Numerical modelling of pore-pressure diffusion in a reservoir-induced seismicity site in northeast Brazil

A. Nascimento, R.J. Lunn, P. Cowie

Research output: Contribution to journalArticle

27 Citations (Scopus)

Abstract

A 3-D fluid-flow model is used to investigate pore-pressure diffusion as a mechanism for reservoir-induced seismicity (RIS) at the Açu reservoir in NE Brazil. The Açu dam is a 34-m high earth-filled dam constructed in 1983 on an area of Precambrian shield. Seismic activity in this area has been monitored over a 10-yr period (1987-1997). The frequency of earthquakes clearly varies with seasonal fluctuations of the reservoir level. Information on the hydrological regime of the Açu reservoir (rainfall, vegetation, surface and subsurface storage, etc.) is used to set up a regional groundwater-flow model in order to obtain boundary conditions for a more detailed study of the area of seismic activity. To explain the observed time lag between maximum reservoir level and peak seismic activity we calculate the magnitude and timing of the maximum piezometric head in the depth range of observed seismic activity. By assuming that individual earthquake ruptures occur when the local piezometric head is at a maximum, values of bulk permeability, K, and storativity, S, are derived. If a 3-D homogenous subsurface permeability structure is assumed then the values of K and S obtained are not self-consistent and are physically unrealistic. However, if a high-permeability fault is embedded into, and explicitly coupled with, the surrounding lower-permeability matrix, then our estimates of subsurface hydraulic properties agree well with other field and laboratory measurements. The inclusion of a discrete fault plane in the model is consistent with the results of high-resolution seismic monitoring using a local digital network of stations, which show that the earthquake hypocentres define a set of steeply dipping NE-striking fault planes beneath the reservoir.
LanguageEnglish
Pages249-262
Number of pages13
JournalGeophysical Journal International
Volume160
Issue number1
DOIs
Publication statusPublished - 2005

Fingerprint

reservoir-induced seismicity
Pore pressure
Brazil
pore pressure
Earthquakes
porosity
Dams
permeability
modeling
Groundwater flow
dams
fault plane
earthquakes
Rain
Flow of fluids
earthquake hypocenter
Earth (planet)
earthquake rupture
Hydraulics
earth dam

Keywords

  • modelling diffusion
  • permeability
  • seismicity
  • numerical modelling
  • civil engineering

Cite this

@article{d37183bdde1e4ec98634e6a6e221f252,
title = "Numerical modelling of pore-pressure diffusion in a reservoir-induced seismicity site in northeast Brazil",
abstract = "A 3-D fluid-flow model is used to investigate pore-pressure diffusion as a mechanism for reservoir-induced seismicity (RIS) at the A{\cc}u reservoir in NE Brazil. The A{\cc}u dam is a 34-m high earth-filled dam constructed in 1983 on an area of Precambrian shield. Seismic activity in this area has been monitored over a 10-yr period (1987-1997). The frequency of earthquakes clearly varies with seasonal fluctuations of the reservoir level. Information on the hydrological regime of the A{\cc}u reservoir (rainfall, vegetation, surface and subsurface storage, etc.) is used to set up a regional groundwater-flow model in order to obtain boundary conditions for a more detailed study of the area of seismic activity. To explain the observed time lag between maximum reservoir level and peak seismic activity we calculate the magnitude and timing of the maximum piezometric head in the depth range of observed seismic activity. By assuming that individual earthquake ruptures occur when the local piezometric head is at a maximum, values of bulk permeability, K, and storativity, S, are derived. If a 3-D homogenous subsurface permeability structure is assumed then the values of K and S obtained are not self-consistent and are physically unrealistic. However, if a high-permeability fault is embedded into, and explicitly coupled with, the surrounding lower-permeability matrix, then our estimates of subsurface hydraulic properties agree well with other field and laboratory measurements. The inclusion of a discrete fault plane in the model is consistent with the results of high-resolution seismic monitoring using a local digital network of stations, which show that the earthquake hypocentres define a set of steeply dipping NE-striking fault planes beneath the reservoir.",
keywords = "modelling diffusion, permeability, seismicity, numerical modelling, civil engineering",
author = "A. Nascimento and R.J. Lunn and P. Cowie",
year = "2005",
doi = "10.1111/j.1365-246X.2005.02473.x",
language = "English",
volume = "160",
pages = "249--262",
journal = "Geophysical Journal International",
issn = "0956-540X",
number = "1",

}

Numerical modelling of pore-pressure diffusion in a reservoir-induced seismicity site in northeast Brazil. / Nascimento, A.; Lunn, R.J.; Cowie, P.

