Tailings Dam Failure

fj United Nations Environment Programme
ji gv
Programme des Nations Unies pour I'environnement
UNEP Industry and Environment / Industrie et Environnement
Tour Mirabeau, 39-43, quai André Citroén, 75739 Pans Cedex 15, France
Tel. (33-1) 44 37 1450 • Facstmilé: (33-1) 4437 1474 • Telex 204 997 F • E-mail : [email protected]
Environmental and Safety Incidents
concerning
Tailings Dams at Mines
Results of a Survey for the years 1980-1996

by
Mining Journal Research Services

The Mining Journal Ltd
60 Worship Street
London EC2A 2HD, UK
May 1996
0
w u
0
A report prepared for
\ 'eRP oj
ANO
AT
Joint UNEPIDHA Environment Unit

United Nations Environment
Programme
Industry and Environment Department of Humanitarian Affairs

Tour Mirabeau, 3943 Quai André Citroën Palais des Nations
CH 1211 Geneva 10, Switzerland
75739 Paris Cedex 15, France
100 % recyc'ed paper

PREFACE
Several recent incidents concerning tailings dams were regarded as environmental emergencies
by the countries in which they occurred. Both UNEP and DHA were called upon to provide
an evaluation of the extent of the disasters, and to give technical assistance and advice in the
follow-up phase. During this work UNEP and DHA became aware that environmental
impacts from tailings dams occur more frequently than is sometimes thought, and that the
consequences are not always well understood.
In order to put these incidents into a clearer perspective the two organizations carried out a
survey of recent incidents with a view to:
- identifying the environmental and safety impacts most commonly encountered,

- the actual frequency of such incidents,
- the environmental and human consequences of such incidents.
The task of undertaking the survey was entrusted to the Mining Journal Research Service,
which augmented information in its database by way of a questionnaire to national authorities
around the world. Organizations concerned with tailings dam safety volunteered their
assistance, notably the International Commission on Large Dams (ICOLD). The additional
information thus provided proved invaluable.
We are now able to provide better researched answers to the three questions above.
However despite the useful results obtained in this study a complete quantification of tailings
dam incidents remains elusive. Data from some countries has been difficult to obtain.
Different interpretations persist as to what constitutes an "incident". UNEP and DHA, with
the assistance of other organizations, will soon decide how to follow up this study and refine
the information.
In the meantime the current report provides a useful overview of the situation in major mining
countries, and allows companies and government authorities to better identify the potential
impacts from their tailings operations. UNEP intends to incorporate the conclusions into its
ongoing work in producing technical publications, and recommending training curricula, and
in providing to government officials information to develop their policies to address mining
and environment issues.
ACKNOWLEDGEMENTS
UNEP and DHA would like to thank the staff of the Mining Journal Research Service,
especially Dr. W. Prast and Mr. S.Walker for their thorough work in completing this survey.
Thanks also go to the many respondents and reviewers who assisted in this work, including
in particular Dr. A. Penman and the ICOLD Committee on Tailings Dams.
The Environmental Health & Safety Division of OECD assisted in the project administration.
For further information contact UNEP IE, 39 Quai André Citroën, 75739 Paris Cedex 15, France, or
Fax (33-1) 44 37 14 74, or e-mail [email protected]

TABLE OF CONTENTS
1.0 Introduction ..............................................1

1.1 Scope and methodology ...................................4
2.0 Review of worldwide mining activities ............................ 7

2.1 Copper ............................................... 7
2.2 Gold ................................................ 7
2.3 Lead and zinc .......................................... 9
2.4 Bauxite ............................................. 10
2.5 Fertilizer minerals ...................................... 10
2 .6 Coal ............................................... 11
2.7 Uranium ............................................. 11
2.8 Global context ........................................ 12
3.0 Review of construction methods ................................13

3.1 Tailings dam types .....................................13
3.1.1 Cross-valley .....................................14
3.1.2 Valley side ......................................14
3.1.3 Valley bottom ....................................14
3.1.4 Ring ..........................................14
3.2 Methods of tailings dam construction .........................15
3.2.1 Upstream construction ..............................15
3.2.2 Downstream construction ............................15
3.2.3 Centreline construction .............................15
3.2.4 Water retention ...................................16
3.3 Key factors in tailings dam construction .......................16
3.3.1 Water control ....................................16
3.3.2 Foundation material ................................17
4.0 Causes of tailings dam failure ..................................19

4.1 Breached retaining walls ............................19
4.2 Earthquake effects .................................19
4.3 Flooding .......................................20
4.4 Decant failure ....................................20
5.0 Environmental effects ........................................23

5.1 Groundwater contamination ..........................23
5.2 Dust emissions ...................................23
5.3 Other environmental impacts .........................24
5.4 Community location ...............................24
I
TABLE OF CONTENTS
6.0 Significant failures and other incidents since 1980 ...................27

6.1 Summaries of responses .................................. 27
6.1.1 United States .................................... 27
6 .1.2 Canada ........................................ 33
6.1.3 Latin America ................................... 34
6.1.4 Europe ......................................... 34
6 .1.5 Africa ......................................... 34
6 .1.6 Asia .......................................... 35
6.1.7 Australia and New Zealand .......................... 36
6.2 Tailings dam populations ................................. 37
6.3 Types of incidents ...................................... 39
6.4 Attitudes towards tailings dams ............................. 42
7.0 Conclusions ...............................................45
8.0 Recommendations ..........................................49
9.0 Bibliography ..............................................51
Appendix
Specific incident details
LIST OF TABLES
1 Selected major mining districts by commodity ........................8

2 Selected tailings dam incidents 1965-1979 .......................... 28
3 Selected tailings dam incidents 1980-1996 ..........................29
4 Selected tailings dam populations ................................38
5 Incident type by location ......................................40
EXECUTIVE SUMMARY
This study has been prepared by Mining Journal Research Services at the request of the
United Nations Environment Programme (UNEP). It reviews the occurrence of reported
failures and other similar incidents involving mine tailings dams in a number of countries
during the period from 1980 to early 1996. The study contains an overview of dam designs
and construction techniques in current use.
Placed in a global context, the study is based on a survey of tailings dams in countries that
account for 65% of global gold production, some 75% of established market economy country
(EMEC) newly mined copper output, and around 60% of EMEC lead and zinc. The countries
covered are also responsible for significant proportions of world production of industrial (non-
metallic) minerals and coal. In total, 52 organisations were contacted in 18 countries. Of this
group, 23 agreed to participate and were sent questionnaires; 20 replies were received.
The term "tailings" is defined as waste products that are generated during the recovery of
mineral commodities from their ores or other sources. The term "tailings dam" is used in its
widest sense, to include both the retaining structure and the impounded material. The
recovery of most metals and non-metallic minerals invariably results in the production of
large amounts of unwanted waste material, often ground to particle sizes of below 100
microns in order to release the valuable constituents. The production of non-metallic minerals
can also result create tailings, as does the upgrading of coal.
The tonnage of tailings generated each year is huge but tends to be concentrated into specific
areas of the world. World copper production in 1995, for example, required the treatment of
upwards of 1,500 Mt of rock, virtually all of which was stored in tailings dams. Annual gold
output requires the production of around 500 Mt of tailings, plus obsolescent leaching heaps.
For potash, over 140 Mt of potash-bearing ore was mined, and around 120 Mt of tailings,
mainly consisting of salt and insoluble clays, would have been produced during 1995.
Tailings dams and other impoundments are essentially a 20th Century invention. Earlier,
residues from mineral recovery operations were generally discharged either on to the surface
close to the mine concerned, or into the nearest watercourse for disposal downstream. In
some cases, river or marine disposal still occurs, but is increasingly a target for environmental
concern. In contrast to earlier practice, the modern tailings impoundment is usually of highly
sophisticated construction based on firm geotechnical foundations. The design and
construction of tailings dams has evolved considerably over the last 20 years. There is also
a large population of tailings impoundments that date from earlier times, however, some of
which are still in use while others are now abandoned.
In such circumstances, incidents occur from time to time that have some impact on the
surrounding environment. Such cases may range from the release of small amounts of dirty
water, or dust emissions from dry tailings surfaces, to full-scale collapses. Fortunately, large-
scale incidents are infrequent. The research for this study has identified fewer than ten major
failures or similar incidents that were reported between 1980 and 1996.
The level of public reporting and discussion following a major failure is very variable. There
is currently no centralised source for information about tailings dam incidents, either
111
internationally or, for the most part, nationally. The professional interest of organisations
such as the International Commission on Large Dams (ICOLD), set up in the late 1920s to
monitor water-retention structures, has been extended to encompass tailings impoundments
as well.
There are a number of important factors that have a direct bearing on the stability of tailings
dams, and hence on their long-term retention capability. Most are concerned with water
handling in one form or another. Principal amongst these are the water control methods and
the foundation materials used.
Virtually all failures occur as a result of the presence of water. This may be as a result of
percolation through the dam wall, internal erosion, overtopping or flooding, and may be
exacerbated by natural phenomena such as earthquakes or persistent heavy rain. The majority
of the incidents reported in this study were attributed by respondents to such natural
processes, which in some circumstances may have provided a trigger for the collapse of
already weakened structures.
As few major mining countries have a single authority responsible for maintaining records
on tailings dams, it is rarely possible to estimate the true population of impoundments on a
national basis, and even less so world-wide. To illustrate the scale of the numbers,
respondents reported total (active and abandoned) dam populations in some individual
countries or regions as being 350 in the state of Western Australia, with a further 43 in
Tasmania. The Canadian provinces of Quebec, British Columbia and New Brunswick
reported 164, about 130 and seven dams respectively, while for the Republic of South Africa,
the dam population is estimated to be in the order of 400.
Collating data on tailings dam performance remains largely a matter for individual agencies,
with a wide variation in the amount and quality of information held. Reorganisation of
government departments has in some cases also left a vacuum regarding information gathering
and the maintenance of records, and breaks in the continuity of the collation process.
Incidents that occur in more remote areas, or in countries that remain closed to outside
inspection, are much less likely to be recorded than those involving industrialised counties
or where companies with international operations are involved.
The information presented here has the potential to form the foundation of a comprehensive
store of tailings dam performance data. It is not complete but does contain a range of
information from numerous, usually reliable sources and may be used by the UNEP to foster
an improved information-sharing effort in this field. Other specific areas in which UNEP's
position could be used to advantage in addressing safety issues pertinent to tailings dams
include continuing the statistical compilation of dam numbers on a world-wide basis, and
assisting national authorities in advising communities perceived to be at risk from potential
dam incidents.
iv
1.0 INTRODUCTION
This study has been prepared by Mining Journal Research Services at the request of the
United Nations Environment Programme (UNEP). It reviews the occurrence of reported
failures and other similar incidents involving mine tailings dams in major mining regions
during the years since 1980.
"Tailings" are defined as waste products that are generated during the recovery of mineral
commodities from their ores or other sources. Typically, the original rock has been crushed
or ground to small particle sizes in a wet recovery process. The term does not encompass
waste rock produced during the mining process in order to uncover or access the ore; such
material is usually handled at as large a particle size as possible and is normally stored
separately from process tailings, close to the mining area. Waste rock may, however, be used
to construct retention dam walls for tailings impoundments, for example, but is generally not
mixed with the finer material. "Slimes" is sometimes used as a synonym for tailings,
particularly in southern Africa.
The terms "tailings dam" and "tailings impoundment" have different connotations in various
countries. In some areas, tailings are retained in an impoundment, pond or lagoon that is
formed by a dam wall; in others, a tailings dam refers to the dam structure and the
impoundment area as well. For the purposes of this study, "tailings dam" is used in its widest
sense, to include both the retaining structure and the impounded material.
The UNEP view is that an "incident" consists of an unplanned event that has some effect on
humans, the environment and property. This includes both dramatic structural failures, which
are usually short-term in occurrence (although the effects may be long-lasting), and persistent
low-level discharges and emissions, such as groundwater seepage and dust. This report
focuses mainly on larger-scale incidents that occurred over a brief time period and had a
greater impact on people and the environment. Smaller incidents, such as minor structural
slips in a dam wall, or short-term discharges of process water, are often contained within the
mine property and are addressed by the company involved, with little outside impact or
publicity.
In a development that is tangential to this study, the use of acid or cyanide leaching on lower-
grade ores ("heap leaching") has increased significantly since 1980. Here, run-of-mine ore
is crushed and then stockpiled for leaching on an impervious liner; when the leaching process
has reached its economic limit, the heap is no longer of economic value. Despite being
I
process waste, such obsolescent heaps are not "tailings" per se on account of the larger
particle size of rock they contain. Moreover, the free-draining characteristics of these heaps
means that there is minimal water retention after leaching has finished, and they are thus
much more stable than impoundments containing finely ground tailings; the latter may be
saturated with water long after abandonment.
The recovery of most metals and non-metallic minerals from their ores or other sources
usually involves some type of mechanical or chemical treatment. Given that the proportion
of any mineral within its natural host rock is usually low (ranging from a few parts per
million for rare metals such as gold to below 10% for most other metals), recovery through
mineral processing invariably results in the generation of large amounts of unwanted material,
often ground to particle sizes of below 100 microns in the course of releasing the valuable
constituents from the ore.
It is evident that recovering gold from ores that contain a few grammes of metal in a ton of
rock results in the need for the long-term storage of the total amount of rock mined, to all
intents and purposes, albeit in ground form. A copper ore grading 0.5% metal will similarly
require the storage of at least 99.5% of the mined rock. Bulk commodities such as iron ore,
which naturally have a higher metal content of between 45 and 65%, are also upgraded before
shipment, resulting in a lesser proportion of finely ground process waste, but a tonnage that
is still very significant.
Processing industrial, or non-metallic minerals can also create tailings. Phosphate fertilizer,
for example, is extracted from phosphate ore by a two-stage chemical process, both of which
involve the production of tailings. Even a simple item such as construction aggregates often
involves washing as-dug gravels to yield a graded product and to remove clays and soil from
the raw material.
Similarly, the upgrading of coal through the removal of non-combustible mineral matter, or
through sizing to produce a specification product, also produces process tailings. In this case,
the tailings contain both finely divided waste and fine coal. Tailings disposal practice in the
coal industry can differ from other branches of the minerals industry in that process waste is
sometimes re-combined with other mining waste in order to minimise the risk of spontaneous
combustion and to enhance dump stability.
In contrast to earlier practice, the modern tailings impoundment is usually the result of
sophisticated civil engineering and construction techniques based on firm geotechnical
I
foundations. The design and construction of tailings dams have evolved considerably over
the last 20 years. There are, of course, many tailings impoundments that date from earlier
times, some of which are still in use while others are now abandoned, either on active mine
sites, or in areas where mining is no longer practised.
In such circumstances, incidents will occur from time to time that impact on the surrounding
environment. These may range from the release of small amounts of dirty water, or dust
emissions from dry tailings surfaces, to a full-scale catastrophic collapse that may have major
environmental and economic impact, cause loss of human life, and be widely reported in the
media, as well as having the potential for loss of human life.
Fortunately, large-scale incidents are infrequent. Research during this study has identified
fewer than ten major failures or similar incidents reported during the period from 1980 to
early 1996. Of these, only four for which detailed information is available - at Stava in Italy,
Jinduicheng in China, Placer in the Philippines and Harmony in South Africa - each involved
the loss of more than ten lives. One further incident, in Brazil, resulted in seven fatalities,
apart from which there are no other records of deaths. In the 15 years between 1965 and
1980, however, there were at least five other reported large-scale incidents that involved mine
tailings or other waste, four of which incurred significant casualties while the fifth caused
extensive low-level radioactive contamination.
It became evident during this research that the level of public reporting and discussion
following a major incident is also very variable. The most recent well-publicised discharge
from of a major tailings impoundment, at Omai (Guyana) in August 1995, focused much
attention on the safety of such structures, although it did not cause the loss of human life.
By contrast, the collapse of a tailings dam wall at the Harmony gold mine in the Republic
of South Africa in 1994, which resulted in 17 deaths, received far less publicity despite
having a greater human impact.
As varying regulations apply from country to country regarding construction methods for
tailings dams, so the requirements for reporting incidents involving tailings also vary widely.
Even if a comprehensive control policy did exist, implementation is only as effective as the
inspection organisation responsible for it. Some countries still have no effective control over
tailings disposal, while others permit discharges into rivers or to the sea rather than tailings
impoundment. Examples include fluvial tailings disposal at the Ok Tedi copper-gold mine
in Papua New Guinea and from silver and base metal mining in Bolivia, and marine disposal
from the (currently shut) Atlas copper mine in the Philippines.
3
1.1 Scope and methodology
This study aims to identify and to annotate major incidents involving mine tailings dams since
1980. Comprehensive, global information was not readily available, and it is conceivable that
a major incident that occurred in the former Soviet Union or in China, for example, may have
gone unreported in the international media. It is much less likely that incidents of lower
significance in such countries reached world attention.
Although mining in some form takes place in almost every country, it was not feasible, within
the time constraints of the study schedule, to investigate every country with a mining industry
of a sufficient size to require major tailings dams. The emphasis has therefore been on
researching as completely as possible the occurrence of incidents in a cross-section of
countries, including major minerals producers and other states in which mining is also
important but is of lower economic significance.
There is currently no centralised source for information about tailings dam incidents, either
on an international basis, or for the most part, nationally. The closest attempts to this type
of collation of information are to be found in the U.S.A. at the United States Committee on
Large Dams (USCOLD) in Denver, Colorado and at Stanford University, Palo Alto,
California. Both have comparatively comprehensive listings of reported (and in some cases,
anecdotal) incidents, although the emphasis at both organisations is on tailings darns in the
U.S.A.
General information about incidents has been obtained from the records of The Mining
Journal and from other literature searches, as well as from the American sources noted above.
More specific details of individual incidents has been sought from state authorities in
Australia and the U.S.A., from provincial authorities in Canada, and from the appropriate
government departments in 14 other, selected countries. In some cases, mining companies
and engineering consultants have been asked to provide details of incidents, although
responses obtained from interviews have in some cases been guarded in view of unresolved
regulatory issues. Details on methodology and the organisations that were contacted for
information are included in the Appendix volume of this report.
Much of the information that is contained in this study has been generated from
questionnaires that were completed by identified respondents in the appropriate agency or
government organisation. It should be noted that the level of detail provided varies
considerably from reply to reply, and may reflect the detail that individual government
departments are required to hold. In some cases, different organisations may hold information
4
on different aspects of an incident, making the search for a complete record that much more
complex.
The information obtained from completed questionnaires and other sources forms the basis
for the data presented in tabular and text form in Chapter 6 of this report, and for the 37
individual incident reports contained in the Appendix. A discussion of the level of confidence
that may be placed in this sample of incidents for which detailed information is available is
presented in Chapter 2.

