UPDATE OF THE RESERVOIR INDUCED SEISMICITY OF BRAZIL - 2008
George Sand França
Adjunct Professor
Seismological Observatory – University of Brasilia
Brazil
Mônica Giannoccaro Von Huelsen
Visiting Professor
Seismological Observatory – University of Brasilia
Brazil
Marcelo Assumpção
Full Professor
Institute of Astronomy, Geophysics and Atmospheric Sciences - University of São
Paulo
Brazil
Luís Carlos Ribotta
Researcher
Institute from São Paulo of Technology
Brazil
Cristiano Naibert Chimplignond
Researcher
Seismological Observatory – University of Brasilia
Brazil
1.
INTRODUCTION
Reservoir Induced Seismicity (RIS), also referred to as Reservoir Triggered
Seismicity (RTS), is the triggering of earthquakes by the physical processes that
accompany the impoundment of large reservoirs. Currently, there are diverse
purpose of hydroelectric projects and Small Hydroelectric Station (SHS), than it is
necessary to understand the nature of the RIS that has been observed to date,
particularly in Brazil. Towards that end, in this paper we review the lithology of
hypocentral areas where RIS has been observed, their filling history and reservoir
characteristics.
In 1984, Berrocal et al., shown the first compilation of reservoir induced seismicitty
(RIS) in Brazil with 7 cases (In Portuguese). In 1992, Gupta shown for world nine
cases, Gomide (1999) analyzed 15 cases and Assumpção et al (2002) analyzed
16 cases of RIS in Brazil with discussion for Precambrian Shield. They shown that
the RIS didn't uniform and there are no correlation between RIS hazard and the
lever natural seismicity.
The last information about RIS from Brazil was in 2000, after that the several
cases RIS occurred in Brazil without information. The following has the same idea,
this present update summary of Brazilian cases may reinforce or the last the
discussion about RIS hazard levels.
2.
REVIEW
Gupta (2002), in his review of studies on RIS, he highlights the following
points: There are over 90 identified sites of earthquakes triggered by the filling of
water reservoirs. The largest and most damaging earthquake triggered by a manmade reservoir was in 1967 in Koyna, India. The magnitude of the earthquake was
6.3. For him, the depth of the water in the reservoir is the most important factor in
RIS. The volume of the water also pays a significant role in triggering an
earthquake. The magnitude of the foreshock is higher than the magnitude of the
aftershock and both values are generally higher than in cases of natural
earthquakes. RIS can be immediately noticed during filling periods of reservoirs.
The effect of RIS can be rapid (following the initial filling of the reservoir) or
delayed (occurring later in the life of the reservoir).
In accordance with Jauhari, 1999, the most widely accepted explanation of how
dams cause earthquakes is related to the extra water pressure created in the
micro-cracks and fissures in the ground under and near a reservoir. When the
pressure of the water in the rocks increases, it acts to lubricate faults which are
already under tectonic strain, but are prevented from slipping by the friction of the
rock surfaces.
3.
RIS IN BRAZIL
Several cases of Brazilian reservoir where induced activity were describers
in table 1 and table 2. Besides the 19 confirmed cases, eight other cases with
doubtful seismicity have been included for the sake of completeness.
The Fig. 1 illustrates the locations of RIS in Brazil. The height of the dam,
maximum water, reservoir volume, dates of impoundments, predominant geology
and Regional stress direction are given in Table 1. The longitude, latitude,
maximum magnitude, reservoir, state, date of the largest event and type of
seismicity are given in Table 2. The magnitudes adopted in this paper is the same
Assumpção et al., 2002: mb for teleseismic events, mR (regional scale) for events
recorded in the range 200 to 1500 km by Brazilian stations (Assumpção, 1983), M L
(local magnitudes) and mD (Duration magnitude) for small events.
The largest induced reservoir in Brazil was magnitude 4.2, associated with two
reservoirs Porto Colômbia and Volta Grande, in November, 1973. In General, the
reservoirs shows the an initial type and sometimes correlating with water level
fluctuations, which could be classified perhaps as delayed behavior (for a detailed
discussion, see Assumpção et al, 2002).
Most of the reservoirs are in granitic region and the strike-slip faults predominant,
but these characteristics are not so real because in general in Brazil has the same
lithology.
There are about 370 reservoirs in Brazil with a volume of 0.2 km 3 or greater, of
these 19 (~ 5 %) are associated with RIS.
Following the RIS near Irapé Reservoir (No. 18; Table 1) (Nascimento P. et al.,
2008).
4.
