantify
and composting plants in the mesoregion of Zona da Mata, Minas Gerais. Eng fluvial sediment and phosphorus concentrations and fluxes in agricultural streams.
Sanitária e Ambiental 21, 547–560 (in portuguese). Sci. Total Environ. 391–402.
Garcia, L.C., Ribeiro, D.B., Roque, F.O., Ochoa-Quintero, J.M., Laurance, W.F., 2017. USGS, 2016. Earth Explore. Retrieved from: http://earthexplorer.usgs.gov/ (Accessed 23
Brazil’s worst mining disaster: corporations must be compelled to pay the actual March 2016).
environmental costs. Ecol. Appl. 27, 5–9. Vrhovnik, P., Dolenec, T., Serafimovski, T., Dolenec, M., Šmuc, N.R., 2013. The occur-
Gbanie, S.P., Tengbe, P.B., Momoh, J.S., Medo, J., Kabba, V.T.S., 2013. Modelling landfill rence of heavy metals and metalloids in surficial lake sediments before and after a
location using geographic information systems (GIS) and multi-criteria decision tailings dam failure. Pol. J. Environ. Stud. 22, 1525–1538.
analysis (MCDA): case study bo, southern Sierra Leone. Appl. Geogr. 36, 3–12.
Gomes, R.B., Tomi, G.D., Assis, P.S., 2016. Iron ore tailings dry stacking in Pau Branco

70

View publication stats

You might also like