2.0 REVIEW OF WORLDWIDE MINING ACTIVITIES
Any mining operation that recovers only a component of the ore mined will produce some
form of tailings. In recent times, the introduction of increasingly sophisticated mineral
recovery techniques has meant that tailings have become finer in size, while the expansion
of mineral output has given rise to the generation of greater annual volumes of tailings
material.
Mining can also be categorised by its location, the size of the operation and the type of
commodity. Selected large-scale mining areas, by country and commodity, are listed in Table
1, which shows that there is a clear geographical concentration of large-scale activity. It
follows, therefore, that the principal concentrations of large tailings dams also occur in these
areas. Estimates of tailings production on a selected commodity-by-commodity basis are
given in the following sections.
2.1 Copper
Western World new mine copper production in 1994 was 7.6 Mt, most of which came from
low-grade "porphyry"-type deposits that contain from 0.4% to perhaps 0.8% copper. A
reasonable average is 0.5-0.6%. Thus the production of 7.6 Mt of copper metal, at an
assumed recovery rate of, say, 80% of the copper contained in the run-of-mine ore, required
the treatment of upwards of 1,500 Mt of rock, virtually all of which was stored in tailings
dams.
Notable exceptions to this are the Ertsberg mine in Irian Jaya (Indonesia) and Ok Tedi in
Papua New Guinea, neither of which has permanent tailings retention facilities. Other mines,
currently closed, that worked on this basis included Bougainville (PNG), Atlas (Philippines)
and Island Copper (B.C.). In each case tailings are or were discharged directly into rivers or
the sea.
2.2 Gold
Mine production was 2,296 t in 1994, from ores that grade a few grammes per tonne.
However, not all of this ore was processed by milling. For example, a considerable
proportion of the U.S.A. output of 331 t was obtained from heap-leach operations.
7
Table 1. Selected major mining districts by commodity
Country/region Districts
Copper
United States ArizonafNew Mexico/Utah

Mexico Sonora
Canada British Columbia
Zambia/Zaire Copperbelt
Chile Central/Northern
Peru Central/Southern
Southeast Asia IndonesialPhilippineslPapua New Guinea
Former Soviet Union Kazakstan
Gold
United States Nevada
Canada Northern Ontario/Quebec
Chile Central
South Africa Gauteng/Free State
Australia Western Australia
Southeast Asia Indonesia/PhilippinesfPapua New Guinea
Lead and zinc
United States Mississippi Valley
Canada British Columbia/Ontario
Peru Central
Australia Queensland
Spain Southern
Sweden Central/Northern
Iron Ore
Australia Western Australia
Brazil Minas Gerais/Amazonia
United States Minnesota
Canada Quebec/Labrador
Sweden Northern
Former Soviet Union Ukraine/Russia -
Mauritania Northern
Table 1. (continued)
Country/region Districts I
Phosphate
United States Florida/North Carolina/Idaho
Middle East Jordan
North Africa Morocco
Uranium
Canada Ontario/Saskatchewan
Australia Northern Territory/South Australia
United States New Mexico/Colorado/Wyoming
Southern Africa Namibia/South Africa
West Africa Niger/Gabon
Coal
United States Wyoming/ColoradolUtahllllinois/
Appalachia
Europe Eastern England/western Germany/Poland-
Czech Republic/northern Spain
Former Soviet Union Ukraine/RussialKazakstan
Southern Africa South Africa/Zimbabwe/Zambia
Asia Eastern India/Indonesia/China
Australia New South Wales/Queensland
Latin America Venezuela/Colombia/Mexico
Assuming an average run-of-mine ore grade of 4 g/t, which may be generous on account of
the large proportion of heap-leach production in the U.S.A. and surface-mined output in
Australia and elsewhere, in 1994 this total would have required the production of around 500
Mt of tailings, plus obsolescent leaching heaps. South Africa alone produces over 100 Mt/y
of gold ore tailings, the largest single concentrated accumulation anywhere.
2.3 Lead and zinc
New mined output of lead was just over 2 Mt in 1994, while zinc production was 6.6 Mt.
The pattern of output of lead and zinc is more complicated than for some other metals, in that

both are often co-products with copper and/or nickel. As a rule of thumb, it would be
unusual to mine a straight lead-zinc orebody unless the combined grade was at least 8%,
although this can be less if economic credits can be obtained from other metals in the ore.
Assuming a global average grade of 8%, which is probably high, the production of 8.6 Mt
of lead and zinc in 1994 would have required the treatment of around 140 Mt of run-of-mine
ore (assuming an 80% metal recovery) and the consequent disposal of some 130 Mt of
tailings in that year.
2.4 Bauxite
Producing aluminium metal involves a two-stage process in which alumina is first reduced
from run-of-mine bauxite ore, and is then smelted to yield metal. Bauxite output in the
established market economy countries was 97.7 Mt in 1994, which was treated to produce
35.4 Mt of alumina. Applying this ratio globally, it is estimated that the 110 Mt mined would
have yielded 40 Mt of alumina. This in turn would have left 70 Mt of tailings, or "red mud"
for disposal.
2.5 Fertilizer minerals
Fertilizer minerals include potash, phosphate and sulphur. Both potash and phosphate are
mined conventionally, and are then upgraded, while most sulphur is solution-mined from the
Frasch liquefaction process or is a by-product from oil and gas processing or metallurgical
industries.
Potash deposits, usually found in association with halite (common salt), typically contain up
to 35-30% K20, of which perhaps 80% is recovered through solution and differential
crystallisation technology. World 1(20 production in 1994 was 22.6 Mt. Thus over 140 Mt
of potash-bearing ore would have been mined, and around 120 Mt of tailings, mainly
consisting of salt and insoluble clays, would have been discarded.
World phosphate rock production was 124 Mt in 1994. Two types of ore are exploited: soft
and hard. Soft, sedimentary ores mined are typically high-grade, containing around 60% bone
phosphate of lime (BPL). There is some upgrading by washing and screening to produce a
phosphate rock concentrate. Tailings from this process, typified by Florida in the U.S.A.,
10
contain very fine clays that can present storage problems on account of their high water-
retention capabilities. Hard phosphate ores, as mined in Russia, are upgraded from an initial
20% to over 80% BPL.
Transforming the concentrate into phosphate fertilizer involves the intermediate stage of
phosphoric acid, of which phosphor-gypsum tailings are a waste product. The annual level
of tailings production (of both types) is not easy to estimate owing to the very individual
characteristics of each phosphate deposit; however, a figure of perhaps 50-60 Mt would be
representative.
2.6 Coal
Total world hard coal production was 3,181 Mt in 1994, plus an additional 1,270 Mt of brown
coal and lignite. Tailings are produced from washing coal, mainly to reduce its ash content
and enhance its heating value. Virtually all lignite and brown coal is used unwashed, while
the proportion of hard coal that is upgraded before use varies markedly from around the
world.
For example, coal mines in Queensland produced 111 Mt of raw coal in 1993-94, of which
85.7 Mt was classed as saleable, giving 24.3 Mt of discard or tailings. In New South Wales,
the corresponding figures were 84 Mt saleable from 101 Mt of raw coal, generating 17 Mt
of discard. However, in the Ruhr coalfield in Germany, the discard rate reaches around 50%
at some mines, meaning that half the run-of-mine output is set aside for disposal.
Coal washery tailings fall into two categories, however. The standard washery discard is
often of a comparatively large particle size (several millimetres and above) compared to
metalliferous tailings, for which a better comparison is the fine coal produced by screening
or spiral treatment to give a better sized product. The removal of coal fines often requires
their storage in lagoons or other impoundments, where they can sometimes form an energy
resource for later recovery with improved technology.
2.7 Uranium
Uranium is increasingly being produced from in-situ leaching operations, in which solution
mining has replaced conventional excavation and milling. As world uranium demand is flat,
11
the amount of tailings being created from uranium mining has been falling. Nonetheless,
there remain large volumes of uranium tailings from mining since the 1950s, which in some
countries, the U.S.A. for example, have become subject to extensive clean-up operations.
Given the potential for longer-term environmental impacts owing to radioactivity, uranium
tailings dams have received increased attention from all quarters. More emphasis has been
placed on adequate retention, curbing dust emissions and preventing fluid seepages, while in
some cases worked-out open pits have been used for tailings impoundment.
In addition to primary deposits, uranium is also produced as a by-product of phosphate mining

in the U.S.A., and as a co-product at some South African gold mines.
2.8 Global context
This study has drawn information from selected sources, and does not attempt to provide a
comprehensive list of tailings dam incidents on a world-wide basis. Nonetheless, the
countries and regions covered in the study represent a major proportion of world mineral
production. For example, between them Australia, Canada, the U.S.A. and Zambia, from all
of which information was received, represented over 40% of established market economy
country (that is, non-C.I.S. and China) copper mine production in 1994; if Chile, Papua New
Guinea and South Africa (from which information was requested but was not received) are
included, the proportion rises to over 75%.
For lead and zinc, the countries covered in the study represent nearly 60% of established
market economy country output. The comparative proportion is 65% of global newly mined
gold, while for phosphate rock the figure is around 35% of global production. Massive
production from China and the C.I.S. countries, for which direct information remains scarce,
skews the proportions for world coal output away from the countries covered, although as
noted above, coal tailings represent a different situation to those generated by metal mining.
12
3.0 REVIEW OF CONSTRUCTION METHODS
Tailings dams and other impoundments are essentially a 20th Centuty invention. Before that,
residues from mineral recovery operations were generally discharged either on to the surface
close to the mine concerned, or into the nearest watercourse for disposal downstream. In
some cases, river or marine disposal still occurs. Recent examples of a lack of tailings
retention include the Ok Tedi and Bougainville mines in Papua New Guinea. Mines such as
the Kilo-Moto gold operations in northeast Zaire still use local rivers to wash away their
residues. Tin producers in western England also traditionally discharged tailings to the sea,
while the former copper mines on Vancouver Island, Canada, and the Black Angel lead-zinc
mine in Greenland all used deep-water deposition for tailings disposal.
Tailings dam construction methods have advanced with time, incorporating better knowledge
of the geotechnical and hydrological parameters that influence dam performance. The study
of soil mechanics has been transferred to dam design, and the professional interest of
organisations such as the International Commission on Large Dams (ICOLD), set up in the
late 1920s to monitor water-retention structures, has been extended to encompass tailings
impoundments as well.
The earliest record in the USCOLD review of tailings dam incidents dates to 1917, when
reference is made to a failed tailings dam in South Africa. That this was noted in a paper
entitled "The Construction and Maintenance of Slimes Dams" suggests that South Africa took
an early lead in this area, probably on account of the very concentrated nature of the
Witwatersrand gold mines and the lack of surface water for tailings disposal. On reflection,
that has been of long-term economic benefit, given the success of the Ergo and other tailings
retreatment schemes that have recovered significant amounts of gold from these old tailings
dams.
3.1 Tailings dam types
Tailings dams are classified into four general types: cross-valley, valley bottom, valley side
and ring, or enclosed. A brief summary of each type follows.
Location of a tailings impoundment depends to a large extent on the natural topography.

