1

Reservoir Induced

Seismicity

Induced Seismicity

Many human activities are known to induce or increase seismic activity (Simpson,1986): Fluid injection for various purposes:

waste disposal,

solution mining,

geothermal power generation, and

secondary oil recovery;

Deep underground mining;

Removal of large volumes of rock during quarrying;

Fluid extraction in petroleum production; and

Impoundment of large reservoirs behind high dams (Simpson, 1986).

A dam is a structure which prevents the flow of

water and accumulates it in a reservoir

Impoundment facility

2

Needs for Dam Construction

1. Drinking and domestic water supply

2. Flood control

3. Irrigation

4. Industrial water supply

5. Hydroelectric energy production

6. Retention and control of sediments

and Inland navigation, Improvement of water quality, Fish Farming, Recreation facilities

REGION THEORETICAL

POTENTIAL (TWh)

TECHNICAL

POTENTIAL (TWh)

AFRICA 10118 3140

N. AMERICA 6150 3120

LATIN AMERICA 5670 3780

ASIA 20486 7530

OCEANIA 1500 390

EUROPE 4360 1430

WORLD 44280 19390

Continent Wide distribution

COUNTRY POWER

CAPACITY (GWh)

INSTALLED

CAPACITY (GW)

TAJIKISTAN 527000 4000

CANADA 341312 66954

USA 319484 79511

BRAZIL 285603 57517

CHINA 204300 65000

RUSSIA 160500 44000

NORWAY 121824 27528

JAPAN 84500 27229

INDIA 82237 22083

FRANCE 77500 77500

Top ten countries (in terms of capacity) The Indian Scenario

The potential is about 84000 MW at 60% load factor spread across six major basins in the country.

Pumped storage sites have been found recently which leads to a further addition of a maximum of 94000 MW.

Annual yield is assessed to be about 420 billion units per year though with seasonal energy the value crosses 600 billion mark.

The possible installed capacity is around 150000 MW

3

The proportion of hydro power increased from 35%

from the first five year plan to 46% in the third five

year plan but has since then decreased continuously to

25% in 2001.

The theoretical potential of small hydro power is 10071

MW.

Currently about 17% of the potential is being

harnessed

About 6.3% is still under construction.

India’s Basin wise potential

Rivers Potential at 60%LF

(MW)

Probable installed

capacity (MW)

Indus 19988 33832

Ganga 10715 20711

Central Indian

rivers

2740 4152

West flowing 6149 9430

East flowing 9532 14511

Brahmaputra 34920 66065

Total 84044 148701

Region wise status of hydro development

REGION POTENTIAL

ASSESSED

(60% LF)

POTENTIAL

DEVELOPED

(MW)

%

DEVELOPED

UNDER

DEVELOPMENT

NORTH 30155 4591 15.2 2514

WEST 5679 1858 32.7 1501

SOUTH 10763 5797 53.9 632

EAST 5590 1369 24.5 339

NORTH

EAST

31857 389 1.2 310

INDIA 84044 14003 16.7 5294

Major Hydropower generating units

NAME STATA CAPACITY (MW)

BHAKRA PUNJAB 1100

NAGARJUNA ANDHRA PRADESH 960

KOYNA MAHARASHTRA 920

DEHAR HIMACHAL

PRADESH

990

SHARAVATHY KARNATAKA 891

KALINADI KARNATAKA 810

SRISAILAM ANDHRA PRADESH 770

4

Installed Capacity

REGION HYDRO THERMAL WIND NUCLEAR TOTAL

NORTH 8331.57 17806.99 4.25 1320 27462.81

WEST 4307.13 25653.98 346.59 760 31067.7

SOUTH 9369.64 14116.78 917.53 780 25183.95

EAST 2453.51 13614.58 1.10 0 16069.19

N.EAST 679.93 1122.32 0.16 0 1802.41

INDIA 25141.78 72358.67 1269.63 2860 101630.08

Region wise contribution of Hydropower

REGION PERCENTAGE

NORTH 30.34

WEST 13.86

SOUTH 37.2

EAST 15.27

NORTH-EAST 37.72

INDIA 24.74

Small Hydro in India

STATE TOTAL CAPACITY (MW)

