Sections

Services

Information

UK United Kingdom

Earthquakes and mining - how humans create seismic activity

This week’s 5.3 magnitude earthquake that struck near Moe in Victoria’s brown-coal mining region of the La Trobe Valley brings to mind the 5.6 magnitude quake of 1989 in another coal-mining heartland…

Coal’s toll? A Newcastle church cupola damaged in the 1989 earthquake serves as a memorial for the 13 people killed. Flickr/OZinOH

This week’s 5.3 magnitude earthquake that struck near Moe in Victoria’s brown-coal mining region of the La Trobe Valley brings to mind the 5.6 magnitude quake of 1989 in another coal-mining heartland: NSW’s Hunter Valley.

Mining cannot be blamed for the Moe quake in the La Trobe Valley’s coalfields. However, it can be argued that mining played a large part in the destruction wrought in 1989 upon Newcastle, with its proximity to underground black-coal mining.

Newcastle’s first coalfield opened in 1801. While the region has prospered economically since the 1950s as a result of exporting black coal to China and Japan, it has also experienced an undesirable consequence of booming mining operations in the form of frequent damaging seismic activity which includes earthquakes in 1841, 1868, 1925, 1989 and 1994.

Due to its unusually high seismic activity, this mining region is referred to as the “Newcastle Triangle Seismic Hazard Zone.” Earthquakes with a magnitude of greater than 5 on the Richter scale occur, on average, every 40 years in this locale. Other areas of New South Wales — such as the Tasman Sea Margin Zone, for example — experience such seismic events, on average, every 170 years.

Newcastle in 1989. AAP

The 1989 Newcastle earthquake has been attributed to deep coal mining in the region. In addition to leaving 13 people dead, more than 160 in hospital, and thousands of buildings damaged, the magnitude 5.6 earthquake resulted in economic losses equivalent to almost A$5 billion, adjusted for inflation. The monetary losses totalled 3.4% of Australia’s Gross Domestic Income (GDI) or 80% of the nation’s GDI per capita growth, illustrating how destructive mining-induced earthquakes can be when urbanisation practices and mining activities are not pursued in a risk-balanced manner.

Could it be a coincidence?

Is the elevated seismic hazard in Newcastle really a result of black coal mining? Or is it a pure coincidence, with the region simply being tectonically active? Research data, analysed over decades, suggest that mining has caused seismic activity in and around Newcastle.

The horizontal deformation (or sideward drift) rates of the local tectonic plate in the Lake Macquarie region, where Newcastle is situated, are lower than what current GPS instruments are able to record accurately (GPS instruments are used to measure horizontal velocities of the moving plates of the earth’s crust). The velocities of the tectonic plates in southern Australia are extremely small. Therefore we talk about this being a Stable Continental Region. The only existing deformation is vertical movement down. Such deformation, called subsidence, is the result of the compaction of old mine shafts deep underground.

Subsidence is an effect that occurs above a deep mine. Another effect occurs below a deep mine, due to the redistribution of the existing tectonic stress field disturbed by the excavation process. High stresses exist inside tectonic plates and they are sensitive to any perturbations. The perturbation, or disturbance, of the stress field reaches beneath the excavations, mainly due to the removal of coal as well as the water that is extracted in order to keep the deep shafts dry.

This mass removal causes an “unload” of the entire continental crust. In the case of mining, an unload is the reduction of weight in a certain area, e.g., black coal and water in a colliery. If it has enough time, the tectonic plate reacts to this unload by forming a small deformation. If it does not have enough time because the rate of mass removal is too high, a fault zone can be reactivated or newly created and an abrupt reaction is induced - an earthquake is produced.

For a seismic event to take place, the stress perturbations need to be above the daily stress changes caused by the sun and the moon, plus the tectonic stress field in the crust and the pre-existing fault zone need to be oblique (at an angle that is neither a right angle nor a multiple of a right angle). Newcastle has the smallest stress vertical and the largest horizontal, and there is an oblique fault. This combination brings a fault closer to failure because the vertical component gets even smaller due to the unload. This reduction causes a reaction (e.g., relaxation) of the rock mass beneath the colliery. If a fault is right near this colliery it can set off earthquakes as part of the “relaxation” process.

If we had the smallest component horizontal and the largest vertical, it would reduce the likelihood of inducing a seismic event. This happened in Pakistan in 1974, when the impoundment of the Tarbela water reservoir locked a seismically active part of the crust at a depth of up to 70 km.