Sites deemed suitable on this criterion must then be examined in detail to establish the nature
of the underlying rocks, their permeability and other factors.
13
3.1.1 Cross-valley
The simplest form of tailings dam, this involves the construction of a dam wall from side to
side of a valley. Provision must be made for the natural watercourse either to be diverted,
or to be piped beneath the dam wall. Such dams are clearly at increased risk in the case of
flash floods, or if the watercourse diversion system becomes clogged. As in all the valley-
type situations, several impoundments can be constructed, one topographically above the
other, to increase the total tailings retention capacity.
The Stava tailings impoundments in Italy that failed in 1985 were of this type, as is the
original tailings dam servicing the Los Bronces mine in central Chile, where it is reported that
a major failure was narrowly averted when the river diversion tunnel became blocked.
3.1.2 Valley side

This approach is used where the topography is gently sloping, the valley side or hillside
impoundment requires the construction of both front and side dam walls. While less at risk
from flash floods, an extensive surface water diversion system is needed on the higher side
of the impoundment to prevent excess water collecting behind the dam.
3.1.3 Valley bottom

A combination of the cross-valley and valley side types, these are used where a valley is too
wide or flat for a single dam wall to be constructed economically. Valley bottom
impoundments feature two dam walls, the longer lying parallel to the fall of the watercourse.
If constructed too close to this, there is the risk of erosion of the toe of the dam wall, as
occurred in the early 1990s at the Kojkovac tailings dam in Montenegro, former Yugoslavia.
3.1.4 Ring
Used in flat-lying topography, this type requires the construction of a complete retaining dam
wall, within which the tailings are stored. Internal dam walls may also be constructed to give
multiple deposition possibilities. South African and Western Australian gold mine tailings
dams are typically of this type, reflecting the flat-lying countryside overlying the goldfields.
14
3.2 Methods of tailings dam construction
A brief review of tailings dam construction techniques may be useful.
3.2.1 Upstream construction

The earliest system to be used, "upstream" construction, involved the formation of a low bund
or retaining wall from coarse material, behind which the tailings were deposited. As the pond
level approached the top of the wall, another layer was added on top of both the existing wall
and the edge of the tailings beach, and the filling process re-started. The process of
increasing the dam height could be repeated several times.
The disadvantage of this system is that there is a line of potential weakness at the base of
each lift, as the new wall is founded on both coarse and fine materials. This could enable the
wall to fail under the influence either of water springing through it, and eroding the dam, or
from the effects of a natural phenomenon such as an earthquake, which could cause
liquefaction of the impounded tailings. Nonetheless, upstream construction is still used, in
particular in dryer areas of the world where water levels in the impoundment are kept to a
minimum level through recycling, and where the risk of seismic activity is low.
3.2.2 Downstream construction

In view of these shortfalls, "downstream" construction of the dam wall has become more
widely used. The system has the advantage that part of the coarsest fraction of the mill
tailings can be used for dam wall construction, with the finer material being retained in the
impoundment.
An impervious barrier is often constructed on the pond side of the dam wall, and internal
drainage provided to ensure that the phreatic surface within the wall is kept well away from
its downstream face. The thickness of the dam wall increases in relation to the height,
providing additional stability, but requiring increasing amounts of material for its gradual
enlargement.
3.2.3 Centreline construction

This technique combines the advantages of both upstream and downstream construction
methods, while providing better seismic stability than the former and requiring less
construction material than the latter. Its water-retention capabilities are not as good as those
of a dam formed using the downstream system, however.
15
3.2.4 Water retention
Based on designs originally intended for water storage, this type of dam includes the use of
an impervious core in the wall, as well as filter and drainage layers within its downstream
face. Water seepage is thus minimised, but at the disadvantage of much higher construction
costs. It is also more difficult to increase the height of this type of structure should a larger
tailings retention volume be needed in the future.
3.3 Key factors in tailings dam construction
There are a number of important factors that have a direct bearing on the stability of tailings
dams, and hence on their long-term retention capability. Most are concerned with water
handling in one form or another.
3.3.1 Water control

Unlike water-retention dams, which are made of compacted natural fill material with an
impervious core, such as concrete, mine tailings dams are typically formed from either coarse
tailings or from locally excavated fill. Both materials permit water to permeate, and the
prevention of water erosion, either on the dam face or within the dam wall is paramount.
Thus modern tailings dam construction practice includes filter and drainage layers within the
dam wall.
Water contained within the dam wall and the impoundment behind will find a natural level
(the phreatic surface) that may vary with the seasons, the volume of tailings disposal, and
other factors. If the phreatic surface intersects the downstream face of the dam, a spring line
will develop, leading to rapid erosion of the dam wall material. Erosion in turn leads to
slippage of the face as the angle of repose of the now-saturated material is exceeded. This
moves the intersection point between the face and the phreatic surface further back into the
dam wall, and the cycle is repeated with the possibility of a breach being formed unless
prompt action is taken to secure the dam face.
Reasons for the initial springing can be many, but mostly concern inadequate drainage of the
impoundment and hence a higher phreatic surface within the tailings. Examples of causes of
this could include a blocked decant system, sudden heavy rain that overwhelms the decant
capacity, blockage to a spillway or watercourse diversion, poorly managed or unbalanced
tailings discharges, or the retention of too much water in the impoundment pond, perhaps on
account of its chemical contents.
16
Controlled seepage is often necessary to manage the water level behind and within the dam
wall, and to dewater the impoundment after tailings deposition is complete. The water-
retention capacity of finely ground material is high, and dewatering to a stable water content
takes years to achieve, even in and climates. Seepage can be contained through the use of
catchment structures downstream from the dam, or by pumping groundwater, with the effluent
being returned to the dam or purified before release. An example is the capture of saline
water that seeps from the tailings dams at potash operations in Saskatchewan, with borehole
pumps used to capture seepage that has entered local aquifers from beneath the impoundment.
Decanting from the impoundment pond is used either to recycle water to the processing plant
or to reduce the water level behind the darn. Decants take two principal forms; either a fixed
structure that maintains the water at a level corresponding to its top opening, or a pontoon-
mounted pumping system. Solid decant towers must be raised in height with each increase
in the dam wall, but have no operating costs. Floating pump stations can be used to regulate
the water level more effectively, but incur operating costs. A break in either the decant
outflow, which generally lies beneath the dam wall, or in the tailings inflow pipe, can lead
to rapid erosion of the wall. A number of dam wall failures have been attributed to internal
erosion caused by damaged or blocked decant pipes, as water swirling from the break caused
extensive internal potholing and eventual breakthrough of the impounded material.
3.3.2 Foundation material

The selection of an appropriate site for a tailings dam calls for favourable topography and the
presence of suitable foundation materials. Unsuitable rock material beneath a dam can lead
to failure of the overlying structure. Thus deep soil or sub-soil that are liable to become
saturated, or clays that could develop slips when wet, are not suitable foundation materials
for dam wall foundations, although clays could provide a naturally impervious barner beneath
the impoundment area. Similarly, the presence of excessive jointing, faulting or fractures in
the underlying rock could lead to long-term weakness and the potential for dam failure.
17
18

4.0 CAUSES OF TAILINGS DAM FAILURE
This study gathered information not only on catastrophic failures, such as dam wall collapses,
but also on lesser incidents. These include over-topping of the dam wall by water or fluid
tailings, accidental or uncontrolled discharges of fluids, and other situations where effluent
of one form or another from a tailings dam affected the environment either near the dam or
further afield.
Virtually all failures occur as a result of the presence of water. Examples of such water
action are percolation through the dam wall, internal erosion, overtopping or flooding. The
water movement may be triggered by natural phenomena such as earthquakes or persistent
heavy rain. The majority of the incidents recorded in this study are a direct result of such
processes.
4.1 Breached retaining walls
Breaches in dam walls occur for a number of reasons, most of which are situations in which
the phreatic surface within the wall rises until it intersects the downstream dam face (see
Section 3.3.1). Other possibilities include piping or 'rat-holing' within the dam wall, as a
result of leakages from decant pipes or spillways, for example, where water under pressure
causes internal damage that eventually weakens the wall structure to the extent that it can no
longer retain the tailings.
Rapid erosion of the dam wall material, which is usually graded sand, will result in breaches
through which the semi-fluid tailings can flow, assisted in the process by the water ponded
on the impoundment surface. This in turn causes further wall erosion, accelerating the rate
of escape.
4.2 Earthquake effects
Shock waves generated by earthquakes or tremors can devastate relatively unconsolidated dam
wall materials, as well as causing the impounded tailings to liquify. Some major tailings dam
disasters have been caused in this way, notably the incidents in Chile in 1965 and at
Mochikoshi in Japan in 1978.
19

The effect of such shock waves on semi-fluid materials has been well-documented. As well
as causing the tailings to liquify, repeated shocks can cause waves to form through the
impoundment, increasing the risk of overtopping the dam wall and giving rise to cyclical
higher pressures on the wall structure. Particularly at risk in these circumstances are walls
constructed using the upstream method, where the joint between individual lifts can form a
weak point in the structure.
4.3 Flooding
Torrential rain, such as occurs with typhoons, or even persistent heavy rain, can flood the
lower reaches of valleys as the increased flow from numerous watercourses combines. In the
upper valley reaches, flowrates in watercourses can increase significantly, leading to rapid
erosion of both natural material and man-made structures.
Most active tailings impoundments have a settling pond area in which there is standing water,
the level of which is maintained by some form of decant system. Water inflows to the pond
that exceed the capacity of the decant, or which result in blockage to the system, carry the
risk of overtopping the dam wall and its subsequent erosion. A case in point is a Chilean
copper operation in which a by-pass tunnel became blocked during heavy rain. Anecdotal
reference suggests that the security of a major tailings impoundment, situated above
residential and agricultural land, was only assured through the rapid construction of a new
concrete spiliway that permitted the storm water to be carried safely over the dam wall.
Other incidents in which flooding played a major role include the Stava dam failure in Italy
in 1985, the Harmony failure in South Africa in 1994, and several failures or potential failures
in the Philippines since 1980, in all of which rains associated with typhoons were a
contributory factor.
4.4 Decant failure
As noted in Section 3.3.1, decants can be either fixed or floating. Fixed structures present
a higher risk because inspection of the pipes beneath the tailings and the dam wall is not
usually possible. Thus any fracture in the pipes, perhaps as a result of movement of the
whole structure or through earthquake damage, increases the possibility of internal erosion
and 'rat-holing'.
20
Blockages to decant systems can also mean an increased risk of overtopping. In general, this
appears to be less of a problem with active dams, from which water is often recycled for
mineral processing, than with dams that have been abandoned and for which regular
inspection is less assured. Thus some incidents of water contamination have been reported
from Canada, although in each case the dam involved was in a remote location where mining
had ceased. Subsequent investigation of the downstream effluent drew attention to decant
systems blocked by natural debris or by pipe collapse.
21
ON
5.0 ENVIRONMENTAL EFFECTS
In addition to the direct results of wall failures, overflows and other incidents essentially of
a structural nature, gradual emissions from tailings impoundments may affect the surrounding
environment. Such impacts include long-term seepage that results in groundwater
contamination, dust emissions, and disturbance of wildlife habitat. It should be noted that
structural failures generally result in a rapid release of material over a short timespan.
Emissions, on the other hand, can continue for years, both during active use of the dam and
after mine closure.
5.1 Groundwater contamination
The movement of possibly contaminated water from an impoundment into underlying strata
depends on the relative permeability of the materials. In cases where there is a highly
permeable zone beneath the tailings, it has been known for fluids from dams to enter
underlying aquifers, forming a 'contamination plume' that spreads in the direction of aquifer
flow. The solution has been to install a ring of wells at an appropriate distance from the
dams, and to pump the groundwater back to the settling ponds.
It should be noted that seepage assists in drying the tailings in an impoundment, especially
after abandonment. In the dry climate of Western Australia, for example, seepage is regarded
as being inevitable and is perceived as having little effect on groundwaters that are already
highly saline. In some circumstances, however, seepage has been recorded as having caused
the groundwater table to rise, bringing saline water closer to the surface and damaging
vegetation.
5.2 Dust emissions
Normally only a problem when a dam has been abandoned, and also climate-dependent, dust
emissions can be a significant nuisance to the surrounding countryside as well as having the
potential for spreading hazardous materials such as asbestos fibre or heavy metals. Uranium
tailings pose a special concern in this context.
Solutions generally revolve around revegetation to stabilise the dam surface, but the success
of this step often depends on the nature of the tailings material, the water-retention capability
23
of the surface layer, and long-term husbandry of the project. Incidents in which dust has
played a role include the cobalt-rich pyrite dumps and tailings from the former copper mine
at Kilembe, Uganda, the surfaces of which have not revegetated successfully, and the dams
associated with the Xihuashan, Dangping, Piaotang and Xialong tungsten mines in China,
where dust emissions are believed to have caused cadmium contamination of the food chain
in neighbouring communities.
5.3 Other environmental impacts
Although not directly related to dam failure, ancillary impacts include the effect of strongly
mineralised or cyanide-bearing water on wildlife. Reference has been made to wildfowl
deaths at the Northparkes copper-gold mine in New South Wales as a result of a temporary
increase in cyanide levels in the tailings pond water.
The provision of safety closures over openings such as boreholes used for monitoring
groundwater levels and quality has also been highlighted as becoming of increasing
importance to prevent wildlife ingress.
It is also relevant to note that tailings dams can provide additional habitat for wildlife, with
abandoned impoundments forming a valuable wetland resource provided that the water present
is of suitable quality.
5.4 Community location
The location of housing near to tailings dam structures can present increased risk should a
failure or other incident occur. Examples include the Le Cobre failures in Chile in 1965, in
which tailings released following earthquake-induced dam wall collapses inundated a mining
camp below, while casualties that resulted from the Stava disaster in 1985 were also sustained
by the population living downstream from the dams.
The proximity of the South African township of Merriespruit to the Harmony tailings dam
undoubtedly increased the potential for casualties when the dam wall failed, and those
reported following the Jinduicheng breach in China were due to material flowing into
residential areas downstream. In fact, in each case within the review period where casualties
24
have been reported in the general population, rather than within the mining workforce, the
dead and injured were people living or working close to the failed structure.
Mining companies and government authorities vary in their ability to regulate where
communities are located. The existence of squatter camps set up by people seeking work
presents a particular problem, especially in countries with limited housing and social
infrastructure. Where the community at risk is directly employed by the mine, its relocation
is simpler, a recent example being the resiting of a residential camp serving the Argyle
diamond mine in Australia. Once it was realised that the mine's tailings dams posed a
potential hazard to the camp in the event of an earthquake, its complete relocation to higher
ground was achieved within three months.
25
FT.
6.0 SIGNIFICANT FAILURES AND OTHER INCIDENTS SINCE 1980
This section notes the tailings dam failures and other incidents that have been reported as
having occurred since 1980. Summaries of responses received are also included. Specific
details of incidents are listed in the Appendix volume.
Table 2 lists selected major incidents recorded in the period 1965-1979. Table 3 lists
incidents since 1980 that were identified during this study. Full information is not, however,
available on all of the incidents included in Table 3.
As might be expected, Table 3 shows that the number of recorded incidents for which details
are available increases in the 1990s. Specific information on dam incidents in the early 1980s
is limited to those that had the greatest impact, while in recent years not only is there more
information but the number of records is also significantly higher. This reporting reflects
increased environmental awareness as well as the adoption of expanded recording systems.
6.1 Summaries of responses
Initial approaches to collect data for this study were made mainly to government organisations
in 17 selected countries with an established mining industry. In some cases, the system for
maintaining records on tailings dams is fragmented, with several agencies or other
organisation holding partial information. In others, reorganisation of government departments,
or changing personnel, has meant that responsibilities have changed or records of incidents,
have been lost. Respondents who agreed to participate were
usually those in the 1980s,
provided with a questionnaire relating to each incident that occurred during the review period
for which records exist.
The following sections contain information on the various national agencies with
responsibility for tailings dams, and review the responses received.
6.1.1 U.S.A.
Responsibility for monitoring tailings dams is split between various agencies. The Mine
Safety and Health Administration (MSHA) has regulatory responsibility for active dams. The
Office of Surface Mining has responsibility for abandoned dams from the coal industry.
Abandoned dams from other sectors of mining are the responsibility of various state
27
9 9
V V
bo b0 bO' Os
(Q (O O
c0u C0.
2ts •
-
V —
. -tn • 0
E
•
-
<O 0
uj
E
U
U 1+-.c
° -2
V o g°
-r
.— . •0 —
Qo
ON 00
Ot
1U
=_D O <Oi
ON
2 00 — —
.- raw
cJ
4-
E O
0
0 U
PC
0 big
0
2
rA i
o U
—
0 V
•C
bOV
V Cb0 E • C . C)
PC
.E .
0
U U
- -O
cn U U
E(1 oo E 04
0
C.)
0
0 0 U 0
O Cl) 0
0 .0
a
.U
2 2
(4)
O • .0
.0 0
0 0
-.
00
'U
0 .
U . 0
040
V
00 00 O\ U
N N N N
O a
— — — — — — — z
= = = = =
cc o 0
o 0
•0 -c
— —
—
4)
EE
4)
; -
- -
0 0 s • 0 0 i Q 0
E 4) 00
00
P C
Cl)
tw 0 E 0 0 — w 4)
w .dE 4)
0 v C)
II!
0
I
-
- • i • L C.)
o C C
..2 cr
— CO
Cl)
co
a CO Cl) CI
Cl) L)
CO - . -
C.) OS) " Cl) Cl,
II
Sc.., - -
z CI)
4) •— 0 0 - — 0
4) CO _ , 0 = 4)
— -c 0
Cl) 4)4; — cO_ Z
- • 4) - - . 00 0
.E
4)
0 • . . . . S • S •
U
4)
0 0 0 00 00
00 00 00 00 00 00 00 00 00 00
ON o'o' C' 0\ 0
14 (4 21. — — — — — —
COOL
E
E U)
S S
cr
. s_°lu
E
-
to U) U)
E0
0e
'
-Q -
S S S S S
- > :
00
S 0 0 N 00 5 5 0 I 5 0
I
a
a
4! 10
00
.2
U
.
40
bOO
.2.
h
C
.- '-
00
4-. U
cqj
00
4.
U
=
0
00
(0
.2
.2
00
(0 .
o
0
—
00
.2
0
(0 .
0
ba
2
-
10
0
0.
3
(0
.2 CO
U • (5
* I tJ
-:
J
o o
.4!
o 0
I
U) 0
• • S • S • • • S • S S S
U
IrS S$ Ir 0 'O O NO '.0 '.0 N N N 00