ARUNACHAL PRADESH 1059.03

HIMACHAL PRADESH 1624.78

UTTAR PRADESH &

UTTARANCHAL

1472.93

JAMMU & KASHMIR 1207.27

KARNATAKA 652.51

MAHARASHTRA 599.47

Sites (up to 3 MW) identified by UNDP

STATE TOTAL

SITES

CAPACITY

NORTH 562 370

EAST 164 175

NORTH EAST 640 465

TOTAL 1366 1010

5

Hydroelectric Power Plants in India

Baspa II Binwa Gaj Nathpa Jakri

Rangit Sardar Sarovar

ACCORDING to the SIZE of the DAM

1. Large (Big) dam

2. Small dam

International Commision on Large Dams, (ICOLD) assumes a dam as big when its height is bigger than 15m.

If the height of the dam is between 10m and 15m and matches the following criteria, then ICOLD accepts the dam as big:

If the crest length is bigger than 500m

If the reservoir capacity is larger than 1 million m3

If the flood discharge is more than 2000 m3/s

If there are some difficulties in the construction of foundation

6

ACCORDING to HEIGHT of DAM

High Dam or Large Dam If the height of the dam is bigger than 100m

Medium Dam If the height of the dam is between 50m and

100m

Low Dam or Small Dam If the height of the dam is lower than 50m

ENVIRONMENTAL IMPACTS of

RESERVOIRS

Loss of land

Habitat Destruction : The area that is covered by the reservoir is destroyed, killing

whatever habitat existed there beforehand.

Loss of archeological and histrorical places

Loss of mineral deposits

Loss of special geological formations

Aesthetic view reduction

Sedimentation

Change in river flow regime and flood effects

Reservoir induced seismicity

Change in climate and plant species

Benefits of Dam Environmental Benefits of Dam

• No operational greenhouse gas emissions

• Savings (kg of CO2 per MWh of electricity):

– Coal 1000 kg

– Oil 800 kg

– Gas 400 kg

• No SO2 or NOX

Non-environmental benefits

– Flood control, irrigation, transportation, fisheries and

– Tourism.

Dam Uses

Direct Water Usage

Private / Domestic - Household purposes, Drinking water and

landscape irrigation

Commercial - Restaurants, hotels, golf courses, etc.

Irrigation – Crop use. Water needs at the scale that large dams

provide most often feed industrial farming practices.

Livestock – Use for animal raising as well as other on-farm needs

Industrial – Cooling water (power generation, refineries, chemical

plants), processing water (manufacturing; pulp and paper, food, high

tech, etc.)

Mining – hydraulic mining, various processes, settling ponds

General public supply – Firefighting, public parks, municipal office

buildings

7

Dam Uses

Indirect Uses

Hydroelectric Power – Power generation is one of the most

common purposes for the construction of large dams. It is promoted

as a totally “clean” form of electricity.

Flood Control – Dams even out the peaks and lows of a rivers

natural flow cycle by calming seasonal flooding, then storing that

water for gradual release year round.

Transportation – Dam locks are used to move ships past large

dams. This in conjunction with flood control make transportation

feasible on rivers that were traditionally wild.

Disadvantages

The loss of land under the reservoir.

Interference with the transport of sediment by the

dam.

Problems associated with the reservoir.

Climatic and seismic effects.

Impact on aquatic ecosystems, flora and

fauna.

Climatic and Seismic effects

It is believed that large reservoirs induce have the potential to induce earthquakes.

In tropics, existence of man-made lakes decreases the convective activity and reduces cloud cover. In temperate regions, fog forms over the lake and along the shores when the temperature falls to zero and thus increases humidity in the nearby area.

Reservoir-Induced Seismicity There is a correlation between the creation of a large

reservoir, and an increase in seismic activity in an area

The physical weight of unnatural reservoirs can cause seismic activity. While not the direct cause of earthquakes, the weight of reservoirs can act as a trigger for seismic activity.

Although not much direct research is available on the subject, the proposed explanation is that “when the pressure of the water in the rocks increases, it acts to lubricate faults which are already under tectonic strain, but have been prevented from slipping by the friction of the rock surfaces”.