In: Geophysical Journal International, Vol. 160, No. 1, 2005, p. 249-262.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Numerical modelling of pore-pressure diffusion in a reservoir-induced seismicity site in northeast Brazil

AU - Nascimento, A.

AU - Lunn, R.J.

AU - Cowie, P.

PY - 2005

Y1 - 2005

N2 - A 3-D fluid-flow model is used to investigate pore-pressure diffusion as a mechanism for reservoir-induced seismicity (RIS) at the Açu reservoir in NE Brazil. The Açu dam is a 34-m high earth-filled dam constructed in 1983 on an area of Precambrian shield. Seismic activity in this area has been monitored over a 10-yr period (1987-1997). The frequency of earthquakes clearly varies with seasonal fluctuations of the reservoir level. Information on the hydrological regime of the Açu reservoir (rainfall, vegetation, surface and subsurface storage, etc.) is used to set up a regional groundwater-flow model in order to obtain boundary conditions for a more detailed study of the area of seismic activity. To explain the observed time lag between maximum reservoir level and peak seismic activity we calculate the magnitude and timing of the maximum piezometric head in the depth range of observed seismic activity. By assuming that individual earthquake ruptures occur when the local piezometric head is at a maximum, values of bulk permeability, K, and storativity, S, are derived. If a 3-D homogenous subsurface permeability structure is assumed then the values of K and S obtained are not self-consistent and are physically unrealistic. However, if a high-permeability fault is embedded into, and explicitly coupled with, the surrounding lower-permeability matrix, then our estimates of subsurface hydraulic properties agree well with other field and laboratory measurements. The inclusion of a discrete fault plane in the model is consistent with the results of high-resolution seismic monitoring using a local digital network of stations, which show that the earthquake hypocentres define a set of steeply dipping NE-striking fault planes beneath the reservoir.

AB - A 3-D fluid-flow model is used to investigate pore-pressure diffusion as a mechanism for reservoir-induced seismicity (RIS) at the Açu reservoir in NE Brazil. The Açu dam is a 34-m high earth-filled dam constructed in 1983 on an area of Precambrian shield. Seismic activity in this area has been monitored over a 10-yr period (1987-1997). The frequency of earthquakes clearly varies with seasonal fluctuations of the reservoir level. Information on the hydrological regime of the Açu reservoir (rainfall, vegetation, surface and subsurface storage, etc.) is used to set up a regional groundwater-flow model in order to obtain boundary conditions for a more detailed study of the area of seismic activity. To explain the observed time lag between maximum reservoir level and peak seismic activity we calculate the magnitude and timing of the maximum piezometric head in the depth range of observed seismic activity. By assuming that individual earthquake ruptures occur when the local piezometric head is at a maximum, values of bulk permeability, K, and storativity, S, are derived. If a 3-D homogenous subsurface permeability structure is assumed then the values of K and S obtained are not self-consistent and are physically unrealistic. However, if a high-permeability fault is embedded into, and explicitly coupled with, the surrounding lower-permeability matrix, then our estimates of subsurface hydraulic properties agree well with other field and laboratory measurements. The inclusion of a discrete fault plane in the model is consistent with the results of high-resolution seismic monitoring using a local digital network of stations, which show that the earthquake hypocentres define a set of steeply dipping NE-striking fault planes beneath the reservoir.

KW - modelling diffusion

KW - permeability

KW - seismicity

KW - numerical modelling

KW - civil engineering

UR - http://dx.doi.org/10.1111/j.1365-246X.2005.02473.x

U2 - 10.1111/j.1365-246X.2005.02473.x

DO - 10.1111/j.1365-246X.2005.02473.x

M3 - Article

VL - 160

SP - 249

EP - 262

JO - Geophysical Journal International

T2 - Geophysical Journal International

JF - Geophysical Journal International

SN - 0956-540X

IS - 1

ER -