(0 00 00 00 00 00 00 00 00 00 00 00 00 00
—
0' Q
— 0
— 0\
— C'
— 0\
— 0
— 0
— C'
— 0
— ,
—
0' C'.
— C'
—
E
• • • • • • • • I • • 2
PC
0
-.
H
4)
I
bo
S I I S $ I S S I I
. .g
• • • •
1-1
PC
00
= -
U I I
0
bo 4)
— ft If I 0
4.
Co
U,
4)
as
0 U
. .
'cc •
I
bb 00
4) - _ (I
. -
4)
Ow
— - 0 '*'
o
-
- c 2. 0 cc .3,
<QF- ta
• • • S S • •
4)
00 0\ 0 C9 r' c'4
44 00 00 00 00 Q\ ON 0% 0% 0%
c a o o a 0% 0% ON
- — — — — - — — * —
0%
— —
0%
— — —
2
2• .
I I S
P o
c •
E E oE
cn
• E
• o =
.-)ZZ
2
o o 0
SIIIS
2
e.3 .
. . •
=o
{I) -.
—
; 0
• f—i U Z
I I 0 00
bo V
C •—
cJ V 0
—
= 0 LV V •
—
2 u •
tI
E 2 =• 0 0 =
0
I_. bo
. _ I
>•
I 0
. •
cc
cn
cc
cn i2
E c
•Xo
S • • • S S
V
Irs 'rs irs .O

ON
agencies, with no federal organisation providing overall consistency of inspection procedures.
The Environmental Protection Agency could become involved in the case of a significant
incident, or if effluent from a tailings dam reaches watercourses. In addition, the Federal
Emergency Management Agency might also have a role in the case of a tailings dam-related
disaster, as might the U.S. Army Corps of Engineers.
Non-government groups that also hold information include the United States Committee on
Large Dams (USCOLD) and the Center on the Performance of Dams at the Department of
Civil Engineering, Stanford University, Palo Mto, California. Outline information was
received from MSHA, the National Mine Health and Safety Academy, USCOLD and Stanford
University. Detailed information has not been provided in most cases. The USCOLD
publication Tailings Dam Incidents is referenced to primary information sources where
possible.
The Buffalo Creek, West Virginia, incident in 1972 remains one of the most widely reported,
although it was not actually a tailings dam that was washed away, but rather coal mine waste.
Other failure types have included foundation collapse, overtopping, uncontrolled seepage and
structural weakness in dam walls; earthquake effects have been noted in water-retention dams
in California, but do not appear to have resulted in recorded damage to tailings dams.
6.1.2 Canada
Regulatory responsibility for tailings dams is held at Provincial government level, the Federal
government only having responsibility for those derived from uranium production. Responses
were sought from the appropriate authorities in the six major mining provinces of British
Columbia, New Brunswick, Newfoundland, Ontario, Quebec and Saskatchewan, and were
received from four: British Columbia, New Brunswick, Quebec and Saskatchewan.
None of the respondents noted any major failures since 1980, although there have been a few
structural failures in dam walls, and some cases of seepage, as at the Rocanville potash
operations in Saskatchewan and at Heath Steele in New Brunswick. Newfoundland reported
three minor incidents, all involving water damage to abandoned dams where the decant or
spiliway systems became blocked.
Responsibility for dams in New Brunswick is held by the provincial Department of the
Environment and the Department of Natural Resources and Energy. Tailings dam incidents
must be reported, as in British Columbia, where responsibility is held by the provincial
Department of Employment and Investment, Energy and Minerals Division, both for active
33
dams and for abandoned dams where care and maintenance has been neglected by the owners,
if known. The responsible agency in Quebec is the Ministry of Environment, to which
incidents must be reported.
6.1.3 Latin America

Requests for information were made to the authorities in Chile, but no response was received.
Situated in an active tectonic area, Chile has had a number of reported incidents in the past,
the most severe being in 1965 when the La Ligua earthquake caused damage to around 15
dams, resulting in over 200 fatalities from the collapse of the El Cobre Old Dam.
Chile's mountainous terrain has also favoured the past construction of cross-valley dams,
which are particularly vulnerable to flood damage both when active and when abandoned.
6.1.4 Europe
Requests for information were made to the authorities in four countries with a well-
established mining sector: France, Germany, Poland and Sweden. Information on Italy was
sought at the Department of Civil Engineering at Imperial College, London. No information
was received from France, Germany and Poland. No incidents were reported by the
respondents in Sweden, or by Imperial College.
With the exception of the Stava incident in Italy in 1985, no records have been identified of
tailings dam incidents in Europe. The Aberfan incident in Wales in 1966 involved mine
waste, not tailings, while a sand slide into a Spanish open pit iron ore mine in the mid- 1 980s
resulted from the failure of an overburden dump. In general, mining in Europe is dominated
by coal and construction raw materials, neither of which typically call for large tailings dams.
In terms of metallic minerals, tailings from French uranium mining probably hold the greatest
potential for environmental concern, but the lack of records of any incident involving this
sector suggests that controls are both adequate and well enforced. In terms of tonnage,
Sweden has the largest metallic mineral output in Europe on account of its iron ore
production, but no incidents were identified.
6.1.5 Africa
Requests for information were made to Ghana, South Africa, Zambia and Zimbabwe.
Responses were received from Zambia and Zimbabwe.
South Africa has the most extensive mining industry on the continent, and consequently has
34
$
a large population of tailings impoundments. Only two major incidents have been recorded
since 1970, at Impala Platinum in 1974 in which a breach in a dam wall caused tailings to
flood an operating mine shaft, and at Merriespruit (Harmony) in 1994. However, a press
report (Mining Journal, March 4, 1994, p.163) notes that on average there are at least two
slimes dam failures each year, most of which are small and do not involve loss of life.
Gold mining produces extensive slimes dams, many of which have been reclaimed and
retreated to recover residual gold. Coal washing also produces large amounts of coarser
waste, for which retreatment techniques are being developed. Past asbestos mining has left
both waste and tailings dumps that are perceived to present a health hazard from dust
emissions.
In Zimbabwe, which has smaller-scale and more scattered mines, reported tailings dam
incidents have been rare. With the exception of its asbestos industry, Zimbabwean tailings
dams are small. This may change once the Hartley platinum project comes on stream. The
country's most serious tailings incident, which occurred at the Arcturus gold mine in 1978,
was overshadowed in terms of publicity by the death of the mine manager during a payroll
robbery the same day. Responsibility for monitoring active dams is held by the Chief
Government Mining Engineer, and for abandoned dams by the Natural Resources Board.
The 1972 Mufulira disaster in Zambia, while involving tailings entering the mine through an
unstable surface area, was not the result of a dam or impoundment failure, but rather the
unforeseen consequences of tailings disposal in an inappropriate position above the mine.
Incidents that have occurred subsequently have predominantly been associated with heavy
seasonal rainfall. The Mines Safety Department and the state mining company, ZCCM, have
responsibility for monitoring both active and abandoned dams.
6.1.6 Asia
Requests for information were made to the authorities in India, China, Papua New Guinea and
the Philippines. No response was received from China or the Philippines, while the response
from India was that there have been no tailings dam incidents recorded during the review
period. The authorities in Papua New Guinea claim that information on tailings dams is
confidential and declined to respond.
Papua New Guinea hosts a number of major copper and gold mines of which two, Ok Tedi
and Bougainville, do not have tailings impoundments. The same is true at the Ertsberg
copper-gold mine in Irian Jaya, Indonesia. At Bougainville, currently closed, tailings disposal
35
was via a flume to the sea, while at Ok Tedi and Ertsberg, it is into local rivers. The original
tailings dam site at Ok Tedi was destroyed by a landslide in 1984, and subsequent studies
concluded that the construction of a permanent facility would be uneconomic.
Being in an area of significant tectonic activity, Papua New Guinea's terrain may be largely
unsuitable for tailings dam construction, as earthquakes and landslides could lessen long-term
dam stability. The same applies to the Philippines, where records would indicate that a
disproportionate number of incidents have occurred during the review period. Earthquake
activity is reported as being significant in a number of these incidents, while torrential rain
associated with typhoons has also been cited.
Information on failures in China is limited, and is often only reported well after the event,
when an incident has had an effect on international markets for a particular material for which
the country is a significant supplier.
Mining in Japan, meanwhile, has been decreasing. The most recent dam incident of
significance was the earthquake-induced collapse of two dams at the Mochikoshi gold mine
in 1978.
6.1.7 Australia and New Zealand

As in Canada, regulatory responsibility for tailings dams in Australia is at State government
level, with the Commonwealth government responsible only for tailings impoundments
derived from uranium mining. Information was requested from the appropriate agencies in
seven Australian states: New South Wales, Northern Territory, Queensland, South Australia,
Tasmania, Victoria and Western Australia. Responses were received from all but Queensland.
Neither Northern Territory, South Australia nor Victoria recorded any incidents during the
review period. In New South Wales, there have been two cases of seepage and one
'unplanned discharge', although in each case the effects were contained within the mine site.
A respondent from Western Australia, the state with the most active mines, noted that there
have been several small-scale incidents since 1980, consisting of localised embankment
failures, shallow groundwater contamination and wind erosion, all of which were successfully
managed by the mining companies.
Other positive points identified as being significant in minimising tailings dam incidents in
Western Australia include the low rainfall and high evaporation rates experienced in the main
36
mining areas; that the majority of dams are free-standing structures built on virtually flat
ground with minimal catchment areas; that most dams are located in the upper parts of
watercourse catchments; and that virtually all starter walls are constructed from selected mine
waste, on which subsequent minor failures have little impact.
The agencies responsible for monitoring both active and abandoned tailings dams in Western
Australia are the state Department of Minerals and Energy and the Department of
Environmental Protection. Any incidents must be reported, and in 1993 the Department of
Minerals and Energy established a database to record details of tailings dams in general.
In Tasmania, regulatory responsibility for active dams is held by the Department of Industry
Safety and Mines. Together with the Department of Environment and Land Management, this
body is responsible for abandoned dams until they are rehabilitated to the satisfaction of the
Chief Inspector of Mines. They then become the responsibility of the leaseholder or
landowner. There is no legal responsibility to report dam incidents unless they are life-
threatening.
Mining activity in New Zealand is limited, and the responsible agency reports no incidents
involving dam failures. The continuing slippage being exhibited by the dam at Golden Cross
is unusual, and while the potential for failure is reported by the mining company involved to
be minimal, the long-term stability of the impoundment has yet to be proven.
6.2 Tailings dam populations
With very few major mining countries having a single authority responsible for maintaining
records on tailings dams, it is difficult to estimate accurately the true population of
impoundments on a national basis, and even less so world-wide. A representative sample
from responses received and other references is shown in Table 4.
Table 4 has been compiled mostly from information supplied by questionnaire respondents.
Differences in recording requirements used, and in the scale of the mining industry in each
country, province or state, are reflected in the level of detail available. To illustrate, the
Canadian province of New Brunswick provided data that are more precise than British
Columbia, with its larger, more diversified and longer established mining sector.
With no direct response received from South Africa, the total number of tailings dams shown
37
in Table 4 is based on literature source estimates, and reflects the scale of that country's
mining industry. Many of the dams are also of large capacity. By contrast, a respondent in
Zimbabwe reported a total of some 500 dams, of which around 450 are in the gold industry.
Given the small-scale nature of this sector in Zimbabwe, the capacity of most of these dams
will be low.
Table 4. Selected tailings dam populations
Country/region Active Abandoned Total
Canada
British Columbia c.30 c.100 c.130
New Brunswick 4 3 7
Quebec 27 137 164
Africa
South Africa c.400