As of now, it is not accurately possible to predict which large dams will produce RIS or how much activity will be produced. Earthquakes that are produced as the result of dams are not usually major, but they still pose a major threat to dam stability and the safety of people living downstream.

8

Of all causes, reservoir impoundment has produced the largest earthquakes.

There is evidence linking earth tremors and reservoir operation for more than 70 dams.

Reservoirs are believed to have induced five out of the nine earthquakes on the Indian peninsula in the 1980s which were strong enough to cause damage.

Reservoir induced seismicity (RIS) is well documented but relatively poorly understood.

The mechanisms of RIS are not sufficiently well understood to predict accurately which dams will induce earthquakes or how strong the tremors are likely to be.

Most of the strongest cases of RIS have been observed for dams over 100 metres high - but dams just half this height are also believed to have induced quakes.

Reservoirs filling can Increase the frequency of earthquakes in areas of

already high seismic activity and

Cause earthquakes to happen in areas previously thought to be seismically inactive.

Reservoir induced seismicity

Depth of the water

The volume of the water

Type of local geology and the region’s historic seismic stress patterns.

Increase the frequency of earthquakes.

An increased rate of activity in RIS cases occurs within 10-15 kilometers of impounded 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).Minor cases of RIS can occur immediately during the filling periods.

Reservoir Filling

Groundwater discharge to a flooded valley is usually inhibited

as a reservoir fills.

Recharge continues unaffected by flooding.

Flooded

Natural

Dams are generally

constructed in groundwater

discharge areas (because

aquifers predominantly

discharge to river valleys).

Reservoir heads are

generally greater than the

aquifer heads.

9

Transient Readjustment

The flow regime adjusts by filling groundwater storage

until a new steady state is established.

Where the water table was near the surface, new

discharge zones become established. (e.g. Flathead

Reservoir, Mt.)

Flow direction reversals in the subsurface are likely to

occur.

Flooded

Natural

Reversal

New Springs

Valley Bottom Stability

Reservoir impoundments can also lead to stability problems.

Beneath the reservoir, increased pore-pressures are partially

compensated by the total stress increase due to the water loading.

Downstream of the impoundment, pore-pressures are increased to

similar levels with no total stress compensation.

Post-Reservoir Head

Pre-Reservoir Head

Zone of potential uplift and slope failure

Aquitard

Aquifer

Valley Wall Stability

Pore-pressures can reactivate bedrock shears, faults and gouge

(mylonitic) zones

Increased uplift pressure can cause heave of the valley floor.

Bedrock slide blocks and landslides can be reactivated or initiated by

large changes in pore-pressures in valley walls.

Regional Aquifer

Slide Debris

Slide Block

Water Table

Piezometric Surface

River

Effects of Dams

10

Effects on Faults Rock effect

Fault effect Evidence

For most well-studied cases of RIS, the intensity of seismic activity increased within around 25 km of the reservoir as it was filled.

The strongest shocks normally occurred relatively soon - often within days but sometimes several years - after the reservoir reached its operating level.

After the initial filling of the reservoir, RIS events normally continued as the water level rose and fell but usually with lower frequency and magnitude than the initial events.

The pattern of RIS is, however, unique for every reservoir.

11

Interpretation of Evidence

The evidence is consistent with a mechanism involving stress-relief.

The early events release the initial stresses more quickly the nearer they are to the critical level for slip.

Later changes in stress trigger less violent releases as the fault plane weakens (c,f approaches residual ) with each successive event.

Another Perspective Seismologists have published a list of about 100 cases of RIS.

These cases show that after the completion of a dam, the reservoir area experienced earthquakes of micro-level magnitude - 2.0 or 3.0 on the Richter scale.

Dense seismic networks have increased the detection potential and increased the number of cases cited as instances of RIS.

The earthquakes that the Indian peninsula has so far experienced may not be attributable to dams.

Construction of dams should be done in such a way as to withstand anticipated seismic activity and minor stress adjustments are inevitable.

Seismicity of India

The map shows

the location of

the Koyna and

Killari

earthquakes in

the largely

aseismic Indian

penninsula.

The recent M7.9

Gujarat quake is

also shown.