RECENT CASE OF RIS - IRAPÉ
Irapé Reservoir has the highest Brazilian dam, width 208 meter high, and is
located at Minas Gerais State, in the Jequitinhonha River. Since September 2003,
nine months before the impoundment of the lake, the seismicity in the region
started to be monitored. During this two years pre-impoundment monitoring no
local event was detected at the area of the lake. The filling of the reservoir started
in December 7, 2005, and was followed by microearthquakes just one day after.
Until October 2006 more 300 events wore detected by this local network (Fig. 2).
The largest microearthquake occurred in May 14, 2006 with 3.0m b magnitude. The
clear time correlation between the start of impoundment of the lake and the
occurrence of seismicity shows a causative relationship for this seismicity (Fig. 3).
Also the spatial distribution of the epicenters let us to conclude that this is another
case of reservoir-triggered seismicity in Brazil of the initial response type.
5. DISCUSSION AND CONCLUSIONS
Brazil has 19 confirmed cases of RIS (Tables 1 and 2). The Fig. 1 is the
update RIS (similar to Assumpção et al., 2002).
Assumpção et al. 2002 and Marza et al. 1999, are list several characteristics,
despite the small number of cases, p. e. RIS hazard in Brazil is not uniform, some
areas to be prone to induced seismicity while the RIS hazard in other areas are
much lower and there isn't correlation between RIS hazard, level natural seismicity
ca be observed with present data and the deep reservoirs should not be used to
reduce studies of RIS in small reservoirs (see Fig. 4).
After update, we can see in table 2 that the initial type is predominant, another
characteristics is the RIS, generally occur in boundary province, although the
number of dams and hydroelectric power is larger than in the provinces.
The scientific exploration of RIS is still not well understood and is therefore not
possible to predict and this work shows that we need more effort to understand
RIS in Brazil, until now we have only seismicity studies with companies
collaboration, but we need more support for scientific aspects for example “the
earthquake forecast studies” or “correlations with geophysics and geologic
methods”, these studies can help and understand the RIS in Brazil. These studies
are our goals to do in future studies.
We just have a statement: Not only, the seismic frequency can increase in areas
whose activity seismic occurrence was previously low, but also the frequency can
decrease in areas with seismic activity occurrence previously high.
Acknowledgments
We thank all Brazilian hydroelectric power companies (CEMIG, CESP,
CHESF, COPEL, ELETRONORTE, ELETROSUL, FURNAS, ITAIPU) for allowing
us to publish results of many years of seismic monitoring of their reservoirs. GSF
and MA thank CNPq for their research grants. We are grateful to Kate Moura and
Marina Dalla Costa for their valuable comments which enhanced the paper.
REFERENCES
[1]
[2]
[3]
[4]
[5]
ASSUMPÇÃO, M. (1983), A Regional Magnitude Scale for Brazil, Bull.
Seismol. Soc. Am. 73, 237-246.
ASSUMPÇÃO, M. (1992), The Regional Intraplate Stress Field in South
America, J. Geophys. Res. 97, 11.889-11.903.
ASSUMPÇÃO, M., BARBOSA, J. R., BERROCAL, J., BASSINI, A.,
VELOSO, J. A. V., MARZA, V., HUELSEN, M., AND RIBOTTA, L. C. (1997),
Seismicity Patterns and Focal Mechanisms in SE Brazil, Rev. Bras.
Geofísica 15, 119-132.
ASSUMPÇÃO, M. (1998), Focal Mechanisms of Small Earthquakes in SE
Brazilian Shield: A Test of Stress Models of the South American Plate,
Geophys. J. Int. 133, 490-498.
ASSUMPÇÃO M., V.I. MARZA, L.V. BARROS, C.N. CHIMPLIGANOND,
J.E.P. SOARES, J.M. CARVALHO, D. CAIXETA, A. AMORIM AND E.
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
CABRAL,(2002) Reservoir induced seismicity in Brazil, Pure Appl.
Geophys. 159, pp. 597–617.
BERROCAL, J., ASSUMPÇÃO, M., ANTEZANA, R., DIAS NETO, C. M.,
ORTEGA, R., FRANÇA, H., AND VELOSO, J. (1984), Sismicidade do Brasil.
IAG-USP/CNEN, São Paulo, Brazil, 320 pp.(In Portuguese).
CBGB (1999), Registro Nacional de Barragens, Brazilian Committee on
Large Dams, CD-ROM.