Zambia 21 17 38
Zimbabwe c.500
Australia
Tasmania 26 15 43
Western Australia 210 140 350
The data presented in Table 4 represent a small sample obtained from questionnaire responses.
Major mining countries such the U.S.A. and Chile are not included, but clearly each has a
substantial number of tailings impoundments, as do the countries of the C.I.S. and, in all
probability, China. However, it is unlikely that a comprehensive census has ever been taken
of all tailings storage facilities in the U.S.A., for example.
It would be unwise to attempt an estimate of total world, or regional tailings dam populations
on the basis of such a small sample. Even within individual countries, the number of dams
found in one area is no guide to the complete population; thus the number of dams in
Western Australia (where the mining industry is centred on gold, iron ore and nickel), cannot
be used as the basis for extrapolation, given that the mining sector in other Australian states
is of a different capacity and produces different minerals.
6.3 Types of incidents
38
Respondents were asked to indicate the types of incidents that were recorded within their
jurisdiction. Information given is shown in Table 5, together with details of incident types
from other sources. Dust emissions are most prevalent in drier climates, while earthquake
damage is typically restricted to circum-Pacific countries.
In the case of damage caused by rainfall or river action, it is usually unclear from the
information available whether failure resulted from the water flow as a primary cause or if
the wet conditions exacerbated previously existing flaws in a dam's design, construction or
operation. It is unlikely that such a technical judgement can be made without extensive
investigation and interpretation of these findings. The task becomes particularly problematic
when dealing with old or abandoned impoundments, which may have been designed to past
criteria that were considered satisfactory at the time of construction, but which may not meet
current standards.
Even in cases where internal erosion appears to have been the principal cause of failure, it
may be impossible to specify the mechanism by which it occurred. A case in point is the
Omai (Guyana) incident in 1995. The official report on the failure of the dam wall concludes
that while it is clear that piping occurred in the dam core as a result of improper construction
methods used for the rockfill, the water and slurry remaining in the pond make it impossible
to determine when in the construction process the faulty work took place. Given that the
Omai dam was constructed within the last five years to current design criteria, this puts into
perspective the difficulty of determining exact causes of failures that have occurred in dams
built many years ago.
A further important consideration is that while mining companies now accept the
internalisation of costs associated with the treatment and storage of wastes, this was not the
case in the past. Societal requirements have changed. Whereas tailings dams built earlier this
century may have been constructed simply as a limited-life retention facility for the duration
of the operation it served, this is not now the case. More and more, authorities expect full
life-cycle costings and mine plans to include exit strategies that incorporate post-mining costs
such as are represented by tailings dam inspection programmes.
39
Table S. Incident type by location
Country/region Types of incident recorded

United States Dam wall failure
Decant failure
Flood damage
Groundwater pollution
Overflow
Potential dam instability
Surface water pollution
Uncontrolled seepage
Canada
British Columbia Dam wall failure

Flood damage
Overflow
New Brunswick Dam wall failure
Dust or gas emissions
Groundwater pollution
Newfoundland Decant failure
Saskatchewan Groundwater pollution
Latin America
Brazil Dam wall failure

Chile Dam wall failure
Earthquake damage
Flood damage
Africa
South Africa Dam wall failure
Decant failure
Overflow
Uganda Dust or gas emissions
Zambia Dam wall failure
Decant failure
Overflow
Zimbabwe Decant failure
Overflow
40
Table 5. (continued)
Country/region Types of incident recorded

Europe
Italy Dam wall failure
Montenegro Potential dam instability
Asia
China Dam wall failure
Indonesia Surface water pollution
Japan Dam wall failure
Earthquake damage
Papua New Guinea Dam wall failure
Flood damage
Philippines Dam wall failure
Earthquake damage
Flood damage
Australasia
New South Wales Groundwater pollution
Northern Territory Seepage
Tasmania Dam wall failure
Decant failure
Flood damage
Overflow
Western Australia Groundwater contamination
New Zealand Potential dam instability
41
6.4 Attitudes towards tailings dams
Public and official attitudes towards tailings dams differ from place to place, often reflecting
the importance and age of the mining industry. Overall, the environmental movement has
increased public awareness of mining. The following quotations are from specific responses
provided by various countries on attitudes towards the safety and potential environmental
impact of tailings dams.
Canada, British Columbia

"The public is very concerned about impacts from and the presence of mining. Government
is committed to promoting and regulating environmentally acceptable mining."
Canada, New Brunswick

"The potential to negatively impact the environment is high. From a general perspective,
most residents of New Brunswick are not aware of the specific safety and potential impacts
of tailings dams. However, they generally do not trust the position taken by government and
industry and require significant education to alleviate fears."
Canada, Quebec
"Safety of active tailings dams is very good and potential environmental impact is low. Some
abandoned tailings dams, more than 25 years old, present some safety risks and have a
moderate environmental impact. Many of them are currently being reclaimed and restored.
About C$8 million were invested in 1995 by the government and industry to reclaim and
secure abandoned mining sites."
Zambia
"Special dumps regulations exist under the Mines and Minerals Act. In recent years, there
has been increased pressure for harmonisation and implementation of environmental laws."
Zimbabwe
"Generally accepted as part of a mining-related economy. Mining activities only affect 1%
of the country's total area."
Australia, Tasmania
"The construction of tailings dams has always been controlled by the Mines Inspection Act,
where specifications must be approved by the Chief Inspector of Mines. The construction,
commissioning and decommissioning of tailings dams is also covered in any Environmental
42
Management Plan that is required prior to commencement of operations. The public is very
concerned about the environment, and standards are forever being revised as public perception
and expectation changes."
Australia, Western Australia

"The general public and government officials are well aware of the potential for both public
safety and adverse environmental impact of poorly constructed and operated tailings storage.
The response has been to make resources available to prevent such poor performance and
ensure suitable legislation is enforced."
43
ME
7.0 CONCLUSIONS
Tailings dams are an innovation of the 20th Century; before the advent of flotation technology
and the treatment of low-grade ores, metallurgical processes were carried out at much coarser
particle sizes and waste material was not impounded. By current standards, annual production
rates of both metals and industrial minerals were low.
Tailings dam design has evolved from rudimentary construction to current, highly engineered
standards. Improved knowledge in geotechnology, soil mechanics and associated fields has
assisted in reducing the risk of dam failure markedly, although incidents still occur as
unexpected or inadequately investigated conditions become apparent after construction. An
area for concern continues to be the potential for inadequate supervision of construction, as
noted by the New South Wales Mines Inspectorate:
"In regard to the construction of tailings dams we are concerned that while their design may
be adequate their construction may not be to spec (fications; construction is often peiformed
other than by the designer. We are considering imposing a mining lease condition requiring
certification that the construction meets design requirements."
Major failures of tailings dams are rare and are often attributed by respondents, at least in
part, to natural forces: earthquakes, flooding or exceptional rainfall. Of the most serious
incidents that have occurred since the mid-1970s (those involving loss of human life), natural
phenomena were associated with virtually all. In some cases, the effects may have been
contributory to other factors. The possibility of determining the degree to which natural
forces were the primary cause or exacerbated some previously existing weakness depends on
the evaluation of geotechnical information that may not be available.
Dam embankments are typically constructed from non-cohesive materials such as cycloned
tailings, earthfill or waste rock, all of which can be easily eroded by water. Failure of decant
systems, spillways and diversion conduits is a major factor in smaller incidents that have a
less extensive environmental impact, and can be a prime cause for localised dam wall erosion
through overtopping or springing. Provision of inadequately sized spillway or decant systems
to handle exceptional runoff from upstream catchment areas can also lead to wall failure.
This applies to both active and abandoned dams, damage to the latter often only becoming
apparent at some time after the event unless routine monitoring is carried out.
45
Recorded incidents involving uncontrolled seepage have become more common with the
increased use of cyanide in the recovery of gold, both in parts of the world where it had not
been used before, and in established markets. Being readily detectable in groundwater, the
presence of cyanide beyond the confines of an impoundment is a clear indication of seepage
beneath or through the dam. This may be caused by poor embankment construction or by
permeation through the foundation strata. In general, however, the installation of cut-off wells
and recycle pumps can prevent the contamination plume from moving far from the
impoundment.
Dust emissions can also present localised nuisance. It is uncommon for this to persist over
large areas, although examples such as dust blowing from the Kilembe tailings dams and
pyrite stockpiles in Uganda, and the widespread cadmium contamination that is understood
to have been emitted from Chinese tungsten mine waste indicate that this can occur. For the
most part, revegetation of the dam walls and former pond surface provides an adequate
remedy, although the tailings material must be made sufficiently fertile to permit plant
growth. There is also a cost involved in long-term maintenance.
This investigation noted a considerable amount of confusion about the true nature of tailings
dams and incidents involving them. For example, statements made in 1995 by advocates
opposing gold mining in western Turkey cited cyanide pollution as having resulted from the
dam failures associated with the La Ligua earthquake in Chile in 1965, and the Buffalo Creek
disaster in West Virginia in 1972. In fact, the Chilean dams were all formed during copper
ore processing, while Buffalo Creek involved coal mine waste; neither had any connection
with cyanide.
A further example is provided by one respondent to this study who cited the pollution caused
in the early 1990s by the Summitville, Colorado, gold mine as being from a tailings dam.
This is also incorrect, as the emissions from the mine derived from its heap-leach operations,
not from tailings. If these errors arise among professionals who are directly involved with
tailings dams, it is easy to appreciate even greater misunderstanding by the general public.
The collation of data concerning tailings dam performance remains largely a matter for
individual agencies, with a resultant wide variation in the amount and quality of information
held. Reorganisation of government departments has in some cases also left gaps in
information gathering and the maintenance of records. Similarly, personnel changes within
the relevant authorities can break the continuity of collation.
46
Incidents that occur in remote areas, or in countries that are selective in their outside contacts,
such as China, are much less likely to be reported than those in industrialised areas or where
companies with international operations are involved. The failure of Manila Mining's Placer
tailings dam in 1995 received little attention outside the Philippines; conversely, the escape
of tailings from Marcopper's open pit impoundment in 1996 was widely reported, principally
because a major Canadian mining company was involved.
The formation of databases by organisations such as the United States Committee on Large
Dams and the Center on the Performance of Dams at Stanford University is still at a
relatively early stage, tailings dams having taken a secondary role to their main task of
monitoring dams used for water storage and hydroelectric generation. Nonetheless, records
of tailings dam incidents have improved markedly since the mid-1980s, and such
organisations make a valuable contribution to the availability of information.
Tailings dams are a necessary constituent of modern mining practice. Incidents of varying
degrees of severity have occurred in the past, and will in the future. The task ahead is to
ensure that the risk of major incidents is minimised in relation to the benefits to be obtained
from mineral extraction. Improvements in design, construction and operating practice will
assist in achieving this, requirements increasingly being stipulated in the relevant regulations.
Australia's Northern Territory provides a suitable example:
"Present requirements for tailings dams in the Northern Territory require the application of
best practice principles, the use of qualed engineers for the design and construction phase
and an approved quality control program."
47
U-11
8.0 RECOMMENDATIONS
The information contained in this report has been provided principally by government
agencies and international organisations that collect data on tailings dam performance. As
mining is a truly global practice, in geographical terms the information presented in this study
is somewhat restricted. Furthermore, the ability of reporting agencies to provide reliable
information for parts of the world outside their areas of prime interest is hampered, in some
cases by a lack of historical data and in others by a reluctance on the part of authorities to
release what may be available.
The UNEP is in a special position to gather information that might otherwise be denied to
other researchers, particularly in conducting a census of tailings dams, identifying the agencies
responsible for their inspection, and gathering details pertaining to recent incidents that might
otherwise go unreported. This report is a step in that process.
The information presented in this report has the potential to be the foundation of a permanent
record of tailings dam performance data. However, these data must not used in isolation from
other sources, several of which have contributed to its formation.
It is also important to encourage international agreement on the action to be taken based on

the information acquired. For example, it may be that once sufficient data are accumulated,
analysis will indicate that it is better to direct resources to addressing perceived shortcomings
in the design and operation of tailings facilities in specific areas of the world, such as those
prone to earthquakes or tropical storms. There is also the potential for technology transfer
on this topic.
The principal international organisations that currently gather dam data are interested
primarily in water-retention facilities rather than tailings dams. It is strongly recommended
that the further collation of information in this area should include input from organisations
or individuals who have first-hand knowledge of the minerals industry, who can determine
the types of information that have most relevance to the overall strategy being adopted, and
who have the professional experience to be able to analyse the information collected in order
to ensure that it can be put to good use.
Great importance should be attached to ensuring the relocation of communities that have a
high risk of inundation should a tailings dam nearby fail. The fact that the largest number
of casualties arising from failures recorded within the study period were sustained by the
49
general population, and not mining company workers directly, underscores the potential
vulnerability of these people. Liaison between the UNEP and appropriate authorities in those
countries perceived to be most at risk from earthquakes or other natural triggers to dam
failures would help to save lives.
As part of this effort, the UNEP could advise authorities on appropriate measures to ensure
that communities considered to be at risk are adequately informed about the possible
consequences from potential spillages.
50
9.0 BIBLIOGRAPHY
Detailed information on specific tailings dam incidents is normally held in the archives of the
individual government agencies that hold regulatory responsibility for such structures. The
detail varies, reflecting current policy, the level of environmental awareness at the time of the
incident, and other factors. Official inquiries are usually convened only in the case of major
incidents, in which case a detailed report will be placed in the public domain. A recent
example is the Guyanese report on Omai:
Report of Conunission of Inquiry into Discharge of Cyanide and Other Noxious

Substances into the Omai and Essequibo Rivers. Guyana Geology and Mines Commission,
Georgetown, Guyana.
Collated information on tailings dam incidents is contained in:
Tailings Dam Incidents. United States Committee on Large Dams, 1616 Seventeenth Street,
Suite 483, Denver, CO 80202, USA.
Background information to tailings dam construction and performance can be obtained from
publications such as:
Design and Evaluation of Tailings Dams. Report No. EPA 530-R-94-038 (NTIS PB94-
201845). U.S. Environmental Protection Agency, Office of Solid Waste, Special Waste
Branch, 401 M Street SW, Washington, DC 20460, USA.
Guidelines on the Safe Design and Operating Standards for Tailings Storages.
Department of Minerals and Energy, 100 Plain Street, East Perth, WA 6004, Australia.
Ritcey, G M (1989). Tailings Management, Problems and Solutions in the Mining

Industry. Process Metallurgy Vol.6, Elsevier Science Publishers, ISBN 0-444-87374-0.
Tailings and Mine Waste 194. Proceedings of the first international conference on tailings
and mine waste '94, Fort Collins, CO. USA, 19-21 January 1994. A A Balkema, ISBN 90-
5410-364-7.
Vick, S G (1990). Planning, Design, and Analysis of Tailings Dams. BiTech Publishers
Ltd, ISBN 0-921095-12-0.
51
APPENDIX
List of 37 records included
Record No. Location
100 Arcturus, Zimbabwe

