A review of Tailings dam failures15 May 2001
The history of engineering failures teaches us that we ignore history at our peril. But have tailings dam failures prompted subsequent improvements in design, operation and maintenance?
Mine owners and those associated with the extractive industries, particularly with project development and permitting, will be familiar with the list of tailings dams events that have been a feature of recent anti-mining publicity. Frequent reference is made to the inventory of tailings dam failures published on the internet, in NGO publications and in various other media. This identifies failures over the last 20 or so years and highlights such details as fatalities, volume of discharge etc. These data have been used, particularly in Europe, to influence delay and encourage prevention of the development of new mining operations.
At face value this list reflects badly on the mining industry globally and the inference is that few technical developments and improvements in design and construction techniques have been made over this period. The extractive industries, despite the political emphasis on minerals recycling, remain major contributors to the world economy and as a result it is important to assess whether the inference that tailings dams performance has deteriorated is justified.
To knowledgeable practitioners the impression given by the bare statistics is misleading and takes no account of the growth in demand for minerals or the geographical development over the period, and hence of the number of new and old operating mines globally. Interpretation of the statistics requires detailed technical assessment of failure type, geographical location as well as local regulatory and environmental regime.
This article will provide a brief review of this data and assess the implications with respect to technical performance, particularly with regard to tailings dam design, management and operation. It is based on a more detailed analysis of mining incidents and tailings dam failures presented at the Institute of Mining and Metallurgy in London in January 2001.
The internet list of events associated with embankment dams constructed to confine waste arising from mineral extraction appears to represent the most recently published and probably the most quoted inventory. However a more appropriate review of performance can be obtained from a comparison of two studies, undertaken during the latter part of the last decade by icold, in which statistics for the failure of both water dams and tailings embankments were reported. Though there are significant differences between water and tailings dams, much of the technology and design approaches are common. Direct comparison of the records for both dam types is considered appropriate. Analysis of the ICOLD reports, which list events from the beginning of the 20th century, has the advantage of enabling a direct comparison of the performance of water and tailings dams over a significantly longer period.
ICOLD’s statistics highlight the problems of reporting dam failures. Both studies were based on questionnaires and the responses received from both authorities and owners and were limited in certain circumstances. Mining companies in particular were often loath to publish any data — for understandable but misguided reasons. However, despite being incomplete, the data show a more positive trend than reported elsewhere. If the background to the mining statistics are disregarded – ie trends in mining volume, location, regulatory environment and extracted mineral, all of which are of direct relevance to the statistics – a number of conclusions can be drawn from an overview of the data.
Firstly the records, summarised in the table below, indicate fewer tailings dam events in the last decade of the 20th century. When viewed as a whole and on a decade-by-decade basis, the trend is similar to that for water dams over the same period. Although the record is known to be incomplete, it can be surmised that any error due to under reporting will be greater for early periods and that this trend is likely to be enhanced rather than diminished.
Secondly, the studies show significant trends in failure mode. For water dams a significant proportion of failures occur in the first five or so years of operation. There appears to be no similar correlation with age for tailings dam. The statistics do however indicate that, as for water dams, there are apparent deficiencies in design, operation and management which are repeated. The ICOLD study of water dam failures identified a number of common themes in terms of failure mechanisms and causation. The results of the study of tailings dams events show similar trends. However a large proportion of recent events implicate operation and management practice rather than poor or inadequate initial design.
Unfortunately, and disturbingly for the industry, records also indicate certain common features amongst the major failures which imply that lessons are not being learnt from the past. The history of engineering failures teaches engineers of all disciplines that we ignore history at our peril.
The case for regulation
One very clear message from the tailings dam data has regulatory rather than technical implications. Analysis of the limited geographical data for tailings dam failures over the 20th century indicates that there have been clear benefits to the industry and to the local environment from appropriate changes in legislation.
There is much debate within Europe, North America and Australia concerning the plethora of legislation aimed at the mining industry – much of which has been poorly thought out and has resulted, arguably, in over prescriptive standards. However, as with water dams and the reservoirs they retain, the development of specific legislation regarding the safety of these structures has resulted in significant improvements in the long term integrity of the facilities.
Consideration of failure rates for tailings dams shows a similar trend in those countries that have enacted specific legislation. These data indicate that the failure rate can be dramatically reduced by the introduction of an appropriate regulatory system, which reviews and monitors these structures.