FERREIRA, J. M., OLIVEIRA, R. T., ASSUMPÇÃO, M., MOREIRA, J. A. M.,
PEARCE, R. G., AND TAKEYA, M. (1995), Correlation of Seismicity and
Water Level in the Acu Reservoir - An Example from NE Brazil, Bull.
Seismol. Soc. Am. 85(5), 1483-1489.
FERREIRA, J., OLIVEIRA, R. T., TAKEYA, M. K., AND ASSUMPÇÃO, M.
(1998), Superposition of Local and Regional Stresses in NE Brazil: Evidence
Form Focal Mechanisms around the Potiguar Marginal Basin, Geophys. J.
Int. 134, 341-355.
FERREIRA J. M., G.S. FRANÇA, C.S. VILAR, A.F., DO NASCIMENTO,
F.H.R. BEZERRA, M. ASSUMPÇÃO, (2008) Induced seismicity in the
Castanhao reservoir, NE Brazil -- Preliminary results, Tectonophysics, V 456,
I 1-2, The Monitoring of Induced Seismicity: Observations, Models and
Interpretations,
Pages
103-110,
ISSN
0040-1951,
DOI:
10.1016/j.tecto.2006.11.011.
GISMAP - www.gismaps.com.br/
GOMIDE, L. C. (1999), Nature and History of Reservoir-induced Seismicity
in Brazil, M.Sc. Thesis, Univ. of South Carolina, 60 pp
GUPTA, H. K., 2002. “A review of recent studies of triggered earthquakes by
artificial water reservoirs with special emphasis on earthquakes in Koyna,
India.” Earth-Science Reviews 58 279–310.
JAUHARI, V.P., 1999. Prepared for Thematic Review IV.5. “Options
Assessment- Large Dams in India Operation, Monitoring and
Decommissioning of Dams” www.dams.org
LIMA, C., NASCIMENTO, E., AND ASSUMPÇÃO, M., (1997), Stress
Orientations in Brazilian Sedimentary Basins from Breakout Analysis Implications for Force Models in the South American Plate, Geophys. J. Int.
130(1), 112-124.
MARZA, V. I., BARROS, L. V., SOARES, J. E. P., CARVALHO, J. M.,
FONTENELLE, D. P., CHIMPLIGNOND, C, N., CAIXETA, D. F., GOMES, I.
P., FURTADO, G. O., CARIM, A. L., SOUZA, I. F., CALIMAN, E. H., AND
BARROS J. B. (1999), Some Aspects of Reservoir-induced Seismicity in
Brazil, Proc. 23rd Brazilian Congr. on Large Dams, Belo Horizonte, Brazil,
vol. 1, 199-211 (in Portuguese with English abstract).
MENDIGUREN, J. A. (1980), A Procedure to Resolve Areas of Diferent
Source Mechanisms When Using the Method of Composite Nodal Plane
Solution, Bull. Seismol. Soc. Am. 70, 985-998.
NASCIMENTO P. F. ; CHIMPLIGANOND C. N. ; FRANÇA, G. S.; ROIG H. L.
(2008) Sismicidade Induzida por Reservatório de UHE Irapé. In: 44
Congresso Brasileiro de Geologia, 2008, Curitiba.
RIBOTTA L. C., MIOTO J. A., MANUZZI J. L., CARVALHO A. M. B. &
VINCIPROVA G. (2008) Sismicidade na área do reservatório de Barra
Grande, SC/RS. In: III Simpósio Brasileiro de Geofísica, 2008, Belém.
[20] VELOSO, J. A. V. (1992), Terremotos induzidos pelo homem, Ciencia Hoje
14, 66-72.
[21] VILAR, C. S. ; FERREIRA, J. M. ; DO NASCIMENTO, A. F. ; BEZERRA, F.
H. R. ; FRANÇA, G. S. . Atividade sísmica em Apodi-RN, natural ou
induzida?. Ciência Sempre - Revista Fapern, v. 6, p. 58-63, 2007.
Table 1
Reservoirs Characteristics
References: (A) Assumpção (1992). (B) Assumpção et al. (1997). (C) Assumpção
(1998b). (D) Ferreira et al. (1995). (E) Ferreira et al. (1998). (F) Lima et al.(1997).
(G) Mendiguren (1980). (H) Ribotta et al. (2008) , (I) Nascimento et al. (2008) , (J)
Ferreira et al. (2008) , (L) Ribotta (communication pessoal) (M) Vilar et al. (2007) ,
(N) Caixeta D. (Communication pessoal), (O) Veloso, (1992).