110 Church Rock, NM, USA
120 Heath Steele, NB, Canada
130 Tyrone, NM, USA
140 Sipalay, Philippines
150 Gray Eagle, CA, USA
160 Golden Sunlight, MT, USA
170 Aurora, NC, USA
180 Marga, Chile
190 Olinghouse, NV, USA
200 Cerro Negro, Chile
210 Veta de Agua, Chile
220 LaBelle, PA, USA
230 Stava, Italy
240 Mineral King, BC, Canada
250 Brazil (iron ore)
260 Rossarden, TAS, Australia
270 Story's Creek, TAS, Australia
280 Pico de Sao Luis, Brazil
290 Mankayan, Philippines
300 Montana Tunnels, MT, USA
310 Xishimen, China
320 Grays Creek, TN, USA
330 Jinduicheng, China
340 Silver King, ID, USA
350 Iron Dyke, BC, Canada
360 Padcal, Philippines
370 Kojkovac, Montenegro
380 Itagon Suyoc, Philippines
390 TD7, Chingola, Zambia
400 Harmony, South Africa
410 Riltec, TAS, Australia
420 Middle Arm, TAS, Australia
430 Omai, Guyana
440 Placer, Philippines
450 Golden Cross, New Zealand
460 Marcopper, Philippines
Note: This system of numbering records has been adopted to simplify the insertion of
additional information as it becomes available.
Al
Methodology
Collecting information for this study involved a three-stage process. Appropriate countries
were firstly selected on the basis of the contribution made to their national economy by
mining, and potential respondents were identified from the Mining Journal's own lists, and
from lists of names supplied by UNEP. Efforts were then made to contact these potential
respondents, with the aim of seeking either their agreement to complete the third stage of the
process, the questionnaire, or to assist in addressing enquiries to better qualified sources.
Initial contacts were made by either telephone or fax to 46 potential respondents in:
Australia (seven states), Brazil, Canada (six provinces), Chile, China, France, Germany,
Ghana, Guyana, India, Italy, New Zealand, Papua New Guinea, Philippines, Poland, South
Africa, Sweden, U.S.A. (state and federal agencies), Zambia and Zimbabwe. Other
organisations contacted included the International Labour Organisation in Geneva,
Switzerland, and the Mineral Policy Center, Washington, D.C., and the geotechnical
engineering consultants, Klohn Crippen, Golders Associates, and Steffen, Robertson and
Kirsten. Contacts were initially requested to indicate whether any tailings dam incidents had
occurred in their country or region since 1980, and if so, whether information existed that
could form the basis for a questionnaire response.
Questionnaires were sent by fax to those respondents who indicated that such data is
available. Nine respondents who replied initially that no incidents had occurred in their area
of responsibility did not receive a further questionnaire. Those parties who did not respond
to the initial invitation to participate were recontacted, usually by telephone. Questionnaires
were not sent in cases where no further interest to take part was shown.
The following organisations were contacted in the initial stage:

Australia
Commonwealth Department of Primary Industries and Energy
New South Wales Department of Mineral Resources
Northern Territory Department of Mines and Energy
Queensland Department of Minerals and Energy
South Australia Department of Mines and Energy
Tasmania Development and Resources
Victoria Department of Agriculture, Energy and Minerals
Western Australia Department of Minerals and Energy
Brazil
Departamento Nacional de Producao Mineral
Canada
Mining Association of Canada
Centre for Resource Studies, Kingston, Ontario
Mine Environment Neutral Drainage Program
British Columbia Ministry of Energy, Mines and Petroleum Resources
New Brunswick Department of Environment
Newfoundland Department of Mines and Energy
Ontario Department of Northern Development and Mines
Ministère du l'Energie et des Ressources (Quebec)
Saskatchewan Industrial Branch, Environment and Resource Management
A2
Chile
Steffen, Robertson and Kirsten, Santiago
China
Research Centre for Geo-Environmental Sciences
France
Department of Environment and Energy
Germany
Federal Institute of Geosciences and Natural Resources
Ministry of Environment
Ghana
Minerals Commission
Guyana
Golden Star Resources Inc
India
Central Mine Research Institute
Italy
Department of Civil Engineering, Imperial College, London
New Zealand
Ministry of Commerce, Energy and Resources Division
Coeur d'Alene Mines Inc
Papua New Guinea
Department of Mining and Petroleum
Philippines
Mines and Geosciences Bureau
Poland
Central Mining Institute
Polish Geological Institute
South Africa
Chamber of Mines of South Africa
Department of Mineral and Energy Affairs
Department of Civil Engineering, University of Witwatersrand
Sweden
Environmental Protection Board
Geological Survey
U.S.A.
Environmental Protection Agency
Mine Safety and Health Administration
National Mine Safety and Health Academy
Office of Surface Mining
Center on the Performance of Dams, Stanford University
US Committee on Large Dams
Zambia
SADC, Mining Sector Coordinating Unit
Zimbabwe
Chamber of Mines of Zimbabwe
A3
Further to responses received from this initial list, questionnaires were subsequently sent to
the following 23 organisations:
Australia
New South Wales Department of Mineral Resources
Northern Territory Department of Mines and Energy
Queensland Department of Minerals and Energy
South Australia Department of Mines and Energy
Victoria Department of Agriculture, Energy and Minerals
Western Australia Department of Minerals and Energy
Canada
British Columbia Ministry of Energy, Mines and Petroleum Resources
New Brunswick Department of Environment
Newfoundland Department of Mines and Energy
Ontario Department of Northern Development and Mines
Ministère du l'Energie et des Ressources (Quebec)
Saskatchewan Industrial Branch, Environment and Resource Management
Chile
Steffen, Robertson and Kirsten, Santiago
Ghana
Minerals Commission
Guyana
Golden Star Resources Inc
Philippines
Mines and Geosciences Bureau
South Africa
U.S.A.
Mine Safety and Health Administration (MSHA)
National Mine Safety and Health Academy
Center on the Performance of Dams, Stanford University
Zambia
SADC, Mining Sector Coordinating Unit
Zimbabwe
Chamber of Mines of Zimbabwe
Completed questionnaires were received from all the Australian states apart from Queensland,
all the Canadian provinces apart from Ontario, Golden Star Resources Inc, MSHA and the
National Mine Safety and Health Academy in the United States, Zambia and Zimbabwe. Of
23 questionnaires sent out, responses were received from 17 organisations, with a further three
(Chile, Ghana and Ontario, Canada) indicating that information might be forthcoming in the
future.
A4
UNEP STUDY OF INCIDENTS INVOLVING MINE TAILINGS DAMS
SPECIFIC INCIDENT DETAILS
Background information:
Record No. 100
Country Zimbabwe
Name of tailings dam Arcturus
Location of tailings dam N/a
Date of incident January 1978
Time of incident Early morning
Mine served by the tailings dam Arcturus gold mine
Mine operating company Corsyn Consolidated Mines
Dam construction comnanv

--------------------------L
.'
N/a
Dam active or inactive at the time Active

of the incident
Dam details:
Type of dam Valley side
Age of dam when incident occurred 12-15 years
Retention capacity of dam 1.7-2.0 Mt
Height of dam wall 15 m
Other dam dimensions 150 x 110 m (approx)
Dam wall construction materials Tailings
Method of dam construction Downstream with evaporation paddocks
Criteria or standards applied during N/a

dam construction
AS
Incident details:
Type of incident Slurry overflow
Estimated volume/tonnage of N/a

tailings/water/slurry released
Primary cause of incident Continuous rain over several days.

Seasonal total rainfall above average
Other contributing factors One faulty decant line
Result of incident Public waterway partially blocked and

contaminated
Number of fatalities One
Extent of property damage Minor damage to local village
Extent of environmental damage Extensive siltation to the waterway and

to adjoining rough pasture
Length of time damage persisted Still evident
Estimated cost of remediation N/a
Legal or other action resulting Remedial work required to clear public

from the incident waterway
Remedial action taken by the dam owner Excavation of permanent diversion

waterway and construction of silt-
containing gabions downstream
Media coverage WC
Literature references N/a
Contact details for further information Mr J Eagling

Chamber of Mines
P0 Box 712
Stewart House
4 Central Avenue
Harare
Zimbabwe
Tel: +263 4 702841
Fax: +263 4 707983
A6
Record No. 110
Country USA, New Mexico
Name of tailings dam Church Rock
Location of tailings dam Church Rock
Date of incident 16th July 1979
Time of incident N/a
Mine served by the tailings dam Church Rock uranium mine
Mine operating company United Nuclear
Dam construction company Raymond Kaiser Engineers

of the incident
Dam details:
Type of dam Water retention
Age of dam when incident occurred N/a
Retention capacity of dam 370,000 m3
Height of dam wall 11
Other dam dimensions N/a
Dam wall construction materials Earthfill
Method of dam construction N/a

dam construction
A7
Incident details:
Type of incident Breached dam wall led to discharge of

contaminated water and slurry into local
watercourse
Estimated volume/tonnage of 370,000 m3 of radioactive water

tailings/water/slurry released 1,000 t of contaminated sediment
Primary cause of incident Differential foundation settlement
Other contributing factors Inadequate foundation material strength;

absence of beach between pond and dam
wall
Result of incident Contamination of watercourse sediments

by radioactive material over a distance of
up to 110 km from the dam site
Number of fatalities Nil
Extent of property damage N/a
Extent of environmental damage N/a
Legal or other action resulting Litigation for damages by local inhabitants

from the incident
Remedial action taken by the dam owner Repair. Operation closed two years later
in response to depressed uranium market
Media coverage Yes
Literature references USCOLD
Contact details for further information N/a
W .
Record No. 120
Country Canada, New Brunswick
Name of tailings dam Heath Steele main dam
Location of tailings dam Bathurst, New Brunswick
Date of incident Continuing since 1970s
Mine served by the tailings dam Heath Steele
Mine operating company Noranda
Dam construction company N/a

of the incident
Dam details:
Type of dam N/a
Age of dam when incident occurred 2 years at commencement of leakage
Retention capacity of dam N/a
Dam wall construction materials Rock and glacial till, clay core

dam construction
A9
Incident details:
Type of incident Persistent uncontrolled seepage
Estimated volume/tonnage of Low volume, constant flow

Primary cause of incident Construction of dam on fractured

bedrock, with no liner or grouting
Other contributing factors Acid generating rock on embankment

above water table
Result of incident Seepage of water containing copper and

zinc through dam wall
Extent of property damage Nil
Extent of environmental damage Localised impact immediately

downstream from the dam
Legal or other action resulting N/a

from the incident
Remedial action taken by the dam owner Collection of seepage, which is pumped
back to tailings pond
Media coverage No
Contact details for further information Department of the Environment

P0 Box 6000
364 Argyle Street
Fredericton
NB E3B5H1
Canada
Tel: +1 506 457 4848
Fax: +1 506 453 2265
AlO
Record No. 130

--
Country United States, New Mexico
Name of tailings dam No. 3 tailings dam
Location of tailings dam Tyrone, NM
Date of incident I 3th October 1980
Time of incident Night
Mine served by the tailings dam Tyrone copper mine
Mine operating company Phelps Dodge

of the incident
Dam details:
Retention capacity of dam 2.5 Mm3
Height of dam wall 66m
Dam wall construction materials Cycloned sand tailings
Method of dam construction Upstream

dam construction
All
Incident details:
Type of incident Breach in dam wall
Estimated volume/tonnage of 2.0 Mm3

Primary cause of incident Rapid increase in dam wall height

causing high internal pore pressures
Other contributing factors Possible leakage from pipeline
Result of incident Tailings flowed through the breach,

measuring 215 in wide by 35 m deep,
down the slope below, up the opposite
valley side, then down the valley for a
distance of 8 km
Extent of property damage Inundation damage to farm land
Extent of environmental damage Localised

from the incident
Remedial action taken by the dam owner Redesign and repair
Media coverage N/a
Contact details for further information Phelps Dodge

2600 N Central Avenue
Phoenix
AZ 85004
USA
Tel: +1 602 234 8100
Al2
Record No. 140

Country Philippines

Name of tailings dam No.3 tailings pond

Location of tailings dam Sipalay, Negros Occidental

Date of incident 8th November 1982

Time of incident 16.00

Mine served by the tailings dam Sipalay copper mine

Mine operating company Marinduque Mining & Industrial Corp.


of the incident
Dam details:
Type of dam N/a
Age of dam when incident occurred 4 years
Retention capacity of dam 37 Mt
Height of dam wall N/a
Dam wall construction materials Mixed mine waste

dam construction
A13
Incident details:
Type of incident Failure of a section of embankment
Estimated volume/tonnage of 27 Mt
Primary cause of incident Slippage of foundations on clayey soils

and surface materials that were not
removed before construction of the
embankment began
Other contributing factors Inadequate anchoring of starter

embankment; use of mixed mine waste
of very variable particle size
Result of incident Widespread inundation of surrounding

agricultural land with up to 1.5 m
thickness of tailings
Number of fatalities N/a
Extent of property damage Extensive to crops, buildings and livestock
Extent of environmental damage Localised
Legal or other action resulting Official inquiry

from the incident
Remedial action taken by the dam owner N/a; Assets of MMI now held by
Marinduque Mining Corp.
Media coverage N/a
A14
Record No. 150
Country USA, California
Name of tailings dam Gray Eagle
Location of tailings dam Happy Camp, CA
Date of incident 1983
Mine served by the tailings dam Gray Eagle
Mine operating company Centurion Gold Ltd

of the incident
Dam details:
Type of dam N/a
Age of dam when incident occurred 1 year
Method of dam construction Downstream with clay core

dam construction
A15
Incident details:
Type of incident Uncontrolled seepage

Primary cause of incident Seepage of effluent through jointed rock

beneath dam
Other contributing factors Higher than expected rainfall; increased

pumping of seepage back to the pond
also put the fluid balance of the dam out
of balance
Result of incident Contamination of groundwater with

cyanide-bearing tailings water

from the incident
Remedial action taken by the dam owner Improved surface diversion and seepage
pumping capability
Media coverage N/a
Contact details for further information Centurion Gold Ltd

410, 750 West Pender Street
Vancouver
B.C. V6C2T7
Canada
Tel: +1 604 681 8466
Fax: +1 604 683 9219
A16
Record No. 160
Country USA, Montana
Name of tailings dam Golden Sunlight
Location of tailings dam Whitehall, MT
Date of incident From January 1983
Time of incident Continuing
Mine served by the tailings dam Golden Sunlight
Mine operating company Golden Sunlight Mines
Dam cnnstniction comnanv

- ------ ------- - N/a

of the incident
Dam details:
Type of dam N/a
Age of dam when incident occurred New
Method of dam construction Centreline
Criteria or standards applied during Installation of 20 m-deep bentonite slurry

dam construction cut-off wall to contain seepage
A17
Incident details:
Type of incident Persistent seepage of contaminated water

Primary cause of incident Permeable plane in material below dam

foundations, and beneath foot of cut-off
wall
Other contributing factors N/a
Result of incident Persistent seepage of cyanide-containing

water into groundwater below dam
Extent of environmental damage Localised to immediately downstream

from dam

from the incident
Remedial action taken by the dam owner Repair to seepage cut-off wall, and
pumping seepage back to tailings pond
Media coverage N/a
Contact details for further information Placer Dome Inc

1600 Bentall IV
1055 Dunsmuir Street
Vancouver
B.C. V6C 1A8
Canada
Tel: +1 604 682 7082
Fax: +1 604 682 7092
A18
Record No. 170
Country USA, North Carolina
Name of tailings dam Texasguif 4B pond
Location of tailings dam Aurora, NC
Date of incident 1st April 1984
Mine served by the tailings dam Lee Creek phosphate mine
Mine operating company Texasguif Inc
Thm ennstniction comnanv

-------- 1.'
N/a

of the incident
Dam details:
Type of dam N/a
Method of dam construction Water retention

dam construction
A19
Incident details:
Type of incident Slippage of part of dam wall
Estimated volume/tonnage of Nil

Primary cause of incident Raised phreatic surface led to instability

of the dam wall
Other contributing factors Clay spoil piled at darn toe prevented

seepage to keep phreatic surface at a safe
level
Result of incident A 60 rn-long section of the dam wall

suffered slippage

from the incident
Remedial action taken by the dam owner Improved drainage installed to dam wall
Media coverage N/a
Literature references
Contact details for further information Texasgulf Inc

P0 Box 30321
3101 Glenwood Avenue
Raleigh
NC 27622-032 1
USA
Tel: +1 919 881 2700
Fax: +1 919 881 2847
A20
Record No. 180
Country Chile
Name of tailings dam Marga
Mine served by the tailings dam N/a
Mine operating company N/a
Dam active or inactive at the time Inactive

of the incident
Dam details:
Type of dam Cross-valley
Dam wall construction materials N/a

dam construction
A21
Incident details:
Type of incident Overtopping failure

Primary cause of incident Lack of spiliway to handle water flows

from catchment area upstream from the
dam
Other contributing factors Inadequate decant capacity
Result of incident Damage to dam wall

from the incident
Remedial action taken by the dam owner N/a
Media coverage N/a
A22
Record No. 190
Country USA, Nevada
Name of tailings dam Olinghouse
Location of tailings dam Wadsworth, NV
Mine served by the tailings dam Olinghouse gold mine
Mine operating company Olinghouse Mining Co.

of the incident
Dam details:
Type of dam Water retention
Dam wall construction materials Unconsolidated earthfill
Criteria or standards applied during No engineering supervision

dam construction
A23
Incident details:
Type of incident Embankment collapse
Estimated volume/tonnage of 25,000 m3

Primary cause of incident Saturation of embankment
Other contributing factors Lack of compaction in dam wall
Result of incident Outflow of tailings over a distance of 1.5

km from the dam site

from the incident
Remedial action taken by the dam owner N/a. Property subsequently acquired by
Green Hill Mining Venture
Media coverage N/a
Contact details for further information Cliff Resources Corp.