The UK is rarely cited as an example of a country with industrial scale mineral extraction, with the current exception of aggregates. This belies the truth as the UK, despite the current minimal output, has been a major world scale producer of copper, tin, lead and most recently of coal. Associated with industrial scale mining in the 19th century was an increasing social awareness, which led to the development of legislation from the 1850s onwards. As a result, mining laws within the UK have been in development over a period of some 150 years and probably represent some of the earliest examples of mining health and safety legislation in the world.
Much of this was reactive. For instance, the establishment of mine water control legislation at the beginning of the 20th century was in response to an increasing death toll underground by flooding. Of most significance to this paper however is the legislative aftermath to the 1966 Aberfan disaster in Wales. This led to the enactment of specific regulations for all mine and quarry tips. The amendment of the UK Mining Act included specific legislation and regulations governing the planning and operation, as well as the statutory inspection and monitoring, of tips and tailings lagoons. The Act further recognised the different requirements of tailings and water dams, further reinforced in the UK 1975 Reservoirs Act. This enforced similar regulation of the inspection and monitoring of water dams but specifically excluded tailings dams.
The UK failure statistics show that in the post-regulation period the number of tailings dam/mine tip incidents reduced to a negligible number, by inference, as a result of the introduction of the Mines & Quarries Regulations in 1971. To put this into context, in 1972 the UK had more than 2500 tips and tailings dams containing well in excess of 2000M tonnes of mine waste. The regulations instigated the requirement for statutory audits of all mine/quarry tips by ‘competent’ persons and the regular preparation of statutory reports. These regulations, though not perfect, had a dramatic effect on the performance of the UK mining industry and showed that appropriately drafted and enforced legislation, specifically for mine structures, can have a beneficial impact on their performance. Similar legislation was enacted in Zambia in 1972 following the Mufulira disaster. The impact of legislation in Zambia was by inference similar in reducing the number and frequency with which reportable events occurred.
The importance of independent expert review processes for such structures has been recognised by major mining houses and lending institutions alike, such that in countries where no appropriate legislation exists, the companies themselves have instigated independent annual inspection and reporting on their tailings disposal facilities.
However, recent events in Europe indicate that inappropriate legislation that does not take account of the specific problems of the mining industry can have a negative effect on tailings dam safety. It may even lead to a false sense of security. In countries where there has been no historic mining legislation but where strong legislation exists for water dams, there has been a tendency for legislators to adopt these regulations. This often has disastrous consequences – primarily because the differences between water and tailings dams was neither recognised nor accepted as of relevance.
The characteristics of a tailings dam and the associated appurtenant works are significantly different from those of a water dam and its reservoir. A lack of understanding of these differences has led to both operational and design problems with structures. The differences are summarised in the table on the left which highlights the design, operation and inspection factors for the two types of structure.
It is imperative that designers and regulators understand these differences in order that appropriate standards and conditions are applied. Fail-ure to distinguish the differences between tailings dams and water reservoirs have, in the author’s opinion, led to many of the failures that have been reported in recent media releases.
Analysis of tailings dam events over the last 100 years indicate that the data presented by the media are misleading and that the rate of occurrence is not on an in-creasing trend. The primary conclusion from review of the records is that improvements in design and operational techniques have led to a reduction in the failure rate during the last ten years, contrary to publicity. Clearly a much more detailed analysis is required to remove industry variables and provide a clear summary and precise comparison. This analysis is in hand.
Although the data show that greater emphasis on the key design aspects has improved performance, primarily through increased scrutiny by regulators, this has not been matched by improvements in management and operation. Disturbingly, in a number of instances, historic failures have not led to subsequent improvements in design for operation.
Of particular significance is the fact that where local mining regulations are in place, and which require ongoing inspection, auditing and monitoring of tailings dams, the incidence of failure has been reduced. Such statutory inspection and continual expert input to the operational design and monitoring automatically addresses the cause of many of the events recorded by ensuring that standards applied during staged construction are appropriate and that the management and operational approach meet design criteria.
The mining industry is just in its complaints of over regulation of the sector, primarily because legislation is often enacted piecemeal and is reactive. In the case of tailings dams, appropriate legislation, based for instance on the UK Mines and Quarries (Tips) Regulations, could be beneficial to both mining companies and permitters in reducing their exposure to failure.