Gujarat

Killari

Koyna

Koyna Dam Earthquake

The area between the Koyna and the Warna dams, in the vicinity of the Shivaji Sagar and Vasant Sagar reservoirs, is unique for its ongoing, high level of seismic activity.

Seismicity at Koyna has close correlation with the filling cycles of the Koyna reservoir.

The 1967 Koyna event, in the watershed of the Krishna River in

Maharashtra state, is a classic example of earthquake activity triggered by reservoir.

The world's worst confirmed reservoir-induced earthquake was triggered by the Koyna Dam.

Nearly 200 were killed in the magnitude 6.3 tremor.

12

Koyna Dam Background

The height of the Koyna-Dam is 103 m, reservoir volume is 2.78×109 m3.

Seasonal fluctuations of the lake level are typically 30 to 35 m and are dominated by monsoon rainfalls.

The site is now highly instrumented and the subject of active research

Since its first impoundment in 1962, more than 150 earthquakes of magnitude 4.0 have been recorded.

Events are mostly restricted to an area 40 × 25 km2 south of the Koyna-Dam.

This marks the area as probably the best in the world to study the phenomenon of reservoir induced/triggered seismicity (RIS).

Killari Event

The most puzzling event in Peninsular India is the Killari earthquake.

The devastating magnitude 6.4 earthquake struck Killari, Maharashtra in 1993, killing 10,000 people.

The event was totally unexpected as it was located in the Deccan Trap-covered stable Indian Shield. There was no record of any historical earthquake in the region.

The Killari earthquake is considered the most devastating SCR (Stable Continental Region) event in the world.

Some seismologists believe that the Killari event was triggered by a nearby (Tirna) reservoir.

Tirna Reservoir The Killari earthquake was about 10 km from the Lower

Tirna Reservoir.

The maximum water depth is about 20m, which is at the low end of the range of depths of reservoirs where induced seismicity has been documented.

The reservoir level was low at the time of the main shock, which is consistent with the expected negative effect of the loading by the reservoir on an underlying thrust fault.

Several other recent earthquakes in peninsular India appear to be located close to reservoirs.

Whether the Killari earthquake was triggered by the Lower Tirna reservoir is not known, but it cannot be ruled out at this time.

13

Narmada Valley Indian seismologists have noted an increase in seismic activity in

the Narmada Valley over the past 20 years, which may be linked to reservoir impoundment.

In the Narmada Valley, a series of tremors were felt soon after the completion of the Sukta Dam.

A strong earthquake hit the Narmada Valley on May 22, 1997, killing around 50 people and injuring 1,000 in the city of Jabalpur in the state of Madhya Pradesh.

The epicentre of this magnitude 6.0 earthquake is believed to have been about 20-40 kilometers from Bargi Dam, which completed filling in 1990.

The recent earthquake has focused attention on the seismic risks faced by the large dams planned for the Narmada Valley, and on the risk of reservoir-induced earthquakes.

Seismic Hazard Assessment Seismic hazard assessments are an integral part of site

investigation for large dams and reservoirs.

In order to interpret the recorded seismicity of a region, a thorough review of the available previous seismicity and seismo-tectonic studies is performed.

The analysis is further deepened through the integration of three-dimensional velocity structures and inversion studies beneath this area.

The compilation of all these data makes it possible to define and gain considerable insight concerning the major seismic sources active in the region.

Some major/minor induced earthquakes

DAM NAME COUNTRY HEIGHT (m) VOLUME OF

RESERVOIR

(m3)

MAGNITUDE

KOYNA INDIA 103 2780 6.5

KREMASTA GREECE 165 4650 6.3

HSINFENGKIANG CHINA 105 10500 6.1

BENMORE NEW

ZEALAND

118 2100 5.0

MONTEYNARD FRANCE 155 240 4.9

M>5 Induced Earthquakes.