P0 Box 14
1902, 95 Wellington Street West
Toronto
Ontario M5J 2N7
Canada
Tel: +1 416 941 9440
Fax: +1 416 941 9450
A24
Record No. 200
Country Chile
Name of tailings dam Cerro Negro No.4
Location of tailings dam Cerro Negro
Date of incident 3rd March 1985
Mine served by the tailings dam Cerro Negro copper mine
Mine operating company Cia Minera Cerro Negro

of the incident
Dam details:
Retention capacity of dam 2Mm3
Method of dam construction Upstream and centreline

dam construction
A25
Incident details:
Type of incident Breach of dam wall

Primary cause of incident Liquefaction of material during earthquake
Result of incident Outflow of tailings, which travelled for 8

km downstream from the dam site

from the incident
Remedial action taken by the dam owner Repair
Media coverage N/a
Contact details for further information Cia Minera Cerro Negro
A26
Record No. 210
Country Chile
Name of tailings dam Veta de Agua No.1
Date of incident 3rd March 1985

of the incident
Dam details:
Type of dam N/a
Method of dam construction Upstream and centreline

dam construction
A27
Incident details:
Type of incident Failure of dam wall

Primary cause of incident Liquefaction of material during earthquake
Result of incident Outflow of tailings, which travelled for 5

km below the dam Site

from the incident
Media coverage N/a
A28
Record No. 220
Country USA, Pennsylvania
Name of tailings dam LaBelle
Location of tailings dam Fayette County, PA
Mine served by the tailings dam LaBelle coal mine
Mine operating company LaBelle Processing Company

of the incident
Dam details:
Type of dam N/a
Dam wall construction materials Mine waste
Method of dam construction Downstream

dam construction
A29
Incident details:
Type of incident Slippage along dam face

Primary cause of incident Movement of foundations along clay

layer beneath
Other contributing factors Rainfall
Result of incident A 240 rn-long section of the dam wall

slumped
Extent of environmental damage Nil

from the incident
Remedial action taken by the dam owner Construction of rock drains to dam face,
and buttress of large waste rock to dam
toe
Media coverage N/a
Contact details for further information LaBelle Processing Company

PU Drawer 1187
I
Uniontown
PA 15401
USA
Tel: +1 412 438 0536
Fax: +1 412 438 1322
A30
Record No. 230
Country Italy
Name of tailings dam Stava
Location of tailings dam Stava, Trento region
Mine served by the tailings dam Prestavel fluorite mine
Mine operating company Prealpi Mineraria
Dam construction company Montedison/Prealpi Mineraria

of the incident
Dam details:
Type of dam Twin cross-valley
Height of dam wall 29.5 m (Upper dam)

26.0 m (Lower dam)
Other dam dimensions Crest length - 200 m (Upper), 180 m

(Lower); design downstream slopes 40°
Dam wall construction materials Cycloned tailings and fluvio-glacial

material
Method of dam construction Upstream (Lower); centreline (Upper)

dam construction
A31
Incident details:
Type of incident Sequential collapse of Upper then Lower

dams

Primary cause of incident Inadequate decant pipe construction
Other contributing factors Inadequate dam stability; poorly drained

site; inadequate drainage to dam wall
constructed by the upstream method;
high pond levels caused by run-off
collection; toe of Upper dam rested on
tailings in Lower dam; high natural
groundwater level
Result of incident Catastrophic failure and mud rush down

valley below the dams
Number of fatalities 268
Extent of property damage Extensive
Extent of environmental damage Extensive
Legal or other action resulting Official inquiry; arrest of mine owners

from the incident
Media coverage Extensive
Literature references Chandler R J and Tosatti G. The Stava

tailings dam failure, Italy, July 1985.
Geotechnical Engineering; 113(2); 67-
79 (April 1995)
Contact details for further information Professor R J Chandler

Department of Civil Engineering
Imperial College
London SW7 2BU
Tel: +44 171 594 6080
Fax: +44 171 594 6053
A32
Record No. 240
Country Canada, British Columbia
Name of tailings dam Mineral King
Location of tailings dam Invermere, B.C.
Date of incident 20th March 1986
Mine served by the tailings dam Mineral King
Mine operating company Mountain Minerals

of the incident
Dam details:
Age of dam when incident occurred 10 years approx.
Retention capacity of dam Small
Dam wall construction materials Coarse tailings sand
Method of dam construction Centre line

dam construction
A33
Incident details:
Type of incident Dam breach

Primary cause of incident High pond overtopped crest
Other contributing factors Diversion ditch blocked by ice during

onset of spring snow melt
Result of incident Permanent closure of dam
Extent of property damage Minimal
Extent of environmental damage Minimal
Length of time damage lasted 2-3 months

from the incident
Media coverage N/a
Literature references Ministry internal memos
Contact details for further information Mr T Eaton

Manager, Geotechnical Engineering
Ministry of Employment and Investment
Energy and Minerals Division
4th Floor, 1810 Blanshard Street
Victoria V8V 1X4
B.C.
Canada
Tel: +1 604 952 0462
Fax: +1 604 952 0481
A34
Record No. 250
Country Brazil
Name of tailings dam N/a
Location of tailings dam Itabirito, Minas Gerais
Date of incident Late May 1986
Mine operating company Itaminos Comercio de Minerios
Dam active or inactive at the time N/a

of the incident
Dam details:
Type of dam N/a
Other dam dimensions Dam length - 150 m
Dam wall construction materials Bricks made from clay and iron ore tailings
Method of dam construction Masonry construction

dam construction
A35
Incident details:
Type of incident Dam wall burst
Estimated volume/tonnage of 100,000 t

Primary cause of incident Saturation of brickwork
Result of incident Release of tailings, which flowed for 12

km down the valley below the dam
Extent of property damage Loss of mining equipment and

infrastructure

from the incident
Media coverage N/a
Literature references Engineering News-Record

June 5, 1986
A36
Record No. 260
Country Australia, Tasmania
Name of tailings dam Rossarden
Location of tailings dam Rossarden
Date of incident 16th May 1986
Mine served by the tailings dam Aberfoyle
Mine operating company Aberfoyle Ltd

of the incident
Dam details:
Height of dam wall 7.5 m, perched on valley edge 190 m

above river
Dam wall construction materials Layered earth and tea-tree matting
Method of dam construction Uncontrolled

dam construction
A37
Incident details:
Type of incident Dam wall partially washed out

Primary cause of incident Water breached the main side of the

impoundment area, swept though the dam
and eroded a section of the front wall
Other contributing factors Poor drainage
Result of incident Dam wall failure
Extent of environmental damage River polluted by tailings
Length of time damage persisted 12 months
Estimated cost of remediation A$5,000
Legal or other action resulting Mines department investigation;

from the incident Government inquiry
Remedial action taken by the dam owner Government repair funded by contractor
Media coverage Local complaints only
Contact details for further information MrJGLas

Inspector of Mines
Industry Safety and Mines Division
P0 Box 56
Rosny Park
Tasmania 7018
Australia
Tel: +61 02 338318
Fax: +61 02 338338
A38
Record No. 270
Name of tailings dam Story's Creek
Location of tailings dam Story's Creek
Date of incident 16th May 1986
Mine served by the tailings dam Story' Creek
Mine operating company Aberfoyle Ltd
Dam construction company Aberfoyle Ltd

of the incident
Dam details:
Other dam dimensions Crest length - 130 m; crest width - I m;

downstream slope - 1:
Dam wall construction materials Mainly tailings
Method of dam construction Uncontrolled

dam construction
A39
Incident details:
Type of incident Wall collapse; discharge of water and

slimes
Estimated volume/tonnage of N/a (assumed minimal)

Primary cause of incident Large influx of water
Other contributing factors Unstable wall, and I in 100 year flood
Result of incident Wall failure; spillway shifted; slimes

escaped; pipeline washed out, causing
further pollution of waterway
Number of fatalities nl
Extent of environmental damage River polluted by tailings/slimes for

approximately 10 km downstream

Remedial action taken by the dam owner Government repair funded by contractor
Media coverage Local complaints only

Inspector of Mines
P0 Box 56
Rosny Park
Tasmania 7018
Australia
Tel: +61 02 338318
Fax: +61 02 338338
A40
Record No. 280
Country Brazil
Name of tailings dam Pico de São Luis
Location of tailings dam Minas Gerais
Date of incident 2nd October 1986
Mine served by the tailings dam Pico de São Luis
Dam active or inactive at the time N/a

of the incident
Dam details:
Type of dam N/a

dam construction
A41
Incident details:
Type of incident Failure of dam wall

Primary cause of incident Erosion at dam toe by water flowing over

spiliway
Other contributing factors Soft clay foundation
Result of incident Release of tailings

from the incident
Media coverage N/a
Contact details for further information ICOLD
A42
Record No. 290
Country Philippines
Location of tailings dam Mankayan, Luzon
Date of incident 17th October 1986
Mine served by the tailings dam Mankayan copper mine
Mine operating company Lepanto Consolidated Mining Co.

of the incident
Dam details:
Type of dam N/a

dam construction
A43
Incident details:
Type of incident Embankment failure

Primary cause of incident Additional loading on original

embankment from 8 m increase in height
Other contributing factors Location of embankment in ancient slide

area; slope of embankment too steep;
decant tower too close to embankment
Result of incident N/a

from the incident
Media coverage N/a
Contact details for further information Lepanto Consolidated Mining Co.

Bank of America - Lepanto Building
8747 Paseo de Roxas
Makati
Metro Manila 3117
Philippines
A44
Record No. 300
Country USA, Montana
Name of tailings dam Montana Tunnels
Location of tailings dam Helena, MT
Mine served by the tailings dam Montana Tunnels gold mine
Mine operating company Pegasus Gold Inc

of the incident
Dam details:
Type of dam N/a
Age of dam when incident occurred 1 year
Criteria or standards applied during Impoundment floor fitted with soil/

dam construction bentonite liner
A45
Incident details:
Type of incident Uncontrolled seepage

Primary cause of incident Breaches in impoundment liner
Other contributing factors Inadequate repairs to liner
Result of incident Contamination of groundwater by

cyanide-bearing tailings water
Extent of environmental damage Localised to immediately downstream

from the dam

from the incident
Remedial action taken by the dam owner Increased seepage pumping capacity
installed
Media coverage N/a
Contact details for further information Pegasus Gold Inc

400 North 9 Post Street
Spokane
WA 99201
USA
Tel: +1 509 624 4653
Fax: +1 509 838 8317
A46
Record No. 310
Country China
Name of tailings dam Xishimen
Location of tailings dam Xishimen
Date of incident 21st March 1987
Mine served by the tailings dam Xishimen iron ore mine

of the incident
Dam details:
Type of dam N/a

dam construction
A47
Incident details:

Primary cause of incident Pond water level allowed to rise too high
Other contributing factors Blocked decant
Result of incident Outflow of tailings through breach

from the incident
Media coverage N/a
A48
Record No. 320
Country USA, Tennessee
Name of tailings dam Tennessee Consolidated No.1
Location of tailings dam Grays Creek, TN
Date of incident 19th January 1988
Mine served by the tailings dam Tennessee Consolidated No.1
Mine operating company Tennessee Consolidated Coal Co.
Dam construction comnanv N/a

of the incident
Dam details:
Type of dam N/a

dam construction
A49
Incident details:
Type of incident Internal erosion followed by failure of a

section of the dam wall

Primary cause of incident Failure of an abandoned outlet pipe
Result of incident Water flowing from abandoned pipe

saturated dam wall, leading to erosion
and breaching; the entire contents of the
impoundment escaped

from the incident
Media coverage N/a
Literature references USCOLD, MSHA
Contact details for further information Tennessee Consolidated Coal Co.

Rt.3
P0 Box 343A
Whitwell
TN 37397
USA
Tel: +1 615 658 5115
Fax: +1 615 658 6890
A50
Record No. 330
Country China, Shaanxi province
Name of tailings dam Jinduicheng
Location of tailings dam Jinduicheng
Date of incident 30th April 1988
Mine served by the tailings dam Jinduicheng molybdenum mine

of the incident
Dam details:
Type of dam N/a

dam construction
A51
Incident details:
Type of incident Breach of dam wall, causing release of

tailings

Primary cause of incident Spillway blockage caused pond level to

rise too high
Other contributing factors High phreatic surface within dam wall
Result of incident Catastrophic failure of centre section of

dam wall
Number of fatalities 20 approx.

from the incident
Media coverage Noted in specialist metals industry press
A52
Record No. 340
Country USA, Idaho
Name of tailings dam Silver King
Location of tailings dam Adams county, ID
Date of incident 5th August 1989
Mine served by the tailings dam Copper Cliff copper-silver mine
Mine operating company Alta Gold Co.

of the incident
Dam details:
Type of dam N/a

dam construction
A53
Incident details:
Type of incident Overtopping of dam wall

Primary cause of incident Slump of waste material into tailings pond
Result of incident Waste material displaced an estimated

1,200-2,400 m3 of water, which
overtopped the dam causing minor
damage to the wall
Extent of property damage Slight

from the incident
Media coverage N/a
Literature references
Contact details for further information Alta Gold Co.