N
Dam name
Dam height
(state)
(m)
Maximum
water
depth(m)
Reservoir
Start of
Predominant
Volume (km3)
Impoundment
geology
Regional Stress
Shmax direction
(reference)
Porto Colombia
40
35
1.46
1973/Apr
Basalt
X
Volta Grande (MG/SP)
45
37
2.30
1973/Sep
Basalt
X
2
Nova Ponte (MG)
142
132
12.8
1993/Oct
Basalt Gneiss
3
Cajuru (MG)
23
20
0.20
1954
Granite Gneiss
4
Capivara (PR/SP)
60
55
10.5
1976/jan
Basalt
X
5
Tucuruí (PA)
106
90
45.8
1984/Sep
Metamorphic
X
6
Balbina (AM)
42
35
17.5
1987/Oct
Sediments
Compression?;NN
granite-gneiss
W-SSE? (A, F)
7
Miranda (MG)
85
82
1.14
1997/Aug
Basalt Gneiss
Paraibuna-
98
90
4.74
1974
Granite Gneiss
Paratinga (SP)
104
90
4.74
1976
Granite Gneiss
Jaguari
67
53
1.5
1969/Dec
Granite Gneiss
X
61
58
0.18
1970/Jul
Gneiss
X
1
Compression; NNESSW (B)
Strike-slip;
E-W
(B,C)
Compression; NNESSW? (B)
Strike-slip;
NE-SW
(G)
8
9
10
Capivari-Cachoeira
(PR)
11
Assu (RN)
31
31
2.4
1985
Granite Gneiss
12
Serra da Mesa (GO)
150
146
54.4
1996/Oct
Granite
13
Marimbondo (MG/SP)
90
86?
6.15
1975
Basalt
14
Sobradinho (BA)
43?
40
34.1
1977
15
Emborcação (MG/GO)
158
154
17.5
1981/Aug
Granite
Gneiss/Schist
Gneiss
Strike-slip;
NE-SW
(G)
Sthike-slip; E-W (D,
E)
Compression;
NNW-SSE (A)
X
X
X
16
17
Xingo (SE/AL)
140
110?
3.8
1994/Jun
Metamorphic
X
Machadinho (SC/RS)
79
126
3.34
2001/Aug
Basalt
X (H)
Barra Grande (SC/RS)
93.4
186
5.2
2005/Jun
Basalt
X (H)
34.6
188
1.48
2005/Oct
Basalt
X (H)
Campos
Novos
(SC/RS)
Quartzite
18
Irapé (MG)
208
137
??
205/Dec
19
Castanhão (CE)
60
??
4.5
2002
Gneiss Migmatites (J)
MetaBasalt
Strike-slip (I)
Doubtful cases
01
Peti (MG)
43
42
0.042
1946
Gneiss
X(O)
02
Furnas (MG)
127
106
23
1963
Quartizite
Strike-slip; E-W (A)
03
Três Marias (MG)
70
60?
21
1962
Sediment
X
04
Quebra-Queixo (SC)
5.6
75
0.136
2003/mar
Andesito Dacito
(L)
05
Três Irmãos (SP)
77
46
13.4
1999/Aug
X
(L)
57.5
???
0.6
2002
???
(M)
30
4.982
???
06
Santa Cruz do Apodi
(RN)
07
Lajeado (TO)
43
08
Itapebi (MG)
60
Granit
Granodiorito
Gnaisse Migmatito
Ortognaisse
Strike-Slip(N)
Strike-slip
Table 2
Characteristics of the induced seismicity: H_water is meter
long.
-48.47
-47.70
-44.75
-51.01
-49.63
-59.6
-47.96
-45.60
-46.06
-48.9
-36.9
-48.31
-48.9
-41.0
-47.9
-37.8
-42.58
-51.78
-51.22
-52.55
-42.61
-37.799
-38.5
-52.55
-51,22
lat.