__________________________________________

Dam / Reservoir DP or Hydrocarbon Field

Magnitude Seismicity Induced Earthquakes

_____________________________________________________________

___________________________________

Gazli field, EIS 7.3 low horizontal midplate

Uzbekistan

Koyna, RIS 6.5 low horizontal midplate

India Coalinga field, EIS 6.5 high horizontal plate boundary

USA

Kremastaa, RIS 6.3 high vertical back arc extension

Greece

Hsinfengkiang, RIS 6.1 low horizontal midplate

China

Kettleman field, EIS 6.1 high horizontal plate boundary

USA

Montebello field, EIS 5.9 high horizontal plate boundary

USA

Oroville, RIS 5.9 low vertical Sierra Nevada foothills

USA

USA

Kariba, RIS 5.8 low vertical midplate

Zambia/Zimbabwee

Marathona, RIS 5.7 high n.a plate boundary

Greece

Aswana, RIS 5.5 low vertical midplate

Egypt

Eucumbene, RIS 5.5 low n.a midplate

Australia

Hoover, RIS 5.5 low vertical Colorado plateau

USA

Denver, IIS 5.5 low vertical Colorado plateau

USA

Caviaga, EIS 5.5 low horizontal midplate

Italy

Lake County, IIS 5.3 low horizontal midplate

USA

Monteynard, RIS 5.3 low vertical Alps foothills

France

El Reno, EIS 5.2 low horizontal midplate

USA

Snipe Lake, EIS 5.1 low horizontal midplate

Canada

14

The most well-known are the instances

of seismicity induced by reservoirs

behind the Hoover and Oroville dams

in USA, Kariba dam on the Zambia–

Zimbabwe border, Nurek dam in

Tadzhikistan, Hsinfengkiang dam in

China and Koyna dam in India.

Dam! What a disaster

33 sure cases of reservoir-induced seismicity or M>4 (7 of them

were over 5.5)

Most know event is the Koyna Reservoir (India) earthquake in

1967. M=6.3 Over 200 fatalities and more than 1500 injured.

Heated debates on the responsibility of the Zipingpu reservoir in

the 2008 Sichuan (China) earthquake. Over 68000 fatalities and

has been felt all over China. During the initial impoundment,

notable number of earthquakes (M 3.5)

Dams and Earthquakes

Sichuan Earthquake May Be Dam-

Induced

Talembote Case History

The assessment of seismic hazard within the Talembote

area, Morocco, is a study of a dam located within the

actively deforming intermountain belt of the Rif region,

considered the most active zone in Morocco.

The historical seismic data available on Morocco extend

to about 11 centuries back in history.

Of more importance is the 20th century seismicity data,

which reveals the occurrence in 1909 of a M6.4 event

about 50 km away from the dam.

15

Talembote Seismic Setting

Of particular importance are shallow surface features; mostly normal and strike-slip faults, which are identified as local faults that are running right next to the dam-site.

However, most of the seismic activity seems to be related to reverse faults along Rif-nappes connected to a detachment surface at about 20 km-depth.

This detachment runs right underneath the dam-site. The detachment zone may coincide with a low strength layer that decouples the overlying sediments from the basement of the African Plate.

As a result, there is a high level of small magnitude earthquakes.

Talembote Seismic Analysis

The analysis of seismic hazard of the site of the Talembote dam has shown that the Maximum Credible Earthquake (MCE) is in the order M6.8, risking to produce a maximum acceleration of 0.5g.

This event could possibly be generated once every ten thousand years by one of the faults passing in the immediate proximity of the dam.

When considering the much shorter design life for the dam-structure, it is normal to use an earthquake return period 7 or 8 times the 75-year design life.

An acceleration of 0.085 g, corresponds to a return period of 550 years. This acceleration is rounded to predict an operational basic earthquake of 0.1 g.

There are 19 cases of RIS in China, including the Xinfengjiang Reservoir which

was associated with a Ms 6.1 event in 1962. Most of the cases of RIS occurred in

South China and are predominantly in karst terrane. The cases of RIS in granitic

rocks, e.g., Xinfengjiang Reservoir appear to be caused by pore pressure

diffusion in fractured rocks. That lithology controls the location of seismicity is

illustrated by the example of RIS in Danjiangkou Reservoir. The temporal

association of RIS with filling showed that in some cases, shallow, small

earthquakes are associated with reservoir impoundment (Skempton’s effect).

Several examples illustrate that the chemical effect of water in dissolution is

responsible for RIS.

The presence of faults in the granitic core where the Three Gorges Project is

under construction, and the presence of outlying carbonate rocks upstream,

suggest the possibility of moderate earthquakes when the reservoir is

impounded.