140, 2319 Foothill Drive
Salt Lake City
UT 84109
USA
Tel: +1 801 483 1116
Fax: +1 801 467 2137
A54
Record No. 350

Country Canada, British Columbia

Name of tailings dam Iron Dyke

Location of tailings dam Kimberley

Date of incident 23rd August 1991


Mine served by the tailings dam Sullivan

Mine operating company Cominco


of the incident
Dam details:
Type of dam Ring
Age of dam when incident occurred Foundation - 40 years

Dykes - 16 years
Height of dam wall 21 m maximum
Other dam dimensions 1,500 m of ring dykes

dam construction
A55
Incident details:
Type of incident Rotational slump of 300 m of ring dyke

Primary cause of incident Generation of excess pore pressure in old

tailings foundation
Other contributing factors Construction equipment travelling over

the site during the 1991 annual dam wall
height increase
Result of incident Temporary closure of hon Pond until

interim dyke and operating plan in place
Extent of property damage Damage to dyke only
Length of time damage lasted 1 year
Estimated cost of remediation Over C$1 million

from the incident
Remedial action taken by the dam owner Geotechnical evaluation and

rehabilitation for closure
Media coverage Yes
Literature references Geotechnical reports from KJohn Crippen

Inc
Contact details for further information Cominco Ltd

200 Granville Street
Vancouver V6C 2R2
B.C.
Canada
Tel: +1 604 682 0611
Fax: +1 604 844 2516
A56
Record No. 360
Country Philippines
Location of tailings dam Padcal, Luzon
Date of incident January 1992
Mine served by the tailings dam Padcal copper mine
Mine operating company Philex Mining Corp.

of the incident
Dam details:
Type of dam N/a
Age of dam when incident occurred estimated 10 years

dam construction
A57
Incident details:
Type of incident Collapse of dam wall
Estimated volume/tonnage of 80 Mt
Primary cause of incident Foundation failure
Other contributing factors Suspected weakening of structure

following major earthquake in July 1990
Result of incident Outflow of tailings

from the incident
Media coverage N/a
Contact details for further information Philex Mining Corp.
A58
Record No. 370
Country Montenegro
Name of tailings dam Kojkovac
Location of tailings dam Majkovac
Date of incident November 1992
Mine served by the tailings dam Inactive lead-zinc mines

of the incident
Dam details:
Type of dam Valley bottom

dam construction
A59
Incident details:
Type of incident Erosion of dam wall toe by adjacent river

Primary cause of incident Flood water in river
Other contributing factors Original placing of dam
Result of incident Dam wall toe was eroded by fast-flowing

river over a distance of 150 m.
Extent of environmental damage Nil; potential for widespread pollution

into Tara river, thence via Danube to
Black Sea

from the incident
Remedial action taken by the dam owner Toe of dam supported by gabions to
prevent further erosion under UN
emergency project
Media coverage Yes
Literature references ICOLD
Contact details for further information UNDRO, Geneva
A60
Record No. 380
Country Philippines
Name of tailings dam Itogon-Suyoc
Location of tailings dam Baguio gold district, Luzon
Date of incident 26th June 1993
Mine served by the tailings dam Itogon-Suyoc gold mine
Mine operating company Itogon-Suyoc Mines

of the incident
Dam details:
Type of dam N/a

dam construction
A61
Incident details:
Type of incident Flood water overtopped dam wall,

leading to partial failure

Primary cause of incident Blockage to diversion tunnel carrying

original river around dam
Other contributing factors Typhoon
Result of incident Collapse of part of dam wall

from the incident
Media coverage N/a
Contact details for further information Itogori-Suyoc Mines
A62
Record No. 390
Country Zambia
Name of tailings dam TD 7
Location of tailings dam Chingola
Date of incident August 1993
Mine served by the tailings dam Nchanga copper mine
Mine operating company ZCCM
Dam construction company Knight Piesold (Zambia)

of the incident
Dam details:
Type of dam Ring
Other dam dimensions Area: 2,500 m2
Dam wall construction materials Concentrator tailings and borrow material

dam construction
A63
Incident details:
Type of incident Slurry overflow following rain storm;

dam wall collapsed
Estimated volume/tonnage of 100

Primary cause of incident Design inadequate for worst-case weather

conditions
Other contributing factors Increased rate of tailings deposition

following plant shut-down
Result of incident Slurry overflow into spiliway
Extent of property damage Blockages to water supply pipes for

domestic usage
Extent of environmental damage Minimal
Estimated cost of remediation Minimal
Legal or other action resulting Official enquiry

from the incident
Remedial action taken by the dam owner Diversion of tailings to disused open pit
Media coverage No
Contact details for further information Mr J Masinga

Head, Group Environmental Services
ZCCM Ltd
Kalulushi
Zambia
Tel: +260 2 749108
Fax: +260 2 733123
AM
Record No. 400
Country South Africa
Name of tailings dam Merriespruit
Location of tailings dam Merriespruit, OFS
Date of incident 22nd February 1994
Mine served by the tailings dam Harmony
Mine operating company Harmony Gold Mines
Dam construction company Fraser Alexander

of the incident
Dam details:
Type of dam Ring
Other dam dimensions Area: 154 ha
Dam wall construction materials Cycloned tailings

dam construction
A65
Incident details:
Type of incident Breach through wall of ring dam released

tailings slurry into nearby housing area
Estimated volume/tonnage of 2.5 Mt

Primary cause of incident Surface water overtopped dam wall

following persistent heavy rain
Other contributing factors Possible increase in phreatic surface in

dam wall as a result of the rainfall;
weakened dam wall from springing;
poor natural drainage beneath dam
Result of incident Escaped tailings covered an area of 2 km

by 400 m
Extent of property damage Extensive
Extent of environmental damage Limited
Legal or other action resulting Judicial inquiry; prosecution of mine

from the incident operator and dam engineer
Literature references Official inquiry report
Contact details for further information SA Government Mining Engineer

Private Bag X59
Pretoria 0001
South Africa
Tel: +27 12 317 9125
Fax: +27 12 322 0810
A66
Record No. 410
Name of tailings dam Riltec
Location of tailings dam Mathinna
Date of incident 6th January 1995
Mine served by the tailings dam Riltec Mathinna
Mine operating company Riltec (Tasmania) Pty Ltd
Dam construction company R Gerke
Dam active or inactive at the time Active (pre-commissioning)

of the incident
Dam details:
Age of dam when incident occurred 3 months
Other dam dimensions Crest - 370 m; crest width - 4 m;

downstream slope - 1:2
Dam wall construction materials Random fill, clay lined
Criteria or standards applied during Consultants' specifications; layered

dam construction compaction at optimum moisture content
A67
Incident details:
Type of incident Leakage of cyanide-contaminated water
Estimated volume/tonnage of 40,000 m3 of cyanide-contaminated water

Primary cause of incident Water table exposed in bottom of dam
Other contributing factors Lack of supervision during construction
Result of incident Pollution of local waterways; cessation

of operations
Extent of environmental damage Polluted streams and local fish kill

Remedial action taken by the dam owner Owner bankrupt; environmental bond
used for remedial work
Literature references Department archives

Inspector of Mines
P0 Box 56
Rosny Park
Tasmania 7018
Australia
Tel: +61 02 338318
Fax: +61 02 338338
Record No. 420
Name of tailings dam No. I tailings dam
Location of tailings dam Middle Arm, Launceston
Date of incident 25th June 1995
Mine served by the tailings dam Beaconsfield
Mine operating company Golconda Ltd
Dam construction company WLPU Consultants (Australia) Pty Ltd

of the incident
Dam details:
Age of dam when incident occurred 12 months
Other dam dimensions Crest - 315 m
Dam wall construction materials Borrow pit; mixed grey loam and stiff
yellow clay
Criteria or standards applied during 90% compaction at optimum moisture

dam construction content
A69
Incident details:
Type of incident Erosion of top of wall, consisting of

tailings
Estimated volume/tonnage of 5,000 m3 of water released into the Tamar

tailings/water/slurry released river
Primary cause of incident Wave action on wall
Other contributing factors Buildup of tailings above height of dam

wall
Result of incident Water containing 95mg/litre free cyanide

released into the Tamar river
Extent of environmental damage Minimal on the property, and localised

fish kill off the property
Length of time damage persisted 3 weeks
Estimated cost of remediation A$20,000-30,000 estimated

from the incident
Remedial action taken by the dam owner Reduce the water level
Media coverage Yes
Literature references Press reports
Contact details for further information MrJG Las

Inspector of Mines
P0 Box 56
Rosny Park
Tasmania 7018
Australia
Tel: +61 02 338318
Fax: +61 02 338338
A70
Record No. 430
Country Guyana
Name of tailings dam Tailings dam No.1
Location of tailings dam Omai
Date of incident 19th August 1995
Mine served by the tailings dam Omai gold mine
Mine operating company Omai Gold Mines Ltd
Dam construction company Knight Piesold/Golder Associates

of the incident
Dam details:
Type of dam Valley bottom
Age of dam when incident occurred 3.5 years
Height of dam wall 44 in
Other dam dimensions 700 in face length, 300 in toe
Dam wall construction materials Waste rockfill wall with a sand filter
between the wall and a compacted
saprolite core
Method of dam construction Rear deposition with water at the face of

the dam
Criteria or standards applied during Standard Knight Piesold design and

dam construction specifications
A71
Incident details:
Type of incident Piping failure leading to failure of the core
Estimated volume/tonnage of 4.2 Mm3

Primary cause of incident N/a
Result of incident Discharge of cyanide-tainted water and

sediments resulting from the washing of
the core through the dam. No tailings
slimes exited the impoundment.
Extent of environmental damage Minor fish kill on the Omai river

adjacent to the mine. No damage to the
main Essequibo river where Canadian
safe drinking water standards were not
exceeded
Estimated cost of remediation Nil
Legal or other action resulting Formal praliamentary commission of

from the incident inquiry, including dam review committee,
process review committee and socio-
economic/environmental committee
Remedial action taken by the dam owner Clean-up of sedimentation in the Omai
river that resulted from washing out the
core of the dam.
Literature references Official report
Contact details for further information Guyana Geology and Mines Commission
Georgetown
Guyana
A72
Record No. 440
Country Philippines
Name of tailings dam Placer
Location of tailings dam Surigao Del Norte
Date of incident 2nd September 1995
Mine served by the tailings dam Placer
Mine operating company Manila Mining Corp.
Dam active or inactive at the time Inactive as a tailings dam, but used for
of the incident waste rock storage on top of old tailings
Dam details:
Type of dam N/a
Age of dam when incident occurred Probably at least 12 years
Other dam dimensions Length - 300 m

dam construction
A73
Incident details:

Primary cause of incident Failure of dam wall led to formation of

100 rn-long breach, probably resulting
from slippage in underlying foundation
materials (marine sediments?)
Other contributing factors Possible connection with magnitude 3.4

earthquake that occurred on 28th August
1995; top of dam wall was used as a
haul road for off-highway trucks
Result of incident Mud flow towards shoreline, with

subsequent sea water ingress into the
dam impoundment
Extent of property damage Loss of mining equipment
Extent of environmental damage Coastal pollution

from the incident
Media coverage N/a
Contact details for further information Manila Mining Corp.
A74
Record No. 450
Country New Zealand
Name of tailings dam Golden Cross
Location of tailings dam Waitekauri Valley
Date of incident Continuing since December 1995
Mine served by the tailings dam Golden Cross gold mine
Mine operating company Coeur Golden Cross
Dam construction company Previous owner

of the incident
Dam details:
Retention capacity of dam 3 Mt (current)
Height of dam wall 25-30 m
Dam wall construction materials Waste rock fill
Criteria or standards applied during Fully engineered

dam construction
A75
Incident details:
Type of incident Slide involving complete dam structure
Estimated volume/tonnage of None

Primary cause of incident Movement on plane of weakness at some

50 m depth between ancient lava flows
and later material, on which the dam was
constructed
Result of incident Ground movement monitoring equipment

installed; construction of dewatering
tunnels from existing underground mine
workings; future of mining operations
under review
Estimated cost of remediation NZ$5 million

from the incident
Remedial action taken by the dam owner Dewater through tunnels beneath dam;
construct rock buttress at dam toe
Media coverage Extensive in local press
Contact details for further information Mr Richard Weston

Mine Manager
Coeur Golden Cross
Tel: +64 7 863 6507
Fax: +64 7 863 6112
A76
Record No. 460
Country Philippines
Name of tailings dam Marcopper
Location of tailings dam Marinduque island
Date of incident March 1996
Mine served by the tailings dam Marcopper
Mine operating company Marcopper Mining Corp
Dam construction comnanv N/a

of the incident
Dam details:
Type of dam Tailings storage in worked-out pit
Method of dam construction Tailings dredged from previous dam

dam construction
A77
Incident details:
Type of incident Tailings escaped through old drainage

tunnel beneath worked-out open pit

Primary cause of incident Failure of existing concrete plug
Result of incident Outflow of water and tailings through

drainage tunnel, with deposition of
tailings in waterways downstream

from the incident
Remedial action taken by the dam owner Install plug in tunnel through which
tailings escaped
Media coverage Yes
Contact details for further information Placer Dome Inc

P0 Box 49330
1600 - 1055 Dunsmuir Street
Vancouver V7X 1P1
B.C.
Canada
Tel: +1 604 682 7082
Fax: +1 604 682 7092
A78

Tailings Dam Failure

Uploaded by

Copyright:

Available Formats

Tailings Dam Failure

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tailings Dam Failure

Uploaded by

Copyright:

Available Formats

fj United Nations Environment Programme

Environmental and Safety Incidents

Tailings Dams at Mines

Results of a Survey for the years 1980-1996

Mining Journal Research Services

Joint UNEPIDHA Environment Unit

100 % recyc'ed paper

- identifying the environmental and safety impacts most commonly encountered,

1.0 Introduction ..............................................1

2.0 Review of worldwide mining activities ............................ 7

3.0 Review of construction methods ................................13

4.0 Causes of tailings dam failure ..................................19

5.0 Environmental effects ........................................23

6.0 Significant failures and other incidents since 1980 ...................27

7.0 Conclusions ...............................................45

8.0 Recommendations ..........................................49

9.0 Bibliography ..............................................51

1 Selected major mining districts by commodity ........................8

2.0 REVIEW OF WORLDWIDE MINING ACTIVITIES

United States ArizonafNew Mexico/Utah

2.3 Lead and zinc

2.5 Fertilizer minerals

In addition to primary deposits, uranium is also produced as a by-product of phosphate mining

2.8 Global context

3.1 Tailings dam types

Location of a tailings impoundment depends to a large extent on the natural topography.

3.1.2 Valley side

3.1.3 Valley bottom

A brief review of tailings dam construction techniques may be useful.

3.2.1 Upstream construction

3.2.2 Downstream construction

3.2.3 Centreline construction

3.3 Key factors in tailings dam construction

3.3.1 Water control

3.3.2 Foundation material

4.0 CAUSES OF TAILINGS DAM FAILURE

4.1 Breached retaining walls

4.2 Earthquake effects

4.4 Decant failure

5.1 Groundwater contamination

5.2 Dust emissions

5.3 Other environmental impacts

5.4 Community location

6.1 Summaries of responses

IrS S$ Ir 0 'O O NO '.0 '.0 N N N 00

Irs 'rs irs .O

6.1.3 Latin America

6.1.7 Australia and New Zealand

6.2 Tailings dam populations

Table 4. Selected tailings dam populations

Country/region Active Abandoned Total

South Africa c.400

6.3 Types of incidents

Country/region Types of incident recorded

British Columbia Dam wall failure

Brazil Dam wall failure

Country/region Types of incident recorded

Canada, British Columbia

Canada, New Brunswick

Australia, Western Australia