-20.04
-19.17
-20.28
-22.84
-3.86
-1.9
-18.96
-23.42
-23.17
-25.1
-5.8
-13.79
-20.3
-9.4
-18.45
-9.6
-16.72
-27.52
-27.77
-26.67
-16.76
-5.768
-5.5
-26.67
-20,67
mb
4.2
4.0
3.7
3.7
3.6
3.7
3.3
3.0
3.0
2.8
3.0
2.2
2.0
2.0
2.0
2.0
3.0
1.8
2.5
1.8
3.0
2.3
1.4
0.1
0.1
H_water
40
132
20
55
90
35
82
88
53
54
33
134
86
41
154
110
200
79
93.4
34.6
208
57.5
60
75
46
Reservoir
Porto Colombia
Nova Ponte
Cajuru
Capivara
Tucurui
Balbina
Miranda
Paraibuna
Jaguari
Capivari-Cachoeira
Assu
Serra da Mesa
Marimbondo
Sobradinho
Emborcacao
Xingo
Irape
Machadinho
Barra Grande
Campos Novos
Irapé
Santa Cruz do Apodi
Castanhão
Quebra-Queixo
Três Irmãos
State
MG/SP
MG
MG
SP/PR
PA
AM
MG
SP
SP
PR
RN
GO
MG/SP
BA
MG/GO
SE/AL
MG
SC/RS
SC/RS
SC/RS
MG
RN
CE
SC
SP
Date
1974.02.24
1998.05.22
1972.01.23
1979.03.27
1998.03.02
2001.05.27
2000.05.06
1977.11.16
1985.12.17
1971.05.21
1994.08.26
1998-06.13
1978.07.25
1979.07.05
1984.05.18
1994.07.20
2006.05.14
2008
2008
2008
2006.05.14
2005.05.27
2004
???
???
Type
Initial
Initial&Delayed
Delayed
Initial&Delayed
Initial&Delayed
Initial&Delayed
Initial&Delayed
Initial
Delayed
Initial
Delayed
Initial
Initial
Initial
Initial
Initial
Initial
Initial
Initial
Initial
Initial
Delayed
Initial&Delayed
???
???
Figure 1: Location of the reservoirs with induced seismicity. The lines refer
to the main geological provinces in Brazil.
Figure 2: Seismicity in Irapé (No. 18), State of Minas Gerais. Epicentral
distribution: triangles denote seismographic stations; circles denote epicenters and
square denote cities.
Figure 3: Temporal evolution: number of monthly events (histogram), monthly
averaged water level (thick line) in Irapé Reservoir.
Figure 4: Natural and Induced Seismicity of Brazil (Assumpção et al. 2002 and
Marza et al. 1999). Thin lines refer to main geological provinces in Brazil. AM –
Amazon basin, BP – Borborema Province, CBS – Central Brazil Shield, GS –
Guyana shield, PB – Parnaíba basin, PR – Paraná basin, SFC – São Francisco
craton
6. SUMMARY AND KEYWORDS
The first compilation of Brazilian seismicity was had in 1984, ever since, more
than 40 cases were registered and compiled.
This objective research is to understand the nature of the RIS that has been
observed to date, particularly in Brazil. It was exposed here a synopsis of the
historical cases of RIS in Brazil, a review the lithology of hypocentral areas where
RIS has been observed, their filling history and reservoir characteristics. Joining
to cases of Brazil RIS was presented recent case, in Irapé Reservoir.
In a general way, we can observe that, at Brazil the effect of RIS is mainly of the
fast type (following the initial filling of the reservoir). Another characteristic is: the
RIS generally occur in boundary province, although the number of dams and
hydroelectric power is larger than in the provinces. The scientific exploration of
RIS is still not well understood and this work shows that we need more effort to
understand RIS in Brazil. Finally it’s important observe that seismic frequency can
change after the build of a dam.
Keywords: Reservoir Induced Seismicity; RIS hazard; Brazilian seismicity;
7. RÉSUMÉ E MOTS CLEFS
La prima compilazione di sismicità brasiliana era avuta in 1984, da allora,
più di 40 casi erano iscritti ed erano compilati.
Questa ricerca oggettiva è capire che la natura del RIS che è stato
osservato per datare, particolarmente in Brasile. Era qui esposto un
sommario dei casi storici di SIGNORE in Brasile, una revisione il lithology
di aree di hypocentral dove RIS è stato osservato, le loro caratteristiche di
storia e serbatoio riempendo. Unire ai casi di Brasile RIS era presentato il
caso recente, nel Serbatoio di Irapé.
In una maniera generale, possiamo osservare ciò, a Brasile l'effetto di RIS
è principalmente del tipo veloce (seguendo il riempire iniziale del
serbatoio). Un'altra caratteristica è: il RIS accade generalmente nella
provincia di limite, sebbene il numero di dighe ed il potere idroelettrico è
più grande di nelle province. L'esplorazione scientifica di RIS bene non è
tuttavia capita e queste mostre di lavoro che abbiamo bisogno di più sforzo
per capire RIS in Brasile. Finalmente è importante osserva quella
frequenza sismica può cambiare dopo che il costruisce di una diga.
Mots clefs : Il serbatoio Ha Indotto la Sismicità; il pericolo di RIS; la
sismicità brasiliana