Ruins from China’s Sichuan earthquake of 2008, which killed tens of thousands of people. The quake has been linked to geoengineering. AAP/EPA/Michael Reynolds

A global perspective

The 1989 Newcastle earthquake is not the only seismic event associated with geoengineering (such as mining, water reservoir impounding, fluid extraction of hydrocarbons or injection of liquid carbon dioxide, deep geothermal power generation, and coastal management). Other notable examples include the 2008 magnitude 7.9 earthquake in China, which occurred 2.5 years after the impoundment of the Zipingpu water reservoir, as well as the 2006 magnitude 3.8 earthquake in Basel, Switzerland, which was caused by deep geothermal power generation and resulted in millions of dollars in damage.

A new study, soon to be published in a special issue, Triggered and Induced Seismicity: Probabilities and Discrimination, of the Journal of Seismology, shows that there is a statistically significant relationship between man-made mass shifts in the Earth’s crust and earthquakes observed in close vicinity (fewer than 50 kilometres) of the geoengineering activities. Most of the earthquakes examined in the study were induced or triggered several years after the geoengineering activities had started. In concluding this special issue of the Journal, recommendations are provided for identifying human-made earthquakes.

Locations of geoengineering operations should be monitored and the seismic events should be analyzed based on the following model types a) physics-based, b) statistical, c) source mechanisms. Such models can discriminate between human-caused and natural seismic activity. At the same time hazard mapping in space and time is crucial to provide information to authorities urban planners. In the case of Newcastle, mining was, is, and will be a part of this region. The associated hazards need to be addressed and risk-mitigation measures need to be implemented, including reinforcement of buildings or changes in urban planning.

The risks associated with human-made earthquakes need to be addressed when large-scale geoengineering projects are planned. Ensuring sustainable geoengineering practices is especially important in the regions of high population densities. This is, in particular, important for deep geothermal power generation or carbon sequestration as hot spots for earthquakes crafted by human action.

Comments welcome below.

Sign in to Favourite

Join the conversation

15 Comments sorted by

  1. Glenn Albrecht

    Murdoch University

    The possibility of anthropogenic earthquakes should be an integral part of environmental impact assessment (EIA). Mine and fracking proponents should have to prove that their activities will not cause seismic activity beyond background levels. The precautionary principle should be applied if there is doubt.

    report
  2. Andrea Young

    Master Data Analyst

    "Mining cannot be blamed for the Moe quake in the La Trobe Valley’s coalfields."
    Given the explanation of the Newcastle earthquake, why can't human mining activity be a factor in the Moe Quake?

    report
    1. Matthew Thompson

      Editor at The Conversation

      In reply to Andrea Young

      Dear Andrea,

      The La Trobe coalfields are open cut, and while Christian does not blanket exclude open-cut operations from causing seismic activity, he certainly does argue that underground coal mining - as Newcastle has a history of - played a big role in the 1989 quake.

      Yours,

      Matt.

      report
  3. Mike Sandiford

    Professor of Geology and Director of Melbourne Energy Institute at University of Melbourne

    … "The velocities of the tectonic plates in southern Australia are extremely small. Therefore we talk about this being a Stable Continental Region"

    Southern Australia does not comprise a number of plates, but is part of one plate and its velocity is not small. In fact the Australian continent is the fastest moving of all continents (6.5 -7 cm/year) and the south-east is the fastest part of our continent. 

    What is relevant is that the velocity gradients across Australia are extremely small…

    Read more
    1. Christian Klose

      Adjunct Research Scientist at Northwest Research Associates

      In reply to Mike Sandiford

      The so called Kaiser effect, that the rock mass can "memorize" its latest peak stress load is not necessary a general effect applicable for earthquake triggering per se, because may physical processes are involved that some times work against each other. Overall, human-triggered or induced earthquakes can rupture after a decade (e.g., Thompson reservoir). Every fourth event, however, nucleates after two years and the seismicity can last for decades, as the Koyna reservoir seismicity in India has…

      Read more
    2. Jonti Horner

      Vice Chancellor's Senior Research Fellow at University of Southern Queensland

      In reply to Mike Sandiford

      re: "Southern Australia does not comprise a number of plates, but is part of one plate and its velocity is not small. In fact the Australian continent is the fastest moving of all continents (6.5 -7 cm/year) and the south-east is the fastest part of our continent. "

      - I think the point is that all continental masses are made up of smaller "microplates", and it is the relevent motion of those plates on the buried/dormant faults that lie between them, that is often linked to earthquakes in the interior…

      Read more
  4. Keith Bradby, Gondwana Link

    logged in via email @gondwanalink.org

    Good article Christian, Thank you

    One good Australia earthquake was missed from your analysis, which I gather some experts in the field suggest may have been mine related. On 20 April 2010 the City of kalgoorlie Boulder suffered a 5.0 earthquake, no lives lost but severe damage to some grand old pubs. It has been suggested the close proximity to the kalgoorlie Super Pit may have been part of the cause, particularly as the main damage was very localised in Boulder. Leads to some articles below, curious to know if there are any refereed papers out by now
    http://www.australiangeographic.com.au/journal/earthquake-rocks-kalgoorlie.htm
    http://au.news.yahoo.com/thewest/business/a/-/diggers/7093175/super-pit-linked-to-quake-cause/

    report
    1. Christian Klose

      Adjunct Research Scientist at Northwest Research Associates

      In reply to Keith Bradby, Gondwana Link

      Edward Cranswick works on this event. He presented at the Australian Earthquake Engineering Society 2011 Conference last year "Coincidence of Mines and Earthquakes in Australia. For his efforts, I hope he/someone gets tax payer's science-dedicated money for conducting a research study with a final peer-reviewed paper.

      report
  5. Chris Sanderson

    CEO

    Bill McGuire has written a good book on this subject called: 'Waking the Giant - How a changing climate triggers earthquakes, tsunamis and volcanoes'. Recommended reading....../Chris

    report
  6. Peter Boyd Lane

    geologist

    A most interesting and relevant article.
    Based mainly on the assessed risk to both ground and surface water a proposal to mine coal near Margaret River (WA) has recently been rejected and we are now trying to get legislation in place to protect the region from both mining and coal seam gas operations.
    Although proposed as a “boutique” underground mine of just 1.2mmt/yr coal, based on analogies we estimate water influx could be up to 10 gigalitres/yr, total mass removal of say 2mmt/yr. Much of the water would be replaced by aquifer recharge etc, but assume net 1.5mt/yr mass removal, over 20 yrs some 30mmt could be removed … and this would be just from one mine. Although seismic activity is rare, seismically mapped faults are common and there is evidence of fault planes being open.
    I believe your discussion can materially assist us in getting protective legislation for our region.

    report
  7. Peter Boyd Lane

    geologist

    Did I make an error by a factor of 10? 10 Gigalitres/yr = 10x10^9 litres = 10x10^6 tonnes, ie total mass removal of 11.2 mmt/yr, not "say 2mmt/yr".
    I would be grateful if someone could check my maths (you can understand why I did geology and not one of the stricter sciences)

    report
  8. Kayt Davies

    Senior Lecturer, Journalism at Edith Cowan University

    Great article :) I have wondered if ConocoPhillips Bayu Undan oil field near Timor had anything to do with the 2004 Aceh Tsunami but have yet to find a sensible answer why it couldn't have been. It seems logical that the process of replacing millions of barrels of undersea oil with water may have caused a bit of movement on the sea floor. Do you know if the possibility was researched?

    report
    1. Mark Duffett
      Mark Duffett is a Friend of The Conversation.

      logged in via Twitter

      In reply to Kayt Davies

      For a start, Bayu Undan would have to be something like 1000 km from Aceh. But in any case, no. 'Millions of barrels' still only equates to a minuscule proportion of the total rock mass. The Aceh tsunami can be entirely accounted for within the framework of natural processes (which is why there have been tsunamis just as large and larger long before there were such things as oil fields).

      report
  9. Blair Donaldson

    logged in via Facebook

    Coal mining being responsible for seismic activity makes much sense, Korumburra in South Gippsland has been a hotspot for tremors in recent years, it was also a site for coal mining, it was an important source of black coal in the late 1800's.
    http://www.theage.com.au/environment/big-victorian-tremor-rocks-gippsland-suburbs-20110705-1gzzf.html

    Unfortunately the dinosaur who poses as our Premier thinks coal mining and CSG is the state's future while the rest of the world is moving to clean energy sources.

    report